reserve bank of s eview may une olume umber · 2019. 3. 18. · federal reserve bank of st.louis...

FEDERAL RESERVE BANK OF ST. LOUIS

R E V IE WVOLUME 83, NUMBER 3MAY/JUNE 2001

The Practice of Central Bank Intervention:Looking Under the Hood

Christopher J. Neely

Forecasting Inflation and Growth: Do Private Forecasts Match Those of Policymakers?

William T. Gavin and Rachel J. Mandal

Toward a New Paradigm in Open Economy Modeling:Where Do We Stand?

Lucio Sarno

A Simple Model of Limited Stock Market Participation

Hui Guo

Director of Research Robert H. Rasche

Associate Director of ResearchCletus C. Coughlin

Review EditorWilliam T. Gavin

BankingWilliam R. EmmonsR. Alton GilbertFrank A. SchmidDavid C. Wheelock

InternationalChristopher J. NeelyMichael R. PakkoPatricia S. Pollard

Macroeconomics Richard G. AndersonJames B. BullardMichael J. DuekerHui GuoKevin L. KliesenMichael T. OwyangDaniel L. Thornton

RegionalRuben Hernandez-MurilloHoward J. WallAdam M. Zaretsky

Managing EditorGeorge E. Fortier

Assistant EditorLydia H. Johnson

Graphic DesignerDonna M. Stiller

Review is published six times per year by theResearch Division of the Federal Reserve Bank of St. Louis. Single-copy subscriptions are availablefree of charge. Send requests to: Federal ReserveBank of St. Louis, Public Affairs Department, P.O. Box 442, St. Louis, Missouri 63166-0442, or call (314) 444-8808 or 8809.

The views expressed are those of the individualauthors and do not necessarily reflect official positions of the Federal Reserve Bank of St. Louis, the Federal Reserve System, or the Board of Governors.

© 2001, The Federal Reserve Bank of St. Louis. Articlesmay be reprinted, reproduced, published, distributed,displayed, and transmitted in their entirety if this copy-right notice is included. Please provide the ResearchDivision, Federal Reserve Bank of St. Louis, P.O. Box442, St. Louis, Missouri 63166-0442, with a copy ofany reprinted, published, or displayed materials.Please note: Abstracts, synopses, and other derivativeworks may be made only with prior written permis-sion of the Federal Reserve Bank of St. Louis. Pleasecontact the Research Division at the above address torequest permission.

R E V I E W


MAY/JUNE 2001 i

ContentsVolume 83, Number 3

1 The Practice of Central Bank Intervention:Looking Under the Hood

Christopher J. Neely

This article first reviews methods of foreignexchange intervention and then presents evi-dence—focusing on survey results—on themechanics of such intervention. Types ofintervention, instruments, timing, amounts,motivation, secrecy, and perceptions of effica-cy are discussed.

11 Forecasting Inflation and Growth: DoPrivate Forecasts Match Those ofPolicymakers?

William T. Gavin and Rachel J. Mandal

Federal Open Market Committee (FOMC) pro-jections are important because they provideinformation for evaluating current monetarypolicy intentions and because they indicatewhat FOMC members think will be the likelyconsequence of their policies. Knowing theFed’s objectives, their forecasts, and recentdeviations of the economy from the forecastsshould be sufficient to understand how theFed is making monetary policy. Results hereshow that the Blue Chip consensus forecastsare a good proxy for the FOMC views. Forexample, they match the policymakers’ viewsas closely as do the Board staff forecasts pre-sented at FOMC meetings. Using alternativeforms of the Taylor rule, the authors showthat the Blue Chip consensus and the Fedpolicymakers’ forecasts have almost identicalimplications for the monetary policy process.

21 Toward a New Paradigm in OpenEconomy Modeling: Where Do We Stand?

Lucio Sarno

This paper provides a selective, up-to-datesurvey of the recent, fast-growing literatureon new open economy macroeconomics.Lucio Sarno begins with a review of the semi-nal paper in this literature, describing thebaseline model proposed therein. He thencovers a number of variants and generaliza-tions of the baseline model involving theallowance for nominal rigidities, pricing tomarket, alternative preference specifications,and alternative financial markets structures.The author also discusses the recent stochas-tic extensions of these models, especiallyfocusing on their implications for the linkbetween risk and exchange rates and on newdirections for the relevant literature.

37 A Simple Model of Limited Stock MarketParticipation

Hui Guo

Stocks have outperformed government bonds,on average, by a large margin in historicaldata. However, most U.S. households do notown stocks, either directly or indirectly. Also,stocks are highly concentrated in the handsof relatively few wealthy people. In this arti-cle, Hui Guo describes some aspects of stockownership. He then uses an overlapping-generations model to help explain why stockmarket participation is so limited and dis-cusses some implications of limited stockmarket participation.

i i MAY/JUNE 2001

R E V I E W


MAY/JUNE 2001 1

The Practice of CentralBank Intervention:Looking Under theHoodChristopher J. Neely

There has been a long and voluminous litera-ture about official intervention in foreignexchange markets. Official intervention is

generally defined as those foreign exchange trans-actions of monetary authorities that are designedto influence exchange rates, but can more broadlyrefer to other policies for that purpose. Many papershave explored the determinants and efficacy ofintervention (Edison, 1993; Sarno and Taylor, 2000)but very little attention has been paid to the morepedestrian subject of the mechanics of foreignexchange intervention like choice of markets, typesof counterparties, timing of intervention duringthe day, purpose of secrecy, etc. This article focuseson the latter topics by reviewing the motivationfor, methods, and mechanics of intervention.Although there apparently has been a decline inthe frequency of intervention by the major centralbanks, reports of a coordinated G-7 intervention tosupport the euro on September 22, 2000, remindus that intervention remains an active policy instru-ment in some circumstances.

The second section of the article reviews foreignexchange intervention and describes several meth-ods by which it can be conducted. The third sectionpresents evidence from 22 responses to a survey onintervention practices sent to monetary authorities.

TYPES OF INTERVENTION

Intervention and the Monetary BaseStudies of foreign exchange intervention gen-

erally distinguish between intervention that does

or does not change the monetary base. The formertype is called unsterilized intervention while thelatter is referred to as sterilized intervention. Whena monetary authority buys (sells) foreign exchange,its own monetary base increases (decreases) by theamount of the purchase (sale). By itself, this typeof transaction would influence exchange rates inthe same way as domestic open market purchases(sales) of domestic securities; however, many cen-tral banks routinely sterilize foreign exchangeoperations—that is, they reverse the effect of theforeign exchange operation on the domestic mon-etary base by buying and selling domestic bonds(Edison, 1993). The crucial distinction betweensterilized and unsterilized intervention is that theformer constitutes a potentially useful indepen-dent policy tool while the latter is simply anotherway of conducting monetary policy.

For example, on June 17, 1998, the FederalReserve Bank of New York bought $833 millionworth of yen (JPY) at the direction of the U.S.Treasury and the Federal Open Market Committee.In the absence of offsetting transactions, thistransaction would have increased the U.S. mone-tary base by $833 million, which would tend totemporarily lower interest rates and ultimatelyraise U.S. prices, depressing the value of the dol-lar.1 As is customary with U.S. intervention, how-ever, the Federal Reserve Bank of New York alsosold an appropriate amount of U.S. Treasurysecurities to absorb the liquidity and maintaindesired conditions in the interbank loan market.Similarly, to prevent any change in Japanesemoney market conditions, the Bank of Japanwould also conduct appropriate transactions tooffset the rise in demand for Japanese securitiescaused by the $833 million Federal Reserve pur-chase. The net effect of these transactions wouldbe to increase the relative supply of U.S. govern-ment securities versus Japanese securities held bythe public but to leave the U.S. and Japanesemoney supplies unchanged.

Fully sterilized intervention does not directlyaffect prices or interest rates and so does not influ-ence the exchange rate through these variables asordinary monetary policy does. Rather, sterilizedintervention might affect the foreign exchangemarket through two routes: the portfolio balancechannel and the signaling channel. The portfolio

Christopher J. Neely is a senior economist with the Federal ReserveBank of St. Louis. The author thanks the monetary authorities ofBelgium, Brazil, Canada, Chile, the Czech Republic, Denmark, France,Germany, Hong Kong, Indonesia, Ireland, Italy, Japan, Mexico, NewZealand, Poland, South Korea, Spain, Sweden, Switzerland, Taiwan,and the United States for their cooperation with this study. Theauthor thanks Michael Melvin and Paul Weller for discussionsabout the survey and Hali Edison, Trish Pollard, and Lucio Sarnofor helpful comments. Mrinalini Lhila provided research assistance.This article was originally published in Central Banking (November2000, XI(2), pp. 24-37; <http://www.centralbanking.co.uk>).

1 Empirically, it has been very difficult to establish the reaction ofexchange rates to changes in economic fundamentals.

2 MAY/JUNE 2001

R E V I E W

balance channel theory holds that sterilized pur-chases of yen raise the dollar price of yen becauseinvestors must be compensated with a higher ex-pected return to hold the relatively more numerousU.S. bonds. To produce a higher expected return,the yen price of the U.S. bonds must fall immedi-ately. That is, the dollar price of yen must rise. Incontrast, the signaling channel theory assumesthat official intervention communicates informa-tion about future monetary policy or the long-runequilibrium value of the exchange rate.

Spot and Forward Markets forIntervention

The previous example implicitly assumed thatthe Federal Reserve Bank of New York conductedits purchase of yen in the spot market—the mar-ket for delivery in two days or less. Interventionneed not be carried out in the spot market, how-ever; it also may be carried out in the forwardmarket.2 Forward markets are those in whichcurrencies are sold for delivery in more than twodays. Because the forward price is linked to thespot price through covered interest parity, inter-vention in the forward market can influence thespot exchange rate.

To understand covered interest parity, con-sider the options open to an American bank thathas capital to be invested for one year. The bankcould lend that money at the interest rate on U.S.dollar assets, earning the gross return of (1+it

USD)on each dollar. Or, it could convert its funds to aforeign currency (e.g., the euro), lend that sum inthe overnight euro money market at the eurointerest rate, and then convert the proceeds backto dollars at the end of the year. If, at the begin-ning of the contract, the bank contracts to convertthe euro proceeds back to dollars, it will receive1/Ft,t+365 dollars for each euro, where Ft,t+365 isthe euros-per-dollar forward exchange rate. Thegross return to each dollar through this secondstrategy is

,

where St is the euros-per-dollar spot exchange rateon day t. If the return to one strategy is higherthan the other, market participants will invest inthat strategy, driving its return down and the otherreturn up until the strategies have approximately

SF

it

t tteuro

, ++( )

365

1

equal return. Covered interest parity (CIP) is thecondition that the strategies have equal return:

(1)

As equation (1) must approximately hold allthe time, intervention that changes the forwardexchange rate must also change the spot exchangerate.3 For example, a forward purchase of eurosthat raises Ft,t+365 must also raise St.

Forward market interventions—the purchaseor sale of foreign exchange for delivery at a futuredate—have the advantage that they do not requireimmediate cash outlay. If a central bank expectsthat the need for intervention will be short-livedand will be reversed, then a forward market inter-vention may be conducted discreetly, with noeffect on foreign exchange reserves data. For ex-ample, published reports indicate that the Bank ofThailand used forward market purchases to shoreup the baht in the spring of 1997 (Moreno, 1997).4

Both the spot and forward markets may beused simultaneously. A transaction in which acurrency is bought in the spot market and simul-taneously sold in the forward market is known asa currency swap. While a swap itself will have lit-tle effect on the exchange rate, it can be used aspart of an intervention. The Reserve Bank ofAustralia (RBA) used the swaps market to sterilizespot interventions. In these transactions, the spotleg of the swap is conducted in the opposite direc-tion to the spot market intervention, leaving thesequence equivalent to a forward market interven-tion. The RBA uses the spot/swap combinationrather than an outright forward transactionbecause the former permits more flexible imple-mentation of the intervention.

1 1365

+( ) = +( )+

iS

Fit

USD t

t tteuro

,

.

2 Exchange rate markets and practices are described in detail in theBank for International Settlements Central Bank Survey of ForeignExchange and Derivatives Market Activity (1999).

3 Of course, equation (1) could continue to hold with a change initUSD or iteuro instead of Ft,t+365, but the interest rates are held fixedby conditions in the U.S. and euro money markets, respectively.

4 Not all spot or forward market transactions are interventions, ofcourse. For example, to limit the costs of capital controls that madeit hard to hedge foreign exchange exposure, the Reserve Bank ofSouth Africa (RBSA) used to provide forward cover for firms withforeign currency exposure. That is, it would buy dollars forwardfrom foreign firms with dollar accounts receivable and sell dollarsto foreign firms with dollar accounts payable in the future. As capi-tal controls have been reduced, the RBSA has reduced its net openposition in the forward market.


MAY/JUNE 2001 3

The Options Market and Intervention

The options market has also been used bycentral banks for intervention. A European-stylecall (put) option confers the right, but not the obli-gation, to purchase (sell) a given quantity of theunderlying asset on a given date. Usually, theoption contract specifies the price for which theasset may be bought or sold, called the strike orexercise price.

Monetary authorities seeking to prevent depre-ciation or devaluation of their currency may sellput options on the domestic currency or calloptions on the foreign currency.5 While the priceof options has no direct effect on spot exchangerates, speculators often purchase put optionsinstead of shorting a weak currency. The writers(sellers) of these put options attempt to hedgetheir position by taking a long position in theweak currency, adding to the downward pressureon its price. By writing put options on the weakcurrency—adding liquidity to the options mar-ket—the central bank provides dealers with a syn-thetic hedge; dealers need not go into the spotmarket to take short positions in the weak curren-cy. This arrangement creates the same type offinancial risk for the central bank—if the currencyis devalued—as would the direct purchase of theweak currency in spot or forward markets. Likeforward market intervention, it does not, however,require the monetary authority to immediatelyexpend foreign exchange reserves. In fact, thestrategy generates revenues upon the sales of theoptions. The Bank of Spain reportedly used thisstrategy of selling put options on the peseta tofight devaluation pressures during 1993 (TheEconomist, 1993), though the institution denied itemphatically (The Financial Times, 1993).

In another intervention strategy using options,the Banco de Mexico has employed sales of putoptions on the U.S. dollar to accumulate foreignexchange reserves since August 1, 1996 (GalanMedina, Duclaud Gonzalez de Castillo, GarciaTames, 1997). The put options give the bearer theright to sell dollars to the Banco de Mexico at astrike price determined by the previous day’sexchange rate, called the fix exchange rate. Theoption may be exercised only if the peso hasappreciated over the last month, if the fix pesoprice of dollars is no higher than a 20-day movingaverage of previous strike prices. This restriction isdesigned to prevent the Banco de Mexico from

having to buy dollars (sell pesos) during a periodof peso depreciation.

The sales of these put options may be consid-ered foreign exchange intervention because they aredesigned to prevent the necessity of intervention topurchase dollar reserves that might affect the ex-change rate in undesirable ways. Because the mech-anism is totally passive—the public decides when toexercise the options—the use of these options effec-tively lessens the signaling impact of Banco de Mex-ico purchases of foreign exchange reserves.

Indirect Intervention

Recall that although official intervention isgenerally defined as foreign exchange transac-tions of monetary authorities that are designed toinfluence exchange rates, it can also refer to other(indirect) policies for that purpose. In addition todirect transactions in various instruments, Taylor(1982a, b) recounts a number of methods by whichcountries intervene indirectly in foreign exchangemarkets. For example, he reports that in the 1970sgovernments manipulated the currency portfolioof nationalized industries in France, Italy, Spain,and the United Kingdom to influence exchangerates. This practice was allegedly used to “disguise”intervention, as was the French and Italian prac-tice of transacting through undisclosed foreignexchange accounts held at commercial banks.

There are innumerable methods of indirectlyinfluencing the exchange rate that do not fit inthe narrow definition of intervention as foreignexchange transactions of monetary authoritiesdesigned to influence exchange rates. These meth-ods involve capital controls—taxes or restrictionson international transactions in assets like stocksor bonds—or exchange controls—the restriction oftrade in currencies (Dooley, Mathieson, and Rojas-Suarez, 1993; Neely, 1999), rather than transactions.Sometimes such methods are substituted for moredirect foreign exchange intervention, especially bythe monetary authorities of countries without along history of free capital movements. For example,Spain, Ireland, and Portugal introduced capitalcontrols—including mandatory deposits againstthe holding of foreign currencies—in the exchangerate mechanism (ERM) crises of 1992-93, in re-sponse to speculation against their currencies.

5 Blejer and Schumacher (2000) discuss the implications of the use ofderivatives for central banks’ balance sheets.

4 MAY/JUNE 2001

R E V I E W

SURVEY RESULTS

To investigate the practice of foreign exchangeintervention, questionnaires on the topic were sentto the 43 institutions that participated in the Bankfor International Settlements (BIS) (1999) survey offoreign exchange practices and the European Cen-tral Bank. Of those 44 authorities, 22 responded tosome or all of the questions asked. Table 1 lists thesurveyed authorities who responded. A cover letterexplained that the survey covered practices over1990-2000, so that authorities that no longer inter-vene—or even no longer have independent curren-cies—could report on past practices. The ReserveBank of New Zealand was the only authority to re-port that it had not intervened in the last 10 years.6To respect the confidentiality of the respondents,the responses of specific institutions will not beidentified unless the information is clearly in thepublic domain. Table 2 shows statistics summariz-ing the distribution of the responses to surveyquestions on the mechanics of, motivation for, andthe efficacy of intervention.

The Mechanics of Intervention

Question 1 inquires about the frequency ofintervention. Of the 14 authorities that respondedto the question, the percentage of business dayson which they report intervening—using eithersterilized or unsterilized transactions—rangedfrom 0.5 percent to 40 percent, with 4.5 percentbeing the median. While there might be someselection bias because authorities that interveneare more likely to respond to the survey, it does

appear that official intervention is reasonablycommon in foreign exchange markets.

In responding to question 2, 30 percent of au-thorities report that all their foreign exchangetransactions change the monetary base, 30 percentthat such dealings sometimes change the base, and40 percent that they never change the base. Forsome issues, such as motivation or time horizon ofeffectiveness, this conflation of responses aboutsterilized and unsterilized intervention is poten-tially unfortunate.

When monetary authorities do intervene(question 3), they seem to have some preferencefor dealing with major domestic banks but will alsotransact with major foreign banks. This should notcome as a complete surprise as banks tend to spe-cialize in trading their own national currencies(Melvin, 1997). Approximately half of authoritieswill sometimes conduct business with other enti-ties, such as other central banks (23.5 percent) orinvestment banks (25 percent); 6.3 percent will al-ways transact with investment banks.

Intervention transactions over the last decadehave almost always been conducted at least par-tially in spot markets according to the answers toquestion 4: 95.2 percent of authorities report thattheir official intervention activities always includespot market transactions and another 4.8 percentsometimes include spot transactions. A total of52.9 percent of authorities report sometimes usingthe forward market, perhaps in conjunction withthe spot market to create a swap transaction. Noauthority reports always using the forward market.Finally, one authority reports having used a futuresmarket to conduct intervention.

There is no clear pattern as to the method ofdealing with counterparties (question 5). Directdealing over the telephone is most popular, beingused sometimes or always by 100 percent of au-thorities. Direct dealing over an electronic networkis used by 43.8 percent of authorities sometimes oralways. Live foreign exchange brokers are usedsometimes or always by 63.2 percent of the re-spondents. Finally, electronic brokers such as EBSare used by 12.5 percent of the authorities. There

6 The Reserve Bank of New Zealand’s response on this point is a mat-ter of public record. See Deputy Governor Sherwin’s May 9, 2000,speech to the World Bank Treasury at<http://www.rbnz.govt.nz/speeches/0092115.html>. This link wascurrent as of April 20, 2001. The Appendix to this article provideslinks to descriptions of foreign exchange markets and/or interven-tion policies in the Web pages of a number of monetary authorities.

Responding Monetary Authorities

Country/authority

Belgium ItalyBrazil JapanCanada MexicoChile New ZealandCzech Republic PolandDenmark South KoreaFrance SpainGermany SwedenHong Kong SwitzerlandIndonesia TaiwanIreland United States

NOTE: The table identifies the 22 monetary authorities thatresponded to the survey.

Table 1


MAY/JUNE 2001 5

Summary of Intervention Survey Responses

1. In the last decade, on approximately what percentage of business days has your monetary authority conducted intervention?

2. Foreign exchange intervention changes the domestic monetary base.

3. Intervention transactions are conducted with the followingcounterparties:

Major domestic banksMajor foreign banksOther central banksInvestment banks

4. Intervention transactions are conducted in the following markets:SpotForward FutureOther (please specify in margin)

5. Intervention transactions are conducted by:Direct dealing with counterparties via telephoneDirect dealing with counterparties via electronic communicationLive FX brokers Electronic brokers (e.g., EBS, Reuters 2002)

6. The following strategies determine the amount of intervention:A prespecified amount is traded Intervention size depends on market reaction to initial trades

7. Intervention is conducted at the following times of day:Prior to normal business hoursMorning of the business dayAfternoon of the business dayAfter normal business hours

8. Is intervention sometimes conducted through indirect methods, such as changing the regulations regarding foreign exchangeexposure of banks?

9. The following are factors in intervention decisions:To resist short-term trends in exchange ratesTo correct long-run misalignments of exchange rates from

fundamental valuesTo profit from speculative tradesOther

10. Intervention transactions are conducted secretly for the following reasons:

To maximize market impactTo minimize market impactFor portfolio adjustment Other

11. In your opinion, how long does it take to observe the full effect of intervention on exchange rates?

A few minutesA few hoursOne dayA few daysMore than a few daysIntervention has no effect on exchange rates

No. of responses

14

No. of responses

20

21181716

21171615

20161916

1720

16212017

21

19

181716

17141112

181818181818

Maximum

40.0

Always

30.0

71.411.10.06.3

95.20.00.00.0

70.012.510.512.5

11.840.0

0.014.310.00.0

0.0

47.4

22.20.0

12.5

41.20.00.08.3

Median

4.5

Sometimes

30.0

28.672.223.525.0

4.852.96.36.7

30.031.352.60.0

70.655.0

43.885.790.064.7

23.8

42.1

44.40.0

25.0

35.357.10.0

16.7

Minimum

0.5

Never

40.0

0.016.776.568.8

0.047.193.893.3

0.056.336.887.5

17.65.0

56.30.00.0

35.3

76.2

10.5

33.3100.062.5

23.542.9

100.075.0

38.922.20.0

27.811.10.0

Table 2

NOTE: Question 1 shows the minimum, median, and maximum responses (from 0 to 100) on the percentage of days intervention wasconducted in the last decade. Questions 2 through 10 show the percentage of responses of “Never,”“Sometimes,” and “Always” to thosequestions. Question 11 shows the percentage of responses indicating that the full effects of intervention were felt at each horizon.

6 MAY/JUNE 2001

R E V I E W

was some evidence that responding institutions aremoving increasingly toward electronic methods,along with other market participants.

There has been very little research on factorsthat determine the magnitude of intervention—though reaction function research touches on thisissue—but the responses to question 6 indicatethat the size of interventions frequently dependson market reaction to initial trades. Ninety-fivepercent of monetary authorities report that mar-ket reaction sometimes or always affects the sizeof total trades. Because the size of the interventionis endogenous to market reaction, determiningthe interaction of intervention and exchange ratechanges at high frequencies might require carefulevaluation.7

Most intervention takes place during the busi-ness day, but almost half the banks report thatthey sometimes intervene prior to business hoursand more than half intervene after businesshours.8 For example, the Reserve Bank ofAustralia publicly states that it is willing to inter-vene outside of normal business hours (Rankin,1998). Some after-hours intervention might be insupport of other authorities. About a quarter ofcentral banks report that they always intervene inthe business morning (14.3 percent) or businessafternoon (10 percent), however.

In answering question 8, 23.8 percent ofauthorities report sometimes intervening withindirect methods. The authorities cite changingbanking regulations on foreign exchange expo-sure and moral suasion as methods of indirectintervention. Not surprisingly, indirect methodsseem to be used predominantly by central bankswithout a long history of free capital movementsor a convertible currency.

The Motivation for Intervention

The motivation for intervention decisions hasbeen widely researched and often discussed.Research and official pronouncements supportthe idea that monetary authorities with floatingexchange rates most often employ interventionto resist short-run trends in exchange rates, theleaning-against-the-wind hypothesis.9 Anotherpopular hypothesis is that intervention is used tocorrect medium-term “misalignments” of exchangerates away from “fundamental values.” Question 9inquires about these possibilities. The responsesgenerally support these hypotheses with 89.5 per-cent of monetary authorities sometimes or always

intervening to resist short-run trends and 66.7 per-cent seeking to return exchange rates to “funda-mental values.” Some countries specified “other”reasons that might be interpreted as variations onthe leaning-against-the-wind or misalignmenthypotheses. Still other countries note macroeco-nomic goals such as limiting exchange rate pass-through to prices, defense of an exchange ratetarget, or accumulating reserves as motivatingintervention.

One hypothesis that has received some atten-tion in the last few years is that profitability is aconsideration in intervention. A series of papershave examined the profitability of intervention(Leahy, 1995; Sweeney, 1997; Saacke, 1999), therelationship between intervention profitability andtechnical analysis (Neely, 1998, 2000), andwhether past profits influence intervention (Kimand Sheen, 1999). While the early evidence(Taylor, 1982b) indicated that central banks werelosing money on their intervention, the laterpapers have been much more supportive of thehypothesis that central banks have at least brokeneven on floating rate intervention, with some evi-dence that they have made large profits.10

The notion that profitability is a considerationin intervention decisions is uniformly rejected,however, by the survey respondents. Not onerespondent to question 9 reports that profitabilitywas ever a motivation for intervention. Despitethis, private conversations with individualsinvolved in intervention decisions suggest thatprofitability is a useful gauge of their success ascareful stewards of public resources. In addition,the Reserve Bank of Australia argues that prof-itability attests that its intervention has stabilizedthe exchange rate (Rankin, 1998). This RBA claimrelies on Friedman’s (1953) claim that stabilizingspeculation is equivalent to profitable speculation.If speculators consistently buy (sell) when theasset price is below (above) its equilibrium value,they will both tend to drive the asset price toward

7 The author thanks Lucio Sarno for pointing this out.

8 Fischer and Zurlinden (1999) examine the affect of interventionusing high frequency data and the time of data.

9 Indeed, the published statements of several central banks specifi-cally cite the desire to counter trends in exchange rates as motivat-ing intervention. See Board of Governors (1994, p. 64) or Rankin(1998).

10 Sweeney (1997, 2000) argues that risk adjustment is crucial inassessing profits or losses from official intervention.


MAY/JUNE 2001 7

its equilibrium value and to profit from thesetransactions.11 The link between profitability andstabilizing speculation is tenuous, however. Salant(1974), Mayer and Taguchi (1983), and De Long,Schleifer, Summers, and Waldmann (1990) pro-vide counterexamples.

The Role of Secrecy in Intervention

The role of secrecy in intervention is not wellunderstood. Most monetary authorities usuallychose to intervene secretly, releasing actual inter-vention data with a lag, if at all. Some authorities,the Swiss National Bank, for example, alwayspublicize interventions at the time they occur.Does secret intervention maximize or minimizethe impact of the transaction? If authorities inter-vene to convey information to markets, why dothey conceal these transactions? Dominguez andFrankel (1993) recount several possibilities: Whenfundamentals are inconsistent with intervention,monetary authorities would prefer not to drawattention to the intervention. Or, the monetaryauthority might have poor credibility for sendingtrustworthy signals. Or, the monetary authoritymight wish to simply adjust the currency holdingsof its portfolio. Bhattacharya and Weller (1997)provide another possible explanation for secrecy.They present a model in which small amounts ofintervention reveal the authority’s information toprivate parties, thus influencing exchange rates.This secrecy issue has not been satisfactorilyresolved in the literature.

Consistent with the confusion in the academicliterature, the answers to question 10 reflect dis-agreement among the respondents about the pur-pose of secrecy. More authorities report sometimesor always intervening secretly to maximize mar-ket impact (76.5 percent) than report sometimesor always intervening secretly to minimize marketimpact (57.1 percent). Such disagreement is sig-nificant. No central bank cites portfolio adjustmentas a reason for secret intervention, contrary to thereasonable conjecture of Dominguez and Frankel(1993). Of course, the central banks might notconsider transactions for portfolio adjustment tobe intervention.

The Horizon of Intervention Effects

Perhaps the most important question in theliterature on central bank intervention is whethercentral bank intervention is effective in influenc-

ing the exchange rate. For many years, the biggesthurdle to answering this question was the paucityof data. More recently, even as more data havebecome available, it is manifest that two barriersto answering the question remain. First, whatwould the exchange rate have been in the absenceof intervention? Second, over what horizon shouldwe measure the effectiveness of intervention andhow large and long-lasting an effect can be con-sidered a success?

The academic literature has been ambivalentabout the efficacy of official intervention in theforeign exchange market. The Jurgensen Report(Jurgensen, 1983) was pessimistic about the effectsof intervention. Dominguez and Frankel (1993)—using then-recently released intervention data—argued that intervention can work by changingexpectations of future exchange rates. Edison(1993) concluded that, while the evidence mightbe consistent with some short-run effect, there isno evidence for a lasting effect from intervention.Sarno and Taylor (2000), in contrast, concludethat the recent consensus of the profession is thatintervention is effective through both the signal-ing and portfolio balance channels.

Despite skepticism on the part of academics,central banks continue to intervene—though per-haps less frequently than in the past—implyingthat policymakers do think that intervention can bean effective tool. Question 11 asks the monetaryauthorities whether intervention has an effect onexchange rates and, if so, over what horizon onemight see the full effect. All of the respondents indi-cate that they think intervention has some effecton exchange rates.12 Most of the respondentsbelieve in a relatively rapid response, over a fewminutes (38.9 percent) or a few hours (22.2 per-cent). Still, a substantial minority think that inter-vention’s full effect is seen over a few days (27.8percent) or more (11.1 percent). The dispersion inthe survey is substantial, indicating almost as muchdiscord among central bankers as among academics.

DISCUSSION AND CONCLUSION

This article has examined the mechanics ofintervention—instruments, counterparties, timing,

11 Friedman (1953) was referring more generally to speculation in for-eign exchange and discussed government speculation (intervention)as a special case.

12 Of course, having an effect on exchange rates at some horizonmight not imply that intervention is successful.

8 MAY/JUNE 2001

R E V I E W

and amounts—as well as related issues like secre-cy, motivation, and the perceived efficacy of suchtransactions. A survey of monetary authorities’intervention practices reveals that a number ofmonetary authorities do intervene with some fre-quency in foreign exchange (mostly spot) markets.The desire to check short-run trends or correctlonger-term misalignments often motivates inter-vention, whereas the size of intervention oftendepends on market reaction to initial trades.Although intervention typically takes place duringbusiness hours, most monetary authorities willalso intervene outside of these hours, if necessary.And, while there is unanimous agreement thatintervention does influence exchange rates, thereis much disagreement about the horizon overwhich the full effect of this influence is felt, withestimates ranging from a few minutes to morethan a few days.

The topic of the practice of official interven-tion is very broad. To simplify this study, issueslike coordinated versus unilateral intervention,choice of intervention currency, and distinguish-ing between intervention in fixed and flexibleexchange rate regimes have been left for furtherstudy. Other issues that merit further considera-tion are motivations for secrecy and the metric forjudging the success of intervention.

REFERENCES

Bank for International Settlements. “Central Bank Surveyof Foreign Exchange and Derivatives Market Activity.”May 1999.

Bhattacharya, Utpal and Weller, Paul A. “The Advantage toHiding One’s Hand: Speculation and Central BankIntervention in the Foreign Exchange Market.” Journal ofMonetary Economics, July 1997, 39(2), pp. 257-77.

Blejer, Mario I. and Schumacher, Liliana. “Central BanksUse of Derivatives and Other Contingent Liabilities:Analytical Issues and Policy Implications.” WorkingPaper No. 00-66, International Monetary Fund(Washington, DC), March 2000.

Board of Governors of the Federal Reserve System. TheFederal Reserve System: Purposes and Functions.Washington, DC, 1994.

De Long, J. Bradford; Shleifer, Andrei; Summers, LaurenceH. and Waldmann, Robert J. “Noise Trader Risk inFinancial Markets.” Journal of Political Economy, August1990, pp. 703-38.

Dominguez, Kathryn and Frankel, Jeffery. “Does ForeignExchange Intervention Work?” Washington, DC: Institutefor International Economics, 1993.

Dooley, Michael P.; Mathieson, Donald J. and Rojas-Suarez,Liliana. “Capital Mobility and Exchange MarketIntervention in Developing Countries.” Working PaperNo. 6247, National Bureau of Economic Research(Cambridge, MA), October 1997.

Economist. “Spanish Peseta; Adios.” 1 May 1993, pp. 84.

Edison, Hali J. “The Effectiveness of Central-BankIntervention: A Survey of the Literature After 1982.”Special Papers in International Economics No. 18,Department of Economics, Princeton University, 1993.

Financial Times. “Peseta Devalued by 8% in ERM: LisbonRealigns Escudo as Madrid Plans 1 1/2 Point Rate Cut.”14 May 1993, pp. 1.

Fischer, Andreas M., and Zurlinden, Mathias. “ExchangeRate Effects of Central Bank Interventions: An Analysisof Transaction Prices.” Economic Journal, October 1999,109(458), pp. 662-76.

Friedman, Milton. Essays in Positive Economics. Chicago:University of Chicago Press, 1953, pp. 157-203.

Galan Medina, Manuel; Duclaud Gonzalez de Castilla, Javierand Garcia Tames, Alonso. “A Strategy for AccumulatingReserves Through Options to Sell Dollars.” Unpublishedmanuscript, Banco de Mexico, 1997. <http://www.banxico.org.mx/siteBanxicoINGLES/bPoliticaMonetaria/FSpoliticaMonetaria.html>.

Jurgensen, Philippe. “Report of the Working Group onExchange Market Intervention.” Washington, DC: U.S.Department of the Treasury, March 1983.

Kim, Suk-Joong and Sheen, Jeffery. “The Determinants ofForeign Exchange Intervention by Central Banks:Evidence from Australia.” Unpublished manuscript, TheUniversity of New South Wales, December 1999.

Leahy, Michael P. “The Profitability of US Intervention inthe Foreign Exchange Markets.” Journal of InternationalMoney and Finance, December 1995, 14(6), pp. 823-44.

Mayer, Helmut and Taguchi, Hiroo. “Official Interventionin the Exchange Markets: Stabilising or Destabilising?”Bank for International Settlements Economic Paper 6,March 1983.


MAY/JUNE 2001 9

Melvin, Michael T. International Money and Finance. 5thEd. Reading, MA: Addison Wesley, 1997.

Moreno, Ramon. “Lessons from Thailand.” Federal ReserveBank of San Francisco Economic Letter No. 97-33, 7November 1997.

Neely, Christopher J. “Technical Analysis and theProfitability of U.S. Foreign Exchange Intervention.”Federal Reserve Bank of St. Louis Review, July/August1998, 80(4), pp. 3-17.

___________. “An Introduction to Capital Controls.” FederalReserve Bank of St. Louis Review, November/December1999, 81(6), pp. 13-30.

___________. “The Temporal Pattern of Trading RuleReturns and Central Bank Intervention: InterventionDoes Not Generate Trading Rule Profits.” Working Paper2000-18B, Federal Reserve Bank of St. Louis, August 2000.

Rankin, Bob. “The Exchange Rate and the Reserve Bank’sRole in the Foreign Exchange Market.” Reserve Bank ofAustralia, 1998. <http://www.rba.gov.au/publ/pu_teach_98_2.html>.

Saacke, Peter. “Technical Analysis and the Effectiveness ofCentral Bank Intervention.” Unpublished manuscript,University of Hamburg, 1999.

Salant, Stephen W. “Profitable Speculation, Price Stability,and Welfare.” International Finance Discussion Paper54, Board of Governors of the Federal Reserve System,November 1974.

Sarno, Lucio and Taylor, Mark P. “Official Intervention inthe Foreign Exchange Market.” Unpublished manuscript,University College, Oxford, 2000.

Sweeney, Richard J. “Do Central Banks Lose on Foreign-Exchange Intervention? A Review Article.” Journal ofBanking and Finance, December 1997, 21(11-12), pp.1667-84.

___________. “Does the Fed Beat the Foreign ExchangeMarket?” Journal of Banking and Finance, May 2000,24(5), pp. 665-94.

Taylor, Dean. “The Mismanaged Float: Official Interventionby the Industrialized Countries,” in Michael B. Connolly,ed., The International Monetary System: Choices for theFuture. Westport, CT: Praeger Publishers, 1982a, pp. 49-84.

___________. “Official Intervention in the Foreign ExchangeMarket, or, Bet Against the Central Bank.” Journal ofPolitical Economy, April 1982b, 90(2), pp. 356-68.

10 MAY/JUNE 2001

R E V I E W

Monetary authority

Reserve Bank ofAustralia

Central Bank of Brazil

Bank of Canada

Bank of England

European Central Bank

German Bundesbank

Hong Kong Monetary Authority

Bank of Japan

Banco de Mexico

National Bank of Poland

Monetary Authority of Singapore

Reserve Bank of South Africa

Swiss National Bank

Bank of Thailand

Federal Reserve of theUnited States

Uniform Resource Locator on intervention orforeign exchange

http://www.rba.gov.au/publ/pu_teach_98_2.html

http://www.bcb.gov.br/ingles/himf1900.shtm#destino6

http://www.bankofcanada.ca/en/backgrounders/bg-e2.htm

http://www.bankofengland.co.uk/factfxmkt.pdf

http://www.ecb.int

http://www.bundesbank.de/en/presse/pressenotizen/pdf/mp100197.pdf

http://www.info.gov.hk/hkma/eng/speeches/speechs/joseph/speech_030398b.htm

http://www.boj.or.jp/en/faq/faqkainy.htm

http://www.banxico.org.mx/siteBanxicoINGLES/bPoliticaMonetaria/FSpoliticaMonetaria.html

http://www.nbp.pl/en/publikacje/index.html

http://www.mas.gov.sg/resource/index.html

http://www.resbank.co.za/ibd/fwdcover.html

http://www.snb.ch/e/publikationen/publi.html

http://www.bot.or.th/BOTHomepage/BankAtWork/Monetary&FXPolicies/EXPolicy/8-23-2000-Eng-i/exchange_e.htm

http://www.federalreserve.gov/pf/pdf/frspurp.pdf

Notes on the pages

Excellent descriptive page on mar-kets and intervention.

Foreign exchange policy.

Concise description of interven-tion policy.

Description of foreign exchangemarkets.

No work directly describing inter-vention policy but speeches andworking papers may relate.

1997 description of the ECB’sresponsibility in foreign exchangeintervention.

Speech by the CEO describingexchange rate policy.

Outline of the Bank of Japan’s for-eign exchange intervention operations.

Description of monetary andexchange rate policies.

Some descriptions of interventionin annual reports.

See the pamphlet, EconomicsExplorer #1.

Describes South Africa’s use offorward cover in somewhat tech-nical terms.

Some intervention descriptions inannual reports.

Very brief description of currentexchange rate policy.

See page 64 for a very briefdescription of intervention policy.

Appendix

Monetary Authority Web Pages Describing Intervention or Foreign Exchange Markets


MAY/JUNE 2001 11

Forecasting Inflationand Growth: Do PrivateForecasts Match Thoseof Policymakers? William T. Gavin and Rachel J. Mandal

Generally, we value forecasts for their accu-racy. In some cases, however, the forecaststhemselves are interesting because of

what they reveal about the forecaster. Monetarypolicymaker forecasts are important because theypartially reveal what policymakers believe will fol-low from their decisions.

Forecasts of inflation and real output (whethermade by Federal Reserve officials or private sectoreconomists) contain information that is importantfor changing the stance of monetary policy. Marketparticipants generally believe that Fed policymakerswill change their policy stance if the economyappears to be headed in a different direction fromwhat was expected at the time policy was adopted.Svensson (1997) and Svensson and Woodford (2000)explain why a central bank might want to target itsinflation forecast. The intuition in their explanationis that policymakers should look at everything thatis relevant when deciding to change the policystance. The trouble with looking at everything isthat there is so much information to process, oneneeds an organizing framework such as a fore-casting model. Forecasting models are developedto monitor incoming information and to weigheach piece appropriately. Forecasting modelsrange from the very largest, with over a thousandequations, to small models that are no more thansimple rules of thumb. Whether using a largeeconometric model or a simple rule of thumb,forecasters rarely use the values that come directlyfrom the model. Rather, they typically make judg-mental adjustments before reporting the forecasts.

In this article, we examine the role of forecasts

in the monetary policy process. Our focus is onthe forecasts of inflation and economic growth,the main policy objectives. Economic forecasts areimportant because they reflect incoming informa-tion about the current state of the economy,including the forecasters’ beliefs about monetarypolicy objectives. In the United States, there are noexplicit numerical objectives for output and infla-tion. Thus, policymaker forecasts are particularlyinteresting because they may reveal informationabout long-run policy goals.

Fed forecasts, unfortunately, are not readilyavailable to the public. We show that the Blue Chipconsensus forecasts, made by a group of privateeconomists, are a good stand-in for the policy-makers’ forecasts. This is important because thepolicymakers in the Federal Reserve, the membersof the Federal Open Market Committee (FOMC),reveal their forecasts only sparingly and after policydecisions are made. First, we show how well theforecasts match. We find that the forecasts of eco-nomic growth are very similar and appear to beabout equal on average. The result for inflationforecasts is more interesting. Here we see that theprivate sector economists generally predictedhigher inflation than did Fed policymakers,especially in the 1980s. The Blue Chip economistsdid not believe that the FOMC would achieve andmaintain such a low inflation rate in the 1980s.Since 1995, the forecasts have converged. Evi-dently, the FOMC has achieved some credibilitywith the Blue Chip economists.

When researchers want to know the history ofpolicymakers’ forecasts, they typically go to theFed’s briefing documents to extract the forecastsof the research staff at the Board of Governors. Weshow that the Blue Chip forecasts for output are asgood a proxy for Fed policymakers’ views as arethe research staff forecasts. In the case of inflation,the results vary with the time horizon. Generally,the Blue Chip consensus forecasts for inflationmatch the policymakers’ forecasts at shorter hori-zons while the research staff forecasts are closer atthe longest horizon.

Finally, we examine the use of alternative fore-casts in a version of the Taylor rule, a popularcharacterization of monetary policy actions. It ispopular because it is a simple summary of a com-plicated policy process. It is expressed as:

(1) ,

where FFtA is the federal funds rate target chosen

FF r y ytA e

t tT

t tF= + + - + -- - - -p p p1 1 1 10 5 0 5. .( ) ( )

William T. Gavin is a vice president and senior economist andRachel Mandal is a research associate at the Federal Reserve Bankof St. Louis. The authors thank Dean Croushore and Dinah Macleanfor helpful comments. The original version of this article, winner ofthe Edmund A. Mennis Contributed Paper Award for 2000, appearedin the January 2001 issue of Business Economics, the journal of theNational Association for Business Economics (NABE). For moreinformation, visit NABE’s Web site at <www.nabe.com>.

12 MAY/JUNE 2001

R E V I E W

by the FOMC, re is the long-run equilibrium realinterest rate (assumed by Taylor to be equal to 2percent per year), pt–1 is the average inflation rateobserved over the previous four quarters, pT is theinflation target (which Taylor assumed to be equalto 2 percent per year), yt–1 is last period’s realgross domestic product (GDP) measured inlogarithms, and yt

F–1 is last period’s potential real

GDP measured in logarithms. The term in thebracket, (yt–1 – yt

F–1), is approximately equal to the

percentage deviation of GDP from the perceivedlevel of potential GDP.

This backward-looking rule prescribes settingsfor the federal funds rate, the Fed’s short-termpolicy instrument, according to the deviation ofthe past year’s inflation from a 2 percent targetand the deviation of last period’s GDP from a mea-sure of potential GDP. We begin by showing thathistorical analysis of the Taylor rule should usereal-time data; that is, data that were availablewhen the federal funds rate target was being set.We show that the forward-looking rule based onpolicymaker forecasts is virtually identical to onebased on Blue Chip consensus forecasts. Neitherdoes quite as well as the backward-looking ruleusing real-time data; however, all three versions ofthe Taylor rule do much better at explaininghistorical movement in the federal funds rate thando rules based on the current revised data.Because purely forward-looking rules may beinherently unstable, we also examine a combina-tion rule that includes both lagged values ofinflation and the output gap using real-time dataand the Blue Chip forecasts of the current-yearinflation and output gap.1 This rule with bothbackward- and forward-looking elements matchesthe actual federal funds rate slightly better thanthe rule based on real-time data.

FOMC AND BLUE CHIP FORECASTS

FOMC members prepare forecasts for Congres-sional testimony twice a year.2 This testimony wasmandated by the Full Employment and BalancedGrowth Act of 1978. Section 108 of this act explic-itly required the Fed to submit “written reportssetting forth (1) a review and analysis of recentdevelopments affecting economic trends in thenation; (2) the objectives and plans … with respectto the monetary and credit aggregates …; and (3)the relationship of the aforesaid objectives andplans to the short-term goals set forth in the most

recent Economic Report of the President …” Inorder to satisfy the third item, the Federal ReserveChairman began reporting a summary of Fed policy-makers’ forecasts to Congress in July 1979. Sincethen, similar summaries of forecasts have beenreported every February and July.3 Forecasts aremade of annual, fourth-quarter-over-fourth-quarter growth rates for nominal GDP, real GDP,and inflation.4 Fed policymakers also forecast theaverage level of unemployment for the fourthquarter of the year. In February, the forecasts per-tain to the current calendar year (referred to belowas the 12-month-ahead forecast). In July, forecastsare updated for the current calendar year (6-month-ahead forecasts) and preliminary projections aremade for the next calendar year (18-month-aheadforecasts).

We focus on the forecasts of real output growthand inflation because they best capture monetarypolicy objectives. We use the output price deflatoras the measure of inflation primarily because ithas been consistently forecasted throughout theentire period. Even when the Fed was reportingthe forecast for inflation based on the consumerprice index (from 1989 through 1999), there wasalso a forecast for both nominal and real output,so there was always an implied forecast for theoutput deflator.

Individual Federal Reserve officials submittheir economic forecasts based on their judgmentabout the appropriate policy to be followed overthe coming year. These individual projections maybe revised after the FOMC adopts a specific policy.The revised projections are then reported as arange, listing the high and low values for each item,and as a central tendency that omits extreme fore-casts and is meant to be a better representation of

1 See Woodford (2000) for a summary of the argument that purelyforward-looking rules may lead to instability.

2 The FOMC is the policymaking committee of the Federal ReserveSystem. When the Board is full, the Committee consists of the 7governors of the Board, the president of the Federal Reserve Bankof New York, and 4 of the remaining 11 Federal Reserve Bank presi-dents who serve on a rotating basis. All 12 presidents attend everymeeting, contribute to the discussion, and provide forecasts thatare summarized in testimony to the Congress. The Green Book is abriefing document with macroeconomic forecasts prepared by staffeconomists at the Board of Governors about three workdays beforeeach FOMC meeting.

3 This reporting requirement has now expired, but the Fed providedforecasts to Congress on July 20, 2000, and February 13, 2001.These data are not included in this study.

4 The Fed switched from GNP to GDP in 1992.


MAY/JUNE 2001 13

the consensus view. In this article, we define theconsensus FOMC forecast as the midpoint of thiscentral tendency range.

The Blue Chip consensus forecasts are takenfrom the February and July reports. These forecastsare collected on the first three working days of themonth, and the information available to privatesector economists is approximately the same asthe information available to the FOMC memberswhen they make their forecasts. Most importantly,both groups usually had the latest information onthe price indexes from the Bureau of Labor Statis-tics and the most recent report on actual GDPfrom the Bureau of Economic Analysis.

Figure 1 is a scatter diagram with trianglesshowing the relation between the consensus GDPgrowth forecasts made by the FOMC between1983 and 1994 and those made by the Blue Chipeconomists during the same period. We start in1983 because that is when the Federal Reservefirst began to report the central tendency of theforecasts. It was also the first year that theyreported forecasts for all the participants: FOMCmembers and nonvoting Federal Reserve Bankpresidents.5

If the FOMC and Blue Chip forecasts wereexactly the same, they would lie on the 45-degreeline shown. As Figure 1 shows, the forecasts werequite similar and seem to be distributed evenlyabove and below the 45-degree line. That is, there

does not seem to be any tendency for the BlueChip economists to systematically forecast moreor less output growth than the FOMC.

The same cannot be said of the inflation fore-casts. The triangles in Figure 2, where most of thepoints lie above the 45-degree line, show that theBlue Chip economists usually forecasted higherinflation than did the FOMC. The period from1983 to the present has been a period of moderateand falling inflation. Throughout, the FederalReserve has had a goal of eliminating inflation. Ingeneral, the FOMC’s forecasts of inflation havebeen lower than the Blue Chip forecasts. However,as inflation became lower in the 1990s, theforecasts have converged, indicating that theprivate sector has gained confidence in the Fed’sability to deliver low inflation. So, although theBlue Chip inflation forecasts have not always beenunbiased indicators of the FOMC’s inflationforecasts, they have been better in recent years.

GREEN BOOK FORECASTS

The Green Book forecast is put together by alarge staff of economists at the Board of Governorsin Washington, D.C. It is prepared for the FOMCmembers who read it in advance of the meetings

5 In July 1979, the Fed reported a range of Board member forecasts(governors only). From 1980 through 1982, the Fed reported arange of forecasts for FOMC members.

Figure 1

Output Forecasts (1983 to 1994)

0

1

2

3

4

5

6

7

0 1 2 3 4 5 6 7Midpoint of FOMC Central Tendency

Blue Chip

Green Book

45 line

Figure 2

Inflation Forecasts (1983 to 1994)

0

1

2

3

4

5

6

0 1 2 3 4 5 6Midpoint of FOMC Central Tendency

Blue Chip

Green Book

45 line

14 MAY/JUNE 2001

R E V I E W

and receive an oral presentation of this forecast atthe meeting. These forecasts are only available tothe public five years after they are made.

Romer and Romer (2000) compare the GreenBook forecasts with private sector forecasts usingquarterly data from 1965 through 1991 and fore-casts over several horizons (usually from forecastsof the current quarter out to seven quarters ahead).They present convincing evidence that the GreenBook inflation forecasts have been more accuratethan the private forecasts, including the Blue Chipconsensus (for the period from 1980 to 1991).They also report that the Green Book forecasts ofoutput were better than private sector forecasts,but the evidence for output forecasts is weaker.

The Green Book forecasts from 1983 through1994 are depicted as circles in Figures 1 and 2.Casual observation suggests that the Green Bookforecasts and the Blue Chip consensus representthe policymakers’ consensus equally well. Thesescatter diagrams combine forecasts across thethree horizons of 6, 12, and 18 months ahead.

Table 1 gives more detailed information abouthow well the Blue Chip consensus and the GreenBook forecast match the FOMC consensus. Resultsare reported for the combined forecasts (combinedover the three forecasting horizons) and for thethree separate horizons. The forecast error inTable 1 is defined as the difference between thealternative forecast (Blue Chip consensus or GreenBook) and the midpoint of the FOMC centraltendency forecast. We report root-mean-squarederrors (RMSE) for both inflation and outputforecasts.

The results are interesting. On average, the dif-ferences in errors between the Green Book andBlue Chip are larger for the real output forecaststhan they are for the inflation forecasts. For both

real output and inflation, the Blue Chip consensusis closer to the FOMC forecast than is the GreenBook. For the first 12 years after the FOMC beganreporting the central tendency, the Blue Chip fore-cast has provided a good measure of the FOMC’sview of the future, as least as good as one wouldget by knowing the Green Book forecast.

RELATIVE ACCURACY

1983 Through 1994

Table 2 reports the relative accuracy of realoutput forecasts to the real-time data from 1983through 1994. For the separate and combinedhorizons, we compare the individual forecasts tothe value that was first reported by the Bureau ofEconomic Analysis.6 The Blue Chip forecasts arebest (lowest RMSE) for the 12- and 18-monthhorizons. The FOMC’s forecasts have the lowestRMSE at the 6-month horizon. In none of thesecases are the Green Book forecasts of real outputbest.7

The Green Book fares better, however, forinflation forecasts from 1983 through 1994, asshown in Table 3. Earlier, we saw that the BlueChip inflation forecasts were generally above theFOMC’s forecasts in the 1980s. Here we see that allthree forecasts, on average, predicted higher thanactual inflation, with the FOMC forecasts sand-wiched between the Blue Chip forecasts on the

6 We used the vintage data sets from the Federal Reserve Bank ofPhiladelphia described in Croushore and Stark (1999).

7 This is surprising given the conclusions in Romer and Romer (2000).They examined an earlier and longer sample with more frequentforecasts over more horizons. We examine only those dates andforecast horizons for which the central tendency of FOMC members’forecasts were reported to Congress.

Blue Chip Versus Green Book as a Proxy for FOMC Forecasts (1983 to 1994)

RMSE of output forecast RMSE of inflation forecast

Blue Chip Green Book Blue Chip Green Book

All 3 horizons 0.22 0.36 0.32 0.38

6-Month horizon 0.17 0.35 0.21 0.25

12-Month horizon 0.25 0.32 0.32 0.38

18-Month horizon 0.24 0.40 0.40 0.47

NOTE: Bold typeface indicates a better proxy for the midpoint of the FOMC tendency.

Table 1


MAY/JUNE 2001 15

high end and the more accurate Green Book fore-casts on the low end.

1995 Through 1999

Table 4 examines the accuracy of the BlueChip and FOMC real output forecasts from 1995through 1999. Again, we report results based onthe combined data sets and also separately foreach forecast horizon. For these five years, boththe Blue Chip and the FOMC policymakers’ fore-casts for real output growth were about 1 percentbelow actual. The large bias in the mean errorreflects the ongoing surprise about the strength ofeconomic growth and upward revisions toestimates of the underlying trend. We find that inthe last five years, on average, the FOMC has beenmore accurate, as measured by the RMSE, than theBlue Chip at all forecast horizons.

We saw in Figure 2 that the FOMC and BlueChip forecasts converged as inflation came downin the 1990s. Table 5 looks at the accuracy of theBlue Chip and FOMC inflation forecasts over thelast five years of the sample. Both the FOMC and

Blue Chip forecasts predicted higher than actualinflation from 1995 through 1999. The FOMCinflation forecasts have been slightly moreaccurate than the Blue Chip forecast for all threeforecast horizons.

Although the FOMC forecasts were more accu-rate than the Blue Chip forecasts, the forecastswere not far apart. On average for all three horizons,the Blue Chip consensus for GDP growth was atenth of a percentage point below the FOMC’s, andthe Blue Chip consensus for inflation was one-tenth higher than the FOMC’s. The five yearsreported in Tables 4 and 5, 1995 through 1999,have been characterized by surprisingly high realGDP growth and surprisingly low inflation, as isseen by the negative mean errors for output growthand the positive mean errors for inflation.

USING FORECASTS IN TAYLOR-TYPERULES

In this section we use a simple policymakingframework to see whether the differences betweenthe Blue Chip and FOMC forecasts are economically

Table 2

Accuracy of Output Forecasts (1983 to 1994)

Mean error RMSE

Blue Chip FOMC members Green Book Blue Chip FOMC members Green Book

All 3 horizons 0.04 0.06 –0.06 0.94 0.96 1.05

6-Month horizon 0.02 0.05 –0.02 0.76 0.74 0.80

12-Month horizon –0.11 –0.08 –0.15 1.05 1.11 1.23

18-Month horizon 0.22 0.22 –0.02 0.99 1.00 1.06

NOTE: Best forecast indicated by bold typeface.

Table 3

Accuracy of Inflation Forecasts (1983 to 1994)

Mean error RMSE


All 3 horizons 0.69 0.46 0.35 0.92 0.80 0.65

6-Month horizon 0.45 0.33 0.21 0.64 0.55 0.36

12-Month horizon 0.60 0.41 0.26 0.79 0.74 0.61

18-Month horizon 1.01 0.65 0.57 1.23 1.05 0.88


16 MAY/JUNE 2001

R E V I E W

significant. Taylor (1993) proposed characterizingpast Fed policy as if it were made according to aformula similar to equation (1), which has come tobe known as the Taylor rule.8

Rotemberg and Woodford (1999) show that arule of this form can be derived as an optimalpolicy under certain conditions. Clarida, Gali, andGertler (1999) show that a rule of this type can beoptimal in a dynamic, forward-looking IS/LMmodel in which the central bank’s loss function isquadratic in deviations of inflation from target andoutput from potential. Even if the central bankcares only about the inflation objective, the nom-inal interest rate target may be set as a function ofthe state of the economy. If the real interest rate isprocyclical, adjusting the federal funds rate targetfor changes in the gap between potential andactual GDP may be a method for taking intoaccount the cyclical deviation of the real interestrate from the long-run equilibrium value.9

While clearly not advocating that any centralbank follow any such simple rule slavishly, Taylorrecommended his rule as a reference point in

debates about whether a policy change might beneeded. Indeed, that has happened as manycentral banks now regularly monitor variations ofthe original Taylor rule. Figure 3 shows the quarterlyaverage federal funds rate and our calculation ofthe federal funds rate target implied by the Taylorrule for the period from 1983 to 1999.

We begin by showing the federal funds ratetarget implied by equation (1).10 As Figure 3 shows,the rule does not do particularly well during theperiods before 1990 or after 1994. Table 6 showsthat the federal funds rate target, predicted byusing current revised data, is, on average, 166 basispoints below the actual federal funds rate.

8 In his 1993 paper, Taylor used current year values for the GDP gapand inflation. Since current year data are unknown at the time poli-cy is made, we have used lagged values.

9 For recent evidence suggesting that the real interest rate is procycli-cal, see Dotsey and Scholl (2000).

10 Note that the usefulness of the Taylor rule has been questioned bymany researchers, including recent articles by Hetzel (2000), Kozicki(1999), McCallum (1999), and Orphanides (1998).

Table 4

Accuracy of Output Forecasts (1995 to 1999)

Mean error RMSE


All 3 horizons –1.13 –1.02 NA 1.46 1.35 NA

6-Month horizon –0.52 –0.53 NA 0.81 0.73 NA




Table 5

Accuracy of Inflation Forecasts (1995 to 1999)

Mean error RMSE


All 3 horizons 0.59 0.48 NA 0.72 0.64 NA

6-Month horizon 0.36 0.29 NA 0.43 0.39 NA





MAY/JUNE 2001 17

Figure 3 also includes the Taylor rule for thefederal funds rate target using real-time data forGDP and inflation and a forecast for potential GDPfrom a recursive model that fits a quadratic timetrend to the real-time data. As the figure shows,there is an important difference in the targetcalculated for the federal funds rate when we usethe real-time data. Contrary to the case using cur-rently available revised data, the real-time Taylorrule generally lies above the actual federal fundsrate. The right-most column in Table 6 shows thatthe average deviation was 34 basis points. Theseresults show that ex post policy rules based onrevised data may do a poor job of replicatingactual policy choices.

Figure 4 includes two versions of a forward-looking Taylor rule where we modify Taylor’sgeneral specification by replacing the backward-looking measures of inflation and output withFOMC and Blue Chip forecasts for the calendaryear. The modified Taylor rule used is

(2) ,

where pte is the forecast of fourth-quarter-over-

fourth-quarter inflation for the current year and( yt

e – ytF ) is the output gap expected for the current

year. We use the real-time data and our quadratictime trend to predict potential GDP in the fourthquarter of each year. We construct a fourth-

FF r y ytB e

te

te T

te

tF= + + - + -p p p0 5 0 5. .( ) ( )

quarter forecast of the level of GDP using theactual real-time value of the previous fourth-quarter level of GDP and the fourth-quarter-over-fourth-quarter forecast of GDP for the current year.Whether we use forecasts from the FOMC or BlueChip, the implications for the federal funds ratetarget are almost identical.

In Table 6, the RMSE between the actualfederal funds rate and the target predicted by thealternative Taylor rules are given along a diagonalin parentheses. For this period, using these fore-casts, the backward-looking rule using real-timedata predicts the actual federal funds rate slightlymore accurately than do the forward-looking rules.The forward-looking version using the Blue Chipconsensus forecasts is more accurate than theversion using FOMC forecasts. However, the meanerror for the FOMC version is closest to zero. As wesaw in Figure 4, the Blue Chip and FOMC versionsof the Taylor rule seem to move in tandem. Thecorrelation between these versions of the Taylorrule is 0.99.

Bernanke and Woodford (1997) have arguedthat purely forward-looking Taylor rules may notbe practical. Chari (1997) explains simply,

Suppose, for instance, that the central bankwants to stabilize inflation rates and privateforecasters have information that is notavailable to the central bank about future

Figure 3

Taylor Rules: Current Versus Real-Time Data

Taylor Rule with Current Vintage Data

1983 1985 1987 1989 1991 1993 1995 1997 19990

2

4

6

8

10

12

Taylor Rule withReal-Time Data

Actual Federal Funds Rate

Figure 4

Taylor Rules: Blue Chip Versus FOMC Forecasts (Semi-Annual Data)

1983 1985 1987 1989 1991 1993 1995 1997 19990

2

4

6

8

10

12 Actual FF

Blue Chip

Fed Policymakers

18 MAY/JUNE 2001

R E V I E W

inflation. The central bank could use pri-vate forecasts of inflation to choose itspolicy instrument. The problem is that ifthe central bank is completely effective inusing its policy instrument to stabilize in-flation, private forecasts of inflation shouldrationally be the central bank’s inflationtarget in which case, private forecasts pro-vide no information about inflation! Thisparadox arises because market forecastsof a goal variable depend upon the centralbank’s policy rule and if the central bankused the information well, market fore-casts will not be informative. (p. 685)

Woodford (2000) recommends policies thatinclude both backward- and forward-lookingelements. We create a combination rule that usesboth the lagged values of inflation and the outputgap as well as the Blue Chip forecasts for the cur-rent year. It is equivalent to taking an average of thereal-time Taylor rule (FFt

A) and the forward-lookingrule using Blue Chip forecasts (FFt

B). The results forthis combination rule are given in the bottom rowof Table 6. The federal funds rate target that comesout of this rule has the highest correlation withthe actual federal funds rate (0.88) and the lowestRMSE (1.02) of all the rules that we considered.

CONCLUSION

We have found that the Blue Chip consensusappears to have been closely matched to the mid-

point of the FOMC’s central tendency forecasts.During the period from the beginning of 1983through the summer of 1994, the Blue Chipforecasts for output were not only more closelyrelated to the FOMC’s output forecasts, but theywere slightly more accurate than the forecasts inthe Green Book. The Green Book forecasts of infla-tion were much more accurate than were the BlueChip’s during the period between 1983 and 1994.Nevertheless, the Blue Chip forecasts were still asclosely related to the FOMC forecasts as were theGreen Book forecasts.

In the period since 1994, the FOMC consen-sus has been more accurate than the Blue Chipconsensus for both inflation and output, but notby much. During the period from 1995 through1999, inflation has been lower than expectationswhile the real economy has been unexpectedlystrong.

For the entire period, the differences betweenthe Blue Chip consensus forecasts and the mid-point of the central tendency are not statisticallyor economically relevant for the policymakingprocess, at least not as that process has beencharacterized by Taylor (1993). We should not besurprised to learn that the Blue Chip forecasts ofinflation and output are highly correlated withFOMC forecasts. Both the FOMC members andthe economists who contribute to the Blue Chipconsensus observe the same statistical releasesand use similar economic theories to interpretthe data.

Alternative Versions of the Taylor Rule

Actual Current FOMC Combination: federal revised Real-time Blue Chip members’ real time and Mean

funds rate data data forecast forecasts Blue Chip error

Current revised data 0.73 (1.42) –1.66

Real-time data 0.87 0.82 (1.04) 0.34

Blue Chip forecast 0.84 0.67 0.92 (1.16) 0.15

FOMC members’ 0.82 0.67 0.91 0.99 (1.23) –0.03forecasts

Combination: real time 0.88 0.76 0.98 0.98 0.97 (1.02) 0.24and Blue Chip

NOTE: Correlations among the alternative predictions of the Taylor rule and the actual federal funds rate are shown in bold. RMSEare shown in parentheses (Taylor rule minus actual federal funds rate). Right column shows the mean error for each version of theTaylor rule.

Table 6


MAY/JUNE 2001 19

REFERENCES

Bernanke, Ben S. and Woodford, Michael. “InflationForecasts and Monetary Policy.” Journal of Money, Creditand Banking, November 1997, 29(4, Part II), pp. 653-84.

Chari, V.V. “Comment on Inflation Forecasts and MonetaryPolicy.” Journal of Money, Credit and Banking, November1997, 29(4, Part II), pp. 685-86.

Clarida, Richard; Gali, Jordi and Gertler, Mark. “TheScience of Monetary Policy: A New KeynesianPerspective.” Journal of Economic Literature, December1999, 37(4), pp. 1661-707.

Croushore, Dean and Stark, Tom. “A Real-Time Data Set forMacroeconomists.” Working Paper 99-4, Federal ReserveBank of Philadelphia, June 1999.

Dotsey, Michael and Scholl, Brian. “The Behavior of theReal Rate of Interest Over the Business Cycle.”Unpublished manuscript, Federal Reserve Bank ofRichmond, 27 February 2000.

Hetzel, Robert L. “A Critical Appraisal of the Taylor Rule.”Unpublished manuscript, Federal Reserve Bank ofRichmond, 11 February 2000.

Kozicki, Sharon. “How Useful Are Taylor Rules for MonetaryPolicy?” Federal Reserve Bank of Kansas City EconomicReview, Second Quarter 1999, 84(2), pp. 5-33.

McCallum, Bennett T. “Recent Developments in theAnalysis of Monetary Policy Rules.” Federal Reserve

Bank of St. Louis Review, November/December 1999,81(6), pp. 3-12.

Orphanides, Athanasios. “Monetary Policy Rules Based onReal-Time Data.” Finance and Economics DiscussionSeries No. 1998-3, Board of Governors of the FederalReserve System, January 1998.

Romer, Christina D. and Romer, David H. “Federal ReservePrivate Information and the Behavior of Interest Rates.”American Economic Review, June 2000, 90(3), pp. 429-57.

Rotemberg, Julio J. and Woodford, Michael. “Interest RateRules in an Estimated Sticky Price Model,” in John B.Taylor, ed., Monetary Policy Rules. Chicago: TheUniversity of Chicago Press, 1999, pp. 57-119.

Svensson, Lars E.O. “Inflation Forecast Targeting: Imple-menting and Monitoring Inflation Targets.” EuropeanEconomic Review, June 1997, 41(6), pp. 1111-46.

___________ and Woodford, Michael. “Indicator Variablesfor Optimal Policy.” Presented at Structural Change andMonetary Policy Conference, Federal Reserve Bank ofSan Francisco, 2000.

Taylor, John B. “Discretion Versus Policy Rules in Practice.”Carnegie-Rochester Conference Series on Public Policy,December 1993, 39, pp. 195-214.

Woodford, Michael. “Pitfalls of Forward-Looking MonetaryPolicy.” American Economic Review, May 2000, 90(2), pp.100-4.

20 MAY/JUNE 2001

R E V I E W


MAY/JUNE 2001 21

Toward a NewParadigm in OpenEconomy Modeling:Where Do We Stand? Lucio Sarno

In the last few decades, there have been a num-ber of important developments, both theoreti-cal and empirical, in open economy macro-

economics and exchange rate economics (see, forexample, Sarno and Taylor, 2001a, b). Also, theincreasing availability of high-quality macroeco-nomic and financial data has stimulated a largeamount of empirical work. While our understand-ing of exchange rates has improved as a result,many challenges and questions remain. This paperselectively surveys the recent literature on “new”open economy macroeconomics. This literature,stimulated by the work of Obstfeld and Rogoff(hereafter OR) (1995), reflects the attempt byresearchers to formalize exchange rate determina-tion in the context of dynamic general equilibriummodels with explicit microfoundations, nominalrigidities, and imperfect competition.1

The main objective of this research program isto develop a new workhorse model for open econ-omy macroeconomic analysis. Relative to the stillubiquitous Mundell-Fleming-Dornbusch (MFD)model (Mundell, 1962, 1963; Fleming, 1962;Dornbusch, 1976), new open economy modelsoffer a higher standard of analytical rigor comingfrom fully specified microfoundations; they offerthe ability to perform welfare analysis and rigor-ously discuss policy evaluation in the context of aframework that allows for market imperfectionsand nominal rigidities. On the other hand, themain virtue of the MFD model is its simpler ana-lytical structure, which makes it easy to discuss in

policy circles. Because the predictions of newopen economy models are sensitive to the partic-ular specification of the microfoundations, policyevaluation and welfare analysis depend on thespecification of preferences and nominal rigidi-ties. In turn, this generates a need for the profes-sion to agree on the “correct” or at least “prefer-able” specification of the microfoundations.

The present paper reviews the key contribu-tions in new open economy macroeconomics inthe last five to six years, also assessing how theintellectual debate stimulated by OR has led tomodels that reflect reality more satisfactorily overtime. The paper also discusses some of the mostcontroversial issues that currently still prevent anyof the models in this area to emerge as a new para-digm for open economy modeling and describesthe directions taken by the latest literature.

The remainder of the paper is set out as fol-lows. The first section provides a review of theseminal paper in this literature, proposing the so-called redux model, while the second section cov-ers a number of variants and generalizations ofthe redux model that permit allowance for alter-native nominal rigidities, pricing to market, alter-native preference specifications, and alternativefinancial markets structures. I then discuss somestochastic extensions of these models, focusing ontheir implications for the relationship betweenuncertainty and exchange rates in the third sec-tion. Some new directions taken by the latest liter-ature on stochastic open economy modeling aredescribed in the fourth section. A final sectionpresents some concluding remarks.

THE REDUX MODEL

The Baseline Model

OR (1995) is the study often considered ashaving initiated the literature on new open econo-my macroeconomics (see, for example, Lane, 1999,and Corsetti and Pesenti, 2001). However, a pre-cursor of the OR (1995) model that deserves to benoted here is the model proposed by Svensson

Lucio Sarno is a reader in economics and finance at the WarwickBusiness School, University of Warwick, and a research affiliate ofthe Centre for Economic Policy Research, London. This paper waswritten in part while the author was a visiting scholar at the FederalReserve Bank of St. Louis. The author thanks the United KingdomEconomic and Social Research Council (ESRC) for providing finan-cial support (grant No. L138251044) and Gaetano Antinolfi, JamesBullard, Giancarlo Corsetti, Brian Doyle, Fabio Ghironi, Peter Ireland,Marcus Miller, Chris Neely, Michael Pakko, Neil Rankin, Mark Taylor,and Dan Thornton for constructive comments. Paige Skiba providedresearch assistance. The views expressed are those of the authorand should not be interpreted as reflecting those of any institution.

1 An early draft of this paper covered some of the models discussedbelow in a more technical fashion. The preliminary technical ver-sion is available from the author upon request (Sarno, 2000). Walsh(1998) also provides an excellent treatment of the redux model,especially focusing on monetary issues. See also the comprehensivetextbook treatment of the early new open economy literature byOR (1996) and its selective coverage by Lane (1999). For a treatmentof the role of imperfect competition in macroeconomic models, seethe survey by Dixon and Rankin (1994).

22 MAY/JUNE 2001

R E V I E W

and van Wijnbergen (1989). They present a stochas-tic, two-country, neoclassical rational-expectationsmodel with sticky prices that are optimally set bymonopolistically competitive firms, where possibleexcess capacity is allowed for to examine interna-tional spillover effects of monetary disturbances onoutput. In contrast to the prediction of the MFDmodel that a monetary expansion at home leadsto a recession abroad, the paper suggests thatspillover effects of monetary policy may be eitherpositive or negative, depending on the relative sizeof the intertemporal and intratemporal elasticitiesof substitution in consumption. It is also fair to saythat the need for rigorous microfoundations inopen economy models is not novel in new openeconomy macroeconomics and has been empha-sized by several papers prior to OR (1995); notableexamples are Lucas (1982), Stockman (1980, 1987),and Backus, Kehoe, and Kydland (1992, 1994, 1995),among others.

The baseline model proposed by OR (1995) isa two-country, dynamic general equilibrium modelwith microfoundations that allows for nominalprice rigidities, imperfect competition, and a con-tinuum of agents who both produce and con-sume. Each agent produces a single differentiatedgood. All agents have identical preferences, char-acterized by an intertemporal utility function thatdepends positively on consumption and realmoney balances but negatively on work effort;effort is positively related to output. The exchangerate is defined as the domestic price of the foreigncurrency. The two countries are called Home andForeign, respectively.

Because the model assumes no impedimentsto international trade, the law of one price (LOOP)holds for each individual good and purchasingpower parity (PPP) holds for the internationallyidentical aggregate consumption basket. PPP is theproposition that national price levels should beequal when expressed in a common currency; theLOOP is the same proposition applied to individu-al goods rather than a consumption basket. Sincethe real exchange rate is the nominal exchangerate adjusted for relative national price levels, vari-ations in the real exchange rate represent devia-tions from PPP. Hence, the LOOP and continuousPPP imply a constant real exchange rate, whilelong-run PPP (where temporary deviations fromPPP are allowed for) implies mean reversion inthe real exchange rate.

OR also assume that both countries can bor-row and lend in an integrated world capital mar-

ket. The only internationally traded asset is a risk-less real bond, denominated in the consumptiongood. Agents maximize lifetime utility subject totheir budget constraints (identical for domesticand foreign agents). Utility maximization thenimplies three clearly interpretable conditions. Thefirst is the standard Euler equation, which impliesa flat time path of consumption over time. Thesecond condition is the money market equilibriumcondition that equates the marginal rate of substi-tution of consumption for the services of realmoney balances to the consumption opportunitycost of holding real money balances (the nominalinterest rate); the representative agent directly bene-fits from holding money in the utility functionbut loses the interest rate on the riskless bond aswell as the opportunity to eliminate the cost ofinflation. (Note that money demand depends onconsumption rather than income in this model.)The third condition requires that the marginal util-ity of the higher revenue earned from producingone extra unit of output equals the marginal dis-utility of the needed effort, and so can be inter-preted as a labor-leisure trade-off equation.

In the special case when net foreign assets arezero and government spending levels are equalacross countries, OR solve the model for incomeand real money balances. Because this model isbased on a market structure with imperfect com-petition where each agent has some degree ofmarket power arising from product differentia-tion, the solutions of the model imply that steady-state output is suboptimally low. As the elasticityof demand (say q ) increases, the various goodsbecome closer substitutes and, consequently, themonopoly power decreases. As q approachesinfinity, output increases, tending to the level cor-responding to a perfectly competitive market.

The main focus of OR (1995) is the impact ofa monetary shock on real money balances andoutput. Under perfectly flexible prices, a perma-nent shock produces no dynamics and the worldeconomy remains in steady state (prices increaseby the same proportion as the money supply).That is, an increase in the money supply has noreal effects and cannot remedy the suboptimaloutput level. Money is neutral.2

2 Note that in the redux model and in a number of subsequentpapers, monetary shocks are discussed without a formalization ofthe reaction functions of the monetary authorities. However, somerecent studies have formally investigated reaction functions in newopen economy macroeconomic models; see, for example, Ghironiand Rebucci (2000) and the references therein.


MAY/JUNE 2001 23

With prices displaying stickiness in the shortrun, however, monetary policy may have realeffects. If the money supply increases, becauseprices are fixed, the nominal interest rate decreasesand hence the exchange rate depreciates. This isbecause, due to arbitrage in the foreign exchangemarket, uncovered interest parity holds. Foreigngoods become more expensive relative to domes-tic goods, generating a temporary increase in thedemand for domestic goods and inducing anincrease in output. Consequently, monetary shocksgenerate real effects on the economy. But how canone ensure that producers are willing to increaseoutput? If prices are fixed, output is determinedby demand. Because a monopolist always pricesabove the marginal cost, it is profitable to meetunexpected demand at the fixed price. Noting thatin this model the exchange rate rises less than themoney supply, currency depreciation shifts worlddemand toward domestic goods, which causes ashort-run rise in domestic income. Home resi-dents consume some of the extra income, but,because they want to smooth consumption overtime, they save part of it. Therefore, although inthe long run the current account is balanced, inthe short run Home runs a current account sur-plus. With higher long-run wealth, Home agentsshift from work to leisure reducing Home output.Nevertheless, because Home agents’ real incomeand consumption rise in the long run, the ex-change rate does not necessarily depreciate.3

Unlike the scenario in a Dornbusch-typemodel, the redux model does not yield exchangerate overshooting. The exchange rate effect issmaller the larger the elasticity of substitution, q;as q approaches infinity, Home and Foreign goodsbecome closer substitutes, producing larger shiftsin demand with the exchange rate changing onlyslightly.

Finally, a monetary expansion leads to a first-order welfare improvement.4 Because the priceexceeds the marginal cost in a monopolistic equi-librium, global output is inefficiently low. Anunanticipated money shock raises aggregatedemand stimulating production and mitigatingthe distortion.

Summing up, in the redux model, monetaryshocks can generate persistent real effects, affect-ing consumption and output levels and theexchange rate, although both the LOOP and PPPhold. Welfare rises by equal amounts at home andabroad after a positive monetary shock, and pro-

duction is moved closer to its efficient (perfectlycompetitive market) level. Adjustment to thesteady state occurs within one period, but moneysupply shocks can have real effects lasting beyondthe time frame of the nominal rigidities becauseof the induced short-run wealth accumulation viathe current account. Money is not neutral, even inthe long run.

A Small Open Economy Version of theBaseline Model

The baseline redux model and most of thesubsequent literature on new open economymacroeconomics are based on a two-countryframework, which allows an explicit analysis ofinternational transmission channels and theendogenous determination of interest rates andasset prices. Nevertheless, similar, simpler modelsmay be constructed under the assumption of asmall open economy rather than a two-countryframework. In the small open economy version itis also easier to allow a distinction between trad-able and nontradable goods in the analysis. OR(1995) provide such an example in their Appendix.In this model, monopolistic competition charac-terizes the nontradable goods sector. The tradablegoods sector is characterized by a single homoge-nous tradable good that sells for the same priceall over the world, perfect competition, and flexi-ble prices. The representative agent in the smallopen economy, called Home, has an endowmentof the tradable good in constant quantity in eachperiod and monopoly power over the productionof one of the nontradable goods.

In this setup, a permanent monetary shockdoes not generate a current account imbalance.Because output of tradable goods is fixed, currentaccount behavior is determined by the time pathfor tradables consumption, which, under log-separable preferences and a discount rate equal to

3 Again, note that in this model money demand depends on con-sumption rather than income. Thus, an increase in consumptiondue to an increase in the nominal money supply raises moneydemand by the same proportion.

4 In order to produce more, Home agents have to work harder. Theeffects from reallocating consumption-production and leisure overtime are second-order, and the excess demand that leads to anincrease in production outweighs these effects. Of course, welfareresults depend upon the welfare function assumed. In the presentcontext, for example, it is important to note that inflation costs(obviously generated by an expansionary monetary policy) are notmodeled explicitly.

24 MAY/JUNE 2001

R E V I E W

the world interest rate, implies a perfectly flatoptimal time path for consumption. Hence, thecurrent account remains in balance. Unlike thescenario in the baseline redux model, however,exchange rate overshooting may occur in thismodel. Since the monetary shock does not producea current account imbalance, money is neutral inthe long run and the nominal exchange rate risesproportionately to the money stock. Because theconsumption elasticity of money demand is lessthan unity (by assumption), the nominal exchangerate overshoots its long-run level.5

Lane (1997) uses this small open economymodel to examine discretionary monetary policyand the impact of openness (measured by the rel-ative size of the tradables sector) on the equilibri-um inflation rate. A more open economy (with alarge tradables sector) gains less from “surprise”inflation because the output gain from a monetaryexpansion is exclusively obtained in the nontrad-ables sector and is relatively low. Since the equi-librium inflation rate under discretion is positivelyrelated to the gains from “surprise” inflation (Barroand Gordon, 1983), the model predicts that moreopen economies have lower equilibrium inflationrates (see also Kollmann, 1997; Velasco, 1997).

Lane (2001) further extends this model byconsidering an alternative specification of the util-ity function under which monetary shocks gener-ate current account imbalances. The sign of thecurrent account response is ambiguous, however;in fact, it depends on the interplay between theintertemporal elasticity of substitution, s , and theintratemporal elasticity of substitution, q. s governsthe willingness to substitute consumption acrossperiods, while q governs the degree of substitut-ability between traded and nontraded consump-tion. If s<q, a positive monetary shock generatesa current account surplus; however, a currentaccount deficit occurs if s>q, whereas the cur-rent account remains in balance if s=q. Hence,this model clearly illustrates how the results stem-ming from this class of models are sensitive to thespecification of the microfoundations.

The implications of small open economymodels of this class seem plausible. While the rel-evant literature (and consequently the rest of thissurvey) largely uses a two-country global econ-omy framework, I think it might also be worth-while to pursue research based on the small openeconomy assumption. Indeed, the small openeconomy assumption is plausible for most coun-

tries, except the United States. Furthermore, testingthe empirical implications of the small open econ-omy models discussed in this section represents anew line of research for applied economists.

RETHINKING THE REDUX MODEL

Nominal Rigidities

As mentioned earlier, subsequent work hasmodified many of the assumptions of the reduxmodel. In this section, I discuss modifications basedon the specification of nominal rigidities. The openeconomy literature surveyed here provides somenovel thoughts in this context and might generateevidence for choosing among alternative specifica-tions of stickiness in macroeconomic models.Whether the extension from closed to open econ-omy models does help to achieve consensus onthe specification of nominal rigidities remains to beseen (see, for example, the arguments presented byOR, 2000a, discussed below).

With respect to nominal rigidities, the reduxmodel assumes that prices are set one period inadvance, which implies that the adjustment toequilibrium is completed after one period. AsCorsetti and Pesenti (2001) emphasize, however, ifprice stickiness is motivated by fixed menu costs,firms have an incentive to adjust prices immediate-ly after a shock if the shock is large enough to vio-late their participation cost by raising the marginalcost above the price. Hence, the redux analysis maybe seen as plausible only within the relevant rangeof shocks.

Hau (2000) generalizes the redux model inthree ways to investigate the role of factor price(wage) rigidities and nontradables for the interna-tional transmission mechanism. First, followingBlanchard and Kiyotaki (1987), the model allowsfor factor markets and for nominal rigidities origi-nating from sticky factor prices (wages). Second,Hau assumes flexible price setting in local curren-cy and does not assume international goods arbi-trage. While the LOOP still holds because of opti-mal monopolistic price setting, nontradables inthe consumer price index produce deviationsfrom PPP. Third, unlike the scenario in the reduxanalysis, Hau also allows for nontradable goods.The main result of the paper is that factor pricerigidities have similar implications to rigid domes-

5 Indeed, this is exactly the same overshooting condition derived inthe Dornbusch (1976) model.


MAY/JUNE 2001 25

tic producer prices. In some sense, the results ofthe redux analysis are confirmed in the context ofa market structure with factor price rigidities. How-ever, nontradables modify the transmission mech-anism in important ways. A larger nontradablesshare implies that exchange rate movements aremagnified, since the money market equilibriumrelies on a short-run price adjustment carried outby fewer tradables. This effect is interesting since itmay help explain the observed high volatility of thenominal exchange rate relative to price volatility.

Within the framework of price level rigidities,however, a more sophisticated way of capturingprice stickiness is through staggered price settingthat allows smooth, rather than discrete, aggregateprice level adjustment. Staggering price models ofthe type developed by, among others, Taylor (1980)and Calvo (1983) are classic examples. Kollmann(1997) calibrates a dynamic open economy modelwith both sticky prices and sticky wages and thenexplores the behavior of exchange rates and pricesin response to monetary shocks with predeterminedprice and wage setting and Calvo-type nominalrigidities. His results suggest that Calvo-type nomi-nal rigidities match very well the observed highcorrelation between nominal and real exchangerates and the smooth adjustment in the price level,but they match less well correlations between out-put and several other macroeconomic variables.

Chari, Kehoe, and McGrattan (CKM) (1998,2000) link sticky price models to the behavior ofthe real exchange rate in the context of a new openeconomy macroeconomic model. They start bynoting that the data show large and persistent devi-ations of real exchange rates from PPP that appearto be driven primarily by deviations from the LOOPfor tradable goods. That is, real and nominalexchange rates are about six times more volatilethan relative price levels and both are highly persis-tent, with first-order serial correlations of about0.85 and 0.83, respectively, at annual frequency.CKM then develop a sticky price model with price-discriminating monopolists that produces devia-tions from the LOOP for tradable goods. However,their benchmark model, which has prices set forone quarter at a time and a unit consumption elas-ticity of money demand, does not come close toreproducing the serial correlation properties of realand nominal exchange rates noted above. A modelin which producers set prices for six quarters at atime and with a consumption elasticity of moneydemand of 0.27 does much better in generating

persistent and volatile real and nominal exchangerates. The serial correlations of real and nominalexchange rates are 0.65 and 0.66, respectively, andexchange rates are about three times more volatilethan relative price levels.

In a closely related paper, Jeanne (1998)attempts to assess whether money can generatepersistent economic fluctuations in a dynamicgeneral equilibrium model of the business cycle.Jeanne shows that a small nominal friction in thegoods market can make the response of output tomonetary shocks large and persistent if it is ampli-fied by real-wage rigidity in the labor market. Healso argues that, for plausible levels of real-wagerigidity, a small degree of nominal stickiness may besufficient for money to produce economic fluctua-tions as persistent as those observed in the data.6

OR (2000a), discussed in detail later in thispaper, develop a stochastic new open economymacroeconomic model based on sticky nominalwages, monopolistic competition, and exporters-currency pricing. Solving explicitly the wage-settingproblem under uncertainty allows the analysis ofthe welfare implications of alternative monetaryregimes and their impact on expected output andterms of trade. To motivate their model, OR showthat observed correlations between terms of tradeand exchange rates appear to be more consistentwith their assumptions about nominal rigiditiesthan with the alternative specification based onlocal-currency pricing.

I now turn to a discussion of the reformulationsof the redux model based on the introduction ofpricing to market.

Pricing to Market

While the redux model assumes that the LOOPholds for all tradable goods, a number of re-searchers have questioned the model on theground that deviations from the LOOP across inter-national borders appear to be larger than can beexplained by geographical distance or transportcosts (see, for example, Engel, 1993, and Engel andRogers, 1996). Some authors have therefore ex-tended the redux model by combining internation-al segmentation with imperfectly competitivefirms and local-currency pricing (essentially pric-ing to market or PTM). Krugman (1987) used theterm PTM to characterize price discrimination for

6 See also Andersen (1998), Benigno (1999), and Bergin and Feenstra(1999, 2000).

26 MAY/JUNE 2001

R E V I E W

certain types of goods (such as automobiles andmany types of electronics) where international ar-bitrage is difficult or perhaps impossible. This maybe due, for example, to differing national standards(for example, 100-volt light bulbs are not used inEurope and left-hand-side-drive cars are not popu-lar in the United Kingdom, Australia, or Japan).Further, monopolistic firms may be able to limit orprevent international goods arbitrage by refusingto provide warranty service in one country forgoods purchased in another. To the extent thatprices cannot be arbitraged, producers can dis-criminate across different international markets.

Studies allowing for PTM typically find thatPTM may play a central role in exchange rate de-termination and in international macroeconomicfluctuations. This happens because PTM acts tolimit the pass-through from exchange rate move-ments to prices, reducing the “expenditure switch-ing” role of exchange rate changes and potentiallygenerating greater exchange rate variability thanwould be obtained in models without PTM. Also,nominal price stickiness, in conjunction with PTM,magnifies the response of the exchange rate tomacroeconomic fundamentals shocks. Further, bygenerating deviations from PPP, PTM models alsotend to reduce the comovement in consumptionacross countries while increasing the comovementof output, fitting some well-known empirical regu-larities (see Backus, Kehoe, and Kydland, 1992).Finally, the introduction of PTM has important wel-fare implications for the international transmissionof monetary policy shocks, as discussed below.

Betts and Devereux (2000b), for example, char-acterize PTM by assuming that prices of manygoods are set in the local currency of the buyerand do not adjust at high frequency. Consequently,real exchange rates move with nominal exchangerates at high frequency. These assumptions alsoimply that price/cost markups fluctuate endoge-nously in response to exchange rate movementsrather than nominal prices (see also Knetter, 1993,on this point). In the Betts-Devereux framework,traded goods are characterized by a significant de-gree of national market segmentation and trade iscarried out only by firms. Households cannot arbi-trage away price differences across countries, andfirms engage in short-term nominal price setting.Therefore, prices are sticky in terms of the localcurrency.7

The Betts-Devereux model is based on aneconomy with differentiated products and assumes

that firms can price-discriminate across countries.With a high degree of PTM (that is, when a largefraction of firms engages in PTM), a depreciation ofthe exchange rate has little effect on the relativeprice of imported goods faced by domestic con-sumers. This weakens the allocative effects of ex-change rate changes relative to a situation whereprices are set in the seller’s currency; in the lattercase, pass-through of exchange rates to prices isimmediate. Hence, PTM reduces the expenditureswitching effects of exchange rate depreciation,which generally implies a shift of world demandtoward the exports of the country whose currencyis depreciating. Because domestic prices show lit-tle response to exchange rate depreciation underPTM, the response of the equilibrium exchangerate may be substantially magnified and, consis-tent with well-known observed empirical regulari-ties, exchange rates may vary more than relativeprices.

PTM also has implications for the internationaltransmission of macroeconomic shocks. In the ab-sence of PTM, for example, monetary disturbancestend to generate large positive comovements ofconsumption across countries but large negativecomovements of output. However, PTM reversesthe ordering: the deviations from PPP induced byPTM make consumption comovements fall. At thesame time, the elimination of expenditure switch-ing effects of the exchange rate enhances comove-ments of output across countries.

In terms of welfare, recall that the frameworkbased on the LOOP and PPP generally suggeststhat an unanticipated monetary expansion raiseswelfare of all agents at home and abroad. WithPTM, however, a domestic monetary expansionraises home welfare but reduces foreign welfareand monetary policy is a “beggar-thy-neighbor”instrument. Therefore, the PTM framework, un-like the framework based on the LOOP and PPP,provides a case for international monetary policycoordination.

Overall, the PTM framework suggests thatgoods market segmentation might help explain in-ternational quantity and price fluctuations andmay have important implications for the interna-tional transmission of economic shocks, policy,and welfare.

7 The model of Betts and Devereux (2000b) is used as a representa-tive of this class of PTM models in this section. Other examples ofmodels adopting PTM are Betts and Devereux (1996, 1997, 1999,2000a); CKM (1998, 2000); and Bergin and Feenstra (1999, 2000).


MAY/JUNE 2001 27

The Indeterminacy of the Steady State

In the framework proposed by OR (1995), thecurrent account plays a crucial role in the trans-mission of shocks. However, the steady state is in-determinate and both the consumption differentialbetween countries and an economy’s net foreignassets are nonstationary. After a monetary shock,the economy will move to a different steady stateuntil a new shock occurs. When the model is log-linearized to obtain closed-form solutions of theendogenous variables, one is approximating thedynamics of the model around a moving steadystate. This makes the conclusions implied by themodel questionable. In particular, the reliability ofthe log-linear approximations is low because vari-ables wander away from the initial steady state.

Many subsequent variants of the redux modelde-emphasize the role of net foreign assets accu-mulation as a channel of macroeconomic inter-dependence between countries. This is done byassuming that (i) the elasticity of substitution between domestic and foreign goods is unity or (ii) financial markets are complete. Both of theseassumptions imply that the current account doesnot react to shocks (see, for example, Corsetti andPesenti, 2001, and OR, 2000a).8 While this frame-work achieves the desired result of determinacyof the steady state, it requires strong assumptions—(i) or (ii) above—to shut off the current account,which is unrealistic. In a sense these solutions cir-cumvent the problem of indeterminacy, but theydo not solve it.

Ghironi (2000a) provides an extensive discus-sion of the indeterminacy and nonstationarityproblems in the redux model. Ghironi also pro-vides a tractable two-country model of macro-economic interdependence that does not rely oneither of the above assumptions in that the elasticityof substitution between domestic and foreigngoods can be different from unity and that finan-cial markets are incomplete, consistent with reali-ty. Using an overlapping generations structure,Ghironi shows how there exists a steady state, en-dogenously determined, to which the world econ-omy reverts following temporary shocks. Accumu-lation of net foreign assets plays a role in thetransmission of shocks to productivity. Finally,Ghironi also shows that shutting off the currentaccount may lead to large errors in welfare com-parisons, which calls for rethinking of several re-sults in this literature.

The issue of indeterminacy of the steady state

deserves further attention from researchers in thisarea.

Preferences

While the explicit treatment of microfounda-tions is a key advantage of new open economymacroeconomic models relative to the MFD model,the implications of such models depend on thespecification of preferences. One convenient as-sumption in the redux model is the symmetry withwhich home and foreign goods enter preferencesin the constant-elasticity-of-substitution (CES) utili-ty function. Corsetti and Pesenti (2001) extend theredux model to investigate the effects of a limiteddegree of substitution between home and foreigngoods. In their baseline model, the LOOP still holdsand technology is described by a Cobb-Douglasproduction function, with a unit elasticity of sub-stitution between home and foreign goods andconstant income shares for home and foreignagents. The model illustrates that the welfare ef-fects of expansionary monetary and fiscal policiesare related to internal and external sources of eco-nomic distortion, namely, monopolistic supply inproduction and monopoly power of a country. Forexample, an unanticipated exchange rate deprecia-tion can be “beggar-thyself” rather than “beggar-thy-neighbor” since gains in domestic output areoffset by losses in consumers’ purchasing powerand a deterioration in terms of trade. Also, open-ness is not inconsequential: smaller and more openeconomies are more prone to inflationary conse-quences. Fiscal shocks, however, are generally“beggar-thy-neighbor” in the long run, but theyraise domestic demand in the short run for giventerms of trade. These results provide a role for in-ternational policy coordination, which is not thecase in the redux model.9,10

An important assumption in the redux modelis that consumption and leisure are separable. This

8 This is a problem often encountered in the international real busi-ness cycles literature. Note, however, that the role of currentaccount dynamics in generating persistent effects of transitoryshocks has often been found to be quantitatively unimportant inthis literature. See the discussion on this point by Baxter andCrucini (1995) and Kollmann (1996).

9 Recall that the redux model has the unrealistic implication that theoptimal monetary surprise is infinite, which is of course not thecase in the Corsetti-Pesenti model.

10 Devereux (1999), Doyle (2000), Tille (1998a, b), Betts and Devereux(2000a), and Benigno (2001), among others, represent attempts tomodel explicitly international policy coordination in variants of theCorsetti-Pesenti model. See also OR (2000b).

28 MAY/JUNE 2001

R E V I E W

assumption is not compatible, however, with a bal-anced growth path if trend technical progress isconfined to the market sector. As a country be-comes richer, labor supply gradually declines, con-verging to a situation in which labor supply is zero,unless the intertemporal elasticity of substitutionis unity. CKM (1998), for example, employ a prefer-ence specification with nonseparable consump-tion and leisure (which is fairly standard, for exam-ple, in the real business cycles literature). Thispreference specification is compatible with a bal-anced growth path and is also consistent with thehigh real exchange rate volatility that is observedin the data. A more elastic labor supply and agreater intertemporal elasticity of substitution inconsumption generates more volatile real exchangerates. Hence, this preference specification providesmore plausible implications for the short-run dynamics of several macroeconomic variables relative to the redux model and better matchessome observed regularities.11

While the discussion in this subsection has fo-cused on only two issues with regard to the specifi-cation of preferences (the degree of substitutabilityof home and foreign goods in consumption andthe separability of consumption and leisure in util-ity), the results of models with explicit microfoun-dations may depend crucially on the specificationof the utility function in other ways. Relaxing thesymmetry assumption in the utility function andallowing for nonseparable consumption andleisure, for example, would yield more plausibleand more general utility functions. Of course, thereare other related important issues and, in this re-spect, the closed economy literature can lend ideason how to proceed; see, for example, the large andgrowing closed economy literature on habit forma-tion and home production.

Financial Markets Structure

The redux model assumes that there is inter-national trade only in a riskless real bond, andhence financial markets are not complete. Devia-tions from this financial markets structure havebeen examined in several papers. CKM (1998)compare, in the context of their PTM model, theeffects of monetary shocks under complete mar-kets and under a setting where trade occurs onlyin one noncontingent nominal bond denominatedin the domestic currency. Their results show thatthe redux model is rather robust in this case. Infact, incompleteness of financial markets appears

to imply small differences for the persistence ofmonetary shocks.12

A related study by Sutherland (1996) analyzestrading frictions (which essentially allow for a dif-ferential between domestic and foreign interestrates) in the context of an intertemporal generalequilibrium model where financial markets areincomplete and the purchase of bonds involvesconvex adjustment costs. Goods markets are per-fectly competitive and goods prices are subject toCalvo-type sluggish adjustment. Sutherland showsthat barriers to financial integration have a largerimpact on output the greater the degree of priceinertia. With substantial price inertia, outputadjusts slowly and more agents smooth their con-sumption pattern via international financial mar-kets. Sutherland’s simulations suggest that finan-cial market integration increases the volatility of anumber of variables when shocks originate fromthe money market but decreases the volatility ofmost variables when shocks originate from realdemand or supply; these results also hold in thegeneralization of Sutherland’s model by Senay(1998). For example, a positive domestic monetaryshock induces a domestic interest rate declineand, therefore, a negative interest rate differentialwith the foreign country. In turn, the negativeinterest rate differential produces a smallerexchange rate depreciation and a larger jump inrelative domestic consumption. This implies thatdomestic output rises less in this model than inthe baseline redux model.

OR (1995) defend their assumptions regardingthe financial markets structure of the redux modelstating that it would seem incoherent to analyzeimperfections or rigidities in goods markets, whileat the same time assuming that international capi-tal markets are complete. Indeed, one may arguethat, if there were complete international risksharing, it is unclear how price or wage rigiditiescould exist. Nevertheless, the assumption of fullinternational capital integration is very controver-sial. While many economists would agree that thedegree of financial integration has increased over

11 A further modification of the redux model considered byresearchers involves the introduction of nontradables in the analy-sis, which typically implies an increase in the size of the initialexchange rate response to a monetary shock; see, for example,Ghironi (2000b), Hau (2000), and Warnock (1999).

12 Note, however, that none of the models discussed in this paper havecomplete markets in the Arrow-Debreu sense, with the possibleexception of CKM (1998).


MAY/JUNE 2001 29

time (at least across major industrialized countries),it is perhaps fair to say that there are frictions infinancial markets (see Obstfeld, 1995). Given thecontroversies over what may constitute a realisticfinancial markets structure, the analysis of theimpact of barriers to financial integration remainsan avenue of research in its own right.

The Role of Capital

The literature has largely neglected the role ofcapital in new open economy models. For example,competitive models with capital can deliver effectsof supply shocks similar to those typically foundin monopolistically competitive models withendogenous utilization of capital (see, for exam-ple, Finn, 2000).13 CKM (1998, 2000) also arguethat capital (omitted in the redux model and mostsubsequent variants of it) may play an importantrole because monetary shocks can cause invest-ment booms by reducing the short-term interestrate and hence generate a current account deficit(rather than a surplus, as in the redux model).Explicitly allowing for capital in new open econo-my models is an important immediate avenue forfuture research.

STOCHASTIC NEW OPEN ECONOMYMACROECONOMICS

Recently, the certainty equivalence assump-tion that characterizes much of the literature dis-cussed above (including the redux model) hasbeen relaxed. While certainty equivalence allowsresearchers to approximate exact equilibriumrelationships, it “precludes a serious welfare ana-lysis of changes that affect the variance of output”(Kimball, 1995, p. 1243). Following this line ofreasoning, OR (1998) first extend the redux modeland the work by Corsetti and Pesenti (2001) to astochastic environment. More precisely, the inno-vation in OR (1998) involves moving away fromthe analysis of only unanticipated shocks.14

Risk and Exchange Rates

The OR (1998) model may be interpreted as asticky-price monetary model in which risk has animpact on asset prices, short-term interest rates,the price-setting decisions of individual produc-ers, expected output, and international tradeflows. This approach allows OR to quantify thewelfare tradeoff between alternative exchangerate regimes and to relate such tradeoff to a coun-

try’s size. Another important finding of this modelis that exchange rate risk affects the level of theexchange rate. Not surprisingly, as discussedbelow, the model has important implications forthe behavior of the forward premium and for theforward discount bias.

The setup of the OR (1998) model adds uncer-tainty to the redux model. Most results are stan-dard and qualitatively identical to those of theredux model. However, one of the most originalresults of this approach is the equation describingthe equilibrium exchange rate. To obtain the equi-librium exchange rate, OR (1998) assume thatHome and Foreign have equal trend inflation rates(equal to the long-run nominal interest ratesthrough the Fisher equation) and use conventionallog-linearizations (in addition to the assumptionthat PPP holds) to obtain an equation of nominalexchange rate determination. This equation maybe interpreted as a monetary-model-type equa-tion where conventional macroeconomic funda-mentals determine the exchange rate. Also, thisexchange rate equation is the same as in theredux model, except for a time-varying risk pre-mium term. Under the assumption of no bubbles,the solution of the model suggests that a level riskpremium enters the exchange rate equation. Insome sense, this model may explain the failure ofconventional monetary models of exchange ratedetermination in terms of an omitted variable inthe exchange rate equation, namely, exchangerate risk; a similar result was obtained by Hodrick(1989) in the context of a cash-in-advance flexible-price exchange rate model. For example, less relative risk of investments in the Home currencyinduces a fall in the domestic nominal interestrate and an appreciation of the domestic currency,capturing the idea of a “safe haven” effect on theHome currency.

For reasonable interest rates, a rise in Homemonetary variability induces both a fall in the

13 Finn (2000) demonstrates that a theory of perfect competition,which views capital utilization as the avenue through which energyenters into the model economy, can explain the observed effects ofenergy price increases on economic activity, which Rotemberg andWoodford (1996) and several subsequent studies defined as inexpli-cable without a theory of imperfect competition.

14 Note, however, that I am using the term stochastic loosely here.Even in approximated dynamics with certainty equivalence, modelsare stochastic. Evaluations of the first-order effects of secondmoments (noncertainty equivalence) recognize an aspect ofstochastic models that is often neglected, but this does not by itselfdefine a stochastic model.

30 MAY/JUNE 2001

R E V I E W

level of the exchange rate risk premium and a fallin the forward premium (the latter fall is shown tobe much larger in magnitude). This result contra-dicts the conventional wisdom that financial mar-kets attach a positive risk premium to the currencywith higher monetary volatility. The intuition isexplained by OR (1998) as follows:

[A] rise in Home monetary volatility maylead to a fall in the forward premium, evenholding expected exchange rate changesconstant. Why? If positive domestic mone-tary shocks lead to increases in global con-sumption, then domestic money can be ahedge, in real terms, against shocks to con-sumption. (The real value of Home moneywill tend to be unexpectedly high in statesof nature where the marginal utility ofconsumption is high.) Furthermore—andthis effect also operates in a flexible-pricemodel—higher monetary variability raisesthe expectation of the future real value ofmoney, other things equal. (p. 24)

This result provides a novel theoretical expla-nation of the forward premium puzzle. Not onlyshould high interest rates not necessarily be asso-ciated with expected depreciation, but the oppo-site may also be true, especially for countries withsimilar trend inflation rates.

Nevertheless, the results produced by thismodel may well depend critically on the specifi-cation of the microfoundations and are, therefore,subject to the same caveats raised by the literaturequestioning the appropriateness of the reduxspecification. Thus, it is legitimate to wonder howadopting the other specifications (alternativespecifications of utility, different nominal rigidi-ties, etc.) described earlier would affect the resultsof the OR (1998) stochastic model. The next sub-section discusses, for example, the changesinduced by the introduction of PTM in this model.

Related Studies

The OR (1998) analysis described above isbased on the following assumptions: (i) that pro-ducers set prices in their own currency, (ii) that theprice paid by foreigners for home goods (and theprice paid by domestic residents for foreign goods)varies instantaneously when the exchange ratechanges, and (iii) that the LOOP holds. Devereuxand Engel (1998) extend the OR (1998) analysis by

assuming PTM and that producers set a price inthe home currency for domestic residents and inthe foreign currency for foreign residents. Hence,when the exchange rate fluctuates, the LOOP doesnot hold. The risk premium depends on the typeof price-setting behavior of producers. Devereuxand Engel compare the agent’s welfare betweenfixed and flexible exchange rate arrangements andfind that exchange rate systems matter not onlyfor the variances of consumption, real balances,and leisure but also for their mean values once riskpremia are incorporated into pricing decisions.Since PTM insulates consumption from exchangerate fluctuations, floating exchange rates are lesscostly under PTM than under producer currencypricing. Consequently, a flexible regime generallydominates a pegged regime.15

Engel (1999) makes four points in summariz-ing the evidence on the foreign exchange risk pre-mium in this class of general equilibrium models.First, while the existence of a risk premium inflexible-price general equilibrium models dependson the correlation of exogenous monetary shocksand aggregate supply shocks, the risk premiumarises endogenously in sticky-price models. Second,the distribution of aggregate supply shocks doesnot affect the foreign exchange risk premium insticky-price models. Third, given that the risk pre-mium depends on the prices faced by consumers,when the LOOP does not hold there is no uniqueforeign exchange risk premium since producersset prices in consumers’ currencies. Fourth, stan-dard stochastic dynamic general equilibriummodels do not usually imply large risk premia.

The common denominator in these models isthat the exchange rate risk premium is an impor-tant determinant of the equilibrium level of theexchange rate. It remains an open questionwhether one could build a sticky-price modelcapable of convincingly explaining the forwardpremium puzzle. Nevertheless, this seems apromising avenue for future research.

NEW DIRECTIONS: THE SOURCE OFNOMINAL RIGIDITIES AND THECHOICE BETWEEN LOCAL AND FOREIGN CURRENCY PRICING

OR (2000a) may have again set new directionsfor stochastic open economy models of the class

15 See also Bacchetta and van Wincoop (1998).


MAY/JUNE 2001 31

discussed in this paper. They start by noting thatthe possibilities for modeling nominal rigiditiesare more numerous in a multicurrency interna-tional economy than in a single-money closedeconomy setting and that, in an international set-ting, it is natural to consider the possibility of seg-mentation between national markets. OR addressthe empirical issue of whether local currency pric-ing or foreign currency pricing is closer to reality.OR argue that, if imports are invoiced in the im-porting country’s currency, unexpected currencydepreciations should be associated with improve-ments (rather than deteriorations) in the terms oftrade. They then show that this implication is incon-sistent with the data. Indeed, their evidence sug-gests that aggregate data may favor a traditionalframework in which exporters largely invoice inhome currency and nominal exchange rate changeshave significant short-run effects on internationalcompetitiveness and trade.

The main reservations of OR about the PTM–local currency pricing framework employed byseveral papers in this literature are captured by thefollowing observations. First, a large fraction ofmeasured deviations from the LOOP results fromnontradable components incorporated in con-sumer price indices for supposedly traded goods(for example, rents, distribution services, advertis-ing, etc.); it is not clear whether the extreme mar-ket segmentation and pass-through assumptions ofthe PTM–local currency pricing approach are nec-essary to explain the close association betweendeviations from the LOOP and exchange rates. Sec-ond, price stickiness induced by wage stickiness islikely to be more important in determining persis-tent macroeconomic fluctuations since trade invoic-ing cannot generate sufficiently high persistence.(Invoicing largely applies to contracts of 90 daysor less.) Third, the direct evidence on invoicing islargely inconsistent with the view that exportersset prices mainly in importers’ currencies (see, forexample, ECU Institute, 1995); the United States is,however, an exception. Fourth, international evi-dence on markups is consistent with the viewthat invoicing in exporters’ currencies is the pre-dominant practice (see, for example, Goldberg andKnetter, 1997).

OR (2000a) build their stochastic dynamicopen economy model with nominal rigidities inthe labor market (rationalized on the basis of thefirst two observations above) and foreign currencypricing (rationalized on the basis of the last two

observations above). They consider a standard two-country global economy where Home and Foreignproduce an array of differentiated tradable goods(Home and Foreign have equal size). In addition,each country produces an array of differentiatednontraded goods. Workers set next period’sdomestic-currency nominal wages and then meetlabor demand in the light of realized economicshocks. Prices of all goods are completely flexible.

OR provide equilibrium equations for presetwages and a closed-form solution for each endoge-nous variable in the model as well as solutionsfor variances and for utility. In particular, the solu-tion for the exchange rate indicates that a relativeHome money supply increase that occurs afternominal wages are set would cause an overshoot-ing depreciation in the exchange rate. A fullyanticipated change, however, causes a preciselyequal movement in the wage differential and inthe exchange rate.

In this setup, OR show welfare results on twofronts. First, they show that constrained-efficientmonetary policy rules replicate the flexible-priceequilibrium and feature a procyclical response toproductivity shocks.16 For example, a positive pro-ductivity shock that would elicit greater labor sup-ply and output under flexible wages optimallyinduces an expansionary Home monetary responsewhen wages are set in advance. The same shockelicits a contractionary Foreign monetary response,but the net global monetary response is alwayspositive. Also, optimal monetary policy allows theexchange rate to fluctuate in response to cross-country differences in productivity shocks. Thisconclusion is similar to the result obtained by Kingand Wolman (1996) in a rational expectationsmodel where monetary policy has real effectsbecause imperfectly competitive firms are con-strained to adjust prices only infrequently and tosatisfy all demand at posted prices. In the King-Wolman sticky-price model, it is optimal to setmonetary policy so that the nominal interest rateis close to zero (that is, neutralizing the effect ofthe sticky prices), replicating in an imperfectlycompetitive model the result that Friedman foundunder perfect competition. Under a perfect infla-

16 These monetary policy rules are (i) constrained since they arederived by maximizing an average of Home and Foreign expectedutilities subject to the optimal wage-setting behavior of workers andprice-setting behavior of firms described in the model, and (ii) effi-cient since the market allocation cannot be altered without makingone country worse off, given the constraints.

32 MAY/JUNE 2001

R E V I E W

tion target, the monetary authority makes themoney supply evolve so that a model with stickyprices behaves much like one with flexible prices.

Second, OR calculate the expected utility foreach of three alternative monetary regimes,namely, an optimal floating rate regime, worldmonetarism (under which two countries fix theexchange rate while also fixing an exchange rate–weighted average of the two national money sup-plies), and an optimal fixed rate regime. The out-come is that the expected utility under an optimalfloating-rate regime is highest. This result is intu-itively obvious given that optimal monetary policyin this model involves allowing the exchange rateto fluctuate in response to cross-country differ-ences in productivity shocks. Fixed-rate regimeswould only be worthwhile if productivity shocksat home and abroad were perfectly correlated.17

The OR (2000a) model addresses several theo-retical and policy questions, including welfareanalysis under alternative nominal regimes. Theassumption that nominal exchange rate move-ments shift world demand between countries inthe short run, which plays a crucial role in thetraditional MFD model, is shown to be consistentwith the facts and can reasonably be used as abuilding block in stochastic open economy mod-els. Needless to say, this approach warrants fur-ther generalizations and refinements. In particu-lar, note that the current account is shut off in OR(2000a) to avoid the indeterminacy problem dis-cussed earlier. However, shutting off the currentaccount makes the model less plausible from anempirical point of view since it distorts thedynamics of the economy being modeled.

It is worth noting that the new open economymacroeconomics literature to date has (implicitlyor explicitly) assumed that there are no costs ofinternational trade. Nevertheless, the introductionof some sort of international trade costs (includ-ing, among others, transport costs, tariffs, andnontariff barriers) may be key in understandinghow to improve empirical exchange rate modelsand in explaining several unresolved puzzles ininternational macroeconomics and finance. Whilethe allowance of trade costs in open economymodeling is not a new idea and goes back at leastto Samuelson (1954), OR (2000c) have recentlystressed the role of trade costs in open economymacroeconomics. Indeed, OR (2000c) presentsomething of a “unified theory” that helps eluci-date what the profession may be missing when

trying to explain several puzzling empirical find-ings using trade costs as the fundamental model-ing feature, with sticky prices playing a distinctlysecondary role. It is hoped that future research innew open economy macroeconomics follows thesuggestion of OR (2000c) to make explicit allow-ance for non-zero international trade costs.

CONCLUSIONS

In this paper, I have selectively reviewed therecent literature on new open economy macro-economics, which has been growing exponentiallyin the last five years or so. The increasing sophis-tication of stochastic open economy models allowsrigorous welfare analysis and provides new expla-nations of several puzzles in international macro-economics and finance. Whether this approachwill become the new workhorse model for openeconomy macroeconomics, whether a preferredspecification within this class of models will bereached, and whether this approach will provideinsights on developing better-fitting empiricalexchange rate models are open questions.

Although the theory in the spirit of new openeconomy macroeconomics is developing veryrapidly, there is little effort at present to test thepredictions of new open economy models.Theorists working in this area should specifyexactly which empirical exchange-rate equationsthey would have empiricists estimate. If there is tobe consensus in the profession on a particularmodel specification, this theoretical apparatus hasto produce clear estimable equations.18

Agreeing on a particular new open economymodel is hardly possible at this stage. This is thecase not least because it requires agreeing onassumptions which are often difficult to testdirectly (such as the specification of the utilityfunction) or because they concern issues onwhich economists have strong beliefs on whichthey have not often been willing to compromise(such as whether nominal rigidities originate fromthe goods market or the labor market or whether

17 Indeed, the results suggest that the difference between the expectedutility under an optimal floating-rate regime and the expected utili-ty under an optimal fixed-rate regime may not be too large if thevariance of productivity shocks is very small or the elasticity of util-ity with respect to effort is very large.

18 A first step toward new open economy macroeconometrics hasbeen made, for example, by Ghironi (2000c). I am also currentlyinvestigating empirical exchange rate equations inspired by thenew open economy macroeconomics literature.


MAY/JUNE 2001 33

nominal rigidities exist at all). Achieving a newparadigm for open economy modeling is, howev-er, a major challenge which lies ahead for the pro-fession. While the profession shows some conver-gence toward a consensus approach in macroeco-nomic modeling (where the need for microfoun-dations, for example, seems widely accepted), itseems very unlikely that a consensus model willemerge in the foreseeable future.

REFERENCES

Andersen, Torben M. “Persistency in Sticky Price Models.”European Economic Review, May 1998, 42(3-5), pp. 593-603.

Bacchetta, Philippe and van Wincoop, Eric. “DoesExchange Rate Stability Increase Trade and CapitalFlows?” Discussion Paper 1962, Centre for EconomicPolicy Research, September 1998.

Backus, David K.; Kehoe, Patrick J. and Kydland, Finn E.“International Real Business Cycles.” Journal of PoliticalEconomy, August 1992, 100(4), pp. 745-75.

___________; ___________ and ___________. “Dynamics ofthe Trade Balance and the Terms of Trade: The J-Curve?”American Economic Review, March 1994, 84(1), pp. 84-103.

___________; ___________ and ___________. “InternationalBusiness Cycles: Theory and Evidence,” in Thomas F.Cooley, ed., Frontiers of Business Cycle Research.Princeton: Princeton University Press, 1995.

Barro, Robert J. and Gordon, David B. “Rules, Discretionand Reputation in a Model of Monetary Policy.” Journalof Monetary Economics, July 1983, 12(1), pp. 101-21.

Baxter, Marianne and Crucini, Mario. “Business Cycles andthe Asset Structure of Foreign Trade.” InternationalEconomic Review, November 1995, 36(4), pp. 821-54.

Benigno, Gianluca. “Real Exchange Rate Persistence withEndogenous Monetary Policy.” Unpublished manuscript,University of California, Berkeley, 1999.

Benigno, Pierpaolo. “Optimal Monetary Policy in aCurrency Area.” Unpublished manuscript, New YorkUniversity, 2001.

Bergin, Paul R. and Feenstra, Robert C. “Pricing to Market,Staggered Contracts and Real Exchange Rate Persistence.”

Working Paper No. 99/01, University of California, Davis,February 1999.

___________ and ___________. “Staggered Price Setting,Translog Preferences, and Endogenous Persistence.”Journal of Monetary Economics, June 2000, 45(3), pp.657-80.

Betts, Caroline and Devereux, Michael B. “The ExchangeRate in a Model of Pricing-to-Market.” EuropeanEconomic Review, April 1996, 40(3-5), pp. 1007-21.

___________ and ___________. “The International MonetaryTransmission Mechanism: A Model of Real ExchangeRate Adjustment Under Pricing-to-Market.” Unpublishedmanuscript, University of British Columbia, 1997.

___________ and ___________. “The International Effectsof Monetary and Fiscal Policy in a Two-Country Model.”Unpublished manuscript, University of British Columbia,1999.

___________ and ___________. “International MonetaryPolicy Coordination and Competitive Depreciation: A Re-Evaluation.” Journal of Money, Credit and Banking,November 2000a, 32(4), pp. 722-45.

___________ and ___________. “Exchange Rate Dynamicsin a Model of Pricing-to-Market.” Journal of InternationalEconomics, February 2000b, 50(1), pp. 215-44.

Blanchard, Olivier J. and Kiyotaki, Nobuhiro. “MonopolisticCompetition and the Effects of Aggregate Demand.”American Economic Review, September 1987, 77(4), pp.647-66.

Calvo, Guillermo A. “Staggered Prices in a Utility-MaximizingFramework.” Journal of Monetary Economics, September1983, 12(3), pp. 383-98.

Chari, V.V.; Kehoe, Patrick J. and McGrattan, Ellen R.“Monetary Shocks and Real Exchange Rates in StickyPrice Models of International Business Cycles.”Unpublished manuscript, Federal Reserve Bank ofMinneapolis, 1998.

___________; ___________ and ___________. “Sticky PriceModels of the Business Cycle: Can the Contract MultiplierSolve the Persistence Problem?” Econometrica, September2000, 68(5), pp. 1151-79.

Corsetti, Giancarlo and Pesenti, Paolo. “Welfare and

34 MAY/JUNE 2001

R E V I E W

Macroeconomic Interdependence.” Quarterly Journal ofEconomics, 2001 (forthcoming).

Devereux, Michael B. “Do Fixed Exchange Rates InhibitMacroeconomic Adjustment?” Unpublished manuscript,University of British Columbia, 1999.

___________ and Engel, Charles. “Fixed vs. FloatingExchange Rates: How Price Setting Affects the OptimalChoice of Exchange-Rate Regime.” Working Paper No.6867, National Bureau of Economic Research, 1998.

Dixon, Huw and Rankin, Neil. “Imperfect Competition andMacroeconomics: A Survey.” Oxford Economic Papers,April 1994, 46(2), pp. 171-99.

Dornbusch, Rudiger. “Expectations and Exchange RateDynamics.” Journal of Political Economy, December1976, 84(6), pp. 1161-76.

Doyle, Brian M. “Reputation and Currency Crises (or‘Countries of a Feather Devalue Together’).” Unpublishedmanuscript, Board of Governors of the Federal ReserveSystem, 2000.

ECU Institute. International Currency Competition and theFuture Role of the Single European Currency. London:Kluwer Law International, 1995.

Engel, Charles. “Real Exchange Rates and Relative Prices:An Empirical Investigation.” Journal of MonetaryEconomics, August 1993, 32(1), pp. 35-50.

___________. “Accounting for U.S. Real Exchange RateChanges.” Journal of Political Economy, June 1999,107(3), pp. 507-38.

___________ and Rogers, John H. “How Wide Is theBorder?” American Economic Review, December 1996,86(5), pp. 1112-25.

Finn, Mary G. “Perfect Competition and the Effects ofEnergy Price Increases on Economic Activity.” Journal ofMoney, Credit and Banking, August 2000, 32(3), pp. 400-16.

Fleming, J. Marcus. “Domestic Financial Policies UnderFixed and Under Floating Exchange Rates.” InternationalMonetary Fund Staff Papers, November 1962, 9(3), pp.369-80.

Ghironi, Fabio. “Macroeconomic Interdependence Under

Incomplete Markets.” Unpublished manuscript, FederalReserve Bank of New York, 2000a.

___________. “U.S.-Europe Economic Interdependence andPolicy Transmission.” Unpublished manuscript, FederalReserve Bank of New York, 2000b.

___________. “Towards New Open Economy Macro-econometrics.” Staff Report 100, Federal Reserve Bankof New York, 2000c.

___________ and Rebucci, Alessandro. “Monetary Rules forEmerging Market Economies.” Unpublished manuscript,Federal Reserve Bank of New York and InternationalMonetary Fund, 2000.

Goldberg, Pinelopi K. and Knetter, Michael M. “GoodsPrices and Exchange Rates: What Have We Learned?”Journal of Economic Literature, September 1997, 35(3),pp. 1243-72.

Hau, Harald. “Exchange Rate Determination: The Role ofFactor Price Rigidities and Nontradables.” Journal ofInternational Economics, April 2000, 50(2), pp. 421-47.

Hodrick, Robert J. “Risk, Uncertainty, and ExchangeRates.” Journal of Monetary Economics, May 1989, 23(3),pp. 433-59.

Jeanne, Olivier. “Generating Real Persistent Effects ofMonetary Shocks: How Much Nominal Rigidity Do WeReally Need?” European Economic Review, June 1998,42(6), pp. 1009-32.

Kimball, Miles S. “The Quantitative Analytics of the BasicNeomonetarist Model.” Journal of Money, Credit andBanking, November 1995, 27(4), pp. 1241-77.

King, Robert G. and Wolman, Alexander L. “InflationTargeting in a St. Louis Model of the 21st Century.”Federal Reserve Bank of St. Louis Review, May/June1996, 78(3), pp. 83-107.

Knetter, Michael M. “International Comparisons of Price-to-Market Behavior.” American Economic Review, June1993, 83(3), pp. 473-86.

Kollmann, Robert. “Incomplete Asset Markets and theCross-Country Consumption Correlation Puzzle.” Journalof Economic Dynamics and Control, May 1996, 20(5), pp.945-61.


MAY/JUNE 2001 35

___________. “The Exchange Rate in a Dynamic-OptimizingCurrent Account Model with Nominal Rigidities: AQuantitative Investigation.” Working Paper WP/97/07,International Monetary Fund, January 1997.

Krugman, Paul R. “Pricing to Market When the ExchangeRate Changes,” in Sven W. Arndt and J. David Richardson,eds., Real-Financial Linkages Among Open Economies.Cambridge, MA: MIT Press, 1987.

Lane, Philip R. “Inflation in Open Economies.” Journal ofInternational Economics, May 1997, 42(3-4), pp. 327-47.

___________. “The New Open Economy Macroeconomics:A Survey.” Working Paper No. 2115, Centre for EconomicPolicy Research Discussion Paper, March 1999; forth-coming in Journal of International Economics.

___________. “Money Shocks and the Current Account,” inGuillermo Calvo, Rudiger Dorbusch, and MauriceObstfeld, eds., Money, Capital Mobility, and Trade: Essaysin Honor of Robert Mundell. Cambridge, MA: MIT Press,2001.

Lucas, Robert E., Jr. “Interest Rates and Currency Prices ina Two-Country World.” Journal of Monetary Economics,November 1982, 10(3), pp. 335-59.

Mundell, Robert A. “The Appropriate Use of Monetary andFiscal Policy for Internal and External Stability.”International Monetary Fund Staff Papers, March 1962,9(1), pp. 70-79.

___________. “Capital Mobility and Stabilization PolicyUnder Fixed and Flexible Exchange Rates.” CanadianJournal of Economics and Political Science, November1963, 29(4), pp. 475-85.

Obstfeld, Maurice. “International Capital Mobility in the1990s,” in Peter B. Kenen, ed., UnderstandingInterdependence: The Macroeconomics of the OpenEconomy. Princeton: Princeton University Press, 1995.

___________ and Rogoff, Kenneth. “Exchange RateDynamics Redux.” Journal of Political Economy, June1995, 103(3), pp. 624-60.

___________ and ___________. Foundations ofInternational Macroeconomics. Cambridge, MA: MITPress, 1996.

___________ and ___________. “Risk and Exchange Rates.”

Working Paper No. 6694, National Bureau of EconomicResearch, August 1998.

___________ and ___________. “New Directions forStochastic Open Economy Models.” Journal of InternationalEconomics, February 2000a, 50(1), pp. 117-53.

___________ and ___________. “Do We Really Need a NewInternational Monetary Compact?” Working Paper No.7864, National Bureau of Economic Research, August2000b.

___________ and ___________. “The Six Major Puzzles inInternational Macroeconomics: Is There a CommonCause?” Working Paper No. W7777, National Bureau ofEconomic Research, July 2000c; in Ben Bernanke andKenneth Rogoff, eds., National Bureau of EconomicResearch Macroeconomics Annual 2000. Cambridge, MA:National Bureau of Economic Research and MIT Press,2001 (forthcoming).

Rotemberg, Julio J. and Woodford, Michael. “ImperfectCompetition and the Effects of Energy Price Increaseson Economic Activity.” Journal of Money, Credit andBanking, November 1996, 28(4), pp. 550-77.

Samuelson, Paul. “The Transfer Problem and TransportCosts, II: Analysis of Effects of Trade Impediments.”Economic Journal, June 1954, 64(254), pp. 264-89.

Sarno, Lucio. “Towards a New Paradigm in Open EconomyModeling: Where Do We Stand?” Unpublishedmanuscript, University of Warwick, 2000; available fromBrian Doyle’s New Open Economy Macroeconomics,<http://www.geocities.com/brian_m_doyle/open.html>.

___________ and Taylor, Mark P., eds. New Developments inExchange Rate Economics, Volumes I-II, Critical Writingsin Economics series. Northampton, MA: Edward Elgar,2001a (forthcoming).

___________ and ___________. Exchange Rate Economics.Cambridge, MA: Cambridge University Press, 2001b(forthcoming).

Senay, Ozge. “The Effects of Goods and Financial MarketIntegration on Macroeconomic Volatility.” ManchesterSchool, Supplement 1998, 66, pp. 39-61.

Stockman, Alan C. “A Theory of Exchange RateDetermination.” Journal of Political Economy, August1980, 88(4), pp. 673-98.

36 MAY/JUNE 2001

R E V I E W

___________. “The Equilibrium Approach to ExchangeRates.” Federal Reserve Bank of Richmond EconomicReview, March/April 1987, 73(2), pp. 12-30.

Sutherland, Alan. “Financial Market Integration andMacroeconomic Volatility.” Scandinavian Journal ofEconomics, December 1996, 98(4), pp. 521-39.

Svensson, Lars E.O. and van Wijnbergen, Sweder. “ExcessCapacity, Monopolistic Competition, and InternationalTransmission of Monetary Disturbances.” EconomicJournal, September 1989, 99(397), pp. 785-805.

Taylor, John B. “Aggregate Dynamics and StaggeredContracts.” Journal of Political Economy, February 1980,88(1), pp. 1-23.

Tille, Cédric. “Substitutability and Welfare.” Unpublishedmanuscript, Federal Reserve Bank of New York, 1998a.

___________. “The Welfare Effects of Monetary ShocksUnder Pricing to Market: A General Framework.”Unpublished manuscript, Federal Reserve Bank of NewYork, 1998b.

Velasco, Andres. “Multiplicity and Cycles in a Real Modelof the Open Economy.” Unpublished manuscript, NewYork University, 1997.

Walsh, Carl E. Monetary Theory and Policy. Cambridge, MA:MIT Press, 1998.

Warnock, Francis E. “Idiosyncratic Tastes in a Two-CountryOptimizing Model: Implications of a StandardPresumption.” Discussion Paper No. 631, InternationalFinance Discussion Papers, Board of Governors of theFederal Reserve System, April 1999.


MAY/JUNE 2001 37

A Simple Model ofLimited Stock MarketParticipationHui Guo

The 1998 Survey of Consumer Finance datashows that only 48.8 percent of U.S. house-holds owned stocks, either (i) directly or (ii)

indirectly through mutual funds. In addition, thereis a close relationship between shareholding andwealth. In 1998, 93 percent of the richest 1 per-cent of the population owned stocks; the richest10 percent owned 85 percent of total stocks andmutual funds, compared with 51 percent of totalsavings deposits. Meanwhile, the average stockreturn is “abnormally” higher than the averagegovernment bond return.1 In this paper, I try toexplain this shareholding puzzle—why many peo-ple do not hold stocks given that stocks outper-form government bonds by a large margin.

It is costly to collect and process informationabout stock markets. Bertaut (1997) finds thatbetter-educated people are more likely to holdstocks, even after controlling for variables such aswealth, current income, and unemployment risk.He interprets education as a measure of the abilityto process information about the market andinvestment opportunities. However, informationcosts are not the only reason for limited stockmarket participation. Rather, recent researchemphasizes that people tend to hold fewer riskyassets such as stocks in their portfolio if they aremore vulnerable to income shocks. For example,borrowing constraints (Guiso, Jappelli, andTerlizzese, 1996), labor income risks (Vissing-Jorgensen, 1998b), home ownership (Fratantoni,1998), and entrepreneurial risks (Heaton andLucas, 2000) are found to deter stock market entry.Moreover, these factors have smaller effects onpeople who have larger wealth. Holtz-Eakin, Joulfaian, and Rosen (1994) find that people arewilling to take more risks if they receive a largeinheritance.

In this paper, I develop a life-cycle model toshow how market imperfections may interact with

heterogeneous wealth to generate limited stockmarket participation. Many factors, such assuccessful entrepreneurial effort, life-cyclesavings, precautionary savings, and inheritance,explain wealth inequality. To keep the model man-ageable, I focus on three key elements, namely,different investment opportunities (stocks andbonds), credit market imperfections, and inheri-tance. In the model, although the stock return ishigher than the bond return, only people withwealth over a certain threshold own stocks. Thisoccurs for two reasons. First, there is a fixed costto entering the stock market. Second, people facea borrowing rate that is higher than the saving rateso that they cannot arbitrage by selling bonds andbuying stocks. As a result, wealthy householdsaccumulate more wealth and pass on a greaterinheritance to their families than poor householdsdo. In the long run, wealth is unequally distributedand wealthy households own almost all stocks.

Some other mechanisms have explained lim-ited stock market participation. Becker (1980)shows that the most patient agent owns all capitalin the long run. Allen and Gale (1994) argue thatthe less risk-averse person is more likely to holdstocks. Constantinides, Donaldson, and Mehra(2000) stress the life-cycle pattern of shareholdings.

Asset returns and limited stock market partici-pation are two closely related issues. However,recent research (i.e., Constantinides, Donaldson,and Mehra, 2000; Polkovnichenko, 2000; andYaron and Zhang, 2000) has had difficulty explain-ing the two simultaneously in general equilibriummodels. Therefore, I address asset returns and lim-ited stock market participation separately in thispaper. First, the asset return is accepted as givenwhen I explain why there is limited stock marketparticipation. Then limited stock market participa-tion is accepted as given when I discuss its effecton the asset return. Nevertheless, we ultimatelyneed to develop a general equilibrium frameworkthat explains both simultaneously, but this isbeyond the scope of this paper.

Limited stock market participation may havelarge effects on asset prices. The risk of the stockmarket return is measured by its covariance withshareholders’ consumption growth in the standardframework, the consumption-based Capital Asset

Hui Guo is an economist at the Federal Reserve Bank of St. Louis.Bill Bock provided research assistance.

1 Mehra and Prescott (1985) argue for an equity-premium puzzle: theobserved equity premium is too large to be explained by existingtheories.

38 MAY/JUNE 2001

R E V I E W

Pricing Model. Mehra and Prescott (1985) calculatethis covariance using aggregate consumption dataand find that it is too small to explain the observedequity premium unless we believe that investorsare extremely risk averse.2 This is the so-calledequity premium puzzle. In their calculation, Mehraand Prescott assume that everyone holds stocks sothat they can use aggregate consumption insteadof shareholders’ consumption. This assumption isinconsistent with the data because not everyoneholds stocks. Recent research finds that limitedstock market participation does help explain theequity premium puzzle. Using the Panel Study ofIncome Dynamics data, Mankiw and Zeldes (1991)find that shareholders’ consumption is morevolatile and more positively correlated with stockmarket returns than non-shareholders’ consump-tion. Brav and Geczy (1996) and Vissing-Jorgensen(1998a) document a similar phenomenon usingthe Consumption Expenditure Survey data. In con-trast, Guo (2000) explores the connection betweenlimited stock market participation and asset pricesby calibrating a heterogeneous agent model inwhich only one type of agent holds stocks. Underreasonable parameterizations, the simulated datamatch the first two moments of the risk-free rateand the stock market return.

A related issue is whether the most recent bullmarket is brought about by the increase in stockmarket participation. According to the Survey ofConsumer Finance data, the stock market partici-pation rate has increased from 31.7 percent in1989 to 48.8 percent in 1998. However, stockhold-ings remain extremely concentrated. For example,the wealthiest 10 percent of U.S. householdsowned 85 percent of total stocks and mutual fundsin 1998, only slightly lower than the 86 percentowned in 1989. Wolff (2000) also reports that theparticipation rate drops sharply if small share-holders are excluded. Therefore, there is littlechange in the concentration of stock ownershipand the most recent bull market is unlikely to beexplained by the increase in the stock market par-ticipation rate.3 Given that stock prices fluctuatewidely in historical data, the most recent run-upin stock prices may be deviations from the trend.

Limited stock market participation might alsohelp reconcile some macroeconomic anomalies.For example, although rich people own almost allstocks, their consumption share is relatively small.This explains why aggregate consumption is notvery responsive to the stock price fluctuation. How-

ever, the effects of limited stock market partici-pation on business cycles have not been fullyexplored yet. Future research along this line shouldimprove our understanding of the economy.

The paper is organized as follows. I firstpresent some stylized facts and then use an over-lapping-generations model to help explain limitedstock market participation and wealth inequality.In the last section I discuss the implications forasset prices.

SOME STYLIZED FACTS

In this section, I summarize some stylizedfacts about financial markets and stock marketparticipation.

• The stock return is persistently higher thanthe risk-free rate over long horizons.

• Very wealthy households own almost allstocks and other investment assets.

• The share of wealth held by very wealthyhouseholds is positively correlated withstock prices.

• The intergenerational transfer is an impor-tant channel through which wealthinequality is preserved over time.

Stocks Outperform Risk-Free AssetsOver Long Horizons

The stock return is much higher than the risk-free rate. During the period 1871-1998, the realcontinuously compounding stock market returnwas about 7.0 percent per year and the risk-freerate was only 2.4 percent.4 The difference isdramatically amplified by the compounding effect.If you invested one dollar in large company stocksat year-end 1925, you would have had $1,370.95by year-end 1996. On the other hand, investmentsof one dollar in short-term government bondsgrew to only $13.54 over the same period (IbbotsonAssociates, 1997).

2 See Kocherlakota (1996) for a survey of recent research on thisissue.

3 Heaton and Lucas (1999) make a similar argument. They also exploresome other explanations for the most recent stock price run-up.

4 These returns were calculated from the historical data constructedby Robert Shiller, which is available from his homepage<http://aida.econ.yale.edu/~shiller/>. The risk-free rate may beoverestimated because it is the return on primary commercialpaper in Shiller’s data. The annual real return on treasury bills is0.6 percent for the period 1926-96 (Ibbotson Associates, 1997).


MAY/JUNE 2001 39

Fama and French (1988), among many others,also document a mean-reverting process in stockprices. This is often interpreted as meaning thatstocks are not as risky in the long run as they arein the short run. In Figure 1, I plot the annualizedequity premium over a 30-year horizon. Forexample, the value corresponding to the year 1997is the average equity premium over the period1968-97. The equity premium over a long horizonis always positive, even for periods that includethe 1929 stock market crash.

However, it is not my intention to advocatestock market investing. First, Samuelson (1994),among others, is skeptical about these finitesample results. He argues that “if you adhere tothe dogma that stocks must beat bonds in thelong-enough run, there is no P/E level that themarket averages out to at which you will take insail. A Ponzi bubble is ever possible, and givenpast psychologies of boom and bust, ever-higherP/E ratios become a self-fulfilling prophecy…”(Samuelson, 1994, p.19). Second, the stockmarket could be very volatile in the short run, andit is rational for people not to participate in stockmarkets if their wealth is too little to absorb largeshocks to stock prices. Heaton and Lucas (2000)argue that even wealthy households should holdfewer stocks if they have significant proprietaryincome.

Very Wealthy Households Own MostStocks

Stocks are highly concentrated in the hands ofvery wealthy households. In 1998, the richest 10percent of U.S. households owned 85 percent oftotal stocks and mutual funds. They also held mostother investment assets, including financial securi-ties, trusts, business equity, and non-home realestate. In fact, they owned 86 percent of totalinvestment assets. However, they had only 44 per-cent of the other assets, including principalresidence, deposits, life insurance, and pensionaccounts. As shown in Figure 2, their shares ofstocks and mutual funds, total investment assets,and other assets are relatively stable over theperiod 1983-98.

Portfolio compositions are also quite differentbetween the very wealthy and the average house-holds, as shown in Figure 3. The principal residenceis the most important asset for average U.S. house-holds, which accounted for 65.9 percent of totalassets for the poorest 80 percent of U.S. householdsin 1995. They also allocated 11.1 percent in liquidassets, 8.5 percent in pension assets, and 12.2 per-cent in investment assets. Conversely, the richest 1percent put 78.5 percent of their total wealth ininvestment assets, 6.4 percent in their principal resi-dence, 7.7 percent in liquid assets, and 4.7 percent

Figure 1

Annualized Equity Premium Over a 30-YearHorizon

1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 20000.00

0.02

0.04

0.06

0.08

0.10

0.12Percent

SOURCE: Shiller’s Data. See footnote 4.

Figure 2

Assets Held by the Richest 10 Percent,1983-1998

1983 1986 1989 1992 1995 199882

85

88

91

40

43

46Percent

SOURCE: Wolff (2000).

Stocks and Mutual FundsOther Assets

InvestmentAssets

ScaleScale

Scale

Percent

40 MAY/JUNE 2001

R E V I E W

in pension assets. Therefore, very wealthy house-holds have a larger share of risky assets in theirportfolios than average households do.

Wealth Distribution Over Time

Wolff (1995) finds that wealth inequality movesclosely with stock prices. His results are reproducedin Figure 4, which plots the share of wealth held bythe richest 1 percent of U.S. households and thedetrended stock prices for the period 1922-98.

It is clear that these two variables movetogether, with a correlation coefficient of about0.61. There are two reasons. First, the wealthiest 1percent own almost all stocks, whereas theprinciple residence is the most important assetheld by the average U.S. households. Second, stockprices are much more volatile than the prices ofother wealth components, including principle resi-dence. Changes in the valuation of existing assetsare thus dominated by fluctuations in the stockmarket (Ludvigson and Steindel, 1999).

Inheritance and Wealth Inequality

There are many factors that explain wealthinequality, such as successful entrepreneurialeffort, life-cycle savings, precautionary savings, andinheritance. Here, I want to stress the empirical rel-evance of inheritance, which is a key element ofthe model presented in the next section.

Inhaber and Carroll (1992) argue thatinheritance is one of the most important sourcesof wealth for the richest people, while it is a minorsource of assets for most others. For example, 80percent of the U.S. population claims never tohave inherited any assets, and only 1 percent ofthe population admits to having inherited assets of$110,000 or more (Inhaber and Carroll, 1992, p.73). Moreover, in both 1988 and 1989, more thanone third of the 400 wealthiest Americans listedtheir primary source of wealth as inheritance,according to Forbes magazine.

Kotlikoff and Summers (1981) argue that inter-generational transfers account for the vast majorityof aggregate U.S. capital formation. Dynan, Skinner,and Zeldes (2000) also find that inheritance is cru-cial in explaining the different saving patternbetween the rich and the poor.

A LIMITED STOCK MARKET PARTICIPATION MODEL

For simplicity, I adopt an overlapping genera-tion model with bequest motives, which is similarto the model studied by Galor and Zeira (1993).While Galor and Zeira emphasized the importanceof different education opportunities in explainingwealth inequality, I assume that households havedifferent investment opportunities—they caninvest in either stocks or bonds. The stock return

Figure 3

Portfolio Compositions: Top 1% Richest andBottom 80% Poorest

Percent

SOURCE: Wolff (1998).

Principle Liquid Pension InvestmentPrinciple Liquid Pension Investment0

20

40

60

80

Bottom 80%

Top 1%

Assets AssetsAssetsResidence

Figure 4

Wealth Distribution and Detrended StockPrices

1920 1930 1940 1950 1960 1970 1980 1990 200015

25

35

45

-2

0

2

4Percent

SOURCE: Wolff (1995) and Wolff (2000) for wealth share;Shiller’s data for stock prices.

Wealth Share of Top 1%

Detrended Stock Price

Scale

Scale

U.S. Dollars


MAY/JUNE 2001 41

is higher than the bond return; however, there is afixed stock market entry cost. In the credit market,banks accept deposits and make loans and thehousehold faces a borrowing rate that is higherthan the saving rate. I show that, initially, onlyhouseholds with endowments over a certainthreshold find it optimal to hold stocks because ofthe fixed entry cost and the wedge between thesaving rate and the borrowing rate. Moreover,some households that initially hold stocks eventu-ally leave the stock market because their relativelysmall endowments do not allow them to leave alarge bequest. Rich households, however, alwayshold stocks and accumulate wealth faster thanpoor households do because they enjoy a higherrate of return on their assets. As a result, wealth isunequally distributed and rich people will hold allstocks in the long run.

Model Setup

There is a continuum of households in aneconomy that persists forever. At time t, eachhousehold, say i, has a new cohort born, ht

i. Thenew cohort receives a bequest, Mt

i, from a parent,ht

i–1, which can be invested in stocks or bonds. At

time t+1, he receives labor income, L, and thepayoff from his earlier investments.5 After leavingbequest Mt

i+1 to his one child, ht

i+1, he consumes

the rest of his wealth and exits the economy.It is costly to enforce loan contracts. Cohorts

can save or borrow only through banks, which havethe lowest enforcement costs. Banks raise moneyby issuing bonds, which promise a gross rate ofreturn, Rb. I assume that the enforcement cost isproportional to borrowers’ leverage ratio and acohort can borrow only at the rate Rb(1+D/W),where D is his outstanding debt and W is his networth.6 A cohort can also invest in stocks, whichoffer a higher rate of return, Rs, than bonds do.However, there is a fixed stock market entry costF>0. The fixed entry cost can be interpreted asinformational costs and factors that affect stockmarket participation decisions, as discussed in theintroduction.

For simplicity, I assume that the gross stockreturn, Rs, and the gross bond return, Rb, are con-stant and that Rs>Rb in the baseline model.However, adding noise to stock returns does notchange the results in any qualitative way as longas stocks are better investments than bonds, in thesense that mean returns to stocks are larger thanmean returns to bonds.

Lastly, there is a progressive tax, t b, on inheri-tance, which will be discussed in more detail inthe next section.

Maximization Problem

Since cohorts differ only in their endowmentsand bequests, I ignore the superscript i and sub-script t. Instead, I denote Mt

i, the bequests receivedby a t-cohort, as M; and I denote Mt

i+1, the bequests

left by a t-cohort, as B.Cohorts have identical preferences, which

depend on consumption, C, and bequests, B. Theutility function is

(1)

where a is the relative weight given to consumption.The maximization of equation (1) is subject to

budget constraints. If a cohort decides to stay outof stock markets and invest all his endowments inbonds, his budget constraints are

(2)

Otherwise, he chooses to pay the fixed entry costF to invest in stocks and his budget constraintsare

(3)

where D is the amount that he borrows from thebond market and Rb[1+D/(M–F)] is the rate atwhich he can borrow because his net wealth isM–F after he pays the fixed cost.

The progressive estate tax is defined by equa-tion (4), for which only bequests that exceed amaximum level B– are taxed:

(4)

The maximization is done in two steps. Acohort first maximizes his total wealth, We, whichis the sum of the payoff to the first period’s invest-ments and labor income. He then decides how to

ttb

B B

B B=

£>

0 if

if .

C B M D F R L DRD

M Fs b+ £ + - + - +-

( ) ( ),1

C B MR Lb+ £ + .

Max C BC B b{ } ( ) ( ) [( ) ],, log loga a t+ - -1 1

5 Galor and Zeira (1993) show that heterogeneity in labor incomeleads to wealth inequality. Here I want to stress the importance ofheterogeneity in investment opportunities on wealth distribution.To ensure a clear demonstration, I assume that all cohorts receivethe same labor income, L. Adding Galor and Zeira’s heterogeneouslabor income should not change these results in any qualitativeway.

6 In a general equilibrium setting, Bernanke, Gertler, and Gilchrist(1999) show that the cost of external funds depends negatively onfirms’ net worth relative to the gross value of capital.

42 MAY/JUNE 2001

R E V I E W

allocate total wealth between consumption andbequests.

Maximization of Wealth. A cohort does oneof two things: either (i) invests all his endowmentsin bonds and his second period total wealth isgiven by

(5) ,

or (ii) pays the fixed entry cost, F, to participate inthe stock market and his second period total wealthis given by

(6)

Because We is a linearly increasing function ofendowments, M, in equations (5) and (6), thereexists a threshold level of endowment,

,

at which the cohort is indifferent to these invest-ment strategies. Moreover, he chooses to invest inbonds if his endowment is below M* and choosesto invest in stocks otherwise. The total wealth, We,is then given by equation (7), which is a monotoni-cally increasing function of M:

(7)

Consumption and Bequests. A cohort choosesconsumption, C, and bequests, B, to maximizeequation (1), subject to the estate tax (equation (4))and the reduced budget constraints (equation (8)):

(8) ,

where We is defined by equation (7). Clearly, thereexist We

* and We**, such that We

*<We**, and the

optimal after-tax bequest is as follows:

(9)

Denote We–1 as the inverse function of We(M). It

is well defined because We(M) is a monotonically

B

W

B

W

W W

W W W

W W

e

e

e e

e e e

e e

*

*

* **

**.

=-

- -

££ £

>

( ) ,

,

( )( ) ,

if

if

if

1

1 1

a

a t

C B We+ £

W

MR L M M

R R M FR

M F R L M M

e

b

s b

bs

=

+ £- - + - + ≥

, if( ) ( )

( ) , if .

*

*2

4

MR R

R R R Rs b

s b s b

* = +- +( )

( )( )

2

3

W M D F R L DRD

M FR R M F

RM F R L

e D s b

s b

bs

= + - + - +-

= - - + - +

max

.

{ } ( ) ( )

( ) ( )( )

1

4

2

W MR Le b= +

increasing function of M. Define M**=We–1(We

*)and M***=We

–1(We**). It is clear that M*** is greater

than M**.After substituting equation (7) into equation

(9), the optimal after-tax bequest is a function ofthe endowment M as in equation (10). Here weassume that M** is greater than M* or that non-shareholders do not have to worry about the estatetax.

(10)

if M<M*;

if M* ≤ M < M**;

if M**≤ M ≤ M***; and

otherwise.

The Dynamics of Wealth

Equation (10) is a first-order difference equa-tion of bequests. The phase diagram is plotted inFigure 5 under the following conditions:

1. ,

2.

3.

The first condition ensures at least one stablesteady state, which is the same as in Galor and Zeira(1993). The second condition ensures that thewealth diverges in the long run. The third conditionis required so that rich people’s wealth has a well-defined steady state; otherwise, it goes to infinity.

These conditions hold under reasonableparameterizations. Let us assume that there are 30years in each period. According to Shiller’s data, Rb

is about 200 percent and Rs is about 760 percent.The current highest marginal estate tax rate is 60percent. Under this parameterization, conditions 1

( )( )( )

1 14

12

- - - +È

ÎÍ

˘

˚˙ <a t R R

RRs b

bs .

( )( )

,14

12

- - +È

ÎÍ

˘

˚˙ >a R R

RRs b

bs

( )1 1- <a Rb

( )( )[( ) ( )

( ) ],1 14

2

- - - - + - +a t R R M FR

M F R Ls b

bs

B,

(1 )[

( ) ( )( ) ],- - - + - +a R R M F

RM F R Ls b

bs

2

4

B

MR Lb

* =- +( )( ),1 a


MAY/JUNE 2001 43

through 3 hold for 0.78<a<0.91. If the popula-tion grows at 2 percent per year, conditions 1through 3 hold for 0.60<a<0.84.7

The following results are shown clearly inFigure 5. First, cohorts with initial endowmentsless than M* do not hold stocks. Second, there aretwo stable steady states ML and MH. Rich house-holds with initial endowments greater than MM

converge to MH, and the remaining poor house-holds converge to ML. Therefore, wealth isunequally distributed and only rich people holdstocks in the long run. Third, reductions in thefixed entry cost move M* toward the origin andtherefore increase stock market participation.

Due to its simplicity, the baseline model doesnot provide a complete description of data. In par-ticular, the actual wealth distribution is not bimodal,people do move up and down the economic scale,and wealthy households also own a significantamount of bonds in the data. This is because ran-dom factors are assumed away in the baselinemodel. For example, entrepreneurial success orfailure can generate mobility in wealth; rich peoplehold bonds to diversify risks. Incorporations ofthese considerations should improve our model’sprediction. As an example, I will show in the nextsection of the article that our model can generatea more realistic wealth distribution if stock returnsare stochastic.

Stochastic Stock Returns. For simplicity and

without loss of generality, I assume that stockreturns are random realizations of two values andare not serially correlated. Also, the investmentdecision is made before the stock return is real-ized. Each cohort now maximizes the expectedutility in his first period and his consumption/bequest decision is the same as in the case of cer-tainty. If conditions 1 through 3 hold in each state,the dynamic of bequests with stochastic stockreturns is shown in Figure 6. Note, the portfoliodecision is independent from the state becausestock returns are not serially correlated.

In the long run, the households with initialendowments less than MM converge to ML. Thehouseholds with initial endowments greater thanMM2 converge to the MH2/MH region. The otherhouseholds’ wealth may fluctuate between MM

and MM2, depending on the realizations of thestock return. The stochastic return model thusgenerates an additional middle class who owns

7 If there is a population growth, conditions 1 through 3 become asfollows (where Rp is the growth rate of population):

1.

2.

3. ( )( )( )

1 14

11

2

- - - +È

ÎÍ

˘

˚˙ <a t R R

RR

Rs b

bs

p

.

( )( )

,14

11

2

- - +È

ÎÍ

˘

˚˙ >a R R

RR

Rs b

bs

p

( ) ,11

1- <a RRb

p

Figure 5

Dynamics of Bequest

B

MML M* M** M***MM MH

Figure 6

Dynamics of Bequest with Uncertainty

B

MML MM2 MH2MM MH

State 1

State 2

44 MAY/JUNE 2001

R E V I E W

some stocks. Moreover, it predicts that the wealthinequality, or the share of wealth held by the rich,increases with stock prices, as observed in the data.

IMPLICATIONS FOR ASSET PRICES

The asset return is taken as given in the limitedstock market participation model presented in theprevious section. In this section, I discuss theeffect of limited stock market participation on theasset return.

Agents are usually assumed to be homogeneousin economic models for the sake of simplicity.However, asset pricing models with homogeneousagents do not provide a good description for theasset return. They fail to explain why the equitypremium is so high (the equity premium puzzle)and why the stock price is so volatile (the excessvolatility puzzle).

In modern asset pricing models, agents arerisk averse and prefer smooth consumption. Thereturn to an asset thus depends on how well itcan be used to smooth agents’ consumption. Intu-itively, one dollar is more valuable in bad stateswhen consumption is low than one dollar in goodstates when consumption is high. A stock is thusunattractive, and shareholders demand a positivepremium to hold such a stock if its return is low(high) when shareholders’ consumption is low(high). Also, the more risk-averse shareholders are,the larger the risk premium they require. It can beshown that the equity premium is equal to gsr,s ina frictionless economy, where g is a measure ofrelative risk aversion and s r,s is the covariancebetween the stock return and the shareholder’sconsumption growth. This is the so-called con-sumption-based Capital Asset Pricing Model.Assuming that everyone holds stocks, Mehra andPrescott (1985) calculate this covariance usingaggregate consumption and find (i) that it is toosmall to explain the observed equity premium or(ii) that there is an equity premium puzzle.

It also can be shown that asset price Pt isequal to the sum of the expected cash flow, Dt+i,weighted by stochastic discount factor Rt+i or

.

In the homogeneous agent model, Rt+i is equal tothe intertemporal marginal rate of substitution

,b i t i

t

U CU C( )( )

+ ¢¢

P E R Dt t t i t ii

= Â + +=

•[ ]

1

where b is the time discount factor, U ´ is the mar-ginal utility, and Ct is the aggregate consumptionat time t. Variations in asset prices thus come fromtwo sources: shocks to the cash flow, Dt+i, andshocks to the stochastic discount factor, Rt+i, whichin turn are caused by aggregate consumptionshocks. Shiller (1981) finds that dividends are toosmooth to explain many variations in stock prices.Similarly, Campbell (1991) finds that most vari-ations in stock prices come from innovations inthe stochastic discount factor. However, the aggre-gate consumption is too smooth to generate thevolatile stochastic discount factor implied by thefinancial data. This is the excess volatility puzzle.

The preference is assumed to be time separablein the examples given above. Constantinides (1990)shows that it is possible to use aggregate consump-tion to generate a volatile stochastic discount factoras well as a large and volatile equity premium in ahabit formation model, in which utility depends onboth current and past consumption. However, therisk-free rate is very volatile in his model because itis also priced by the same volatile stochastic dis-count factor. This is easy to understand. Cash flowsare stochastic on stocks and are predetermined onbonds. Given that dividends are smooth in the data,this difference is rather small. Stocks and bondsshould thus exhibit similar properties, i.e. meansand/or variance if they are priced by the samestochastic discount factor. However, the stock returnis much higher and much more volatile than thebond return in the data. Therefore, stocks and bondsare not likely to be priced by the same stochasticdiscount factor. This poses a serious challenge tothe homogeneous agent model.8

Recent research by Guo (2000) suggests thatthese puzzles might be related to the fact that onlya few wealthy people own almost all stocks. Heshows that a heterogeneous agent model oflimited stock market participation can replicatethese phenomena.

There are two types of agents in his model:one is a shareholder and the other is a non-shareholder. They both receive labor incomes, butonly the shareholder receives dividends. Labor in-comes and dividends follow stochastic processes.Both agents use bonds to diversify the income risk;

8 One exception is Campbell and Cochrane (1999), who avoid thisproblem by choosing a particular habit form so that the risk-freerate is constant. However, they need a very large risk-aversion toexplain the puzzles mentioned above. Therefore, they do not reallysolve the equity premium puzzle.


MAY/JUNE 2001 45

for example, an agent buys (sells) bonds when hisincome is above (below) the trend. However, theycan borrow from the bond market only up to alimited amount or there are borrowing constraints.The model is calibrated using income processesestimated by Heaton and Lucas (1996), and thesimulated data match the mean and variance ofstock returns and bond returns under reasonableparameterizations.

Unlike the homogeneous agent model, stocksand bonds may be priced by different stochasticdiscount factors, as in the model by Guo (2000),because of limited stock market participation. In particular, while bonds are priced by the inter-temporal marginal rate of substitution of a non-constrained agent(s), stocks are always priced bythe shareholder’s intertemporal marginal rate ofsubstitution. Stocks and bonds are thus priced bydifferent intertemporal marginal rates of substitu-tion when the shareholder’s borrowing constraintsare binding. The stochastic discount factor forbonds is

and for stocks it is

,

where Cts and Ct

n are consumption of the share-holder and the non-shareholder, respectively. It isclear that the former is larger and smoother thanthe latter because borrowing constraints put alower bound on the discount factor of bonds, butnot stocks. Therefore, in Guo’s model, the bondreturn is low and smooth while the stock return ishigh and volatile, as observed in the data.

Intuitively, bonds are desirable and are pricedat a premium because they can be used todiversify income shocks. Stocks are not desirablebecause they cannot be used to diversify incomeshocks. The precautionary saving motive thuslowers only the risk-free rate, but not the stockreturn. This echoes Weil’s (1989) argument thatthe equity premium puzzle is indeed a risk-freerate puzzle. To see this, the equity premium in Guo(2000) is equal to gsr,s+rs

f–min{rsf,rn

f}, where g isthe relative risk aversion coefficient, s r,s is thecovariance between shareholder’s consumptionand the stock return, and rs

f(rnf) is the shadow

risk-free rate of the shareholder (non-shareholder).The equity premium is larger in Guo’s model than

b i t is

ts

U C

U C

( )( )

+ ¢¢

max {( )( )

,( )( )

}b bi t is

ts

i t in

tn

U C

U C

U C

U C+ +¢

¢¢¢

in the representative agent model for two reasons.First, there is an extra nonnegative term,rs

f–min{rsf,rn

f}, reflecting the fact that bonds(stocks) can (cannot) be used to hedge incomeshocks. This term can be interpreted as a liquiditypremium. Second, the covariance between share-holders’ consumption and the stock return islarger than the covariance between aggregate con-sumption and the stock return.

CONCLUSION

The recent Survey of Consumer Finance datashow that stocks are highly concentrated in thehands of a few wealthy people. In this paper, Iused an overlapping-generations model to helpexplain this limited stock market participation anddiscussed its effect on asset prices. However, otherimplications, such as its effect on business cycles,have not been fully explored yet. Future researchalong this direction should improve our under-standing of the economy.

REFERENCES

Allen, Franklin and Gale, Douglas. “Limited MarketParticipation and Volatility of Asset Prices.” AmericanEconomic Review, September 1994, 84, pp. 933-55.

Becker, Robert. “On the Long-Run Steady State in a SimpleDynamic Model of Equilibrium with HeterogeneousHouseholds.” The Quarterly Journal of Economics, 1980,95, pp. 375-82.

Bertaut, Carol. “Stockholding Behavior of U.S. Households:Evidence from the 1983-1989 Survey of ConsumerFinances.” Review of Economics and Statistics, May 1998,80(2), pp. 263-75.

Bernanke, Ben; Gertler, Mark and Gilchrist, Simon. “TheFinancial Accelerator in a Quantitative Business CycleFramework.” Working Paper No. 6455, National Bureauof Economics Research, 1998.

Brav, Alon and Geczy, Christopher. “An EmpiricalResurrection of the Simple Consumption CAPM withPower Utility.” Memo, University of Chicago, 1996.

Campbell, John. “A Variance Decomposition for StockReturns.” Economic Journal, March 1991, 101, pp. 157-79.

___________ and Cochrane, John. “By Force of Habit: AConsumption-Based Explanation of Aggregate Stock

46 MAY/JUNE 2001

R E V I E W

Market Behavior.” Journal of Political Economy, April1999, 2(107), pp. 205-51.

Constantinides, George. “Habit Formation: A Resolution ofthe Equity Premium Puzzle.” Journal of Political Economy,June 1990, 98(3), pp. 519-43.

___________; Donaldson, John and Mehra, Rajnish. “JuniorCan’t Borrow: A New Perspective on the Equity PremiumPuzzle.” Working Paper No. 6617, National Bureau ofEconomics Research, 2000.

Dynan, Karen; Skinner, Jonathan and Zeldes, Stephen. “Dothe Rich Save More?” Working Paper No. 7906, NationalBureau of Economics Research, 2000.

Fama, Eugene and French, Kenneth. “Permanent andTemporary Components of Stock Prices.” Journal ofPolitical Economy, April 1988, 96(2), pp. 246-73.

Fratantoni, Michael. “Homeownership and Investment inRisky Assets.” Journal of Urban Economics, July 1998,44(1), pp. 27-42.

Galor, Oded and Zeira, Joseph. “Income Distribution andMacroeconomics.” Review of Economic Studies, 1993,60(1), pp. 35-52.

Guiso, Juigi; Jappelli, Tullio and Terlizzese, Daniele.“Income Risk, Borrowing Constraints and PortfolioChoice.” American Economic Review, March 1996, 86(1),pp. 158-72.

Guo, Hui. “Business Conditions and Asset Prices in aDynamic Economy.” Working Paper 2000-31A, FederalReserve Bank of St. Louis, 2000.

Heaton, John and Lucas, Deborah. “Portfolio Choice andAsset Prices: The Importance of Entrepreneurial Risk.”Journal of Finance, June 2000, 55(3), pp. 1163-98.

___________ and ___________. “Stock Prices andFundamentals.” NBER Macroeconomics Annual, 1999,pp. 213-42.

___________ and ___________. “Evaluating the Effects ofIncomplete Markets on Risk Sharing and Asset Pricing.”Journal of Political Economy, June 1996, 104(3), pp. 443-87.

Holtz-Eakin, Douglas; Joulfaian, David and Rosen, Harvey.“Entrepreneurial Decisions and Liquidity Constraints.”Rand Journal of Economics, July 1994, 25(2), pp. 334-47.

Inhaber, Herbert and Carroll, Sidney. How Rich is Too Rich?Income and Wealth in America. New York: PraegerPublishers, 1992.

Kocherlakota, Narayana. “The Equity Premium: It’s Still aPuzzle.” Journal of Economic Literature, March 1996,34(1), pp. 42-71.

Mankiw, Gregory and Zeldes, Stephen. “The Consumptionof Stockholders and Nonstockholders.” Journal ofFinancial Economics, March 1991, 29(1), pp. 97-112.

Kotlikoff, Laurence and Summers, Lawrence. “The Role ofIntergenerational Transfers in Aggregate CapitalAccumulation.” Journal of Political Economy, August1981, 89(4), pp. 706-32.

Ludvigson, Sydney and Steindel, Charles. “How Importantis the Stock Market Effect on Consumption?” FederalReserve Bank of New York Economic Policy Review, July1999, 5(2), pp. 29-51.

Mehra, Rajnish and Prescott, Edward. “The Equity Premium:A Puzzle.” Journal of Monetary Economics, March 1985,15(2), pp. 145-61.

Polkovnichenko, Valery. “Heterogeneous Labor Incomeand Preferences: Implications for Stock MarketParticipation.” Working Paper, University of Minnesota,2000.

Samuelson, Paul. “The Long-Term Case for Equities.”Journal of Portfolio Management, October 1994, 21(1),pp. 15-24.

Shiller, Robert. “Do Stock Prices Move Too Much to BeJustified by Subsequent Changes in Dividends?” AmericanEconomic Review, June 1981, 71(3), pp. 421-36.

Schwert, William. “Why Does Stock Market VolatilityChange Over Time?” Journal of Finance, 1989, 44(5), pp.1115-53.

Vissing-Jorgensen, Annette. “Limited Stock MarketParticipation.” Working Paper, MIT, 1998a.

___________. “An Empirical Investigation of the Effect ofNon-Financial Income on Portfolio Choice.” WorkingPaper, MIT, 1998b.

Weil, Philippe. “The Equity Premium Puzzle and the Risk-Free Rate Puzzle.” Journal of Monetary Economics,November 1989, 24(3), pp. 401-21.


MAY/JUNE 2001 47

Wolff, Edward. “Recent Trends in Wealth Ownership,1983-1998.” Working Paper No. 300, Jerome LevyEconomics Institute, 2000.

___________. “Recent Trends in the Size Distribution ofHousehold Wealth.” Journal of Economic Perspectives,July 1998, 12(3), pp. 131-50.

___________. Top Heavy: A Study of the IncreasingInequality of Wealth in America. New York: TheTwentieth Century Fund Press, 1995.

Yaron, Amir and Zhang, Harold. “Fixed Costs and AssetMarket Participation.” Revista de Analisis Economico,2000, 15, pp. 89-109.

48 MAY/JUNE 2001

R E V I E W


MAY/JUNE 2001 49

COMING IN THE NEXT ISSUE OF REVIEW

In Honor of Darryl Francis

Darryl Francis: Maverick in the Formulation of Monetary Policy

Jerry L. Jordan

Money and Monetary Policy: An Essay in Honor of Darryl Francis

Allan H. Meltzer

Monetary Policy in Theory and Practice

Expectations, Open Market Operations, and Changes in the Federal Funds Rate

John B. TaylorCommentary: Athanasios Orphanides

Identifying the Liquidity Effect at the Daily Frequency

Daniel L. ThorntonCommentary: Simon Gilchrist

Assessing Simple Policy Rules: A View from a Complete Macroeconomic Model

Eric M. Leeper and Tao ZhaCommentary: Kenneth D. West

Measuring Systematic Monetary Policy

Kevin D. Hoover and Oscar JordáCommentary: Valerie A. Ramey

Monetary Policy Analysis in Models Without Money

Bennett T. McCallumCommentary: John V. Leahy

Monetary Transmission Lags and the Formulation of the Policy Decision on Interest Rates

Charles A.E. GoodhartCommentary: Laurence H. Meyer

50 MAY/JUNE 2001

R E V I E W

1. FRED (Federal Reserve EconomicData), a database providing U.S. economic and financial data andregional data for the Eighth FederalReserve District. You can access FREDthrough the Internet atwww.stls.frb.org/fred.

2. Research Department, FederalReserve Bank of St. Louis, P.O. Box442, St. Louis, Missouri 63166-0442.You can request data and programs on either disk or hard copy. Pleaseinclude the author, title, issue date,and page numbers with your request.

3. Inter-university Consortium forPolitical and Social Research(ICPSR). Member institutions canrequest data through the CDNet Orderfacility. Nonmembers should write to:ICPSR, Institute for Social Research,P.O. Box 1248, Ann Arbor, Michigan48106, or call (734) 998-9900.

All nonproprietary and nonconfidential data and programs for the articles writtenby Federal Reserve Bank of St. Louis staff and published in Review are available toour readers and can be obtained from three sources:

www.stls.frb.org

Federal Reserve Bank of St. LouisPost Office Box 442St. Louis, Missouri 63166-0442

reserve bank of s eview may une olume umber · 2019. 3. 18. · federal reserve bank of st.louis...

Documents