Speech Exchange Rate Regimes for Emerging Markets

At the outset, I should explain what motivates my interest in this topic. It goes without saying that none of what I say today implies any dissatisfaction with the Australian exchange rate regime. The float has worked well in Australia and there is no reason to reassess or reconsider. Today's talk is intended to be part of a vigorous debate currently going on in the international community about which exchange rate regime should (and, more specifically, should not) be adopted in emerging countries, with this particularly aimed at countries which are undergoing financial deregulation. These countries are being exhorted to adopt exchange rate regimes at the ends of the spectrum – either a pure free floating rate, or a strongly fixed rate (preferably with institutional backing in the form of a currency union or even dollarisation).

One argument with superficial attractiveness is that the old regimes failed these countries, and therefore something else – preferably very different from the old regime – will solve the problem. A principal purpose of this paper is to argue that the appropriate regime will differ between countries and perhaps also over time – there are no simple rules and no easy solutions. Whatever regime is chosen, it has pluses and minuses, and any regime will require careful nurturing and support from other policies. The crisis countries should learn from their experience, but should not assume that, because one regime failed, another will be trouble-free. It is certainly true that the countries at the centre of the Asian crisis – Thailand, Indonesia and Korea – had rates that were de facto soft fixes – a high degree of stability in the good times, but not any serious institutional defences when they came under pressure. One clear lesson is that these exchange rate regimes did not perform well in the crisis. Some have also noted that Hong Kong and Argentina, with hard pegs, were able to withstand the pressure, in the sense that their exchange rate regimes did not break down. Others have observed that countries such as China and India were able to maintain parities by virtue of the application of capital controls, so an equally vigorous debate on the role of capital controls is intertwined with the debate on regimes. This paper attempts to draw together this diverse experience.

What Have We Learnt from the Recent Crisis?

As a preface to this section, I should note that, as usual, the crisis has been used by various people to advance their own theories on the way the world works and the way it should work. As with many debates, the coalitions which form to advocate a particular outcome often have diverse motives, and not everyone comes from the same starting point or uses the same analytical framework. Any sensible observer of the crisis understands that there were a variety of causes, and identifying one single cause is unlikely to move us to a world of perfection. With this preamble, let us look at the immediate and obvious lessons for exchange rates from the crisis.

  • Exchange rates overshoot, resulting in substantial and sustained shifts of real exchange rates not explicable in terms of fundamentals. Indonesia entered the crisis with its exchange rate only modestly overvalued, but the rate fell to one-fifth (!) of its pre-crisis value. Markets are prone to exhibiting a herd mentality and, once an exchange rate starts to move, or a peg is broken, it can be difficult to limit adjustments to modest amounts.
  • Contagion is a problem. While markets did discriminate among countries to a degree, this was more noticeable in the longer run than in the short run.
  • There has been a lack of players willing to take contrarian foreign exchange positions in newly-emerging countries. This is a self-reinforcing problem. There are few stabilising speculators so the exchange rate moves a long way; but because the rate moves a long way, there are few who are willing to take on the risk of stabilising speculation. This lack of contrarian position-taking might reflect some relative lack of information about these economies, which should be straightforward to remedy. But more often the problem is more deep-seated and the remedy more difficult. Part of the problem here is that the market's view of exchange rate fundamentals (‘the model’) is nebulous and fragile. As well, there may be a lack of policy credibility, which can only be redressed over time.
  • The transition from regulated to liberalised financial systems is a vulnerable period. Careful sequencing may help, but it is difficult to create the necessary infrastructure quickly.
  • Massive capital flows (inflows prior to the crisis, and outflows after) were a dominant factor. For example, Thailand experienced inflows equal to 13 per cent of GDP in 1996. Choice of regime needs to take into consideration not only current account issues but also the greatly increased mobility of capital and the international integration of financial markets.

Choice of Regime

Some high-profile commentators drew from the Asian Crisis experience the conclusion that countries should opt for either a very ‘hard’ fix or a very free float. ‘Thus the new language of speculative attacks, multiple equilibria and moral hazard is in many ways simply a new overlay on an old debate. And yet, a genuinely new element has recently been thrown into the mix. This is the proposition that countries are – or should be – moving to the corner solutions. They are said to be opting either, on the one hand, for full flexibility, or, on the other hand, for rigid institutional commitments to fixed exchanges, in the form of currency boards or full monetary union with the dollar or euro. It is said that the intermediate exchange rate regimes are no longer feasible – the target zones, crawls, basket pegs, and pegs-adjustable-under-an-implicit- escape-clause – are going the way of the dinosaur' (Frankel, Schmukler and Servén 2000). The Economist (1999, pp 15-16) put the same point more briefly: ‘Most academics now believe that only radical solutions will work: either currencies must float freely, or they must be tightly tied (through a currency board or, even better, currency union)’.

The rationale is rather different from the traditional textbook arguments about the choice of exchange rate regime.[1] The new argument is a response to the volatility of capital flows and the threat of self-fulfilling speculative attacks: both strong fixes and free floats are immune from attack, at least in a definitional sense. At the fixed end, the emphasis is on credibility, so irrevocability is an important element of the fix – hence the interest in currency boards and the ultimate fix – dollarisation. At the other end of the spectrum, the pure float also has the attraction of its ability to withstand an attack (or large volatile capital flows) without the exchange rate regime collapsing.

Considering the depth and sophistication of traditional arguments, pro and con, for a wide range of different regimes, the new slant on this seems at first sight rather superficial, not far from a truism (if an attack on a fixed rate succeeds, the ‘fix’ wasn't strong enough:[2] and an attack on a floating rate leaves it immune only to the extent that the authorities can put up with whatever movement in the rate occurs). If this new slant on choice of regime is to make a useful contribution to the debate, it has to be seen as another layer on top of the other (still valid) arguments about regimes and has to be seen to address, more specifically, why the perceived advantages of intermediate regimes (‘the middle of the spectrum’) are overridden.

One flawed argument in support of the ‘disappearing middle’ of the spectrum goes under the shorthand of the ‘Impossible Trinity’: it is not possible to have, simultaneously, an open capital market, a fixed exchange rate, and an independent monetary policy. While the trinity may indeed be impossible, it relates only to a fixed rate and therefore says nothing about the possibility of having intermediate positions – a flexible but managed rate, a reasonable degree of monetary independence, and an open capital account. For that matter, the criticisms arising from the Asian Crisis could be properly directed only at the ‘fixed but adjustable’ segment of the spectrum, where reserves were spent in futile defence of a rate which markets no longer believed was viable. Even here, the argument is far from clear-cut: Indonesia did not defend its exchange rate when it came under pressure, and yet suffered the greatest fall (and the most serious consequences from this fall).

The valid elements of the case against a ‘fixed but adjustable’ regime are two-fold:

  • that the fixed element may have encouraged borrowers to be too confident in taking out foreign-exchange-denominated loans;
  • that when pressure came on the exchange rate, it created the same incentives which drive bank runs – the one-way bet which makes rational investors close out their position ahead of a possible collapse, and creates the incentive on the part of the authorities to defend the status quo to avoid triggering the run.

The prescription seems easy, but these ‘ends of the spectrum’ may not be comfortable or easy to maintain. Only a limited number of countries will find the policy constraints of a currency board or the loss of sovereignty involved in dollarisation politically palatable. Smaller countries with a suitable pegging partner may find this an acceptable and viable solution, but – leaving aside the special case of the Euro – medium-sized countries will only be attracted to a hard fix if there are very compelling constraints on their choice, and they will find the degree of adaptation needed in the domestic economy can be painful indeed. Where relative prices cannot change through exchange rate adjustment, domestic prices must adapt.[3] Even super-flexible Hong Kong has found this hard, with GDP falls as large as Korea's during the Asian crisis. Argentina's choice of a strong fix may well be correct policy, but it is clearly painful.[4]

If the fixed end of the spectrum does not facilitate the sorts of adjustment which the underlying fundamental may require, then the free float end of the spectrum may well deliver much more variability than the fundamentals require.

Views on how a free float would work in practice have changed over the past thirty years, in the light of experience. Early protagonists were confident that floating exchange rates would be stable and would reflect fundamentals. Harry Johnson (1972, p 208) assured us that: ‘A freely flexible exchange rate would tend to remain constant so long as underlying economic conditions (including government policies) remain constant; random deviations from the equilibrium level would be limited by the activities of private speculators’. Friedman (1953) believed: ‘… instability of exchange rates is a symptom of instability in the underlying economic structure … a flexible exchange rate need not be an unstable exchange rate. If it is, it is primarily because there is underlying instability in the economic conditions …’.

In practice, the general experience might be summed up this way:

  • nominal exchange rates are, unaccountably, closely correlated with real exchange rates;[5]
  • the switch from fixed to floating rates has produced much more variability (‘an order of magnitude more’), even when fundamentals are not more variable (Flood and Rose 1999);
  • fundamentals cannot explain the behaviour of the exchange rate over a short/medium-term horizon;[6]
  • exchange rates have at times exhibited long-lived swings, with no apparent changes in fundamentals significant enough to justify them. The US dollar appreciated by about 90 per cent against the Deutschemark in the first half of the 1980s, only to completely unwind this appreciation by 1988. The yen appreciated by about 75 per cent against the US dollar in the first half of the 1990s, and unwound this appreciation by 1998.

As a result of this experience, a textbook free float with interest rates set exclusively for the needs of the domestic economy (i.e. ‘benign neglect’) is out of favour. The exchange rate is routinely a factor in setting interest rates. Many countries intervene (if only occasionally) in foreign exchange markets, and when this is done with finesse and sensibly (i.e. not to defend a particular exchange rate), it has often been helpful.[7]

The Intrinsic Problems of Exchange Rates in Emerging Countries

If all this produces volatility and misalignment in developed countries' exchange rates, how much more serious is it for emerging countries which have:

  • much less well-defined trade-based fundamentals;
  • no long historical experience of market-determined exchange rates;
  • rapidly evolving production structures;
  • few Friedmanite stabilising speculators;
  • much larger and more volatile capital flows, in relation to the size of their domestic capital markets and economies more generally.

On top of this, we have the evidence from Hausmann, Panizza and Stein (1999) that those Latin American countries which embraced floating during the 1990s experienced substantially larger changes in interest rates than fixed-rate countries (i.e. they were not, in fact, able to direct monetary policy to the needs of the domestic economy, but had to use it to defend their exchange rates).

But the main problem faced by emerging countries which chose a free float is how the rate would behave in the face of the large and volatile capital flows which these countries experienced in the 1990s. In due course, these flows seem likely to resume for many, if not all, of the countries which experienced them in the 1990s. They are driven by two powerful factors. First, the desire for diversification on the part of the managers of the huge stock of investment funds in America and Europe, whose weight of foreign assets is well below the theoretical desideratum. Secondly, by the intrinsically high profit prospects in the Asian emerging countries (more on this later).

It is worthwhile spending some time to see just how difficult a problem this is. In standard versions of the exchange rate story, the real exchange rate is determined by productivity developments in the real economy, and capital flows tend to play a rather secondary role. They are often treated, essentially, as a residual. Implicitly, there is a ready supply of world capital, so that the current account is determined by a country's saving/investment balance, and the capital account is a residual to fund this. There is also a presumption that there is a ready supply of stabilising speculators, so any significant departures from fundamentals will be ironed out promptly. The standard model for incorporating capital flows into the analysis is the portfolio balance view, where the main action is with interest differentials. With some interest differential in place (usually reflecting differing cyclical positions), enough capital flows to the country to push up the exchange rate so that expected returns are equalised internationally (risk-adjusted, of course) by the prospect of a subsequent reversal of the exchange rate. The higher exchange rate helps to open up a current account deficit, which provides the real transfer counterpart of the financial flows.

But if we try to apply this to the emerging countries, the fit is not good. Perhaps most fundamentally, the countries which received huge capital inflows in the first half of the 1990s offered high interest rates (real and nominal), not as a temporary cyclical policy response to the phase of the cycle, but on a continuing basis. While these countries are making the transition towards the technological frontier, it is quite likely that higher returns will be available to capital, so a real interest rate differential will persist over the medium term – decades rather than years. Capital inflow cannot immediately reduce this interest differential. In the meantime, portfolio equilibrium could, in theory, be maintained by the real exchange rate being bid up, so that the higher domestic interest rate is balanced by the prospect of subsequent depreciation.[8],[9]

How should the exchange rate behave in these circumstances? Real interest differentials of, say, 3 per cent might persist for a decade or more. If these numbers are realistic, the portfolio balance model would suggest that the exchange rate has to appreciate initially by some 30 per cent (and will appear seriously and persistently uncompetitive for trade in goods and services), before depreciating by 3 per cent per year over the following decade. So the potential swings in real exchange rates, even if well-behaved in terms of the model, are much greater for emerging markets. Add to this some extrapolative expectations, some herding, and above all a risk premium which varies with the latest wave of euphoria or pessimism, and the potential both for volatility, and for significant and sustained misalignment – in both directions – is clear. It seems unlikely, to say the least, that an exchange rate could follow this portfolio-balance path of appreciation followed by depreciation without the market baulking at the large shifts and the overvaluation during the long transition.

Is it not surprising, in such a world, that countries have sought to limit the extent of the swings (misalignments). They have sought to resist the appreciations of the exchange rate during the periods of large capital inflow partly because of an intuition (which gains some support from the Japanese experiences in the 1950s and 1960s) that strong international competitiveness provided beneficial price signals for the most dynamic sector of the economy – tradeables, especially exports. Elements of old-fashioned mercantalism may be present also.

Perhaps more important still, the large foreign capital inflow creates a destabilising feedback loop – while the inflow is strong, the rising exchange rate reduces the cost of borrowing in foreign currency, and encourages more borrowing. At the first sign of some weakness in inflow or the exchange rate, domestic borrowers (including, sometimes, banks) are exposed to greater credit risk, and their lenders (understandably) will want to withdraw funding. This has many of the characteristics of a domestic bank run. If confidence could be maintained (in this case, in the stability of the exchange rate), then the withdrawal of funds will not occur. If confidence weakens, the process is self-reinforcing.

This self-reinforcing instability cannot be easily removed. Some have suggested hedging as the answer. There is, first, the question of whether the country as a whole can hedge its foreign exchange risk. While an individual can shift the risk to another party, a country in aggregate can only hedge its risk if it can persuade foreigners to take on the foreign exchange risk – in effect, lend in the domestic currency. Even if this could be achieved, one party in foreign capital flow transactions (either the borrower or the lender) will be exposed to foreign-currency risk, and will have the incentive to unwind the transaction in the face of a threat to the exchange rate. In this rather unstable world, it is hardly surprising that countries have resisted the upward pressure on exchange rates that accompanies big capital inflows, because they feel (with some justification) that it makes them vulnerable to later sharp depreciation.

When, after the crisis, many commentators attribute the crisis to overvalued exchange rates, and others urge these countries to hold reserves equal to their short-term debt (the so-called Guidotti Rule), is it surprising that authorities intuitively feel justified in resisting appreciations? But if the correct lessons are to be taken from the crisis, then those countries which do not have strong reasons to go to a hard fix should move decisively away from ‘fixed-but-adjustable’, to adopt a version of what has become widely used best practice: allowing the market to determine the rate, leaving considerable volatility (i.e. not leaning against the wind through frequent intervention), and even when the rate moves significantly away from what are judged to be the ‘fundamentals’, being prepared to put up with this, and stay the resort to intervention, until the case is overwhelming and the prospect of success good. Interest rate changes in support of the exchange rate are a legitimate part of this regime, while acknowledging that feasible interest rate settings may not be enough to counter extreme market pessimism. This is clearly distinct from the totally passive position of a free float (‘benign neglect’), but it is what is generally done in most developed countries. And, most importantly, it is a long way from the day-by-day activism that has been discredited by the crisis.

Unless we take into account why countries have a ‘fear of floating’, and have often gravitated to soft fixes in the face of capital inflows, we will be in no position to provide persuasive arguments for the adoption of greater flexibility.

Perhaps the basis of persuasion might be three-fold:

  • that some flexibility will be helpful in absorbing the capital inflow, in buffering external shocks, and responding to the changing productive capacity of their economies;
  • that this flexibility (a.k.a. short-term volatility) may inhibit some short-term flows, by serving as a constant reminder that exchange rate volatility can outweigh the interest-rate advantage of foreign-currency borrowings; [10]
  • allowing (even encouraging) a fair degree of volatility around a real exchange rate which is stable over time (or moves only slowly, in response to changing fundamentals) provides the opportunity for the authorities to have the best of flexibility, while leaving open the possibility of intervention (both via interest rates and directly in foreign exchange markets) when the rate has already moved quite some distance away from the fundamentals. There are obviously difficult practical issues regarding the operation of intervention, but the intuitive idea is straightforward enough – the further the actual exchange rate has departed from the equilibrium, the more damage the misalignment will do; the more confident the authorities can be that they will be acting as profitable stabilising speculators (buying cheap and selling dear); and the greater likelihood of success of any intervention on the part of the authorities (see Volcker (1995)).

In short, there is scope for emerging economies to operate flexible exchange rate regimes without them having to adopt a textbook type of pure float. Indeed, it would be odd if the international debate urged emerging markets to adopt more pure forms of floating than the industrial countries have been able to sustain, especially when the conditions necessary for a successful pure float are less likely to be present in such economies. Even this strategy will be uncomfortable, and is not guaranteed to succeed. Large reserves and uncomfortable interest rates might be needed to mount an effective defence, and it will be tempting to mount this defence too early (not to allow enough flex).

If a concrete example is needed to illustrate the point, Singapore provides it. It was, arguably, the most successful country exposed to the full force of the crisis, with a managed exchange rate (managed both in the sense of intervention and through well-designed active capital account policies) (see Lee Hsien Loong (2000)).

Back-up or Support for the Exchange Rate Regime

What more could be done by emerging markets to enhance the possibility of a successful exchange rate regime? If we accept that one of the central factors presenting pressures on exchange rate regimes in emerging markets is large and volatile capital flows, then the obvious issue is whether measures can be taken to reduce the size and volatility of the flows.

The first possibility here is capital controls. Mere mention of this possibility creates deep rifts in the debate, with views strongly held and a fair bit of heat generated in the process. The process of transition (financial deregulation) is tricky and time-consuming. Perhaps the common-sense way to approach this is to accept the possibility that Chilean-style controls (taxes on short-term inflows) may be useful for some countries during the transition, but not too much should be expected of them (see the conclusions on Chile itself, which suggest that the controls managed to lengthen the maturity of the debt, without being able to prevent the exchange rate from appreciating during the phase of capital inflow) (see Edwards (1998)).[11] It may also be useful to adopt the techniques used successfully by Singapore (and now others) to limit the ability of domestic banks to lend to foreigners in domestic currency, thus making it much harder for foreign speculators to take large short positions on the currency.

Other than these specific types of capital control, there are a set of issues which can be put in terms of prudential supervision, rather than capital controls. The sorts of prudential controls which might be used are to limit the opportunities for residents to borrow in foreign currency (i.e. to prevent a repeat of the Bangkok International Banking Facility) and to monitor them when they do; and to keep very tight (indeed, unashamedly intrusive) constraints on banks' ability to have open foreign exchange positions or indirect exposure through foreign exchange loans. The brief lesson here is that emerging countries should be very ready to put on tough prudential controls, and should ignore those who claim (as happened in Australia during the 1980s) that such prudential controls are ‘re-regulation by the back door’. The doctrinal protagonists of free markets (usually arguing their own self-interest above all) must not be allowed to intimidate regulatory authorities into ignoring the case for ‘rules of the game’.

On a wider canvas, the emerging countries should press forward with the efforts to put in place best-practice accounting, legal and corporate systems and regulations. The quality of credit and investment standards, and the strength of management of corporate risks and balance sheets will determine how closely capital flows match fundamental investment opportunities. This will determine how effectively a country captures the benefits of access to world capital markets. Paradoxically, marked improvement in these areas may well encourage even greater inflows: but it should, at least, lessen the likelihood of sudden reversals of capital.

A Tentative Operational Classification

Countries might be categorised this way:

  1. Those whose size or market development justify floating (i.e. those who will want to have an independent monetary policy, and have the institutions to make this feasible). While credible monetary and fiscal policy will contribute to exchange rate stability, even these countries will, at times, find that exchange rates move too far relative to fundamentals. There may be a legitimate role for intervention in these circumstances, but it should be used sparingly to have the desired impact upon expectations and the behaviour of market participants. In a flexible rate regime, what is the appropriate nominal anchor to pin down market expectations (e.g. an inflation target)?
  2. Those in transition to this state. While accepting the advantages of flexibility, these countries may want a greater degree of fixity for a time, to help anchor expectations where these are fragile. These countries may reinforce their exchange rate stability by policies that reduce the attractiveness and the potentially harmful effects of short-term, easily reversible capital flows. Countries in these categories might need to focus on systems that encourage long-term financing and on maintaining stable current account deficits. These countries may want to contemplate Chilean-style capital controls (price-based aimed at short-term flows) and to moderate the pace of financial liberalisation, to keep in step with the pace of development of financial infrastructure. Adoption of sound and consistent economic policies may gradually build credibility, but the economy may remain vulnerable during this building process. Financial infrastructure will develop and gain depth over time. As this happens and markets gain knowledge of and confidence in the regime, exchange rates should become less vulnerable and less in need of defensive capital controls. In the meantime, countries should be ready to move quickly away from defence of an unsustainable rate (which obviously requires fine judgment about ‘sustainability’). Allowing exchange rate movement (even if excessive) will, generally, be preferable to exhausting foreign exchange reserves defending a particular level of the exchange rate.
  3. Those countries where the case for a fixed rate is compelling might want to consider the rival merits of some form of strong policy commitment – a currency board or dollarisation. This would include small countries with a similar economic structure to the potential anchor country; countries (including larger countries) with a high degree of economic integration, measured by trade or investment flows; or those with a history of poor economic management and currency crises, where governments cannot borrow long-term domestically, or in domestic currency offshore. For many small, and typically open, countries, the cost of an independent monetary policy will often be greater than the stabilising benefit of a flexible exchange rate. But there are costs, too, in these strongly fixed systems – if lender of last resort is to be provided, it would be a fiscal rather than central bank matter; and dollarisation may imply loss of seigniorage. If a hard fix is elected, what is the most appropriate anchor currency?


The Harry Johnson and Milton Friedman quotes cited above now seem touchingly naïve, or at best outdated advocacy, overtaken by the experience of three decades of floating rates. Flexible rates have turned out to flex more than expected. Developed countries have learned to live successfully with this, but it is harder going for emerging countries with big capital flows. The best-practice regime commonly used in developed countries allows wide latitude for the market to set the rate, supported by occasional intervention and rather more frequent help from interest rates. Many emerging countries could benefit from moving to this generic type of regime, and to help this process, we need a vigorous debate about when to intervene (and, more importantly, when not to); what role interest rates and inflation targets might play; and what additional measures might help to handle large and volatile capital flows. This will take the debate far away from the fixed-but-adjustable regimes that failed during the crisis, towards the floating end of the spectrum.

But no good case is improved by exaggeration: ‘benign neglect’ is not best practice anywhere, and a rhetoric which fuzzes this issue is unhelpful. To truncate the debate by excluding, from the outset, the valid modifications to this ‘end spectrum’ solution runs the risk that countries will not only fuzz the debating points, but fuzz their own understanding of the issues, and leave them unprepared for the coming challenges.

The last word could be given to a veteran of this debate, Richard Cooper (1999, pp 16-17): ‘What is less obvious is that floating rates, independent monetary policy, and freedom of capital movements may also be incompatible, at least for countries with small and poorly developed domestic capital markets, i.e. for most countries. That would leave a more limited menu of choice for such countries: between floating rates with capital account restrictions and some monetary autonomy, or fixed rates free of capital restrictions but with loss of monetary autonomy. Put bluntly, two prescriptions regularly extended to developing countries by the international community, including the IMF and the US Treasury, namely to move toward greater exchange rate flexibility and to liberalize international capital movements, may be in deep tension, even deep contradiction’.


The traditional case for a fixed rate rested on one of two grounds. The first could be broadly termed ‘optimal currency area’ arguments – when there are close trade ties and similarity of productive structure, a fixed rate will encourage beneficial trade and capital flows. The second traditional strand of arguments in favour of fixed rates sees the exchange rate regime as the anchor for monetary policy. The traditional argument in favour of flexible rates was that it freed a country to apply an independent monetary policy, directed at the needs of its own domestic economy. The exchange rate could then be a buffer to soften the impact of external shocks such as changes in the terms of trade. [1]

And Obstfeld and Rogoff (1995) note that few fixes last longer than five years. [2]

Friedman, arguing the case for floating, drew an analogy with daylight-saving time – it is easier to move to summer time than to co-ordinate everyone to move the time of their activities. [3]

Those countries which adopted a fixed exchange rate in order to anchor prices in the face of persistent inflation usually found that this worked well in reducing inflation, but at the cost of loss of international competitiveness, leaving a legacy which had to be addressed by other painful remedies. It is worth noting, also, that even hard fixes are subject to attack: it just takes a different form – a run on domestic banks which drives up interest rates. [4]

See Cooper (1999, p 16). [5]

The classic reference, Meese and Rogoff (1983), showed that existing exchange rate models based on economic fundamentals could not reliably out-predict the na├»ve alternative of a ‘no-change’ forecast for year-to-year changes in major-industrial-country exchange rates. Some more recent models can out-predict a ‘no-change’ forecast (for example, MacDonald and Taylor (1993)) but the basic empirical fact remains largely intact. No-one has yet been able to uncover macroeconomic fundamentals that explain more than a modest fraction of year-to-year changes in industrial-country floating exchange rates. Frankel and Rose (1995, p 1707) summarise the dismal state of exchange rate empirical research: ‘… the case for macroeconomic determinants of exchange rates is in a sorry state. With the exception of some significance in bits of statistical innovation and announcements at very short horizons, and some hazy predictive power at long horizons, there is little support for standard macroeconomic models’. Flood and Rose (1999, p F668) find that the switch to a floating rate produces much more variability, even if the fundamentals are no more variable: ‘The policy switch between fixed and flexible exchange rates entails an essential shift in market structure across regimes … Expectations are policy-dependent’. [6]

It is not the purpose of this paper to address the issue of stability among the G3 currencies, but it might be noted that greater stability would clearly have been very helpful for the other countries. When the yen moved from 80 yen per US dollar in April 1995 to almost 150 in the middle of 1998, this was clearly quite disruptive for countries with liabilities denominated in either US dollar or yen. A case can be made that a trigger or catalyst for some of the crisis has, in fact, been movements in foreign exchange rates or interest rates – with the tightening of US interest rates in early 1994 being a factor in Mexico, and the depreciation of the yen after April 1995 being a factor in causing yen-denominated borrowers to focus on their yen-carry borrowing strategies. [7]

Whether these persistently high rates were a response of the underlying fundamentals, or a period of prolonged euphoria, is not the point. I favour the first explanation, but whatever the reason, domestic borrowers had strong incentives to seek foreign-currency-denominated (i.e. lower interest rate) funding overseas, and foreigners had incentive to provide funds. [8]

Keynes (1980) had identified this problem in 1942: ‘In my view the whole management of the domestic economy depends upon being free to have the appropriate rate of interest without reference to rates prevailing elsewhere in the world. Capital control is the corollary of this’. [9]

The value of this should not be overstated. Much of the foreign borrowing in the Asian countries before 1997 was in currencies other than US dollars (mainly yen – see Table 11 of Goldstein and Hawkins (1998)), and so the volatility of exchange rates would have been abundantly plain to borrowers. This did not seem to inhibit their borrowing. [10]

The other much rarer possible use of capital controls is that, in extremis, a crisis might be sufficiently bad to justify a standstill on repayment of debt, in the context of ‘private sector involvement’ (‘bailing-in the private sector’). [11]


Cooper RN (1999), ‘Exchange Rate Choices’, Harvard Institute of Economic Research Discussion Paper No 1877.

Edwards S (1998), ‘Capital Flows, Real Exchange Rates and Capital Controls: Some Latin American Experience’, NBER Working Paper 6800.

Flood RP and AK Rose (1999), ‘Understanding Exchange Rate Volatility Without the Contrivance of Macroeconomics’, Economic Journal, 109, pp F660-F672.

Frankel JA and AK Rose (1995), ‘Empirical Research on Nominal Exchange Rates’, in G Grossman and K Rogoff (eds), Handbook of International Economics, Vol. III, North-Holland, Amsterdam, pp 1689–1729.

Frankel JA, S Schmukler and L Servén (2000), ‘Verifiability and the Vanishing Intermediate Exchange Rate Regime’, NBER Working Paper 7901.

Friedman M (1953), ‘The case for flexible exchange rates’, in Essays in Positive Economics, University of Chicago Press, Chicago, pp 157–203.

Goldstein M and J Hawkins (1998), ‘The Origin of the Asian Financial Turmoil’, Reserve Bank of Australia Research Discussion Paper No 9805. Available at <URL:http://www.rba.gov.au>.

Hausmann R, U Panizza and E Stein (1999), ‘Why Do Countries Float the Way they Float?’, Inter-American Development Bank Research Department Working Paper No 418. Available at <URL:http://www.iadb.org/OCE/pdf/418.pdf>.

Johnson HG (1972), ‘The Case for Flexible Exchange Rates, 1969’, Further Essays in Monetary Economics, Allen and Unwin, London, pp 198–228.

Keynes JM (1980), ‘Letter to R.F. Harrod, 19 April 1942’, in The Collected Writings of John Maynard Keynes, Volume XXV, Activities 1940–1944 Shaping the Post-war World: The Clearing Union, Macmillan, London, pp 146–151.

Lee HL (2000), ‘Post Crisis Asia – The Way Forward’, William Taylor Memorial Lecture, 21 September. Available at <URL:http://www.mas.gov.sg/newsarchive/sp_20000921-c.html>.

MacDonald R and MP Taylor (1993), ‘The Monetary Approach to the Exchange Rate: Rational Expectations, Long-Run Equilibrium and Forecasting’, IMF Staff Papers, 40, pp 89–107.

Meese RA and K Rogoff (1983), ‘Empirical Exchange Rate Models of the Seventies: Do They Fit Out of Sample?’, Journal of International Economics, 14, pp 3–24.

Obstfeld M and K Rogoff (1995), ‘The Mirage of Fixed Exchange Rates’, NBER Working Paper 5191.

The Economist (1999), ‘Global Finance: Time for a Redesign?’, 30 January, pp 1–18.

Volcker PA (1995), ‘The Quest for Exchange Rate Stability: Realistic or Quixotic’, The 50th Anniversary Stamp Lecture, London University, 29 November. Available at <URL:http://www.iie.com/TESTMONY/volcker.htm>.