jpm fixed in
TRANSCRIPT
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 1/236
Global Fixed Income Markets2012 Outlook
Pavan WadhwaAC
(44-20) 7777-3370
Fabio Bassi
(44-20) 7325-8615
www.morganmarkets.com J.P. Morgan Securities Ltd.
See page 233 for analyst certification and important disclosures.
Contents
Overview 3
Economics 31
Euro Cash 38
European Derivatives 57
United Kingdom 77
US Cross Sector 90
US Treasuries 107
US Interest Rate Derivatives 131
Japan 158
Inflation-linked Markets 176
Australia
New Zealand 219
…………………………………………………….
Interest rate forecasts 226
Recent curve movements 227
Recent sov cash spread
movements 228
Recent sov CDS spreadmovements 229
Sov & bank redemptions 230
Euro area sov ratings / SMPpurchases / Election calendar 231
Euro area fact sheet 232
European Rates StrategyJ.P. Morgan Securities Ltd.November 24, 2011
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 2/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) [email protected]. Morgan Securities Ltd.
2
Key 2012 Global Fixed Income Markets Trading Themes
Overview Pavan Wadhwa, Kedran Panageas The negative feedback loop between sovereigns, banks, and the real economy is exacerbating the peripheral debt crisis.
Technicals are poor for peripheral sovereign debt; the existing investor base has been impaired and distressed credit funds
are keeping their distance due to the highly politicised nature of sovereign debt restructurings. We expect further rounds
of forced capital increases for European banks as the Euro zone slips into recession. The ECB is unlikely to step up its
pace of buying significantly, absent a catastrophe. Position for lower German yields and a flatter curve. Underweight
intra-EMU debt relative to Germany. Australian and Chilean duration is attractive. We highlight trades that are positively
convex in peripheral spreads, and others that are uncorrelated to the sovereign debt crisis.
Euro Pavan Wadhwa, Fabio Bassi, Gianluca Salford We see the refi rate at 0.50% and 10Y Bund yields at 1.25% by mid-2012; go long duration and initiate 2s/10s flatteners
with a 100bp target. Underweight peripherals and position for flatter credit curves. EONIA fixings will fall to 30bp as
excess liquidity stays elevated; go long 6Mx6M EONIA. Position for 1s/5s outright and conditional bull flatteners, and
buy receiver structures on 6Mx5Y swaptions. Swap spreads are likely to trade wider than their Lehman peak; we target
2Y and 10Y swap spreads at 145bp and 90bp, respectively, and favour 2Y wideners. Conditional swap spread wideners
offer better risk/reward than outright wideners. In EUR and GBP gamma, buy intermediate-tails and sell short-tails; we
target EUR 3Mx10Y at 9.4bp/day. Favour longs in Bund volatility vs. swaption volatility.
UK Francis Diamond We expect QE gilt purchases to be upsized to a whopping £425bn in 2012. QE, low non-domestic ownership of gilts, and
a flexible currency should limit any increase in gilt yields should the UK fall under the sovereign risk spotlight. 2Y gilts
should trade in a 50-60bp range. 10Y and 30Y gilt yields are likely to decline: we target 1.50% in 10Y gilts by 2Q12 and
expect the 2s/10s gilt curve to flatten to 100bp. Position for wider 5Y and 10Y swap spreads in 1H12.
US Terry Belton, Srini Ramaswamy
Stay long duration in early-2012, targeting 1.70% in 10Y Treasuries. Avoid consensus trades. Synthetic Treasuriescreated by asset swapping cheap foreign bonds should outperform in 2012. FRA/OIS and intermediate swap spreads will
widen initially before narrowing back. Initiate “synthetic” conditional curve trades by replacing swaptions at the front end
with YCSOs. Approach 2012 with a long gamma bias, but look for 3Yx10Y to decline to 6bp/day by 1Q12 end.
Japan Takafumi Yamawaki, Yuya Yamashita We expect 10Y JGBs to trade in a 0.8-1.1% range in 1H12 and 0.9-1.3% in 2H12, with risks biased towards lower JGB
yields. We have a flattening bias on the JPY swap curve. Large negative USD/JPY cross currency basis makes it
attractive to buy JGBs and swap them into USD. 3s/6s basis is likely to widen.
Inflation Jorge Garayo, Francis Diamond, Kimberly Harano
We expect significant declines in headline and core inflation across the board in 2012, although inflation expectations will
remain anchored. Real yields will test new lows, turning negative in Europe and staying close to 0% in 10Y TIPS.
Breakevens will be biased narrower. Expect cash breakevens to underperform relative to inflation swaps in early 2012.
Euro area: Position for flatter breakeven curve and higher peripheral linker yields. UK: Be long intermediate real yields,
and short inflation breakevens in the 10Y sector. US: Declining inflation and nominal yields, along with fiscal tightening,
should cause breakevens to narrow.
Australia / New Zealand Sally Auld Buy AUD duration on dips as offshore demand is likely to stay elevated, the RBA will ease policy, and domestic investors
are short their benchmarks. Position for curve steepening. We are biased towards wider swap spreads. In NZD, we
expect yields to make new lows in 2012 and the curve to be biased steeper.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 3/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Kedran Panageas (44-20) 7777-0326J.P. Morgan Securities Ltd
3
Overview
• The negative feedback loop between sovereigns,
banks and the real economy is exacerbating the
peripheral crisis in the Euro zone
Sovereigns
• Fallen angels: peripheral sovereign debt has too
much credit risk to remain in DM portfolios, but
is too large to be easily absorbed in EM
portfolios…
• …while the ‘benchmarking benefit’ that typically
accrues to corporate ‘fallen angels’ will not accrue
to distressed sovereign debt…
• …and, given the highly politicized nature of
sovereign debt restructurings, distressed credit
funds have been slow to invest in this space
• Non-domestic investors, who own nearly half of
all peripheral sovereign debt, have been steadily
selling and are likely to have limited appetite to
add risk in the near term
• France is at risk of losing its AAA rating by mid-
2012, which would cause EFSF capacity to fall by
one-third
Banks
• Spreads on senior bank debt and covered bonds
are near, if not above, Lehman highs, while Euro
zone bank debt issuance has plummeted to below
Lehman levels
• The exposure of core country banks to all
peripheral assets tops €1tn, or 90% of bank
capital and reserves…
• …and the EBA’s €104bn estimate of bank capital
shortfall for Euro zone banks is €150bn below
J.P.Morgan estimates. We expect further rounds
of forced capital increases for Euro zone banks
Economy
• J.P.Morgan economists expect Euro area growth
to significantly undershoot official forecasts…
• …as further fiscal consolidation by Euro area
sovereigns pressures GDP growth lower…
• …and forced bank deleveraging puts 1-2%-pts of
downward pressure on GDP growth
Recommended Trades
• Position for lower yields and a flatter German
curve
• We recommend underweighting peripheral debt
in the first half of 2012; however, we expect
peripheral yields to fall in 2H12 in response to a
concerted ECB/EU/IMF response to the crisis
• Global markets that offer high real yields,
potential policy rate cuts, and high sensitivity of
GDP growth to Europe are attractive; go long
10Y duration in Australia and Chile
• We highlight trades that are positively convex
when peripheral spreads widen/narrow…
• …as well as trades that offer diversification from
the peripheral debt crisis
The sovereign pandemic
If 2010 was the year that the EU sovereign crisis erupted,
2011 was the year that it morphed into a dangerous new
phase. No longer did it limit itself to small nations such
as Greece, Ireland, and Portugal. Rather, the contagion
spread to Italy and Spain and to a certain extent, core
countries such as France. Italy and Spain constitute
nearly one-third of European debt markets, with acombined €2tn of marketable debt outstanding.
This dangerous new phase began around mid-year, when
Greece started to slip on its fiscal targets and the threat of
private-sector burden-sharing became real (Exhibit 1).
Despite making good progress on fiscal targets and
structural reforms in 2010, Greece hit a wall in 2011. By
mid-May, it was massively behind target and the IMF
needed assurance of future funding in order to continue
disbursing promised monies. This led to a second Greek
bailout in July that stipulated a bond exchange on private
holdings that would extend maturities to 15–30 years and
lower coupon rates modestly. This broke the taboo thatsovereign debt was riskless and reinforced the notion
that peripheral debt was now a credit product.
This debt restructuring, although never actually
consummated, as well as the anticipated need for wide-
scale bank recapitalisation and unhelpful political
rhetoric in Italy, were the three main factors that allowed
the crisis to jump containment. By August Italian and
Spanish 10Y bond yields had breached 6% and the ECB
was forced to start buying Italian and Spanish bonds in
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 4/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Kedran Panageas (44-20) 7777-0326J.P. Morgan Securities Ltd
4
their Securities Market Programme (SMP). The crisis
took another turn for the worse in September and
October when Greece’s backsliding on reforms
deepened, the EU negotiated a new and more severe
Greek debt restructuring, and Greek Prime Minister
George Papandreou called for a public referendum on the
bailout package.1 In response, Angela Merkel and
Nicolas Sarkozy raised the prospect of a Greek exit
from the Euro zone which reinforced the risk of a
hard default and/or a currency switch in investors’
minds.
All five peripheral countries have now been forced to go
to early elections or technocratic governments since the
onset of the crisis, with Greece and Italy being the most
recent victims. This indicates the difficulty in enacting
needed reforms in the face of economic slowdown, as
well as the high political and implementation risk of fiscal austerity.
Going forward, we think the crisis will continue to
escalate in 1H12. Our baseline view for 2012 includes
the following developments:
1) The market faces a bimodal distribution of
outcomes going forward: either the ECB will step
1 For reference, Appendix-1 presents a detailed timeline of key crisisevents since 2009.
up its bond-buying pace significantly, or another
country may eventually lose access to capital
markets. Our basecase view is that the ECB (and
Germany) will remain deeply reluctant to greatly
increase the size of the its purchases at this time,
which will contribute to an escalation of the crisis in
1H12. This is due to the ECB’s desire to 1) maintain
oversight and aid conditionality, in order to mitigate
moral hazard, and 2) keep monetary and fiscal policy
distinct. However, we expect a concerted
ECB/EU/IMF policy response in 2H12 which will
help to ultimately stabilize yields.
2) Greece may continue to fail to meet its fiscal and
structural reform targets. We think Greece will
eventually need a more severe debt restructuring
than the 50% haircut agreed to at the 26 October
EU summit, which does not go far enough to restoredebt sustainability.2
3) The Euro area will dip into recession, with the
biggest contraction seen in the periphery (see
Economics). A recession dynamic will make it more
difficult for sovereigns to meet fiscal and structural
reform targets.
2 Please see Overview, Global Fixed Income Markets Weekly, 28October 2011.
Exhibit 1: The sovereign debt crisis engulfed Italy and Spain in 2011, morphing into a dangerous new phase. This was spurred by 1) political risk inGreece and Italy, 2) Greek debt restructuring, and 3) concerns around bank capital needsTimeline of the EU sovereign debt crisis; 10Y weighted* peripheral spreads vs. key events; bp
0
100
200
300
400
500
600
700
800
Jan-10 Apr-10 Jul-10 Oct-10 Jan-11 Apr-11 Jul-11 Oct-11
Greekcrisis
EFSFconcerns
EBA 1st stress test
Irishconcerns
EU plans mandatoryPSI with ESM aid;
Irish bailout
Discussions onEFSF2 / ESM
Portuguese bailout;discussions of Greek PSI
EBA 2nd stress test;2nd Greek bailout
Italian crisis and ECBbuys IT/ES
Greek backsliding /renegotiation of PSI; referendum and Italian politicalgridlock
0
100
200
300
400
500
600
700
800
Jan-10 Apr-10 Jul-10 Oct-10 Jan-11 Apr-11 Jul-11 Oct-11
Greekcrisis
EFSFconcerns
EBA 1st stress test
Irishconcerns
EU plans mandatoryPSI with ESM aid;
Irish bailout
Discussions onEFSF2 / ESM
Portuguese bailout;discussions of Greek PSI
EBA 2nd stress test;2nd Greek bailout
Italian crisis and ECBbuys IT/ES
Greek backsliding /renegotiation of PSI; referendum and Italian politicalgridlock
* Spreads to Germany of Greece, Ireland, Portugal, Italy, and Spain. Weighted by proportional contribution to the J.P. Morgan EMU Bond Index.Note: For reference, Appendix-1 presents a detailed timeline of key crisis events since 2009.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 5/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Kedran Panageas (44-20) 7777-0326J.P. Morgan Securities Ltd
5
4) Faced with recession and disinflation, the ECB will
cut the refi rate to 50bp by June 2012, in
successive 25bp moves occurring at its December,
March, and June meetings. We expect the deposit
rate to be reduced to 25bp at the December meeting
(see Economics).
5) Portugal and Ireland will need to renegotiate
funding packages in mid-to-late 2012, which will
likely require private sector involvement (PSI).3
6) France is at risk of losing its AAA rating given
contingent commitments to other sovereigns and
French banks, especially if economic growth
disappoints. This could reduce EFSF capacity by up
to one-third (see Euro Cash).
Why is this crisis so pernicious?
Exiting this crisis has proven extremely difficult for
several reasons. First, we are in a negative feedback loop
that is proving difficult to stabilize (Exhibit 2). The
banking crisis and recession of 2008/2009 damaged
sovereign balance sheets, and today’s sovereign crisis
(and the policy response) are in turn damaging bank
balance sheets and stunting economic growth. As with
3 Please see Overview, Global Fixed Income Markets Weekly, 8 July2011.
most negative feedback loops, a circuit-breaker is needed
in the form of some external factor such as outside
money, central bank intervention, or a positive growth
shock from abroad. In the sections below, we discuss
each of the three legs of the crisis in more detail –
sovereigns, banks, and the economy – with regard to howthey will drive outcomes in the periphery in 2012.
Second, and critically, the scale of the problem is much
larger than anything experienced before. The Russian
and Argentine defaults in 1998 and 2001, respectively,
only affected between €100–€125bn of marketable debt
each. This was considered to be large at the time, but it
pales in comparison to the nearly €2.5tn of marketable
debt outstanding today across Italy, Spain, Greece,
Ireland, and Portugal.4 Given the scale of the problem, it
is difficult to find a credible and/or willing lender of last
resort.
Third, EU policy response to date has been fractured and
behind the curve. The financially sounder core countries
and the ECB have been reluctant to support at-risk
countries to the extent necessary, for fear of socialising
debt and exacerbating moral hazard. This problem will
4 The marketable debt figure includes T-bills, zero-coupon notes, and
bonds, but does not include official sector loans, private sector loans,commercial credits, etc. We estimate total debt outstanding of the five
peripherals is between €3.0–3.5tn, based on current GDP and debt/GDPratios.
Exhibit 2: The negative feedback loop between sovereigns, banks, and the real economy is exacerbating the peripheral crisisTransmission mechanism between the three pillars of the EU sovereign debt crisis
Banks
Sovereign
Economydeleveraging
credit losses
PROBLEM:
• Overleverage
• Loss of capital market access
• Solvency crisis
PROBLEM:
• Exposure to peripheral assets
• Overleverage
• Loss of capital market access
PROBLEM:
• Slow growth
• Recession risk
• Social unrest
f i s c a l c
o n s o l i d
a t i o
n w r i t e
d o w
n s
b a i l o
u t s
r e c e s s i o n /
s o c i a l u
n r e
s
t
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 6/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Kedran Panageas (44-20) 7777-0326J.P. Morgan Securities Ltd
6
be difficult to overcome until there is more control, more
trust, and more homogenisation of fiscal policies across
Europe. The fiscal integration process will take years if not decades to achieve, and will require profound EU
treaty changes.5 We now discuss each of the three pillars
of the EU crisis in more detail.
1. The sovereign debt and funding crisis
First, since the onset of the financial crisis in 2008,
sovereign balance sheets have been severely damaged: it
is, in the words of German Chancellor Angela Merkel, at
least a 10-year project to restore fiscal health. Debt-to-
GDP ratios amongst the peripherals are up 37%-pts on
average (Exhibit 3). Even stronger countries now appear
stretched.
Second, rating agencies have been relentless in
downgrading peripheral sovereign debt. The average
peripheral rating has fallen from A+/Aa2 two years ago
to BBB-/Ba1 today (Exhibit 4). This stands in stark
contrast to the rest of the developed market (DM)
sovereign universe, the vast majority of which has
minimal credit risk and is rated AA or higher (more on
this later).6
Third, the crisis has severely damaged market liquidity.
This is evident in wide bid/offers, thin trading volumes,
and skyrocketing market volatility for peripheralsovereigns (Exhibit 5).
Together, the heightened credit risk and market
volatility have severely impaired the investor base for
peripheral sovereign debt. The natural investor base
consists mainly of rates-based investors, including
central banks, commercial banks, insurance companies,
pension funds, etc. These investors begin to shy away at
the first sign of credit risk or investment volatility.
Further, these investors also face heightened regulatory
scrutiny. This consists of additional disclosure
requirements, capital stresses on assets held in banks’
held-to-maturity books, a possible move to non-zero risk-weightings, and the need for deleveraging or asset sales.
This treatment makes them less able to withstand the
5 See Overview, Global Fixed Income Markets Weekly, 19 August 2011for a discussion on the lengthy process required to change EU treaties.6 In fact, Moody’s classifies only three DM issuers (outside of the
peripherals) as single-A (Israel, the Czech Republic, and South Korea).
These are stronger single-A credits, all of them rated A1 at Moody’sand two split-rated at S&P. Specifically, Israel is rated AA- at S&P,
the Czech Republic is rated AA at S&P, and South Korea is rated A+ atS&P.
Exhibit 3: Peripheral sovereign balance sheets have been severelyimpaired…Debt-to-GDP ratios across selected Euro zone countries, end-2011E* vs. end-2008; %
2011E* 2008 Change
Greece 158 111 47Ireland 112 44 68
Portugal 102 72 30
Italy 120 106 14
Spain 68 40 28
Average**, peripherals 112 75 37
Belgium 97 90 7
France 85 68 17
Germany 82 66 16
Euro area 88 70 18 * Official estimate as provided by Eurostat.** Simple average.
Source: Eurostat
Exhibit 4: …and rating agencies have been relentless in downgradingperipheral sovereign debt…Number of notches that peripheral countries have been downgraded by Moody’s andS&P over the past two years
18-Nov Dec 2009
Downgrade
(notc hes ) 18-Nov Dec 2009
Downgrade
(notches)
Greece CC NEG BBB+ (WN) 12 Ca DEV A2 14
Ireland BBB+ AA 5 Ba1 NEG Aa1 9
Portugal BBB- NEG A+ 5 Ba2 NEG Aa2 9
Italy A NEG A+ 1 A2 NEG Aa2 3
Spain AA- NEG AA+ 2 A1 NEG Aaa 4
Av erage* BBB- A+ 5 Ba1 Aa2 8
S&P Moody's
* Simple average across the peripherals based on a unit ratings scale (AAA=1, AA+ = 2, AA = 3, etc.).Note: NEG, POS, DEV indicates the ratings outlook is negative, positive, or developing, respectively. WN (WP) indicates the rating is on negative (positive) watch.Source: Bloomberg, Moody’s, S&P
Exhibit 5: …while market liquidity has been severely impaired, leadingto wide bid-offers and skyrocketing volatility in peripheral yields6M standard deviation of daily yield changes in Italian 10Y b/m vs. average monthlybid-offer spread on Italian 10Y bonds*bp/day bp of yield*
2
4
6
8
10
12
14
16
Nov 08 May 09 Nov 09 May 10 Nov 10 May 11 Nov 11
1
2
3
4
5
67
8
9
6M yield vol
bid-offer spread
* Estimated from the average bid-offer spread reported in price terms by MTS andprice and duration statistics of the J.P.Morgan Global Bond Index for Italy 7-10Ybonds.Source: MTS, J.P.Morgan
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 7/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Kedran Panageas (44-20) 7777-0326J.P. Morgan Securities Ltd
7
MTM volatility that they previously would have been
able to weather through fair value accounting in held-to-
maturity portfolios. As a result, the traditional peripheralsovereign investor base has grown significantly more
risk-averse over the past year, leading to a constant
stream of media reports of large institutional investors
paring their peripheral debt holdings.
This selling has pushed yields significantly higher:
peripheral sovereign debt now has too much credit
risk to remain in DM portfolios, but is too large to be
easily absorbed in EM portfolios. First, outstanding
debt from EU peripherals comprises over half of
outstanding EM sovereign debt, making it a huge
challenge to absorb (Exhibit 6). Second, downgraded
sovereign debt will most likely not be added to EMsovereign indices (which have fairly low country wealth
and income thresholds) or to high-yield or EM corporate
credit indices. Thus, the ‘benchmarking benefit’ that
typically accrues to corporate fallen angels when they
migrate from investment-grade to high-yield territory is
unlikely to accrue to distressed sovereign debt. For
instance, we estimate that the yield on the typical auto
industry ‘fallen angel’ fell an average of 150bp after
crossing into high-yield territory, due to demand from
benchmarked investors in high-yield space (Exhibit 7).
Third, even though credit-focused hedge funds or
distressed investors may theoretically be interested,
the highly politicised nature of sovereign
restructurings makes them more difficult to analyse
as compared to the better-defined corporate
restructuring protocol.7 For instance, GM is routinely
pointed to as an example of a politicised bankruptcy in
corporate space, where the government’s involvement
led to the typical recovery priority being discarded, to the
detriment of bond investors and to the benefit of
unionized employees. Although several investors
explored legal options, they ultimately settled for the
government-negotiated outcome as it was too
difficult/expensive to fight in court. Such treatment is
likely to be amplified many times over in the sovereign
debt space, where there is no defined priority of
payments in a debt restructuring. This is ultimately to
the detriment of the EU as it is likely to dissuade new
bondholder investment in peripheral sovereigns. For
instance, the market reacted quite negatively when
Merkel and Sarkozy first announced in late 2010 that any
official loans granted out of the ESM after 2013 would
7 See Overview, Global Fixed Income Markets Weekly, 10 December 2010, for sovereign debt legal issues.
have preferred creditor status over private debt
investors8. Partly due to these concerns, distressed
credit funds have been slow to invest in the peripheral
sovereign space and will likely remain wary.
8 For instance, see “Cross-country variations in capital structures: therole of bankruptcy codes”, Viral Acharya et. al., Journal of Financial
Intermediation, 2011, and “Creditor rights and Corporate risk-taking”,Viral Acharya et. al., Journal of Financial Economics, forthcoming.
Exhibit 6: Peripheral sovereign debt has too much credit risk to remainin DM portfolios but is too large to be easily absorbed in EMportfolios…Outstanding developed and emerging markets debt universe* by Moody’s rating**; %of total
Rating DM ex peripherals Peripheral EM Total
Aaa 48% - - 48%
Aa 26%^ - 4% 30%
A 1%^^ 7% 1% 9%
Baa - - 6% 6%
Ba - 1% 3% 4%
B - - 2% 2%
<= Caa - 1% - 1%
Total 75% 9% 16% 100%
* Sample consists of 105 countries of which 27 (78) are developed (emerging) debtmarkets. Total gross debt outstanding is $55tn of which $46tn ($9tn) is from developed(emerging) markets. Debt includes all outstanding bonds, loans, and other obligations.** Local currency long-term debt rating. Note: the US is rated AAA by Moody’s.
^ Consists of three countries: Belgium, Hong Kong, and Japan.^^ Consists of three countries: Israel, South Korea, and the Czech Republic.Source: Moody’s, IMF
Exhibit 7: …and the ‘benchmarking benefit’ that typically accrues tocorporate ‘fallen angels’ will not accrue to distressed sovereign debt.Moreover, given the highly politicised nature of sovereignrestructurings, distressed credit funds have been slow to invest in thisspace
Avg. yield-to-maturity of 10Y GM, Ally Financial, Ford, and Ford Motor Credit* bondsaround the 5 May 2005 downgrade to HY**; %
6.5
7.0
7.5
8.0
8.5
9.0
9.5
10.0
10.5
-3 -2 -1 0 1 2 3# months around downgrade to HY**
* Specific bonds used are: GM 7.125% Jul13, Ally Financial 6.5% Feb16, Ford 7.45%Jul31, and Ford Motor Credit 7% Oct13. We use the Jul31 bond for Ford Motor Co. asit did not have a large/liquid 10Y note outstanding at the time.** S&P downgraded Ford, GM, and their subsidiaries on the same date, 5 May 2005.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 8/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Kedran Panageas (44-20) 7777-0326J.P. Morgan Securities Ltd
8
Additionally, non-domestic investors are likely to
have limited appetite to add risk in the near term.
Non-domestic investors own nearly half of all peripheral
sovereign debt; their share ranges from nearly 40% in
Spain to over 80% in Ireland (Exhibit 8). These
investors are typically amongst the first to sell and, once
having exited the asset class, will likely be reluctant to
return in the near future. We expect them to provide a
steady drip of selling pressure going forward.
Given these poor demand technicals, sovereign marketaccess remains tenuous and there is no firm ceiling for
spreads. Indeed, European sovereign spreads are
currently close to their lifetime wides, despite the steady
stream of bailouts, crisis measures, and backstop ECB
buying (Exhibit 9). Even EFSF paper has widened to
120bp over Libor, indicating an aversion to credit risk,
ratings downgrade risk, and structural complexity
(Exhibit 10).
Exhibit 9: Given these poor demand technicals, sovereign spreadscontinue to trade close to their lifetime wides, despite ECB buying10Y spreads to Germany; 18 November 2011 vs. 2Y maximum; bp
18-Nov-11 Max 18-Nov / Max
Greece 2425 2471 98%
Ireland 625 1142 55%
Portugal 910 1161 78%
Italy 494 575 86%
Spain 469 486 97%
Belgium 282 309 91%
France 147 188 79%
Av erage* 765 905 85% * Simple average.
Exhibit 10: Even EFSF paper has widened to 120bp over Libor,
indicating an aversion to credit risk, ratings downgrade risk, andstructural complexity10Y ASW spreads for EFSF* and France; bp
-20
0
20
40
60
80
100
120
140
Aug Sep Oct Nov
France EFSF
* EFSF 3 3/8 July 2021
Exhibit 8: Additionally, non-domestic investors own nearly half of allperipheral sovereign debt. These investors have been steadily sellingand may be reluctant to return to this asset class in the near future
% of marketable debt outstanding held by non-domestic investors; %% non-domestic
Greece 67%
Ireland 83%
Portugal 80%
Italy 45%
Spain 38%
Wtd av g., peripherals* 49%
France 66%
Germany 81%
* Weighted by marketable debt outstanding.Source: National central banks. Data as of latest available update for each sovereign;
generally 2Q11/3Q11.
Exhibit 11: Although core-country banks have been cutting their exposure to peripherals, their total exposurestill tops €1tn, or almost 90% of bank capital and reservesExposure of core-country European banks to peripheral country assets as % of bank capital and reserves; 2Q11 vs. 4Q09; €bn
Claims on Austria Belgium France Germany Netherlands 2Q11 4Q09 % chg
Greece 2 1 38 15 3 60 101 -41%
Ireland 2 18 22 76 12 129 214 -40%
Italy 17 17 287 111 36 468 576 -19%
Portugal 1 2 18 25 5 50 81 -37%
Spain 6 16 104 122 53 302 420 -28%
Total 28 54 469 349 109 1009 1392 -28%
Capital and reserves (C&R) 92 58 492 391 102 1135 1081 5%
Exposure / C&R 30% 93% 95% 89% 107% 89% 129% --
Origin country Total exposure
Source: BIS, ECB. Data as-of 2Q11.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 9/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Kedran Panageas (44-20) 7777-0326J.P. Morgan Securities Ltd
9
2. Bank funding and equity capital crisis
Sovereign fiscal issues and slowing economic growth
have direct repercussions on banking sectors. Due to
cross-border capital flows, losses are transmitted quickly
from borrower to lender countries. Although core-
country banks have been cutting their exposure to
peripherals, their total peripheral exposure still tops €1tn,
or almost 90% of bank capital and reserves (Exhibit 11).
A sizable portion of this total exposure is peripheral
sovereign debt; German and French banks alone hold
nearly €180bn, which is now imperilled given the PSI
precedent set by Greece (Exhibit 12). In addition,
anticipated recessions in the peripherals will increase the
losses on other assets such as consumer and corporateloans. Non-performing loan ratios (NPLs) are up sharply
across the periphery since the recession of 2009 and will
likely continue rising as the periphery slides back into
recession (discussed further below; Exhibit 13).
Not surprisingly, these exposures have pummeled bank
stock prices and impaired funding access. Bank equity
prices are down 35–40% on the year and spreads on
senior bank debt and covered bonds are at multi-year
highs (Exhibit 14). Additionally, debt issuance has
ground to a halt; rolling quarterly long-term debt
issuance is at a 5-year low, lower than even Lehman
levels (Exhibit 15). Net, we estimate that core-country
banks have between 0-2 quarters of funding coverage at
this time.9 ECB reliance has steadily increased and the
ECB’s balance sheet likely will have to expand
massively if it is to contain the incipient bank funding
crisis (discussed further below). For instance, the Wall
Street Journal recently reported on the growing volume
9 See Overview, Global Fixed Income Markets Weekly, 30 September 2011.
Exhibit 13: The recession and its aftermath have led to sharply higher NPLs in the periphery, even on non-sovereign lendingNon-performing loans as a % of total gross domestic loans, quarterly data for Greece*and Spain**; %
0
2
4
6
8
10
12
2006 2007 2008 2009 2010 2011
Greece
Spain
* For 2006-2007, quarterly data interpolated from year-end data.** Total of loans to household and corporate sector.Source: Bank of Greece, Bank of Spain
Exhibit 14: As a result of large exposures to peripheral assets, spreadson senior bank debt and covered bonds are near, if not above, Lehmanhighs…Senior bank debt and covered bond ASW spreads; current as of 18 Nov 2011; bp
-50
0
50
100
150
200
250
2007 2008 2009 2010 2011
Senior
Covered bonds
Current : 216bp
Current : 198bp
Source: J.P. Morgan Maggie Credit Index
Exhibit 15: …and Euro zone bank debt issuance has plummeted tobelow Lehman levelsRolling 3M sum of gross bank debt and covered bond issuance for Euro zone banks;all currencies/all domiciles; €bn/quarter
0
50
100
150
200
250
300
2007 2008 2009 2010 2011
Bank debt + CB
Bank debt only Avg Total = €130bn
(€90+40bn)
Total = €65bn
(€53+€12bn)
Total = €50bn
(€33+€17bn)
Source: Dealogic
Exhibit 12: German and French banks alone hold nearly €180bn of peripheral public sector debt, which is now imperilled given the PSIprecedent set in Greece
Exposure of core-country European banks to peripheral public sector debt (includingsovereign debt) as of 2Q11; €bn
Claims on Belgium France Germany 2Q11 4Q10 % chg
Greece 1 7 9 19 32 -41%
Ireland 0 2 2 5 6 -18%
Italy 11 74 33 125 131 -4%
Portugal 1 4 6 17 20 -16%
Spain 3 21 20 48 50 -4%
Total 17 108 70 215 240 -11%
Origin country Total exposure
Source: BIS. Data as-of 2Q11.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 10/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Kedran Panageas (44-20) 7777-0326J.P. Morgan Securities Ltd
10
of collateral swaps designed to boost banks’ access to
ECB facilities.10
On the equity front, European banks are overleveraged,
especially given growing recession risk and market-
implied losses on sovereign debt. The recently
announced capital shortfall of €104bn11 for Euro zone
banks is roughly €150bn below J.P.Morgan estimates
(Exhibit 16).12 To be sure, the 9% core Tier 1 hurdle that
the EBA used is aggressive, but in our opinion the rapid
deterioration in economic fundamentals is a more
significant driver. We therefore believe that further
capital increases (above and beyond the €104bn estimate)
will likely be required of banks. Capital needs are likely
to be higher still if other peripheral countries such as
Ireland and Portugal eventually undergo debtrestructuring, which is our basecase expectation.13
Bank stress feeds back into the sovereign crisis by
increasing sovereign contingent liabilities. Particularly at
risk are Spain, Belgium, and France. Spain has large
unrecognised capital shortfalls at its cajas14, while the
prospect of bank guarantee schemes and capital
injections have already been cited as risk factors for
France’s AAA rating.
3. Spillover to the real economy
Sovereign fiscal austerity and bank deleveraging arespilling over to the real economy, stunting economic
growth. Our economists expect Euro area real GDP to
grow at a rate of -0.6% in 2012, with the contraction
concentrated in the periphery (we forecast 2012 GDP-
weighted growth of -1.8% in the periphery) (Exhibit 17).
This is driven in large part by the large and sustained
fiscal consolidation that must be done. Based on prior
episodes of fiscal consolidation, we estimate that each
1%-pt of fiscal tightening in the primary deficit lowers
10
See ‘ Banks face funding stress’, 17 November 2011, Wall Street Journal . In these collateral swaps, a bank pays a spread to swap
collateral which is not eligible for ECB repo in order to receive higher-quality collateral which is eligible. This enables the bank to boost ECB
access and is a sign of the increasing scarcity/cost of market funding.11 Note that the €104bn shortfall estimate is for Euro zone banks only.
The total estimated shortfall of bank capital across the entire Euro area
is €106bn.12 We estimate that the largest discrepancy stems from removing the
original macroeconomic stresses applied in July; a modest portion
(roughly €20bn) stems from leaving out the cajas in Spain13 Please see Overview, Global Fixed Income Markets Weekly, 8 July2011.14 See EBA Stress Test Results Analysis: Must do much better , RobertoHenriques, 8 July 2011.
GDP growth by 0.7%-pts (Exhibit 18).15 Additionally,
we estimate that each 1%-pt of bank deleveraging
depresses economic growth by 0.6–0.8%-pts (Exhibit
19). Extrapolating to the Euro area suggests 1–2%-pts of
downward pressure on GDP growth from bank
deleveraging. Thus, in our opinion, economic risk is
skewed to the downside.
15 See Growth in the Euro area periphery to underwhelm, 16 September 2011, Nicola Mai.
Exhibit 16: The EBA’s €104bn estimate* of bank capital shortfall for Euro zone banks is €150bn below J.P. Morgan estimates. This isunlikely to be the last round of forced capital increases for European
banks, especially if economic growth disappoints and sovereigns suchas Portugal and Ireland eventually undergo debt restructuringCapital shortfalls announced by the European Banking Authority (EBA) for variousEuro zone countries’ banking sectors vs. J.P. Morgan estimates**; €bn
EBA target
capital
J.P.Morgan
estimated capital
shortfall
J.P.Morgan -
EBA estimate
104 254 151
France 9 48 39
Germany 5 43 38
Spain 26 53 27
Italy 15 35 20
Greece 30 40 10
Portugal 8 17 9
Others 11 18 7
Country
Total
* Note that the €104bn shortfall estimate is for Euro zone banks only. The totalestimated shortfall of bank capital across the entire Euro area is €106bn.** See ‘The Great Bank Deleveraging: Banking Sector Outlook 2012 ’, 4 November 2011. Assumptions underlying J.P. Morgan estimates are 1) 9% Tier 1 capitalthreshold; 2) original Basel 2.5 Tier 1 capital ratios as provided by the EBA; 3) sovereign MTM on entire holdings of sovereign debt as of end-October; 4) macrostresses as provided by the EBA in June 2011; and 5) cajas included in Spain.Source: EBA, J.P. Morgan
Exhibit 17: J.P.Morgan economists expect Euro area growth in 2012 tosignificantly undershoot official forecasts…2012 GDP growth forecasts; % YoY
JPM Official EC JPM - Official JPM - EC
Greece -6.6 0.8 -2.8 -7.4 -3.8
Ireland 0.0 1.6 1.1 -1.6 -1.1
Portugal -3.9 -2.8 -3.0 -1.1 -0.9
Spain -1.1 2.3 0.7 -3.4 -1.8
Italy -1.5 0.6 0.1 -2.1 -1.6
Wtd avg., peripherals* -1.8 1.1 0.0 -2.8 -1.8
Belgium -0.4 2.3 0.9 -2.7 -1.3
France -0.1 1.0 0.6 -1.1 -0.7
Germany 0.3 1.0 0.8 -0.7 -0.5
Euro area* -0.6 1.1 0.5 -1.7 -1.1 * Weighted by 2011 GDP.Source: National treasury ministries, European Commission and J.P.Morgan forecasts
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 11/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Kedran Panageas (44-20) 7777-0326J.P. Morgan Securities Ltd
11
We note that J.P.Morgan economic projections are
significantly more bearish than official forecasts, either
by national treasury ministries or by the EuropeanCommission (EC).16 For example, the J.P.Morgan GDP
forecast for the Euro area is over 1%-pt more bearish
than the EC forecast, and the forecast for the periphery is
nearly 2%-pts more bearish (Exhibit 17).
Slower growth feeds back into the sovereign crisis by
increasing the fiscal deficit, forcing the sovereign into a
vicious cycle of more cutbacks which, in turn, negatively
impact growth and damage political and popular support.
For instance, the realised GDP contraction in Greece has
been several %-pts worse than forecast by the EU/IMF,
coming in close to -5.5% for 2011 vs. an estimated -3%
originally. Similarly, unemployment has surged, leadingto widescale social protests and political backsliding
(Exhibit 20). Across the periphery, all five ruling
coalitions have changed since the onset of the crisis.
However, any benefits from a technocratic government
will likely accrue only over the long term. In the short
term, the effect can be more market-negative if it
increases uncertainty or makes it more difficult to
implement reforms. Net, the result of continued GDP
contraction in 2012 is likely to be further slippage on
fiscal targets, higher deficits, increases in debt/GDP
trajectories, and heightened risk of PSI in other
peripherals besides Greece (despite official declarations
that Greece is a one-off).
Analysing the risk posed by peripherals
The catalogue of risks facing the Euro zone is daunting,
especially where Greece and Italy are concerned. As
discussed previously, Italy’s size and cross-border
linkages make it systemically important. If Italy were to
lose market access, it would greatly increase the
contingent liabilities and funding costs of other
sovereigns, and would most likely result in Spain losing
market access as well. We thus view Italy as the most
challenging issue for 2012.
Why is Italy at risk? First, Italy’s debt burden is large.
For instance, we estimate that Italian debt/GDP will peak
near 125% in 2013, and slowly decline towards 114% by
2020 (Exhibit 21). This is based on our economists’
projections of economic growth and takes into account
recently announced austerity measures (and anticipated
countermeasures when growth disappoints). These
projections envision, for instance, that Italy will run a
16 The latter is typically more up-to-date for latest economic trends.
Exhibit 18: …as urther fiscal consolidation by Euro area sovereignsputs significant downward pressure on GDP growthFiscal consolidation targets and their impact on GDP growth*; % of GDP
2011 2012 2013 2011 2012 2013
Greece 6.0 5.1 0.8 -4.2 -3.6 -0.6
Ireland 2.2 1.6 2.8 -1.5 -1.1 -2.0
Portugal 5.7 6.1 2.2 -4.0 -4.3 -1.5
Spain 2.7 2.6 2.5 -1.9 -1.8 -1.8
Italy 0.8 3.5 2.3 -0.6 -2.5 -1.6
Wtd avg., peripherals* 2.1 3.4 2.3 -1.5 -2.3 -1.6
Belgium 0.2 1.2 1.4 -0.1 -0.8 -1.0
France 1.4 1.6 2.0 -1.0 -1.1 -1.4
Germany 0.0 1.0 0.0 0.0 -0.7 0.0
Euro area 1.2 2.1 1.4 -0.8 -1.5 -1.0
Fiscal consolidation target Impact on GDP growth*
* Fiscal tightening defined as the change in the cyclically-adjusted primary balance.We estimate the follow-on change in GDP growth to be -0.7 times the annual fiscaltightening. Please see Growth in the Euro area periphery to underwhelm, 16September 2011, Nicola Mai.Source: National treasury ministries, EC and J.P.Morgan estimates.
Exhibit 19: Further, the US experience suggests that annual GDPgrowth declines by 0.7%-pts for each 1%-pt increase in bank capitalratio; extrapolating to the Euro area suggests 1–2%-pts of downwardpressure on GDP growth from bank deleveragingResidual of US vs. global real GDP growth (YoY) regressed against US banks’ equitycapital ratio*; January 1999–March 2011; quarterly data; %
y = -70.0x + 6.7
R2 = 31%
-3
-2
-1
0
1
2
3
4
8.0% 8.5% 9.0% 9.5% 10.0% 10.5% 11.0% 11.5%
Bank equity capital ratio*
* Bank equity capital ratio defined as the bank equity capital of FDIC-insuredcommercial banks and savings institutions divided by total assets. Tier-1 capital ratiohas roughly a 1:1 beta to this definition of equity capital ratio over the last 20 years.Source: FDIC, J.P. Morgan
Exhibit 20: The unemployment rate has climbed over 2.5%-pts onaverage since the start of the peripheral debt crisis, leading to socialunrest, political backsliding, and slippage on structural reformsUnemployment rate; current vs. end-2009; %
Current Dec-09 Chg.
Greece 17.6 10.2 7.4
Ireland 14.2 12.9 1.3
Portugal 12.5 11.3 1.2
Italy 8.3 8.3 0.0
Spain 22.6 19.1 3.5
Av g., peripherals* 15.0 12.4 2.7
Euro area 10.2 10.1 0.1
* Simple average.Source: Eurostat
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 12/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Kedran Panageas (44-20) 7777-0326J.P. Morgan Securities Ltd
12
sustained primary surplus equivalent to 5% of GDP, and
that real growth will climb to 0.8% per annum, with
annual inflation at 2%.
Second, given that the Italy’s debt burden is second only
to Greece’s, any slippage on fiscal consolidation or debt
costs could quickly turn the liquidity crisis into a
solvency crisis. To illustrate this, we calculate the
sensitivity of debt/GDP ratios projected in 2020 and
2030 to various economic assumptions. Thesesensitivities are shown in Exhibit 22. We find that:
• A -1%-pt shock to estimated nominal GDP growth
rates (vs. J.P. Morgan base case trajectory shown in
Exhibit 21) will boost debt/GDP in 2020 (2030) by
13% (33%)17…
• A -1%-pt shock to the primary surplus each year
(again vs. J.P. Morgan basecase trajectory) will
boost debt/GDP in 2020 (2030) by 10% (26%)
• A 1%-pt higher borrowing rate on new Italian debt
(vs. J.P. Morgan assumption of 6.5%) will boost
debt/GDP in 2020 (2030) by 7% (22%)
Third, while any one of these shocks on its own would
be negative and leave the debt burden at a very high
level, the shocks are likely to be correlated. That is, if
growth surprises to the downside, it will be more difficult
to run the targeted primary surplus, and borrowing costs
will likely rise. Thus, even mild underperformance can
17 Note that a shock to real growth and a shock to inflation have the
same impact on nominal GDP growth. Thus, they would have the sameimpact on debt sustainability.
rapidly spiral out of control. To be sure, a mild
outperformance will have the opposite impact, with
debt/GDP ratios falling more than is reflected in our
baseline scenario.18
How likely is Italy to deviate to the downside from our
basecase scenario? Unfortunately, our assumptions may
already by generous. On the economic front, risk is
skewed to the downside. In the short term, this is driven
18 However, we note that the sensitivity of the debt/GDP trajectory to
interest rate and growth rate shocks is non-linear due the compounding
of financing costs over time. Therefore, the betas are somewhat larger
(in absolute terms) when the situation is deteriorating, and somewhat
smaller when the situation is improving. For instance, for a 2020horizon, the beta for a positive shock (+1%pt to growth, or -1%-pt to
the borrowing rate) is roughly 5-10% smaller than the beta for a
negative shock (in absolute value terms). For the 2030 horizon, the
beta for a positive shock is roughly 15-20% smaller than the beta for anegative shock. On the other hand, the beta to primary surplus
assumption is linear since it is not multiplied by the outstanding debtstock each year but rather is added to it.
Exhibit 21: Fiscal sustainability in Italy and Spain is uncertain…J.P.Morgan’s basecase forecast of debt/GDP trajectory* and key macro assumptions underlying the trajectory for Italy and Spain**;blue selection indicates peak debt/GDP
2011 2012 2013 2014 2015 2016 ...2020 ...2030
Debt/GDP (% of GDP)* 121 124 125 125 123 121 114 100
Nominal grow th rate (% YoY) 2.0 -0.1 -0.3 1.2 2.7 3.0 2.8 2.8
Primary surplus (% of GDP) 0.5 3.2 4.8 5.0 5.0 5.0 5.0 5.0
Wtd av g. coupon rate*** (%) 4.1 4.6 4.9 5.1 5.3 5.4 5.9 6.3
Debt/GDP (% of GDP)*,** 70 75 81 85 86 87 84 77
Nominal grow th rate (% YoY) 1.9 -0.1 0.6 1.6 2.2 2.4 3.4 3.4
Primary surplus (% of GDP) -4.9 -3.2 -1.5 -0.6 0.8 2.0 3.0 3.0
Wtd av g. coupon rate*** (%) 3.7 4.5 4.9 5.2 5.4 5.6 6.0 6.4
Spain
Italy
* Note that J.P.Morgan estimate for 2011 debt/GDP is slightly more bearish than the latest available official forecast (shown in Exhibit 3)which is based on Eurostat data.** Spanish debt/GDP trajectory does not include expected bank recapitalisation needs of roughly €50bn, covering both banks and cajas(J.P. Morgan estimate). 2011 Spanish GDP is estimated near €1.08tn, so this equates to roughly 5% of current GDP. Rating agencyestimates of Spanish bank recap needs run from around €20bn to as high as €120bn in stressed cases.*** Based upon current weighted average coupon rate and average marginal borrowing rate of 6.50% on new issuance.
Exhibit 22: …and highly sensitive to adverse shocks in GDP growth,primary surplus and funding costsChange in projected 2020 and 2030 debt/GDP ratio for a -1%-pt shock to nominalgrowth and primary surplus assumptions*, and for a +1%-pt shock to borrowing rateassumption*; ratios
Year
Nominal
growth rate
Primary
surplus
Marginal borrow ing
rate on debt
2020 13 10 7
2030 33 26 22
2020 9 10 52030 22 25 16
Spain
Variable Shocked
Italy
* Measured around the J.P.Morgan baseline forecast, holding other variables constant.J.P. Morgan base case assumptions shown in Exhibit 21.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 13/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Kedran Panageas (44-20) 7777-0326J.P. Morgan Securities Ltd
13
by the possibility of European policy dysfunction, and in
the long term, by Italy’s deep-seated structural
challenges. For instance, we assume a long-term 0.8% per annum growth rate even though Italy has only
managed a 0.3% annualised growth rate over the past
decade. On the financing front, Italian yields have spiked
past 7% in recent weeks, so that a 6.5% borrowing cost
may also be ambitious.
Fiscal pressures on Spain are also significant. Spain has a
lower debt burden than Italy does, but is running a higher
deficit. Thus, we expect Spanish debt/GDP to peak later,
somewhat shy of 90% of GDP in 2016 (Exhibit 21). This
assumes that the incoming administration in Spain will
pursue additional fiscal tightening relative to what has
been budgeted so far, eventually achieving a primarysurplus of 3% per annum. However, this trajectory does
not include expected bank recapitalisation costs that
could add between 2-12% debt/GDP, depending on the
estimate.19
The sensitivity of Spain’s debt/GDP trajectory
to shocks in macro assumptions is similar to Italy’s, but
somewhat smaller given a lower starting debt burden
(Exhibit 22).
Thus, there is significant slippage potential, especially
for countries whose debt levels are already high. It may
be challenging for Italy to achieve the necessary fiscal
consolidation amidst a recession and a high level of
social resistance. The recently appointed technocraticgovernment may prove more effective than the previous
administration, but the impact of any structural changes
that it implements is not likely to be felt on the economy
for a long time.20 Given these pressure points, Italy, and
possibly Spain, risk losing market access in 2012.
Possible policy response if another
peripheral country loses market access
Given their size, it would be challenging to fund Italy
and Spain in the event that they lose market access.
We estimate total funding needs for Italy and Spain of
nearly €900bn over the next three years (Exhibit 23).
This includes bond redemptions (conventionals and non-
conventionals) and anticipated deficits, but assumes that
both continue to roll T-bills in the private market. Also,
19 J.P. Morgan estimates roughly €50bn of Spanish bank recapitalisation
needs, covering both banks and cajas. Rating agency estimates run from
around €20bn to as high as €120bn in stressed cases.20 See Euro Cash, Global Fixed Income Markets Weekly, 11 November
2011, for an analysis of Italy’s last experience with a technocraticgovernment.
note that funding needs are front-loaded (€350bn in
2012).
Unfortunately, the official resources available to deal
with this outcome are limited. Euro area sovereigns are
already stretched financially, and non-Euro area
sovereigns see little reason to support fundamentally
wealthy countries. We estimate that in extremis, there is a
maximum of roughly €725bn available to support Italy
(€525bn) and Spain (€200bn), as shown in Exhibit 24.
This arises from the following sources:
1) We believe that Italy can raise at most €125–150bn
from one-off measures such as wealth taxes,
privatisations, and sales of gold and FX reserves
over the next 12 months.21 Please see Appendix-2
for our calculations. We assume that anything Spain
21 Because we give Italy the benefit of the doubt by achieving a high
rate of privatisation in the first year, we assume potential privatisationrevenues in the subsequent 1-2 years are negligible.
Exhibit 23: Not counting bills, Italy and Spain have gross funding needsover the next three years of just under €900bn…Yearly Italian and Spanish funding needs; includes bond redemptions and general
government budget deficits*; €bn2012 2013 2014 Total
Italy 233 171 134 539
Spain 113 111 103 327
Italy + Spain 347 282 237 866 * We assume T-bills continue to be rolled. Bond redemptions include both conventionaland non-conventional bond redemptions. Budget deficits based on J.P. Morgan basecase expectations.
Exhibit 24: …while the maximum funding available from varioussources is just over €700bn, or about €150bn short of their 3Y fundingneedsMaximum funding available to Italy and Spain from private and official sources; €bn
One-off wealthtaxes, etc.*
SWFs** IMF*** EFSF^
Italy 135 40 210 140 525
Spain -- 20 110 70 200
Italy + Spain 135 60 320 210 725
Private sources
Total
Official sources
* Includes one-off wealth taxes, privatisation, and sale of gold and FX reserves. Pleasesee Appendix-2 for details.** We estimate sovereign wealth funds would be willing to contribute no more than
€60bn to purchases of peripheral sovereign debt. Please see Appendix-3 for details.*** Please see “IMF 101”, Global Fixed Income Markets Weekly , 21 October 2011 for discussion of IMF lending capacity.^ Please see Overview, Global Fixed Income Markets Weekly , 14 October 2011 and28 October 2011 for discussion of EFSF spare capacity and leverage prospects. Weassume that €123bn of the EFSF’s current unallocated capacity is allotted to Greece
for its revised bailout package and that €61bn will be used for a second bailout of Ireland and Portugal in 2H12.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 14/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Kedran Panageas (44-20) 7777-0326J.P. Morgan Securities Ltd
14
manages to raise from one-off measures will be
small and likely absorbed by further bank
recapitalisation costs.
22
2) We believe that sovereign wealth funds (SWFs) will
be willing to contribute no more than €60bn, or
roughly one-third of the amount currently allocated
to European fixed income in their portfolios. Please
see Appendix-3 for our calculations. Of this, we
split the contribution roughly 2:1 between Italy and
Spain.
3) We believe that the IMF can allocate no more than
€320bn to Italy and Spain combined. Even that
would require upsizing IMF capacity through
exercising New Arrangements to Borrow (NABs)and allocating a huge percentage of available
capacity to the two countries.23 It would also likely
be unpopular politically and might require quid pro
quos such as a reshuffling of voting rights. Of this,
we split the contribution roughly 2:1 between Italy
and Spain.
4) Finally, we do not think that the EFSF will be able to
meet its leverage targets, for reasons discussed
previously.24 After allocating funds for Greece’s
second bailout package and assuming another €60bn
is required to support Ireland and Portugal in
2H12,25 only €210bn of spare EFSF capacity would be left. This assumes France manages to keep its
AAA rating (France provides roughly one-third of
EFSF AAA guarantees and hence one-third of EFSF
issuance capacity). We also think it is unlikely that
the European Stability Mechanism (ESM), the
€500bn planned successor to the EFSF, will be
brought forward early and allowed to run
concurrently, as has been mooted in the media. This
would require greatly increasing the German bailout
contribution, which goes against Chancellor
Merkel’s promises to the German parliament and
would likely be challenged by the German
constitutional court.26
22 Also, note that Spain owns considerably less gold and FX reserves
than Italy does. 23 Please see “IMF 101”, Global Fixed Income Markets Weekly, 21
October 2011.24 See Overview, Global Fixed Income Markets Weekly, 14 October
2011 and 28 October 2011.25 See Overview, Global Fixed Income Markets Weekly, 8 July 2011. 26 See Overview, Global Fixed Income Markets Weekly, 9 September 2011.
Under these assumptions, total available funding of
€725bn is sufficient to fund Italy and Spain for a little
over two years. This may be enough to buy some time,
but is unlikely to keep yields in check. In the long run,
we believe that the ECB will eventually need to step in to
stabilize markets by monetizing EMU debt.
The ECB’s role – How much would the ECB need to
buy to stabilise markets?
Most observers, including ourselves, believe that the
crisis has progressed to such an extent that the ECB is
now the only institution remaining with enough
firepower to short-circuit the contagion and stabilise
markets. However, it is reluctant to do so as it would
mean socialising a large amount of debt, exacerbating
moral hazard, and potentially damaging its credibility by
blurring the line between fiscal and monetary policy.
Additionally, debt monetisation is not a long-termsolution: there is no guarantee that debt capital markets
will reopen for peripheral sovereigns once the ECB stops
buying.
In this section, we assess just how active the ECB would
need to be in order to deliver ‘market stability’, and
whether they would be able to intervene to the extent
required. We focus on the immediate pressures facing the
market, specifically in the next one year. As mentioned
in the previous section, Italy and Spain have combined
Exhibit 25: Selling pressure on Italian and Spanish bonds could be aslarge as €650-700bn…Estimated secondary market selling pressure in Italian and Spanish bonds that mature
in 2013 or after; €bn
Investor type
Fraction
held (%)
Par value
held (€bn)
Fraction
(%)
Par value
(€bn)
Domestic 55% 675 20% 135
Non-domestic 45% 552 66% 364
All -- 1227 -- 499
Domestic 62% 268 20% 54
Non-domestic 38% 165 66% 109
All -- 433 -- 162
Domestic 57% 943 20% 189
Non-domestic 43% 717 66% 473
All -- 1660 -- 662
Debt stock that
matures post-2012*
Amount of debt
sold
Italy + Spain
Spain
Italy
* We include all conventionals, international bonds, linkers, and zero-coupon andfloating rate notes that mature in 2013 or later. We assume that 1) 2/3rd of all non-domestic investors and 1/5th of all domestic investors wish to sell their holdings of Italian/Spanish paper and 2) T-bill markets remain open to Italy and Spain. Domesticand non-domestic ownership shares taken from Exhibit 8 above.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 15/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Kedran Panageas (44-20) 7777-0326J.P. Morgan Securities Ltd
15
primary market funding needs of €350bn over the next
year, and we believe that demand for new Italian and
Spanish bonds will be limited. In addition, we estimate
between €650-700bn of secondary market selling
pressure over the next 12 months (Exhibit 25), based on
the following assumptions:
1) €1.6tn of Italian and Spanish bonds outstanding that
mature in 2013 or later,…
2) …total weighted-average non-domestic ownership
of 43%, and…
3) …an estimated sell-rate of 2/3rds for non-domestic
investors and 1/5th for domestic investors.27
Combined, this nets to over €1tn of funding demand/
selling pressure over the next year. Given the private
sector resources available to Italy and Spain of roughly
€135bn, we estimate that the supply/demand imbalance
would be nearly €900bn in Italian and Spanish paper
over the next year (Exhibit 26). This equates to around
€75bn per month of supply/demand imbalance.
Even if Italian and Spanish gross issuance needs were
funded from other sources, the supply/demand imbalance
would still be around €450-500bn of Italian and Spanish
27 The estimate of investors looking to sell is purely a J.P.Morgan
expectation, but we believe that our assumed sell rates are reasonablegiven anecdotal investor discussions and the empirical pattern of the
past few months: Italian yields have risen over 200bp since August,despite sizable ECB buying.
paper over the next year, or roughly €40bn per month
pre-SMP buying.
To put these numbers in context, we note that the ECB
has been purchasing about €35bn of peripheral paper per
month since August. While the ECB does not identify
individual countries, we assume the split has been
roughly 3:1 between Italy and Spain. Over this same time
period, net issuance has actually been negative for Italy
(Exhibit 27). Despite the negative net issuance and
Exhibit 26: …suggesting that the supply/demand imbalance in Italian/Spanish paper over the next 12 months would be around €75bn/month, or animbalance of around €40bn/month assuming that the SMP continues its current rate of purchase of €35bn/monthEstimated supply/demand imbalance* in Italian and Spanish bonds over the next one year, with and without IMF/EU/SWF commitments**; €bn
Without
IMF/EU/SWF
With
IMF/EU/SWF
Without
IMF/EU/SWF
With
IMF/EU/SWF
Without
IMF/EU/SWF
With
IMF/EU/SWF
Gross issuance / funding needs 233 233 113 113 347 347
Domestic inv estor selling 135 135 54 54 189 189
Non-domestic inv estor selling 364 364 109 109 473 473
One-off w ealth tax es, asset sales 135 135 -- -- 135 135
SWFs -- 40 -- 20 -- 60
IMF/EFSF -- 233** -- 113** -- 347**
€bn total 598 324 276 142 873 467
€bn/month 50 27 23 12 73 39Supply/demand imbalance*
Italy Spain Italy + Spain
Selling pressure on bonds*
Funds available**
* Estimates of funding needs and secondary market selling pressure taken from Exhibits 23 and 25 above. We assume T-bills continue to be rolled.
** Estimates of amounts raised from one-off wealth taxes and SWFs are taken from Exhibit 24 above. However, we assume that the IMF and EFSF contribute just enough tocover gross issuance and funding needs over the next 1Y, rather than providing three years funding in advance.
Exhibit 27: Despite negative net supply in Italy, and significant ECBbuying since August,…Gross bond issuance*, redemptions*, and SMP purchases in Italy and Spain** since 8
August 2011; €bn
52
21
61
14
92
31
0
10
20
30
4050
60
70
80
90
100
Italy Spain
Gross issuance
Redemptions
SMP purchases**
* Excluding T-bills. Bond redemptions include conventional and non-conventional bondredemptions.** The SMP has purchased a total of €123bn of peripheral paper in the roughly 3 1/2months since they initiated purchases on 8 August 2011. We assume that these SMPpurchases are split 3:1 between Italy and Spain.Source: Bloomberg, ECB, J.P.Morgan
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 16/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Kedran Panageas (44-20) 7777-0326J.P. Morgan Securities Ltd
16
sizable ECB buying, Italian yields have risen over 200bp
since August (Exhibit 28). This indicates the strength of
the selling pressure currently going through thesecondary markets.
Will the ECB be willing to double its current rate of
purchases to eliminate the supply/demand imbalance?
We make several points here. On the one hand, there are
no hard capacity constraints that would prevent it from
doing so.28 On the other hand, while the ECB has
increased its balance sheet less than other central banks,
it has arguably taken on a good deal more credit risk. The
lion’s share of the growth in the ECB’s balance sheet
stems from increased support for peripheral country
banks and peripheral sovereigns (please see Appendix-
4). Finally, based on recent official commentary, thereappears to be growing reluctance to expand the size of
SMP purchases. The German contingent remains firmly
opposed to this and insists that monetary financing to
sovereigns is against the ECB’s mandate. Further, recent
media reports suggest that this view is spreading. For
instance, there is reported to be a limit on purchases of
no more than €20bn per week. This is said to be a
reduction from the original limit, and a limit that the
ECB Council discussed lowering further, as late as last
week.29 To be sure, such limits are self-imposed and can
always be changed if market conditions change.
However, it appears that at least over the near term,
the ECB remains reluctant to step up the size of its
buying.
Going forward, our baseline view is that the ECB will
not be willing to purchase debt in large quantities,
unless the market forces its hand. This is due to 1) the
need to maintain oversight and aid conditionality in order
to mitigate moral hazard; 2) the desire to avoid blurring
monetary and fiscal policy; and 3) a reluctance to take
any steps that might damage credibility or independence.
The 2012 outlook
The long-term solution to the EU sovereign crisis isconsistent fiscal and structural reform, better governance
in the Euro zone, more fiscal integration, and potentially
jointly guaranteed debt issuance. While some small steps
28 The only hard restriction is that the ECB cannot purchase bonds
directly from a sovereign (i.e. on the primary market) or set up credit
facilities for sovereigns.29 See ‘Upper limit for bond purchases/Obergrenze fur Anleihekaufe’,
Frankfurter Allgemeine, 17 November 2011; and ‘ ECB hits back at
calls for intervention’, FT, 18 November 2011
have been taken, these efforts will take far too long to
bear fruit in order to deliver market stability in 2012.30
In the near term, as in all negative feedback loops or
crises of confidence, outside intervention is required
in the form of a lender of last resort. With the ECB
reluctant to play that role, the path of least resistance
is for the crisis to worsen. Thus, our baseline view is
that the crisis stays on a slow boil over the next fewmonths, possibly resulting in another country losing
market access. In the event that such a risk plays out, we
believe that the ECB will be compelled to start
monetizing debt since the alternative would be
significantly worse.
There are a large number of risks that could go wrong in
other countries as well (Exhibit 29). These include, for
instance, the prospect of
• another round of PSI in Greece,…
• …a hard default in Greece,…
• …potential PSI in Ireland and Portugal,…
• …another country needing to tap the EFSF,…
• …and a French downgrade.
30 See Overview, Global Fixed Income Markets Weekly, 19 August2011, for discussion of the process required to institute Eurobonds.
Exhibit 28: …peripheral yields have continued to rise after initiation of the SMP programPortuguese and Italian 10Y yields around the s tart of ECB bond purchases for eachcountry*; %
4.00
5.00
6.00
7.00
8.00
-3 -2 -1 0 1 2 3 4 5
Months around first day of SMP purchases*
Italy
Portugal
* The SMP started buying Portuguese and Italian bonds on 10 May 2010 and 8 August2011, respectively. Portugal was eventually bailed out by the IMF/EFSF on 3 May2011.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 17/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Kedran Panageas (44-20) 7777-0326J.P. Morgan Securities Ltd
17
Given the large number of risks on the horizon, one or
more of these threats are likely to play out in 2012. The
combination of a steady stream of negative news and
market disappointments will keep liquidity poor, spreads
wide, and investors skittish.
Trading strategies
In our view, the crisis is likely to worsen in 1H12,
before the ECB feels compelled to step in and stabilize
markets by monetizing debt. We therefore expect
peripheral yields to continue to climb going into 1H12,
especially if another country loses market access. Note
that bond yields of a country that loses market access do
not immediately stabilise; as demonstrated in Exhibit 30,
official intervention has thus far provided only a
temporary respite. Thus, short positions in peripheral
debt have been profitable even when initiated post- bailout.
Additionally, in case of loss of market access, the debt of
a country is likely to be downgraded which would
exacerbate the selling pressure. Recent rating agency
commentary around Italy and Spain, and the pattern of
Greece, Ireland, and Portugal when they lost market
access, suggests that Italy would be downgraded if it lost
market access (Exhibit 31). Indeed, sovereign CDS
spreads are trading significantly wider than their
similarly-rated corporate brethren, suggesting that the
risk of further sovereign downgrades is high (Exhibit
32).
Of course, the ECB will be quick to realize the severity
of the situation if a large country loses market access,
suggesting that this will be the tipping point where it will
be compelled to start monetizing sovereign debt in large
Exhibit 29: The number of risks on the horizon is extremely large; we expect the peripheral crisis to escalate in 1H12Risks in 2012, and possible policy responses / positive surprises
Effect Risk
Greece fails to get further funding, leading to a hard default and potential ex it from the Euro zone
Italy and/or Spain lose access to capital markets
Italian coalition collapses, leading to early elections
More sev ere PSI in Greece, triggering contagion fears in Italy and Spain
French downgrade
Ireland / Portugal require PSI
Another country requires EFSF assistance (such as Cyprus)
Italy drags its feet on fiscal consolidation / structural reforms
Bank deleveraging depresses economic grow th
Euro area in deep recession prev ents planned fiscal c onsolidation
LCH continues to boost margin requirements, leading to further delev eraging in peripheral paper
ECB w illing to monetise debt in large quantities
Germany agrees to fund ESM alongside EFSF
Significant new money commitments from IMF/BRICs/SWFs
Upside surprise on economic grow th
Upside surprise on implementation of austerity measures
Potential policy
responses / positive
surprises
Risks
Severe
Significant
Long term
Exhibit 30: Bond yields in bailed-out European countries havecontinued to soar even after the bailout request
Average of Greek, Irish, and Portuguese 10Y benchmark yields around the date whenbailout was first requested*; %
months around bailout request*
6.00
7.00
8.00
9.00
10.00
11.00
-3 -2 -1 0 1 2 3
* Greece, Ireland, and Portugal formally requested a bailout on 2 May 2010, 21
November 2010 and 6 April 2011, respectively.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 18/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Kedran Panageas (44-20) 7777-0326J.P. Morgan Securities Ltd
18
quantities. We expect that the eventual response from
the ECB, IMF and EU will be on a scale sufficient to
ultimately stabilise peripheral yields in 2H12,unlike
the limited attempts made thus far with Greece, Ireland,
and Portugal (Exhibit 33). We also see 10Y Bund yields
falling to 1.25% by 2Q12 in a flight-to-quality, before
rebounding to 1.75% by end-2012. Thus, we
recommend underweighting peripheral duration and
buying German duration going into 1H12.
Below, we also present three different types of trading
strategies that allow investors to take advantage of our bearish view on both the European economy and the
peripheral situation. These strategies, in turn, offer 1)
significant upside in a global recession, 2) positive
convexity to peripheral risk, and 3) diversification away
from the peripheral sovereign debt crisis.
1. Identifying the best long-duration trades in a low-
yield world
Valuations for fixed income are clearly stretched in the
major markets. For instance, US, UK, and European 10Y
real yields are currently below zero, and central bank
policy rates are near lower bounds. Thus, we look for global markets where long duration positions offer more
upside, without compromising too much on credit risk. In
Exhibit 34, we highlight sovereigns that are rated A+ or
above and, at the same time, offer positive real yields,
attractive policy rate downside, and strong economic
linkages to Europe (based on 10Y regression of annual
nominal GDP growth). As a simple metric of
attractiveness, we average each country’s cross-sectional
z-scores across the three measures listed above. This
captures countries that are most likely to be impacted by
the slowdown in growth in Europe and where
policymakers have the most leeway to respond by
slashing rates. Based on this metric, long Australian
and Chilean 10Y duration appear to be the most
attractive trades. We particularly like longs in
Australia: the RBA is expected to cut rates and domestic
technicals are strong (see Australia).
Exhibit 31: Rating agencies have stated that Italy will be downgradedfurther if it loses access to debt capital markets. We expect another three-notch downgrade (from A to BBB) in that event
Number of notches that bailed-out sovereigns were downgraded by Moody’s in the 3months prior and 6 months after the bailout request*
-3M to aid
request
Aid request
to +6M
Greece 1 4 14
Ireland 0 7 9
Portugal 3 4 9
Average 1 5 11
Italy 3 - 3
Downgrade (notches) Total downgrade
since Jan-10
(notches)
* Italy is currently rated A2 by Moody’s and A with a negative outlook by S&P.Source: Bloomberg, Moody’s, S&P
Exhibit 32: Indeed, sovereign CDS spreads are trading significantlywider than their corporate brethren, suggesting that the risk of further sovereign downgrades is high
CDS spreads of iTraxx components regressed against corporate Moody’s rating vs.CDS spreads of sovereigns regressed against Moody’s rating*; triangles showsovereigns and circles show corporates; bp
0
500
1000
1500
2000
Moody's rating
IE
PT
IT
ES
Aaa Aa2 A2 Baa2 Ba2 B2 Caa2
* Using a unit scale for Moody’s rating (AAA=1, Aa1 = 2, Aa = 3, etc.), we model thefollowing relationships based on current cross-sectional data:
5Y Sov. CDS = 76.6*(Moody's) + 38.7; R2 = 88%Sample based on 16 developed market sovereigns (US, UK, Japan, Australia,Sweden, Denmark, and 10 Euro area countries excluding Greece).
5Y Corp. CDS = 7.2*(Moody's)^2 - 73.8*(Moody's) + 358.9; R2 = 66%Sample based on nearly 200 corporates from the iTraxx High-grade, High-yield, andCross-over indices.
Source: Moody’s, Bloomberg, Markit
Exhibit 33: Spreads and yield forecast: We expect the crisis to getworse initially, before concerted policy action pushes peripheral yieldslower in 2H12J.P. Morgan 2012 forecast for 10Y yields and benchmark spreads to Germany
18-Nov 2Q12 4Q12 18-Nov 2Q12 4Q12
Germany -- -- -- 1.97 1.25 1.75
France 147 225 190 3.46 3.50 3.65
Greece 2425 2400 2400 26.32 25.25 25.75
Ireland 625 850 900 8.15 9.75 10.75
Italy 494 775 625 6.95 9.00 8.00
Portugal 910 1200 1300 11.04 13.25 14.75
Spain 469 700 600 6.70 8.25 7.75
Wtd av g., peripherals* 699 955 855 -- -- --
Spread to Germany (bp) Yield (%)
* Weighted by proportional contribution to the J.P. Morgan EMU Bond Index.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 19/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Kedran Panageas (44-20) 7777-0326J.P. Morgan Securities Ltd
19
2. Identifying trades that are positively convex in a
risk-off (or risk-on) world
Given the bimodal distribution of risks, trades that offer
positive convexity to the crisis are attractive. We define a
positively convex ‘risk-off trade’ as one which performs
well if the crisis worsens but does not lose much if the
crisis stabilises. Similarly, we define a positively convex
‘risk-on trade’ as one which performs well if the crisis
abates but does not lose much if the crisis worsens.
To identify such trades, we look at a wide selection of global assets. We regress asset levels against 10Y
weighted peripheral spreads since 2010. Assets that have
a sufficiently strong and convex relationship are
presented in Exhibits 35–36. Specifically, these are
trades where the R-squared exceeds 65% and the t-
statistic on the squared peripheral spread term exceeds
+/- 4. Using the regression betas, we project gains and
losses to the recommended position from a +/-100bp
shock to peripheral spreads. The convexity in each
position is defined as the additional gain that one makes
in the targeted scenario due to the convex performance
profile, divided by the average absolute trade P/L for a
+/-100bp move.
For risk-off positioning, we find that Japanese 2Y swap
spread wideners, long Aussie duration, and short France
vs. Germany offer the greatest positive convexity
(Exhibit 35). However, long Aussie duration is our
preferred trade in that it is amongst the very few that
offer positive carry and it is expected to benefit from
policy rate cuts and strong technicals (see Australia).
For risk-on positioning, short Kiwi 2Y duration, UK 10Y
swap spread narrowers, and short gold offer the greatest
positive convexity (Exhibit 36). As before, there are few
assets that simultaneously offer positive carry and
positive convexity. One that does is long US municipal
credit.
Exhibit 34: Look to global markets for higher real yields, potential policy rate cuts, and high sensitivity of GDP growth to Europe. Long 10Yduration in Australia and Chile appear to be the most attractive tradesStatistics on global benchmark nominal and real yields, central bank policy rates, and beta of GDP growth to Europe; %
Continen t C ountr y 18-Nov 2H12 FC – 18-Nov L10Y, % YoY Beta to Europe Corr to Europe
Americas Chile AA 4.8 3.1 1.7 5.25 -1.25 10.0 1.3 66% 1.5
Americas Canada AAA 2.1 2.1 0.1 1.00 0 4.9 1.3 90% 0.0
Americas USA AA+ 2.0 2.5 -0.5 0.13 0 4.2 1.0 91% -0.7
Europe Czech AA 3.9 2.4 1.4 0.75 0 5.1 1.1 84% 0.5
Europe Sw eden AAA 1.7 1.8 -0.05 2.00 -0.75 4.1 1.1 91% -0.1
Europe Germany AAA 2.0 2.0 -0.04 1.25 -0.75 3.2 1.0 100% -0.3
Europe Denmark AAA 2.0 2.1 -0.2 1.20 -0.75** 3.0 1.0 86% -0.3
Europe UK AAA 2.3 2.5 -0.2 0.50 0 4.1 1.0 92% -0.5
Europe Sw itzerland AAA 0.9 0.8 0.1 0.13 0 2.9 0.8 81% -0.6
Asia/Pacific Australia AAA 4.1 2.9 1.2 4.50 -0.75 7.0 0.8 86% 0.6
Asia/Pacific China AA- 3.7 2.7 1.0 6.56 0 14.8 0.2 16% 0.3
Africa/ME Israel AA- 4.1 2.4 1.7 3.00 -0.50 5.7 0.3 26% 0.1
Asia/Pacific New Zealand AA+ 4.0 2.9 1.0 2.50 0 5.3 0.6 80% 0.1
Asia/Pacific Japan AA- 1.0 -0.2 1.2 0.05 0 -0.3 0.9 89% 0.0
Asia/Pacific Taiw an AA- 1.3 1.0 0.3 1.88 0 3.8 0.8 58% -0.2
Asia/Pacific S Korea A+ 3.8 3.3 0.5 3.25 0 7.1 0.4 43% -0.3
Av erage -- 2.7 2.1 0.6 2.12 -0.30 5.3 0.8 74% --
Average Z-
score*
Avg nominal GDP growthPolicy Rate (%)Rating
(S&P)
10Y Yield
(%)
Avg Inflation,
L10Y (% YoY)
Real yield
(%)
* Defined as the average of three cross-sectional z-scores: 1) 10Y real yield 2) level of central bank policy rate and 3) 10Y GDP beta to Europe. Z-score defined as (level – cross-sectional sample average)/cross-sectional sample standard deviation.** J.P.Morgan does not forecast central bank rates for Denmark but here we assume that they cut rates in line with the ECB, as is their typical pattern.Note: Yields shown for benchmark government bonds. Real yield defined as nominal yield – realised annual headline inflation over the past 10 years. Beta and correlation of GDPgrowth to Euro area GDP growth based on a 10Y regression of annual nominal GDP growth rate data.
Source: Bloomberg, J.P.Morgan
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 20/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Kedran Panageas (44-20) 7777-0326J.P. Morgan Securities Ltd
20
Exhibit 35: Trades that offer the greatest positive convexity to peripheral spread widening include 1) Japanese 2Y swap spread wideners; 2) longAussie duration; and 3) short France vs. Germany. Long Aussie duration offers positive carry as wellGlobal ‘risk-off’ trades (trades with a positively convex performance profile when peripheral spreads are widening)
Class Region Item
Trade
Positioning Curr R2 * Beta* Beta_sq*
Risk-
off**
Risk-
on**
Difference
(Risk-off - Risk-on)***
%
Convexity^^
Yields Australia/NZ Aussie 2Y Long duration 320 65% 26.1 -6.6 67 -54 13 22%
Yields Australia/NZ Aussie 10Y Long duration 405 71% -9.3 -2.1 39 -35 4 12%
Sw ap spreads UK UK 2Y SS Wideners 96 75% -3.5 1.2 14 -11 2 20%
Sw ap spreads Japan JPY 2Y SS Wideners 26 66% -8.8 0.9 4 -2 2 65%
EMU spreads Europe France - Germany Wideners 147 90% -7.6 3.1 35 -29 6 19%
Credit Europe Itrax x Wideners 189 76% 6.8 1.5 27 -24 3 11%
Credit Europe Maggie Banks Senior Wideners 216 75% 1.5 2.3 34 -29 5 15%
Credit UK UK Fins Senior Wideners 315 71% -6.0 3.9 49 -41 8 17%
Credit Asia Itrax x Asia Wideners 212 65% -11.0 3.8 42 -35 8 20%
Equities Europe EU banks Short 125 82% -4.8 -1.7 29 -26 3 13%
Equities Europe Italy MIB Short 15233 72% -589.2 -109.3 2120 -1901 219 11%
Regression of asset against
wtd per iph. spreads^ P/L for +/-100bp change in wtd per iph. spreads^
* Trades based on polynomial regression vs. 10Y weighted peripheral spreads since January 2010. Trades shown have R-squared >65% and statistically significant convexity(T-stat vs. squared peripheral spread term > +/- 4). Beta = the regression beta vs. 10Y weighted peripheral spreads measured in % and Beta_sq = the regression beta vs. 10Yweighted peripheral spreads^2, again measured in %. A positive (negative) value of beta_sq indicates a convex (concave) asset.** Modelled gain (loss) for a +100bp (-100bp) spread move, assuming peripheral spreads at 650bp. Gains and losses are calculated with respect to the recommended trade (e.g.reversing the betas where appropriate, or where beta_sq<0).*** Defined as the absolute value of gain in the risk-off scenario minus absolute value of gain in the risk-on scenario.^ 10Y weighted peripheral spread computed against Germany for Greece, Ireland, Portugal, Italy, and Spain (weighted by the size of their outstanding bond market).^^ % Convexity defined as the difference in the absolute value of gains divided by the average of the absolute values of gains in both the risk-off and risk-on scenarios.Units: yields, curve, swap spreads, credit spreads shown in bp of yield. Equities and commodities shown in index points. FX are in ratios.
Exhibit 36: Trades that offer the greatest positive convexity to peripheral spread narrowing include 1) short Kiwi 2Y duration 2) UK 10Y swap
spread narrowers; and 3) short gold. Long US muni credit is one of the few that offers positive carry as wellGlobal ‘risk-on’ trades (trades with a positively convex performance profile when peripheral spreads are narrowing)
Class Region Item
Trade
Positioning Curr R2 * Beta* Beta_sq*
Risk-
off**
Risk-
on**
Difference
(Risk-on - Risk-off)***
%
Convexity^^
Yields Australia/NZ Kiw i 2Y Short duration 250 78% -72.5 4.8 -6 15 10 90%
Curv e UK 2s/10s Steepeners 178 75% -30.0 1.3 -12 14 3 20%
Sw ap spreads UK UK 10Y SS Narrow ers 31 83% 12.4 -0.7 -3 4 1 41%
Sw ap spreads Scandis Denmark 10Y SS Narrow ers 65 80% 12.6 -0.7 -3 4 1 36%
Credit USA 10Y Muni Narrow ers 216 82% 7.9 -0.3 -3 4 1 18%
Commodities Global Gold Short 1720 89% 225.0 -13.1 -41 67 26 49%
FX FX CHF/HUF Short CHF 246 89% 20.6 -1.2 -3 6 2 53%
Regression of asset against
wtd periph. spreads^ P/L for +/-100bp change in wtd periph. spreads^
*, **, ^,^^: See footnotes in Exhibit 35 above. *** Defined as the absolute value of gain in the risk-on scenario minus absolute value of gain in the risk-off scenario.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 21/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Kedran Panageas (44-20) 7777-0326J.P. Morgan Securities Ltd
21
3. Identifying trades that are orthogonal to the
peripheral debt crisis and therefore offer
diversification benefits
One of the difficulties of investing in the current market
environment is that return correlation is extreme. Indeed,
one could be forgiven for assuming that we live in a one-
factor world (with that factor being sovereign risk).
Trades that are relatively uncorrelated with peripheral
spreads can add value in a portfolio context, especially if they carry positively. We screen nearly 200 asset classes
globally and find that just a few, in fact around 5% of
those sampled, demonstrate a significant lack of
correlation. Specifically, Exhibit 37 highlights asset
classes for which the R-squared in a polynomial
regression vs. peripheral spreads is both 1) less than 20%
in 2010-2011, and 2) less than 30% in 2011. Of these,
we note that prime RMBS and Auto ABS in the US carry
positively. Moreover, J.P.Morgan strategists expect
spreads in these sectors to be stable to firmer in 2012.
Exhibit 37: The elusive quest for diversification from the peripheral crisis in Europe: US prime RMBS and Auto ABS carry positively and spreadsare expected to be stable to firmer in 2012Trades with low correlation* to EU peripheral spreads
USA:
Prime RMBS; AAA Auto ABS; CDS-cash basis;
UK:
Rate Vega
Japan:
10s/30s;
Canada:
5s/10s; EUR/CAD
Denmark:
2s/10s;
Brazil:
Real/CLP
Korea:
USD/KRW; EUR/KRW;
Notes:* Trades based on polynomial regression vs. 10Y weighted peripheral spreads since January 2010, as described in footnotes in Exhibit 35 above. Trades shown here have R2 <20% over entire sample period and R-squared < 30% since January 2011. Around 5% of the nearly 200 trades sampled meet these criteria.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 22/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Kedran Panageas (44-20) 7777-0326J.P. Morgan Securities Ltd
22
Trading themes
• Be long German and UK duration; enter 2s/10s
curve flatteners
We expect the peripheral crisis to escalate in 1H12,
driving peripheral yields higher and German and UK
yields lower in a flight-to-quality dynamic. Further, a
Euro area recession, slow growth in the UK, and
disinflation will add downwards pressure to
intermediate yields, flattening the 2s/10s curve.
• Underweight intra-EMU paper vs. Germany
We expect intra-EMU spreads to widen as the
peripheral crisis escalates in 1H12. Also, ratings
downgrades and poor technicals will contribute to
upwards pressure on peripheral yields.
• Go long 10Y duration in Australia and Chile
Global markets that offer high real yields, potential
policy rate cuts, and high sensitivity of GDP growth to
Europe are attractive. Long Australian and Chilean
10Y duration appear the most attractive; we
particularly like longs in Australia given advantageous
technicals and expected central bank rate cuts.
• Position in trades that are positively convex to
peripheral spread widening
Trades that perform well when peripheral spreads arewidening but give up little when they are narrowing
are attractive ‘risk-off’ trades. We find that Japanese
2Y swap spread wideners, long Aussie duration, and
short France vs. Germany offer the greatest positive
convexity to peripheral spread widening . Long Aussie
duration is also attractive as it is one of the few trades
which offer positive carry as well. On the other hand,
short Kiwi 2Y duration, UK 10Y swap spread
narrowers, and short gold offer the greatest positive
convexity to peripheral spread narrowing .
• US Prime RMBS and Auto ABS offer
diversification away from the peripheral crisis
We screen nearly 200 asset classes globally and find
that just a few, around 5%, demonstrate a significant
lack of correlation to the sovereign debt crisis. Of
these, we note that US prime RMBS and Auto ABS
carry positively; J.P.Morgan strategists expect spreads
in these sectors to be stable to firmer in 2012.
Exhibit 38: Bird’s eye view of our major trade recommendations by currencyEuro area UK US Japan
Duration Long 10Y Long 10Y Long 10Y Long bias
Long 6Mx6M EONIA
Curve 2s/10s flatteners 2s/10s flattener Long end steepening bias Swap curve flattening bias
1s/5s bull flatteners 10s/30s flattening bias
Swap spreads 2Y wideners 10Y wideners 10Y wideners Short end wideners
FRA/OIS w ideners Long end narrow ers
TIBOR/LIBOR narrowers
3s/6s w ideners
Swap spread curve Flattening bias Neutral Flattening bias
Gamma Long gamma in 10Y tails Long gamma in 10Y tails Long Neutral
Short gamma in 2Y tails Short gamma in 2Y tails
Vega Neutral Neutral Short Neutral
Inflation 1s/10s HIC P sw ap curv e flattener Long intermediate real yields Short breakev ens N eutral
Long French CPI-linked vs. Euro HICP linked Short 10Y breakevens
Cross-market Wider intra-EMU spreads
Underweight Italy and France vs. Germany
Credit curve flatteners
Short Euro vs. US breakevens
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 23/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Kedran Panageas (44-20) 7777-0326J.P. Morgan Securities Ltd
23
Appendix-1: Timeline of key events in the EU sovereign debt crisis2009
04-Oct-09 George Papandreou leads Socialist Pasok Party to landslide victory in Greek elections
20-Oct-09 New Greek Finance Minister Papaconstantinou says deficit will rise to 12.5 percent of GDP this year, triple the prior forecast
2010
13-Jan-10 European Commission report accuses Greece of statistics fraud; new Finance Minister says: "There is no skeleton in the closet"
14-Jan-10 Greece adopts three-year plan to bring the European Union’s biggest budget deficit to w ithin the EU limit in 2012
02-Feb-10 Greek government announces austerity package to get deficit to 3 percent of GDP in 2012
11-Feb-10EU leaders hold first emergency summ it on Greece. EU agrees to take “determined and c oordinated action” to protect financial stability of euro area, w ithout
giv ing further details
08-Mar-10 Portuguese government announces new budget cuts, more asset sales and a freeze on public wages
16-Mar-10 Euro region finance ministers lay groundwork for making emergency loans av ailable to aid Greece
18-Mar-10Papandreou calls on EU partners to come up w ith specific aid measures w ithin a week to help Greece, hints he m ight seek support from IMF if EU partners
don’t act
24-Mar-10 Fitch cuts Portugal’s credit rating to AA-
25-Mar-10Trichet say s that the ECB w ill continue to accept bonds rated as low as BBB- as collateral. An EU summit agrees to general principles behind a Greek bailout,
including IMF inv olvement, but again provides no details30-Mar-10 Ireland says country ’s banks need to raise an additional €31.8bn of capital
12-Apr-10 Euro area finance ministers agree to provide up to €30bn of loans to Greece over the next y ear with the IMF agreeing to put up another €15bn in funds
21-Apr-10 The EU says Greece’s 2009 budget deficit was w orse than it previously forecast and could top 14 percent of GDP as “off- market swaps” c loud its estimates
23-Apr-10 Papandreau asks the EU for aid
27-Apr-10 S&P cuts Greece rating to junk (BB+), downgrades Portugal to A-
02-May-10 Euro area agree to upsize its rescue package to Greece to €110bn. Greece agrees to €30bn in austerity cuts over the next 3 years in ex change for the aid
03-May-10 The ECB says it will indefinitely accept Greek collateral regardless of the country’s credit rating
05-May-10Protests in Athens against the gov ernment’s austerity plans turn violent and 3 people are killed w hen they bec ome trapped in a bank set ablaze by
demonstrators
06-May-10 Flash crash in US stocks
07-May-10European leaders agree in principle to set up a €440bn emergency temporary fund to stem the sovereign crisis, w hich will com e to be know n as European
Financial Stability Facility (EFSF)
07-May-10 In Spain, the Government and the main opposition party reach agreement over mergers betw een cajas and the need of sav ings-bank regulation reform
10-May-10
EU finance chiefs agree to establish a €750bn permanent aid facility to succ eed the EFSF (the European Stabilisation Mechanism , or ESM). Also, the ECB 1)
extends unlimited € liquidity for banks, 2) re-opens $ sw ap lines w ith the Fed, and 3) announces a gov ernment bond purchase program to focus on peripheral
sov ereign debt (the SMP). This is the biggest attempt y et to solv e the peripheral crisis
12-May-10 Spain announces public-wage cuts and a pension freeze
18-May-10 Germany surprises the markets with a ban on naked short selling of debt securities, CDS and shares in 10 financial institutions.
21-May-10 German parliament approves its share of the EFSF commitments
26-May-10 Italian Prime Minister Silvio Berlusconi's ministers approve €24bn in budget cuts for 2011-2012
28-May-10 Fitch cuts Spain’s AAA rating one level to AA+
14-Jun-10 Moody’s cuts Greek rating to junk (Ba1)
15-Jun-10 Chinese sign multibillion euro contracts with Greece hours after Moody 's debt downgrade
16-Jun-10 The French government Wednesday unv eiled a controversial plan to gradually raise the age at which workers can retire with pension benefits
23-Jul-10 Europe publishes the results of the first round of bank stress tests. 7 banks failed with a total capital shortfall of €3.5bn
18-Oct-10
German Chancellor Angela Merke l and French President Nicolas Sarkozy agree that priv ate investors may be required to contribute to future EU bailouts v ia a
mandatory solv ency rev iew of countries requesting aid; they als o declare that official loans granted out of the ESM after 2013 should hav e preferred creditor
status over private debt investors
01-Nov-10 Portugal’s gov ernment and biggest opposition party agree to let nex t year’s budget pass, aiming to stem the euro region’s fourth-biggest fiscal shortfall
01-Nov-10Greek minister Pangalos sugges ts to Greek media that: "Demonizing debt restructuring is wrong. Debt exis ts to be restructured. We may pursue it ourselv es or
it may be proposed to us and it may be too adv antageous to turn it dow n."
10-Nov-10 LCH.Clearnet increases haircuts for Irish bonds
16-Nov-10 December aid tranche for Greece is delayed until January, after the Austrian and Finnish ministers say Greece hasn't met EU criteria
21-Nov-10 Ireland says it will apply for a bailout
28-Nov-10 Ireland gets €85bn bailout. European leaders scale back proposals to inflict losses on bondholders in future EU bailouts
23-Dec-10 Fitch cuts Portugal to A+ rating Source: Bloomberg, Reuters, WSJ, FT, NY Times, Telegraph
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 24/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Kedran Panageas (44-20) 7777-0326J.P. Morgan Securities Ltd
24
Timeline of key events in the EU sovereign debt crisis, continuedDate Event
2011
24-Jan-11 Spain announces new capital requirements for banks
11-Feb-11 Axel Weber resigns from Bundesbank after opposing the ECB’s crisis policy
25-Feb-11Fine Gael wins elec tions and knocks out the current party from the Irish government, establishing a coalition with Labour. Backtracks from demands for bank
debt burden-sharing but continues its demands for lower interest rates on official loans
10-Mar-11 Bank of Spain announces €15bn capital shortfalls in national bank stress tests
11-Mar-11 EU summit agrees to expand powers of EFSF by 1 ) allowing it to buy debt in primary m arkets and 2) upsizing guarantees to tap its full €440bn in firepow er
15-Mar-11 EcoFin meeting agrees on new Stability & Growth competitiveness pact
21-Mar-11EU finance ministers decide on mec hanisms for allowing the region’s permanent bailout mechanis m, the ESM, to lend €500bn in 2013, Greece's loan rates are
low ered by 100bp and loan maturity is ex tended to 7.5y rs; Ireland's loan terms are unchanged
23-Mar-11 Portuguese prime minister resigns over loss of support stemming from austerity measures
25-Mar-11 Agreement on capital structure and lending terms of the ESM at the EC heads of state meeting
27-Mar-11 Merkel's coalition loses in state elections in Baden-Wurttemberg, narrowing her support in the upper house of Parliament.
31-Mar-11 Irish bank stress tests announce €24bn shortfall across the remaining 4 un-nationalized banks; €5bn will come from sub debt writedowns
01-Apr-11 ECB agrees to drop its ratings requirement on Irish-backed paper at its tender operations06-Apr-11 Portugal requests aid from the EU
15-Apr-11 Papandreou announces €76bn of austerity measures, later increased to €78bn, running through the end of 2015
17-Apr-11 In the Finnish national election, True Finns, a Euro-skeptic party, polls third in a tight 4-way race, taking roughly 20% of seats in the legislature
03-May-11 Portugal reaches agreement on a €78bn aid package
16-May-11 Eurogroup meeting endorses Portuguese bailout package
16-May-11 Eurogroup meeting endorses Draghi as the next ECB chief
17-May-11 European finance ministers propose talks w ith bondholders around extending Greece’s debt-repay ment schedule
24-May-11 Greece releases a new medium-term plan with more aggressiv e fiscal consolidation and privatization targets
01-Jun-11IMF holds off releasing the nex t tranche of Greece's aid package until a new bailout package, w hich meets necc essary funding needs for at least 1 y ear, is
approved
02-Jun-11 The ECB and EU to come to truce around Greek PSI v ia a "Vienna initiative" approach
05-Jun-11Centre-right party w ins conv incingly in Portuguese parliamentary election; mos t Portuguese parties have already signaled their support for required austerity
measures
15-Jun-11 A parliamentary c risis in Greece sends markets into a tailspin. Fac ed w ith party defections, the Greek prime minister calls for a no-confidence vote to be held
on 21 June. EU em phasises that the disbursement of the nex t tranche of aid package requires approv al of all new austerity meas ures
16-Jun-11 The IMF agrees to release the next tranche of Greek aid w ithout formal agreement on a new package in place
17-Jun-11 Merkel and Sarkozy agree in principle to a "Vienna-initiative" debt rollover for Greece
21-Jun-11 Greek prime minister wins no-confidence vote
24-Jun-11 EU finance ministers agree on details around EFSF upsizing and remove preferred creditor status from ESM loans to Greece, Ireland and Portugal
29-Jun-11 Greek parliament approves austerity measures (principles)
29-Jun-11 French banks propose a Greek debt rollover w ith modest haircuts
30-Jun-11 Greek parliament approves austerity measures (implementation measures)
02-Jul-11 EU releases next tranche of Greek aid package
04-Jul-11 S&P announces that the French proposal for the Greek debt rollover would garner an SD rating
06-Jul-11Follow ing rating agency comments that Vienna initiative approach w ould most likely garner an SD rating, German and Dutch officials return to calls for a formal
debt restructuring in Greece. The ECB remains adamantly opposed
07-Jul-11 ECB agrees to drop the ratings requirement on Portuguese-backed paper at its tender operations
08-Jul-11 IMF releases next tranche of Greek aid package
08-Jul-11 Contagion spreads to Italy . Negative headlines around finance minister and banks under stress (Unicredit trading halted)
11-Jul-11Munitions blast at Cy prus nav al base leads to sev eral multi-notch rating downgrades and fears of a potential Cy priot bailout. Sev eral ministers resign and
protests call for the dissolution of the ruling gov ernment coalition
15-Jul-11 2nd round of EBA bank stress test results released; net €2.5bn capital shortfall estimated across 8 banks
21-Jul-11 Details of 2nd Greek aid package announced; IIF banking federation launches a voluntary Greek debt rollover
02-Aug-11 Italian 10Y yield closes above 6%, at highest levels since joining the EMU
05-Aug-11Italian Prime Minis ter Berlusconi announces new reforms and promises to front-load budget cuts in an apparent quid pro quo; the ECB begins buy ing Italian and
Spanish debt the week after
05-Aug-11 S&P downgrades the US to AA+ Source: Bloomberg, Reuters, WSJ, FT, NY Times, Telegraph
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 25/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Kedran Panageas (44-20) 7777-0326J.P. Morgan Securities Ltd
25
Timeline of key events in the EU sovereign debt crisis, continuedDate Event
08-Aug-11 ECB starts buy ing Italian and Spanish debt in its SMP purchase program
16-Aug-11Finland's deal to get collateral in exchange for loans granted to Greece in Greece's 2nd aid package becomes public and sparks controversy amongst other
Euro area countries
02-Sep-11IMF/EU/ECB officials end Greece's 5th quarterly rev iew early , unhappy w ith the pace of structural reforms. They delay decis ion on granting the September aid
tranche to Greece
6-8 Sep 2011 A wide sw athe of German politicians declare that Greece will not get its next aid tranche w ithout a positive IMF rev iew and sufficient progress on reforms
06-Sep-11 Swiss c entral bank announces it will target a 1.2 floor for Swiss/ Euro exchange rate
07-Sep-11German constitutional court rules that the EFSF structure is constitutional but requires m ore strict ov ersight form Parliament (mandatory Parliamentary Budget
committee approval before aid is distributed). The ruling has bearish implications for Eurobonds
11-Sep-11 Greece announces it will speed up reforms and collect a new property tax to help bridge the budget gap
14-Sep-11 Italian parliament gives final approval in a confidence vote to a €54bn austerity package to balance the budget by 2013
19-Sep-11 S&P downgrades Italy 1 notch to A
20-Sep-11 Slovenian government loses a confidence v ote w hich threatens to delay EFSF ratification
29-Sep-11 Bundestag votes to approve EFSF reforms
02-Oct-11 Greece’s gov ernment approves the draft budget for 2012
3-4 Oct 2011EU finance ministers w ork out a rev amped deal on collateral for Greek loans that satisfies Finnish demands and those of other euro-region gov ernments
opposed to a bilateral deal for Finland
07-Oct-11 Fitch cuts Spain rating to AA- and Italy rating to A+
11-Oct-11 Troika releases statement on the fifth review of Greek economy and suggests that the sixth tranche of the bailout pay ments, w orth €8bn, w ill be paid
13-Oct-11Slovak ia becomes the final country to approve the upsized EFSF facility. The ruling government lost a confidence vote tied to the initial EFSF v ote which w ill
lead to new parliamentary elections
21-Oct-11 Papandreou wins parliamentary approval of the latest austerity bill in Greece, w hich includes w age and pensions cuts and plans to lay off 30,000 state workers
23-Oct-11 European leaders say a summit on the euro crisis w on’t produce decisions and set another meeting for 26 October
26-27 Oct 2011EU leaders agree to 1) leverage the EFSF to boost its firepower to €1tn, 2) force private inv estors to accept a 50 percent haircut on Greek bonds, 3) push Euro
zone banks to raise €104bn in new capital, and 4) ex tend a new aid package w orth €130bn for Greece
31-Oct-11 Surprising his ow n party , EU officials, and markets, Papandreou calls for a public referendum on the second bailout agreement in Greece
02-Nov -11 European leaders suspend aid payments to Greece and say Greece must decide w hether it w ants to stay in the euro
03-Nov-11 Papandreou backs down on referendum
06-Nov -11 Papandreou agrees to step aside to make way for a government of national unity in Greece
08-Nov-11Berlusconi fails to secure a majority in a v ote on public finance in the low er house of the Italian parliament. He later announces that he will res ign once he has
secured passage of the upcoming budget bill. Italian 10Y yield c rosses the 7% mark
08-Nov -11 LCH.clearnet increases margin requirements on Italian paper by 3.5-5.0% points across the curv e
09-Nov -11 Greek political leaders agree on a new government, Papandreou steps down
10-Nov -11 Former ECB v ice-president Lucas Papademos is tapped to head Greece's new coalition government
12-Nov -11 The Italian parliament passes the budget bill and Berlusconi resigns
13-Nov -11 President Napolitano nominates former European Commissioner Mario Monti to become prime minister of Italy , heading a new technocratic gov ernment
16-Nov-11 The Papademos-led Greek government wins a v ote of confidence
16-Nov-11 Mario Monti is sw orn in as Italian Prime Minister
17-18 Nov 2011 Mario Monti wins c onfidence v ote in both houses of the Italian Parliament
Source: Bloomberg, Reuters, WSJ, FT, NY Times, Telegraph
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 26/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Kedran Panageas (44-20) 7777-0326J.P. Morgan Securities Ltd
26
Appendix-2: What is the maximum
amount of funds that Italy can raise via
one-off wealth taxes and asset sales?
There has been speculation in the media around the
extent of funding that Italy can raise via one-off wealth
taxes and asset sales. In this appendix, we attempt to
quantify the maximum amount that Italy can potentially
raise via three sources of funding:
1. A one-off tax on bank deposits, securities, and
property,
2. Privatisation of financial and non-financial assets,
and
3. Partial sale of gold + FX reserves held at the central bank
Exhibit A1 shows the wealth in Italy as of 2009, as
estimated by the Bank of Italy. Around half of the total
wealth of around €9.4tn is in housing, and another almost
one-third is held in cash deposits, Italian government
bonds, and other securities. We assume that, in extremis,
Italy will be able to impose a one-off tax of 100bp on
housing and Italian government bonds,31 and 60bp on
cash deposits/other securities and mutual funds.32 Such a
move would raise €60–70bn, or around 3 months of
Italian funding needs.
On the privatisation front, it is difficult to get accurate
data on holdings of financial and non-financial assets
held by the Italian government.33 Additionally, it may be
hard to make assumptions on the possibility of large-
scale privatisations given the difficult economic
environment. We therefore look at historical precedent
to estimate how much money can be raised via
privatisations, since Italy has been privatising assets over
the past several years. Exhibit A2 shows the amount of
financial and non-financial assets that Italy has privatised
on average over the past several years, as a percentage of
GDP.34 Also shown is the amount of financial assets
privatised in the top 3 years. Although data on peak non-
31 In the US, an annual tax on property is levied by the local (town)
authority to fund the local school system. This tax ranges from 0.2% to
4.0% (average 1%) of the value of the property. We therefore assume
that Italy can impose such a one-off tax. For more information, see:http://www.taxfoundation.org/publications/show/1913.html.32 There is historical precedent for a 60bp tax on cash deposits. In 1992,
Italy imposed such a tax under Prime Minister Giuliano Amato.33 For example, a recent presentation on the Italian Treasury’s websitehas data on Italian government holdings going no further than 2005. 34 Note that since we use two different data sources, the data periods for
privatisation of financial and non-financial assets do not match.
financial asset privatisation is not available, we estimate
it from the ratio of ‘Top 3 years’ to ‘average’ financial
assets privatised. We give Italy the benefit of the doubt
and assume that it can privatise assets at the maximum
rate achieved over the past few years. Since Italy has
historically achieved peak annual privatisation of 1.9%
of GDP for financial assets, and around 0.9% of GDP for
non-financial assets, we conclude that a similar high rate
Exhibit A1: Even assuming onerous one-off wealth taxes can beimposed on Italian households, the maximum that the government canraise is €60–70bn, or 3 months of Italian funding needs
J.P.Morgan expectation of maximum one-off taxes that may be raised by ItalyAssumed one-off tax
€bn % of total bp €bn
Housing 4830 51% 100 48
Land 241 3%
Non-residential buildings 335 4%
Valuable items 124 1%
Other 347 4%
Deposits, cash, post office 1062 11% 60 6
Italian gov ernment bonds 189 2% 100 2
Other securities and mutual funds 1576 17% 60 9
Insurance technical reserv es 631 7%
Commercial credits 107 1%
Total 9442 100% 66
Wealth
Source: Household wealth in Italy 2009, No. 67, Supplements to the Statistical Bulletin,Banca D'Italia, Eurosystem
Exhibit A2: Historically, Italy has been able to achieve peak annualprivatisation of 1.9% of GDP for financial assets, and an estimated 0.9%of GDP for non-financial assets. Achieving a similar high rate of privatisations in 2012 will likely raise €40–45bn of funding*, or 2 monthsof Italian funding needs
Average and maximum level of privatisation receipts achieved by Italy in past years
Privatisation receipts;
% of GDP
Financial assets (1987-09)
Av erage 0.7%
Top 3 y ears** 1.9%
Non-financial assets (2000-06)
Av erage 0.3%
Top 3 y ears*** 0.9% * Italian GDP is around €1.5tn, so 2.8% of GDP works out to a round €40–45bn.** Top 3 years for financial asset privatisation were 1997, 1999, and 2005 when
€19bn, €27bn, and €26bn were privatised, respectively. Source: Privatisationbarometer.*** Italy privatised non-financial assets worth 0.3% of GDP between 2000-2006.
Although data on peak non-financial asset privatisation is not available, we estimate itfrom the ratio of ‘Top 3 years’ to ‘average’ financial assets privatised. Source: Should Italy Sell Its Nonfinancial Assets to Reduce the Debt ?, Stefania Fabrizio, IMF PolicyDiscussion Paper, April 2008.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 27/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Kedran Panageas (44-20) 7777-0326J.P. Morgan Securities Ltd
27
of privatisations in 2012 will likely raise €40–45bn of
funding,35 or around 2 months of Italian funding needs.
Additionally, the Bank of Italy holds gold + FX reserves
in excess of 8% of GDP, compared to a Euro area
average of just under 6% (Exhibit A3). Bringing them in
line with the average would generate funding of around
2% of GDP (€30bn), or around 1-1/2 months of funding
needs.36,37,38 Note that the Bank of Italy also owns around
€44bn of European sovereign debt (excluding BTPs), a
part of which may also be saleable.
Overall, we find that, in extremis, Italy may be able to
raise around €125–150bn over the next year via a
wealth tax and aggressive asset sales, allowing it to
avoid tapping capital markets for as long as 6–8months. In the long run, however, it is going to be
difficult for Italy to raise taxes significantly in order to
achieve a primary surplus as it is already one of the most
heavily taxed nations in the OECD world (Exhibit A4).
Instead, it will need to rely on increasing the tax base and
boosting growth.
35 Italian GDP is around €1.5tn, so 2.8% of GDP works out to around
€40–45bn.36 Technically, governments cannot use gold for funding needs since
central banks are legally independent and free from governmentinfluence. Indeed, according to an FT article entitled Italy to use gold
reserves to cut national debt , 1 August 2007, “Italy’s governmentcannot unilaterally infringe the central bank’s independence and
autonomy, although the bank could freely agree to contribute some of its reserves”. Given the severity of this crisis, we believe that the
central bank will not hesitate in making such a contribution, especially
since it would contribute to financial stability.37 The ECB will need to approve the sale of gold and FX reserves by
the Bank of Italy. Since Italy will be selling such assets merely to bringthem in line with the Euro area average, we believe that such approval
will be forthcoming.38 In 2009, central banks in the Eurosystem (together with Sweden and
Switzerland) agreed to not sell more than 400 tonnes of gold per year intotal, or around €17bn at current market prices. Given the severity of
the crisis, however, we believe that such restrictions may be relaxed inorder to help Italy.
Exhibit A3: Italy’s holdings of gold + FX reserves are significantlyhigher than the Euro area average; bringing these assets in line withthe average should yield around 2% of GDP (€30bn), or around 1-1/2
months of funding needs*Gold and FX reserves of various Euro zone countries as a % of GDP
Tonnes €bn €bn %GDP
Austria 280 12 5 301 17 5.7%
Belgium 228 10 6 370 16 4.2%
Finland 49 2 4 190 6 3.3%
France 2,435 103 22 1,988 125 6.3%
Germany 3,401 143 29 2,567 172 6.7%
Greece 112 5 0 218 5 2.2%
Ireland 6 0 0 156 1 0.4%
Italy 2,452 103 26 1,586 129 8.1%
Netherlands 613 26 7 607 33 5.5%
Portugal 383 16 1 172 17 9.7%Spain 282 12 12 1,075 23 2.2%
Total 10,239 431 112 9,231 543 5.9%
GDP (€bn)Gold + FX res.Gold** FX reserves
(€bn)
* Note that the Bank of Italy also owns around €44bn of European sovereign debt(excluding BTPs), a part of which may also be saleable.** Gold spot level used: €1310/oz.Source: World Gold Council, ECB
Exhibit A4: It may be difficult for Italy to raise taxes significantly on anongoing basis, given that it is already one of the highest taxedcountries in the OECD worldTotal tax revenue as % of GDP*; %
24
2829
31
34 3537 39
42 43 43 43 4445
20
25
30
35
40
45
50
U S
I r e l a n d
G r e e c e
S p a i n
U K
P o r t u g a l
G e r m a n y
N e t h e r l a n d s
F r a n c e
A u s t r i a
F i n l a n d
B e l g i u m
I t a l y
* * S c a n d i
* Data as of 2008 for Netherlands and Portugal, and as of 2009 for the rest of thecountries.** Scandi average includes Norway, Denmark, and Sweden.
Source: OECD
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 28/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Kedran Panageas (44-20) 7777-0326J.P. Morgan Securities Ltd
28
Appendix-3: What is the maximum
amount of funds that sovereign wealth
funds (SWFs) might be willing to invest
in EU peripherals?
EU officials are targeting emerging market countries
such as Brazil, China, and Russia, countries with large
amounts of FX reserves such as Japan, and countries
with significant natural resources such as Norway and
Qatar to help peripheral EU sovereigns. In particular, the
plan to leverage EFSF capacity increases the importance
of finding equity or debt investors from amongst this
pool. Despite stepped-up entreaties from EU officials,
we do not think these nations will be willing to
contribute a large amount of funds.
We base our view on several factors.
1) While sovereign wealth funds do control a large
amount of resources, nearly €2tn in total, the pool of
resources available to EU sovereigns is much smaller
than widely believed. This is because of several reasons.
First, SWFs are typically highly concentrated in equity.
Based on 1H12 data from Monitor, a consulting group,
we estimate that less than a quarter of SWF assets are
allocated to fixed-income. Second, SWF assets are
diversified across currencies; we estimate that a little
over one-third is allocated to European currencies. Thisimplies that in total, just around €170bn is currently
allocated to European fixed income (€170bn = €2tn *
24% * 36%, as shown in Exhibit A5). Third, this
amount includes non-Euro area currencies such as
sterling and Swiss franc, and non-sovereign exposures
such as corporate credit, structured finance, covered
bonds, etc. Although few SWFs break out portfolio
composition in much greater detail than this, it is likely
that the pool of resources available to euro area sovereign
debt is smaller still.
2) Recent commentary suggests that these countries are
reluctant to provide large amounts of aid. This is
especially true given that EU officials have indicated
they do not plan to make political concessions in order toobtain commitments. Exhibit A6 highlights recent
statements by high-ranking officials in various countries.
Exhibit A5: Despite having nearly €2tn of resources, very little of sovereign wealth fund resources are allocated to European fixedincome (€170bn). We assume that this allocation to European fixed
income could be increased at most by one-third, or €55–60bnSWF holdings and amount allocated by region and sector as of 2Q11; €bn
Total Europe Americas Asia/EM
Sector Sector w eight (%) 100% 36% 38% 26%
Total 100% € 2,012 € 720 € 771 € 520
--Fix ed Income 24% € 476 € 170 € 183 € 123
--Equity 67% € 1,340 € 479 € 514 € 347
--Other 8% € 165 € 59 € 63 € 43
--Cash 1% € 30 € 11 € 12 € 8
Region w eight (%) and allocation (€bn)
Note: Converted from USD using 1.35 EUR/USD FX rate.Source: Monitor Company Group, J.P. Morgan estimates
Exhibit A6: Recent commentary suggests that most non-Euro area countries are reluctant to provide aid, particularly countries that are lesswealthy than the EU peripheralsSelection of recent comments regarding SWF or FX reserve investments in peripheral sovereign debt
Country Speaker Date Comment Pos/Neg
RussiaSergey Ignatiev, Central bank
chairman18-Nov
Russia’s central bank isn’t planning to purchase debt issued by the EFSF until there’s more clarity
regarding the rescue fund.Negative
02-Nov It’s now “too soon” for his country to contribute. Negativ e
28-Oct"We, of course, mus t wait until its structure is ex tremely c lear. And moreover, this inv estment must be
decided on after serious, technical discussions."Negative
Oeystein Olsen, Central BankGovernor
28-Oct "It's not on the agenda to contribute any ex tra investments in the special purpose inv estment vehic le(SPIV) that have elements of help or aid. That's outside of the mandate."
Negative
Yngve Sly ngstad, Chief
Executive of the Norway oil fund28-Oct
"We do not have sufficient detail really to comm ent on what it would do for us in terms of investment
opportunity. .. Of course w e are committed to Europe although we have signalled that over time the fund
w ill have less of its investment in Europe."
Negative
Jens Stoltenberg, Prime Minister 26-Oct“N orway w ill not participate in any aid package...That is not our task. I believ e it’s w rong that Norw ay
should take part in such support measures.”Negative
Yoshihiko Noda, Prime Minister 02-NovWe welcome the [EU summit] package, and... "from the standpoint of supporting such efforts, w e w ill
continue to study the purchase of EFSF bonds."Positive
Senior Japanese finance minister 01-Nov "I told [Mr. Regling] that w e w ill continue to purchase EFSF bonds." Positiv e
Norway
Japan
ChinaZhu Guangyao, Vice Finance
Minister
Source: WSJ, Bloomberg, Reuters
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 29/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Kedran Panageas (44-20) 7777-0326J.P. Morgan Securities Ltd
29
For instance, Norway looks opposed to making any
investments on concessionary terms, and China and
Russia have also emphasised that any investment willdepend on the structure of the investment, how attractive
the investment is, and how extensive EU co-investment
is. It is especially unlikely that countries which are less
wealthy will see a compelling reason to support
relatively wealthy EU countries.
Thus, we believe that the €170bn allocated to
European fixed income will at most be increased by
one-third (or €55–60bn) as a contribution towards
peripheral sovereign debt purchases. Even this would
represent a large portfolio concentration of risk and
would likely require considerable arm-twisting and/or
political quid pro quos.
Appendix-4: The extent of ECB aid to
peripheral countries
Since the onset of the global financial crisis in mid-2007,
the ECB has grown its balance sheet by a substantial
amount, over €1.1tn. This is about half the rate of
increase seen in other central bank balance sheets
(Exhibit A7). However, the lion’s share of the increase
(nearly €700bn) comes from increased support to
peripherals, via
• repo operations to banks in peripheral countries,…
• …emergency liquidity assistance (ELA39) to banks
in peripheral countries,…
• …purchases of covered bonds (largely from Spanish
banks), and…
• …purchases of peripheral sovereign debt (Exhibit
A8).
The rest of the increase comes from items such as gold
reserves (up around €250bn), other securities and assets
excluding ELA (up €250bn), and claims on non-euroarea residents (up €100bn). Thus, while the ECB has
expanded its balance sheet less than its counterparts
have, it has arguably taken on a good deal more
credit risk in doing so.
Given this substantial increase in balance sheet risk, the
ECB is reluctant to further expand the size of its
39 See Overview, Global Fixed Income Markets Weekly, 1 April 2011for more description of ELA.
peripheral sovereign debt purchases. This is confirmed
by recent comments made by EU and ECB officials. For
example, while some German advisors have begun to
speak out in favour of a lender-of-last resort
commitment, key German leaders such as Merkel and
Schäuble, as well as the German Bundesbank, appear toremain firmly opposed (Exhibit A9).
Exhibit A7: The ECB has grown the size of its balance sheet by asmaller percentage than the Fed and BoE over the past few years...Total assets held on central bank balance sheets, current vs. 4Q06
3Q11 4Q06
bn %
ECB 2,289 1,134 1,154 102%
Fed 2,871 863 2,008 233%
BoE 243 81 162 199%
Change
bn of ntl ccy
Source: ECB, Federal Reserve, Bank of England
Exhibit A8: …but asset growth is heavily skewed towards riskier exposures such as peripheral bank repo and ELA* and SMP purchasesof peripheral sovereign debtECB balance sheet composition since 2007**; €bn
600
800
1000
1200
1400
1600
1800
2000
2200
2400
2007 2008 2009 2010 2011
repo to core banks
repo to periphery
banks/ELA*
SMP
Cov ered bond purchases
All other balance-
sheet items**
* ELA, or emergency liquidity assistance, is typically included in “Other assets” or “Other claims” on central banks’ balance sheets. We estimate the current amount of ELA outstanding based on Irish and Greek central bank data. Since the onset of thefinancial crisis in 2009, we estimate ELA reliance has risen to nearly €90bn in thesetwo countries. See Overview , Global Fixed Income Markets Weekly , 1 April 2011 for more description of ELA.** Includes items such as gold and FX reserves, claims on non-Euro area residents,and other assets and securities. We subtract the estimated ELA outstanding from thisfigure.Source: ECB, Central Bank of Ireland, Bank of Greece. Data as of Sep 2011.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 30/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Kedran Panageas (44-20) 7777-0326J.P. Morgan Securities Ltd
30
Exhibit A9: Indeed, comments from senior officials in Germany and at the ECB suggest that they are not ready to monetise peripheral debt in largequantities just yetSelection of recent policymaker comments regarding ECB purchases of peripheral sovereign debt
Country Speaker Date Comment
18-Nov
“Losing credibility c an happen quickly -- and history show s that regaining it has huge economic and social costs." Keeping prices s table
“is the major contribution we can make in support of sustainable growth, employ ment creation and financial stability . And w e are making
this contribution in full independence.”
03-Nov"Our securities market programme has three characteristics: it is temporary; it is limited; it is justified in restoring the functioning of
monetary transmission channels."
03-NovThe [ECB] w ill "not be forced by anybody " to buy bonds. .. it w as pointless to think sov ereign bond yields c ould be brought down for a
protracted period by outside intervention.
18-Nov
“T he economic cos ts of any form of monetary financing of public debts and deficits outweigh its benefits so c learly that it will not help to
stabilize the current situation in any sustainable way . The lack of success i n containing the crisis does not justify overstretching the
mandate of the central bank and making it responsible for solv ing the crisis. "
12-Nov"We have a mandate and w e have to stick to our mandate. Fix ing an interest rate for a country is c ertainly not compatible with our
mandate... y ou could not argue that this w as not monetary financing."
12-Nov"This w ould violate Article 123 of the EU treaty . I cannot see how y ou can ensure the stability of a monetary union by v iolating its legal
provisions."
Jürgen Stark, ECB Ex ecutive Board
Member 11-Nov "The ECB will never be lender of last resort... The ECB can't compensate for deficits in crisis nations."
11-Nov“T he ECB is last resort lender for banks because that is its function, but not for governments because the union’s treaty explicitly forbids
it."
Angela Merkel, Prime Minister 16-Nov "The way w e see the (EU) treaties is that the ECB does not have the possibility of solving these problems here."
Wolfgang Schäuble, German Finance
minister 18-Nov
“If w e did that, the consequence w ould be that for some m onths we’d hav e a certain quiet. But in the longer term, the financial markets
would assume that the euro isn’t a stable currency, w hether financial markets w ould then think that it’s not that bad, I doubt it. That’s w hy
i don’t think it’s the right solution.”
Mario Draghi, ECB President
ECB/
Eurosystem
Jose Manuel Gonzalez-Paramo, ECB
Executive Board Member
Jens Weidman, head of the
Bundesbank
Germany
Source: Bloomberg, Reuters, Telegraph, Businessweek
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 31/236
Economic ResearchGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
David MackieAC
(44-20) 7325-5040
[email protected] Chase Bank N.A., London Branch
31
Economics
• Sovereign stress is expected to drive Euro area
into a mild recession,…
• …but there is significant risk that the downturn
could be deep
• The ECB main policy rate is expected to be cut to
0.50%
• We do not forecast ECB QE, but SMP purchases
will push the agenda for a fiscal union
• The rest of the world escapes recession, but
growth will be lackluster• Fiscal tightening will be a significant headwind in
the US
• The UK MPC will continue to expand its gilt
purchase programme, possibly extending that to
other assets
• EM growth forecasts have been cut in line with
the historical relationships between DM and EM
The outlook for the global economy in
2012
Since the global recovery began in mid-2009, there has
been a fair amount of tension between the forces of
cyclical lift (easy monetary policy, low levels of
spending on durables, low inventories and healthy non-
financial corporates) and the structural drags
(deleveraging by households, banks and governments,
elevated uncertainty about fiscal and regulatory issues,
and a decline in underlying growth potential).
As the recovery began to unfold, we expected the forces
of cyclical lift to dominate and ensure above-trend global
growth. Even so, the structural drags were always
expected to ensure that the return to full employmentwould be a long journey. Indeed, this was the experience
of the global economy in the first 18 months of the
recovery, from mid-2009 to the end of 2010, when GDP
growth averaged 3.8%ar. However, 2011 has not
followed this script at all (Exhibit 1). It now looks like
the pace of global growth in 2011 will be around two-
thirds of what we expected earlier in the year (a quarterly
average growth rate of 2.3%ar compared with our earlier
expectations of 3.9%ar). Further, our expectations for
the global economy in 2012 show a much more
moderate pace of growth than we expected earlier: an
average quarterly annualised rate of 2.1%ar. Essentially,
after the above-potential performance of the first 18
months of the recovery, the global economy is now
expected to experience below-potential growth through
2011 and 2012.
There are a number of reasons why the global economy
is now performing at a sub-par pace. There were clearly
two huge unexpected shocks in 1H11: the surge in
commodity prices, which dramatically reduced
purchasing power for commodity consumers, and theJapanese earthquake and tsunami, which not only dealt a
huge blow to the Japanese economy but also disrupted
the global manufacturing supply chain. Although the
impact of these shocks started to fade in the summer,
other drags entered the picture. Essentially, the structural
drags have become bigger and are now dominating over
the cyclical lift. The debt ceiling debate in the US has led
to more fiscal tightening next year and beyond.
Meanwhile, the Euro area sovereign crisis has pushed the
region back into recession, reflecting the direct effects of
fiscal consolidation and the indirect effects of a
tightening of financial conditions. The tipping point for
the Euro area was the spread of sovereign stress to Italy,which led to a sharp fall in confidence and asset prices,
and a significant tightening in bank funding conditions.
Even though monetary policy has eased further in both
the US and the Euro area, and emerging market growth is
holding up relatively well, the structural drags in the
US and the Euro area appear to be sufficient to
restrain global growth to a sub-par pace through the
end of 2012.
Exhibit 1: Global GDP growthRealised real GDP growth and J.P. Morgan forecast; %oya
-10
-5
0
5
10
2007 2008 2009 2010 2011 2012 2013
Global
Emerging markets
Developed markets
forecast
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 32/236
Economic ResearchGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
David MackieAC
(44-20) 7325-5040
[email protected] Chase Bank N.A., London Branch
32
Deleveraging shifts from households to sovereignsand banks
Over the past few years, there has been much focus onhousehold deleveraging. However, as far as the broader
economy is concerned, the drag from household
deleveraging is now very limited. The sharp move in
household financial positions ended some time ago, and
we are currently in a situation where households are
generating free cash flow and using that to repay debt
(Exhibit 2). The big deleveraging stories of this year
and the next concern sovereigns and banks. After
easing fiscal policy significantly during the crisis, fiscal
authorities in the US and Europe are turning towards
tightening. Having already begun this year, it is due to
get much more intense in 2012 (Exhibit 3). Meanwhile,
the sovereign crisis in the Euro area and the policyresponse encouraging recapitalisation looks to be
accelerating the deleveraging process in Euro area banks.
It is hard to know exactly where equilibrium
configurations of primary positions and debt levels stand,
but the sovereign deleveraging is likely to be a long and
hard journey, most probably longer and harder than the
journey for households. The key difference is that
sovereigns hope to take more time to complete this
journey, although across the periphery of the Euro area,
financial market impatience is accelerating the
adjustment with inevitable consequences for growth.
Fiscal tightening is a key reason why global growth is
expected to be sub-par in 2012, and this basic picture is
unlikely to change in 2013 and beyond. Sovereign
deleveraging is likely to dominate the macro landscape
for years to come.
Euro area slides into recession
Recent data confirm that the Euro area moved into a
recession in the autumn. When gauging whether an
economy is transitioning into a recession, we look at both
changes and levels of key variables. Over the past 6
months, the area-wide composite PMI has fallen by more
than during any other 6-month period except in theimmediate aftermath of the Lehman bankruptcy (Exhibit
4). Further, the level of the PMI is now well below where
it stood at the start of the 2008-09 recession (April 2008)
(Exhibit 5). The labour market data also are sending a
compelling message of recession. Over the past three
months, unemployment has risen by an average of
125,000, well ahead of the average in the three months
up to the start of the 2008/09 recession (Exhibit 6).
Exhibit 2: US household financial position* and stock of debtoutstanding% of GDP % of GDP
-4
-2
0
2
4
6
8
1980 1985 1990 1995 2000 2005 2010
40
50
60
70
80
90
100Financial position Debt stock
* US households financial position: ((disposable income - personal outlays)-( grossprivate investment -consumption of fixed capital))/GDP
Exhibit 3: US fiscal policy changes for 2012$bn
Current law Net change
Payroll tax -110 +65
Other taxes -25 +45
Infrastructure, S&L support -125 +15
Unemployment relief -50 +12
Other spending -40 -40
Total -350 +97
% of GDP -2.3 +0.6
Imapct on GDP (%pt) -1.8 0.5 Note: Obama plan incorporates changes that will not be fully implemented in 2012
Exhibit 4: Change in Euro area composite PMI6M change in Euro area composite PMI; points
-15
-10
-5
0
5
10
15
1998 2000 2002 2004 2006 2008 2010 2012
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 33/236
Economic ResearchGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
David MackieAC
(44-20) 7325-5040
[email protected] Chase Bank N.A., London Branch
33
At the moment, our forecast is of a mild recession in
the Euro area. The contraction is expected to last
through next autumn and to involve a peak-to-troughdecline in the level of GDP of just over 1%. In the
recessions of the mid-1970s, early 1980s, and early
1990s, the Euro area saw peak-to-trough declines in the
level of GDP of 2.5%, 0.5%, and 1.9%, respectively.
And in the 2008/09 recession, GDP fell by a staggering
5.5%.
Given the relatively unusual nature of the current
environment, and the difficulty in gauging the behaviour
of policymakers, we would emphasise that there is a lot
of uncertainty around our central projection for the Euro
area real economy. Given the stress in the financial
system, the recession could easily be deeper than ourcentral projection, especially if policymakers do not
manage the situation well.
What usually causes a recession?
Recessions occur when one or more negative shocks hit a
vulnerable economy. These shocks can take a variety of
forms, but they involve either a squeeze in current real
disposable income (interest rates, commodity prices,
fiscal policy) or a shift in the forces that influence current
spending relative to current real disposable income
(interest rates, asset prices, wealth, credit availability,
uncertainty). Meanwhile, an economy becomesvulnerable to such shocks for a number of reasons:
households or non-financial corporates have
overextended their spending on durable goods and
inventory relative to their income (through excessive
borrowing, partly driven by elevated asset prices); or
banks have over-levered their balance sheets relative to
their capital positions (again partly driven by elevated
asset prices); or corporate profits have been squeezed to
such an extent that drastic action is required to restore
profitability.
The Euro area recessions of the 1970s, 1980s, and 1990s
were driven by monetary policy shocks that hit a privatesector with overextended levels of spending and a
corporate sector where profitability had been depressed
by an overheated economy. The 2008/09 recession too
involved a monetary policy shock, but there was also a
huge asset-price shock; the overextension was in the
financial sector as well as in the non-financial sector. In
2008/09, the classic recessionary adjustment in the non-
financial private sector was amplified by a severe
adjustment in the financial sector.
Exhibit 5: Euro area composite PMIIndex
35
40
45
50
55
60
2007 2008 2009 2010 2011 2012
Lev el at the start of 2008/09
recession: 51.9
Exhibit 6: Euro area unemploymentMonthly change; 3M average; 000s
-200
-100
0
100
200
300
400
500
2007 2008 2009 2010 2011 2012
Lev el at the start
of 2008/09
recession: 29
Exhibit 7: Euro area bank lending standardsNet tightening of loans to non-financial corporations and households; %
-20
0
20
40
60
80
2003 2005 2007 2009 2011
Net tightening of loans to non-fin.corporates
Net tightening of loans to households
Source: ECB
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 34/236
Economic ResearchGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
David MackieAC
(44-20) 7325-5040
[email protected] Chase Bank N.A., London Branch
34
What does the current Euro area recession look like?
At first blush, the Euro area recession that we are
sliding into now looks more like the 2008/09experience. Instead of a monetary policy shock, the
region is experiencing a confidence shock – confidence
in sovereign solvency – which is being amplified by a
still-vulnerable financial system. As asset prices have
fallen, bank funding stress has risen dramatically.
Although banks have access to unlimited funding from
the central bank, funding stress does lead to higher retail
borrowing rates, tighter lending standards (Exhibit 7),
and a desire on the part of banks to de-lever their balance
sheets. This suggests that the Euro area will experience
something of a credit crunch in the coming quarters. As
we saw in 2008/09, tight credit availability can take a
significant toll on the real economy.
However, a key difference between now and early
2008 is that levels of spending on durable goods and
inventory are currently low, and financial positions in
both the household and corporate sectors are
currently good. There is still a deleveraging process
under way in some of the peripheral economies – as
households generate free cash flow to reduce the size of
their balance sheets – but conditions are not nearly as
stretched as they were at the start of the 2008/09
recession. Moreover, the significant fiscal consolidations
are concentrated in the peripheral economies, and it ishard for the Euro area to have a deep recession
without a significant fall in Germany and France.
Moreover, the global backdrop looks better now than it
did in 2008/09: what we are experiencing at the moment
is an asymmetric Euro area shock, rather than a common
global shock. The differences between now and early
2008 – and the uneven spread of fiscal consolidation –
are why we think the present recession will be relatively
mild.
The distinguishing feature of recessions tends to be depth
rather than duration. The 2008/09 recession in the Euro
area was almost three times deeper than the early 1990s’recession, but it was only one-quarter longer in duration.
This suggests that policy plays a key role in arresting the
negative dynamic in recessions. However, in contrast to
the experience of previous recessions, when
policymakers were adding stimulus to the economy, the
policy action that we anticipate in the coming months –
asset purchases by the EFSF and the ECB, loans to
sovereigns and bank recapitalisation – is more about
crisis containment. This leads to the question of how
effective policy will be in containing the negative
dynamic in the real economy and creating a base for the
ultimate recovery.
All of this indicates that it is difficult to have a lot of
conviction about what the present Euro area recession
will look like. The conditions are unusual and it is
unclear how confidence, asset prices, and bank funding
stress will interact, and how that will impact the real
economy. There is also uncertainty about policy. This is
not simply about whether policymakers will respond in a
timely and effective manner. It is also about the limited
scope for the addition of outright stimulus. Our
inclination is to think that downside risks prevail
around our central projection.
Recession makes fiscal and structural adjustmentsharder to achieve
As the Euro area slides into a recession that looks likely
to be particularly deep in parts of the periphery,
budgetary positions will slip due to the automatic
stabilisers. Given the need for peripheral sovereigns to
keep moving towards more sustainable fiscal positions,
there will be enormous pressure to offset some of the
budgetary slippage with additional tightening measures,
even though these will weigh additionally on demand.
Our forecast does incorporate some additional fiscal
Exhibit 8: Government budget balance projectionsOfficial* and J.P. Morgan budget balance projections**; % of GDP
2011 2012 2013 2014
Greece Official -7.3 -5.6 -4.4 -2.2JPM, no extra aust. -9.1 -10.3 -9.0 -8.2
JPM, more austerity -9.1 -8.3 -6.9 -6.7
Ireland Official -10.0 -8.6 -7.2 -4.7
JPM, no extra aust. -9.6 -9.3 -9.3 -7.9
JPM, more austerity -9.6 -8.9 -8.3 -6.4
Portugal Official -5.9 -4.5 -3.0 -1.8
JPM, no extra aust. -6.8 -5.9 -5.4 -5.3
JPM, more austerity -6.8 -5.2 -4.2 -3.6
Spain Official -6.0 -4.4 -3.0 -2.1
JPM, no extra aust. -7.1 -6.6 -6.0 -5.9
JPM, more austerity -7.1 -5.9 -4.5 -4.0
Italy Official -3.9 -1.6 -0.1 0.2
JPM, no extra aust. -4.3 -2.8 -2.1 -2.3
JPM, more austerity -4.3 -2.2 -1.0 -0.9
Belgium Official -3.6 -2.8 -1.8 -0.8JPM, no extra aust. -3.8 -4.1 -3.6 -3.1
JPM, more austerity -3.8 -3.5 -2.6 -1.9 * Official projections are from the Stability and Growth Programmes for Greece,Ireland, Spain and Belgium (the Greek one was published in August rather than in thespring). Portuguese projections are from their fiscal strategy and Budget 2012documents, published in October. Italian projections are from their updated strategydocument published in September.** “JPM, no extra austerity” assumes that governments do not correct the cyclical fiscalslippage due to growth undershooting the official forecasts“JPM, more austerity” assumes additional fiscal tightening worth 50% of the cyclicalslippage
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 35/236
Economic ResearchGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
David MackieAC
(44-20) 7325-5040
[email protected] Chase Bank N.A., London Branch
35
tightening next year, beyond the plans already
announced. Essentially, we assume that governments
seek to offset half of the cyclical slippage (Exhibit 8).
In addition to putting significant pressure on individual
sovereigns, the recession dynamic is also affecting the
rescue mechanism put in place by the region. As
France’s AAA rating looks less secure, due to both the
discussion about leveraging the EFSF and the evidence
that France itself may be falling into a meaningful
recession, the EFSF’s own situation looks less robust.
The response of the ECB to the macro landscape
In early November, the ECB cut the main policy rate
from 1.50% to 1.25%. Although this move was not
widely expected, it is not hard to understand given the
deterioration in the macro landscape over the past few
months. Indeed, the new ECB President, Mario Draghi,
spoke of the region heading toward a mild recession,
which would exert downward pressure on wage and price
inflation. We expect another rate cut to follow in
December, and for the main policy rate to ultimately
reach 0.50%. If the recession turns out to be a deep one,
there will be talk about whether the ECB should engage
in quantitative easing, which we would interpret as a
targeted amount of bond purchases spread across the
region in order to ease the monetary stance further by
lowering longer-term interest rates. We would contrastthis with the Securities Markets Programme (SMP),
which is intended to improve the transmission of the
prevailing monetary stance rather than add more
stimulus. At the moment, we do not anticipate the
ECB engaging in QE, although the central bank’s
balance sheet will still expand further due to SMP
purchases.
The response of the ECB to sovereign stress
As the stress in the sovereign bond markets has spread to
core countries such as France and Austria, it has become
clear that something more dramatic from policymakers is
needed to stabilise markets and ensure financial stabilityacross the region. Not surprisingly, all eyes are on the
ECB as the only institution able to step in quickly and
decisively.
The obvious next step would be for the ECB to announce
its intention to use the SMP aggressively to ensure
financial stability in the region. In our view, there is no
legal or technical limit to the amount of explicit
support that the ECB can give secondary markets
and the amount of implicit support that the central
bank can give to primary markets (by purchasing
around the primary auctions to help dealers get short and
shed unwanted positions, but not at the auction itself).
This more-aggressive stance could be adopted with
explicit yield targets, although we doubt that the ECB
will go that far.
Whether this stabilises the situation remains unclear.What lies at the heart of the crisis is a fear that the
process of sovereigns returning to solvency via austerity
alone, i.e. without local currency depreciation and a
locally orientated monetary stance, is dynamically
unstable. Adding to the prospect of a dynamically
unstable process is the widely held perception that
peripheral economies are profoundly uncompetitive and
that some of the structural reforms needed to improve
competitiveness will depress growth and damage social
cohesion before they deliver any macroeconomic
Exhibit 9: Italian debt-to-GDP under different scenarios% of GDP*
90
100
110
120
130
140
2000 2005 2010 2015 2020
J.P. Morgan
baseline
Official plan
Mild shock case
* See Exhibit 10 for assumptions under different scenarios
Exhibit 10: Key variables for Italian debt dynamicsEvolution of macro variables under alternative scenarios*; % of GDP, unless otherwisespecified
Official
plan
J.P. Morgan
baseline
Mild shock
case
Marginal borr. rate (from now), % 5.8 6.5 6.5
Real long run growth pot., % 1.2 0.8 0.6
Maximum sustained prim. balance 5.7 5.0 3.5
Average borrowing rate, % 5.0 5.4 5.4
Nominal GDP growth, % 3.0 2.0 1.7
Government debt 104.8 120.8 130.2Primary balance 5.4 4.8 3.3
Yearly change in debt -3.4 -0.7 1.4
Key assumptions
Key variable averages over the period 2012-20
* Official projections are partly the result of our assumptions, as they are not availablebeyond 2014.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 36/236
Economic ResearchGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
David MackieAC
(44-20) 7325-5040
[email protected] Chase Bank N.A., London Branch
36
benefits. This fear of dynamic instability has increased
with the growing evidence that the region is sliding into
recession, possibly a deep one. Greece provides a clear example of how badly things can turn amid political and
social instability and sovereign debt restructuring.
At some point, sizable ECB interventions would stabilize
the system. But, the situation for Italy remains a key
uncertainty (Exhibit 9 and 10). In our view, the SMP is
not an appropriate mechanism for funding Italy’s
gross financing need over the medium term. Thus, the
region does need to create some kind of l iquidity
hospital to accommodate Italy. The EFSF as currently
constructed cannot do this. Thus, it is not surprising that
France has put the idea of turning the EFSF into a bank
back onto the agenda and that there are ongoingdiscussions about how to involve the IMF more
substantially.
Alongside increased ECB intervention and the creation
of a liquidity hospital for Italy, the Germans will
continue their campaign for further fiscal integration.
As this crisis has continued to unfold, it has become
increasingly evident that the current blueprint for the
future – more peer surveillance and pressure but no de
jure loss of fiscal sovereignty, and with a permanent
liquidity hospital based on the idea of shared fiscal
capacity – is unlikely to work, unless sovereign debt
levels are much lower. The risk of sovereign liquiditycrises is just too great. But, in terms of dealing with this
fundamental issue in the functioning of the Euro area,
policymakers are moving incrementally. In the near term,
the Germans want limited treaty change to introduce
automatic sanctions in the workings of the stability and
growth pact, the possibility of referring sovereigns to the
European Court of Justice, and the possibility of greater
interventions in national budgets. It seems that they
continue to want an orderly sovereign default mechanism
as part of the permanent crisis resolution framework. It is
not clear, however, that this will be sufficient.
Impact of the Euro area recession on the rest of theworld
In the middle of 2011, there was a serious growth scare
in the US, but the economy has settled at a close-to-
potential growth rate in the second half of the year. Even
so, two risks loom over the 2012 outlook for the US.
The first risk relates to spill-overs from the economic and
financial crisis in Europe. Around 20% of US exports are
destined for Europe, and the economic contraction will
surely impact demand from that market. If that were the
only transmission channel, the risks would appear to be
contained. However, the potential financial spill-overs
could be larger. Thus far, most measures show only a
modest impact of the European crisis on domestic credit
availability. Nonetheless, if conditions in Europe worsen,
the US economy would not be immune from a seizing up
of global financial markets.
The second risk is a tightening of domestic fiscal policy.
The recession produced an unprecedented fiscal policy
response as the Federal deficit increased to 10% of GDP.
The political tolerance for such large deficits has
lessened recently, and several temporary fiscal stimulus
measures are set to fade in 2012, which could subtract
1%-2%-points from GDP growth next year. There has
been some talk about extending the temporary measures,
but nothing has been agreed upon so far. The
anticipated drag from fiscal policy is the main reason
behind our view that growth will be below potential in
2012, when we anticipate US growth will average
1.75%.
This expected tightening of fiscal policy is likely to hit
the economy where it is weakest: the household sector.
Slack conditions in labour markets have eroded labour’s
bargaining position, resulting in very tepid wage gains
(Exhibit 11). This outcome has been positive for
corporate profits – which have received a significant lift
from restrained unit labour costs – but has held back
consumers’ purchasing power.
Exhibit 11: US labour market3M moving average of monthly change in non-farm payrolls and unemployment ratefor 25 years and older
% %
-1.0
-0.5
0.0
0.5
2006 2007 2008 2009 2010 2011 2012
3
4
5
6
7
8
9
Montlhly
payrolls
Unemployment rate
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 37/236
Economic ResearchGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
David MackieAC
(44-20) 7325-5040
[email protected] Chase Bank N.A., London Branch
37
The low rate of inflation has given the Federal Reserve
leeway to be creative in supporting the recovery. The two
experiments this year were the explicit mid-2013 fundsrate guidance, and Operation Twist. We think the most
likely next step for the Fed is to replace the mid-2013
rate guidance with language that is more explicitly
dependent on a limited number of observable economic
variables, particularly the unemployment rate. While we
are not forecasting more asset purchases in 2012, the risk
of this is high and rising (subjectively, we would put it at
around 40%). Moreover, Fed officials have shown an
increasing openness to considering returning to MBS
purchases. A significant setback in the Euro area, or
an unwelcome fall in inflation expectations, could be
all that is required for QE3 in the US.
The UK has conspicuously chosen to front-load its fiscal
adjustment. While that has shielded it from direct attack
on the gilt market, public sector job losses, tax increases,
and spending restraint have weighed on growth. Through
noisy data, growth has run at a near-1% pace since mid-
2010. As we look forward, fiscal consolidation is
ongoing, while confidence and bank funding conditions
are caught in the back wash of Euro area turmoil.
Leaning against this is the fact that the direct drags on
household purchasing power from rising VAT, energy,
and food bills are dissipating. And as inflation begins to
ease, the MPC will continue to expand its gilt
purchase programme, plausibly extending that to
other assets (such as bank term debt) as 2012
progresses. All told, the forecast shows the UK seeing
near stagnation in output through what will remain
volatile GDP data, with unemployment continuing to
rise.
Given expected sluggish developed market growth,
the emerging market outlook has also markedly
weakened. The drag from DM to EM operates through
two channels. The first is through trade linkages. A
slowing in EM growth from weaker exports to DM is
unavoidable. We estimate that EM export growth will
slow to just 4% in 1H12, which is weaker than at any
time during the 2000s’ expansion.
The second channel of transmission is through financial
markets. As risk appetite wanes, capital flows to the EM
slow or even reverse, equity markets sell off, credit
conditions tighten, currencies fall, and the prices of
commodities – a source of revenue for many EM
countries – decline. In turn, domestic demand in the EM
is hit. This time around, the financial transmission
channel could be magnified by the rapid rise in credit
since the start of the recovery in 2009. While domestic
funding sources have increased, deleveraging of
European banks could still pose a risk as their foreignclaims in EM are sizable – particularly in EMEA and
Latin America.
The interaction of these complex forces has generated
a unit-beta response in the EM to shifts in US/Euro
area growth over the past decade. In other words, a
1%-pt decline in US/Euro area growth has translated, on
average, into a 1%-pt decline in EM growth. EM Asia
has a smaller beta of 0.6 (owing to very low betas in
China and India), Latin America has a beta of 1.2, and
EMEA has the highest beta of 1.4.
Currently, the channels of transmission are operating asusual. Sluggish DM import demand has produced a sharp
slowdown in EM export growth. Risky assets have sold
off across the world and commodity prices have fallen.
In turn, economic activity has decelerated in most EM
countries. J.P. Morgan’s recently lowered GDP growth
forecasts for the next five quarters align well with the
implied beta relationships for the three EM regions, with
the average projected at a sub-trend pace of 5% in 2012.
Should the US/Euro area slide into a deeper downturn
than currently forecast, the greater flexibility of policy in
the EM could lead to a relative outperformance in the
region. This outcome, however, would require a changein EM policymakers’ reaction function. On the downside,
increased leverage in the region could exacerbate
financial stresses, leading to a worse-than-unit beta
outcome.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 38/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Gianluca SalfordAC
(44-20) 7325-4334
Aditya Chordia (44-20) 7777-9841J.P. Morgan Securities Ltd
38
Euro Cash
• 2011 saw a further escalation of the Euro area
sovereign crisis as stress reached Italy and core
countries
• In our central scenario the crisis will worsen in
the first part of 2012 on the back of Euro-wide
recession and further market pressure on
sovereigns…
• … forcing policymakers to react aggressively to
stabilise the situation in the second part of the
year
• The ECB will be forced to cut the refi rate to0.50% and expand its balance sheet even further
• Although our fair value model suggests that 10Y
Bund yields could drop below 1.00%, as the
PMIs/core inflation drop and peripheral spreads
widen, we forecast a floor in 10Y Bunds at 1.25%
• We recommend long 10Y duration exposure; but
this view is predicated on policy makers being
able to ultimately contain the crisis; if they fail
Germany’s role as a safe haven will be challenged
• The short end of the curve has little scope to rally
under our assumption that the depo rate will not
fall below 0.25%: we therefore recommend 2s/10s
flatteners targeting around 100bp by 1H12
• With regard to intra-EMU spreads, we forecast
generalized widening vs. Germany
• We also recommend credit curve flatteners: in
line with the experience of Greece, Ireland and
Portugal, investors will switch from focussing on
yields to prices in case of severe sovereign stress
• The empirical price relationship between short
and long dated bonds is convex: we recommend
flatteners weighted by the empirical beta to get
protection from tighter spreads and steepercurves
• We present a framework to analyse the yield
spread of high- and low-coupon bonds of similar
maturity under sovereign stress: Bonos 2032 are
rich vs. Bonos 2037
• We analyse the investor base of Euro area,
German, French, Italian and Spanish bonds. The
reliance on foreign banks and foreign central
banks is high for all of them, except Spain
• On the supply side, higher redemptions will be
generally offset by lower deficits. We expect
€710bn of conventional bond issuance in 2012,
slightly below 2011 levels
• Italy faces heavy bond redemptions in 2012,
especially in the February-April period, whereas
Spain’s redemption calendar is more manageable
2011 recap: Shattered hopes
“It is going to get worse before it gets better” was
investors’ mantra into 2011, a view that we shared. The
beginning of the year caught many investors on the
wrong foot as policymakers seemed to calm fears with
little effort and macro data continually surprised to
the upside, leading to stronger equities and higher
yields, with 10Y bund yields peaking at 3.50% in mid-
April (Exhibit 1). In addition, higher inflation prints
prompted the ECB to surprise market participants and
increase interest rates twice after being on hold for
almost two years. Declining excess cash in the systemadded to the perception of normalisation. It was not
meant to last.
The sovereign crisis raised its ugly head again in mid-
April. Portugal’s request for aid was met with a yawn,
but once again Greece was the catalyst: amid clear signs
of fiscal slippage, the request for private sector
involvement in exchange for further official support
broke a taboo and hit investors’ confidence hard. In less
than 6 months, Bund yields halved on the back of
Exhibit 1: 2011 recap: markets and data peaked in the spring and havebeen on a downward trend ever since
10Y Bund yield and Euro area composite PMI% index
1.50
1.75
2.00
2.25
2.50
2.75
3.00
3.25
3.50
3.75
Jan-11 M ar-11 M ay -11 Jul-11 Sep-11 Nov -11
46
48
50
52
54
56
58
60PMI
10Y Bund
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 39/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Gianluca SalfordAC
(44-20) 7325-4334
Aditya Chordia (44-20) 7777-9841J.P. Morgan Securities Ltd
39
pressure from peripherals, a collapse in the PMIs and the
August downgrade of the US by S&P. Since October,
yields have been range-bound and hostage to headline
news related to the sovereign crisis.
We are ending the year with 2s/5s significantly flatter YTD, 5s/10s almost unchanged and 10s/30s steeper. The
2s/5s and 5s/10s curves exhibited typical directionality in
the first part of the year, flattening in the bearish move,
but they lost directionality in the second part. On the
contrary, the 10s/30s Bund curve showed poor
directionality in the first part of the year refusing to
flatten further from already stretched levels, but bull-
steepened with decent directionality in the second half of
the year (Exhibit 2).
Sovereign risk escalated to yet new highs in 2011. In
the first part of the year, the problem remained isolated to
the smaller peripheral countries (Greece, Ireland, andPortugal). However, policy makers’ decision to include
private sector involvement (PSI) in the new package for
Greece spooked investors, spreading contagion to the
larger peripheral countries (Italy and Spain) and beyond.
It is interesting to highlight that in relative terms, France
or supranational issuers such as the EFSF
underperformed more than Italy and Spain did in the
latter part of the year (Exhibit 3), making the distinction
between core and peripheral debt markets somewhat
obsolete.
Although all intra-EMU spreads to Germany reached
their all-time highs since the Euro inception, France and
Italy were the worst performers in relative terms,
whereas Ireland was the best performer (Exhibit 4).
Rating agencies were very busy with downgrades
(Exhibit 5), adding to the pressure.
Once again, policy makers were forced to take
unprecedented actions: 1) ECB purchases: In August,
the ECB was forced to intervene by buying Italian and
Spanish paper through the SMP after a 4 month hiatus
(Exhibit 6). We estimate that the ECB has purchased
Exhibit 2: The curve directionality to the level of the market shifted fromthe short and medium end to the ultra-long end in the second part of the year 2s/5s, 5s/10s, and 10s/30s regressed against 2Y German government par rates; 1 Jan
2011 to 11 Apr 2011 and 12 Apr 2011 to 18 Nov 2011;2s/5s 5s/10s 10s/30s
Jan-April -0.27 -0.30 -0.01
Apr-Nov 0.09 -0.14 -0.19
Jan-April 75% 93% 1%
Apr-Nov 42% 65% 68%
Beta
R-squared
* Peak in 10Y Bund yields.
Exhibit 3: Sovereign risk escalated to yet new highs in 2011YTD % change in 10Y spreads to Germany for France, Greece, and Italy; %
-50%
0%
50%
100%
150%
200%
250%
300%
350%
Jan-11 M ar-11 M ay -11 Jul-11 Sep-11 Nov -11
France Greece Italy
Exhibit 4: Intra-EMU spreads widened aggressively after mid-April 201110Y benchmark spreads to Germany; bp
18-Nov-11 11-Apr-11* 30-Dec-1030-Dec to
Current
11-Apr to
Current
30-Dec to
11-Apr
Austria 147 31 46 216% 372% -33% 189 31
Belgium 282 72 104 170% 291% -31% 309 72
Finland 67 25 28 138% 168% -11% 75 24
France 147 30 42 249% 394% -29% 188 30
Netherlands 56 22 24 131% 152% -8% 65 19
Greece 2425 933 948 156% 160% -2% 2471 747
Ireland 625 561 611 2% 11% -8% 1142 529
Italy 494 125 186 166% 297% -33% 575 125
Portugal 910 521 374 144% 75% 39% 1 161 349
Spain 469 170 248 89% 177% -32% 486 170
% change from
2011
high
2011
low
* Peak in 10Y Bund yields.
Exhibit 5: Rating agencies had a busy year cutting peripheral sovereigndebt ratings further
Average* long-term local currency debt rating for peripheral countries at the end of 2009, 2010 and 18 Nov 2011;
Greece Ireland Italy Portugal Spain
2009 2010 2011
AAA
AA
BBB
BB
CCC
A
B
* Average of Moody’s, S&P and Fitch. AAA/Aaa=1, AA+/Aa1=2, etc.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 40/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Gianluca SalfordAC
(44-20) 7325-4334
Aditya Chordia (44-20) 7777-9841J.P. Morgan Securities Ltd
40
more than €120bn worth of Italian and Spanish paper
since it restarted the SMP (Exhibit 7); 2) Expansion of
EFSF and ESM lending capacity to €440bn and €500bn,respectively, with media speculation of combined fire
power 1 and still pending projects to leverage the EFSF
firepower through different mechanisms;2 3) A second
bailout package for Greece with PSI in July, which is
being renegotiated; and 4) After a mild stress test in the
summer, authorities pushed for a €100bn plan to
recapitalise banks in October.
What has been the impact of this flurry of
announcements? Since the beginning of the crisis,
policymakers’ promises of future action and surprise
ECB purchases have had a large impact both on
German yields and on peripheral spreads, buttypically the impact has been short-lived as evidenced
by sovereign spreads trading close to their widest
levels ever (Exhibit 8).
2012: Flirting with danger
We expect that the outlook for the Euro area fixed
income market will be dominated by a combination of 1)
macro developments, and 2) evolution of the sovereign
crisis and policy making response.
On the macro front, our economists are more negative on
growth than either the latest set of European Commissionforecasts or consensus, forecasting an area wide
recession that only Germany will avoid. Even though
we are far from forecasting deflation, we believe that in
2012 the ECB will be forced to do all it takes to support
the region, cutting the refi to 0.50% (Exhibit 9) in
quarterly 25bp steps.
As discussed at length in the Overview, we fear that, as
was the case in 2011, in the first part of 2012 domestic
political constraints will ensure that policymakers will
continue to be reactive rather than proactive. Italy and
Spain may lose access to the primary market, in which
case our view is that policy makers will realise themistake, step up the effort and manage to contain the
1We think it is unlikely to have the EFSF and the ESM in place
together (with combined capacity) as the German Supreme Court ruledthat Germany’s exposure should be limited and that the German
government must now obtain the approval of the Parliamentary Budget
committee before giving any guarantees. See Overview, Global Fixed Income Markets Weekly, 9 September 2011. 2 See Overview,Global Fixed Income Markets Weekly, 14 October 2011
and 28 October 2011.
Exhibit 6: The ECB started to buy Italian and Spanish bonds on8 August 2011 and currently holds almost €200bn of peripheral countrydebt in the SMP on a cash basisWeekly and cumulative ECB bond purchases through SMP*; official purchases data on
a cash basis; €bn
0
25
50
75
100
125
150
175
200
May-10 Aug-10 Nov-10 Mar-11 Jun-11 Sep-11
0
5
10
15
20
25
Weekly bond purchases (rhs)
Cumulativ e bond purchases (lhs)
* Excludes roughly €6.7bn of matured bonds.Source: ECB
Exhibit 7: We estimate that the ECB’s SMP purchases account for roughly 20% of the bond market in Greece, Ireland, and Portugal, and7% in Italy and SpainJ.P.Morgan estimate of notional amounts of peripheral bonds purchased through theECB SMP* and their current bond market size; €bn
Greece Ireland Portugal I taly Spain
Notional bought** 50 18 20 95 32
Bond market size 247 85 104 1359 474
% 20% 21% 19% 7% 7% * We estimate notional amounts using official total weekly purchases, cash prices andour estimate of country and maturity split.** Includes roughly €6.7bn of matured bonds, mostly Greek ones.Source: J.P.Morgan, ECB
Exhibit 8: Although policymakers’ announcements had a positiveimpact on markets, the effect was short-lived, as evidenced bysovereign spreads currently trading close to their widest levels ever Behaviour of the 10Y Bund and 10Y weighted peripheral spread* after significantpolicy announcements or actions; detailed announcements in grey
Date EventDays of
sell off
Peak sell
off (bp)
Days of
tightening
Peak %
tightening
03-May-10 First bailout plan for Greece 0 n/a 0 n/a
10-May-10 EFSF announcement, SMP 7 16 20+ -46%
23-Jul-10 EBA stress test 4 6 12 -8%
28-Nov-10 Irish bailout 20+ 35 0 n/a
10 -Jan-11 Discussion of comprehensive plan 20+ 42 20+ -27%
25-Mar-11 Comprehensive plan announcement 15 24 14 -8%
04-May-11 Portuguese bailout 1 2 1 -1%
15-Jul-11 EBA second stress test 0 n/a 0 n/a
21-Jul-11 EFSF 2, second Greek bailout 2 11 6 -12%
08-Aug-11 SMP 2 0 n/a 19 -17%
04-Oct-11 Discussion of comprehensive plan 2 8 46 4 -6%
26-Oct-11 Announcement of comprehensive plan 2 2 20 2 -5%
Max 20+ 46 20+ -46%
10Y wtd. peri. spreads*10Y Bund
* 10Y weighted peripheral spread computed against Germany for Greece, Ireland,Italy, Portugal, and Spain (weighted by the size of their outstanding bond market).
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 41/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Gianluca SalfordAC
(44-20) 7325-4334
Aditya Chordia (44-20) 7777-9841J.P. Morgan Securities Ltd
41
crisis. If everything plays out as expected, 2H12 should
see a marginal improvement in financial conditions.
Based on these views, we recommend investors:
1) Go long duration. All the drivers of our 10Y Bund
model point to lower yields in 2012 (see below). We
target 10Y Bunds yields at 1.25% (Exhibit 10).
2) Enter 2s/10s flatteners. We see limited scope for
much lower short-term yields and recommend flatteningexposure in 2s/10s, targeting around 100bp by mid-2012.
3) Position for further sovereign stress,
underweighting all intra-EMU countries vs.
Germany. We expect the credit curve to flatten with
wider spreads, but we favour weighting trades for the
empirical convexity of the price curve to minimise risk.
Our forecast for intra-EMU spreads is shown in
Exhibit 11.
Short-end yields: Not far from the bottom
Even with our downbeat view, we struggle to find
value at the short end of the German curve. Short-
term German yields are close to historical lows, already
pricing in 1) an ECB deposit rate at 0.25% for a
prolonged period of time, 2) extra liquidity in the system
to be abundant (Exhibit 12), and 3) a significant
premium for the perceived safety of German government
bonds.
It is unlikely that German 2Y yields will move much in
either direction in the first part of the year if our macro
forecast of declining GDP for four consecutive quarters
Exhibit 10: We forecast a trough in 10Y Bund yields at 1.25%J.P.Morgan interest rate forecast and spread vs. forwards; German benchmarksunless otherwise stated; %
18-Nov-11 1Q12 2Q12 3Q12 4Q12
2Q12 vs.
fwd (bp)
4Q12 v s.
fwd (bp)
ECB refi 1.25 0.75 0.50 0.50 0.50 n/a n/a
1M EONIA 0.65 0.38 0.32 0.32 0.32 -13 -30
2Y 0.46 0.35 0.30 0.40 0.50 -37 -44
5Y 1.10 0.75 0.70 0.85 1.05 -69 -57
10Y 1.97 1.55 1.25 1.50 1.75 -86 -48
30Y 2.61 2.15 1.95 2.25 2.50 -70 -18
2s/10s (bp) 151 120 95 110 125 -49 -4
10s/30s (bp) 64 60 70 75 75 16 30
Exhibit 11: Intra-EMU spreads will reach new highs by mid-2012Forecast of 10Y curve-adjusted spread to Germany; bp
18-Nov -11 Q2 Q4 18-Nov -11 Q2 Q4
Germany - - - 1.97 1.25 1.75 Austria 147 200 175 3.50 3.25 3.50
Belgium 282 400 350 4.80 5.25 5.25
Finland 67 100 80 2.61 2.25 2.55
France 147 225 190 3.46 3.50 3.65
Greece 2425 2400 2400 26.32 25.25 25.75
Ireland 625 850 900 8.15 9.75 10.75
Italy 494 775 625 6.95 9 .00 8.00
Netherlands 56 85 70 2.52 2.10 2.45
Portugal 910 1200 1300 11.04 13.25 14.75
Spain 469 700 600 6.70 8.25 7.75
Weighted spread* 461 625 555 - - -
Weighted peripheral spread** 699 955 855 - - -
Spread to Germ any (bp) Yield (%)
* Weighted spread computed against Germany for the 10 largest Euro area countries(weighted by the size of their outstanding bond market).
** Weighted peripheral spread computed against Germany for Greece, Ireland, Italy,Portugal and Spain (weighted by the size of their outstanding bond market).
Exhibit 12: High liquidity and significant premium for German bondsleave limited scope for a further decline in short-dated German yields
Average monthly seasonally adjusted excess liquidity* and 3M German GC-3M EONIAOIS; monthly data
€bn
0
50
100
150
200
Jan-11 Apr-11 Jul-11 Oct-11
-30
-25
-20
-15
-103M GC-OIS
Excess liquidity ; sa
* Seasonally adjusted excess liquidity = OMOs – Autonomous factors – Reserverequirements.
Exhibit 9: Our economists forecast a mild Euro area recession,declining inflation and official rates troughing at 0.50%Forecasts for 2012 real GDP growth, HICP inflation, and end-of-2012 ECB refi rate; %
-1.00
-0.50
0.00
0.50
1.00
1.50
2.00
Grow th Inflation ECB refi
J.P. Morgan Consensus EC
n/a
Source: J.P.Morgan, European Commission, Consensus Economics, Reuters
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 42/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Gianluca SalfordAC
(44-20) 7325-4334
Aditya Chordia (44-20) 7777-9841J.P. Morgan Securities Ltd
42
plays out. However, an improvement in 2H12 should
push yields higher as market participants might start to
see some light at the end of the tunnel. We forecast 2Yyields at 30-35bp until 2Q12 and at 50bp by the end of
2012. The 50bp year-end yield target for 2Y is consistent
with our expectations of 3M GC rates around 10bp and a
2Y benchmark/3M GC rate spread in line with average
since the beginning of 2010 of around 40bp (Exhibit 13).
10Y Bunds to drop to record low: Go long
In 2011, quite a few investors complained that low
German yields did not reflect fundamentals in their view.
Indeed, a decelerating but still-growing economy and
core inflation not far from the ECB target are difficult to
reconcile with 10Y Bund yields almost 200bp lower than
the 10-year average. We find that incorporating intra-
EMU spreads in a macro model explains the dramatic
decline in Bund yields over the past 2 years quite well
(Exhibits 14 and 15).
Looking ahead, all the explanatory variables point to
falling 10Y Bund yields over the next 6-12 months. On
the macro side, to be consistent with our growth forecast,
the composite PMI should fall further by 1H12, before
rebounding above 50 by the end of the year, whereas
core inflation should decline gently in 1H12, and more
aggressively in 2H12. We expect peripheral spreads to
widen further in 1H12 before retracing a bit. Pluggingthe inputs into the model would give a fair value for the
10Y Bund around 1.00% by the middle of the year
(Exhibit 16). We expect non-linearities to eventually
emerge, preventing the Bund from reaching such an
extreme level, but 10Y Bund yields at 1.25% certainly
seems possible. As discussed in detail in the demand
section below, we believe that Germany’s position as
safe haven is at risk in case of an escalation of the crisis
if policy makers are not able to contain the fallout from
the crisis.
Given our views on the short and long end of the curve,
we recommend 2s/10s flatteners, targeting around100bp mid-year. However, we have repeatedly found
that bouts of risk appetite can hurt flatteners, as was the
case in 4Q10 and in September-October 2011. When the
market is pricing in low-for-long, any short-term
improvement in sentiment is felt first at the long end of
the curve, creating a convex behaviour as the 2Y rate
catches up with a lag. Investors who seek protection from
bearish steepening should consider 1s/5s conditional bull
flatteners in swaps (see Euro Derivatives).
Exhibit 13: Under the assumption of 3M GC repo at 10bp, we forecast2Y rates at around 50bp by the end of 2012, 40bp above 3M GC repo, inline with the average of past two yearsEvolution of 2Y German par govt rate minus 3M general collateral repo rate; %
-0.40
-0.20
0.00
0.20
0.40
0.60
0.80
1.00
Jan10 Apr10 Jul10 Oct10 Jan11 Apr11 Jul11 Oct11
Average: 41bp
Exhibit 14: In our fair value model, 10Y Bunds yields are explained by acombination of macro variables and sovereign stress…10Y Bund yields regressed against composite PMI, HICP ex food and energy and 10Yweighted peripheral spread*; daily interpolated data over past 10Y; %
Beta T-stat
Intercept 0.716 9.3
Composite PMI (units) 0.031 25.0
Core inflation (%oya) 0.901 70.1
10Y weighted peripheral spread (bp) -0.003 -47.7
R-squared 83%
Standard error 0.28 * Weighted peripheral spread computed against Germany for Greece, Ireland, Italy,Portugal and Spain (weighted by the size of their outstanding bond market).
Note: In the presence of the above factors, addition of front-end yields as another explanatory variable does not materially improve the goodness of the fit.
Exhibit 15: … with intra-EMU spread widening accounting for around175bp of decline in 10Y Bund yields over the past two yearsMarginal contribution to 10Y Bund yields*: deviation from average; %
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
PMIInflation10Y wtd. peripheral spread
* See Exhibit 14 for 10Y Bund yield model details.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 43/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Gianluca SalfordAC
(44-20) 7325-4334
Aditya Chordia (44-20) 7777-9841J.P. Morgan Securities Ltd
43
Given our forecast for record-low yields, we look at the
example of Japan to analyse the behaviour of the
2s/5s/10s fly under extreme conditions. Typically, in a
scenario of unchanged monetary policy stance, the fly is
fairly directional, allowing investors to choose between
outright longs or barbell trades based on RV and relativecarry and slide considerations. However, even with
unchanged monetary policy, when interest rates drop
significantly the directionality of the 2s/5s/10s can flip,
with the 5Y underperforming a combination of 2Y
and 10Y amid falling yields; this was the case in Japan
(Exhibit 17), and for a few days in the Euro area during
mid-September. We therefore do not recommend
buying 5Y in a 50:50 2s/5s/10s fly, despite our bullish
outlook.
At the ultra-long end of the curve, we anecdotally
witnessed a reduction in client interest in 10Y+ core
bonds, with an impact on relative liquidity comparedwith shorter segments of the curve. This forced primary
dealers to increasingly use the Buxl futures as a hedging
instrument (Exhibit 18). We recommend neutral
exposure on 10s/30s curve.
Finally, on RV, we highlight that the 5s/10s curve is
trading close to the cheapest level of the past 10 years vs.
2s/30s (Exhibit 19). We recommend 5s/10s flatteners
against weighted 2s/30s steepeners.
Intra-EMU spreads: Wider in 1H12
Given the view highlighted in the Overview and above,
we believe investors should underweight all intra-
EMU countries vs. Germany in the first part of the
year. However, stronger policy action should promote a
mild reversal in most countries in 2H12 (Exhibit 11).
To refine our views, we analyse the bond market along
two axes: 1) Country selection, and 2) Maturity/coupon
selection.
Exhibit 16: Under our baseline scenario, a further drop in the PMI, lower core inflation and wider peripheral spreads should pressure Bundyields below 1.00%; however, given non-linearities, we expect Bundyields to bottom around 1.25% in 1H1210Y Bund fair value yield based on the model described in Exhibit 14 and J.P.Morganforecasts for the explanatory variables; 1H12
Level Beta
Composite PMI (units) 43.75 0.031
Core inflation (%oya) 1.50 0.901
10Y weighted peripheral spread (bp) 955 -0.003
Fair value 10Y Bund yield (%) 0.85
Exhibit 17: We do not recommend buying 5Y in 2s/5s/10s fly as abullish duration trade because the directionality can change whenyields drop substantially6M beta of 50:50 2s/5s/10s vs. 5Y JGB yield; 1 Jan 2001 – 18 Nov 2011;
-1.2
-1.0
-0.8
-0.6
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
0.00 0.50 1.00 1.50 2.00
5Y JGB yield; %
Exhibit 18: Buxl futures volumes have increased amid decliningliquidity in the cash market3M rolling average of Bund and Buxl futures volumes; ‘000s of contracts
400
600
800
1,000
1,200
1,400
1,600
2006 2007 2008 2009 2010 2011
2
3
4
5
6
7
8
9
Bund
Buxl
Exhibit 19: 5s/10s remains close to the cheapest level of the past 10years vs. 2s/30sResidual of 5s/10s regressed against 2s/30s*; German government par rates; past 10years; bp
-20
-15
-10
-5
0
5
10
15
20
2001 2003 2005 2007 2009 2011
Current: 15bp
* 5s/10s = 0.35*2s/30s – 2.94; R-squared: 96%
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 44/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Gianluca SalfordAC
(44-20) 7325-4334
Aditya Chordia (44-20) 7777-9841J.P. Morgan Securities Ltd
44
In our view, country selection hinges less than in
previous years on fundamental, medium-term
analysis and more on technical factors, due to theprogressive deterioration of market liquidity. We
define technical factors as events that change market
dynamics in a non-linear fashion, such as rating
downgrades and loss of market access among the
negatives, and explicit or implicit forms of market
support on the positive side.
We expect France to continue to underperform in relative
terms on the back of rating-sensitive selling by foreign
accounts. The importance of the rating is not confined to
France. If Italy loses market access and gets downgraded
to BBB space, as stated by rating agencies under such an
occurrence, there is likely to be further pressure on theinflation-linked segment due to potential index-inclusion
rules (see Inflation-linked Markets section), in addition to
further erosion of its investor base. As discussed in the
Overview, we expect any request for support to be
market-negative: in Greece, Ireland, and Portugal, selling
pressure accelerated after bailouts (see Exhibit 30 in the
Overview). On the positive side, positive technical
pressure can come from aggressive SMP-buying, which
could offset the market signal, and from idiosyncraticissues such as the Irish re-investment of the promissory
note proceeds.3
In terms of valuations, our updated sovereign risk index
(Exhibits 20 and 21) shows that:
1) Austria, Portugal and Spain are cheap vs. fair
value. We agree with Austria, but not on Portugal
and Spain. We believe the case for an Austrian
downgrade is dubious if the authorities take appropriate
action (see below). On Portugal, we believe that there is
a high probability that investors will be at least asked to
extend maturities of existing Portuguese bonds later inthe year, making current valuations (especially at the
short end) unattractive. Despite a low debt/GDP ratio, we
believe that Spanish bonds can come under significant
pressure if the country loses market access.
2) France, Greece and Netherlands trade on the
3 See Euro cash, Global Fixed Income Markets Weekly, 14 October 2011 for details.
Exhibit 20: Sovereign risk index: nobody is perfectJ.P.Morgan sovereign risk index* and its components; lower-than-average risk in grey;
2012
debt/GDP
(%)
2012
deficit/
GDP (%)
Total
private
debt/GDP
2012 current
account
balance (% of
GDP)
Global
competitiveness
index
Political
risk index
Industrial
action index
2012
GDPpc
(EU=100)
Data Austria 73 -3.1 150 2.8 5.14 2 0 124
Belgium 99 -4.6 131 2.1 5.20 7 79 116
Finland 52 -0.7 153 0.0 5.47 2 73 123
France 89 -5.3 138 -3.3 5.14 2 132 104
Germany 81 -1.0 111 4.4 5.41 3 6 108
Greece 198 -7.0 122 -7.9 3.92 4 132 63
Ireland 118 -8.6 341 1.5 4.77 2 38 119
Italy 121 -2.3 126 -3.0 4.43 4 35 90
Netherlands 65 -3.1 219 7.0 5.41 4 6 126
Portugal 111 -4.5 223 -5.0 4.40 1 11 54
Spain 74 -5.9 211 -3.0 4.54 0 60 80
Cross sectional z-score Sov. risk index
Austria -0.6 -0.4 -0.4 -0.7 -0.5 -0.4 -1.1 -0.9 -0.6
Belgium 0.0 0.2 -0.6 -0.6 -0.6 2.2 0.6 -0.6 0.1
Finland -1.2 -1.4 -0.3 -0.1 -1.1 -0.4 0.4 -0.9 -0.6
France -0.2 0.5 -0.5 0.7 -0.5 -0.4 1.7 -0.1 0.1
Germany -0.4 -1.3 -0.9 -1.1 -1.0 0.1 -1.0 -0.3 -0.7
Greece 2.5 1.1 -0.8 1.7 1.9 0.6 1.7 1.5 1.3
Ireland 0.5 1.8 2.4 -0.4 0.2 -0.4 -0.3 -0.7 0.4
Italy 0.6 -0.8 -0.7 0.6 0.9 0.6 -0.4 0.4 0.2
Netherlands -0.8 -0.4 0.6 -1.7 -1.0 0.6 -1.0 -1.0 -0.6
Portugal 0.3 0.1 0.7 1.0 1.0 -1.0 -0.9 1.8 0.4
Spain -0.6 0.7 0.5 0.6 0.7 -1.5 0.2 0.8 0.2 * Calculated as weighted average of cross-sectional z-spreads; debt/GDP ratio accounts for 25% of the weight, other variables have equal weight.Source: Eurostat, EC, ECB, National Central Banks, World Economic Forum, European Industrial Relation Online
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 45/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Gianluca SalfordAC
(44-20) 7325-4334
Aditya Chordia (44-20) 7777-9841J.P. Morgan Securities Ltd
45
expensive side. We agree on all accounts. We have
been highlighting for a while that France’s fundamentals
are weaker than its AAA peers. 10Y Greek bonds arefairly priced for the second PSI (with 50% haircut), but
we expect another debt restructuring in the not too distant
future. With regards to the Netherlands, we believe that
domestic support is likely to help only at the ultralong
end of the curve.
We cover the most interesting issues for the major Euro
area countries in the section below.
Austria
We believe most macro variables for Austria look
very good, and certainly consistent with a AAA
rating. However, we see two sources of weakness for Austria that need to be addressed: 1) The bond market is
small, with limited domestic participation (less than 25%
according to financial accounts data) and therefore
potentially vulnerable to the vagaries of the market; and
2) Austrian banks’ net exposure to countries outside
the Euro area, in particular to Eastern Europe,
remains the country’s Achilles’ heel (Exhibit 22). Our
working assumption is that the government has the
resources and the will to provide a stronger support
framework in place: S&P recently expressed the desire to
see non-voting participation capital that some banks
raised during the crisis transformed into more solid
ordinary capital.
Belgium
Despite investors’ focus on higher-than-average
debt/GDP ratio, most of Belgium’s economic variables
are pretty good, especially the net international
investment position. Political issues have been in the
spotlight for more than a year now, and despite the
country working well on ‘autopilot’ for quite some time,
there is a need for a consensus on roughly 3% of GDP
worth of fiscal consolidation in 2012. Discussions are
ongoing but have been so far unsuccessful. The fragility
of financial institutions is the second issue that concerns
us, especially given the amount of support that was
needed in the 2008-09 crisis, and more recently, with
Dexia (Exhibit 23).
Finland
Finland is a very solid AAA country from a macro point
of view, but is likely to be penalised vs. Germany by the
relative lack of liquidity. However, low borrowing
requirements and the potential support coming from its
extensive general government assets should limit
excessive underperformance. Finland is in a unique
Exhibit 21: We find that Austria, Portugal, and Spain trade on the cheapside vs. fundamentals, whereas Greece, Netherlands and France are onthe expensive side
10Y spread to Germany vs. fair value implied by sovereign risk index*; bp;Index Current Model Difference % difference
Austria -0.6 147 81 66 81%
Belgium 0.1 282 299 -18 -6%
Finland -0.7 67 71 -4 -6%
France 0.1 147 304 -156 -51%
Greece 1.5 2,425 4,128 -1703 -41%
Ireland 0.4 625 565 60 11%
Italy 0.2 494 399 95 24%
Netherlands -0.6 56 84 -28 -33%
Portugal 0.4 910 550 360 65%
Spain 0.1 469 298 172 58% * 10Y spread = exp (1.88*sov risk index + 5.6); R-squared: 86%; see Exhibit 20 for thecomputation of the sovereign risk index.
Exhibit 22: Austria’s main weakness lies in the net exposure of itsbanking sector to extra Euro area countries, second only to IrelandNet external assets of the banking sector; as of September 2011; %GDP
-20%
0%
20%
40%
60%
80%
A u s t r i a
B e l g i u m
F i n l a n d
F r a n c e
G e r m a n y
G r e e c e
I r e l a n d
I t a l y
N e t h e r l a n d s
P o r t u g a l
S p a i n
Source: ECB
Exhibit 23: Belgian financial institutions have received considerablestate support since the beginning of the crisisDetails of state support to major financial institutions with significant activities inBelgium; €bn
Capital Guarantees*
Asset
protection
scheme Capital Guarantees
Asset
protection
schemeDexia Belgium 3.0 91 - 4.0 54 -
Others 3.4 59 - - 36 -
KBC Belgium 7.0 - 20 - - -
Fortis** Belgium 4.7 - - - - -
Others 6.5 - - - - -
Ethias Belgium 1.5 - - - - -
2008-09 2011
* Total guarantees agreed for Dexia were €150bn in 2008-09, peak was <€100bn,currently <€30bn.** The original plan for Fortis (capital injection) was eventually changed, withnationalisation of the Dutch arm and sale of the Belgian arm to BNB Paribas.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 46/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Gianluca SalfordAC
(44-20) 7325-4334
Aditya Chordia (44-20) 7777-9841J.P. Morgan Securities Ltd
46
position among the major Euro area countries, with
general government assets exceeding liabilities by a
wide margin (Exhibit 24).
France
We believe that the main issue that investors have with
France relates to the country’s ability to hold to its AAA
rating.
Moody’s and S&P have laid the ground for an action on
France with their recent comments. Moody’s focused on
“the deterioration in debt metrics and the potential for
further contingent liabilities to emerge are exerting
pressure on the stable outlook of the government’s Aaa
debt rating”. S&P instead focused on macro
performance, suggesting that in case of a double-dip in
the Euro area, France would likely be downgraded.
We review rating actions on AAA countries over the past
five years (Exhibits 25 and 26). We find that:
1) Negative outlooks and negative watches are unlikely
to be reversed.
2) In all instances, a downgrade has been preceded by
either a negative outlook or a negative watch.
3) Downgrades have occurred with an average lag of
roughly 3 months from a negative outlook.
4) Fiscal problems are (unsurprisingly) the most quoted
reason for action, followed by deteriorating economic
outlook, problems in the financial sector which lead to
Exhibit 24: The Finnish government balance sheet is very strong asgovernment assets are significantly higher than liabilitiesNet general government assets (financial assets minus financial liabilities) of Euro areacountries; % of GDP; latest data available
-125%
-100%
-75%
-50%
-25%
0%
25%
50%
75%
A u s t r i a
B e l g i u m
F i n l a n d
F r a n c e
G e r m a n y
G r e e c e
I r e l a n d
I t a l y
N
e t h e r l a n d s
P o r t u g a l
S p a i n
Source: National financial accounts
Exhibit 25: Downgrades of a AAA country have always been precededby either a negative outlook or a negative watchNegative outlook/watch and rating actions on AAA countries over the past 5 years;
Country Agency Negativeoutlook
Negativewatch
Downgrade
Outlook -
downgrade
(months)
Watch -
downgrade
(months)
Currentrating
Iceland Moody 's 05-Mar-08 20-May -08 2.5 Baa3
Ireland Moody ' s 30-Jan-09 17-Apr-09 02-Jul-09 5.0 2.5 Ba1
Ireland S&P 09-Jan-09 30-Mar-09 2.7 BBB+
Spain Moody 's 30-Jun-10 30-Sep-10 3.0 A1
Spain S&P 12-Jan-09 19-Jan-09 0.2 AA-
UK* S&P 21-May -09 AAA
US Moody 's 02-Aug-11 13-Jul-11 Aaa
U S S&P 1 8-Apr-11 14-J ul-11 05-Aug-11 3. 6 0. 7 AA+
Av erage - - - - 3.5 1.6 * The UK outlook was changed back to stable on 26 October 2010.Source: Bloomberg, Moody’s, and S&P
Exhibit 26: Fiscal problems have been the most quoted reason for areview/watch/rating action on a AAA rated countryReasons quoted by rating agencies for review/watch/rating action on AAA countriesover the past 5 years;
Country Agency Economic FiscalFinancial
sector
Funding
costsPolitical
External
position
Technical
issues
Iceland Moody's X X
Ireland Moody's X X X X
Ireland S&P X X X
Spain Moody's X X X
Spain S&P X X X
UK S&P X X
US Moody's X X
US S&P X X X
4 6 4 2 3 1 2
X X X X X
Problems
Grand total
Potential issues for
France
Source: Moody’s and S&P
Exhibit 27: In the event of a French downgrade, we expect the 10-15Ypart of the French curve to find better support than other sectors,thanks to the large domestic insurance sector Total assets (financial and non-financial) of insurance corporations of the Euro areacountries; 2010;
€bn % GDP
Austria 119 40%
Belgium 250 67%
Finland 52 27%France 1910 96%
Germany 1549 60%
Greece 16 7%
Ireland - -
Italy 525 33%
Netherlands 414 68%
Portugal - -
Spain 274 25%
Total 5108 100%
Insurance corp. assets
Source: ECB
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 47/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Gianluca SalfordAC
(44-20) 7325-4334
Aditya Chordia (44-20) 7777-9841J.P. Morgan Securities Ltd
47
contingent liabilities, in addition to question marks about
political resolve.
5) Only the UK has been able to return to a neutral
outlook thanks to a combination of lower-than-expected
cost of contingent liabilities, strong political resolve, and
improving economic outlook.
In terms of macro and fiscal variables, France starts from
a weaker position than other AAA countries (Exhibit 20).
A further deterioration in the macro outlook and in
financial stress is likely to add to the pressure on France.
Worries about contingent liabilities, increasing cost of
funding, and political uncertainty associated with
presidential elections are likely to prompt a rating
action in 2012.
We expect pressure to come from foreign central banks’
diminished appetite for short-end paper. We are more
sanguine about the prospects for the 10-15Y part of
the curve, given the massive size of the domestic
insurance industry (Exhibit 27) and the likelihood of
asset rotation into long-dated French bonds. Official data
showed that domestic insurance companies hold around
€220bn in French government securities, out of €1.9tn of
assets.
Germany
At the moment of writing, Germany is the only Euro areacountry that has always traded like a safe-haven asset,
thanks to solid macro fundamentals but also helped by
the liquidity that its size and bond futures provide. In our
central scenario this will continue, with yields expected
to drop even further on the back of intra-Euro area
reallocation flows. However, policy makers’ inability to
provide an effective backstop and an escalation of the
crisis beyond our central scenario would likely challenge
our view: at that point, a run on the currency by extra-
Euro area investors would dent Germany’s safe-
haven status.
Greece
The appointment of a technocratic government in Greece
should pave the way for successful implementation of the
second attempted PSI even though the NPV of the new
structure might be below 30c. We expect the voluntary
bond exchange to go through, but we believe that
bond holders will eventually be asked for a further
contribution, although it is difficult to predict the
timing. The bond price curve has flattened aggressively
since the announcement of the first PSI (Exhibit 28). We
expect further selling pressure into the end of the year as
European financial institutions continue to clean their
balance sheets.
Ireland
Ireland is the success story of 2011, with 10Y spreads to
Germany close to unchanged on the year despite a selloff
at the short end. We believe the reasons for this relative
outperformance are both fundamental and technical: 1)
For the first time in 4 years, there appears to be some
stabilisation in the debt/GDP ratio trajectory; 2) The peak
in Irish outperformance coincided with heavy domestic
purchases on the back of the capital injections in the
Exhibit 28: The Greek bond price curve is very flat as investors attach ahigh probability to a debt restructuringClean-price GGB curve after announcement of first and second PSI and current*;
points
20
30
40
50
60
70
80
90
2012 2016 2020 2024 2028 2032 2036 2040
Bond maturity date
PSI 1
PSI 2
Current
* Dates used: First PSI: 22 July 2011, Second PSI: 27 October 2011 and current:18 November 2011.
Exhibit 29: We remain sceptical about Ireland’s ability to deliver theneeded fiscal consolidation from a deficit/GDP above 10% in 2011General government budget balance for Ireland and average of peripheral countries exIreland; realised numbers for 2007-2010 and forecasts for 2011-2012; % of GDP
0.1
-7.3
-14.2
-11.2-10.3
-8.6
-2.3
-4.9-6.3
-8.6
-10.6
-5.2
-16
-14
-12
-10
-8-6
-4
-2
0
2
2007 2008 2009 2010* 2011 2012
Ireland Avg. peripheral ex Ireland
*Irish deficit number for 2010 exclude the 20.1% one-off banking sector supportmeasures adjustment.Source: 2012 EC autumn forecasts
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 48/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Gianluca SalfordAC
(44-20) 7325-4334
Aditya Chordia (44-20) 7777-9841J.P. Morgan Securities Ltd
48
summer. Irish banks increased their bond portfolio by
more than 20% in August and September alone. Going
forward, we remain sceptical about Ireland’s ability todeliver the required fiscal consolidation from a
deficit/GDP above 10% (Exhibit 29), still-enormous
net external liabilities, and still-falling house prices.
We are medium-term negative on short-dated bonds as
we believe they do not discount a high enough
probability of restructuring, even in the milder form of
maturity extensions.
Italy
We cover Italy at length in the Overview and in the
demand/supply section below. The bottom line is that we
see a risk that access to the market may become
impaired, prompting significantly higher yields anddowngrades, with technicals especially negative for
inflation-linked bonds.
The Netherlands
Dutch bonds have remained, until recently, remarkably
stable compared with Germany, thanks to their strong
perceived creditworthiness. Amid overall sound macro
fundamentals, the ongoing decline in house prices raises
some concerns in the context of a heavily leveraged
household sector, as highlighted by the Dutch National
Bank in the Financial Stability Report. Going forward,
we see the emergence of further market segmentation
in the DSL market (Exhibit 30), with domesticinsurance companies and pension funds likely to keep the
long and ultra-long ends of the curve supported, and
more pressure on the short end of the curve due to lower
structural domestic demand.
Portugal
Our assessment of the latest troika’s Portugal review is
mixed: the 5.9% 2011 deficit/GDP target will only be
met through an account expedient due to expenses
overruns, but the troika expressed confidence about
hitting the 4.5% 2012 target. The need for structural
reforms was also stressed. We believe that a long period
of fiscal tightening and internal devaluation to restore
competitiveness in a highly leveraged economy is
unlikely to prove successful.
An ex-official Treasury official recently stated that
Portugal might need additional €20-25bn in addition to
the €78bn packaged agreed in the spring as public
companies are unlikely to be able to roll their debt, an
event not contemplated when the package was approved.
Given the risks of PSI in the latter part of the year
when a new package will be negotiated, we are
generally negative on Portuguese bonds, especially in
the 2-3Y part of the curve.4
Spain
Spain’s attempt at rebalancing its economy is impressive,
especially in terms of current account balance, which
moved from -9.6% of GDP in 2008 to a more
manageable -3.4% in 2011. The fiscal journey has also
been laudable, with more than 4% points of deficit
reduction between 2009 and 2011, but Spain continues to
4See Overview, Global Fixed Income Markets Weekly, 4 July 2011 for
details.
Exhibit 30: We expect domestic insurance companies to keep the longand ultra-long ends of the Dutch curve well supported, whereas theshort/intermediate part of the curve will remain under pressure due to
lack of structural domestic demandDutch bonds spread to interpolated German curve; bp
0
20
40
60
80
2012 2018 2024 2030 2036 2042
Bond maturity date
Exhibit 31: Spain’s strong domestic investor base provides support toits bonds markets% of domestic investors in central government securities in selected Euro areacountries; %
34%
19%
33%
17%
55%
20%
62%
0%
10%
20%
30%
40%
50%
60%
70%
F r a n c e
G e r m a n y
G r e e c e
I r e l a n d
I t a l y
P o r t u g a l
S p a i n
Source: J.P.Morgan estimates, national central banks, debt management agencies
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 49/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Gianluca SalfordAC
(44-20) 7325-4334
Aditya Chordia (44-20) 7777-9841J.P. Morgan Securities Ltd
49
suffer from 1) poor fiscal discipline at the local level; and
2) uncertainty about the final cost of measures to clean
banks’ balance sheets of non-performing assets. Amongthe positives for Spain, it is worth highlighting the high
percentage of domestic investors (Exhibit 31). However,
we see a risk that Spain may lose market access at
some point during the course of 2012, pushing yields
higher.
Peripheral duration neutral curve flatteners and
weighted flatteners are attractive
After a few years of a sovereign debt crisis, it is
empirically well established that credit spread curves
are directional to the level of spreads: the wider the
spread to Germany, the flatter or inverted the credit curve
(Exhibit 32). The process is very common in credit
markets when bonds transition from high-grade to high-
yield. Investors can analyse the relative merits of curve
positions (boxed vs. Germany) vs. outright spread
positions looking at relative value and/or carry and slide.
Based on our view that spreads are going to widen across
the board and our flattening view on the German curve,
we recommend duration-neutral curve flatteners in
France, Italy, and the Netherlands. Selling high-price,
short-dated bonds and buying low-price, long-dated ones
can make the strategy more attractive (see below).
Price convexity between short and long dated bondsFor investors who want to be protected in case of a
tightening of spreads to Germany that would result in a
steepening of the credit curves, a source of relative
value can be found in the empirically convex
relationship between prices at the short and long ends
of the curve. As an example, we look at 3Y and 30Y
Portuguese bonds (Exhibit 33) to highlight that 1)
initially, 30Y bond prices fall more forcefully than prices
at the short end of the curve, but 2) eventually, the short
end of the curve catches up as the probability of a short-
dated credit event increases. Based on the empirical
regression, when 3Y Portuguese prices were at 100, 90,
and 80, the sensitivity of 30Y prices were 1.7, 1.2, and
0.6, respectively. At an extreme, under most debt
restructurings, bond prices across maturities tend to
converge to similar levels.
The Italian and Spanish price curves have, until now,
exhibited limited convexity, but we expect this to
change in case of further stress, in line with the
experience of Greece, Ireland, and Portugal. Empirically,
30Y Italian bond prices are exhibiting roughly twice as
much price volatility as 3Y ones (Exhibit 34). Selling
Exhibit 32: Credit spread curves are directional to the level of spreads:the wider the spread to Germany, the flatter or inverted the credit curve.Given our view that spreads are going to widen across the board, werecommend duration-neutral curve flatteners in all countries3s/10s spread curve to Germany for the Euro area countries ex Greece, regressedagainst 3Y spread to Germany; bp
y = -0.0002x 2 - 0.014x + 7.4
R2 = 98%
-500
-400
-300
-200
-100
0
100
0 500 1000 15003Y spread to Germany; bp
Ireland
PortugalItaly
Spain
Belgium
Exhibit 33: Under stress, the price curve exhibits convex behaviour asthe short end finally catches up with the long end in terms of losses30Y Portuguese bond price vs. 3Y Portuguese bond price; s ince May 2010*; cleanprices; points
y = 0.021x 2 - 2.7x + 134.5
R2 = 94%
40
50
60
70
80
90
100
60 70 80 90 100 1103Y Portugal bond price; points
Slope: 1.7
Slope: 1.2
Slope: 0.6
* The ECB started the SMP on 10 May 2010.
Exhibit 34: We recommend selling €100mn of BTPs Jun14 vs. €50mn of BTPs Feb37 to exploit the convexity of the price curve relationshipRecommended trade weight on short BTP Jun14-long BTP Feb37 based on empiricalprice relationship since the beginning of the first ECB SMP*;
BTP Jun14 BTP Feb37 Ratio
Dirty price 95.6 69.9 0.73Empirical price beta 1 2.01
Nominal weight 100.0 49.7
Cash weight 100.0 36.3
Yield (%) 6.23 6.67
Modified duration 2.25 12.94
PVBP 2.15 9.04 * BTP Feb37 = 235.3 - 5.64*(BTP Jun14) + 0.041*(BTP Jun14)^2; R-squared: 95%;calculated using clean prices.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 50/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Gianluca SalfordAC
(44-20) 7325-4334
Aditya Chordia (44-20) 7777-9841J.P. Morgan Securities Ltd
50
€100mn of BTPs Jun14 into €50mn of BTPs Feb37
would result in a trade that is locally hedged for small
price changes and would start to be profitable in case of further sovereign stress.
How to price high- low-coupon bond yield spreads
under sovereign stress
As the sovereign crisis evolved from liquidity to
potential solvency issues, investors have been switching
out of high-coupon, high-price bonds into low-coupon,
low-price bonds with similar maturity. Exhibit 35 shows
how the yield spread of the pair with the highest coupon
mismatch (BTP Aug23-Nov23) has moved from virtually
zero to around 60bp.
Although it seems intuitive to recommend switches
between bonds with similar maturity and yield, and
different coupon and price under sovereign stress, it
is not straightforward to estimate the fair value once
yield spreads have already moved as in the example
above. In the current environment of relatively flat yield
curves like the Italian or Spanish ones, traditional
cheap/dear analysis based on bootstrapping would
suggest that wide yield spreads between high- low-
coupon bonds with similar maturities are not justified
and have a source of relative value.5
One needs to incorporate in the analysis risk free rates,the probability of a credit event throughout the bond life,
and the recovery rate in order to justify higher yields for
higher coupon bonds.
We analyse how bond maturity and coupon rates
impact yield spreads for same-maturity bonds in a
simplified world (with constant conditional probability
of default per year, constant recovery rate and constant
zero risk-free rate). We calculate NPVs for theoretical
bonds with different characteristics and compare their
fair value IRRs. Typically, for bonds with the same
maturity and fixed coupon differential: 1) the longer the
maturity of the bond the wider the yield spread (Exhibit36), 2) the lower the coupons the wider the yield spread,
3) the higher the conditional probability of debt
restructuring, the higher the impact on the yield spread
(Exhibit 37).
5In the typical textbook world of risk free government bonds and
upward sloping yield curves, high-coupon, high-price bonds should
trade with a higher yield than low-coupon, low-price bonds with similar maturities.
Exhibit 35: The yield spread between high- and low-coupon bonds hasrisen with higher Italian yieldsYield spread between BTP 9% Nov23 (current clean price 111.6) and BTP 4.75%
Aug23 (current clean price 82.6) regressed against yield of BTP 4.75% Aug23; last sixmonths; %
0
10
20
30
40
50
60
70
4.00 5.00 6.00 7.00 8.00 9.00BTP 4.75% Aug23 yield; %
y =20.1x + -87.5
R 2 = 81%
Exhibit 36: The high- low-coupon impact on yield spread is positivelycorrelated with bond maturity but negatively with the coupon rateTheoretical yield spread between two bonds with the same maturity and coupon of thefirst bond 2% lower than the coupon of the second bond under different maturities andcoupon assumptions; NPV and IRR calculations are based on 0% risk free rate,conditional default rate of 5% per year, 40% recovery rate; %
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.00 2.00 4.00 6.00 8.00 10.00
Coupon rate of low er-coupon bond; %
30Y
10Y
Exhibit 37: The high- low-coupon impact on yield spread is positivelycorrelated with the expected probability of restructuringTheoretical yield spread between two bonds with the same maturity and coupon of thefirst bond 2% lower than the coupon of the second bond under different couponassumptions and conditional default probabilities per year; NPV and IRR calculations
are based on 30Y bonds, 0% risk free rate, 40% recovery rate; %
0.00
0.20
0.40
0.60
0.80
1.00
1.20
1.40
1.60
1.80
0.00 2.00 4.00 6.00 8.00 10.00
10% conditional prob. of default
5% conditional prob. of default
Coupon rate of low er-coupon bond; %
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 51/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Gianluca SalfordAC
(44-20) 7325-4334
Aditya Chordia (44-20) 7777-9841J.P. Morgan Securities Ltd
51
We can now apply this stylised framework to two real
cases: 12Y BTPs and 20Y-25Y Bonos (Exhibit 38).
What is the fair value spread between 12Y BTPs withcoupons of 4.75% and 9.00% when the low coupon bond
is yielding around 7.00%? We assume a fixed recovery
rate of 40% and fixed risk free rate of 2%, and we
calculate the annual conditional probability of default
that is consistent with the market yield of the low coupon
bond. We then use the same parameters to calculate the
theoretical yield on the high coupon bonds6. Based on
our model, the spread should be around 75bp vs. current
market level of 60bp, making the market spread broadly
fair. What is the fair value spread between the 20Y
5.75% Bono and the 25Y 4.20% Bono when the latter is
yielding around 7.00%? Based on our model, the spread
should be around -80bp vs. current market level of -7bp,suggesting a significant mispricing. We therefore
recommend investors sell the Bono Jul32 and buy the
Bono Jan37.
In conclusion, under sovereign stress, the standard
cheap/dear analysis is not useful to deal with bonds
issued by the same issuer, with similar maturities but
with significantly different coupons. The fair value
yield spread between bonds with different coupons is
positively influenced by maturity and default probability,
and negatively influenced by the coupon rate for a
constant coupon difference. We believe the BTP Aug23-
Nov23 yield spread is not far from fair price, but theBono Jul32-Jan37 spread is too tight.
Bond demand: In whose hands is the
hot potato?
Given the increasingly technical nature of the Euro area
government bond market, we look at its investor base,
first in aggregate (Euro area) and then for the four major
countries. We conclude that despite different investor
bases, no country is in the position of Japan, where
95% of the country’s domestic investor base acts as a
potent dampener of sovereign stress (see Japan).
Euro area
According to ECB data, in mid-2011, Euro area banks
(monetary and financial institutions [MFIs], to be
precise) held roughly 23% of Euro area (general)
government securities, whereas the Euro system held 7%
of the total. It is important to highlight that SMP holdings
account for less than 50% of that amount. Euro area
6 Using inputs that are extracted from market prices is computationally
more complex, but does not change the results significantly.
insurance companies and pension funds had sizeable
holdings (18%), whereas investors outside the Euro area
held almost 1/3 of the total (Exhibit 39). We estimatethat central banks constitute the lion’s share of external
holdings. For more specific cross-border exposure in the
banking sector, please refer to Exhibits 11 and 12 in the
Overview.
Flow-wise (Exhibit 40), Euro area banks were decent
buyers of bonds in 1H11, but consistent with
anecdotal evidence, they became heavy sellers in
3Q11 (and likely in 4Q11, based on individual bank
updates). In the recent words of the president of the
Exhibit 38: We believe that the BTP 4.75% Aug23-9% Nov23 spread isclose to fair value, but the Bono 5.75% Jul32 is too expensive vs. theBono 4.2% Jan37
Market and theoretical* yields on BTPs 4.75% Aug23 and 9% Nov23 and Bonos4.20% Jan37 vs. Bono 5.75% Jul32; %
Maturity Coupon Dirty price Mkt yield Theoretical yield
BTP Aug-23 4.75 84.1 7.06 7.06
BTP Nov-23 9.00 112.2 7.65 7.81
Spread (bp) 59 75
Bono Jan-37 4.2 71.3 7.01 7.01
Bono Jul-32 5.75 88.7 7.08 7.80
Spread (bp) -7 -79 * We adjust a flat conditional probability of default under 40% recovery rate and 2%risk free rate to match the market yield of the first bond and calculate the theoreticalyield of the second bond under the same parameters.
Exhibit 39: Euro area banks and extra-Euro area investors are thelargest holders of Euro area government bonds…Split of Euro area general government bond holdings; as of 2Q11
bn %
MFIs 1,579 23%
Eurosystem (ECB, NCBs)* 469 7%
Investment funds 671 10%
Insurance companies + pension funds 1,228 18%
Others 633 9%
Rest of the World 2,183 32%
Total 6,763 100%
2Q11
* The number includes holdings other than SMP.Source: ECB
Exhibit 40: … but banks turned net sellers of bonds in 3Q11Change in Euro area general government bond holdings*; €bn
2009 2010 1H11 3Q11 4Q11*MFIs 228 41 53 -43 n/a
Eurosystem (ECB, NCBs)** 42 104 15 81 33
Insurance co. + pension funds 100 101 13 n/a n/a
Others -13 199 47 n/a n/a
Rest of the World 248 154 150 n/a n/a
Total 604 598 278 n/a n/a * SMP purchases from 1 October 2011 till 18 November 2011.** Including revaluation/devaluation changes up to 3Q11.Source: ECB
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 52/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Gianluca SalfordAC
(44-20) 7325-4334
Aditya Chordia (44-20) 7777-9841J.P. Morgan Securities Ltd
52
European Banking Federation: “The banks are doing
exactly what they should be doing: they are reducing
their risk … We can see that clearly as now Italian bondsare being sold off.” The country wise split shows that in
3Q11, French banks were the most aggressive in
reducing their government bond portfolios (-€16bn),
followed by Spanish and German ones, while Italian
banks were still adding to their government bond
portfolios.7
After a relatively quiet 1H11, the ECB had to
significantly increase purchases: the central bank releases
on a weekly basis their SMP holdings, providing the
timeliest information. We estimate that between the end
of 3Q11 and mid-November, another €33bn was bought
by the ECB in addition to €81bn in 3Q11. For the other investors, data is available only until 2Q11. It is worth
highlighting the strong buying from extra-Euro area
investors and subdued purchases by insurance companies
and pension funds in 1H11.
Germany
Given the relatively large amount of local debt in
Germany (almost 40% of the total, between securities
and loans), we focus on central government securities. In
trying to estimate central government holdings, we faced
the problem that, based on different sources, there do not
seem to be enough German government bonds and T-
bills around to satisfy all the reported demand. Weestimate that domestic holdings of central government
securities are a small percentage of the total, and foreign
central banks constitute more than one third of the
investor base (Exhibit 41).
Germany’s role of an international safe haven is
confirmed by the 6% jump in foreign holdings of general
government debt in 2010 to 60%, with a further increase
in 1Q11. However, as discussed previously, we believe
that strong international participation, in particular
from overseas investors, makes German bonds
vulnerable to an escalation of the crisis. If the fabric of
the Euro zone is perceived to be under threat, even
German bonds might lose their safe-haven status and be
sold by foreign central banks.
France
Exhibit 42 shows our best estimate of the investor
holding of French government securities.8 Nearly two
7 ECB data shows changes due to outright buying/selling and to
revaluations/devaluations. 8
See Euro cash, Global Fixed Income Markets Weekly, 21 October
2011 for more details.
thirds of French government securities are held by
foreign investors, of which the major chunk is owned by
foreign central banks, based on our estimates (28% of the
total). We believe that, from a technical point of view,
France is fragile due to the large percentage of
foreign investors who are likely to be sensitive to its
AAA rating. Official data shows that foreign holdings
have been declining since mid-2010 after steady
increases since the introduction of the Euro.
Exhibit 41: We estimate that foreign central banks alone hold more thanone third of German bonds and T-bills, making even the German bondmarket vulnerable to a severe escalation of the sovereign crisis
J.P.Morgan estimate of German central government securities holdings; as of 1Q11; €bn
Domestic Foreign Total Domestic Foreign Total
MFIs 107 210 317 10% 19% 29%
Insurance companies 40 150 190 4% 14% 17%
Mutual funds 39 116 155 4% 11% 14%
Central banks 4 395 399 0% 36% 37%
Others 20 10 30 2% 1% 3%
Total 210 881 1091 19% 81% 100%
€bn %
Source: Bundesbank, ECB, BIS, EFAMA, IMF
Exhibit 42: We estimate that almost 30% of French bonds and T-billsare held by foreign central banks that are likely to be sensitive to a
rating downgradeJ.P.Morgan estimate of French central government securities holdings; as of 1H11; €bn
Domestic Foreign Total Domestic Foreign Total
MFIs* 147 141 288 11% 11% 22%
Insurance companies 221 100 321 17% 8% 25%
Mutual funds 21 133 154 2% 10% 12%
Central banks n/a 363 363 n/a 28% 28%
Others 53 129 181 4% 10% 14%
Total 442 866 1308 34% 66% 100%
€bn %
* Domestic central bank might be under MFIsSource: AFT, ECB, BIS, EFAMA, IMF
Exhibit 43: We estimate that foreign banks are by far the largest foreign
players in Italian bonds, putting pressure on prices as they de-lever their portfoliosJ.P.Morgan estimate of Italian central government securities holdings; as of May 2011;
€bn
Domestic Foreign Total Domestic Fore ign Total
MFIs 188 189 377 13% 13% 25%
Insurance companies 165 85 250 11% 6% 17%
Mutual funds 64 77 141 4% 5% 9%
Central banks 67 66 133 4% 4% 9%
Others* 335 258 593 22% 17% 40%
Total 820 675 1494 55% 45% 100%
€bn %
* 5% of general government securities are held abroad, but is attributable to Italiansavers (4.3% of total debt).
Source: Bank of Italy, ECB, BIS, EFAMA, IMF
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 53/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Gianluca SalfordAC
(44-20) 7325-4334
Aditya Chordia (44-20) 7777-9841J.P. Morgan Securities Ltd
53
Italy
Banca d’Italia publishes lagged data on the split between
various domestic investors (MFIs, central banks, mutualfunds, insurance companies, pension funds, and others)
and foreign investors in government securities. They do
not provide the holding split among foreign investors,
who hold nearly 45% (€675bn, Exhibit 43) of
government securities as per the most recent data.
However, they do estimate that 5% of general
government securities held abroad are attributable to
Italian savers9.
We collect information from various sources to get an
idea of the foreign investor split in Italy. BIS data shows
that non-Italian banks have exposure of around €200bn
to the Italian general government, and we assume thataround 95% of that is in Italian government securities.
Euro area investment funds hold nearly €670bn of Euro
area government securities, and we estimate that
holdings of Italian bonds are around €140bn (20% of
total), based on Italy’s contribution to the JPM EMU
government bond index. IMF (CPIS) data shows that
foreign central banks held nearly €50bn of Italian
securities on their books at the end of 2009. Given that
the IMF survey captures only a small sub-sample of
reporting entities, and reserves have increased since
2009, we estimate that total numbers might be 30%
higher (around €65bn). Lastly, based on our insurance
companies’ equity analysts’ database, we estimate thatnearly €85bn of Italian government securities is held by
non-Italian insurance companies.
Flow-wise, available data on government bond securities
shows stable foreign holdings for the first 5 months of
the year, but timelier Balance of Payments data shows a
sharp decline in Italy’s foreign liabilities between July
and August (Exhibit 44). The €11bn increase in
domestic banks’ holding of government bonds in 3Q11
and aggressive purchases of Italian bonds through the
ECB SMP (we estimate €95bn of purchases of BTPs) are
also consistent with domestics and the ECB offsetting
large international selling. As discussed in the Overview,
we expect selling pressure from international investors to
continue in coming months.
9 In the latest Financial and Stability report, Banca d’Italia analysts
stated, “An estimated 4.3% (of total debt) is held by individually
managed portfolios and investment funds administered by foreignintermediaries but attributable to Italian savers.”
Spain
Tesoro Publico provides detailed domestic and foreign
investor holding data for government securities (Exhibit
45), but we use our own estimates for the split of foreign
investor types based on other sources.10 Spain is the
country with the lowest share of foreign investors (latest
data, 38%). Domestic investors, especially the domestic
banking sector, are the most important players in the
Spanish government bond market and based on availableinformation, they have continued to support the segment
in 2011. Domestic banks increased their exposure by
nearly 6%-pts, from around 28% at the end of 2010 to
33% at the end of August 2011. From a technical point
of view, Spain looks less vulnerable than Italy.
10The foreign split by investor type published by the Spanish Treasury
is based on non-resident withholding tax refunds.
Exhibit 44: Foreign investors have reduced exposure to Italian financialassets by €40bn over July-AugustCumulative changes in foreign holdings of Italian assets*; €bn
-20
0
20
40
60
80
100
120
Aug-09 Dec-09 Apr-10 Aug-10 Dec-10 Apr-11 Aug-11
* Italian government securities account for roughly 2/3rds of foreign liabilities.Source: Bank of Italy Balance of Payments data
Exhibit 45: Spanish banks are the key players in the Bono market,making Spain less vulnerable to selling pressure from internationalinvestorsJ.P.Morgan estimate of Spanish central government securities holdings; as of August2011; €bn
Domestic Foreign Total Domestic Foreign Total
MFIs 175 51 227 33% 10% 42%
Insurance companies 36 25 61 7% 5% 11%
Mutual funds 23 45 67 4% 8% 13%
Central banks 63 73 136 12% 14% 25%
Others 32 11 43 6% 2% 8%
Total 328 205 534 62% 38% 100%
% €bn
Source: Bank of Spain, ECB, BIS, EFAMA, IMF
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 54/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Gianluca SalfordAC
(44-20) 7325-4334
Aditya Chordia (44-20) 7777-9841J.P. Morgan Securities Ltd
54
Bond supply: Expected to decline
Under the assumption that no other country loses market
access, we expect around €710bn of conventional
supply in 2012, around €20bn lower than in 2011
(Exhibit 46), whereas net conventional supply would
decline by almost €75bn due to higher redemptions. Our
economists forecast Euro area central government
deficits to decline by €50bn to around €275bn in 2012
(Exhibit 47), despite their view that non-core countries
will fail to hit the targets.
The most important conclusions on conventional
supply are:
1) In core countries, we expect gross supply to remainstable or decline marginally
2) We project Italian gross conventional issuance to
decline by €5bn to around €145bn. On the other hand, we
expect Spanish gross supply to remain stable as the
increase in redemption is offset by a decline in central
government deficit.
3) Even though Irish officials have expressed their
intention to return to the issuance market in 2012, we
believe this is unlikely and expect them to remain out of
the bond market, along with Greece and Portugal, in
2012.
Other supply
We estimate that floaters, inflation-linked, and zero-
coupon bonds add another €60-65bn to Italy’s supply;
the numbers for the other countries is significantly
smaller. Italy may also increase its T-bill issuance to
meet its funding needs. For the rest of the Euro area
countries, we forecast zero net T-bill issuance due to
governments’ reluctance to further shorten the maturity
of the debt after cutting the T-bill outstanding by nearly
€50bn in 2011. €-denominated, non-conventional net
issuance would likely remain close to zero, but we expecta decline in non-€-denominated issuance, given the
declining appetite of international investors.
Most countries will announce issuance guidelines in
December, and we will publish the Euro area 2012
supply update in early January.
Italy and Spain funding in 2012
Italian redemptions in 2012 are very heavy, with around
€200bn of marketable bonds maturing (Exhibit 48).
Redemptions concentration is particularly high between
February and April (around €90bn). Based on the last
official release, Italy had around €14.5bn in Treasury’s
payment account (conto disponibilità) at the end of
September 2011. We estimate the number to be in the
range of €40-50bn by the end of 2011, based on Italy’s
net issuance since the end of September and our estimate
of Italian deficit seasonal pattern. This amount alone will
not be sufficient to cover their funding requirements in
1H12, but see Overview Appendix-2 for other sources of
funding for Italy.
Exhibit 46: Government conventional bond supply will declinemarginally in 2012Gross conventional bond issuance history and J .P.Morgan 2012 forecast*; €bn
512
584
535 533506
487
562
820 839
733711
400
450
500
550
600
650
700
750
800
850900
2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012f
* Under the assumption that no other country loses market access.
Exhibit 47: 2012 net conventional supply will decline by around €75bndue to an increase in conventional bond redemptionsGross and net conventional bond issuance* and deficit for 2011 and J.P.Morganforecasts for 2012; €bn
Gross bond
issuance
Net bond
issuanceDeficit
Gross
bond
issuance
Net bond
issuanceDeficit
Austria 20 10 10 2 0 -1
Belgium 40 16 13 2 -3 -2
Finland 13 7 7 1 0 -1
France 180 95 82 -4 -10 -10
Germany 175 18 26 -6 -16 4Greece 0 -30 18 0 -8 2
Ireland 0 -6 14 0 -1 -2
Italy 145 35 37 -5 -27 -26
Netherlands 48 18 12 -5 -6 -7
Portugal 0 -10 9 -7 -8 -1
Spain 90 49 45 1 4 -3
Total 711 202 274 -22 -75 -47
2012 v s. 2011
* Gross conventional issuance is not equal to the sum of net conventional issuanceand deficit as a part of the deficit is met by non-conventional supply and other sourcesof funding.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 55/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Gianluca SalfordAC
(44-20) 7325-4334
Aditya Chordia (44-20) 7777-9841J.P. Morgan Securities Ltd
55
In contrast, Spain has a very light redemption schedule in
1H12, giving it some time to front-load. Further, Spain’s
most liquid cash balances have remained in the €50-60bn
range since 2009, with the latest published balance of
€59bn at the end of 2Q11.
EFSF/EFSM supply
The EFSF and EFSM have issued around €45bn of bonds
so far in 2011 to fund the EU share of the Portuguese and
Irish aid packages (Exhibit 49). The bulk of the issuance
came from the EFSM as the EFSF remained out of the
markets for most of 2H11 as the new version of the EFSF
awaited parliamentary approvals. We expect the EFSM
to tap the market one last time before the year ends.
EFSF/EFSM funding requirements for 2012 to
support Ireland and Portugal are quite manageable,
at around €32bn, out of which €10bn would be covered
by the EFSM and the remaining €22bn would be
provided by the EFSF. However, the EFSF issuance
numbers may get revised significantly higher once thesecond bailout package for Greece gets finalised.
Given the EFSF’s several guarantee structure (no joint
guarantees) and significant uncertainty around Greece
PSI and the leveraged structures, EFSF bonds have
recently come under market pressure. EFSF funding
costs, which used to be around 5-10bp above LIBOR,
increased dramatically to 104bp when they last issued in
early November. In contrast, EFSM bonds, which have a
joint and several guarantee structure, have outperformed
their peers (Exhibit 50).
Trading themes
• Enter long 10Y duration trades in Germany
Outright recession in the region, falling core inflation
and a worsening of the sovereign crisis will push
German yields down in the first part of the year. We
target a trough in 10Y Bunds at 1.25% but the failure
to contain the crisis might eventually hurt even
German bonds, the last safe haven asset in the region.
Exhibit 48: Italy has a highly concentrated redemption schedule for 2012, while Spain has a light redemption schedule, which gives it anopportunity to front-loadMonthly redemption data for marketable bonds* issued by the Euro area countries;
€bn
A TS B EF F IM F RF D EM NL G GR D I EP I TL PTE ESP
Jan12 2 0 0 15 25 14 0 0 0 0 0
Feb12 0 0 0 0 0 0 0 0 36 0 1
Mar12 1 4 0 0 19 0 14 6 27 0 1
Apr12 0 0 0 18 16 0 0 0 28 0 12
May 12 0 0 0 0 0 0 9 0 1 0 0
Jun12 0 0 0 0 19 0 0 0 2 10 0
Jul12 10 0 0 28 27 15 0 0 17 1 13
Aug12 0 1 1 0 0 0 8 0 12 0 0
Sep12 0 12 6 12 21 0 0 0 11 0 2
Oct12 0 0 0 19 16 0 0 0 20 0 20
Nov 12 1 0 0 0 0 0 0 0 13 0 0Dec12 0 8 0 5 17 0 2 0 30 1 0
Total 15 25 7 98 160 30 33 6 198 13 49
Core Peripheral
* Marketable bonds include conventionals, floaters, zero-coupons, linkers, andinternational bonds.
Exhibit 49: The EFSF/EFSM has issued around €45bn of bonds so far i2011, with the bulk of the issuance coming from EFSMDetails of the EFSF and the EFSM issuances in 2011; EFSF supply highlighted ingrey; €bn
Date Issuer Maturity Size
Issuance level
(spread to mid-
swap; bp)
EMU beneficiary
05-Jan-11 EFSM 04-Dec-15 5.0 12 Ireland
25-Jan-11 EFSF 18-Jul-16 5.0 6 Ireland
17-Mar-11 EFSM 04-Apr-18 4.6* 8 Ireland
24-May-11 EFSM 04-Jun-21 4.8 14 Portugal/Ireland
25-May-11 EFSM 03-Jun-16 4.8 0 Portugal
15-Jun-11 EFSF 05-Jul-21 5.0 17 Portugal
22-Jun-11 EFSF 05-Dec-16 3.0 6 Portugal
14-Sep-11 EFSM 21-Sep-21 5.0 20 Portugal
22-Sep-11 EFSM 04-Sep-26 4.0 40 Ireland/Portugal
29-Sep-11 EFSM 04-Oct-18 1.1 15 Ireland/Portugal
07-Nov-11 EFSF 04-Feb-22 3.0 104 Ireland
Total EFSM 29.2
Total EFSF 16.0
Grand total 45.2 * Out of €4.6bn issued, €1.2bn was used for BoP and the remaining €3.4bn was givento Ireland.
Exhibit 50: EFSM bonds have outperformed their peers whereas EFSFbonds have been under market pressure, given the high level of uncertainty and the several but not joint guarantee structure
ASW spread for EFSF Jul16, EIB Jul16 and EFSM Dec15 bonds; bp
-20
0
20
40
60
80
100
May -11 J un-11 Jul-11 Sep-11 Oct-11 Nov -11
EIB Jul16
EFSF Jul16
EFSM Dec15
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 56/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Gianluca SalfordAC
(44-20) 7325-4334
Aditya Chordia (44-20) 7777-9841J.P. Morgan Securities Ltd
56
• Enter 2s/10s flatteners
The short end of the curve has little scope to rally
even if the ECB cuts the refi to 0.50%: we thereforerecommend 2s/10s flatteners, with a target of around
100bp by mid-year.
• Enter intra-EMU spread wideners
We forecast generalised spread widening vs. Germany
in 1H12.
• Enter flatteners in non-German curves
We also recommend credit curve flatteners: in line
with the experience of Greece, Ireland and Portugal,
in case of sovereign stress, we expect investors to
move from yield to price considerations. Technicals
point to flatter curves also in core countries such as
France and Netherlands, as we expect domesticsupport to the ultralong end of the curve.
• RV on peripherals: weighted flatteners and high-
coupon low coupon
We also recommend flatteners weighted by the
empirical price beta to exploit the convexity between
short-dated and long-dated bond prices. We present a
framework to analyse the yield spread of high- and
low-coupon bonds with similar maturity under
sovereign stress: We recommend selling Bonos 2032
vs. Bonos 2037 as the current yield spread is not
consistent with the coupon differential.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 57/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Fabio Bassi (44-20) 7325-8615
Khagendra Gupta (44-20) 7777-1980J.P. Morgan Securities Ltd
57
European Derivatives
• The EONIA curve was driven by ECB policy
rates and liquidity conditions in 2011
• We expect excess liquidity to remain elevated in
2012…
• …as peripheral banks continue to rely on the
ECB for funding, given limited market access and
large bank debt redemptions in 1H12
• The EONIA/refi bias is likely to remain around
70-75% of the corridor…
•
…and, with the ECB expected to cut refi anddeposit rates to 0.50% and 0.25% respectively,
we expect EONIA fixings to fall to 30bp
• We recommend longs in 6Mx6M EONIA with a
target of around 30bp…
• …and bullish option structures on Dec12 Euribor
• With peripheral sovereign stress expected to
escalate in 1H12, we expect the swap curve to bull
flatten…
• …and recommend carry efficient flatteners which
are a proxy for bullish positions but a with better
risk profile,…• …and receiver structures on 6Mx5Y swaptions
• We present a framework to analyse the
attractiveness of long-dated forward steepeners…
• …but refrain from recommending these trades
due to the prevalence of technical factors in this
part of the curve
• The 10s/30s swap curve will continue to be driven
by technical factors and is expected to flatten as
peripheral spreads widen
• Implied curve directionality has frequently
underestimated delivered directionality during
2011…
• …making conditional structures attractive at the
very front end of the curve; we recommend 1s/5s
bull flatteners
• Swap spreads will continue to be driven by
peripheral risk and are likely to eventually trade
wider than their Lehman peak
• We target 2Y and 10Y swap spreads at 145bp
and 90bp, respectively, by mid-2012
• Delivered directionality of swap spreads hashistorically exceeded implied directionality…
• …as options markets have failed to price in
greater delivered counter directionality in risk-off
environments,…
• …suggesting that conditional swap spread
wideners offer better risk/reward than outright
swap spread wideners
• Although spread markets are being driven
largely by a single-factor, i.e. sovereign risk,…
• …some spread markets such as OIS swap
spreads and 3s/6s basis are still trading far belowtheir local/Lehman peaks,…
• …suggesting that investors should express risk-
off views in these markets
• EUR volatility in the belly of the curve will
increase as the peripheral crisis escalates; we
target 3Mx10Y at 9.4bp/day, up from its current
level of 8bp/day
• Front-end volatility, however, will decline as the
ECB eventually goes on hold and excess liquidity
stays high…
• …and we recommend fading flare-ups in 2Y tail
volatility
• Favour long positions in Bund volatility vs.
swaption volatility
• On a cross-market basis, we prefer buying EUR
gamma to GBP gamma
EONIA curve
The EONIA curve in 2011 was driven primarily by
ECB activity and liquidity conditions. Early in the
year, EONIA yields drifted higher on declining excessliquidity and a hawkish ECB. The ECB hiked the refi
rate in April and July, and phased out some of the
extraordinary liquidity measures to reduce European
banks’ reliance on ECB funding. In the summer,
however, further escalation of the peripheral crisis,
declining inflationary pressures, and a deteriorating Euro
area macro outlook led the ECB to initially extend
liquidity measures and later cut refi by 25bp at its
November meeting (Exhibit 1).
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 58/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Fabio Bassi (44-20) 7325-8615
Khagendra Gupta (44-20) 7777-1980J.P. Morgan Securities Ltd
58
The dynamic of excess liquidity in 1H11 was broadly in
line with our 2011 outlook 1 of declining excess liquidity.
Indeed, non-seasonally adjusted excess liquidity
averaged €40bn and €20bn in 1Q and 2Q, respectively,
close to our forecast of €50bn in 1Q and €20bn in 2Q.
Our outlook for a further decline in excess liquidity in
2H11 was severely challenged, however. At its August
meeting, the ECB, faced with increasing concerns
surrounding peripheral sovereign debt and slowing
economic growth, provided unlimited liquidity until
1Q12. At its October meeting, it extended liquidity
measures even further and re-introduced two new 1Y
LTROs, covering liquidity funding until the beginning of
2013.
As a result of the extension of these extraordinary
liquidity measures, excess liquidity increased to around
€200bn (this is the peak 1M average of non-seasonally
adjusted excess liquidity). The increase in excess
liquidity has widened the bias between EONIA fixings
and refi once again, broadly following the dynamic
captured in our non-linear model of the EONIA/refibias.2 However, segmentation in the EONIA market has
reduced the impact of excess liquidity on the EONIA/refi
bias in recent months.
Over the past couple of years, whenever excess liquidity
crossed €120bn, the EONIA/refi bias tended to widen to
about 85% of the corridor between refi and deposit
facility (which would currently be equivalent to -65bp).
1 See Global Fixed Income Markets 2011 Outlook , 26 November 2010.2 See Global Fixed Income Markets 2011 Outlook , 26 November 2010.
More recently, however, despite high excess liquidity,the EONIA/refi bias has averaged around -55bp. For
instance, the EONIA/refi bias averaged around -45bp in
the September maintenance period and around -55bp in
the October and November maintenance periods
(Exhibit 2). In our view, this decline in the EONIA/refi
bias is driven by segmentation in the EONIA market,
resulting from peripheral sovereign stress. Indeed, the
residual from our long-term model of EONIA/refi bias
increases (EONIA/refi bias becomes less negative) as
peripheral spreads widen (Exhibit 3). Going forward,
Exhibit 2: As excess liquidity increased to over €200bn, EONIAfixings declined but remained above the level implied by our modelof EONIA/refi bias…
EONIA/refi bias* regressed against non-seasonally adjusted excess liquidity; past2Y; bp
-80
-60
-40
-20
0
20
40
60
0 100 200 300 400
Oct & Nov11 Maint period
Excess liquidity (NSA); €bn
y = -9E-06x 3 + 0.006x 2 - 1.1x
R2 = 60%
* EONIA/refi bias defined as EONIA fixings – refi rate.
Exhibit 3: …likely because peripheral stress resulted in weaker banks finding it more difficult to access funding marketsResidual* from Exhibit 2 regressed against 10Y weighted peripheral spread**; past3M; bp
y = -0.0004x 2 + 0.47x - 142.1
R2 = 37%
-30
-20
-10
0
10
20
30
300 400 500 600 700 800
10Y weighted peripheral spread; bp
18-Nov -11
* Residual obtained by regressing EONIA/refi bias against non-seasonallyadjusted excess liquidity over the past 2Y.** Weighted peripheral spread computed against Germany for Greece, Ireland,Portugal, Italy, and Spain (weighted by the size of their outstanding bond markets).
Exhibit 1: EONIA yields rose in the first half of 2011 on the back of ECB rate hikes and declining excess liquidity. Yields have fallensince early August as the ECB cut rates and re-introduced
extraordinary liquidity measures3M EONIA and ECB refi rate; 1 January 2011 – 18 November 2011; %
0.40
0.60
0.80
1.00
1.20
1.40
1.60
04-Jan 08-Mar 11-May 13-Jul 15-Sep 18-Nov
3M EONIA
Refi rate
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 59/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Fabio Bassi (44-20) 7325-8615
Khagendra Gupta (44-20) 7777-1980J.P. Morgan Securities Ltd
59
we expect this bias to stay at 70-75% of the deposit/refi
corridor (currently equivalent to -55bp) in 2012.
We believe excess liquidity will remain elevated in
2012. Peripheral banks have funded an increasingly
higher proportion of their assets at the ECB over the past
few months, and we do not expect this to change next
year. On the contrary, we expect an escalation of the
peripheral crisis, which will make wholesale funding
even more difficult for European banks, compelling them
to further increase their reliance on the ECB for funding.
The borrowing needs of peripheral European banks
for 2012 are quite heavy as there are significant
senior unsecured redemptions, especially in 1H12
(Exhibit 4). Absent a full re-opening of the unsecured bond market, we expect that most of these redemptions
will be covered by the ECB’s MRO/LTRO, and only
partially by alternative sources of funding such as retail
bonds, private placement and covered bonds.
The last 1Y tender will take place in December. With
about €140bn expiring at the 3M tender, we expect
€75bn to be rolled and about €100-125bn to be tendered
at the 1Y (13M tender), which will extend over both
2011 and 2012 year-ends. Overall, we expect an increase
in excess liquidity from these tenders of about €30-60bn
and a significant increase in the average duration of the
existing tenders.
Where does that leave the ECB in terms of policy rate?
We expect the ECB to cut the refi rate to 1.00% at the
December meeting and then to cut it a further 25bp
each quarter, down to 0.50% by June 2012, while
leaving the deposit facility rate at 0.25% after the
December cut. In that scenario, we expect EONIA to
trade as low as 32bp, with the EONIA/refi bias at 70-
75% of the deposit/refi corridor. Therefore, we see value
in long positions in mid-to-late 2012 such as 6Mx6M
EONIA (Exhibit 5). Although there is a risk that the
ECB will not cut rates below 1%, we believe long
positions at the front end are likely to offer an option-like
profile, with significant downside only if rates are left
unchanged at 1.25%.
Recent widening of the FRA/OIS basis at the front-end
of the curve has worsened the carry in long Euribor
positions, which is now negative for the first year of the
Euribor curve. Additionally, we expect funding
pressures to increase, driving FRA/OIS close to its
Lehman highs by mid-2012, before it starts to decline in
2H12. From both a carry and FRA/OIS perspective, we
see value in bullish structures only further out theEuribor curve.
In Exhibit 6, we plot our projected level of EONIA,
under the assumption that the refi rate will be cut to
0.50% by June, and the projected Euribor levels obtained
by adding to the projected EONIA curve the maximum
and minimum levels of the FRA/OIS basis since
Exhibit 4: With bank funding markets unlikely to open fully inearly 2012, we expect a reasonable portion of European banks’borrowing needs to be funded by the ECB
Cumulative total redemptions of senior unsecured bonds for peripheral*country banks in 2012; €bn
59
114
135
153
40
60
80
100
120
140
160
1Q12 2Q12 3Q12 4Q12
* Greece, Ireland, Italy, Portugal, and Spain.Source: Dealogic
Exhibit 5: We are bullish on front-end EONIA as ECB refi anddeposit rate cuts are expected to drive EONIA yields into the low30s; receive 6Mx6M forward EONIAProjected upside from outright longs across the EONIA curve; %
Market
Date ECB OIS Refi Depo EOIN A/ refi bias ECB OIS
Dec11 0.55 1.00 0.25 -0.55 0.45 0.10
Jan12 0.47 1.00 0.25 -0.55 0.45 0.02
Feb12 0.45 1.00 0.25 -0.55 0.45 0.00
Mar12 0.48 0.75 0.25 -0.37 0.38 0.10
Apr12 0.47 0.75 0.25 -0.37 0.38 0.08
May 12 0.47 0.75 0.25 -0.37 0.38 0.08
Jun12 0.51 0.50 0.25 -0.18 0.32 0.19
Jul12 0.50 0.50 0.25 -0.18 0.32 0.18
Aug12 0.51 0.50 0.25 -0.18 0.32 0.19
Sep12 0.53 0.50 0.25 -0.18 0.32 0.22
Oct12 0.55 0.50 0.25 -0.18 0.32 0.24
Nov 12 0.57 0.50 0.25 -0.18 0.32 0.26
Dec12 0.61 0.50 0.25 -0.18 0.32 0.29
J.P.Morgan Upside from
longs
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 60/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Fabio Bassi (44-20) 7325-8615
Khagendra Gupta (44-20) 7777-1980J.P. Morgan Securities Ltd
60
1 January 2009.3 We then show the current and
projected Euribor curves consistent with our FRA/OIS
assumptions. We draw the following conclusions from
this analysis. First, although the FRA/OIS is trading
close to its widest levels since 1 January 2009, we do not
see value in outright longs at the very front end of the
curve, as we expect FRA/OIS to widen further. Second,
Euribor fixings are expected to decline towards the end
of 2012, with our forecast for Dec12 Euribor at 1.00%.
Third, we see more value in bullish option structures
than outright Euribor longs.
We analyse the projected P&L of various bullish option
structures on Dec12 Euribor under various yield
projections in Exhibit 7. We draw the followingconclusions. First, call spreads are relatively expensive
and offer only modest upside in our central scenario of
Dec12 Euribor at 1%. Second, symmetric structures
such as call flies are relatively cheap and offer upside for
a wide range of Euribor fixings (0.785% to 1.215%).
Third, 1x2s are the most attractive structures as they can
be implemented at a credit and are expected to be
profitable as long as Dec12 Euribor remains above
0.68%, or a mere 3bp above the minimum Euribor fixing
level reached between June 2009 and March 2011. We
therefore recommend 1x2s on Dec12 Euribor.
Swap curve
The swap curve was driven by ECB activity and the
peripheral crisis in 2011. The ECB hiked the refi rate
in April and July, driving the curve flatter in a selloff.
Monetary policy turned at the August meeting when it
3 The period since 1 January 2009 does not include the FRA/OIS peak
witnessed during the Lehman crisis, when front FRA/OIS widened to150bp.
became clear that the ECB had gone on-hold. The 2s/10s
swap curve steepened aggressively as the market started
pricing in ECB easing. Since then, it has mostly bull–
flattened (Exhibit 8) going into the ECB cut in
November. The 10s/30s swap curve remained
uncorrelated with front-end yields, driven mostly by
technical factors. Since the beginning of the year, the
2s/10s swap curve flattened 55bp, about 35bp more
than was implied in the forward curve early in the
year, whereas 10s/30s flattened only 8bp, 12bp less
than the forward.
Exhibit 6: We expect bank funding pressures to increase,driving the FRA/OIS basis close to its Lehman highs in 1H12before declining in 2H12. On balance, we expect December 2012Euribor to trade below its current level…1) Projected EONIA levels under the assumption that the refi rate is cut to50bp, 2 ) min/max Euribor threshold based on the min/max* FRA/OIS since1 January 2009, and 3) 18 November 2011 vs. forecast y ield on front Euribor strip; %
0.20
0.40
0.60
0.80
1.00
1.20
1.40
1.60
Dec-11 Mar-12 Jun-12 Sep-12 Dec-12
Projected EONIA
EONIA + Min FRA/OIS
EONIA + Max F RA/OIS
Euribor 18 Nov
Euribor forecast
* Min/max since 1 January 2009. See footnote 3 further discussion.
Exhibit 7: …and therefore favour bullish option structures in December 2012 Euribor Upfront cost and projected P&L under different Euribor projections for various bullish option structures on Dec12 Euribor futures; bp
Projected P&L (bp) at m aturity when Euribor is
70 80 90 100 110 120 130 65* 130**
Call spread 98.750/99.000 +1/-1 14.0 11.0 11.0 11.0 11.0 1.0 -9.0 -14.0 11.0 -14.0
Call spread 98.750/99.250 +1/-1 24.5 25.5 20.5 10.5 0.5 -9.5 -19.5 -24.5 25.5 -24.51x 2 98.750/99.000 +1/-2 -7.0 2.0 12.0 22.0 32.0 22.0 12.0 7.0 -3.0 7.0
1x 2 98.750/99.250 +1/-2 14.0 31.0 31.0 21.0 11.0 1.0 -9.0 -14.0 26.0 -14.0
Ladder 98.750/99.000/99.250 +1/-1/-1 3.5 16.5 21.5 21.5 21.5 11.5 1.5 -3.5 11.5 -3.5
Fly 98.750/99.000/99.250 +1/-2/+1 3.5 -3.5 1.5 11.5 21.5 11.5 1.5 -3.5 -3.5 -3.5
Condor 98.750/99.000/99.250/99.500 +1/-1/-1/+1 7.5 12.5 17.5 17.5 17.5 7.5 -2.5 -7.5 7.5 -7.5
Dec 2012 Euribor futures; Forward = 126bp;
Cost (bp)Ty pe Strike Contracts
* Minimum Euribor fixing between June 2009 to March 2011, a period when the refi rate was 1%.** Maximum Euribor fixing between June 2009 to March 2011, a per iod when the refi rate was 1%.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 61/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Fabio Bassi (44-20) 7325-8615
Khagendra Gupta (44-20) 7777-1980J.P. Morgan Securities Ltd
61
A notable characteristic of the swap curve during
2011 has been frequent changes in curve
directionality observed on the back of central bank
activity and the peripheral crisis. The change in swap
curve directionality may be seen in the evolution of PCA
factor loadings (Exhibit 9). Early in the year, when the
market was expecting the ECB to remain on hold, the
curve was driven by the intermediate sector, flattening in
rallies and steepening in selloffs. This resulted in higher first-factor loading on 10Y yields relative to 2Y yields.
This dynamic changed in 2Q and 3Q as the ECB started
to hike rates and the curve became driven by the front
end. Consequently, the first factor loading on 2Y yields
increased relative to 10Y. This directionality has flipped
once again over the past three months, when
intermediates started driving the curve.
Going forward, we expect the ECB to eventually cut the
refi rate to 0.50% in 1H12, while keeping the deposit
facility rate at 0.25%. With the market already pricing
much of the expected easing (see EONIA curve), we
expect the front end to remain tethered and the curveto exhibit positive directionality in 2012, flattening in
a rally and steepening in a selloff .
2s/10s flatteners have been profitable on average during
2011 and their P&L performance has been correlated
with the carry at inception. Exhibit 10 shows the
success ratio and the average 3M P&L on 2s/10s
flatteners initiated at different levels of ex ante carry
during 2011. The analysis shows that flatteners have
performed better when initiated at positive carry, and
Exhibit 8: A year of flattening: the swap curve bear flattenedearly in the year on the back of ECB rate hikes, and bull flattenedlater due to escalation of the peripheral crisis
2Y swap yields and 2s/10s EUR swap curve; 1 January 2011 – 18 November 2011;% bp
1.20
1.40
1.60
1.80
2.00
2.20
2.40
2.60
Jan 11 Mar 11 May 11 Jul 11 Sep 11 Nov 11
80
100
120
140
160
1802Y sw ap rate
2s/10s swap curve
Exhibit 9: Changes in swap curve directionality may be seen in theevolution of PCA factor loadings. Going forward, with the front endexpected to remain tethered, we expect the swap curve to continueto flatten in a rally and steepen in a selloff 3M rolling 2Y and 10Y first factor loading obtained from PCA*; %
0%
10%
20%
30%
40%
50%
Jan 11 Mar 11 May 11 Jul 11 Sep 11 Nov 11
10Y
2YBear
flattening
Bull
flattening
* Principal component analysis (PCA) is a tool that recombines observed variablesinto factors such that a smaller number of these factors explain a large percentageof variation observed in the original data set. We run PCA on demeaned levels of 1Y, 2Y, 3Y, 5Y, 7Y, 10Y, 12Y, 15Y, 20Y, 25Y, and 30Y swap yields.
Exhibit 10: Over the past year, flatteners have been more profitablewhen initiated at high levels of carry; by this metric, flattenerscurrently appear attractiveSuccess ratio and statistics on P&L of 2s/10s flatteners held for three monthsrelative to the 3M carry at inception*; 1 January 2011 – 18 November 2011; bp
3M Carry <-8bp >-8bp and <0bp >8bp
# trades 88 48 25
# profitable trades 34 32 25
Success Ratio 39% 67% 100%
Av erage -3 7 29
Min -38 -28 8
Max 24 30 53
SD 17 17 15
* We split the sample of trades into three categories based on the top, the central,and the bottom third of carry level at trade inception.
Exhibit 11: Further escalation of the peripheral crisis is expected toput flattening pressure on the swap curve…2s/10s EUR swap curve regressed against 10Y weighted peripheral spreads*;1 January 2011 – 18 November 2011; bp
y = -0.08x + 156.5
R2 = 50%
80
100
120
140
160
180
200 300 400 500 600 700 800
18-Nov-11
10Y weighted peripheral spread; bp
* Weighted peripheral spread computed against Germany for Greece,Ireland, Portugal, Italy, and Spain (weighted by the size of their outstanding
bond markets).
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 62/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Fabio Bassi (44-20) 7325-8615
Khagendra Gupta (44-20) 7777-1980J.P. Morgan Securities Ltd
62
have averaged negative returns if implemented when the
carry was in the bottom third of the sample.
Another significant driver of the curve has been the
peripheral sovereign crisis. We believe that this crisis
will likely worsen in 1H12 (see Overview and Euro
Cash), resulting in wider peripheral spreads. This
will likely put flattening pressure on the Euro swap
curve (Exhibit 11). Empirical analysis over the past
year indicates that the 2s/10s swap curve tends to flatten
about 8bp for every 100bp of widening in 10Y weighted
peripheral spreads. We expect about 250bp of widening
in 10Y weighted peripheral spreads (see Euro Cash),
which will put around 25bp of flattening pressure on the
2s/10s swap curve. Additionally, 3M carry on 2s/10sflatteners is +5bp; we therefore recommend that
investors implement swap curve flatteners.
Next, we further analyse carry in the swap curve to find
the most attractive sectors to implement flatteners. Since
early 2011, carry in long positions at the front end of the
curve has flipped from positive to negative as the
fronts/reds curve has inverted. This has turned carry in
flattening trades positive, especially when anchored at
the very front end of the curve (Exhibit 12). On a
carry-to-risk basis, we believe 1s/3s and 1s/5s are the
most attractive flatteners on the swap curve. These
trades have a strong bullish bias and sport a better risk profile than outright longs.
Investors wishing to express a bullish view on the
intermediate sector may also consider call structures.
With our 5Y German benchmark yield target at 0.65%
(see Euro Cash), we see value in receiver structures on
5Y swaps despite our view that 5Y swap spreads will
widen to 130bp (see Swap spreads below). In Exhibit
13, we analyse the projected upside for various bullish
structures implemented with 6M receivers. Among
them, we prefer symmetric structures such as receiver
butterflies or condors, given their limited downside.
However, investors less concerned about 5Y swap rates
declining below 1.50%-1.60% may consider asymmetric
Exhibit 12: …and investors wishing to position for this shouldconsider flatteners that offer attractive risk-adjusted carry…
Risk-adjusted carry in outright long positions in swaps and swap curve flatteners;bp
18-Nov-11 18-Nov-11 30-Jun-11 04-Jan-11
1Y 1.62 -8 14 9 43 -0.7
2Y 1.60 -1 12 12 71 -0.1
3Y 1.71 4 13 14 82 0.2
5Y 2.09 6 11 12 93 0.3
10Y 2.68 4 7 8 102 0.2
15Y 2.96 3 6 6 107 0.1
30Y 2.82 1 3 2 116 0.0
1Y/2Y -2 7 -2 4 32 0.9
1Y/3Y 10 12 -2 5 43 1.1
1Y/5Y 47 14 -4 3 59 0.9
2Y/5Y 49 7 -1 0 39 0.7
2Y/10Y 108 5 -5 -4 60 0.4
2Y/30Y 122 2 -9 -10 78 0.1
5Y/10Y 59 -2 -3 -4 29 -0.2
Current ann. risk
adj. carry**
3M Carry (bp)Trade
Ann Risk*
(bp)
* Annualised risk is defined as sqrt(251)*3M standard deviation of daily changes.** Annualised risk-adjusted carry is defined as 4*3M Carry/Annualised risk.
Exhibit 13: …or bullish option structures on intermediate tailsCost and projected payoff of various bullish option structures on 5Y EUR swap rates; bp
Ty pe Strike Contracts Low er Upper Upside Dow nside
Call spread ATM/ATM-1C +1/-1 6.1 13.2 - 216.1 7.1 6.1
Call spread ATM/ATM-2C +1/-1 11.2 13.2 - 211.0 15.2 11.2
1x 2 ATM/ATM-1C +1/-2 -16.4 13.2 179.3 238.5 29.6 -1x 2 ATM/ATM-2C +1/-2 -6.2 13.2 163.2 228.3 32.6 -
Ladder ATM/ATM-1C/ATM-2C +1/-1/-1 -11.3 13.2 171.3 233.4 24.5 -
Ladder ATM/ATM-1C/ATM-3C +1/-1/-1 -7.1 13.2 162.3 229.2 20.3 -
Ladder ATM/ATM-1C/ATM-4C +1/-1/-1 -3.7 13.2 152.4 225.8 16.9 -
Ladder ATM/ATM-2C/ATM-4C +1/-1/-1 1.4 13.2 144.3 220.7 25.0 -
Butterfly ATM/ATM-1C/ATM-2C +1/-2/+1 1.0 13.2 196.7 221.2 12.2 1.0
Butterfly ATM/ATM-2C/ATM-4C +1/-2/+1 3.6 13.2 172.9 218.5 22.8 3.6
Condor ATM/ATM-1C/ATM-2C/ATM-3C +1/-1/-1/+1 1.9 13.2 184.4 220.2 11.3 1.9
Condor ATM/ATM-1C/ATM-3C/ATM-4C +1/-1/-1/+1 2.7 13.2 172.0 219.4 10.5 2.7
Max P&L (bp)
6Mx5Y Swaptions; Carry = 13.2bp; ATMF = 222.1bp; Spot = 208.9bp; Spot 1Y Min = 184.1bp
Yield bounds (bp)**Cost (bp)
Projected payoff
at maturity (bp)*
Note: ATM-1C refers to the swaption strike that is 1-carry (13bp) away from ATM swap rate.* As the forward rate is expected to slide to spot rate, the projected payoff at maturity is equal to the current slide.** Yield levels above and below which the trade becomes unprofitable.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 63/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Fabio Bassi (44-20) 7325-8615
Khagendra Gupta (44-20) 7777-1980J.P. Morgan Securities Ltd
63
structures such as receiver ladders, which may be
implemented at a credit.
Another set of trades that have captured investor interest
over the past year has been long-dated forward
steepeners on the swap curve. We analyse the
attractiveness of such trades by building a ‘carry
efficient frontier’ of steepener trades that: 1) offer
high risk-adjusted carry, and 2) are trading flat from
a historical perspective (Exhibit 14). In addition, given
our view that the spot 2s/10s curve will flatten, we
eliminate trades that have large positive correlation to the
2s/10s curve. The best trades are in the top right
quadrant of Exhibit 14 which shows forward curves that
are relatively flat and offer high risk-adjusted carry.
Unfortunately, however, we found that most trades havetheir long leg anchored in the 20Y+ area of the curve.
Since long dated forward curves are positively correlated
with 10s/30s, which is itself driven by technical factors
(see below), we remain cautious on long-dated
forward steepeners and refrain from recommending
trades in this sector.
Long-dated swap curve
The 10s/30s swap curve has been in a 25bp range since
the beginning of the year and is now trading close to the
bottom of the range. Various technical factors have, over
the past five years, overwhelmed the typical relationshipof the 10s/30s curve with front end rates. For example,
in mid- and late-2008, the flattening in 10s/30s was
driven by the hedging of non-inversion notes from the
dealer community.4 Additionally, hedging flows from
pension funds have frequently been an important driver
of the 10s/30s curve. Over the past couple of years,
however, 1-way CSA hedging has dominated long end
flows due to the peripheral debt crisis.5
We develop a relative value model that captures the
sensitivity of the 10s/30s curve to both macro and
technical factors using the following variables: 1) 10Y
weighted peripheral spread, 2) 2Y swap rate, 3) pension
fund coverage ratio, and 4) dummy variables for hedging
of non-inversion notes in mid-2008 and early 2009
(Exhibit 15). The betas of the regression model are
significant and have the correct signs. The model
indicates that 10s/30s tends to flatten:
4 See Global Fixed Income Markets Weekly, 25 April 2008 and 06 June
2008.5 See Global Fixed Income Markets Weekly, 30 July 2010.
Exhibit 14: Although flatteners appear attractive on the spotcurve, steepeners appear attractive on the forward curve based onvaluations and carry, especially when they have little correlation
with the spot 2s/10s curve; however, we remain cautious…Risk-adjusted carry* of various forward steepeners** vs. its 5Y z-score of relativesteepness*** (reverse axis); unitless
0.0
0.2
0.4
0.6
0.8
1.0
-1.5-1.0-0.50.00.5
3Yx5s/20s2Yx5s/20s
2Yx10s/20s
3Yx10s/20s
4Yx5s/20s
5Yx5s/20s
5Yx10s/ 20s
Relative steepness; z-score
* Risk-adjusted carry defined as 1Y carry divided by 2*SD of quarterly changeson the forward curve.** Black dots indicate curve segments with correlation lower than 20% to 2s/10s;grey dots indicate curve segments with correlation higher than 20% to 2s/10s.*** 5Y z-score; negative numbers (right side of X-axis) indicate a flat curve.We include 1Y to 5Y forward steepeners on 5s/10s, 5s/20s, 5s/30s, 10s/20s,and 10s/30s.
Exhibit 15: …as technical factors such as 1-way CSA hedgingmay drive the 10s/30s swap curve flatter if the crisis gets worse10s/30s EUR swap curve regressed against 1) 10Y weighted peripheral spread*,2 ) 2Y EUR swap yield, 3) pension fund coverage ratio, and 4) two dummy
variables for the 31 May 2008 to 15 August 2008 and the 15 September 2008 to15 January 2009 periods; past 5Y; bp
y = -0.03*(Peri sprd)-13.0*(2Y)+42.6*(PF ratio)
- 6.8(I1) - 28.3*(I2)+15.7
R2 = 74%
-40
-20
0
20
40
60
0 200 400 600 80010Y we ighted peripheral spread; bp
18-Nov -11
* Weighted peripheral spread computed against Germany for Ireland, Portugal,Italy, Spain and Greece (weighted by the size of their outstanding bondmarkets).
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 64/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Fabio Bassi (44-20) 7325-8615
Khagendra Gupta (44-20) 7777-1980J.P. Morgan Securities Ltd
64
• 3bp for every 100bp widening in the 10Y weighted
peripheral spread,
• 13bp for every 1% increase in 2Y yields,
• 4.2bp for every 10% decline in pension fund
coverage ratio, and
• 7bp and 28bp in the two episodes of hedging of non-
inversion notes of mid-2008 and early 2009,
respectively.
Anecdotal evidence suggests pension fund and exotic
desk hedging has declined recently, leaving 1-way CSA
hedging as the main driver of this segment of the curve.
With peripheral spreads expected to widen significantly
from current levels, we see risk of further flattening in
10s/30s. Specifically, we expect the 10Y weighted
peripheral spread to widen 250bp, putting around 6-
8bp of flattening pressure on the 10s/30s swap curve.
Indeed, the correlation between weighted peripheral
spreads and 30Y swap yields has become more negative
in 2H11, as the peripheral crisis has spread to Italy(Exhibit 16). Thus, our negative view on peripherals
leads us to a 10s/30s swap curve flattening bias.
1-way CSA hedging on the swap curve has richened the
30Y sector. Indeed, 30Y is trading rich on a level neutral
10s/30s/50s fly. However, historical analysis indicates
that 30Y richness is a function of peripheral stress, with
the residual of the 10s/30s/50s level neutral fly being
strongly correlated with peripheral spreads (Exhibit 17).
Given our negative view on the crisis, we believe that the
30Y sector could richen further and we recommend that
investors do not fade the move.
Implied and delivered curve directionality
With the ECB unlikely to cut rates below 0.50%, we
believe the curve will remain driven by the intermediate
sector, flattening in rallies and steepening in selloffs.
During 2011, this directionality was not fully priced in
the curve and offered several opportunities for
Exhibit 17: …since we believe that 30Y swap yields could richenfurther if peripheral sovereign spreads continue wideningResidual of 10s/30s/50s EUR swap fly regressed against 30Y EUR swap yields
vs. 10Y weighted peripheral spreads*; past 2Y; bp
y = -0.02x + 6.26
R2 = 54%
-15
-10
-5
0
5
10
0 200 400 600 800
10Y weighted peripheral spread; bp
18-Nov-11
* See Exhibit 15 for definition of weighted peripheral spread.
Exhibit 18: Implied curve directionality has frequentlyunderestimated delivered directionality in 1s/10s, makingconditional structures attractive in the past…Implied* and 3M forward looking delivered directionality** of 1s/10s EUR swapcurve; %
0%
20%
40%
60%
80%
100%
Jan 10 Apr 10 Jul 10 Oct 10 Jan 11 Apr 11 Jul 11 Nov 11
Implied direc tionality
Delivered directionality
* Implied directionality for swap curve calculated as ratio of implied volatility of 3MxLong leg/3MxShort leg -1.** Delivered directionality for swap curve calculated as 1 - 3M beta of dailychanges in Short leg regressed against daily changes in Long leg.
Exhibit 16: With the crisis spreading to Italy, the correlationbetween peripheral spreads and 30Y swap yields has becomemore negative on the back of technical flows such as 1-way CSA
hedging. We have a 10s/30s swap curve flattening bias…2M correlation of daily changes in 10Y weighted peripheral spreads to dailychanges in 30Y EUR swap yields; %
-100%
-80%
-60%
-40%
-20%
0%
J an 10 Apr 10 J ul 10 Oc t 10 J an 11 Apr 11 J ul 11 N ov 11
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 65/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Fabio Bassi (44-20) 7325-8615
Khagendra Gupta (44-20) 7777-1980J.P. Morgan Securities Ltd
65
conditional curve trades. For example, implied
directionality in the 1s/10s swap curve has been, until
recently, in a 20-50% range whereas delivered
directionality has rarely fallen below 50% (Exhibit 18).
Moreover, since delivered directionality is autocorrelated
(Exhibit 19), investors may use historical delivereddirectionality as a measure of future expected delivered
directionality, thereby providing them with a way to
judge the attractiveness of conditional curve trades at any
point in time.
For example, implied directionality is significantly below
delivered directionality for curve trades anchored at the
very front end (1Y), making such conditional curve
trades attractive. However, trades anchored further out
(2Y+) are unattractive due to recent richening of implied
directionality in such structures (Exhibit 20). Therefore,
we favour bull flatteners in 1s/2s and 1s/5s, while
avoiding other structures.
Swap spreads
Like most asset classes, swap spreads were held hostage
to market risk aversion stemming from the burgeoning
peripheral debt crisis (Exhibit 21). After staying in a
narrow 20-35bp range in the first half of the year, 10Y
swap spreads embarked on a widening trend in the
second half in response to a sharp deterioration in risk
markets, reaching a peak of 75bp. Subsequently, spreads
declined below 50bp in October as equity markets
rallied, before widening sharply in response to the threat
of a Greek referendum and significant political/market
stress in Italy. Since the end of last year, 2Y, 10Y and30Y swap spreads have widened 40, 45 and 15bp,
respectively.
Historically, we have used 5 factors to explain 10Y swap
spread movements:
• German government bond issuance,
• Amount of liquidity in the banking system, proxied
by EONIA rates,
• Directionality to 10Y Bund yields,
Exhibit 20: Implied directionality is below delivered directionalityfor curve trades anchored at the very front end, makingconditional curve trades attractive
Current implied and delivered directionality* for various curve trades; %Trades Implied Deliv ered Implied-Deliv ered
1s/2s 1% 42% -41%
1s/5s 21% 59% -38%
1s/10s 47% 67% -19%
1s/30s 65% 73% -8%
2s/5s 20% 28% -8%
2s/10s 46% 41% 5%
2s/30s 63% 51% 12%
5s/10s 22% 12% 10%
5s/30s 36% 26% 9%
10s/30s 12% 14% -2%
* See Exhibit 18 for definition of implied and delivered directionality.
Exhibit 21: Although the primary driver of swap spreads in 2011was risk aversion,*…10Y German b/m swap spreads vs. DJ Euro Stoxx 600 index (inverted); since1 January 2011 – 18 November 2011;bp points (inverted axis)
20
30
40
50
60
70
80
Jan 11 Mar 11 May 11 Jul 11 Sep 11 Nov 11
200
220
240
260
280
300
10Y ASW
Euro Stoxx 600(inverted axis)
* The correlation between DJ Euro Stoxx 600 index and 10Y swap spreads sincethe beginning of the year is around 95%.
Exhibit 19: …especially since delivered directionality isautocorrelatedSpread of ex post 3M delivered and implied 1s/10s curve directionality*
regressed against the spread of ex ante 3M delivered and implied 1s/10scurve directionality; past 2Y; %
y = 0.76x + 0.08
R2 = 57%
-20%
0%
20%
40%
60%
80%
-20% 0% 20% 40% 60% 80%3M lagged (delivered - implied) directionality ; %
18-Nov-11
* See Exhibit 18 for definition of implied and delivered directionality.Y variable: Delivered directionality at time t minus implied directionality at timet – 3M.X variable: Delivered directionality at time t – 3M minus implied directionality attime t – 3M.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 66/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Fabio Bassi (44-20) 7325-8615
Khagendra Gupta (44-20) 7777-1980J.P. Morgan Securities Ltd
66
• Risk aversion, proxied by swaption implied volatility,
and
• Swapped issuance.
Much of the widening this year, however, may be
explained by sovereign spreads. Indeed, a regression of
10Y swap spreads against 10Y weighted peripheral
spreads over the past year shows an R-squared of 80%,
up from just over 40% in 2010. This suggests that, much
like every other asset class, swap spreads are living in a
single-risk factor world.
Although swap spreads appear to be driven solely by
sovereign risk, in actuality it is impossible to distinguish
between this (relatively) new driver of swap spreads
versus traditional drivers. This is because sovereign risk is highly correlated with typical high frequency drivers
of swap spreads. Indeed, as peripheral spreads have
widened, Bund yields have fallen, implied volatility has
shot up, and swapped issuance has plummeted. Since the
beginning of the year, the correlation between peripheral
spreads and these three drivers is 70-90% (Exhibit 22).
Given these high correlations, the change in 10Y swap
spreads since end-2010 can be explained reasonably well
with traditional drivers. Between December 2010 and
November 2011, for example, 10Y German b/m swap
spreads (averaged over a 1M period) widened 26bp, of
which 25bp can be explained by high frequency factors(Exhibit 23).
Going forward, we expect swap spreads to continue to
be driven by the peripheral debt crisis which is likely
to get worse before it gets better (see Overview). Over
the past 2 years, 10Y swap spreads have widened an
average of 9bp for each 100bp widening in 10Y weighted
peripheral spreads (Exhibit 24). Given our expectation
that 10Y weighted peripheral spreads are likely to widen
from their current level of 700bp to a peak of 955bp in
1H12, we believe that 10Y swap spreads will widen to a
Exhibit 22: …high cross correlations make it is difficult todisentangle the effect on swap spreads of peripheral sovereign
spreads versus more traditional drivers…Correlations between traditional high frequency drivers of swap spreads and10Y weighted peripheral spreads; data since 1 January 2011; %
10Y swap
spreads
10Y Bund
yield
3Mx10Y
implied
v ol
Swapped
issuance*
10Y
peripheral
spread**
10Y swap spreads 100%
10Y Bund y ield -95% 100%
3Mx 10Y implied v ol 91% -96% 100%
Sw apped issuance* -77% 72% -71% 100%
10Y peripheral spread** 90% -93% 86% -68% 100%
* We use fixed-rate, €-denominated issuance by financial institutions (includingcovered bonds) and supras/agencies, plus one-half of corporate bond issuance,as an indicator of potential swapped issuance.
** Weighted peripheral spread computed against Germany for Greece, Ireland,Portugal, Italy, and Spain (weighted by the size of their outstanding bondmarkets).
Exhibit 23: …and, indeed, much of the swap spread widening in 2011 may be explained by lower Bund yields and higher implied volatility,which themselves reflected an escalation of the peripheral debt crisisReturn attribution of 10Y German b/m swap spreads between 30 December 2010 and 18 November 2011*; bp
20-day MA as of:
18-Nov -11 30-Dec-10 Chg Beta Ex p. Chg
10Y sw ap spread; bp 62 36 26
Low freq regression; 1 Jan 00 - 1 Jan 08
IMM1 EONIA; % 0.56 0.73 -0.17 11.0 -2
1Y sum of German govie issuance**; €bn 170 194 -24 -0.12 3
Exp. chg from low freq factors 1
High freq regression; 1 Jan 08 - Present
Yields (10Y Bund); % 1.91 2.97 -1.07 -19.0 20
3Mx10Y sw aption v olatility ; bp/day 7.5 6.0 1.5 4.7 7
20-day MA of sw apped issuance*** (€bn/day ) 1.3 0.5 0.8 -2.6 -2
Exp. chg from high freq factors; bp 25
Exp. chg from all factors; bp 26
* J.P.Morgan’s low frequency model regresses 10Y swap spreads against first IMM date EONIA and gross annual issuance of German government bonds over the 8Yperiod ending 1 January 2008. Residual from this low-frequency regression is computed over the period 1 January 2008 – present and regressed against the three highfrequency factors.** We use an exponential model of govie issuance as an explanatory variable because the sensitivity of swap spreads declines with increasing govie issuance. For simplicity, however, we show the overall (linear) sensitivity of spreads to issuance in the table, at current issuance levels.*** We use fixed-rate, €-denominated issuance by financial institutions (including covered bonds) and supras/agencies, plus one-half of corporate bond issuance, as anindicator of potential swapped issuance.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 67/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Fabio Bassi (44-20) 7325-8615
Khagendra Gupta (44-20) 7777-1980J.P. Morgan Securities Ltd
67
peak of 90bp. Thereafter, as the peripheral debt crisis
starts to wane, we expect 10Y swap spreads to narrow to
around 80bp by end-2012.
We also expect 2Y, 5Y and 30Y swap spreads to widen
concurrently with 10Y spreads, in line with their
historical pattern. For example, over the past two years,
2Y spreads have widened around 60bp while 10Y
spreads have widened 40bp. This suggests a ratio of 1.5
between 2Y and 10Y spreads. We use this ratio to
project 2Y spreads. Exhibit 25 shows our swap spread
forecast. Given that we believe the sovereign debt crisis
will continue to worsen in 1H12, swap spreads across the
curve are likely to widen, with the front end being the
most susceptible to spread widening. We therefore
recommend that investors position for wider swap
spreads in anticipation of further peripheral stress,
especially in the front part of the swap spread curve.
Correlations between various spread markets have
skyrocketed as the peripheral crisis has intensified.
Exhibit 26 shows 10Y weighted peripheral spreads
versus the % of variability explained by the first factor in
a rolling 6M PCA6 performed on various spread
6 Principal component analysis (PCA) is a tool that recombinesobserved variables into factors such that a smaller number of these
factors explain a large percentage of variation observed in the originaldata set.
markets.7 The % of variability explained by the first
factor has jumped from a low of 50% to its current levelof around 90%, suggesting that spreads have increasingly
been driven by a single factor – sovereign risk. Given
this dynamic, investors who are bearish on the
peripheral crisis should consider positioning for
wider spreads in markets which have lagged the
recent widening in peripheral spreads, and are trading
far from their peak levels. Such spreads have more
7 We run PCA on the correlation matrix of spread levels shown in
Exhibit 27 below (not including bank CDS and Euro Stoxx bank index).
Exhibit 25: …and, given our view that the peripheral debt crisiswill get worse before it gets better, we believe that swap spread
widening has more to goJ.P.Morgan swap spread forecast*; bp
18 Nov 11 1Q12 2Q12 3Q12 4Q12
10Y w td peripheral** spread 699 835 955 900 855
2Y Sw ap spreads 110 130 145 135 130
5Y Sw ap spreads 96 115 130 120 115
10Y Sw ap spreads 67 80 90 85 80
30Y Sw ap spreads 19 25 30 25 25
* We use the model in Exhibit 24 above and our projections on 10Y weightedperipheral spreads (see Euro Cash) to forecast 10Y swap spreads. Swapspreads at other maturities (such as 2Y) are projected by assuming that theywiden proportionally based on their widening relative to 10Y swap spreads over the past 2 years.** See Exhibit 22 for definition of weighted peripheral spreads.
Exhibit 26: Although correlations between various spreads haveskyrocketed as the peripheral debt crisis has intensified…10Y weighted peripheral spreads* versus % of variability explained by the firstfactor in a rolling 6M PCA** on various spreads***;bp %
200
300
400
500
600
700
800
Jan 11 Mar 11 May 11 Jul 11 Sep 11 Nov 11
40
50
60
70
80
90
100% v ariability explained
by 1st PCA factor
Weighted peripheral
spread
* See Exhibit 22 for definition of weighted peripheral spreads.** Principal component analysis (PCA) is a tool that recombines observedvariables into factors such that a smaller number of these factors explain a largepercentage of variation observed in the original data set.*** We run PCA on the correlation matrix of spread levels shown in Exhibit 27below (not including bank CDS and Euro Stoxx bank index).
Exhibit 24: Going forward, we expect swap spreads to continue tobe driven by the peripheral debt crisis, which has caused 10Y
swap spreads to widen an average of 9bp for each 100bp wideningin peripheral sovereign spreads…10Y German b/m swap spreads regressed against 10Y weighted peripheralspreads*; past 2Y; bp
y = 0.09x + 7.7
R2 = 82%
0
20
40
60
80
0 200 400 600 800
10Y weighted peripheral spread*; bp
* See Exhibit 22 for definition of weighted peripheral spreads.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 68/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Fabio Bassi (44-20) 7325-8615
Khagendra Gupta (44-20) 7777-1980J.P. Morgan Securities Ltd
68
potential to widen than those that are trading at, or close
to, their wides.
Exhibit 27 shows how far various market variables are,
on average, from their local and Lehman peaks.8 We
examine FRA/OIS basis, Libor and OIS swap spreads,
interest rate bases, FX bases, sovereign spreads,
European bank CDS, and European bank equities.
Markets that appear to be farthest away from their
local/Lehman peaks are OIS swap spreads and 3s/6s
basis. We therefore recommend that investors consider
positioning for a widening of these spreads. On the other
hand, FRA/OIS, sovereign spreads and European bank
CDS spreads are close to their wides; investors who
believe that the sovereign debt crisis will subside should
consider positioning in these markets.
As the sovereign debt crisis has progressed, the optionsmarket has frequently underpriced swap spread
directionality. As Exhibit 28 shows, the implied
directionality of 10Y swap spreads has consistently
stayed above delivered directionality. This is because
markets have failed to account for the fact that swap
spread counter directionality increases significantly
during stressed environments as investors rush into the
8 Distance from peak is defined as (Peak minus current level) / (Peak
minus 1 June 2011 level). 1 June 2011 was roughly the start of this legof the sovereign debt crisis.
Bund contract (Exhibit 29). Since market stress has
increased almost monotonically in the past few months,
swap spread directionality has increased concurrently
(that is, become more negative). Given that we expect
market stress to increase in 2012, we recommend
Exhibit 28: As the sovereign debt crisis has progressed, theoptions market has consistently underpriced swap spreaddirectionality…2-week MA of implied directionality of front Bund swap spreads versus 3Mdelivered directionality*; past 1Y; %
-40
-30
-20
-10
0
10
Jan 11 Mar 11 May 11 Jul 11 Sep 11 Nov 11
Implied direc tionality
Delivered directionality
* Implied directionality computed as (implied vol on maturity-matched swaption /implied vol on Bund) – 1. Delivered directionality computed as rolling 3M beta of daily change in swap spread regressed against daily change in CTD spot yield.Contract is rolled 10-days before option expiration.
Exhibit 27: …some spread markets remain further away from their local and Lehman peaks than others. Investors who are bearish on theperipheral crisis, should consider risk-aversion trades where levels are far from peak levels achieved during periods of stress. On this basis,
OIS swap spread wideners appear attractive, followed by 3s/6s basis wideners Average distance of various spreads from local and Lehman peaks; %
Peak level* Distance from peak**
18 Nov 11 01 Jun 11 Since 1-Jun-11 LEH times Since 1-Jun-11 LEH times Av erage
1 IMM2 FRA/OIS 77 24 77 93 0% 24% 12%
2 IMM6 FRA/OIS 60 28 60 62 0% 8% 4%
3 2Y Libor sw ap spreads 110 53 112 118 2% 11% 7%
4 10Y Libor Sw ap spreads 67 33 78 82 24% 30% 27%
5 2Y OIS sw ap spreads 18 10 25 41 46% 74% 60%
6 10Y OIS sw ap spreads 13 -2 35 40 59% 63% 61%
7 1Y 1s/3s basis 33 14 33 54 1% 53% 27%
8 1Y 3s/6s basis 29 15 35 44 26% 51% 39%
9 1Y EUR/USD Fx basis -87 -26 -88 -123 2% 38% 20%
10 2Y EUR/USD Fx basis -72 -26 -73 -87 3% 25% 14%
Average 16% 38% 27%
11 10Y w td peri spread 699 353 739 - 10% - 10%
12 Bank CDS*** 312 123 322 322 5% 5% 5%
13 Euro Stox x Bank Index 125 193 119 90 8% 34% 21%
* Peak level between 1 June 2011 and 18 November 2011 and past 5Y to capture Lehman peaks.** Distance from peak defined as (Peak minus current level) / (Peak minus 1 June 2011 level).*** Average of 5Y CDS on Deutsche, UBS, BNP, Barclays, Santander, and Unicredit.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 69/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Fabio Bassi (44-20) 7325-8615
Khagendra Gupta (44-20) 7777-1980J.P. Morgan Securities Ltd
69
conditional bull wideners as a less risky alternative to
outright swap spread widening trades.
Euro volatility
Much like 2010, 2011 was not a year to buy swaption
gamma since such a strategy would have yielded flat
returns with large P&L swings (Exhibit 30). On
average, losses from buying volatility during risk-on
episodes offset gains collected during risk-off
episodes, resulting in high P&L volatility (Exhibit 31).
We analyse returns from buying 3Mx10Y swaption
gamma during various risk-on and risk-off episodes from
January 2010, a period during which the sovereign debt
crisis dominated market sentiment. Exhibit 32 lists the
date ranges used for this analysis. We find that, on
average, buying gamma during risk-off periods is indeed profitable.9 For instance, buying 3Mx10Y gamma has
returned an average of 40bp of notional during these risk-
9 We note that there are episodes when volatility markets initially do
not respond in an expected manner at the onset of a risk-off regime. Nevertheless, on the back of continued stress, they eventually do catch
up, resulting in positive returns to long volatility positions. For
example, during the risk-off episode in May 2011, volatility markets
were unmoved by discussions surrounding the Greek PSI even thoughequity markets were trended lower. Long gamma positions resulted in
losses during this period. However, volatility shot up as soon asconcerns surrounding Italy came to the fore in July.
off episodes. On the other hand, buying volatility during
risk-on episodes has resulted in an average loss of 75bp
of notional (Exhibit 31).
Drivers of volatility in 2011
The sovereign debt crisis and the ECB were the
dominant drivers of the volatility surface in 2011
(Exhibit 33). Volatility in the belly of the curve headed
lower in 1Q11 as markets expected the peripheral crisis
Exhibit 30: Despite heightened uncertainty and peripheral stress,2011 turned out to be a mediocre year for outright long EURgamma positions…
Average 1M returns* and annualised information ratio (IR) from buying EUR3Mx10Y swaption straddles;bp of notional information ratios
-60
-40
-20
0
20
40
2001 2003 2005 2007 2009 2011
-6
-4
-2
0
2
4 Av erage (left)
Information ratio (right)
* Trades entered daily and held for 1M. Returns calculated using J.P.Morganvolatility indices, which assume daily delta-hedging and zero transaction cost.Options are re-struck at the beginning of each month.
Exhibit 31: …as positive returns from buying gamma during risk-off periods was offset by losses during risk-on episodes
Average* cumulative return** from buying 3Mx10Y EUR swaption straddles; bpof notional
-80
-60
-40
-20
0
20
40
60
0% 20% 40% 60% 80% 100%
% days into risk on/off period; %
Risk-off
Risk-on
* Averaged over risk-on/off episodes (see Exhibit 32 below for definition of dates).** See Exhibit 30 for definition of returns.
Exhibit 29: …as markets have failed to take into accountincreasing counter directionality of swap spreads in stressedenvironments. Going forward, we recommend conditional swap
spread trades as a less risky alternative to outright trades3M delivered directionality* of 10Y swap spreads versus 20D MA of 10Yweighted peripheral spreads (inverted axis);% bp (inverted axis)
-35
-30
-25
-20
-15
-10
-5
Jan 11 Mar 11 May 11 Jul 11 Sep 11 Nov 11
200
300
400
500
600
700
Delivered directionality
Weighted peripheral
spread (inv erted axis)
* Delivered directionality computed as rolling 3M beta of daily change in swapspread regressed against daily change in CTD spot yield. Contract is rolled 10-days before option expiration.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 70/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Fabio Bassi (44-20) 7325-8615
Khagendra Gupta (44-20) 7777-1980J.P. Morgan Securities Ltd
70
to be contained to Greece, Ireland and Portugal.
However, as discussions around Greek PSI gained
momentum, implied volatility inched higher and finally
spiked up as contagion spread to Italy and Spain.
The front end of the volatility curve was, however,
equally impacted by both economic data and the
peripheral crisis. An improving outlook for globalgrowth and rising inflation expectations led the ECB to
start hiking policy rates in April. Higher rates and the
escalating peripheral crisis drove front end volatility
higher. Later on, as the Euro area economy started to
deteriorate, the ECB reacted by extending extraordinary
liquidity measures and eventually cutting policy rates in
November. At that point, front-end volatility decoupled
from peripheral stress as it became clear that the ECB
was likely to remain in a low-for-long mode.
Trade the belly of the volatility curve from the long
side
Our baseline view for 2012 is that the peripheral crisis
will worsen leading to higher implied volatility in the
belly of the curve (see 2012 targets below). While
implied volatility will likely increase in tandem with
peripheral spreads, delivered volatility in the belly of the
curve is also expected to remain high for two reasons.
First, event risk remains high. Even if policy makers
come up with a resolution to the crisis, implementation
risks of such measures will remain high. Second, bank
funding pressures and thin markets will support delivered
volatility. Thus, we recommend trading the belly of
the volatility curve from the long side in 2012.
The ex-ante level of 1) implied volatility, 2) weighted
peripheral spread, and 3) 2s/10s swap curve were the
best predictors of volatility returns in 2011. Buying
gamma at lower levels of implieds yielded better returns.
Similarly, buying gamma when peripheral spreads are
high and the 2s/10s curve is steep resulted in better
Exhibit 32: Risk-on/off episodes from January 2010 to presentStart and end dates for various risk-on/off episodes since 1 January 2010, and% change in Euro Stoxx50 between start and end dates (%);
Risk-on
Start End # months Start End % change*
05-Feb-10 15-Apr-10 2.3 2632 3013 14%
30-Nov -10 18-Feb-11 2.7 2651 3068 16%
16-Mar-11 02-May -11 1.6 2721 3009 11%
12-Sep-11 28-Oct-11 1.6 1995 2462 23%
Av erage 2.1 16%
Risk-off
04-Jan-10 05-Feb-10 1.1 3018 2632 -13%
15-Apr-10 24-May -10 1.3 3013 2558 -15%
18-Feb-11 16-Mar-11 0.9 3068 2721 -11%
02-May -11 12-Sep-11 4.5 3009 1995 -34%
28-Oct-11 18-Nov -11 0.7 2462 2237 -9%
Av erage 1.7 -16%
Euro Stoxx 50Date
* % Change of Euro Stoxx 50 defined as (End level – Start level) / Start level.
Exhibit 33: Implieds on longer tails were boosted by the sovereigndebt crisis, whereas implieds on shorter tails were negativelyimpacted by high excess liquidity
EUR 3Mx10Y and 3Mx2Y implied volatility versus 10Y weighted peripheralspreads*; 1 January 2011 – 18 November 2011;bp/day bp
4
5
6
7
8
Jan 11 Mar 11 May 11 Jul 11 Sep 11 Nov 11
200
300
400
500
600
700
800
3Mx2Y
3Mx 10Y Peripheral spread
* 10Y weighted peripheral spread computed against Germany for Greece,Ireland, Portugal, Italy, and Spain (weighted by the size of their outstanding bondmarkets).
Exhibit 34: Long EUR volatility returns have been driven by thelevel of implieds and peripheral spreads at trade initiation…Rolling 1M return* from buying EUR 3Mx10Y volatility regressed against 1Mlagged levels of 1) EUR 3Mx10Y implied volatility, 2 )10Y weighted peripheralspread**, and 3) 2s/10s EUR swap curve; 1 February 2011– 18 November 2011;bp of notional
y = -52.3*(3Mx10) + 0.67*(Peri sprd) + 2.1*(2s/10s) - 207
R2 = 68%
-700
-650
-600
-550
-500
-450
-400
4 5 6 7 8
1M lagged 3Mx10Y implied vol; bp/day
* See Exhibit 30 for definition of returns.
** See Exhibit 33 for definition of weighted peripheral spreads.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 71/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Fabio Bassi (44-20) 7325-8615
Khagendra Gupta (44-20) 7777-1980J.P. Morgan Securities Ltd
71
returns. The intuition behind this is that wide peripheral
spreads and a steep curve results in high subsequent
delivered volatility, leading to improved returns.Together these three factors explain about 2/3rds of the
total variability in returns in 2011 (Exhibit 34).
Using the above relationship we develop a framework for
identifying trading opportunities in 2012 (Exhibit 35).
Over the past few months, the 2s/10s swap curve has
exhibited a strong relationship to peripheral spreads (see
Exhibit 11). Thus, in our model, we replace the 2s/10s
curve with peripheral spreads. We then use the modified
2-factor model to calculate the level of implied volatility
that generates flat returns over the next month given
current level of peripheral spreads. We find that buying
gamma when the combination of implied volatility and peripheral spreads is below the -1 standard deviation line
is generally profitable. Strong sell signals are rarely
generated because of the overall bias towards higher
volatility and wider peripheral spreads over the past few
months. For 2012, we therefore recommend initiating
gamma trades if the combination of implied volatility
and peripheral spread is outside the 1 standard
deviation line.
With the peripheral debt crisis expected to escalate, and
10Y weighted peripheral spreads targeted to widen by
250bp, we target implied volatility on 3Mx10Y at
9.4bp/day (Exhibit 36).10
Fade volatility flare-ups at the front end in 2012
With excess liquidity high and the ECB expected to be
on perma-hold starting in 2H12 (see EONIA curve),
front-end yields are likely to stay rangebound, leading to
a low volatility environment (Exhibit 37). We,
therefore, target 3Mx2Y implied volatility at 4bp/day
and recommend short gamma positions on 2Y tails.
A risk to this outlook comes from FRA/OIS volatility
which may stay elevated due to bank funding pressures
(see Swap spreads). Front-end volatility is likely to
flare-up during episodes of FRA/OIS widening. We
recommend fading such moves.
10 We use the following model to arrive at our target for 10Y implied
volatility: 3Mx10Y = 0.0061*(10Y weighted peripheral spread) + 3.56;R-sqr = 73%; 1 Jan 2010 – 18 November 2011.
Exhibit 35: …leading us to recommend buying volatility when thecombination of implied volatility and peripheral spreads is 1SDbelow the zero iso-return line
3Mx10Y EUR volatility versus 10Y weighted peripheral spreads* at initiation of long gamma trades in 2011; solid line is t he zero iso-return line and light lines
are ±1SD lines**; bp/day
4
5
6
7
8
9
10
300 400 500 600 700
Iso-return line
<0 returns
>0 returns
Current
Sell volatility
Buy v olatilityStay
neutral
Weighted peripheral spread; bp
18-Nov-11
* See Exhibit 33 for definition of weighted peripheral spreads.** Based on regression beta and standard errors in Exhibit 34.Note: we reduced our 3-factor model to a 2-factor model as follows. The 2s/10sswap curve has exhibited strong directionality with peripheral spreads over thepast few weeks (2s/10s swap curve = -0.0014*(10Y wtd peripheral spread) +1.94; 8 August 2011–18 November 2011; R2 = 71%). We use this relationshipto replace the 2s/10s curve with peripheral spreads. We then generate the zeroiso-return line based on implied volatility and peripheral spreads alone.
Exhibit 36: EUR implied volatility targets for 2012EUR implied volatility targets for 1Q12 and 2Q12*; bp/day
18 Nov 11 1Q target 2Q target 2Q target - current
3Mx 1Y 5.4 4.0 3.5 -1.9
3Mx 2Y 5.5 4.5 4.0 -1.5
3Mx 5Y 6.6 6.6 6.7 0.1
3Mx 10Y 8.0 8.7 9.4 1.4
3Mx 30Y 8.9 9.8 10.6 1.7
2Yx 2Y 5.9 5.5 5.2 -0.7
2Yx 5Y 6.1 6.0 6.1 0.0
2Yx 10Y 6.4 6.6 6.7 0.4
5Yx 5Y 5.6 5.5 5.6 0.0
5Yx 10Y 5.6 5.7 5.7 0.0
10Yx 10Y 5.0 5.0 5.0 0.0
* Target for 3Mx10Y is based on its relationship with the 10Y weightedperipheral spread and our forecast of this spread in 2012 (see Euro Cash). For 30Y, we assume that the beta between 30Y and 10Y remains at 120%. 2Yimplieds in 2Q is the average implied of 3Mx2Y for 1H10 – a period when theECB was firmly on hold and excess liquidity was high. 5Y is computed as anaverage of 2Y and 10Y.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 72/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Fabio Bassi (44-20) 7325-8615
Khagendra Gupta (44-20) 7777-1980J.P. Morgan Securities Ltd
72
With front end swap yields range-bound (Exhibit 38),
mean reversion may be an attractive source of return as
delta-hedging short volatility positions infrequently could
boost performance. Indeed, 2Y swap yields have been
about 30% less volatile measured on a periodic basis than
on a daily basis.11 Exhibit 39 shows statistics on 1M
rolling returns from selling 3Mx2Y EUR gamma using
time-based and delta-based hedging strategies. We draw
the following conclusions from this analysis. First,delta-hedging less frequently provides higher risk-
adjusted returns than daily delta-hedging. Second, re-
setting hedges based on a delta threshold is better than
using a time-based rule. Thus, going into 2012, we
recommend infrequent delta hedging of short
volatility positions at the front end of the volatility
curve.
Stay neutral on ultra long tails
The 30Y sector remains the ‘forbidden fruit’ of European
volatility markets because volatility on 30Y tails has
historically been impacted by technical factors such asnon-inversion note hedging, pension fund receiving, and
1-way CSA hedging. Due to these technical factors,
implied volatility on 30Y tail tends to increase in excess
of 10Y tails whenever we encounter a crisis
(Exhibit 40).
11 For instance, 1M delivered volatility of 2Y swaps measured on 5D
changes is currently at 3.5bp/day (after adjusting by sqrt(5)) while thatmeasured on daily changes is at 5bp/day.
Exhibit 37: Look for delivered volatility at the front end to declinedue to exceedingly high excess liquidity; fade volatility flare-upsin 2Y tails
1M delivered volatility on 3Mx2Y EUR swaps regressed against 1M deliveredvolatility of 3Mx3M EONIA; 1 January 2009 – 18 November 2011; bp/day
y = 0.71x + 1.78
R2 = 64%
1
2
3
4
5
6
7
8
0 1 2 3 4 5 6 71M deliv v ol of 3Mx 3M EONIA; bp/day
18-Nov-11
Exhibit 38: High excess liquidity is also expected to keep frontend rates in a range,…EUR 2Y swap yield; %
1.35
1.40
1.45
1.50
1.55
1.60
1.65
05-Aug 26-Aug 16-Sep 07-Oct 28-Oct 18-Nov
Exhibit 39: …and we recommend enhancing returns of a shortvolatility position in 2Y tails by delta-hedging infrequentlyStatistics on 1M rolling returns* from selling 3Mx2Y EUR swaptions usingvarious time-based and delta-based hedging strategies; 8 August 2011– 18November 2011; bp of notional
Av erage SD Information Ratio
Days based hedging rule
Daily 9.3 8.8 3.7
Weekly 16.3 8.6 6.5
Fortnightly 18.0 10.8 5.8
Delta threshold based hedging rule
10% 9.5 8.8 3.7
30% 15.0 9.9 5.2
50% 20.2 9.6 7.3
* See Exhibit 30 for definition of returns.
Exhibit 40: Implieds on 30Y tails have lost some of their sensitivity to 10Y tails as technical factors have declined; weexpect technicals to remain a driver of 30Y tails, however, andsuggest a neutral view in this sector Rolling 3M beta from regressing daily changes in EUR 3Mx30Y implied volatilityagainst daily changes in EUR 3Mx10Y implied volatility; unitless
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
Jan 08 Oct 08 Jul 09 Apr 10 Jan 11 Nov 11
1-way CSA
hedging
Non-inversion
note hedging
Current
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 73/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Fabio Bassi (44-20) 7325-8615
Khagendra Gupta (44-20) 7777-1980J.P. Morgan Securities Ltd
73
The sensitivity of 30Y tail volatility to 10Y tail volatility
appears to have declined somewhat over the past few
months. This may be because some of the technicalflows supporting 30Y tails have declined recently (for
example, pension funds refuse to receive at these low
levels of yields). Unfortunately, however, other hedging
flows such as 1-way CSA flows have continued to drive
30Y yields lower (and volatility higher). Going forward,
it is difficult to judge the evolution of these technical
flows and therefore we recommend staying away from
30Y tails.
Buy Bund volatility vs. swaption volatility as a source
of relative value
2011 has been a year of govie volatility. Regular boutsof flight-to-quality have helped long gamma positions in
govie options produce high risk-adjusted return,
especially relative to swaption gamma. Exhibit 41
shows the average monthly returns and annualised
information ratio from several such trades. We draw
three main conclusions from this analysis. First, buying
govie gamma outright was profitable across the curve,
with Bunds outperforming Bobl and Schatz. Second,
buying govie gamma vs. maturity-matched swaption
gamma was also profitable. Here again a strategy
involving Bunds was the clear winner. Finally, returns
from buying govie gamma vs. swaption gamma were
more stable than buying outright govie gamma, i.e., theyyielded superior information ratios.
Swap spread directionality was consistently negative in
2011 and generally exceeded implied directionality (see
Swap spreads). This disparity, which was most
prominent in Bunds, was the primary reason for the
outperformance of govie vs. swaption gamma. We
expect this disparity to continue in 2012 and recommend
buying Bund gamma vs. swaption gamma as an
efficient way to position for an escalation of the
peripheral crisis.
To help generate a trading signal, we develop a
framework akin to the one in Exhibit 35. As a first step,
we estimate that 60% of the variability in returns is
explained by peripheral spreads and implied volatility
differential (defined as govie implied minus swaption
implied) at trade initiation (Exhibit 42). The regression
signs are intuitive; buying govie vs. swaption gamma at
low implied volatility differential and wide peripheral
spreads is attractive. This is because wide peripheral
spreads indicate stress in the system which typically
results in high subsequent delivered volatility differential
as asset swaps become increasingly counter directional.
Exhibit 41: Buying govie volatility vs. maturity matched swaptionvolatility was a profitable strategy in 2011. Implementing thestrategy in Bunds would have resulted in higher P&L and
information ratio than buying Bobl or Schatz volatility outright Average 1M returns* (bp of notional) and annualised information ratio (unitless)from buying govie volatility and selling a gamma-equivalent amount of maturitymatched swaption volatility, vs. buying volatility outright; all trades scaled tohave the same gamma as Bund; 1 January 2011– 18 November 2011;bp of notional Information ratios
0
3
6
9
12
15
18
Bund Bobl Schatz Bund Bobl Schatz
Long gov ie vs . mat matched
swpt
outright
0.0
0.5
1.0
1.5
2.0
2.5
3.0
* See Exhibit 30 for definition of returns.
Exhibit 42: Outperformance of govie vs. swaption volatility can beexplained by the implied volatility differential and peripheralspreads at trade initiation…Rolling 1M returns* from buying Bund volatility vs. selling a gamma equivalentamount of maturity matched swaptions regressed against 1M prior levels of 1)10Y weighted peripheral spreads, and 2 ) implied volatility differential; 1 January
2011– 18 November 2011; bp of notional
y = 0.23*(Peripheral sprd) - 24.4*(Imp vol s pread) - 57.2
R2 = 63%
-20
0
20
40
60
80
100
120
200 250 300 350 400 450 500 550 600 650
1M ago 10Y w td peripheral spreads; bp
* See Exhibit 30 for definition of returns.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 74/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Fabio Bassi (44-20) 7325-8615
Khagendra Gupta (44-20) 7777-1980J.P. Morgan Securities Ltd
74
In Exhibit 43, we present the zero iso-return line and ±1
standard deviation lines from buying Bund gamma vs.maturity-matched swaption gamma. Buying Bund
volatility vs. swaptions when the combination of implied
volatility differential and peripheral spreads is below the
zero iso-return line is generally profitable. The table
within Exhibit 43 shows statistics on trades initiated
following the signal from this model in 2011. In addition
to a higher average P&L (overall average is shown in
Exhibit 41), this strategy has a success ratio of 85%. Of
course, these statistics are based on in-sample data and so
need to be viewed accordingly. However, given that the
fundamental drivers are not likely to change in 2012, we
recommend buying govie vs. swaption gamma when
the combination of implied volatility differential and
peripheral spreads is below the -1 standard deviation
iso-return line.
Stay neutral on vega
Unlike gamma, buying vega would have been
profitable in 2011 (Exhibit 44). Outright long vega
positions would have resulted in higher returns with
smaller swings. For instance, buying 5Yx10Y swaption
straddles would have produced an average P&L of 21bp
of notional per month compared to flat returns for
3Mx10Y. This outperformance of vega is mainly due to
the increase in implied volatility and reduced impact of
lower delivered volatility. For instance, 5Yx10Y implied
volatility has increased from 5.0bp/day to 5.6bp/day in2011 while 3Mx10Y implied volatility has increased
from 5.7bp/day to 8bp/day. However, subsequent
delivered volatility on 10Y swaps has been relatively low
compared to implied volatility at inception. For
example, average 3Mx10Y implied volatility at inception
during 2011 is 5.8bp/day, while subsequent 1M delivered
volatility has been around 5.1bp/day. Thus, for short
dated options, vega gains (from increasing implieds)
have been largely offset by gamma losses.
Exhibit 43: …suggesting that investors should buy govie vs.swaption volatility when the combination of implied volatility
differential and peripheral spreads is 1SD below the zero iso-return lineImplied volatility spread* vs. 10Y weighted peripheral spread; solid line shows
the zero iso-return line and light lines are ±1SD** lines; table shows statistics of trades initiated in 2011***; 1 January 2011– 18 November 2011; bp/day
-1
0
1
2
3
4
5
200 300 400 500 600 700
Iso-return line
<0 returns
>0 returns
Current
1SD
line
10Y weighted peripheral spread; bp
18-Nov-11
Av erage 22
IR 3
% profitable 85%
Light (dark) dots represent trades that were profitable (unprofitable) in 2011 over the next month, without considering the trading rule discussed here.* Bund implied – maturity-matched swaption implied volatility; futures rolled 20Dbefore option expiry.** Based on regression beta and standard errors in Exhibit 42.*** 1M statistics of trades entered when the combination of implied volatilitydifferential and peripheral spreads was below the zero iso-return line.
Exhibit 44: Vega outperformed gamma in 2011 as low deliveredvolatility was a drag on the latter
Cumulative returns from buying €100mn notional EUR 3Mx10Y vs. 5Yx10Yswaption straddles; 1 January 2011 – 18 November 2011; bp of notional
-200
-100
0
100
200
300
Jan 11 Mar 11 May 11 Jul 11 Sep 11 Nov 11
5Yx10Y
3Mx10Y
Exhibit 45: Vega has exhibited a concave profile to gamma. Thus,despite our bullish view on gamma, we stay neutral on vega5Yx5Y EUR implied volatility regressed against 3Mx5Y EUR implied volatility;past 2Y; bp/day
y = -0.08x 2 + 1.20x + 0.99
R2 = 73%
4.0
4.5
5.0
5.5
6.0
3 4 5 6 7 8
18-Nov-11
3Mx 5Y implied vol; bp/day
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 75/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Fabio Bassi (44-20) 7325-8615
Khagendra Gupta (44-20) 7777-1980J.P. Morgan Securities Ltd
75
Additionally, vega has exhibited a concave profile to
its gamma counterparts (Exhibit 45). Thus, despite
our bullish view on gamma, we stay neutral on vega.
We present our targets for various vega points in Exhibit
36.
Sterling volatility
The story of Sterling volatility in 2011 has been eerily
similar to its EUR counterpart. Despite changing
inflation expectations and BoE policy reaction, GBP
volatility in the belly of the curve has been largely driven
by peripheral concerns. Thus, it is not surprising to see
that buying GBP swaption gamma in 2011 would have
also yielded mixed results, i.e., low monthly average
returns with large swings (Exhibit 46).
We do not expect this to change in 2012 and
recommend buying GBP swaption gamma in 10Y
tails. At the front end of the curve, we recommend
fading flare-ups by selling 2Y tail volatility and
enhancing returns in a range-bound environment by
hedging infrequently. Indeed, the argument for selling
volatility in GBP is even stronger despite the recent
widening of FRA/OIS at the front end. Our economists
forecast the BoE to engage in further QE purchases (see
United Kingdom) and the base rate to remain at 0.50%
throughout 2012, given the gloomy economic backdrop
and declining inflation expectations. A flat SONIA
curve will likely keep delivered volatility at the front-end
low (Exhibit 47). We present our target for various
points on the GBP volatility surface in Exhibit 48.
Given a choice between buying EUR and GBP gamma,
what should an investor do? Investors facing this
Exhibit 46: Similar to EUR volatility, buying GBP volatility in 2011would have generated mixed returns, since positive returns during
risk-off periods were offset by losses during risk-on periods Average 1M returns* and annualised information ratio achieved by buying GBP3Mx10Y swaption straddles;bp of notional ratios
-40
-30
-20
-10
0
10
20
30
40
2008 2009 2010 2011
-3
-2
-1
0
1
2
3
Av erage (left)
Information ratio (right)
* Trades entered daily and held for 1M. See Exhibit 30 for definition of returns.
Exhibit 47: We expect front-end volatility in GBP to declinemarginally as BoE remains on hold in 2012 and the policy
expectations curve remains flat…3Mx2Y GBP implied volatility regressed against 3M/15Mx3M GBP OIS curve;past 1Y; bp/day
y = 0.02x + 4.04
R2 = 76%
3
4
5
6
7
-20 0 20 40 60 80 100 120 140
18-Nov -11
3M/15Mx 3M OIS curve; bp
Exhibit 48: …even as volatility on 10Y tails increases in responseto an escalation in the peripheral debt crisisGBP implied volatility targets for 2012 vs. 18 November 2011; bp/day
18 Nov 11 1Q target 2Q target 2Q target - current
3Mx 2Y 4.8 4.5 4.0 -0.8
3Mx 5Y 5.9 6.0 6.1 0.2
3Mx 10Y 7.2 7.5 8.2 0.9
Exhibit 49: Given our view that the peripheral crisis will get worse,
we prefer buying EUR gamma over GBP gamma… Average* cumulative return** from buying 3Mx10Y EUR and GBP swaptionstraddles; bp of notional
-80
-60
-40
-20
0
20
40
60
0% 20% 40% 60% 80% 100%
EUR
GBP
EUR
GBP
Risk-off
Risk-on
% days into risk on/off period; %
* Averaged over risk-on/off episodes (see Exhibit 32 for risk-on/off dates).** See Exhibit 30 for definition of returns.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 76/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) 7777-3370
Fabio Bassi (44-20) 7325-8615
Khagendra Gupta (44-20) 7777-1980J.P. Morgan Securities Ltd
76
choice should consider buying EUR gamma over GBP
gamma. Although GBP volatility has been driven by
peripheral spreads, EUR gamma remains at the fore front
of the crisis and is more likely to outperform GBP.
Indeed, long gamma positions in EUR have, on
average, outperformed long GBP gamma positions
during various risk-off episodes in the past two years
(Exhibit 49). We also note that selling EUR gamma
during risk-on episodes is better than selling GBP
gamma. One reason for superior returns of EUR gamma
is that GBP implieds tend to decline with respect to EUR
implieds during periods of stress and vice versa (Exhibit
50). We use this relationship to arrive at our targets for
GBP gamma for 2012, listed in Exhibit 48.
Trading themes
• We are bullish on EONIA and high excess
liquidity; receive 6Mx6M EONIA on the back of
ECB policy rate cuts; implement bullish option
structures on Dec12 Euribor
− With the ECB expected to cut refi and deposit rates to0.50% and 0.25% respectively, we expect EONIA
fixings to fall to 30bp by mid-2012. We recommend
longs in 6Mx6M EONIA with a target of around 30bp,
and bullish option structures on Dec12 Euribor.
• We see value in bullish option structures on 5Y
swaps and recommend 1s/5s outright or
conditional swap curve flatteners
− With peripheral sovereign stress expected to escalate
in 2012, we expect the swap curve to flatten and
recommend carry efficient flatteners such as 1s/5s,which are a proxy for bullish positions but with better
risk profile, along with receiver structures on 6Mx5Y.
• 10s/30s is likely to remain driven by technical
flows; we have a flattening bias
− The 10s/30s swap curve is likely to remain driven bytechnical factors and is expected to flatten as
peripheral spreads widen. We present a framework toanalyse the attractiveness of long-dated forward
steepeners but refrain from recommending this trade
due to the prevalence of technical factors.
• We favour 2Y wideners and have a swap spread
curve flattening bias
− Swap spreads will likely widen beyond their Lehman peak in 1H12. We target 2Y and 10Y swap spreads at
145bp and 90bp, respectively, by mid-2012.
Conditional swap spread wideners offer better risk/reward than outright swap spread wideners.
• Position for further widening in spread markets,
such as OIS swap spreads and 3s/6s basis, that
have lagged the recent widening in swap spread
− Although spread markets are being largely driven by a
single-factor, i.e. sovereign risk, some spread markets
such as OIS swap spreads and 3s/6s basis are stilltrading far below their local/Lehman peaks,
suggesting that investors should express risk-off views
in these markets.
• Trade EUR gamma from the long side in the belly
of the curve, while fading spikes in volatility at the
front end; buy Bund volatility vs. swaption
volatility
− EUR volatility in the belly of the curve will increaseas the peripheral crisis escalates; we target 3Mx10Y at
9.4bp/day by 2Q12. Buy gamma in the belly of the
volatility curve and fade flare-ups in volatility at the
front end. Favour long positions in govie volatility vs.
swaption volatility.
− On a cross-market basis, buy EUR gamma vs. GBP
gamma.
Exhibit 50: …as GBP implieds will likely underperform EURimplieds during such episodes
3Mx10Y GBP-EUR implied volatility spread regressed against 3Mx10Y EURimplied spread; risk-off periods* from 1 January 2010 – 18 November 2011;bp/day
y = -0.44x + 2.57
R2 = 55%
-2
-1
0
1
2
3.4 4.0 4.6 5.2 5.8 6.4 7.0 7.6 8.2
EUR 3Mx 10Y implied vol; bp/day
18-Nov -11
* See Exhibit 32 for dates of risk-off episodes.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 77/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Francis DiamondAC
(44-20) 7325-3541
[email protected]. Morgan Securities Ltd
77
United Kingdom
• Gilt yields fell during 2011 on the back of
peripheral spread widening and reached all-time
lows across the curve. The BoE restarted QE gilt
purchases despite headline CPI breaking the 5%
level
• The macro picture for the UK is worrying, with
2012 growth expected to be just 0.5%. The BoE
will likely keep rates on hold and we expect QE
gilt purchases to increase to £425bn in 2012. As a
result we estimate the BoE will own just over
40% of the total secondary market of outstanding
gilts
• UK banks have relatively small exposures to
peripheral sovereign debt but they could be
vulnerable if financing markets freeze up given
their exposure to wholesale funding markets and
2012 refinancing needs
• We expect 2Y gilts to trade in a 50-60bp range
and recommend fading any richening move below
the 50bp level
• Wider peripheral spreads will drive 10Y and 30Y
gilt yields lower – we target 1.50% in 10Y gilts
and 2.25% in 30Y gilts by 2Q12. Should the UK
fall under the sovereign risk spotlight, we expect
ongoing QE purchases, a relatively low
proportion of non-domestic ownership and an
independent currency to limit any fiscal-driven
rise in gilt yields
• The 2s/10s curve will remain highly directional
with 10Y yields. We position for flattening and
expect 2s/10s to reach the 100bp level by mid-
2012
• We have a bias for a flatter 10s/30s curve as
directionality with 10Y gilts should weaken and
increasing QE purchases should drive the curve
flatter. We target the 75bp level by 2Q, 15bpflatter than currently implied by the forwards
• Position for wider 5Y and 10Y swap spreads in
1H12, driven by wider peripheral spreads, lower
gilt yields and a £150bn increase in QE gilt
purchases
2011 – The year of the long bond
2011 saw the European sovereign risk crisis move up
another notch. The year started off relatively benignly,
with the market pricing in rate-tightening expectations as
inflation moved higher and surprised to the upside. The
SONIA curve was pricing 100bp of tightening over a 1Yforward horizon in the first couple of months of the year.
10Y gilt yields briefly touched the 4% level in February
and traded in a broad range until the sovereign risk crisisreignited in the spring. From then on, gilt yields steadily
fell almost in lockstep with Bunds to make new all-time
historic lows, with 10Y real yields moving into negativeterritory.
Domestic data took a backseat as peripheral spreads
were the main driver of gilt yields and of the 2s/10s
Exhibit 1: The story of 2011 – it’s a one-factor worldCorrelation of UK par yields and curve with 10Y weighted peripheral spreads*, last 6M;%
74% 72%78% 79%
73%
62%
37%
66%
36%
45%
0%
20%
40%
60%
80%
100%
2 Y
5 Y
1 0 Y
3 0 Y
5 0 Y
2 s / 5 s
5 s / 1 0 s
2 s / 1 0 s
1 0 s / 3 0 s
3 0 s / 5 0 s
* 10Y weighted peripheral spread computed against Germany for Greece, Ireland,Portugal, Italy, and Spain (weighted by the size of their outstanding bond markets).
Exhibit 2: 2011 was the year of the long bondTotal return performance, 1 January – 18 November 2011, government bonds;%
0%
5%
10%
15%
20%
25%
30%
35%
40%
2Y 5Y 10Y 30Y
UK US Germany
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 78/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Francis DiamondAC
(44-20) 7325-3541
[email protected]. Morgan Securities Ltd
78
curve (Exhibit 1). The large fall in 30Y yields resulted
in long-end bonds outperforming on the curve in total
return terms, a theme that was common to both USTreasuries and Bunds (Exhibit 2). The BoE increasedQE by £75bn in October on the back of heightened
downside growth risks despite headline CPI inflation
reaching the highest level since mid-2008 (5.2% oya),
although the market impact was more limited than whenQE1 first started. The 2s/10s curve was highly directional
and flattened almost 70bp whilst 10s/30s steepened from
April to September only to flatten back following the QEextension in October.
Sterling funding markets reflected the various facets of
the sovereign risk crisis as Libor/OIS spreads reached
their widest levels since May 2009 whilst the flight-to-
quality demand for gilts and the increase in QE gilt purchases drove gilt GC below SONIA for short dates1.
2012 looks bleak
The macro outlook for the UK in 2012 is predominantly
a function of the global outlook and the ongoing
sovereign risk issues in Europe. Our economists expect
0.8% growth in the UK for this year and 0.5% oya in
2012, with the likelihood of two consecutive quarters
of mild contraction estimated at close to 50%. We believe that that the UK will avoid a deep recession
provided growth outside of Europe holds up.
Empirically, the UK PMI has demonstrated larger sensitivity to the global ex-Euro area PMI, but other
factors such as bank funding concerns and the ongoing
fiscal contraction present downside risks to UK growth.
There are worrying signs that UK output couldremain at anemic levels for several years. Corporates
have been running financial surpluses (i.e. have been net
lenders) for several years but since 2007, households
have shifted from being net borrowers to net lenders possibly indicating signs of a balance sheet recession
(Exhibit 3). As a result, the government stepped in with
fiscal supports in order to prevent the economy from
shrinking, with total government net borrowing moving
from 2.5% of GDP in 2007 to 10% of GDP in 2010.Fiscal contraction has prevented this pace of government
support from continuing, but the private sector is in debt-
reduction mode and appears unlikely to take up themantle to drive the economy forward for some time (we
estimate consumption will not begin recovering until at
least 2H12). The growth rate of loans to the householdsector has recently fallen to negative levels, and both
1 This has been the case in Euro cash markets for some time.
Exhibit 3: There are signs of a balance sheet recession in the UKFinancial balance by sector (negative is net borrower, positive is net lender); % GDP
-12%
-8%
-4%
0%
4%
8%
1990 1995 2000 2005 2010
Financial Corporates
Gov t Household
Exhibit 4: Private sector loan growth in the UK is in the doldrumsGrowth rate in MFI lending to private sector; %oya
-20%
-10%
0%
10%
20%
30%
40%
50%
60%
1998 2000 2002 2004 2006 2008 2010
Financial
Corporates
Household
Exhibit 5: Several more years of fiscal correction are needed to bringthe deficit downFiscal thrust* and Public Sector Net Borrowing (PSNB); % of GDP
-6
-4
-2
0
2
4
6
8
10
1972 1982 1992 2002 2012
Fis cal thrus t N et borrow ing
forecast
2% fiscal drag is max imum
likely achiev able in one year
* Fiscal thrust/drag is defined as year on year change in PSNB.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 79/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Francis DiamondAC
(44-20) 7325-3541
[email protected]. Morgan Securities Ltd
79
corporate and financial loan growth remain at very
depressed levels historically (Exhibit 4).
We expect the base rate to be kept on hold at 50bp
during 2012, with QE purchases increasing by
another £150bn by the end of the next year, taking
total QE purchases up to £425bn (around 30% of
GDP and 40% of total outstanding of gilts). Headlineinflation is forecast to fall to 1.9% oya by the end of
2012, but we expect this to have little impact on either
gilt yields or inflation breakevens (see Inflation Linked Markets), as markets will remain focused on the risks
around peripheral Europe and the growth outlook.
The UK budget deficit remains historically high, and we
expect this to come in at 5.6% of GDP for FY11/12, but
several more years of fiscal correction of around 1.5–2%of GDP per year are needed to reduce the deficit to more
appropriate levels (Exhibit 5). The Chancellor has statedthat the government’s self-imposed rules are to achieve a
balanced, cyclically adjusted deficit and have a falling
trajectory of debt/GDP by 2015/16. The recent downshiftin UK growth will likely mean that the 2012 Budget may
just show these rules being missed and that the
Chancellor will have to pencil in some additional modest
tightening in the next few years.
Summary themes
The evolution of peripheral spreads will likely be thekey driver of gilt yields for 2012. Macro, fiscal and
technical factors will all be important, but will likely
have a larger impact on the evolution of the curve andswap spreads than the outright level of yields. We
describe our trading themes and gilt market views indetail in the following pages, but in summary we think:
1) 2Y gilt yields will remain anchored and will likely
trade within a tight range. We believe the BoE isunlikely to cut base rates below the current level and
think 2Y gilt yields are floored at close to 50bp.
Exhibit 6: We expect 10Y and 30Y yields to fall substantially in 1H12, with the 2s/10s curve flatteningCurrent and J.P. Morgan gilt forecasts for 2012; % unless stated
18-Nov -11 1Q12 2Q12 3Q12 4Q12Q2 vs. fwd
(bp)
Q4 vs. fwd
(bp)
BoE base 0.50 0.50 0.50 0.50 0.50 n/a n/a
1M SONIA 0.52 0.50 0.50 0.50 0.50 4 0
2Y 0.48 0.50 0.50 0.55 0.65 4 -1
5Y 1.13 1.00 1.00 1.05 1.15 -22 -17
10Y 2.26 1.80 1.50 1.75 1.95 -89 -61
30Y 3.19 2.70 2.25 2.40 2.60 -103 -75
2s/10s (bp) 178 130 100 120 130 -93 -60
10s/30s (bp) 93 90 75 65 65 -14 -14
Exhibit 7: Gross gilt sales likely to rise to close to £200bn in FY12/13,but net sales will be lowGross gilt issuance and net gilt issuance adjusted for redemptions and QE*, fiscal year
basis; £bn
0
50
100
150
200
250
0 4 / 0 5
0 5 / 0 6
0 6 / 0 7
0 7 / 0 8
0 8 / 0 9
0 9 / 1 0
1 0 / 1 1
1 1 / 1 2
1 2 / 1 3 ( J P M )
Net of QE and redemptions Gross gilt sales
* We expect £150bn of additional QE taking total gilt purchases to £425bn by the endof 2012.
Exhibit 8: Gilt issuance in the belly of the curve is likely to increase inFY12/13Gilt issuance by sector as proportion of total issuance, FY11/12 and J.P.Morganforecast for FY 12/13; %
0%
5%
10%
15%20%
25%
30%
35%
40%
Short nom inal M edium
nominal
Long nom inal* Linkers *
FY 11/ 12 F Y 12/13
* Includes syndicated issuance.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 80/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Francis DiamondAC
(44-20) 7325-3541
[email protected]. Morgan Securities Ltd
80
2) Wider peripheral spreads will likely drive 10Y and
30Y gilt yields lower – we target 1.50% in 10Y gilts
and 2.25% in 30Y gilts by 2Q12 (Exhibit 6). We seelimited risk from sovereign concerns in the UK. Indeed,if the UK were to fall under the sovereign risk spotlight,
we expect ongoing QE purchases, a relatively low
proportion of non-domestic ownership and an
independent currency to limit any fiscal-driven rise ingilt yields.
3) The 2s/10s curve will remain highly directional
with 10Y yields. Position for flattening as we expect
2s/10s to reach the 100bp level by mid-2012.
4) Bias for a flatter 10s/30s curve. We expect
directionality with 10Y gilts to weaken as yields fall
further. We think 30Y gilts will modestly outperform asQE gilt purchases increase to the £425bn level.
5) Position for wider swap spreads in the 5Y and 10Y
sectors.
Gilt supply, QE purchases and bank
funding
We expect gross gilt issuance of around £190bn in
FY12/13, some £25bn above planned gilt sales for the
current fiscal year. Net borrowing for FY12/13 willlikely be some £10bn higher than forecast by the OBR
(£122bn) given a small degree of fiscal slippage, and weexpect any increase in the current year net borrowingrequirement to be reflected in higher borrowing needs for
FY12/13. Gilt redemptions are also higher next fiscal
year at just under £50bn.
Issuance net of redemptions and QE purchases is
expected to be just below £25bn in the current fiscal
year, which will then rise to just under £60bn in FY12/13
(Exhibit 7). In terms of maturity split, we expect a £17bnincrease in issuance in the 5-15Y sector, to partially
counter the expected increase in QE gilt purchases
(Exhibit 8). We expect long nominal gilt issuance to
increase around £5bn, but remain roughly unchanged as a proportion of total gilt issuance. On a calendar year basis
we expect £185bn of gilt supply in 2012 to be fully offset
by £43bn of redemptions and £150bn of QE gilt purchases.
We think that the demand impact through QE gilt
purchases could be a significant driver of gilt yields.
Our central view is for the BoE to announce QE
increases of £75bn in both February 20122 and May2012, taking total QE asset purchases to £425bn. We
expect these purchases to be entirely in gilts as no other sterling fixed income markets are large enough to
accommodate the scale of QE purchases, although we
don’t rule out that the BoE could extend purchases into
bank debt on a much smaller scale. The BoE currently
2 Note: The £75bn of purchases which began in October 2011 will be
completed in early February 2012, taking total QE purchases to£275bn.
Exhibit 9: Total QE BoE gilt holdings are expected to peak at around40% of the secondary market next year, but the BoE could own almost60% of the 10-25Y sector
Evolution of BoE holdings as % of gilt secondary market outstanding* based onadditional QE** and gilt issuance profile for 2012; %
10%
20%
30%
40%
50%
60%
Nov 11 Mar 12 Jun12 Dec12
3-10Y 10-25Y 25Y+ Total
* Excludes DMO collateral holdings.** We assume a further £150bn of QE gilt purchases taking total QE gilt holdings to£425bn.
Exhibit 10: If BoE QE gilt holdings were to reach £450-500bn, then theliquidity of the gilt market could be severely impacted as the BoE wouldown large segments of the gilt marketEstimates for distribution of BoE QE gilt ownership by sector by end 2012incorporating our issuance forecast; £bn
37%40%
52%59%
66%
60%
45%50%
33%
43%
52%
40%
0%
10%
20%
30%
40%
50%
60%70%
80%
400 450 500
Hypothetical total BoE QE gilt holdings; £bn
3-10Y 10-25Y 25Y+ Total
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 81/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Francis DiamondAC
(44-20) 7325-3541
[email protected]. Morgan Securities Ltd
81
owns around 25% of the gilt market and around 38% of
the secondary market outstanding of bonds in the 10-25Y
sector.
Given our QE forecast and supply estimates for the
next fiscal year, we estimate that the BoE could own
just over 40% of the entire gilt market and close to
60% of the 10-25Y sector (Exhibit 9). In our view, thisis close to the limit at which QE gilt purchases will affect
gilt market liquidity, and we think that the BoE will be
unable to increase QE gilt purchases above the £450-500bn level as holdings in some sectors of the curve
would become prohibitively large (Exhibit 10).
Concerns around the impact of the sovereign risk crisis
on banks’ increased in the past month. UK banks have
relatively small exposures to peripheral governmentbonds (Exhibit 11), but they could be vulnerable if
financing markets freeze up given their exposure to
wholesale funding markets and 2012 refinancing
needs. UK banks have around €100bn of refinancing
needs in 2012, which is large in absolute terms comparedwith other European countries, but is relatively small in
terms of total banking sector assets (Exhibit 12).
Nonetheless, should conditions in financing markets
deteriorate if sovereign risk stresses increase, then this
could well pose some problems. We would expect the
BoE to respond by increasing the amount of fundingavailable via the 3M and 6M LTROs, which are currently
small in size at £5bn and £2.5bn, respectively, as wasdone in 2009.
The BoE is unlikely to reopen the SLS3, which will close
in 2012 as other mechanisms such as the discountwindow facility are now in place. UK banks have raised
capital and reduced leverage, compared with European
banks, but capital remains below the 10% Basel III
requirement which needs to be implemented by 2019.
Our equity analysts expect them to meet this targetorganically, but in the event of a severe recession, they
are uncertain as to whether regulators would intervene
and force capital-raising at some point over the next few
years.
Front end duration view
2011 was almost a re-run of 2010 in SONIA space, asrate-tightening expectations priced in during the first few
months of the year were steadily unwound from April
onwards, although the magnitude of the re-pricing was
3 Special Liquidity Scheme (SLS) was launched in April 2008 andallowed institutions to lend a wide range asset-backed bonds to the BoE
in exchange for 9M T-bills.
Exhibit 11: UK bank exposures to peripheral government bonds arerelatively small compared with total bank reservesUK bank exposures to Euro zone sovereigns, public sector claims*; $bn
$bn %
% bank
reserves
Austria 2 1% 0%
Belgium 6 3% 1%
Finland 4 2% 1%
France 56 31% 10%
Germany 62 34% 11%
Greece 4 2% 1%
Ireland 4 2% 1%
Italy 17 10% 3%
N etherlands 20 11% 3%
Portugal 2 1% 0%
Spain 8 4% 1%Peripheral 34 19% 6%
Ita + Spa 25 14% 4%
Public sector
* Public sector claims as proportion of all Euro zone public sector claims. Derived fromBIS data 2Q11.
Exhibit 12: UK bank refunding needs for 2012 are large in absolute termsbut are comparatively small as a proportion of banking sector assetsTotal securitised, secured and unsecured bank refinancing requirements for 2012 €bn proportion of total banking sector assets; %
0
20
40
60
80
100
120
140
A u s t r i a
B e l g i u m
F i n l a n d
F r a n c e
G e r m a n y
G r e e c e
I r e l a n d
I t a l y
N e t h e r l a n d s
P o r t u g a l
S p a i n
U K
0.0
1.0
2.0
3.0
4.0 Amount
Proportion of bank assets
Exhibit 13: We don’t expect the spread between O/N SONIA and the baserate to fall to the levels seen during QE1, as access to the BoE reservefacility was expanded in late 2009O/N SONIA – base rate spread and periods of QE purchases; bp
-15
-10
-5
0
5
10
Mar 09 Sep 09 Mar 10 Sep 10 Mar 11 Sep 11
QE1
QE2
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 82/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Francis DiamondAC
(44-20) 7325-3541
[email protected]. Morgan Securities Ltd
82
larger than that seen in 2010. We expect the BoE to
keep base rates on hold at 50bp, with additional
monetary easing coming from further QE rather thana lower base rate. The BoE views 50bp as an effectivefloor and has previously stated that its preferred tool for
further easing monetary conditions is more QE. To this
extent, we think that SONIA is effectively floored at
50bp, and we do not expect it to trade much below thislevel. Following on from the start of QE in March 2009,
O/N SONIA fell around 15bp below the base rate, but
this was due to fragmentation in the SONIA marketgiven the limited number of participants who had access
to the BoE reserve facilities. In the autumn of 2009, the
BoE lowered the requirements for institutions to become
reserve scheme members and access to the BoE reserve
accounts was expanded; the move saw the O/N SONIA
vs. base rate spread gradually increase (Exhibit 13).Given that access to the reserve facilities remains
unchanged, we don’t think that any additional QE in
itself should push O/N SONIA below the level of base
rates.
Given this, we expect SONIA rates up to 12M
forward to hover around 50bp, and we would look to
fade any cheapening on a forward basis as we think
the probability of rate hikes over the next 12 months is
very small. Libor rates have widened to reflect increased
funding pressure concerns stemming from the sovereignrisk crisis, and we think that Libor/OIS spreads can
widen further in the UK as peripheral spreads widen as per our forecast.
Historically, the level of 2Y par yields can be well
explained by the slope of the SONIA curve (we use18Mx1M – 6Mx1M), and this has been the case over
most of the 2011 (Exhibit 14). Given our view that base
rates will be on hold for 2012 as QE is increased further,
we expect the SONIA curve to remain at very flat levels,
with a limited amount of term premia priced in. The6Mx1M/18Mx1M curve has traded in a range around the
10bp level over the past few months, and we expect this
dynamic to continue in 1Q12 as we see little chance of
the curve pricing in any BoE rate tightening expectations.We think the forward SONIA curve should not trade
below 0bp (as the BoE will not lower base rates further in our view) which effectively means 2Y gilt yields
should not fall much below the 50bp level. We expect
2Y gilt yields to range-trade around the 50-60bp level
for much of 2012.
Be long 10Y and 30Y duration….
In the belly of the curve, 5Y and 10Y gilt yields made
all-time lows in 2011, and are far below the levels that
Exhibit 14: 2Y gilts are driven by the slope of the SONIA curve andshould not trade much below 50bp in our view given that the base rateis floored at 50bp
2Y par gilt yield regressed against level of SONIA curve, November 2010 – November 2011; %
y = 0.0004x 2 + 0.005x + 0.58
R2 =98%
0.40
0.60
0.80
1.00
1.20
1.40
1.60
-20 0 20 40 60 80 100 120SONIA curve*; bp
current
* 6Mx1M/18Mx1M SONIA curve.
Exhibit 15: Peripheral spreads are more significant in explaining 10Ygilt yields than macro factors…10Y par gilt yield regressed against 10Y peripheral spread*, UK composite PMI, CPIand QE dummy; monthly data March 2007–October 2011 Factor Coefficient T-Stat
Peripheral Spread, bp -0.01 -12.7
Composite PMI 0.03 3.7
CPI, %oy a 0.24 3.7
QE dummy -0.36 -3.0
R-squared 83%
Std. error, bp 32
* 10Y weighted peripheral spread computed against Germany for Greece, Ireland,Portugal, Italy, and Spain (weighted by the size of their outstanding bond markets).
Exhibit 16: …and the model suggests 10Y yields are some 60bp abovefair value
Actual and model predicted 10Y par yields, monthly data March 2007–November 2011; %
1.00
2.00
3.00
4.00
5.00
6.00
Mar 07 Mar 08 Mar 09 Mar 10 Mar 11
Actual Predicted
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 83/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Francis DiamondAC
(44-20) 7325-3541
[email protected]. Morgan Securities Ltd
83
macro-economic models would suggest. However, the
low level of yields is less a function of the UK macro
landscape and more a function of the sovereign risk crisisin Europe. We present a simple regression model tohighlight this point (Exhibit 15) and whilst the level of
the composite PMI and level of headline CPI are
important drivers, their statistical significance is much
lower than 10Y weighted peripheral sovereign spreads.We also include a QE dummy to control for periods of
BoE gilt-buying. This model suggests that 10Y yields
are some 60bp above fair value level (Exhibit 16). We
expect 10Y yields to gradually fall to the 1.50% level
over the first half of 2012 as peripheral spreads widen
substantially, with the weighted peripheral spread
expected to reach the 955bp level, from the current
700bp (see Euro Cash for our forecasts). A tightening
back in peripheral spreads should result in a rise in 10Ygilt yields in 2H12.
We note that over recent weeks gilt yields have not fallen
as much as the model suggests given the move in
peripheral spreads, but we think this is reflection of alack of investor appetite into year end to enter fresh
duration positions. We expect the usual linear
relationship between gilt yields and spreads to resume in
the New Year. We think that 5Y yields will become
sticky in a rally and will fall less than 10Y yields. We
expect 30Y yields to outperform, driven by the
ongoing QE gilt purchases and weakening
directionality (See curve section below).
Gilts have acted as a safe haven in recent months as
European sovereign risks have increased, and we expect
this to continue into 2012. In the event that the situationin Europe escalates far enough so that Germany looses its
safe-haven status then Bund yields could rise. In that
environment, we think gilt yields will also rise but may
well outperform bunds due to flight-to-quality flows.
…as we don’t expect fiscal pressures to
drive gilt yields higher
In presenting our duration view, we are assuming that thecurrent sovereign risks do not envelop the UK in the
coming months. Although the UK is one of the worst
AAA countries in terms of its debt dynamics and still hassome way to travel on its fiscal journey, our base case
scenario is that UK gilt yields do not rise materially in
response to fiscal concerns. Through the large-scale
purchases of gilts the BoE appears to be sacrificing long-
term inflation certainty for near-term fiscal stability. QEgilt purchases are expected to rise to £425bn by the end
of 2012, which should limit the potential for substantial
fiscally-driven increases in gilt yields. If fiscal stresses
were to become severe enough then the currency could
also act as a pressure valve (Exhibit 17), limiting the
magnitude of any rise in gilt yields (note our central viewis for relatively stable trade weighted sterling in 2012)
4.
The impact of weaker growth on the UK’s debt/GDP
dynamic is important. Exhibit 18 shows our estimates of the evolution of this metric under several different
economic scenarios. In our central case of 0.5% GDP in
2012 and a slow recovery to the 2% level over thesubsequent few years, debt/GDP (excluding the impact
of financial interventions) will peak just under 77% on a
4 See Global FX Strategy 2012
Exhibit 17: The currency can provide an escape valve when UK fiscalpressures buildTrade-weighted sterling index regressed against PSNB*, yearly 1976–2011
y = -2.7x + 106.3
R2 = 41%
70
80
90
100
110
120
-2.0 0.0 2.0 4.0 6.0 8.0 10.0
PSNB, % GDP
2011 est
* Public Sector Net Borrowing.
Exhibit 18: We expect debt/GDP to reach the 75%–85% level in fiveyears’ timeJ.P. Morgan’s debt/GDP forecasts under various economic scenarios* on a fiscal year basis, dashed line is end of government forecast period; %
60
70
80
90
100
110
2011/12 2012/13 2013/14 2014/15 2015/16 2016/17
Baseline
Severe Recession
Mild Recession
95% lev el
* Baseline is our central GDP forecast; mild recession assumes mildly negative GDP in2012 followed by a modest recovery back to the 2% level; severe recession assumesa repeat of 2008/09, with 2012 GDP at -1.5% and 2013 GDP at -3.5%.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 84/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Francis DiamondAC
(44-20) 7325-3541
[email protected]. Morgan Securities Ltd
84
five-year horizon, some 10% higher than the estimate for
FY11/12. In a mild recession scenario where output is
mildly negative for 2012 this peak is predicted to bearound 85% with a steeper trajectory. It would take a
recession of the severity of 2008/09 to push debt/GDP
to close to 100% in five years time, something we
think is not likely at the moment.
If either the baseline or mild recession scenarios play
out then gilt yields will not rise materially due to
fiscal concerns as long as the government maintainsits fiscal commitments and credibility. However, if the
severe recession scenario plays out then we expect rating
agencies to act, not only because of the weaker growth
outlook but also because of an expected increase in
contingent liabilities as the UK government may have to
re-introduce its guarantee programmes, and financialmarkets will likely price an increase in UK sovereign
risk. In this environment safe-haven purchases of giltswould abate and non-domestic holdings of gilts would
probably fall.
In addition, as non-domestic holdings are just less than
one third of the total gilt market which should limit an
impact from foreign investors taking flight from the UK
gilt market. This compares favourably with the main
Euro zone countries where non-domestics own between
40 and 65% of the total bond markets and domesticholdings in the UK are not highly concentrated in the
banking sector (Exhibit 19). It is worth noting thatforeign central bank ownership has been very stable ataround 6-8% of the market and that the increase in non-
residents holdings has come from other investors. In fact,
the increase in other foreign investors (ex central banks)holdings of gilts has been well correlated with gilt
supply, suggesting that non-domestic investor
participation has increased as the size of the gilt market
has expanded (Exhibit 20).
In our view, market pressure is only likely to come to
bear if the government’s fiscal plans lose credibility.
So far the coalition has shown no signs in wavering on its
fiscal tightening commitment but the Chancellor’s self-imposed rules are close to being breached. If output
weakened to the point where fiscal contraction needed to be slowed then we think that the Government can back
load tightening further into the current parliament term.
However, we think that the maximum tightening
achievable in any one year would be 2% of GDP.
Attempting to do any more than this amount would risk
straining the government’s fiscal credibility in our view.
Our base case view is that the UK retains its AAA
rating and stable outlook but rating agency action
Exhibit 19: Non-domestic investors’ holdings of gilts are lower thannon-domestic holdings in the main European bond marketsForeign and domestic bank* holdings of government bonds as % of total marketoutstanding; %
66
81
4538
30
46
611 10 13
33
12
4
44
0
10
20
30
40
50
60
70
80
90
F r a n c e
G
e r m a n y
I t a l y
S p a i n
U K
U S
J a p a n
Non-domestic Domestic bank*
* Excludes Central Bank holdings.
Exhibit 20: Non-domestic investors’ gilt holdings have increased as giltissuance and market liquidity have increasedForeign investor ex-Central Banks gilt holdings regressed against gross gilt issuance,annual data 1976–2011; %
y = -0.0001x 2 + 0.002x + 0.10
R2 = 74%
10%
15%
20%
25%
30%
0 50 100 150 200 250
Gross gilt issuance, £bn
Exhibit 21: Our central view is that the UK’s sovereign rating will not bedowngraded, but should rating agencies put the UK on negativeoutlook, we expect only a modest 20bp rise in 10Y gilt yields
Adjusted 10Y gilt yield* around the time when S&P placed the UK on negative outlookin May 2009; bp
-25
-20
-15
-10
-5
0
5
1015
20
Mar 09 Apr 09 May 09 Jun 09 Jul 09
S&P places UK on negative outlook
* 10Y par gilt = 0.7* 10Y par Bund yield + 0.76 * 12Mx3M (SONIA – EONIA) + 1.2. R-
squared: 82%, std. error: 9bp.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 85/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Francis DiamondAC
(44-20) 7325-3541
[email protected]. Morgan Securities Ltd
85
cannot be ruled out if output weakens substantially.
However, if the UK were to be put on review for a
potential downgrade then we think the gilt marketreaction will be muted. When the UK was placed onnegative outlook by S&P on 21 May 2009 the adjusted
10Y gilt yield rose by less than 20bp (Exhibit 21) and
the impact dissipated quite quickly. Downgrades of AAA
sovereigns have typically had a limited impact on yieldshistorically with the Japanese downgrades in November
1998 and February 2001 resulting in less than a 15bp rise
in yields in the week following the move. The USdowngrade earlier this year had a similarly limited
reaction on the day of the announcement with Treasury
yields falling some 30bp in the following week on flight
to quality flows.
Curve view – position for 2s/10s
flattening
The 2s/5s curve is flatter since the start of the year but
has been trading in a 30bp range since mid-August and
has shown little correlation with European peripheral
spreads over this period. We can model the par 2s/5scurve as a function of three variables: 1) the forward
SONIA curve (24Mx1M – 12Mx1M), to capture
changing base rate expectations; 2) 5Y RPI to capture
changing inflation expectations; and 3) a QE dummy
variable to take into account QE gilt purchases (whichhave been in the 3Y+ sector of the gilt curve). The model
fit has been good over the past couple of years (Exhibit22), and currently the 2s/5s curve is fairly valued.
Going forward, we expect this range-trading dynamic
to continue. Under the assumption that QE gilt purchases continue for the next 6 months, we outline
how the curve could change under different assumptions
for the slope of the SONIA curve and the level of 5Y RPI
swaps (Exhibit 23). In our view, the SONIA curve is
unlikely to trade below zero (as we think the chances of base rate cuts are low), and we don’t expect this curve to
trade much above 35bp, given our low-for-long view.
We think 5Y RPI swaps could modestly fall to the 2.50%
level at worst, and with the SONIA curve likely toremain close to current levels, we expect very modest
flattening of the 2s/5s curve. Any steepening is likely to
be limited, and we think the 2s/5 curve can trade in a
50-70bp range over 1H12.
The 2s/10s curve has been highly directional with 10Y
yields when the 10Y rate is below 3.20% (Exhibit 24).
We expect this to remain the case over 2012 as 2Y yieldsare anchored, and we find it hard to envisage a scenario
in which 10Y gilt yields will rise above this 3.20% level
in the next couple of quarters. We don’t expect the
Exhibit 22: The 2s/5s curve can be well explained by the slope of theSONIA curve, the level of 5Y RPI, and QE purchases
Actual 2s/5s par gilt curve vs. model curve*; March 2009–November 2011; %
0.60
0.80
1.00
1.20
1.40
1.60
1.80
Mar 09 Sep 09 Mar 10 Sep 10 Mar 11 Sep 11
Actual Model
* 2s/5s par = 0.25 * 5Y RPI inflation swap + 0.49*OIS curve – 0.18 * QE dummy +0.028. R-squared: 79%, std. error: 10bp.
Exhibit 23: We think the 2s/5s curve cannot flatten much more, and anysteepening will struggle to beat the forwards – we recommend playingthe rangeProjected changes* in 2s/5s curve over the next 6 months, given various levels of 5YRPI swaps and OIS curve**; shaded indicates our expected outcomes; bp
OIS curve**; bp 2.50 2.75 3.00 3.25 3.50
75 15 25 30 35 40
50 5 10 15 25 30
25 -10 0 5 10 150 -20 -15 -10 0 5
-25 -30 -25 -20 -15 -10
5Y RPI swap, %
* Calculated using regression equation in Exhibit 22. Assumes QE buying is in placeover the next 6 months. Current OIS curve = 17bp, 5Y RPI swap = 3.04%.**OIS curve = SONIA OIS 24Mx1m – SONIA OIS 12Mx1M.
Exhibit 24: With the front end anchored, the 2s/10s curve is likely toremain highly directional with 10Y yields during 20122s/10s par curve regressed against 10Y par yields, last 12M, %
y = 0.77x - 0.03
R2 =93%
y = -0.04x + 2.6
R2 = 0%
1.60
1.80
2.00
2.20
2.40
2.60
2.80
2.00 2.20 2.40 2.60 2.80 3.00 3.20 3.40 3.60 3.80 4. 00 4. 20
10Y par rate, %
10Y par > 3.2%10Y par < 3.2%
current
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 86/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Francis DiamondAC
(44-20) 7325-3541
[email protected]. Morgan Securities Ltd
86
increase in QE to generate an increase in forward
inflation expectations and we think it is too early for this
dynamic to be reflected via a steeper curve. Inflationexpectations are only likely to become unanchored to theupside when excess demand has recovered and the
excess central bank liquidity is being reflected in strong
credit and loan growth. This scenario is several years
away yet in our view. We forecast the 2s/10s curve to
flatten to the 100bp level by the end of 2Q,
substantially more than priced into the forwards.
The 2s/5s/10s fly is typically viewed as a directional fly,
and this has been the case for most of the past 12 months,
although we note that the directionality has weakened
substantially when 5Y par yields have been below 1.5%
(Exhibit 25). This convexity phenomenon is likely due
to investors attributing some degree of risk premia tolonger maturity rates when the base rate is anchored (i.e.
some upward slope) and has been observed in Japan (see Euro Cash) in the past when yields get to very low
levels. We think this convexity phenomenon will remain
in place as 5Y and 10Y yields fall further.
The 10s/30s curve has also been highly directional with
10Y yields (Exhibit 26), although the curve has
persistently traded too flat vs. the level of yields over the
past couple of months. We think this is a reflection of the
resumption of QE gilt purchases in early October as the10s/30s curve, adjusted for the level of 10Y yields, has
been broadly flattening as the BoE has been buyingrelatively more 25Y+ gilts compared with 10-25Y gilts,taking into account DMO supply (Exhibit 27). Any
implied steepening from lower 10Y gilt yields will likely
be gradually offset by the BoE purchasing relativelymore 25Y gilts compared with 10-25Y gilts (in 10Y
equivalent terms).
We note that since QE restarted in October that the BoE
has had to pay relatively more compared with pre-auction levels for 25Y+ gilts and that cover ratios for
buybacks in this sector have been much lower than in the
other two sectors (Exhibit 28). We expect this dynamic
to continue as QE purchases increase. The directionalityof the 10s/30s curve with the level of yields is likely to
become weaker as 10Y yields fall further as investors rollalong the curve in search of higher yields. Overall, these
factors should result in modest flattening in the
10s/30s curve and we target the 75bp level by 2Q,
some 15bp flatter than currently implied by the
forwards.
Exhibit 25: 2s/5s/10s fly directionality has weakened as yields havefallen and is likely to remain low2s/5s/10s par yield fly (50:50) regressed against 5Y par yield, last 24M; %
y = 0.16x - 0.4
R2 =6%
y = 0.26x - 0.61
R2 = 90%
-0.40
-0.30
-0.20
-0.10
0.00
0.10
0.20
0.30
0.50 1.00 1.50 2.00 2.50 3.00 3.50
5Y par yield; %
5Y par < 1.5%
5Y par > 1.5%
Exhibit 26: The 10s/30s gilt curve has been driven by 10Y yields…10s/30s par gilt curve regressed against 10Y par yields, January 2011–November 2011; bp
y = -35.2x + 196.6
R2 = 80%
40
60
80
100
120
140
2.00 2.50 3.00 3.50 4.00 4.50
10Y par yield; %
current
Exhibit 27: …and the excess flattening since October can be explainedby BoE gilt purchases in the 25Y+ sector 10s/30s curve adjusted for level of 10Y gilt yields* vs. cumulative excess of QE giltpurchases in the 25Y+ sector compared with the 10-25Y sector adjusted for DMOsupply **; weekly data 10 October 2011–18 November 2011bp £bn, 10Y equivalents
-24
-22
-20
-18
-16
-14
-12
-10
10 Oct 17 Oct 24 Oct 31 Oct 07 Nov 14 Nov
-4
-2
0
2
4
6
8
10
Cum. ex cess of 25Y+ purchases (£bn**)
Adjusted 10s/30s
* Residual from linear regression of 10s/30s curve vs. 10Y yields over the past 12months, weekly average.** Expressed in £bn of 10Y equivalents, cumulative purchases from 6 October 2011when QE restarted minus DMO supply. Scale inverted.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 87/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Francis DiamondAC
(44-20) 7325-3541
[email protected]. Morgan Securities Ltd
87
Market technicals – curve-trading rule
and trading short-dated gilts
The 5s/10s curve has generally been non-directional with
the level of yields for most of the past year, although thisdirectionality has increased in the past couple of months.
We think this sector of the curve is the most likely to be
affected by momentum dynamics as regular DMOsupply, QE purchases and reasonable liquidity can result
in opportunities for investors wishing to express short-
term tactical views.
To this extent, we have developed a momentum-trading
rule model for the 5s/10s par gilt curve, based on a
moving average indicator 5. The intuition is that when the
signal is large and positive, then the curve should flattenin subsequent days, and when the signal is large and
negative, the curve should steepen in subsequent days.
We test a trading rule based on different signal thresholds
for initiating and exiting trades since the start of 2010
and since the start of 2011. A rule that initiates curve
trades when the signal threshold is +/- 2.5 and exitstrades when the threshold is +/- 0.25 has performed the
best over the two periods, generating a total P/L since
2010 and 18bp P/L since 2011 (Exhibit 29), with a
success ratio around 75% in both periods. The
performance of this rule has been lower in 2011compared with 2010 as general risk appetite has fallen as
a result of the worsening sovereign risk crisis. In anenvironment in which both regular supply and BoE
buying continue to take place in the 5Y and 10Y
sectors, we expect the 5s/10s curve momentum model
to continue delivering positive results.
We also look at the behaviour of very short-end gilts
from a technical perspective. As bonds fall below 1Y
remaining maturity, they become less liquid and cease to
become part of most investors’ portfolios. We wouldexpect gilts to start to cheapen vs. surrounding lines at
some point before they fall below the 1Y point. Looking
at examples over the past couple of years and using a
yield spread adjusted for the level of 2Y gilts we find thatthis starts to occur, on average, 50 days before the bonds
become sub 1Y in maturity (Exhibit 30). Gilt 4.5%
Mar13 falls below the 1Y remaining maturity point
5 We first generate an MACD (Moving Average Convergence
Divergence) measure. This is calculated as the expected moving
average (EMA) of the 5s/10s curve over a 5-day period minus the EMA
of the 5s/10s curve over a 25-day period. We then generate a signal
measure which is the 10-day EMA of the derived MACD measure. Thetrading signal is defined as the MACD measure minus the signal
measure.
Exhibit 28: The BoE has had to pay more vs. market levels to buy 25Y+bonds – we think this will continue
Average coverage ratio and average yield of bonds purchased vs. market levels for
QE gilt buybacks conducted from 6 October 2011 onwardsRatio bp
1.5
2.0
2.5
3.0
3.5
3-10Y 10-25Y 25Y+
0.0
0.1
0.2
0.3
0.4 Avg. cover ratio
Yield vs. market
Exhibit 29: A momentum trading rule for the 5s/10s gilt curve hasgenerated positive P/L since 2010Total P/L and number of signals generated by 5s/10s curve momentum model *trading rule ** since 2010 and 2011; bp
18
46
4
11
0
10
20
30
40
50
Since 2011 Since 2010
P/L Total no. s ignals
* We first generate a MACD (Moving Average Convergence Divergence) measure.This is calculated as the expected moving average (EMA) of 10Y yields over a 5-dayperiod minus the EMA of 10Y y ields over a 25-day period. We then generate a signalmeasure which is the 10-day EMA of the MACD measure. The trading signal is definedas the MACD measure minus the signal measure.** The trading rule is calibrated based upon a threshold for the signal at which toinstigate and unwind trades. We use a signal of + (-) 2.5 as the threshold to enter
trades and a signal of + (-) 0.25 as the threshold to close open trades.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 88/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Francis DiamondAC
(44-20) 7325-3541
[email protected]. Morgan Securities Ltd
88
towards the end of 1Q12 and we recommend
underweighting this line vs. gilts 8% Sep14 and 2.25%
Mar14.
Position for wider 5Y and 10Y swap
spreads
Sterling swap spreads have widened over 2011, with the bulk of this occurring in the past few months of the year.
10Y swap spreads were relatively stable during 1H11,
trading in an 8bp range from January to July, beforewidening out during 2H11 to reach levels not seen since
mid-2009 (Exhibit 31). The move wider occurred across
the swap spread curve, with 2Y swap spreads widening
the most (53bp up to 18 November 2011), driven by
increased European sovereign risk fears and renewed bank funding concerns. The resumption of QE in October
resulted in 30Y swap spreads moving out of the trading
range that had been in place for most of 2011.
Swap spreads in the 2Y sector have broadly mirrored
the moves in European peripheral spreads, as Libor/OIS
spreads have trebled with the increase in bank fundingconcerns, and we expect this dynamic to continue in
2012.
Further out, 5Y and 10Y swap spreads have generally
been directional with the level of yields, although thisrelationship has started to become non-linear as yields
have reached extremely low levels. The resumption of BoE gilt purchases has also affected swap spreads. Over the past three years, 10Y swap spreads can be reasonably
well explained by the level of 10Y yields, monthly QE
purchases in the 10Y sector net of DMO supply, andimplied volatility (Exhibit 32).
Given our views of lower 10Y gilt yields and an increase
in QE to £425bn, we expect 10Y swap spreads to widen
to around the 55bp level in 1H12 (Exhibit 33) beforeslowly tightening back in 2H12 as yields rise modestly.
Our expected increases in net gilt supply for FY12/13 is
relatively small, and we think that the market is already
expecting a slight worsening in public finances.However, the scope for the government to increase
borrowing is limited, and we expect any tightening
impact on 10Y swap spreads from increased borrowingforecasts from the OBR this year or increased gilt
issuance in FY12/13 to be limited.
30Y swap spreads can be explained by the level of 10Y
swap spreads and net QE buying of 30Y gilts (Exhibit
34). As discussed, we don’t expect the BoE to alter the
maturity split of gilt purchases if QE is extended, and we
expect 30Y swap spreads to gradually widen over 1H12,
Exhibit 30: As bonds approach 1Y remaining maturity they tend tounderperform vs. 18M and 2Y gilts
Adjusted yield curve* for selected 1Y vs. 18M and 2Y bond pairs** in the businessdays around gilts falling below 1Y remaining maturity; bp
-20
-15
-10
-5
0
-90 -60 -30 0 30
Days around bond falling sub 1Y
* Yield spread adjusted for 2Y par gilt yield.** Yield spreads used: Mar11/Mar12 yield spread around 7 March 2010; Dec11/Jun12yield spread around 7 December 2010; Mar12/Mar13 spread around 7 March 2011;Jun12/Mar13 spread around 6 June 2011.
Exhibit 31: Following a period of relative stability, 2Y and 10Y swapspreads have substantially widened back to mid-2009 levels2Y and 10Y maturity matched swap spreads; bp
-50
0
50
100
150
Nov 08 May 09 Nov 09 May 10 Nov 10 May 11 Nov 11
2Y 10Y
QE1 QE2
Exhibit 32: 10Y swap spreads can be modeled as a function of giltyields and net QE purchases10Y swap spread regression model*, monthly data November 2008–October 2011 Variable Coefficient T-stat
10Y y ield, % -9.7 -2.2
Monthly Net QE purchases, £bn 2.3 5.2
3Mx 10Y v ol, bp 3.4 2.2
R-squared 59%
Std error 11
* Maturity matched benchmark gilt swap spread.
Exhibit 33: We expect wider swap spreads across the curve, with 30Yspreads breaking the 0bp levelCurrent maturity matched swap spreads and J.P. Morgan forecasts for 2012
Sw ap spreads 18-Nov -11 1Q12 2Q12 3Q12 4Q12
2Y 96 110 120 110 100
5Y 71 85 95 85 65
10Y 31 45 55 45 25
30Y -9 0 5 0 -10
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 89/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Francis DiamondAC
(44-20) 7325-3541
[email protected]. Morgan Securities Ltd
89
reaching and then modestly breaking through the 0bp
level. However, we note that 30Y swap spreads globally
are much tighter than their pre-crisis levels (Exhibit 35),and whilst we expect 30Y UK swap spreads to widen, wedon’t expect a move back to the levels seen in 2007,
given the current global backdrop. In addition, 30Y swap
spreads are more affected by investor flows and range-
trading behaviour, and the relative cheapness of linkerson ASW compared with nominal gilts (see Inflation
Linked Markets) could encourage asset swap investors to
shift out of holding nominal gilts into inflation-linkedgilts on ASW. Given these technical factors, we think
that swap spread widening to 0bp will be a slow
process interspersed by bouts of tightening.
Trading themes
• Be long duration in 10Y and 30Y gilts in 1H12
We expect wider peripheral spreads to drive gilt
yields lower and we forecast 10Y gilts to reach
1.50% by 2Q. The fiscal environment is challenging
but we expect ongoing QE gilt purchases and a
relatively low proportion of non-domestic investors
to prevent UK yields from rising due to a spill over
of sovereign risk fears into the UK
• Enter 2s/10s curve flatteners
The curve will likely remain highly directional with
10Y yields. We forecast almost 90bp of flatteningcompared to what is priced by the forwards by the
middle of 2012
• Bias for a flatter 10s/30s curve
We think the 10s/30s curve can flatten despite our
view of lower nominal yields. As gilt yields fall
further we expect the directionality of the 10s/30s
curve to weaken as investors move out the curve in
search of yield. In addition, further QE gilt purchases should support 30Y gilts relative to 10Y
gilts
•
Position for wider 5Y and 10Y swap spreadsSwap spreads are expected to widen further driven
by wider peripheral spreads, lower nominal yieldsand a total of £425bn of QE gilt purchases by the
end of 2012
Exhibit 34: 30Y swap spreads can be broadly explained by the level of 10Y swap spreads and net QE buying in the 25Y+ sector…
Actual 30Y maturity matched swap spread and model predicted 30Y swap spread*;
weekly data, past 12 months; bp
-30
-25
-20
-15
-10
-5
Nov 10 Feb 11 May 11 Aug 11 Nov 11
Actual Predicted
* 30Y ASW = 0.45 * 10Y ASW + 0.62 * net QE buying of 25Y+ gilts – 25.3. R-squared:58%, std error: 4bp.
Exhibit 35: …and we expect further widening but not back to the levelsseen prior to 2008Maturity matched 30Y swap spread for the UK and weighted global 30Y swap spread*;bp
-75
-50
-25
0
25
50
75
Nov 06 Nov 07 Nov 08 Nov 09 Nov 10 Nov 11
UK Global ex UK
* Calculated as the average of US, German and Japanese 30Y maturity matched swapspreads, weighted by total government bond market value.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 90/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Kimberly L. Harano (1-212) 834-4956J.P. Morgan Securities LLC
90
US Cross Sector
• We look for growth to slow in 1H12 as past fiscal
stimulus programs roll off, but accelerate
thereafter. With the unemployment rate
remaining elevated, core inflation should slow
• Although our baseline forecast does not yet call
for an imminent QE3, we think there is a
significant chance that the Fed resumes balance
sheet expansion in 2012, most likely by buying
MBS
• The European crisis will likely worsen before it
gets better. Key developments in 2012 are likely
to include further funding pressures on Italy, a
Euro-area recession, and increased risks of a
downgrade of France
• The crisis should pressure Treasury yields lower
into 1Q12, but a deteriorating supply/demand
balance, rich valuations, and a decoupling from
Europe will limit the rally
• Despite near-term risks from a worsening crisis,
spread products are likely to benefit from
significant offsets such as solid credit
fundamentals, negative net supply (in securitized
products), and attractive valuations—cautiously
overweight ‘B’ rated high yield bonds, new issue
‘AAA’ and subordinate CMBS relative to legacy
paper, select ABS, and Agencies; stay overweight
EMBIG versus CEMBI
• Overweight mortgages given significant risks of
QE3 and attractive valuations
• Tail risk is not what it used to be: the European
crisis is far from the only “tail” scenario investors
will need to consider in 2012. We characterize
four distinct risks—the European crisis, an EM
Asia hard landing, and deflationary and
inflationary scenarios—and explore the best
hedging strategies
• We find that synthetic strategies designed to
match the duration of liabilities or hedge the
funding gap for defined-benefit pension plans
significantly lower the tracking error and
volatility
The year that was
The past year was an unforgettable one, even if we might
wish we could forget it. It is impossible to forget the
human tragedy in Japan from the earthquake, tsunami
and the nuclear reactor crisis in the first quarter. The
earth appeared to shift under market participants’ feet as
well, with a number of crises coming out of left field.
The year started off strongly, with forecasts for robust
growth pushing equities and bond yields higher and
Exhibit 1: Another wild ride in markets11/17/11 level, quarter-to-date change, quarterly changes over 2011, and changes over 2010, 2009, and 2008 for various market variables
CurrentQTDchg
3Q11chg
2Q11chg
1Q11chg
2010chg
2009chg
2008chg
Global Equities (level)
S&P 500 1216.1 84.7 -189.2 -5.2 68.2 143 212 -565
E-STOXX 2242.8 63.1 -668.9 -62.4 118.1 -172 517 -1952
FTSE 100 5423.1 294.7 -817.2 36.9 8.8 487 979 -2023
Nikkei 225 8479.6 -220.7 -1115.8 61.0 -473.8 -318 1687 -6448
Sovereign par r ates (%)
2Y US Treasury 0.290 -0.006 -0.175 -0.301 0.197 -0.58 0.57 -2.43
10Y US Treasury 2.015 -0.007 -1.238 -0.277 0.148 -0.59 1.09 -1.32
2Y Germany 0.373 -0.135 -1.058 -0.162 0.939 -0.55 -0.46 -2.21
10Y G ermany 1.963 0.024 -1.111 -0.346 0.391 -0.44 0.33 -1.23
2Y JGB 0.132 -0.012 -0.026 -0.023 0.034 -0.02 -0.19 -0.33
10Y JGB 0.895 -0.100 -0.166 -0.143 0.138 -0.17 0.14 -0.28
5Y Sovereign CDS (bp)
Greece 4964 -975 3827 1019 86 721 61 205
Spain 478 94 122 29 -116 235 4 92
Portugal 1173 -42 417 206 101 403 -7 77
Italy 570 96 303 24 -86 126 -58 145
Ireland 763 22 -50 128 57 441 -5 156
Funding spreads (bp)
2Y EUR par swap - par gov't spd 119.6 20.4 38.0 -0.7 -18.3 28.5 -42.3 39.6
2Y USD par swap - par gov't spd 50.9 22.9 5.7 3.8 -3.8 -5.4 -61.2 8.4
EUR FRA-OIS spd 81.4 18.0 34.0 3.1 -7.2 3.3 -39.9 27.0
USD FRA-OIS spd 67.7 24.3 22.6 -0.2 1.7 4.6 -60.9 27.6
1Y EUR-USD xccy basis -88.2 -15.6 -46.5 0.1 22.7 -22.0 22.6 -36.0
Currencies
EUR/USD 1.352 0.009 -0.107 0.031 0.092 -0.103 0.016 -0.056
USD/CHF 0.917 0.011 0.065 -0.073 -0.023 -0.103 -0.012 -0.077
USD/JPY 76.96 0.21 -4.29 -1.73 1.30 -10.81 1.49 -23.00
JPM Trade-weighted USD 81.27 -0.74 4.41 -1.46 -1.71 -3.20 -4.66 6.38
Spreads (bp)
30Y CC MBS L-OAS 47.7 -7.0 17.0 9.6 -2.4 35 -66 59
10Y AAA CMBS spd to swaps 315.0 -50.0 145.0 30.0 -50.0 -260 -375 793
JULI (ex-EM) Z-spd to Tsy 235.6 -13.6 96.7 14.9 -17.4 -11 -363 333
JPM US HY index spd to worst 741.4 -66.6 241.0 50.1 -66.0 -74 -1068 1128
EMBIGLOBAL spd to Tsy 416.6 -48.4 176.5 -10.4 10.3 -6 -430 470
MAGGIE (Euro HG spd to govies) 76.2 9.8 18.1 6.3 -6.8 14 -27 38
US Financials spd to Tsy 325.6 -8.5 153.0 21.1 -19.2 -11 -359 336
Euro Financials spd to gov ies 316.3 18.5 127.9 20.1 -29.4 47 -164 208
10Y AAA muni/Tsy ratio (lev el) 114% 3.8% 25.4% -5.8% -3.0% 18% -47% 36%
30Y AAA muni/Tsy ratio (lev el) 126% 6.0% 21.5% -7.4% -0.9% 18% -104% 97%
Commodities
Gold futures ($/t oz) 1774.30 157.00 106.90 86.60 17.90 313 223 27
Oil futures ($/bbl) 98.82 19.49 -16.09 -11.30 15.34 12.0 34.8 -51.4
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 91/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Kimberly L. Harano (1-212) 834-4956J.P. Morgan Securities LLC
91
causing credit spreads to narrow (Exhibit 1). By the
middle of the year, however, expectations for US growth
plunged sharply as the run-up in commodity prices posed
a threat to growth and the supply shocks from Japan
appeared more severe than anticipated. At the same time,
the sovereign crisis in Europe ballooned from impacting
only smaller, peripheral countries into a much larger
issue, spreading to Spain and Italy, and threatening the
banking system. Third, political deadlock in the USnearly produced a crisis in markets as Congress played a
game of chicken with the vote to increase the debt
ceiling. In part as a result of that political maneuvering,
S&P downgraded the US sovereign rating from AAA to
AA+ with a negative outlook. All of these events led to a
sharp underperformance of risky assets in the second and
third quarter.
With the economy looking vulnerable, the Fed was
forced to ease again, this time launching “Operation
Twist” instead of another round of balance sheet
expansion. Finally, in the fourth quarter, economic data
turned more upbeat, and with some positivedevelopments coming out of Europe after the October 26
summit, US risky asset markets have managed to rally
this quarter. All in all, given crises erupting left and right,
2011 was a year with trading ranges in yields and credit
spreads that were only surpassed by the financial crisis
periods in the US (Exhibit 2).
Given the high volatility, it is unsurprising that investors
were generally not rewarded for taking on additional risk
this year, particularly in high-beta sectors. Unlike years
past, when the riskiest/most illiquid sectors such as
emerging markets and CMBS produced the best returns,
this year CMBS, high yield, and especially equities
sharply underperformed, while longer-maturity US and
German government bonds posted the best returns
(Exhibit 3).
Exhibit 3: In 2011, investors were generally not rewarded for
taking on more risk in high-beta sectorsRolling 3-month total returns (not duration adjusted), averaged over 3/31-10/31/11 versus the standard deviation of those returns for various assetclasses; %
-4
-2
0
2
4
6
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0
Standard deviat ion of 3M returns; %
EM
HY
JULIMBS
Agy
US 1-3Y
DEM 7-10Y
ABS
CMBS
US 7-10Y
DEM 1-3Y
S&P
Exhibit 2: This crisis-filled year produced trading ranges in yields
and credit spreads that were only surpassed by the 2008-09 crisisand recoveryYearly range in 10-year Treasury yields and JULI ex-EM I-spread to Treasury;bp
0
100
200
300
400
20112010200920082007200620052004200320022001
10Y UST yield JULI
Exhibit 4: Risky asset spreads have been very well correlated tomeasures of European exogenous riskCross-sector spread index* versus average CDS for French banks**;Level bp
50
100
150
200
250
300
350
400
-1.0
-0.8
-0.6
-0.4
-0.2
0.0
Nov 10 Jan 11 F eb 11 Apr 11 Jun 11 Jul 11 Sep 11 Nov 11
Cross-sector spread index
French bank CDS
* Average of 3-year z-scores for 5-year swap spreads, JULI spread to Tsy, JPMHY index spread to worst, 10Y AAA CMBS spread to swaps, 2Y AAA card ABSspread to swaps, and EMBIGLOBAL strip spread to Tsy.** Average of 5-year CDS spreads for Societe Generale, Credit Agricole andBNP Paribas.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 92/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Kimberly L. Harano (1-212) 834-4956J.P. Morgan Securities LLC
92
Lifting the covers off 2012: What lies
beneath
Although it has been nearly two years since Greece’s
fiscal problems entered the spotlight, the European crisis
continues to impact US markets in highly significant
ways. On a fundamental basis, Europe now risks falling
back into recession, posing a threat to global growth as
well. In addition, the flight-to-quality premium has been
a factor keeping US Treasury yields well below fair
value. Indeed, we have often noted that yields, adjusted
for their usual long-term drivers as well as short-term
technical factors such as positions, have been very well
correlated to measures of European exogenous risk. This
has been true for risky assets as well: as Exhibit 4
shows, our cross-sector spread index has been very wellcorrelated to French bank CDS spreads. As a result,
proxy variables for the European crisis have become part
of our frameworks for thinking about valuations in
various sectors.
Amidst this backdrop of weak fundamentals and high
geopolitical risk, regulatory and litigation risks have also
been growing. In general, the pace of macro-prudential
regulation has been slower than we anticipated, causing
regulatory uncertainty to persist. For example, the
implementation of Dodd-Frank has been slow, and the
rulemaking process continues to suffer from considerable
delays, perpetuating uncertainty. However, severalcritical changes have come to pass: US bank ratings have
been downgraded as a result of lower systemic support,
and the FDIC changed its deposit insurance assessment
to shift more of the burden onto larger banks.
One area where some progress has been made is money
fund reform. Recently, SEC Chairman Mary Shapiro
indicated that the Financial Stability Oversight Council is
close to proposing rules for money fund reform, which
we expect to be implemented by the end of 2012. At this
point, we think the preferred route for regulation will be
a capital buffer requirement, likely in combination with
gating restrictions. However, we caution that in the
current environment of very low yields it will likely be
difficult for money funds to raise capital (see Short-Term
Fixed Income).
Litigation risk around banks has also taken a toll on asset
markets, particularly in securitized product markets
where banks have a large footprint. To see this, we note
that in an environment where banks are very well
capitalized, litigation risk should negatively impact bank
equity prices more than bank credit. Thus, we define our
legal risk metric as the residual of the log of bank equity
prices regressed against the average CDS level for the
USD Libor bank panel. (We multiply the residual by -1
such that higher values of our index indicate heightened
legal risk.) As Exhibit 5 shows, our index of legal risk
has been well-correlated to MBS spreads, with MBS
cheapening when legal risk rises.
All of these factors will continue to drive markets in
2012. In particular, Europe will very much continue to be
a factor. To help gauge where we are in the crisis, we
find it is interesting to compare the European crisis with
the 2007-08 financial crisis in the US and in Japan in the
early 1990’s. Although the events that may have sparked
the crises are not exactly the same (the bursting of real
estate bubbles in the case of Japan and the US, and the
bursting of a sovereign credit bubble in the case of
Europe), the subsequent propagation and contagion, as
well as the eventual measures that will be necessary to
end the crisis, are similar. In Exhibit 6, we present a
combined crisis timeline, which compares events in theJapan banking crisis to similar events in the US financial
crisis and European sovereign crisis.
As in the US and Japan, the European crisis began with
“localized” losses (on subprime mortgages in the US and
on peripheral sovereign bonds in Europe). As bank
capital positions weakened in an environment of
declining risky asset prices, banks began deleveraging.
These actions now appear likely in Europe, as several
banks have indicated that they expect to delever. In fact,
Exhibit 5: Litigation risk for banks has taken a toll on securitizedproductsLegal risk index* versus 1-week average of 30-year MBS Libor OAS;
bp
10
20
30
40
50
60
70
80
-0.20
-0.15
-0.10
-0.05
0.00
0.05
0.10
0.15
May 10 Aug 10 Nov 10 Mar 11 Jun 11 Sep 11
Legal risk index
MBS OAS
* Legal risk index calculated as the residual of the log of the KBW bank stockindex regressed against the average 5-year CDS spread for USD Libor panelbanks** over the past two years (1-week averages are used). The residual ismultiplied by -1, such that increasing values of the index indicate heightened risk.** Excludes Bank of Nova Scotia.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 93/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Kimberly L. Harano (1-212) 834-4956J.P. Morgan Securities LLC
93
our European bank credit analysts estimate that the 28
large banks they cover will reduce assets by anywhere
from around €830bn to almost €2tn over the next twelve
months in order to raise capital ratios (see “The Great
Bank Deleveraging: Banking Sector Outlook 2012,”
Roberto Henriques et al, 11/4/11).
Eventually, forced recapitalization through public capital
was employed in the US and Japan. Although the EBA
has already mandated that some banks raise €106bn of
capital (rather than de-lever) to improve their capital
ratios, we think another round of capital raising may beneeded given that the EBA’s estimate of capital needs is
nearly €200bn below our estimates. Moreover, definitive
public programs to enforce recapitalization (rather than
deleveraging) have yet to be announced in Europe.
However, in the US case, bank recapitalization alone was
not enough to reopen broken primary markets and short-
term funding markets. The Fed established the CPFF to
provide liquidity for CP issuers, and supported
securitized product markets through TALF. In addition,
Treasury established a guarantee program for money
market funds, and the FDIC established the TLGP to
temporarily guarantee the senior debt of FDIC-insured
institutions. Although some guarantee and deposit
insurance schemes from the 2007-08 financial crisis are
still in play in Europe, comprehensive programs
spanning the Euro area have yet to emerge, and
policymakers are currently focused on other mechanisms
(such as providing monoline insurance) to allow for
primary issuance by affected sovereigns.
Of course, the rest is history—ZIRP and quantitativeeasing followed suit in the US as well as in Japan. It is as
yet unclear if and when this might occur in Europe, but
the ECB did cut rates this month to 1.25%, and we look
for the ECB to cut rates to 0.5% by June 2012.
One last thing to note from our three-crisis comparison is
the timescales. The pace with which policy measures are
unfolding in Europe is clearly slower than was the case
in the US, but hopefully faster than in the Japanese
experience.
Exhibit 6: Side-by-side comparison of Japan’s banking crisis, the US financial crisis, and the European sovereign crisisTimeline of key events in Japan’s banking cr isis and comparable events in the US financial crisis and ongoing European sovereign c risis
US Euro area
1991-92 Peak in land prices Jul-06 Peak in housing prices Aug-09 Trough in sovereign CDS
1996
Diet passes six laws and establishes Housing Loan
Administration Corporation and Resolution Collection Bank to
cope with liquidation and recovery of assets of failed jusen
and cooperatives. Deposit Insurance Scheme strengthened 10/03/08
Congress passes TARP/EESA to buy troubled assets from
financial institutions 05/10/10
EU and IMF launch a €750bn stabilization scheme with
government-backed loan guarantees and a commitment to
buy European sovereign bonds
Nov-97
Government to guarantee full amount of deposits in yen and
other currencies, bank debentures and certain trusts offered
by trust banks 10/03/08
Under TARP/EESA, deposit limit increased to $250K for
individuals; government still working on plan for unlimited
guarantee for small businesses
(Guarantee schemes from the 2008 financial crisis are still in
place)
Feb-98
Two new laws passed including a provision of ¥30tn to
recapitalize banks 10/14/08 TARP modified to prov ide up to $250bn of capital for banks 10/26/11
EU officials call for a €106bn recapitalization of banks and
plan to leverage the EFSF v ia a monoline insurance scheme
or via a leveraged CDO-style investment vehicle
M ar-98 Government injec ts ¥1. 8tn into large banks 10/ 14/ 08
Government injects $125bn into 9 banks; offers another
$125bn to others Jun-12 Recapitalization of banks to be completed
Late
1998 and
Sep-99
Government offers ¥20tn for credit guarantee schemes, and
then introduces scheme to guarantee bonds of small and
medium-sized compnaies Oct-08
FDIC to temporarily guarantee the senior debt of all FDIC-insured institutions and their holding companies, as well as
deposits in non-interest bearing deposit transaction accounts;
Treasury establishes guarantee program for money market
funds; CPFF created to provide a liquidity backstop for CP
issuers 10/26/11
EU officials discussed exploring a coordinated guarantee
scheme, but nothing concrete has been proposed
Feb-99-
Aug-00 BoJ employs zero interest rate pol icy 12/16/08 The Fed introduces a zero interest rate pol icy
EC B cut the refi rate by 25bp to 1.25% on 11/3/11, and we
expect the refi rate to be cut to 0.5% by June 2012
Mar-01 BoJ introduces quantitative easing 03/18/09
The Fed announces Treasury purchases and increased
purchases of Agency debt and MBS (QE1)
ECB has been purchasing sovereign debt, but it is not
"quantitative easing" since they have been sterilizing
purchases
Jun-03 BoJ starts puchasing ABS (including ABCP )
Oct-Nov
08
MMIFF created to purchase assets from money market
funds; TALF created to provide liquidity to ABS investors
JAPAN
Source: Akihiro Kanaya and David Woo, “The Japanese Banking Crisis of the 1990s: Sources and Lessons,” IMF Working Paper, January 2000. Masaaki Kanno, “How Japan’s bankingcrisis ended in a lost decade,” J.P. Morgan Economic Research note, 10/31/08.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 94/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Kimberly L. Harano (1-212) 834-4956J.P. Morgan Securities LLC
94
Look for further deterioration in Europe over the
near term
Our global strategists look for the followingdevelopments in Europe in 2012. First, we expect
funding spreads on the semi-peripherals to worsen early
next year. Second, Greece is likely to continue to fail to
meet austerity targets, which could eventually lead to a
more severe debt restructuring. Third, we think the Euro
area will dip into recession, with the peripheral area hit
hardest. This could make it difficult for EU sovereigns to
meet their fiscal and structural reform targets. Fourth, we
expect increasing risks of Portugal and Ireland needing to
renegotiate their funding packages in mid- to late-2012,
which will likely require private sector involvement.
Finally, France is at risk of losing its AAA rating,
especially if the Euro area economy weakens more than
expected.
Now that sovereign debt of many European countries is
too low-rated to remain in portfolios seeking “risk-free”
assets, these bonds will slowly but surely need to find
new homes in risk-seeking investors’ portfolios. The US
experience with GSE debt in 2008-09 (when foreign
investors sharply scaled back on holdings of this asset
class) makes for a useful comparison, and highlights the
staying power of bearish technicals when such shifts are
underway. Moreover, although the Fed was willing to
intermediate the transfer, by stepping in to buy Agencydebt, the ECB has thus far been reluctant to intermediate
in the size required to stabilize spreads. Thus, we remain
bearish on intra-EMU spreads and expect them to widen
going into 1H12, before concerted policy action pushes
them narrower in 2H12.
Separate from our baseline view, it is interesting to note
that our J.P. Morgan Fixed Income Markets Investor
Survey indicates that 33% of investors believe Greece
will exit the EMU by the end of 2012, while 25% of
investors believe a country other than Greece will exit by
the end of 2013 (see Appendix at end of piece for
detailed results). Thus, with all these risks on the
horizon, we expect European developments to remain a
major driver for markets in 2012.
The outlook for the economy and Fed policy
Beyond European factors, the US growth outlook will
also be an important driver. Although the super
committee failed to agree on a deficit reduction proposal,
given that the mandatory cuts will not begin until 2013,
there is considerable uncertainty around the nature of
fiscal policy adjustments (both potential stimulus
packages and potential deficit reductions) that may occur
over the next year.
In our baseline economic forecast, we assume no
additional fiscal stimulus will be passed by the end of
2011. This is roughly in line with the results of our
investor survey, which shows that 49% of investors
believe no additional stimulus will be passed. However,if any fiscal stimulus is passed, we think it is unlikely to
be large enough to completely offset the drag from past
stimulus programs rolling off, which we estimate will
reduce GDP by 1.7% in 2012 on a year-over-year basis.
As a result, we look for growth to slow substantially in
the first half of 2012 and then recover in the second half
(Exhibit 7). With growth likely to weaken, we expect the
unemployment rate to remain elevated, and high resource
slack should push core and headline inflation lower.
Thus, with the economy likely to weaken further, what
will the Fed do? The results of our investor survey show
that 80% of investors believe the Fed will embark on
another round of balance sheet expansion via asset
purchases (QE3) in 2012. We think the next policy action
the Fed will undertake will most likely be the so-called
“Evans plan,” which calls for specifying the economic
conditions that need to be met for the Fed to tighten. That
said, we also think there is a significant chance that QE3
will be deployed, especially in the form of MBS
purchases, if inflation expectations fall enough. Thus,
QE3 remains on the horizon, even if not yet a force in
markets.
Implications for US fixed incomemarkets
As mentioned above, Europe is likely to remain the
biggest driver of US markets in 2012. In Treasuries, we
think the worsening crisis will sustain the flight-to-
quality bid, pushing yields lower in early 2012 despite
already-rich valuations. Some evidence of these flows
can already be seen in data on foreign purchases:
Europeans have increased purchases of Treasuries, likely
reflecting asset allocation shifts out of riskier EU
Exhibit 7: We expect growth to slow substantially in 1H12 due tofiscal drags and then recover in 2H12J.P. Morgan forecast for real GDP growth, the unemployment rate and core CPI
4Q11 1Q12 2Q12 3Q12 4Q12Real GDP (% q/q, saar) 3.0 0.5 1.5 2.5 2.5
Unemployment rate (% ) 9.0 9.0 9.0 9.0 9.0
Core CPI (% q/q, saar) 1.2 1.2 1.0 1.2 1.2
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 95/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Kimberly L. Harano (1-212) 834-4956J.P. Morgan Securities LLC
95
sovereign debt. However, such effects are unlikely to be
permanent. Correlations between US rates and EU
sovereign spreads will decline eventually, and such portfolio rebalancing will eventually wane, likely later in
1H12. As this European demand subsides, the supply-
demand imbalance in US Treasuries is likely to shift in
an unfavorable direction, and we expect poor technicals
and rich valuations to eventually cause yields to rise over
the course of 2012 (see Treasuries).
At the other end of the risk spectrum, emerging markets
will also remain hostage to European developments
thanks to heightened correlations. However, we think a
number of factors make EM a relatively defensive asset
class. In contrast to 2008, EMBIG yields and spreads are
well above those of—say—high grade corporates,making relative valuations attractive. In addition, stable
growth, greater policy flexibility, low refinancing needs,
a diversified investor base and an upward ratings
trajectory are all supportive factors. Thus, we look for 5-
10% total returns in 2012.
High yield investors face a similar backdrop. On one
hand, Europe remains a risk, but on the other hand, we
think default rates will remain low over the next few
years due to strong liquidity and the health of corporate
balance sheets. We project default rates of 1.5% and
2.0% for high yield bonds and loans, respectively, in
2012; even in a US recessionary scenario, we see defaultrates peaking near 5-6%, versus 10-15% peaks in prior
recessions. High carry and modest spread tightening
should produce attractive total return—we expect returns
to reach 9.6% in 2012. Within the sector, however, we
remain biased towards greater tiering, and recommend
underweighting ‘CCC’s versus higher rated credits. In
particular, we find ‘B’s to be the most attractive rating
category as they offer the best balance between
heightened macro risks, sound credit fundamentals and
low default risk.
In higher quality spread product, we recommend
overweights for various reasons. In CMBS, we are
cautiously optimistic given cheap fundamental credit risk
and negative net issuance of private label CMBS. We
favor new issue ‘AAA’s and subordinates relative to
legacy CMBS. Although early 2012 will very likely
experience ongoing spread volatility, over a 12-month
horizon, we look for legacy benchmark ‘AAA’s to
outperform corporates, tightening by 70bp by mid-year.
Similarly, in ABS, we think credit fundamentals will
remain solid despite stubbornly-high unemployment and
look for technicals to remain favorable due to a persistentsupply shortage. Although we believe ABS spreads have
limited incremental tightening potential from here, we do
think some sectors offer value. Specifically, our top picks
in ABS heading into next year in the ‘AAA’ space are
non-prime auto, retail cards and cross-border ABS. In
addition, we like subordinate credit card and auto ABS.
In Agencies, we are overweight—albeit cautiously so.
Agency spreads remain vulnerable to a worsening of the
European crisis, and our investor survey points to a
desire to reduce exposure to Agencies; however, supply
technicals are strong, as net issuance in 2012 will once
again be significantly negative.
In MBS, the prospect of the Fed buying MBS under a
QE3 program is a powerful wildcard, and should limit
the downside in the asset class. Given attractive spreads
currently, we recommend heading into 2012 with an
overweight.
In high grade, the broad outlines of our view are similar
to our view on rates; prospects for a near-term worsening
in Europe keep us underweight in the near term, but the
significant divergence between strong corporate credit
metrics and wide spread valuations will eventually cause
a decoupling, and likely cause spreads to narrow stronglyat some point in the future. We forecast high grade bond
spreads at 175bp at year-end 2012, versus spreads at
235bp today. Within the high grade market, we
recommend underweighting Financials versus Non-
Financials and underweighting high beta sectors and
those most closely tied to Europe.
Contemplating risks in 2012 …
Although we are generally positive on risky assets over a
medium-term horizon in 2012, given that markets face
unusually high risks, we think it is useful to quantify how
much various spreads might move in an environment of weakening growth and a worsening European crisis. To
do this, we present a stylized model for spreads in
various sectors. We use four factors to model spreads
over the past two years. First, we use 4-week average
jobless claims as a proxy for economic conditions.
Second, we use the size of the Fed’s securities holdings
as a proxy for QE. Third, we use the average CDS spread
for French banks to account for the impact of the
European crisis. Finally, we use an interaction term
defined as the product of French bank CDS and our
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 96/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Kimberly L. Harano (1-212) 834-4956J.P. Morgan Securities LLC
96
flight-to-liquidity index to account for the negative
feedback loop between crisis-induced de-risking and
declining liquidity. Exhibit 8 presents the statistics for
our model.
Although our model does not account for the
idiosyncratic drivers of each market, it does allow us tomake broad projections for spreads based on projections
for the underlying drivers, and—even more
importantly—assess the risk exposure of each sector to
the underlying macroeconomic undercurrents. Exhibit 8
shows our model’s fair value estimate, based on current
levels of the independent variables and also assuming the
Fed’s security holdings increase by $500bn under QE3.
In addition, we use the betas from our model to
characterize the risk exposure in each sector to economic
deterioration (proxied by jobless claims worsening by,
say, 25K), the exposure to a worsening crisis in Europe
(proxied by a 100bp rise in French bank CDS spreads),
and the exposure to QE3 (assuming a $500bn increase inthe Fed’s security holdings). To put the different asset
classes on a comparable footing, we divide the projected
changes by the 2-year standard deviation of 3-month
changes in spreads.
Based on these risk exposure metrics, we find that
CMBS, high yield, and equities are most exposed to a
weakening of the economy. In contrast, CMBS seems to
be least exposed to European risks, whereas CDX.IG, US
Financials, and equities are most exposed. Finally,
CMBS and equities appear poised to benefit the most if
QE3 occurs.
… and tail risks
As investors contemplate the tail risks to their portfolios
in the year ahead, the European crisis perhaps looms
largest in their minds; however, we think there are three
other key risk scenarios that also merit focus. These
include a “hard landing” in China and emerging Asia, a
return of deflation fears in the US, and (in the opposite
direction) the potential for high future inflation (or
inflation expectations) as a result of easy monetary
policy now.
In order to assess the impact of these four scenarios on
markets, we first pick proxy variables that would be
expected to move significantly in these crises—variables
that we could view as almost definitional in
characterizing each of the risk scenarios that are to behedged. For example, we use an average of French bank
CDS as a metric for the EU crisis and define our tail risk
scenario with respect to this metric. We use a basket of
EM currencies to proxy the health of EM Asian
economies and in specifying the risk associated with that
scenario. Third, we use 10-year Treasury yields as a
reference variable for our deflation and high inflation
scenarios.
Exhibit 8: Quantifying the risks for spreads
Current level of spreads; our strategists’ targets for mid-year 2012*; statistics for spreads regressed against 4-week average of jobless claims (000s), size of the Fed’ssecurities holdings ($bn), average French bank CDS** (bp), and average French bank CDS multiplied by flight-to-liquidity index*** using weekly data over past 2 years;projection for spreads if $500bn of QE3 occurs; 2-year standard deviation of 3-month changes, and normalized risks****
Beta T-stat Beta T-stat Beta T-stat Beta T-stat R-sq
10Y AAA CMBS spd to swaps 320 250 3.09 8.8 -0.103 -2.4 0.24 1.2 0.365 3.5 71% 317 81 0.95 0.30 -0.64
JULI (ex-EM) spd to Tsy 235 213 0.46 6.4 -0.025 -3.0 0.35 8.1 0.085 3.9 80% 237 31 0.37 1.11 -0.40
JPM US HY index spd to worst 743 700 2.29 10.7 -0.085 -3.3 1.02 7.9 0.270 3.9 82% 789 95 0.60 1.07 -0.45
EMBIGLOBAL spd to Tsy 415 375 0.69 5.0 0.025 1.5 0.54 6.4 0.074 1.8 74% 446 48 0.36 1.12 0.26
US Financials spd to Tsy 346 300 0.77 7.3 -0.018 -1.5 0.63 10.0 0.139 4.3 86% 357 50 0.39 1.26 -0.18
5Y CDX.IG spd 136 118 0.28 6.0 0.004 0.6 0.28 9.6 -0.005 -0.3 76% 142 18 0.38 1.49 0.10
5Y CDX.HY spd 772 650 2.92 9.6 -0.008 -0.2 1.52 8.3 0.228 2.3 78% 811 150 0.49 1.01 -0.03
S&P 500 (points) 1216 1475 -2.13 -8.6 0.182 6.0 -1.16 -7.6 0.044 0.6 81% 1245 86 -0.62 -1.35 1.06
Projection
with QE3 (bp)
Mid-year
target (bp)
Jobless claims Fed sec holdings French bank CDS FTLIxFrench bank CDSCurrent
level (bp)
Econ
risk
QE3
risk
EU
risk
SD
(bp)
* For JULI, CDX.IG, and CDX.HY, mid-year targets are interpolated based on year-end 2011 targets or current levels and year-end 2012 targets. For US Financials, targetis based on the JULI target and the historical relationship between the two spreads. For the S&P 500, this is the year-end 2011 target.** Average 5-year CDS spread for Societe Generale, Credit Agricole, and BNP Paribas.*** Average of the 2-year Z-scores of (i) spread between 3-month Fed funds strip and 3-month T-bill, (ii) market depth, which is calculated as the half the size of the sum of the top three bids and offers for the 5-year hot run Treasury note, averaged between 8:30am and 10:30am daily, and (iii) 5-year muni pre-refunded bond yields spread to 5-year on-the-run Treasury yield.**** Econ risk assumes a 25K rise in the 4-week average of jobless claims. EU risk assumes a 100bp widening in French bank CDS. QE3 risk assumes the Fed’s balancesheet increases by $500bn. Normalized risks are calculated as the projected change in each of these scenarios is divided by the 2-year standard deviation of 3-monthchanges.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 97/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Kimberly L. Harano (1-212) 834-4956J.P. Morgan Securities LLC
97
Next, we define our tail risk scenarios based on changes
in these variables. For the Europe and EM scenarios, we
define the tail risk as a two standard deviation move in
the reference variables, which works out to a 125bp
widening of French bank CDS and a 2.6 point rise in our
EM currency index. For the deflation risk scenario, we
looked at the beta between 10-year yields and 5Yx5Y
inflation swap rates during periods when both declined
sharply. Then, if we assume that forward inflation swap
rates fall 1.25%, we project that 10-year yields fall 80bp;
this is our specification for the deflationary scenario.
Finally, for the high inflation scenario, we looked at the
relationship between 10-year yields and headline CPI in
1971-74 and 1977-80, two periods when CPI surged over
8%-points. We then estimate that if headline inflation
doubles from current levels, 10-year Treasury yields
should rise 110bp.
In general, the best ways to position for tail risk scenarios
involve taking advantage of heightened correlation in
these crisis periods. Therefore, in order to identify the
best hedges for these four tail risk scenarios, we
calculated empirical betas between a number of market
variables and our proxy variables, both in normal periods
and “stress” periods, which are periods when our proxy
variables moved significantly. Exhibit 9 presents the
results of our analysis, as well as the projected moves for
the various market variables in each risk scenario. We
also show the ratio of the stress move to the normal move
for the market variables that show substantial increases
in correlation in crises.
We can make several observations based on this table.
First, we find that in general, currencies are poor hedges
for crises since “stress betas” were not meaningfully
higher than normal period betas, perhaps reflecting the
impact of FX intervention. Second, for the European
crisis scenario, we find that longs in 10-year Treasuries
and 2s/10s curve flatteners offer attractive “beta pick-
up.” Equities, commodities, and high yield also show
increased correlation in an intensifying European crisis,
but interestingly, high grade credit does not.
Exhibit 9: The tail wagging the dog? Projected market moves in four tail risk scenariosStatistics for various market variables regressed against our proxy variables, in normal periods and stress periods*; projected moves in our risk scenarios**, and ratio of stress move
to normal move
Stress
beta
Normal
beta
Stress
move
Normal
move Ratio
Stress
beta
Normal
beta
Stress
move
Normal
mov e Ratio
Stress
beta
Normal
beta
Stress
move
Normal
mov e Ratio
Stress
beta
Normal
beta
Stress
move
Normal
mov e Ratio
10Y Tsy yields (bp) -0.847 -0.501 -105.8 -62.5 1.69 -20.3 -15.4 -52.2 -39.6 1.32 110 -80
2s/10s Tsy curve (bp) -0.644 -0.171 -80.5 -21.3 3.78 -6.22 -6.02 -16.0 -15.5 46.2 41.2 50.8 45.3 1.12 74.4 41.2 -59.5 -33.0 1.81
10s/30s Tsy curve (bp) 0.021 0.070 2.6 8.8 -1.64 -2.07 -4.2 -5.3 -14.3 -14.9 -15.8 -16.4 4.2 -14.9 -3.4 11.9
S&P 500 (points) -2.22 -1.19 -278 -148 1.87 -99.7 -83.6 -256 -215 1.19 46.9 96.1 51.6 105.7 241 96 -193 -77 2.51
MSCI G7 Index (points) -1.95 -1.15 -244 -144 1.70 -96.4 -82.0 -248 -211 1.18 42.3 83.2 46.6 91.5 206 83 -164 -67 2.47
MSCI EM Index (points) -1.93 -1.30 -241 -163 1.48 -127.5 -109.6 -328 -282 1.16 48.2 88.1 53.0 96.9 188 88 -150 -70 2.13
Nikkei Index (points) -18.1 -10.3 -2264 -1288 1.76 -1024.7 -827.7 -2636 -2129 1.24 337.7 894.2 371.4 983.6 1235 894 -988 -715 1.38
EUR/USD 0.00015 -0.00044 0.019 -0.055 -0.042 -0.046 -0.107 -0.118 0.0157 0.0036 0.017 0.004 0.0563 0.0036 -0.045 -0.003
USD/JPY -0.01176 -0.01156 -1.47 -1.44 -2.487 -1.349 -6.40 -3.47 1.84 2.72 4.08 3.00 4.49 1.48 4.08 -1.19 -3.26
AUD/USD -0.00062 -0.00038 -0.077 -0.047 -0.062 -0.054 -0.160 -0.139 1.15 0.0286 0.0303 0.031 0.033 0.0899 0.0303 -0.072 -0.024
Gold futures ($/t. oz) 1.959 0.933 244.69 116.52 2.10 -49.40 -37.87 -127.1 -97.4 1.30 25.9 -12.5 28.4 -13.8 -2.07 -113.0 -12.5 90.4 10.0
WTI oil futures ($/bbl) -0.149 -0.053 -18.55 -6.57 2.83 -9.59 -7.02 -24.7 -18.1 1.37 9.47 8.01 10.4 8.8 1.18 19.01 8.01 -15.2 -6.4 2.373M fw d copper price ($/m ton) -15.51 -4.12 -1937.45 -514.84 3.76 -714.5 -566.0 -1838.1 -1456.1 1.26 378.8 614.0 416.7 675.4 1131.9 614.0 -905.5 -491.2 1.84
5Y CDX.IG spread (bp) 0.32 0.49 40.27 60.72 31.6 25.2 81.2 64.9 1.25 -19.1 -26.3 -21.0 -28.9 -53.3 -26.3 42.6 21.0 2.03
5Y iTraxx Main spread (bp) 0.44 0.57 54.87 70.70 29.2 23.4 75.2 60.2 1.25 -20.9 -28.4 -22.9 -31.3 -43.6 -28.4 34.9 22.7 1.53
5Y CDX.HY price (points) -0.098 -0.069 -12.26 -8.65 1.42 -2.70 -4.23 -6.95 -10.89 4.21 5.30 4.63 5.83 4.05 5.30 -3.24 -4.24
European crisis EM Asia hard landing High inflation Deflation/low y ields
* For the European crisis, the stress beta is calculated as monthly changes in market variables regressed against monthly changes in average French bank CDS*** over 7/1/11-9/10/11, while the normal beta is calculated over the past 5 years. For the EM Asia hard landing scenario, the stress beta is calculated as weekly changes in market variablesregressed against weekly changes in an EM currency basket**** over 8/1/08-11/30/08 and 8/1/11-10/31/11, while the normal beta is calculated over the past 5 years. For the highinflation scenario, the stress beta is calculated as week ly changes in market variables regressed against weekly changes in 10-year Treasury yields (%) over 5/16/07-6/22/07,5/5/09-8/7/09, and 6/10/11-6/29/11, while the normal beta is calculated over the past 5 years. For the deflation/low yield scenario, the stress beta is calculated as weekly changes inmarket variables regressed against weekly changes in 10-year Treasury yields (%) over 4/29/10-9/1/10 and 7/25/11-10/3/11, while the normal beta is calculated over the past 5years.** For the European crisis scenario, we assume the French bank CDS average rises 125bp. For the EM As ia hard landing scenario, we assume the EM FX basket cheapens 2.6points. For the high inflation scenario, we assume 10-year Treasury yields rise 110bp, and for the deflation scenario, we assume 10-year yields fall 80bp. Then we calculateprojected moves using both the stress beta and normal beta.*** Average 5-year CDS spread for Societe Generale, Credit Agricole, and BNP Paribas; in bp.
**** Average of USD/BRL, USD/INR and USD/MXN.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 98/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Kimberly L. Harano (1-212) 834-4956J.P. Morgan Securities LLC
98
Third, in the EM Asia hard landing scenario, we find
increased correlation for 10-year Treasuries, equities and
commodities, but not for high yield. This scenario is the
only one in which currencies serve as effective hedges:
the yen typically strengthens as Japanese investors
unwind long-EM trades, and the Aussie dollar typically
weakens along with other EM/commodity currencies.
Fourth, we find that hedging against high inflation is
difficult, with very few market variables displaying
increased correlation in this scenario. Only 2s/10s curve
steepeners and longs in gold or oil show up as attractive
hedge trades. In contrast, several market variables show
increased correlation in a deflation scenario, with selling
equities and commodities among the most attractive
trades. It is also interesting to note that the beta between
equities and yields in a deflation scenario is higher than
the beta in inflation scenarios. This is likely because a
weak growth/deflation scenario is clearly negative for
equities, whereas the effect of high inflation on equities
is mixed—on one hand, equities are real assets, so equity
prices should rise with inflation, but on the other hand,
increased risks of Fed tightening should constrain prices.
Now that we have defined our scenarios and our
projected market moves, in order to evaluate which
hedges are currently most attractive, we look at digital
and dual digital option prices for these market variables.
Digital options, which give a payoff of 1 if a certain
condition is met, are insightful because they may be
interpreted as the market-implied probability of a given
scenario. In Exhibit 10, we show indicative prices for
several digital options and dual digital options that are
potential hedge trades in our four scenarios. We have
chosen the strikes for the single digital options based on
the expected “stress moves” of market variables, while
the strikes for the dual digital options are based on the
“normal moves” as implied by Exhibit 9 and reflect
moves that are highly likely to occur in a given risk
scenario. The rationale for choosing less out-of-the-
money strikes in dual-digitals is this: although each move
may be more probable individually, as well as jointly in a
given tail risk scenario, the dual digital can often trade
Exhibit 10: Indicative prices for digital and dual digital options that we expect to pay out under our stress scenariosInstrument, strike*, payoff if market is below or above the strike, and indicative premium for 1-year digital or dual digital option, with USD payouts (where applicable); the
most attractive trades for a given panel are highlightedEUROPEAN CRISIS
Instrument Strike Payoff if Premium
10Y Tsy yields (bp) -105 Below 17.6%
2s/10s Tsy curve (bp) -80 Below 11.4%
S&P 500 (pts) -275 Below 22.0%
Gold futures ($/t. oz) 245 Above 30.0%
WTI futures ($/bbl) -18.5 Below 35.3%
Copper fut. ($/m ton) -1950 Below 33.5%
Instrument #1 Strike Pay off if Instrument #2 Strike Pay off if Premium
Gold futures ($/t. oz) 115 Above S&P 500 (pts) -145 Below 15.6%
WTI futures ($/bbl) -6.5 Below S&P 500 (pts) -145 Below 23.8%
Copper fut. ($/m ton) -515 Below S&P 500 (pts) -145 Below 23.5%
EM ASIA HARD LANDING
Instrument Strike Payoff if Premium
10Y Tsy yields (bp) -50 Below 42.4%
Nikkei Index (points) -2600 Below 10.7%
USD/JPY -6.5 Below 19.0%
Gold futures ($/t. oz) -125 Below 50.0%
WTI futures ($/bbl) -25 Below 29.1%
Copper fut. ($/m ton) -1840 Below 34.1%
Instrument #1 Strike Pay off if Instrument #2 Strike Pay off if Premium
Nikkei Index (points) -2100 Below USD/JPY -3.5 Below 12.5%
Nikkei Index (points) -2100 Below Gold futures ($/t. oz) -95 Below 11.3%
Gold futures ($/t. oz) -95 Below 10Y Tsy y ields (bp) -40 Below 25.0%
Copper price ($/m ton) -1450 Below USD/JPY -3.5 Below 23.0%
HIGH INFLATION
Instrument Strike Payoff if Premium
Gold futures ($/t. oz) 30 Above 46.0%
WTI oil futures ($/bbl) 10.5 Above 44.7%
10Y Tsy yields (bp) 110 Above 13.3%
Instrument #1 Strike Pay off if Instrument #2 Strike Pay off if Premium
Gold futures ($/t. oz) 15 Abov e WTI futures ($/bbl) 9 Abov e 34.6%
DEFLATION/LOW YIELDS
Instrument Strike Payoff if Premium
2s/10s Tsy curve (bp) -60 B elow 15.4%
S&P 500 (pts) -190 Below 28.1%
WTI futures ($/bbl) -15 Below 37.5%
Copper fut. ($/m ton) -900 Below 39.3%
CDX.IG spd (bp) 40 Above 42.0%
10Y Tsy yields (bp) -80 Below 27.4%
Instrument #1 Strike Pay off if Instrument #2 Strike Pay off if Premium
S&P 500 (pts) -75 Below WTI futures ($/bbl) -6.5 Below 27.8%
S&P 500 (pts) -75 Below Copper fut. ($/m ton) -490 Below 28.9%
S&P 500 (pts) -75 Below CDX.IG spd (bp) 20 Above 42.0%
* Strike is ATM spot + indicated value for equity indices, and ATM forward + indicated value for all others.
Note: Swaptions are used instead of Treasury options.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 99/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Kimberly L. Harano (1-212) 834-4956J.P. Morgan Securities LLC
99
cheap because the correlation is underpriced in normal
markets. As such, cheap dual digitals—based on less
extreme strikes—will give an indication of which pair-
wise correlations are underpriced currently in markets,
relative to what can be expected in a tail risk scenario.
For example, our analysis projects that the S&P 500 will
fall 278 points if the European crisis worsens
significantly, compared to a 148-point fall using normal
betas. One way to hedge for such a move in the S&P 500
is to buy a 1-year digital option that pays out if the S&P
500 falls 275 points. We estimate that the premium on
this option is approximately 22.0%. Alternatively, one
can buy a dual digital option, which pays out only if two
conditions are met. In this case, we use two weaker
conditions, both of which would be very likely to occur
in the risk scenario. For example, one can buy an option
that pays out only if the S&P 500 falls to a certain level
and gold futures rise to a certain level. Since we projectthat the S&P 500 will fall 148 points and that gold
futures will rise $116.5 based on typical betas, we
consider a dual digital option that pays out only if the
S&P 500 falls at least 145 points and gold futures rise as
least $115. Interestingly, even though the dual digital has
strikes that are closer to at-the-money, the premium on
the dual digital option is cheaper—15.6% versus 22.0%
for the single S&P 500 option. This suggests that the
correlation between S&P 500 and gold in a European
crisis scenario as predicted by our analysis is not priced
into markets. Put differently, tail risk is priced in to a
greater extent in the S&P implied distribution, but less
priced into the S&P/gold correlation market.
Indeed, we find that the correlations between
commodities and other variables appear underpriced
versus what our model predicts for all four scenarios. In
each case, a dual digital structure that pairs a commodity
with another variable is cheaper than a single digital
option on that commodity.
We can draw several other conclusions from the table.
First, we find that floors on the 2s/10s Treasury curve
appear to be the most cost-efficient way to position for a
deterioration of the European crisis. The dual digital with
gold and S&P 500 is the second cheapest way to hedge
the European crisis, and is cheaper than all the other
single digitals we considered.
Second, the best ways to position for economic weaknessin EM Asia tend to involve exploiting correlations with
other liquid markets, particularly the Nikkei Index. We
find that puts on the Nikkei are the most cost-effective
way to hedge for EM Asia weakness, followed closely by
dual digitals involving the Nikkei and gold futures or the
Nikkei and USD/JPY.
Third, for the two inflation scenarios, rate hedges are the
most attractive—for the high inflation scenario, outright
payer swaptions are most attractive, and for the deflation
scenario, floors on the 2s/10s Treasury curve are
cheapest. This is likely because changes in inflation
expectations have the greatest impact on rates as opposedto other asset classes.
Exhibit 11 presents a summary of the best trading
themes for each risk scenario.
Trends in supply and demand in fixed
income markets
Net issuance across fixed income markets averaged
around $1.5tn over 2010 and 2011, down from the nearly
Exhibit 11: Summary of the best tail risk hedgesSummary of the best trading themes for each risk scenario based on Exhibit 10Scenario Best trading themes
European crisis 2s/10s curve floors, yield floors, S&P puts
EM A sia hard landing Nikkei puts, USD puts/JPY calls
H igh inflation P ay er sw aptions
Deflation/low y ields 2s/10s curve floors, yield floors, S&P and commodity puts
Exhibit 12: We expect total net issuance across fixed incomemarkets to decline in 2012Historical data and J.P. Morgan forecast for long-term (>1Y) net issuance in2011 and 2012; $bn
Net ($ bn) 2004 2005 2006 2007 2008 2009 2010 2011 E 2012 E
IG corporates 215 169 326 408 103 95 247 353 303
HY corporates -17 -25 15 6 -25 76 151 100 85
EM corporates 54 76 102 117 17 80 150 141 129
Municipals 166 156 173 38 29 108 34 -71 3
Non-Agency MBS 423 616 539 201 -359 -379 -274 -218 -186
Agency MBS -24 97 269 510 519 524 -80 0 -55
CMBS 79 97 165 175 -40 -47 -57 -46 -17
ABS 1 50 38 47 -50 -25 -96 -50 -50
CLOs 25 43 86 76 5 -17 -15 -8 -13
Agency Debt 116 54 86 -64 -39 -19 -119 -162 -185
Treasuries 296 263 177 135 396 1,549 1,611 1,364 1,036
Total 1,334 1,596 1,976 1,649 556 1,945 1,552 1,403 1,050
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 100/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Kimberly L. Harano (1-212) 834-4956J.P. Morgan Securities LLC
100
$2tn peak reached in 2009. Next year, we expect net
issuance to take another leg down, driven primarily by a
sharp decline in net issuance of long-term Treasuries(Exhibit 12). Although we expect gross issuance of
coupon Treasuries to be virtually unchanged in 2012, the
net number will likely fall about $328bn due to increased
redemptions. Muni issuance, on the other hand, is
projected to rebound from -$71 to $3bn. Finally, we
expect net issuance across corporates and securitized
products to remain roughly the same at $520bn and
-$320bn, respectively.
How does our supply projection compare to expected
demand? In Exhibit 13, we present our projections for
the sources of “organic” demand for fixed income assets
in 2012 from six major classes of investors: mutualfunds, foreigners, insurance companies, commercial
banks, pension and retirement funds, and the Federal
Reserve. Clearly, prices in markets will adjust so that
demand will meet supply, but we think it is useful to
estimate the demand flows that might be expected given
the current state of markets.
First, we can think of mutual fund purchases of fixed
income assets as being driven by two factors: on one
hand, as the economy weakens, mutual funds will likely
increase their purchases of bonds, and do the opposite if
economic prospects improve. Second, as front-end rates
fall, investors will likely shift out of money market fundsand into bond mutual funds in order to pick up yield.
Thus, we find that using initial jobless claims as a
barometer for the economy and 3-month T-bill yields as
a proxy for money market rates does a good job of
explaining changes in mutual fund holdings of fixed
income over the past 25 years. If we assume that initial
jobless claims rise slightly as the economy weakens in
1H12 and that T-bill yields remain near current levels,
then we project that mutual funds will add about $230bn
of fixed income assets in 2012, close to the average
annual purchase over 2006-10.
Second, we model changes in foreign holdings of fixed
income assets by regressing annual changes in holdings
versus the annual trade deficit of the US. Given that we
expect the trade deficit to widen modestly next year, to
about $582bn, we project $585bn of net purchases by
foreigners in 2012. This is slightly below the net
purchases we have seen over 2006-10.
Third, we model flows from insurance companies using
the slope of the 10s/30s curve (with a steeper curve
leading to greater purchases) and the backward-looking
performance of the S&P 500 (with positive past
performance leading to lower fixed income purchases).
Given that the 10s/30s curve has averaged around 1.15%
over the past year, and given that quarterly changes in the
S&P have averaged around 140 points, we project that
insurance companies will purchase around $165bn fixed
income assets in 2012. This is substantially higher than
the average over 2006-10, but slightly below the $200bn
Exhibit 13: Expected 2012 fixed income purchases by variousinvestor classesStatistics for regression models for quarterly changes (1-year average) or
annual changes in fixed income holdings for various investor classes, estimatedpurchases based on forecast for variables, expected purchases in 2012, andaverage actual annual purchase over 2006-10 ; $bnMUTUAL FUNDS
Period 3Q86 -2Q11
R-sq 58%
Factor Beta T-stat Forecast
Initial jobless claims (1Y avg; 000s) 0.12 3.7 420
3-month T-bill y ield (1Y avg; % ) -5.82 -6.5 0.11
Intercept 8.22 0.6
Estimated 2012 purchases ($bn) 228
Actual avg purchase over 2006-10 255
FOREIGNERS
Period 1992-2010
R-sq 66%
Factor Beta T-stat Forecast
Annual trade balance ($bn) -0.94 -5.7 -582
Intercept 39.67 0.6
Estimated 2012 purchases ($bn) 585
Actual avg purchase over 2006-10 647
INSURANCE COMPANIES
Period 3Q86-2Q11
R-sq 55%
Factor Beta T-stat Forecast
1Y lag 10s/30s Tsy curve (1Y avg; % ) 23.46 5.2 1.10
1Y lag 1Y chg in S&P 500 (1Y avg; pts) -0.06 -6.8 140
Intercept 23.35 11.1
Estimated 2012 purchases ($bn) 163 Actual avg purchase over 2006-10 86
COMMERCIAL BANKS
Ex pected 2012 purchases ($bn) 80
Actual avg purchase over 2006-10 113
PENSION AND RETIREMENT FUNDS
Ex pected 2012 purchases ($bn) 135
Actual avg purchase over 2006-10 138
FEDERAL RESERVE
Ex pected 2012 purchases ($bn) 0 But some chance of QE3
Actual avg purchase over 2006-10 282
T OT AL EXPECT ED PURCHASES ($bn) 1191
Note: Insurance companies consist of life insurance and property & casualtyinsurance companies. Pension and retirement funds include private pension
funds, as well as state, local, and federal government retirement funds.Source: Federal Reserve Flow of Funds
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 101/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Kimberly L. Harano (1-212) 834-4956J.P. Morgan Securities LLC
101
and $188bn of purchases seen in 2009 and 2010,
respectively.
For banks and pension funds, we assume that they
maintain their pace of purchases over recent years. Banks
added $83bn of fixed income securities in 2010 and are
on track to add a similar amount in 2011, so we estimate
that banks will purchase around $80bn of fixed income
assets in 2012. Similarly, pension funds have been
steadily adding around $135bn of fixed income assets
each year, so that is our estimate for 2012 purchases.
Finally, our baseline view is that the Federal Reserve will
not (yet) embark on QE3, but there is a significant risk
that it will sometime in 2012, particularly involving
mortgages. All in all, we estimate net organic demand for fixed-income assets based on current conditions to be
about $1200bn in 2012, with additional upside risk due
to the Fed. This is comfortably above our net supply
forecast.
Aggregate measures of net demand and net supply do not
tell the whole story, however. The majority of fixed
income investor classes discussed in our table are natural
buyers of fixed income spread product ; in contrast, net
supply in 2012 will come almost entirely from
Treasuries. With Fed purchases of Treasuries likely to be
zero on a net basis in 2012 (based on our assumption that
if QE3 occurs, it will be in MBS), the Treasury marketwill likely witness a deteriorating supply/demand
imbalance as we move through 2012 (see Treasuries).
Investors’ asset allocation plans for
2012
In our J.P. Morgan US Fixed Income Investor Survey,
we also asked investors about the changes they plan to
make to their positions over the next year. In Exhibit 14,
we show the weighted average response for expected
changes in portfolio currency allocations. Unsurprisingly,
investors generally plan to reduce exposure to European
assets and add exposure elsewhere. In particular, most
investors indicated that they planned to add exposure to
emerging market assets (denominated in local currencies)
and US assets.
Exhibit 15 shows how investors plan to change their
asset class exposures over the next year. According to
our survey, investors generally plan to reduce exposure
to Agency debt and cash, with a smaller fraction
planning to reduce duration risk on net. At the opposite
end of the spectrum, investors generally plan to increase
exposure to high yield and investment grade corporates.
They also plan to add exposure to Agency MBS, perhaps
in anticipation of outperformance due to additional Fed
purchases of mortgages. It is interesting to note that only
a small fraction of investors plan to add exposure to
equities and commodities this year, since real assets like
equities and TIPS were investors’ top picks last year (see
US Fixed Income Markets 2011 Outlook , 11/24/10).
Exhibit 14: Investors plan to reduce exposure to European assetsand add exposure to US and EM local currency assetsWeighted expected change in exposure for assets in various currencies
according to our J.P. Morgan US Fixed Income Markets Investor Survey
-0.07
0.07 0.08
0.16 0.17
-0.10
-0.05
0.00
0.05
0.10
0.15
0.20
EUR ass ets Other DM EM USDassets USD assets EM local FXassets * To calculate the weighted expected change in exposure for each asset class,we assigned values of 1, 0, and -1 to responses of add, maintain, and reduceexposure, respectively, and then we summed the values and divided by thenumber of responses.
Exhibit 15: Investors plan to add exposure to corporate credit riskand reduce exposure to Agency debt and cashWeighted expected change in exposure* for various asset classes according toour J.P. Morgan US Fixed Income Markets Investor Survey
-0.19 -0.19
-0.04
0.040.07 0.09 0.09
0.160.19 0.20 0.20
0.24
0.30
-0.2
-0.1
0.0
0.1
0.2
0.3
A g y
d e
b t
C a s
h
D u r ' n r i s
k
C o m m o
d i t i e s
E q u
i t i e s
N o n - A
g y
M B S
A B S
T I P S
C M B S
E M
A g y
M B S
I G H Y
* To calculate the weighted expected change in exposure for each asset class,we assigned values of 1, 0, and -1 to responses of add, maintain, and reduceexposure, respectively, and then we summed the values and divided by the
number of responses.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 102/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Kimberly L. Harano (1-212) 834-4956J.P. Morgan Securities LLC
102
ALM strategies for pension funds
Over the first half of 2011, the funding gap for defined
benefit pension funds fluctuated around -$200bn, then it
deteriorated sharply as equities and rates both tumbled in
the second half of the year. As of the end of October, the
funding gap stood at -$398bn, nearly double its value at
the start of the year (Exhibit 16). This dramatic
deterioration highlights the risks of running a significant
asset-liability duration mismatch.
Indeed, the most recent Flow of Funds data on defined
benefit pension fund assets showed that funds had 49%
of their assets invested in equities and mutual funds and
only 38% invested in fixed income securities
(Exhibit 17). If we assume that fixed income assets
mirror the composition of an aggregate index like the J.P.
Morgan Global Aggregate Index (GABI), then given the
38% fixed income allocation, the estimated duration of
assets would be only 1.9 years as of the end of October.
On the other hand, if we regress monthly changes in D.B.
pension fund liabilities against monthly changes in 30-year swap rates and AA rated corporate bond spreads, we
estimate an empirical duration of $128bn per 1%-pt
change in 30-year swap rates (Exhibit 18). Given
aggregate liabilities of about $1615bn currently, this
relationship implies the duration of liabilities is around
7.9 years, substantially greater than the estimated asset
duration of 1.9 years.
One way to address this asset-liability duration gap is to
increase allocations to fixed income, and specifically
long-duration assets, since the duration of pension fund
liabilities is longer than the aggregate duration of the US
fixed income universe. However, the prospects for better
ALM-matching using cash assets may be difficult given
our expectation of low long-end fixed income supply. As
Exhibit 19 shows, long-end gross issuance fell sharply in
2011 due to declines in municipal and high grade
corporate supply. Although we expect long-end gross
issuance to increase in 2012, it will likely remain below
Exhibit 18: Estimated empirical duration of pension fund liabilitiesMonthly changes in top 100 D.B. pension fund liabilities regressed against monthlychanges in 30-year swap rates (%) and monthly changes in JULI AA spread to
Libor (bp); monthly data over 11/08-10/11; $bn
-100
-50
0
50
100
150
-1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8
1-mo chg in 30Y swap rate; %
Y = -127.8(swap rate chg) - 1.07(JULI spdchg) + 0.59R-sq = 75%
Source: Milliman 100 monthly Pension Funding Index
Exhibit 19: Issuance of long-term debt will likely increase in 2012,but remain below historical averagesGross issuance of long-term (>10Y) debt, actual and forecast for 2011 and 2012, anddebt currently outstanding with >10Y to maturity*; $bn
2005 2006 2007 2008 2009 2010 2011 E 2012 E
Treasuries 14 41 42 49 143 183 191 194 830
Corporates 67 104 162 95 118 120 85 115 1080
Municipals 334 320 373 332 303 301 135 169 1678
TOTAL 415 465 578 477 564 605 411 478 3588
Current
outstanding
Gross issuance of long-term debt
* For Treasuries, we exclude the amounts held by the Fed. Corporate issuance isfixed-rate only.
Exhibit 16: The defined benefit pension funding gap has nearlydoubled during 2011
Assets minus liabilities for top 100 defined benefit pension plans; $bn
-500
-400
-300
-200
-100
0
Oct 08 May 09 Nov 09 Jun 10 Dec 10 Jun 11
Source: Milliman 100 monthly Pension Funding Index
Exhibit 17: Pension fund asset breakdown Assets of private defined benefit pension funds as of 2Q 2011; $bn
Fixed income 848
Tsys 353
GSE-backed securities 121
Corporates/foreign bonds 374
Equities 1095
Other 272
Total assets 2215
Source: Federal Reserve Flow of Funds Table L.118.b. Note “Equities” includesequities and mutual fund holdings.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 103/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Kimberly L. Harano (1-212) 834-4956J.P. Morgan Securities LLC
103
the average levels seen over 2006-10. Furthermore, the
Fed’s Operation Twist will likely take another $100bn of
long-end Treasuries out of the market in December
through June.
In such an environment of constrained long-term debt,
we think synthetic overlay strategies to increase duration
and spread exposure are attractive. We consider two sets
of strategies. The first set of strategies is designed to
match the duration exposure of the liabilities. This can be
thought of as the “end game” position that pension funds
undertaking ALM strategies aim to reach, and represents
the benchmark for any “new” investments that a pension
fund manager might use. The second set of strategies is
designed to hedge the funding gap itself, and could be
useful to a fund manager seeking to hedge all or part of a pension fund’s current funding gap.
For the liability-matching strategies, we examine three
variations of overlay strategies: 1) buying 30-year
(classic bond) Treasury futures, 2) receiving fixed in 30-
year swaps, and 3) receiving fixed in 30-year swaps and
buying risk (selling protection) in 5-year CDX.IG. We
calculated the monthly returns on these three strategies
assuming monthly rebalancing, whereby the duration and
credit exposures were given by the rolling 3-year betas of
changes in liabilities regressed against changes in 30-
year swap rates and AA rated corporate bond spreads (as
shown in Exhibit 17). We then compared the returns on
these strategies to the actual changes in liabilities, and in
Exhibit 20, we show the monthly tracking error over
January 2010-October 2011. As the exhibit shows, all
three synthetic strategies showed substantially lower
magnitudes of tracking error. Unsurprisingly, the strategy
Exhibit 20: Strategies to match the duration of liabilities hadsignificantly lower tracking error than the actual assetsStandard deviation of the net P/L of three strategies* and of the actual change inassets minus liabilities**; 1/10 – 10/11; $bn
29.5
26.524.5
50.1
20
25
30
35
40
45
50
55
Futures Swaps Swaps + CDX Actual assets
* Net P/L defined as monthly P/L on assets plus any overlays, net of monthlychange in value of liabilities. Strategies match the duration and spread exposureas given by rolling 3-year regressions of monthly changes in liabilities versusmonthly changes in 30-year swap rates and AA corporate spreads to Libor.They involve 1) buying 30-year (classic bond) Treasury futures, 2) receivingfixed in 2Mx30Y swaps, and 3) receiving fixed in 2Mx30Y and buying risk(selling protection) in 5-year CDX.IG. We assume monthly rebalancing.** Change in assets minus change in liabilities for top 100 defined benefit plansas given by Milliman 100 monthly Pension Funding Index.
Exhibit 21: The pension funding gap is exposed to rates, creditspreads, and equitiesMonthly changes in top 100 D.B. pension fund funding gap regressed againstmonthly changes in 30-year swap rates (%),S&P 500 (points), and JULI AA spreadto Libor (bp); monthly data over 11/08-10/11; $bn
-150
-100
-50
0
50
100
-1.5 -1.0 -0.5 0.0 0.5 1.01-mo chg in 30Y swap rate; %
Y = 134.9(swap chg) + 0.61*(S&P chg) +1.36(JULI spd chg)+3.95R-sq = 85%
Source: Milliman 100 monthly Pension Funding Index
Exhibit 22: Strategies to hedge the funding gap substantiallylowered the volatility of the gapStandard deviation of monthly changes in gap plus P/L on hedge strategy* and of actual monthly change in funding gap**; 1/10-10/11; $bn
25.1
28.1
50.1
20
25
30
35
40
45
50
55
Swaps + C DX+ S&P Fut ures + C DX+ S&P C hg in gap
* Strategies hedge the duration, equity and spread exposure as given by rolling3-year regressions of monthly changes in the funding gap versus monthlychanges in 30-year swap rates, the S&P 500, and AA corporate spreads toLibor. They involve 1) buying 30-year (classic bond) Treasury futures or receiving fixed in 2Mx30Y swaps, 2) buying risk (selling protection) in 5-year
CDX.IG, and 3) selling S&P 500 futures. We assume monthly rebalancing.** Change in assets minus change in liabilities for top 100 defined benefit plansas given by Milliman 100 monthly Pension Funding Index.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 104/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Kimberly L. Harano (1-212) 834-4956J.P. Morgan Securities LLC
104
involving swaps and CDX.IG had the lowest tracking
error, since it hedged both the duration and credit
exposure of the liabilities. The outperformance of thestrategy using swaps versus the strategy using futures
may be in part explained by the fact that we used 30-year
swap rates in our models to determine the duration
exposure, while the bond futures contract (currently the
most liquid long-end Treasury futures instrument) creates
exposure to the 15-year point.
To hedge the funding gap, we first have to identify the
drivers of the gap. Exhibit 21 shows that changes in the
funding gap have been well explained by changes in 30-
year swap rates, AA rated corporate bond spreads, and
the S&P 500, with increases in these factors leading to an
improvement in the gap, and decreases leading to aworsening of the gap. Thus, the strategies to hedge the
funding gap involve positions with the opposite
exposures, i.e., a long duration position (either via 30-
year swaps or bond futures), a long risk position in
CDX.IG, and a short in S&P futures. We calculated the
returns on these strategies assuming monthly rebalancing
based on the betas from rolling 3-year regressions, and
we combined them with the actual changes in funding
gap. As Exhibit 22 shows, using these hedge strategies
significantly lowered the monthly volatility of the
funding gap. As with the first set of strategies, the
outperformance of the strategy using swaps is likely
partly because 30-swap rates are used in our empiricalestimation of the duration exposures.
Cross sector trading themes
• Stay long duration early in 2012 targeting 10-year
yields to reach 1.70% in 1Q12
A worsening of the European sovereign debt crisis
should accelerate the flight-to-quality bid into
Treasuries early next year pushing yields lower; rich
valuations, a deteriorating supply/demand imbalance,
and poor technicals will limit the upside, however, and
keep rates range bound later in 2012.
• Look for TIPS breakevens to narrow sharply over
the next few months and then widen modestly over
2012
Breakevens will hit a number of headwinds in early
2012: we expect nominal yields to plummet, fiscal
policy to tighten, and headline inflation to fall. Thus,
our fair value model projects that breakevens should
narrow over the next few months, though breakevens
should widen over the remainder of the year as
nominal rates rise. QE3 also poses an upside risk to
our targets for breakevens.
• Position for wider 10-year maturity matched swap
spreads in the near term, but look to initiate
narrowers towards the end of 1Q12
In the near term, FRA-OIS spreads will likely
continue to widen, and banking stock valuations will
likely remain under pressure, both of which should
pressure 10-year swap spreads wider. We expect 10-
year swap spreads to hover near 27bp towards the end
of this year and in early 1Q12. Beyond that point,
issuance-related swapping is likely to pick up, and the
growing likelihood of concerted action by central
banks should help spreads narrow. Look for narrowing
to 18bp by mid-year.
• Stay long gamma going into year end and in 1H12
A persistent crisis in Europe and poor risk appetite
should cause market depth to stay depressed for much
of 4Q11 and 1H12, creating conditions favorable for
long gamma positions.
• Modestly overweight Agency bullets tactically
hedged with swap spread wideners
With 5-year Agency spreads wider than our T+33bp
fair value target for 2Q12, we recommend an
overweight position. However, given the significant
impact of the European crisis on Agency spreads, werecommend tactically overlaying swap spread
wideners as a hedge. In particular, we like adding to
front-end bullets (2- to 3-year) on dislocations. We
also recommend that investors use short-dated
callables to enhance excess returns versus lockout-
matched bullets. Given greater clarity around short-
term rates due to the Fed’s low-for-long policy, we
would look at structures with lockouts as far out as 12-
15 months.
• Modestly overweight MBS going into the new year
We recommend beginning the new year with an
overweight given solid fundamentals and the risk of
QE3 in mortgages. Technicals should also be
supportive given negative net supply. However, we
only look for 10bp of tightening in the basis due to
lower leverage and higher capital requirements.
• Add CMBS versus corporates
We position relatively defensively and favor new issue
‘AAA’s and subordinates relative to legacy CMBS.
For investors looking for yield in legacy paper, add
exposure to higher-quality 2005/2006 AJs but avoid
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 105/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Kimberly L. Harano (1-212) 834-4956J.P. Morgan Securities LLC
105
2007 AJs given higher downside in worse-than-
expected economic scenarios. While early 2012 will
experience ongoing spread volatility, over a 12-monthhorizon, look for legacy benchmark ‘AAA’s to
outperform corporates, tightening by 70bp by mid-
year.
• We are neutral on ABS
ABS spreads likely have limited incremental
tightening from here and tiering will be pronounced.
Within ABS, our top picks in ABS heading into next
year in the ‘AAA’ space are non-prime auto, retail
cards and cross-border ABS. In addition, we like
subordinate credit card and auto ABS.
• Stay underweight high grade corporates but look for spreads to tighten later in 2012
Although we are currently underweight, we look for
spreads to tighten sharply at some point in the future
given the divergence between strong corporate credit
metrics and wide spreads. We forecast high grade
bond spreads at 175bp at year-end 2012, versus
spreads at 235bp today. On a sector basis, we
recommend underweighting Financials versus Non-
Financials and underweighting high beta sectors and
those most closely tied to Europe.
• Remain bearish on CDX in the near term but look
for spreads to tighten over the longer term
Similar to high grade cash markets, we expect strong
credit fundamentals and technical forces to drive
spreads tighter. At year-end 2012, we expect CDX.IG
to trade at 100bp (versus 136bp today) and CDX.HY
at 550bp/$98.0 (versus 772bp/$90.0 today), in line
with our 175bp year-end HG bonds forecast. In
addition, we believe that investors should take
advantage of opportunities where CDS-bond basis is
more negative than -75bp in HG and -100bp in HY.
• Overweight ‘B’s, be neutral on ‘BB’s, and
underweight ‘CCC’s in high yield
We think default risk is likely to be negligible over the
next few years due to strong liquidity and the health of
corporate balance sheets. Thus, we forecast high-yield
bond and loan spreads will tighten to T+705bp and
L+660bp by year-end 2012. In particular, we find ‘B’s
to be the most attractive rating category as they offer
the best balance between heightened macro risks,
sound credit fundamentals and low default risk.
• Stay overweight Senior CLOs (AAA/AA) but
neutral CLO Mezzanine to Subordinates
(A/BBB/BB)Given the macro risks and lack of liquidity, we stay
neutral CLO Mezzanine to Subordinates (A/BBB/BB)
but overweight Seniors (AAA/AA). Trading themes
include contraction upside as reinvestment periods
end, spread tiering in mezzanine bonds, and some
opportunities to take advantage of the high current
carry in equity
• Look for muni ratios to fall and for spreads to
tighten in early 2012
We look for ratios to fall in early 2012, but remain
elevated versus historic norms due to strong
reinvestment capital, high volatility, modest inflows,and some flight-to-quality bid. Over the first half,
spreads may also tighten as yields remain ultra low
and both market returns and the credit environment
remain unremarkable. As a relative value trade, we
like overweighting ‘A’ and ‘BBB’ rated spreads given
investors’ grab for yield and wide ‘A’ and ‘BBB’
spreads.
• Remain marketweight EMBIG but underweight
CEMBI
EMBIG remains the more defensive asset class with
cash flows of $64bn next year versus projected
sovereign issuance of $60bn. Within the EMBIG,underweight EMEA EM sovereigns (through
underweights in Croatia and Ukraine) as we see the
much lower growth environment due to a European
recession contributing to a higher level of sovereign
stress in this region. We are also underweight
idiosyncratic downside risks in the Middle East, where
we see continued political instability and worsening
fundamentals as underpriced; we are underweight
Egypt (through quasi-sovereign Nile Finance) and
Lebanon.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 106/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Kimberly L. Harano (1-212) 834-4956J.P. Morgan Securities LLC
106
Appendix: Results of the J.P. Morgan Fixed Income Markets Investor Survey
Yes 80%
No 20%
None 49%
$1-100bn 19%
$100-200bn 19%
>$200bn 14%
Yes 52%
No 48%
Yes 33%
No 67%
Yes 25%
No 75%
<1.5% 6%
1.5% - 2.0% 22%
2.0% - 2.5% 48%
2.5% -3.0% 20%
3.0% -3.5% 5%
>3.5% 0%
1) Do you think the Fed will expand its balance sheet via
large-scale asset purchases (QE3) in 2012?
2) How much new fiscal stimulus do you think Congress will
pass by the end of the 2011?
3) Do you think any ratings agencies will lower the US
sovereign rating by one or more notches by the end of 2012?
4) Do you think Greece will exit the EMU by the end of
2012?
5) Do you think any country other than Greece will exit the
EMU by the end of 2013?
6) Where do you expect 10-year yields to be at the end of 2Q
2012?
Reduce M aintain Add N /A
USD assets 2% 51% 18% 29%
EUR assets 16% 27% 9% 47%
Other developed markets 2% 35% 9% 54%
EM local currency assets 4% 17% 20% 59%
EM USD denominated assets 2% 21% 9% 68%
Reduce Maintain Add N/A
Duration risk 27% 38% 23% 13%
TIPS 7% 27% 24% 42%
Agency debt 21% 30% 2% 47%
Agency MBS 11% 17% 31% 41%
Non-agency MBS 7% 30% 17% 46%
CMBS 6% 24% 24% 46%
ABS 7% 28% 17% 48%
IG Corporates 9% 19% 33% 39%
High Yield 2% 17% 31% 50%
EM 4% 17% 24% 56%
Equities 7% 18% 15% 60%
Commodities 4% 13% 7% 76%Cash 22% 31% 4% 43%
I already have 11%
Yes 4%
No 37%
N/A 48%
7) How do you expect to change your portfolio currency
allocations over the next year?
8) Are you looking to reduce, maintain or add exposure to the
following asset classes over the next year?
9) In 2012, are you planning to modify your benchmark to
exclude Fed holdings of Treasuries or MBS?
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 107/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Terry BeltonAC
(1-212) 834-4650
Meera Chandan (1-212) 834-4924
Kimberly L. Harano (1-212) 834-4956
J.P. Morgan Securities LLC
107
US Treasuries
• A worsening of the European sovereign debt crisis
will accelerate the flight-to-quality bid into
Treasuries early next year; rich valuations, a
deteriorating supply/demand imbalance, and poor
technicals will limit the upside, however, and keep
Treasury rates broadly range bound in 2012
• Foreign investors will need to double their
Treasury purchases next year to clear the market;
while the bid from the EU has been strong,
demand from EM has weakened as they have
slowed their accumulation of USD reserves• Disintermediation of money funds, limited supply,
and Fed policy support lower front-end yields
• With poor liquidity, whipsaw risk in the Treasury
market will be elevated next year; we discuss the
importance of avoiding consensus trades in 2012
and highlight a measure of investor positioning
that helps identify turns in Treasury rates
• Positive carry yield curve trades in the front end
are likely to perform poorly given negative curve
convexity; we favor negative carry curve trades
anchored in the front end and positive carry
trades anchored in intermediates
• Synthetic Treasuries created by asset swapping
cheap foreign bonds should outperform in 2012
• Reconstitutions of STRIPS should be strong in
2012, causing short maturity Cs richening versus
similar maturity Ps
• The US fiscal outlook improved in 2011, although
debt levels are still on an upward trajectory. We
expect ratings agencies to maintain their existing
ratings through 2012; another $1.8tn of deficit
reduction is still needed for the US to achieve fiscal
sustainability
Fall from grace
Treasury rates fell sharply in 2011 with 10-year yields
reaching an all-time low of 1.70% in 3Q11. With great
irony, the record low occurred in the aftermath of a major
fall from grace; for the first time ever, the US lost its
‘AAA’ sovereign rating as S&P downgraded the US in
recognition of the sharp deterioration in US debt metrics
over the last few years. Despite a worsening sovereign
credit outlook, Treasury rates moved sharply lower,
with 10-year yields falling 135bp since the start of the
year to their current level of 1.96%. With front-end
yields approaching the zero-rate boundary, the rally
also caused the curve to flatten. Year to date, 2-year
yields have declined 30bp, 5-year yields have declined
110bp, and 30-year yields have declined 140bp.
Exhibit 2: The Fed’s biggest impact this year was arguably on 30-year rates; the long end flattened 50bp after the Fed began signalingOperation Twist in August10s/30s Treasury yield curve; %
0.9
1.0
1.1
1.2
1.3
1.4
1.5
Feb 11 May 11 Aug 11 Nov 11
Exhibit 1: Yields tracked US economic data in 1H11 beforedecoupling in 2H11 as the European debt crisis moved to the core
Average Italy and France CDS spreads (bp) versus 10-year Treasury yields (%)and 3-month rolling average of J.P. Morgan EASI*bp %
50
100
150
200
250
300
350
400
450 1.5
2.0
2.5
3.0
3.5
4.0
Feb 11 May 11 Aug 11 Nov 11
10-year Treasury yieldsItaly / France CDS spreads (inverted)EASI
* EASI is the J.P. Morgan Economic Activity Surprise Index and measures thenumber of positive surprises minus the number of negative surprises divided bythe total number of data releases over the past six weeks. This index has beenscaled to fit on the chart as 3.2+J.P. Morgan Economic Surprise Index/10.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 108/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Terry BeltonAC
(1-212) 834-4650
Meera Chandan (1-212) 834-4924
Kimberly L. Harano (1-212) 834-4956
J.P. Morgan Securities LLC
108
The downtrend in rates in 2011 masks two very different
dynamics driving the Treasury market this year. In 1H11,
rates were comparatively range-bound, with yields largely
driven by changing perceptions around the US economic
outlook. Increased optimism on better growth following
the payroll tax cut helped push 10-year yields to the high
of the year (3.72%) in 1Q11; these moves were reversed
in 2Q11 as economic data disappointed to the downside.
During both quarters, volatility was comparatively low,
with rates mostly trading within 30bp of their 1H11
average of 3.30%.
The dynamic in 2H11 was quite different, as the European
debt crisis spread from the smaller peripherals to the much
larger bond markets of Italy and core Europe. In response,volatility spiked, Treasury rates plunged, and correlations
with European sovereign spreads moved to extreme levels
(Exhibit 1). Despite some analysts arguing that the US
was in recession, the 2H rally actually occurred as
economic data surprised to the upside, highlighting that
the rally primarily reflected a strong flight-to-quality bid
into Treasuries, as well as unwinds of bad positions.
The Treasury market was also supported by Fed policy
this year, with stimulus coming from both $640bn of asset
purchases under QE2 and another $175bn (10-year
equivalents) of duration buying in 2H11 as part of
Operation Twist. The overall impact on yields from allthat buying was small, however; most estimates including
our own suggest that QE2 lowered intermediate rates by
15-30bp (see US Fixed Income Markets Weekly, July 15,
2011). Surprisingly, Fed policy this year likely had its
biggest impact on the long end of the curve; after
steepening nearly 50bp during the first seven months of
the year, the 10s/30s yield curve flattened all that and
more after the Fed began signaling Operation Twist in
August (Exhibit 2).
Fed policy was also modestly effective in helping reduce
term premium in the front end of the curve. Two-year
rates fell sharply after the August FOMC meeting as theFed adopted a new communication strategy and clarified
its intention to keep the funds rate near zero until mid-
2013. Since then, 1-year forward 3-month OIS has
averaged 11bp compared to a 24bp average in the 1-month
period leading up to the August FOMC meeting
(Exhibit 3). To be sure, low front-end yields are also
being supported by a collateral shortage in the Treasury
bill market, and by the European-led flight to quality into
Treasuries.
Finally, we note that the large decline in Treasury rates in
2011 occurred in an environment where long-term
inflation expectations remained relatively stable. For the
first time since 1997, 10-year Treasury yields moved
significantly below long-term inflation expectations, thus
implying negative 10-year real returns. With the exceptionof the UK, these real yields compare unfavorably to
virtually every other developed bond market in the world,
including Japan, where real rates are currently 140bp
above those in the US (Exhibit 4). Such low real rates
highlight both the strength of demand for Treasuries
during this year’s debt crisis as well as the significant
challenges Treasury investors face to earn reasonable
returns going forward.
Exhibit 4: The 2011 rally has pushed US 10-year real rates 140 bpbelow those in Japan10-year government bond yield minus 10-year inflation swap rate; %
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
2009 2010 2011
USJapan
Exhibit 3: Term premium in the front end declined sharply this year in part due to the Fed’s new guidance on policy rates2-year Treasury yields versus 1Yx3M OIS rate; %
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
Feb 11 May 11 Aug 11 Nov 11
2-year Treasury yields1yx3m OIS rate
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 109/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Terry BeltonAC
(1-212) 834-4650
Meera Chandan (1-212) 834-4924
Kimberly L. Harano (1-212) 834-4956
J.P. Morgan Securities LLC
109
The outlook for Treasury yields
Looking into 2012, our outlook is for Treasury volatilityto remain high but with rates ultimately remaining range-
bound as further deterioration in Europe competes with
rich valuations. Within this range, we look for rates to
move lower early in the year as the crisis in Europe
deteriorates further and investors add to the risk-off trade.
At the same time, we see the flight-to-quality bid
eventually fading, as rich valuations, crowded investor
positions, and still-heavy supply allow the Treasury
market to decouple from the ongoing stress in Europe. In
the discussion below, we assess the major factors likely to
drive Treasury yields in 2012, present our interest rate
forecast, and highlight the many risks around that forecast.
The case for lower Treasury yields in 2012 really rests on
only two factors, but they are important ones that, at least
early next year, are likely to cause Treasury yields to trade
at even richer levels than those reached in 2011. The first
is Europe. Our baseline view is that the sovereign debt
crisis in Europe is now set to take another turn for the
worse and is likely to fuel a renewed flight-to-quality bid
into Treasuries early next year. While the details are more
fully discussed in our Global Fixed Income Markets
Outlook publication (see Global Fixed Income Markets
2012 Outlook , November 24, 2011), we would highlight
three of the most worrisome aspects of that outlook withimplications for the Treasury market.
• We look for intra-EMU spreads to widen
significantly early next year. The widening will
likely be driven by increased selling from
unlevered investors who thought they owned
risk-free government bonds but are now
overweight high yielding risky assets. With
yields elevated, Italy’s access to private market
funding may become less secure, creating some
risk that they are forced to seek external funding
support. Italy is the world’s third largest bond
market, however, and with €250bn of funding
needs in 2012 alone (including €200bn of
redemptions), they will stretch the limits of the
EU’s and IMF’s ability to meet their funding
needs.
• Policymakers will likely continue to move too
slowly to get ahead of the deepening crisis. The
ECB will likely remain a reluctant lender of last
resort and, with heavy investor selling likely to
continue, is unlikely to buy enough to stabilize
spreads near current levels. While we do not
expect a EU breakup, statements from officials in
Germany, the Netherlands, and other creditor
countries about EU exit strategies are likely to
increase and provide ongoing support for
investors to short semi-peripheral bond markets.
• The crisis is likely to broaden further beyondItaly, especially if the Euro area recession proves
severe. In that case, risks would rise of a France
sovereign downgrade; this would lower the EFSF
funding capacity and further increase the
resources required from the creditor countries.
We note that these views represent our base case outlook;the headline risk around Europe is extraordinarily highand a permanent solution to the crisis could certainly befound just when the stress reaches its deepest point.
How low could intermediate Treasury yields go if Europe
deteriorates further? Correlations between Treasury yieldsand European sovereign spreads remain high and shouldstay that way into the early part of next year. Using our fair value model (see box), we estimate that for everyadditional widening of 100bp in 10-year Italy spreads toGermany, 10-year Treasury yields should fall by 16bp, allelse equal. If 10-year Italy yields reach 9% next year, wewould expect 10-year Treasury yields to move below therecord 1.70% reached last year. To be sure, US rates willeventually decouple from Europe as rich valuations bring
Exhibit 5: Correlations between 10-year Treasuries and European
sovereign spreads remain elevated and were especially high when10-year yields hit record lows1-month rolling average of 10-year Treasury yields (%) versus the 1-month rollingcorrelation between weekly changes in 10-year Treasury yields and semi-peripheral* Europe sovereign CDS spreads% correlation; inverted
1.9
2.0
2.12.2
2.3
2.4
2.5
2.6
2.7
2.8
2.9
3.0 -1.0
-0.8
-0.6
-0.4
-0.2
0.0
0.2
Sep 11 Oct 11 Nov 11
Treasury yieldsCorrelation; inverted
* Average 5-year CDS spreads of France, Spain and Italy.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 110/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Terry BeltonAC
(1-212) 834-4650
Meera Chandan (1-212) 834-4924
Kimberly L. Harano (1-212) 834-4956
J.P. Morgan Securities LLC
110
sellers. But for now, correlations (and betas) remain quiteelevated and were especially high when rates made their
record lows back in September (Exhibit 5).
After Europe, the second most important factor supportinglower Treasury yields in 2012 is likely to be Fedsponsorship. With growth likely to remain sluggish in1H12 as fiscal drags come into play and unemploymentremains high, we look for the Fed to announce another round of asset purchases in MBS. While the eventualimpact on Treasury yields may be modest, the initialsignal by the Fed that more QE is coming is likely to leadto an outsized rally as investors rebalance their portfolios.In each of the previous four instances where the Fedannounced large scale asset purchases, Fed signaling inadvance of the formal announcement caused 10-year
Treasury yields to fall sharply (by an average of 50bp) aslevered investors significantly increased duration longs(Exhibit 6).
Beyond Europe and the Fed, other factors supportive of lower yields in 2012 are small by comparison. Theyinclude a worsening of the front-end collateral shortagehelping push GC rates lower (see GC repo discussion below), a modest weakening in growth and inflationexpectations early next year, and further changes in the
Flight to quality and fair value of 10-year TreasuriesIn February 2011, we introduced a new short-term fair value
model for 10-year Treasury yields that combines both economicand technical factors likely to drive yields in the near term.Since then, we have modified the model somewhat byintroducing better measures of investor positioning, and addedEuropean sovereign debt spreads as a proxy for flight-to-qualityflows from Europe. The model now explains 10-year Treasuryyields as a function of five variables including 5-year forwardinflation expectations derived from the inflation swap market,the 1-year ahead J.P. Morgan forecast for real GDP growth,near-term expectations on policy rates measured by the 3-monthforward on 3-month OIS, the average 5-year CDS spreads of France, Italy, and Spain, and a measure of levered investor
positions (see box below on measuring investor positions). Themodel is estimated using weekly averages for data since June2008. All five variables are highly significant and explain about
80% of the variability in 10-year rates during this period. Thecoefficient on the flight-to-quality variable indicates that every100bp increase in semi-peripheral spreads lowers 10-year yields
by 33bp on average. Alternatively, because average semi- peripheral CDS spreads have been moving only half as much asspreads on Italian government bonds recently, we estimate 10-year yields should fall by roughly 16bp for every 100bp increasein spreads on Italian government bonds. This estimate holdsother variables constant; we expect the actual change in 10-year rates to be larger early next year as increased sovereign stresswill also be accompanied by an increase in levered investor duration longs.
10-year Treasury yield model parameters:*
Variable Current level Beta T-stat
Intercept - 1.983 14.2
Inflation expectations; % 2.72 0.328 7.3
GDP growth forecast; % 1.75 0.184 24.9
3Mx3M OIS; % 0.10 0.476 24.9
Sovereign CDS; % 4.01 -0.326 -26.5
Levered investor positions -0.64 -0.296 -19.1
*Estimated from 6/1/08-11/10/11.N=864, SER=0.229, R2=0.824
10-year Treasury yields versus model fair value; %
1.5
2.0
2.5
3.0
3.5
4.0
2009 2010 2011
ActualModel
Exhibit 6: Buy the rumor, sell the fact; 10-year Treasuries haverallied an average of 50bp leading into the last four Fed QE
announcements10-year Treasury yields (%) and J.P. Morgan index of levered investor bondpositions*% positions index; inverted axis
2.3
2.4
2.5
2.6
2.7
2.8
2.9
3.0 -0.6
-0.5
-0.4
-0.3
-0.2
-0.1
0.0
0.1
0.2
0.3
0.4
-30 -20 -10 0 10 20 30Business days around quantitative easing
10-year Treasury yields
Positions index; inverted
* Levered investor positions index is the weighted average of aggregate net longsheld by non-commercial investors as provided by CFTC and the partial beta of hedge fund returns versus 10-year Treasury yields, converted to a z-score. QEannouncement dates are 11/25/08, 3/18/09, 9/21/10, and 9/21/11.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 111/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Terry BeltonAC
(1-212) 834-4650
Meera Chandan (1-212) 834-4924
Kimberly L. Harano (1-212) 834-4956
J.P. Morgan Securities LLC
111
Fed’s policy rate guidance that should help lower term premium in the yield curve next year.
Against these bullish factors for the Treasury market,there are also some important offsets that we believe willultimately push Treasury yields higher in 2012. The firstis valuations. Ignoring the impact of the (presumably
temporary) flight-to-quality bid into Treasuries, our fair value model for 10-year yields suggests Treasuries are80bp expensive (Exhibit 7). While this richness can persist when Europe remains under stress, the magnitudeof the mispricing highlights the significant cheapeningthat is likely to occur in the Treasury market as soon as thesituation in Europe stabilizes.
Second, we view the supply/demand outlook in theTreasury market as unfavorable next year, with durationsupply rising and Fed buying likely to be significantlyweaker than in 2011. As discussed in more detail below,we expect overall duration supply (net of Fed purchases)across US fixed income markets to increase 11% nextyear, assuming a moderately sized QE3 program in MBS.Treasury supply net of Fed purchases is projected to morethan double to $1tn, requiring a significant increase in buying from foreign investors. This increased supply isoccurring at the same time that the largest foreign holdersof Treasuries have slowed their accumulation of USDreserves with Fed custody holdings of Treasuriesessentially flat since June (Exhibit 8).
Third, with Treasury rates very rich and most investorsalready extremely pessimistic about Europe, we would
expect Treasury yields to begin decoupling from Europeanheadlines at some point next year. In part, this reflectsinvestor positioning: correlations tend to rise early on inthe crisis as the news forces investors to rebalance but fallonce those rebalancing trades have occurred. Indeed, prior to the broadening of the crisis in August to include Italyand France, US rates had already begun decoupling from peripheral spreads; correlations between 10-year Treasuryyields and the average spreads of Greece, Portugal andSpain had fallen to zero this summer, even though
Exhibit 9: Correlations between US rates and spreads inGreece/Portugal/Spain had fallen to zero this summer despitecontinued spread widening
Average Greece, Portugal and Spain CDS spreads versus the 3-month rollingcorrelation between 1-month changes in 10-year rates and average Greece,Portugal and Spain CDS spreadsbp correlation; inverted axis
400
500
600
700
800
900
1000
1100
1200 -0.6
-0.4
-0.2
0.0
0.2
0.4
0.6
Feb 11 Apr 11 Jun 11 Aug 11
Average CDS spreadsCorrelation; inverted
Exhibit 8: The largest foreign holders of Treasuries slowed their accumulation of USD reserves this year with Fed custody holdingsof Treasuries flat since JuneTreasuries held at custody at the Fed excluding repo versus foreign exchangereserves of the largest Treasury holders*;$bn $bn
5100
5200
5300
5400
5500
5600
5700
2560
2580
2600
2620
2640
2660
2680
Feb 11 Apr 11 Jun 11 Aug 11 Oct 11
Treasuries held incustody at the Fed
FX reserves
* Includes FX reserves of South Korea, Brazil, Japan, China, Russia and Taiwan.Dollar is held constant at March 2009 levels.
Exhibit 7: 10-year Treasury rates are rich by 80bp10-year fair value* excluding the impact of the flight to quality bid into Treasuriesversus actual 10-year Treasury yields; %
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
2009 2010 2011
Model Actual
* 10-year Treasury model yield = 0.34 + 0.66 * 5yx5y inflation swap rates (%) +0.85 * 3mx3m OIS rate (%) + 0.2 * 1-year ahead J.P. Morgan real GDP growthforecast (%) – 0.33 * J.P. Morgan investor positions index in bonds; Estimatedover 6/1/08-11/11/11 period.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 112/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Terry BeltonAC
(1-212) 834-4650
Meera Chandan (1-212) 834-4924
Kimberly L. Harano (1-212) 834-4956
J.P. Morgan Securities LLC
112
peripheral spreads continued to move significantly wider (Exhibit 9). To be sure, rebalancing flows out of Italy are
far from done, suggesting the Treasury rally is not over.But at some point next year, we expect the short risk tradeto become crowded again limiting the upside in Treasurieseven if Italy continues to widen.
The offsetting factors of further deterioration in Europeand already rich valuations in the Treasury market willmake it difficult for Treasuries to trend for an extended period in either direction next year. Along with our outlook for near-trend growth and for policy rates to stayanchored near zero at least until 2014, we expect range- bound Treasury rates in 2012. Within that range, we look for rates to move lower early next year as Europe
deteriorates and investors reset short risk positions. Our 1Q12 target for 10-year rates is 1.70% (Exhibit 10),which we view as consistent with another 100-200bpincrease in Italian 10-year yields combined with anincrease in duration buying from levered investors earlynext year as they reset the risk-off trade. Our year-endtarget for 10-year yields is 2.5%. This assumes only a partial retracement back to fair value as we expect Europeto remain unsettled, resulting in some lingering flight toquality-related demand.
At the front end of the curve, our forecast tracks theoutlook for 10-year rates (adjusted for the lower volatility
of short rates) but incorporates some additional richeningdriven by an anticipated decline in GC repo and OIS (see“The front end and the outlook for GC” below). Weexpect the Fed to further strengthen its forward guidanceon policy rates early in 2012 helping further flatten theOIS curve. In addition, front-end Treasuries are cheap toOIS and we expect 3-5bp of richening in 2-year yieldseven if OIS remains unchanged.
We are biased to a modestly steeper 10s/30s yield curve in2012 targeting 110bp by mid-year. Following theannouncement of Operation Twist in September, the longend of the curve had significantly flattened relative to fair value before retracing much of the mispricing over the past month. Our 10s/30s yield curve model, which is based on the level of front-end rates, inflationexpectations, variable annuity (VA) hedging needs, andlong-end supply (see U.S Fixed Income Markets Weekly,September 23, 2011) currently shows the curve as 7 bp tooflat (Exhibit 11). While we could see some modestflattening in 1Q12 as stresses in Europe increase long-endVA-related duration buying, increased prospects that theFed will expand its balance sheet as we move through theyear provides an offset. On balance, with Treasury rateslargely range bound into mid-year, we look for a modestly
steeper 10s/30s curve and target 30-year yields to reach3.60 by mid-year.
Finally, we would note that this is a year where volatilityis likely to be extraordinarily high even if rates ultimately prove range-bound. With poor liquidity and balance sheetsof many financial institutions becoming more constrainednext year, the market is likely to continue to experiencefrequent whipsaws in Treasury rates driven as much bytechnicals and investor rebalancing flows as byfundamentals. In the next section, we discuss why
technicals are likely to be so important next year and makethe case that contrarian duration trading strategies arelikely to be especially profitable in 2012. We also discussimproved metrics we have developed this year for trackinginvestor technicals that have proven to be a good leadingindicator of turns in the Treasury market.
Exhibit 11: After a significant overshoot following Operation Twist,the 10s/30s curve has moved back to fair value
Actual versus model* for the 10s/30s Treasury yield curve; %
0.7
0.8
0.9
1.0
1.1
1.2
1.3
1.4
1.5
1.6
May 10 Nov 10 May 11 Nov 11
ActualModel
* 10s/30s curve modeled as 1.147 - 0.003292 * index of variable annuity hedging($bn 20-year equivalents) + 0.1169 * 5yx5y inflation swap rates (%) – 0.2446 * 3-
year yields (%) + 0.001028 * Amount of Treasuries outstanding in the 10- to 30-year sector net of Fed purchases via Operation Twist ($bn)
Exhibit 10: J.P. Morgan interest rate forecast
% Current 4Q11 1Q12 2Q12 4Q12
21 Nov 11 31 Dec 11 31 Mar 12 30 Jun 12 31 Dec 12
2-yr Treasury 0.28 0.22 0.17 0.30 0.30
5-yr Treasury 0.91 0.90 0.75 1.25 1.25
10-yr Treasury 1.96 2.00 1.70 2.50 2.50
30-yr Treasury 2.94 3.00 2.70 3.60 3.60
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 113/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Terry BeltonAC
(1-212) 834-4650
Meera Chandan (1-212) 834-4924
Kimberly L. Harano (1-212) 834-4956
J.P. Morgan Securities LLC
113
Avoiding crowds in 2012
While our baseline forecast is for yields to move lower early in 2012 as contagion spreads in Europe, markets are
very unlikely to move in a straight line; we expect
volatility to remain extremely high with frequent turns in
US interest rates during the course of the year driven by
poor liquidity and technicals around investor positioning.
While technicals have always been an important factor
driving yields, a combination of cyclical and structural
factors have amplified their impact since the end of the
2008 financial crisis, causing intermediate Treasury rates
to be extraordinarily volatile despite stable policy rates
(Exhibit 12). These factors include both declining market
depth and a growing disparity between the size of dealer
and end-user balance sheets.
The importance of technicals on Treasury rates in the last
couple of years is highlighted in Exhibits 13 and 14
which compare 10-year Treasury yields to a measure of
investor duration positioning we have developed (see grey
box). The exhibits highlight four major turns (whipsaws)
in the Treasury market since 2Q10 that were triggered or
at least amplified by investor unwinds of crowded
consensus trades. This includes: 1) the Treasury rally in
2Q10 following a period where improving growth and Fed
exit strategy discussions resulted in significant duration
shorts by levered investors; 2) the sell-off in Treasuries
following the start of QE2 in October 2010; 3) the 3Q11rally in Treasuries following the end of QE2; and, 4) the
4Q11 sell-off in Treasuries following the EU driven flight-
to-quality bid in Treasuries. While the initial catalyst for
the unwinds differed, the whipsaw in Treasury rates
occurred in each case just after our investor position
measure reached a fairly extreme level with unwinds
amplifying the market move. Over the four episodes, the
change in 10-year Treasury yields (with Treasury yields
moving in the opposite direction of the consensus view in
all four cases) averaged 70bp. The largest move (97bp
sell-off) followed the start of QE2 when our position
measure indicated levered investors held their largest long
duration position since 2007.
Exhibit 13: The importance of avoiding crowdsConsensus View
Duration Rationale
Short Growth/Fed exit 5/18/10 10/21/10 5.0 -85
Long QE2 coming 10/21/10 4/12/11 5.6 97
Short QE2 ending 4/12/11 6/23/11 2.3 -59
Long EU crisis 9/23/11 10/18/11 0.8 34
Average move (abs value) 3.4 69
10y yield
change; bp
Time
(months)
Positions
reversed by
Peak
positons by
Exhibit 12: Going nowhere fast: a structural decline in liquidity has
caused Treasury volatility to spike despite stable policy ratesRolling 1-year standard deviation of daily changes in 10-year Treasury yieldsdivided by the rolling 1-year standard deviation of daily changes in the 3Mx3M OISrate
0
2
4
6
8
10
12
1995 1999 2003 2007 2011
Exhibit 14: Crowded trades have led to four major rate whipsawssince 2Q1010-year Treasury yields (%) versus the J.P. Morgan levered investor position indexin bonds*
1.5
2.0
2.5
3.0
3.5
4.0 -1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
May 10 Nov 10 May 11 Nov 11
10-year Treasury yields; invertedPositions index
* Positions index is the weighted average of aggregate net longs held by non-commercial investors as provided by CFTC and the partial beta of hedge fundreturns versus 10-year Treasury yields, converted to a z-score.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 114/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Terry BeltonAC
(1-212) 834-4650
Meera Chandan (1-212) 834-4924
Kimberly L. Harano (1-212) 834-4956
J.P. Morgan Securities LLC
114
Measuring investor positionsOne of the more active areas of research for us in 2011 has been developing alternative measures of investor positioning that better explain
Treasury yield movements. While survey data (e.g., the J.P. Morgan Treasury Client Survey) are often used as a barometer of investor positioning, we have found it to be less useful for explaining and predicting Treasury yields than actual position measures available fromfutures markets. In addition, position measures calculated by reverse engineering return data (i.e., return betas) appear more reliable to us thansurvey measures.The table below provides a comparison of alternative measures of investor positioning based on how well they explain yield movementswhen added to our short-term fair value model for 10-year Treasury yields. The investor position measures we compare include (durationweighted) CFTC data on net speculative positions in interest rate futures, survey data from our weekly duration survey, and return betascalculated using daily returns published by the largest core bond funds as well as hedge funds. The fair value model (see grey box above),which is estimated using weekly data since June 2008, explains 10-year Treasury yields as a function of inflation expectations, the J.P.Morgan 1-year ahead growth forecast, the level of 3-month forward OIS, a flight-to-quality measure proxied by European semi-peripheralspreads, as well as a measure of investor positions. While most position indicators are statistically significant and improve the overall fit of the model, we find levered investor position measures explain Treasury yield movements much better than unlevered investor
positions. The best-fitting models are the ones that use CFTC data on speculative positions followed by a position measure based on macrohedge fund return betas. These models have the lowest regression standard error, highest t-statistic on the position variable, and highest R-squared (see Table). By contrast, the J.P. Morgan duration survey measure and the return betas for core bond fund managers have the lowest
explanatory power with the bond fund return beta statistically insignificant.We also find that the 10-year model can be improved by combining some of these position measures. In particular, we have created a leveredinvestor index that converts the CFTC and hedge fund return betas to a z-score, and takes a weighted average based on regression weights.This index has the best fit in our 10-year yield model when compared to other position measures and has been stable through time; the R-squared increases to 0.82, and the standard error of the regression falls to 0.23. This index is currently at -0.5 suggesting levered investors arenow modestly short duration.Beyond helping understand the factors driving yield changes, we also find position indicators to be useful contrarian indicators that help usforecast future yield changes. Reflecting the fact that levered investor positions tend to be mean-reverting, we find our position index to be auseful leading indicator of yield changes. This is highlighted in the last exhibit below which compares monthly changes in 10-year yields tothe ex ante level of our J.P. Morgan positions index.10-year Treasury yield model summary statistics*Model includes: Current value Coefficient T-stat RSQ SER
CFTC spec longs -1.1 -0.175 -11.8 78.3 0.25
Hedge fund beta -0.4 -0.090 -11.3 78.0 0.26
JPM duration survey 0.6 -0.059 -6.2 75.9 0.27
Mutual fund beta 0.3 -0.006 -0.5 74.8 0.27
CFTC/HF Index -0.8 -0.295 -21.5 82.3 0.23 *All models estimated from 6/1/08-11/8/11. In addition to a measure of investor positions, the model includes 5Yx5Y forward inflation expectations, the JPM 1-year aheadgrowth forecast, the level of 3-month forward OIS, and an average of France/Italy/Spain 5-year CDS.
JPM index of levered investor positions:* 1M change in 10Y UST yields vs. one-month ago JPM index of leveredpositions; 5/10-11/11
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
May 10 Nov 10 May 11 Nov 11
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
-1.0
-0.8
-0.6
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
Nov 10 May 11 Nov 11
1m ago Positions Index1m change in yields; %
* Defined as 0.66 * CFTC net spec longs z score +.33 * hedge fund return beta z-score. Hedge fund betas are calculated using daily returns on the IQ Global Macro Beta Index.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 115/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Terry BeltonAC
(1-212) 834-4650
Meera Chandan (1-212) 834-4924
Kimberly L. Harano (1-212) 834-4956
J.P. Morgan Securities LLC
115
While markets often zigzag, the magnitude of the recent
whipsaws appear unusually large, given that the economic
outlook has been comparatively stable (slow but positive
growth) and the Fed is expected to keep rates near zero for
a prolonged period of time. In our view, the volatilityreflects a secular decline in market liquidity as a declining
pool of dealers face increasingly larger capital
requirements and stricter leverage ratios. Some evidence
of this can be seen in the decline of dealer reported VAR
levels since 2010 and the corresponding decline in market
depth (Exhibit 15). We would note that this reduced
ability of dealers to provide liquidity is occurring at the
same time that the balance sheets of the largest asset
managers are growing, creating an imbalance in the
supply and demand for liquidity (Exhibit 16).
Because poor market depth is unlikely to improve next
year as US and especially European banks de-lever further, technicals are likely to again be a dominant driver
of Treasury yields with investor rebalancing flows having
an outsized impact on yields. Thus, despite our 1Q12
outlook for lower yields driven by renewed stress in
Europe, our bias in 2012 will be to trade duration
tactically with an emphasis on fading crowded trades
and positioning for mean reversion. This strategy
proved especially profitable in 2011; a simple rule of
going long duration each day our levered investor
positions index was below -0.75 and short duration each
day our index was above 0.75 had a success rate of over
95% (based on 71 trades) and average P/L for 1-month
holding periods of 25bp (Exhibit 17). With the index
currently at -0.5, we are currently biased to holding long
duration trades expecting investors to build long duration
positions as the risk-off trade gains momentum early in
2012.
The front end and the outlook for GCThe past year has been characterized by increased demand
for front-end Treasuries from money market funds. As the
European crisis escalated and risk aversion increased, the
assets under management (AUM) in prime money market
funds fell dramatically and investors instead favored the
relative safety of government money market funds. Since
the end of May, the AUM in prime money market funds
have fallen $200bn, while the AUM in government money
market funds have increased by $130bn. As a result,
Treasury holdings of money market funds (including repo)
Exhibit 15: Market depth has declined as dealers continue to lower VAR and shrink their balance sheets…Reported average quarterly VAR for the 9 largest investment banks* versusmarket depth**; $mn$mn $mn
90
95
100
105
110
115
120
125
80
100
120
140
160
180
200
220
240
May 10 Nov 10 May 11 Nov 11
VARMarket depth
* Average VAR reported by JPM, GS, MS, BAC, C, UBS, CS, Soc Gen and DB.3Q11 VAR is an estimate, projected from the results reported to date (whichexcludes C).** Market depth is calculated as the 3-month moving average of average size of the top three bids and offers, in $mn, for the on-the-run 10-year Treasury note,averaged between 8:30 a.m. and 10:30 a.m. daily.
Exhibit 16: …while end-user demand for liquidity has increased AUM for the largest asset managers*; $tn
14.5
15.0
15.5
16.0
16.5
17.0
17.5
18.0
1Q10 2Q10 3Q10 4Q10 1Q11 2Q11 3Q11 *Includes BlackRock, Allianz Group, State Street Global, Deutsche Bank,Vanguard Group, Fidelity Investments, J.P. Morgan Chase, BNP Paribas, AXAGroup and BNY Mellon.Source: Bloomberg, Company websites, P&I/Towers Watson Global 2010 ranking
Exhibit 17: A simple contrarian trading strategy based on our indexof investor positioning was highly profitable in 2011Number of trades, success rate and average P/L from back-testing a trading rule*based on the J.P. Morgan positions index
TradeNumber of
tradesSuccess rate (%) Average P/L (bp)
Long 10s 39 95 22.9
Short 10s 32 100 28.3
Total 71 97 25.3 * The rule involved going long duration each day our levered investor positionsindex was below -0.75 and short duration each day our index was above 0.75.Positions were held for a one month period.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 116/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Terry BeltonAC
(1-212) 834-4650
Meera Chandan (1-212) 834-4924
Kimberly L. Harano (1-212) 834-4956
J.P. Morgan Securities LLC
116
have increased dramatically (Exhibit 18). This increased
demand has come in a period of falling T-bill supply. The
combination of these two factors proved to be very
supportive of valuations: T-bills across the curve richened
to historical levels, with yields at or near-zero levels.
Supportive technicals notwithstanding, GC repo rates have
stayed stubbornly high, with 1-month term GC currently
at around 11bp.
Looking ahead, we expect GC repo rates to head lower
targeting 1-week GC to average 5bp in 1H12 for the
following reasons. First, supply technicals in the front-end
are very supportive. Our short-term fixed income
strategists expect issuance in most front-end markets to be
negative in 2012, with Non-Financial CP being a notable
exception (Exhibit 19). Moreover, net T-bill issuance for
FY2012 will be close to zero or negative, depending on
the magnitude of fiscal stimulus that is passed. We expect
the budget deficit to total $1.1tn in FY2012, assuming that
a $100bn stimulus is passed. With coupon net issuance
expected to be $1.09tn, the net issuance of T-bills will
total a mere $10bn over that period. However, if nostimulus is approved, then T-bill issuance will total
-$90bn. Second, demand for T-bills in particular is likely
to stay elevated despite shrinking supply. Investors will
likely stay risk-averse since headline risk from Europe
will stay elevated in 1H12, and demand for government
paper will persist as assets continue to migrate out of
prime and into government money market funds.
Moreover, unattractive net yields in prime funds (a mere
3bp) relative to government funds (1bp) will also be
supportive of this trend. Finally, possible regulatory
changes in money market fund capital requirements and
the constant NAV model in late 2012 could further
accelerate this trend (see Short-term Fixed Income).
The combination of these factors is likely to put
downward pressure on GC repo rates as we head through
the year. Historically, the outstanding balance of T-bills
net of money market holdings of Treasuries has been a
significant driver of repo rates, with every $100bn of
additional supply biasing 1-week GC repo rates 4bp
higher (Exhibit 20). Given our expectation of close-to-
zero net bill issuance in 1H12, combined with our
expectation that money market funds will purchase $20bn
of Treasuries per quarter, we expect that 1-week GC repo
Exhibit 19: We expect net issuance in the front end to be negative
in 2012Projected 2012 net issuance in short-term fixed income markets by sector; $bn
-200
-150
-100
-50
0
50
Non-fin CP T-bills ABCP Agencydiscos
FinancialCP
YankeeCDs
Exhibit 20: We look for 1-week GC repo rates to average near 5bpin 1H given our outlook for supply and money market funddemand…1-week Treasury GC repo rate (bp) modelled as a function of outstanding T-bills
net of money market fund holdings of Treasuries ($bn)
0
5
10
15
20
25
950 1000 1050 1100 1150 1200 1250 1300 1350 1400
Effective T-bills outstanding; $bn
Y = 0.0421 X1 - 36.6723R² = 65.92%standard error = 3.9300period = Nov 19,09 - Nov 19,11
Exhibit 18: The perfect storm: high money market demand in a
period of falling supply resulted in lower bill yieldsTreasury bills outstanding ($bn) versus Treasuries held by money market funds($bn)
1450
1500
1550
1600
1650
1700
1750
1800
430
440
450
460
470
480
490
500
510
520
Feb 11 May 11 Aug 11 Nov 11
T-bills outstandingTreasuries held by money market funds
Source: Treasury, J.P. Morgan, iMoneyNet
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 117/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Terry BeltonAC
(1-212) 834-4650
Meera Chandan (1-212) 834-4924
Kimberly L. Harano (1-212) 834-4956
J.P. Morgan Securities LLC
117
rates will eventually average around 5bp in 1H12, or 6bp
lower than current levels.
This environment will be broadly supportive of lower
2-year yields as well, and will push them closer to 17bp
in 1Q. In addition, to lower financing rates, our positive
outlook on 2-year Treasuries is driven by the following
factors.
First, the expansion of the Fed’s balance sheet under QE3
will increase the supply of reserves and help lower the
effective funds rate. Historically, every $100bn of
additional reserves in the system has lowered 3-month
OIS rates by 1.5bp, as shown in Exhibit 21. Thus, if the
Fed announces quantitative easing of $500bn, it couldsignificantly lower front-end OIS rates, bringing them
close to zero.
Second, changes in the Fed’s communication policy will
be supportive as well, since it will further reduce term
premium from the front end of the curve. We expect the
Fed to modify its communications on policy rates during
1Q12 by linking future interest rate hikes to specific
economic conditions along the lines suggested by Chicago
Fed President Evans. This should lengthen the market’s
expectation of the Fed-on-hold horizon beyond mid-2013.
Assuming that this outcome causes forward OIS rates to
narrow back to levels reached when the Fed firstintroduced the mid-2013 language on August 9 suggests
1Yx1Y OIS rates could fall by almost 10bp (Exhibit 22).
We estimate that this would lower 2-year Treasury yields
by another 8bp.
Finally, we expect 2-year Treasuries to richen relative to
OIS as we approach mid-year when the Fed is scheduled
to complete Operation Twist. Fed selling of the front end
has helped Treasuries cheapen relative to OIS with the
spread between 2-year Treasury yields and 2-year OIS
currently at 13bp, or 6bp wider than its 1-year average.
In sum, given our outlook for GC, term premium, and anexpansion of the Fed’s balance sheet, we expect 2-year
Treasuries to richen on an outright basis and relative to
OIS. We look for 2-year rates to trade back to 17bp, or
around 10bp richer than current levels.
Finally, we also note that changes to the Fed’s
communication policy are also likely to remove term
premium even further out the curve, such as in the 3-year
sector. As shown in Exhibit 22, at close to 35bp, the
1Yx1Y/2Yx1Y OIS curve is quite steep and has the
potential to flatten significantly further if the market starts
to price in a longer period of Fed-on-hold.
Exhibit 21: The expansion of the Fed’s balance sheet under QE3will increase the supply of reserves and help lower the effective
funds rate1-month rolling average of 3-month OIS rate (bp) regressed against US bankreserves ($bn)
6
8
10
12
14
16
18
20
22
24
1000 1100 1200 1300 1400 1500 1600 1700
Reserves; $bn
Y = -0.0145 X1 + 33.8150R² = 78.67%standard error = 1.9268period = Nov 21,09 - Nov 21,11
Exhibit 22: OIS forwards in the front end of the curve havesubstantial room to fall if the Fed adopts the Evans planMinimum 1-year spot, 1Yx1Y and 2Yx1Y OIS rates since the FOMC meeting on
August 9 versus current levels; %
0.12
0.19
0.55
0.06 0.08
0.46
0.0
0.1
0.2
0.3
0.4
0.5
0.6
1y spot 1yx1y 2yx1y
Current
Min since 8/9
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 118/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Terry BeltonAC
(1-212) 834-4650
Meera Chandan (1-212) 834-4924
Kimberly L. Harano (1-212) 834-4956
J.P. Morgan Securities LLC
118
Supply Outlook
Supply technicals improved during 2011 as gross duration
supply across fixed-income markets fell 10% and duration
supply net of Fed purchases fell 24% (Exhibit 23). The
decline in gross issuance was fairly broad-based and came
primarily from spread products led by Municipals (down
29%), Agency debt (down 28%), MBS (down 14%), and
high-grade credit (down 9%).
Issuance within the Treasury market also fell in 2011
driven by a combination of lower deficits and the
suspension of SFP bills issuance. For calendar year (CY)
2011, we project net Treasury issuance to equal $1.09tn
versus $1.59tn in CY2010 (Exhibit 24). Notably, even
though nominal coupon sizes were left unchanged during
the year, net issuance of nominal coupon Treasuries fell
by $318bn as redemptions were higher (Exhibit 24). 2011
also became the third consecutive year of negative bill
issuance, causing the share of T-bills in the Treasurymarket to fall further to 15.2% currently from 19.6% at the
beginning of the year. As a result, the weighted average
maturity of outstanding debt increased to 63 months, its
longest since August 2002, and up from 55 months at the
end of 2009 (Exhibit 25).
Looking ahead, we expect overall duration supply in
fixed-income markets to increase 8% in 2012 approaching
the record levels set in 2009-2010 (Exhibit 23). This
increase is primarily driven by higher supply in the MBS
and Municipal markets with mortgage refinancing
Exhibit 24: Look for net issuance of coupon Treasuries to fall
further in 2012 Net issuance in Treasuries by sector and by calendar year; $bn
Sector CY 2010 CY 2011* CY 2012*
Bills -19 -274 -8
Nominal coupons 1563 1245 936
TIPS 48 120 100
Total 1592 1090 1028
Total ex-SFP bills 1397 1290 1028
Nominal coupons+TIPS 1611 1364 1036 * 2011 and 2012 are J.P. Morgan projections.
Exhibit 25: The share of T-bills in the Treasury market fell againthis year, further lengthening the weighted average maturity of theTreasury market Weighted average maturity of outstanding Treasuries (months) versus the shareof T-bills in the Treasury market (%); dashed lines are projections*Months %
10%
15%
20%
25%
30%
35%
45
50
55
60
65
70
75
Jan 00 Aug 02 Feb 05 Aug 07 Feb 10 Sep 12
Weighted average
maturity
Share of T-bills
* Assumes unchanged coupon sizes and FY budget deficit of $1.1tn
Exhibit 26: We look for the budget deficit to fall 15% in FY2012Budget deficit by fiscal year; $bn
-237-127
158
377 413
319 248162
455
141612941298
1100
-500
0
500
1000
1500
2000 2003 2006 2009 2012*
* 2012 is J.P. Morgan forecast.
Exhibit 23: Gross duration supply fell by 10% in 2011, but is poised
to increase by 8% in 2012 Annual duration supply via MBS, Municipal, investment grade corporates,Treasuries and Agency debt markets, gross and net of Fed purchases; $bn 10-year equivalents
1500
2000
2500
3000
3500
2007 2008 2009 2010 2011 2012
Gross duration supply
Supply ex-Fed
* 2011 and 2012 are J.P. Morgan forecasts. We assume that the Fed purchasesa total of $500bn MBS 2012 via QE3.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 119/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Terry BeltonAC
(1-212) 834-4650
Meera Chandan (1-212) 834-4924
Kimberly L. Harano (1-212) 834-4956
J.P. Morgan Securities LLC
119
projected to increase and municipal finances staying weak.
Within the Treasury market, the outlook is more
favorable, with the deficit expected to be on a declining
trajectory. Our forecast assumes a FY2012 budget deficit
of $1.1tn, lower than the $1.3tn in FY 2011 (Exhibit 26).
Thus, net issuance will be lower, too, totaling $1.04tn
(Exhibit 24).
The composition of issuance will likely be similar to thatin recent years. We look for nominal coupon sizes to stay
unchanged in 2012, but look for TIPS issuance to increase
modestly (Exhibit 27). Given our deficit projection, this
implies that Treasury bill net issuance will be close-to-
zero in 2012, with the share of bills in the Treasury market
falling modestly to 14.6% (Exhibit 25). This projection
assumes Congress passes $100 bn of fiscal stimulus; if no
stimulus is passed, net issuance of bills is projected to fall
to -$90bn, bringing the share of bills to 13.7% of the
market by the end of FY2012. In this case, there is some
possibility that Treasury will cut coupon sizes later in the
year in order to stabilize the size of the bill market.
Demand Outlook
At the same time that duration supply is set to increase
across fixed income markets, demand technicals are likely
to be more challenging for the Treasury market in 2012.
The primary reason is a decline in Fed sponsorship.
While the Fed purchased 60% of net Treasury issuance in
2011, we expect limited net buying of Treasuries by the
Fed next year with any increase in its balance sheet likely
to be targeted in MBS. As a result, private investors and
foreign central banks will need to more than double (from
$454bn in 2011 to $1,000bn in 2012) the amount of
Treasuries they buy in order for the Treasury market to
clear in 2012.
After the Fed, the single largest buyer of Treasuries in
2011 was foreign investors. Net purchases in 2011 are on
track to reach $290bn with most of the buying occurring
in 3Q11 after the end of QE2. While this is significantly
lower than the magnitude of purchases in the last three
years of $640bn per year, it is higher as a percentage of
effective supply. Foreigners took down 64% of effective
Exhibit 27: Nominal coupon sizes should be stable in 2012 while
TIPS issuance increases modestly J.P. Morgan gross issuance forecast; $bn
2s 3s 5s 7s 10s 30s
5y
TIPS
10y
TIPS
30y
TIPS Subtota l
Jan 12 35 32 35 29 21 13 14
Feb 12 35 32 35 29 24 16 10
Mar 12 35 32 35 29 21 13 12 537
Apr 12 35 32 35 29 21 13 16
May 12 35 32 35 29 24 16 12
Jun 12 35 32 35 29 21 13 8 537
Jul 12 35 32 35 29 21 13 14
Aug 12 35 32 35 29 24 16 14
Sep 12 35 32 35 29 21 13 12 541
Oct 12 35 32 35 29 21 13 8
Nov 12 35 32 35 29 24 16 12Dec 12 35 32 35 29 21 13 14 535
Total 420 384 420 348 264 168 44 76 26 2150
Chg from 2011 0 0 0 0 0 0 6 6 3 15
Exhibit 28: While the magnitude of buying by foreign investors was
lower in 2011 than that in recent years, they took down a larger portion of effective supplyForeign purchases of Treasuries ($bn) versus foreign purchases as a percentageof Treasury net issuance excluding Fed purchases (%)% $bn
0
200
400
600
800
20%
40%
60%
80%
100%
120%
140%
2001 2003 2005 2007 2009 2011*
%ge of netissuancebought byforeigners
Amount of foreignTreasury purchases
Source: Federal Reserve Z.1, TICNote: 2011 is an estimate of the total foreign purchases reported by the FederalReserve in Flow of Funds data. 1H11 uses actual Flow of Funds data, Q3 modelsthe Flow of funds purchases as -3.7 + 0.975 * Monthly net purchases reported byTreasury in the TIC report, while Q4 is estimated by a combination of: (a) Flow of funds modeled as 3.8 + 1.34 * Monthly change in Treasury custody holdings and(b) the YTD monthly average purchases of Treasuries by foreigners.
Exhibit 29: A flight-to-quality bid: EU countries were net buyers of Treasuries in 3Q11 as the crisis in Europe escalated, more thanoffsetting selling by EM countries
Average monthly purchases of coupon Treasuries by foreigners in August andSeptember according to TIC data versus its prior 6-month average; $bn
Region Aug-Sep av erage 6m av erage
EU 44.4 10.9
UK 33.7 15.1
France 10.9 -1.6
Japan 21.8 4.2
China -9.8 4.4
EM ex China* -9.4 2.0
* Includes Brazil, Russia, African countries, Taiwan, South Korea and Mexico.Countries in grey are also included in the EU total.Source: TIC
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 120/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Terry BeltonAC
(1-212) 834-4650
Meera Chandan (1-212) 834-4924
Kimberly L. Harano (1-212) 834-4956
J.P. Morgan Securities LLC
120
supply (i.e., net issuance excluding Fed purchases), higher
than the 3-year average of 51% (Exhibit 28). The
composition of demand was favorable and foreigners
heavily favored coupon securities to bills.
While overall foreign demand for Treasuries has been
strong since the end of QE2, Treasury data for the third
quarter shows a significant shift in the composition of foreign buying. Demand from EM reserve managers
weakened during the quarter as they slowed their building
of USD FX reserves, while demand from the EU increased
in response to the sovereign debt crisis (Exhibit 29).
While we view the recent slowing in EM demand as
mostly a cyclical phenomenon, it is also likely a sneak
preview of a longer-term trend underway for many EM
countries to gradually slow their reserve accumulation.
Officials in China have become increasingly vocal about
the need to slow FX reserve building and diversify away
from the US dollar (see US Fixed Income Markets Weekly,
September 19, 2011). At the same time, this
diversification has been slow-moving and is likely to
remain so (Exhibit 30). On balance, we view weaker
sponsorship by EM reserve managers as a risk to the
Treasury market rather than our base case. In the
meantime, softer demand from the EM is likely to be
offset by the flight-to-quality bid from the EU.
Exhibit 31 shows our 2012 projections for Treasury
demand by investor type. The estimates are based on a set
of simple empirical models discussed in more detail below
(see box and Exhibit 32). For foreign investors, we
project total purchases of $540bn or 53% of net Treasury
issuance; by comparison, foreign purchases amounted to
64% of net issuance (net of Fed purchases) in 2011 and
48% in 2009-2010 (Exhibit 31). With the curve flat and
rates low, we expect demand from pension funds and
insurance companies to be modestly below average at
around $150bn (Exhibit 32). We also expect commercial
Exhibit 32: Models for estimating Treasury buying by pensionfunds, insurance companies and commercial banksEmpirical models for quarterly purchases by pension funds/ insurance companiesand commercial banks; $bn
Pension funds/ insurance companies
Variable Coefficient T-stat Projection
Intercept 35.9 1.7 1
30-year Treasury yields; % -7.8 -2.1 3.25
10s/30s curve; % 18.5 2.7 1
Quarterly effective net issuance of Treasuries; $bn 0.04 2.6 260
Projected purchases in 2012: 150
Quarterly data regressed over 15-years; R2=54%; Standard error = 62
Commercial banks
Variable Coefficient T-stat Projection
Intercept -5.4 -1.23
30-year MBS CC Libor OAS; bp -0.6 -4.7 43
2s/5s Treasury curve; % 7.0 1.4 0.8
Quarterly effective net issuance of
Treasuries; $bn 0.06 5.4 260
Projected purchases in 2012: -35
Quarterly data regressed over 5-years; R2=74%; Standard error = 10. Projections
are for the full year 2012.
Source: J.P. Morgan, Federal Reserve H.Z.1
Exhibit 31: Despite our expectation of increased foreign demand,
the overall supply/demand imbalance will worsen next year giventhe lack of Fed buyingNet issuance of Treasuries per calendar year versus purchases by differentinvestor types; $bn
Treasury net issuance; $bn 1476 1592 1093 1028
Investor
Foreign Investors 555 654 290 540
Commercial banks 92 113 -75 -35
Pension funds/ Insurance companies 211 224 130 150
Broker dealers -68 -29 100 -20
Mutual Funds 68 42 45 60
Federal Reserve 300 244 639 0Other 318 344 -36 333
2009 2010 2011* 2012*
Source: J.P. Morgan, Federal Reserve* 2011 and 2012 are J.P. Morgan estimates.
Exhibit 30: The share of USD in FX reserves has stabilized at
around 65%Estimated* share of USD in foreign exchange reserves; $bn
64%
66%
68%
70%
72%
74%
Jun 00 Jun 02 Jun 04 Jun 06 Jun 08 Jun 10
Source: IMF, Bloomberg* Assumes that the allocation of Japan’s and China’s FX reserves mirror thedistribution of FX reserve currency allocation of advanced economies andEmerging economies, respectively.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 121/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Terry BeltonAC
(1-212) 834-4650
Meera Chandan (1-212) 834-4924
Kimberly L. Harano (1-212) 834-4956
J.P. Morgan Securities LLC
121
banks to be net sellers of coupon Treasuries next year,
likely preferring MBS given its cheapness and the
relatively flat front end of the Treasury yield curve.
Finally, we expect mutual fund demand and demand from
broker-dealers to be in line with their recent averages.
In total, demand across the six largest investor types sums
to $695bn. This suggests that a comparatively large
amount must be funded by other investors, highlighting a
significantly worse supply/demand imbalance than in
2011.
One potential offset is the proceeds of maturing FDIC
guaranteed bank debt: $45bn is due to mature in
December 2011, and another $165bn is due to mature in
2012 (Exhibit 33). While little data is available on the buyers of this product, our best guess is that they were
primarily Treasury/Agency debt buyers and that the
proceeds will likely make their way back into the front
end of the either the Treasury or the Agency debt market.
Exhibit 33: Redemptions of FDIC guaranteed bank debt will beheavy in 2012 providing some support for Treasuries
FDIC guaranteed bank debt redemptions; $bn
0
10
20
30
40
50
60
Nov 11 Jan 12 Mar 12 May 12 Jul 12 Sep 12 Nov 12
Source: Bloomberg
Forecasting foreign demand for TreasuriesIn order to forecast foreign demand for Treasuries, we model average monthly foreign Treasury purchases (available from the Federal ReserveFlow of Funds release) as a function of the US trade deficit with the rest of the world and net issuance of Treasuries (i.e., Treasury net issuance
minus Fed Treasury purchases via quantitative easing). The trade deficit of the US with the rest of the world is a significant driver of foreign purchases of US securities, since it leads to an accumulation of USD reserves abroad. In addition, the magnitude of foreign purchases of Treasuries also depends on available supply as foreign investors have shown a willingness to increase their buying in response to larger USdeficits.
Exhibit A1 shows the details of this model, and Exhibit A2 shows that the model has tracked actual purchases reasonably well since 2002. Themodel shows that each additional $10 bn increase in the monthly trade deficit results in an additional $2bn of foreign buying. Similarly, eachadditional $10 bn increase in monthly Treasury supply results in an additional $4.3 bn of foreign buying. Given our expectation that the tradedeficit and effective net issuance will average $48bn and $85bn per month in 2012, respectively, we look for foreign buying of Treasuries toaverage $45bn per month or $540bn for the year, representing 53% of Treasury net issuance next year.
Exhibit A1: A model for forecasting foreign purchases of Treasuries Exhibit A2: Actual versus predicted monthly average of Treasurypurchases by foreigners; $bn
Net purchases by foreigners modeled as:
Variable Coeff T-stats Proj. 2012
Intercept -0.3 -0.1
Trade surplus; $bn -0.20 -2.5 -48.0
Net issuance of Treasuries, net of Fed
purchases; $bn 0.43 19.1 85.0
Projected foreign purchases of Treasuries; $bn/ month 45
R Square: 91%
Std. Error: 5.6
Regression uses rolling 1-year monthly averages and is fitted on quarterly
data over 10-years.
Source: Federal Reserve Z.1, US Census Bureau, Treasury
0
10
20
30
40
50
60
70
2002 2005 2008 2011
ActualModel
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 122/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Terry BeltonAC
(1-212) 834-4650
Meera Chandan (1-212) 834-4924
Kimberly L. Harano (1-212) 834-4956
J.P. Morgan Securities LLC
122
With the Agency debt market shrinking every year,
Treasuries are likely to be the beneficiary of this demand.
Ratings Risk and the US Fiscal Outlook
One of the great ironies in the Treasury market this year is
the fact that 10-year yields reached all-time lows in the
immediate aftermath of S&P’s downgrade of the US
sovereign rating. The loss of the US ‘AAA’ rating
followed three consecutive years of record deficits that
caused net federal debt to increase to 73% in 2011 from
40% of GDP in 2008. In August, with a divided Congress
unable to reach a consensus on how to stabilize US fiscal
metrics and debt levels projected to continue to grow,
S&P fired a warning shot to the US about the need to get
its fiscal house in order.
While a lot more work lies ahead for the US to stabilize its
finances, the good news is that the debate changed
dramatically in 2011, and a down payment on spending
cuts was made with the passage of the Budget Control Act
(BCA) of 2011. These cuts include $900bn agreed to in
August and another $1.2tn in automatic spending cuts
required by the bill, with the Deficit Reduction Committee
failing to reach agreement. With these cuts, federal net
debt is projected to grow from 73% of GDP currently to
84% by 2021; while still on an uptrend, the trajectory is
much flatter than estimates prior to the passage of BCA,
with July CBO projections for 2021 equal to 95% of GDP(Exhibit 34). In order to stabilize debt levels, we estimate
Congress still needs to reduce the structural deficit by an
additional 1%-pt of GDP per year, or $1.8tn over the 9-
year period. These additional cuts are not likely in an
election year but are certainly possible in 2013.
Exhibit 35 presents a short menu of alternatives that could
be used to reduce the structural deficit further; any
combination that totaled an additional 1% (beyond the
BCA) would stabilize the ratio of US debt to GDP.
To be sure, these projections are sensitive to a number of
assumptions including economic growth. CBO’s baseline
projections, upon which Exhibit 34 is based, assume real
GDP growth will average 2.8% from 2012-2021. If
growth averaged 1% per year lower, we estimate debt
levels would deteriorate by 5%-pts, reaching 89% of GDP
by 2021 (Exhibit 36). In order to stabilize debt-to-GDP
under this weaker growth outlook, Congress would need
to find an additional $2.3tn in savings beyond the $1.2tn
in automatic cuts triggered by the Budget Control Act.
With some, albeit modest, progress made on budget
reform this year, we expect S&P and Moody’s to maintain
their existing US sovereign ratings in 2012, although Fitch
Exhibit 36: 2021 debt levels will deteriorate 5%-pts if growth slipsby 1% per year Projected Net Federal debt to GDP ratio in different growth scenarios; %
70
75
80
85
90
2012 2013 2014 2015 2016 2017 2018 2019 2020 2021
Baseline 2.8% growth
1.8% real GDP growth
Exhibit 34: Half way home: Debt to GDP in the US is still on anupward trajectory; another $1.8tn in cuts are needed to stabilize debtlevelsProjected Net Federal debt to GDP ratio*; %
70
75
80
85
90
95
100
2012 2013 2014 2015 2016 2017 2018 2019 2020 2021
BCA baseline
BCA+ 1.8 tn
Pre BCA baseline
* CBO’s August alternate baseline and J.P. Morgan estimates
Exhibit 35: Alternatives to reduce the structural deficit: anycombination that totals 1% of GDP will stabilize the ratio of US debtto GDPVarious alternatives to reduce the structural deficit; as percentage of GDP and $bn
Cumulative
deficit reduction*
Sunset all Bush tax cuts 1.4 2459
Sunset tax cuts on wealthy (per Obama**) 0.4 700
Feldstein 2% itemized deduction limit 0.2 350
Cap itemized deductions at 28% 0.2 293
Asset and spectrum sales; user fees 0.1 200
Tax international income 0.1 133
% of GDPOptions
*Cumulative deficit reduction relative to BCA baseline for 2013-2021** Extend cuts for families with income under $250,000
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 123/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Terry BeltonAC
(1-212) 834-4650
Meera Chandan (1-212) 834-4924
Kimberly L. Harano (1-212) 834-4956
J.P. Morgan Securities LLC
123
is likely to move to a negative outlook. In their recent
statements, S&P and Moody’s both noted that the US
rating would not be lowered in response to the failure of
the Joint Select Committee, as long as the automatic
spending cuts of $1.2tn are implemented (Exhibit 37).Fitch has been more negative in its comments around the
failure of the Super Committee and we expect them to
place the US on negative outlook while maintaining the
‘AAA’ rating.
Finally, we note that derivative markets also appear
sanguine about the US fiscal outlook and the near-term
risk of further downgrades. CDS on US Treasuries have
been remarkably stable since August, outperforming most
other AAA-rated sovereigns. With the exception of
Norway, US CDS is now tighter than every other AAA-
rated sovereign; spreads are currently 45bp tighter than
Germany and 180bp tighter than France (Exhibit 38).
Trading themes for 2012
While Treasury yields are expected to remain low in 2012,
volatility will still be high, creating regular opportunities
for active investors to enhance the return on their
investment portfolios. In the discussion below, we discuss
four broad areas of opportunities for Treasury investors in
2012 including yield curve carry trading, supply cyclicals,
relative value opportunities around Fed purchase
operations, and cross-currency arbitrage that involves
creating synthetic Treasuries with cheap foreign bonds.
Yield curve trading
While yield curve carry trades typically perform well
when the Fed is on hold and rates are range-bound, the
current environment will likely prove more challengingwith front-end rates so low. In 2012, our approach to yield
curve trading will generally be driven by carry
considerations. But, in contrast to conventional wisdom,
our bias will be to favor negative carry curve trades in the
front end of the curve where convexity is attractive. In
intermediates, however, we expect positive carry trades to
outperform and will generally concentrate long carry
exposure in the intermediate to long end of the curve.
Some support in favor of avoiding positive carry yield
curve trades in the front end of the curve is highlighted in
Exhibit 38: US sovereign CDS spreads are below those of countries
with ‘AAA’ ratings5-year sovereign CDS spreads; bp
Country Rating Outlook Rating Outlook Rating Outlook
Norway AAA Stable Aaa Stable AAA Stable 45
United States AA+ Neg Aaa Neg AAA Stable 52
Sweden AAA Stable Aaa Stable AAA Stable 66
Finland AAA Stable Aaa Stable AAA Stable 70
Australia AAA Stable Aaa Stable AAA Stable 90
Britain AAA Stable Aaa Stable AAA Stable 92
Germany AAA Stable Aaa Stable AAA Stable 97
Netherlands AAA Stable Aaa Stable AAA Stable 117
Denmark AAA Stable Aaa Stable AAA Stable 123
Austria AAA Stable Aaa Stable AAA Stable 215
France AAA Stable Aaa Stable AAA Stable 230
Fitch
5y CDS
S&P Moody's
Source: Bloomberg
Exhibit 37: Comments by rating agencies on further downgrades
Rating
Agency
CurrentUS
sovereign
rating
Comment on Super Committee
S&P
AA+;
Negative
outlook
"The Fiscal Committee's inability to agree on fiscal measures
that would stabilize U.S. government debt as a share of GDP is
consistent with our Aug. 5 decision to lower our rating to 'AA+'.
However, we expect the caps on discretionary spending as laid
out in the Budget Control Act of 2011 to remain in force. If these
limits are eased, downward pressure on the ratings could
build."
Moody's
Aaa;
Negativeoutlook
"The Aaa government bond rating for the United States is
unaffected by the lack of a deficit reduction agreement by the
Joint Select Committee on Deficit Reduction...While a change in
the composition of the spending cuts would not be a major ratingconsideration, a reduction in the total amount that would increase
the projected increase in federal debt over the coming decade
could have negative rating implications"
Fitch
AAA;
Stable
outlook
"Failure by the Super Committee to reach agreement would likely
result in a negative rating action -- most likely a revision of the
rating Outlook to Negative...Less likely would be a one-notch
downgrade. Fitch now expects to conclude its review of the US
sovereign rating by the end of November."
Exhibit 39: Location, location, location: carry was a contrarianindicator of performance for front-end curve trades in 2011 3-month holding period P/L from going long the belly (5s) versus a level andcurve neutral combination of wings (2s/10s and 2s/7s) in 2011 vs. the 3-monthcarry and roll in the trade at time of initiation between 1/1/11 and 8/12/11; bp
-30
-20
-10
0
10
20
30
-6 -4 -2 0 2 4 6
2s5s10s
2s5s7s
3m carry and roll at initiation; bp
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 124/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Terry BeltonAC
(1-212) 834-4650
Meera Chandan (1-212) 834-4924
Kimberly L. Harano (1-212) 834-4956
J.P. Morgan Securities LLC
124
Exhibit 39, which compares the performance of 2s/5s/7s,
and 2s/5s/10s weighted butterfly trades this year to ex ante
carry and roll on the trade. The weights on each trade were
designed to make the butterfly curve and level neutral and
returns are calculated over 3-month holding periods. The
Exhibit highlights that while carry is a good leading
indicator of P/L on these trades, it is a contrarian
indicator; the more negative the carry, the larger the
subsequent P/L. Initiating these weighted curve tradeswhen carry was positive in 2011 has resulted in a loss
80% of the time, with an average 3-month P/L of -8bp.
The poor performance of front-end butterflies during 2011
arises from the negative curve convexity on these trades.
As shown in Exhibit 40, positive carry 2s/5s/10s weighted
butterflies that are long the belly tend to become
steepeners as the curve flattens and become flatteners as
the curve steepens. This behavior reflects the changing
volatility of 2-year yields with 2s becoming sticky, as
rates fall and the curve flattens, and more volatile as rates
rise and the curve steepens. To offset these changing
volatilities, rebalancing the position would involve addingflattening risk (i.e., selling more 2s versus 5s) when the
curve flattens to offset the falling volatility of 2s.
Similarly, rebalancing involves adding steepening risk
(i.e., scaling back the short in 2s versus 5s) when the curve
steepens to offset the increasing volatility of 2s.
In contrast to carry trades anchored in the front end, we
favor positive carry yield curve trades anchored in the
intermediate sector where convexity is less of an issue.
Exhibit 41 shows that the 3-month holding period returns
from being long 7s versus a risk-weighted 5s/10s barbell
have been well correlated with the carry and roll on the
trade.
In addition to carry, we note that butterflies anchored in
the long and intermediate sector of the curve have been
more mean reverting than front-end butterflies. To
determine how mean reversion of various butterfly trades
performed in 2011, we tested a simple trading rule that
involved initiating a long position in the belly versus the
wings when the belly appeared cheap by more than 1.5
standard deviations on a curve and level neutral basis. The
reverse was done when the belly appeared rich. The trade
was then held for a 1-month period. Exhibit 42 shows the
results of such a trading rule and shows that intermediate
and long-end butterflies had higher risk-adjusted returns
by following such a strategy. By contrast, front-end curve
Exhibit 40: A long position in 5s versus 2s and 10s increasingly
becomes a steepener in a rally and a flattener in a sell-off 3-month holding period P/L from initiating a long position in a level and curveneutral 2s/5s/10s butterfly when it carries positively (bp), regressed against thechange in the 2s/10s curve during the holding period (%)
-35
-30
-25
-20
-15
-10
-5
0
5
10
15
-1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8
Change in 2s/10s curve over the holding period; %
Y = -4.03 X1 - 19.43 X1̂ 2 - 5.34R² = 29.0%standard error = 8.26
period = Oct 12,10 - Aug 12,11
Exhibit 41: The performance of intermediate butterflies is well
correlated with carry at inception 3-month holding period P/L of a long position in 7s versus 5s and 10s versus 3-month carry and roll in the trade at inception; bp
-6
-4
-2
0
2
4
6
8
10
-2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.03m carry and roll in weighted trade at time of initiation; bp
Y = 3.0385 X1 + 4.1926R² = 63.35%standard error = 2.0730period = Jan 01,11 - Aug 12,11
Exhibit 42: Butterflies in the long end of the curve are more likely tobe mean-reverting than front-end butterflies 1-month holding period returns from initiating level and curve neutral butterflieswhen the belly is mispriced by at least 1.5 standard deviations in 2011; bp unlessotherwise specified
Bfly Avg P/L Min Max # Trades Hit rate Std Dev Risk adj returns
10y15y20y 1.5 1 3 20 100% 0.5 3.00
10y12y15y 0.5 0 1 21 100% 0.2 2.88
5y10y15y 5.0 -7 8 20 95% 3.3 1.50
5y7y10y 3.8 -2 7 8 88% 2.8 1.36
7y15y20y 1.8 0 4 34 85% 1.4 1.285y10y20y 6.2 -10 12 34 91% 5.8 1.06
3y5y7y 2.1 -4 7 14 71% 3.8 0.55
2y5y10y 1.5 -11 12 17 65% 6.7 0.22
2y5y7y 0.7 -9 6 16 63% 5.0 0.15
1y3y5y 0.2 -5 28 16 13% 8.0 0.03
10y20y30y -0.2 -4 9 5 20% 5.3 -0.04
2y3y5y -0.3 -4 6 19 37% 2.4 -0.13
1y2y3y -0.6 -3 1 2 50% 2.8 -0.20
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 125/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Terry BeltonAC
(1-212) 834-4650
Meera Chandan (1-212) 834-4924
Kimberly L. Harano (1-212) 834-4956
J.P. Morgan Securities LLC
125
trades exhibited weak mean reversion, with the simple
trading rule producing poor risk-adjusted returns.
In sum, our yield curve bias in 2012 will be to favor
negative carry curve trades in the front end of the
curve where convexity is attractive. In intermediates,
we expect positive carry trades to outperform and will
look to add to long carry positions as high carry points
become cheap on the curve.
Supply cyclicals
Trading strategies around Treasury auction concessions
had a mixed track record in 2011, with returns weakest intrades around 2-, 3-, and 5-year auctions. The largest and
the most consistent supply concessions have been in the
very long end of the curve. Exhibit 43 shows the
performance of various trades in the days before and after
auctions in different sectors. As shown in the exhibit,
hedged for the level of rates, the 7s/30s curve has
steepened going into the bond auction in 9 out of 11
months this year by an average of 4bp. Notably, this
steepening has persisted despite Fed purchases in the very
long end via Operation Twist. Exhibit 44 shows that even
though the magnitude of the supply concession has
been smaller since Operation Twist began, it is still
sizable.
With liquidity poor, we expect these concessions to persist
in 2012, and we favor curve steepeners going into bond
auctions. Our outlook is motivated in large part by our
expectation that market depth will likely remain poor in
1H12, as bank deleveraging continues. Historically,
supply concessions have increased in periods of poor
market liquidity, as shown in Exhibit 45. Thus, we
Exhibit 43: Supply concessions have been the largest and most
consistent going into the bond auction, a trend we expect willcontinue in 2012 Average P/L in select weighted curves* from being short the sector beingauctioned in the days prior to its auction versus being long the sector in the dayspost auction calculated between January and October 2011; bp unless otherwisespecified
Avg Chg Hit rate Avg Chg Hit rate
2s3s 1.5 80% 1.0 60% 2y
2s10s 2.4 50% 1.7 60% 2y
2s3s 1.0 80% -0.6 50% 3y
3s5s 1.0 40% 1.6 60% 3y
3s7s 1.4 40% 0.7 80% 3y
3s5s -0.8 50% -0.3 60% 5y
5s10s 0.9 40% 0.7 60% 5y
5s30s 3.0 50% 2.5 70% 5y
3s7s 2.5 70% 4.6 70% 7y
7s10s 0.3 50% 0.5 60% 7y
7s30s 2.7 60% -1.2 50% 7y
2s10s 0.4 70% 1.1 50% 10y
5s10s 1.1 70% 0.6 30% 10y
7s10s 1.6 70% 0.0 40% 10y
2s30s 1.6 80% 1.1 60% 30y
5s30s 4.1 80% 1.1 40% 30y
7s30s 4.0 80% 3.6 50% 30y
Wtd.
Trade
Pre auction Post auctionCentered
around
auction in
* Weighted trades include 0.57*3s-2s, 0.31*10s-2s, 0.57*5s-3s, 0.58*7s-3s,0.85*10s-5s, 0.64*30s-5s, 0.95*10s-7s, and 30s-0.85*7s. Curves are weighted to
be level neutral.
Exhibit 45: Poor liquidity tends to amplify auction cyclicals30-year – 0.85 * 7-year Treasury rates averaged in the business days around the30-year auction in periods of high market depth* and low market depth (Data over last 3-years has been used); %
1.65
1.70
1.75
1.80
1.85
1.96
2.01
2.06
2.11
2.16
-12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12
Business days around 30-year auction
High marketdepth
Low marketdepth
* Market depth is calculated as the 10-day moving average of average size of thetop three bids and offers, in $mn, for the on-the-run 10-year Treasury note,averaged between 8:30 a.m. and 10:30 a.m. daily. Periods with market depth of greater than $100mn are classified as high market depth periods.
Exhibit 44: While supply concessions ahead of the bond auction
have fallen since Operation Twist began, they are still sizable Average change in 30-year - 0.85 *7-year curve in eight business days before thebond auction during QE2 (January to June 2011), post QE2 (July and August)and Operation Twist (October and November); bp
0
2
4
6
8
10
QE2 Jul-Aug Operation Twist
Average in different periods
2011 YTD average
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 126/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Terry BeltonAC
(1-212) 834-4650
Meera Chandan (1-212) 834-4924
Kimberly L. Harano (1-212) 834-4956
J.P. Morgan Securities LLC
126
recommend that investors actively position for a steeper
curve ahead of bond auctions, particularly in periods in
which risk aversion is elevated and balance sheet is
constrained.
Trading around the Fed
The Fed’s purchase operations still continue to provideshort-term trading opportunities in the Treasury market.
Its preference to consistently purchase securities that are
cheap to a fitted curve has resulted in further compression
in the dispersion of valuations along the Treasury curve,
as measured by the root mean square error of a fitted
Treasury curve (RMSE). Even though this has been a
well-identified trading theme for some time (through QE1
and QE2), the strategy of owning issues cheap on the
curve in anticipation of Fed buying has remained
profitable, even under Operation Twist. Exhibit 46
presents the results from following a simple daily strategy
of buying issues that are cheap on the curve and selling
similar maturity issues that are trading rich; the strategyhas been most profitable in the 6- to 8-year sector, and 25-
to 30-year sector. As a result, the dispersion of valuations
in these sectors has compressed to their lowest level in a
year (Exhibit 47). On the other hand, as also shown in the
exhibit, dispersion in the 8- to 10-year sector still has
room to compress further relative to its recent history.
Typically, the richest issue in this sector is the on-the-run
10s, and these have historically tended to cheapen relative
to the older securities as they get reopened in the weeks
Exhibit 48: New-issue 10s have cheapened relative to old 10s inthe weeks after they have been auctioned
Adjusted* and unadjusted on-the-run / triple olds yield spread in the days after anew 10-year is auctioned; %
-0.070
-0.065
-0.060
-0.055
-0.050
-0.045
-0.040
-0.035
-0.030
0.120
0.125
0.130
0.135
0.140
0.145
0.150
0.155
0.160
0 10 20 30 40 50 60
# of business days after a new 10-year auction
New/ olds spread adj. for 7s10s curveNew/ olds spread
* Adjusted for the 7s/10s yield curve. Equals on-the-run 10-year yield – tripleolds yield – 0.25 * 7s/10s curve
Exhibit 49: The yield pickup of synthetic assets constructed bybuying JGBs/ Bunds and swapping back to USD is substantial…Yield on synthetic asset* minus Treasury yield for various government bondsand sectors; bp
JGB Bunds JGB Bunds UST
2y 117 30 2.1 0.2 0.3
3y 124 23 1.6 0.2 0.2
5y 133 15 1.1 0.4 0.4
Sector
Spread pick of synthetic
versus US Treasuries; bp
Annualized risk-
adjusted return**
* 2-year synthetic yield equals JGB 2-year yield – basis swap spread + (USDswap yield – Yen swap yield) + 6M/3M Libor basis.** 3-month return/ risk. 3-month return is one-fourth the yield plus the roll. Risk is(5-year standard deviation of 3-month changes in yield)*(modified duration).
Exhibit 46: A strategy of buying issues with the highest yield errors
versus issues with the lowest yield errors has been profitable duringOperation Twist Results from back-testing a simple trading rule*; table shows average change**,standard deviation of changes, minimum change, average change/ standarddeviation and hit rate*** for par asset swap spread difference and yield differencefor each bond pair
Avg chg Min chg Avg/Std.Dev Hit Rate Avg chg Min chg Avg/Std.Dev Hit Rate
6-7 1.3 -1.3 0.8 80% -1.0 -4.3 -0.6 40%
7-8 2.1 0.3 1.9 100% 0.9 -2.1 0.6 67%
8-9 -0.7 -2.2 -0.6 20% 0.8 -1.0 0.7 73%
9-10 -2.0 -4.9 -1.0 13% 0.9 0.4 2.6 100%
10-15 1.3 -4.1 0.5 80% -1.9 -5.2 -0.9 13%
15-20 -1.6 -3.2 -1.5 7% -0.4 -2.8 -0.3 53%
24-27 1.2 0.6 3.2 100% 0.8 0.1 1.6 100%
27-30 0.3 -0.6 0.6 80% 0.5 -0.3 0.9 80%
Yrs to
mat
Par asset swap Yields
* Trading rule involves buying the issue with the highest yield error in each maturitybucket versus issue with the lowest yield error in each bucket on a daily basisstarting September 21, 2011. Holding period is one month.** Positive number for average change implies that bond with highest yield error outperformed the bond with the lowest yield error *** Hit rate equals number of trades in which spread widened/ total number of trades in each maturity bucket
Exhibit 47: Mispricings are largest in the 8- to 10-year sector of the
curve RMSE of Treasuries in select parts of the Treasury curve; bp
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
May 11 Aug 11 Nov 11
6y-8y8y-10y20y-30y
Latest observation
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 127/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Terry BeltonAC
(1-212) 834-4650
Meera Chandan (1-212) 834-4924
Kimberly L. Harano (1-212) 834-4956
J.P. Morgan Securities LLC
127
after they are first auctioned (Exhibit 48). In general, we
would favor overweighting off-the-run 10s versus newly
issued on-the-runs to benefit from supply-related
cheapening as well as the Fed’s preference for purchasing
cheap off-the-run issues.
Creating front-end synthetic Treasuries
The cross currency basis of USD with JPY and EUR was pushed further into negative territory this year, as the
pullback by US money funds lending to European banks
resulted in increased demand for USD funding. At its
current level of -80bp, the 2-year JPY/USD cross-currency
basis is near the lows of the last decade, while the
EUR/USD basis is at its lowest since December 2008.
This has made it very attractive for investors to create
synthetic USD assets constructed by buying foreign
government bonds that are swapped back to USD (see
grey box).
Synthetic Treasuries and cross currency basis swapsA cross-currency basis swap is an agreement to exchange floating rates in a foreign currency against USD Libor. The transaction involvesan exchange of foreign currency up front with floating payments exchanged over time. The floating payments equal USD Libor versusforeign Libor plus a spread where, by convention, the spread on the foreign Libor is referred to as the basis swap spread ( Exhibit A).Currently, spreads for 2-year USD/JPY basis swaps equal -80bp meaning an investor can pay 3-month Yen Libor minus 80bp versusreceiving 3-month USD Libor flat for the next eight quarters.The large negative basis spread currently makes it attractive for a USD floating-rate investor to create synthetic floating rate USD assets by
buying floating rate assets in Yen and swapping them to USD floating-rate assets with a cross-currency basis swap. For example, a USinvestor can synthetically create a 2-year floating rate USD asset that earns USD Libor + 80 bp by 1) investing in a Yen floating-rate assetthat pays Yen Libor flat and 2) entering in a cross-currency basis swap at a spread of -80.The yield pick-up embedded in the negative basis swap also creates opportunities for fixed-rate investors. An investor benchmarked to 2-
year Treasuries, for example, can create a synthetic 2-year Treasury note with a much higher yield by instead buying 2-year JGBs and usingthe basis swap and interest rate swap markets to convert the cash flows back into fixed-rate USD flows. To convert the JGB cash flows intofixed-rate USD flows, the investor 1) buys the JGB and asset swaps it into a semiannual Yen floater, 2) uses a 3s/6s basis swaps to convertJPY semiannual cash flows into quarterly cash flows, 3) enters a cross currency basis swap (paying Yen Libor + spread versus receivingUSD Libor) to convert it to a USD floater, and 4) converts the floating rate USD cash flows into fixed semiannual cash flows using plainvanilla interest rate swaps. Of course, these trades can be collapsed into a single trade to effectively convert the 2-year JGB cash flows intoa 2-year semiannual fixed-rate USD asset. The cash flows on this trade have the investor paying a Yen coupon equal to the yield on theJGB (offsetting JGB cash flows) and receiving a USD coupon (Exhibit B). The yield on the synthetic 2-year Treasury can be calculated as:Synthetic yield = JGB asset swap spread + 6M/3M Libor basis – basis swap spread + USD fixed swap rateAt current levels, this implies
Synthetic 2-year yield = -25 + 12 - (-80) + 77 = 1.44% or 117bp higher than 2-year Treasury notes.
Exhibit A: Basic cross currency swap Exhibit B: 2-year fixed semiannual currency swap
JPY 7,716 mm initial exchange
Investor Dealer Yen Libor + spread
USD Libor
US $ 100 mm initial exchange
JPY 7,716 mm initial exchange
Investor Dealer JPY 0.12%
USD 1.44%
US $ 100 mm initial exchange
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 128/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Terry BeltonAC
(1-212) 834-4650
Meera Chandan (1-212) 834-4924
Kimberly L. Harano (1-212) 834-4956
J.P. Morgan Securities LLC
128
JGBs in particular look very attractive when swapped
back into US dollars. A trade constructed in this fashion in
the 2-year sector currently results in a pick-up of 117bp
versus similar maturity Treasuries (Exhibit 49). Not only
is this incremental yield high relative to recent history
(Exhibit 50), but the unlimited USD swap lines provided
by the Fed limit the potential widening in the EUR/JPY
basis. Currently, the 3-month JPY/USD FX basis is
trading at -70bp. Converting this basis (which is quoted
versus Libor) to an OIS basis (OIS basis = Libor basis +
JPY 3M Libor/OIS spread – USD 3M Libor/OIS spread)
translates to an OIS basis of -100. Because this spread is
equivalent to the penalty rate the Fed charges banks to use
its currency swap line (OIS+100), we should not expect
any further widening in the 3-month basis. This limits the
downside risks in buying JGBs swapped to US dollars.
Outlook on STRIPS
The size of the P-STRIPS market is unchanged this year,
with $197bn outstanding. However, the intra-year pattern
exhibited more volatility: net stripping activity picked up
in Q2 bringing the market to its largest size in a decade,
but fell in 2H11. The quarter ending October 2011 saw the
largest 3-month decline in the size of the market since
January 2009, as the very long end of the curve flattened
massively and Operation Twist started taking long-end
supply out of the market (Exhibit 51). Investor demand
for STRIPS remained largely in the very long end of the
curve. P-STRIPS with more than 17 years to maturity now
comprise 66% of the market versus 60% at the beginning
of the year (Exhibit 52).
Exhibit 53: Low yields, a flat curve, and Operation Twist are likelyto keep net stripping activity muted in 1H12
A model for quarterly changes in P-STRIPS outstanding; $bn
Variable Coefficient T-statistics
Intercept -55.3 -5.7
10-year yields; % 10.5 5.4
10s/30s curve; % 24.9 5.8
Change in S&P 500; points 0.01 1.7Quarterly Treasury duration purchases
by the Fed; $bn of 10-year equivalents -0.0209 -1.5
R2 = 50%; Std. error= 5
Model fitted for quarterly data over a 5-year period. 10-year Treasury rates andthe 10s/30s curve are quarterly averages.
Exhibit 52: STRIPS in the very long-end of the curve continued togain market share in 2011P-STRIPS outstanding by sector as a percentage of total; %
0
10
20
30
40
50
60
70
2002 2005 2008 2011
<5.5-yrs5.5-to 10-yrs10- to 17-yrs17+ yrs
Exhibit 50: …and at multi-year highsSynthetic 2-year Treasury yield* minus 2-year US Treasury yield; %
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
May 10 Nov 10 May 11 Nov 11
* Synthetic 2-year Treasury yield equals JGB 2-year yield – basis swap spread +(USD swap yield – Yen swap yield) + 6M/3M Libor basis.
Exhibit 51: The size of the STRIPS market fell in 2H11P-STRIPS outstanding; $bn
170
175
180
185
190
195
200
205
210
Nov 09 May 10 Nov 10 May 11 Nov 11
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 129/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Terry BeltonAC
(1-212) 834-4650
Meera Chandan (1-212) 834-4924
Kimberly L. Harano (1-212) 834-4956
J.P. Morgan Securities LLC
129
Looking ahead, we expect net stripping activity to stay
muted and for the market to continue to shrink in 1H12.
This view is in part motivated by our outlook that the
market environment will be characterized by high risk
aversion as the crisis in Europe escalates, which will keep
yields low and the very long end of the curve flat causing
risky assets to underperform. Historically, a flat curve,
low yields, and weak equities have all caused the STRIPS
market to shrink (Exhibit 53).
Fed purchases of Treasuries via Operation Twist will also
have a significant impact and is likely to result in
increased reconstitution activity for the purpose of
delivering whole bonds into the Fed. As a rule of thumb,
quarterly Treasury duration purchases of $100bn of 10-
year equivalents by the Fed have lowered P-STRIPS
outstanding by around $2bn per quarter. In 1H12, we
expect duration buying by the Fed to average $180bn of
10-year equivalents per quarter, suggesting that the market
should shrink by $4bn per quarter due to this factor alone.
With 32% of Fed purchases in maturities greater than 20-
years and the bulk of the STRIPS market concentrated in
that sector of the curve, we expect reconstitutions to be
particularly high. Indeed, there was already evidence of
this in October (the first month of Operation Twist).
Exhibit 54 shows that issues that were purchased in size
by the Fed in October were also the ones that were
reconstituted the most.
In sum, we look for the STRIPS market to shrink
further in 1H12. Whole bonds that are trading cheap
relative to the fitted Treasury curve will likely lead the
way in reconstitution activity (Exhibit 55), since the
Fed concentrates its purchases in such issues (almost
75% of its purchases have occurred in positive yield
error bonds thus far in Operation Twist). This has two
implications for the STRIPS market. First, shorter
maturity Cs should richen versus similar maturity
Ps, since reconstitution of longer maturity bonds will
reduce the relative supply of shorter maturity Cs.
Indeed, this has been one driver of the C-P spread
recently (Exhibit 56). This richening of Cs will likely
occur most in Cs that correspond to coupon dates of
highly stripped whole bonds that have positive yield
Exhibit 55: Issues with high yield errors like Aug-40s and May-40s
will likely lead reconstitution activity given the Fed’s preference for issues trading cheap on the curve Amount stripped by issue ($bn) versus current yield error (bp)
F-36
F-37
M-37F-38
M-38
F-39
M-39 A-39
N-39F-40
M-40
A-40
N-40F-41
M-41 A-41
N-41
0
5
10
15
20
25
30
-2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5
Yield error; bp
Exhibit 54: Operation Twist Impact: Issues that were purchased in
size by the Fed in October were also reconstituted the most duringthe monthChange in stripped amount for the top 10 issues purchased by the Fed in the 10-to 30-year sector in October versus Fed purchases of the bond in October; $bn
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
0.0 0.5 1.0 1.5 2.0 2.5 3.0
Fed purchases in Oct; $bn
Exhibit 56: Increased reconstitutions are likely to drive C-P spreadsnarrower in the coming monthsCoupon minus Principal STRIPS yield spread in the 2022-23 sector* (%) versusthe total amount of P-STRIPS outstanding ($bn)% $bn
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
170
175
180
185
190
195
200
205
210
2010 2011
Average C-P spreads; %P-STRIPS outstanding; $bn
* Average C-P spread between issues maturing on 8/15/22, 11/15/22, 2/15/23and 8/15/23.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 130/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Terry BeltonAC
(1-212) 834-4650
Meera Chandan (1-212) 834-4924
Kimberly L. Harano (1-212) 834-4956
J.P. Morgan Securities LLC
130
errors in the very long end of the curve. Second, Ps
corresponding to bonds that are likely to be reconstituted
should richen relative to other surrounding Ps.
In sum, we expect the size of the STRIPS market to
decline further in 2012. Short maturity Cs should richen
versus similar maturity Ps, as should Ps on long-dated
bonds that are currently cheap to the par curve.
Trading themes
• Stay long duration in early-2012 targeting 10-year
yields to reach 1.70% in 1Q12
A worsening of the European sovereign debt crisis
should accelerate the flight-to-quality bid into
Treasuries early next year pushing yields lower; rich
valuations, a deteriorating supply/demand imbalance,
and poor technicals will limit the upside, however, and
keep rates range bound later in 2012.
• Position for lower front-end yields in 2012
Front-end rates should trend lower in 2012 driven by
declining front-end supply, further policy rate guidance
by the Fed that flattens term premium, an increase in
the supply of excess reserves following QE3, and the
end of Operation Twist; look for GC repo to average
near 5bp, and 2-year yields to fall to 17bp.
•
Avoid crowds in 2012A secular decline in liquidity will lead to heightened
whipsaw risk in 2012 and cause Treasury yields to be
volatile within a broad trading range; trading strategies
based on our measure of investor positions that fade
crowded consensus trades are likely to perform well in
2012 as they did this year.
• Create synthetic 2-year Treasury notes by buying
JGBs and currency swapping them back to USD
Synthetic Treasuries constructed by buying foreign
bonds and currency swapping them back to US dollars
are at the most attractive levels of the last decade. The
yield on a synthetic 2-year note created by assetswapping a 2-year JGBs currently equals 1.44% or 117
bp higher than 2-year Treasury yields.
• We expect positive carry yield curve trades in the
front end to perform poorly in 2012; favor convexity
over carry in the front end of the curve and carry in
the intermediate sector of the curve
Positive carry yield curve trades in the front end are
likely to perform poorly reflecting the extreme negative
convexity in these trades. We recommend negative
carry curve trades in the front end of the curve where
convexity is attractive. In intermediates, we expect
positive carry trades to outperform and will look to add
to long carry positions as high carry points become
cheap on the curve.
• The long end of the curve is biased slightly steeper;
our mid-year target for the 10s/30s Treasury curve
is 110bp
The 10s/30s curve is slightly flat to fair value; we favor
positioning for a steeper curve in advance of long-end
supply to take advantage of the large supply concession
in the long end.
•
An increase in reconstitutions in the STRIPS marketshould cause shorter maturity Cs to outperform Ps
We expect the size of the STRIPS market to decline
further in 2012. Short maturity Cs should richen versus
similar maturity Ps, as should Ps on long-dated bonds
that are currently cheap to the par curve.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 131/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Praveen Korapaty (1-212) 834-3092
Alberto Iglesias (1-212) 834-5116
J.P. Morgan Securities LLC
131
US Interest Rate Derivatives
• There is a new normal in the swaps market—
convexity-hedging flows and issuance-related
swapping matter less, but European markets and
Fed purchases matter more
• FRA-OIS spreads should remain under pressure
into 1Q12, but concerted central bank policies
should cause it to narrow thereafter—look for
FRA-OIS to widen in the short term, but
eventually narrow to 50bp by mid-year.
Intermediate maturity swap spreads are similarly
biased wider in the near term, but should end
1H12 close to current levels
• Thanks to sticky front-end Treasury yields,
short-expiry swaptions on short tails are
effectively like options on front-end spreads. We
recommend utilizing 1Mx1Y payer / receiver
swaptions to initiate asymmetric exposure to
front-end spread widening / narrowing
• Fed purchases of longer-end Treasuries will offer
opportunities in 1H12—trade the impact of intra-
month swings in purchase pace on swap spreads,
and look to trade maturity-matched swap spread
switches based on the Fed’s metric of value
• Initiate steepeners between intermediates and the
long end hedged with Eurodollars, to position for
yield curve normalization as well as carry
• Look to initiate positive carry, belly richening
butterflies as attractive ways to gain empirically
convex exposure to higher front-end yields
• Yield curves between the front end and
intermediates will remain directional in 2012, but
conditional curve trades are not cheap—initiate
“synthetic” conditional curve trades, created by
replacing swaptions at the front end with YCSOs
• Approach 2012 with a long gamma bias, but look for 3Yx10Y to decline to 6bp/day by 1Q12 end
• Bermudan receiver swaptions offer directional
vega exposure, as well as exposure to cheap
implied correlation and forward volatility. We
introduce a novel approach to measure relative
value in Bermudan swaptions, based on using
approximating Canaries
Swaps
Swap spreads across the curve traded in significantly
wide ranges over the past year, and have recentlywidened to the upper end of those ranges (Exhibit 1).
The move wider late in the year has been mainly due to
the steadily deteriorating sovereign debt crisis in Europe,
which has now spread well beyond its origins in Greece.
The resulting moves in swap spreads have been
reminiscent of 2008, although trading ranges this year
remain small in comparison; 10-year spreads, for
instance, traded in a 75bp range in 2008.
The volatility in swap spreads is a reflection of the fact
that spreads have been influenced by numerous factors in
significant ways. Intermediate spreads narrowed going
into the second quarter as markets priced in the end of
QE2, but worsening conditions in Europe steadily have
pressured spreads wider since then; more recently, the
insufficient nature of policy measures announced after
the recent EU summit has sparked renewed widening in
spreads, bringing them closer to the wides of the year in
the 2- to 10-year sector of the curve.
In short, spreads have been largely driven by the Fed’s
Treasury purchases and by the European banking system
Exhibit 1: Swap spreads are near the upper end of a relatively wide
trading range in 2011Statistics regarding maturity matched swap spreads in various benchmark sectors;bp
* As of 11/17/2011
Exhibit 2: Collateralization rates have been on the rise across all OTCderivative marketsPercent of trade volume subject to credit support agreements
Source: ISDA Margin Survey
Current Start Average High Low Range
2Y 53.2 18.1 23.7 53.2 12.6 40.6
3Y 50.1 23.9 28.2 50.1 20.4 29.7
5Y 44.3 15.9 25.0 44.3 16.2 28.1
7Y 34.1 9.6 20.0 34.1 10.5 23.5
10Y 19.3 5.5 11.5 22.7 3.9 18.8
30Y -27.3 -21.6 -26.3 -19.8 -39.0 19.2
2011 YTD statistics
2003 2004 2005 2006 2007 2008 2009 2010 2011
All OTC derivatives 30 51 56 59 59 63 65 70 70
Fixed Income derivatives 53 58 58 57 62 68 63 79 79
Credit derivatives 30 45 59 70 66 74 71 93 93
FX derivatives 21 24 32 37 36 44 36 57 58
Equity derivatives 27 45 45 46 51 52 52 71 72
Commodities 17 25 33 44 40 39 39 62 60
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 132/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Praveen Korapaty (1-212) 834-3092
Alberto Iglesias (1-212) 834-5116
J.P. Morgan Securities LLC
132
crisis—factors that swap market participants would not
normally view as endogenous. Indeed, endogenous
factors that have historically exerted considerable
influence on swap spreads—mortgage market hedging
flows and the swapping of corporate issuance, as well as
longer-term drivers such as budget deficit expectations
and funds rate expectations—have played a less
significant role this year. With deficit expectations
remaining largely stable and with the Fed on hold for
several years, longer-term drivers have not experienced
enough volatility to influence swap spreads in a
significant way. Also, the numerous well-understood
hurdles to refinancing mortgages have made mortgages
ever less negatively convex; this fact, together with
falling portfolios at the two major GSEs, has largelyhelped to explain the decline of mortgage-hedging flows
as a driver of swap spreads. Somewhat less permanently,
low and sticky yield levels appear to have lessened the
imperative to swap fixed rate debt issuance on the part of
many high grade corporate issuers, including those who
traditionally swap their issuance.
Such behavioral shifts on the part of market participants
are only one among numerous changes impacting the
swaps market. Another key transformative development
is of course the ongoing implementation of the Dodd-
Frank reforms, as they pertain to the derivatives markets.
Much uncertainty remains in this regard, and we discussthe current status of regulatory implementation; however,
one key change that has already taken place in the swaps
market is a move towards OIS discounting as a quasi-
standard for valuing off-market swaps. To be sure, this is
not exactly a new development, and fixed income
derivatives markets have led the broader trend towards
higher levels of collateralization in recent years
(Exhibit 2). Given this trend, the move towards OIS
discounting (which is arguably closer to the funding rates
for commonly used collateral) has been in the making for
sometime now. Nonetheless, there is growing evidence
that markets have gravitated towards value measures that
are more consistent with OIS discounting. For instance, par asset swap spreads for premium/discount bonds
appear to serve as better signals of value when computed
using OIS discounting, as opposed to Libor discounting.
In other words, swap market participants will need to
adjust to a world where the reference curve (with respect
to which asset valuations are measured) and the discount
curve are different; we discuss this further in a later
section.
Swap spread drivers: the new normal
As we consider the likely outlook for intermediate
maturity swap spreads over the course of next year, it is just as important to recognize what will not be driving
swap spreads. These include several traditional and long-
standing factors that influenced swap spreads in the past,
such as (most notably) budget deficit expectations. To be
sure, should the path of fiscal policy deviate from
expectations in a sufficiently large way, budget deficit
expectations will undoubtedly re-emerge as a key driver
Exhibit 3: No convexity to hedge—the magnitude of convexity in current
coupon mortgages is near historical lows Absolute value of FNMA 30-year current coupon TBA Libor OA-convexity;
Exhibit 4: The declining footprint of mortgage hedgers in the swapsmarket is also due to mandated reductions in the GSEs’ retainedportfoliosTotal outstanding swaps notional amount held by Fannie Mae and Freddie Mac,versus their total retained portfolio size;$bn $bn
0.0
0.5
1.0
1.5
2.0
2.5
3.0
Feb 00 Oct 02 Jul 05 Apr 08 Jan 11
Current
1400
1450
1500
1550
1600
1650
1700
800
1000
1200
1400
1600
1800
2000
Dec 06 May 08 Sep 09 Feb 11
Retainedportfolio
Swapnotional
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 133/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Praveen Korapaty (1-212) 834-3092
Alberto Iglesias (1-212) 834-5116
J.P. Morgan Securities LLC
133
of swap spreads. In this regard, it is still possible (but
perhaps not likely) that fiscal policy adjustments will be
enacted sometime next year; despite the initial failure of
the Congressional super committee, mandatory cuts do
not take effect until 2013, and thus significant time
remains for Congressional action on fiscal policy.
A tactical model for benchmark 10-year maturity matched swap spreadsIt is no exaggeration to say that the world of swap spreads is experiencing its own “new normal.” Traditional, long-term drivers have simply become lessvolatile and more range-bound—Fed funds expectations are virtually assured to remain stable for a long period of time, and budget deficit expectationshave become quite stable. Traditional tactical factors have become less relevant too; declining portfolio sizes at the GSEs, and structural impediments torefinancing mortgages have made mortgage-hedging flows relatively insignificant in terms of driving swap spreads. In addition, anecdotal evidence
points to reduced swapping of debt issuance, thanks to low outright yield levels.
Long-term structural factors remain relevant, and budget deficit expectations could still prove significant in 1H12 if fiscal policy were to deviatematerially from current expectations. That said, trading swap spreads in the 10-year sector now requires a different, more tactical model to assess fair value, which is outlined here. This model is more short term in nature out of necessity, given the recent nature of the developments discussed above. Thefactors incorporated in our model are (i) 6-month FRA-OIS spreads (to capture bank funding pressures and Libor expectations), (ii) the bank stock index(a broader measure of financial system health), (iii) weekly open market gross purchases (i.e., not net of sales at the front end) of Treasuries by the Fed(which matters for swap spreads even under a balance sheet neutral construct as with Operation Twist), and (iv) weekly swapped corporate issuance.
The first three factors are relatively straightforward; the last—swapped issuance—has been more interesting, because low-yield levels have likelydeterred the amount of fixed rate issuance that has been swapped this year. Understanding the aggregate swapping activity of high grade debt issuers isdifficult given sketchy data, but necessary nonetheless since moves in yields could alter swapping behavior by material amounts. Therefore, here we
attempt to estimate the fraction of high grade issuance that gets swapped as a function of rates. Our approach is as follows. For a representative sample of non-financial companies over a 3-year period, we examined data on the aggregate amount of fixed-rate issuance (as a percentage of overall fixed rateissuance) that was swapped to floating, using information from financial statements. As seen in Exhibit A1, the fraction of overall non-financialcorporate debt issuance that gets swapped to floating has indeed been sensitive to yield levels, falling sharply as 10-year swap yields fall below 3%. Alsointeresting is that the data suggests that swapping activity may rise again in very low yield environments. While the data in these ultra-low yield regimesis admittedly limited, it stands to reason that swapping activity might begin to rise again as yields fall below a second threshold, perhaps reflectingexpectations of the Fed being on hold for a very long time.
Unfortunately, such data cannot be compiled for financial issuers; it is nonetheless important to account for potential changes in their swapping behavior too, given that these institutions are both large issuers and also tend to swap considerable fractions of their issuance. Moreover, anecdotal evidencesuggests that these issuers too have curtailed their swapping activity as yields have fallen. To estimate their swapping behavior as a function of yields, wetake the leap of faith that the fraction of financial issuance that is swapped may be modeled as a linear function of the corporate issuance swap-fraction.That is, we assume that financial issuance fractions are a similar non-linear function of yields, but could differ in magnitude since financials typicallytend to swap a greater portion of their fixed-rate debt to floating. We estimate this function to be consistent with broad anecdotal evidence regarding the
percentage of financial issuance that gets swapped. To be specific, our linear transformation causes financial issuance swapping percentages to fall to30% if corporate swapping fractions go to zero, and financial issuance swapping fractions go to 100% when corporate swapping fractions go to 100%.
Exhibit A2 presents a time series of our estimate for total swapped monthly high grade issuance, calculated by multiplying actual monthly financial andnonfinancial issuance by our estimate for swapping fractions (based on yield levels at the time of issuance).
Armed with an approach to estimate issuers’ debt swapping percentages, we can now present our model for benchmark 10-year maturity matched swapspreads. This new, high-frequency model for swap spreads is detailed in Exhibit A3. As a rule of thumb, each 10bp of FRA-OIS widening implies a2.5bp widening in 10-year spreads; $10bn in weekly Treasury gross purchases is worth 1bp of widening in spreads; $10bn in swapped high gradeissuance (estimated per our framework) would narrow swap spreads by 2.5bp; and a 10 point rise in the BIX bank stock index would cause swap spreadsto narrow 1.3bp.
Exhibit A1: Issuance swapping fraction vs. yields Exhibit A2: Estimated swapped issuance Exhibit A3: A model for 10-year swap spreads
0%
5%
10%
15%
20%
25%
2.0 2.5 3.0 3.5 4.0
Low yieldperiods
High yieldperiods
Overall fit
0
10
20
30
40
50
Nov 08 May 09 Dec 09 Jul 10 Jan 11 Aug 11
Factor Coefficient T-stat Current
Intercept 20.4 6.3
FRA-OIS 0.26 10.2 72.3
Wkly Treasury purchases by Fed 0.10 4.6 10Swapped HG issuance -0.25 -4.1 1.7
BIX -0.13 -6.5 119.0
R^2
Projected 10Y swap spread
Actual
73%
23.9
19.3
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 134/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Praveen Korapaty (1-212) 834-3092
Alberto Iglesias (1-212) 834-5116
J.P. Morgan Securities LLC
134
However, barring such a surprise, the fact remains that 1-
year ahead budget deficit expectations have remained
(and are likely to remain) very stable, moving in a
$200bn range. As a result, assuming there are no
unforeseen developments on the fiscal policy front,
deficit expectations will likely not particularly impact 10-
year swap spreads in 1H12. Similarly, given a
historically low level of negative convexity in mortgages
(Exhibit 3), mortgage hedging flows have not been, and
will likely not be, a significant driver of spreads in 2012.
The footprint of mortgage hedgers in the swaps market is
also declining for a much more obvious reason—
mandated caps on the retained portfolios of Fannie Mae
and Freddie Mac point to a decline in the swap hedging
activity of these hedgers (Exhibit 4), who havetraditionally been sizeable players in the swaps market.
In order to reflect this altered market environment that
has now become the “new normal,” we have augmented
our more traditional longer-term benchmark 10-year
swap spread model with a considerably revised tactical
model. Not only is this model more short term in nature
(out of necessity given the recent nature of the
developments discussed above), but it also addresses
newly emergent tactical forces impacting swap spreads—
Fed purchases of Treasuries at the longer end, behavioral
shifts with respect to swapping of debt issuance, and of
course the European crisis. A detailed description of thistactical model is included in the grey box on benchmark
10-year swap spreads.
Swap spread trading themes in 2012
Such a model has several implications for swap spreads
looking ahead into 1H12. First and foremost, our outlook
on the various drivers points to widening in intermediate
swap spreads heading into year-end and in 1Q12, and an
eventual narrowing (to slightly below current levels) by
mid-year 2012; this is laid out in our swap spread
forecast in a later section.
Second, swap spreads—even in the 10-year sector—willlikely remain quite vulnerable to exogenous risk from the
European crisis. Exhibit 5 presents an estimate of the
partial impact on 10-year swap spreads, if each of the
drivers moves by an amount equal to one standard
deviation of rolling 3-month changes over the past year.
Not only are FRA-OIS spreads the single most important
driver of intermediate spreads—as noted in Exhibit 5, but
they are also significantly inversely correlated to the BIX
(whose partial beta is negative), implying that the two
most significant drivers of swap spreads are likely to
move in correlated fashion, amplifying the effect of
European developments on swap spreads. Hedging swapspread positions with 35% risk (rather than the 25%
partial beta) in FRA-OIS spread positions would help
mitigate the effective risk exposure to the European
crisis.
Third, although the Fed’s Treasury purchases are likely
to be stable on average (with Operation Twist scheduled
to continue through the end of June 2012, and with QE3
unlikely to include Treasuries until after Operation Twist
is concluded), the actual schedule of Treasury purchases
Exhibit 5: Ranking swap spread drivers—FRA-OIS spreads and bank
equity prices are likely to be the dominant drivers of swap spreadsPartial impact of a 1 standard deviation move* in each of the drivers in our model for benchmark 10-year maturity matched swap spreads; bp
* Estimated as the standard deviation of rolling 3-month changes over the past year.Impact estimated as the 1 SD move in each driver multiplied by its beta in our model.
Exhibit 6: The weekly pace of Treasury purchases by the Fed can exhibitsignificant tactical variability, even as it remains stable on averageduring Operation Twist …Weekly gross purchase of Treasuries by the Fed; $bn
-3
-2
-1
0
1
2
3
FRA-OIS Fed purchases Corp. issuance BIX
0
5
10
15
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 135/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Praveen Korapaty (1-212) 834-3092
Alberto Iglesias (1-212) 834-5116
J.P. Morgan Securities LLC
135
by the Fed will likely create attractive tactical trading
opportunities in swap spreads. It is interesting to note
that there can be considerable variation in the pace of
gross Treasury purchases on a week-to-week basis; for
instance, the weekly pace of gross purchases declined
from a mid-October high of $14bn to under $5bn in just
two weeks (Exhibit 6). Such tactical swings in the
weekly pace of Fed purchases tend to impact swap
spreads (adjusted for FRA-OIS spreads). To illustrate
this tactical effect, we identified the three dates that saw
the biggest 2-week upward change in rolling weekly
purchases, as well as the three dates that saw the biggest
downward change. As seen in Exhibit 7, spreads
(adjusted for FRA-OIS) narrowed in the periods that saw
sharp declines in the pace of buying, while periods that
saw a sharp pickup in the pace of purchases were
characterized by a widening in swap spreads. Given that
the Fed will be publishing schedules at the end of each
month for its purchases in the following month, shifts in pace will be known ahead of time, and this tactical effect
can therefore be traded.
Fourth, barring considerable disruptions in primary
issuance markets due to the European crisis, typical intra-
year seasonal patterns in issuance will likely offer
attractive opportunities as well. Over the past five years,
high grade swapped issuance has tended to be far below
average in the months of February and July, while
reaching highs in the months of May and September
(Exhibit 8). The best way to position for this intra-year
seasonal pattern is to initiate swap spread narrowers inearly March. Benchmark 10-year swap spreads, adjusted
for FRA-OIS, have narrowed significantly in the
March/April period in each of the past four years, and
narrowed by a modest amount in 2007. Given the
significant narrowing in spreads in this period on average
(Exhibit 9), we would look to initiate spread narrowers
in early March.
Exhibit 8: Our estimate of swapped high grade issuance also exhibitsconsiderable intra-year seasonal patterns, with the Feb–May period
seeing the biggest rise in pace …Weekly swapped high grade issuance*, averaged by month over the past five years;$bn
* J.P. Morgan estimate, based on a model for the fraction of financial andnonfinancial debt issuance that is swapped
Exhibit 9: … causing swap spreads to be biased narrower in that periodon averageCumulative change in 10-year maturity matched swap spread adjusted for 6-monthforward FRA-OIS in the 2-month period from the first business day of March,averaged over the past four years; bp
Business days after the beginning of March
Exhibit 7: … which should make for tactical trading opportunities, sincespreads have tended to narrow in periods when the Fed’s purchase
pace declines sharply, and widen when the opposite is trueCumulative change in 10-year maturity matched swap spread, net of a 35% risk-weighted 6-month forward FRA-OIS hedge, averaged in periods when the weeklypace of Treasury purchases by the Fed fell sharply* and rose sharply**; bp
# business days around selected date* Dates used are 1/5/11, 7/15/11 and 11/2/11. ** Dates used are 11/18/10, 1/20/11and 5/11/11.
3
4
5
6
7
8
9
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
-15
-10
-5
0
5
0 5 10 15 20 25 30 35 40
-4
-3
-2
-1
0
1
2
3
-10 -8 -6 -4 -2 0 2 4
Rise in pace
Decline inpace
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 136/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Praveen Korapaty (1-212) 834-3092
Alberto Iglesias (1-212) 834-5116
J.P. Morgan Securities LLC
136
Modeling the duration exposure in variable annuities and its impact on 30-year swap spreads
Since the start of QE2 in 4Q10, long-end swap spreads have been mainly driven by three factors—the slope of the 2s/30s Treasury curve, the Fed’s gross purchases of Treasuries, and the shifting duration of the variable annuity universe.
Estimating the last of these three factors is itself a challenge. Pricing variable annuities (henceforth referred to as VAs) is an exceedingly complex undertaking, requiringthe joint modeling of long-term interest rates as well as equities; in addition, actuarial risks stemming from life insurance related guarantees, and other features make the
product path dependent, adding to the complexity. It is nonetheless important to capture the nonlinear dependencies of VA duration exposure with respect to equities andlong-term swap yields, since VAs have been a popular product for insurance companies for many years with over $1.5tn estimated to be outstanding, and the interest ratein these products can be highly nonlinear.
A common product is to pay policy holders a variable return based on the performance of the S&P 500 (or another benchmark) but provide downside protection byguaranteeing a minimum income stream for a specified time period. In rising equity markets, the product creates little risk to insurance companies, since equity returns aremerely passed through to policyholders. However, in falling equity markets, the minimum guaranteed income stream becomes more binding. Effectively, as the“moneyness” of the embedded equity put increases, insurance companies increasingly become short a fixed income annuity, requiring them to add long duration hedges asa result.
Recognizing that our objective is not to price variable annuities accurately, but merely to capture the trend in their duration exposures as well as their nonlinear
relationship with equities/yields, we devise a simpler approach. Our approximation approach is based on three principles. First, we start with the assumption that theduration of the VA universe may be approximated by a weighted combination of the durations of a set of much simpler “VA lite” instruments, which we refer to as “VAkernels.” This is not unlike “series approximation” techniques commonly used in mathematics to solve difficult problems. One example of a VA kernel is a simple producwhere a policyholder pays $100 on (say) January 1, 2007, intended to be invested in the equity market for a 10-year (fixed) horizon, with a guaranteed withdrawal amountof $5 per year for the subsequent 20 years. Several VA kernels may be created by varying the start date, the length of time of the intended equity investment, and theminimum guarantee amount. Effectively, we price the complex VA universe as a linear combination of simpler VA kernels, each of which is priced in a manner thatcaptures the nonlinearities with rates and equities. Second, we price the present value of this instrument by ignoring actuarial risks and using an option pricing framework.We use implied distributions from the swaptions market as well as long-term S&P vols and correlation estimates to calculate the price. We also use numerical tweaks tocalculate the partial exposure with respect to long-term swap rates (i.e., duration). Third, we use a calibration approach to solve for the appropriate weights on the variousVA kernels. Our calibration relies on the anecdotally known fact that VA risk exposures were significant influences on long-end swap spreads in certain periods of time,such as 4Q08; we may thus solve for non-negative coefficients that maximally explain the portion of long-end swap spread behavior not explained by other factors inthose select periods of time. In order to mitigate circularity (since we plan to use VA duration estimates to model long-end swap spreads), no data after 2008 has been usein calibration, and out-of-sample performance has been tested and found to be reasonable. The aggregate duration of the VA universe is shown in Exhibit B1; as can beseen this remains near historical highs and could worsen if yields and equities were to decline in 1Q12 on the back of continuing deterioration in Europe.
Armed with this estimate for VA duration, we may now model 30-year swap spreads as a function of the three factors outlined above. Exhibit B2 presents the statisticsfrom regressing 30-year benchmark maturity matched swap spreads versus the 2s/30s curve, cumulative gross purchases of Treasuries by the Fed (i.e., not net of front-en
sales, and since the start of QE1), and the duration of the VA universe in 20-year equivalents, over the past 15 months (i.e., since QE2 expectations became a significantfactor, resulting in altered dynamics at the long end of the Treasury curve). As a rule of thumb, each 10bp flattening in the 2s/30s Treasury curve would widen 30-year spreads by 2.8bp, a fall in VA duration by $10bn 20-year equivalents would cause a 5bp widening in spreads, and the cumulative effect of the Fed’s long-end purchasesunder Operation Twist ($400bn in all) should be about 3.6bp (not including indirect effects due to the impact of such purchases on the curve). As seen in Exhibit B3, thismodel has been successful in fitting the observed behavior of 30-year swap spreads.
Exhibit B1: Estimated duration of VA universe Exhibit B2: 30-year swap spread model Exhibit B3: Actual spreads versus model fair value
100
125
150
175
200
225
250
275
Jan 07 May 08 Sep 09 Feb 11
Factor Coefficient T-stat Current
Intercept 160.7 18.4
2s/30s Treasury curve, % -27.7 -20.7 2.897
VA aggregate duration; $bn 20s -0.5 -24.1 240.4
Cum gross Tsy purchases by Fed; $bn 0.009 17.0 1200.5
R^2
Projected 30Y swap spread
Actual
-25.2
-27.1
74%-45
-40
-35
-30
-25
-20
-15
Nov 10 May 11 Nov 11
30-year maturity matched swap spread; bpModel
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 137/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Praveen Korapaty (1-212) 834-3092
Alberto Iglesias (1-212) 834-5116
J.P. Morgan Securities LLC
137
Long end swap spreads: does VA hedging still
matter?
Swap spreads at the long end of the curve are broadly at
the mercy of two large technical forces—Fed purchases
of Treasuries (particularly at the long end of the curve),
and the shifting duration hedging needs of life insurance
companies’ variable annuity portfolios (VA duration, for
short). Although the significance of the latter has been
diminishing, as insurance companies broaden their use of
Treasury-based derivatives and/or cash Treasuries, we
believe it is too soon to dismiss VA duration needs as a
driver of long-end swap spreads.
Estimating the duration of the VA universe is a complex
undertaking, and we have outlined an approximate
scheme for doing this in previous research (see Interest
Rate Risk in Variable Annuities, J.P. Morgan Research
Note, September 28, 2011). A short summary of our approach to estimating VA duration, as well as its use in
our revised model for 30-year benchmark maturity
matched swap spreads, is presented in the grey box on
long-end swap spreads.
Several points are interesting to note with respect to our
long-end spread model. First, the partial sensitivity of
long-end spreads to VA duration has indeed been
declining in magnitude (Exhibit 10). That said, reports
of its demise as a factor impacting long-end spreads are
exaggerated, at least for now, and it is equally
important not to overstate the extent of this decline.
To all appearances, it would seem that VA hedging
has become insignificant, or even a contrarian
indicator of long-end swap spreads. Indeed, in recent
months, long-end spreads have widened as VA
duration has risen (Exhibit 11). However,
Exhibit 12: … masks a more subtle message—the recent strong equity-rate correlation has caused the 2s/30s curve to become negativelycorrelated with VA duration, producing offsetting impacts on long-endspreadsRolling 3-month correlation between weekly changes in the S&P and 30-year swapyields, versus the effective beta* of 30-year swap spreads versus VA duration;
bp per $bn 20-year equivalents
* Calculated as partial beta of 30-year spreads with respect to VA duration (asdetailed in the footnote to exhibit 10), plus the partial beta of long end spreads withrespect to the 2s/30s curve (from the J.P. Morgan 30-year spread model) times therecent beta of the 2s/30s Treasury curve with respect to VA duration. Recent betacalculated over rolling 3-month periods.
Exhibit 11: Appearances can be deceptive—the seemingly wrong-way
correlation between VA duration and long end swap spreads …30-year maturity matched swap spread versus estimated VA duration*$bn 20-year equivalents bp
* J.P. Morgan estimate
Exhibit 10: The partial sensitivity of long-end swap spreads to VA
duration has indeed been declining, but reports of its demise (as afactor impacting spreads) are as yet exaggeratedRolling partial beta of 30-year maturity matched swap spreads versus VA duration*;bp per $bn 20-year equivalents
* Based on rolling 15-month regressions of 30-year benchmark maturity matchedswap spreads versus VA duration, the 2s/30s Treasury curve and cumulative grossTreasury purchases by the Fed. Regressions are over 15 months because our current model for 30-year swap spreads is designed to cover the period from
August 2010, when QE2 became a significant factor affecting long end spreads.
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
-0.50
-0.45
-0.40
-0.35
-0.30
-0.25
-0.20
-0.15
-0.10
-0.05
0.00
0.05
May 11 Jul 11 Sep 11 Nov 11
S&P/30Y yield correlation
Effective VA beta
220
225
230
235
240
245
250
255
260
-38
-36
-34
-32
-30
-28
-26
-24
-22
-20
-18
Sep 11 Oct 11 Nov 11
VA duration30-year swap spread
-0.75
-0.70
-0.65
-0.60
-0.55
-0.50
-0.45
Feb 11 May 11 Aug 11 Nov 11
VA beta
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 138/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Praveen Korapaty (1-212) 834-3092
Alberto Iglesias (1-212) 834-5116
J.P. Morgan Securities LLC
138
appearances can be deceiving, and this masks a more
subtle message, which is that in recent months VA
duration and the 2s/30s curve have become much more
negatively correlated than before, causing their impacts
to offset each other; moreover, this negative correlation
is itself a consequence of the elevated positive correlation
between long-end yields and the S&P. In other words,
increased equity-rate correlation has caused VA duration
to rise sharply as long-end yields fall and the 2s/30s
curve flattens. As a result, the partial impact of increased
VA duration needs has been mitigated by the counter-
directional exposure of long-end spreads to the curve
(Exhibit 12).
Such correlations could persist for some time, producinglimited net impact on long-end swap spreads. However,
this should be seen as a happy coincidence, and is
unlikely to persist in the medium term. Should equities
and rates decouple in 1H12, as indeed we expect them to,
such a canceling effect will be less pronounced, and the
impact of VA duration on long-end spreads will likely be
apparent once again. Therefore, we continue to rely on a
fair-value framework for long end spreads that includes
VA duration as well as the 2s/30s curve.
Front-end spreads
At the front end of the curve, swap spreads have become pure plays on the European crisis, exhibiting strong
correlation to FRA-OIS spreads but little else. This will
likely remain the case in 1H12, and opportunities in this
sector will likely revolve around trading the mispricing
in front-end spreads relative to FRA-OIS on a hedged
basis.
Thus, a view on FRA-OIS is essential to determine the
likely evolution on front-end swap spreads in 2012. To
this end, we present a simple model that attempts to
explain the behavior of FRA-OIS differentials in recent
months. USD FRA-OIS spreads have been driven by two
factors recently—the EUR/USD OIS FX basis, and semi- peripheral European sovereign CDS spreads. (Although
the commonly quoted EUR/USD FX basis is with respect
to a Euribor/Libor basis swap, it is more useful to adjust
this quoted spread to be with respect to an EONIA/OIS
basis swap. We do this because (i) EONIA and OIS rates
are more reflective of the marginal cost of funds for
banks in the Eurozone and the US, and (ii) the existence
of the Fed’s dollar swap lines at a 100bp penalty bounds
this EUR/USD OIS FX basis at -100bp. For a more
detailed discussion of cross-currency basis swaps, see Decoding the FX basis market for signs of $ funding
stress, Pavan Wadhwa et al , August 23, 2011.
A more negative basis implies a higher USD borrowing
rate at which a European borrower would be indifferent
to borrowing in USD versus borrowing in euros and
swapping to USD, and may thus be thought of as a
fundamental funding market factor. We also use semi-
peripheral spreads as a second factor to capture the broad
correlations between credit spreads and metrics of
European stress.
Exhibit 13 presents statistics regarding such a model,
and also uses the model parameters to project 3-month
forward FRA-OIS spreads by year end and by mid-year
2012. To make these projections, we are assuming that in
the near term, conditions deteriorate further into year end
and in 1Q12, and the EUR/USD OIS FX basis reaches -
100bp, which should act as a near-term floor thanks to
unlimited central bank funding. We also make
projections for European stress metrics based on
estimates from our European rates strategists, and look
for a near-term widening in our measure of semi-
Exhibit 13: Our model for FRA-OIS spreads suggests that FRA-OIS
could widen into year-end, but narrow going into mid-year Statistics from regressing 3-month forward constant maturity FRA-OIS differential (bp)versus the 3-month EUR/USD FX OIS basis (bp) and semi-peripheral sovereign CDSspreads* (bp), and projections for FRA-OIS at end of 4Q11 and 1H12 based onassumptions for the drivers.
* Average of Italy, France and Spain 5Y CDS spreads. Regression based on 6 monthsof history, and are as of COB 11/18/2011. Shaded values represent projections.
Exhibit 14: Swap spread forecastJ.P. Morgan projections for maturity matched swap spreads in various benchmarksectors; bp
* Current values as of COB 11/18/2011. Projections assume 3-month forward FRA-OIS spreads evolve according to the trajectory shown in the previous exhibit.
Factor Coefficient T-stat Current 1Q12 1H12
Intercept 4.6 3.8
EUR/USD FX OIS basis -0.43 -11.9 -73 -100 -40
Semiperipheral spreads 0.06 7.7 406 450 475
R^2
FRA-OIS fair value 60 74 50
Actual 63
95%
Current 1Q12 1H12
2Y swap spread 49 53 38
5Y swap spread 40 45 34
10Y swap spread 17 27 18
30Y swap spread -29 -23 -23
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 139/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Praveen Korapaty (1-212) 834-3092
Alberto Iglesias (1-212) 834-5116
J.P. Morgan Securities LLC
139
peripheral spreads. Finally, we assume that policy actions
to be enacted sometime in 1H12 will include concerted
actions by central banks (alluded to by Rosengren lastweek) such as dropping the penalty on the Fed’s dollar
swap lines to 50bp from 100bp. Under such a scenario,
FRA-OIS spreads may be expected to widen to nearly
75bp in the near term, but narrow back to 50bp by mid-
year 2012 despite a steady widening in semi-peripheral
spreads.
Given these projections for FRA-OIS spreads, which is a
key determinant of spreads in the 2- to 10-year sector,
and based on our projections for other inputs, our
estimates for maturity matched swap spreads in various
sectors as of the end of 1Q12 and 1H12 are shown in
Exhibit 14. Broadly speaking, we see the potential for swap spreads to widen going into 1Q12; however, since
front-end spreads are generally wide to fair value
currently, this upside is greatest in the 10-year sector.
Heading into 2Q12, we would look for swap spreads to
narrow, a view that is premised upon favorable policy
actions in Europe as discussed earlier. In the 30-year
sector, we expect spreads to widen initially and then
remain largely range-bound.
Front-end spread… options?
The correlation between front-end swap spreads and
FRA-OIS spreads is a reflection of a broader
phenomenon at the front end of the yield curve.
Subsequent to the Fed announcing its commitment to low
rates until at least mid-2013, OIS rates and Treasury
yields have fallen and the OIS curve has become both
flat and sticky; as a result, the volatility in front-end swap
spreads is now almost entirely due to the volatility of the
swap rate itself. Put differently, bearish positions at the
front end are effectively the same as front-end swap
spread wideners. This is strikingly evident in Exhibit 15,
which shows that 2-, 3- and 5-year swap spreads have all
become highly correlated to front-end swap rates, with
little differences between the three sectors.
This is quite useful—options on front-end yields do exist,and are effectively options on front-end swap spreads.
Short expiry payer swaptions and/or payer swaption
spreads on front-end swap yields (such as 3Mx1Y payer
swaptions, for instance) can be used to create asymmetric
exposure to wider front-end swap spreads. As we have
noted often in recent weeks, given their limited-risk
nature, such trades can offer better risk-reward than
outright swap spread wideners or narrowers. Indeed, in
the past month, we have recommended payer swaption
spreads as proxies for front-end swap spread wideners, as
well as receiver swaption spreads in place of spreadnarrowers (see US Fixed Income Markets Weekly,
Interest Rate Derivatives, dated September 23, 2011 and
November 18, 2011, respectively).
This state of affairs is likely to persist into next year.
Going into 2012, a key part of our strategy with respect
to front-end swap spreads will be to trade them using
payer and receiver swaption spreads as proxies.
Exhibit 16: Sectors where the Fed is actively buying Treasuries are likelyto see a convergence in yields between issues of similar durationDispersion* of bond yields in the 3- to 6-year sector (where the Fed is not buying)and the 20- to 30-year sector (where the Fed is buying); bp
* Calculated as the root mean squared y ield error of bonds in that sector relative toan overall par Treasury fitted curve
Exhibit 15: Swap spreads across much of the front end have become
highly correlated to front-end swap yields with little distinction acrosssectorsMaturity-matched swap spreads in 2-, 3- and 5-year sectors versus the 3Mx1Yforward swap yield; past three months; bp
3Mx1Y forward swap yield; %
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
Sep 11 Oct 11 Nov 11
3- to 6-year 20- to 30-year
20
25
30
35
40
45
0.4 0.45 0.5 0.55 0.6 0.65 0.7
2Y 5Y 3Y
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 140/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Praveen Korapaty (1-212) 834-3092
Alberto Iglesias (1-212) 834-5116
J.P. Morgan Securities LLC
140
O-spreads and spread curve relative value
With Operation Twist already underway and slated tocontinue until the end of 1H12, relative value spread
switch trades will likely prove to be a profitable trading
strategy as we head into 2012. The logic underlying this
is quite simple; experience from the Fed’s Treasury
purchases under QE1 as well as QE2 suggest that the
Fed’s issue selections are based on yield differentials of
various bonds with respect to a fitted curve—i.e., yield
errors (see Treasuries). Thus, it is reasonable to expect
convergence between rich and cheap issues, measured
through metrics that mirror the Fed’s, particularly in
sectors where the Fed is actively purchasing Treasuries.
Such convergence is indeed already evident since the
commencing of Operation Twist; yield errors are
converging at the longer end of the curve, producing
little dispersion across bonds in those sectors, while a
similar fall in dispersion is not evident in the shorter end
(Exhibit 16).
Of course, yield errors are not directly tradable, but
switch trades on a maturity matched swap spread basis
are a close proxy. Such trades should prove attractive in
periods and sectors characterized by active open market
Treasury purchases by the Fed. But what relative value
metrics are suitable for sectors of the curve or periods of
time where the Fed is not actively purchasing Treasuries?
Historically, asset swap spreads (some times called a
proceeds asset swap spread) have served as a useful
metric of value. The logic behind this measure is this:
suppose that an investor puts up $100 to purchase a
Treasury bond regardless of its actual full price, finances
(or lends) the difference between the bond’s full price
and par via a swap where the investor pays the bond’s
coupon cash flows versus receiving Libor plus a spread
on the floating leg. This spread is called the proceeds
asset swap spread (or simply, asset swap spread), and is
a clean metric for comparing bonds with different
coupons and prices (see Exhibit 17 for an illustration).
Traditionally, these asset swap spreads were solved for
by requiring that cash flows from the associated swap,
discounted using swap curve zero rates, should produce
an NPV that offsets the amount being financed (or lent).That is, the swap curve served as both the reference
curve as well as the discount curve. In principle,
however, the discount curve should depend on the
funding curve for the type of assets that would be used to
collateralize the mark-to-market variations in the swap.
Since US dollar interest rate swaps are commonly
Exhibit 17: An illustration of the roceeds asset swap spread and O-
spread A schematic illustration of the proceeds asset swap spread for a given bond
Exhibit 18: Might is right—the Fed’s preferred metric displaces the market’s value metric in sectors and times where the Fed is actively buying Treasuries,but O-spreads are likely to be the best value metric otherwiseStatistics regarding the back-testing of trading strategies* using three metrics of value: yield error, asset swap spreads (based on swap curve discounting) and o-spreads (or assetswap spreads to the swap curve, calculated using OIS discounting)
* Trading strategy is as follows: every day, the cheapest and richest bonds in each sector are identified, based on each of the three value metrics. The rich bond is then sold and thecheap bond is bought, on a maturity matched swap spread switch basis. Trades are held for a month and unwound thereafter. We back-tested these strategies from the start of QE2through mid-November.
Dirty Bond
price
For bond: 8% Nov 2021
Spread: -12.9bp under Libor discounting
-17.4bp under OIS discounting
Investor Swap counterparty
Bond coupons
3M Libor + spread
Dirty price minus par
During Jul - Oper. During Jul - Oper. During Jul - Oper. During Jul - Oper. During Jul - Oper. During Jul - Oper.
Overall QE2 Sep Twist Overall QE2 Sep Twist Overall QE2 Sep Twist Overall QE2 Sep Twist Overall QE2 Sep Twist Overall QE2 Sep Twist
# trades 230 146 64 20 230 146 64 20 230 146 64 20 230 146 64 20 230 146 64 20 230 146 64 20
# wins 202 135 53 14 196 119 60 17 208 125 63 20 139 103 18 18 137 86 43 8 139 87 42 10
# losers 28 11 11 6 34 27 4 3 22 21 1 0 91 43 46 2 93 60 21 12 91 59 22 10
Hit ratio 88% 92% 83% 70% 85% 82% 94% 85% 90% 86% 98% 100% 60% 71% 28% 90% 60% 59% 67% 40% 60% 60% 66% 50%
Avg Gain 4.2 4.2 4.7 2 4 3.7 4.9 3.1 4.7 3.9 6.6 3.1 2.5 2.9 1.5 1.5 3.7 3.2 5.1 1.2 4 3.8 5.2 0.9
Avg Loss -2.4 -0.7 -4.7 -1.3 -1.2 -1.4 -0.7 -0.5 -1.1 -1.1 -0.8 N/A -1.6 -1.6 -1.6 -0.4 -3.1 -3.2 -3.6 -1.9 -3.5 -3.4 -4.2 -2.6
Overall avg 3.4 3.8 3.1 1 3.2 2.7 4.5 2.5 4.1 3.2 6.5 3.1 0.9 1.5 -0.7 1.3 0.9 0.6 2.2 -0.7 1 0.9 1.9 -0.8
7- to 15-year sector
Value metric: ASW Value metric: OSWValue metric: Yield error Value metric: Yield error Value metric: ASW Value metric: OSW
2- to 5-year sector
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 141/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Praveen Korapaty (1-212) 834-3092
Alberto Iglesias (1-212) 834-5116
J.P. Morgan Securities LLC
141
collateralized by Treasuries, it would be more correct to
use OIS-curve based zero rates to discount future cash
flows (since Treasuries fund at GC rates, which arerather close to OIS rates, and since a well developed OIS
curve exists). With regulatory reform forcing almost all
swaps to be collateralized going forward, a more correct
way to define a proceeds asset swap spread would be to
still define a spread versus the Libor curve (as in the
previous case), but using the OIS curve for discounting
cash flows. We introduce new terminology to reference
this new measure of value, and will refer to it as
proceeds asset swap o-spread (or simply, o-spread )
heretofore in our research. O-spreads will be understood
as a reference to spreads on the floating leg of a swap,
where the spread has been solved for using OIS zero
rates for discounting purposes.
Which of these metrics works well in practice? To
answer this question, we back-tested trading strategies
based on three different value metrics, in two sectors of
the curve, over the past year. Every day, we identified the
richest and cheapest bonds in each of the two sectors,
based on each of three value metrics (yield error, asset
swap spread and o-spread). We then initiated a long in
the cheap bond and a short in the rich bond, but on a
maturity-matched swap spread basis (for the purpose of
lessening P/L noise due to curve movements). All trades
were held for a month and unwound thereafter. The twocurve sectors we chose are the 7- to 15-year sector (a
beneficiary of Fed purchases during QE2 as well as
Operation Twist), and the 2- to 5-year sector (which
benefited from Fed purchases during QE2, but not during
Operation Twist). Statistics regarding our results are
presented in Exhibit 18.
Three points are interesting to note from our results.
First, yield errors are the best metric for sectors/periods
where Fed purchases are a factor. Put differently, Fed
purchases distort normal value metrics, and essentially
impose the Fed’s value metric in its place. This is evident
from the fact that hit ratios were highest in both curvesectors during QE2 when using yield errors as the value
metric. More recently under Operation Twist, using yield
errors once again results in the highest hit ratios in the 7-
to 15-year sector.
Second, when Fed purchases cease to be a factor, the
resulting distortion effect on value metrics is quick to
dissipate. This is seen in the fact that during the brief 3-
month hiatus in Fed purchases in the 7- to 15-year
sectors, yield errors were a poor indicator for
convergence trades, underperforming asset swap spreads
as well as o-spreads as a value metric. This is also
evident in the fact that since the ending of Fed purchases
in the 2- to 5-year sector of the curve, yield error has
underperformed the other two value metrics.
Third, there is some evidence to suggest that o-spreads—
rather than traditional asset swap spreads—are a better
value metric when Fed purchases are not a factor. This is
seen by looking at the performance of the three differentvalue metrics in the 2- to 5-year sector, which has been
free of Fed purchase distortions since the end of QE2.
Looking ahead to 2012, we anticipate that the general
trend in these results will continue to hold—convergence
trades at the long end of the curve will continue to be
attractive when chosen using yield errors as a measure of
value; for other sectors of the curve, convergence trades
indicated by o-spreads as a value metric will likely prove
profitable. With Fed purchases likely to be a mainstay of
Exhibit 19: The yield curve has flattened significantly this year in all
sectorsYear-to-date statistics regarding the slope of the curve in various sectors; bp
* As of 11/17/2011 COB
Exhibit 20: Investors reaching for carry will only find it further out on thecurve
3-month carry and slide in various sectors of the yield curve; bp
* As of 11/17/2011 COB
Current Start Average High Low YTD chg
1s/2s 4.6 35.8 25.5 56.4 0.4 -31.2
2s/3s 9.9 48.3 36.1 59.8 9.9 -38.4
3s/5s 42.7 90.8 76.8 98.0 42.7 -48.1
5s/7s 42.6 65.5 59.8 68.2 40.8 -22.9
7s/10s 40.1 57.0 54.7 63.2 36.9 -17.0
10s/30s 55.1 76.7 77.0 93.5 51.3 -21.7
2011 YTD statistics
0
2
4
6
8
10
12
14
ED2 1Y ED6 2Y ED10 3Y 5Y 7Y 10Y 30Y
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 142/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Praveen Korapaty (1-212) 834-3092
Alberto Iglesias (1-212) 834-5116
J.P. Morgan Securities LLC
142
markets in 2012, such relative value switch trades will be
a key part of our market strategy in the year ahead.
Swap yield curve
The yield curve has been on the move in 2011, flattening
significantly in all sectors of the curve (Exhibit 19). The
flattening mostly occurred in 3Q, leading into and after
the August FOMC meeting, as a number of factorsconspired to flatten the curve. At the very front end of
the curve, the biggest driver has been the Fed’s
conditional commitment, announced at the August
FOMC meeting, to hold the funds rate near current levels
until mid-2013. Although falling short of a firm
commitment, the Fed’s indication that the current
economic outlook would likely be consistent with a low
funds rate for at least two years had a significant
impact—front-end term premium collapsed, OIS rates
fell, the OIS curve flattened sharply, and yield levels
declined sharply across much of the front end of the
curve, out to the 5-year sector. The very front end of theswaps curve has been additionally biased flatter by the
widening in Libor/OIS spreads. Further out the curve, the
key driver has been the Fed’s Operation Twist.
Much of this backdrop is likely to stay unchanged as we
head into next year. Operation Twist will continue to
dominate rates markets through the end of 1H12, and the
Fed’s mid-2013 conditional commitment is likely to
remain valid for quite some time, which should keep the
very front end of the curve very flat for the foreseeable
future. In addition, the overhang from Europe will likely
persist, weighing on funding spreads and thus the front
end of the swap curve.
The implications for yield curve trading strategies in
2012 are three-fold. First, investors will need to reach
further along the curve for carry, as the flatness of the
very front end has erased carry in the front and red
Eurodollar sectors (Exhibit 20). Second, the very frontend of the swap curve—1-year swaps and front
Eurodollars—will likely be the least attractive sector to
earn carry, since longs in these sectors are vulnerable to a
worsening of the European crisis. Carry in this sector is
more optical than real, as spot Libor rates continue to
rise, making forwards more likely to be realized than
spot. Ironically, wider FRA-OIS spreads can lead to
better carry/slide optically, making front-end longs look
more attractive, even as rising spot Libor rates means
that carry in this sector is merely a mirage.
Empirical evidence supports this view. Exhibit 21 shows
that the usefulness of carry as an indicator (measured asthe beta between the 1-month P/L on a cross section of
Eurodollar sector curve and butterfly trades and the ex-
ante carry/slide on those trades) has declined as spot
Libor has risen; in other words, slide on the Eurodollar
curve ceases to predict future spot yields when funding
pressures are at work in pressuring Libor higher.
Thus, going into 1H12, we will not look to earn carry at
the front end of the Libor swap curve (unless of course if
policy actions were to cause a more sanguine outlook on
Exhibit 22: Yield curves are now highly directional with yield levels
thanks to ultra low—and sticky—front-end yieldsRolling 3-month beta between weekly changes in various yield curves and weeklychanges in the corresponding longer end y ield;
Exhibit 21: At the front end of the swap curve, carry is now a mirage—
rising spot Libor rates at the front end has decreased the usefulnessof carry as a measure of the attractiveness of Eurodollar trades1-week moving average of carry-beta* versus spot 3-month Libor;
%, inverted axis
* Beta between the 1-month P/L on a c ross section of Eurodollar sector curve andbutterflies versus the 1-month ago slide on those trades
-0.2
0.0
0.2
0.4
0.6
0.8
May 11 Jul 11 Aug 11 Oct 11
2s/5s
5s/10s
0.28
0.30
0.32
0.34
0.36
0.38
0.40
0.42-1
0
1
2
3
09 Aug 19 Aug 29 Aug 08 Sep 18 Sep 28 Sep 08 Oct
3-month Libor (inverted)
Carry beta
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 143/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Praveen Korapaty (1-212) 834-3092
Alberto Iglesias (1-212) 834-5116
J.P. Morgan Securities LLC
143
the European crisis at some point). Indeed, as we noted
in an earlier section, rather than owning the front end of
the Eurodollar curve for the sake of carry, being short thesector via short-expiry payer swaptions offers the most
asymmetric way of positioning for deterioration in
Europe.
Third, the yield curve is likely to exhibit a very high
degree of directional exposure to yield levels, an artifact
of ultra-low yields in the less-than-5-year maturity
sector. This is seen in the elevated positive betas between
various yield curves and yield levels (Exhibit 22).
Hedge curve steepeners with Eurodollars
Based on these observations, three types of tradingthemes are likely to prove attractive in 2012. First,
forward curve steepeners (or spot curve steepeners with
the front end being anchored in sectors that offer the best
carry) hedged with shorts in the Eurodollar sector are
likely to be attractive ways to earn carry while also
protecting against exogenous risk from Europe. To
identify the best trades, we identified the potential reward
and risk in numerous such combinations. We estimated
the risk weight on the Eurodollar hedge as the 1-year
average of the rolling 3-month beta between weekly
changes in each curve and the selected Eurodollar yield.
Based on this hedge ratio, we then estimate the potential
gain as the 3-month carry/slide, plus 50% of the potentialupside, which in turn is defined as the average level of
the weighted yield spread since the start of 2010 minus
its current level. In order to properly assess risk, we
focus on a downside-vol measure (in addition to a more
typical standard deviation based volatility estimate).
Specifically, we estimate the downside risk by assuming
that semi-peripheral spreads (an average of Spain, France
and Italian sovereign CDS spreads) widen by 100bp, and
estimate the likely move in the weighted spread, and we
additionally assume a 1-standard error downward move
from this relationship. Thus, our downside risk measure
reflects an attempt to capture the effect of deteriorating
conditions in Europe. Finally, we calculate an efficiency
ratio, which is the ratio of potential gain to the downside
risk.
Exhibit 23 presents a visual snapshot of results, and
charts the potential gain versus the magnitude of the
downside risk, while Exhibit 24 presents more detailed
statistics regarding numerous such trades, ranked
according to their efficiency ratios. As can be seen,
10s/30s and 5s/30s steepeners, hedged with shorts in red
Exhibit 24: Detailed statistics regarding the potential reward anddownside risk in yield curve steepeners hedged with shorts in red
EurodollarsBeta of various yield curves against Eurodollar yields*, the current weighted yieldspread corresponding to each trade**, 3-month carry and slide, potential upside***, 3-month standard deviation of weekly changes, downside risk**** and efficiencyratio***** for various yield curve s teepener trades hedged with red Eurodollar shorts.
* 1-year average of 3-month beta of weekly changes in the curve versus weeklychange in ED yield.** Current value as of 11/14/2011. *** See footnote in previous exhibit. **** Seefootnote in previous exhibit.***** Efficiency ratio defined as the ratio of 3-month carry plus 50% of upside todownside volatility.
Exhibit 23: Steepeners such as 10s/30s and 5s/30s, hedged with shorts
in red Eurodollars, are attractive ways to earn carry and position for yield curve normalization, while hedging European exogenous risksPotential gain* versus the magnitude of downside risk** for various curve steepenershedged with short Eurodollars; bp
Downside risk; bp* Upside defined as average of the weighted yield spread corresponding to eachtrade since 1/1/2010, minus its current value as of 11/14/2011. Potential gain definedas carry/slide plus one half of the upside.** Downside risk estimated as the expected decline in the weighted yield spread for a100bp rise in semi-peripheral spreads, plus a 1-standard error downward move inaddition. Beta with respect to semi-peripheral spreads estimated based on aregression of monthly changes since 1/1/2010.
Curve ED Beta
Spread
(%)
Carry
(bp)
Upside
(bp)
Vol
(bp)
Downside
vol (bp)
Efficiency
ratio
3M fwd 10s/30s 6th -0.24 0.73 4.4 26.5 11.0 20.8 0.85
1Y fwd 10s/30s 6th -0.25 0.61 4.1 22.2 9.2 18.4 0.83
6M fwd 10s/30s 6th -0.24 0.69 4.0 25.0 10.3 20.0 0.82
2Y fwd 10s/30s 6th -0.24 0.42 4.2 17.3 7.5 15.9 0.81
2Y fwd 5s/30s 6th -0.47 1.16 9.0 32.3 16.7 32.1 0.78
3M fwd 5s/30s 6th -0.37 1.71 7.7 51.6 25.5 43.2 0.78
1Y fwd 5s/30s 6th -0.46 1.53 8.0 42.1 21.5 38.0 0.77
3Y fwd 5s/30s 6th -0.39 0.75 7.5 25.8 12.9 26.9 0.76
3Y fwd 5s/10s 6th -0.20 0.53 4.0 12.1 7.9 14.4 0.702Y fwd 5s/10s 6th -0.23 0.74 4.8 15.0 10.4 17.6 0.69
3Y fwd 3s/10s 6th -0.32 0.97 7.9 18.2 13.8 25.5 0.67
3Y fwd 2s/10s 6th -0.40 1.30 9.6 22.9 17.8 32.5 0.65
2Y fwd 3s/10s 6th -0.36 1.34 7.4 24.4 18.4 30.6 0.64
1Y fwd 5s/10s 6th -0.21 0.92 3.9 19.9 13.3 21.6 0.64
2Y fwd 5s/10s 5th -0.26 0.76 5.0 11.2 10.0 16.8 0.63
2Y fwd 2s/10s 6th -0.42 1.71 7.2 32.2 24.0 37.6 0.62
5
10
15
20
25
30
35
10 20 30 40 50
10s/30s
5s/30s
2s/10s and 3s/10s
5s/10s
1s/5s, 2s/5s and 3s/7s
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 144/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Praveen Korapaty (1-212) 834-3092
Alberto Iglesias (1-212) 834-5116
J.P. Morgan Securities LLC
144
Eurodollars are among the best such trades; variations
involving 3- and 2-year forward 2s/10s steepeners (again
hedged with red Eurodollars) are also attractive, but
slightly less so. As we head towards 2012, we would
look to a framework such as this to identify the best ways
to position for a normalizing in yield curves.
Take advantage of non-linearities in the yield curve
due to sticky front-end yields
Our second yield curve trading theme takes advantage of
yield curve nonlinearity in low yield regimes. One
consequence of yields being very close to the zero bound
at the front end of the curve is that yield curve butterflies
will likely exhibit nonlinear behavior with respect to
yield levels. This is strikingly evident by looking at—for
instance—a curve neutral 2s/10s/30s butterfly (25:75
weights on the wings) versus 2-year yields (Exhibit 25).
This suggests that curve-neutral belly-richening
butterflies locally hedged for exposure to front-end yields
should experience empirically convex upside, in either a
front-end rally or a sell-off. Moreover, since the
underlying reason for this nonlinearity—the stickiness of
yields when near the zero boundary—is also affecting
swaption implied volatility skews, we may use
information from swaption skews to infer which
butterflies are likely to exhibit the most nonlinear
behavior (see Implied Weights in Yield Curve Relative
Value Trades, J.P. Morgan Research note, May 25, 2011,
for a detailed discussion of our framework). Exhibit 26
outlines several butterflies that are likely to exhibit the
most asymmetric behavior with respect to rates under
different rate scenarios, thanks to such nonlinearities.
Some of these trades have positive (or very slightly
negative) carry, which also makes them attractive ways
of positioning for a large, unanticipated, move higher infront-end yields.
Use payer swaptions to position for an eventual
normalization in rates
Third, with yields extremely low at the front end, and
with steep implied volatility skews, payer swaption
spreads (1:1 weighted) now offer attractive reward-to-
premium ratios (Exhibit 27), and are attractive as
limited-risk ways of positioning for an eventual
normalization in rates towards higher levels, should that
occur in 2Q12 as our rates strategists expect. Although
we expect yield levels to fall in the early part of 2012, we
do expect such a move to higher rates in the middle of
next year (see Treasuries). Given the highly uncertain
nature of policy developments in Europe, and the
resulting risks in outright short-duration positions, we
view 1:1 weighted limited-risk payer swaption spreads as
the preferred way to position for the normalization in
yield levels that we expect as we head towards the
middle of next year. Moreover, with yields expected to
drift initially lower in 1Q12, we would expect to find
even better entry levels for such trades in the early part of
next year, and would look to initiate payer swaption
Exhibit 25: Yield curve butterflies begin to exhibit non-linear behavior as front end yields become sticky near the zero bound
A curve neutral (25:75 weighted) 2s/10s/30s swap yield butterfly spread, versus 2-year swap yield; %
2-year swap yield; %
Exhibit 26: The nonlinearity implied by the skew is pronounced for various benchmark butterfliesProjected evolution of left- and right-leg weights for several butterflies under various left leg swap yield shifts* and projected upside** in each scenario and carry (bp).
* Our methodology is as follows. For each left leg rate shift, we calculate the right leg rate shift based on the current ratio of ATM vols. We then combine the left and right rateshifts with the current implied weights to calculate a belly rate shift for that scenario. Third, we use the current implied vol skews to estimate the change in then-ATM implied volsfor those rate shifts in each leg. Last, armed with estimates for then-ATM implieds in all three legs, we calculate new implied left and right weights.** The projected upside is estimated as sum over left and right leg values of half the change in weight from the un-shifted scenario times the corresponding shift in rate.
-0.3
-0.2
-0.1
0.0
0.1
0.2
0.4 0.5 0.6 0.7 0.8 0.9 1 1.1
Projected Projected Projected
upside from upside from upside from
Butterfly -25bp 0bp +25bp +75bp -25bp 0bp +25bp +75bp -25bp scenario +25bp scenario +75bp scenario 3M Carry
3s/7s/15s 1.32 1.03 0.85 0.69 0.34 0.36 0.34 0.30 3.3 2.1 10.4 1.22s/7s/15s 0.86 0.60 0.46 0.35 0.62 0.64 0.62 0.54 2.7 1.3 4.8 -3.7
3s/10s/30s 0.78 0.60 0.48 0.37 0.63 0.66 0.63 0.55 1.5 0.8 3.1 -0.4
5s/7s/15s 1.07 1.02 0.99 0.96 0.15 0.15 0.15 0.13 0.6 0.3 1.7 2.6
Left weight given a shift in left rate of Right weight given a shift in left rate of
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 145/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Praveen Korapaty (1-212) 834-3092
Alberto Iglesias (1-212) 834-5116
J.P. Morgan Securities LLC
145
spreads on the back of declines in yield levels.
Conditional trades … with a twist
The fourth yield curve trading theme that we will attempt
to take advantage of in 1H12 involves exploiting the high
degree of directionality between yield curves and yield
levels. With yield curves anchored at the front end
expected to be highly directional with yield levels,
conditional curve trades (i.e., conditional flatteners in a
rally and conditional steepeners in a sell-off) will likely be an ongoing trading theme in 1H12.
Options markets are pricing in this directional nature of
yield curves to some extent, but not fully. Although
swaption implied volatility on short tails (such as 2s) is
well below implied volatility on longer tails (such as
10s), the implied beta between the curve and long end
yields is well below empirical betas (Exhibit 28). This
allows for the construction of premium-neutral weighted
conditional curve trades at the current time; however,
with yield curves expected to remain highly directional
for quite some time, we would not expect the options
markets to persistently allow for attractive entry levels on
such trades.
How else can one exploit yield curve directionality oncethe swaptions markets have priced in the relative vol
differentials fairly? To answer this question, we explore a
variant of the “usual” conditional curve trades. Our
variant is based on two observations: first, the 10s/30s
curve is highly correlated (in magnitude, but with a
negative beta) to front-end yields, and second, implied
volatility on the 10s/30s curve in the YCSO market has
tended to trade rich. This suggests that YCSOs on the
Exhibit 28: Yield curves have become highly directional with rates, and
options markets are pricing this in to some extent but not yet fullyImplied* and empirical** betas between the 3-month forward 2s/10s curve and3Mx10Y forward swap yield;
*Implied beta calculated as 1-(3Mx2Y / 3Mx10Y) implied bp vol ratio. ** Empiricalbeta calculated as the 3-month beta between weekly changes in the 3M fwd 2s/10scurve and 3Mx10Y forward swap yield.
Exhibit 27: Payer swap spreads (1:1 weighted) are attractive as limited-
risk ways of positioning for an upward drift in rates by mid-year Statistics regarding 1:1 weighted payer swaption spreads (buy the ATMF strike, sellthe A+50 strike) with a 6/30/2012 expiry on various underlying tails
* As of COB 11/17/2011
Exhibit 29: Conditional curve trades with a twist—since the 10s/30s curve is well correlated to front-end yields, and since YCSOs on 10s/30s are rich,2-year tail swaptions can be replaced with options on the 10s/30s curve to create “synthetic” conditional curve trades at better entry levelsImplied volatilities (bp/day), notionals ($mn) and premium($) for conditional curve trades and corresponding values for trades in which the left leg option is replaced by a YCSOcurve cap.
0.4
0.5
0.6
0.7
0.8
0.9
Aug 11 Sep 11 Oct 11 Oct 11 Nov 11
Empirical beta Implied beta
Slide Premium Max
Tail Spot Forward bp bp yield gain Ratio
1Y 0.752 0.846 9.3 12.1 37.9 3.1
18M 0.780 0.859 7.9 12.1 37.9 3.1
2Y 0.799 0.897 9.8 12.2 37.8 3.1
3Y 0.897 1.080 18.3 13.3 36.7 2.8
5Y 1.324 1.570 24.6 16.1 33.9 2.1
7Y 1.750 1.963 21.3 17.6 32.4 1.8
10Y 2.151 2.317 16.6 18.6 31.4 1.7
15Y 2.503 2.615 11.2 19.2 30.8 1.6
30Y 2.701 2.764 6.3 20.1 29.9 1.5
Yields, %
Net prem
Net Beta wrt on modified
Expiry Left Right Type Left Right Left Right Premium Expiry YCSO Imp. Vol Left leg Notional Type Premium trade
3m 2Y 5Y Rec 3.46 5.26 -100 40.8 111,655$ 3m 10s/30s 4.6 -0.41 -485.9 Cap 14.4 (373,100)$
3m 2Y 7Y Rec 3.46 6.61 -100 29.8 195,317$ 3m 10s/30s 4.6 -0.41 -485.9 Cap 14.4 (289,438)$
3m 2Y 10Y Rec 3.46 7.86 -100 21.7 272,424$ 3m 10s/30s 4.6 -0.41 -485.9 Cap 14.4 (212,331)$
3m 2Y 30Y Rec 3.46 8.96 -100 9.6 339,559$ 3m 10s/30s 4.6 -0.41 -485.9 Cap 14.4 (145,196)$
6m 2Y 5Y Rec 3.36 5.28 -100 40.7 167,224$ 6m 10s/30s 4.0 -0.34 -582.5 Cap 17.7 (569,328)$
6m 2Y 7Y Rec 3.36 6.41 -100 29.7 265,495$ 6m 10s/30s 4.0 -0.34 -582.5 Cap 17.7 (471,057)$
6m 2Y 10Y Rec 3.36 7.46 -100 21.6 355,936$ 6m 10s/30s 4.0 -0.34 -582.5 Cap 17.7 (380,616)$
6m 2Y 30Y Rec 3.36 8.26 -100 9.6 423,282$ 6m 10s/30s 4.0 -0.34 -582.5 Cap 17.7 (313,270)$
Trade Notionals
Desired Trade
Implied vol
Replacement for left leg
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 146/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Praveen Korapaty (1-212) 834-3092
Alberto Iglesias (1-212) 834-5116
J.P. Morgan Securities LLC
146
10s/30s curve may be used in place of swaptions on 2-
year tails; specifically, receiver swaptions on 2-year tails
may be replicated via similar expiry caps on the 10s/30s
curve, and payer swaptions on 2-year tails may be
replaced with floors on the 10s/30s curve.
Exhibit 29 presents details for a sample set of
conditional curve trades, with the 2-year leg being
replaced by options on the 10s/30s curve; for instance,
the short receiver swaption position in a conditional
bullish flattener would be replaced with short position in
1-look caps on the 10s/30s curve with the same expiry.
The short payer swaption leg in a conditional bearish
steepener would similarly be replaced with a short floor
position on the 10s/30s curve. As can be seen in the
table, such modified “synthetic” conditional curve trades
are interesting because they can mimic an unweighted
conditional curve trade and also take in premium
currently, which cannot be done right now in the
swaptions market. A visual confirmation that such a trade
would indeed track the yield curve (which, after all, isthe ultimate intent of a conditional curve trade) is seen in
Exhibit 30. The yield spread corresponding to a 6-month
forward 2s/10s curve trade, where the 2-year leg has
been replaced with a beta weighted position in the
10s/30s curve, is shown in this exhibit; as can be seen,
this spread does indeed closely track the 2s/10s curve,
which was the original desired trade in this example.
Exhibit 31: A tale of two different vol markets in 2011—being short
volatility was profitable until July, and long volatility positions haveprevailed afterwards1-month Gamma and Vega P/L* from long straddles in 3Mx10Y swaptions ; bp of notional
* Returns calculated using the J.P. Morgan Volatility Index, which assumes dailydelta rebalancing and zero transaction costs. Options are re-struck at the start of each month.** November-to-date P/L.
Exhibit 32: Expect liquidity to remain poor: risk appetite continues todecline as the European crisis spreads into semi-peripheralcountries …Reported average quarterly VAR for the 9 largest investment banks* versus 3-monthmoving average of semi peripheral spreads**; $mn
* Average VAR reported by JPM, GS, MS, BAC, C, UBS, CS, Soc Gen and DB.3Q11 VAR is an estimate, projected from the results reported to date (excluding C).** Average of 5-year CDS spreads on France, Italy and Spain.
Exhibit 30: “Synthetic” curve trades—where the front end leg has been
replaced with a beta-weighted position in the 10s/30s curve—are likelyto track the desired original trade6-month forward 2s/10s swap curve, versus the yield spread corresponding toreplacing the 2-year leg with a beta weighted 6-month forward 10s/30s curve position
*Calculated as 6Mx10Y minus (6Mx10s/30s curve divided by -0.34)
-60
-40
-20
0
20
40
60
J a n
F e
b
M a r
A p r
M a y
J u n
J u
l
A u g
S e p
O c
t
N o v
* *
G amma PL Vega PL
50
100
150
200
250
30090
100
110
120
130
140
Dec 09 Apr 10 Jul 10 Oct 10 Feb 11 May 11 Aug 11
Semi-peripheralspreads
VAR
3.63.8
4.0
4.2
4.4
4.6
4.8
5.0
1.41.5
1.6
1.7
1.8
1.9
2.0
2.1
Sep 11 Oct 11 Nov 11
Yield spread for synthetic trade; %6M fwd 2s/10s; %
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 147/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Praveen Korapaty (1-212) 834-3092
Alberto Iglesias (1-212) 834-5116
J.P. Morgan Securities LLC
147
Options
Short volatility strategies have thus far managed to
remain profitable in 2011 year-to-date, but it has been a
clear case of two different market environments. From
the start of the year through the end of July, selling
gamma proved considerably profitable overall. Incontrast, since August, the opposite has been the case,
and long gamma strategies have steadily gained ground
(Exhibit 31).
While the outperformance of long gamma strategies in
August is a typical seasonal phenomenon related to
declines in market depth, heightened macro uncertainty
stemming from Europe has thus far forestalled a recovery
in market depth, helping preserve conditions that are
favorable to long gamma positions.
Looking ahead towards the remainder of this year and
1Q12, our best guess is that risk appetite will remain low,and liquidity conditions will remain poor. Declining risk
appetite is perhaps most evident in the daily VARs
reported by major investment banks in their quarterly
financials. As seen in Exhibit 32, VARs remain on the
downtrend, reflecting growing caution as the European
crisis has spread into semi-peripheral countries.
In the near term as we head into year end, seasonals are
unfavorable as well and could make for worsening risk
appetite and market depth. As seen in Exhibit 33, market
depth has exhibited a tendency to decline to annual lows
in December. This should prove unhelpful to marketliquidity and could support elevated realized volatility.
One way to see this is to estimate the width of the market
at which a given fixed size can be transacted by a price-
taker. For instance, an investor needing to (say) buy
$150mn 10s might simply lift the highest offer in a
period when market depth is high, but might need to lift
(say) the top three or four offers in periods of lower
market depth, assuming a static order book. Examining
the implied width of the market for a given fixed size (we
Exhibit 35: A model for seasonally-adjusted market depthModel* for market depth in the 10-year sector, and forecast for 2012
* Model based on regression of seasonally-adjusted market depth against semi-peripheral spreads and cross asset correlation, which we proxy using the rolling 3-month correlation between weekly changes in the S&P and 30-year swap yields.Semi-peripheral spreads are an average of 5-year CDS spreads on France, Spainand Italy, and are used to capture broader risk aversion. Regression estimated over two years of history.
Exhibit 33: Intra-year seasonals are likely to be unhelpful to market
liquidity, as evidenced by the tendency of market depth to reach itslows in December Percentage deviation of market depth in the 10-year sector from average over the pastfive years
* Market depth is calculated as the average size of the top three bids and offers, in$mn, for the on-the-run 10-year Treasury note, averaged between 8:30 a.m. and10:30 a.m. daily.
Exhibit 34: Larger market moves become likely in periods of low
market depth, as a price-taker will need go deeper down the order stack to trade a given amountRegression of market width* spread against market depth** in the 10-year sector; bp
Market depth; $mn* Market width calculated as the spread between mid-market and the level at whicha fixed size ( $150mn) of the on-the-run 10-year Treasury note can be traded,assuming a static order book. Bid-side and offer-side widths are averaged here.** see footnote in previous exhibit
Factor Coefficient T-stat 1Q12 2Q12 4Q12
Intercept 303.1 97.0
Semi-peripheral spreads -0.47 -27.8 70 125 70
Cross-asset correlation -76.1 -19.2 -0.1 -0.3 -0.3
R-squared
Std error
Impact on market depth -26 -36 -10
Projected market depth 70 60 86
77%
26.4
Chg in driver, from current
-60%
-40%
-20%
0%
20%
40%
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
0.5
1.0
1.5
2.0
2.5
3.0
50 100 150 200 250
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 148/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Praveen Korapaty (1-212) 834-3092
Alberto Iglesias (1-212) 834-5116
J.P. Morgan Securities LLC
148
use $150mn) is insightful. As Exhibit 34 shows, this
implied market width grows nonlinearly with depth,
highlighting the fact that the market is much more prone
to violent moves in periods of low market depth—as
even small-sized trades produce sizeable moves.
Looking past year end, the key question then is whether
market depth will begin to improve with a renewal of
risk appetite next year. While this has historically been
the case seasonally, there are good reasons not to expect
such a pick-up after year-end 2011. In order to better
understand the drivers of market depth, which has
become a key determinant of realized volatility, we
analyze the drivers of market depth on a seasonally-
adjusted basis (given significant intra-year variation dueto seasonals). As seen in Exhibit 35, market depth
(adjusted for intra-year seasonals) has been well
explained by two factors over the past two years. The
first factor is a measure of cross asset correlations,
proxied here by the rolling equity-rate correlation. All
else equal, rising cross-asset correlations effectively
increase the risk in an investor’s portfolio, in turn leading
to cuts in notional amounts of risk-taking in individual
markets. As short-term cross-asset correlations have risen
in markets, measured VARs in portfolios have likely
risen as well, prompting risk reduction; market depth
would likely suffer in such periods.
A second factor that has also biased market depth lower
has been the steady contagion from the European crisis,
which we measure by using an average of semi-
peripheral spreads. Finally, although not formally in our
model, we note that some of this decline in market depth
is also structural, reflecting stricter capital and leverage
requirements on banks.
Looking ahead, cross-asset correlations are likely to
remain elevated or even increase further in the near term
going into year end, but decouple thereafter in 2012.
Semi-peripheral spreads, however, are likely biased
wider, in the view of our European rates strategists.Incorporating their views into the framework of our
model for market depth suggests that the typical new-
year rebound in market depth may be unlikely; we
project that market depth is likely to average similar or
lower levels for much of 1H12, as also seen in
Exhibit 35.
What does this mean for delivered volatility? To answer
this, we turn to a model for realized volatility we have
used in recent months. This model explains realized
volatility in the 10-year sector using yield levels, the
curve, market depth and an index of hedge fund leverage
as independent variables. It is presented in Exhibit 36
together with the details of our projections for 2012. As
can be seen, we generally project elevated realized
volatility over much of 2012 (i.e., levels comparable to
current implieds), and would thus look for an
Exhibit 36: Look for realized volatility to decline only modestlyafter year end, and remain elevated in 1H12Statistics regarding a model* for 1-month ahead realized volatility on 3Mx10Yforward swap yields, and forecast for 2012
* Model based on 1-year regress ion of 1-month 3Mx10Y realized volatilityagainst ex-ante levels of the 2s/10s swap yield curve, market depth in the 10-
year sector**, 3M forward 10-year swap yields and the hedge fund leverageindex***.** Market depth is calculated as the average size of the top three bids andoffers, in $mn, for the on-the-run 10-year Treasury notes, respectively,averaged between 8:30 a.m. and 10:30 a.m. daily.*** Leverage defined as the sum of the absolute value of each of the six betacoefficients in the following multiple regression: daily excess returns on the IQHedge Global Macro beta index regressed against daily excess returns on 1)J.P. Morgan global bond index ex-US, 2) J.P. Morgan US 7-10Y bond index, 3)J.P. Morgan US 7-10-year minus 1-3-year bond index, 4) MSCI G7 index, 5)USD (J.P. Morgan USD cash index minus J.P. Morgan global cash index), and6) GS Commodity index. Higher values imply more leverage.
Exhibit 37: Mortgage market convexity needs are near historiclowsConvexity exposure in Fannie Mae and Freddie Mac’s portfolios*;$bn 10s per 10bp
* Estimated from their reported interest rate sensitivities.
Factor Coefficient T-stat Current 1Q12 2Q12 4Q12
Intercept -0.6 -0.6
3M fwd 2s/10s curve (%) 4.0 5.0 1.49 -0.3 0.4 0.5
Market depth ($mn) -0.017 -11.1 98 -26 -36 -10
3Mx10Y yield (%) -1.5 -2.9 2.29 -0.35 0.45 0.65
H.F. leverage index 5.85 15.6 1.4 -0.1 -0.4 -0.5
R-squared
Std error
Impact on del. Vol (bp/day) -0.8 -0.8 -1.7
Projected realized vol 8.4 7.6 7.6 6.7
72%
0.97
Chg in driver, from current
-10
-9
-8
-7
-6
-5
-4
-3
-2
2009 2010 2011
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 149/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Praveen Korapaty (1-212) 834-3092
Alberto Iglesias (1-212) 834-5116
J.P. Morgan Securities LLC
149
environment that is largely favorable to long gamma
positions. In the early part of the year, the main driver of
this is likely to be heightened risk aversion and low
market depth that combine to create illiquid conditions
that support high realized volatility. Later in the year,
however, the driver is likely to be higher rates—we
expect US yields to decouple from the European crisis
later in 2012, and rising yields will likely emerge as themain driver sustaining elevated realized volatility.
Long-dated volatility
As we contemplate the outlook for longer-dated swaption
implied volatility in the year ahead, it is useful to revisit
the key drivers of implieds in this sector in recent years.
In addition to the overall level of volatility, which would
of course drive implieds all across the vol surface, there
have been two important drivers. The first is mortgage
market convexity, which has of course steadily declined
in magnitude, resulting in declining demand for
intermediate expiry swaption volatility. We use theimputed convexity of the two GSEs’ portfolios net of
hedges (which they report monthly) as our preferred
variable to capture mortgage market convexity needs in
the aggregate (Exhibit 37).
A second factor that has been relevant in recent years is
long-term inflation expectations. As we note in the Cross
Sector Overview, there are not many attractive
alternatives to position for a sharp rise in long-term
Exhibit 38: Longer-dated swaption implied volatility has been wellexplained recently by three factors—the level of volatility overall,mortgage market convexity, and inflation expectations3Yx10Y implied volatility versus in-sample model fair value *
* Model based on 3-year regression of 3Yx10Y implied vol against GSEportfolio convexity (see Exhibit 37), 10-year inflation swap yields (%) and3Mx10Y implied vol (bp/day).
Exhibit 40: Implied correlation has plunged in anticipation of Operation Twist …
Implied correlation between 6M forward swap yields from YCSO market. Averagevalue from 8/18/2010 to 8/18/2011, value as of 8/18/2011 and drop during the period8/18/2011-11/18/2011; %
.
Exhibit 39: Look for 3Yx10Y swaption volatility to decline to 6bp/day in1Q12
Statistics regarding a model* for 3Yx10Y implied volatility and forecast for 2012
* See Exhibit 38
Exhibit 41: … and that, coupled with an inverted volatility curve, hasled to cheap forward volatilityBeta weighted differential between forward and spot volatility, current level and 2-year Z-score
* Beta weighted.
5
6
7
8
9
10
Nov 08 Jun 09 Dec 09 Jul 10 Jan 11 Aug 11
3Yx10Y implied vol
Model
-100%
-50%
0%
50%
100%
6 M x
5 Y
6 M x
1 0 Y
6 M x
3 0 Y
1 Y x
5 Y
1 Y x
1 0 Y
1 Y x
3 0 Y
2 Y x
5 Y
2 Y x
1 0 Y
2 Y x
3 0 Y
Avg 8/18/2011 Chg 8/18/2011-11/18/2011
Factor Coefficient T-stat Current 1Q12 2Q12 4Q12
Intercept 1.4 9.7
GSE portfolio cvx ($bn 10s/10bp) -0.17 -17.1 -2.6 -2.1 -3.3 -3.6
10Y inflation swap rate; % 1.0 21.3 2.26 1.81 1.96 1.96
3Mx10Y implied vol; bp/day 0.30 37.9 7.65 7.6 7.6 6.7
R-squared
Std error
Actual implied vol; bp/day 6.9
Projected implied vol, bp/day 6.4 5.9 6.2 6.0
80%
0.335
Projections
2-year average
Forward Spot (bp/day) 2-year Z-score
6Mx5Yx5Y 5Yx5Y 1.02 -0.6 -0.3 -3.46Mx5Yx10Y 5Yx10Y 1.00 -0.4 -0.2 -3.11Yx5Yx5Y 5Yx5Y 0.93 -0.2 0.2 -2.9
1Yx5Yx5Y 5Yx5Y 0.93 -0.2 0.2 -2.91Yx5Yx30Y 5Yx30Y 1.03 -0.9 -0.5 -2.76Mx1Yx10Y 1Yx10Y 1.12 -1.4 -0.9 -2.21Yx1Yx10Y 1Yx10Y 1.10 -1.4 -0.7 -2.11Yx1Yx10Y 1Yx10Y 1.10 -1.4 -0.7 -2.16Mx6Mx10Y 6Mx10Y 0.96 -0.7 0.4 -1.91Yx6Mx10Y 6Mx10Y 0.90 -0.2 0.7 -1.91Yx6Mx10Y 6Mx10Y 0.90 -0.2 0.7 -1.9
Richness/cheapness of forward vol
Forward minus
spot vol spread*
(bp/day)
2-year beta
of forward
vs spot vol
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 150/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Praveen Korapaty (1-212) 834-3092
Alberto Iglesias (1-212) 834-5116
J.P. Morgan Securities LLC
150
inflation expectations, other than outright bearish trades
or payer swaptions. Indeed, recent years have seen
considerable demand for trades designed to protect
against a rise in long-term inflation, via payer swaptions/
payer swaption spreads (or CMS caps or cap spreads),
although this demand has ebbed and flowed with market
sentiment in this regard.
These two factors, together with 3Mx10Y swaption
implied volatility (as a proxy for the overall level of
volatility) do indeed successfully explain most of the
variation in implied volatility levels in the intermediate
expiry sector, as seen in Exhibit 38. We also project the
likely evolution of the drivers in Exhibit 39. In
particular, we project the convexity of the GSEs’
portfolios based on the most recent known data point, the
recent relationship of this variable with yield levels, and
our projections for yield levels in 2012; we project 10-year inflation swap yields based on our projections for
TIPS breakevens (see TIPS); and we draw upon our
forecast for short expiry swaption implied volatility
detailed earlier.
Putting it all together, our projections point to declines in
3Yx10Y implied volatility to 6bp/day by the end of
1Q12, with range-bound levels thereafter.
Bermudan Swaptions
As investors look ahead to devising their volatility
strategies in 2012, several underlying principles will
likely be important.
First, given extremely low rates across much of the
curve, implied volatility across much of the volatility
surface has become highly directional with rate levels.
Exhibit 42: A Bermudan swaption may be thought of as spanning
numerous European swaptions, corresponding to each of thevarious possible exercise dates …Graphic representation of a 1Yx5Y Bermudan swaption with semi-annual exercisepoints, and its spanned European swaptions.
Exhibit 43: … also, it can be thought of as spanning numerous
Canary swaptionsGraphic representation of a 1Yx5Y Bermudan with annual exercise points, and its 10spanned Canary swaptions
Exhibit 44: The BC basis is consistently narrower than the BE basisunder different yield curve scenarios1Yx5Y ATM receiver Bermudan/Canary basis and Bermudan/European basis for the CTD divided by the Bermudan vega* for various parallel shifts of the yieldcurve**; %
Shift in yields; %* Bermudan vega in bp of notional per bp shift in daily vol** COB 11/10/2011
t=0 t=1Y t=6Y
─ ─ ─►
t=1Y (first exercise) ─ ─ ─► 5Y swap
t=1.5Y ─ ─ ─ ─ ─►
─ ─ 2Y expiry ─► 4Y swap
─ ─ ─ ─ ─ ─ ─ ─ ─►
─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─►t=3.5Y
─ ─ ─ ─ 3.5Y expiry ─ ─ ─ ─ ─► 2.5Y swap
─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─►
─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─►t=5Y
─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─►
─ ─ ─ ─ ─ ─ 5.5Y expiry (last exercise) ─ ─ ─ ─ ─► 6M
Spanned European swaptions
1Yx5Y Bermudan swaption
t=0 t=1Y t=6Y
─ ─ ─►
t=1Y t=2Y ─ ─ ─► 1Y(2Y)x5Y Canary
t=3Y ─ ─ ─► 1Y(3Y)x5Y Canary
t=4Y ─ ─ ─►
t=5Y ─ ─ ─► 1Y(5Y)x5Y Canary
t=2Y ─ ─ ─ ─ ─ ─ ─► 2Y(3Y)x4Y Canary
─ ─ ─ ─ ─ ─ ─► t=4Y ─ ─ ─ ─ ─ ─ ─►
t=3Y ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─► 3Y(4Y)x3Y Canary
─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─►t=4Y
─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─►
1Yx5Y Bermudan swaption
Spanned Canary swaptions
20%
40%
60%
80%
100%
120%
-150 -75 -25 0 25 75 150
BC basis/Berm. vega BE basis/Berm. vega
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 151/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Praveen Korapaty (1-212) 834-3092
Alberto Iglesias (1-212) 834-5116
J.P. Morgan Securities LLC
151
All else equal, this argues for structuring portfolios to
have directional vega exposure—i.e., positions that
become long vega in a sell-off are attractive.
Second, one consequence of the Fed’s recent initiation of
Operation Twist has been the plunge in correlation
(Exhibit 40). While this is most noticeably true for
correlations between front-end and back-end rates, it is
also true more broadly. Last, with liquidity likely to
remain poor, realized volatility is likely to remain high as
well and preserve an inverted implied volatility curve
across expiries—i.e., short expiries are likely to remain
elevated relative to longer expiries. One corollary of the
last two points—cheap implied correlation and an
inverted volatility curve—is that forward volatility
appears cheap (Exhibit 41).
Thus, vega strategies for 2012 should seek to leverage
the current cheapness of implied correlations and forward
volatility, while creating positions that gain vega in aselloff. Interestingly, Bermudan receiver swaptions offer
a way to do all of those in one package. Since Bermudan
swaptions can be exercised at several points in time, the
value of near-term exercise grows with the moneyness of
the option. Thus, Bermudan receivers, for instance,
should increasingly track short-expiry receiver swaptions
in a rally, but effectively behave more like longer-dated
vega instruments in a sell-off (see Introduction to
Bermudan Swaptions and a Framework for Analysis,
Exhibit 45: The Canary CTD mimics the Bermudan riskcharacteristics better than individual European counterparts,
both for delta …Deviation of the delta profile of selected European and Canary swaptions fromthe Bermudan delta profile plotted against parallel shifts of the yield curve; %
The delta profile is obtained as a % of the original delta (that is, for no changein the yield curve). Each curve is obtained by subtracting the European or Canary delta profile from the Bermudan delta profile. The 1Y(2.5Y)x5Y Canaryand the 1Yx5Y European are the CTDs.
Exhibit 46: … and vegaDeviation of the vega profile of selected European and Canary swaptions from the
Bermudan vega profile plotted against parallel shifts of the yield curve; %
The vega profile is obtained as a % of the original vega (that is, for no change in theyield curve). Each curve is obtained by subtracting the European or Canary vegaprofile from the Bermudan vega profile. The 1Y(2.5Y)x5Y Canary and the 1Yx5YEuropean are the CTDs.
Exhibit 47: Sources of value in the 5Yx30Y Bermudan receiver swaptionProperties* of Bermudan swaption, Canary CTD and volatility curve as of COB11/01/2011
* Premia quoted in bp of notional and vega in bp of notional by 1/10 bp shifts in dailyvolatility.
-0.3
-0.2
-0.1
0.0
0.1
0.2
0.3
-150 -100 -50 0 50 100 150
1Y(2.5Y)x5Y Canary
1Yx5Y European
2.5Yx3.5Y European
-3%
-2%
-1%
0%
1%
2%
3%
-150 -100 -50 0 50 100 150
1Y(2.5Y)x5Y Canary
1Yx5Y European
2.5Yx3.5Y European
Premiums from a Bermudan model
5x30 Berm receiver 1891.5 bp
5Y(14Y6M)x30Y Canary 1741.1 bp
Vega 33.1 bp of notl per 0.1 bp/day
Alternate fair values for Canary based on:
Implied correlation 1848.6 bp
Cheapness of Canary (& Berm)
In price terms 107.5 bp of notional
In vol units 0.32 bp/day
Mispricing in correlation markets
Canary price using implied correlation 1848.6 bp
Canary price using realized correlation 1851.7 bp
Correlation driven cheapness of Berm
In price units 3.1 bp of notional
In vol units 0.009 bp/day
Mispricing of 5x30/14.5x20.5 vol curve 0.001 bp/day
Total mispricing 0.335 bp/day
plus
plus
equals
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 152/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Praveen Korapaty (1-212) 834-3092
Alberto Iglesias (1-212) 834-5116
J.P. Morgan Securities LLC
152
J.P. Morgan Research Note, July 1, 2003 for more
details). In addition, Bermudan swaptions are also
naturally long exposure to correlation and forward
volatility.
However, the complexity of Bermudan swaptions (due to
several possible exercise dates) means that they have
exposure to a number of different correlations and
forward volatilities. It is therefore unclear exactly which
implied correlations and which forward volatilities a
given Bermudan structure is chiefly exposed to, if such
simplification is even possible. More broadly, while
Bermudan receiver swaptions may offer a slate of
desirable characteristics, it is unclear if they offer cheap
means to add such exposures since developing a relativevalue view on Bermudan swaptions is challenging.
In this section, we build on our previous research on
Bermudan swaptions to develop exactly such a relative
value metric for each Bermudan structure. Our approach
is as follows.
First, we identify a simpler instrument that is “spanned”
by a Bermudan swaption that most closely mirrors the
Bermudan itself. In our earlier work, we noted that each
Bermudan swaption may be thought of as spanning a
basket of European swaptions—one for each possible
exercise date. Moreover, the premium of the Bermudanswaption must be higher than each of the European
swaptions spanned by it, and the European swaption with
the highest premium in that basket may be thought of as
the best European approximation to the more complex
Bermudan, at least locally (Exhibit 42). (We label that
particular European swaption as the Bermudan’s “CTD,”
motivated by the Treasury futures analogy and
notwithstanding the fact that it is the most expensive
European swaption in the basket, not the cheapest).
While that observation is useful in developing some
intuition regarding the behavior of Bermudan swaptions,
the relative instability of the Bermudan swaption’sEuropean CTD has in practice meant that it does not
easily lead to a value metric. In this piece, we examine a
natural generalization of that idea that does lead to a
value metric. Much like the Bermudan swaption may be
thought of as spanning a basket of one-exercise date
European swaptions (one for each possible date on which
the Bermudan can be exercised), we may also think of
the Bermudan swaption as spanning a basket of two-
exercise date Canary swaptions, with one Canary for
each possible pair of dates on which the Bermudan
swaption can be exercised (Exhibit 43). We recall here
that a Canary option is one that can be exercised on
exactly two dates; for instance, a 1Y(2Y)x5Y has
possible exercises one and two years from now into
swaps that end in six years. Moreover, just as was the
case earlier, the price of the Bermudan swaption must be
at least as large as the highest premium canary that is
spanned by the Bermudan—we refer to that particular
Exhibit 48: The relative value metric shows several sectors ascurrently mispricedComponents* of our relative value metric for Bermudan receiver swaptions andtotal in various sectors expressed in price terms (bp of notional) and Bermudanvolatility terms (bp/day) as of COB 11/01/2011
* Canary CTD premium using implied correlation minus Premium of the CanaryCTD implied by Bermudan price.** Canary CTD premium using realized correlation minus premium using impliedcorrelation.*** Canary CTD premium using the second forward rate volatility shifted by volcurve mispricing (from a 2-year regression of vol curve versus second vol level)minus premium using current implied volatilities and realized correlation.
Exhibit 49: Total mispricing of the 1Yx10Y Bermudan receiver swaptionover the past four yearsTotal mispricing in the 1Yx10Y Bermudan receiver swaption; bp/day
* Negative values indicate cheapness of Bermudan swaptions
Sector
Price Vol Price Vol Price Vol Price Vol
5Yx30Y 107.5 0.32 3.1 0.01 19.3 0.06 129.8 0.39
6Mx30Y 93.5 0.33 -1.1 0.00 13.3 0.05 105.7 0.37
1Yx30Y 91.0 0.31 -0.9 0.00 13.7 0.05 103.9 0.36
3Yx30Y 93.2 0.30 0.7 0.00 15.9 0.05 109.7 0.35
2Yx30Y 86.1 0.28 -0.4 0.00 14.6 0.05 100.4 0.33
6Mx2Y 2.6 0.24 -0.2 -0.02 0.3 0.03 2.8 0.25
6Mx10Y 7.7 0.09 7.3 0.09 1.7 0.02 16.6 0.20
3Yx10Y 13.8 0.12 5.1 0.04 0.3 0.00 19.2 0.17
5Yx10Y 12.0 0.09 8.3 0.06 1.3 0.01 21.5 0.17
3Yx7Y 4.5 0.06 6.0 0.08 1.0 0.01 11.5 0.14
1Yx10Y 7.2 0.08 3.5 0.04 0.9 0.01 11.6 0.13
2Yx10Y 8.2 0.08 4.5 0.04 0.4 0.00 13.1 0.13
5Yx5Y 6.4 0.10 -0.3 -0.01 1.6 0.02 7.7 0.12
6Mx3Y 1.6 0.09 1.9 0.11 -1.6 -0.09 1.9 0.11
6Mx5Y 1.6 0.05 3.8 0.11 -3.3 -0.10 2.0 0.06
Bermudan / Canary
basis cheapness*
Correlation
cheapness** Total
Vol curve
cheapness***
-0.6
-0.4
-0.2
0.0
0.2
Nov 07 May 08 Dec 08 Jun 09 Jan 10 Aug 10 F eb 11 Sep 11
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 153/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Praveen Korapaty (1-212) 834-3092
Alberto Iglesias (1-212) 834-5116
J.P. Morgan Securities LLC
153
Canary as the Canary CTD, in keeping with our earlier
terminology.
But why is it useful to think of a Bermudan swaption as
spanning a set of Canaries? Because the Canary CTD is
much more stable than the European CTD, across a range
of rate, curve and volatility moves, making it a useful
simplification (Exhibits 44-46). Moreover, canaries also
are exposed to exactly one correlation and forward
volatility, which will be useful to us. For instance, the
5Yx30Y Bermudan currently has the 5Y(14.5Y)x30Y as
its Canary CTD, with resulting exposure to 5Y forward
9.5Yx20.5Y correlation and forward volatility.
We may now develop a relative value metric. In doing
so, we note that there are three possible sources of value.First, the price of the canary CTD in the Bermudan
swaptions market (or from a suitably calibrated
Bermudan swaption pricing model) may differ from the
price of a canary that can independently be calculated
from vanilla implied volatilities and implied correlations
from the YCSO market. Second, using an empirical
framework, we may conclude that implied correlations in
the YCSO market are themselves rich or cheap, based on
comparisons to realized correlations over time. Third,
again using an empirical framework, we may take a view
on the volatility curve corresponding to a given canary.
Last, we may total up the mispricing from all threesources, and divide by the vega of the Bermudan
swaption to compute the richness or cheapness of the
Bermudan swaption, expressed in basis points per day.
This is illustrated in detail below for the case of the
5Yx30Y Bermudan receiver swaption in Exhibit 47. The
sources of value are also illustrated for the currently most
mispriced Bermudan receiver swaptions in Exhibit 48,
while Exhibit 49 shows a time series of the total
mispricing in the past four years for the particular case of
the 1Yx10Y Bermudan receiver. Large negative values
of our total mispricing should be taken as a signal to buy
the Bermudan.
As a way to analyze the usefulness of our metric we
evaluated the performance of buying several Bermudan
receiver swaptions at previous points in time in 2011
when our metric suggested cheapness. Three strategies
were tested: (i) hedging only the delta, (ii) hedging the
delta as well as the vega by selling the European “CTD”
swaption, and (iii) hedging the delta as well as vega by
selling a portfolio of replicating European swaptions.
Our results in this regard are preliminary, and much more
analysis is needed. However, our initial results are
promising, and suggest that such a relative value metric
is in fact useful in signaling cheapness of Bermudan
swaptions (Exhibit 50); our (limited) results currently
appear to suggest that buying the Bermudan versus the
European CTD swaption (and dynamically delta-hedging
the switch) is the best way to monetize the cheapness of
Berms.
Regulatory updateUnder the Dodd-Frank Act, several government agencies
such as the CFTC, SEC, FDIC, and the Federal Reserve
were tasked with rule-making responsibilities. We review
developments this year on this front, as well as some
potential changes in markets that could result from the
broader regulatory reform effort.
Proposed and finalized rules
The CFTC, the key rulemaking body as it pertains to
derivatives markets, was tasked under the Dodd-Frank
Act to regulate Swap Dealers (SD) on capital and margin
requirements, as well as on more stringent recordkeeping
and reporting. The act also required the CFTC to specify
rules related to the migration of trading standardized
derivatives to regulated exchanges or Swap Execution
Facilities (SEF).
The process of arriving at final rules has proved fairly
time intensive. In July, the CFTC issued an order
clarifying the effective date of the provisions in the swap
regulatory regime established by Title VII of the Dodd-
Exhibit 50: More empirical analysis is needed, but our results currentlysuggest that our metric does indeed identify value in Bermudanswaptions, and also that the best way to monetize this metric is to buythe Bermudan swaption against the European CTD on a delta hedgedbasis3-month return of long Bermudan positions delta hedged*, vega hedged with respect tthe European CTD, and vega hedged to a replicating portfolio**;
* Options are delta hedged daily and rolled every month. Returns assume zerotransaction costs.
** Replicating portfolio refers to a weighted combination of European swaptions that ar“spanned” by the Bermudan swaption. The weights are solved for using a scheme toreplicate the Bermudan swaption’s characteristics under a range of scenarios. SeeIntroduction to Bermudan Swaptions and a Framework for Analysis, J.P. MorganResearch Note, 7/1/2003 for details.
bp of notl. bp/day bp of not l. bp/day notl. bp/da
1/3/2011 -0.44 88.5 0.4 25.0 0.1 8.5 0.0
5Yx30Y 3/1/2011 -0.44 -105.4 -0.4 311.3 1.3 28.5 0.1
4/1/2011 -0.43 -125.8 -0.5 61.9 0.3 -25.9 -0.1
5/2/2011 -0.38 -217.4 -0.9 185.8 0.8 -152.6 -0.6
Trade start
dateBermudan
Delta hedged returns over 3-months
Vega hedged to
European CTD
Vega hedged to
replicating portfo
Mispricng at
inception
Bermudan
vol terms
(bp/day)
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 154/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Praveen Korapaty (1-212) 834-3092
Alberto Iglesias (1-212) 834-5116
J.P. Morgan Securities LLC
154
Frank Act. The order provides temporary relief from
provisions that that would have become effective July 16,
2011 up to the effective date of final rules or December
31, 2011. Specifically, the CFTC provided relief from
certain provisions in the Commodities and Exchange Act
(CEA) that do not require rule-making but reference one
or more terms regarding swap entities and instruments
(such as the terms “swap,” “swap dealer,” “major swap
participant,” or “eligible swap participant”) that the act
requires be further defined. The CFTC has, jointly with
the SEC, issued notices of rulemaking to further define
these terms. In its latest amendment to its July order, the
Commission is proposing to extend the latest date of
expiry of relief from December to July 16, 2012 as it has
not yet finalized its definition of these terms. As such,several aspects of the rule-making are still in various
stages of either being open to comments or under
discussion by the commission. Rules in this stage
include, but are not restricted to, key items affecting
OTC derivative markets as relating to margin
requirements for uncleared swaps, capital requirements
for Swap Dealers (SD) and Major Swap Participants
(MSP), and collateral type and custody rules. Other items
yet to be finalized include rules on Derivatives Clearing
Organizations (DCOs) with regards to financial resources
and the resource waterfall in case of a default by a
clearing member, as well as end-user exceptions from the
mandatory clearing requirement. Several rules were infact finalized, such as those related to swap data
repositories, on particular products such as agricultural
swaps, and certain enforcement-related items, but as
noted earlier, key rules remain outstanding.
The Dodd-Frank Act also required the Federal Reserve to
implement several provisions, sometimes in conjunction
with other agencies. As part of this inter-agency effort
(between the Fed, FDIC, FCA, FHFA and OCC), a
proposal seeking comment was issued in April that
would require swap entities regulated by the five
agencies to collect minimum amounts of initial margin
and variation margin from counterparties to non-clearedswaps and non-cleared, security-based swaps (with a
commercial end-user exception as long as the entities’
margin exposure is below an appropriate credit exposure
limit). The amount of margin required under the
proposed rule would vary based on relative risk of the
counterparty and of the swap. The proposal would apply
to new, non-cleared swaps entered into after the proposed
rule’s effective date. Existing capital standards that apply
to swap entities as part of their prudential regulation
already address non-cleared swaps, and therefore are not
separately addressed in the rule.
The Fed also requested public comment on the “Volcker
Rule” requirements, developed jointly with the FDIC,
OCC, SEC and CFTC. Briefly, the Volcker Rule prohibits insured depository institutions, BHCs, and their
subsidiaries or affiliates from engaging in short-term
proprietary trading of any security, derivative, and
certain other financial instruments for the entity’s own
account, subject to certain exemptions. The act also
prohibits exempt transactions that may result in a conflict
of interest with customers or in a material exposure to
high-risk assets or trading strategies as defined by the
rule. The proposal clarifies the scope, as well as provides
Exhibit 51: Cleared interest rate swap volumes have risen significantlyover the last few yearsPercent of notional outstanding cleared via LCH
Source: ISDA, BIS, LCH
Exhibit 52: Cleared volumes of other interest rate derivatives have risenover the year Percent of notional outstanding centrally cleared in 2011
Source: Trioptima
20%
30%
40%
50%
60%
2007 2008 2009 2010
10%
13%
16%
19%
22%
25%
28%
31%
58%
60%
62%
64%
66%
68%
70%
72%
Dec 10 Feb 11 Apr 11 Jun 11 Aug 11 Oct 11
IRSOISIR Basis Swap (rhs)
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 155/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Praveen Korapaty (1-212) 834-3092
Alberto Iglesias (1-212) 834-5116
J.P. Morgan Securities LLC
155
certain exemptions to prohibitions in the act.
Transactions in certain instruments, such as obligations
of the US government or a US government agency, theGSEs, and state and local governments are exempt from
the statute’s provisions. As expected, it also exempts
market-making, underwriting, and risk-mitigation
hedging. It also provides commentary to assist entities in
distinguishing market making-related activities from
proprietary trading.
Finally, we note that next year brings the observation
period for the Liquidity Coverage Ratio (LCR) proposed
under Basel III. Relevant regulators will start collecting
data to assess LCRs, with mid-2013 being the deadline
before which any changes to the current LCR proposal
may be announced. Our short-term analysts note that inits existing form, LCR is likely to be fairly onerous on
banks. In its current form, this requirement would result
in increased demand for scarce high-quality collateral.
Market developments and outlook
Despite extensions in the rule-making process, markets
have moved ahead towards a central clearing model. This
trend towards a clearing model predates the new
regulations (Exhibit 51). As of October 2011, around
half of the $511tn interest rate derivatives outstanding
were cleared. A majority of the cleared derivatives were
interest rate swaps (87%), which are about 60% of notional of interest rate derivatives outstanding. Over the
last year, roughly two-thirds of interest rate swaps
outstanding (based on USD-equivalent notionals) have
been centrally cleared, and the percentages of other
products currently cleared (OIS and basis swaps) have
grown substantially (Exhibit 52).
Another trend that predates post-crisis regulations is the
rise in the level of OTC derivative collateralization. As
seen in Exhibit 2, margin coverage in fixed income
derivatives is high, and covered over half the trade
volume well before the crisis.
One outcome of the move to central clearing and more
stringent collateral requirements has been the move to
OIS discounting as discussed earlier. The move would
result in a one-time accounting gain (loss) depending on
whether the counterparty was long (short) an in-the-
money swap position. More interesting is the question of
whether there would be a significant switch from Libor-
based hedges to OIS-based ones. Given the much higher
levels of liquidity in Libor swaps, we think this is
unlikely in the near term.
Another potential consequence of regulatory changes is a
change in issuer behavior. Corporate issuers that swap
their fixed-rate debt issuance may incur higher costs
irrespective of whether they may avail of the end user
exemption from mandatory clearing because their dealer
counterparties will incur higher capital costs on
uncleared swaps. The proposals could also impact
callable issuance. Many issuers of callables routinely sell
the embedded option on issuance to monetize favorable
pricing and achieve a lower funding cost. While
swaptions are not yet a cleared product, they will
nevertheless require their own collateral. Even though
this should not be an issue for some larger issuers like
FNMA and FHLMC that carry a significant amount of
eligible collateral, other issuers like FHLB that do not
(and non-US issuers such as KFW that do not typically
fully collateralize their derivative positions) would have
to evaluate the cost savings from callable issuance
against additional collateral costs. On the margin, this
could result in lower callable issuance and lower vol
supply. We note, however, that the rules are not finalized
with regards to which of these entities will be eventually
subject to mandatory clearing and margin requirements,
even as entities seek exemptions from various provisions
of the act.
Trading themes
• Look for FRA-OIS spreads to stay wide or widen
further into 1Q12, but begin to narrow thereafter
In the near term, worsening conditions in Europe are
likely to bias the EUR/USD cross currency basis
narrower, bias the semi-peripheral European sovereign
spreads wider, and preserve upward pressure on FRA-
OIS spreads. However, looking past this likely near-
term intensification, policy actions (a potential
reduction in the penalty on the Fed’s USD swap lines
to 50bp from 100bp) are likely to eventually dominate, pushing FRA-OIS spreads narrower to 50bp by the
end of 1H12.
• Swaptions on short tails are effectively options on
front-end swap spreads—use receiver/payer
swaptions in place of front-end spread
narrowers/wideners
With the Fed’s conditional commitment to low rates
until mid-2013 helping to keep front-end Treasury
yields low and sticky, front-end swap spreads have
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 156/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Praveen Korapaty (1-212) 834-3092
Alberto Iglesias (1-212) 834-5116
J.P. Morgan Securities LLC
156
become highly correlated to front-end swap yields.
Thus, options on front-end yields can effectively serve
as options on front-end swap spreads. Buy payer
swaptions or 1:1 weighted payer swaption spreads in
place of front-end swap spread wideners, and buy
receiver swaptions or receiver swaption spreads as
proxies for front-end spread narrowers.
• Position for wider 10-year maturity matched swap
spreads in the near term …
In the near term, FRA-OIS spreads will likely
continue to widen, and banking stock valuations will
likely remain under pressure, both of which should
pressure 10-year swap spreads wider. We expect 10-
year swap spreads to hover near 27bp towards the endof this year and in early 1Q12.
• … but look to initiate narrowers towards the end of
1Q12
High grade corporate issuance tends to exhibit strong
intra-year seasonal patterns, and issuance-related
swapping is likely to pick up over the March-May
period. Intermediate maturity swap spreads have
consequently tended to narrow in this period. In
addition, as we look into next year, the growing
likelihood of concerted action by central banks
(including perhaps lower penalties on USD swap lines
with the Fed) should also help swap spreads to turn thecorner. Look for narrowing to 18bp by mid-year.
• Trade intra-month swings in the pace of Fed
purchases of longer-end Treasuries
While the Fed tends to purchase similar amounts each
month in total, the actual schedule of purchases
(which is known ahead of the month) can produce
significant intra-month variations in pace.
Intermediate maturity swap spreads have tended to
narrow in periods characterized by a sharp slowing in
pace, while widening when the opposite is true.
• Take advantage of relative value opportunities insectors subject to Fed purchases
A simple strategy of buying the cheapest issue
(measured using yield error as a metric) and selling
the richest issue on a maturity matched swap spread
switch basis has been attractive in periods and sectors
where the Fed is actively purchasing Treasuries. With
Operation Twist slated to continue until the end of
1H12, we would look to make such switches a key
part of our relative value trading strategy in 1H12.
• Position for a steeper yield curve between
intermediates and the long end, hedged with red
Eurodollar shorts
Steepeners anchored in the 5-year sector and beyond
offer attractive carry and slide in addition to being
considerably mispriced relative to front-end yields.
Also, thanks to the flattening of the front end of the
curve, shorts in red Eurodollars have small carry costs,
and also help mitigate exogenous risks.
• Take advantage of yield curve nonlinearities
induced by the stickiness of front-end yields near
the zero bound
As front-end yields become highly sticky near the zero
bound, their relative volatility versus points further outthe curve falls, causing locally level and curve neutral
belly-richening butterflies to exhibit empirical
convexity versus front end yields. Butterflies that are
likely to exhibit the greatest asymmetry can be
identified by examining implied volatilities and skew
information. In addition, some of these butterflies
have positive carry and slide, making them
particularly attractive ways to position for an
unanticipated front end selloff.
• The yield curve should remain highly directional
with rates in 1H12; combine YCSOs with
swaptions to create “synthetic” conditional curvetrades that offer better entry levels than regular
conditional curve trades
Sticky front-end yields are likely to cause the yield
curve between the front end and intermediate yields to
remain highly directional with yield levels for the
foreseeable future. However, swaptions markets have
priced such directionality in (albeit to varying
degrees), and entry levels are often unattractive. By
exploiting the correlation of the 10s/30s swap curve
with front-end yields, and given the richness of
YCSOs on the 10s/30s curve, replacing swaptions at
the front end of the curve with suitably weighted
YCSOs results in a “synthetic” conditional curve tradethat can be constructed at much better entry levels.
• Stay long gamma going into year end and in 1H12
A persistent crisis in Europe and poor risk appetite
should cause market depth to stay depressed for much
of 4Q11 and 1H12, creating conditions favorable for
long gamma positions.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 157/236
US Fixed Income Strategy
US Fixed Income Markets 2012 Outlook
November 24, 2011
Srini RamaswamyAC
(1-212) 834-4573
Praveen Korapaty (1-212) 834-3092
Alberto Iglesias (1-212) 834-5116
J.P. Morgan Securities LLC
157
• Look for intermediate-expiry swaption implied
volatility to drop into 1Q12, but stay range-bound
thereafterIntermediate expiry swaption implied volatility
currently looks rich; look for 3Yx10Y swaption
implied volatility to fall to 6bp/day by the end of
1Q12, and remain range-bound thereafter.
• Look to add exposure to Bermudan receiver
swaptions
Cheap implied correlations and an inverted implied
volatility curve have made forward volatility cheap. In
addition, the strong relationship between implied
volatility and rates that emerges in low yield regimes
makes it desirable to hold positions that become
longer vega in a sell-off. One attractive way to initiatedirectional vega exposure, while also positioning for a
rise in implied correlations and forward volatility, is to
replace European receiver swaptions with Bermudan
receiver swaptions.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 158/236
Japan Rates ResearchGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Takafumi YamawakiAC
(81-3) 6736-1748
Yuya Yamashita, CFA (81-3) 6736-1493J.P. Morgan Securities Japan Co., Ltd
158
Japan
• 2011 was a noteworthy year for the JGB market
in three respects: 1) The Great East Japan
Earthquake of 11 March 2011 2) JGB-UST
decoupling, and 3) A weakening of the ultra-long
sector. However, JGB stayed within a narrow
range of 0.95-1.35% in 2011
• Our main scenario points to a continuation of
range-bound trading while our risk scenarios
suggest that JGB yields are more likely to fall
than rise. The downside in yields could be tested
if the global economy loses momentum in 1H2012
while a gradual cyclical recovery in 2H12 should
trigger only moderate rises in yen interest rates.We are therefore anticipating ranges of 10Y JGB
yields as 0.8-1.1% for 1H12 and 0.9–1.3% for
2H12
• This report considers why Japanese sovereign
debt can be considered relatively ‘safe’ by
international standards while also identifying
some cause for concern in the longer term
• We would consider any rise in JGB yields that
might be triggered by the FY12 issuance plan to
be little more than extremely short-lived noise
• We expect short-end swap rates to remain sticky
under the BoJ’s interest rate policy. We also have
in mind a risk scenario where LIBOR rises on a
possible worsening of the crisis in Europe. We
have a flattening bias on JPY swap curve as a
whole
• We have a widening bias on short-end swap
spreads. We think the cheapness of JGB vs. swap
rates in the super-long sectors will be sustained
into 2012
• Large negative cross-currency basis spreads
make it attractive to create synthetic USD bonds
by buying JGBs and swapping them to USD.
With higher FX hedge cost, Japanese investorsmay accelerate the repatriation of USD funds
back to JGBs
• We have a tightening bias on the TIBOR/LIBOR
fixing spread on positive convexity. We expect
more tightening if LIBOR increases than
widening if LIBOR falls
• We think the short-to-medium-term 3s/6s basis
spreads are biased wider
The JGB market in 2011
Three key points – What sort of year has 2011
been for the JGB market?
US Treasury yields rose sharply in January and February
as equity prices were driven higher by the Fed's late-2010
QE2 announcement and the Obama administration's
decision to extend the so-called ‘Bush tax cuts’. The 10Y
JGB yield hit its year-to-date high of 1.35% as yen bonds
were caught up in this global selloff, but the domestic
market then turned unusually volatile in the wake of the11 March 2011 Great East Japan Earthquake, with losses
on stock positions apparently triggering a certain amount
of profit taking in JGBs towards the 31 March end of the
Japanese fiscal year (Exhibit 1)
Yen interest rates followed UST yields lower from mid-
April through July, but then failed to keep pace with
further declines from August onwards. The year as a
whole has thus been characterised by a rather gradual
downtrend in JGB yields since May.
2011 was a noteworthy year for the JGB market in three
respects.
1: The Great East Japan Earthquake of 11 March 2011
The Great East Japan Earthquake of 11 March 2011
appears set to be one of the defining events of Japan's
already long history.
While the initial earthquake ended up causing relatively
little direct damage, the subsequent tsunami destroyed a
number of towns and villages and triggered a major crisis
at the Fukushima Daiichi nuclear power plant. Japan is
Exhibit 1: JGB stayed in a narrow range of 0.95-1.35% in 201110YJGB yield history; %
0.80
0.90
1.00
1.10
1.20
1.30
1.40
1.50
Jan-10 Jul-10 Jan-11 Jul-11
QE2
Earthquake
Decoupling
from UST
high correlation
with UST
1.35%
0.95%
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 159/236
Japan Rates ResearchGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Takafumi YamawakiAC
(81-3) 6736-1748
Yuya Yamashita, CFA (81-3) 6736-1493J.P. Morgan Securities Japan Co., Ltd
159
by no means unaccustomed to earthquakes and tsunami,
but the virtually unprecedented size of this tsunami
nevertheless shocked the nation, with significant negativeramifications for consumer spending, manufacturing
supply chains, and electricity supply capacity. The
collapse and rebound of industrial production provides a
particularly clear picture of the extent to which economic
activity was impacted by the disaster (Exhibit 2).
The following couple of months saw JGB market
participants focus on: 1) the potential for post-quake
reconstruction to drive a V-shaped economic recovery;
and 2) the prospect of an increase in JGB issuance to pay
for post-quake rebuilding. The subsequent economic
recovery fell somewhat short of initial expectations,
however, and with the JGB market asked to absorb onlya limited amount of additional issuance, a significant rise
in yen interest rates was averted.
2: The Euro zone crisis and JGB-UST decoupling
UST-JGB correlations declined markedly as the
European sovereign debt crisis dominated headlines from
August onwards. UST and German Bund yields fell
sharply as the US economy too began to lose momentum,
but JGB yields dropped only slightly during this period.
Exhibit 3 illustrates the extent to which yen interest rates
were ‘left behind’ by the fall in UST yields.
3: A weakening of the ultra-long sector
The performance of ultra-long JGBs was a key difference
between 2010 and 2011. The 30s/40s curve steepened
sharply this year (Exhibit 4), as demand for these
longest-dated bonds dropped dramatically while demand
for other parts of the super-long sector remained strong.
The 40s have a very thin investor base compared with
that of the 20s, however, and we will be monitoring this
recent slackening of the supply/demand in the 40s to see
if it continues to put steepening pressure to the JGB yield
curve in 2012.
Exhibit 3: UST-JGB correlations declined markedly from Augustonwards.UST 10Y vs. JGB 10Y since October 2010; %
0.70
0.80
0.90
1.00
1.10
1.20
1.30
1.40
1.50
1.50 2.00 2.50 3.00 3.50
JGB 10y r
UST 10y r
Oct10 - July11
Aug11-Nov11
Exhibit 2: The collapse and rebound of industrial production dataprovide a particularly clear picture of the extent to whicheconomic activity was impacted by the disaster Industrial production seasonal adjusted; CY 2005 = 100
75
80
85
90
95
100
Jul-10 Oct-10 Jan-11 Apr-11 Jul-11
Earthquake
Exhibit 4: The 30s/40s curve steepened sharply this year JGB 30s/40s curve history; bp
0
5
10
15
20
25
30
Oct-10 Jan-11 Apr-11 Jul-11 Oct-11
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 160/236
Japan Rates ResearchGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Takafumi YamawakiAC
(81-3) 6736-1748
Yuya Yamashita, CFA (81-3) 6736-1493J.P. Morgan Securities Japan Co., Ltd
160
Outlook for the JGB market in 2012
Six factors support a relatively narrow trading
range
While we do not expect volatility to remain as low as it
has been in 2H11, we nevertheless expect to see a
continuation of relatively range-bound trading in 2012.
Looking at the key factors summarised in Exhibit 5, we
see little prospect of significant yield movements in
either direction. Below, we discuss our main scenario
and risk scenarios for each of these factors.
1: Trade balance
Main scenario: Large current account surplus despite
near-zero trade balance (neutral bias on yields: 75%)Exhibit 6 shows Japan’s trade balance, which remained
in deficit this year as exports fell in the wake of the
earthquake, and import value was driven up by high
commodity prices. The trade deficit exceeded ¥400bn
(seasonally adjusted) in both April and May, but had
narrowed to ¥21.8bn as of September. Japan may find it
difficult to book a consistent trade balance surplus in
2012, given the prospect of exports being held back by a
global economic slowdown, but we expect it to remain
roughly in equilibrium (i.e. near zero) over the year as a
whole. This would see Japan's massive income surplus
translate into a similarly large current account surplus,
implying that the market should have little problem
absorbing even relatively high levels of JGB issuance.
Risk scenario: Severe overseas economic slowdown or
rise in commodity prices (upward bias on yields: 20%)
Japan’s trade balance would likely plunge deep into
deficit under a severe global recession, thereby fueling
speculation that a current account surplus might be
difficult to sustain in the longer term and potentially
triggering a selloff of JGBs by foreign investors.
Exhibit 6: Japan's trade balance remained in deficit this year and thecurrent account surplus has been gradually shrinking recentlyCurrent account surplus in Japan; ¥tn
-10
-5
0
5
10
15
20
25
30
35
96 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11
Income Trade Balance
C urrent T rans fer Serv ic e
Exhibit 5: Six factors support for the JGB market moving in a relatively narrow range, but with a downward bias in yieldsSix factors for the JGB market, their bias towards yield move and our subjective probabilities assigned to each scenario
ProbabilityYield
Down
Yield
Neutral
Yield
UpComments
1 Trade Balance 5% 75% 20%If trade balance stays in the negative for a long time, some investors
may have concerns about JGB sustainability.
2 BoJ Additional Easing 35% 60% 5%If the BoJ indroduce further easing such as increasing Rinban,
extending the maturity of the Asset Purchase Program, and lowering the IOER
3 Domestic Politics 10% 65% 25%If Japanese economy goes into a recession, tax increase might
be extended.
4 Lifer's Investment Activity 40% 50% 10%If US rates stay around the current levels for a long time, lifers will
shift their investment money from USTs to JGBs.
5 Bank Activity 30% 60% 10%If banks face difficulties in non-performing loans, they will buy
JGBs more aggressively.
6 UST Rates 30% 50% 20% If UST stay below 2%, JGB yields will follow USTs.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 161/236
Japan Rates ResearchGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Takafumi YamawakiAC
(81-3) 6736-1748
Yuya Yamashita, CFA (81-3) 6736-1493J.P. Morgan Securities Japan Co., Ltd
161
2: Additional BOJ easing
Main scenario: Additional easing limited to further
expansion of Asset Purchase Program (APP) (neutral bias on yields: 60%)
The Bank of Japan eased monetary policy on three
occasions following the 11 March disaster.
14 March 2011: Increase in the amount of APP: + ¥5tn
4 August 2011: Increase in the amount of APP: + ¥10tn
27 October 2011: Increase in the amount of APP + ¥5tn
However, each of these actions was limited to an
expansion of the central bank’s APP, with the BoJ
refraining from exercising any of its other options (2-4
below):
1) Further expansion of APP
2) Lengthening of duration of APP JGB purchases
(from 2 years to 5 years)
3) Lowering of the interest rate paid on excess
reserve balances
4) Increase in Rinban outright JGB purchases
With overseas central banks seemingly reluctant to lower
IOER (interest on excess reserves) rates or increase their
outright bond purchases, we see little prospect of the BoJ
opting for any of the options (2-4) unless the domestic
economy were to slow dramatically. Central banks are
increasingly prone to criticism for both direct
underwriting of government debt and excessive
purchases via the secondary market. With the BoJ
understandably keen to avoid any negative surprises on
the monetary policy front, we do not expect to see any
announcements with sufficient impact to knock JGB
yields out of relatively narrow trading ranges.
Risk scenario: Yen driven higher by aggressive Fed
easing (downward bias on yields: 35%)
There is at least some chance that the BoJ will bow to
pressure from markets and politicians by announcing
slightly more aggressive measures, with (2) perhaps themost likely option (other than a further APP expansion)
by virtue of its relative ease of implementation. If
USD/JPY were to drop below 70 following a QE3
announcement by the Fed, then we would expect the BoJ
to respond by lengthening the duration of its APP JGB
purchases, thereby generating downward pressure on the
curve as a whole and the 5Y yield in particular.
3: Domestic political situation
Main scenario: Decision on consumption tax hike helps
to maintain a minimum level of fiscal discipline (neutral bias on yields: 65%)
Fiscal discipline became a major policy priority after
Yoshihiko Noda took over from Naoto Kan as Prime
Minister, with the need for sharply higher spending in the
wake of the 11 March disaster seemingly
counterbalanced by a strong commitment to responsible
fiscal management. The Noda administration also
appears determined to hike the consumption tax rate. In
fact, the administration has set its guidline to hike the
consumption tax rate from 5% to 10% in two stages from
October 2013. Fiscal policy thus appears unlikely to be
an Achilles heel for the JGB market provided that the Noda government can remain in office.
Risk scenario: Political squabbling sees fiscal discipline
thrown out the window (upward bias on yields: 25%)
A severe global economic downturn could potentially see
the fiscally hawkish Noda administration overthrown, in
which case the new government could be expected to
loosen the fiscal reins and relegate a consumption tax
hike.
4: Investment by life insurers
Main scenario: Shift out of foreign bonds into JGBs(neutral bias on yields: 50%)
Japan has also seen an increase in USD funding costs
(Exhibit 7), with 3M roll cost rising to around 0.8%
annualised. This means that less than 1.5% of carry is
now earned by buying 10Y USTs, as a result of which
domestic investors have started to shift back into JGBs.
The focus is life insurance companies as their activities
are the key for the demand and supply balance of the
long end of the curve. Their buying has been already
modestly increasing over the past few years (Exhibit 8).
Some investors may also have replaced a portion of their
FX-hedged foreign bond holdings with unhedged
positions, but full substitution appears unlikely given the
risk it would entail. We therefore expect to see a gradual
shift into (super-long) JGBs that should help to limit any
rise in yields towards the far end of the curve.
Risk scenario 1: Renewed investment in foreign bonds
following a resolution of the Euro zone crisis (upward
bias on yields: 10%)
Confidence in USD funding may be boosted if a
significant improvement in the Euro zone situation drives
10Y UST and German bund yields substantially higher,
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 162/236
Japan Rates ResearchGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Takafumi YamawakiAC
(81-3) 6736-1748
Yuya Yamashita, CFA (81-3) 6736-1493J.P. Morgan Securities Japan Co., Ltd
162
in which case European bonds and USTs could start to
look more attractive to domestic lifers from a relative
value perspective (versus JGBs).Risk scenario 2: Increased investment in JGBs as the
Euro zone situation continues to deteriorate (downward
bias on yields: 40%)
A further escalation of the Euro zone crisis would
probably make lifers increasingly risk averse, with a
selloff in domestic equities – and the corresponding need
to reduce risk exposure – likely to accelerate purchases
of super-long JGBs.
5: Investment by banks
Main scenario: Little change in earnings, little need for
aggressive trading (neutral bias on yields: 60%)
Recent financial results indicate that bad-debt disposal
costs remain well under control, suggesting that there is
little need for banks to chase relatively risky returns in
the JGB or the broader securities markets. We therefore
expect to see only moderate trading activity, with banks
likely to take profits when yields fall and buy on (price)
dips when yields rise. This should help limit overall
market volatility.
Risk scenario: Economic downturn increases reliance on
trading profits (downward bias on yields: 30%)
It is possible that a further escalation of the Euro zonecrisis could have negative ramifications for Japan and the
rest of Asia, in which case domestic banks may be forced
to step up their bond trading with a view to covering
higher bad loan costs. This could result in a slightly more
volatile market, with JGB yields likely to decline at least
temporarily.
6: Correlation with US Treasuries
Main scenario: JGB yields unlikely to rise sharply if the
10Y UST yield remains somewhere around 2.5% (neutral
bias on yields: 50%)
As should be evident from Exhibit 3, the JGB marketwas much more sensitive to movements in UST yields
prior to August 2011. A return to pre-August correlations
would imply that a 10Y UST yield of around 3.0%
should be commensurate with a 10Y JGB yield of around
1.20%. (Our UST research team is forecasting that the
10Y UST will stay below 3.0% throughout 2012.)
Our main scenario for 2012
Our main scenario for each of the above six factors
points to the continuation of range-bound trading, while
our risk scenarios suggest (on balance) that JGB yields
are more likely to fall than rise. The downside could be
tested if the global economy loses momentum in 1H12,
while a gradual cyclical recovery in 2H12 should trigger
only moderate rises in yen interest rates.
We are therefore anticipating ranges of 10Y JGB
yields as 0.80%–1.10% for 1H12 and 0.90%–1.30%
for 2H12.
Exhibit 8: Lifers’ activities are the key for the demand and supplybalance of the long end of the curve. Their buying has been modestlyincreasing over the past few years3M moving average of monthly purchase amount in super long end from lifers; ¥bn
0
200
400
600
800
1,000
1,200
1,400
Apr-08 Apr-09 Apr-10 Apr-11
3M moving average
Exhibit 7: The increase in USD funding cost has reduced theattractiveness of USTs. We therefore expect to see a gradual shift into(super-long) JGBs that should help to limit any rise in yields towards
the far end of the curve3M FX hedge cost in USD/JPY: % annualised
0.00
0.20
0.40
0.60
0.80
1.00
Oct-10 Jan-11 Apr-11 Jul-11 Oct-11
* 3M FX hedge cost is calculated as annualized 3M forward discount vs. spot FX level.Source: Bloomberg
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 163/236
Japan Rates ResearchGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Takafumi YamawakiAC
(81-3) 6736-1748
Yuya Yamashita, CFA (81-3) 6736-1493J.P. Morgan Securities Japan Co., Ltd
163
Japanese sovereign risk
Still relatively safe, but some longer-termconcerns remain
A continued escalation of the European debt crisis has
focused substantial attention on Japanese sovereign risk,
but the stability of domestic interest rates over the past
few years suggests that a ‘sovereign shock’ is still quite a
remote prospect. This report considers why Japanese
sovereign debt can be considered relatively ‘safe’ by
international standards, while also identifying some
cause for concern in the longer term.
Why the JGB market is still relatively safe?
1: Massive domestic savings: A healthy balance sheetfor the nation as a whole
Exhibit 9 shows (inward and outward) financial assets
and liabilities for the various sectors of the Japanese
economy, with equities counted separately owing to their
significant price volatility. These figures should help
give an overall picture of Japanese financial flows and
leverage.
The household sector is seen to hold net financial assets
totaling about ¥1,048tn, while the social security
system’s pension reserves amount to about ¥175tn in net
assets. The (central + local) government sector has netliabilities totaling ¥840tn, while non-financial private
corporations have around ¥39tn in net liabilities.
Foreigners’ net liabilities total ¥348tn, implying an
equivalent amount of Japanese claims against foreign
entities.
Overall, we believe that Japan's balance sheet is healthy,
with the existence of massive domestic asset holdings a
key difference from the current situation in Europe.
2: High levels of issuance digested with little difficulty
The past behaviour of these various sectors should help
explain why the market as a whole has had little
difficulty in absorbing continued increases in JGB
issuance.
• Foreigners’ net liabilities (= Japan’s net assets) tend to
decline when the yen is strengthening and increasewhen the domestic currency is weakening. In the
absence of particularly dramatic exchange rate
fluctuations, however, it seems likely that Japan will
continue to gradually accumulate net assets by virtue
of its consistent current account surplus.
• The existence of a current account surplus means that
the government sector’s fiscal deficit is being more
than offset by increases in the net assets of the private
sector (households + corporate). The corporate
sector’s net financial liabilities have declined sharply
in recent years owing to a diminished need for
borrowing, while the household sector has continuedto build up its net assets. Some analysts have
suggested that Japan’s adverse demographics – an
Exhibit 9: Given massive domestic asset holdings, we believe that Japan's balance sheet is healthyFinancial assets and liabilities by sectors; ¥tn
Inward Asset Inward Liability Inward Asset Inward Liability Inward Asset Inward Liability
2,430 2,642 607 480 1,392 443
Outward Asset Outward Liability Outward Asset Outward Liability Outward Asset Outward Liability
244 30 106 3 9 0
Net Asset Net Asset Net Asset
10 -39 1,048
Equity Asset Equity Liability Equity Asset Equity Liability Equity Asset Equity Liability113 105 138 407 90
Inward Asset Inward Liability Inward Asset Inward Liability Inward Asset Inward Liability
128 1,109 147 24 212 217
Outward Asset Outward Liability Outward Asset Outward Liability Outward Asset Outward Liability
88 2 31 0 34 472
Net Asset Net Asset Net Asset
-840 175 -348
Equity Asset Equity Liability Equity Asset Equity Liability Equity Asset Equity Liability
73 18 21 95
Financial Institution Non Fin Corporate Household
Government Social Security Foreigners
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 164/236
Japan Rates ResearchGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Takafumi YamawakiAC
(81-3) 6736-1748
Yuya Yamashita, CFA (81-3) 6736-1493J.P. Morgan Securities Japan Co., Ltd
164
aging and shrinking population – will eventually push
the household sector’s net assets into a downtrend, but
this has thus far been prevented by significanttransfers of income from the government sector to
households (Exhibits 10&11).
The resulting decline in corporate demand for funds and
increase in deposits have left banks awash with surplus
funds to invest in JGBs, thereby helping to keep yen
interest rates low and relatively stable. We expect that
this mechanism will continue to function for as long as
Japan’s current account balance remains in surplus.
3: Current account surplus has declined slightly but is
unlikely to disappear
It is of course necessary to recognise that Japan’s current
account surplus has been gradually shrinking recently as
shown in Exhibit 6. The trade balance deteriorated
sharply in the wake of the Great East Japan Earthquake
of 11 March, owing to supply chain disruptions and a
surge in oil imports. We expect to see only a moderate
surplus going forward given the likelihood that Japan
will continue to import fossil fuels while exporters are hit
by a global economic downturn.
Japan continues to book a large income surplus,
however, and we expect this to remain somewhere in the
order of ¥10tn per annum unless there is a significantdecline in net claims against foreigners. This should also
provide substantial support for the JGB market.
4: Difficult to envisage a default scenario
While there has indeed been some discussion of the
possibility that Japan might default on its debt, we
consider this an extremely unlikely scenario for the
following reasons.
Shifting out of JGBs during a ‘flight to quality’ will
entail foreign currency risk, making this a very difficult
option to contemplate for most market participants.(Shifting from Greek bonds into German bunds involves
no such risk). We expect support from domestic
investors not willing to take currency risk.
• Japan has its own central bank, which should be able
to purchase JGBs in the event of a crisis. This makes it
theoretically possible for Japan to avoid a default
indefinitely, i.e. the BoJ is theoretically able to
underwrite JGBs in the primary market with necessary
law changes.
• Even if the yen were to weaken as a consequence of
JGB purchases by the BoJ, Japan’s ample foreign
reserves mean that this would actually help to reduce
the state’s net debt. (In the case of Argentina,
depreciation of the home currency triggered a sharp
increase in USD-denominated liabilities.)
• A depreciation of the yen would also be likely to boost
exports and thereby bring about a significant increase
in the current account surplus, with positive
ramifications for the JGB market.
We therefore see very little risk of Japan facing a Greece-
like sovereign debt crisis.
Exhibit 10: Despite adverse demographics, the household sector hascontinued to build up its net assets…History of net asset in household sector; ¥tn
700
750
800
850
900
950
1,000
1,050
1,100
1997 1999 2001 2003 2005 2007 2009
Exhibit 11: … on the back of significant transfers of income from thegovernment sector to householdsHistory of net liability by sectors; ¥tn
0
100
200
300
400
500
600
700
800
900
1997 1999 2001 2003 2005 2007 2009
Government
Foreigners
Corporate
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 165/236
Japan Rates ResearchGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Takafumi YamawakiAC
(81-3) 6736-1748
Yuya Yamashita, CFA (81-3) 6736-1493J.P. Morgan Securities Japan Co., Ltd
165
5: Some gradual progress towards more stable fiscal
footing
There were concerns that the three supplementary budgets made necessary by the Great East Japan
Earthquake of 11 March would require a sharp increase
in JGB issuance, but the government has been able to
deplete frontloaded JGB issuance to the point where the
market will only be asked to absorb an additional ¥2tn
(or thereabouts), which suggests that any impact on the
overall supply/demand balance should be minimal.
Moreover, the government appears to be making slow-
but-steady progress in its efforts to return Japan to a
more sustainable fiscal footing.
• The government’s medium-term fiscal framework
covering the period through FY2014 (endorsed by the
Cabinet in August) seeks to limit annual general
account spending (excluding debt-servicing costs) to
¥71tn and new JGB issuance to ¥44tn. Cabinet Office
projections indicate that meeting these targets would
still leave Japan unable to achieve its longer-term goal
of wiping out the primary fiscal deficit by FY2020,
but it is nevertheless impressive that Japan has clearly
announced its intention to maintain a certain level of
fiscal discipline despite the massive shock of the 11
March disaster.
• The government has also indicated that the post-quake
rebuilding process is to be managed outside thegeneral budget, with the necessary funds to be repaid
via temporary tax hikes – including higher personal
income and residential taxes – and non-tax revenues.
• The Noda administration has set its guidline to hike
the consumption tax rate from 5% to 10% in two
stages from October 2013.
Cause for concern in the longer term
6: Massive imbalance between the government and
private sectors
Japan faces a severe imbalance where the governmentsector’s massive net liabilities (currently exceeding
¥800tn) are covered in their entirety by the domestic
private sector (households + non-financial private
corporations). It is also somewhat troubling that the
banking sector holds such a large proportion of the
private sector’s net assets. This lack of diversification has
the potential to exacerbate any upward movement in yen
interest rates.
As should be evident from Exhibit 12, there have been
two particularly sharp rises in JGB yields over the past
decade:
• June–August 2003, when a surge in VaR for the
banking sector triggered an across-the-board reduction
of interest rate risk, and
• 2005–2006, when banks sought to reduce their
exposure to medium-term JGBs ahead of a BoJ rate
hike.
Insufficient diversification on the part of investors was a
contributing factor in both cases.
7: Super-long issuance outstanding continues to rise
A sharp increase in outstanding issuance of super-long
JGBs is also worthy of some attention. Total outstanding
issuance beyond the 11Y sector has risen by around
¥40tn over the past two years and is projected to exceed
¥140tn by end-FY2011 (Exhibit 13). This extremely
rapid pace of supply reflects large and regular offerings
via 20Y, 30Y, and 40Y tenders as well as Auctions for
Enhanced Liquidity.
Japanese life insurers are already preparing for the
launch of a domestic version of the Solvency II
framework (currently being implemented in Europe)
around 3 – 4 years from now, and their associated ALM
measures appear likely to ensure relatively strong
demand for super-long JGBs over the next 2 – 3years.
Beyond that point, however, it may be necessary to find a
new investor base for super-long offerings. The current
debate over a possible lengthening of the average
Exhibit 12: The lack of diversification in the banking sector has beenthe main driver of the sharp rises in JGB yields over the past decadeLong history of JGB 10Y yield; %
0.20
0.40
0.60
0.80
1.00
1.20
1.40
1.60
1.80
2.00
00 01 02 03 04 05 06 07 08 09 10 11
Banks reduced
their rates exposure
in a rate hike cycle
Bank
TradingBank
Trading
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 166/236
Japan Rates ResearchGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Takafumi YamawakiAC
(81-3) 6736-1748
Yuya Yamashita, CFA (81-3) 6736-1493J.P. Morgan Securities Japan Co., Ltd
166
duration of JGB holdings for public pension portfolios
could be a step in the right direction.
Long-term outlook for the JGB market
As indicated in Exhibit 14, we expect the 10Y JGB yield
to remain below 1.5% over the coming 3 – 4 years before
shifting into a 1.5%–2.0% range for the next few years,
with fiscal risk unlikely to become a major market factor
until at least a decade from now.
Exhibit 13: Total outstanding issuance beyond the 11Y sector isprojected to exceed ¥140tn by end-FY2011, increasing the need tofind a new investor base for super-long offerings.
Total market outstanding of 10Y+ JGB excluding 15Y CMT, nominal bonds only; ¥tn
0
50
100
150
200
03 04 05 06 07 08 09 10 11 12 13 14
Forecast
Exhibit 14: We expect the 10Y JGB yield to remain below 1.5% over the coming 3–4 years before shifting into a 1.5%–2.0% range for thenext few years, with fiscal risk unlikely to become a major marketfactor over the next 10 yearsLong-term forecast of 10Y JGB yield; %
0.00
0.50
1.00
1.50
2.00
2.50
3.00
11 12 13 14 15 16 17 18 19 20 21 22 23 24
Short Term
Rangebound
10Y=0.8-1.5%
Medium Term
High Volatility
10Y=1.5-2.0%
Long Term
Gradual Yield Rise
10Y>2.0%
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 167/236
Japan Rates ResearchGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Takafumi YamawakiAC
(81-3) 6736-1748
Yuya Yamashita, CFA (81-3) 6736-1493J.P. Morgan Securities Japan Co., Ltd
167
Forecasts for FY2012 calendar basemarket issuance
The Ministry of Finance tends to announce its JGB
issuance plan for the following fiscal year around
Christmas. As indicated in Exhibit 15, we expect the
FY2012 issuance plan (which is also likely to be
announced before the end of 2011) to show an increase
of ¥300-400bn per month in calendar base market
issuance (supply at scheduled auctions from April
through March) relative to the offering amounts that will
be seen from December 2011 onwards. We would expect
this to break down into a ¥100bn/month increase in
issuance of 1Y, 2Y and 5T JGBs and a further
¥100bn/month increase in the 20Y sector (Exhibit 16).
Increases of this magnitude are unlikely to have a
significant impact on overall market sentiment, and we
would therefore consider any rise in JGB yields that
might be triggered by the announcement to be little more
than extremely short-lived noise.
Our key assumptions are as follows.
We currently expect the FY2012 issuance plan to show:
• No more than ¥44tn in new financial resource bonds
(thereby meeting the target set out in the government’s
medium-term fiscal framework)
We expect this to break down as follows:
• Around ¥150tn in calendar base market issuance
(offerings to regular auction participants)
• ¥4.2tn in Non-price Competitive Auction II issuance
• ¥2tn in adjustments by depleting front-loaded issuance
• Around ¥3.5tn in issues to individual (retail) investors,
which would be on a par with the third FY2011
supplementary budget
• Around ¥16tn in long-term JGBs held by the BoJ are
due for redemption during FY2012
Exhibit 16: … and we would expect this to break down into a¥100bn/month increase in issuance of either 2Y or 5TY JGBs anda further ¥100bn/month increase in the 20Y sector Calendar base JGB issuance in FY2012; JPYtn
Source: J.P. Morgan
Exhibit 15: FY2012 issuance plan (which is also likely to be announced before the end of 2011) to show an increase of ¥300-400bn per month incalendar base market issuance…JGB issuance plan (JPM estimate) in FY2012; JPYtn
FY2011 FY2012
3rd Suple Lowest Forecast Highest
Legal Grounds 181.7 173.9 ~ 176.1 ~ 179.2
New Resource JGBs 44.3 43.9 ~ 44.0 ~ 44.1 Government Target
Refunding JGBs 109.3 113.5 ~ 114.0 ~ 114.5 Around 114 trJPY
FILP JGBs 16.5 12.0 ~ 13.0 ~ 14.5 Redemption of FILP bonds is smaller than FY2011
Reconstruction JGBs 11.6 2.0 ~ 2.5 ~ 3.5 3.5trn at a maximum JGBs for govt pension 0.0 2.5 ~ 2.5 ~ 2.6 Temp issuance for Govt pens ion
FY2011 FY2012
3rd Suple Lowest Forecast Highest
Issuance Methods 181.7 176.1
Calendar Base 145.7 148.0 ~ 149.7 ~ 152.0 Calendar base will be around 150trn
Non-Price Comp 5.9 4.2 ~ 4.2 ~ 5.0 3.75% of total amount of 2Y + JGBs
Depleting Front Loaded 14.8 1.0 ~ 2.5 ~ 4.0
For Households 3.5 2.5 ~ 3.5 ~ 4.0 Same size as 3rd supplementary budget of FY11
Bank of Japan 11.8 15.0 ~ 16.2 ~ 18.0 Same size as the redemption in BoJ portfolio in FY12.
Comment
Comment
FY11 FY2012
Dec~ Amt Total
6mTB 0.9 0.9 1 0.9
1Y TB 2.5 2.5 12 30.0
2Y JGB 2.7 2.7 12 32.4
5Y JGB 2.5 2.6 12 31.2
10Y JGB 2.2 2.2 12 26.4
20Y JGB 1.1 1.2 12 14.4
30Y JGB 0.7 0.7 8 5.6
40Y JGB 0.4 0.4 4 1.6Liquidity 0.6 0.6 12 7.2
149.7
Calendar
Base
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 168/236
Japan Rates ResearchGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Takafumi YamawakiAC
(81-3) 6736-1748
Yuya Yamashita, CFA (81-3) 6736-1493J.P. Morgan Securities Japan Co., Ltd
168
Interest Rate Derivatives
Swap Curve
Over the past year, the swap curve bull-flattened up to
10Y sectors while the curve in the super-long sectors was
shifted down in a parallel-to-modest bull-steepening
fashion (Exhibit 17). The direction and magnitude were
generally in line with the movement of the JGB yield
curve. Even after markets recovered from the post-
earthquake turmoil, the BoJ continued to provide
abundant liquidity in response to the deepening debt
crisis in Europe and global economic slowdown. In
addition, the Fed introduced an explicit time frame of
interest rate policy until mid-2013 in August and then the
so-called Operation Twist in September. The BoJ is
expected to hold the policy rate at a low level even
longer than the Fed. The bull-flattening of the JPY swap
curve up to 10Y sectors this year is consistent with such
global low-for-long environment.
6M LIBOR effectively lays a floor to the swap rates up
to 5Y. Therefore, in a bullish environment, the level of
the shorter-term swap rate tends to be well supported
while longer end rates have more room for downshift
given a slow fall in 6M LIBOR. Therefore, in the course
of a downshift of the curve, short-end flatteners begin to
show convexity from a certain point. And butterflies,
after the belly richens as rates decline, begin to cheapenfrom a certain point. In fact, as the curve flattens further
out, the 2s/5s/10s fly did not richen in a rally this year.
On the other hand, the 5s/10s/20s spread has been on a
richening trend for about a year (Exhibit 18) and may
start to show convexity going forward. The long end of
the curve remained flat. The inversion in the long end of
the forward swap curve, which had first taken place after
Lehman shock in 2008, was not corrected either this
year, although the extent of inversion declined. The
10Yx10Y / 20Yx10Ycurve is still in negative territory at
-36bp (Exhibit 19).
The supply and demand balance in 30Y sector remainedone-sided in 2011 with strong receiving of the long end
of the curve. Firstly, declining FX rates and risky assets
put a flattening pressure on the long end of the curve.
While the issuance sum of structured notes remained
small, hedging activities for existing positions continued.
Secondly, asset swap investors stayed away from the
30Y sector. Domestic investors buy ASW up to 20Y.
Overseas investors, who had once invested in 30Y JGBs
as a LIBOR+ floating asset by using ASW, did not come
back to the market. Lastly, hedge funds sporadically paid
Exhibit 17: Swap curve bull-flattened up to 10Y sectors while thecurve in the super-long sectors was shifted down in a parallel tomodest bull-steepening fashion over the past year Change over the past year;
% bp
0.00
0.50
1.00
1.50
2.00
0 5 10 15 20 25 30
-30
-25
-20
-15
-10
-5
0
5Change (RHS)
Term structure (LHS)
Exhibit 18: As the curve flattens further out, the 2s/5s/10s fly did notrichen in a rally this year whereas the 5s/10s/20s has richenedconsistentlySwap 2s/5s/10s and 5s/10s/20s 50;50 butterfly since 2011; bp
-16
-14
-12
-10
-8
-6
-4
-2
0
Jan-11 Apr-11 Jul-11 Oct-11
-50
-45
-40
-35
-30
-25
-20
2s/5s/ 10s Butterfly (RHS)
5s/10s/20s Butterfly (LHS)
Exhibit 19: The inversion in the long end of the forward swap curveremained, though the extent of inversion declined. The10Yx10Y/20Yx10Y curve is at -36bp10Yx10Y / 20Yx10Y curve since 2011; bp
-60
-50
-40
-30
-20
-10
0
Jan-11 Apr-11 Jul-11 Oct-11
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 169/236
Japan Rates ResearchGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Takafumi YamawakiAC
(81-3) 6736-1748
Yuya Yamashita, CFA (81-3) 6736-1493J.P. Morgan Securities Japan Co., Ltd
169
super-long swap rate and bought long-dated payer
swaptions as a so-called ‘Japan fiscal premium trade’
with the anticipation of an increased Japanese fiscaldeficit leading to higher long-term yields. However, they
generally preferred benchmark 10Y tail to express the
view compared with 20Yx10Y for a liquidity reason
because the trade was essentially more macro-driven.
Going forward, for short-end rates, we expect them to
remain sticky under the BoJ’s interest rate policy as a
main scenario. At the same time, we also have in mind a
risk scenario that LIBOR rises on a possible worsening
of the crisis in Europe, which we think is getting to be
more of a reality. Short-term swap rates will likely rise if
LIBOR rises. We have a flat-to-upward bias on short-
term swap rates. On the other side of the curve, weexpect the flatness of the long-end swap curve to
continue into 2012. The crisis in Europe still has a long
way to go before it approaches a resolution. The crisis
puts a downward pressure on long-term swap rates via
two routes. The supply and demand balance is unlikely to
see a large change. Risk appetite of foreign (European)
investors is unlikely to recover enough for them to come
back to 30Y ASW in the near term. Also, a worsening of
the crisis would lead to a selloff in risky assets and a JPY
appreciation, which would flatten the long end of the
curve through dealers’ hedging activities. In sum, we
have a flattening bias on JPY swap curve as a whole.
With short-end rates held at a low level, we focus on
convexity trades such as weighted flatteners and
weighted butterflies. In addition, if the markets price in a
longer time frame of a low interest-rate policy, the effect
will likely spill over from shorter ends to longer ends, as
investors seek carry and move along the curve. In that
scenario, we expect a cheapening of the belly on shorter
dated butterflies such as 2s/5s/10s and then possibly
5s/10s/20s spreads as a medium-term trend. For example,
the 5s/10s/20s 50:50 fly has a convex profile vs. 5s/20s
curve, therefore a curve neutral 5s/10s/20s fly is expected
to show a convex P&L (Exhibit 20). In the long end,
long-dated steepeners are attractive in carry
considertations..
Swap spreads
Swap spreads tightened over the past year (Exhibit 21).
For one, we think this was related to a rise in fiscal
premium for JGBs. As mentioned earlier, the increase in
JGB issuance associated with the aftermath of the 11
March earthquake has continued to be in focus. On the
other hand, from a swap perspective, the swap curve
Exhibit 20: The 50:50 5s/10s/20s butterfly has historically showedpositive convexity to the 5s/20s swap curveSwap 5s/10s/20s butterfly vs. swap 5s/20s spread since 2010; bp
y = 0.02 x 2 - 3.52 x + 188.77
R2 = 0.46
-30
-25
-20
-15
-10
-5
0
5
100 110 120 130 140 150
5s/20s sw ap curv e; bp
Exhibit 21: Swap spreads have broadly tightened across the curvewith the exception of below 2Y over the past year Change of swap spreads over the past year; bp
-30
-20
-10
0
10
20
30
1 2 3 4 5 6 7 8 910 12 15 20 25 30
Maturity
-30
-20
-10
0
10
20
30
Change (RHS)
Term structure (LHS)
Exhibit 22: A PCA* indicates that swap spreads tended to move inthe same direction, but the move tended to be larger in the longer end. There was also another factor that impacted the 7Y sector and super-long sectors in the opposite directionJanuary 2011– November 2011; first two PCA factors, first factor explains 73%,second factor explains 15% unit
-0.60
-0.40
-0.20
0.00
0.20
0.40
0.60
2Y 5Y 7Y 10Y 20Y 30Y
1st factor
2nd factor
* The PCA is conducted on the daily changes in swap spreads across the curve.The data period is from January – November 2011.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 170/236
Japan Rates ResearchGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Takafumi YamawakiAC
(81-3) 6736-1748
Yuya Yamashita, CFA (81-3) 6736-1493J.P. Morgan Securities Japan Co., Ltd
170
faced a bull-flattening pressure, led by super-long sectors
on the back of a selloff of risky assets and JPY
appreciation.
To get a better understanding of this year’s trend, we
conducted a PCA. Exhibit 22 shows the result of the
PCA analysis of the daily changes in swap spreads over
this year. The first and second PCA factors explain 73%
and 15% of the total variance, respectively. The result
indicates that swap spreads generally moved in the same
direction but the move was larger in the longer end. This
is in line with the above explanation that a rise in fiscal
premium caused this year’s swap spread changes. The
second factor, which is independent of the first one and
smaller in its impact, had an effect on the 7Y sector and
super-long sectors in the opposite direction. We can think of this factor as somehow related to ‘flight-to-quality’
movement with richening of 7Y on JGB futures buying
and cheapening of 30 on swap curve receiving.
In general, JPY swap spreads are largely affected by 1)
repo/LIBOR spread as a determinant of baseline values,
and 2) JGB yield levels. Since December 2010, the
repo/LIBOR spread has stayed within the 21-23bp range
(Exhibit 23), therefore had only a limited impact on
swap spreads. In fact, in a simple multiple regression
analysis over this year, the repo/LIBOR spread does not
have a meaningful explanatory power and JGB yield
levels, in turn, have higher t-values.
Partly mentioned in the Swap Curve section, we have a
flat-to-upward bias on LIBOR. From a monetary policy
perspective, we do not expect the BoJ to hike a rate
ahead of the Fed. Neither do we expect the BoJ to step in
to an IOER cut. However, a cut in IOER is still possible.
In this scenario, it is likely that with LIBOR falling faster
than repo rates, the repo/LIBOR spread will tighten. A
more likely risk scenario is that markets face stress and
LIBOR rises. In that scenario, LIBOR rises while repo
rates are kept low and the repo/LIBOR spread widens.
All in all, we have a widening bias on short-end swap
spreads. For the super-long sectors, we think the
cheapness of JGB vs. swap rates will remain into
2012, given our view that the flatness in the long end of
the JPY swap curve will continue.
From a relative value and shorter-term perspective, swap
spread curve trades are worth considering. One way to
see such relative values is to run simple multiple
regression. For example, Exhibit 24 shows the residuals
of swap spreads from the regression analysis where
swap spreads are regressed on underlying JGB yields and
1Y LIBOR/OIS basis (which we think better reflects
day-to-day moves than the repo/LIBOR spread) as
explanatory variables. It indicates that front end spreads
are too wide while long end spreads are too narrow. We
would use this model to assess the relative
richness/cheapness on the swap curve, however given
our view of swap curve flattening we recommend
investors not to chase this relative value.
Exhibit 23: Since December 2010, repo/LIBOR spreads havestayed within the 21-23bp range, therefore had only a limitedimpact on swap spreads.
Spread between the S/N repo rate and 6M JPY LIBOR since 2010; bp
20
22
24
26
28
30
32
34
36
38
Jan-10 Jul-10 Jan-11 Jul-11
Source: Bloomberg, J.P. Morgan
Exhibit 24: Our JGB swap spreads model on underlying JGByields and 1Y LIBOR/OIS basis* indicates that front end spreadsare too wide while long end spreads are too narrow. However, wewould not fade this relative value.Residuals of swap spreads as of 17 November 2011; bp
-6
-4
-2
0
2
4
6
8
2 3 4 5 6 7 8 910 12 15 20 25 30
Maturity
JGB cheap
JGB rich
* The swap spreads are regressed on the underlying JGB yields and 1YLIBOR/OIS basis. The data period is from January – November 2011.The regression equation for 10Y swap spread is, for example,10Y swap spreads = 8.9 – 19.0 x 10Y JGB Yield - 0.48 x 1Y LIBOR / OIS
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 171/236
Japan Rates ResearchGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Takafumi YamawakiAC
(81-3) 6736-1748
Yuya Yamashita, CFA (81-3) 6736-1493J.P. Morgan Securities Japan Co., Ltd
171
Cross-Currency Basis Swap
USD/JPY cross-currency basis spreads gradually
narrowed in 1H11. However, they widened back (moved
–ve) in 2H11. In addition, we had two sharp widening in
August and November (Exhibit 25).
A medium-to-long-term driver is dollar funding pressure.
Concerns about European banks have heightened. Some
banks that found it difficult to raise dollar funding via
standard funding markets probably headed for cross-
currency basis markets. USD/EUR cross-currency basis
spreads widened, and USD/JPY cross-currency spreads
followed. On a flow side, coupon swap (forward USD
buying) flows from Japanese investors, which had once
counteracted a widening of the spreads in the past, were
substantially reduced as the USD/JPY FX rate continued
to break record lows this year. Meanwhile, periodic
samurai issuance put pressure in a negative direction to
the spreads.
Short-term and sharp widening in August and November
is explained by the FX intervention by the MoF and BoJ,
which created a significant number of short USD
positions in markets. Our FX team estimates the amount
of intervention to be around ¥4.5tn in August and ¥7.4tn
in October. Some market participants covered those
positions in the cross-currency basis market as well as
FX markets. August, in particular, saw some investors
spread tightening positions as a carry trade. The FX
intervention triggered stop-loss activities on such
positions, which accelerated further widening. In fact, the
magnitude of widening was milder over October –
November. Probably a large part of speculative positions
had been already cleared after the intervention in August.
It is possible to estimate the break-even level of short-
term FX basis spreads. If a bank funds JPY cash at
overnight rate and swaps it into USD by FX OIS basis
swap, it effectively funds USD cash at USD OIS + X
(Exhibit 26, left). Moreover, short-term FX LIBOR
basis should move interactively with 3M LIBOR/OIS
basis in each currency and FX OIS basis (Exhibit 26,
right). Even if the overnight call market is not available,
banks can resort to the BoJ’s liquidity providing
operation which lends JPY cash at 0.1% p.a. for 3M or
6M terms in exchange for certain eligible collaterals.
The cost involved in the above method should
theoretically not exceed the cost via other dollar funding
means. Currently, the BoJ’s dollar lending facility is in
place. It lends unlimited USD at approximately USD
overnight rate + 100bps in exchange for certain eligible
collaterals with the additional haircut of about 12.5% for
equal to or less than the 1M term and 20% for more than
1M and equal to or less than the 3M term. Taking the
cost of the haircut into account, the effective cost of USD
funding via the BoJ’s dollar facility is higher than USD
FF + 100bps. Therefore, FX OIS USD/JPY basis spreads
Exhibit 25: USD/JPY FX basis spreads gradually narrowed in 1H11.However, they widened back (moved –ve) in 2H11Cross currency basis spreads since 2011; bp
-100
-90
-80
-70
-60
-50
-40
-30
-20
-10
0
Jan-11 Apr-11 Jul-11 Oct-11
1Y FX LIBOR basis
(bp)
3M FX OIS basis
Source: Bloomberg, J.P. Morgan
Exhibit 26: (1) If a bank funds JPY cash at overnight rate and swapsit into USD by FX OIS basis swap, it effectively funds USD cash atUSD OIS + X. (2) Short-term FX LIBOR basis should moveinteractively with 3M LIBOR/OIS basis in each currency and FX OISbasis. The cost involved in the above method should theoreticallynot exceed the cost via other dollar funding means…How to calculate the break-even level of short-term FX basis spreads
Funding JPY from
overnight marketJPY O/N
Funding JPY from overnight
marketJPY O/N
Fixing via JPY OIS JPY OISConverting it to JPY LIBOR
via JPY LIBOR/OIS basis
- O/N +
3M JPY LIBOR- Y
Swapping JPY cash
into USD by FX OIS
basis
- ( JPY OIS – X)
+ USD OIS
Swapping JPY cash into
USD by FX LIBOR basis
- (3M JPY LIBOR - Z)
+ 3M USD LIBOR
Convert it to USD OIS via
USD LIBOR/OIS basis
- 3M USD LIBOR
+ USD OIS + W
Net USD OIS + X Net USD OIS - Y + Z + w
(1) (2)
Exhibit 27: …but 1Y FX OIS spread at the right is already below –100bp. The implied FX OIS USDJPY spreads exceed those seen inthe actual FX OIS USD/JPY basis market.Breakeven USD funding cost; bp
JPY 3M
LIBOR/OIS
basis (Y)
FX LIBOR
USD/JPY
basis (Z)
USD 3M
LIBOR/FF
basis (W)
USD OIS +
(-Y+Z+W)
JPY OIS +
(Implied
FX OIS)
FX OIS
USDJPY
basis
1Y 16.9 66.0 49.3 109 -109 -110
2Y 17.0 75.5 58.9 120 -119 -120
3Y 16.8 82.3 66.2 124 -123 -125 Source: Bloomberg, J.P.Morgan
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 172/236
Japan Rates ResearchGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Takafumi YamawakiAC
(81-3) 6736-1748
Yuya Yamashita, CFA (81-3) 6736-1493J.P. Morgan Securities Japan Co., Ltd
172
can go under -100. But the divergence should not be
large, assuming that the two OIS rates are equal to the
overnight rate in each currency and constant. In addition,if the spread charged on the dollar lending facility by the
central banks is cut from the current 100bps, FX OIS
USD/JPY basis spreads are expected to move in a
positive direction accordingly,
Exhibit 27 shows the current level of LIBOR/OIS basis
spreads in each currency, FX LIBOR USD/JPY basis
spreads and the FX OIS spreads implied from those
numbers. 1Y FX OIS spread at the right is already below
–100bp. Further, the implied FX OIS USDJPY basis
spreads exceed those seen in the actual FX OIS USDJPY
basis market. We have to be reminded that it starts from
the assumption that a bank funds JPY at the overnightrate. In addition, it depends on many other variables such
as how long the BoJ’s lending facilities (in JPY and
USD) would be available, how much the collateral and
additional haircut actually costs, and bid-offer. Lastly,
there could be the possibility that markets overshoot at
extreme stress, irrespective of these fundamentals.
There are some implications from large negative
USD/JPY cross-currency basis spreads. First, large
negative basis spreads make it attractive for a USD
investor to create synthetic USD bonds by buying JGBs
and swapping them into USD with a cross-currency basis
swap. The level of yield pick-up appears to at ahistorically attractive level (Exhibit 28). This applies
even when compared with the difference of CDSs of the
two countries (Exhibit 29). Although we need to bear in
mind that CDSs of JGBs and USTs are mainly traded in
USD and EUR, respectively, the comparison still gives
us some guidance on the level of the pickups.
Secondly, the return of FX-hedged foreign bonds has
become less attractive (Exhibit 30). Some investors,
when investing in foreign bonds, hedge their FX risk.
They often use rolling 3M FX forward. As a result of
sharp widening in the 3M term FX OIS basis spread, the
cost of 3M FX hedge has become higher. If short-term
basis spreads are stabilised at a large negative level,
Japanese investors may accelerate the repatriation of
USD funds back to JGBs. The focus is on life insurance
companies because these investors would increase
investment into super-long sectors with extra JPY cash.
The return of FX-hedged 10-year USTs is now even
lower than in August, when we saw ¥550bn net selling of
medium- and long-term foreign bonds from life
insurance companies, according to MoF data.
Exhibit 28: The level of yield pick-up appears to be at a historicallyattractive level on 1YPick-up in yield of USD-asset-swapped 1Y JGB over 1Y USTs since 2010; bp
0
20
40
60
80
100
120
Jan-10 Jul-10 Jan-11 Jul-11
Source: Bloomberg, J.P. Morgan
Exhibit 29: The level of yield pick-up appears to be attractive evenwhen compared with the difference of CDSs of the two countriesPick-up in yield of USD-asset-swapped JGBs over USTs; bp
JGB JGB ASWJPY 3s/6s
Basis
FX LIBOR
USD/JPY
basis
3M USD
LIBOR +
Pickup vs
UST
JGB/UST
CDS spread
1Y -25.5 13.5 -60.0 48 107 8
2Y -23.9 14.0 -70.0 61 110 23
3Y -20.6 14.1 -78.0 73 119 31
* For Exhibits 27&29, all the numbers are based on mid levels. For relevant basisspreads in each currency, cross-currency basis spreads, ASWs and CDSs, weused the nearest benchmark points. We calculated these numbers, with theassumption thatUSD/JPY conversion factors are the ratio of modified durations for simplification.
Exhibit 30: The return of FX-hedged foreign bonds has become lessattractive with higher FX hedge cost, which could lead to further demand to the super-long sectors from life insurance companies.20Y JGB yields and 10Y UST yields with FX hedge since 2010; %
0.50
1.00
1.50
2.00
2.50
3.00
3.50
Jan-11 Apr-11 Jul-11 Oct-11
10Y UST hedged by 3M roll
20Y JGB
FX hedge cost is calculated as 3-month forward point divided by spot FX rateSource: Bloomberg, J.P. Morgan
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 173/236
Japan Rates ResearchGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Takafumi YamawakiAC
(81-3) 6736-1748
Yuya Yamashita, CFA (81-3) 6736-1493J.P. Morgan Securities Japan Co., Ltd
173
Volatility
The implied volatility surface as a whole was shifted
downward with a steepening bias over the past year
(Exhibit 31). Implied volatilities soared last December,
as QE2 pushed up 10Y UST yields to above 3.50%.
JGBs were sharply sold off and 10Y JGB yields moved
higher from 0.85% in October to 1.3% in December. The
selloff also triggered unwinding of short volatility
positions, and volatilities spiked. Implied volatilities rose
sharply soon after the 11 March earthquake. The BoJ
provided abundant liquidity to ease the post-earthquake
turmoil and maintained liquidity at a high level. Interest
rates gradually declined thereafter. The 10Y swap rate
and 10Yx10Y forward swap rate moved lower to 0.96%
and 2.43%, respectively, from 1.21% and 2.67% at the beginning of this year. Steadily declining yields also
lowered realised volatilities and then implied volatilities
(Exhibit 32).
From an investor flow perspective, first, the fall in the
surface was led by constant selling flows in gamma
sectors such as target buying and covered call from
domestic investors. Second, issuance volume of
structured notes remained small, but domestic banks kept
selling callable structured deposit to mass retail. Banks
generally sell options to hedge long vega positions. Their
selling tends to take place in the middle of the surface as
structured deposits often have a call option after 1–3
years and a maturity of 5–10 years. Such volatility
selling flows are another factor for cheapening of the
vega sector. On the other hand, there were sporadic
buying flows in longer tail vega sectors from
macro/leveraged investors. They bet on a long-term
fiscal deficit deterioration scenario of the Japanese
government. These flows pushed up and supported
longer ends of the surface.
Into 2012, we think that the current shape of the surface
is likely to remain. We are neutral on gamma. Under a
low-for-long regime, short-term interest rates will likely
remain low and move within a narrow range and
volatility selling flows are expected to continue. On theother hand, we cannot discount the possibility of a rise in
gamma sector volatilities under a stressed scenario, given
the situation in Europe. In long-tail vega sectors, we
expect volatilities to remain relatively high. With
absolutely low interest rates and volatilities in the JPY
market, we do not expect selling demand to outweigh
buying demand in these sectors. Rather, since an
issuance increase of JGBs continues to be a major topic
of debate, concerns about the long-term fiscal deficit of
Japan are likely to be priced in and out from time to time.
Exhibit 31: The implied volatility surface as a whole was shifteddownward with a steepening bias over the past year.Implied volatility changes over the past year; bp/day
-1.8
-1.6
-1.4
-1.2
-1.0
-0.8
-0.6
-0.4
-0.2
0.0
3M 1Y 2Y 5Y 10Y
1Y tail 2Y tail 5Y tail 10Y tail 20Y tail
Exhibit 32: Implied volatilities soared last December as QE2pushed up 10Y UST yields to above 3.50% They also rose sharplysoon after the earthquake in March before declining later in theyear.1Yx10Y normal implied volatility since 2010; bp/day
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Jan-10 Jul-10 Jan-11 Jul-11
Earthquake
UST Yield Rise
after QE2
Exhibit 33: TIBOR remained sticky to fall relative to CP rateTIBOR 3M and a-1 rated CP rate; %
0.00
0.10
0.20
0.30
0.40
0.50
0.60
Nov -09 May -10 Nov -10 May -11 Nov -11
CP rate a-1
3M D Tibor
Source: Bloomberg, JASDEC, J.P. Morgan
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 174/236
Japan Rates ResearchGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Takafumi YamawakiAC
(81-3) 6736-1748
Yuya Yamashita, CFA (81-3) 6736-1493J.P. Morgan Securities Japan Co., Ltd
174
TIBOR/LIBOR Spread
There are three categories of TIBOR-LIBOR spreads: 1)
spot fixing TIBOR-LIBOR spread; 2) IMM 3M TIBOR-
LIBOR spreads whose liquidity runs up to the 8th
contract; and 3) Term TIBOR/LIBOR spreads which are
the difference between swap rates usually against TIBOR
6M and LIBOR 6M, but for short maturities also against
TIBOR 1M and LIBOR 1M.
TIBOR fixing remained sticky to fall while CP rates
were stable at a low level (Exhibit 33). We think the
reason for the stickiness has remained unchanged over
the past few years, i.e. market practice is that TIBOR is
used as a reference rate to corporate lending. As the
credit spread charged to domestic companies continuedto be tight (Exhibit 34), TIBOR tends to be maintained
at a high level relative to the policy rate. As a result, the
3M TIBOR/LIBOR fixing spread remains positive.
There were two periods in the past when TIBOR/LIBOR
fixing spread was negative (Exhibit 35). One was just
after the end of the BoJ’s quantitative easing in 2006
when LIBOR priced in the future rate hike earlier, and
the other was after the subprime shock in summer 2007
as LIBOR jumped. We think that the former case is not
impossible, but very unlikely. Since the Fed introduced
an explicit time frame of interest-rate policy until mid-
2013 in August, the BoJ is expected to continue thecurrent interest-rate policy even longer than the Fed. The
latter case is more likely given the current situation in
Europe. But even if it happens, the magnitude may be
milder as central banks are committed to avoiding a
liquidity crisis.
With regards to term TIBOR-LIBOR, we think that
negative spreads are unlikely unless the spot fixing
TIBOR-LIBOR spread becomes negative. Short-term
spreads are generally driven by the spot fixing spread.
For longer tenors, the spreads are largely driven by
domestic banks’ hedging activities. They use TIBOR as
their internal pricing and TIBOR swaps fit the purpose.However, as LIBOR swaps are more liquid, they often
use LIBOR swap and hedge the TIBOR/LIBOR risk
from time to time. Their choice of hedge timing is the
source of term TIBOR-LIBOR fluctuation. However,
term TIBOR-LIBOR spreads generally moved in a very
tight range with limited flows and a sticky spot-fixing
spread this year. (Exhibit 36)
Looking ahead, we have a tightening bias on the
TIBOR-LIBOR fixing spread. Historically, LIBOR
Exhibit 34: As the credit spread charged to domestic companiescontinued to be tight, TIBOR tends to be maintained at a high levelrelative to the policy rate.
Average rate charged to corporate in bank lending; %
1.00
1.20
1.40
1.60
1.80
2.00
2004 2005 2006 2007 2008 2009 2010 2011
Source: BoJ, J.P. Morgan
Exhibit 35: LIBOR fixing was higher than TIBOR fixing at the endof QE in 2006 and after the subprime shock in 20073M TIBOR and LIBOR fixing; %
0.00
0.20
0.40
0.60
0.80
1.00
1.20
2005 2006 2007 2008 2009 2010 2011
3M LIBOR
3M Euroyen T IBOR
Source: Bloomberg
Exhibit 36: Term TIBOR-LIBOR spreads generally moved in a verytight range with limited flows and sticky spot-fixing spread thisyear 5Y term TIBOR/LIBOR spread and spot-fixing spread; bp
-4
-2
0
2
4
6
8
10
12
Jan-10 Jul-10 Jan-11 Jul-11
5Y Term T/L
TIBOR-LIBOR fix ing spread
Source: Bloomberg
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 175/236
Japan Rates ResearchGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Takafumi YamawakiAC
(81-3) 6736-1748
Yuya Yamashita, CFA (81-3) 6736-1493J.P. Morgan Securities Japan Co., Ltd
175
tended to move faster and be more volatile both in a rise
and in a fall. Therefore, the TIBOR-LIBOR fixing spread
has had a tendency to narrow in a rise and to widen in afall. As mentioned in the previous sections, there are
some risk scenarios. One is a fall of LIBOR on a cut in
IOER. Another scenario is that LIBOR rises under
extreme stress, which we think is more likely. Besides,
the level of LIBOR is absolutely low and lower than
TIBOR. The magnitude of the spread move is likely to be
asymmetric, i.e. we expect more tightening if LIBOR
increases than widening if LIBOR falls.
3s/6s spread
The Term LIBOR 3s/6s spread curve steepened this year
(Exhibit 37). The spreads moved in a parallel fashionuntil 2010. However, although the spreads widened, the
widening was larger at the longer ends. Widening can be
explained by one-sided demand in the 3s/6s market. With
relatively tight credit spreads in Japan, samurai issuance
continued in 2011. When a non-resident entity issues
JPY bonds (e.g. EMTN or samurai) and swaps into USD,
it enters a paying position of 3s/6s basis swap (pay 3M
vs. rec 6M) as well as a receiving position in a cross-
currency swap (Exhibit 38), driving 3s/6s basis higher
and FX basis wider (more –ve). Such flows tended to
take place in 5–10Y sectors. Conversely, if a non-
resident investor buys a JPY bond with a cross-currency
swap as a synthetic USD bond, it has the opposite impacton the spreads. However, there was little such demand
this year. Rather, those investors who had such positions,
particularly in the super-long sectors, might have
unwound them for the purpose of reducing balance
sheets. This could be a reason for the widening on
longer term 3s/6s spreads.
We think the short-to medium-term spreads are
biased wider. Tight credit spreads in Japan are still
attractive for non-resident entities to raise funding, even
after large negative cross-currency basis spreads and
extra premium for non-resident entities are considered.
Issuance by non-residents is expected to continue in 2012and such flows would be a widening factor in the
medium-term spreads. There is also the possibility that
LIBOR will rise if the crisis in Europe deepens. If the
scenario realises, the spot fixing 3s/6s spread will widen
as it had amid the Lehman shock and lead to a widening
of term 3s/6s spreads at the short ends.
Exhibit 37: Term LIBOR 3s/6s spread curve widened and steepenedthis year Spot fixing 3s/6s and 5Y and 20Y term 3s/6s basis; bp
10
15
20
25
30
Jan-10 Jul-10 Jan-11 Jul-11
Spot fixing 3s/6s LIBOR
5Y term 3s/6s
20Y term 3s/6s
Source: Bloomberg, J.P. Morgan
Exhibit 38: Flows of non-resident entity issuing JPY bonds andswapping into USD drive 3s/6s basis higher and FX basis wider (more –ve)How 3s/6s is involved in swapping JPY into USD
Issuer Samurai/EMTN
JPY 3M L + 3s/6s basisJPY 6M
USD 3M L
JPY 3M L +FX LIBOR basis
JPY Fixed
JPY FixedJPY 6M
3s / 6s
marktet
Swap marktet
FX basis
marktet
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 176/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Jorge GarayoAC
(44-20) 7325-4820
[email protected]. Morgan Securities Ltd
176
Inflation-linked Markets
• 2011: Linkers underperformed as nominal yields
declined substantially on the back of an
escalation of the sovereign risk crisis in the Euro
area and a deteriorating global macro outlook
• 2012 outlook for inflation: Substantial falls in
headline and core inflation across the board
• We believe that the market will focus more on
downside risks from a deeper recession than on
medium-term inflation, as inflation expectations
remain anchored
• German and UK 10Y real yields to test new lowsand fall well below zero in Europe but expected
to stay close to 0% in 10Y TIPS…
• …while breakevens should be biased lower as the
sovereign crisis continues in 1H12
• Expect cash breakevens to underperform relative
to inflation swaps in early 2012
• Euro area: Position for flatter breakeven curve,
as inflation should be sticky and longer-dated
breakevens remain highly directional
• UK strategy: Be long intermediate real yields,
and short inflation breakevens in the 10Y sector
• UK tactic: Go short UK real yields into linker
syndications as a tactical trading strategy
• US: Fair-value model suggests falls in inflation
and nominal yields, combined with fiscal
tightening, should cause breakevens to narrow
into early 2012
• Beyond 1Q12 longer maturity TIPS breakevens
should widen as nominal rates rise
• Supply: UK and US to see modest increases in
gross supply, whereas Euro area issuance should
be below average, with Italy being the mainquestion mark
2011: Linkers outpaced by nominals in a
relatively high inflation environment
2011 was dominated by large declines in nominal yields,
as the European sovereign debt crisis accelerated from
the summer. Throughout the year, market-implied
inflation breakevens were initially driven by a spurt of
Exhibit 1: Inflation breakevens were supported by energy price inflationin 1Q11 but started to decline in the summer Inflation breakevens for 10Y benchmarks in US TIPS, Germany and the UK; %
1.00
1.50
2.00
2.50
3.00
3.50
Nov 10 Feb11 May 11 Aug11 Nov 11
UK Germany US
Exhibit 2: Headline inflation rose steadily over the past year, withinflation in 1Q11 being lifted by energy prices…Inflation measures, in oya terms; %
-4.00
-2.00
0.00
2.00
4.00
6.00
Nov -06 Nov -07 Nov -08 Nov -09 Nov -10 Nov -11
Euro HICP US CPI UK RPI
Exhibit 3: …and core inflation also moving up significantly over theperiodCore inflation measures, in oya terms; %
0.00
1.00
2.00
3.00
4.00
Nov 06 Nov 07 Nov 08 Nov 09 Nov 10 Nov 11
HICP core US CPI core UK CPI core
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 177/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Jorge GarayoAC
(44-20) 7325-4820
[email protected]. Morgan Securities Ltd
177
energy price inflation in 1Q11, but they entered a clear
downward phase from May onwards (Exhibit 1). From
this point, inflation breakevens were driven lower by a prevailing risk-off environment and worries of a
worldwide recession as data printed particularly weak
globally.
The evolution of inflation prints would suggest a
positive return year for inflation-linked markets.
Headline inflation rose steadily since November 2010,
peaking at the end of the year (Exhibit 2). Even though
energy inflation played a significant part in 1Q11, core
inflation also rose throughout the year (Exhibit 3). This
was contrary to our January predictions that a high
degree of slack would push core inflation lower (with the
caveat that the move in UK core CPI was heavilyimpacted by VAT effects).
However, inflation prints were not the main driver
and inflation expectations were pushed lower by the
macro landscape. The downgrading of growth
expectations in the Euro area and globally from the
summer was likely exacerbated by sovereign risk
dynamics in the periphery. Commodities reversed their
upward trend, and equity markets fell over 15% in the
space of two weeks in August. Central banks recognised
a shift in the macro fundamentals and risks around them,
dismissing the contemporaneous ‘relatively high’
inflation prints. The ECB eased rates in November after having tightened policy in April and July, whereas the
Fed committed to keeping rates unchanged up to mid-
2013 and initiated ‘Operation Twist’. The BoE
announced another £75bn of QE purchases and is
expected to announce another round of QE in the coming
months.
Survey-based measures of medium-term inflation
expectations remained well anchored throughout the
year despite the significant deterioration in the macro
outlook . Inflation expectations from 2013 onwards were
broadly unchanged according to economists’ consensus
forecasts (Exhibit 4), supporting the message of
anchored inflation expectations that central banks have
been highlighting thus far.
So how did inflation-linked portfolios fare relative to
conventional government bonds? Their returns
(adjusted for duration differences) were negative in the
Euro area and the UK and just about matched those
of Treasuries in the US (Exhibit 5). In other words, a
strategy of being long inflation breakevens across the
curve would have given investors a return of -3.3% and
Exhibit 4: Medium-term inflation expectations are well anchoredaccording to economists’ surveys, with the Euro area and the US likelyclose to target and UK CPI expected to be 0.8% above target
Long term inflation forecasts from economists*; %
Oct10 Oct11 C hg Oct10 Oct11 C hg Oct10 Oct11 Chg
2011 1.9 3.1 1.2 2.6 4.4 1.8 1.6 2.6 1.0
2012 2.2 2.1 -0.1 2.1 2.7 0.6 1.5 1.8 0.3
2013 2.1 2.1 0.0 2.6 2.5 -0.1 1.7 1.8 0.1
2014 2.2 2.3 0.1 3.0 2.7 -0.3 1.9 1.9 0.0
2015 2.2 2.3 0.1 2.6 2.6 0.0 2.0 2.0 0.0
2016 2.3 2.2 -0.1 2.7 2.8 0.1 2.1 2.0 -0.1
2017-21 2.3 2.2 -0.1 2.7 2.8 0.1 2.1 2.1 0.0
US CPI UK CPI Euro HICP
* Source: Consensus Economics, October 2011 long-term forecasts.
Exhibit 5: Despite the upward trend in inflation in 2011, buyingEuropean linkers vs. conventional bonds would have beenunprofitable, with Italian linkers being the main drag for the Euro area;the same strategy in the US would have posted returns close to zeroTotal return between 15 November 2010 and 18 November 2011; %
Total return
Euro linkers (JPM ELSI) -10.7%
Euro gov comparator* portfolio -5.8%
Long inflation B/E (JPM EBEX) -3.3%
Euro linkers ex Italy ex Greece -2.7%
Italian linkers -23.5%
US TIPS (JPM JUSTINE) 12.5%
USTs comparator* portfolio 10.4%
Long inflation B/E (JPM UBEX) 0.0%UK Linkers (FTSE) 18.7%
Gilts 7Y+ (JPM GBI) 20.4%
Long inflation B/E (est.**) -5.7%
* Comparator portfolios are based on the same weights used by linker indices for comparable nominal bonds, adjusting for the relative duration.** Calculated by replicating a strategy of being long FTSE All-Share index linkedagainst being short the JPM Gilts 7Y+ index in a duration-adjusted amount, re-weighted weekly.
Exhibit 6: The deterioration in liquidity was an important driver of performance. Our estimated bid-offers widened to levels above thoseseen in 4Q08 for Italian linkers, and to similar levels in French linkerswhereas other linker markets and inflation swaps bid-offers held inremarkably wellIndicative bid/offers across different inflation products*; bp of yield
Assumptions
Prior 4Q 08 1H 11 2H 11
Italy linkers 2 15 2.5 20 €50mn size
France linkers 2 15 2.5 15 €25-50mn size
German linkers 2 15 2.5 10 €25-50mn size
Euro HICP sw aps 2 15 2 4 €50mn size
UK linkers 2 7 2 2 £10mn size
UK RPI sw aps 4 18 4 4 £25mn size
US TIPS 2.5 7.5 3 4 $50mn size
US CPI sw aps 5 15 5 5 $25-50mn size
2008 2011
* Source: J.P. Morgan
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 178/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Jorge GarayoAC
(44-20) 7325-4820
[email protected]. Morgan Securities Ltd
178
-5.7% in the Euro area and the UK, respectively, and a
flat return in the US.
As in 2008, the deterioration in linkers liquidity was
an important driver of performance, particularly in
the Euro area. Bid/offers reached similar or larger
dislocations than those seen in 2008. In particular
Italian BTPeis bid/offers widened on average 20bp vs.
15bp at their worst point in 2008. But even bid-offers for
French linkers widened to levels reached in late 2008
(Exhibit 6). Supply in Euro area linkers was heavily
affected by the sovereign debt dynamics. Germany
cancelled its linker programme in 2Q11 and 3Q11, not
tapping the market between April and November. Italy’s
net issuance of linkers in 2H11 was just around €1.5bn
(vs. usually €2–4bn per quarter). Unstable marketconditions resulted in reduced auctions, and the Treasury
conducted buybacks of heavily dislocated lines. It is
interesting to note that US and UK linkers, as well as
inflation swaps, held up much better in terms of bid-
offers, despite some evident deterioration in liquidity
dynamics.
Inflation outlook for 2012
The inflation outlook for 2012 can be characterised
by an expected decline in both core and headline
inflation (Exhibit 7). In terms of headline inflation, we
expect more significant declines, in great part because of energy and food base effects from substantial price
increases in the past year. In sequential terms (seasonally
adjusted q/q changes in CPI indices), we have already
seen a clear declining trend in Euro area headline
inflation, and we expect significant falls in the UK and
the US in the coming quarters (Exhibit 8). Euro area
inflation should drop to an average of 1.9% oya during
2012, in line with the ECB’s target. In the UK, we expect
headline inflation to fall sharply but average 2.5% (above
the BoE’s target of 2%) whereas we see US inflation
averaging 1.4%.
The assumptions behind our forecasts are as follows:
• Stable oil prices, assuming Brent around
$110/bbl, most likely stuck in a $100–115/bbl
range. We see OPEC supporting prices above
$100 (barring a very significant recession,
which could push it below that figure), while
marginally supportive fundamentals could push
it to $115/bbl.
Exhibit 7: J.P. Morgan 2012 inflation forecasts: core measures todecline across the board; headline inflation to fall but average ‘abovetarget’ in the UK, ‘around target’ in the Euro area, and ‘well below
target’ in the USUS= US CPI-U nsa and core CPI, Euro = Euro HICP all items, core HICP (excludingall food and energy), UK= UK RPI, CPI and core CPI; oya%
CPI Core HICP Core RPI CPI Core
Oct 11 3.5 2.1 3.0 1.6 5.4 5.0 3.4
Nov 11 3.5 2.1 3.0 1.6 5.2 4.8 3.2
Dec 11 3.1 2.2 2.8 1.7 4.9 4.2 3.0
Jan 12 2.7 2.1 2.5 1.7 4.4 3.6 2.6
Feb 12 2.4 1.9 2.4 1.7 3.7 3.2 2.1
Mar 12 1.6 1.9 2.0 1.6 3.5 3.2 2.1
Apr 12 1.2 1.7 1.8 1.5 3.0 2.5 1.5
May 12 1.0 1.5 1.9 1.6 2.9 2.5 1.6Jun 12 1.3 1.3 2.0 1.5 3.1 2.7 2.0
Jul 12 1.2 1.1 1.9 1.6 3.2 2.5 1.9
Aug 12 1.0 1.0 1.9 1.5 2.9 2.3 1.6
Sep 12 0.9 1.1 1.7 1.4 2.6 1.9 1.6
Oct 12 1.1 1.1 1.5 1.4 2.7 2.0 1.6
Nov 12 1.2 1.1 1.5 1.5 2.6 1.9 1.5
Dec 12 1.3 1.2 1.5 1.5 2.6 1.9 1.5
Av e 2012 1.4 1.4 1.9 1.5 3.1 2.5 1.8
US Euro UK
Exhibit 8: Euro area sequential inflation has already initiated adownward trend and should stabilise around 1%; we expect significantfalls in the US and the UK in the next few quartersEuro area HICP inflation Q/Q saar, US CPI inflation Q/Q saar and UK CPI Q/Q saar,actual and J.P. Morgan forecasts; %
-2.00
-1.000.00
1.00
2.00
3.00
4.00
5.00
6.00
7.00
Mar 09 Sep 09 Mar 10 Sep 10 Mar 11 Sep 11 Mar 12 Sep 12
Forecasts
Euro area
UK
US
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 179/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Jorge GarayoAC
(44-20) 7325-4820
[email protected]. Morgan Securities Ltd
179
• Stable currencies in the major crosses, in line
with their recent behaviour this year
• 50% chance of a technical recession in the UK
(base case is one-quarter of GDP contraction,
annualised 0.5% growth in 2012), a moderate
recession in the Euro area (peak-to-trough
contraction in GDP just over 1%) and the US
growing at sub-par pace of 1.8% in 2012.
One of the main forces that poses significant
downside risks to inflation is the prospect of a more
severe and prolonged recession in the Euro area,
which could significantly impact the region and the rest
of the world. A global spillover could potentially bring
oil prices well below the $100/bbl level and pushheadline inflation two to three tenths lower, purely
through energy components, as well as impact long-term
inflation expectations. This is not our central scenario,
however, and there are also upside risks if core
inflation remains sticky on the way down if we
misjudge the slack in the economy. As always, the
behaviour of currencies could also have a significant
impact on imported inflation. In the Euro area, ongoing
fiscal retrenchment in key countries may prove
inflationary over the near term, as governments are
likely to increase taxes for the consumer (in the form of
VAT, administered prices, and other duties). Our forecast
does not incorporate any of these factors, and thereforewe should be conscious of them.
We show a summary of some of these factors’
sensitivities in Exhibit 9. We also discuss some of these
risks in some detail in the individual market sections.
Are we approaching deflation? Not quite
The quick deterioration in the macro data and a gloomy
outlook for coming years has raised questions about the
possibility of entering a very low inflation world. In our
view, a zero-inflation world would come with a
significant decline in medium term inflation
expectations. We are far from that, as inflation
expectations have remained firmly anchored in the
three main markets. The prospect of a global recession is
pushing expectations lower in the Euro area, but not to a
worrying degree.
Exhibit 9: Our inflation view is sensitive to assumptions aboutvolatile elements in baskets and model risk for core inflation; we seedownside risks coming from a deeper recession (lower oil) and
upside risks from tax increases (VAT) or currency depreciationImpact on headline inflation; % points
10%
move in
oil price^
10%
move in
food
prices^
10% drop
in
currency*^
Core
models
error
State
controlled
prices / VAT
increase**
Euro HICP 0.10 0.15 0.25 +/- 0.7 0.25
UK RPI 0.20 0.10 0.65 +/- 0.7
US CPI 0.50 0.15 0.25 +/- 0.5
J.P.
Morgan
view
$100-120
range -Downside
in a
deeper
recession
Near termdownside
to food
prices
Central
view of
stable
currencies -
EUR at risk
Risk of
deeper recession
/
Misjudge
slack
Most likely
increases in
Euro area
(-) (-) (+) (-) / (+) (+)
* Based on drop in Trade Weighted currency.** Euro area based on combined 1% moves in Italy, Spain and France, assuming a65% pass through on 65% of baskets, for more granular information see Exhibit 26.^ For simplicity we assume broadly equal betas on the downside and upside.(-) Downward impact on inflation (+) Upward impact on inflation.
Exhibit 10: Investors expect central banks to do a good job atavoiding deflation; the Fed and the BoE are seen as giving significant
weight to growth, while the ECB is more inflation-oriented butremains the most credible in terms of keeping inflation close totargetInvestor perception of weights (median response) given to growth and inflation for each central bank*; %
80
35
70
20
65
30
0%
20%
40%
60%
80%
100%
US Euro area UK
Growth / employment Inflation
* See: “J.P. Morgan inflation expectations survey, November 2011” by Garayo &Chordia
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 180/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Jorge GarayoAC
(44-20) 7325-4820
[email protected]. Morgan Securities Ltd
180
With central banks watching these carefully, there is no
reason to believe in a changing inflation environment,
which would only be consistent with the world economy
sinking into a prolonged depression (not our current
expectation). As we showed in our J.P. Morgan
November Inflation Expectations Survey (11 November
2011), an overwhelming 45% of investors expect
inflation to be close to target in the Euro area over the
medium term. In the case of the BoE and the Fed, they
have both been extremely vocal about the risks of low
inflation and continue to be seen as giving a lot of weight
to growth at the current cross roads (Exhibit 10).
Inflation expectations 1 year out remain remarkably high
according to our latest Inflation Expectations survey
(November 2011), even if there has been a shift lower inthe Euro area (Exhibit 11). Our own expectation is that
Euro area inflation will average 1.9% in 2012, which is
likely to keep expectations relatively high. Furthermore,
the risks of low inflation over the medium term seem to
be pretty much counterbalanced by perceived risks of
very high inflation. This is most clear in the US and the
UK, where investors attach a particularly high
probability of inflation being above 4% in 2 out of the
next 5 years (Exhibit 12). Admittedly, investors’
perceived probability of a deflation in the Euro area has
increased recently (to 15%), but this is counterbalanced
by an equal probability of very high inflation. Looking at
inflation volatility markets, the implied probability of
deflation in 2 or 5 years’ time is not significantlydifferent from that of inflation being above 4% in our
survey (Exhibit 13).
Exhibit 11: Inflation expectations 1 year out do not point toimmediate disinflation; investors were likely much more worriedabout low inflation in July 2010 than in our latest survey
Percentage responses to the question: “Where do you expect core inflation to be ayear from today?” from the J.P. Morgan inflation expectations client survey,November 2011 and previous surveys (from July 2010); % applies to responsesfrom investors across different asset classes
US Euro area UK
Nov 2011 survey
Below 0.5% 4% 6% 4%
0.5-1.5% 24% 42% 13%
1.5-2.5% 58% 38% 37%
2.5%+ 15% 14% 47%
July 2010 survey
Below 0.5% 20% 22% 12%
0.5-1.5% 53% 55% 30%
1.5-2.5% 24% 20% 39%
2.5%+ 4% 4% 18%
Mean (Nov 11) 1.86 1.61 2.34
Chg v s. prev -0.01 -0.16 -0.06
Mean (Jul 11) 1.87 1.77 2.40
Mean (Mar 11) 1.80 1.80 2.40
Mean (Nov 10) 1.40 1.35 1.90
Mean (Jul 10) 1.10 1.00 1.60
48%
84%
77%
57%
Exhibit 12: The Euro area shows the highest probability of deflationamong investors, but it is still counterbalanced by an equalprobability of high inflation; probabilities remain heavily skewed to
the upside in the UK and the US% probability; J.P. Morgan inflation expectations client survey* November 2011
0
5
10
15
20
25
30
35
US Euro UK US Euro UK
Prob of deflation Prob of high inflation
Jul 11
surveyNov 11
survey
* Probability perception of deflation/high inflation as per the question below:“What is, in your view, the likelihood of getting deflation (CPI annual inflation below0%) or high inflation (CPI annual inflation above 4%) for two years out of the next five years in each of these regions?”.
Exhibit 13: The implied probability of deflation in 2 or 5 years’ time isnot significantly different from that of inflation being above 4%; theUK prices in high probability of high inflation in 5 years’ time; over the near term US is pricing a relatively higher probability of deflation% probability of inflation <0% or above 4% in 2 and 5 years’ time*
2Y 5Y 2Y 5Y 0% floor 4% cap
Euro HICP 8.5% 12.8% 5.2% 11.8% 35 44UK** 10.7% 10.9% 8.9% 35.9% 52 67
US 14.3% 14.6% 12.4% 23.5% 39 74
Prob Inf>4% p.a. 5Y ZC (bp of notional)Prob Inf<0% p.a.
* Probability is derived from pricing caps and floors spreads using the J.P. Morganvol surface calibrated with current zero coupon style caps and floors as well as y/ ycaps and floors.** For UK RPI, we report the probability of inflation > 4.8% and <0.8%, roughlyequivalent to UK CPI > 4% or <0% p.a. if we applied the historical spread betweenRPI and CPI of 0.8%.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 181/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Jorge GarayoAC
(44-20) 7325-4820
[email protected]. Morgan Securities Ltd
181
Euro area market dynamics
Sovereign risk to dominate inflation
breakevens in 2012
Euro area breakevens have been intrinsically linked to
the sovereign risk dynamics during 2011, with a very
clear correlation between the risk-off/risk-on dynamics
and linker breakeven performance (Exhibit 14). This
dynamic is unlikely to change in 2012 and the
performance of peripheral spreads is likely to continue to
dominate while inflation fundamentals will take the back
seat.
Italian linkers to continue to feel the pressure
The main question is whether the underperformance of
BTPei linkers vs. BTPs will continue, and whether it will
spread to French linkers vs. OATs as we have seen
recently. The large declines in Italian BTPei breakevens
relative to inflation swaps have been significantly larger
than what we saw in the 2008 financial crisis, while
breakevens for French OATeis have already moved by a
similar degree vs. 2008 (Exhibit 15). This has led to
negative breakevens in Italian linkers out to 2019 and a
complete decoupling from inflation expectations.
Our central view is that Italian breakevens willcontinue to underperform as spreads widen further.
There are risks that Italy may be downgraded and if this
is severe enough it could impact inflation indices. Based
on our understanding of current index rules of one of the
most widely followed benchmarks, a downgrade to BBB
will result in investors reallocating BTPei holdings
(Exhibit 16). These numbers are based on our estimate
that 20–30% of investors are benchmarked, holding
around €20bn of the €77bn of current market value. With
such a large percentage of the market potentially being
dropped from indices, we believe other non-
benchmarked investors may also be reluctant to hold
onto their BTPei investments, resulting in
underperformance ahead of the event and exacerbating
the impact.
We note the following could well limit any potential
impact to Italian linkers:
1) The index rules could be changed if a majority
of investors consider them to be very harmful
for the market.
Exhibit 14: Inflation breakevens have been extremely correlated withequity markets over the past year, rising in risk-on and falling in risk-off environments
Euro Stoxx 50 and DBRi 2020 inflation breakevens vs. risk-on and risk-off *;Index points %
1800
2000
2200
2400
2600
2800
3000
3200
Nov10 Jan11 Mar11 May11 Jul11 Sep11 Nov11
1.20
1.40
1.60
1.80
2.00
2.20
2.40
10Y B/E
Euro STOXX 50
o n
o f f
o f f
o n
o f f
o n
* Risk-on and Risk-off periods defined with reference to sustained trends in EuroSTOXX 50 index – See European Derivatives for definition.
Exhibit 15: BTPei linker breakevens have underperformed vs. HICPswaps by more than twice as much as seen in late 2008; Frenchbreakevens have already moved by a similar degree as in 2008BTPei 2019 breakeven minus HICP swaps, OATei 2020 breakevens minus HICPswaps; bp
-250
-200
-150
-100
-50
0
2008 2009 2010 2011
BTPei 19 b/e minus
HICP s w aps
OATei 20 b/e minus HICP swaps
Financial
crisis late
2008
Exhibit 16: In the event that Italian linkers leave linker indices, wewould see a reallocation out of BTPeis by benchmarked investors;
France should be the main beneficiary in terms of inflows, given thatit constitutes 54% of Euro benchmarksRelative market value of constituents of Euro linker indices and expected activityfrom benchmarked investors; €bn
MV % split MV ex Ita Chg 20% 30%
Italy 77.1 28% 0% -28% -15.4 -23.1
France (€) 74.4 27% 37% 10% 5.7 8.5
France (F) 75.3 27% 37% 10% 5.7 8.6
Germany 52.5 19% 26% 7% 4.0 6.0
Total 279 100.0%
Euro l inker index* Index ex Ita ly Indexed activ ity**
* Based on Euro area government linkers with maturity over 1Y excluding Greece** Expected selling/buying under two assumptions for percentage held bybenchmarked investors in €bn.
(€) Linked to Euro HICP ex tobacco (F) Linked to French CPI ex tobacco.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 182/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Jorge GarayoAC
(44-20) 7325-4820
[email protected]. Morgan Securities Ltd
182
2) Investors could switch to a new linker
benchmark that includes Italy regardless of the
rating. This is more of an administrative process, and would only be likely adopted by
Italian domestic investors (which we estimate
own around 50% of the outstandings).
Taking all risks into account, we see a risk that Italian
real yields go significantly higher. If nominal yield
spreads were to widen by 250bp (broadly consistent with
our projection of 7.75% for 10Y Italy-Germany spread,
see Euro cash), Italian breakevens could trade very
negative out to the 10Y sector, with real yields between
9.50% and 11% in the sub-10Y sector possible. Our
model uses recent betas of breakevens to peripheral
spreads, which are then adjusted according to dirty pricedynamics of linkers. Linkers whose dirty prices reached
levels of 60–70 have seen buying flows, resulting in a
better performance in breakeven terms (as has been the
case in BTPei 2041 and BTPei 2035).Exhibit 17 shows
the moves in inflation breakevens during the first two
weeks of November vs. the dirty prices of linkers, we
think this behaviour could continue.
The beta of Italian breakevens to peripheral spreads has
been approximately 40% out to the 15Y sector over the
past 6 months (i.e 4bp of breakeven falls for every 10bp
of widening of spreads, Exhibit 18) and around 25% for
the very long end. Two factors drove the BTPei real yieldunderperformance relative to nominal BTPs. First, linker
real yields rose reflecting liquidity concerns and balance
sheet de-leveraging, and, second, the ECB bought
conventional BTPs in its SMP, focused on maturities out
to the 10Y sector of the curve. While the illiquidity factor
is unlikely to result in further rises in real yields, the
SMP has continued to expand, and we expect increased
buying volumes in BTPs. We see a very small
probability that the SMP starts to buy linkers or BTPs at
the very long end of the curve, so we would continue to
expect high betas with nominals in sub-10Y breakevens.
Another factor we must take into account is the deflation
floor embedded in the principal of BTPeis,
particularly relevant in a negative inflation breakeven
environment. The fact that principal payments of linkers
cannot go below par creates a theoretical floor for
breakevens. This floor is relatively close for the BTPei
September 2016, which only has around 3% of accrued
inflation but reached the breakeven lows of around
-0.50% in mid-November. We can calculate the
breakeven level which is consistent with the principal
hitting the deflation floor, which is somewhere between
-0.60% and -0.70% (which on a compounded basis
would completely offset the 3% of inflation accrual).
Exhibit 19 projects the nominal cash flows and future
HICP index levels applied to BTPei 2016 using the
current inflation breakeven expectation, as well as under
the scenario where we reach the threshold at which the
principal floor kicks in. In this scenario, the final HICP
index level expected is at the same level as the base CPI
value of 109.5227.
Putting all the information together, we have modeled the
expected breakeven moves for all BTPeis under a
significant increase in BTP nominal yield spreads to
Bunds (ranging from 220bp to 290bp) in Exhibit 20. Our
Exhibit 17: Italian linkers with a high dirty price saw significantdeclines in breakevens in November, whereas those with low dirtyprices outperformed – We must allow for this when projecting
performance going forwardInflation breakeven change 1 November–18 November vs. cash dirty price of BTPeilinkers on 1 November; bp
14
16 1719
212326
3541
-150
-100
-50
0
50
100
60 70 80 90 100 110 120Dirty price, points
B / E c h a n g e
Exhibit 18: BTPei breakevens have declined 4bp per every 10bpwidening in the nominal yield spread to Germany and are likely todecline further given our view on peripheral spreadsBTPei 2021 breakeven vs. nominal yield spread in 10Y nominal benchmarks over the past 6M; %
y = -0.43x + 2.79
R2 = 79%
0.0
0.5
1.0
1.5
2.0
2.5
1.0 2.0 3.0 4.0 5.0 6.0
Italy -Germany nom y ld spread, %
B T P e i 2 1
b r e a k e v e n
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 183/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Jorge GarayoAC
(44-20) 7325-4820
[email protected]. Morgan Securities Ltd
183
beta for breakevens is based on the factors highlighted
above. The BTPei 2016 is expected to hit the threshold
breakeven of -0.65%, at which point the deflation floor on the principal starts to kick in, and therefore
outperforms other lines. BTPei 14s and BTPei 17s would
suffer most under our assumption while we would expect
more limited falls in breakevens at the very long end of
the curve.
While we have not modeled alternative widening
scenarios, we should expect broadly proportional
relative performance. Clearly a scenario where Italian
spreads tighten (unlikely in our view) would result in
normalisation in Italian real yields and significant
performance, both in absolute and relative terms.
French linkers should prove more resilient than
Italian to a worsening of the sovereign risk dynamics
Is France likely to follow the footsteps of Italian
linkers? As Italian and Spanish yields spreads widened
significantly, French lines started to underperform in
November as France widened vs. Germany. However,
we believe there are a series of factors that put French
linkers in a better position than Italian ones.
• Sales of Italian linkers (28% of the linker
market) could result in switches into French
(54% of the market if we include both Frenchand Euro HICP-linked lines). With German
linkers only around 20% of the market and
trading at negative real yields, French linkers
should be reasonably attractive alternatives.
Looking at the relative weights of linkers within
indices, the bonds that should attract more
inflows are (in this order) OATei 20, OATi 17,
DBRi 16 & DBRi 20, and OATi 13.
• We expect domestic demand to remain strong
on the back of French CPI hedging (at least in
1Q12) and interest from Libor-based domestic
investors to resurface. Inflows into Livret A
accounts have been particularly strong in 2011
(Exhibit 21), and we predict the Livret A rate
will increase to 2.50% in February 2012 (before
falling back to 1.75% later in the year, see
Exhibit 22). This means that we are likely to
see some French CPI hedging demand in the
early stages of 2012. Domestic Libor-based
demand for linkers should provide some support
to the market in the event of significant
Exhibit 20: Possible scenario for BTPei breakevens if BTP nominal yield spreads widened by an average 250bpCurrent linker metrics for Italian BTPeis and projected levels under nominal yield widening assumptions
Line
Dirty
price
Cum
Infl
Real
Yld
B/E
Infl
Threshold
B/E*
Nom Yld
widening
assume
Beta real
to nom
ylds
Beta to
B/E
Expected
Real Yld
Expected
B/E Infl
Expected
Move (bp)
15/09/2014 103.23 17% 7.12% -0.49% -5.51% 288 1.40 40% 11.1% -1.64% -11515/09/2016 82.58 3% 7.13% -0.34% -0.61% 275 1.10 24% 10.1% -0.61% -27
15/09/2017 85.51 12% 7.20% -0.30% -1.90% 260 1.36 36% 10.7% -1.23% -94
15/09/2019 78.19 6% 6.87% 0.11% -0.80% 250 1.24 24% 10.0% -0.49% -60
15/09/2021 70.96 4% 6.63% 0.31% -0.42% 250 1.24 24% 9.7% -0.29% -60
15/09/2023 73.21 10% 6.81% 0.59% -0.79% 240 1.14 14% 9.5% 0.26% -34
15/09/2026 68.29 2% 6.72% 0.74% -0.13% 230 1.12 12% 9.3% 0.47% -28
15/09/2035 70.20 15% 5.27% 2.36% -0.58% 220 1.10 10% 7.7% 2.14% -22
15/09/2041 63.74 5% 5.16% 2.22% -0.15% 220 1.08 8% 7.5% 2.04% -18 * Threshold breakeven level at which the principal deflation floor would be hit.
Exhibit 19: The breakeven floor-threshold for BTPei is not far fromcurrent levels given current market volatility; breakevens should notgo below -0.66% as this would imply hitting the floor on the principal
Cash flow calculation under current breakeven assumption and thresholdassumption for BTPei Sep 2016 (Base CPI value 109.5227);Nominal cash flows and Index points
Date Current B/E= -0.20%p.a. Threshold= -0.66% p.a.
Cash Flows HICP index Cash Flows HICP index
21/11/2011 -83.988 112.796667 -82.176 112.79667
15/03/2012 1.083 112.999 1.082 112.863
17/09/2012 1.082 112.855 1.078 112.482
15/03/2013 1.081 112.715 1.075 112.111
16/09/2013 1.079 112.574 1.071 111.736
17/03/2014 1.078 112.432 1.068 111.361
15/09/2014 1.077 112.293 1.064 110.996
16/03/2015 1.075 112.152 1.061 110.623
15/09/2015 1.074 112.013 1.057 110.258
15/03/2016 1.073 111.872 1.054 109.890
15/09/2016 103.088 111.732 101.050 109.5227
IRR 6.64% 6.71%
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 184/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Jorge GarayoAC
(44-20) 7325-4820
[email protected]. Morgan Securities Ltd
184
cheapening of linkers in asset swap terms,
though the timing of such flows is difficult to
predict.
Outlook for real yields, breakevens, and
linker asset swaps
Real yields can go more negative
Real yields for German linkers are negative out to the 7Y
sector and a mere 30bp in the DBRi 2020 line. Real
yields for French linkers are slightly higher given the
current credit risk dynamics. The main driver for real
yields has been a deteriorating macro outlook,
combined with the flight-to-quality resulting from the
sovereign risk crisis. We think these factors are likely tocontinue to drive real yields into more negative territory
while the ECB easing the real refi rate should also
contribute to lower real yields at the front end (with our
view not fully priced in, see European Derivatives). With
a forecast of 10Y Bunds reaching 1.25% by 2Q12 and
Italian spreads widening 250bp (see Euro Cash), we
believe real yields will go negative in the 10Y sector. If
we use the beta of real to nominal yields of the past 12
months, the DBRi 2020 should decline around 40bp, to
negative -0.10% real yield, while 10Y breakevens would
fall by a similar extent, to around 1.10%.
Upside risks to our inflation call
The J.P. Morgan inflation assumption is consistent with
inflation staying a bit stickier than the path implied by
inflation swaps (Exhibit 23). Moreover, it does not
factor in upside risks from rises in state-administered
prices (VAT and duties), which we see as a potential
upside risk for the beginning of 2012. With Italian,
Spanish, and French public finances under close scrutiny,
we believe that any shortfalls in austerity plans are likely
to be met with rises in taxes or state-controlled prices.
Italy and Spain have just formed new governments,
which will be very much focused on providing a deficit-
reduction plan, while France is being closely watched by
markets.
We have already seen a rise in ‘administered-prices’
inflation in Italy and Spain, where these correspond to
around 8% of the respective inflation baskets (Exhibit
24). In Exhibit 25, we set out the impact state-controlled
prices may have on Euro HICP inflation, looking at VAT
rises, VAT reclassifications, general rises in administered
prices (which include services such as refuse collection,
water supply, as well as pharmaceutical products), and
Exhibit 21: Livret A inflows have been significant during 2011, weexpect continued inflows in the first months of 2012…Inflows into Livret A accounts, centralised by CDC; €bn
-2
-1
0
1
2
3
4
5
J a n - 0
9
A p r - 0 9
J u l - 0 9
O c t - 0 9
J a n - 1
0
A p r - 1 0
J u l - 1 0
O c t - 1 0
J a n - 1
1
A p r - 1 1
J u l - 1 1
2 0 0 8 *
2 0 0 7 *
18.3YTD €16.2bn2010 €7.8bn
Exhibit 22: … as we expect the Livret A rate to go up to 2.50% inFebruary before moving down to 1.75% in the following fixing (giventhe expected downward trajectory in inflation)Livret A rate*, French CPI ex tobacco inflation oya, previous month average of 3MEuribor and EONIA; %
-1.00
0.00
1.00
2.00
3.00
4.00
5.00
Jan07 Jan08 Jan09 Jan10 Jan11 Jan12
Inflation
Avg of 3M Euribor & EONIA (1M)
Liv ret A Rate
Current: 2.25%
Forecasts
* Livret A rate is set by Bank of France every 15 January and 15 July with referenceto an average of inflation and rates (also depicted in the Exhibit), floored at inflationplus 25bp.
Exhibit 23: Inflation swaps assume a larger fall in inflation than theJ.P. Morgan forecast, even if this does not include upside risks fromadministered prices, state-controlled duties and VATEuro HICP ex-tobacco J.P. Morgan forecasts and implied by HICP swaps; % oya
1.0
1.5
2.0
2.5
3.0
Oct11 Dec11 Feb12 Apr12 Jun12 Aug12 Oct12 Dec12
JP Morgan EURO HICP sw aps
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 185/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Jorge GarayoAC
(44-20) 7325-4820
[email protected]. Morgan Securities Ltd
185
duties on energy prices. Our conclusion is that the
potential impact from these measures on Euro HICP
can be quite substantial, somewhere between 0.2%and 0.5%, on inflation.
Inflation breakevens offer value at the front end…
Traditional fundamental models for inflation breakevens
would fail to explain recent performance. Breakevens
were driven by nominal yields, peripheral spreads and
equities more than by inflation fundamentals. As a
measure of value in inflation breakevens, we prefer
looking at the excess nominal returns that sub-5Y linkers
would provide over conventional bonds, using a blend of
our inflation scenario to 2013 and the path implied by
inflation swaps thereafter (Exhibit 26). Linkers offer above-average returns vs. conventionals in the case of
France (except OATei 12), but even the German OBLi
13 offers a substantial 65bp pickup vs. Bunds. Even if
the current illiquidity dynamics justify positive excess
returns in linkers, we think front end breakevens offer
good value at these levels.
Long end breakevens to be driven by nominal yields
and a flight to quality dynamic – Favour breakeven
flatteners
Inflation breakevens further out are much more
likely to be driven by the performance of Bund yieldsand a flight-to-quality dynamic. The inflation
breakevens implied by the French HICP-linked lines
are only a touch lower than German breakevens (Exhibit
27), and we see risks that they underperform in
relative terms, driven by a general widening of
French nominal yields vs. Bunds.
Exhibit 24: “Administered prices” inflation has sky rocketed in Italyand Spain, as one would expect with countries engaging in fiscaltightening, thus we expect more pressure on this front
Administered price inflation as defined by Eurostat*; % oya
-4%
-2%
0%
2%
4%
6%
8%
10%
2002 2004 2006 2008 2010
France Italy Spain
* Items correspond to around 8% of basket in Italy & Spain, 13.5% in France.
Exhibit 25: Fiscal austerity programmes can impact short-term inflation significantly, with numerous challenges putting pressure to reducedeficits: optimistic growth forecasts and increased visibility will force countries to hike VAT, administered prices, and other duties. Italy, Spain,and France could push Euro HICP by up to half a percentage pointImpact on Euro HICP inflation from hypothetical changes in government-controlled measures
Euro area
HICP
weight
VAT rate
(standard
%)
Admin
prices
weight
>>
% duties
on
gasoline
1% higher
VAT*
VAT
reclass**
Admin
prices***
2.5% rise
Rise
duty on
gas^
Total
potential
Germany 26% 19.0 12.9 59.2 0.11 0.06 0.08 0.25
France 21% 19.6 13.6 57.3 0.09 0.06 0.07 0.22
Italy 18% 21.0 8.1 55.8 0.08 0.04 0.04 0.01 0.16
Spain 13% 18.0 8.0 49.0 0.05 0.03 0.03 0.06 0.16
Netherlands 5% 19.0 13.6 60.3 0.02 0.01 0.02 0.05
Greece 4% 23.0 9.8 59.3 0.02 0.01 0.01 0.03
Euro area 19.8 11.3 56.9 0.37 0.21 0.24 0.07 0.88
France + Italy + Spain impacts combined 0.22 0.12 0.13 0.07 0.55
Euro HICP inflation sens itivities
>> Weight of items whose prices are fully or mainly administered by the State, as defined by the European Commission.* Effect from a 1% rise in s tandard VAT rate. Assumes a pass through of 65% on 65% of the total basket, or around a 0.40% increase in domestic inflation.** Assuming a VAT reclassification of 2% of basket items from the reduced rate of VAT to a standar rate of VAT.*** Assuming a 2.5% rise in all administered prices in country baskets (average administered price inflation has been 3.5% over the last 7 years).^ Assuming convergence in duties as a % of overall price of gasoline to the Euro area average, affecting the item “Fuels for personal transport equipment” in the basket.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 186/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Jorge GarayoAC
(44-20) 7325-4820
[email protected]. Morgan Securities Ltd
186
Overall, being short breakevens is not necessarily the
best risk-adjusted trade, given the expected volatility
with sovereign risk dynamics. Therefore, we preferexpressing this view as a curve trade through
inflation breakeven flatteners. This is consistent with
our expectation of significant flattening in the nominal
yield curve (in 2s/10s), plus our view that inflation can
surprise on the upside in 1H12. There are different
variants in which we can express this view: 1)1s/10s
flatteners in HICP swaps (more of a play on inflation
surprising on the downside), 2) overweighting OBLi 13
vs. BTPei 19 in breakeven terms (incorporating a
negative view on Italian breakevens), or 3) BTPei 16s vs.
BTPei 21 (incorporating the value of the deflation floor
in BTPei 16).
Inflation swaps are likely to outperform cash
breakevens as peripheral spreads continue to widen,
and in general, we wouldn’t fade the relative historical
cheapness in the first months of 2012. A further
cheapening may provide good medium-term
opportunities to buy linker breakevens vs. swaps.
Regarding Euro HICP-linked vs. French CPI linkers,the latter have underperformed significantly in November and present opportunities for investors to
overweight OATis vs. OATeis over the medium term.Whereas OATis have traded historically at higher
breakevens than OATeis, they are currently tradingaround similar breakeven levels, around 20bp too cheapvs. a regression since January 2010 in the sub-10Y sector (Exhibit 28). While this may reflect the relatively lower liquidity of French CPI-linked lines, we expect a decentamount of Livret A hedging to take place early in 2012,which should favour French CPI lines. Any demand for French linkers from Libor-based investors should alsofavour French CPI-linked lines, given they provide amore attractive pickup.
Supply/demand dynamics
Euro area linker supply: risk of very low supply fromItaly, lower volumes than in 2011
Supply of inflation-linked bonds in 2011 was below our
initial expectations (€41bn vs. our initial forecast of
€51bn), as Italy issued only around €1.5bn of linkers (net
of buybacks) in 2H11. This was clearly a reflection of the
limited demand for Italian linkers during the sovereign
debt crisis that began unfolding in the summer. France
continued to issue at regular auctions, supplying around
Exhibit 26: Sub 5Y linkers will provide reasonable pickup vs.conventionals under our inflation scenario, which suggestsbreakevens offer value
Current levels, Nominal IRR under J.P. Morgan inflation assumption and pickuprelative to conventional bonds; % and bp
Lines Real Yld Inf B/E
Nom Rtn
(JPM*)
Pickup vs.
Noms (bp)
Euro HICP lines
Germany OBLI 2.250 0413 -0.60 0.86 0.97 65
Germany DBRI 1.50 0416 -0.35 1.20 1.31 35
France OAIE 3.000 0712 -1.52 2.43 0.84 7
France OAIE 1.600 0715 1.30 0.91 3.22 88
France CPI lines
France OAIE 2.500 0713 0.28 1.12 2.14 55
France BTNS 0.45 16 IL 1.59 0.87 3.50 91
France OAIE 1.000 0717 1.75 1.04 3.70 86
* Projected nominal return using J.P. Morgan forecast out to December 2012 andinflation forwards implied in Euro HICP swaps thereafter with J.P. Morgan seasonalmodel.
Exhibit 27: Inflation breakevens for French linkers are trading closeto those of German lines, and could under perform, though we do notexpect them to behave like Italian breakevensInflation breakevens for Euro HICP linked lines (x-axis denotes redemption year); %
-1.00
-0.50
0.00
0.501.00
1.50
2.00
2.50
12 14 16 18 20 22 24 26 28 30 32 34 36 38 40
Germany
France
Italy
Exhibit 28: French CPI-linked breakevens are cheap vs. French EuroHICP-linked breakevensResidual from regressing OATi 2019 vs. OATei 2020 breakeven, since January2010; %
-30
-20
-10
0
10
20
Jan10 Apr10 Jul10 Oct10 Jan11 Apr11 Jul11 Oct11
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 187/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Jorge GarayoAC
(44-20) 7325-4820
[email protected]. Morgan Securities Ltd
187
€3.5bn during this period, while Germany did not hold
any auctions between April and November.
With sovereign risk dynamics likely to continue to
impact markets, we expect supply in 2012 to be a
below-average year. The total supply to the market will
very much depend on Italy’s strategy towards linker
issuance in a dislocated market. While we are calling for
wider BTP spreads, the timing of such widening may
impact the issuance strategy from the Italian Treasury,
which will remain as flexible as possible. We have
therefore forecast Italian linker supply to be of the order
of €12bn, assuming they continue to tap the market in an
opportunistic manner. We expect Germany to issue €9bn
while we have penciled €18bn for France (with a bit
more issued in Euro HICP lines through a new 5Y benchmark in Q1). There are clear downside risks to
Italian linker issuance, which could result in a bit more
issued from France and Germany. Overall, we expect the
supply of linkers to be at best around €39bn, slightly
below what was issued in 2011 (Exhibit 29).
On the demand side, inflation-hedging at the long end
of the curves has been very limited in 2011, and we do
not expect a huge change in this dynamic in 1H12.
The pension reform agenda in the Netherlands can be a
significant development for inflation-hedging, but we do
not expect a quick implementation of the proposed
measures, which give more leeway for Dutch pensionfunds to be more active in inflation hedging programmes.
Cross market: relative flattening vs. the
US inflation curve
Our view for Euro area breakevens contrasts with our
outlook for US TIPS, where we see short dated
breakevens most at risk and expect the breakeven curve
to steepen (around 40bp in 5s/30s by 2Q 12, see US TIPS
section). This is on the back of significant expected falls
in US inflation prints and further impact from fiscal
tightening on short dated breakevens.
Thus we generally recommend overweighting Euro
area breakevens at the front end vs. US TIPS, whilst
doing the opposite in 10Y+ maturities. In swaps space,
the 5s/30s curve in HICP swaps is broadly at the same
level as in US CPI swaps, whereas on average the latter
has traded 25bp steeper (Exhibit 30).
Exhibit 29: Euro area linker supply forecasts for 2012: lower volumesthan an already low 2011Gross supply (net of buybacks); €bn
Q1 Q2 Q3 Q4 Total % vs 2011Euro HICP 10.5 8.0 7.5 6.0 32.0 82% +0.8
France (Euro) 5.3 2.8 1.5 1.5 11.0 28% +1.6
Italy (Euro) 3.3 3.3 3.0 2.5 12.0 31% (1.8 )
Germany (Euro) 2.0 2.0 3.0 2.0 9.0 23% +1.0
France CPI 2.8 2.8 0.8 0.8 7.0 18% (3.6 )
Total 13.3 10.8 8.3 6.8 39.0 (2.8 )
% of total 32% 26% 20% 16% 93%
vs 2011 (3.5 ) (4.6 ) +4.6 +0.7 (2.8 )
By Maturity
Short (<3yrs) 1.3 1.5 1.3 1.0 5.0 11% +5.0
Medium (3-7yrs) 6.8 4.3 0.0 2.0 13.0 28% (5.7 )
Long (7-12yrs) 2.5 2.8 2.5 2.3 10.0 22% (5.6 )
Ultra long (13yrs+) 2.8 2.3 4.5 1.5 11.0 24% +3.6
Exhibit 30: The 5s/30s inflation curve should be steeper in US CPIswaps compared to Euro HICP swaps5s/30s HICP swap and US CPI swap curves; bp
0
20
40
60
80
100
120
140
Nov 09 May 10 Nov 10 May 11 Nov 11
US CPI 5s/30s
Euro HICP 5s /30s
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 188/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Jorge GarayoAC
(44-20) 7325-4820
Francis Diamond (44-20) 7325-3541J.P. Morgan Securities Ltd
188
UK inflation-linked markets
2011 reviewIndex-linked gilts underperformed nominals during 2011.
Despite the fact that both CPI and RPI inflation both
broke the 5% level (5.2% and 5.6%, respectively), the
linker market was driven by the collapse in nominal
yields as investors sought the safe haven of UK gilts as
the situation in peripheral Europe deteriorated. On the
real yield curve, 5Y real yields fell the most but were
also the most volatile (Exhibit 31), while at the long end
of the curve, 30Y and 50Y real yields have moved
broadly in lockstep. Real yields are barely above 0bp at
the long end, and although this part of the curve is much
less affected by fundamental valuations and macro
drivers than the short end, there has been a high
correlation between 30Y and 10Y real yields.
Breakevens are lower on the year with the movement
driven entirely by the fall in nominal gilt yields. The
resumption of QE purchases in October had little impact
on cash breakevens, despite the fact that linkers are not
included in the QE purchase basket. This was not the
case when QE was first started in March 2009 when 10Y
breakevens initially fell 70bp following the
announcement, only to reverse this over the following
weeks. 2011 saw the issue of new ILG29, ILG34, and
ILG62 gilts via syndication with the latter, attracting anorder book close to £10bn.
2012 inflation view
2011 saw CPI inflation rise to the highest levels since
September 2008, driven by higher food and energy
prices. Core CPI rose above 3.5% oya during 1H11
before subsiding. Looking forward into 2012, we think
headline and core inflation will fall markedly to 2.7%
oya and 2.0% oya, respectively by the end of 2Q12 (see
Exhibit 7 in the earlier section Inflation outlook for
2012). The drivers are lower food and energy prices on
the back of base effects, a downshift in core goods pricesdriven by a weaker growth outlook and the impact of the
VAT hike falling out of oya inflation (Exhibit 32). We
forecast the contribution of core goods inflation to
headline CPI to be slightly negative by the end of 2012.
The main risks to our inflation forecast stem from
downside risks to growth. Our baseline view is that the
UK generates anaemic growth of 0.5% oya in 2012 but
there is clearly the risk that output is flat or negative over
the year, dragging core goods and services inflation
down. Aside from the usual monthly volatilities
surrounding the CPI comments, it is difficult to envisage
a scenario in which inflation does not fall from current
levels. The impact on inflation of QE increasing to
£425bn as per our forecast is difficult to gauge, but the
BoE estimates that the impact of £200bn QE on headline
Exhibit 31: 5Y real yields outperformed in 2011 and correlation withnominal yields was high across the curvePerformance of UK index linked gilts from 1 January 2011–17 November 2011; bp
unless stated Real y ield Current High Low Chg.
Std.
Dev*
R-sq vs.
10Y real
R-squ
vs. nom
ILG16 -140 6 -153 -132 5 93% 96%
ILG22 -30 87 -36 -96 5 100% 95%
ILG32 5 91 3 -63 4 97% 91%
ILG40 5 88 5 -61 4 95% 88%
ILG55 4 72 3 -48 3 88% 73%
Exhibit 32: Our forecast for lower headline inflation in 2012 is drivenby falls in all the main componentContribution to headline CPI by component, %oya
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
Jan-11 May -11 Sep-11 Jan-12 May -12 Sep-12
FoodEnergy
Core goods
Services
Exhibit 33: An increase in QE gilt purchases will push 5Y real yieldslower– we forecast a 50bp fallEstimated change in 5Y real yields based on two factor model* under variousscenarios for additional QE and level of RPI by mid 2012; bp
4.50 4.00 3.50 3.00 2.50 2.00
50 -20 -10 -5 5 10 20
75 -35 -25 -20 -10 -5 5
100 -45 -40 -35 -25 -20 -10
125 -60 -55 -45 -40 -35 -25
150 -75 -70 -60 -55 -45 -40
Mid 2012 RPI, %oya
A d d i t o n a l Q E b y
m i d 2 0 1 2 ,
£ b n
* we model 5Y real y ields as a function of the base rate adjusted for QEpurchases** and the level of RPI.5Y = 0.48 * effective policy rate – 0.15* RPI. R-squ: 81%, std. error: 50bp** we assume £25bn of QE lowers base rate by 0.3%. Current effective rate is -2.8%.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 189/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Jorge GarayoAC
(44-20) 7325-4820
Francis Diamond (44-20) 7325-3541J.P. Morgan Securities Ltd
189
CPI has been between 0.75% and 1.50%, and we don’t
discount that a further increase in QE could prevent
inflation falling as low as we forecast. The monetaristargument that excess reserves will generate inflation
will only become relevant once economic recovery is
achieved and demand for credit increases – a
situation which is several years away, in our view.
Real yields view – be long 10Y real
yields
Real yields are at historically low levels, but we think
that they can fall further in 2012, particularly in the sub-
10Y sector of the curve. The QE increase in October
2011 and the very dovish set of forecasts in the
November Inflation Report suggest that the BoE isfocused on managing downside risks to growth, with
more QE likely in the near future in our view. Our model
of 5Y real yields as a function of the base rate adjusted
for QE purchases and the level of RPI (to measure
demand for inflation protection) suggests that 5Y real
yields can fall around 40bp-50bp by mid-2012
(Exhibit 33) from current levels. Given the strong
correlation between 5Y and 10Y real yields we expect
ILG22 real yield to fall around 30-40bp by mid 2012.
We note that, long positions in 5Y and 10Y linkers offer
attractive carry and slide vs. the forwards (Exhibit 34).
During 2010, 30Y real yields traded in a very pronounced 60bp–90bp range (ILG40) and many end
investors tailored their investment strategies around
playing this range. This has not been the case this year as
30Y yields have trended downwards (Exhibit 35). One
of the key questions for 2011, in our view, is the
direction of 30Y real yields. Will 30Y yields rise back
up to the 2010 levels or will a new trading range be
established around current levels?
To answer this question, we need to take a look at the
drivers of 30Y real yields, namely nominal 30Y yields
and the supply/demand outlook. We note that that
correlation between 30Y real and nominal yields has not
been constant over the past few years, particularly during
periods of QE gilt purchases when the relationship has
tended to be weaker (Exhibit 36). This correlation also
tends to increase when nominal yields are trending
higher or lower (Exhibit 37).
We can model 30Y real yields as a function of 30Y
nominal yields and the size of pension fund liabilities
(Exhibit 38) where an increase in liabilities tends to
result in lower real yields. The intuition here is that as
Exhibit 34: Long positions in 5Y and 10Y real yields look attractivecompared to the forward real yield curveImplied carry and slide for being long real par rates over 3M, 6M, 9M, and 12M
horizon; bp
7
14
21
28
6
11
17
22
13
45
0
5
10
15
20
25
30
3M 6M 9M 12M
5Y 10Y 30Y 50Y
Exhibit 35: Long-end real yields trended lower in 2011 after range-trading for most of 2010Par 30Y real yield; %
0.00
0.20
0.40
0.60
0.80
1.00
1.20
Feb 10 Aug 10 Feb 11 Aug 11
Exhibit 36: The correlation between 30Y real and nominal yields isweaker when the BoE is conducting QE gilt purchasesBeta and r-squared between 30Y real par and 30Y nominal par yields, January2009–October 2011
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
Beta R-squ
QE Non-QE
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 190/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Jorge GarayoAC
(44-20) 7325-4820
Francis Diamond (44-20) 7325-3541J.P. Morgan Securities Ltd
190
liabilities, which are inflation-linked, increase, then
pension funds are more likely to buy index-linked gilts to
hedge these, although we note that changes in liabilitiesmay not always result in renewed hedging flows as they
can also increase due to increases in longevity and tend
to increase as the point of payment increases. This
model suggests that 30Y index-linked gilt yields
currently look cheap (Exhibit 39), and our forecast for
a 100bp fall in 30Y gilts would imply 30Y real yields
rallying 30bp to around the -25bp level, although we
think that in practice any move to this level would be
very slow and gradual. The 0bp level may represent a
psychological barrier around which investors may
attempt to fade any move to negative 30Y real yields. At
the other extreme, we think any move in 30Y real yields
back up to the 1% level is highly unlikely to occur over 2012 as 30Y nominal gilt yields would need to rise to
around 5% based on the current betas.
Breakeven view
Breakevens have fallen over the past year by 50bp in the
10Y and 30Y sectors, and by nearly 100bp in the 5Y
sector, despite the rise in CPI inflation to 5.2% and RPI
inflation reaching the 5.6% level. In fact, over the past
few years 5Y cash breakevens have show a weak
relationship with spot inflation prints unless RPI is below
the 2% level (Exhibit 40) and over the past couple of
years, nominal yields have been the dominant drivers of breakevens (Exhibit 41).
Our breakeven view for 2012 is as a straightforward one
– lower nominal yields will drive breakevens lower in the
5Y and 10Y sectors of the curve. Our forecast for a
75bp fall in 10Y gilt yields during 1H12 (to the
1.50%) level would broadly translate into a 30bp fall
in ILG22 breakevens, to the 2.30% level. Our
expectations for CPI and RPI inflation may, at the
margin, also contribute to lower breakevens. In the 30Y
sector, we note that non-linearities are starting to appear
in the relationship between breakevens and nominal
yields (Exhibit 42), which is likely a result of demand-specific drivers. Pension funds and ALM-driven
investors tend to have specific trigger levels at which
they look to hedge inflation which may limit the
magnitude of any down move in 30Y breakevens.
The 5s/10s inflation curve has been broadly range-bound,
particularly in the RPI swap space, and we think this can
continue. We think it is far too early to position for
sustained steepening on the back of unanchored inflation
expectations from an increase in QE. The monetarist
Exhibit 37: The relationship between 30Y real yields and nominals isweaker when nominal yields are range-boundStatistics from regressing 30Y par real yields vs. 30Y par nominal yields
Period 30Y y ield Beta Rsq
19/11/09 - 15/2/10 sell-off 0.8 64%
16/2/10 - 16/12/10 range 0.2 37%
10/2/11 - 18/11/11 Rally 0.5 89%
Exhibit 38: 30Y real yields can be explained by nominal yields andthe size of pension fund liabilities…Statistics from regressing 30Y par yield vs. 30Y nominal par yield and the PPFmeasure of total liabilities for DB pension funds, monthly data, March 2003–Oct11
Variable Coefficient T-stat
30Y nominal y ield (%) 0.308 3.6
PPF liabilities (£bn) -0.003 -18.2
R-squ 81%Std. error (bp) 21
Exhibit 39: …and currently look some 30bp cheap Actual and predicted level* of 30Y par real yields, monthly data March 2003–October 2011; %
-0.50
0.00
0.50
1.00
1.50
2.00
2.50
Mar-03 Mar-05 Mar-07 Mar-09 Mar-11
Predicted
Actual
0bp level
* predicted using model in Exhibit 38.
Exhibit 40: 5Y breakevens are generally non-directional with inflationwhen RPI is above 2%5Y par cash breakeven* regressed against RPI print, 2005–2011**; %
y = 0.1x + 2.55
R2 = 3%
y = 0.31x + 2.3
R2 = 29%
1.00
1.50
2.00
2.50
3.00
3.50
4.00
-2.0 -1.0 0.0 1.0 2.0 3.0 4.0 5.0 6.0
RPI, %oya
* Derived from BoE fitted par curve.** Excludes period from December 2008–February 2002.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 191/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Jorge GarayoAC
(44-20) 7325-4820
Francis Diamond (44-20) 7325-3541J.P. Morgan Securities Ltd
191
arguments that, for additional QE, translates into higher
inflation is not easy to demonstrate empirically, and we
think that inflation expectations run the biggest risk of becoming unanchored when the macro backdrop has
improved. This scenario is several years away, in our
view, and we see little benefit in positioning for this
outcome now.
Cash breakevens vs. RPI swaps
The past few months have seen a marked
underperformance of index-linked cash breakevens
compared with RPI swaps, with the relative spread now
at the widest levels ever for ILG55 and at its widest level
since March 2009 for ILG37 (Exhibit 43). This spread
essentially reflects the relative ASW levels betweenindex-linked gilts and conventional gilts, and indicates
that linkers offer an attractive pick-up vs. Libor
compared with conventional gilts.
Going forward, we think this relative spread can
widen further, and we would not advocate fading this
apparent cheapness. We think Libor-based investors may
look to take advantage of the relative cheapness in
linkers compared with nominals, but over the past 6
months, this relative spread has been driven by nominal
gilt ASW levels and European peripheral spreads,
particularly in the 10Y+ sector (Exhibit 44). We think
cash breakevens will continue to underperform swapbreakevens, given our view that peripheral spreads will
widen further in the coming months and that nominal
ASW levels may fall further (i.e. gilts outperform
swaps), driven by additional QE and lower nominal gilt
yields. For investors who do want to fade this cheapness,
we think the sub-10Y part of the curve is the best place
to position, given the lower correlation with peripheral
spreads and nominal ASW levels.
Market technicals
Issuance
Our fiscal outlook and issuance forecasts for gilts are
described in the United Kingdom section, and for
FY12/13, we expect linker issuance to break through the
£40bn level (Exhibit 45). The low level of spot and
forward real yields, and the continued backdrop of strong
demand from institutional investors warrant an increase
in index-linked gilt supply, in our view. Total gilt supply
is expected to rise to £190bn, and we forecast linker
issuance remaining around 23% of total gilt supply. As is
typically the case, we expect the bulk of issuance to
come in the long end (15Y+) of the real yield curve, with
taps to ILG34, ILG40 and ILG62 likely. We also see a
reasonable chance of a new linker in the 35Y sector of
the curve, as well as the possibility of a short-dated linker
in the 10-12Y sector, although this is less likely now that
the ILG29 has been issued.
Exhibit 41: Breakevens have been driven by nominal yields over thelast couple of years. We expect the strong relationship to continue …Statistics from regressing breakevens vs. nominal yields over the last 12M and 24M
Beta R-squ Beta R-squ
ILG22 0.4 87% 0.4 68%
ILG32 0.5 91% 0.5 77%
ILG40 0.5 92% 0.5 81%
ILG55 0.6 93% 0.7 84%
12M 24M
Exhibit 42: …but the long end of the BE curve is becoming non-linear ILG40 regressed against nominal 30Y yields, last 3M; %
y = 0.64x 2 - 3.9x + 9.0
R2 = 83%
2.90
3.00
3.10
3.20
3.30
3.40
3.50
3.0 3.2 3.4 3.6 3.8 4.0
30Y nominal y ield, %
18 Nov
Exhibit 43: Cash breakevens are at cheap levels vs. RPI swaps
across the curve…Cash breakeven – RPI swap for selected linkers; bp
-50
0
50
100
150
Oct 08 Jun 09 Feb 10 Oct 10 Jun 11
ILG17 ILG37 ILG55
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 192/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Jorge GarayoAC
(44-20) 7325-4820
Francis Diamond (44-20) 7325-3541J.P. Morgan Securities Ltd
192
The results of the consultation process on whether the
DMO should issue CPI-linked gilt should be available in
the coming months. In our view, the decision as towhether CPI-linked gilts will be issued will be very much
driven by anticipated end-investor demand, particularly
pension funds and other investors who will be required to
provide CPI-linked payments. We expect around 15% of
pension schemes with indexation linked to RPI will be
able to or will want to make the switch from RPI to CPI
indexation, as mandated by new legislation. However,
given that pension fund schemes hedge only around 60%
of their inflation risk, they may prefer to continue to use
RPI-linked instruments and run the CPI-RPI basis risk.
If CPI-linkers were to be issued, we think they would
likely be in the 15-30Y maturity sector of the curvewhere demand is likely to be the highest. We would also
expect a flexible issuance programme to enable the DMO
to tailor issuance needs in either the CPI-linked or RPI-
linked space to meet demand, with the allocation of
supply between the two instruments likely to vary during
the fiscal year.
Real yield behaviour into syndications
For the past couple of fiscal years, linker syndications
have risen to around 40%–50% of total index-linked gilt
sales, and we expect a similar proportion of linkers to be
issued via syndications in FY12/13, equating to £18bn– 22bn of syndicated linker issuance, based on our
forecast. Since the DMO introduced syndications as a
permanent feature of the issuance programme in April
2009, there have been 9 index-linked gilt syndications,
raising a total of £38.75bn, with an average size of
£4.3bn.
These syndications have given rise to some interesting
behaviour on the linker curve. Looking at the
performance of the closest maturity bond to the linker
being issued, we note that, on average, real yields have
risen 8bp in the 15 business days prior to the syndication
(Exhibit 46). This is then quickly reversed in the fivedays following the syndicated supply. This phenomenon
has occurred into 7 out of the 9 syndications, with the
largest increase close to 25bp. We think that this can be
explained by the fact that 1) syndications are often
multiples of the size of typical index-linked gilt auctions
and represent large concentrated amounts of linker
duration supply into the market. This creates uncertainty
around the actual potential demand for the bond; and 2)
the actual details of the bond maturity are not announced
until a few weeks before the syndication date.
Exhibit 44: … and peripheral spreads and nominal ASW levelsexplain the bulk of this cheapening for 10Y+ l inkersR-squ from separately regressing* cash BE – RPI spread vs. nominal ASW and
peripheral spreads, last 6M; %
0%
20%
40%
60%
80%
100%
ILG17 ILG22 ILG37 ILG55
Nom ASW Peripheral spread
* Single factor regressions.
Exhibit 45: We expect over £40bn of linker issuance in FY12/13Index linked gilt total issuance and supply as % of total gilt issuance£bn %
0
10
20
30
40
50
2 0 0 0
2 0 0 1
2 0 0 2
2 0 0 3
2 0 0 4
2 0 0 5
2 0 0 6
2 0 0 7
2 0 0 8
2 0 0 9
2 0 1 0
2 0 1 1
2 0 1 2
( f )
0%
10%
20%
30%
40%
Amt, £bn Proportion, %
Exhibit 46: Real yields rise prior to linker syndications on average… Average real yield of closest maturity linker around index linked gilt syndications*,July 2009–October 2011; bp
55
60
65
70
-20 -15 -10 -5 0 5 10 15 20
Business day s around linker sy ndications
* Dates and linkers used: 23/07/09 ILG42, 24/09/09 ILG50, 27/01/10 ILG40,26/05/10 ILG50, 27/07/10 ILG40, 27/01/11 ILG55, 24/05/11 ILG34, 26/07/11 ILG34,25/10/11 ILG62.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 193/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Jorge GarayoAC
(44-20) 7325-4820
Francis Diamond (44-20) 7325-3541J.P. Morgan Securities Ltd
193
We expect this trend to continue and we think
investors can exploit this dynamic by selling the
closest maturity bond to the linker being issued viasyndicate in 15 business days ahead of linker
syndications. Instigating a trading rule in which the
closest maturity linker to the syndicated bond is sold on
the 15th day prior to the syndication and unwound on the
day of the syndication itself would have generated a total
of 24bp of P/L in yield terms in 2011, and on average,
6bp per syndication over the past three years (Exhibit
47).
Demand for linker syndications is almost entirely from
domestic investors, and the size of the book as well as
the actual amount of the bonds issued can be affected by
large asset allocation orders. Predicting the size of the potential demand for linker syndication in advance is not
easy, but looking at the relative valuations between
equities and real yields may indicate potential demand.
The idea here is that pension funds and asset managers
are more likely to switch from equities into linkers when
equities are relatively rich compared with linkers on a
real-yield basis. This is because pension funds
asset/liability mix is not managed based on specific
funding level triggers but more holistic, long-term
funding plans, and on aggregate, funds are reluctant to
crystallise equity losses.
We can measure the relative valuations by looking at theratio of equity E/P (as a measure of equity real yield) vs.
the real yield on long linkers. Looking at this valuation
metric and its evolution in the few days prior to
syndications there appears to be a loose relationship i.e.
syndication demand is stronger if equities have richened
vs. linkers, although this is skewed by the strong demand
for the ILG62 syndication in October 2011. It is worth
noting that the syndication of ILG62 coincided with the
richest levels of equity/linkers for 2011, and we think
that looking at the performance of this ratio ahead of
future linker syndications may provide some indication
of potential investor demand.
Exhibit 47: …creating attractive trading opportunities for investorsTotal and average P/L from trading rule which goes short closest maturity linker onT-15 and unwinds the trade on T*; bp
0
5
10
15
20
25
30
35
2009 2010 2011
Total P/L Average P/L
* T = date of linker syndication.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 194/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Jorge GarayoAC
(44-20) 7325-4820
Kimberly Harano (1-212) 834-4956J.P. Morgan Securities Ltd, J.P. Morgan Securities LLC
194
US inflation-linked markets
Up, down, then back aroundThe path of breakevens over 2011 looks remarkably
similar to the 2010 experience. First, breakevens steadily
widened over the first few months of the year thanks to a
surge in energy prices (Exhibit 48). In late spring,
however, European concerns resurfaced, the US
economy appeared to hit a speed bump, and the resulting
risk-off trade pushed breakevens narrower. Finally, in the
fall, the Fed was forced to step in with Operation Twist,
and this action, along with some positive news out of the
October EU summit, helped breakevens reverse some of
their narrowing. On a total return basis, the sizable
increases in headline CPI helped TIPS outperform
nominals early in the year, but TIPS sharply
underperformed Treasuries in 3Q11 as nominal yields
declined more than real yields (Exhibit 49).
Real yields, in turn, refreshed their all-time lows this year
(Exhibit 50). Five-year yields traded negative for most
of 2011, hitting a low of -1.12% in early November. As a
result, the Apr-16 TIPS became the first Treasury
security issued with the new minimum 0.125% coupon
rate, despite a clearing yield of -0.18%. Even 10-year
real yields turned negative in 3Q11, reaching -0.15% at
the lows.
TIPS supply and demand in 2012
After Treasury announced modifications to the TIPS
calendar at the November 2010 refunding, 2011 became
the first year with a TIPS auction every month. At this
year’s November refunding, Treasury wrote that “TIPS
are an important part of Treasury’s overall debt
management strategy” and indicated that it “expects to
continue to gradually increase gross issuance of TIPS in
2012.” Based on these comments, we look for modest
increases in auction sizes across the curve, for a total
of $146bn of gross issuance in 2012 (Exhibit 51)
compared to $131bn of expected issuance in 2011. With$46.3bn of redemptions in 2012 and net Fed purchases of
zero, effective net issuance in 2012 will likely rise to
$100bn versus $79bn (net of redemptions and Fed
purchases) in 2011.
Alternatively, we can evaluate changes in TIPS supply
on a duration-weighted basis, since the Fed will
remain a source of duration demand in 2012, even if
its holdings of TIPS will likely remain unchanged in par
notional terms. Over two rounds of purchases and sales
operations, the Fed has bought about $2.3bn of 10-year
TIPS equivalents per month. If we assume the Fed
maintains this pace, then it will likely purchase about
$14bn 10-year Treasury equivalents in 2012. However,
these purchases will only partially offset the expected
Exhibit 48: Breakevens widened early in the year thanks to a surgein energy prices, but narrowed sharply in the summer as growthprospects waned and risk aversion rose
5-year, 10-year, and 30-year TIPS breakevens (all left axis) versus rolling frontWTI futures contract price (right axis);bp $/bbl
70
80
90
100
110
120
100
150
200
250
300
Nov 10 Jan 11 Feb 11 Apr 11 Jun 11 Jul 11 Sep 11 Nov 11
30Y
10Y
5Y
Oil prices
Exhibit 49: TIPS outperformed nominals on a total return basis in1H11, but sharply underperformed in 3Q11Quarterly total returns on J.P. Morgan TIPS index (JUSTINE) and GBI-US indexby maturity bucket; duration of the indices, and TIPS excess returns*; %
1-10Y 10Y+ 1-10Y 10Y+ 1-10Y 10Y+ 1-10Y 10Y+ 1-10Y 10Y+
1Q11 -0. 1% -0. 9% 3. 9 13. 0 2. 3% 1. 3% 4.7 14. 4 2. 4% 2. 3%
2Q 11 2.3% 3.3% 3.9 13.2 2.9% 5.0% 4.9 14.3 0.0% 1.4%
3Q11 3.8% 23.5% 4.0 14.7 1.8% 11.7% 5.1 14.8 -3.0% -12.0%
QTD 0.1% -0. 3% 4.1 14.8 1.4% 3.8% 5.1 15.2 1.3% 4.0%
Nominal returns Nominal dur'n TIPS returns TIPS dur'n TIPS-nom returns*
* Calculated as TIPS return-(Tsy duration)/(TIPS duration)*(Tsy return).
Exhibit 50: Real yields refreshed their lows in 20115-year and 10-year TIPS real yields; %
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
Nov 10 Feb 11 Jun 11 Sep 11
10Y
5Y
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 195/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Jorge GarayoAC
(44-20) 7325-4820
Kimberly Harano (1-212) 834-4956J.P. Morgan Securities Ltd, J.P. Morgan Securities LLC
195
rise in TIPS duration supply. As Exhibit 52 illustrates,
even with Fed demand, we expect TIPS duration
supply to rise to $140bn in 2012 from $108bn 10-yearTIPS equivalents in 2011.
Although supply should increase in 2012, we expect
end-user demand to rise to meet it. As Exhibit 53
shows, while the average offering size for TIPS has
steadily increased over recent years, so has end-user
demand, with direct and indirect bidders taking down
even greater fractions of (even larger) issues. The effect
of larger auction sizes can also be seen in primary dealer
transaction data: as Exhibit 54 shows, transaction
volumes in TIPS have steadily increased since Treasury
committed to increase auction sizes and improve
liquidity in late 2009. Although increased transactionvolume by itself is not necessarily indicative of improved
liquidity, anecdotal reports suggest that the increased
frequency and size of auctions have improved the
depth of the TIPS market and attracted a broader
base of investors1.
Outside of auctions, we can gauge investor demand for
TIPS based on flows into inflation-protected mutual
funds. In our 2011 Outlook, we noted that mutual fund
flows have historically been correlated to changes in
energy prices, with higher prices leading to greater flows.
In addition, since the Fed began its QE programs, long-
term inflation concerns have increased—a development
1 See “US Treasury’s Makeover on TIPS Ushers in Boom Era; China
Buys,” Min Zeng, Wall Street Journal, 11/15/11
that has been captured in the rise in the 5Yx10Y payer
swaption skew. As a result, we found that the level of theswaption skew has helped explain flows into inflation-
protected mutual funds over the past three years. With
the skew holding fairly constant this year, however,
energy prices were the dominant driver of mutual fund
flows (Exhibit 55). As our oil strategists look for modest
gains in oil prices next year (see below), we expect to
see continued inflows into inflation-protected mutual
funds.
Exhibit 51: We expect gross issuance of TIPS to increasemodestly in 2012…J.P. Morgan forecast for gross issuance; reopenings shaded in grey; $bn of
real principal5yr TIPS 10yr TIPS 30yr TIPS Subtotal
Jan-12 14
Feb-12 10
Mar-12 12 36
Apr-12 16
May-12 12
Jun-12 8 36
Jul-12 14
Aug-12 14
Sep-12 12 40
Oct-12 8
Nov-12 12
Dec-12 14 34
TOTAL 44 76 26 146
Exhibit 52: …and duration supply will rise substantially evenwith Operation TwistGross issuance of TIPS versus duration supply net of Fed purchases*; $bn of real principal in 10-year TIPS equivalents
41
51
77
61
86
68
41
47
54 54
7268
40
50
60
70
80
90
1H10 2H10 1H11 2H11 1H12 2H12
Gross issuance Dur'n supply net of Fed
* Assumes pace of duration buying over October and November 2011 ($2.3bn 10-year TIPS equivalents per month) is maintained.
Exhibit 53: If you sell it, they will comeOffering size for TIPS auctions averaged by calendar year, versus average of sum of direct and indirect bidder percentage*;$bn %
35%
40%
45%
50%
55%
5
6
7
8
9
10
11
2006 2007 2008 2009 2010 2011
Avg auction sizeEnd user %
* 2011 average is over January-November .Source: US Treasury
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 196/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Jorge GarayoAC
(44-20) 7325-4820
Kimberly Harano (1-212) 834-4956J.P. Morgan Securities Ltd, J.P. Morgan Securities LLC
196
In sum, although we look for TIPS supply to increase
next year, we expect demand at auctions and retail
demand to remain robust. Indeed, our J.P. Morgan USFixed Income Investor Survey indicates that 24% of
respondents plan to increase their exposure to TIPS next
year, versus just 7% that plan to reduce exposure (see US
Cross Sector Overview).
The outlook for US inflation
Last year, heading into 2011, we looked for headline and
core inflation to end the year below 1%. Instead, monthly
gains in headline inflation averaged 0.6% over the first
five months of the year, reaching 3.9% (year-over-year
ago) by September. Similarly, core CPI surged to 2.1%
in October from 0.65% in December 2010.
Looking ahead, we expect both headline and core
inflation to moderate. As Exhibit 56 shows, the upside
surprises in core CPI were driven primarily by strong
gains in apparel and vehicle prices as well as some
firming in owners’ equivalent rent (OER). The gains in
the first two categories were driven by temporary
factors—high cotton prices and supply shocks from
Japan—which we do not expect to recur in 2012. In
addition, with incomes constrained by the weak labor
market, gains in OER should remain muted. Thus, we
look for core CPI to decline to 1.0% and headline CPI to
fall to 1.4% by year end. Exhibit 57 shows our forecastfor the monthly changes in headline CPI. We look for
prices to rise modestly over 1H12 and be roughly flat
over 2H12.
Two factors that pose risks to our inflation forecast are
energy prices and the dollar. Our commodity strategists
currently look for modest declines in Brent and WTI oil
prices early in the year followed by increases in 2H12(Exhibit 58). In addition, our currency strategists look
for the dollar to weaken modestly over the next year.
Inflation expectations and dispersion
How does our forecast for inflation compare to investors’
expectations? To answer this question, we examine the
November 2011 results of our J.P. Morgan Inflation
Expectations Survey, which surveys a wide range of
investors across the globe. As Exhibit 59 shows,
Exhibit 54: Transaction volumes in TIPS have steadily increasedsince Treasury committed to improving liquidityRolling 3-month moving average of primary dealer transaction volume in TIPS
(daily average for each week); $bn
4
6
8
10
12
Oct 06 Mar 08 Jul 09 Nov 10 Source: Federal Reserve Bank of New York.
Exhibit 55: Flows into inflation-protected mutual funds weredriven by energy prices this year Monthly flows into inflation-protected mutual funds* versus 1-year percentagechange in oil futures prices, lagged 1 month;$bn %
-5
0
5
10
15
20
25
30
35
40
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
Aug10
Sep10
Oct10
Nov10
Dec10
Jan11
Feb11
Mar 11
Apr 11
May11
Jun11
Jul11
Aug11
Sep11
Mutual fund flows
1Y chg in oil prices
* Includes both weekly and monthly reporters.Source: EPFR Global
Exhibit 56: The upside surprise in core CPI this year was driven bygains in OER, apparel and vehicle prices, which we do not expectto recur in 2012Year-over-year-ago percent change in owners’ equivalent rent, apparel prices,and vehicle prices; %
-4%
-2%
0%
2%
4%
6%
Oct 08 May 09 Nov 09 Jun 10 Dec 10 Jul 11
OER Apparel Vehicles
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 197/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Jorge GarayoAC
(44-20) 7325-4820
Kimberly Harano (1-212) 834-4956J.P. Morgan Securities Ltd, J.P. Morgan Securities LLC
197
investors’ expectations for near-term core inflation are
higher than they were a year ago. Our November 2011
survey shows that over half of investors expect coreinflation to be between 1.5%-2.5% next year, while a
similar number expected inflation to range between 1-2%
in 2011. In addition, the “tails” of the distribution have
shifted: last year, 10% of investors expected core
inflation to be below 0.5% while 8% expected inflation
to exceed 2.5%. The corresponding figures for this year’s
survey are 4% and 15%, respectively, indicating that
investors now appear to be more concerned about upside
risk in inflation than downside risk.
The message is similar for medium-term expectations.
Sixty percent of respondents think headline inflation will
be above target or significantly above target over themedium term, while 8% think it will be below target
(Exhibit 60). In contrast, in last year’s survey 57% of
respondents thought headline inflation would be above
target, while 13% thought it would be below target. It is
interesting to note that the tails of the distribution have
fallen: last year, about 30% of investors expected
inflation to be either below target or significantly above
target, while only 20% of investors held those extreme
views this year.
This decline in the dispersion of medium-term inflation
expectations can be seen in other survey measures. For
example, Blue Chip collects economists’ forecasts for various economic indicators over the long term. As
Exhibit 61 shows, the dispersion of economists’
forecasts for medium-term inflation (over the next five
years) rose sharply around the financial crisis, but has
now declined to levels last seen in early 2008.
The outlook for breakevens
Last year, we modified our long-term model for
breakevens to use an “illiquidity cost metric” as one of
the explanatory variables in the model, instead of trying
to model liquidity-adjusted breakevens. This year, we
make two additional modifications to our model. First,given that the Fed’s quantitative easing programs have
had substantial impacts on inflation expectations and
breakevens, we have added the size of the Fed’s
securities holdings as an additional explanatory variable.
Second, instead of using Nov-27 P-STRIPS asset swap
spreads minus a barbell of 10-year and 30-year Treasury
asset swap spreads as our metric for the cost of
illiquidity, we are now using the spread between 5-year
pre-refunded muni yields (taxable-equivalent) and 5-year
Treasury yields. Pre-refunded munis are fully
collateralized with Treasuries, so they have effectively
the same credit risk as Treasuries, but they trade cheaper
because of lower liquidity. Thus, the pre-refunded
muni/Treasury spread serves as a metric for liquidity.
The other factors in our model are unchanged. We model
3-month forward breakevens to take out the impact
of near-term carry, and we use four additional
Exhibit 57: We look for headline inflation to moderate in 2012J.P. Morgan forecast for monthly % changes in headline CPI and year-over-year ago rate;
% %
0.02
0.200.23
0.260.24
0.16
-0.02
0.09
0.050.08
0.15
-0.06
0.5
1.0
1.5
2.0
2.5
3.0
-0.2
-0.1
0.0
0.1
0.2
0.3
Jan12 Feb12 Mar 12 Apr 12 May12 Jun12 Jul12 Aug12 Sep12 Oct12 Nov12 Dec12
Monthly change
Year-over-year ago rate
Exhibit 58: We expect modest increases in oil prices and aweakening of the dollar over the course of 2012Our forecast for WTI (period average) and the J.P. Morgan USD index (nominalnarrow effective exchange rate)
Current 1Q12 2Q12 3Q12 4Q12
Brent oil futures; $/bbl 107.56 108.00 105.00 115.00 120.00
WTI oil futures; $/bbl 97.41 90.00 90.00 100.00 110.00
JPM Dollar index; level 81.4 82.4 80.5 80.5 80.0
Exhibit 59: Most investors expect higher core inflation than theydid at this time last year Distribution of expectations for US core inflation over the next year from the J.P.
Morgan Inflation Expectations Client Survey as of November 2011 andNovember 2010; %
1 03
5
18
33
24
96
1 1
7
19
31
20
12
4 4
0
10
20
30
40
B e l o w -
0 . 5 %
- 0 . 4
9 -
0 %
0 - 0 .
4 9 %
0 . 5 - 0 .
9 9 %
1 . 0 - 1 .
4 9 %
1 . 5 -
1 . 9 9 %
2 - 2 .
4 9 %
2 . 5 - 2 .
9 9 %
A b o v e
3 %
Nov 11 Nov 10
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 198/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Jorge GarayoAC
(44-20) 7325-4820
Kimberly Harano (1-212) 834-4956J.P. Morgan Securities Ltd, J.P. Morgan Securities LLC
198
factors: 1) the level of nominal yields to account for the
directionality between breakevens and nominal yields; 2)
the unemployment rate, as a measure of slack in theeconomy; 3) the one-year-ahead forecast for the
budget surplus as a percentage of GDP, as a proxy for
fiscal policy-driven inflation pressures, and 4) realized
headline CPI, since inflation expectations tend to be
anchored to realized inflation itself. Exhibit 62 presents
the statistics for our model for 5-, 10-, and 30-year
breakevens.
We can make several observations about breakevens
based on this model. First, as expected, breakevens are
positively correlated with nominal rates, with breakevens
widening as nominal rates rise. Interestingly, long-end
breakevens appear to be most sensitive to nominal rates,widening about 3.6bp for every 10bp rise in long-end
nominal yields, while front-end breakevens are less
sensitive to rates. On the other hand, front-end
breakevens are most sensitive to the current level of the
unemployment rate and realized inflation, as these
factors have a greater impact on near-term inflation
expectations.
Forward-looking budget surplus expectations have a
similar impact across the curve, with every 1% reduction
in the annual deficit (roughly $160bn) causing
breakevens to narrow about 9-10bp. We can use these
betas to estimate the impact of future deficit reduction proposals on breakevens. For example, if market
expectations shift to expect an additional $1tn of cuts
over 10 years (above and beyond baseline expectations),
that would equate to $100bn of cuts per year, or about
0.5% of GDP. Based on the betas in Exhibit 62, these
expected cuts would lower the fair value for breakevens
by about 4.5-5bp.
As for the Fed’s balance sheet, we find that increases
(due to quantitative easing) lead to wider breakevens,
Exhibit 62: Our updated model for breakevens
Statistics for 3-month forward breakevens regressed against nominal yields, the unemployment rate, the 3-month average of 1-year ahead budget surplus expectations asa percentage of GDP, year-over-year ago headline CPI, the size of the Fed’s balance sheet, and our illiquidity cost metric*; monthly data over 12/2002-10/2011
Nominal
yields (% )
Unemployment
rate (%)
Budget surplus as
% of GDP (% )
Headline CPI,
yoy rate (% )
Fed's securities
holdings ($bn)
Illiquidity cost
metric (bp)
Intercept
(bp) R-sq
Beta 20.97 -29.3 -8.7 15.9 0.034 -0.46 237.7 85%
T-stat 4.5 -4.2 -2.4 6.8 4.4 -7.1
Beta 25.67 -21.4 -9.6 11.4 0.029 -0.23 172.7 83%
T-stat 6.1 -5.2 -4.0 7.2 5.6 -5.1
Beta 35.66 -21.6 -8.8 6.8 0.023 -0.16 161.4 85%
T-stat 9.0 -7.1 -4.4 5.0 5.1 -3.830Y
5Y
10Y
* 5-year muni pre-re yield (taxable equivalent) minus 5-year hot-run Treasury yield
Exhibit 60: Fewer investors expect extreme outcomes for medium-term inflation compared to last year…Distribution of expectations for US headline inflation over the next 2-5 years
from the J.P. Morgan Inflation Expectations Client Survey; %
8
33
48
1213
30
41
16
0
10
20
30
40
50
60
below target around target above target s ig. above target
Nov 11 Nov 10
Exhibit 61: …and this decline in the dispersion of inflationexpectations can be seen in other surveys
Average of top 10 responses minus average of bottom 10 responses for long-range consensus projection for year-over-year ago percent change in headlineCPI (average rate for the next five years); %
0.6
0.8
1.0
1.2
1.4
1.6
Mar 03 Jul 04 Dec 05 Apr 07 Sep 08 Jan 10 May 11
Source: Blue Chip Economic Indicators
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 199/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Jorge GarayoAC
(44-20) 7325-4820
Kimberly Harano (1-212) 834-4956J.P. Morgan Securities Ltd, J.P. Morgan Securities LLC
199
with the effect strongest on the front end. Using the betas
from our model, we can also estimate the impact of QE1
and QE2 on breakevens. Given that the Fed’s securityholdings have grown by about $2.14tn since late 2008,
when the Fed first began purchasing MBS, we estimate
that the fair value of TIPS breakevens have risen by
72bp, 62bp, and 49bp, in the 5-, 10-, and 30-year sectors,
respectively. Our model also implies that 5-year
breakevens will benefit the most if the Fed embarks on
QE3, with every $200bn in net purchases increasing the
fair value of breakevens by 6.7bp.
Five-year breakevens also appear likely to benefit the
most if liquidity improves, or underperform the most if
liquidity deteriorates. This suggests that in flight-to-
liquidity periods, the breakeven curve is likely tosteepen, ceteris paribus.
Putting it all together, we project the level of breakevens
in 2012 based on our forecasts for the underlying
variables and our expectation that our illiquidity cost
metric declines to its average level over the past year. We
also assume that the Fed does not expand its balance
sheet via QE3, though there is upside risk to our fair
value projections for breakevens if it does. As Exhibit 63
shows, we expect breakevens to be hit by quite a few
negative factors: the unemployment rate will remain
at elevated levels, budget deficit expectations should
trend lower as fiscal tightening takes hold, and
headline CPI should decline sharply from its current
level. In addition, we look for nominal rates to plunge
in 1Q12 as the European crisis worsens, and then rise
over the rest of the year. Thus, early next year, negative
factors will dominate, and we expect breakevens to
narrow sharply. As the year progresses, however,
breakevens should widen out as rates rise. Thus, based on
this analysis, we recommend underweighting TIPSversus nominals over the next few months. Looking
beyond 1Q12, we would look to initiate breakeven
wideners.
A model for TIPS asset swaps
How are TIPS asset swaps likely to perform in 2012? To
answer this question, we present a simple model for TIPS
asset swaps in Exhibit 64. Since TIPS asset swaps have
no inflation risk due to the embedded inflation swap,
Exhibit 65: Because of seasonality in inflation, the value of theinflation “stub” for TIPS varies over timeValue of the inflation stub at mid-month points over the calendar year for TIPSwith different maturity months; %
For month
starting
Seasonal factor,
with lag applied Jan Feb Apr Jul
15 Jan -0.28% 0 -0.28% -0.25% 0.48%
15 Feb -0.11% 0.27% 0 0.03% 0.76%
15 Mar 0.14% 0.38% 0.10% 0.14% 0.87%
15 Apr 0.26% 0.24% -0.04% 0 0.73%
15 May 0.31% -0.02% -0.30% -0.27% 0.47%
15 Jun 0.16% -0.33% -0.61% -0.58% 0.16%
15 Jul 0.06% -0.49% -0.77% -0.74% 0
15 Aug -0.10% -0.55% -0.82% -0.80% -0.06%
15 Sep -0.16% -0.45% -0.72% -0.70% 0.04%
15 Oct -0.09% -0.29% -0.57% -0.54% 0.19%
15 Nov -0.07% -0.20% -0.48% -0.45% 0.28%
15 Dec -0.13% -0.13% -0.41% -0.38% 0.35%
Inflation stubs for TIPS lines (by maturity month)
Note: We estimate the seasonal factor as the difference between non-seasonally adjustedchanges in headline CPI minus seasonally-adjusted changes, averaged by calendar monthover 2001-10. This series is lagged by two months to account for the lag in TIPS indexationand interpolated to get mid-month values.
Exhibit 63: Our targets for breakevens in 2012J.P. Morgan forecast for the drivers of forward breakevens and projected level of breakevens based on the model in Exhibit 15; bp
Factors Current 1Q12 2Q12 4Q125Y UST y ield, % 0.92 0.75 1.25 1.25
10Y UST y ield, % 2.01 1.70 2.50 2.50
30Y UST y ield, % 3.00 2.70 3.60 3.60
Unemployment rate, % 9.1 9.0 9.0 9.0
Budget surplus as % ge of GDP, % -7.5 -6.4 -6.1 -5.4
Headline CPI, oya % 3.72 1.46 1.10 1.40
Fed's security holdings ($bn) 2625 2625 2625 2625
Illiquidity metric, bp 91.8 80.3 68.7 45.7
Breakeven Targets Current 1Q12 2Q12 4Q12
3Mx5Y breakevens, bp 174.9 120 130 140
3Mx10Y breakevens, bp 205.3 160 175 180
3Mx30Y breakevens, bp 218.4 175 205 205
Exhibit 64: TIPS asset swaps have been negatively correlated withthe level of breakevens10-year TIPS asset swap spread regressed against 10-year TIPS breakevensand Nov-27 P-STRIPS asset swap spread; all data are 1-week averages;weekly data over past three years; bp
-20
0
20
40
60
80
100
0 50 100 150 200 250 30010-year TIPS breakeven; bp
Y = -0.31(10Y BE)+0.46(20Y P-STRIPS) + 79.0R-sq = 77%
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 200/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Jorge GarayoAC
(44-20) 7325-4820
Kimberly Harano (1-212) 834-4956J.P. Morgan Securities Ltd, J.P. Morgan Securities LLC
200
they should trade similarly to asset swaps on other less-
liquid sectors of the Treasury market. Thus, we use 20-
year P-STRIPS asset swaps as one of the explanatoryvariables in our model.
In addition, although TIPS asset swaps should not have
inflation risk, we find that they have been correlated with
TIPS breakevens over recent years, with asset swaps
narrowing (i.e., TIPS richening versus swaps) when
breakevens widen. This is likely because demand for
TIPS—which will impact asset swaps—has been
correlated to inflation expectations and thus breakevens.
Given our expectation that breakevens will narrow over
the next few months and that liquidity will likely remain
constrained, we look for TIPS asset swaps to widenmodestly, though longer-maturity asset swaps will
likely retrace that widening over the course of next
year.
Trading seasonality in TIPS
Since TIPS are indexed to headline CPI, which is not
seasonally adjusted, TIPS are exposed to seasonality.
Furthermore, for a given TIPS issue, the value of
inflation seasonality varies at different points in the year,
even if we assume the seasonal pattern in inflation
remains the same over time. To see this, we consider a
hypothetical TIPS with a January 15 maturity date. If welook at this TIPS on November 15, we can think of
inflation over the remaining life of the bond as n full
years of November 15-November 15 inflation (with a net
seasonality component of zero), plus an inflation “stub”
for a November 15-January 15 period. Similarly, if we
look at the same bond on March 15, we can think of it
accruing n-1 full years of March 15-March 15 inflation
(with a net seasonality component of zero), plus a March
15-January 15 inflation stub.
In Exhibit 65, we show the value of those inflation stubs
for the different TIPS maturity months over the course of
a calendar year. Here we estimate the seasonal factors bytaking the difference between non-seasonally adjusted
changes in headline CPI minus seasonally-adjusted
changes and averaging the differences by calendar month
over 2001-10. We also lag the series by two months to
account for the lag in TIPS indexation, and then we
interpolate to get mid-month values. As the table shows,
some TIPS lines benefit from seasonality more than
others—in particular, the inflation stubs for July TIPS are
almost always positive, while the stubs for February
TIPS are almost always negative.
Although the inflation stubs for individual TIPS vary
over time, the difference between stubs for TIPS with
different maturity months (i.e., January versus April)
Exhibit 66: We estimate that the April-July seasonality differentialis worth about 36bp of yield for 2-year TIPSDifference in value of the inflation stub for various pairs of TIPS (in price terms)
as given in Exhibit 23, and in yield terms for various maturitiesJan-Apr Jan-Jul Apr-Jul
Price impact (% ) 0.25% -0.49% -0.73%
Yield impact, by maturtiy (bp)
2Y 12.2 -23.9 -36.1
5Y 4.8 -9.4 -14.2
10Y 2.5 -4.8 -7.3
20Y 1.3 -2.5 -3.8
Exhibit 67: Projected value of seasonality differentials for TIPSwith different maturities over timeDifference in value of the April-July inflation stub for bonds with different maturityyears; bp of yield
-60
-50
-40
-30
-20
-10
0
May 10 Aug 10 Dec 10 Mar 11 Jun 11 Sep 11
201320152017201920212031
Exhibit 68: The flattening of the Apr-13/Jul-13 TIPS yield curve hasbeen roughly in line with what we would expect given the changein the value of the seasonality differential
Apr-13/Jul-13 TIPS real yield curve versus projected value of the April-Julyseasonality differential; bp of yield
-55
-50
-45
-40
-35
-30
-25
-20
-30
-25
-20
-15
-10
-5
0
5
10
15
May 10 Aug 10 Nov 10 Mar 11 Jun 11 Sep 11
Apr-13/Jul-13 TIPS curve
Value of seasonality differential
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 201/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Jorge GarayoAC
(44-20) 7325-4820
Kimberly Harano (1-212) 834-4956J.P. Morgan Securities Ltd, J.P. Morgan Securities LLC
201
stays constant, assuming that the seasonal pattern is
unchanged. Thus, we can estimate the difference in the
value of seasonality for various pairs of TIPS, as shown
in Exhibit 66. As the table shows, April TIPS are most
adversely affected by seasonality, accruing 0.25%-pt less
inflation than January TIPS, and 0.73%-pt less inflation
than July TIPS. We can also translate this seasonality
differential into yield terms by dividing by duration. As
the table shows, the April-July seasonality differential is
worth 36bp of yield for 2-year TIPS, but only about 4bp
of yield for 20-year TIPS.
To see how the value of this seasonality differential (in
yield terms) changes as bonds age, we project the
differential for April and July TIPS with various maturity
years in Exhibit 67. For longer-maturity TIPS, the
change in the value of the seasonality differential over
time is small: for example, for TIPS maturing in 2019,
the differential was worth 8.4bp of yield 18 months ago,
and is currently worth 9.5bp. In contrast, for TIPS
maturing in 2013, the differential was worth about 25bp
of yield 18 months ago, and is currently worth 51bp. We
can see the impact of this seasonality effect on the Apr-
13/Jul-13 TIPS real yield curve—as Exhibit 68 shows,the flattening of this curve over the past 18 months has
been roughly in line with what we would expect given
the change in the value of the seasonality differential.
In practice, it may be difficult to isolate the value of
seasonal differentials, since the spread between two
issues will also be impacted by the slope of the overall
curve as well as liquidity differentials. For example, even
though April TIPS have a seasonal disadvantage
compared to July TIPS, in the 2016 sector, the April
TIPS is the benchmark 5-year and enjoys greater
liquidity, while the Jul-16 TIPS is an old 10-year issue
that is less liquid. As a result, the fact that the spread between these two issues is almost zero likely indicates
that the liquidity premium for Apr-16 TIPS counters the
seasonality disadvantage. In addition, at the very front
end of the TIPS curve, investors’ near-term inflation
forecasts are likely to have a greater influence on spreads
than backward-looking estimates of seasonality.
With these constraints in mind, we think it is still
possible to find instances where investors can take
advantage of seasonal differentials. For example, we
estimate that Jul-16 TIPS should trade about 11bp rich to
Jan-16 TIPS based on seasonals, but they are currently
about 1bp cheap. In addition, this curve currently lookstoo steep relative to the broader Jan-15/Jan-19 curve, and
liquidity differentials should be a minor issue, since both
Exhibit 69: TIPS breakevens have closely tracked equities over thepast two years10-year TIPS breakevens versus S&P 500;
bp level
1000
1100
1200
1300
1400
140
160
180
200
220
240
260
280
N ov 09 Feb 10 J un 10 Sep 10 Dec 10 Mar 11 J ul 11 Oc t 11
10-year TIPS breakevens
S&P 500
Exhibit 70: Going short breakevens versus the S&P 500 whenbreakevens look wide has been more profitable than going longbreakevens when breakevens look narrowStatistics for long or short breakeven trades initiated when the residual* reached acertain level and hedged with the opposite position in S&P 500**; 11/09-10/11
Long BE if
residual (bp) <
#
trades
Avg P/L
(bp)
#
winners
Hit
ratio
Avg gain
(bp)
Avg loss
(bp)
-10 125 -0.6 64 51% 9.5 -11.2
-15 85 -0.5 45 53% 8.7 -10.8
-20 421.9
2662%
8.8 -9.4
-25 11 6.4 10 91% 8.3 -11.8
Short BE if
residual (bp) >
#
trades
Avg P/L
(bp)
#
winners
Hit
ratio
Avg gain
(bp)
Avg loss
(bp)
5 130 6.8 82 63% 15.8 -8.6
10 70 9.3 49 70% 17.3 -9.4
15 32 16.3 27 84% 20.2 -4.4
20 10 18.0 10 100% 18.0 N/A
* Backward-looking 3-month residual of the level of breakevens versus the level of S&P 500.** The hedge ratio is given by the backward-looking 3-month beta of weekly changes inbreakevens versus weekly changes in the S&P 500.Note: P/L calculated as change in breakevens minus (hedge ratio)*(change in S&P) plus 1-month ex-ante breakeven carry for long breakeven positions, and the inverse of that quantityfor short breakeven positions.
Exhibit 71: Going long breakevens when carry is attractive andbreakevens look narrow has been profitable…Statistics for long 5-year breakeven trades (held for one month) initiated when ex-ante 1-month carry was above a certain level and the rolling 6-month residual of 5-year breakevens regressed against 5-year nominal yields and WTI oil futuresprices was below a certain level; 1/09-10/11
Carry
(bp) >
Residual
(bp) <
#
trades
Avg carry
(bp)
Avg P/L
(bp)
#
winners
Hit
ratio
Avg
gain (bp)
Avg loss
(bp)
3 0 88 8.0 6.9 56 64% 19.1 -14.4
3 -2 77 7.7 6.5 48 62% 19.4 -14.9
5 0 56 10.4 6.2 34 61% 17.7 -11.6
5 -2 47 10.3 5.6 28 60% 17.4 -11.9
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 202/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Jorge GarayoAC
(44-20) 7325-4820
Kimberly Harano (1-212) 834-4956J.P. Morgan Securities Ltd, J.P. Morgan Securities LLC
202
issues are old 10-year TIPS. Thus, we think
overweighting Jul-16 TIPS versus Jan-16 TIPS is
currently an attractive way to take advantage of seasonality differentials. Such trades, based on the theme
of identifying seasonality mispricings, will be a key
element of our TIPS strategy in 2012.
Relative value trading themes for 2012
Trading breakevens versus the S&P 500
Although we expect breakevens to narrow in early 2012
as the European crisis worsens, one way we can protect
against being wrong is to trade breakevens on a hedged
basis—namely, by trading breakevens versus equities.
Over recent years, the correlation between breakevens
and risky assets has been striking, with breakevenswidening in “risk-on” trades and narrowing in “risk-off”
trades. In particular, breakevens have closely tracked the
S&P 500 over the past two years (Exhibit 69). Given this
strong correlation, we would expect trading breakevens
versus a beta-weighted amount of equities when
breakevens look either very narrow or very wide relative
to equities to be an attractive trading theme.
We test this hypothesis by back-testing trades over the
past two years as follows: each day, we calculate the
backward-looking 3-month residual of the level of
breakevens versus the level of S&P 500. If the residual is
below (above) a certain threshold, we initiate a long(short) breakeven position and hedge with the opposite
position in the S&P 500. The hedge ratio is given by the
backward-looking 3-month beta of weekly changes in
breakevens versus weekly changes in the S&P 500.
Exhibit 70 presents the results of our analysis. We would
expect the strategy to work equally well when
breakevens look either too wide or too narrow, but we
find that over this period, going long breakevens hedged
for the S&P when breakevens look narrow has generally
been unattractive, unless breakevens look extremely
narrow. In contrast, going short breakevens on a hedged
basis when breakevens look too wide has been
profitable—these trades have had high hit ratios andattractive P/Ls. This asymmetry could be a result of our
observation that equities and inflation expectations have
been more correlated during deflationary periods than
inflationary periods in recent years (see US Cross Sector
Overview).
Thus, one trading theme we like for 2012 is
opportunistically selling breakevens versus a hedged
amount of S&P 500 futures when breakevens look too
rich versus equities.
Earning carry when breakevens look narrow
A second trading theme we like is opportunistically
going long breakevens when carry is attractive and
breakevens look narrow relative to fair value. To see this,
we back-tested a strategy under which long breakeven
trades were initiated for a 1-month period when 1) 1-
month carry exceeded a certain level and 2) the
backward-looking 6-month residual of breakevens
regressed against nominal yields and oil prices fell below
a certain threshold. In Exhibit 71, we present statistics
for these trades over January 2009-October 2011. As the
exhibit shows, this strategy generated hit ratios around
60% and average 1-month P/L of 5-7bp.
We found that this strategy was even more effective
using inflation swaps—specifically, receiving
inflation/paying fixed when carry was attractive and
inflation swap rates looked too low relative to fair value.
As Exhibit 72 shows, using inflation swaps produced
significantly higher hit ratios and average P/Ls than the
strategy using breakevens. Even if we account for the
wider bid-offer in inflation swaps, the derivatives version
of this strategy still appears more attractive than the cash
version.
Thus, a second trading theme we like is receiving
inflation/paying fixed in swaps when carry is
attractive and inflation swap rates look too low.
Exhibit 72: …and this strategy works even better with inflationswapsStatistics for receiving inflation/paying fixed positions in 5-year or 3-year inflation
swaps (held for one month) when ex-ante 1-month carry was above a certain leveland the rolling 6-month residual of 5-year or 3 -year inflation swaps regressedagainst 5-year or 3-year nominal yields and WTI oil futures prices was below acertain level; 1/09-10/11
Carry
(bp) >
Residual
(bp) <
#
trades
Avg carry
(bp)
Avg P/L
(bp)
#
winners
Hit
ratio
Avg
gain (bp)
Avg loss
(bp)
5-year swaps
3 -6 63 7.6 15.2 54 86% 19.3 -9.4
3 -14 25 7.2 15.7 22 88% 19.9 -14.7
5 -6 45 9.1 12.6 38 84% 16.4 -7.8
7 -6 29 10.9 10.3 23 79% 15.4 -9.1
3-year swaps
7 -6 42 16.5 13.0 32 76% 19.5 -8.1
7 -10 25 17.2 14.8 20 80% 20.4 -7.8
9 -8 27 18.3 14.7 22 81% 20.1 -8.9
11 -8 20 21.2 12.7 15 75% 19.9 -8.9
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 203/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Jorge GarayoAC
(44-20) 7325-4820
[email protected]. Morgan Securities Ltd
203
Medium term themes for inflation
markets
Euro area
• Favour breakeven flattening trades
Even if inflation should fall, we see upside risks to
prints in 1H12 vs. expectations, while the long end
likely to be driven by sovereign risk dynamics which
should flatten nominal yield curves. Favour 1s/10s
flatteners in HICP swaps, overweight OBLi 13 vs.
BTPei 19 breakevens or BTPei 16s vs. BTPei 21
breakevens.
• Overweight French CPI-linked vs. Euro HICP-
linked lines into early 2012French-CPI linked should perform well in 1Q12 when
we expect further hedging from Livret A accounts and
support from Libor-based domestic investors.
• Italian real yields to be under pressure
Real yields can reach 10% on the back of our
peripheral spread views as well as risks that Italian
linkers are dropped from linker benchmarks.
UK
• Be long real yields in intermediate part of the curve
Our model predicts 5Y real yields to fall 30-40bp fromcurrent levels. Long positions in 5Y and 10Y linkers
offer attractive carry and slide vs. the forwards.
• Short 10Y breakevens
Nominal gilts will continue to be the major driver of
breakevens. Our predicted 75bp fall in 10Y gilt yields
during 1H12 (to the 1.50%) level translates into a
30bp fall in ILG22 breakevens, to the 2.30% level.
• Go short real yields into linker syndications as a
tactical trading strategy
A trading rule selling the closest maturity bond to the
linker being issued via syndicate 15 business daysahead of syndications has been profitable.
US
• Look for TIPS breakevens to narrow sharply over
the next few months and then widen modestly over
2012
Breakevens will hit a number of headwinds in early
2012: we expect nominal yields to plummet, fiscal
policy to tighten, and headline inflation to fall. Thus,
our model projects that breakevens should narrow
over the next few months, though breakevens should
widen over the remainder of the year as nominal rates
rise. QE3 also poses an upside risk to our targets.
• Overweight TIPS with attractive seasonals versus
TIPS with unattractive seasonals when mispricings
existWe recommend taking advantage of seasonality
differentials among TIPS that are close in maturity
when mispricings exist.
• Opportunistically trade breakevens versus the S&P
500 when breakevens look too wide
With correlations expected to remain high as long as
European risks remain significant, trading breakevens
hedged for moves in the S&P 500 is likely to prove
attractive, as it has in recent years.
• Look to tactically trade breakevens and inflation
swaps for carryWe find that strategies involving receiving inflation
when carry is attractive and swap rates look low have
been profitable over recent years and recommend that
investors use these strategies to earn carry.
Cross market themes
• Generally overweight US inflation breakevens vs.
Euro area equivalents in 10Y+, vs. doing the
opposite at the front end
This is based on our relative inflation views, the risk
of sovereign dynamics impacting Euro breakevens
more than US equivalents.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 204/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Sally AuldAC
(61-2) 9220-7816
[email protected]. Morgan Securities Australia Limited
204
Australia
• J.P. Morgan economists expect the Australian
economy to grow 3.0% in 2012, but risks are
biased to lower growth outcomes given the fragile
global environment. Against this backdrop, J.P.
Morgan economists expect the RBA to deliver
50bps of easing in Q1 next year
• While the domestic economic outlook might
suggest the potential for modestly higher yields,
the experience of 2011 suggests that domestic
drivers may not be the main influence on term
yields
• As such, the bearish outlook we retain on Europe
means that we recommend extending duration on
dips in AUD fixed income. Look to lighten
duration in 2H12, given the risks that the market
prices in a more optimistic global outlook
• We target the 3s/10s bond curve spread to make
new highs in 2012. Investors trading long
duration should implement cash bond curve
flatteners in order to gain some portfolio
protection
• We think swap spreads will be biased wider in
2012, with the spread curve likely to flatten. We
like owning cash spread product against swap
rather than bond, and prefer NSWTC in the
semi-government space
• Look to buy the NSWTC Nov-2035i against the
ACGB Sep-2030i. The ACGB Aug-2020i looks
cheap against the UST Jul-2021i; real yield
spreads look wide given AUS-US growth
differentials
Economy expected to improve in 2012,
but global headwinds are fierce
2011 was a year of disappointment on the
macroeconomic front, at least relative to expectations.
Consensus forecasts now expect that Australia will likely
record growth of 1.5-2.0% in 2011, well below trend.
This was a lower number than most expected at the
beginning of the year, but the disappointment was largely
due to the impact of severe floods in Queensland and a
slower-than-anticipated recovery from the natural
disaster.
The outlook for Australia in 2012 is more optimistic,
with the ongoing recovery in coal exports expected to see
solid growth momentum in 1H12. J.P. Morgan expects
growth of 3.0% for the calendar year; the RBA expects
4.0% growth, although we would argue that this looks a
little ambitious given risks surrounding the European
(and global) outlook. While Australia’s direct trading
links with Europe are small (around 8% of Australianexports go to Europe), the indirect links via Asia are
reasonably strong. As the RBA noted in its most recent
quarterly Statement on Monetary Policy:
“The possibility of a sharp economic deterioration in
Europe represents a downside risk for the Australian
economy. Given strong trade links with Asia, it is likely
that Australia would be less directly affected than some
other countries by a deterioration in Europe, although
the economy would still be affected through falls in asset
Exhibit 1: Australia’s (export weighted) trading partner growth –forecast to be below trend in 2011 and 2012Calendar year % growth
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
1995 1998 2001 2004 2007 2010
Average
Exhibit 2: Movements in the RBA cash rate generally correlate well withglobal GDP growth%-pts Qtrly. annualised %
-5.5
-4.5-3.5
-2.5
-1.5
-0.5
0.5
1.5
2.5
Mar-98 Dec-00 Sep-03 Jun-06 Mar-09 Dec-11
-10.0
-8.0
-6.0
-4.0
-2.0
0.0
2.0
4.0
6.0
8.0
RBA cash rate, annual
change, lhs
Global GDP, rhs, pushed
fwd 2q
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 205/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Sally AuldAC
(61-2) 9220-7816
[email protected]. Morgan Securities Australia Limited
205
prices and weaker household and business confidence.
Commodity markets could be expected to weaken and
growth in domestic incomes would be lower .”
The clear message is that Australia will not be immune to
a severe economic contraction in Europe. Against this
backdrop, J.P. Morgan economists are forecasting the
RBA to deliver modest easing in early 2012 and expect a
cash rate of 3.75% by March next year. The risks to this
view are clearly biased to the downside, given fragility in
the global growth environment and risks around an
escalation in the European crisis. Indeed, our estimate of
Australia’s trading partner growth in 2012 suggests little
change from 2011 (Exhibit 1), with trading partner
growth at 4.2% (below the average of 4.9%). It is also
worth noting that movements in the RBA cash rategenerally correlate well with the global growth outlook
(Exhibit 2).
Of course, the main medium-term influence on the
Australian macro-economic outlook continues to be
China, and while our economists have recently shaved
their growth forecasts for China, they still expect a
reasonably positive outcome in 2012 (growth of 8.3%).
As Exhibit 3 illustrates, Australian and Chinese GDPs
have been well correlated over the past 5-6 years, with
growth in China leading Australian growth by around 3
months. Although our expectation for Chinese economic
growth would be historically consistent, with Australiangrowth around 2.0%, the consensus expects strong capex
growth and a rebound in flood-affected exports to
temporarily drive a divergence between Australian and
Chinese growth outcomes (hence the disconnect between
Australian and Chinese growth forecasts in Exhibit 3).
Overall, our forecasts for Australian growth next year
represent a cautiously optimistic outlook, especially
given global risks and our forecasts for another year of
below-trend trading partner growth. Accordingly, we
believe risks are biased to the downside to our growth
forecasts. This is especially the case given a reasonably
subdued pulse in the domestic dataflow of late.
Can the RBA validate front-end
pricing…another 165bp of rate cuts?
As noted above, J.P. Morgan economists expect the cash
rate to reach 3.75% by March 2012. We still think the
risks to this view are to the downside, given the situation
offshore. And while the RBA may not yet believe it has
begun a traditional easing cycle, we would expect the
RBA’s policy stance to retain a degree of optionality in
the months ahead. But the market is already pricing in
this risk; Exhibit 4 illustrates that the OIS curve is
pricing in 165bp of rate cuts over the next 12 months.
Exhibit 5 illustrates that since early August, the markethas oscillated between pricing in 100bp and 165bp of
rate cuts from the RBA (over a 12-month horizon). Thisis a distinct change from the period January–August2011, when the market priced anywhere from 0-50bp of rate hikes.
Investors may be tempted by the attractive carry in paid
positions in the very front end of the Australian curve;
for example, a paid position in AUD 1Y swap currently
offers 30bp of carry and roll over a 3-month horizon. But
assuming the RBA does not validate current market
pricing anytime soon and with no ‘end-game’ solution in
Exhibit 3: Chinese and Australian GDP growth have been wellcorrelated over the past 6 yearsDotted lines represent J.P. Morgan forecasts
oya % oya %
6.0
8.0
10.0
12.0
14.0
16.0
Dec-04 Apr-06 Aug-07 Dec-08 Apr-10 Aug-11 Dec-12
0.5
1.6
2.7
3.8
4.9
6.0Chinese GDP, lhs, pushed fw d 1q
Australian GDP, rhs
Exhibit 4: The market is pricing in over 150bp of rate cuts from theRBA…
AUD OIS curve, %
2.50
2.75
3.00
3.25
3.50
3.75
4.00
4.25
4.50
4.75
Current Feb-12 Apr-12 Jun-12 Aug-12 Oct-12
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 206/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Sally AuldAC
(61-2) 9220-7816
[email protected]. Morgan Securities Australia Limited
206
Europe on the horizon, we see little reason for current
market pricing to change materially into 2012 given:
1) Ongoing concerns surrounding Europe, China
and global growth;
2) Data suggesting that while the Australian
economy is not on the verge of a recession, nor
is it showing any sign of a significant uptick in
momentum;
3) Technical aspects to the OIS market which
currently bias yields lower (domestic banks are
structural receivers of OIS given their need to
hedge deposit books); and
4) The fact that the RBA could – if needed – cut
rates by a lot, relative to other developed market
central banks.
Interestingly, the cycle trough in 3Y yields has often
been a good indicator of the subsequent trough in the
cash rate (2001 and 2008). Exhibit 6 illustrates that with
3Y government bonds currently yielding 3.25%, this
would be consistent with a further 125bp of easing from
the RBA. It also implies that there is probably 40–50bp
of risk premium in the very front end of the AUS yield
curve at present.
The outlook for outright yields – lower-
yield highs and lower-yield lows in 2012
as domestic fundamentals continue to
take a back seat
Broadly speaking, the trend over the course of 2011 was
towards lower yields in Australia, with a rally in global
bond markets, some disappointment on the domestic and
global growth fronts, and a November rate cut from the
RBA being the main drivers of lower yields. A persistent
bid from offshore investors also helped to suppress yields
(see our analysis below for more detail on this theme).
Australian 3Y bond yields have traded a 215bp range
over the course of 2011, and at 3.23% at the time of
writing, sit at the lows of the year (Exhibit 7).
Remarkably, the calendar year yield high for 3Y bonds
has been very similar over the past 3 years, at around
5.35%. Exhibit 7 illustrates the structurally lower trading
range for 3Y yields post the financial crisis; with
offshore risks mounting, we think it is likely that
Australian yields post new lows in 2012.
Exhibit 5: …and has been pricing in cuts since early Augustbp of tightening* priced in over the next 12 months
-175
-125
-75
-25
25
75
Jan-11 Apr-11 Jul-11 Oct-11
* negative number implies easing
Exhibit 6: Even term yields suggest the possibility of more easing – thelevel of 3-year bond yields suggest the risk of rate cuts, if history is anyguide%
2.0
3.0
4.0
5.0
6.0
7.0
8.0
Jan-00 Feb-02 Mar-04 Apr -06 May-08 Jun-10
3-year bond yield
RBA cash rate
Exhibit 7: 3-year government bond yield calendar year trading rangeshighlight a structurally lower trading range post crisis and the risk of new yield lows in 2012Trading range for 3Y government bond yields over the course of 2011; %
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.06.5
7.0
7.5
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 207/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Sally AuldAC
(61-2) 9220-7816
[email protected]. Morgan Securities Australia Limited
207
Even the 10Y bond has traded within a 175bp range over
the course of the year, with a yield high for the year of
5.75% also close to the yield highs seen in 2009 and2010 (Exhibit 8).
With outright yields at the lows for the year (and not far
from the yield-lows seen since 2009) and with J.P.
Morgan economists expecting a pick-up in growth
momentum next year, it might be intuitive that short-
duration strategies are the correct view on Australian
bonds in 2012.
However, the message from 2011 has been that
domestic fundamentals have not been the main driver
of yields over the course of this year – indeed, Exhibit
9 illustrates the divergence between domesticfundamentals and the level of 3Y bond yields. Against
this backdrop, we think there are a number of reasons
why investors are not likely to be rewarded for short-
duration strategies in the coming year:
1) Our forecast for US 10Y yields does not
envisage a significant selloff, at least for the
first half of calendar year 2012. After an early
rally to 1.7%, we expect the UST 10Y yield to
oscillate around the 2.5% level for much of the
year, limiting the extent to which Australian
bond yields are likely to rise.
2) Domestic real money accounts are already
running short-duration positions and have done
so for much of 2011 (Exhibit 10), and as such,
we believe there will be little appetite to shorten
the duration of portfolios further from current
levels.
3) Our European rates strategy team remains
extremely cautious on the outlook for the
European peripherals in 2012. We think
developed market bond yields will struggle to
selloff significantly in such an environment.
Furthermore, this backdrop will reinforce the
attractive risk/reward in owning AAA-rated
Australian sovereign risk at (relatively)
attractive yield levels.
4) Following on from point 3, we expect offshore
demand for Australian government bonds to
remain solid in 2012. As we note below, 71% of
ACGBs are owned by foreign investors (as of
June 2011).
Exhibit 8: The 10Y bond has traded within a 175bp range over thecourse of the year, with a yield high for the year of 5.75% also close tothe yield highs seen in 2009 and 2010Trading range for 10Y government bond yields over the course of 2011; %
3.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
Exhibit 9: Australian bond yields have not traded in line with domesticfundamentals in 2011…% oya %
2.0
3.0
4.0
5.0
6.0
7.0
Mar-99 Dec-01 Sep-04 Jun-07 Mar-10
0.0
3.0
6.0
9.0
3y bond yield, lhs
Real GDP + CPI (2q ave. ), lagged
2 qtrs, rhs
Exhibit 10: …but this has not prevented domestic real money managersfrom trading short duration for most of 2011
Actual portfolio duration relative to benchmark index, %
86
90
94
98
102
106
110
Nov -06 Dec-07 Jan-09 Feb-10 Mar-11
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 208/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Sally AuldAC
(61-2) 9220-7816
[email protected]. Morgan Securities Australia Limited
208
Of course, our fair value models for both 3Y and 10Y
government bonds continue to suggest that current yield
levels for AGCBs should be regarded as rich. Exhibit 11 shows our fair value for 10Y bond yields. For 2012, we
see fair value around 4.75%.
But with higher-than-usual levels of uncertainty
persisting offshore and ongoing demand for yields from
foreigners, the valuation richness of the ACGB market
may be a constant theme in 2012. In 2011, investors have
not been paid to position portfolios on domestic
fundamentals alone, and we suspect this will continue to
be the case in 2012 given our global strategy views.
Indeed, movement in a weighted European peripheral
10Y swap spread has accounted for almost 86% of the
moves in AUD 1y/1y swap over the course of 2011.
Our preferred duration strategy for 2012 would be to
use any selloff in bond markets to enter long-duration
portfolio structures. Indeed, we expect the next couple
of months to provide good opportunities for investors to
initiate long-duration positions. Short-risk trades are
likely to be unwound going into the year-end as money
managers, hedge funds, and dealers alike pare back their
balance sheet. Towards 2H12, we would look to lighten
duration given the risk that the market could start to price
in a more optimistic global outlook.
Against the US, we see Australian 10Y bonds as cheap at present. Our medium-term model of the AUS-US 10Y
bond spread suggests that the fair value for the spread is
probably around the 175bp level for 2012 (Exhibit 12).
The spread is currently trading at 205bp, and we suspect
that AUD 10Y bonds will be well bid on any widening of
this spread. Increasingly, relative (and not absolute) yield
levels are becoming the driver of cross-border bond
market flows.
Offshore investment becomes a key
feature of Australian fixed income
markets and is here to stay
One of the features of the Australian government bond
market in recent years has been strong demand for
ACGBs by offshore investors. Indeed, foreign buyers
have accounted for much of the additional supply in
ACGBs (Exhibit 13). Offshore holders now own 71% of
ACGBs, just shy of all-time highs (Exhibit 14). The
solid demand for ACGBs in recent years reflects a
number of factors:
Exhibit 11: Our fair value model suggests current 10Y governmentbond yields are rich. For 2012 we see fair value around 4.75%10Y government bond yield, actual and model*; %
2.0
4.0
6.0
8.0
10.0
12.0
14.0
16.0
Sep-88 Mar-93 Sep-97 Mar-02 Sep-06 Mar-11
Actual
Model
+/- 2 s.d.
* model is 10yr yield = 2.52 + 0.60*90-day rate + 0.31*moving avg. of nominal GDP –4.49*% of ACGBs held offshore; R-squared is 0.92, standard error is 0.72ppts.
Exhibit 12: Australian 10Y bonds still look cheap vs. US 10Y notes onour medium-term modelQuarterly data for spread between 10Y Australian government bonds and USTreasuries, actual and model*; bp
0
50
100
150
200
250
300
Jun-96 Apr-99 Feb-02 Dec-04 Oct-07 Aug-10
Actual
Model
* model is 10yr spread = 58.13 + 5.32*real GDP growth spread (pushed fwd. 4q) +30.98*1y swap yield differential (%-pts); R-squared is 0.83, standard error is 0.29pptsLatest read for the actual spread is 30-Sep-11.
Exhibit 13: Foreign buying of ACGBs has accounted for much of theadditional supply in recent years
AUDbn
0
50
100
150
200
Dec-88 Jun-92 Dec-95 Jun-99 Dec-02 Jun-06 Dec-09
ACGBs (offshore holdings)
Total ACGBs outstanding
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 209/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Sally AuldAC
(61-2) 9220-7816
[email protected]. Morgan Securities Australia Limited
209
1) Reserve diversification by central banks into
AUD;
2) A largely non-leveraged investor base with
longer investment horizons (central banks and
offshore pension funds) and a small level of
holdings in absolute terms pre-crisis ($34.4bn as
of December 2007);
3) Wide yield differentials with most developed
market economies and outperformance by the
Australian economy (suggesting potential for
AUD appreciation); and
4) A strong fiscal situation in Australia.
When we look at the relative rankings of AAA-rated
sovereigns1, Australia should definitely be considered a
‘strong’ AAA-rated sovereign. And when one considers
the additional yield pick-up, it is not difficult to see why
offshore capital has found a home amongst ACGBs
(Exhibit 15).
As we have noted above, ongoing foreign demand for
ACGBs is one reason we do not expect a significant sell-
off in AUD rates. Indeed, we can estimate the extent to
which changes in offshore participation can impact the
level of bond yields. As we note in the footnote to
Exhibit 11, we include the proportion of ACGBs held byoffshore investors as an explanatory variable in our
medium-term, fair-value model for Australian 10Y bond
yields.
Our other inputs are a moving average of nominal GDP
growth (to capture the influence of domestic economic
fundamentals on long-end yields) and the 90-day rate (to
capture the influence of short-term interest rates on the
level of term yields). The model statistics are outlined in
Exhibit 16.
The model suggests that a 1-ppt rise in offshore
ownership will see 10Y yields decline 4.5bp. This
suggests that the rise in offshore holdings though 2010
and 2011 has lowered 10Y yields by around 34bp, all
other things being equal.
We think that much of rise in offshore demand for
ACGBs has been driven by central bank reserve
diversification. This would be consistent with recent data
1 For more details on the analysis behind the ranking of AAA-rated
sovereigns see S. Auld, The Antipodean Strategist , 18 August 2011
from the IMF’s composition of foreign exchange
reserves, which suggests that non-G4 currencies now
account for 5% of reserves, up from just over 2% in
Exhibit 14: Offshore holders now own 71% of ACGBs, just shy of all-time highsProportion of ACGBs held by offshore investors; %
0
10
20
30
40
50
60
70
80
Sep-88 Jun-93 Mar-98 Dec-02 Sep-07
Exhibit 15: Australia is one of the stronger AAA-rated sovereigns, yet ithas the highest 10Y bond yields among all AAA-rated peers*10-year sovereign bond yield vs. standard deviations away from AAA median**
0.51.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
-0.6 -0.3 0.0 0.3 0.6 0.9 1.2
Std. Devs . aw ay from AAA median
CHF
AUSNZ
UKGER
Stronger AAAWeaker AAA
* we include New Zealand even though Moody’s is the only rating agency to give thesovereign an AAA credit rating.** based on a model based on real GDP growth and 1Y swap y ield differentials.
Exhibit 16: Our models for 10Y yields suggest a 1% increase in offshoreownership will result in yields declining 4.5bpModel statistics for our 10Y ACGB model
Model Statistics Value
Intercept 2.5
Nominal GDP (moving ave.) 0.6
90-day rate 0.3
% of offshore holdings -4.5
R-squared 0.92
Standard error (ppts) 0.7
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 210/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Sally AuldAC
(61-2) 9220-7816
[email protected]. Morgan Securities Australia Limited
210
2009. And even with much of the new allocation to AUD
now out of the way, quarterly net buying of ACGBs still
remains well above average.
But the other driver of offshore demand for ACGBs
continues to be the wide yield differential – both in
nominal and real terms – between Australian government
bonds and bonds in most developed market economies.
Exhibit 17 shows that the 10Y real yield spread between
Australia and the US has a leading relationship with
changes in offshore holdings of ACGBs. If anything, the
chart suggests that current levels of foreign demand look
a touch low relative to rate spreads. And with the outlook
for G4 government bond markets consistent with
relatively low yields for a long time, we expect the ‘real
money carry trade’ to support foreign sponsorship of theACGB market.
Curve spreads: the 3s/10s spread has
scope to steepen further
Generally speaking, the shape of the yield curve over
time in Australia is determined by movements in the
RBA’s cash rate. Exhibit 18 illustrates the point –
investors who think the RBA will validate market pricing
should clearly position for a steeper curve spread.
Indeed, our fair value model for the 3s/10s bond spread
suggests that the curve has traded flat relative to fair value for some time (Exhibit 19). There are possibly two
reasons for the persistent cheapness of the curve – it
could be a reflection of strong demand for AUD duration
from offshore (as discussed above), or the negative carry
inherent in curve steepeners may prevent investors from
positioning for a steeper curve.
Given our expectation of rate cuts from the RBA, we
believe the 3s/10s bond spread will make new highs for
the cycle in Q1 next year and target a move to +105bps.
The carry and roll associated with cash bond steepeners
is very negative. Accordingly, we think there is better
value in entering forward starting steepeners; forward
curve spreads are very flat, suggesting that investors
biased towards steeper curves should easily beat the
forwards.
The least carry and roll negative forward starting
steepener which offers some roll-up to spot is the 3s/7s
swap curve steepener, 1Y forward. The 1Y forward
spread is around 4bps flatter than the spot spread, and the
carry and roll on the trade costs 4.8bps over a 3-month
Exhibit 17: Real yield spreads are a key driver of offshore demand for ACGBs%-pts %-pts
-15
-10
-5
0
5
10
15
20
25
30
Sep-97 Aug-00 J ul-03 J un-06 May -09
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0 Annual change in offshore holdings of ACGBs, lhs
10y AUS-US real yield spread, rhs, 6 mths fw d
Exhibit 18: The policy cycle tends to be the medium term driver of movements in the 3s/10s bond curve in Australia%-pts %
-0.8
-0.4
0.0
0.4
0.8
1.2
1.6
Aug-98 J un-01 Apr-04 Feb-07 Dec -09
2.00
3.00
4.00
5.00
6.00
7.00
8.00
AUS 3s10s curve, pushed fwd. 3 months, lhs
RBA cash rate, rhs, inv erted
Exhibit 19: The 3s10s bond curve still looks a touch flat relative to our fair value model3s/10s bond curve actual and model*; %-pts
-1.25
-0.75
-0.25
0.25
0.75
1.25
1.75
Oct-01 Apr-04 Oct-06 Apr-09 Oct-11
Actual
Model
* model is 3s/10s bond spread = 2.20 + 0.07*1s6s AUD OIS curve (%-pts) - 0.41*3ybond yield + 0.18*US2s10s bond spread; R-squared is 0.88, standard error is0.15ppts.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 211/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Sally AuldAC
(61-2) 9220-7816
[email protected]. Morgan Securities Australia Limited
211
horizon. The 1Y forward spread is currently 66bps; we
target a move to +90bps.
Investors already trading long duration
should look to use cash bond flatteners
as a means of portfolio hedging
Given the prospect of difficult trading conditions in
2012, we think it is wise to think about trades which are
1) positive carry, and 2) not so correlated with moves in
European peripheral spreads. In the Australian market,
carry-efficient curve trades offer good opportunities for
investors trading long duration looking for some
portfolio diversification or protection in the event that
peripheral spreads narrow. Exhibit 20 illustrates that the
3s/10s curve has been reasonably invariant to moves in peripheral spreads over the course of the year.
This is probably contrary to expectations – most would
have expected the curve to bull-steepen in a risk-averse
environment and when the RBA is easing policy. As we
note above, we think the most logical explanation for this
is constant demand for AUD duration from offshore.
The positive carry associated with cash bond flatteners
means that curve flattening trades can offer some
protection for investors trading long duration. In the cash
space, the most carry-efficient flatteners involve shorting
April 2012 bonds to buy April 2023 or April 2027 bonds
(around 15bp of carry over a 1-month horizon); based off
the relationship in Exhibit 20, this suggests the flattener
offers protection against around 215bp of 10Y peripheral
spreads narrowing (over 1-month).
Swap spreads should be biased wider in
2012, 3s/10s spread curve likely to
flatten
3Y and 10Y swap spreads have drifted wider over the
course of 2011 (Exhibit 21), and we would expect the
higher trading range to persist into 2012. We expectspreads to make new wides next year, and target 3Y
spreads to trade a 50bp to 85bp range, and 10Y spreads
to trade a 55bp to 90bp range in 2012. We expect spreads
to retain strong directionality next year, with spreads
widening on a rally and narrowing on a selloff.
We still believe that a paid position in AUD 3Y swap
spreads is a good way in which to position for still-wider
peripheral spreads in Europe; Exhibit 22 illustrates the
strong correlation between the two series. Aside from
Exhibit 20: Even though the 3s/10s bond curve has shown somecorrelation to peripheral spreads this has been reasonably limited3s/10s bond curve vs. 10Y weighted peripheral swap spread*, %-pts and bp
y = 0.0007x + 0.2384
R2 = 0.5181
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
200 300 400 500 600 700 800
10y w eighted peripheral swap spread
* A weighted peripheral spread computed against Germany for Ireland, Portugal, Italy,Spain and Greece (weighted by the size of their outstanding bond markets)
Exhibit 21: AUS 3Y and 10Y swap spreads have drifted wider over thecourse of the year Swap spread measures; bp
10
20
30
40
50
60
70
80
Jan-11 May -11 Sep-11
3y swap spread
10y swap spread
Exhibit 22: AUS 3Y swap spreads exhibit strong directionality with 5Yperipheral spreads*bp bp
200
400
600
800
1000
Jan-11 Apr-11 Jul-11 Oct-11
20
35
50
655-year weighted
peripheral spread tobund, lhs3-year sw ap spread,
rhs
* as defined in the footnote in Exhibit 20.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 212/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Sally AuldAC
(61-2) 9220-7816
[email protected]. Morgan Securities Australia Limited
212
reflecting heightened banking system risk, wider swap
spreads should also reflect strong demand for Australia’s
AAA-rated sovereign paper in times of market stress(especially stress that is driven by sovereign debt
concerns). And if peripheral spreads move wider and
markets remain illiquid, then AUD corporate and
Kangaroo issuance will likely be limited in 2012,
implying the absence of issuance-related, receive-side
flow in the swap market. Indeed, year-to-date issuance in
Australia has been the lowest since 2008 for all spread
products (Exhibit 23).
Despite our view of wider swap spreads, there are a
couple of factors which might work to cap the move
wider in spreads next year. The first is further rate cuts
from the RBA. Generally speaking, a lower cash rate has been associated with narrower 3Y swap spreads (less
demand for fixed-rate mortgage products implies less
demand from mortgage hedgers). But we expect this
effect to be limited given the likelihood of an escalation
of the European crisis.
The second is the potential for AUD/JPY-related
receiving in the long end (10Y sector) of the swap curve;
generally speaking, a lower AUD/JPY exchange rate
generates receiving of longer-dated swaps due to the
hedging needs inherent in the issuance of callable power
reverse dual currency notes2.
Taking into account all the varied influences on swap
spreads next year, we are biased towards a flatter 3s/10s
swap spread curve.
Macro-prudential regulation continues
to be the major structural influence on
relative value in AUD cash spread
product
In terms of cash spread product, regulation continues to
be the major structural influence upon the spread product
market in Australia. In 2011, the major development wasgreater clarity on APS210, the regulatory standard which
defines the type of assets that can be held in domestic
bank liquidity books. In addition, the Australian bank
regulator (APRA) and the RBA announced the
establishment of a committed secured liquidity facility
(CLF) in order to facilitate compliance by Australian
2 See S. Auld, Why does a higher AUD/JPY FX rate generate paying in
AUD long end swap yields and AUD/USD cross currency basis
spreads?: Explaining the impact of Callable Power Reverse Dual
Currency Notes, 21 April 2011).
Authorised Deposit-Taking Institutions (ADIs) with the
new macro-prudential framework outlined by Basel III.
The purpose of this facility is to enable an ADI to
establish a committed secured liquidity facility with the
RBA, sufficient in size to cover any shortfall between the
ADI’s holdings of high-quality liquid assets and theliquidity coverage ratio requirement (LCR). APRA has
determined that the only assets that qualify as Level 1
(high quality liquid) assets are cash, balances held with
the Reserve Bank of Australia, and Commonwealth
Government and semi-government securities. APRA has
announced that (at this point in time) there are no assets
that qualify as Level 2 assets. This implies that other
AUD AAA-rated assets, such as supra-national bonds,
covered bonds and RMBS, will only qualify towards the
LCR via the CLF.
Exhibit 23: In 2011, $A issuance volumes are likely to be the lowestsince 2008
AUDbn
0
20
40
60
80
100
120
2008 2009 2010 2011 YTD
Financials SSA Corporate
Exhibit 24: The market has made a clear distinction between Europeanand non-European supra-national paper this year
Asset swap levels (semi/quarterly basis), bp*
-10
20
50
80
110
140
170
200
Apr-2012 Dec-2014 Sep-2017 Jun-2020 Mar-2023
ASIA EIB EUROFIADB IFC KFWRENTEN
*Levels as of 17-November-2011
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 213/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Sally AuldAC
(61-2) 9220-7816
[email protected]. Morgan Securities Australia Limited
213
The ADI will pay a 15bp fee to access the CLF. Given
current market spreads and RBA haircuts/repo charges,
we estimate that spread product – particularly Europeansupra-national paper, RMBS and covered bonds – look
cheap to a bank liquidity portfolio. In terms of AUD
AAA-rated paper eligible for the CLF, an optimally
structured liquidity portfolio should overweight RMBS
and covered bonds3.
More generally, we remain cautious on spread products,
given the risk of further spread widening in peripheral
Europe. In the supra-national space, the market is very
clearly making a geographic distinction, as investors and
dealers reduce exposure to European names in favour of
Washington names (Exhibit 24). As long as European
issues remain largely unresolved, we expect thisdivergence to remain with non-European AAA-rated
assets well sought after (and note that European supra-
nationals – EIB and KfW – represent 35% of 2011 YTD
Kangaroo issuance so far this year).
Semi-government paper – we favour
owning NSWTC bonds against swap
and ACGBs
Given our views on swap spreads, we like owning cash
spread product relative to swap rather than government
bonds. When we look at AAA-rated semi-government
paper using an assessment of relative credit metrics, both
SAFA and NSWTC paper look cheap against swap
(Exhibit 25). However, given recent commentary from
the South Australian Premier, we see a significant risk
that the state of South Australia could loose its AAA
rating. As such, we prefer owning NSWTC paper against
swap.
In contrast, WATC and TCV paper looks rich/fair value
against swap. This relative value assessment holds across
both the 5Y and 10Y sectors, but we prefer to exploit this
relative value in the 5Y sector, given our view on the
swap spread curve (likely biased flatter) and like owning
3 For further detail on this issue, see S. Auld, The RBA's committed
liquidity facility and initial margin changes: A low fee means that European supras, covered bonds and RMBS look cheap, 16 November 2011.
the NSWTC April-15 against the TCV November-
164(Exhibit 26).
As we noted above, we remain cautious on spread
product given 1) the likelihood of an escalation in the
European crisis and 2) the prospect of disappointment on
State government budget outcomes (given a weaker domestic growth backdrop). But we note that the latter
point works in favour of NSWTC paper, given the strong
possibility of an announcement concerning the full
privatisation of New South Wales’ electricity assets. One
implication of any such policy announcement is that the
4 More detail on our relative value methodology can be found in S.
Auld, Strategy Update: Ranking the AAA-rated State Government Issuers – Where is the risk/reward? 4 October 2011.
Exhibit 25: 5-year semi-government spread to swap vs. relative creditranking – NSWTC looks cheap relative to TCV
Asset swap vs. average z-score away from AAA mean; bp
-8.0
0.0
8.0
16.0
24.0
-0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4
Average Z-score aw ay from AAA mean
WATC (Apr-15)
NSWTC (Apr-16)
SAFA (Apr-15)
TCV (Nov-16)
Exhibit 26: We like owning the NSWTC April-15 against the TCVNovember-16 given the cheapness of the asset swap boxSpread in asset swap measures between the NSWTC April 15 and TCV November 16;bp
-2.0
-1.0
0.01.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
May-11 Jun-11 Jul-11 Aug-11 Sep-11 Oct-11
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 214/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Sally AuldAC
(61-2) 9220-7816
[email protected]. Morgan Securities Australia Limited
214
sale of electricity assets could mean significantly reduced
funding requirements for NSWTC; currently, around
30% of the NSWTC net borrowing requirement is usedto fund electricity assets. All other things equal, this
should aid out-performance of NSWTC bonds, especially
in the mid- to longer-dated maturities, and drive
flattening of the NSWTC bond curve5.
Australian inflation-linked bonds: the
outperformer in 2011
Australian inflation-linked bonds have performed very
well over the course of 2011, with real yields declining
considerably over the year (Exhibit 27). The rally has
been most pronounced at the front end of the curve,
helping the real yield curve to bull-steepen. This isconsistent with both downward revisions to the
Australian growth profile over the course of the year and
a global rally in inflation-linked bonds.
Interestingly, Australian 10Y real yields are close to their
all-time lows (since the data series began in the mid-
1980s) (Exhibit 28). Given that real yields are
theoretically meant to encapsulate expectations about
GDP growth over the long run, this performance is
perhaps a little counter-intuitive (especially given the
consensus expectation of strong, medium-term and long-
term growth prospects in Australia thanks to its close ties
with China).
We think one potential explanation for the collapse in
real yields is Australia’s disappointing productivity
performance. In the long run, productivity growth is an
important determinant of an economy’s GDP
performance. Exhibit 29 illustrates that labour
productivity growth and 10Y real yields are reasonably
well correlated. The implication of this is that ongoing
trend declines in productivity might imply a structurally
lower trading range for longer-dated real yields going
forward.
But one important point to remember is that despite the
impressive rally in Australian real yields this year,
spreads to offshore markets continue to remain wide
(Exhibit 30). As we have noted above, this real yield
differential continues to be a driver of offshore demand
for ACGB products in both the nominal and inflation-
linked space.
5 For more detail, see S. Auld, Strategy Update: NSW Electricity
Privatization - implications for the NSWTC bond curve, 2 November 2011.
Exhibit 27: Australian real yields have fallen significantly across thecurve over the course of 2011Real yields across Australian linkers; %
0.75
1.00
1.25
1.50
1.75
2.00
2.25
2.50
2.75
3.00
2015i 2020i 2025i 2030i
3-Jan-11
18-Nov-11
Exhibit 28: The recent rally has pushed Australian 10Y real yields tohistoric lowsMonthly data for 10Y real yields; %
1.5
2.5
3.5
4.5
5.5
6.5
J ul-86 Mar-91 N ov -95 J ul-00 Mar-05 N ov -09
Exhibit 29: The trend decline in productivity may be driving structurallylower real yields in Australiaoya % %-pts
-3.5
-2.5
-1.5
-0.5
0.5
1.5
2.5
3.5
4.5
5.5
6.5
Mar-00 Aug-02 J an-05 J un-07 N ov -09
-1.20
-0.80
-0.40
0.00
0.40
0.80
1.20Labour productivity, lhs
Annual change in 10y real y ield, rhs
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 215/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Sally AuldAC
(61-2) 9220-7816
[email protected]. Morgan Securities Australia Limited
215
Indeed, the growth differential between Australia and the
US correlates well with the 10Y real yield spread; based
on J.P. Morgan growth forecasts for Australia and theUS, the 10Y real yield spread looks too wide at present
(Exhibit 31). Accordingly, we think the better AUS-US
spread contraction trade is to own the ACGB Aug-2020i
against the UST Jul-2021i.
Australian breakevens have not performed well over the
course of the year. This was largely contrary to
expectations, where the consensus anticipated rising
inflation in Australia. However, a softer growth outcome
in 2011 and a lower inflation forecast track from the
RBA has meant that market inflation expectations have
struggled to rally. This has been mirrored in J.P.
Morgan’s November 2011 Inflation Expectations Survey,where expectations of the chance of above-target
inflation in Australia have declined consistently over the
course of the year. The survey suggests that it was
changing perceptions of domestic investors (as opposed
to offshore investors) which, in turn, have driven the
move towards more benign inflation expectations.
J.P. Morgan economists now expect core inflation to
remain within the target band until the end of 2012.
Headline inflation should decline markedly in the near
term (as base effects associated with the impact of the
Queensland floods roll off), but will move higher as the
impact of the carbon tax works its way through theeconomy. J.P. Morgan forecasts for inflation are outlined
in Exhibit 32.
We think one of the better opportunities in the AUD
inflation-linked market is to be found in NSWTC
inflation-linked bonds. NSWTC has been the
predominant issuer in the semi-government inflation
market, largely due to its requirement to fund state-
owned utility assets. However, a recent discussion paper
suggests the strong probability that the NSW
Government will consider the full privatisation of its
electricity assets (generation and transmission and
distribution assets).
In our view, the main consequence should the
government decide to proceed with privatisation is that
there is likely to be far less supply in NSWTC inflation-
linked bonds as a large part of longer-dated inflation
linked issuance is used to fund electricity assets. And
given the recent under-performance of the 2035i
maturity, we think there is good value in owning
NSWTC 2035 linkers vs. bonds. For investors able to
Exhibit 30: Despite the rally in Australian real yields, 10Y cross-marketreal yield spreads remain wideMonthly data for 10Y real yield spreads to TIPS, UK linkers and Canadian Real bonds;
%-pts
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
Mar-00 Nov-01 Jul-03 Mar-05 Nov-06 Jul-08 Mar-10
Spread to US
Spread to UK
Spread to Canada
Exhibit 31: The 10Y real yield spread between Australia and the USlooks too wide, given their GDP differential10Y AUS-US real yield and real GDP growth spreads (J.P. Morgan forecast dashedline);%-pts %-pts
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
Mar-99 May -02 Jul-05 Sep-08 Nov -11
-2.0
-1.0
0.0
1.02.0
3.0
4.0
5.0
6.0
7.010y AUS-US real yield spread, lhs
Real GDP spread, rhs, pushed fwd 4q
Exhibit 32: We expect core inflation to remain within the target band in2012. Headline inflation should decline markedly in the near term onbase effects, but will move higher as the impact of the carbon taxworks its way through the economy
Australian headline and core inflation, actual and J.P. Morgan forecasts; oya %
0.0
1.0
2.0
3.0
4.0
5.0
6.0
03Q1 05Q3 08Q1 10Q3 13Q1
Headline
Core
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 216/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Sally AuldAC
(61-2) 9220-7816
[email protected]. Morgan Securities Australia Limited
216
trade asset swaps, we think that the 2035i also looks
attractive on an asset-swap basis.
Australian bond supply update
The Australian Office of Financial Management
(AOFM) expects that gross Treasury bond issuance will
be around AUD51bn in the current 2011/12 financial
year. In net terms, supply will be around $37bn. So far
this year, the AOFM has issued $44.7bn of ACGBs, with
net issuance at $33.5bn due to the maturity of the Jun-11
benchmark. Four new benchmark bonds have been
introduced, including two new long-end bonds with
maturity dates of 21 April 2023 and 21 April 2027. The
latter bond extended the benchmark ACGB curve by a
further four years. Exhibit 33 shows the ACGB benchmark curve and benchmark issue size as of 31
October 2011.
There is currently $14.4bn of ACGB inflation-linked
bonds on issue across four benchmark lines, and we
expect a further $1.2bn to be issued by June 2012. We
expect gross (and net) issuance of $2.0bn in 2012. As of
October 2011, the AOFM had no plans to extend the
inflation-linked benchmark curve, but in the May budget,
it committed to keeping the size of the inflation-linked
market at 10% to 15% of the total ACGB market.
At present, the outlook for bond supply in 2012 isdifficult to ascertain. One reason is that there is some
uncertainty about whether or not the Federal Government
will be able to meet its commitment to returning the
budget to a small surplus in the 2012/13 fiscal year. J.P.
Morgan economists are doubtful, and do not envisage the
budget returning to surplus until 2014/15. Exhibit 34
highlights the difference between Federal Government
and J.P. Morgan forecasts for the budget in the coming
fiscal years. Over the forecast horizon, the cumulative
difference is $29bn. We should get greater clarity on the
Government’s intentions when it releases the next Mid-
Year Economic and Fiscal Update before the year ends.
Depending on which set of forecasts we use, the outlook
for ACGBs (nominal and inflation-linked) on issue
through to mid-2014 varies from $215bn to $250bn
(Exhibit 35). Note that this estimate does not include T-
notes (usually around $10bn-15bn on issue).
Either way, it is likely that Australia will require the debt
ceiling to be raised at some point in either 2012 or 2013.
Section 5 of the Commonwealth Inscribed Stock Act
(1911) states, “The total face value of stock and
Exhibit 33: The benchmark ACGB curve, maturity and issue sizeOutstanding amount by line in ACGBs; AUDbn
0.0
3.0
6.0
9.0
12.0
15.0
18.0
1 5 - A
p r - 1 2
1 5 - N
o v - 1 2
1 5 - M
a y
- 1 3
1 5 - D
e c - 1
3
1 5 - J
u n - 1 4
2 1 - O
c t - 1 4
1 5 - A
p r - 1
5
2 1 - O
c t - 1 5
1 5 - J
u n - 1
6
1 5 - F
e b - 1 7
2 1 - J
u l - 1 7
2 1 - J
a n - 1
8
1 5 - M
a r - 1
9
1 5 - A
p r - 2
0
1 5 - M
a y
- 2 1
1 5 - J
u l - 2 2
2 1 - A
p r - 2
3
2 1 - A
p r - 2 7
Exhibit 34: Federal and J.P. Morgan government budget forecasts – J.P.Morgan economists are less optimistic on the government’s ability toreturn to surplus in 2012/13Budget balance; AUDbn
-60
-50
-40
-30
-20
-10
0
10
2009-10 (a) 2010-11 (a) 2011-12 (f) 2012-13 (f) 2013-14 (f)
J.P. Morgan
Government
Exhibit 35: Supply of ACGBs depends heavily on budget forecasts for coming fiscal yearsOutstanding amount of ACGBs (nominal and inflation-linked), forecast under FederalGovernment and J.P. Morgan assumptions; AUDmn
0
65,000
130,000
195,000
260,000
Dec-70 Jun-78 Dec-85 Jun-93 Dec-00 Jun-08
Using Federal Governm ent budget forecasts
Using J.P. Morgan budget forecasts
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 217/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Sally AuldAC
(61-2) 9220-7816
[email protected]. Morgan Securities Australia Limited
217
securities on issue under this Act and the Loans
Securities Act 1919 at any time must not exceed $250
billion.” As of 11 November 2011, total CommonwealthGovernment Securities on issue amounted to $215.1bn.
Looking to the end of the forecast horizon, there is still
some debate about the future size of the ACGB market.
This is because if the Federal Government realises its
forecasts for a return to surplus by 2012/13, then this
would imply a negative net bond supply at some point. In
the 2011/12 Budget, the Government noted the
recommendation from a panel of financial market
participants and regulators that the ACGB market be kept
at a constant size of nominal GDP (around 12% to 14%)
(Exhibit 36). If the Government decides to accept this
recommendation, then the AOFM will be required tomanage a portfolio in order to invest surplus cash.
Exhibit 36: There is a proposal to keep the size of the ACGB market asa constant proportion of nominal GDPSize of ACGB market as a proportion of nominal GDP; %
0.0
5.0
10.0
15.0
20.0
25.0
30.0
1983 1988 1993 1998 2003 2008 2013
Proposed target lev el:
12%-14% of GDP
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 218/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Sally AuldAC
(61-2) 9220-7816
[email protected]. Morgan Securities Australia Limited
218
Trading themes
• Extend duration on dips− The lesson in 2011 for investors in AUD fixed income
was that events in Europe – rather than domesticfundamentals – had a major impact on government bond yields.
− Our European strategists remain bearish on Europe,and for this reason we are biased towards long-duration strategies in AUD fixed income in 2012.
− While this is a somewhat difficult call given thatyields are close to their lows for the year, we note thatdomestic investors are still running short of benchmark duration; US 10Y yields should declinefurther and strong demand from offshore investors
should continue to support the market.− Look to lighten duration in 2H12, given the risk that
markets price in a more optimistic global outlook.
• Curve
− The 3s/10s bond curve should steepen in 2012, givenour economists’ expectation of further rate cuts fromthe RBA. But cash bond steepeners have negativecarry and roll; instead, we prefer to express this viewvia forward starting swap steepeners.
− There is very attractive carry and roll in cash bondcurve flatteners. Accordingly, we think these are aneffective portfolio hedge for investors already running
long-duration portfolios as they provide protectionagainst the peripheral spread narrowing.
• Spread product
− We are biased towards wider swap spreads next year
and believe that paying 3Y swap spreads is a goodinsurance trade to hold into 2012. If risk aversiondominates markets, then we would expect the assetswap curve to flatten.
− In the cash spread product, we remain cautious.Domestic positioning still feels overweight spread product and as such, we think there will be better buying opportunities. We favour owning NSWTC paper to swap and AUD European supras against USDEuropean supras.
• Inflation
− We like owning the ACGB Aug-2020i against theUST Jul-2021i; this spread looks too wide given theexpected growth spread between Australia and the US.This trade should also perform given our view that thenominal AUS-US 10Y bond spread is too wide versusour medium term forecasts.
− Our preferred trade in the inflation space is to own the NSWTC Nov-2035i inflation-linked bond against theACGB Sep-2030i bond. Any decision to privatise NSW electricity assets will be extremely positive for longer-dated NSW inflation-linked paper.
• AUS-US 10Y spread
− Our medium-term model suggests the 10Y spreadshould be biased narrower from current levels.
Exhibit 37: AUD interest rate forecastsPercent
Current Dec-11 Mar-12 Jun-12 Sep-12 Dec-12
RBA cash rate 4.50 4.25 3.75 3.75 3.75 3.75
90-day rate 4.60 4.00 3.65 3.90 4.00 4.00
3Y yield 3.22 3.25 2.75 3.30 3.50 3.70
10Y yield 3.99 4.15 3.80 4.30 4.35 4.40
3s/10s curve (bp) 77.0 90.0 105.0 100.0 85.0 70.00
3Y swap yield 3.84 4.00 3.65 4.05 4.10 4.20
10Y swap yield 4.77 4.95 4.65 5.05 5.05 5.05
3Y swap spread 62.0 75.0 90.0 75.0 60.0 50.0
10Y swap spread 75.0 80.0 85.0 75.0 70.0 65.0
AUS-US 10Y spread (bp) 203.0 210.0 210.0 180.0 175.0 170.0
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 219/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Sally AuldAC
(61-2) 9220-7816
[email protected]. Morgan Securities Australia Limited
219
New Zealand
• Despite the fact that both the NZD 2Y and 10Y
swap yields are close to all-time lows, we think
New Zealand yields will likely reach new lows
again in 2012. Accordingly, we prefer to extend
duration on any selloff
• This position should work well for 1H12, and we
would look to lighten duration into the second
half of the year due to the risk of a move towards
higher yields globally
• We think the NZD OIS curve looks cheap relative
to the AUD OIS curve, especially looking at Jun-
12 RBA and RBNZ OIS dates
• Generally, our view is that the 2s/10s swap curve
looks too flat, especially if the RBNZ does not
tighten policy next year. We like entering 2s/5s
swap curve steepeners, 3 months forward. An
alternative expression of this view is to pay the
belly of the 2s/5s/10s swap butterfly, 3 months
forward
• Given our view that 1) New Zealand swap yields
are likely to reach new lows, and 2) that 10Y asset
swap spreads already look too wide, we would
look to underweight the long end of the NZGB
curve against swap.
The economic recovery that wasn’t
New Zealand has faced a plethora of headwinds over the
course of 2011. Indeed, we estimate that the economy
grew by just 2.3% this year, a small improvement on the
1.7% growth recorded in 2010. Indeed, Exhibit 1
illustrates that in level terms, the New Zealand economy
is yet to return to its pre-crisis levels.
The New Zealand economy has been buffeted by
ongoing household sector deleveraging, an elevated NZD, weaker-than-expected trading partner growth
(especially Australia) and the impact of the Canterbury
earthquake (and the subsequent delayed recovery
efforts). Indeed, we expect household deleveraging to be
a permanent feature of the New Zealand economic
backdrop in the years ahead – the experience of the
Scandinavian countries in the early 1990s is quite
illustrative for New Zealand too (Exhibit 2).
Exhibit 1: The New Zealand economy is yet to return to its pre-crisispeak…Real GDP, Index is December 2007 = 100
96
97
98
99
100
Dec-07 Sep-08 Jun-09 Mar-10 Dec-10
Exhibit 2: …and if the Scandinavian experience is any indication,household deleveraging could take a long timeHousehold debt to GDP, %
47
55
63
71
79
87
95
103
-7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7
Years from start of recession
35
40
45
50
55
New Zealand, lhs
Scandinavian average, rhs
Exhibit 3: Low number of building consents suggest that post-earthquake-related activity is yet to have a material impact on theeconomyNumber of building consents issued
500
1000
1500
2000
2500
3000
3500
Jan-90 Oct-93 Jul-97 Apr-01 Jan-05 Oct-08
Seasonally Adjusted
Trend
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 220/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Sally AuldAC
(61-2) 9220-7816
[email protected]. Morgan Securities Australia Limited
220
The one bright spot for the New Zealand economy in
2011 was the Rugby World Cup (RWC), although
sentiment data since the RWC suggests that any boostwas very short-lived.
Accordingly, we only expect a modest uptick in growth
in 2012, with calendar growth likely to be 2.8% during
the year. It has become evident that ongoing seismic
shocks are continuing to delay any rebuilding efforts in
the earthquake-affected regions and as such, the boost to
broader economic activity is likely to be more of a 2013
story. Exhibit 3 shows that so far, there has been no
notable uptick in building consents issued.
The RBNZ responded to the severe earthquake in
Canterbury by cutting the OCR by 50bp to 2.5%, back tothe lows seen during the financial crisis (Exhibit 4).
Initially, the RBNZ characterised this as an emergency
setting and was keen to withdraw the extra stimulus as
soon as possible. It has since become clear that this
setting of monetary policy looks increasingly appropriate
given the lack of momentum in the domestic economy
and mounting risks offshore (Exhibit 5). We see a risk
that RBNZ rhetoric will turn more dovish as we move
into 2012.
J.P. Morgan economists do not expect the RBNZ to
begin normalising policy until September 2012. Clearly,
risks are biased to unchanged policy for longer and possibly, a new low in the OCR. Indeed, the RBNZ’s
decision in early November to delay the introduction of
some macro-prudential regulatory changes (the core
funding ratio for New Zealand banks) because of
offshore instability may go some way to reflecting policy
makers’ current mindset.
Front-end pricing: A new low in the NZ
policy rate in 2012?
Front-end pricing in New Zealand has been very volatile
over the course of 2011, with the market pricing in
anywhere from 20bp of rate cuts over a 12-month
horizon to 110bp of rate hikes over the same period
(Exhibit 6). Currently, the market is expecting the
RBNZ to deliver 20bp of rate cuts by June 2012. In
contrast, the local market economists’ consensus strongly
favours the next move in rates as up in 2012.
While we believe that offshore (and domestic)
fundamentals would need to deteriorate markedly from
here before the RBNZ will countenance the need for a
lower policy rate, we are also cognizant that illiquid
Exhibit 4: The RBNZ returned the policy rate back to lows seen duringthe financial crisis …OCR; %
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
Mar-99 Dec-01 Sep-04 Jun-07 Mar-10
Exhibit 5: …but if global conditions deteriorate further, then the casefor further rate cuts will start to mount – there is a strong correlationbetween moves in the OCR and J.P. Morgan’s Global PMI%-pts Index
-6
-5
-4
-3
-2
-1
0
1
2
Sep-99 F eb-02 Jul-04 Dec -06 M ay -09 Oc t-11
32
37
42
47
52
57
62
Annual difference in
cash rate, lhs
Global PMI, pushed
fwd 6 months, rhs
Exhibit 6: Monetary policy expectations have been quite volatile in NewZealand this year Tightening* priced in over a 12-month period; bp
-40
-20
0
20
40
60
80
100
120
Jan-11 Apr-11 Jul-11 Oct-11
* Negative number implies easing
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 221/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Sally AuldAC
(61-2) 9220-7816
[email protected]. Morgan Securities Australia Limited
221
global funding markets could also pressure the RBNZ to
ease policy again. Exhibit 7 illustrates that there is a
reasonable disconnect between the level of the OCR and bank lending rates. More expensive bank funding will
only pressure these spreads wider.
In a relative value sense, we think the front end of the NZ
OIS curve looks cheap against the AUD OIS curve. For
example, the June 2012 RBNZ meeting date OIS implies
20bp of rate cuts, while the June 2012 RBA meeting date
OIS implies almost 160bp of rate cuts. If the RBA has
cut 160bp by the middle of 2012, then the RBNZ will
have likely cut by more than 20bp. Similarly, if markets
normalise, then there is scope for a more rapid selloff in
the AUD front end than in the NZD front end.
The outlook for outright yields – new
lows in 2012
New Zealand yields have largely moved lower over the
course of the year, with a new historical yield low in the
2Y swap rate made this month. As Exhibit 8 illustrates,
price action has been consistent with the bullish trend
seen in 2Y rates over the past couple of years. This
theme has also been replicated in the longer end of the
swap curve, with the 10Y yield reaching a new historical
yield low in November.
The move lower in yields has been in sympathy with the broader macroeconomic environment in New Zealand
and offshore (disappointing growth outcomes), and a
growing expectation that central bank policy could turn
more dovish. In addition, we think the recent move lower
has been a function of positioning, with most dealers
looking for higher yields into the year-end and not
prepared for a rally.
In the short term, some of the domestic sentiment
indicators suggest the possibility of a bounce in the NZD
2Y swap yields (Exhibit 9). However, as we noted in our
outlook for Australian yields, domestic fundamentals
have not necessarily been a reliable guide to interest-rate
market movements in 2011. And given the risks to the
global backdrop, we would look to use any short-term
sell-off in New Zealand yields as an opportunity to
extend duration.
In the 10Y part of the curve, a similar theme has
prevailed. The NZD 10Y swap has made new, all-time
lows recently but still looks to be trading largely in line
with our estimate for fair value (Exhibit 10).
Exhibit 7: Lending rates have not fallen to the same extent as the OCRin New Zealand, meaning bank funding stress could also be a keydriver of RBNZ policy moves in 2012Various bank lending rates and OCR; %
2
3
4
5
6
7
8
9
10
11
Mar-99 Oct-01 May -04 Dec -06 Jul-09
OCR
Variable business lending rateFloating mortgage rate
Exhibit 8: NZD 2-year swap yields have made new historical lowsrecently, consistent with the bull trend in place since late 20092Y swap yield, %
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
Jan-07 May -08 Sep-09 Jan-11
Exhibit 9: Similar to Australia, domestic indicators suggest the risk of higher yields – but this can easily be swamped by offshoredevelopmentsIndex %-pts
-55.0
-45.0
-35.0
-25.0
-15.0
-5.0
5.0
15.025.0
35.0
45.0
Jun-04 Oct-06 Feb-09 Jun-11
-5
-4
-3
-2
-1
0
1
2
3Domestic economic activity expected, lhs,
pushed fwd. 1q Annual change in 2y swap, rhs
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 222/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Sally AuldAC
(61-2) 9220-7816
[email protected]. Morgan Securities Australia Limited
222
Our view on duration in New Zealand next year looks for
lower yields, at least in 1H12. This is likely to be driven
by:
1) The prospect of lower UST yields in early 2012;
2) The strong likelihood of noticeably more dovish
commentary from the RBNZ;
3) The risk that the domestic data flow in New
Zealand starts to disappoint; and
4) Ongoing flight to quality into New Zealand
fixed income.
Curve spreads should be biased steeper in 2012
The NZD 2s/10s swap curve has traded a wide range
over the course of 2011, reaching a peak of 214bp in late
April and a trough of 127bp in late September. The curve
is currently towards the lower end of this range, at
around 152bp. Generally speaking, our view is that the
curve looks too flat at current levels, especially given the
risks surrounding the OCR in the coming year. Exhibit
11 suggests that the 2s/10s swap curve is already priced
for a modest tightening cycle in 2012.
We think the better curve trade in New Zealand is to position for steeper forward curves in 2012. At present,
we think the forward swap curves look too flat
(especially 2s/5s and 2s/10s) and like paying the 2s/5s
swap curve 3-months forward. Indeed, our models
suggest that the 2s/5s forward curve spread is pricing in a
cash rate close to 3.5% in just six months’ time, an
expectation well above our economists’ view and current
front-end pricing (Exhibit 12).
These trades are attractive because they roll positively up
to spot, especially in 2s/10s (the 1Y forward drop to spot
is around 12bp). Investors who prefer a more defensive
curve steepening position should look to short the front
end against a forward-curve steepener; either the front
end is too rich or the forward curve spreads are too flat.
The shape of the forward curves (in terms of what they
imply for the path of the policy rate) also suggests that
the 2s/5s/10s swap butterfly, 3-months forward, is
starting to look too low. We like paying the belly of the
fly on any move below flat (currently, the spread is
market at 1bp).
Exhibit 10: Fair value model for the NZD 10Y swap yield – according toour forecasts, new yield lows are likely in the new year 10Y swap yield; %
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
8.0
8.5
Mar-98 Oct-00 May-03 Dec-05 Jul-08 Feb-11
Actual
Model
Model is 10-year swap yield = 0.44 + 0.10*90-day rate + 0.65*US 10-year swap yield +0.46*moving average of nominal GDP growth; R-squared = 0.75, standard error is0.33ppts.
Exhibit 11: The 2s/10s curve looks flat, even if the RBNZ executedanother 50bp of rate hikes next year %-pts %
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
Mar-99 Jan-02 Nov -04 Sep-07 Jul-10
0.5
1.5
2.5
3.5
4.5
5.5
6.5
7.5
8.5
9.5
2s10s curve, lhs
NZ OCR, rhs, inverted
Exhibit 12: The cash rate implied by 2s/5s forward curve is too highrelative to our forecast or market pricing. Therefore given the negativecorrelation* between the 2s/5s curve and cash rate, we favour 2s/5sswap curve steepeners 3M forwardCash rate over various horizons; %
2.0
2.2
2.4
2.6
2.8
3.0
3.2
3.4
3.6
3.8
4.0
Current 3m 6m 1y 2y
Cash rate implied by 2s5s forward curv e
Cash rate, JPM forecast
Cash rate implied by market pricing
* Model used for the curve-implied cash rate: 2s5s swap curve = 1.11 - 0.29*real cash
rate; R-squared is 0.86, standard error is 0.22ppts.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 223/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Sally AuldAC
(61-2) 9220-7816
[email protected]. Morgan Securities Australia Limited
223
Swap spreads look close to their wides
in the long end of the curve
New Zealand swap spreads have been quite volatile over
the course of this year, with the 2021 bond trading at
almost 40bp over the swap earlier in the year (Exhibit
13). Since then, NZGBs have generally outperformed
swap, and swap spreads look to be pushing back to the
wides seen earlier in the year.
Generally speaking, NZGB swap spreads tend to exhibit
strong directionality with 2Y swap yields. Indeed, our
model suggests that movements in 2Y swap yields can
explain 86% of the variation in 10Y swap spreads. This
is because the swap market is generally seen as the more
liquid market and will usually outperform bonds in arally and underperform bonds in a selloff, as market
participants look to express duration views through
swaps rather than bonds. Our model of 10Y swap spreads
(Exhibit 14) suggests that swap spreads look wide given
the level of 2Y swap yields.
Given our view that 1) New Zealand swap yields are
likely to reach new lows, and 2) that 10Y swap spreads
already look too wide, we would look to underweight the
long end of the NZGB curve against swap on any out-
performance of NZGBs in Q1 next year. Given the
relative issue size ($6.6bn of Mar-19s and $10.4bn of
May-21s), we would prefer to express this view via the2019 maturity. Look to position for a narrower swap
spread on the May-19 bond on any move to the +30bp
level (currently 24bps).
New Zealand bond supply update
The New Zealand Debt Management Office (NZDMO)
announced that gross bond supply in the 2011/12 fiscal
year will be $13.5bn. In net terms, this amounts to a
supply of $5.9bn, given the redemption of the Nov-2011
bond. Pre-funding in the prior fiscal year (2010/11) has
meant that gross bond supply remains unchanged despite
significant costs associated with the post-earthquakerebuilding. So far this year, the NZDMO has issued
$6.45bn of benchmark bonds and is running ahead of
schedule (on a pro-rata basis) for the fiscal year.
The NZDMO lengthened the maturity of the NZGB
curve this year with the issuance of a 5.5% April 2023
maturity. The NZGB benchmark curve is illustrated in
Exhibit 15; in addition, benchmark line sizes for the
April 2013 and April 2015 bonds were increased to a
maximum of $12.0bn this year. In New Zealand’s
Exhibit 13: NGZBs have generally outperformed swaps since April2011, when they traded 40bp over, and look to be pushing back to their widesSwap spread; bp
-50
-40
-30
-20
-10
0
10
20
30
40
50
Jan-11 Apr-11 Jul-11 Oct-11
NZGB Apr-13
NZGB May-21
Exhibit 14: 10Y swap spread model – spreads look too wide given thelevel of yields10Y swap spread actual and model*; bp
-60
-40
-20
0
20
40
60
80
100
120
140
160
Oct-03 Jul-05 Apr-07 Jan-09 Oct-10
Actual
Model
+/- 2 s.d.
* Model 10Y swap spread = -87.55 + 23.90*2Y swap yield; R-squared = 0.86, standarderror = 12.95bp
Exhibit 15: NZGB benchmark curveOutstanding by line; NZDbn
0.0
2.0
4.0
6.0
8.0
10.0
12.0
1 5 - A p r - 1 3
1 5 - A p r - 1 5
1 5 - D e c - 1 7
1 5 - M a r - 1 9
1 5 - M a y - 2 1
1 5 - A p r - 2 3
.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 224/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Sally AuldAC
(61-2) 9220-7816
[email protected]. Morgan Securities Australia Limited
224
inflation-linked bond space, the NZDMO has postponed
plans for an inflation-linked issue (2023 maturity) until
1H12. There has been no new issuance into the soleinflation-linked February 2016 maturity so far this year.
Total bonds on issue (nominals) in New Zealand now
total $47.4bn (we exclude ‘non-market’ bonds in our
calculations – those held by the RBNZ and EQC). We
expect that gross (and net) issuance in 2012 to likely total
$12.3bn, taking bonds on issue close to $60bn by the end
of 2012 (Exhibit 16).
Exhibit 16: Issuance of NGZBs in 2012 will take outstanding amount toclose to $60bnTotal outstanding of NGZBs; NZDbn
0
20
40
60
80
Jan-00 Apr-03 Jul-06 Oct-09 Jan-13
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 225/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Sally AuldAC
(61-2) 9220-7816
[email protected]. Morgan Securities Australia Limited
225
Trading themes
• Extend duration on dips− Despite the fact that both the NZD 2Y and 10Y swap
yields are close to all-time lows, we think NewZealand yields will likely reach new lows again nextyear. Accordingly, we prefer to extend duration on anyselloff.
− This position should work well for 1H12, and wewould look to lighten duration into the second half of the year due to the risk of a move towards higher yields globally.
− We think the NZD OIS curve looks cheap relative tothe AUD OIS curve, especially looking at the June2012 RBA and RBNZ OIS dates.
• Curve
− Generally, our view is that the 2s/10s swap curvelooks too flat, especially if the RBNZ does not tighten policy in 2012. Carry + roll on 2s10s steepeners islargely neutral.
− We think the forward 2s/5s swap curves look too flatrelative to policy expectations. Implement 2s/5ssteepeners, 3-months forward (3-month carry + roll of 2.5bps). A more defensive expression of this tradewould be to hold steepeners against paid 3-month OIS positions.
− We like paying the belly of the 2s/5s/10s swap curve,
3-months forward. This trade has positive carry + rollof 3.0bp over a 3-month period.
•
Spreads− Given our view that 1) New Zealand swap yields are
likely to reach new lows, and 2) that 10-year assetswap spreads already look too wide, we would look tounderweight the long end of the NZGB curve againstswap on any out-performance of NZGBs in Q1 nextyear.
− Given the relative issue size ($6.6bn of Mar-19s and$10.4bn of May-21s), we would prefer to express thisview via the 2019 maturity. Look to position for anarrower asset swap spread on the May-19 bond onany move to the +35bp level.
Exhibit 17: NZD interest rate forecasts%
Current Dec-11 Mar-12 Jun-12 Sep-12 Dec-12
RBNZ OCR 2.50 2.50 2.50 2.50 2.75 3.00
90-day rate 2.66 2.65 2.65 2.75 3.00 3.25
3-year yield 2.82 2.85 2.50 3.25 3.40 3.50
10-year yield 3.91 4.05 3.80 4.40 4.40 4.40
10-year swap spread (bp) 27.5 20.0 25.0 15.0 10.0 10.0
2-year swap yield 2.71 2.65 2.30 2.95 3.00 3.10
10-year swap yield 4.23 4.25 4.05 4.55 4.50 4.50
2s10s swap curve (bp) 152.0 160.0 175.0 160.0 150.0 140.0
NZ-US 10-year spread (bp) 190.0 200.00 210.0 190.0 180.0 170.0
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 226/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Aditya ChordiaAC
(44-20) 7777-9841
[email protected]. Morgan Securities Ltd
226
Interest rate forecasts
Euro area 18-Nov-11 Mar-12 Jun-12 Sep-12 Dec-12 United States 18-Nov-11 Mar-12 Jun-12 Sep-12 Dec-12
Refi rate 1.25 0.75 0.50 0.50 0.50 Fed funds 0 - 0.25 0 - 0.25 0 - 0.25 0 - 0.25 0 - 0.25
Euribor 3M 1.47 1.40 1.45 1.20 1.00 Libor 3M 0.49 0.60 0.50 0.40 0.35
2Y 0.46 0.35 0.30 0.40 0.50 2Y 0.28 0.17 0.30 0.30 0.30
5Y 1.10 0.75 0.70 0.85 1.05 5Y 0.92 0.75 1.25 1.25 1.25
10Y 1.97 1.55 1.25 1.50 1.75 10Y 2.01 1.70 2.50 2.50 2.50
30Y 2.61 2.15 1.95 2.25 2.50 30Y 3.00 2.70 3.60 3.60 3.60
2s/10s 151 120 95 110 125 2s/10s 173 153 220 220 220
10s/30s 64 60 70 75 75 10s/30s 99 100 110 110 110
2s/30s 215 180 165 185 200 2s/30s 271 253 330 330 330
2Y 110 130 145 135 130 2Y 48 53 38 38 38
5Y 96 115 130 120 115 5Y 40 45 34 34 34
10Y 67 80 90 85 80 10Y 16 27 18 18 18
30Y 19 25 30 25 25 30Y -30 -23 -23 -23 -23
United Kingdom 18-Nov-11 Mar-12 Jun-12 Sep-12 Dec-12 Australia
Base rate 0.50 0.50 0.50 0.50 0.50 Cash rate 4.50 3.75 3.75 3.75 3.75
Libor 3M 1.02 1.05 1.10 1.05 1.00 Govt curve 10Y 3.99 3.80 4.30 4.35 4.40
2Y 0.48 0.50 0.50 0.55 0.65
5Y 1.13 1.00 1.00 1.05 1.15 New Zealand
10Y 2.26 1.80 1.50 1.75 1.95 Cash rate 2.50 2.50 2.50 2.75 3.00
30Y 3.19 2.70 2.25 2.40 2.60 Govt curve 10Y 3.91 3.80 4.40 4.40 4.40
2s/10s 178 130 100 120 130
10s/30s 93 90 75 65 65
2s/30s 271 220 175 185 195
2Y 96 110 120 110 100
5Y 71 85 95 85 65
10Y 31 45 55 45 25
30Y -9 0 5 0 -10
Japan
O/N call rate 0.05 0.05 0.05 0.05 0.05
2Y 0.12 0.15 0.15 0.15 0.15
5Y 0.31 0.30 0.35 0.40 0.45
10Y 0.94 0.90 0.95 1.10 1.15
20Y 1.70 1.75 1.80 1.90 2.00
30Y 1.93 1.95 2.00 2.10 2.15
2s/10s 82 75 80 95 100
10s/30s 99 105 105 100 100
2s/30s 181 180 185 195 200
Govt curve
Swap spreads
Govt curve
Swap spreads
Govt curve
Govt curve
Swap spreads
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 227/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Aditya ChordiaAC
(44-20) 7777-9841
[email protected]. Morgan Securities Ltd
227
Recent curve movements
Govt. curve 18-Nov-11 3Q end 2Q end 1Q end YTD min YTD max 18-Nov-11 3Q end 2Q end 1Q end YTD min YTD max
2Y 0.46 0.56 1.61 1.78 0.31 1.92 0.48 0.56 0.83 1.34 0.47 1.55
5Y 1.10 1.17 2.28 2.66 0.86 2.82 1.13 1.07 1.84 2.25 0.95 2.55
10Y 1.97 1.89 3.03 3.35 1.69 3.50 2.26 2.26 3.38 3.68 2.10 3.88
30Y 2.61 2.64 3.76 3.79 2.46 3.97 3.19 3.55 4.28 4.35 3.12 4.58
2s/5s 63 61 67 88 44 109 64 50 101 91 41 135
2s/10s 151 133 142 157 128 207 178 169 255 234 152 276
10s/30s 64 75 73 44 39 99 93 129 90 67 62 144
Swap curve 18-Nov-11 3Q end 2Q end 1Q end YTD min YTD max 18-Nov-11 3Q end 2Q end 1Q end YTD min YTD max
2Y 1.60 1.50 2.18 2.35 1.37 2.47 1.54 1.28 1.45 1.86 1.08 5.05
5Y 2.09 1.98 2.82 3.08 1.84 3.23 1.86 1.78 2.51 2.97 1.64 3.22
10Y 2.68 2.55 3.44 3.63 2.40 3.79 2.60 2.60 3.55 3.80 2.48 4.05
30Y 2.82 2.74 3.85 3.86 2.60 4.02 3.11 3.34 4.08 4.13 3.04 4.292s/5s 49 48 64 73 41 88 32 50 106 111 -257 125
2s/10s 108 105 126 128 97 165 105 132 211 194 -162 214
10s/30s 14 19 42 23 14 42 51 74 53 32 20 85
Swap spreads 18-Nov-11 3Q end 2Q end 1Q end YTD min YTD max 18-Nov-11 3Q end 2Q end 1Q end YTD min YTD max
2Y 110 92 55 53 47 112 96 67 48 48 42 96
5Y 96 81 49 39 37 100 71 50 40 36 29 71
10Y 67 65 41 26 21 78 31 21 6 6 2 37
30Y 19 9 9 4 0 22 -9 -21 -20 -22 -30 -6
2s/5s -14 -11 -6 -14 -20 2 -25 -17 -8 -12 -26 -7
2s/10s -43 -27 -14 -27 -44 -3 -65 -46 -43 -41 -65 -27
10s/30s -48 -56 -32 -21 -67 -18 -40 -42 -26 -29 -46 -22
EUR GBP
Govt. curve 18-Nov-11 3Q end 2Q end 1Q end YTD min YTD max 18-Nov-11 3Q end 2Q end 1Q end YTD min YTD max2Y 0.28 0.25 0.46 0.79 0.15 0.85 0.12 0.14 0.17 0.21 0.12 0.25
5Y 0.92 0.97 1.75 2.22 0.77 2.40 0.31 0.37 0.43 0.49 0.29 0.62
10Y 2.01 1.93 3.16 3.45 1.71 3.72 0.94 1.03 1.13 1.25 0.94 1.35
30Y 3.00 2.92 4.38 4.51 2.76 4.76 1.93 1.91 2.01 2.18 1.85 2.26
2s/5s 64 71 130 144 57 155 19 23 26 28 17 38
2s/10s 173 167 270 267 151 289 82 89 96 104 82 114
10s/30s 99 99 123 106 95 148 99 89 88 93 86 101
Swap curve 18-Nov-11 3Q end 2Q end 1Q end YTD min YTD max 18-Nov-11 3Q end 2Q end 1Q end YTD min YTD max
2Y 0.78 0.58 0.69 0.96 0.42 1.05 0.38 0.35 0.38 0.39 0.31 0.46
5Y 1.34 1.26 2.02 2.42 1.07 2.62 0.46 0.49 0.56 0.63 0.42 0.78
10Y 2.18 2.11 3.26 3.56 1.90 3.83 0.95 1.03 1.16 1.31 0.95 1.40
30Y 2.70 2.70 4.07 4.29 2.56 4.54 1.72 1.80 2.01 2.17 1.72 2.25
2s/5s 56 69 133 147 53 158 7 14 17 24 7 322s/10s 140 154 257 260 135 279 57 68 78 91 57 98
10s/30s 52 58 81 73 51 93 77 77 85 86 75 88
Swap spreads 18-Nov-11 3Q end 2Q end 1Q end YTD min YTD max 18-Nov-11 3Q end 2Q end 1Q end YTD min YTD max
2Y 48 32 22 16 13 52 26 21 22 19 15 26
5Y 40 29 25 19 16 44 14 11 13 13 6 17
10Y 16 16 8 8 2 22 -1 0 2 5 -3 9
30Y -30 -27 -31 -23 -38 -20 -22 -13 -4 -1 -22 2
2s/5s -9 -3 3 3 -7 9 -12 -10 -9 -5 -13 -5
2s/10s -32 -15 -14 -8 5 33 -27 -21 -19 -13 -27 -12
10s/30s -46 -43 -39 - 31 27 54 -21 -13 -6 -6 -21 -4
USD JPY
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 228/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Aditya ChordiaAC
(44-20) 7777-9841
[email protected]. Morgan Securities Ltd
228
Recent sovereign cash spread movements
Spread to Germany; bp
18-Nov-11 1W ago 2W ago 1M ago 3M ago 1Y min 1Y max 1Y avg 1Y SD 1Y z-score
Austria 111 94 45 47 30 6 130 27 19 4.4
Belgium 344 273 219 212 132 32 357 104 62 3.9
Finland 39 23 13 17 12 -4 51 10 7 4.0
France 119 105 63 69 33 4 140 28 24 3.7
Netherlands 29 19 15 14 11 1 48 9 6 3.3
Greece 9131 8414 8379 6499 3321 1092 9131 3209 2330 2.5
Ireland 773 765 801 732 774 361 2164 808 314 -0.1
Italy 572 543 522 408 283 83 684 225 136 2.6
Portugal 1507 1556 1900 1603 1099 299 1901 962 500 1.1
Spain 484 421 370 322 264 102 492 225 83 3.1
Wtd. peri. spread* 1418 1321 1315 1042 632 229 1418 574 336 2.5
18-Nov-11 1W ago 2W ago 1M ago 3M ago 1Y min 1Y max 1Y avg 1Y SD 1Y z-score
Austria 177 157 110 100 54 30 202 55 30 4.1
Belgium 330 294 283 281 196 78 351 158 63 2.7
Finland 82 63 52 52 33 6 91 26 19 3.0
France 151 135 106 104 47 21 187 49 30 3.4
Netherlands 66 50 41 51 35 5 83 25 17 2.5
Greece 4003 3771 3573 2939 1734 937 4059 1714 763 3.0
Ireland 706 707 704 614 780 495 1509 759 211 -0.3
Italy 533 548 526 435 304 102 675 250 132 2.1
Portugal 1322 1351 1459 1375 1085 292 1527 905 392 1.1
Spain 453 409 388 344 302 155 474 266 72 2.6
Wtd. peri. spread* 892 871 840 709 503 241 951 451 179 2.5
18-Nov-11 1W ago 2W ago 1M ago 3M ago 1Y min 1Y max 1Y avg 1Y SD 1Y z-score Austria 147 149 113 105 60 31 189 58 28 3.1
Belgium 282 256 253 244 176 72 309 141 56 2.5
Finland 67 55 51 51 43 24 75 37 11 2.8
France 147 148 120 110 64 30 188 58 30 3.0
Netherlands 56 47 43 45 38 19 65 33 10 2.3
Greece 2425 2357 2407 2266 1352 747 2471 1342 491 2.2
Ireland 625 611 631 637 761 501 1142 679 128 -0.4
Italy 494 492 471 399 285 125 575 244 113 2.2
Portugal 910 907 986 961 930 306 1161 722 269 0.7
Spain 469 395 374 338 291 170 486 267 65 3.1
Wtd. peri. spread* 699 673 666 603 444 229 739 400 135 2.2
18-Nov-11 1W ago 2W ago 1M ago 3M ago 1Y min 1Y max 1Y avg 1Y SD 1Y z-score
Austria 121 124 95 82 41 28 158 45 22 3.5Belgium 238 238 219 210 147 75 280 129 43 2.5
Finland - - - - - - - - - -
France 152 158 123 109 57 32 199 58 30 3.2
Netherlands 11 6 7 9 11 6 15 10 2 0.6
Greece 1523 1487 1448 1303 761 458 1544 781 305 2.4
Ireland - - - - - - - - - -
Italy 450 444 426 392 287 153 515 254 92 2.1
Portugal 648 652 639 628 544 254 682 443 146 1.4
Spain 431 389 365 343 270 185 448 267 53 3.1
Wtd. peri. spread* 563 545 525 484 345 205 595 320 104 2.3
10Y
30Y
2Y
5Y
*Weighted peripheral spread computed against Germany for Greece, Ireland, Portugal, Italy and Spain (weighted by the size of their outstanding bond market). 30Y doesnot contain Ireland.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 229/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Aditya ChordiaAC
(44-20) 7777-9841
[email protected]. Morgan Securities Ltd
229
Recent sovereign CDS spread movements
CDS spread*; bp
18-Nov-11 1W ago 2W ago 1M ago 3M ago 1Y min 1Y max 1Y avg 1Y SD 1Y z-score
US 29 29 29 50 50 16 50 33 13 -0.3
UK 51 50 49 52 36 21 58 34 9 1.8
Germany 46 45 43 48 30 10 61 24 12 1.8
Austria 149 127 109 112 48 20 159 54 33 2.9
Belgium 306 279 260 275 202 62 320 156 68 2.2
Finland 34 32 30 38 30 12 45 22 9 1.3
France 158 138 121 131 98 29 167 68 38 2.4
Netherlands 55 49 48 41 26 10 60 26 12 2.5
Greece*** 55 52 54 57 37 15 59 31 14 1.7
Ireland 911 930 917 968 1051 531 1606 858 223 0.2Italy 573 551 463 428 336 71 627 229 150 2.3
Portugal 1507 1500 1471 1527 1196 392 1757 934 440 1.3
Spain 451 405 360 362 351 142 464 274 82 2.2
Wtd. peri. spread** 616 591 521 506 429 148 657 316 145 2.1
18-Nov-11 1W ago 2W ago 1M ago 3M ago 1Y min 1Y max 1Y avg 1Y SD 1Y z-score
US 40 40 40 65 65 37 65 52 9 -1.3
UK 90 89 87 92 79 47 104 70 13 1.6
Germany 95 92 87 95 82 36 120 61 21 1.7
Austria 207 185 159 163 120 50 222 98 39 2.8
Belgium 332 305 286 301 235 115 346 201 58 2.3
Finland 68 63 60 76 60 24 90 45 18 1.3
France 221 202 178 191 148 63 234 116 45 2.3Netherlands 113 100 98 98 67 28 122 60 25 2.2
Greece*** 60 58 59 61 45 25 63 41 12 1.7
Ireland 715 730 720 760 785 510 1199 694 121 0.2
Italy 533 530 490 453 360 125 583 269 130 2.0
Portugal 1060 1054 1034 1119 869 390 1235 750 258 1.2
Spain 465 423 383 385 370 197 479 310 68 2.3
Wtd. peri. spread** 557 545 507 492 415 189 595 331 115 2.0
2Y
5Y
* 25bp running coupon used for Finland, France, Germany, Netherlands and US. 100bp running coupon used for UK, Austria, Belgium, Greece, Ireland, Italy, Portugal andSpain. Spreads for all the countries except US are in $ CDS and for US it is in € CDS.** Weighted peripheral spread computed as CDS spread of Ireland, Portugal, Italy and Spain, weighted by the size of their outstanding bond market.*** For 2Y and 5Y Greece CDS, we show upfront premium instead of flat CDS spread for 100bp running coupon.
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 230/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Aditya ChordiaAC
(44-20) 7777-9841
[email protected]. Morgan Securities Ltd
230
Euro area marketable bond* and bank debt** redemptions
€bn
Sov. Banks Sov. Banks Sov. Banks Sov. Banks Sov. Banks Sov. Banks Sov. Banks
Dec-11 0 2 0 0 0 0 0 10 18 3 0 1 18 16
Jan-12 2 3 0 5 0 0 15 5 25 28 14 8 56 50
Feb-12 0 2 0 2 0 0 0 7 0 8 0 12 1 31
Mar-12 1 2 4 1 0 0 0 7 19 13 0 3 24 26
Apr-12 0 2 0 0 0 1 18 6 16 3 0 10 34 22
May-12 0 2 0 2 0 0 0 2 0 2 0 4 0 12
Jun-12 0 1 0 1 0 1 0 5 19 9 0 2 19 20
Jul-12 10 1 0 0 0 0 28 13 27 10 15 2 81 25
Aug-12 0 0 1 1 1 1 0 0 0 5 0 3 2 10
Sep-12 0 2 12 0 6 0 12 5 21 11 0 2 51 21
Oct-12 0 1 0 3 0 0 19 9 16 9 0 1 35 23
Nov-12 1 1 0 1 0 0 0 1 0 5 0 1 1 9
Dec-12 0 2 8 0 0 0 5 3 17 3 0 1 30 9
Total 15 18 25 17 7 4 98 74 178 108 30 52 353 272
Sov. Banks Sov. Banks Sov. Banks Sov. Banks Sov. Banks Sov. Banks
Dec-11 3 0 0 0 0 2 0 1 0 6 3 10
Jan-12 0 2 0 1 0 4 0 4 0 6 0 18
Feb-12 0 1 0 2 36 5 0 1 1 21 38 29
Mar-12 14 2 6 2 27 6 0 1 1 14 48 25
Apr-12 0 0 0 3 28 2 0 0 12 15 40 21
May-12 9 0 0 0 1 6 0 1 0 10 9 17
Jun-12 0 1 0 2 2 3 10 2 0 25 12 33
Jul-12 0 1 0 0 17 2 1 2 13 8 31 12
Aug-12 8 1 0 0 12 2 0 0 0 3 20 5
Sep-12 0 2 0 0 11 4 0 0 2 7 13 12
Oct-12 0 1 0 1 20 6 0 0 20 5 40 12
Nov-12 0 1 0 3 13 4 0 0 0 10 13 18
Dec-12 2 0 0 0 30 10 1 0 0 5 34 15
Total 36 11 6 14 198 55 13 14 49 133 301 227
Austria Belgium
Greece Ireland
Netherlands Core Euro-area** *
Peri. Euro-area****
Finland
Portugal Spain
France Germany
Italy
* Marketable bonds include: conventionals, linkers, floaters zero coupons and international bonds.** Maturities in all currencies and jurisdictions and include secured, unsecured and securitised issuance, including MTNs but excluding short-term (maturity of less than oneyear) and self-funded deals (deals where there is only one bookrunner and it is also the issuer). The data also include any government guaranteed issuance by the banksbut no direct issuance by government or government sponsored institutions.*** Austria, Belgium, Finland, France, Germany and Netherlands**** Greece, Ireland, Italy, Portugal, SpainSource: J. P. Morgan, Dealogic
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 231/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Aditya ChordiaAC
(44-20) 7777-9841
[email protected]. Morgan Securities Ltd
231
Euro area sovereign ratings / SMP purchases /
Election calendar
Sovereign ratings
Austria AAA Aaa AAA
Belgium AA+ NEG Aa1 NEG* AA+ NEG
Cy prus BBB NEG* Baa3 NEG* BBB NEG
Finland AAA Aaa AAA
France AAA Aaa AAA
Germany AAA Aaa AAA
Greece CC NEG Ca DEV CCC
Ireland BBB+ Ba1 NEG BBB+ NEG
Italy A NEG A2 NEG A+ NEG
Netherlands AAA Aaa AAA
Portugal BBB- NEG Ba2 NEG BB+ NEG
Slov akia A+ POS A1 A+
Slov enia AA- Aa3 NEG* AA- NEG
Spain AA- NEG A1 NEG AA- NEG
S&P Moody's Fitch
* represents under watch; grey highlight: below IG;Outlook:NEG - negative outlook, POS – positive outlook, DEV - developing outlookand blank represents stable outlookNotes:1 – For a country’s exclusion, Barclays Capital requires 2 out of 3 credit ratings (S&P,Moody’s and Fitch) to be below IG. Citigroup requires both S&P and Moody’s ratingbelow IG. J.P. Morgan EMU Investment Grade index requires Moody’s, S&P, andFitch each to rate a country above IG to maintain inclusion.
2 – Markit iBoxx uses an average rating methodology (S&P, Moody’s and Fitch) for acountry's exclusion.Source: Bloomberg
SMP purchases
First Last First LastWeekly
(€bn)
Amount
matured
(€bn)
Amount
offered for
sterilization
(€bn)
09-Nov 15-Nov 14-Nov 18-Nov 8.0 0.1 194.5
02-Nov 08-Nov 07-Nov 11-Nov 4.5 0.6 187.0
26-Oct 01-Nov 31-Oct 04-Nov 9.5 0.0 183.0
19-Oct 25-Oct 24-Oct 28-Oct 4.0 0.0 173.5
12-Oct 18-Oct 17-Oct 21-Oct 4.5 0.2 169.5
05-Oct 11-Oct 10-Oct 14-Oct 2.2 0.0 165.0
28-Sep 04-Oct 03-Oct 07-Oct 2.3 0.0 163.0
21-Sep 27-Sep 26-Sep 30-Sep 3.8 0.0 160.5
14-Sep 20-Sep 19-Sep 23-Sep 4.0 0.1 156.5
07-Sep 13-Sep 12-Sep 16-Sep 9.8 0.0 152.5
31-Aug 06-Sep 05-Sep 09-Sep 14.0 0.0 143.0
24-Aug 30-Aug 29-Aug 02-Sep 13.3 0.0 129.0
17-Aug 23-Aug 22-Aug 26-Aug 6.7 1.3 115.5
10-Aug 16-Aug 15-Aug 19-Aug 14.3 0.0 110.5
03-Aug 09-Aug 08-Aug 12-Aug 22.0 0.0 96.0
78.3 4.3 74.0
Trade date Settlement date
Cumulativ e amount till 2 Aug 2011
Note: Every Tuesday ECB sterilizes SMP purchases during the period between the
Wednesday two weeks back (first trade date) and the Tuesday of the previousweek (last trade date). This is equivalent to SMP purchases which settle onMonday (first settlement date) to Friday of the previous week (last settlement date).The ECB started buying Italian and Spanish bonds on Monday, 8 August 2011.Source: ECB
Election calendar Period Date Month Election in
2012 19 February Greece
22 April France
2013 - - Austria
- - Germany
- - Italy
2014 - - Belgium
- - Netherlands
2015 - - Finland
Portugal
- - Spain
2016 - - Ireland
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 232/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Aditya ChordiaAC
(44-20) 7777-9841
[email protected]. Morgan Securities Ltd
232
Euro area fact sheet
Avg debt maturity Non-dom Bank assets % of bank assets in govt % loans/ % of bank assets
Bonds Tbi lls % Tbi lls (years, JPM) Gross (€bn, JPM) Net (€bn, JPM) investors /GDP ratio securi ties (incl . non-dom.) deposi ts rat io funded at the CB
Oct-11 Oct-11 Oct-11 Oct-11 Tot. 2012 Tot. 2012 (%) Sep-11 Sep-11 Sep-11 Latest
Austria 167 1 1 7.6 20 10 79 3.3 2.9 109 0.7
Belgium 264 32 11 6.7 40 16 59 3.2 8.8 59 1.6
Finland 62 7 10 5.2 13 7 84 3.0 1.7 141 0.0
France 1098 163 13 7.3 180 95 66 4.2 3.9 117 1.0
Germany 1021 63 6 6.2 175 18 81 3.2 3.7 84 0.4
Greece 247 12 5 6.7 0 -30 67 2.3 9.4 134 15.9
Ireland 85 0 0 6.3 0 -6 83 4.1 3.3 156 21.8
Italy 1356 130 9 6.9 145 35 45 2.5 6.5 126 2.8
Netherlands 261 40 13 6.2 48 18 68 4.0 4.0 125 0.7
Portugal 104 10 9 6.1 0 -10 80 3.4 4.8 119 8.0
Spain 469 81 15 6.7 90 49 38 3.3 5.0 107 2.4
GDP wtd. avg. 9 6.7 65 3.3 4.7 108 2.1
US* 8246 1480 15 5.3 2150 936 46 - 3.7 81 -
UK* 944 57 6 14.5 186 142 30 5.4 1.8 91 2.0
Marketable debt outstanding (€bn) Conv. bond issuance
GDP GDP growth Inflation** Budget balance^ Prim. Balance Gross debt Curr. acc. bal. GDPpc Unempl.
(€bn) (oya, %) (oya, %) (% of GDP) (% of GDP) (% of GDP) (% of GDP) (EU=100) rate (%)
2012 2012 2012 2012 2012 2012 2012 2012 Latest
Austria 310 0.9 2.2 -3.1 -0.3 73 2.8 124 3.9
Belgium 382 0.9 2.0 -4.6 -1.3 99 2.1 116 6.7
Finland 198 1.4 2.6 -0.7 0.5 52 0.0 123 7.8
France 2,028 0.6 1.5 -5.3 -2.5 89 -3.3 104 9.9
Germany 2,623 0.8 2.0 -1.0 1.3 81 4.4 108 5.8
Greece 212 -2.8 0.8 -7.0 1.0 198 -7.9 63 17.6
Ireland 159 1.1 0.7 -8.6 -4.3 118 1.5 119 14.2
Italy 1,617 0.1 2.0 -2.3 3.1 121 -3.0 90 8.3
Netherlands 623 0.5 1.9 -3.1 -1.2 65 7.0 126 4.8
Portugal 169 -3.0 3.0 -4.5 0.8 111 -5.0 54 12.5
Spain 1,094 0.7 1.1 -5.9 -3.5 74 -3.0 80 22.6
GDP wtd. avg. 9,415 0.5 1.8 -3.4 -0.2 91 0.0 101 10
US 15,495* 1.5 1.9 -8.5 -5.4 105*** -3.1 - 9.0
UK 1,566* 0.6 2.9 -7.8 -4.6 89 -0.9 - 8.3
* Local currency** HICP; National index if not available*** IMF Estimate^ Net lending (+) or net borrowing (-)Source: EC European Economic Forecast, Autumn 2011, IMF, Eurostat and ILO
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 233/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) [email protected]. Morgan Securities Ltd
233
Analyst Certification:
The research analyst(s) denoted by an “AC” on the cover of this report certifies (or, where multiple research analysts are primarilyresponsible for this report, the research analyst denoted by an “AC” on the cover or within the document individually certifies, with
respect to each security or issuer that the research analyst covers in this research) that: (1) all of the views expressed in this reportaccurately reflect his or her personal views about any and all of the subject securities or issuers; and (2) no part of any of the researchanalyst's compensation was, is, or will be directly or indirectly related to the specific recommendations or views expressed by the researchanalyst(s) in this report.
Other Disclosures
J.P. Morgan ("JPM") is the global brand name for J.P. Morgan Securities LLC ("JPMS") and its affiliates worldwide. J.P. Morgan Cazenove is amarketing name for the U.K. investment banking businesses and EMEA cash equities and equity research businesses of JPMorgan Chase & Co.and its subsidiaries.
Options related research: If the information contained herein regards options related research, such information is available only to persons whohave received the proper option risk disclosure documents. For a copy of the Option Clearing Corporation's Characteristics and Risks of Standardized Options, please contact your J.P. Morgan Representative or visit the OCC's website athttp://www.optionsclearing.com/publications/risks/riskstoc.pdf
Legal Entities DisclosuresU.S.: JPMS is a member of NYSE, FINRA, SIPC and the NFA. JPMorgan Chase Bank, N.A. is a member of FDIC and is authorized andregulated in the UK by the Financial Services Authority. U.K.: J.P. Morgan Securities Ltd. (JPMSL) is a member of the London Stock Exchangeand is authorized and regulated by the Financial Services Authority. Registered in England & Wales No. 2711006. Registered Office 125 LondonWall, London EC2Y 5AJ. South Africa: J.P. Morgan Equities Limited is a member of the Johannesburg Securities Exchange and is regulated bythe FSB. Hong Kong: J.P. Morgan Securities (Asia Pacific) Limited (CE number AAJ321) is regulated by the Hong Kong Monetary Authorityand the Securities and Futures Commission in Hong Kong. Korea: J.P. Morgan Securities (Far East) Ltd, Seoul Branch, is regulated by the KoreaFinancial Supervisory Service. Australia: J.P. Morgan Australia Limited (ABN 52 002 888 011/AFS Licence No: 238188) is regulated by ASICand J.P. Morgan Securities Australia Limited (ABN 61 003 245 234/AFS Licence No: 238066) is a Market Participant with the ASX andregulated by ASIC. Taiwan: J.P.Morgan Securities (Taiwan) Limited is a participant of the Taiwan Stock Exchange (company-type) andregulated by the Taiwan Securities and Futures Bureau. India: J.P. Morgan India Private Limited, having its registered office at J.P. MorganTower, Off. C.S.T. Road, Kalina, Santacruz East, Mumbai - 400098, is a member of the National Stock Exchange of India Limited (SEBIRegistration Number - INB 230675231/INF 230675231/INE 230675231) and Bombay Stock Exchange Limited (SEBI Registration Number -INB 010675237/INF 010675237) and is regulated by Securities and Exchange Board of India. Thailand: JPMorgan Securities (Thailand)Limited is a member of the Stock Exchange of Thailand and is regulated by the Ministry of Finance and the Securities and Exchange
Commission. Indonesia: PT J.P. Morgan Securities Indonesia is a member of the Indonesia Stock Exchange and is regulated by the BAPEPAMLK. Philippines: J.P. Morgan Securities Philippines Inc. is a member of the Philippine Stock Exchange and is regulated by the Securities andExchange Commission. Brazil: Banco J.P. Morgan S.A. is regulated by the Comissao de Valores Mobiliarios (CVM) and by the Central Bank of Brazil. Mexico: J.P. Morgan Casa de Bolsa, S.A. de C.V., J.P. Morgan Grupo Financiero is a member of the Mexican Stock Exchange andauthorized to act as a broker dealer by the National Banking and Securities Exchange Commission. Singapore: This material is issued anddistributed in Singapore by J.P. Morgan Securities Singapore Private Limited (JPMSS) [MICA (P) 025/01/2011 and Co. Reg. No.: 199405335R]which is a member of the Singapore Exchange Securities Trading Limited and is regulated by the Monetary Authority of Singapore (MAS) and/or JPMorgan Chase Bank, N.A., Singapore branch (JPMCB Singapore) which is regulated by the MAS. Malaysia: This material is issued anddistributed in Malaysia by JPMorgan Securities (Malaysia) Sdn Bhd (18146-X) which is a Participating Organization of Bursa Malaysia Berhadand a holder of Capital Markets Services License issued by the Securities Commission in Malaysia. Pakistan: J. P. Morgan Pakistan Broking(Pvt.) Ltd is a member of the Karachi Stock Exchange and regulated by the Securities and Exchange Commission of Pakistan. Saudi Arabia: J.P.Morgan Saudi Arabia Ltd. is authorized by the Capital Market Authority of the Kingdom of Saudi Arabia (CMA) to carry out dealing as an agent,arranging, advising and custody, with respect to securities business under licence number 35-07079 and its registered address is at 8th Floor, Al-Faisaliyah Tower, King Fahad Road, P.O. Box 51907, Riyadh 11553, Kingdom of Saudi Arabia. Dubai: JPMorgan Chase Bank, N.A., DubaiBranch is regulated by the Dubai Financial Services Authority (DFSA) and its registered address is Dubai International Financial Centre -Building 3, Level 7, PO Box 506551, Dubai, UAE.
Country and Region Specific DisclosuresU.K. and European Economic Area (EEA): Unless specified to the contrary, issued and approved for distribution in the U.K. and the EEA byJPMSL. Investment research issued by JPMSL has been prepared in accordance with JPMSL's policies for managing conflicts of interest arisingas a result of publication and distribution of investment research. Many European regulators require a firm to establish, implement and maintainsuch a policy. This report has been issued in the U.K. only to persons of a kind described in Article 19 (5), 38, 47 and 49 of the Financial Servicesand Markets Act 2000 (Financial Promotion) Order 2005 (all such persons being referred to as "relevant persons"). This document must not beacted on or relied on by persons who are not relevant persons. Any investment or investment activity to which this document relates is onlyavailable to relevant persons and will be engaged in only with relevant persons. In other EEA countries, the report has been issued to personsregarded as professional investors (or equivalent) in their home jurisdiction. Australia: This material is issued and distributed by JPMSAL inAustralia to "wholesale clients" only. JPMSAL does not issue or distribute this material to "retail clients". The recipient of this material must notdistribute it to any third party or outside Australia without the prior written consent of JPMSAL. For the purposes of this paragraph the terms"wholesale client" and "retail client" have the meanings given to them in section 761G of the Corporations Act 2001. Germany: This material is
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 234/236
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 235/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) [email protected]. Morgan Securities Ltd
235
8/3/2019 Jpm Fixed In
http://slidepdf.com/reader/full/jpm-fixed-in 236/236
European Rates StrategyGlobal Fixed Income Markets 2012 OutlookNovember 24, 2011
Pavan WadhwaAC
(44-20) [email protected]. Morgan Securities Ltd