Printer Friendly

FEERs and the ERM.

A number of economists, including the author, were critical of the central rate that was chosen when sterling entered the ERM in October 1990, on the ground that it overvalued the pound. Specifically, the central rate against the other ERM currencies implied a higher value for the pound than that yielded by calculations of fundamental equilibrium exchange rates' (FEERs).

The present paper aims to explain the concept of the FEER, introduced by the author in Williamson (1983), and argues that it provides the right criterion for assessing whether a currency is correctly valued. It also sketches the evidence for believing the pound's ERM central rate to be above the FEER. A final section considers the policy implications of the finding that sterling is overvalued.

The concept of the FEER

Under the Bretton Woods system, an exchange rate was said to be in fundamental disequilibrium' if its maintenance was inconsistent with the medium-run achievement of macroeconomic balance, meaning both internal and external balance. By analogy, the fundamental equilibrium exchange rate is defined as the exchange rate that is consistent with macroeconomic balance in the medium run. Since the nominal exchange rate consistent with macro balance will change over time when inflation at home differs from that abroad, the FEER is defined as the real exchange rate consistent with macro balance. And since changes in bilateral exchange rates will leave the balance of payments unchanged as long as the effective exchange rate is constant, the FEER is defined in terms of the real effective exchange rate.

In order to estimate a FEER, one needs three things: an interpretation of what is meant by internal balance, an interpretation of what is meant by external balance, and a macroeconometric model which can be solved to yield the exchange rate (trajectory) consistent with the medium-run achievement, and subsequent maintenance, of simultaneous internal and external balance. The part of the macroeconometric model that is critical in this exercise is the trade sector, consisting of equations expressing the dependence of output and the balance of payments on demand and competitiveness (the real exchange rate).

Interpretation of the term 'internal balance' is rather uncontroversial nowadays, given the general acceptance of the notion of a natural rate of unemployment' or NAIRU. The concept clearly means achievement of the highest level of activity consistent with the continued control of inflation. But even that definition leaves some scope for disagreement about how far and how fast inflation should be reduced, resolution of which requires an essentially normative judgment.

Interpretation of what should be meant by 'external balance' is altogether more difficult (and hence leaves substantially more ground for subjective, and partly normative, judgments). The traditional interpretation of external balance as overall balance (a capital flow that finances the current account imbalance with no change in reserves) does not suffice, because different interest rates are consistent with different capital flows and therefore with correspondingly different current accounts and hence exchange rates. One needs a criterion for picking out one of these sets from the others.

A minimum criterion is to require that the current account outcome be sustainable, in the sense that it be part of a path that can be followed without surprises that would make agents wish ex post that they had acted differently. This rules out the possibility of very large current account deficits financed by massive capital inflows attracted by exceptionally high interest rates: such a strategy may be feasible for a time as long as confidence is maintained, but sooner or later confidence will collapse and the strategy will cease to be viable. It may also rule out very large surpluses (though this is much less clear), but, even if that is the case, sustainability alone is unlikely to be sufficient to pin down the current account to a unique value. For that one needs some additional criterion.

One simple and traditional rule of thumb that still seems to have appeal in certain quarters is for all countries to aim at a balanced current account. This is not a sensible objective. Countries can often expect to benefit by exporting or importing capital over a long period of years, when domestic savings are respectively greater or less than domestic investment opportunities evaluated at the world interest rate.

A more appealing approach involves seeking to optimize the rate of capital flow. One cannot, of course, assume that actual capital flows over some short period were optimal: partly because many capital flows are ephemeral, or even reversible, and partly because capital flows may be induced by levels of interest rates that are non-optimal. Nor can one identify an appropriate target by extracting from the balance of payments accounts some sub-aggregate of flows, e.g. those placed in long-term assets. Instead, start with the national income accounts rather than the balance of payments accounts. A familiar identity states that:
 (X-M) = (S-1) (G - T)

Net investment in = Net savings of - Public sector rest of world private sector deficit.

Hence, the optimal capital flow whose obverse provides an appropriate current account target is the difference between the net savings of the private sector and the optimal public sector deficit, both of which should for this purpose be cyclically adjusted to get their values at 'internal balance'. The public sector deficit is derived by posing the intertemporal optimisation problem: given domestic savings behaviour and investment opportunities, what sequence of public sector deficits and implied foreign borrowing would maximize intertemporal welfare?

Critics of this approach observe that the political process does not always deliver budgetary outcomes that approximate optimality, and argue that not much is gained by calculating the FEER that would be consistent with a budgetary outcome that is not going to be realised. Hence they suggest that it is more useful to identify the likely fiscal position, just as one identifies the likely net savings of the private sector, and then treat the residual current balance as the magnitude that needs to be financed by the underlying capital flow. Unfortunately this approach also has a very unattractive feature, in that it may involve calling an unsustainable trajectory the 'equilibrium trajectory' (in circumstances in which fiscal policy and/or the current account are set on a path that is not sustainable in the medium term).

My ongoing effort to estimate FEERs adopts a compromise between these two approaches. I start by taking the second of the two concepts, based on existing and projected fiscal policies in each country, and derive the implied set of current account imbalances when each country is at internal balance. I then ask whether the resulting set of current account positions are individually sustainable, or whether the current account balances of some countries are leading to a debt buildup where the ratio of debt to GNP or exports would not level out (or would level out only at such a high ratio as to jeopardise credit-worthiness) in the medium term.(1) if a country's imbalance is judged unsustainable, it is implicitly required to reduce its fiscal deficit and thus current deficit far enough to achieve sustainability.

A second question also needs to be posed: are the set of projected current account outcomes mutually consistent? That is, do they sum to a figure whose counterpart the rest of the world will be prepared to accept and able to finance? Where the set of plans are judged collectively unsustainable, I assume that surplus countries should reduce their current account targets (and implicitly increase their fiscal deficits) to an equal percentage of GNP. For example, if the target deficits sum to $100 billion and the GNP of surplus countries summed to $10 trillion, then each surplus country would be assigned a surplus equal to 1 per cent of its GNP.

It is worth noting two points about FEERS. First, they contain a normative element, inasmuch as both internal and external balance are to some extent normative constructs (internal balance because of differing views about how far and how fast inflation should be reduced, and external balance because views may differ about what imbalances are safe or optimal). Second, one must expect that in general FEERs will change over time-because of productivity bias' (Balassa 1964) which allows fast-growing countries a real appreciation consistent with continuing payments balance; because countries in imbalance will be accumulating net foreign assets or liabilities, which have to be serviced by changes in the balance of trade; (2) and because of the 'Houthakker-Magee effect' where income elasticities of demand multiplied by growth rates differ across countries (Johnson 1958, Houthakker and Magee 1969). The FEER is therefore to be conceived of as a trajectory, not as a constant level.

The main competitor to the FEER as a method of calculating equilibrium exchange rates is the use of purchasing power parity (PPP) comparisons. This is, for example, the basis of the currency evaluations published by Goldman Sachs (London)(1), which have consistently shown the pound to be undervalued in the ERM. The weak version of PPP, which uses the criterion simply to compensate for differential inflation, is unobjectionable. But the strong version, which claims that the equilibrium exchange rate is that which equalises purchasing power in different countries, (4) is another matter altogether.

This is because there is in general no reason to think that the PPP exchange rate will be consistent with macroeconomic balance. To see this, suppose that by accident PPP did happen to be consistent with macro balance. Suppose that the country in question happened to be an oil importer, and that it was then confronted by a permanent rise in the price of oil. This would create a payments deficit at internal balance and the initial (PPP) real exchange rate, elimination of which would require a more competitive real exchange rate. But the oil price rise would have no effect on the PPP exchange rate, assuming that the oil intensity of production were similar at home and abroad, so that after the oil shock the exchange rate consistent with macro balance would no longer coincide with the PPP rate.

More generally, the equilibrium exchange rate-in the FEER sense of the rate consistent with macro balance-depends on a range of factors that do not influence the PPP rate: the target current account balance, the trade elasticities, the terms of trade, non-price competitiveness, and so on. The only case in which the PPP rate should be expected to coincide with the rate consistent with macro balance is when the trade elasticities are infinite-a case which is definitively rejected by the empirical evidence (Goldstein and Khan 1985).

If the PPP rate is not consistent with macro balance, is there any reason why it might nonetheless be considered an equilibrium rate? The only reason that has been suggested, so far as I am aware, is that over time it might become consistent with macro balance, because low (high) unit labour costs might attract (repel) investment in tradeable goods industries, perhaps particularly by foreign direct investors (Wren-Lewis 1991). This suggestive hypothesis has been investigated by Peter Hooper of the International Finance Section of the Federal Reserve Board, but he has failed to find any empirical support for it.

In order to explore just how misleading a guide the PPP criterion provides, Philip Bagnoli ran a simulation on the trade equations of the INTERMOD macroeconometric model in which the exchange rates of all the G7 currencies were placed at their PPP values as estimated by Goldman Sachs in the June 1991 issue of The International Economics Analyst Output in each country was assumed to remain unchanged, to simulate the assumption of internal balance. The model estimated that under PPP the current account deficit of the United States would increase by 1.3 per cent of GNP by the fifth year, while the current surpluses of Japan and Germany would increase by 5.9 per cent and 4.8 per cent of GNP respectively. Given that the United States was already expected to be in deficit and Japan and Germany in surplus in 1996, such increased imbalances are surely inconsistent with any reasonable interpretation of external balance: indeed, their realisation would seem certain to provoke a protectionist backlash in the United States. Thus, unless one is prepared to dismiss the global macroeconometric models completely, one must dismiss the PPP criterion-not just as conceptually flawed, but as not even providing a useful first approximation to equilibrium exchange rates.

The sterling FEER versus the ERM central rate In the latest (as yet unpublished) version of my ongoing attempt to estimate the FEERs of the G7 currencies, I use the IMF's estimates of recent output gaps to generate estimates of internal balance.

So far as external balance is concerned, I argue that both Canada and the United States can reasonably aim for moderate current account deficits in the next few years (primarily on demographic grounds), that Japan and Germany(5) are natural surplus countries, and that current account balance is a sensible objective for Britain, France and Italy. I set similar objectives for other areas of the world. I limit current account deficit targets to a maximum of 1.5 per cent of GNP (or 1 per cent in the case of the United States) on the grounds that higher levels may not be sustainable: this figure may be on the conservative side by .5 or even 1 per cent of GNP, but the evidence of the debt crisis shows that the costs of forfeiting creditworthiness can be devastating, so it is prudent to err on the side of caution. I then sum the total deficits available, deduct an allowance for the world current account discrepancy, and distribute the available surpluses in proportion to the projected GNPs of the surplus countries. It happens that each surplus country gets assigned a target surplus of 1.5 per cent of GNP, which by pure coincidence is exactly equal to the maximum target for deficits. This gives me my pattern of external balance targets for 1995.

The proprietors of six global macroeconometric models were kind enough to run simulations on their models designed to estimate the exchange-rate trajectories that would approximately achieve these internal and external balance targets by 1995 (and maintain them thereafter). Given that most of the models are set up to trace the impact of changes in fiscal and monetary policy on variables like output and the balance of payments, rather than to find the implications for intermediate variables like exchange rates of differences in targets for what the models treat as endogenous, this is a technically non-trivial exercise. As one might expect, the models yielded fairly different results, so it was necessary to find some way of reducing them to a single set of estimates. I decided to pick the results from one of the models, rather than to average the results from them all, on two grounds: that an averaging procedure could produce results that would be dismissed as inconsistent by all the models, and that choosing an average would make it difficult to calculate variants on the initial figures or to update the estimates.

I used three criteria in selecting one of the models in preference to the others. The first was that the chosen model should produce estimates that were typical of the group of models as a whole. The second criterion was that it should be a priori well-suited to the task of estimating FEERS, e.g. by modelling all the G7 currencies or by avoiding dependence of the results on the quirks of a sophisticated dynamic structure (given that the question of interest concerns quasi steady states). The third criterion was that it should be easy to calculate new FEER estimates by varying or updating the assumptions. On all three criteria it was clear that the most suitable model was Steady State GEM, i.e. the steady state version of the Global Economic Model used and largely developed at the National institute of Economic and Social Research (see Barrell and Wren-Lewis, 1989, for a description).

Steady State GEM yielded a set of estimates of misalignments-i.e., deviations of the FEER from the actual real effective exchange rate-in the first quarter of 1990. For the European currencies, the D-Mark was estimated to be 15 per cent undervalued, the French franc to be 2 per cent overvalued, the Italian lira to be 6 per cent overvalued, and the pound sterling to be 11 per cent overvalued. These misalignments of real effective exchange rates implied a set of misalignments of real bilateral exchange rates and, making the slightly uncomfortable assumption that correction of misalignments would not provoke any offsetting changes in domestic price levels, in nominal bilateral exchange rates. Or, under the same assumption, the set of FEERs implies a set of equilibrium bilateral nominal exchange rates. Those for the intra-European exchange rates are shown above the diagonal in table 1. The corresponding ERM central rates are shown below the diagonal.

As can be seen by comparing the two halves of the table, only the central rate of the French franc against the lira appears to be close to equilibrium. The D-Mark is undervalued against all the other currencies, and the pound is overvalued against them all. Given the general D-Mark undervaluation, however, the size of sterling's overvaluation is much exaggerated by the comparison between the estimated equilibrium of DM 2.24 shown in the table and sterling's ERM central rate of DM 2.95.

The erosion (or disappearance) of the German surplus as a result of German reunification has led some analysts to challenge the proposition that the D-Mark is still undervalued (see, for example, Stephen Marris in Bergsten, 1991 forthcoming). In contrast, my own view is that the case for an upward realignment of the D-Mark within the ERM was strengthened rather than weakened by German reunification, because reunification made the German government willing to undertake the substantial fiscal expansion that was the necessary complement to revaluation. This expansion has created a need for high interest rates in Germany in order to keep inflation at bay- a need that is much greater than in other ERM countries, where the expansionary impulse from increased spending in Eastern Germany is much weaker than it is in Western Germany. Since the high German interest rates limit the ability to cut interest rates elsewhere in the ERM, the effect of German fiscal expansion has been to export recession to the rest of the ERM. The solution is to revalue the D-Mark within the ERM, thus shifting demand from Western Germany to other countries, alleviating both the inflationary pressure in the former and the deflationary pressure in the latter.

Given the uncertainty surrounding the D-Mark FEER following German unification it seems to me that the best estimate of sterling's overvaluation visa-vis the ERM is not given by comparing the D-Mark rates in table 1, but rather by comparing the estimated equilibrium with the pound's central rate against the French franc. This comparison indicates an overvaluation of the order of 10 per cent.

In another study, which was completed prior to the entry of sterling to the ERM, Wren-Lewis et al (1990) sought to estimate the exchange rate at which sterling should join. This study used the NIESR model of the UK economy. It sought to evaluate the sterling FEER, allowing (as my own work has not) for possible interdependence between the NAIRU and the FEER.(6) The study concluded that at the end of 1989 sterling was between 5 and 10 per cent overvalued relative to the FEER. On the implicit assumption that bilateral exchange rates among all currencies other than the pound were in equilibrium and the explicit assumption that the British authorities would be prepared to devalue the pound in the early years after entry until British inflation could be reduced to that prevailing elsewhere in the ERM, this was translated into an optimal entry rate of about DM 2.60 = L1 (with a range between DM 2.50 and DM 2.70). If the pound were to enter as a hard currency that would not be devalued, Wren-Lewis eta/estimated that it would be necessary to enter at a more competitive rate, say DM 2.40 = 21, in order to end up at the FEER after the transitional period.

The accompanying note by Barrell and in't Veld now concludes that the pound is overvalued by about 5-10 per cent in the ERM. This conclusion is consistent with current forecasts of the outlook for the British economy, which seem to agree on the probability of a continuing substantial current account deficit despite the depth of the present recession and the expectation of slower recovery than elsewhere. The sort of evidence presented to challenge the conclusion that the pound is overvalued consists of the PPP comparisons that were dismissed above and simpleminded comparisons of the present real exchange rate with its historical average, which ignore the uncomfortable evidence that, apart from the boost given by North Sea oil coming on stream a decade ago, there has been a long-run tendency for the pound's FEER to depreciate.

Policy implications

Suppose one accepts the evidence that the pound's central rate is overvalued by 10 per cent or so. Does this necessarily imply that the pound should be realigned downward by about 10 per cent within the ERM on the first occasion when this is politically feasible?

Two arguments can be made against such a realignment. One is that realignments threaten the integrity of the ERM, by reviving the opportunities for making speculative profits out of exchange-rate changes, by reminding the public that the ERM is not a monetary union, and by delaying the convergence of interest rates and inflation.

There are elements of truth in these charges, but they do not add up to a very compelling case. Speculators can take a free ride at the expense of the central bank only when the parity change is wider than the width of the band. Fortunately the fact that the government chose wide 6 per cent margins (i.e., a 12 per cent band) gives adequate scope to engineer a realignment of the size that the FEER calculations suggest to be needed without obliging the Bank of England to make presents to speculators.

Second, the ERM is not a monetary union, as the Delors Report emphasised it would not be during Stage 1 of the move to EMU, for the very good reason that realignments might still be needed until convergence had gone far enough to justify a move on to later stages. Unfortunately debate on convergence has focused only on the need for convergence of inflation and fiscal policies, and neglected the equally crucial need for convergence in competitive positions. A successful realignment, i.e. one whose real effects were not significantly eroded by induced inflation, could help to achieve a convergence of competitive positions.

Neither is it clear that the effects of realignments on interest rates need be pernicious. A devaluation that was expected can permit a reduction in interest rates, since the interest rate need no longer include a devalation premium once the devaluation has occurred. Moreover, an expectation that devaluation will not occur can actually make it harder to combat inflation (by attracting a capital inflow that presses interest rates downwards), as Alan Walters (1989) argued would occur and as has happened in Spain.

What cannot be disputed is that a realignment would threaten anti-inflation policy. This issue is so important that it constitutes the second argument against a realignment, quite independent of any impact that realignments may have on the integrity of the EMS. The fact is that British membership in the ERM was sold to the British public largely as a way of making a credible anti-inflation commitment that would reduce the cost of squeezing inflation out of the system, and it is now being claimed that there are already some signs that wage inflation is indeed falling faster than might have been expected on the basis of previous experience. A realignment that rekindled expectations that wage-sefters would always get bailed out by devaluation could do more harm than good.

The issue is not just whether the devaluation would be offset by faster inflation; clearly there is no point in devaluing unless one expects to retain most of the competitive gain, but that is not a sufficient condition to justify a devaluation. The key question is what devaluation would do to the credibility of the commitment to reduce inflation.

The decision to enter the ERM at an exchange rate above the FEER was singularly unfortunate. it means that the British economy is confronted not just by the task of reducing inflation to the German rate, but of going sufficiently below the German rate for several years to compensate for the initial overvaluation (plus the ground that is still being lost this year and the additional ground that, barring a miracle, will again be lost next year). This is a far more onerous task than that achieved by France over the past decade, since the franc was if anything undervalued rather than overvalued when France first committed itself to stability-oriented policies in 1983. Nonetheless, the price paid by France to achieve virtual price stability was half a decade of growth below the growth rate of productive potential, involving a cumulative growth shortfall of over 5 per cent of GNP.

Presumably Britain faces an even more costly adjustment. In other words, the cost of avoiding realignments is likely to be little growth until the late 1990s.

While that cost could surely have been much reduced by a more prudent choice of entry rate to the ERM, it does not necessarily follow that it can be reduced by devaluation. Bygones are bygones. The key issue is whether it would be possible to design an anti-inflation strategy that could be introduced simultaneously with a realignment so as to maintain credible expectations of inflation deceleration. if that proves beyond the wit of policy-makers, the next best may be to avoid realignments and accept the costs of convergence with grim determination. The worst policy of all is surely to start off with portentous declarations of how unthinkable a realignment would be and then to cave in when the going gets rough.

I have concentrated in this note on the implications of FEERs for the transitional phase before the EMS solidifies into a monetary union, since this is the period that poses the major challenge to policy. I agree with the argument of Barrell and in't Veld in the accompanying note that maintaining exchange rates near FEERs within a monetary union will require continuing modest inflation differentials, with most or all of the other large European countries needing to keep their inflation rates lower than in Germany. I also agree that this should not be regarded as impossible or even costly in the long run. The problem is a transitional one, of achieving convergence in competitiveness as well as inflation. But to recognise that a problem is transitional is not to belittle it. Making the transition seems likely to dominate macroeconomic policy in Britain for the rest of the decade.

Balassa, Bela (1964), The purchasing power parity doctrine: a reappraisal', Journal of Political Economy, vol.72, no. 6, December.

Barrell, Ray, and Simon Wren-Lewis (1989), Fundamental equilibrium exchange rates for the G7', CEPR Discussion Paper no. 323.

Bergsten, C. Fred, ed. (1991 forthcoming), International Adjustment and Finance: Lessons of 1985-1990 (Washington: Institute for International Economics).

Goldstein, Morris, and Mohsin S. Khan (1985), 'Income and price effects in foreign trade', in R.W. Jones and P.B. Kenen, eds., Handbook of International Economics, vol. II (Amsterdam: North-Holland).

Houthakker, Hendrik S., and Stephen P. Magee (1969), 'Income and price elasticities in world trade', Review of Economics and Statistics, May.

Johnson, Harry G. (1958), 'Increasing productivity, income-price trends, and the trade balance', Economic Journal, September.

Krugman, Paul (1985), is the strong dollar sustainable?', in The US Dollar: Prospects and Policy Options (Federal Reserve Bank of Kansas City).

Marris, Stephen N. (1985), Deficits and the Dollar (Washington: Institute for international Economics, revised edition 1987).

Walters, Alan (1989), A critical view of the EMS', in J.A. Dorn and W. A. Niskanen, eds., Dollars, Deficits, and Trade (Washington: Cato Institute).

Williamson, John (1 983), The Exchange Rate System (Washington: Institute for International Economics, revised edition 1985).

Wren-Lewis, Simon (1 991), 'On the analytical foundations of the fundamental equilibrium exchange rate', mimeo.

Wren-Lewis, Simon, Peter Westaway, Soterios Soteri, and Ray Barrell (1990), Choosing the rate: an analysis of the optimum level of entry for sterling into the ERM', National Institute Discussion Paper no. 171.


See Krugman (1985) or Marris (1987) for the type of analysis required. In an interesting recent paper, Simon Wren-Lewis (1991) points out that this implies that the FEER is not independent of the path taken to achieve equilibrium. Slower adjustment by a deficit country will imply a larger transitional deficit, a larger debt that needs to be serviced, and hence a more competitive FEER. See their publication The International Economics Analyst.

Thus the PPP exchange rate is L1 = DM 3.14 if L1 will buy the same bundle of goods in Britain as DM 3.14 buys in Germany. (This is the latest Goldman Sachs figure, from the June 1991 issue of The International Economics Analyst, table 1.)

This analysis was undertaken prior to German reunification being factored into the macroeconometric models. Such interdependence arises if higher real wages, which are possible with a less competitive exchange rate, increase labour supply and thus diminish the NAIRU.
COPYRIGHT 1991 National Institute of Economic and Social Research
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1991 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:fundamental equilibrium exchange rate; exchange rate mechanism
Author:Williamson, John
Publication:National Institute Economic Review
Date:Aug 1, 1991
Previous Article:The world economy.
Next Article:FEERs and the path to EMU.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters