consumption smoothin_ current account_france

8/4/2019 Consumption Smoothin_ Current Account_france

1/18

IMF WORKING PAPER 1995 International Monetary Fund

This is a Working Paper and the author(s) would welcomeany comments on the present text. Citations should refer toa Working Paper of the International Monetary Fund, m en-tioning the author(s), and the date of issuance. The viewsexpressed are those of the author(s) and do not necessarilyrepresent those of the Fund.

WP/95/119 INTERNATIONAL MONETARY FUNDEuropean I and Research Departments

Consumptio n Smoothing and the Current Account:Evidence for France, 1970-94

Prepared by Pier re-Richard Agenor, Claude Bismut, Paul Cashinand C. John McDermott

Authorized for distribution by Ulrich Baumgartner and Donald MathiesonNovember 1995

Abstract

This paper estimates a simple consumption-smoothing model of theFrench current account , and examines its capacity to predict recentdevelopments in France's external performance. The model views thecurrent account as a buffer through which private agents can smoothconsumpti on over time in response to temporary disturbances to output,investment, and government expenditure. The empirical resultsindicate that the model performs well overall, and predicts correctlythe sharp turnaround in France's external accounts observed in thepast three years--a feature of the data that conventional models oftrade flo ws, based on income and relative price variables, appearunable to explain.

JEL Classification Numbers:F32, F41, F47

This paper was prepared while Claude Bismut was a SeniorEconomist in the European I Department and Paul Cashin an Economist inthe Research Departm ent. We would like to thank, without implication,Ulrich Baumgartner, Benoit Coeure, and Jean-Luc Tavernier for helpfuldiscussions and comments on an earlier draft.


2/18

- ii -

Contents Page

Summary iiiI. Introduction 1

II. The Model 2III. Estimation Results 6IV. Summary and Conclusions 8Figure 1. France: Actual and Predicted Consumption Smoothing 8aComponent of the Current AccountAppendix: Unit Root Tests 10References 12


3/18

I. IntroductionFrance's external accounts improved markedly in 1992, and

continued to improve in 1993 and 1994. The trade balance recorded asurplus equivalent to 0.5 percent of GDP in 1992 and 1.3 percent in1993 (compared to a deficit of 0.4 percent in 1991), whereas thecurrent account balance recorded a surplus of 0.3 percent in 1992 and0.9 percent in 1993-94. However, conventional models of trade flowsbased on income and relative price variables appear to be unable toaccount for these developments. Movements in competitiveness record edsince the late 1980s and the 1992-93 recession apparently e xplain onlya small fraction of the improvement in France's external accountsbetween 1992-94. 1/

This paper attempts to interpret France's recent externalperformance o n the basis of the consumption-smoothing (orintertemporal) approach to the current account. This approach, whic hhas gained in popularity in recent year s, focuses on the impact ofcurrent and future values of households' long-run expected income onsaving and investment. 2/ The consumption-smoothing model has alsofeatured prominently i n the debate launched by Feldstein and Horioka(1980) on the degree of capital mobility across countries. Asobserved by Ghosh (1995), a formal comparison of the variance of thecurrent account derived from the intertemporal model with the varianceof the actual current account provides a natural procedure forassessing the extent to which a country can borrow freely (subject toa no-Ponzi game condition) at world interest rates .

In the present paper , a simple intertemporal model of the currentaccount is estimated and used as a benchmark for assessing thebehavior of France's external accounts. Section II presents themodel. Section III provides estimates of the model parameters .Estimation results indicate that the model tracks remarkably well thehistorical data over the period 1970 to 1994--despite the fact thatFrance experienced large external shocks during the 1970s and that theperiod prior to the mid-1980s was characterized by capital contr ols.The existence of such controls could have bee n expected to distortlending and borrowing operations of French resi dents, and thus pr eventthem from effectively smoothing consumption spending over time. Acomparison of the actual and simulated paths indicates that the modelalso accounts for the current account surpluses registered betw een1992 and 1994. This result suggests that the surpluses may havereflected a downward revision in households' long-run expec tedresources, rather than a significant improvement in external pric ecompetitiveness.

1/ Developments in competitiveness of the French manufacturingindustry since the early 1980s are reviewed by Agenor (1995) . A"conventional" model of trade flows (based on an an error-correctio nframework) is estimated by Agenor and Bismut (1995).

2/ See Obstfeld and Rogoff (1994) and Razin (1995) for recen tsurveys of the literature.


4/18

1/ This basic framework has been used by a number of autho rs,including Milbourne and Otto (1992), Otto (1992), Sheffrin and Woo(1992) and Ghosh (1995). The presentation here follows more closelythe last two pape rs. Obstfeld and Rogoff (1994) and Razin (1995)discuss various extensions of the basic framework.

- 2 -

II. The Model

The intertemporal approach to the current account is derived fromthe permanent-income theory of consumption and saving. In the contextof a small open economy with access to world capital market s, thepermanent-income theory implies that temporary shocks (which bydefinition have larger effects on current resources than lifetimeresources) may lead to large fluctuations in national saving and thecurrent account.

The model tested in this paper is based on a simple one-goodframework, which can be described as follows. 1/ Suppose that theonly traded asset is a consumption-indexed bond with fixed face valuethat pays net interest at the rate p between two periods . Therepresentative consumer's preferences are separable intertemporally.The expected value of the discounted utility flow of the consumer isgiven by

where a is a discount factor, c consumption, the expectationsoperator (conditional on information available at t = 0 ) , u() theinstantaneous utility function, which possesses all familiarproperties (namely, u' > 0 and u'' < 0 ) . The consumer's b udgetconstraint is

where 6 denotes holdings of indexed bonds (which are freely tradedacross countries) at the beginning of period t, p the worl d interestrate (assumed fixed), yt output, It investment, r lump-sum taxes, andA the first-difference operator. The government maintains a balancedbudget, so that


5/18

- 3 -

+ 7 t - It - c t - gc. (4)

The representative consumer maximizes (1) subject to (4) . As forinstance in Otto (1992) and Sheffrin and Woo (1990), Fisherianseparability is assumed to hol d so that the levels of output andinvestment are treated as exogenous in solving for the optimalconsumption path. Approximating the instantaneous utility function by

2the quadratic function u(ct ) = ct -ac /2 (which requires that c < I/afor the margina l utility of consumption to remain positive) andimposing the standard no-Ponzi game restriction yields: 1/

where E is the expectations operat or, 9 = ap(l+p)/[a(l+p) - 1 ] , andz = y -I -g is output net of private investment and public spending,and is often referred to as net output (Sheffrin and Woo , 1990),net private noninterest cash flow (Obstfeld and Rogoff, 1994), orsimply national cash flow (Ghosh, 1995). The parameter 0 - 1 reflectsthe consumption tilting dynamics of consumption, which results fromdivergences between the world interest rate and the domestic rate oftime preference, (l-a)/a. Whe n o > 1 (or 0 < 1 ) , agents tiltconsumption towards the future (the present), whereas if 0 = 1 thereis no tilting effect (see Sac hs, 1982). In general, the higher theelasticity of intertemporal substitution, the stronger will be thetilting effect (Obstfeld and Rogoff, 1994). Thus, along the optimalpath, private consumption depends on the present value of the expectedfuture stream of the cash flow, as well as the economy's existingstock of net foreign assets. Equivalently, equation (5) shows thatprivate optimal consumption is proportional to wealth , defined as thepresent discounted value of the cash flow plus interest payments onthe existing stock of assets. As pointed out by Milbourne and Otto(1992, p. 372 ) , g in (5) could be interpreted as the present-valuetax cost, using a Ricardian-equivalence argument. Thus, (5) may beviewed as expressing consumption as a function of the present-value ofsavings out of disposable inco me.

By definition, the current account is equal to gross nationalproduct q (GDP plus interest income on the outstanding stock offoreign assets, yt + bt minus private and public expenditure

1/ The quadratic approximation is such that u ' " ( ) = 0, andimplies that the optimal path of consumption does not depend onuncertainty over future consumption or the degree of variability ofincome.


6/18

- 4 -

(ct +It +g t ) . The consumption-smoothing component of the currentaccount, CAt , can be defined as: 1/

CA* - (qt-It-gt) - 8c*.

Noting that the current account is also equal to net foreignasset accumulation (bt+1 -b t), simple manipulations using (5) and (6)yield

whic h shows that the consumption- smoothing component of the currentaccount is equal to minus the present discounted value of expectedchanges in the national cash flow variable . Equation (7) indicatesthat permanent s hocks, which have no effect on expected changes in thecash flow variable, leave the current account unaffected--as theirexpected change is zero. A permanent increase in yt , for i nstance,induces a one-for-one increase in consumption and no change in thecurrent account. 2/ Unfavorable temporary cash flow shocks--such asan unexpected increase in government expenditure or investment--causethe expression on the right-hand side of (7) to fall, that is, thecurrent account to move into smaller surplus or greater deficit, andconversely in the case of favorable shocks. Thus, the current accountacts as a buffer to smooth consumption in the presence of temporarydisturbances. 3/

The main problem wit h implementing (7) empirically is themeasurement of the expected changes in the cash flow vari able. Inline with most of the existing literature, here we will assume thatthe current account itself reflects all information about the future

1/ Removing the consumption-tilting component of the currentaccount (which is nonstationa ry) is necessary to ensure validity ofstandard inference techniques, as discussed below.

2 / However, permanent positive shocks to investment, to theextent that they raise future ou tput, will lead to a deterioration ofthe current account (see Obstfeld and Rogoff, 1994).

3/ Equation (7) also implies that a country will experience apersistent current account imbalance if the expected change innational cash flow has a non-zero mean. Suppose, for instance, thatagents expect national cash flow to fall over time by an average ofFF. 10 million. The steady-state current account surplus woul d begiven by FF. (10/p) million--or, assuming a world interest rate equalto 4 percent, a surplus of about FF. 250 million.


7/18

- 5 -

course of the cash flow variable. 1/ In other words, national saving(net of investment) should help predict subsequent movements in thenational cash flow variable.

We begin therefore by estimating a first-order un restrictedbivariate vector autoregressive (VAR) model in the cash flow variableand the actual consumption-smoothing component of the current account ,defined as CAt = (qt -It -gt)- oct:

ca-1where 1t and c2t

a3 CA -e2t-(8)

t - l - i

are disturbance terms and 6 an estimated value of 6(see below). We then make use of the implication that

't+hCAt+h* CA. CAtJ

(9)

which can be substituted in (7) to give the estimate of the optimalcurrent account:

CA* = -CA^

(10)

where I is the identity matrix. Expression (9) is valid a s long asthe infinite sum in equation (7) converges. This requires that thevariables appearing in the VAR system (8) be stationary. Assumingthat zt is I(1), Lzt will be I(0). And since under the null theactual consumption-smoothing current account (which is then equal toCA ) is a discounted sum of Lz t , it will also be I(0). 2/

The expression for the optimal current account derived from (10)can then be compared to the actual current account to determine, forinstance, the extent to which the surplus recorded in the past threeyears in France can be explained by consumption smoothing behavio r. Avariety of additional tests are also reported below.

1/ See for instance Ghosh (1995) and Sheffrin and Woo (1990).This approach rests on the idea of Campbell and Shiller (1987),according to which a n asset price reflects all the available infor-mation that is useful in predicting future earnings.

2/ Trehan and Walsh (1991) examine the conditions under wh ichthe assumption of stationarity of the current account is necessary forintertemporal budget balance .


8/18

- 6 -

III. Estimation Results

The data used in our empirical investigation are annual series onnational accounts for the period 1970 to 19 94, obtained directly fromthe French National Statistical Office, INSEE. yt is gross domesticproduct, qt gross national product, It gross private and publicinvestment (defined as including inventory accumulation), and gtcurrent government expenditure. All nominal series were deflated bythe implicit GDP deflator to ensure consistency. 1/ The world interestrate was set at 4 percent in all calculations reported below. 2/

The first step in the procedure is to obtain an estimate of 9, inorder to construct the (stationary) consumption- smoothing componentof the current account by removing the nonstationary component of theactual series (or, more specifically, the stochastic trend) associatedwith consumption tilting. This estimate can be obtained from acointegrating regression which examines the relationship b etweenprivate consumption c and qt -It -gt . 3/ This relationship wasestimated using the Phillips-Hansen FM method, which yields anasymptotically correct variance-covariance estimator when estimatingcointegrating vectors in the presence of serial correlation andendogeneity (Phillips and Hansen, 1990). In our cointegratingregression of qt -I t -g t on ct we obtained an estimate 9 of 0.9757, witha long-run standard error (as derived from the FM method ) of 0.005. At-test (with the standard variance estimator replaced by the standarderror derived from the FM method) indicates that the estimated valueis significantly different from unity. This estimate thereforeindicates that France has consumed more than its permanent cash flowand has foregone future consumption in favor of present consumptio n.By definition, as indicated above, the residual series from the

1/ Current account estimates based on national accounts data donot fully reflect nominal capital gains and losses on net foreignassets. In addition, our deflation procedure does not allow us tocorrect for the inflationary erosion of the real value of foreignassets. These limitations are common to most studies in this area.

2/ An annual world interest rate of 4 percent was used bySheffrin and Woo (1990), and by Ostry and Levy (1995) in theiranalysis of the role of financial deregulation and intertemporalfactors in the behavior of saving in France. We also estimated themodel with a world interest rate equal to 2, 3, 5, and 6 percent . Theresults differed only marginally from those reported below. Inparticular, the ratio of the variance of the actual to the predictedcurrent account tends to decrease as p increases.

3/ The Appendix reports the results of unit root tests showingthat the two series are I(1).


9/18

- 7 -

c o i n t e g r a t i n g r e g r e s s i o n o f q t-I t-g t o n ct is e q u a l to CA t , the a c t u a lconsumption-smoothing component of the current account.

The hypothesis of cointegration between the cash flow variable ztand private consumption ct was tested using the LC test of Hansen(1992) , and the stability of the relationship between them wasexamined using Hansen's (1992) mean-F and sup-F tests. 1/ The LC testof the null hypothesis of a cointegrated relationship between zt andc does not reject the hypothesis: a value of 0.41 is the result; the5 percent critical value (in the presence of a constant term) is 0.58.Furthermore, the Hansen stability tests indicate that the relationshipbetween zt and ct is stable--the null hypothesis of constantparameters in the FM cointegrating regression is not rejected. Themean-F and sup-F tests have values of 4.07 and 5.51; the correspondingcritical values at the 5 percent significance level are 4.57 and12.40, respectively.

One of the testable implications of the consumption smoothingmodel is that the current account should Granger-cause subsequentchanges in the cash flow variable (Obstfeld and Rogoff, 1994). Acurrent account deficit today, for instance, should signal an expectedincrease in future cash flows--associated, say, with an expectedreduction in future government expenditure. A standard F-test of thenull hypothesis of the absence of Granger causality from CA t to zt isindeed rejected--it has a value of 5.87 and a p-value of 0.025.

We then estimated the VAR system in terms of Azt and CA t ,including constant terms. The intertemporal model describedpreviously implies two restrictions on the parameter estimates derivedfrom the VAR system under the null, namely, that the coefficient onLz t be equal to zero and that the coefficient on CA t be equal to onein equation (10) . Formally, let T denote the matrix multiplying thevector [Azt CA t ]' in equation (10) . Let r = [rZ TCA] be aconformable partition of the matrix T. The model therefore requiresthat T = 0, and YCA = 1. Acceptance of these restrictions implies

Z CAthat movements of the actual consumption-smoothing current account CA treflect those of the optimal consumption-smoothing current accountCA t , constructed as in equation (10) .

1/ The Appendix presents the results of unit root tests on ztand c .


10/18

- 8 -

These joint restrictions were tested using a Wald test. The2value of the test statistic (which is distributed has a x with 2

degrees of freedom) was Wald = 0.872 (with a p-value = 0.647),indicating strong acceptance of the restrictions. This suggestsFrance had little difficulty in smoothing consumption through foreignborrowing and lending in the face of exogenous shocks. This issomewhat surprising, since during most of the estimation period (fromthe early 1970s to the mid-1980s) capital controls were inplace--which in principle should have limited the ability of domesticresidents to engage in international borrowing and lending. 1/ Actualand predicted values of the current consumption-smoothing currentaccount are shown in Figure 1. The figure shows indeed that the twoseries move closely together. Changes in national cash flows explainfairly well movements in the actual consumption-smoothing currentaccount balance.

A further testable implication of the model is that the varianceof CAt and CAt should be equal. Such a comparison provides a jointtest of the assumption of a high degree of capital mobility and thevalidity of the intertemporal model of the current account (Ghosh,1995). We found the variance of CA t to be significantly larger thanthat of CAt --by more th an 40 percent in relative terms--indicating"excess volatility " in the actual consumption-smoothing component ofthe current account . This result is similar to those obtained bySheffrin and Woo (1990) and by Ghosh (1995) for Canada, Germany,Japan, and the Unite d Kingdom. It suggests that capital flows betweenFrance and the rest of the world may have been more volatile thanwould be justified by expected changes in fundamentals.

IV. Summary and Conclusions

The purpose of this paper has been to estimate a simple inter-temporal model of France's current account over the period 1970-94,and to examine the extent to which it predicts recent develop ments.The key prediction of the model is that a country's current accountwill be in deficit (surplus) whenever national cash flow, defined asgross domestic product less investment and government spending,is expected to rise (fall) over time. Intuitively, if the cash flowis expected to grow over time, the country will find it optimal to

1/ For a discussion of the structure of capital restrictionsprior to the mid-198 0s, see Ministry of Finance (1986). Thedismantling of capital controls bega n in 1984 (within the context of ageneral movement towards capital market integration in Europe) , andwas completed in 1990. For an overview of the liberalization processin recent years, see the annual reports on Exchange Arrangements andExchange Restrictions published by the Fund.


11/18

Fge1

Fa

AuaPeceC

mpoSmohnCmp

ohCeA

WodneeRe=00

(c

a1

boF


12/18


13/18

- 9 -

borrow against future resources (that is, to increase its ext ernaldebt) by running a current account deficit. If, by contras t, nationalcash flow is expected to fall over time--as might be the case ifgovernment spending were expected to increase in the future--thecountry would run a current account surplus (increase its savin gs, orreduce external indebtedness) today in order to maintain consumptionin the future at a level consistent with permanent income.

Estimation results suggested that, despite its simplicity, thebasic consumption-smoothing model explains fairly well fluctuations ofthe current account balance since the early 1970s, which is suggestiveof a high degree of capital mobility between France and the rest ofthe world. This is consistent with recent evidence for otherindustrial countries. Our results are particularly remarkable in viewof the fact that estimation was performed in part over a period duringwhich France experienced large external shocks, and restrictions onoverseas capital transactions (which should have limited, inprinciple , international borrowing and lending activities by domesticresidents) were in place. In addition, the model predicts correctlythe sharp turnaround in France's external accounts observed in 1992--afeature of the data that "traditional" models of trade flows, based o nincome and relative price variabl es, appear to be unable to explain.1/ The improvement in the current account (or, equivalently, in netsaving) appears to reflect a more pessimistic assessment by consumersof their future income prospect s. In turn, the expected deteri orationin income may have been related to the 1992-93 rece ssion (the largestin France since the end of the second World War) and the associatedincrease in unemployment.

1/ The optimal current account predicted by the consumption-smoothing model for the period 1992-94 is lower than the actualcurrent account balance--but the difference is not large.


14/18

- 10 -

APPENDIXUnit Root Tests

This Appendix examines the time series properties of consumption(c ) , national disposable income ( Y t ) , and national cash flow (zt )series. The Phillips-Perron Z(a) and Z(t) unit root tests are usedto determine the order of integration of the series. 1/ The Fejerkernel is used, as is the prewhitening technique of Andrews andMonahan (1992) to aid in estimating the long-run variance of the errorterm in the unit root regression. Andr ews' (1991) data-dependentautomatic bandwidth selector is also invoked to determine the optimallag truncation order.

The tests were sequenced from a maximum of two unit roots down toone unit root. At each stage an ordinary least squares regession ofthe form x = u + fit + ax - + e was run, where x (which is equal toc , z or y ) is in first-difference form in the first stage, and inlevels in the second stage.

The results indicate that all of the series are nonstationary(Table A1). Both statistics testing for one unit root versus no unitroots are insignificant, indicating that we cannot reject the nullhypothesis of a unit root. Conversely, both statistics testingfor two unit roots versus at most one unit root are significant, whichindicates we can reject the hypothesis of two unit roots--the threeseries are integrated of order one, or I(1). 2/ Accordingly, bothvariables appearing in the VAR of equation (8) are stationary.

1/ See Dickey and Rosanna (1994) for a discussion of Phillips -Perron tests of stationarity. These non-parametric tests have anadvantage over the standard Dickey-Fuller tests in that nuisanceparameters are (asymptotically) eliminated, even if disturbances arenot independent and identically distributed.

2/ The Z(a) test for consumption (when =0 , and p=O) marginallyfails to reject the null hypothesis of two unit roots at the 5percent significance level. However, this hypothesis is rejected bythe Z(t) test.


15/18

- 11 -

Table A1Univariate Unit Root Tests, 1970-94

APPENDIX

Deterministiccomponent

Variables2/

Tests 3/Z(t)

Two unit roots vs . at most oneNoneConstantConstantNoneConstantConstantNoneConstantConstant

and

and

and

time

time

time

trend

trend

trend

AyAyAyAcAcAcAzAzAz

0.8049-0.0433-0.18830.85990.26590.22010.8134-0.0560-0.1070

-3.0486-24.3676-27.4075-2.3355

-17.5878-19.3006-2.8999-24.8018-25.9740

-1.6757-5.1036-5.8716-1.3003-3.6994-3.8960-1.5097-5.0191-5.2946

One unit root vs . noneNoneConstantConstantNoneConstantConstantNoneConstantConstant

and

and

and

time

time

time

trend

trend

trend

yyyccczzz

1.02450.97230.78861.02410.98130.70631.02580.98130.6908

0.5884-0.6550-4.64230.5764-0.4783

-13.72530.6181-0.4428-7.8432

7.8368-2.8014-2.01976.0060-1.2710-2.66697.8844

-1.6252-2.3331

1/ The Phillips-Perron tests are based on the following modelfor any series x: x = \i +fit+ ax - + e . The tests rely onrejecting the null hypothesis of a unit root (a=l) in favor ofstationarity. This hypothesis is tests by the Z(a) and Z(t) tests.If the null cannot be rejected (the test statistics are notsignificant), then the series has a unit root.

2 / All series are in logarithms.3/ The critical values for the Z(a) and Z(t) tests at the 5

percent significance level are -7.3 and -1.95 for no deterministiccomponents (fi=P**O) , -12.5 and -3.0 for the constant term only (/x 0,and =0), and (-17.9 and -3.60) for both deterministic components

and


16/18

- 12 -References

Agenor, Pierre-Richard, "Competitiveness and External TradePerformance of the French Manufacturing Industry," unpublished,International Monetary Fund (May 1995).

Agenor, Pierre-Richard, and Claude Bismut, "France: External TradeDevelopments," unpublished, International Monetary Fund (June1995).

Andrews, Donald W. K., "Heteroskedasticity and AutocorrelationConsistent Covariance Matrix Estimation," Econometrica, Vol. 59(May 1991), pp. 817-58.

Andrews, Donald W. K., and J. Christopher Monahan, "An ImprovedHeteroskedasticity and Autocorrelation Consistent CovarianceMatrix Estimation," Econometrica, Vol. 60 (July 1992),pp. 953-66.

Campbell, John, and Robert Shiller, "Cointegration and Tests ofPresent Value Models," Journal of Political Economy, Vol. 95(October 1987), pp. 1062-88.

Dickey, David A., and Robert J. Rossana, "Cointegrated Time Series: AGuide to Estimation and Hypothesis Testing," Oxford Bulletin ofEconomics and Statistics, Vol. 56 (August 1994), pp. 325-53.Feldstein, Martin, and Charles Horioka, "Domestic Saving and

International Capital Flows," Economic Journal, Vol. 90 (June1990), pp. 314-29.

Ghosh, Atish R. , "International Capital Mobility Amongst the MajorIndustrialised Countries: Too Little or Too Much?," EconomicJournal, Vol. 105 (January 1995), pp. 107-28.Hansen, Bruce E., "Tests for Parameter Instability in Regressions withI(1) Processes," Journal of Business and Economic Statistics,

Vol. 10 (July 1992), pp. 321-35.Milbourne, Ross, and Glenn Otto, "Consumption Smoothing and the

Current Account," Australian Economic Papers, Vol. 31 (December1992), pp. 369-83.

Ministry of Finance, Livre Blanc sur le Financement de l'Economie, LaDocumentation Francaise (Paris: 1986).Obstfeld, Maurice, and Kenneth Rogoff, "The Intertemporal Approach tothe Capital Account," Working Paper No 4893, National Bureau ofEconomic Research (October 1994).


17/18

- 13 -

Ostry, Jonathan, and Joaquim Levy, "Household Saving in France," StaffPapers (International Monetary Fund), Vol. 42 (June 1995),pp. 375-97.

Otto, Glenn, "Testing a Present-Value Model of the Current Account:Evidence from US and Canadian Time Series," Journal ofInternational Money and Finance, Vol. 11 (October 1992),pp. 414-30.

Phillips, Peter C. B., and Bruce E. Hansen, "Statistical Inference inInstrumental Variables Regression with I(1) Processes," Review ofEconomic Stu dies, Vol. 57 (January 1990), pp. 99-125.

Razin, Assaf , "The Dynamic Optimizing Approach to the Current Accoun t:Theory and Evidence," in Understanding Interdependence: TheMacroeconomics of the Open Economy, ed by Peter Kenen, PrincetonUniversity Press (Princeton, NJ: 1995).

Sachs, Jeffrey, "The Current Account in the Macroeconomic AdjustmentProcess," Scandinavian Journal of Economics, Vol. 84 (June 1982),pp. 147-59.

Sheffrin, Steven M., and Wing T. Woo, "Present Value Tests of a nIntertemporal Model of the Current Account," Journal ofInternational Economics, Vol. 29 (August 1990), pp. 237-53.

Trehan, Bharat, and Carl E. Walsh, "Testing Intertemporal BudgetConstraints: Theory and Applications to U.S. Federal Budget andCurrent Account Deficits," Journal of Money, Credit, and Banking,Vol. 23 (May 1991), pp. 206-23.


18/18

consumption smoothin_ current account_france

Documents