Are Intermediate Regimes Vanishing?

Speech given at the international conference on "Exchange Rate Regimes in Emerging Market Economies"
Tokyo, Japan

December 17, 1999



I want to start with one of Paul Krugman's remarks last night, where he classified economists into four quadrants. I, and I think many people here in this conference, fall squarely in the quadrant defined by the twin beliefs that monetary policy matters and that markets cannot always be trusted to get things right. If I believed that markets could always be trusted to get it right, I would be a floater and happy to say "Let's all be like Australia", and be done with it. There is, I admit, a different logic than distrust of the market for wanting to peg the exchange rate, namely the belief that a country needs to have an exchange rate peg as a nominal anchor. I am certainly not persuaded by this thesis: Brazil's experience since it devalued last January is sufficient reason for rejecting it. (Nor, for that matter, am I persuaded that there is a need for any other formal nominal anchor, as opposed to macro discipline of the sort that Keynes preached, but let me not get into that.)

The purpose of exchange rate management, as I see it, is to try to ensure that exchange rates behave something like the textbooks say they do. That is, that they do not get carried off to patently overvalued or undervalued levels, which I believe we see actually occurring in the real world with monotonous frequency.

Given that as the background from which I come to the problem, let me turn to the evidence for this thesis that we have two corner solutions left, and we are driven to one or the other. I think this is overwhelmingly because of the belief that intermediate regimes have proven themselves to be highly susceptible to speculative crises. That was already clear as regards the adjustable peg way back. Almost the first thing I wrote in international monetary economics was my Princeton Essay The Crawling Peg in the mid-1960s. One of the points I made there was that the adjustable peg was not going to be viable indefinitely, because it was bound to be swept away by the increasing capital mobility. And it was really that thought which motivated the search for intermediate solutions, which eventually matured into what Rudi Dornbusch has called the BBC (Band, Basket, Crawl) regime. The idea was that with this regime you could prevent misalignments emerging, and thus avoid the need for discrete changes in exchange rates, and thus preempt market expectations of the discrete changes that create scope for large speculative profits, which would permit avoidance of the speculative pressures that swept away Bretton Woods and later the narrow-band ERM.

Three years ago, when many of us met in Seoul in a previous conference, I would have argued that this type of intermediate regime was functioning reasonably well in a number of countries. Since then things have not gone well for the BBC regimes. Colombia and Chile both abandoned their crawling band systems last September. In Indonesia, a well-established crawling peg with a band that had become reasonably wide got swept away in a single day in the East Asian crisis in August 1997. This demonstrated that even a country which manages its own crawling band in such a way as to avoid giving rise to speculative pressures can still get side-swiped by contagion of the sort that we saw in East Asia. Israel has chosen to gradually widen its band to a point where it has become just about irrelevant. Poland seems to be going the same way. And Russia (I confess I did not realize it at the time) turns out to have made the same mistake that Mexico did, of not crawling fast enough to maintain the center of the band in the vicinity of the equilibrium exchange rate. Instead it tried to use the exchange rate as a nominal anchor to drag inflation down, but that brought about a serious overvaluation that crucified the real economy and precluded recovery.

A year ago Latin America had one currency board and two floating rates among the seven major currencies. That is, three countries employed corner solutions and four were using intermediate regimes. Today there is only one left in the middle, and that is Venezuela with its old-fashioned adjustable peg, the worst of all systems. The ones that seemed to be doing something better-Brazil, Chile, and Colombia-have joined the floating currencies. The region has gone to six out of seven in extreme regimes.

Stanley Fischer has argued a number of times that South Africa, Mexico, and Turkey would all have succumbed to the crisis in the last two years had they not had floating exchange rates. I do not see any strong reason to contest that assessment.

Those are the grounds for the belief that intermediate regimes are vanishing. Do they make me join the rush to the corners? The answer is no, I still have reservations on two sorts of grounds. The first ground is that I am not yet convinced (maybe I will be after ten or twenty years, if I see countries at the corners never experiencing monetary crises) that these corner solutions really are crisis-free. About Hong Kong, we discussed last night the sense in which it did not have a proper currency board in place. But even if there is a proper currency board in place, and it is covering M0 with international reserves, theory does not say that that is necessarily enough. To be certain of avoiding a crisis, it is necessary to cover all of M3, because if the authorities were to stop converting M3 on demand into M0, that would constitute a crisis at least as bad as any you could imagine happening in the foreign exchange market. Hence the logic has to be that, if a run starts under a currency board, there is a self-equilibrating mechanism to stop it: interest rates will rise so as to discourage the run. But that is the same logic as with any other fixed-rate system, and it will work just as long as the public finds it credible. If and when there is a loss of confidence in a currency board, the game will be up for this system too.

Similarly, one can imagine how excessive capital inflows may still be sometimes welcomed under a floating exchange rate. Countries may therefore still get into a vulnerable situation with a lot of short-term foreign debt, unless you believe the story that this happened only because everybody thought there was an implicit guarantee from having a pegged exchange rate. That strikes me as implausible: after all, most of the East Asian currencies actually were not guaranteed, and most of those countries actually described their regimes to International Financial Statistics as managed floating. So I would guess that one day we will see a country with a floating currency get over-exposed and suffer an unexpectedly large depreciation in consequence. That could lead to the sort of crisis that we saw in East Asia.

Hence, I am not persuaded that these corner solutions are necessarily crisis-free. At the same time, I recognize that the intermediate regimes do have a greater vulnerability.

The other ground for my reluctance to join the rush to the corners is that avoiding crises is not the only ground on which one should choose an exchange-rate regime. If one believes that misalignments are costly, then the possibility of avoiding them should also be taken into consideration. And on this ground the intermediate regimes retain an overwhelming advantage. In particular, I find it difficult to believe that any country will be able to sustain an economic miracle of the sort that East Asia achieved in the quarter century before the crisis with a floating exchange rate, given the evidence that a competitive exchange rate was a key foundation of the miracle. Any country that gets its policies in order nowadays quickly finds itself inundated with inflows of foreign capital, which appreciate its exchange rate in a way that undermines the attractiveness of investing in tradable goods industries and thus threatens to abort rapid growth. My hunch is that only a policy of managing the exchange rate so as to keep it competitive, which may involve resisting excessive inflows of capital, offers a chance of reviving and sustaining miracle-level growth rates.

And so the question is, can one conceive of alternative versions of the intermediate regimes, which would be less vulnerable to speculative crises but would nonetheless help to avoid misalignments?

Let me start by mentioning but rejecting the idea Peter Bofinger presented yesterday for flexible exchange rate targeting. I think that might be an answer if you think problems arise only because of bad behavior in the private sector. But governments sometimes misbehave as well, and most of us regard exchange rate arrangements as a method of disciplining governments as well as the private sector. And that function is simply not served by his proposal. Indeed, if you have a private sector that worries about the government using this sort of arrangement opportunistically, it seems to me that it could be a very dangerous system. So I do not see that as an answer, though, of course I do agree if a government is going to have managed floating, then it should do it in the sort of way that Peter described.

Secondly, let me draw your attention to something that I wrote back in the mid 1980s. When I first produced my proposals for 'target zones' (as I called crawling bands, before Paul Krugman appropriated the term 'target zones' for the European-type system in which parities were supposed to be fixed), I suggested that they might have 'soft margins' or 'soft buffers'. The idea was that there should not be an absolute government commitment to defend the edges of the zones, but rather that, in the event of strong speculative pressures, the government should have the right to say, "OK, we will let the rate go outside, but we warn you that we will use policies as and when possible to try to bring it back in, because that is where we still think it ought to be." I see some additional merit in that looser sort of system now, and regret that it rather got lost in the discussion of the late 1980s. An analysis of a particular version of this proposal by Bartolini and Prati in Economic Policy (April 1997) suggests that it would be worth reviving.

Or possibly, it would be worth going back even further in intellectual history to the mid-1970s, when there was a proposal called the 'reference rate proposal' floated in the literature shortly after the advent of generalized floating. The idea here was that the authorities should try and estimate what I subsequently called the fundamental equilibrium exchange rate. There would then be an obligation never to intervene, or conduct other exchange-rate policies, in a way that would push the market rate away from this reference rate. This idea was considered in the context of the discussion of the guidelines for floating in the IMF in the mid-1970s. It was rejected at that time essentially because the United States Treasury was run by ideological floaters who thought it wrong for governments to try and think what exchange rates ought to be, but it seems to me the time would now be opportune to reconsider whether this might not provide a way of reintroducing some sort of order into exchange rate arrangements without the type of brittleness that has caused these crises.

Another variant that I want to mention is much more recent and comes from a committee in India chaired by the former Deputy Governor of the Reserve Bank of India (Mr. S.S. Tarapore). This committee was charged with considering the case for India moving to capital account convertibility, and they recommended in May of 1997 that India should try to establish capital account convertibility within three years. But they also laid down some preconditions, which I do not think there is much chance India would have accomplished in three years even in the absence of the East Asian crisis. But what I want to draw your attention to is not the main proposals of the Committee for capital account convertibility, but rather the sort of exchange rate system that they suggested should go along with it, which they called a 'monitoring band'. This, in their conception, would again have required an attempt to estimate the equilibrium exchange rates which would form the center of the band. Within some range around that (they suggested plus or minus five percent), there should be a rule that the central bank would not intervene in the market. But once the rate went outside that band, on either side, they would be allowed but not compelled to intervene. This again gives a flexible intermediate system, which would avoid a commitment to defend a particular margin, which has proved such a problem in the context of speculative attacks.

Of the latter three proposals, I like the monitoring band proposal best. In discussing how monitoring bands might be deployed, let me draw on another strand of Peter Bofinger's work. This is the idea that intervention needs to be done by the strong currency country in order to give it credibility, because there is no automatic limit on how much a strong currency country can intervene. Suppose there was a monitoring band, where beyond five percent, the central bank of the country involved could intervene if it wanted to, and beyond 15 percent, the central banks of other countries would be free to intervene to support a currency if they wanted to. And suppose that arrangement had been in effect in East Asia in 1997. Having depreciated by 15 percent (which was enough for any objective need), suppose that the Thai baht or the Indonesian rupiah or the won had then been brought up enthusiastically by the Bank of Japan and the Federal Reserve, is it conceivable that we would have had the sort of crisis that actually developed?

As I have indicated, the purpose of an intermediate regime is to prevent the market from going off to extremes that make no sense in terms of underlying fundamentals. One is trying to induce the type of stabilizing speculation that Paul Krugman modeled in his classic article. What still remains to be resolved empirically is whether a band does in fact influence speculative behavior in the markets. We know that exchange rates within bands are much more strongly mean-reverting than they are without bands, but there remains controversy as to whether this is due to Krugman-type effects or to the interest rate behavior of the central banks. There is a lot of experience of bands on which one could draw, and I think it should be possible at this stage for some talented econometrician to do a study which would sort this out. If it turns out that bands do contribute to mean reverting behavior on the part of market participants, we would still need to make sure that the looser forms of intermediate regime that I have discussed above could be expected to have the same effect. A reasonable inference from the existing literature is that such an effect would be weaker, but would still exist.

I think we have to concede that these are not the best of times for advocates of intermediate regimes. But it is premature to conclude that the case for them has vanished. We may yet be able to fashion an alternative preferable to either of the extremes that are currently supposed to provide the only options.