Back Matter

Back Matter

G. Kincaid, Martin Fetherston, Peter Isard, and Hamid Faruqee
Published Date:
December 2001
    • ShareShare
    Show Summary Details
    Appendix I Alternative Methodologies for Assessing Equilibrium Exchange Rates

    As discussed in Section II, CGER’s assessments for the industrial countries rely primarily on the macroeconomic balance framework, while also taking account of traditional purchasing power parity perspectives. In discussing this choice of methodology, it may be useful to consider briefly several approaches on which CGER has chosen not to rely.

    Extended Purchasing Power Parity with Balassa-Samuelson Effects

    An important modification or refinement of the long-run PPP hypothesis has come from the observation that the prices of nontradable goods and services, relative to the prices of tradables, tend to be higher in high-income countries than in low-income countries. This observation emerged from attempts to make quantitative comparisons of living standards in different countries in a series of projects sponsored by the United Nations and other international organizations, and spearheaded to a large extent by economists from the University of Pennsylvania.44 These studies have established that the methodology of comparing international standards of living by converting national accounts data at market exchange rates into a common currency unit generally understates the living standards of low-income countries relative to those of high-income countries. Samuelson (1994) has referred to this empirical regularity as the “Penn effect.”

    Balassa (1964) and Samuelson (1964) attempted to explain the empirical regularity, along with an apparent tendency for market exchange rates to deviate systematically from PPP over the long run. They conjectured that the tendency for the relative price of nontradables to be higher in high-income countries reflected a tendency for productivity in the tradable goods sector to rise relative to productivity in the nontradables sector as real incomes expanded.45 Given competitive pressures within each country for workers with similar skills to receive similar wages in the two sectors, relatively rapid productivity growth in the tradables sector would tend, other things being equal, to push up the relative cost of production in the nontradables sector and, hence, the relative price of nontradables. Under conditions in which the relative price of tradable goods across countries remained constant, such an increase in the relative price of nontradables would in turn give rise to an appreciation of the real exchange rate.46

    For purposes of policy analysis, the extended PPP framework provides a basis for refining the traditional PPP formula to take account of trends in the relative prices of nontradables. CGER has not attempted such a refinement, however, partly because PPP considerations play a secondary role in its assessments, partly because its focus has been primarily on the industrial countries, where the case for such refinement is relatively weak, and partly because the prospective gains from such refinement are limited by difficulties in measuring (or finding suitable proxies for) trends in the relative prices of nontradables.47

    Estimated Reduced-Form Exchange Rate Models

    During the decade that followed the breakdown of the Bretton Woods exchange rate system in the early 1970s, international economists devoted considerable attention to the formulation and empirical testing of reduced-from exchange rate models. Contributors to this approach typically started by describing macroeconomic behavior in terms of a small number of key behavioral relationships and then combined the relationships to arrive at a single reduced-form equation for the exchange rate. These reduced-form models often fit the historically observed data fairly well within the sample periods over which they were estimated. As already noted, however, by the early 1980s a careful evaluation of such results had delivered the sobering assessment that a variety of state-of-the-art single-equation reduced-form models, in forecasting beyond the sample periods over which the models were estimated, were unable to outperform the predictions of a simple random walk model at short-run horizons of up to a year or longer.48 Indeed, this result even obtained when the post-sample model forecasts were based on realized values of the explanatory variables. Among other things, such findings drove home the point that in-sample goodness-of-fit is not a sufficient criterion for evaluating exchange rate models and that post-sample testing is important.

    Advances in econometric methodology since the mid-1980s have provided new techniques for seeking to estimate models that capture the long-run relationships between exchange rates and other economic fundamentals.49 This has led to the development of conceptual frameworks that focus simultaneously on long-run equilibrium conditions for both asset stocks and current account flows (more generally, national income account flows).50

    At the stage of empirical implementation, this approach may also involve the estimation of single-equation reduced-form models.51 These specifications have several features that limit their attractiveness. One feature is that they do not yield explicit estimates of equilibrium current account positions. This tends to lessen the appeal of the approach, as it is difficult to judge the plausibility of equilibrium exchange rate estimates unless the associated estimates of equilibrium current account positions can be assessed. A second feature of many of these models is that reducing the conceptual framework to an empirically estimable single equation generally involves simplifying assumptions and precludes identifying the parameters of the fully specified conceptual framework. As a result, the end product typically is significantly less informative than the conceptual framework.52

    General Equilibrium Frameworks

    Approaches based on simulations of general equilibrium models lie at the other end of the spectrum from those based on PPP or estimates of single-equation reduced-form models. The attractive feature of general equilibrium approaches is that the analysis is based on more complete models of macroeconomic behavior; by the same token, how-ever, the added complexity can be a drawback.

    Obviously, the general equilibrium approach cannot be implemented for a given country until a macroeconomic model has been specified and estimated for that country. In addition, the case for implementing such an approach depends on whether the available models have well-defined and conceptually appealing long-run properties. Some macroeconomic models that have been designed primarily for purposes of short-term forecasting do not have carefully specified long-run properties, which limits their appropriateness for analyzing the long-run relationship between the exchange rate and other economic fundamentals. Moreover, even when appropriate models are available, the resource-intensiveness of the general equilibrium approach may be an impediment to its application, particularly when the exercise involves assessing equilibrium exchange rates for a number of different currencies simultaneously.

    The IMF’s global macroeconometric model, MULTIMOD, has carefully specified long-run properties, but other features limit its attractiveness for generating estimates of equilibrium exchange rates.53 These features derive from the fact that MULTIMOD was not designed for purposes of generating a baseline forecast, but was rather intended to analyze the implications of various shocks to the global economy, using as its baseline the medium-term WEO projections, which reflect the detailed knowledge and judgments of the IMF’s country economists. The WEO projections are generated for a five-year horizon, and beyond that various assumptions are imposed to extend MULTIMOD’s baseline into a model-consistent balanced growth path. As such, MULTIMOD’s baseline (or control solution) reflects the assumption that real exchange rates remain constant over the WEO horizon and imposes fairly strong constraints on the paths that exchange rates take beyond the WEO horizon.

    For most of the “shock minus control” experiments conducted with MULTIMOD, the simulated effects of the shocks—and their plausibility—reflect the dynamic properties of the model, but are largely independent of the baseline. Using MULTIMOD to solve for equilibrium exchange rate paths, however, would be a different type of exercise than simulating the effects of exogenous shocks, and would not be very meaningful in light of the prior restrictions placed on the baseline paths for real exchange rates.54

    Appendix II A Model of Current Account Adjustment and Benchmark Comparators for the WEO Projections

    In applying its macroeconomic balance methodology, CGER relies on WEO projections as its estimates of underlying current account balances. For a number of years CGER also used a simple calibrated trade model developed in the IMF’s Research Department to generate alternative estimates of underlying current accounts (see Bayoumi and Faruqee, 1998). This framework—known as the aggregated RES model—employs common equation specifications and parameter values across countries and treats each country’s current account outflows as a homogeneous aggregate. Goods (and nonfactor services) produced in different countries are modeled as imperfect substitutes, with export (import) volumes assumed to depend on both the level of foreign (domestic) activity and the current and lagged values of the real effective exchange rate. In addition, the aggregated RES model assumes that exchange rate changes are fully passed through into import prices, with no effect on (domestic-currency-denominated) export prices.

    While it was helpful to be able to compare the WEO projections with alternative estimates derived from the aggregated RES model, the two sets of estimates were based on somewhat different concepts of the underlying current account. In particular, the aggregated RES model was designed to yield estimates of what base year current account positions would be if all countries were producing at potential output and after adjusting for the full effects of past exchange rate changes. By contrast, the WEO estimates are projections of five-year-ahead current account positions, conditional on all countries producing at potential output and no further changes in exchange rates.

    Over the past year, the Research Department has implemented several changes in the framework of the RES model to make it more useful for identifying cases in which the WEO-based estimates of underlying current account positions might warrant further consideration. In particular, the Research Department has started to develop a more disaggregated framework that can be used to generate benchmark projections for different components of the current account. The effort to extend the RES model is guided by the degree of disaggregation in the WEO projections, which decompose the current account into three components: non-oil trade, oil trade, and other items (i.e., factor income flows and transfers). The extensions that have been made to the RES framework have not altered the long-run responsiveness of current account positions to changes in real exchange rates, and the aggregated RES model continues to be relied upon in the calculations of equilibrium exchange rates.

    The first results of this effort are new equations for modeling the volumes of non-oil exports and imports. The new equations:

    • model non-oil trade volumes in a manner that is conceptually similar to the approach adopted in the aggregated RES model, but are designed to generate five- or six-year-ahead “projections” with explicit allowance for the effects of trends in activity variables (as distinct from simply allowing for output gaps to close);55

    • use measures of real absorption (domestic or weighted-average foreign) as activity variables; and

    • are specified in terms of levels of trade flows rather than ratios of trade flows to domestic GDP (or absorption), since domestic activity would not be an appropriate scale variable in six-year-ahead projections of export volumes.

    The specific forms of the non-oil trade volume equations are:

    where Qx and QM denote the logarithms of export and import volumes, A and Af denote the logarithms of domestic and foreign real absorption, R is the logarithm of the real exchange rate (with an increase in R representing a real appreciation), and the Δ terms refer to changes over six-year intervals.56

    The parameter values and lag patterns that describe the responses of trade volumes to changes in real exchange rates are identical to those in the previous specification of the RES model,57 but the activity elasticities have now been set at 1.9 for non-oil export volumes and 2.1 for non-oil import volumes. This parameterization—with non-oil trade volumes calibrated to grow about twice as fast as economic activity when real exchange rates remain constant—provides a “good fit” to the history of the past 15 years for the industrial countries on average.58 The slight asymmetry between the two elasticities allows for the fact that the trade volume equations focus on a measure of aggregate demand as the sole activity variable, abstracting from supply-side factors that tend to have stronger effects on developing country exports (industrial country imports) than on industrial country exports.59

    The new trade volume equations provide benchmark comparators for the corresponding components of the WEO projections. The benchmarks are not intended as an alternative set of projections: they do not take account of country-specific factors, and history reveals considerable variation over time in observed outcomes relative to retrospective benchmark calculations for individual countries. Nevertheless, the benchmarks provide CGER with a consistent and disciplined approach for considering the possible bias in the corresponding components of the WEO projections. Although the benchmark calculations take as given the WEO projections for absorption, they identify cases in which the projected changes in trade volumes relative to absorption seem inconsistent with the recent history of exchange rate movements.

    In addition to deriving benchmarks for the volumes of non-oil trade flows, CGER has initiated efforts to develop benchmarks for the prices of non-oil exports and imports by modeling these prices in a manner that captures an element of pricing-to-market in the short run while also reflecting the long-run trends in general price levels. It would be desirable, at a later stage, to further extend the benchmarks by modeling oil trade in a systematic manner that captures the sensitivity of oil import (and export) volumes to changes in the relative price of oil, and by systematically linking investment income flows to asset stock positions and the general level of interest rates.

    Appendix III The Saving-Investment Model for Industrial Countries

    The saving-investment (S-I) model relates each country’s current account (saving-investment balance) to four explanatory variables: its stage of development, as represented by its per capita income position; its fiscal position; the gap between its actual and potential output levels; and the level of world interest rates. An earlier version of the equation included the ratio of the dependent-age population to the working-age population as a fifth explanatory variable. The conceptual framework is described by Masson (1998). The model is simplified by imposing the constraint that aggregate saving must equal aggregate investment for the world as a whole, which provides a condition that links the level of the world interest rate to the other variables in the model. This condition can be substituted for the level of the world interest rate to derive an equation that relates each country’s S-I balance to relative levels of its per capita income, fiscal position, and output gap. Substituting out the world interest rate is not equivalent to treating it as a constant, or as irrelevant to the S-I balances of individual countries, but rather amounts to an implicit assumption that the world interest rate is determined by global variables—including, in particular, global measures of income per capita, the fiscal balance, and the output gap.

    Econometric testing of the model focused on both cross-section and panel results, exploring the panel data with both error-correction and partial-adjustment models, taking into consideration a reasonably long list of candidate explanatory variables, and settling in the end on a partial-adjustment equation. The specification had the form:

    where CUR is the current account (as a ratio to GDP), SUR is the fiscal surplus (as a ratio to GDP) relative to the industrial country average, DEM is the dependency ratio relative to the industrial country average, YPCAP is income per capita relative to that of a reference country (the United States), GAP is the output gap, and the μi are country-specific constant terms. The model also included a dummy variable to capture the effects of German unification from 1990 onward.60

    The econometric results that emerged from the original set of tests are described by Faruqee and Debelle (1998). The relative fiscal position and most of the country-specific constant terms were highly significant, as were the lagged dependent variable and the output gap. The econometric testing found that, for the industrial countries as a group, country-specific interest rates did not have significant effects on S-I balances when the output gap was also included as an explanatory variable. The indirect and limited role that interest rates play in the S-I model—and, more generally, the inadequate attention to factors that explain the relative attractiveness of investing in different countries—is an issue that warrants high priority in future efforts to strengthen CGER’s methodology.61

    The partial-adjustment model described above can be transformed to derive the following long-run (steady-state) specification:

    where ci=μi/(1λ) and γ=β/(1λ). The transformation reflects the assumption that output gaps vanish in the long run. This specification corresponds to equation (2) in the main text (but retains the demographic variable).

    The variables in the S-I model are viewed as direct determinants of saving and investment in the medium run, which are different from the explanatory variables of the standard trade-equation models that are used to generate estimates of underlying current account positions. Although one of the variables in the estimated partial-adjustment equation is the relative output gap, this variable does not appear in the long-run relationship. Its inclusion in the model enables a better overall fit of the historical data and better estimates of the parameters that capture the effects of the medium-run determinants.

    The estimated parameters used to generate the saving-investment norms for industrial countries were updated prior to CGER’s March 2001 assessment exercise. The decision to update was motivated in part by the fact that the framework that CGER had been using as a basis for the norms—equations (6.4) and (6.5) in Faruqee and Debelle (1998)—was estimated over the sample period 1971–93, which it had since become possible to extend through 1999. In addition, during the latter part of 1999 a major revision had been made to the historical real GDP data for the United States—the reference country for the relative income variable used in the S-I equation.

    The main implications of updating the data are evident from the five sets of equations reported in Table 2. In each case, the table reports the estimated parameters of the dynamic model and the implied estimates of the long-run coefficients; the panel estimates of the country-specific constant terms (fixed effects) are not reported in the table. In addition to exploring the implications of updating the original estimates (case 1), the new regression results (cases 2–5) are based on a normalized measure of the output gap (GAP) in which each country’s gap is expressed as a deviation from the GDP-weighted industrial country average.62

    Table 2.Saving-Investment Equation for Industrial Countries


    1. Original estimates0.660**0.153**−0.0450.045*−0.251**R2=0.76
    2. Normalized output gap0.590**0.132**−0.0490.049*−0.243**R2=0.72
    1971–9310.32−0.120.12D.W= 1.7
    3. Youth dependency ratio0.588**0.131**−0.076*0.048*−0.246**R2=0.72
    1971–9310.32−0.180.12D.W= 1.7
    4. Extended sample0.686**0.078**0.0240.036*−0.238**R2=0.77
    1971–9910.250.0760.12D.W= 1.8
    5. Extended sample0.685**0.073**0.033*−0.235**R2=0.77
    1971–9910.230.11D.W.= 1.8
    Note: A *(**) indicates significance at the 10 (5) percent level.

    Implied long-run coefficients, derived as γj=βj/(1λ).

    Comparison of cases 1 and 2 in the table isolates the effects of data revisions and normalization of the output gap. The main effect is a reduction in the long-run coefficient on the relative fiscal surplus (SUR) from 0.45 to 0.32.63 Case 3 indicates that replacing the overall dependency ratio with the youth dependency ratio raises the significance of the relative demographic variable (DEM) in the 1971–93 sample. Cases 4 and 5 show the effects of extending the sample period through 1999. Note that the fiscal coefficient falls somewhat further but remains highly significant, while dependency ratios (both the overall ratio and the ratio for youths) have negligible coefficients in the extended sample period; hence they are dropped in case 5.

    These findings motivated CGER to shift to basing its S-I norms on the results reported for case 5. In doing so, adjustments were made to the country-specific constant terms in cases for which the new norms would otherwise have implied uncomfortably large discontinuities from the norms that were used in the previous semiannual assessment exercise.64

    Although the long-run S-I equation includes only two explanatory variables—a situation that calls for continuing research efforts—the limitations of the equation should be evaluated with the following considerations in mind. First, while only two significant explanatory variables appear in the long-run equation, the country-specific constant terms are also highly significant, as are the lagged dependent variable and output gap in the dynamic equation; and the latter equation fits the data fairly well. Second, the S-I norms that CGER adopts are not determined by the equation alone; as noted, in some cases judgments have been incorporated into the norms (via adjustments to the country-specific constant terms) to allow for the effects of omitted factors that are not captured by the equation. Third, the exchange rate assessments do not simply reflect the macroeconomic balance estimates but also take account of PPP-based estimates and provide considerable scope for judgment. And fourth, even in its simple form the S-I equation contributes importantly to maintaining the consistency of CGER’s assessments by providing an agreed formula for changing the norms systematically (and gradually) over time in association with changes in variables that are perceived to be important determinants of medium-run saving-investment behavior.

    Appendix IV An Econometric Study of Saving-Investment Behavior in Developing Economies

    The study by Chinn and Prasad (2000) looked for systematic relationships between saving-investment balances and a fairly long list of possible explanatory variables for developing countries, focusing on both cross-section and time-series data. In contrast to CGER’s econometric analysis of the saving-investment behavior of the industrial countries, the Chinn-Prasad (CP) study gave emphasis to exploring the relevance of countries’ abilities to attract capital from abroad. The objective was to try to find an acceptable way to model the size of the S-I surpluses or deficits that have been historically observed for developing countries, taking into account, inter alia, economic factors that might limit their access to international capital markets.

    Table 3 reports some of the regression findings from the CP study. Column 1 shows results for a sample of 71 developing countries, while columns 2–5 show results obtained after separating the sample into the African countries and the developing countries excluding Africa.65 The regressions focus on explaining the current account to GDP ratio in terms of the variables listed in the rows of the table. The table reports the magnitudes of all coefficients that were estimated to differ significantly from zero, as well as the signs (+ or -) of coefficients that did not differ significantly from zero. The results reported in columns 3 and 5 include the foreign aid/GDP ratio (row 9) among the explanatory variables; the results in columns 1, 2, and 4 do not. Foreign aid receipts have had a significant negative effect on current account balances in Africa, but have not been a significant explanatory variable for the other developing countries as a group.

    Table 3.Results of Panel Regressions(Dependent variable: current account to GDP ratio)
    DevelopingDeveloping Countries
    CountriesExcluding AfricaAfrican Countries
    1. Government budget balance (ratio to GDP)0.39***0.26***0.18***0.60***0.64***
    2. Relative dependency ratio (young)–0.06*–0.16**
    3. Relative dependency ratio (old)+
    4. Relative income0.33*–0.45**
    5. Relative income squared0.25**0.27**+–0.49*+
    6. Financial deepening0.04***0.04**0.03*++
    7. Openness ratio–0.03**–0.02*+
    8. Net foreign asset/GDP ratio0.04***0.04***0.04***0.04***0.03*
    9. Foreign aid/GDP ratio+–0.51***
    10. Terms of trade volatility+0.03*+++
    11. Average GDP growth+
    12. Capital controls (current account)+++++
    13. Capital controls (capital account)+++
    14. Dummy for oil-exporting countries++++
    15. Significant time dummies1981–851981–851981–85None1976–80
    Adjusted R-squared0.440.450.490.490.59
    Number of observations2231551426864
    Note: Ordinary least squares specifications with dummy variables for each time period. The dependent and independent variables are nonoverlapping five-year averages of the corresponding annual variables, except for NFA/GDR, which is the observed value during the initial year of each five-year period. The symbols *,**, and *** indicate statistical significance at the 10 percent, 5 percent, and 1 percent levels, respectively; + and - indicate the signs of insignificant coefficients.

    As was the case in the industrial country regressions reported in the earlier discussion of equation (2), the government budget balance (row 1) has a significant positive effect on the overall saving-investment balance while the dependency ratios (rows 2 and 3) tend to have negative effects on the saving-investment balance. The regression analysis looked for a nonlinear effect of relative per capita income, based on the hypothesis that, at relatively low stages of development, increases in income would lend to improve a country’s access to foreign capital while, at advanced stages of development, the correlation between income and the current account would become positive (or less negative). The latter part of the hypothesis reflects the notion that countries at the highest income levels and most advanced stages of development tend to be capital exporters, an implication of the negative relationship between the abundance of existing capital and the marginal returns on additional investments. The hypothesis suggests a positive coefficient on the relative income squared term (row 5) and finds some support from the results in columns 1 and 2.

    Among the other variables tested in the regression analysis, financial deepening (as measured by a ratio of broad money to GDP), openness (exports plus imports as a share of GDP), and the initial net foreign asset (NFA) position as a ratio to GDP were found to have significant effects on the current account to GDP ratio (rows 6–8). A priori, the latter two variables were viewed as particularly relevant to a country’s ability to attract foreign capital. The finding of a negative correlation between the S-I balance and openness is consistent with the view that a country’s ability to attract foreign capital is enhanced by a relatively large capacity to generate export revenues or compress imports for purposes of meeting debt-service payments.

    The significant positive correlation between the current account and the initial NFA/GDP ratio does not have a clear interpretation.66 Several factors may be contributing to it. One factor is the direct contribution to the current account of the (net) income on the net foreign asset position. It may also be the case that whatever factors determined the relative attractiveness of countries to capital inflows in the past—and hence led to their initial net foreign liability (asset) positions—have continued to explain the relative sizes of their current account positions. Another plausible interpretation of the positive sign is that it is capturing the effect of the NFA/GDP ratio adjusting toward some long-run equilibrium level.67

    It would be desirable to be able to separate the different channels through which the current account is affected by the initial net foreign asset position.68 In principle, this could be done by modeling and estimating simultaneously the equilibrium NFA/GDP ratio and the path of adjustment to long-run equilibrium. Such an undertaking, however, is not straightforward. Similarly, the introduction of country-specific interest rates (or cost of capital measures) into the analysis would present challenges: even if adequate data were available, one would have to deal with two-way causality between the current account and the cost of capital.69

    The plausibility of the CP equation depends not only on the signs, magnitudes, and statistical significance of the individual estimated parameters but also on the implied estimates of historically normal saving-investment balances or “S-I norms,”70 For many of the emerging market economies on which CGER has focused, the implied S-I norms seem reasonable. For a number of others, however, the calculations seem implausible as estimates of equilibrium saving-investment balances.71


      Alberola, Enrique, Susana G. Cervero, J. Humberto Lopez, and Angel Ubide, 1999, “Global Equilibrium Exchange Rates—Euro, Dollar, ‘Ins,’ ‘Outs,’ and Other Major Currencies in a Panel Cointegration Framework,IMF Working Paper No. 99/175, (Washington: International Monetary Fund).

      Artus, Jacques, 1978, “Methods of Assessing the Long- Run Equilibrium Value of an Exchange Rate,Journal of International Economics, Vol. 8, pp. 27799.

      Artus, Jacques, and Malcolm Knight,1984, Issues in the Assessment of the Exchange Rates of the Industrial Countries: A Study by the Research Department, IMF Occasional Paper No. 29 (Washington: International Monetary Fund).

      Balassa, Bela, 1964, “The Purchasing-Power Parity Doctrine: A Reappraisal,Journal of Political Economy, Vol. 72 (December), pp. 58496.

      Barro, Robert J., 1989, “The Ricardian Approach to Budget Deficits,Journal of Economic Perspectives, Vol. 3 (Spring), pp. 3754.

      Bayoumi, Tamim, and Hamid Faruqee, 1998, “A Calibrated Model of the Underlying Current Account,” inExchange Rate Assessment: Extensions of the Macroeconomic Balance Approach, IMF Occasional Paper No. 167, ed. byPeter Isard and Hamid Faruqee (ed) (Washington: International Monetary Fund).

      Berg, Andrew, Eduardo Borensztein, Gian Maria Milesi- Ferretti, and Catherine Pattillo, 1999, Anticipating Balance of Payments Crises: The Role of Early Warning Systems, IMF Occasional Paper No. 186 (Washington: International Monetary Fund).

      Bernheim, B. Douglas, 1989, “Neoclassical Perspective on Budget Deficits,Journal of Economic Perspectives, Vol. 3 (Spring), pp. 5572.

      Breuer, Janice Boucher, 1994, “An Assessment of the Evidence on Purchasing Power Parity,” inEstimating Equilibrium Exchange Rates, ed. byJohn Williamson (ed) (Washington: Institute for International Economics).

      Cassel, Gustav, 1918, “Abnormal Deviations in International Exchanges,Economic Journal, Vol. 28, pp. 41315.

      Cassel, Gustav, 1922, Money and Foreign Exchange After 1914 (New York: Constable & Co.).

      Chinn, Menzie, and Eswar S. Prasad,2000, “Medium- Term Determinants of Current Accounts in Industrial and Developing Countries: An Empirical Exploration,IMF Working Paper No. 00/46 (Washington: International Monetary Fund).

      Clark, Peter B., Leonardo Bartolini, Tamim Bayoumi, and Steven Symansky, 1994, Exchange Rates and Economic Fundamentals: A Framework for Analysis, IMF Occasional Paper No. 115 (Washington: International Monetary Fund).

      Clark, Peter B., and Ronald MacDonald, 2000, “Filtering the BEER: A Permanent and Transitory Decomposition,IMF Working Paper No. 00/144 (Washington: International Monetary Fund).

      Debelle, Guy, and Hamid Faruqee,1996, “What Determines the Current Account? A Cross-Sectional and Panel Approach,IMF Working Paper No. 96/58 (Washington: International Monetary Fund).

      De Broeck, Mark, and Torsten Slok, 2001, “Interpreting Real Exchange Rate Movements in Transition Economies,IMF Working Paper No. 01/56 (Washington: International Monetary Fund).

      Dooley, Michael P. and Peter Isard,1987, “Country Preferences, Currency Values and Policy Issues,Journal of Policy Modeling, Vol. 9 (Spring), pp. 6581.

      Edwards, Sebastian,1989, Real Exchange Rates, Devaluation and Adjustment: Exchange Rate Policy in Developing Countries (Cambridge, Massachusetts: MIT Press).

      Edwards, Sebastian, 1994, “Real and Monetary Determinants of Real Exchange Rate Behavior: Theory and Evidence from Developing Countries,” inEstimating Equilibrium Exchange Rates, ed. byJohn Williamson (ed) (Washington: Institute for International Economics).

      Elbadawi, Ibrahim,1994, “Estimating Long-Run Equilibrium Real Exchange Rates,” inEstimating Equilibrium Exchange Rates, ed. byJohn Williamson (ed) (Washington: Institute for International Economics).

      Engel, Robert F., and C.W. J. Granger,1987, “Co-integration and Error Correction: Representation, Estimation, and Testing,Econometrica, Vol. 55 (March), pp. 25176.

      Faruqee, Hamid,1995, “Long-Run Determinants of the Real Exchange Rate: A Stock-Flow Perspective,Staff Papers, International Monetary Fund, Vol. 42 (March), No. 1, pp. 80107.

      Faruqee, Hamid,1998, “Methodology for Calculating Equilibrium Exchange Rates and Question of Global Consistency,” inExchange Rate Assessment: Extensions of the Macroeconomic Balance Approach, IMF Occasional Paper No. 167, ed. byPeter Isard and Hamid Faruqee (ed) (Washington: International Monetary Fund).

      Faruqee, Hamid, and Guy Debelle,1998, “Saving-Investment Balances in Industrial Countries: An Empirical Investigation,” inExchange Rate Assessment: Extensions of the Macroeconomic Balance Approach, IMF Occasional Paper No. 167, ed. byPeter Isard and Hamid Faruqee (ed) (Washington: International Monetary Fund).

      Faruqee, Hamid, Peter Isard, and Paul R. Masson, 1999, “A Macroeconomic Balance Framework for Estimating Equilibrium Exchange Rates,” inEquilibrium Exchange Rates, ed. byRonald MacDonald Jerome L. Stein (ed) (Boston: Kluwer Academic Publishers).

      Flood, Robert P., and Mark P. Taylor, 1996, “Exchange Rate Economics: What’s Wrong with the Conventional Macro Approach?” inThe Micro structure of Foreign Exchange Markets, ed. byJeffrey A. Frankel, Giampaolo Galli, and Alberto Giovannini (ed) (Chicago: University of Chicago Press).

      Flood, Robert P., and Andrew K. Rose, 1999, “Understanding Exchange Rate Volatility Without the Contrivance of Macroeconomics,Economic Journal, Vol. 109, No. 459 (November), pp. F66072.

      Frankel, Jeffrey A., and Andrew K. Rose, 1995, “Empirical Research on Nominal Exchange Rates,” inHandbook of International Economics, Vol. 3, ed. byGene M. Grossman Kenneth Rogoff (ed) (Amsterdam: North-Holland).

      Froot, Kenneth A., and Kenneth Rogoff, 1995, “Perspectives on PPP and Long-Run Real Exchange Rates,” inHandbook of International Economics, Vol. 3, ed. byGene M. Grossman Kenneth Rogoff (ed) (Amsterdam: North-Holland).

      Gilbert, Milton, and Irving B. Kravis, 1954, An International Comparison of National Products and the Purchasing Power of Currencies: A Study of the United States, the United Kingdom, France, Germany, and Italy (Paris: Organization for European Economic Cooperation).

      Goldman Sachs, 1996, The Foreign Exchange Market (July).

      Goldman Sachs, 1997, The Foreign Exchange Market (September).

      Goldstein, Morris, and Mohsin S. Khan, 1985, “Income and Price Effects in Foreign Trade,” inHandbook of International Economics, Vol. 2, ed. byRonald W. Jones and Peter B. Kenen (ed) (Amsterdam: North-Holland).

      Granger, C. W. J., 1981, “Some Properties of Time Series Data and Their Use in Econometric Model Specification,Journal of Econometrics, Vol. 16 (No. 1), pp. 12130.

      Group of Ten Deputies, 1993, International Capital Movements and Foreign Exchange Markets: A Report to the Ministers and Governors (Rome: Bank of Italy).

      Harrod, R. E, 1939, International Economics (London: Nisbet and Co.).

      International Monetary Fund, 2000, World Economic Outlook, October 2000 (Washington).

      Isard, Peter, 1995, Exchange Rate Economics, Cambridge Surveys of Economic Literature (Cambridge, England: Cambridge University Press).

      Isard, Peter, and Steven Symansky, 1996, “Long-Run Movements in Real Exchange Rates,” inExchange Rate Movements and Their Impact on Trade and Investment in the APEC Region, IMF Occasional Paper No. 145, ed. byTakatoshi Ito, Peter Isard, Steven Symansky, and Tamim Bayoumi (ed) (Washington: International Monetary Fund).

      Isard, Peter, and Hamid Faruqee, (ed)ed., 1998, Exchange Rate Assessment: Extensions of the Macroeconomic Balance Approach, IMF Occasional Paper No. 167 (Washington: International Monetary Fund).

      Kahn, Robert, and Roger Nord,1998, “Applications in IMF Surveillance Over Major Industrial Countries,” inExchange Rate Assessment: Extensions of the Macroeconomic Balance Approach, IMF Occasional Paper No. 167, ed. byPeter Isard Hamid Faruqee (ed) (Washington: International Monetary Fund).

      Knight, Malcolm D., and Paul R. Masson, 1988, “Fiscal Policies, Net Saving, and Real Exchange Rates: The United States, the Federal Republic of Germany, and Japan,” inInternational Aspects of Fiscal Policies, ed. byJacob Frenkel (ed) (Chicago: University of Chicago Press).

      Krajnyák, Kornélia,2000, “Switzerland’s External Position in International Perspective,”in IMF Staff Country Report No. 00/43 (Washington: International Monetary Fund).

      Kravis, Irving B., Alan Heston, Robert Summers,1982, World Product and Income: International Comparisons of Real Gross Product (Baltimore: Johns Hopkins University Press).

      Lane, Philip, and Gian Maria Milesi-Ferretti,2001a, “Long-Term Capital Movements,IMF Working Paper No. 01/107 (Washington: International Monetary Fund); forthcoming in NBER Macroeconomics Annual 2001.

      Lane, Philip, and Gian Maria Milesi-Ferretti,2001b, “External Wealth, the Trade Balance, and the Real Exchange Rate” (unpublished).

      Lane, Philip, and Gian Maria Milesi-Ferretti,2001c, “The External Wealth of Nations: Measures of Foreign Assets and Liabilities for Industrial and Developing Countries,Journal of International Economics, Vol. 55 (December), pp. 26394.

      Laxton, Douglas, Peter Isard, Hamid Faruqee, Eswar Prasad, and Bart Turtelboom,1998, MULTIMOD Mark III: The Core Dynamic and Steady-State Models, IMF Occasional Paper No. 164 (Washington: International Monetary Fund).

      MacDonald, Ronald, 1999, “Exchange Rate Behaviour: Are Fundamentals Important?Economic Journal, Vol. 109 (November), pp. F67391.

      Masson, Paul, 1998, “A Globally Consistent Conceptual Framework,” inExchange Rate Assessment: Extensions of the Macroeconomic Balance Approach, IMF Occasional Paper No. 167, ed. byPeter Isard Hamid Faruqee (ed) (Washington: International Monetary Fund)

      Meade, James E, 1951, The Theory of International Economic Policy, Volume One: The Balance of Payments (London: Oxford University Press).

      Meese, Richard A. and Kenneth Rogoff, 1983a, “Empirical Exchange Rate Models of the Seventies: Do They Fit Out of Sample?Journal of International Economics, Vol. 14 (February), pp. 324.

      Meese, Richard A. and Kenneth Rogoff, 1983b, “The Out-of-Sample Failure of Empirical Exchange Rate Models: Sampling Error or Misspecification?” inExchange Rates and International Macroeconomics, ed. byJacob A. Frenkel (ed) (Chicago: University of Chicago Press).

      Meese, Richard A. and Kenneth Rogoff,1988, “Was It Real? The Exchange Rate-Interest Differential Relation Over the Modern Floating-Rate Period,Journal of Finance, Vol. 43 (September), pp. 93348.

      Meredith, Guy, 1998, “A Dynamic Extension of the Macroeconomic Balance Approach,” inExchange Rate Assessment: Extensions of the Macroeconomic Balance Approach, IMF Occasional Paper No. 167, ed. byPeter Isard and Hamid Faruqee (Washington: International Monetary Fund).

      Metzler, Lloyd A., 1951, “Wealth, Saving and the Rate of Interest,Journal of Political Economy, Vol. 59, pp. 93046.

      Moggridge, Daniel E.,1972, British Monetary Policy, 1924—1931: The Norman Conquest of $4.86 (Cambridge, England: Cambridge University Press).

      Mussa, Michael, 1984, “The Theory of Exchange Rate Determination,” inExchange Rate Theory and Practice, ed. byJohn F. O. Bilson Marston Richard C. (ed) (Chicago: University of Chicago Press).

      Nurkse, Ragnar, 1945, “Conditions of International Monetary Equilibrium,”Essays in International Finance, No. 4 (Princeton, New Jersey: Princeton University Press).

      Obstfeld, Maurice, and Kenneth Rogoff, 2000, “The Six Major Puzzles in International Macroeconomics: Is There a Common Cause?” inNBER Macroeconomics Annual 2000, ed. byB. Bernanke K. Rogoff (ed) (Cambridge, Massachusetts: MIT Press).

      Polak, Jacques J., 1995, “Fifty Years of Exchange Rate Research and Policy at the International Monetary Fund,” Staff Papers, International Monetary Fund, Vol. 42 (December), pp. 73461.

      Ricardo, David, 1821, “On the Principles of Political Economy and Taxation,” reprinted inThe Works and Correspondence of David Ricardo, Vol. I, ed. byPiero Sraffa (ed) (Cambridge, England: Cambridge University Press, 1951).

      Rogoff, Kenneth,1996, “The Purchasing Power Parity Puzzle,Journal of Economic Literature, Vol. 34 (June), pp. 64768.

      Rogoff, Kenneth, 1999, “Monetary Models of Dollar/Yen/Euro Nominal Exchange Rates: Dead or Undead?Economic Journal, Vol. 109 (November), pp. F65559.

      Samuelson, Paul A., 1964, “Theoretical Notes on Trade Problems,Review of Economics and Statistics, Vol. 46 (May), pp. 14554.

      Samuelson, Paul A., 1994, “Facets of Balassa-Samuelson Thirty Years Later,Review of International Economics, Vol. 2 (October), pp. 20126.

      Summers, Robert, and Alan Heston, 1991, “The Penn World Table (Mark 5): An Expanded Set of International Comparisons, 1950–1988,Quarterly Journal of Economics, Vol. 106 (May), pp. 32768.

      Swan, T. W., 1963, “Longer-Run Problems of the Balance of Payments,” inThe Australian Economy: A Volume of Readings, W. Arndt W. M. Corden (ed) (Melbourne: F. W. Cheshire).

      Williamson, John, 1985, The Exchange Rate System (Washington: Institute for International Economics).

      Williamson, John, ed.,1994, Estimating Equilibrium Exchange Rates (Washington: Institute for International Economics).

      Williamson, John, andMolly Mahar, 1998, “Current Account Targets,”Appendix A in Real Exchange Rates for the Year 2000, ed. bySimon Wren-Lewis and Rebecca L. Driver (ed) (Washington: Institute for International Economics).

    In addition, sharp currency appreciations often give rise to political pressures to protect exchange-rate-sensitive sectors of the economy from foreign competition.

    Meese and Rogoff (1983a, 1983b, 1988). Numerous attempts to overturn the Meese-Rogoff results have failed; see Rogoff (1999) and the survey by Frankel and Rose (1995). Moreover, as emphasized by Flood and Rose (1999), the short-run volatility of exchange rates far exceeds the volatility of macroeconomic fundamentals.

    Economists have also had limited success in finding fundamentals-based explanations for the short-run behavior of the prices of assets other than foreign exchange—for example, equity prices.

    The dramatic changes that have occurred in the size of financial asset holdings, and in the amounts of these holdings that can be moved rapidly between countries and/or currencies, have led some to suggest that the short-run behavior of exchange rates can be explained in terms of capital flows and that current account flows have a relatively minor influence. This view can be misleading in two senses. First, although exchange rate movements and capital flows may be closely associated in the short run, it would be misleading to regard capital flows as more than a proximate cause of exchange rate movements; a deeper understanding of the short-run behavior of exchange rates would require an understanding of the factors that drive capital flows. Second, because the capital account and current account are linked by the balance of payments identity, it would be misleading to suggest that the behavior of exchange rates is more closely related to the capital account than to the current account.

    The term purchasing power parity was coined in the early twentieth century by Cassel (1918, 1922).

    Whether one looks at consumer price indices, GDP deflators, wholesale price indices, unit labor costs, or export price indices, there is considerable month-to-month and quarter-to-quarter variation in the associated measures of real exchange rates. Hyperinflations provide exceptional circumstances in which PPP has not been discredited as a description of short-run behavior.

    The fit is not as close for the developing countries or for the industrial countries over the previous quarter century. Indeed, as elaborated in the discussion of the Balassa-Samuelson hypothesis in Appendix I, there is empirical evidence of systematic deviations from PPP.

    See Polak (1995), who emphasizes the focus on external balance and the use of elasticity calculations but does not mention internal balance. The macroeconomic balance approach apparently began to take shape in the writings of Nurkse (1945) and Metzler (1951) and benefited from pathbreaking contributions by Meade (1951) and Swan (1963).

    There are some definitional distinctions between national accounts concepts and balance of payments concepts that need to be taken into account when applying the macroeconomic balance framework, especially with respect to the treatment of net factor income payments and transfers.

    This simplifying assumption is consistent with the fact that most empirically estimated reduced-form models of industrial country saving and investment behavior do not include the exchange rate among the main determinants. As elaborated below, it is also associated with an assumption that the industrial countries have perfect access to international capital markets (apart from any time-invariant interest rate premia). A more complete macroeconomic framework could recognize that exchange rates may influence saving and investment through their effects on the terms of trade, the profitability of the traded-goods sector (or subsectors susceptible to “Dutch disease”), the level of potential output, and the distribution of income.

    As an alternative to treating the goods of different countries as imperfect substitutes, which is the standard approach in models used for projecting trade volumes empirically, some conceptual models distinguish between tradable and nontradable goods, assume that tradable goods are homogeneous across countries, and model current account adjustment as a process of shifting the balance between each country’s production and absorption of tradable goods; see, for example, Dooley and Isard (1987) and Obstfeld and Rogoff (2000). Attempts to apply the latter approach empirically are typically constrained by the paucity of data that adequately decompose production and absorption into their tradable and nontradable components.

    For major industrial countries, the WEO projections are generated to a large extent from formal models, but the structures of these models differ across countries.

    Polak (1995) points to the lack of agreed analytic procedures for modeling equilibrium current account positions as a major weakness in the IMF’s applications of the macroeconomic balance approach during the 1970s and 1980s. Prior to the estimation of a model of saving-investment behavior over the medium run, CGER’s macroeconomic balance assessments for the major currencies were based on ad hoc assumptions about equilibrium ratios of net foreign assets to GDP and associated equilibrium ratios of current accounts to GDP. See Williamson and Mahar (1998) for an alternative exercise in generating S-I norms.

    The original econometric estimates are described in Debelle and Faruqee (1996) and Faruqee and Debelle (1998), and the conceptual framework is further elaborated in Masson (1998) and Faruqee, Isard, and Masson (1999).

    An implicit assumption is that each country can borrow or lend any amount of capital internationally (i.e., any shortfall or excess of domestic saving relative to domestic investment) at a fixed premium above the world rate of interest. This assumption of perfect capital mobility may be an acceptable simplification for industrial countries but would be inappropriate for most other economies.

    There is a long-standing debate on the economic implications of public deficits; see Barro (1989) and Bernheim (1989) for reviews of the Ricardian and neoclassical perspectives.

    To the extent that success in mitigating macroeconomic instability tends to promote per capita income growth (and perhaps reduces the cost of servicing public debt) over the medium run, the framework can be viewed to indirectly capture certain longer-run effects of monetary policy.

    The norms reflect adjustments to the country-specific constant terms for Japan. Australia, Denmark, New Zealand, Norway, and Switzerland. In each case, the adjustment has raised the level of the curve shown in Figure 5 without affecting the time profile of the norm; the implicit counterpart of these adjustments is an increase in the size of the deficit norm for the developing countries. For Japan, the norm has been adjusted upward to offset the influence on the estimated country-specific constant term of the abnormal component of investment during the bubble of the 1980s and early 1990s. For Norway, the norm has been judgmentally raised in light of the effects on national wealth and saving of the general rise in oil prices since the historical sample period. For Switzerland, the norm has been judgmentally raised to better capture the implications of a relatively high net foreign asset position. For Australia, Denmark, and New Zealand, the norms were judgmentally raised (to a higher surplus for Denmark and lower deficits in the other two cases) to avoid discontinuities on the occasion of shifting the calculations to updated estimates of the S-I equation.

    The deterioration of Japan’s relative structural fiscal deficit during the 1990s stems from both the substantial deterioration of Japan’s own fiscal position and significant improvements in structural fiscal positions elsewhere.

    The panel data set used to estimate the S-I model included the members of the euro area as individual countries, partly in recognition of the separate status of their currencies during most of the sample period and also to provide more degrees of freedom. Aggregate euro area variables were then constructed either by adding the variables for the individual member countries or as weighted averages, as appropriate. (Weighted averages make sense for variables expressed as ratios to GDP’, including the country-specific constant terms in the equation that explains the S-I balance as a ratio to GDP.) For purposes of using the RES model to calculate the amounts that current accounts and exchange rates need to adjust to reach medium-run equilibrium levels (see below), intra-euro-area trade was removed from measures of euro area exports and imports. This makes the framework consistent with the assumption that adjustments in euro exchange rates affect trade between the euro area and the rest of the world but do not affect trade among members of the euro area. It also has the effect of treating the euro area in the aggregate as a less open economy than its individual member countries on average.

    Such oversimplification limits the usefulness of the macroeconomic balance approach as a normative tool, just as the usefulness of a PPP-based assessment for normative purposes is limited by sensitivity to the type of price index and averaging period chosen.

    Krajnyák (2000) has developed a portfolio allocation model of S-I behavior for Switzerland.

    As elaborated in Appendix III, the estimates of the S-I model were updated prior to the March 2001 assessment exercise, which changed the 2006 value of the calculated S-I norm for the United States to a deficit of 1.6 percentage points of GDP—an upward revision (in absolute value) of ¾ of 1 percentage point of GDP. That reduced by about 10 percentage points the macroeconomic balance calculation of the amount that the U.S. dollar would need to depreciate in multilateral terms to reach its medium-run equilibrium level.

    As noted earlier, this simplifying assumption could be relaxed in principle (implying a nonvertical S-I line)—for example, to recognize the phenomenon of “Dutch disease.” However, the simplifying assumption is consistent with many other empirical models of saving and investment behavior.

    See Faruqee (1998). When the sum of the underlying current account positions is reasonably similar to the sum of the saving-investment norms, the results of calculating equilibrium exchange rates in a multilateral framework generally do not deviate materially from those derived by focusing on countries individually (as in Figure 4).

    Since the introduction of the euro, CGER has assessed the euro area as a single bloc, effectively reducing the number of currencies that enter its multilateral calculations. The multilateral calculations focus on real exchange rates, and the methodology abstracts from the second-order effects that inflation differentials (and, hence, changes in real exchange rates) within the euro area might have on current account adjustment for the euro area as a whole. Because the euro area as an aggregate is a less open economy than its individual members on average, its consolidation into a single unit can be regarded, in the context of Figure 4, as the substitution of a relatively steep UCUR locus for a set of relatively flat UCUR lines. A relatively flat UCUR line implies that a country’s current account has a relatively high responsiveness to a change in its effective exchange rate vis-à-vis the currencies of all other countries. However, because trade among members of the euro area is a relatively large share of their international trade, the aggregate current account of this group of countries with relatively flat UCUR lines and internally fixed exchange rates will not exhibit a relatively high responsiveness to a change in their exchange rates vis-à-vis external currencies. At the time that CGER modified its methodology to incorporate the euro area as a single bloc (early 1999), its assessments under the old methodology did not suggest a need for substantial adjustments of the individual euro legacy currencies vis-à-vis each other. Calculations of the euro’s medium-run equilibrium exchange rates under the new methodology differed by about 3 percent from the corresponding estimates for the synthetic euro under the old methodology.

    See Goldstein and Khan (1985). The calibrated long-run (and medium-run) elasticity of export (import) volume with respect to the real effective exchange rate is 0.71 (0.92), and it is assumed that exchange rate changes are fully passed through over the medium run into import prices, with no effect on domestic-currency-denominated export prices.

    Outside forecasts appear to exhibit a broadly similar degree of global export pessimism. In particular, as of April 2001, the Consensus Forecasts, which are based on surveys of more than 200 prominent financial and economic forecasters and extend over a two-year horizon, showed the global current account deficit widening to about -$300 billion in 2001 and -$340 billion in 2002, compared with the WEO projections of about -$260 billion and -$300 billion, respectively.

    The uniform 4.4 percentage point “adjustment” is the outcome of taking as given the WEO current account projections for all countries and the saving-investment norms for the industrial countries, and of accepting as plausible the developing country S-I norm that is implied by the assumed norm for the global S-I balance. The methodology does not allow one to directly allocate the adjustment for bias in a nonuniform way. It does, however, permit indirect allocation via nonuniform adjustments to countries’ underlying current accounts or saving-in vestment norms, which would have nonuniform effects on the estimates of equilibrium exchange rates. As an extreme example, if the U.S. current account deficit projected for 2006 had been reduced by the entire $254 billion difference (as of early 2001) between the global discrepancy in underlying current account positions and the global discrepancy in S-I norms—thereby reducing the projected U.S. current account deficit by 1.9 percentage points of GDP—no uniform adjustment for bias would have been required. Such indirect allocation of the entire adjustment to the United States would have raised considerably (by about 25 percent) the macroeconomic balance estimate of the medium-run equilibrium value of the dollar, although it would not have affected PPP-based assessments of the dollar’s overvaluation.

    See Isard and Faruqee (1998). The retrospective applications relied on the RES model to generate estimates of underlying current account positions. They suggested that the U.S. dollar was very substantially overvalued against the Japanese yen and the deutsche mark in early 1985 and substantially undervalued against both of the other key currencies in the spring of 1995. They also suggested that the pound and the lira were substantially overvalued against their partner currencies in the Exchange Rate Mechanism in the months prior to the summer 1992 currency crisis, reflecting large underlying current account deficits in the United Kingdom and Italy.

    Along with the assessments of multilateral exchange rates, the macroeconomic balance calculations yield globally consistent assessments of bilateral exchange rates (not shown in Table 1), and CGER extends its summary judgments to bilateral rates.

    See Kahn and Nord (1998) and IMF News Brief No. 95/12, April 14, 1995.

    For background on the staff’s work on early warning systems, see Berg and others (1999).

    This reflects the general state of the empirical literature on saving and investment behavior, which has not achieved much success in linking movements in S-I balances to structural changes.

    The definition and measurement of internal balance or potential output poses greater difficulties in applications to emerging market economies than in applications to industrial countries, and applications to emerging market economies also call for greater attention to the commodity composition of trade and the role of official financing.

    ln practice, the measurement of NFL/GDP ratios poses several difficulties—in particular, the difficulty of revaluing asset and liability stocks to account for changes in exchange rates and the question of whether to include capital that has flown offshore in a country’s stock of foreign assets. Reported data on international investment positions are not available for many developing countries and do not extend back very far where they are available. In addition, these data are based on national sources that employ a variety of methodologies in calculating the values of foreign asset and liability stocks. The data set used by CGER has been assembled and analyzed by Lane and Milesi-Ferrelti (2001a, 2001c) and represents the most comprehensive effort to date to construct NFL data based on a uniform methodology. Since that data set ends in 1998, it was extended by cumulating current accounts for 1999–2000 and adding them to NFL stocks at the end of 1998. Measurement error in official current account data, especially in recent years, remains a potential problem with these estimates.

    It should be recognized, for example, that in countries with IMF-supported programs, policy regimes and economic structure generally undergo substantial changes that tend to be reflected in the WEO projections but may be inadequately taken into account by the criteria for assessing current account sustainability.

    Most staff reports on Article IV consultations with industrial countries outside the euro area, as well as staff reports on consultations with European Union institutions, now include discussions of CGER (or CGER-type) assessments.

    Harrod (1939) provided an earlier discussion of some of the key arguments made by Balassa and Samuelson, the seeds of which have been traced back to Ricardo (1821).

    See Isard and Symansky (1996) for a formal description of how the PPP formula can be modified to allow for Balassa-Samuelson effects. Like traditional PPP, the extended hypothesis does not provide an empirically valid description of exchange rate behavior in the short run.

    Some economists have used GDP per capita as a proxy; see De Broeck and Sløk (2001) and International Monetary Fund (2000), Box 4.4, pp. 168–69.

    The key advances were the introduction of the concept of cointegration by Granger (1981) and Engel and Granger (1987) and the subsequent development of time-series econometrics.

    Such conceptual frameworks define the equilibrium levels or timepaths of exchange rates as the levels or paths that give rise to current account flows that are consistent with convergence to long-run asset stock equilibrium. An early contribution to the empirical implementation of these models was made by Faruqee (1995), based on a continuous-time version of the stock-flow consistent framework described by Mussa (1984). More recently, Alberola and others (1999) have extended the methodology further. Goldman Sachs (1996, 1997) has relied on the methodology in providing forecasts for its clients. See MacDonald (1999) and Clark and MacDonald (2000) for perspectives on the literature. See also Edwards (1989, 1994) and Elbadawi (1994) for applications of this methodology to developing countries.

    These reduced-form error-correction models, which in some cases are vector-equation frameworks rather than single equations, contain long-run components that embody the PPP hypothesis; in that sense, they can also be viewed as extensions of the traditional PPP hypothesis. In these reduced-form models, however, PPP is embodied not as a time-invariant level of the long-run real exchange rate, but rather as a steady-state condition in which the equilibrium level of the real exchange rate depends on the steady-state levels of various fundamental determinants.

    Another difficulty arises from the fact that the models relate trends in real exchange rates to the observed values of other variables. Accordingly, the derived estimates of equilibrium exchange rates are conditional on assumptions about the equilibrium values of explanatory variables; and for some of these explanatory variables, such as ratios of net foreign liabilities to GDP, economists may not have strong prior beliefs about long-run equilibrium values.

    See Laxton and others (1998) for a description of MULTIMOD.

    Meredith (1998) has modified the Japan block of MULTIMOD to develop a framework in which it does appear valid to interpret the solutions for exchange rates as equilibrium paths. This required strong simplifying assumptions, however, and an effort to extend the framework to the other country blocks in MULTIMOD would be a resource-intensive project.

    The six-year-ahead framework reflects the fact that incomplete or very preliminary data for the current year often makes it desirable to use the previous year as the base year when generating the benchmark projections.

    Under the assumption that real exchange rates remain constant at prevailing levels, the square-bracketed terms in the equations—when the dating of ΔR corresponds to the change from the base year to six years later—are equivalent to the expression:


    where δR represents the (logarithmic) difference between the real exchange rate on which the WEO projections are conditioned and the average real exchange rate during the base year; δR-1 represents the difference between the average real exchange rate in the base year and the average real exchange rate during the previous year; and so forth. The weights on each “incremental”exchange rate change can be interpreted as the proportion of the long-run effect of that change that is yet to be realized.

    The framework reported in Bayoumi and Faruqee (1998) was modified at the end of 1998 based on econometric work that supported the shift from a three-year distributed lag to a five-year lag.

    An elasticity of 2.5, which would be more consistent with the previous version of the calibrated RES current account model (in which ratios of trade volumes to GDP were assumed to exhibit an elasticity of 1.5 with respect to GDP), would show substantially more trade expansion than has been observed historically.

    The deviations of these elasticities from 2.0 generate six-year-ahead projections in which the aggregate current account position for the industrial countries is roughly consistent with the industrial country S-I norm. The asymmetry in activity elasticities tends to (partly) offset the trade-balance effects of absorption growing more slowly in industrial countries than in the rest of the world.

    See Faruqee and Debelle (1998) for details on the adjustment for German unification.

    A number of empirical studies have found strong empirical links between current account flows and country-specific interest rates; see, for example, Obstfeld and Rogoff (2000).

    The output gap is the only country variable that was not normalized in the original estimation work reported by Faruqee and Debelle (1998).

    Not normalizing the output gap—that is, using the exact specification in case 1 but with the revised data—would yield an insignificant coefficient (near zero) on income per capita and a long-run fiscal coefficient of 0.36.

    The adjustments to the norms are described in a footnote that accompanies the discussion of Figure 5 in the main text. Although the desire to avoid significant discontinuities motivated adjustments in cases for which the previous norms had seemed reasonable, in other cases the primary motivation was to adjust (or restore previous adjustments) for important country-specific considerations that are not adequately captured by the S-I model.

    Because of data limitations (short time series), transition economies were not included in the sample of countries. Regional dummy variables designed to test the extent to which the developing countries can be treated as a homogeneous group proved insignificant for all regions except Africa. A dummy variable for oil-exporting countries was significantly positive in same specifications, but not in the “final specifications” shown in Table 3.

    The CP data set covered the period 1971–95, which was broken into nonoverlapping five-year subperiods, with the panel regression analysis focusing on corresponding five-year averages of annual data as a strategy for trying to see through the shorter-run influences on current account behavior. The initial NFA/GDP ratio was then measured as the ratio existing at the beginning of each subperiod.

    However, the correlation between the current account and NFA can be either negative or positive during the transition to long-run equilibrium.

    Lane and Milesi-Ferretti (2001b) have taken a significant step in this direction by focusing on the trade balance (TB) and the relationship between TB/GDP and NFA/GDP.

    Lane and Milesi-Ferretti (2001a) have recently shown that the evolution of NFA positions over the period 1970–98 can be largely explained by shifts in relative output levels, the stock of public debt, and demographic variables, with the latter two factors particularly important for developing countries.

    In calculating S-I norms for future years, it is highly desirable to shift to WEO data (and projections) for the explanatory variables in the CP equation. Overall, the WEO data generally move in line with the corresponding series in the CP data set, but to align the filled values of the equation (based on WEO data) with the sample period observations it is necessary to introduce country-specific constant terms (or one-time level adjustments) into the calculations. The constant terms have been calibrated to imply mean-zero errors between observed and fitted current-account-to-GDP ratios over the period 1980–95.

    A number of factors that seem important in assessing medium-term sustainability—such as political conditions, the role of past defaults, and the climate for and level of foreign direct investment—are precluded from the analysis by the inherent difficulties of quantification or limitations in available data.

    Recent Occasional Papers of the International Monetary Fund

    209. Methodology for Current Account and Exchange Rate Assessments, by Peter Isard, Hamid Faruqee, G. Russell Kincaid, and Martin Fetherston. 2001.

    208. Yemen in the 1990s: From Unification to Economic Reform, by Klaus Enders, Sherwyn Williams, Nada Choueiri, Yuri Sobolev, and Jan Walliser. 2001.

    207. Malaysia: From Crisis to Recovery, by Kanitta Meesook, II Houng Lee, Olin Liu, Yougesh Khatri, Natalia Tamirisa, Michael Moore, and Mark H, Krysl. 2001.

    206. The Dominican Republic: Stabilization, Structural Reform, and Economic Growth, by Alessandro Giustiniani, Werner C. Keller, and Randa E. Sab. 2001.

    205. Stabilization and Savings Funds for Nonrenewable Resources, by Jeffrey Davis, Rolando Ossowski, James Daniel, and Steven Barnett. 2001.

    204. Monetary Union in West Africa (ECOWAS): Is It Desirable and How Could It Be Achieved? by Paul Masson and Catherine Pattillo. 2001.

    203. Modern Banking and OTC Derivatives Markets: The Transformation of Global Finance and Its Implications for Systemic Risk, by Garry J. Schinasi, R. Sean Craig, Burkhard Drees, and Charles Kramer. 2000.

    202. Adopting Inflation Targeting: Practical Issues for Emerging Market Countries, by Andrea Schaechter, Mark R. Stone, and Mark Zelmer. 2000.

    201. Developments and Challenges in the Caribbean Region, by Samuel Itam, Simon Cueva, Erik Lundback, Janet Stotsky, and Stephen Tokarick. 2000.

    200. Pension Reform in the Baltics: Issues and Prospects, by Jerald Schiff, Niko Hobdari, Axel Schimmelpfennig, and Roman Zytek. 2000.

    199. Ghana: Economic Development in a Democratic Environment, by Sérgio Pereira Leite, Anthony Pellechio, Luisa Zanforlin, Girma Begashaw, Stefania Fabrizio, and Joachim Harnack. 2000.

    198. Setting Up Treasuries in the Baltics, Russia, and Other Countries of the Former Soviet Union: An Assessment of IMF Technical Assistance, by Barry H. Potter and Jack Diamond. 2000.

    197. Deposit Insurance: Actual and Good Practices, by Gillian G.H. Garcia. 2000.

    196. Trade and Trade Policies in Eastern and Southern Africa, by a staff team led by Arvind Subramanian, with Enrique Gelbard, Richard Harmsen, Katrin Elborgh-Woytek, and Piroska Nagy. 2000.

    195. The Eastern Caribbean Currency Union—Institutions, Performance, and Policy Issues, by Frits van Beek, José Roberto Rosales, Mayra Zermeño, Ruby Randall, and Jorge Shepherd. 2000.

    194. Fiscal and Macroeconomic Impact of Privatization, by Jeffrey Davis, Rolando Ossowski, Thomas Richardson, and Steven Barnett. 2000.

    193. Exchange Rate Regimes in an Increasingly Integrated World Economy, by Michael Mussa, Paul Masson, Alexander Swoboda, Estehan Jadresic, Paolo Mauro, and Andy Berg. 2000.

    192. Macroprudential Indicators of Financial System Soundness, by a staff team led by Owen Evans, Alfredo M. Leone, Mahinder Gill, and Paul Hilbers. 2000.

    191. Social Issues in IMF-Supported Programs, by Sanjeev Gupta, Louis Dicks-Mireaux, Ritha Khemani, Calvin McDonald, and Marijn Verhoeven. 2000.

    190. Capital Controls: Country Experiences with Their Use and Liberalization, by Akira Ariyoshi, Karl Habermeier, Bernard Laurens, Inci Ötker-Robe, Jorge Iván Canales Kriljenko, and Andrei Kirilenko. 2000.

    189. Current Account and External Sustainability in the Baltics, Russia, and Other Countries of the Former Soviet Union, by Donal McGettigan. 2000.

    188. Financial Sector Crisis and Restructuring: Lessons from Asia, by Carl-Johan Lindgren, Tomás J.T. Baliño, Charles Enoch, Anne-Marie Gulde, Marc Quintyn, and Leslie Teo. 1999.

    187. Philippines: Toward Sustainable and Rapid Growth, Recent Developments and the Agenda Ahead, by Markus Rodlauer, Prakash Loungani, Vivek Arora, Charalambos Christofides, Enrique G. De la Piedra, Piyabha Kongsamut, Kristina Kostial, Victoria Summers, and Athanasios Vamvakidis. 2000.

    186. Anticipating Balance of Payments Crises: The Role of Early Warning Systems, by Andrew Berg, Eduardo Borensztein, Gian Maria Milesi-Ferretti, and Catherine Pattillo. 1999.

    185. Oman Beyond the Oil Horizon: Policies Toward Sustainable Growth, edited by Ahsan Mansur and Volker Treichel. 1999.

    184. Growth Experience in Transition Countries, 1990–98, by Oleh Havrylyshyn, Thomas Wolf, Julian Berengaut, Marta Castello-Branco, Ron van Rooden, and Valerie Mercer-Blackman. 1999.

    183. Economic Reforms in Kazakhstan, Kyrgyz Republic, Tajikistan, Turkmenistan, and Uzbekistan, by Emine Gürgen, Harry Snoek, Jon Craig, Jimmy McHugh, Ivailo Izvorski, and Ron van Rooden. 1999.

    182. Tax Reform in the Baltics, Russia, and Other Countries of the Former Soviet Union, by a staff team led by Liam Ebrill and Oleh Havrylyshyn. 1999.

    181. The Netherlands: Transforming a Market Economy, by C. Maxwell Watson, Bas B. Bakker, Jan Kees Martijn. and Ioannis Halikias. 1999.

    180. Revenue Implications of Trade Liberalization, by Liam Ebrill, Janet Stotsky, and Reint Gropp. 1999.

    179. Disinflation in Transition: 1993–97, by Carlo Cottarelli and Peter Doyle. 1999.

    178. IMF-Supported Programs in Indonesia, Korea, and Thailand: A Preliminary Assessment, by Timothy Lane, Atish Ghosh, Javier Hamann, Steven Phillips, Marianne Schulze-Ghattas, and Tsidi Tsikata. 1999.

    177. Perspectives on Regional Unemployment in Europe, by Paolo Mauro, Eswar Prasad, and Antonio Spilimbergo. 1999.

    176. Back to the Future: Postwar Reconstruction and Stabilization in Lebanon, edited by Sena Eken and Thomas Helbling. 1999.

    175. Macroeconomic Developments in the Baltics, Russia, and Other Countries of the Former Soviet Union, 1992–97, by Luis M. Valdivieso. 1998.

    174. Impact of EMU on Selected Non-European Union Countries, by R. Feldman, K. Nashashibi, R. Nord, P. Allum, D. Desruelle, K. Enders, R. Kahn, and H. Temprano-Arroyo. 1998.

    173. The Baltic Countries: From Economic Stabilization to EU Accession, by Julian Berengaut, Augusto Lopez-Claros, Françoise Le Gall, Dennis Jones, Richard Stern, Ann-Margret Westin, Effie Psalida, Pietro Garibaldi. 1998.

    172. Capital Account Liberalization: Theoretical and Practical Aspects, by a staff team led by Barry Eichengreen and Michael Mussa, with Giovanni Dell’Ariccia, Enrica Detragiache, Gian Maria Milesi-Ferretti, and Andrew Tweedie. 1998.

    171. Monetary Policy in Dollarized Economies, by Tomás Baliño, Adam Bennett, and Eduardo Borensztein. 1998.

    170. The West African Economic and Monetary Union: Recent Developments and Policy Issues, by a staff team led by Ernesto Hernández-Catá and comprising Christian A. François, Paul Masson, Pascal Bouvier, Patrick Peroz, Dominique Desruelle, and Athanasios Vamvakidis. 1998.

    169. Financial Sector Development in Sub-Saharan African Countries, by Hassanali Mehran, Piero Ugolini, Jean Phillipe Briffaux, George Iden, Tonny Lybek, Stephen Swaray, and Peter Hayward. 1998.

    168. Exit Strategies: Policy Options for Countries Seeking Greater Exchange Rate Flexibility, by a staff team led by Barry Eichengreen and Paul Masson with Hugh Bredenkamp, Barry Johnston, Javier Hamann, Esteban Jadresic, and Inci Ötker. 1998.

    167. Exchange Rate Assessment: Extensions of the Macroeconomic Balance Approach, edited by Peter Isard and Hamid Faruqee. 1998

    166. Hedge Funds and Financial Market Dynamics, by a staff team led by Barry Eichengreen and Donald Mathieson with Bankim Chadha, Anne Jansen, Laura Kodres, and Sunil Sharma. 1998.

    165. Algeria: Stabilization and Transition to the Market, by Karim Nashashibi, Patricia Alonso-Gamo, Stefania Bazzoni, Alain Féler, Nicole Laframboise, and Sebastian Paris Horvitz. 1998.

    164. MULTIMOD Mark III: The Core Dynamic and Steady-State Model, by Douglas Laxton, Peter Isard, Hamid Faruqee, Eswar Prasad, and Bart Turtelboom. 1998.

    163. Egypt: Beyond Stabilization, Toward a Dynamic Market Economy, by a staff team led by Howard Handy. 1998.

    Note: For information on the title and availability of Occasional Papers not listed, please consult the IMF Publications Catalog or contact IMF Publication Services.

      You are not logged in and do not have access to this content. Please login or, to subscribe to IMF eLibrary, please click here

      Other Resources Citing This Publication