investing towards your goal · 1 1 dec-13 oil us$ price long-term performance of oil price oil...

Hot Topics | March 2015

Leading opinion: Goals-based investment strategies

INVESTING TOWARDS

YOUR GOAL

HOT TOPICS I MARCH 2015

3

The issues surrounding trustee duties are complex and depend entirely on the particular circumstances facing each fund. Trustees must in all cases take their own decision on issues based on their particular fund’s circumstances at the time. It is for this reason that trustees can’t simply rely on what we’ve discussed here today, neither should they regard our discussions as legal advice. Trustees should get specific assistance where they are uncertain of the consequences or reasonableness of any contemplated action.

Copyright in this material is expressly reserved. This workbook may not be copied, stored, retrieved or in any way reproduced without the express written permission of Alexander Forbes Financial Services (Pty) Ltd. Breach of copyright is a serious offence and can lead to litigation.

This is not a work of final reference. While every attempt is made to ensure that the information published is accurate Alexander Forbes Financial Services (Pty) Ltd, the editors, publishers and printers take no responsibility for any loss or damage that may arise out of the reliance by any person on any of the information contained herein.

For further details of our services please contact our office in Johannesburg:Telephone: +27 11 269 1800Fax: +27 11 269 1111Email: [email protected]

INDEX

Going back to first principles 14

Investing towards a goal – thoughts for retirement funds 22

Sneak peek into demystifying risk budgets 27

Liability-driven investment in a goals-based world 31

Life stage solutions – an example of goals-based investing 35

Giving tax-free savings a purpose 44

Section 1 | Economic landscape

Main themes for 2014 6

Looking ahead at 2015 11

Section 2 | Developing investment strategies with an eye on the prize

4

ALEXANDER FORBES

In 2013, we started by asking: who could blame investors for expecting double-digit real returns to continue indefinitely on the back of three very solid years of growth? We pointed out, however, that this didn’t tell the full story and cautioned that there were significant risks lurking in the system. Much of the market dynamic was and is still being manipulated by central bankers around the world. Although interest rates started rising in 2014 after being at record lows, relatively low rates are expected to persist for some time. Slow global economic growth persists as the International Monetary Fund revised down the Gross Domestic Product growth forecasts for 2015.

After setting new highs in 2013, the strong performance in the first half of 2014 was tempered by the second half. But as we have said before, the indices (or any average for that matter) can hide the truth. Sector performance was a tail of two halves once again, with significant divergence between financials, industrials and resources. In addition, small caps delivered just more than double that of large caps for 2014.

But performance over short time periods can distract trustees and advisers from some of the more fundamental considerations. One such consideration is how our investment strategies have performed relative to a fund’s objective. Especially in times of market uncertainty, it’s important to keep your eye on the goal.

Nerina Visser, in her editorial introduction of the Collective Insight January 2015 publication, stated that: “The investment industry continues to be challenged on how it defines performance, how it delivers value and how it measures success.” And in the same publication, Andrew Bradley commented to say: “When money is managed in the traditional way, the manager is either focused on beating a theoretical benchmark, beating their peers or managing the portfolio within some notional risk appetite. None of these particularly resonate with the investor’s individual lifestyle and investment goals.”

“This is the key difference between the conventional approach to managing money and a goals-based approach. A goals-based approach provides no guarantee of better performance, but it does significantly increase the probability of alignment between the investor’s objectives and their portfolio construction and ultimate performance.”

In considering these issues, we’d like to first set the scene. In the first section of this Hot Topics, we consider the main economic themes of 2014 as well as the themes that have emerged for 2015. We will then share our insights related to retirement funds following the National Budget Speech held in February. We’ll then consider the new tax-free savings accounts and how these can be effectively used by giving savings a purpose to individuals.

Together with this workbook, the 2014 Manager Watch™ Survey of Retirement Fund Investment Managers has been made available to you. As you can see in the survey publication, we also look at historical market performance and the drivers thereof, considering also how these have impacted on the asset manager landscape. To end off this section we go back to first principles by considering the steps to evaluating an investment strategy.

In the second and final section we then get down to considering what it really means to have a goals-based investment philosophy. We’ll explore the practical implementation thereof, by way of a next-generation life stage model. We’ll also discuss liability-driven investments within a goals-based framework and the importance of a risk budgeting process.

We hope that this Hot Topics session proves to be insightful and as always, we welcome your feedback.

RegardsThe Hot Topics Team

PREFACE Keep calm and aim for a goal

We would like to thank the following colleagues who contributed to this workbook:

Anne Cabot-Alletzhauser Amy Underwood Michael Prinsloo Trevino Ramsamy

Vickie Lange Deslin Naidoo Dwayne Kloppers Sudhir Madaree


5

SECTION 1Economic landscape


5

ALEXANDER FORBES

6

0

20

40

60

80

100

120

140

160

Dec

-83

Dec

-85

Dec

-87

Dec

-89

Dec

-91

Dec

-93

Dec

-95

Dec

-97

Dec

-99

Dec

-01

Dec

-03

Dec

-05

Dec

-07

Dec

-09

Dec

-11

Dec

-13

Oil

US$

pric

e

Long-term performance of oil price

Oil price

Source: INET BFA

Macro factors that impact on performanceSince the 2008 global financial crisis, investors globally have operated in a risk-on risk-off framework. Six years on, the intervention policies and strategies applied to stabilise global capital markets and restore confidence remain significant drivers of investment decision-making and valuation. This juxtaposition of weak economies and high levels of artificial liquidity has led to significant rotation between asset classes as managers seek to balance safety with the quest for yield.

The role of primary providers of stimulus shifted in 2014 in response to changes in underlying economies. After a difficult 2013, the US economy is showing strong signs of recovery, allowing it to close its quantitative easing (QE) programme. As a result, US equities and bonds performed very well and the dollar strengthened substantially against other developed currencies.

In contrast, concerns about lacklustre growth and stagflation have brought Japan and the European Union to the fore in stimulus programmes. Consequently, earnings have been low in these markets and interest rates have at times even been negative.

With these evolutions in the ongoing and persistent theme of global stimulus, 2014 also saw a number of its own particular quirks. Concerns around the Chinese economy have been present for a while, but in 2014 the Chinese statistics department confirmed that the economy is slowing. This is primarily due to efforts to shift from an export-driven to a consumption-driven economy. As such, most don’t expect the economy to resume its demand for commodities.

Main themes for 2014

On the commodities front, oil had a dramatic collapse in the second half of the year (down -48.3% for the year as illustrated in the graph below), providing welcome relief to emerging markets but producing jitters in oil-producing nations.


7

Oil’s predicament only exacerbated Russia’s woes after geopolitical tensions within Ukraine have tarnished its global image. In the wake of this, the rouble collapsed and Russian bonds continue to flounder at junk status.

On the local front, commodity concerns in the last part of the year came on top of a mining strike in the first part. Together with a myriad of other economic challenges, the South African outlook weakened in the latter half of the year.

Global market performanceGlobal market performance echoed the economic trends in many ways. Buoyed by a strong performance by US equity (share) markets (+13.7%), the MSCI World Index finished

off 2014 on a positive note in US dollar terms (MSCI World +5.5%). But, the US’s strong performance masked a negative year for other major developed markets with Europe down -5.7% and Japan down -3.7%.

Emerging markets performed a little better at -1.8%, while frontier markets were -9.0% down. Many frontier markets rely on trade with Japan and Europe, which led to their dismal performance. As such, the MSCI All Countries Index – which combines both developed and emerging equity markets – returned only +4.7% in dollar terms.

-80%

-60%

-40%

-20%

0%

20%

40%

60%

80%

Dec

-88

Feb-

90

Apr

-91

Jun-

92

Aug

-93

Oct

-94

Dec

-95

Feb-

97

Apr

-98

Jun-

99

Aug

-00

Oct

-01

Dec

-02

Feb-

04

Apr

-05

Jun-

06

Aug

-07

Oct

-08

Dec

-09

Feb-

11

Apr

-12

Jun-

13

Aug

-14

EM less World Average 1-standard deviation below

Rolling 1-year performance: Emerging versus developed markets

Source: Investment Solutions, MSCI

Global government bonds were relatively flat with the Citibank WGB Index declining -0.5% in dollar terms. From the South African investor’s perspective though, the depreciation of the rand to the dollar meant that the rand returns of these indices were significantly higher – with the MSCI All Countries Index returning +15.7% and the WGB Index up +9.9%.

ALEXANDER FORBES

8

Local market performanceDespite deteriorating local economic conditions, the South African equity (share) market finished the year in positive territory, though slightly down from its peak in July. Domestic consensus is that the local equity market is overpriced and may be heading into an asset price bubble. Another surprise for the year was the divergence in performance between the two major equity benchmarks. The FTSE JSE All Share Index (ALSI) returned +10.9% over the year, while the FTSE JSE Shareholder Weighted All Share Index (SWIX) returned +15.4% over the year. At 4.5%, this divergence is at one of the highest points in its history, as illustrated in the following graph:

Source: INET BFA

-8%

-6%

-4%

-2%

0%

2%

4%

6%

8%

-50%

-40%

-30%

-20%

-10%

0%

10%

20%

30%

40%

50%

Nov

-06

Feb-

07

May

-07

Aug

-07

Nov

-07

Feb-

08

May

-08

Aug

-08

Nov

-08

Feb-

09

May

-09

Aug

-09

Nov

-09

Feb-

10

May

-10

Aug

-10

Nov

-10

Feb-

11

May

-11

Aug

-11

Nov

-11

Feb-

12

May

-12

Aug

-12

Nov

-12

Feb-

13

May

-13

Aug

-13

Nov

-13

Feb-

14

May

-14

Aug

-14

Nov

-14

Ret

urns

Rolling 1-year performance

SWIX-ALSI FTSE/JSE SWIX All Share FTSE/JSE All Share

The South African equity (share) market finished the year in positive territory.


9

The best performing sub-sector within financials was banking (+32.1%) despite the collapse of African Bank and rising interest rates. Other sub-sectors also contributing to performance were real estate (+26.5%) and life insurance (+21.5%). Industrials were bolstered by healthcare (+35.9%), consumer services (+29.7%), and consumer goods (+14.1%) with muted returns from technology (+12.7%), telecoms (+9.7%) and industrials (+7.0%). By contrast, with the exception of forestry and paper (+11.7%), it was red across all theresources sub-sectors.

Which specific investment styles dominated performance? Momentum, minimum volatility and quality strategies all led the pack in 2014. At the same time, value, and in particular, deep value investment styles were seriously penalised. In summary, earnings and price momentum continue to dominate market returns to the detriment of the performance of those South African managers who favour deep value and contrarian styles.

South Africa’s Monetary Policy Committee (MPC) hiked the repo rate twice over the course of 2014.The 50 basis point (bp) hike in January 2014 caught the market unaware, introducing significant volatility to both interest rates and currency markets. The 25bp hike in July met a more muted response, but it did accelerate net foreign bond outflows, which totalled R58 billion for the year as sovereign credit expectations deteriorated.

The drivers behind these hikes remained consistent through the year, including:

■ Concerns about the weak rand ■ Rising inflation ■ Labour conflicts and high wage increases ■ Moderate to weak economic growth ■ A wider current account deficit ■ Stubborn unemployment.

The major difference between the two indices is that the ALSI holds 6% more in resources than the SWIX, 5% less in financials and 11% less in industrials. This super-sector difference contributed to 2.5% of the performance differential as financial shares outperformed resources by 42%. The early part of the year saw resources shares gain ground on the other two sectors, but this trend fell apart in the latter half, with resources ending the year down 14.7%, as illustrated in the following graph:

0

50

100

150

200

250

300

350

400

450

1/31

/200

7 3/

31/2

007

5/31

/200

7 7/

31/2

007

9/30

/200

7 11

/30/

2007

1/

31/2

008

3/31

/200

8 5/

31/2

008

7/31

/200

8 9/

30/2

008

11/3

0/20

08

1/31

/200

9 3/

31/2

009

5/31

/200

9 7/

31/2

009

9/30

/200

9 11

/30/

2009

1/

31/2

010

3/31

/201

0 5/

31/2

010

7/31

/201

0 9/

30/2

010

11/3

0/20

10

1/31

/201

1 3/

31/2

011

5/31

/201

1 7/

31/2

011

9/30

/201

1 11

/30/

2011

1/

31/2

012

3/31

/201

2 5/

31/2

012

7/31

/201

2 9/

30/2

012

11/3

0/20

12

1/31

/201

3 3/

31/2

013

5/31

/201

3 7/

31/2

013

9/30

/201

3 11

/30/

2013

1/

31/2

014

3/31

/201

4 5/

31/2

014

7/31

/201

4 9/

30/2

014

11/3

0/20

14

1/31

/201

5

Industrials Resources Financials

Source: INET BFA

ALEXANDER FORBES

10

Template Performance to 31 December 2014

Indices 3 months 1 years 3 years 5 years 10 years

Local Equities FTSE/JSE ALSI 1.4% 10.9% 19.5% 15.8% 18.0%

Local Equities FTSE/JSE SWIX 3.8% 15.4% 21.6% 17.8% 18.6%

Local Bonds All Bond Index (ALBI) 4.2% 10.1% 8.7% 10.0% 8.5%

Local Cash STeFI Overnight 1.4% 5.3% 5.0% 5.3% 6.9%

Inflation-Linked Bonds

Besa Inflation-linked Bond Index

(BSAGI) 2.2% 11.1% 10.2% 10.9% 11.2%

Local Property SAPY 11.1% 26.6% 23.1% 21.4% 21.5%

Global Equities MSCI All Countries World Index

(ACWI) 2.9% 15.7% 29.3% 20.1% 14.6%

Global Bonds Citi World Government Bond Index (WGBI) 0.9% 9.9% 11.6% 11.3% 10.8%

Global Property UBS Global Investors Index 12.6% 34.7% 30.9% 24.8% 14.9%

Source: Investment Solutions

These drivers also led to credit agencies downgrading South Africa multiple times over the course of the year.

Despite these concerns, bond markets delivered strong returns in line with equity markets. Inflation-linked bonds beat the ALSI with a +11.1% return as real rates remained low. Nominal bonds were not far behind at +10.1%. Factors that significantly drove bond market performance were instrument term (duration) and credit pick-up. The long-end of the South African government bond curve (bonds maturing after more than12 years) flattened during 2014, most noticeably in the fourth quarter as inflation expectations diminished with falling commodity prices. The short-end of the curve (bonds maturing within 3 years) displayed the opposite effect by rising during the quarter

and year due to the interest rate hikes. Fixed-rate bonds remain unattractive compared to floating rate notes in the short end. Cash returns were the weakest asset class with the STeFI returning just +5.3%.

Which asset class produced the strongest performance? Listed property. The local property market shrugged off a poor 2013, and returned to its status as the best performing asset class with a +26.6% return. Much of this strong performance appears over the second half of the year and is a function of lower longer term bond yields and high-income yields relative to prevailing rates.

All in all, market direction proved difficult to call both globally and locally.

Global and local performance as at 31 December 2014

Source: Investment Solutions


11

Subdued global growth expected The global growth outlook for 2015 has taken a knock as the International Monetary Fund (IMF) trimmed the global growth forecast from 3.8% down to 3.5%. Despite the recent fall in the oil price and currency depreciation in the euro, the IMF warned that this would not offset weaker economic conditions in major global markets such as Europe, China and Russia. In contrast, US growth is forecasted for 2015 at 3.4%.

The demand for commodities from China waned over 2014, which has translated into lower prices across most base metals. Commodity-driven economies, including South Africa, would be impacted in multiple ways, including:

■ Falling employment in the sector ■ Lower economic growth due to declining production ■ Widening trade deficit ■ Currency depreciation ■ Falling prices of commodities and associated resource shares.

In South Africa, resources shares on our equity market may remain under pressure over the short to medium term. For long-term investors, this may be an opportunity to buy companies through the weaker part of the cycle. Investors with short-dated liabilities or who are approaching retirement might need to review their portfolio exposures with their managers for appropriateness. Potential retrenchments within the affected sectors can impact on retirement funds with a change in the member profile or benefit structures.

Structural reduction in the oil priceAlthough oil has rebounded from its lows in January (Brent crude US$48 to settle over US$60), lower demand for and increasing supply of the commodity implies a sustainably low oil price. Normally, OPEC’s response to bolster oil prices is to reduce oil supply. Given their current concerns about losing market share, they aren’t reducing supply and thereby upsetting the historical supply–demand balance.

For most consumers, this is welcome news, as oil is a significant factor in inflation. Should the lower price persist, this should reduce global inflation. For emerging-market countries that import oil (like South Africa), a lower inflation rate will help stimulate growth, increase disposable income, stabilise the currency and help reduce the current account deficit. These are all positive factors for the economy and investments. Lower inflation rates will positively impact on fixed-income instruments, but not necessarily cash.

If the oil price fall reverses, the opposite scenario is likely to emerge.

Taking a leap of faith with QELower inflation and deflationary environments are not good for countries in the eurozone and Japan, which already face very low inflation rates. Consumers in these countries are incentivised to defer consumption as they expect prices to be cheaper in the future. This results in subdued local demand for goods and services, which reduces growth expectations.

The European Central Bank (ECB) has taken a leap of faith (as the United States did) by launching a one-trillion euro rescue plan to stimulate the economy. Sovereign debt is expected to be purchased from March this year until September 2016, allowing 60 billion euros to be injected per month into the stagnating region.

This liquidity should result in a depreciation of the euro, which can adversely impact on South Africa as Europe is its biggest trading partner. The low interest rate environment may also lead to a quest for yield by European investors resulting in capital flows to emerging markets. This may artificially inflate the value of these markets. South Africa may be a preferred destination for investors due to the high governance standard and the sophistication of its financial system.

Looking ahead at 2015

The Organisation of the Petroleum Exporting Countries (OPEC) is an intergovernmental organisation unifying the petroleum policies of 12 oil-exporting developing nations.

ALEXANDER FORBES

12

Economists are concerned that as only 20% of debt security purchases will be under the control of the ECB, it fundamentally transfers the majority of the risk of default to national central banks. This creates uncertainty around the unity of the euro, allowing for divergent monetary policies and placing the eurozone on the path of possible fracture. Greece is currently most at risk of falling out of the euro. One possible consequence of a euro fracture is a global market crash.

Interest rate cuts or hikes?Market analysts have long anticipated that the United States would raise interest rates once it successfully closed its QE programme. However, this is no longer an obvious decision. With the US dollar appreciating against most major currencies, its global competitiveness is diminishing. The deflationary and persistently low inflationary environment is not conducive to raising rates as this will further strengthen the dollar and could undermine the significant strides made in the US economy. This may result in the US Federal Reserve Bank delaying the decision to raise rates.

Our local Monetary Policy Committee faces similar issues, as the need to stimulate growth and improve global competitiveness would weigh against a rate hike. The decision to leave rates unchanged at its January meeting was not surprising as the Consumer Price Index managed to fall within the 3%–6% target band and is expected to reach 3.5% later in the year. A stable or declining inflationary environment would be positive for equity, property and fixed income investments. Although not expected by the market, there is a small probability that if the outlook improves, a rate cut might be in order. This would not be beneficial to cash investments. However, there are a number of macro factors that can change outcomes significantly that can’t be controlled but only hedged: the stability of the eurozone after the QE programme as well as responding to changes in Greece, US interest rate decisions and oil prices.

There is a small probability that if the outlook improves, a rate cut might be in order. This would not be beneficial to cash investments.


13

SECTION 2Developing investment strategies with an eye on the prize


13

ALEXANDER FORBES

14

The passive versus active debate endures, year after year. And even as this debate continues to rage, a new tension has emerged over which passive strategy is better– the passive versus passive debate.

But all these debates have become a smokescreen, obscuring the most important questions: What does a good passive investment strategy look like? For that matter, what does a good active strategy look like? How do we go about identifying these ideals?

The answers to these questions are remarkably similar. They hinge on questions of how well the strategy has been built to meet its objective, rather than on questions of absolute performance. Are fees important? Of course they are, but if the investment strategy has not been designed to meet the desired outcome, and if the target is missed, then the issue of fees pales in comparison.

Comparing active strategies is hugely compromised by the fact that one is very rarely comparing apples with apples. Active manager strategies might have different strategic asset allocation benchmarks, employ different indices, have different risk profiles and different investment styles. But the surprise is that comparing passive strategies can be just as complex and riddled with inappropriate comparisons. Not all passive strategies are created equal. So when comparing them, it isn’t sufficient to say that the one with the lowest fees is the winner. There’s more to it than that.

This article goes back to first principles to provide you with a framework to navigate the complexities of choosing an investment strategy. We’ve used life stage strategies as our point of comparison as these are perhaps the most complex. By the end of this article, you should be able to get to the essence of any strategy comparison, whether active or passive.

Going back to first principles

Steps to evaluating an investment strategy

Step one | Understand your objective

Life stage portfolios were designed to ensure that retirement fund members could adequately and sustainably replace a portion (often in the range of 60–75%) of their final pensionable income on retirement. It is against this standard that any investment strategy – passive or otherwise – must be measured.

Note that the goal is not to maximise the wealth of members at the point of retirement. It is to ensure that there will be sustainable income replacement post-retirement. This distinction is critical and has become even starker given recent returns in the market. Recently, investment returns have been excellent and yet members’ ability to replace income has been far from adequate. What’s happening here?

Not all passive strategies are created equal.


15

0

10

20

30

40

50

60

70

80

90

100

Dec

-01

Jun-

02

Dec

-02

Jun-

03

Dec

-03

Jun-

04

Dec

-04

Jun-

05

Dec

-05

Jun-

06

Dec

-06

Jun-

07

Dec

-07

Jun-

08

Dec

-08

Jun-

09

Dec

-09

Jun-

10

Dec

-10

Jun-

11

Dec

-11

Jun-

12

Dec

-12

Jun-

13

Dec

-13

Jun-

14

Dec

-14

Replacement ratio index of members born in different years

1952-born 1962-born 1972-born 1982-born

Inde

x

37.8 40.3

50.1

66.7

The index tracks the projected retirement incomes of several defined contribution fund members, of differing ages, over time, showing how they were on track to replace 75% of their income in 2002 and how their retirement prospects are changing over time. What it illustrates is that while investment returns have been excellent, the cost of buying that income stream has been rising. Absolute wealth might be increasing, but the purchasing power of that wealth has been declining faster.

This is where risk management becomes critical. A robust methodology for translating risk, returns and objectives into a compatible mix for an individual investor is through a risk budget.

The reality is that the optimal strategy for one objective, maximising asset growth, is not necessarily the optimal strategy for another, matching the fluctuating cost of securing an income stream. Consider if you wanted to win a sprint, you’d go to Usain Bolt. But if you wanted to win a marathon, you definitely wouldn’t!

Step two | Understand the process

To evaluate an investment strategy, it’s important to understand how it’s created. A generic investment product design process would go through the following steps:

Determine the investment objective

Calculate the asset mix that meets the

investment objective

Establish the horizon of the objective

Establish a monitoring and

rebalancing process

Define the appropriate risk controls

Choose implementation benchmarks for

the assets

The Alexander Forbes Pension Index illustrates this best in the following graph.

Source: Alexander Forbes Research & Product Development

Flow chart 1: Generic investment strategy design

ALEXANDER FORBES

16

The single most important factor in any investment strategy is in setting its objective. Common investment objectives are:

■ Capital growth. Maximising performance within a given risk tolerance.

■ Capital protection. Protecting capital from losses over a specified time period.

■ Liability matching. Being able to meet liability payments as they arise in the future.

■ Income targeting. Growing capital to create a steady income stream.

Portfolios can have single or multiple objectives. Portfolios with multiple objectives would generally make it less likely that any of the individual objectives would be optimally met. But the key point is that comparing portfolios with different objectives is a meaningless exercise.

Life stage portfolios typically have two objectives: 1. To create as much asset growth as prudence allows over

the course of a member’s employment2. To replace as much of that member’s pre-retirement

income as possible. Most life stage strategies will split these objectives into two components:a. An accumulation phaseb. A de-risking phase.

This means that the life stage strategy needs at least two portfolios with two different strategic asset allocations, as well as a process for moving from the accumulation portfolio to the final portfolio. Sometimes, this process will consist of intermediate portfolios in which case the strategy will include more portfolios. A key point is that each of the fund’s objectives can be assessed in their own right as to whether they have achieved those ends. Invariably, though, trustees only focus on whether the accumulation goal has been met.

Understanding the strategic asset allocationSetting the strategic asset allocation combines the investment objective, the horizon for that objective and the appropriate risk controls. Asset-liability modelling or other similarly rigorous processes should be used in this step.

In 1986, Brinson, Hood and Beebower published the article ‘Determinants of portfolio performance’ and popularised the insight that over 90% of the variation in returns of a portfolio is attributable to its asset allocation. This has been confirmed by numerous studies since then.

Determining your strategic asset allocation is your most important decision, not your decision to use active or passive.

The strategic asset allocation decision for a retirement fund needs to cover three aspects:1. The accumulation portfolio. Although growth assets tend

to have the highest returns through time, this portfolio also needs to manage diversification and find the most efficient ways to extract the benefits from different asset classes through time. This requires the right balance of risk.

2. The final portfolio. This portfolio is what the member will be holding when they retire and will determine what kind of income they can retire on. For this reason, it can’t focus only on returns. Furthermore, the member is likely to only be in this portfolio for a short period of time, so the strategic asset allocation needs to be robust even in stressed conditions.

3. The de-risking process. This is the process whereby the member moves from the accumulation portfolio to the final portfolio and is often a dangerously neglected piece of the puzzle. It affects the accumulated wealth more than the entire accumulation phase as the process of trading from one portfolio to the other can undermine all that has come before.

In the same way that setting the strategic asset allocation for the accumulation and final portfolios should be done using robust methods, so too should the de-risking process. For instance, complex simulation techniques can be used to compare two different processes to determine which is structurally superior.

It’s important to understand how the two investment strategies relate to each other. If the strategies are closely aligned you can use shorter periods with higher frequency of switching to achieve the optimal efficiency in reaching your objective.

For example, when de-risking from the accumulation growth strategy into an income-targeting objective, you can achieve this in different ways. Two main questions will need to be answered:1. Over what period? 2. At what frequency?

Determining your strategic asset allocation is your most important decision, not your decision to use active or passive.


17

An optimisation exercise can simulate expected future outcomes. In one of our trials we tested a seven-year annual framework against a five-year quarterly framework for two objective-based portfolios and established that you can add at least 20 basis points per year over the de-risking period with shorter term re-balancing. However, trustees should also consider the operational costs of implementing high-frequency structures.

What this stresses is the importance of robust decision making at every step. It can’t simply be based on historical returns and intuitive guesses. It requires methods that take into account the wide range of possible outcomes using proper statistical forecasting methods. Because it is a challenging exercise, it’s usually done with the help of a consultant who understands the nature of the liabilities.

Understanding the benchmarkOnce the strategic asset allocation has been determined, the benchmark for each asset class needs to be set. This choice needs to be carefully considered – especially in the case of passive investment where the benchmark ultimately becomes the investment.

Key factors that need to be considered before selecting a benchmark:

■ The structure of the benchmark ■ Public and independent benchmark construction ■ Availability of investment portfolios ■ Costs ■ Liquidity

Market cap indices are typically chosen because they are the most widely accepted as being representative of the economic reality of the market. They are also the most publicly available.

Risk-efficient benchmarks such as minimum variance and equally weighted indices can create liquidity and trading issues. Fundamental indices such as RAFI are vendor-specific with built-in biases and may cost more.

In many cases, practical constraints are dominant factors in choosing a benchmark. The choice of benchmark is an active decision that can materially affect the short-term performance outcomes.

A perfect example of this is the deviation between two of South Africa’s major equity benchmarks, the ALSI and the SWIX. They consist of the same shares and they are both market cap weighted indices. But their risk structures are dramatically different. In fact, over the last few years, their performance has diverged significantly, as illustrated in the following graph:

Figure 1: Comparing the FTSE JSE SWIX All Share against the FTSE JSE Free Float All Share

Rolling 1-year performance

-8%

-6%

-4%

-2%

0%

2%

4%

6%

8%

-8%

-6%

-4%

-2%

0%

2%

4%

6%

8%

Nov

-06

Feb-

07

May

-07

Aug

-07

Nov

-07

Feb-

08

May

-08

Aug

-08

Nov

-08

Feb-

09

May

-09

Aug

-09

Nov

-09

Feb-

10

May

-10

Aug

-10

Nov

-10

Feb-

11

May

-11

Aug

-11

Nov

-11

Feb-

12

May

-12

Aug

-12

Nov

-12

Feb-

13

May

-13

Aug

-13

Nov

-13

Feb-

14

May

-14

Aug

-14

Nov

-14

SWIX-ALSI

ALEXANDER FORBES

18

What is the objective of the accumulation strategy?a. How does the asset allocation achieve the

objectives?b. Why are the underlying benchmarks selected?

What is the objective of the final portfolio?a. What risks are managed explicitly? And what are not?b. How long is the member in this strategy?c. How does the asset allocation achieve the

objectives?d. Why are the underlying benchmarks selected?e. How effective is the strategy under stressed

conditions? Probability of failure? Note: Back-testing based on historical conditions is not sufficient.

How does the de-risking process work?a. When does the de-risking process start?b. How do all the portfolios interact during the

de-risking process?c. How robust is the process behind setting the

process?

How robust is the technology and operational processes supporting the solution?

Similarly, every time a portfolio strays from its strategic asset allocation, a decision needs to be made on when to rebalance. But rebalancing costs money. Rebalancing too often can be costly, while rebalancing too infrequently will cause the portfolio to start deviating from its objective.

Step three | Understand the output

Whatever type of investment strategy one is evaluating, the temptation is to go straight to performance numbers even though all such strategies come with their standard warning: “past performance is no indicator of future performance”.

Box 1: Questions to ask about the design of a life stage

42

31

As most accumulation portfolios will hold more than 50% of their portfolio in local equity, choosing between the ALSI and the SWIX could have resulted in at least a 2% performance difference over one year at the end of 2014. The view is that these divergences should converge over time, but the timing of a member’s retirement date is not usually something we can control.

Similarly, in global equity, benchmarks vary greatly in their exposure to developed versus emerging markets. The misleadingly-named MSCI World Index only provides exposure to developed markets, while the MSCI All Countries Index covers both emerging and developed markets. Clearly, given the divergence between performance in developed and emerging markets, and their varying risk profiles and return premiums, the choice of a global benchmark can also make a big difference.

Bottom line: comparing passive portfolio performances can be hugely problematic if the underlying solutions are not applying the same benchmarks. Performance differentials in indices can far outweigh performance differentials from fee structures.

Only once a strategic asset allocation is set and the benchmark chosen does the choice of active versus passive come into the picture. Both active and passive are simply an implementation method for an investment strategy. Neither choice can save a member who has been invested with a bad strategic asset allocation or a poorly chosen benchmark.

Understanding monitoring and rebalancingThis final step in the process can materially affect the outcomes if executed poorly. It becomes even more material in the case of a passive investment strategy. If the main selling point of passive is its low costs, badly managed operational elements can have a material impact on costs.

How an investor’s cash moves in and out of a passive portfolio can produce significant tracking error outcomes if not managed fluidly.


19

555.46

476.39

438.65

379.42

340.18

75.00 100.00 125.00 150.00 175.00 200.00 225.00 250.00 275.00 300.00 325.00 350.00 375.00 400.00 425.00 450.00 475.00 500.00 525.00 550.00 575.00 600.00

Feb-

04

May

-04

Aug

-04

Nov

-04

Feb-

05

May

-05

Aug

-05

Nov

-05

Feb-

06

May

-06

Aug

-06

Nov

-06

Feb-

07

May

-07

Aug

-07

Nov

-07

Feb-

08

May

-08

Aug

-08

Nov

-08

Feb-

09

May

-09

Aug

-09

Nov

-09

Feb-

10

May

-10

Aug

-10

Nov

-10

Feb-

11

May

-11

Aug

-11

Nov

-11

Feb-

12

May

-12

Aug

-12

Nov

-12

Feb-

13

May

-13

Aug

-13

Nov

-13

Feb-

14

May

-14

Aug

-14

Nov

-14

Cumulative perfromance (Based on R100 investment)

AF Passive Bold AF Passsive Bold Moderate AF Passsive Moderate AF Passsive Moderate Cautious AF Passive Cautious

0.00%

5.00%

10.00%

15.00%

20.00%

25.00%

1 year 3 years 5 years 10 years

AF Passive Bold AF Passsive Bold Moderate

AF Passsive Moderate AF Passsive Moderate Cautious

AF Passive Cautious

Historical returns

Be careful of basing your decisions purely on graphs like these.

Despite their ubiquity, this critical warning often goes ignored. It is warning that historical and back-tested performances, as well as any other historically calculated metrics, are insufficient to draw robust conclusions. The past may look nothing like the future, so a much wider range of scenarios

needs to be tested – this is exactly the point underscored by the financial crisis where many financial models had neglected the full range of outcomes. Trustees need to be wary of basing investment strategy decisions purely on graphs like those in Figure 2:

Figure 2: Returns of the Alexander Forbes passive portfolios to December 2014


ALEXANDER FORBES

20

Box 2: The four primary dangers of focusing on historical returns

Historical returns are based on a very small sample set. Historical returns represent one of an infinite set of outcomes that could have transpired. Even with 50 years of data, the proof is not significant.

Historical returns ignore investment horizons. It is easier to analyse funds over long timelines, but depending on how many portfolios are used in a life stage process, a member may be in a specific portfolio only for a short period. This means that the design needs to be robust over shorter time horizons, which is often a much more challenging task.

Historical returns are not related to objectives. Remember our example earlier of Usain Bolt – good in sprints, poor in marathons. Instead of using annualised risk or returns, tests need to be done to determine the likely success of meeting specific objectives. The historical probability of failure is only one indicator. Much more needs to be done.

Historical returns do not effectively capture the portfolio’s prospective risk. There have been significant advances in risk management over the last decade, particularly for South African assets. Modern risk systems provide insight into how much risk the portfolio is exposed to any given point, given prevailing market conditions.

Comparing any two investment strategies, irrespective of whether they are passive, active or smart beta strategies, is flawed when looked at solely from the perspective of historical returns. It gets even more complicated when looking at life stage solutions.

The success or failure of a life stage solution is only clear over a very long time period. Additionally, most members will change funds and strategies through their lives, so establishing the best way to build such a strategy cannot

be based on historical or idealised conditions. Importantly, the life stage design that had the highest probability of success at the outset may not in the fullness of time actually have provided the best outcome in terms of performance, but remember at the outset it still gave you the highest probability of reaching your goal, which is what you are aiming for.

Evaluating a passive strategy rests on the process, not on historical returns.

1 3

24

Evaluating a passive strategy rests on the process, not on historical returns.


21

Everything should be made as simple as possible but no simpler

The great challenge in assessing investment strategies is to recognise that sometimes the simplest way of assessing the value of a strategy is not necessarily the best – in fact, it can be decidedly misleading.

Performance comparisons are our natural default mode – but the true test of whether one strategy is better than another requires that we first assess the likelihood that a given strategy can actually meet its intended goal. It takes a bit of extra effort to make this analysis but essentially there is no alternative.

In the final assessment, some passive strategies will be more dependably structured than others for meeting a particular funding target. The same can be said of active strategies. In our 2012 Hot Topics investment seminar, we spoke at length about the continuum of performance certainty that can be drawn around investment strategies. That continuum took us from passive through to active specialist and finally on to active balanced strategies as a way of illustrating the likelihood of an increase around performance

variability to our required outcomes. In this Hot Topics, we continue the debate by adding a methodology that allows us to forecast the outcome variability that can come from how the solutions’ asset strategies and asset classes have been structured in reference to that target.

Our advice: When faced with a plethora of marketing material, make life simpler. Avoid performance comparisons – and most importantly ask the questions in Box 1!

In the case of our passive versus passive debate, ensure that you are comfortable with the process behind a passive strategy. Simply because it is passive does not mean it is better, or simpler, or cheaper. Interrogate the process to make sure it has been built rigorously to provide your members with the greatest chance of meeting the fund’s objective.

Whatever proponents of active or passive may say, the suitability of a strategy to your own member requirements lies in how it is constructed, and this needs to be where the focus for evaluation lies.

ALEXANDER FORBES

22

Why are we still locked in a performance race if goals-based investing is so powerful in reinforcing all the right behaviours in both our members and our fiduciaries?

Investing towards a goal – thoughts for retirement funds

Robert Merton, the Nobel Laureate in finance, is emphatic about one point when it comes to retirement investing: “Sustainable income flow, and not the stock of wealth, is the objective that counts for retirement planning.”

As a fiduciary on a retirement fund, which option is preferable to you?

1. An investment strategy that has a relatively high probability of providing the income replacement that your members require when they retire?

2. An investment strategy that either could equally provide high, inflation-beating, competitor-beating returns or could produce inadequate outcomes against either those benchmarks. More importantly, it could fail to provide the necessary income replacement.

Chances are that, while you intuitively understand that option 1 should be the preferred option, you also know that real or perceived performance pressures or lack of clarity around fund objectives will often lead a board of trustees to opt for option 2.

The interesting part is that, when the defined benefit model dominated, option 1 was the ubiquitous choice for trustees. Here the objective was clear and codified in the rules. The question is therefore whether the shift to defined contribution funds should have changed investment strategies – or whether it was an unintended consequence.

Asset management can be a powerful tool. Give an asset manager a target, a time frame, and the degree of certainty required of meeting that target, and a fund manager can generally deliver a reasonable outcome. Turn that requirement into a performance race, by demanding that the fund manager outperform their competitors in perpetuity, and outcomes (both in absolute returns and relative to a specific goal) become decidedly more random.

Bottom line: as an investment strategy, a goals-based framework has a much higher probability of fulfilling client needs by way of meeting an objective, than a performance-based framework. If the focus of retirement funds is to ensure that individuals have an adequate replacement for their income when they cease working, can we really afford to gamble?

What is goals-based investing? Goals-based investing differs from traditional investing in three critical ways:

■ It defines risk as being the risk of an investor falling short of their goal, not as volatility in returns.

■ It employs investment strategies and asset classes that increase the probability of meeting those specific targets as opposed to simply beating an asset mix or peer group benchmark.

■ Performance monitoring and measurement is done in reference to whether the fund or member is on target for meeting their funding requirement.

This sounds sensible, but why are we still locked in a performance race if goals-based investing is so powerful in reinforcing all the right behaviours in both our members and our fiduciaries?

How did performance rankings get so entrenched?Essentially, the shift towards a performance-driven investment model that has occurred over the last 30 years owes much to the technological limitations that made it infeasible to provide customised solutions to a massively expanded consumer base. The rise of the unit trust industry and the shift for employees from a defined benefit to a defined contribution model contributed to an explosion in the democratisation of investing to individual investors. Performance comparisons between managers became the easiest metric on which to base one’s choice of investment portfolio simply because it was easyto communicate, and the information was readily available.

Today, technology has evolved to the point where the industry can offer cost-effective solutions that both dynamically assess asset-liability shifts and then create the right investment mix to provide the best chance of meeting those funding goals over the required time frame. Ultimately, the more individualised the solution, the better the likely outcome.


23

Perhaps, though, the real reason the goals-based concept has battled to gain momentum is that over the last 12 years markets have continued to set new performance records, in spite of the occasional blip (think Global Financial Crisis of 2008). As long as markets continue to power on, the pressure to change the model of ‘top performance as a meaningful measure of success’ is unlikely to change. But the more investors begin to appreciate that fund manager track record successes have very little to do with the investor’s personal financial success, the greater the pressure will be on the industry to change to a model that is more outcomes-based focused.

How would goals-based investing strategies work with retirement funds?Retirement funds provide a natural candidate for the goals-based investment approach. Many trustees may be thinking this is exactly what they are doing now. But let’s put conventional retirement fund investing to the test to see if it properly adheres to a goals-based approach. 1. Have you got defined goals for your fund or its members?

What is the objective of a retirement fund? To provide the highest fund credit possible? To ensure capital protection at all times? Actually, the Pension Funds Act is fairly specific as it defines pension fund organisation to mean:

“(a) any association of persons established with the object of providing annuities or lump sum payments for members or former members of such association upon their reaching retirement dates, or for the dependants of such members or former members upon the death of such members; or...”

Regulation 28 to the Pension Funds Act deals with investment strategies, and says that:“A fund has a fiduciary duty to act in the best interest of its members whose benefits depend on the responsible management of fund assets. This duty supports the adoption of a responsible investment approach to deploying capital into markets that will earn adequate risk adjusted returns suitable for the fund’s specific member profile, liquidity needs and liabilities…”

“Principles (2) (c) A fund and its board must at all times apply the following principles: (iv) ensure that the fund’s assets are appropriate for its liabilities;…”

Most trustees probably feel they have met these requirements. Generally, in defined contribution funds, the goal has been defined as the member’s ability to replace a percentage (often 75%) of their final income on retirement.

It’s not a bad start, but we need to take it further by directly translating this goal into action by the strategies we adopt.

This approach also has a few problems: ■ Not all your members will be able to live on 75% of their final income. Some may need less (for example, those with spouses whose earnings help close any gaps or who have other sources of income) and others who may need much more (single mothers who had children late in their lives and are therefore having to support children into retirement, and so-called sandwich generation retirees who have dependants who are both older and younger than they are).

■ The challenge that Treating Customers Fairly poses to trustees is that they need to ensure that the product they are offering to members is appropriate and meets the needs of the individual customers (members). As such, the more diverse the member population, the more difficult it is to meet this challenge.

■ An individual’s replacement ratio requirements will typically change over the course of their lives, making it a moving target.

Importantly, a truly goals-based strategy is about maximising the probability of achieving the targeted replacement ratio whilst retaining a balance with other protection. And not what ratio is actually achieved as this will only be known afterwards.

ALEXANDER FORBES

24

2. Have you structured your solutions right to meet those goals? Trustees may consider whether they have selected the correct asset management strategy. However, we have highlighted at previous Hot Topics sessions that getting the structure right involves the full value chain: ■ Did the sponsor (employer, unions or individual) establish the right parameters to meet the objectives (in other words, promoting transfer of employee’s fund credits from the previous employer, setting a reasonable pensionable pay, contribution rate, and retirement age, and offering flexibility in benefit offering)?

■ Have the trustees determined the right parameters for monitoring the journey?

■ Do the communication strategies to members provide clarity around what decisions they are accountable for and what choices are in their best interest?

Only once we can tick all those boxes does selecting the right investment strategy become truly meaningful.

3. Have you identified risk in a meaningful way? Some trustees think about risk as volatility of performance only. This manifests in several ways. Risk numbers for investment reporting are typically little more than measures of the volatility (returns relative to some benchmark) of an investment strategy. This is the typical risk versus return efficient frontier that trustees may be familiar with, so-called Modern Portfolio Theory. The problem is that this efficient frontier has little to do directly with the underlying target of a particular retirement income. Risk questionnaires, often used in financial planning or advice, relating to member choice also typically try to determine the level of aggressiveness a member can tolerate in performance volatility. Neither helps trustees with the discussion around how much risk is required to meet members’ goals. Nor do they help trustees structure an investment solution that has the highest probability of meeting a given target. Consider two parameters: ■ An indication of how much shortfall can be tolerated in meeting members’ goals. In this context, risk is defined as the risk of not meeting the member’s liability.

■ A risk budgeting exercise that helps control which investment decisions should direct the performance outcomes. This would be predicated on which decisions have the highest probability of adding value.

4. Have you identified the investment or employee benefit strategy that has the highest probability of meeting your member’s needs? The key here is rigorously testing for probability of success on both these parameters.

Strategies such as traditional life stage investment, whilst certainly at some levels leaning towards a goals-based philosophy, are capable of doing a limited amount on behalf of members. The strategy is only effective if members have a reasonable level of fund credit at the beginning of the pre-retirement glide path that secures their future retirement income. In addition, many life stage strategies require a degree of homogeneity around members’ post-retirement investment requirements. These are rarely the case and often trustees may fail to appreciate whether the life stage strategy they have selected is really the most appropriate for their membership demographic. As such, our evolutionary journey into the future must take these limitations into account. To increase our probability of success, we need to consider how we can improve our life stage approaches and ultimately better individualise the strategies for members. The strategy that has the highest probability of meeting our differentiated member goals is one that will recognise both the different start and end points of members as well as constantly realign movements in the markets and the member’s liabilities.

5. Have you identified the appropriate building blocks to tackle a goals-based approach? A retirement funding goals-based approach may require a range of building blocks: ■ One that maximises long-term growth ■ One that maximises income replacement ■ One that protects against downside loss ■ One that provides growth over the short term (to be used just before and after retirement).

This is a discussion around the mandate intent of each strategy and whether it is adequately structured to deliver on what is required.


25

6. Are we monitoring success correctly?Becausse monitoring is perhaps the most important part of a goals-based process, it should be given due attention.

We have stated that there are two retirement related goals that are meaningful to the member: ■ Do they have adequate means to replace the income they are earning when they retire sustainably?

■ Do they have adequate protection for their income and well-being over the course of their life?

Are members retiring with enough?Simplistically, we could easily measure the first goal by identifying the replacement ratio that members actually end up retiring on. The problem here of course is that, by then, it’s too late to rectify anything. The Alexander Forbes latest Member Watch database suggests that actual retirees have achieved a replacement ratio of 32.8% on pensionable salary from their funds – a figure which is consistent with other findings in the industry. It would be quite a shock to members to find out how little of their earnings they were replacing at retirement age.

Members should be provided with annual statements that inform them of where they are in their journey to that replacement ratio objective, whether they are still on target for success, and if not what could they do that could improve the long-term outcome before it’s too late.

But an even more successful monitoring strategy is one that acknowledges that both the trustees and the members play a role in influencing the final outcomes. In addition to providing members with insights into what they could do to influence outcomes, trustees need a clear understanding of how their decisions impact on outcomes.

LifeGauge provides a superb stress testing and monitoring framework which demonstrates how any trustee, member or employer decisions – from costs and preservation decisions, to investment strategies, to contribution, pensionable pay parameters, and more – impact on each member. Because each member is at a different point in their journey and has a different fund credit base, the impact of each decision (and even market impacts) can vary dramatically between members.

Is your investment strategy performing as required to meet the funding requirements?Monitoring this aspect of the process means we have to move away from the traditional way that manager performance is measured – in other words, did the manager beat some asset class benchmark? This may be of interest, and does have some relevance to trustees in ensuring the managers do their jobs but it has little to do with the actual objective of the fund.

In goals-based investing, what we need to be measuring is whether each strategy element is actually delivering as required:

■ Is the income securing strategy effectively mirroring the cost of procuring an income?

■ Is the capital protection strategy protecting us from downside risk?

■ Is our short-term growth strategy showing enough volatility control to be viable over the short term?

■ Is our long-term growth portfolio achieving the growth required relative to salary inflation in order for our contribution rates to have a reasonable chance of meeting our targeted replacement ratio?

These are clearly very different metrics to monitor, but they are not difficult to measure. What is required is mindset change on how fiduciaries think about their investments, select their investments and measure their investment solutions.

The Alexander Forbes latest Member Watch database suggests that actual retirees have achieved a replacement ratio of 32.8% on pensionable salary from their funds.

ALEXANDER FORBES

26

In summary To get goals-based investing right, which is in members’ interests, there needs to be a change of the norms in the industry, both in asset management and benefit consulting.

What needs to happen?

■ The advisory and asset management framework must change in tandem. It’s not enough to target certain goals and then simply place assets into performance-based pooled portfolios with the hope that these strategies might deliver what’s required. The investment vehicles must change so that targets have a higher probability of being met.

■ Benefit design considerations need to move beyond an approach that simply bases benefit structures on what everyone else is doing.

■ The regulatory environment needs to support these types of solutions and not confound their delivery. While TCF rightly requires us to put members needs first and ensure that investments are appropriate, asset class constraints, risk questionnaire requirements and performance reporting guidelines that service the conventional performance-based model only are problematic.

■ Performance measurement needs to change. Technological innovations now mean that we can provide every member cost-effectively with some notion on whether they are meeting their retirement income investment goals and, if not, what types of interventions could take place.

■ The place and time for an effective digital delivery of the above solutions has come. Digital solutions to engage members personally are likely to become a priority in the near future.

Overall, the industry needs to rethink how they are helping their clients set the right expectations and then manage continuously those expectations.

When all is said and done, hope is not an optimal strategy – whether we are dealing with investment performance or protection for our members. Goals-based strategies say: “Let’s take the randomness out of the outcomes. Let’s focus on whether our members are getting where they need to go.”

And in case you feel that that this is the first time you have heard about goals-based investing, we take you back to Hot Topics March 2014. In that workbook, we were looking at life stage strategies and made the following points, still hugely relevant in this discussion:

“What we can say is that in a risk cognisant world, developing our investment strategy in such a way that it is at inception optimised to deliver our key objective (such as a replacement ratio) with as

high a probability as we can, and then refining this over time to improved techniques, is more than intuitively correct – it is, based on the modelling and the inputs at the time, the most appropriate strategy for that objective because it is optimal. This does not guarantee that in hindsight, a slightly different structure could not have delivered a marginally better result. It is all about maximising the probability of success.”

An important takeaway here is that we are not determining success by comparing the various funds’ returns against each other, nor against market related benchmarks or other unrelated strategies. Instead we are determining the success of the strategy by comparing the experience to the stated objectives.

To get goals-based investing right, which is in members’ interests, there needs to be a change of the norms in the industry.


27 27

It is one thing to argue for a goals-based investment strategy, it’s another thing altogether to know how to implement it. The challenge facing boards of trustees throughout the world is whether their selected investment strategies can achieve the fund’s objectives within acceptable levels of risk.

This challenge became particularly relevant over the last few years when declining real yields saw pension liabilities grow faster than their assets. Many defined benefit (DB) funds went further into deficit with some high-profile failures such as the City of Detroit in the United States.

Members of defined contribution (DC) funds were not immune to this problem either. Retirees in this period received lower than expected monthly incomes when converting their pension savings. Once again, the culprit here was low yields.

What went wrong? 1. Real rates moved to historical lows. This is a major input

in calculating annuity payouts.2. Investment strategies focused largely on capital risks in

the equity market and ignored interest rate risks.3. Investment strategies did not incorporate the kind of

asset classes or asset strategies that could provide a targeted income replacement outcome.

Many blame the financial crisis and the subsequent quantitative easing (QE) programmes that kept interest rates artificially low for creating the existing investment environment. This may have been the catalyst, but it would be myopic to assume that this market structure is temporary. These programmes are not yet out of the system and may well remain a reality for many years to come. Japan and Europe, for example, have initiated new QE programmes while some may term the new infrastructure development initiative in China as a form of QE.

But funding risk isn’t the only concern for trustees. Gradually, investors are beginning to appreciate that downside market events are beginning to take place with far greater frequency and severity than statistical distributions assume. One has only to look at the continuous new highs being set by markets globally to recognise that capital market risk could be an equally important concern. At some point the pundits will be right: equity markets will be overpriced and investors will realise it.

The real problem, though, is that returns over the past 12 years have been so strong that it would be easy to lose sight of the real risks being taken. So how does a board of trustees navigate these tumultuous waters?

Sneak peek into demystifying risk budgets

ALEXANDER FORBES

28

How could we more effectively address risk?“The role of the trustees is to understand the nature of the fund and its liabilities, consider the investment objectives of the fund’s various stakeholders, gain an understanding of the various investment risks that require management and their inter-relationships, as well as the investment strategy and portfolio management techniques used to manage these risks while seeking to enhance returns, and ultimately to develop an investment strategy appropriate to the fund.”

For many retirement funds, a discussion of risk is often a secondary consideration. And yet, as Benjamin Graham, the father of value investing, so eloquently put it: “The essence of investment management is the management of risks not the management of returns”. Effectively, if one concentrates on managing the risk of falling short of your target, instead of chasing performance, your probability of success increases significantly.

Risk management is paramount in any successful investment solution where we require a clear insight and accountability on what is driving performance outcomes. That said, for many funds the investment strategy does not explicitly define the process of integrating risk into its investment strategy framework.

The closest many funds get to this is by conducting an asset-liability modelling (ALM) exercise, which sets the asset allocation that forms the basis of the investment strategy.

However, the ALM process takes place infrequently (annually or even longer), whilst market risk needs to be monitored more frequently, preferably monthly for a pension fund. Further, an ALM process is a probabilistic approach looking at the investment strategy problem over a long periods (>5 years), with assumptions that are statistically smoothed. This has the effect of underestimating actual market risks which are much more severe over shorter periods. As such, the ALM process is a valuable starting point for determining the strategic asset allocation for the fund, but is deficient in measuring and monitoring risk.

Why do funds need a risk budget?An effective mechanism to complement the ALM exercise in the ongoing management of the fund is to construct a risk budget. A risk budget is similar to a financial expenses budget, as illustrated in the table below. The decisions are as follows:

Household expense budget

Investment risk budget

How much is available to spend?

How much risk can be taken?

What can it be spent on? Where should the risk be taken?

What is the expected value from this expenditure?

What is the expected additional return or protection?

We know that to generate additional return, the investment strategy needs to assume risk. But what kind of risks should be taken and what is the right level of risk?

Taking too much risk could generate unwanted (negative) outcomes. But risk has a duality: not taking sufficient risk can lead to an overall failure of the strategy. The key investment risk types that the trustees need to consider in a risk budgeting framework are:

■ Not meeting projected liabilities ■ Losing capital ■ Not generating real growth ■ Strategies that take too much or too little risk to meet goals

■ Benchmark effectiveness


29

The essence of risk budgetingA risk budget is meant to be a pragmatic yet measurable framework. Risk budgeting is a process of measuring, allocating and controlling risk of the individual components (asset classes, individual funds, strategies, and even instrument types) in the context of maintaining an overall risk level for the total portfolio. These measures need to be forward-looking risk metrics to control future outcomes. Historical risk metrics can’t satisfy this effectively. It’s a bit like driving while looking in the rear view mirror. It only works if you are going backwards.

To successfully achieve this, a risk budgeting process seeks answers to the following questions:

■ What types of risks are needed or can be assumed by the fund?

■ What are the horizons of these risks? Short, medium or long term?

■ Which risk criteria are relevant for the fund? How are they measured?

■ What is the tolerance for these measures? ■ How would these measures be monitored and managed? ■ What actions need to take place when these limits are breached?

■ What are the responsibilities to be allocated to authorised individuals?

Sometimes the elephant in the room is: “Why should we change anything when everything is seemingly going well in terms of returns?”. But just because one has not had an accident does not imply that the need for insurance is obsolete.

Many trustees would prefer to take the least risk to achieve the investment objectives. Least risk does not mean no risk! The most significant value of the risk budget process is that it helps establish how much additional return risk is required to meet investment objectives. More importantly, it provides the investment process with a framework for determining which risks are worth taking and which risks should be minimised. Essentially, the framework provides full accountability, transparency and measurability for the value-add of every component of the investment strategy.

The main function of determining the risk budget and principles is to be able to allocate the fund’s various exposures to risk in a manner that maximises the probability of achieving the stated objective for a fund within a given cost structure, whilst adhering to the investment philosophy of the fund.

Each asset class exhibits a range of factors that drive performance outcomes. And each factor exhibits varying degrees of probability of success. Designing the optimal range of strategies or asset managers to use for any given asset class is a function of maximising the high probability performance drivers and minimising the fund’s exposure to low probability drivers within a given risk budget.

There are many key investment decisions, such as ALM, benchmark selection, and strategy choice, made in a fund that have a significant risk impact but they are not observable because these decisions are not measured. When implementing the investment strategy, these decisions have an impact on the efficacy of the solution.

In most South African investment modelling frameworks, the ALM exercise and the core investment strategy will only use the major global and local asset classes (equity, bonds, including inflation-linked bonds, property and cash). However, different portfolio strategies could be used to complement this asset mix: hedge funds, private equity, commodities, and so on. This decision changes the risk to the stated investment objective. Other decisions such as benchmark selection, tactical asset choices, and implementation constraints (liquidity) also change the risk of the investment strategy.

Unfortunately, because most funds measure their risk against the composite broader asset class benchmark, these additional risks will effectively be undetected. The misinterpretation of the quantum of risk the fund actually faces could be significantly understated. What a risk budget allows us to do then is account for all decisions taken across the investment process and provide some measure of their potential impact.

29


30

The power of a well-defined risk budget is that it provides the basis for answering questions such as the optimum allocations to the major asset classes, the optimum allocation between passive and active and the optimal blend of differentiated mandates within an active strategy.

In addition, a well-thought-out risk budget allows you to determine:

■ An appropriate allocation to alternative asset classes ■ Appropriate asset class limits for use in tactical asset allocation

■ Thresholds for a more efficient form of rebalancing ■ Quality of the risk taken across assets, managers and strategies

More importantly, risk budgeting provides a framework incorporating strategies related to broader strategic initiatives of the fund and its members. ESG initiatives, promoting BEE fund managers, and participating in community development projects would be good examples. While it may be unclear exactly what the alpha potential of these strategies might be, the risk budgeting process helps define how these strategies can be integrated into a fund such that any potential downside risk could be controlled and would be tolerable.

Ultimately, the risk budget extends the risk management policy defined in the investment policy statement (IPS) by converting this into a framework that explicitly defines the risk limits, processes, methodologies and calculations to be used to monitor and govern the total and individual

exposures to strategies, funds, asset classes and instruments. This framework is the critical starting point for stress testing the fund against different market scenarios, linking investment performance to decisions (even not making a decision) and understanding how effectively managers and their specific styles are complementing outcomes and risk control. This dramatically simplifies hiring and firing decisions.

A fund’s risk budget is essentially a living document that is incorporated as an addendum to the IPS. This means it needs to be actively reviewed. It provides both the important checks and balances required to deter the temptation to override decisions, and a qualitative and quantitative framework by which trustees can assess every decision in the fund.

A risk budget brings the following significant advantages to a fund:

■ Allows trustees to discuss relevant issues proactively, and may eliminate reactive processes.

■ Provides a more transparent but stricter framework to interrogate the total investment strategy, individual portfolio strategies, managers and implementation decisions.

■ Enables decision making. ■ Provides accurate supporting information for a quicker implementation processes.

■ Strengthens the governance structure. ■ Provides a robust framework for trustees to operate in together with their various advisers and managers.

The illustration below demonstrates the change in risk due to the various investment-related decisions.

Decisions

Cha

nge

in ri

sk

ALMSAA Benchmark selection

Strategy selection

Manager style

Tactical style



31

Some readers might be thinking they have heard this all before. Is goals-based investing not almost the same concept as liability-driven investment (LDI), the buzzword of the pension industry over the last five to ten years?

Broadly speaking, both processes involve the development and monitoring of an investment strategy to meet certain predefined goals. However, LDI strategies have traditionally resided in the defined benefit fund world, where employers needed to derive certainty that they could adequately fund the liability of their member’s post-retirement income requirements. This helped manage the impact on their balance sheets.

As goals-based investing has evolved, though, it has become increasingly clear that LDI investment strategies could provide a critical building block in enabling defined contribution funds to provide more targeted outcomes to individual members.

LDI unpacked LDI asset managers reduce or eliminate unrewarded inflation and interest rate risk for investors with well-defined liabilities or targets. What does this mean? Consider Thabo saving for his retirement in five years. By applying actuarial techniques, we can estimate what salary Thabo will earn across these five years and therefore estimate how much money he will save toward retirement. We also know he doesn’t want to take a large cut to his monthly income when he retires. The trustees of Thabo’s pension fund try to design an investment strategy that will generate a suitable retirement income out of these contributions. Two important risks that could derail this planning (before or after Thabo’s retirement) include:

1. Rising inflation: inflation (increases in the cost of living) could rise rapidly, outpacing investment returns and eroding the real value or purchasing power of Thabo’s income.

2. Lower returns: the return available in the market could fall. Therefore the investment return the trustees have planned on earning for Thabo could erode.

LDI asset managers address these risks for investors.

Investing in assets that guarantee actual inflation and a fixed return until the time Thabo needs each income payment would protect against these risks. An example of such an asset is an inflation-linked bond. Inflation-linked bonds pay a stream of cash flows increased by whatever inflation turns out to be.

An inflation-linked annuity is a second example which would pay Thabo a constant monthly income increased by inflation. The inflation-linked annuity will not be available for purchase by Thabo until he retires. Both assets are guaranteed to keep up with inflation and to deliver a fixed real return in excess of inflation.

Critically, the trustees (or Thabo, after his retirement) need not invest all of his assets into these investments. A portion of his assets could be invested to provide some protection against inflation. Goals-based investing tells us how much Thabo should commit to LDI assets. LDI managers use these concepts to reduce the risk that investors (such as insurers or pension funds) do not keep pace with inflation or earn an adequate return. In fact, technology exists that allows LDI managers to use only a portion of Thabo’s assets to protect his entire portfolio against reductions in available returns. These technologies are not as effective as locking in the required returns with all of his assets, but protect him against most of the risk of reductions in the return available.

These risks are technical and hence many people saving for retirement don’t think about them. They can and do cause major fluctuations in the incomes that pension fund members retiring at different times receive.

We require a different mindset to get our heads around the fact that a negative investment return could still give a higher pension income in retirement if annuity prices reduce by more than that negative return.

The amount and exact type of risk control required depends on the timing and nature of each investor’s targets or liabilities. It has therefore only been economically viable to apply LDI technology to very large investors such as insurance companies promising to pay annuity payments to clients or defined benefit pension funds who have promised to pay inflation-linked pensions to their members. This is likely to change though as providers use automation to apply these technologies to individual DC fund members.

Liability-driven investment in a goals-based world

ALEXANDER FORBES

32

How is LDI being used internationally in DC schemes using goals-based investing?Automation is allowing funds to apply goals-based investing at an individual level to their members. Instead of designing a single investment strategy that would suit the average member, funds can now set replacement ratio or retirement income targets or objectives for their individual members. Globally there is an emerging trend of funds optimising the investment strategy of each member to these targets, developing a mix of assets appropriate to their individual circumstances. For example, a member who chooses a very low contribution rate might justifiably need to take on more investment risk to stand any chance of receiving a reasonable retirement income. Some systems allow further personalisation using member engagement through an automated online process. The member we described might be highly risk averse and may indicate they do not want an aggressive investment

strategy. Instead the member might be satisfied with a limited retirement income.

LDI is used within this framework of targeting an income-related goal to manage risk. LDI makes up one of the asset classes or building blocks available to these investment strategies. More specifically, it is the lowest risk asset class in the context of an income target, much like cash would be the lowest risk if your aim was not to make any capital losses.

The bulk of the value-add in these systems is likely to flow from the individualisation (consideration of each member’s financial circumstances) and the use of appropriate goals, such as retirement income. The use of LDI assets adds a further (marginal) increase in overall investment efficiency relative to the use of conventional bond portfolios.

90%

95%

100%

105%

110%

1 N

ovem

ber 2

012

1 Ja

nuar

y 20

13

1 M

arch

201

3

1 M

ay 2

013

1 Ju

ly 2

013

1 S

epte

mbe

r 201

3

1 N

ovem

ber 2

013

1 Ja

nuar

y 20

14

1 M

arch

201

4

1 M

ay 2

014

1 Ju

ly 2

014

1 S

epte

mbe

r 201

4

1 N

ovem

ber 2

014

Cumulative outperformance of liabilities by LDI industry

90%

95%

100%

105%

110%

1 N

ovem

ber 2

012

1 Ja

nuar

y 20

13

1 M

arch

201

3

1 M

ay 2

013

1 Ju

ly 2

013

1 S

epte

mbe

r 201

3

1 N

ovem

ber 2

013

1 Ja

nuar

y 20

14

1 M

arch

201

4

1 M

ay 2

014

1 Ju

ly 2

014

1 S

epte

mbe

r 201

4

1 N

ovem

ber 2

014

Cumulative outperformance of liabilities by LDI industry

The chart below shows how various asset managers in the LDI industry have outperformed their benchmark liabilities. For each manager, we have assumed they started with a portfolio equal in size to their liability. We then plot their cumulative outperformance of liabilities. Across the industry we see that all composites have outperformed their liabilities despite this covering an incredibly volatile time for interest rate markets. We also see that LDI does not necessary preclude an investor from earning alpha (outperformance), though the primary purpose of these portfolios is to track liability movements. In general, the larger the alpha earned the greater the volatility of the outperformance observed. In all cases we see good downside protection relative to liabilities.

Figure 1: One-month outperformance figures

Source: Alexander Forbes Manager Watch LDI Survey, Alexander Forbes calculation


33

What can LDI teach us about the successful application of goals-based investing in South Africa? Although LDI is an asset management style and goals-based investing an investment framework, the two have a lot in common. Both are technically complex concepts that most individuals may struggle to understand or value. Both deal with an intuitive concept (investing relative to your aim or goal) but both are notoriously difficult to measure or compare in a simple way.

Unfortunately, the take-up of LDI in the DB pension market in South Africa was slow, spanning a number of years. The greatest challenge was the technical complexity of the solution and the lack of complete risk elimination. However, these are two characteristics shared with goals-based investing. LDI taught us some useful lessons that will apply to goals-based investing:

■ It took time for trustees to become comfortable with LDI and it is likely that goals-based investing will be no different. A core of progressive boards with large governance budgets adopted this technology early on, but many boards preferred to wait for the technology to be perceived as more mainstream before taking it

on board. ■ Reliable, comparable return and risk figures were important to most trustees. Early attempts to measure performance were fraught with problems. Inappropriate performance comparisons surfaced, which many trustees used due to the apparent certainty and simplicity of these measures.

■ As with any highly technical product, it was easy for managers to manipulate clients’ views of their skill and the value of this technology and the role it would play in their broader portfolio.

■ Like any other investment management style, performance figures alone are not sufficient to choose a manager. An understanding of what drove that performance, an understanding of the manager’s processes, people and skills and an understanding of what the investor is truly after are required.

■ Obtaining specialist advice early on in the investment process is in members’ best interests and can save a fund from misaligned expectation or poor service delivery. The ultimate provider of a product should not be entrusted with objectively evaluating their own product relative to the market.

What LDI won’t teach usVirtually all goals-based investing relies on a process known as stochastic simulation. Stochastic simulation looks at thousands (or millions) of possible outcomes an investor can face across the rest of their life, calculating the retirement income or other objectives they could receive in each. This allows one to control risk effectively for an investor.

It is important to remember that a goals-based investment strategy is chosen in advance, based on forecasts rather than actual results. The strategy chosen is therefore the right or wrong choice on a prospective basis, not based on whether it outperforms another strategy across a specific period in hindsight. In fact, there is very little we can learn about the strategy we implemented, relative to the objective we set by monitoring performance against traditional benchmarks.

This is not to say you set the strategy and walk away. The same process needs to be repeated to tweak the direction to maintain the optimal strategy.

Monitoring performance of the underlying asset managers being relative to their benchmarks remains important though to evaluate the value these add or detract. The same principles apply as are currently applied in this regard. Within a goals-based investment framework, the principles used to choose between active and passive management or to select a style of active management would also still apply as before.

For example, Sarah might know that she would like to target a retirement income of R10 000 a month, but she also knows she cannot survive on less than R5 000 a month. Stochastic simulation allows a provider to develop the best strategy that meets both of these requirements. For example, the strategy might have an expected average income of R10 000 a month and a negligible probability of an income below R5 000 a month.

ALEXANDER FORBES

34

Different goals-based investing objectives exist. Some approaches aim to minimise the probability you retire with an income below a certain level. Other frameworks aim to prioritise spending objectives (such as current spending, retirement income and healthcare spending) and take on risk aligned with the value of each objective. How does one compare these frameworks and ultimately choose one? NOT by comparing historical investment performance. This would be like choosing between medical and motor insurance based on which you have successfully claimed from in the past. It would be far more useful to consider whether you can afford the premiums for each and whether you could financially tolerate the loss of your car or expensive medical procedures. Trustees should consider what objectives they believe would suit their membership.

In doing so, the following will play an important role: ■ The needs of their members. ■ Philosophical views held by the board might inform one objective being deemed ‘more appropriate’ as a target for members’ savings. For example, should members who are deeply underfunded be directed to ‘go for broke’ or preserve the (inadequate) retirement income they can

look forward to? How will global coverage by a larger state pension change this?

■ The financial acumen, or otherwise, of the membership. ■ Data limitations and access will become increasingly important as investment strategies become more individualised. Funds with limited access to their membership and poor existing data may want to retain an ‘average member’ approach to setting investment strategies.

■ Legislation and minimum standards will also become increasingly important. We have already seen this highlighted in the US where strong recommendations in this regard have been made by the Government Accountability Office.

■ Eventually it will be possible to use stochastic modelling to compare different frameworks or objectives and different solutions to these. This is currently not possible. This approach will also remain problematic for as long as these solutions and frameworks remain proprietary products.

Conclusion

Goals-based investing is not a new idea. The concept of designing investment strategies optimally around one’s liabilities or objectives is well established. The technologies available to do so have evolved over time. For example, twenty years ago the use of stochastic asset-liability modelling was not common in pension funds. Today most DC funds either use asset-liability modelling exercises themselves or use portfolio ranges that have been designed using these techniques. The next major innovation in this regard is taking this technology to the individual DC member. Through advances in technology, it will soon be possible to apply a goals-based investment framework to each individual member, accounting for their financial circumstances.

LDI will form part of this solution. LDI can remove or reduce some of the unrewarded risks DC members are exposed to. This includes protecting them against the risk that available investment returns fall in the future or that very high inflation rates erodes the purchasing power of their savings. LDI assets are therefore an effective building block within a goals-based investment framework that should not be ignored.

LDI will also play a more conceptual role in establishing goals-based investing within the minds of DC members. Thanks to the high hedging efficacy of LDI, it will be

possible to offer members soft guarantees – levels of retirement income they are highly unlikely to fall below. This is simply not possible without the use of LDI and individualised investment solutions. Offering something like this is already realistic for members within the last ten years of retirement. Certainty is something members do understand. The same is not true of the more complex and technical benefits such as investment efficiency brought about by using LDI assets.

In truth, attempting to fully guarantee a retirement income would be too costly for most members. The industry will, however, be able to use this technology to limit downside, offering soft guarantees on lower limits to retirement income. This remains useful and could dramatically reduce risk and anxiety for members. It is our belief that members will better understand what level of retirement income they need and the probability associated with reaching that goal. This could be highly effective in a semi-automated, individualised advice framework.

In closing, the goals-based investing framework in our view better manages member expectations in respect of outcomes, better aligns to members’ actual needs through optimisation and contains implicit advice which is suitable and can take account of changing circumstances.


35

Life stage solutions – an example of goals-based investing

Think differentlyInvesting for retirement demands a very specifi c form of investment option. It is as much about securing a meaningful income post-retirement as it is about maximising the ability to build up members’ savings pre-retirement: two objectives that require two very different investment strategies.

Knowing how and when members should be exposed to which investment strategy in the course of their long journey to retirement is the critical test of an effective retirement funding plan.

For most of a member’s savings life, investing in an aggressive growth portfolio is a perfectly viable strategy. But, securing a monthly post-retirement income stream requires the purchase of an annuity. When interest rates fl uctuate, the cost of these annuities also fl uctuates. This means that the amount of income that can be purchased for every rand of capital at the point of retirement may be very different from the amount anticipated in the fi ve to seven years before retirement. How then can investors

stabilise the certainty of that income enough in advance so that they can adequately plan for their retirement requirements?

To change one’s investment strategy from one that is growth oriented towards one that is more liability focused to secure the expected target income requires disciplined action to make the relevant portfolio changes through to retirement. For the average person this is a diffi cult process to manage independently.

Life stage investing is a convenient single-step investment strategy that takes this into consideration. This form of investing is not a new concept and has been applied in South Africa for some time. In fact, in Hot Topics March 2014 we considered whether life stage models had delivered the expected benefi ts and we took a brief glimpse into their future – which we now look to expand.

Life stage investing, in its earliest form, recognised that the ability of an individual to assume risk in their retirement portfolios decreases as they approach their retirement.

■ Understanding retirement

■ Objectives ■ Accumulating assets

■ Contribution levels ■ Risk tolerance ■ Preservation

■ Establishing needs ■ Consumption ■ Managing income, drawdown

■ Protecting assets ■ Growth ■ Unexpected events

Saving for retirement

Many years 5 to 7 years Hopefully many years

■ Focused objectives ■ Changing risk tolerance

■ Changing risk capacity

■ Replacing income ■ Protecting assets ■ Growth

Preparing for retirement

Managingretirement

Retirement

ALEXANDER FORBES

36

How life stage solves this dilemma?A life stage investment strategy is based on a two-stage framework, where the first phase focuses on saving and accumulating assets for retirement. During this early phase the individual has a long time horizon left to retirement and as such they can assume more investment risk which can potentially increase the long-term returns for the individual. An investment portfolio that contains mainly equity and other growth assets would be appropriate during this phase. Disciplined savings behaviour by the individual, such as regular contributions and preserving retirement savings when leaving a pension fund, will contribute hugely to a successful outcome for the life stage investment strategy.

The second phase of the investment strategy focuses on preparing the individual for retirement. During this phase, behavioural research has shown that the average person will tend to become more risk averse and would seek to avoid losses on their investments. The early life stage strategies recognised this change in risk tolerance and reduced the volatility of the portfolio as the individual approached retirement.

Traditionally the strategy would systematically reduce the asset exposure from the ‘more risky’ equity assets to ‘less risky’ fixed income and cash assets as the retirement date approaches, as depicted in the graph below.

The problem with this traditional approach though is that, while cash may provide high capital protection characteristics, it does not adequately protect against inflationary erosion. As such, it is largely uncorrelated to changes in the price of liabilities and annuities.

Clearly, the focus on reducing total risk (as volatility of returns or capital loss) has had several flaws. It assumes that by maintaining the portfolio’s ability to achieve an above inflation return the portfolio should be able to achieve the goal of replacing income on retirement. That said, the recent past has demonstrated that this assumption can fail, as many individuals in South Africa and especially in the UK found that their retirement

savings were woefully inadequate to maintain their standard of living despite seemingly successful investment outcomes.

The reason for this is that the cost of replacing members’ income in the future depends on the expected future real interest rates and yields, exactly the same factors that drive annuity prices. Once members are close to retirement, changes in yields can have a significant impact on their standard of living unless their investment strategy explicitly manages this risk as well. The risk to the future obligations is often ignored as it is generally difficult to measure and communicate.

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

8 years + 7 years 6 years 5 years 4 years 3 years 2 years Final year

Local equities Local property Local nominal bonds Local inflation-linkedbonds

Local cash

Offshore equities Offshore bonds Offshore cash Gold

Figure 1: Traditional shifts in asset allocation



37 37

Evolved thinkingLife stage products are evolving with the recognition that we need to create investment solutions that go beyond the simple requirement of reducing capital market risk. If income replacement more correctly specifies the problem, then the investment strategy to achieve this specific goal is substantially different.

The problem is, standard investment portfolios built by using Modern Portfolio Theory will generally be inadequate as this approach only models volatility as a risk. Additionally, in all likelihood, the time horizon of the strategy may not be aligned to the remaining investment horizon.

In summary, by replacing our traditional portfolio design approach with a goals-based approach, we should be able to see these qualitative and quantitative differences:

■ Portfolios are now constructed in the context of the specific risks associated with the defined objective(s).

■ Time horizon becomes a critical factor. ■ Strategies are selected based on how well they improve the probability of meeting a specific goal.

■ Success is defined in terms of whether the solution is actually meeting its stated goal.

How does this evolved thinking potentially change our current life stage solution?

The accumulation phaseThe life stage solution in the accumulation phase is not materially impacted by a goals-based approach, except for the recognition that investors can have more freedom in the design of their growth portfolios in terms of growth and risk objectives – a return maximisation objective. This strategy can be aligned to the investor’s preferred investment philosophy. Selecting which growth strategy will

produce the best outcomes in the future is impossible to determine in advance regardless of what past performance data may reveal.

Research has shown that members remain invested in strategies that they understand best and the selection of portfolio is driven largely by philosophy and risk tolerance. The relevant points that investors need to appreciate about a growth portfolio are:

■ The investments are primarily in growth assets such as equities and property to take advantage of long-term risk premiums and, as such, is not suitable for short-term investors as over the short term, returns are likely to be volatile with potential losses. These should not be a cause for concern.

■ There are many ‘growth’ investment strategies that can produce the long-term accumulation the members require for their asset wealth growth. These strategies can be implemented in balanced, specialist active and specialist passive approaches.

The de-risking phaseThis is where the most notable changes will take place. There are two critical changes in this phase:

■ Defining an appropriate investment objective• Desired outcome• Over what timeline• For how much risk?

■ Constructing effective goals-based portfolios that recognise the specific risks and time horizon that the goal is exposed to.

Time is a critical resource that the retiree does not have in this phase. Unlike the accumulation phase where market downturns could be waited out, time in this phase is limited.

ALEXANDER FORBES

38

How time can smooth the probability of loss

The graph above was part of an analysis of capital loss characteristics of different risk return profiles. One of the portfolios used assumes a CPI+3% return per annum with low volatility (similar to bonds). Over a one-year period there is a 20% probability of loss, which erodes to just over 2.5% over a four-year period. Another way of interpreting this graph is that in an independent world the risk-of-loss to retirement from 4 years out increases 8 times as you approach 1 year! It is therefore important to know that your solution was created to manage the right risks explicitly for the defined period of time. A risk budgeting framework

would complement a goals-based approach to build and design the investment strategy.

In a risk budgeting framework one would explicitly define the types, levels and horizon of different risks relevant to the investor and strategy. By introducing risk budgeting into the exercise we can assess not only how to allocate risk, but we can exercise control over the risk of the different components in the context of maintaining an overall risk level for the total portfolio.

Time helps to smooth out risks, as depicted in the graph below.

Figure 2: The probability of losing money over different time periods and product development

Risk-of-loss to retirement from 4 years out increases 8 times as you approach 1 year!

Source: Alexander Forbes Research


39

Capital protectionCapital protection strategies seek to protect investors from losses, particularly when markets become stressed. Funds that take advantage of general asset class behaviour under normal market conditions may deliver favourable outcomes over such times but may fail when it’s really needed. In Figure 2 the risk of loss increases as you reach retirement. If the objective of your retirement portfolio was to protect your capital, the real test of the portfolio has to be viewed over rolling 1-year periods.

Historical performance would fail to capture the risk the fund was exposed to as in most instances the risks did not materialise. However, it is important to know that there is a risk management strategy to control the expected level of loss should a critical market event take place. Banks in particular were brought under the spotlight post the 2008 crisis in terms of managing their capital adequacy for these events. This would not be dissimilar for a member who chooses this objective. This is measured through the Conditional Value at Risk metric – a measurement of how much one expects to lose in adverse market conditions.

By combining investment stategies or building blocks, bespoke individual solutions can be created

Accumulation orgrowth strategies

Income replacement Income replacement annuities

Livingannuities

Other

Capital preservation

Capital growth



Managingretirement

Solutions for meeting the de-risking objectives

What makes the de-risking framework complex here is that we effectively have to solve for multiple risks. These are: income replacement, capital preservation and capital growth.

The challenge is to design a strategy that acknowledges that these goals have differential priorities in our lives, but recognises that we still need to address all three to meet our members’ requirements.

Taken by themselves, these distinctly different goals could potentially demand very different solutions as we show below.

Accumulation orgrowth strategies

Income replacement Income replacement annuities

Livingannuities

Other

Capital preservation

Capital growth



Managingretirement

Many years 5 to 7 years Hopefully many years

Retirement

ALEXANDER FORBES

40

0

10

20

30

40

50

60 20

10/1

1/01

2010

/12/

01

2011

/01/

01

2011

/02/

01

2011

/03/

01

2011

/04/

01

2011

/05/

01

2011

/06/

01

2011

/07/

01

2011

/08/

01

2011

/09/

01

2011

/10/

01

2011

/11/

01

2011

/12/

01

2012

/01/

01

2012

/02/

01

2012

/03/

01

2012

/04/

01

2012

/05/

01

2012

/06/

01

2012

/07/

01

2012

/08/

01

2012

/09/

01

2012

/10/

01

2012

/11/

01

2012

/12/

01

2013

/01/

01

2013

/02/

01

2013

/03/

01

2013

/04/

01

2013

/05/

01

2013

/06/

01

CVaR through time

Implemented portfolio Modelled SAA ILB10 ILB30

Bonds Property Equity Foreign cash (Currency)

Foreign bonds Foreign equity

These types of techniques can be used to explicitly control the downside risk outcomes according to the risk budget. In the figure above, we see that all assets (except cash) are exposed to losses in different adverse markets. Diversification of assets is a means to limit but not eliminate losses except through the use of explicit hedging and insurance mechanisms which come at a cost.

Income replacementIncome replacement as an objective measures the ability of the portfolio to protect the expected income stream that the member could create from the accumulated value of the retirement savings. For purposes of building the investment strategy, the strictest and prudent form of this is implied by the expected monthly income that an inflation-linked annuity would pay given the accumulated fund value of a member.

Portfolio solutions can be constructed using different liability structures:

■ An inflation-linked income stream into perpetuity using market real yields

■ Actuarial annuity forecast models (for example, price an inflation-linked annuity for a male retiring at age 63 with the annuity continuing to pay monthly to his spouse for 10 years after his death)

■ Historical inflation models: Using the Barclays South Africa Inflation Linked Index as a proxy

■ Real-world historical model: Using real-world annuity prices

Therefore the goals-based portfolio needs to be built to reduce the variation around this income replacement objective. It also needs to reduce the risk of not meeting the expected outcome.

The ability to eliminate the risk completely would require an LDI strategy but the efficacy of alternative strategies can be tested against different horizons and risk tolerances.

Figure 3: Conditional Value at Risk of different asset types



41

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

8 years + 7 years 6 years 5 years 4 years 3 years 2 years Final year

Local equities Local property Local nominal bonds Local inflation-linkedbonds

Local cash

Offshore equities Offshore bonds Offshore cash Gold

Immediately one would notice that the portfolio strategy also reduces equity exposure as with a traditional strategy but does not move into cash assets but instead to assets that are more related to liability risk, such as inflation-linked bonds, more property and other fixed income instruments.

The portfolio would also need to choose its underlying asset class strategies in line with the objectives. For example, given that the horizon to retirement is very short, long-term alpha strategies such as equity value investing is not ideal due to the certainty of disinvestment by the member.

Figure 4: An asset allocation more focused on multiple objectives


Goals-based solutions are focused on delivering the outcomes that are sought with a higher probability of success with lesser variation. This makes pure performance-based comparisons archaic.

Goals-based life stage products do not need to explicitly target singular objectives. A board of trustees may set multiple objectives for both the accumulation and de-risking portfolios; each objective would be prioritised and set with appropriate risk levels.

For example, we might ask our next generation de-risking portfolio to achieve the following:

■ Protect capital over a two-year cycle, with a 95% certainty of a value at risk of no more than 2.5% over 1 year.

■ Protect the ability to replace income over a two-year cycle with no more than 95% downside variation of 10%.

■ Whilst targeting a 3% real growth rate.

ALEXANDER FORBES

42

A meaningful monitoring frameworkA monitoring framework that only reflects historical performance and volatility numbers is not effective in determining the efficacy of a fund’s objective. Alternative measures need to be established. Probability of success is one of these measures. The reciprocal element of this is measuring the probability of failure.

In the exercise below trustees asked us how such ade-risking portfolio would compare against a popular absolute return portfolio in meeting the trustees declared objectives.

Let’s examine how the two portfolios stand up against multiple measures of goal success:

In this example we compare a retirement life stage goals-based de-risking portfolio against a more traditional absolute return portfolio from 2003 to 2014. There is a material shift between examining the risk-return attributes of the funds towards measuring success (or in this case failure) of the stated objectives.

Although the failure percentages seem small, let’s consider them in the real world of the member experience. In the absolute return portfolio, 1 in 15 members experienced a loss of capital. On top of that, 1 in 11 experienced a loss of expected income. In the next-generation goals-based solution, 1 in 100 members experienced a loss of capital and 1 in 40 experienced a loss of expected income.

That said, against the traditional measure of performance, the absolute return fund certainly looked like it would be the better portfolio. The absolute return fund would in all

likelihood generate more instances of growth greater than 3%.

The question any solutions provider needs to ask, though, is whether the additional return is worth the loss of certainty to your objective if you are 1 to 2 years from retirement.

What this exercise should demonstrate, though, is that historical performance comparisons simply do not provide a strong basis to make meaningful comparisons between products. Strong interrogation of the investment process as described elsewhere in this workbook, as well as how the portfolios are positioned with respect to appropriate risk measures of the objective, is clearly the key pointof assessment.

Failure of objective over 1 year

Objective Absolute return portfolio Goals-based de-risking portfolio

Capital protection 6.50% 0.81%

Income replacement 8.90% 2.40%

Inflation protection 13.82% 13.82%

Growth>3% 16.26% 19.51%

Figure 5: Exploring the historical probability of failing different objectives when comparing two different portfolios



43

Conclusion

Life stage solutions are particularly effective in automatically adjusting members exposure to risk as they approach retirement. The critical question being addressed here is what risks are really important to the member and how can we go beyond conventional de-risking strategies to better address those risks.

While we can identify new approaches that can address these goals more effectively, the critical

issue is that we can only begin to appreciate this added value if we learn how to assess that value. Conventional performance measures simply don’t provide that answer. There’s a weaning process that will be required here no doubt. But if we can shift our view to whether the member is winning rather than focusing on whether the fund is beating its peers, we believe we can add significantly more value to members’ lives.

ALEXANDER FORBES

44

National Treasury has introduced legislation from March 2015 that will allow discretionary tax-free savings for individuals, in order to encourage savings in South Africa. These accounts, also called tax-free savings accounts, will allow individuals to contribute up to R30 000 per year, subject to a lifetime limit of R500 000, into a variety of savings accounts that will receive preferential tax treatment of the investment returns.

These savings accounts are exempt from the following taxes:

■ Income tax on interest ■ Capital gains tax ■ Dividend withholding tax

but contributions are from after-tax income.

Furthermore, these savings accounts have been introduced with the principles of simplicity, transparency and suitability being paramount.

While the introduction of these accounts is no doubt a step in the right direction, there are a couple of questions you may be asking. Who benefits from these accounts? What are some of the considerations in taking advantage of the tax benefits? What should these accounts be used for?

Let’s look at each of these questions in turn.

Who benefits from these accounts?The initial starting point is the approximately 5 million individuals who have to submit a tax return (although in reality only about 3.5 million would gain a tax advantage as they pay the vast majority of the personal income tax). Those individuals who are currently putting aside discretionary savings which are subject to tax will feel the benefits by redirecting their future savings into these accounts. To consider the overall benefits of these accounts, one needs to allow for the variety of tax exemptions and deductions that are currently offered to individuals.

What are some of the considerations in taking advantage of the tax benefits?Given that these exemptions and deductions already existed to a certain degree, how would this new account provide an incentive that was not previously there? Individuals under the age of 65 are currently allowed R23 800 in interest income tax-free per year. This will no longer be increased in line with inflation and will effectively be replaced in time by the tax-free savings account. Individuals are also currently exempt from paying tax on the first R30 000 of any capital gains. Comparing the level of these exemptions to the annual limit available in the tax-free savings account of R30 000 it is apparent that the account isn’t suitable for all types of savings a person may have (at least in the short term).

Let’s assume an investor opens a tax-free savings account and invests entirely in an appropriate equity portfolio. Furthermore, we assume the investor contributes their full R30 000 allowance each year to this account and has no other investments using up their capital gains and interest allowances. If equities were to provide a return similar to the average return experienced by equities over the past 10 years then the benefits of the tax saved with respect to the capital gains on the investment will only begin after 5 years. A smaller investment of R1 000 a month pushes the length of time until the tax benefits are experienced out to 8 years. The length of time before the benefits of an interest income account would be an even longer wait, although this will reduce as the exemption is effectively phased out as the limit will no longer be increased. Ignoring the fact that this is typically the advice on the minimum investment term for an equity investment anyway, this also shows that these accounts are best suited for a longer term investment.

Of course, if one is already using the existing exemptions, immediate benefits are apparent.

Giving tax-free savings a purpose


45

R 500 000

R 750 000

R 1000 000

R 1250 000

R 1500 000

R 1750 000

R 2000 000

R 2250 000

R 2500 000

Year 10 Year 12 Year 15

Effect of tax on long-term investment

Typical multi-asset class fund (40% marginal tax-rate)

Typical multi-asset class fund (18% marginal tax-rate) Tax-free multi-asset class fund

R 500 000

R 750 000

R 1000 000

R 1250 000

R 1500 000

R 1750 000

R 2000 000

R 2250 000

R 2500 000

Year 10 Year 12 Year 15

Effect of tax on long-term investment

Typical multi-asset class fund (40% marginal tax-rate)

Typical multi-asset class fund (18% marginal tax-rate) Tax-free multi-asset class fund

So although the introduction of these accounts provides the first steps of encouraging a savings culture, how can individuals be further enticed to use these accounts knowing that the real tax benefits to new savers only come to fruition after several years of investment?

What should these accounts be used for?A recent study by University College London has suggested that the secret to a longer life is for individuals to have a sense of purpose and meaning to their lives. The research indicates that those individuals who feel they have a purpose also appear to survive longer. That link can be easily understood and reasoned with and similarly those principles can be easily linked to savings. Many people begin with good intentions to save but at the first signs of difficulty, or perhaps when the first sale signs hang in the shop window, those savings are depleted in favour of immediate spending. But if that savings account had a purpose, a realisable goal for the individual to strive for, this may see the savings survive for longer. No longer is an individual just choosing between adding or keeping

funds in their savings account compared to their latest expenditure, now the decision is between spending now or remaining committed to achieving a long-term goal they value and can visualise. A face has effectively been put on the savings account rather than it being viewed as a delay in consumption. These savings now have a meaning and purpose.

The employer angleThe 2014 edition of Benefits Barometer provided a sobering insight for employers: the retirement savings and employee benefit systems that were an integral part of their compensation packages were failing to provide employees with critical financial protection.

Getting members to appreciate the importance of preservation and the purchase of adequate income protection both before and after retirement has proven to be a monumental challenge.

Assumptions: Investment in a local multi-asset class fund (70% equity, 15% bond, 15% cash) achieving returns similar to those experienced over the last 10 years. The individual has no other discretionary investments and contributes R2 500 per month increased with inflation.

The effect of taxation on a long-term investment is illustrated in the graph below.

ALEXANDER FORBES

46

The reality is that for many employees, securing their day-to-day financial well-being simply takes precedence over saving for the future. Employees are also not fully cognisant of the value of the income protection offered to them as employee benefits. When families become financially stretched, cashing out of their retirement funds is often their first port of call.

The conclusion drawn by the Benefits Barometer research was that if we want South African workers to take their retirement plans and employee benefits seriously – and it’s clearly in the interests of both National Treasury and employers that they do – then we need to help them become resilient to short-term financial crises.

This is where the new tax-free vehicles being promoted by National Treasury can play a critical role. The value of these vehicles can be understood on two levels:1. By providing a parallel investment to a retirement

savings and employee benefits plan, they can provide the all-important ‘savings safety valve’. If a crisis does

materialise, these savings vehicles can provide the first line of defence for a family. However, remaining invested for the long term should be encouraged to get the benefits of compound interest and tax. Any amounts withdrawn cannot be replaced due to the annual and life time limits.

2. By linking these savings vehicles to specific financial goals, research suggests that investors will tend to exhibit a far higher level of commitment to the specific savings effort. Examples of compelling goals-based savings incentives could be:

a. To save for a child’s education at a specific date in the future

b. To provide an additional savings protection against escalating post-retirement healthcare costs

c. To provide for frail-care housing post-retirement d. To secure housing in retirement after one has been

living in company-sponsored housing during one’s work life.

National Treasury has provided a vehicle to help individuals drive towards these goals and provided an incentive. Attaching an appropriate goal to this vehicle will help improve the take-up of the incentive on offer. The real benefits of the tax-free savings accounts are felt in the longer term when thebenefits of compound interest come through. For this reason it is most suited for individuals who can attach the savings account to a long-term goal with savings terms of at least 8 years. These goals may include saving for your child’s university education, a supplement to your retirement funding or

post-retirement medical expenses, or that trip sailing around the world when you finally havethe time to do it.

We have been given that first step and initial incentive to begin saving by government. It’s now up to us to ensure we remain committed to the long-term saving plan to reap the benefits of those incentives. Linking savings to a purpose will help make that long-term savings journey easier, and having a purpose for one’s savings may just lead to having a longer life to enjoy the fruits of those savings as well.

NOTES:

50 Designed by Alexander Forbes Financial Services 9990-2015-03.

investing towards your goal · 1 1 dec-13 oil us$ price long-term performance of oil price oil...

Documents