Response to "The Asian Crisis: Lessons for the Future"

It is a great pleasure to be here, and an honour to be taking part in such an illustrious forum. We have all come to admire the mix of academic rigour, practical common sense, and pure humanity and civility that Stan brought to his time as Deputy Managing Director of the Fund. Even those of us who shouted at the Fund during the Asian crisis had the feeling that Stan was listening and trying to distil some sense out of the cacophony of voices that were aimed at him during that turbulent period. Of course each of us believed that what we were saying was right, but I for one accept that advice was being offered across the whole of the spectrum - from the incisively prescient to the outright silly - and it was hard to distinguish our pearls of wisdom from the dross that others were offering.

Given the time constraint, I want to pick up just one issue from the range of topics covered today - the exchange rate. It is such a vexed and sensitive topic that I need to start with a proviso: that what I say has no direct bearing on Hong Kong and its exchange rate regime. Hong Kong is special in many ways, and its exchange rate regime is one of them. It has served the economy well: and, I might add at the same time, that the role of the HKMA during the crisis exhibited the pragmatic, non-dogmatic approach which has served Hong Kong so well. I hope this audience will find some interest in a comment which, while germane to the Asian Crisis, has no direct policy implications for Hong Kong.

The exchange rate regime is still an issue of contention, both because it is unresolved, and it impinges so widely. I shall argue that, while we have better insights into what won't work with exchange rates, we still have an imperfect knowledge of what will work, reliably and consistently over time, to send the right price signal for resource allocation. We may be older and wiser, but we have still not found an exchange rate regime which can be relied on to help buffer the economy from shocks rather than exacerbate them.

However we look at it, exchange rates were a central element in the story of the crisis. Whatever the weaknesses in the financial sectors and their governance, the initiating shock came via the exchange rate. It reflected the size and overwhelming volatility of capital flows. Looking ahead, we need either a system which smooths the huge variability of foreign capital flows, or we need to find a more effective way of coping with it. I shall argue that because we can do neither particularly well, we need to do both.

The magnitude of these capital reversals need emphasising. Thailand was receiving private capital inflows equal to 12 % of its GDP before the crisis, which turned into an outflows equal to 16% of GDP during the crisis (and which continued at a pace of 10% of GDP per year since the crisis). Paul Volker has emphasised the scale disparities between the financial markets of these countries and those of the world as a whole, with the equity markets minute compared with the volume of international capital flows. Could a sophisticated market with breadth, depth and resilience have coped well with these extraordinary changes from euphoria to blind panic? I doubt it. But in any case these were "emerging markets" - by definition embryonic markets with little experience and inadequate infrastructure of rules and institutions.

Some of you will be thinking that the size and volatility of the flows was simply a product of the exchange rate regime, and that if a better exchange rate regime had been in place, then the flows would not have been large or volatile. This is, in my view, misguided thinking, perhaps driven by too much "book learning". The huge capital inflows in the years leading up to the crisis reflected, initially, the attractive profit opportunities which occurred in these countries as they integrated with the outside world, absorbing technology at a prodigious pace. They were learning quickly to do things much better, and there were large profit opportunities available (reflected in high interest rates) in this journey towards "best international practice". The fact that the inflows ran ahead of the investment opportunities was not something unusual or uniquely associated with quasi-fixed exchange rates: we observe it in every country in every upswing, including in the recent high-tech boom in the US. More to the point, these huge flows were not (or at least not to any extent) driven by some expectation of appreciation of the exchange rate - they began with well-based fundaments and were driven forward, for the main part, by unthinking euphoria. These flows were not just the product of domestic investors relying on a fixed exchange rate - it takes two to tango, and the foreign parties were more than eager to join the dance. The early foreign investors saw the great profit opportunities: the later ones went along (or were taken along) for the ride. This was not the equilibrium world of textbook analysis, but a chaotically dynamic world where capital was flooding to the places which genuinely seems to be the new frontiers, in an attempt to equilibrate returns between these new vibrant markets and the seemingly-sclerotic opportunities in the old economies. This was the market working as markets always do - seeking out non-equilibrium profit opportunities and overdoing the response when it found them. This was globalisation at work.

What were the consequences of the huge inflows? It should have been no surprise that markets had difficulty in "price discovery". It was no wonder that asset prices were bid up. And it was to be expected that upward pressure would occur on exchange rates. No-one should claim any early-warning prescience on account of the upward pressure on exchange rates, or identify this is a harbinger of the crisis - unless they are also going to go on and say that the capital inflow was the source of the problem and offer a solution. Current account deficits had to widen: this was, after all, the mechanism that brings about the transfer of real resources. If the exchange rate had been allowed to appreciate, would this have stemmed the flow? In the antiseptic world of the textbook, it does. In the real world, extrapolative expectations are the powerful dynamic, and fundamentals do not provide a clear and firm anchor for the exchange rate. Could the exchange rate have appreciated enough to create the expectation of future depreciations that would in turn have balanced off the interest differential? Could this hair trigger, knife-edge equilibria have survived not just the swings in opinion, but the correlated errors in forecasting that set the herd (or the lemmings) running in the other direction? It seems an act of extreme faith to expect this process of absorbing enormous capital flows to have been anything other than hugely disruptive.

We can't run the counterfactuals, but we know that countries with quite flexible rates and strong well-based financial sectors - Singapore and Australia - experienced change in the exchange rate of 25-30% during this period. We know in Australia that a floating exchange rate, for all its undoubted advantages, is hard work. How much harder for countries with less-well-established fundamentals, little history to anchor exchange rate expectations, neophyte institutions and weaker financial sectors.

Where does that leave us on exchange rate regimes? I accept whole-heartedly the thrust of Stan's view - that the move towards flexibility has been a move for the good. Some countries will find the fixed rate appropriate, but for most, a greater degree of flexibility than existed in the years leading up to the crisis will be appropriate. I want to explore two aspects of this a little further.

First, just how free a float? Many - Stan included - talk favourably in terms of a bipolar view of exchange rate regimes. Unless this is carefully qualified and explained, there is a risk that it plays to the purist (I would say extremists) in the floating camp. There are still commentators who argue that if governments so much as think about exchange rates, this is a mortal sin (probably related to the sin of conceit). Serious practical policy-makers should, however, note that not since Jimmy Carter's Presidency has anyone mouthed the words "benign neglect", and that the Japanese authorities did very nicely, thank you, out of selling the Yen at 80 and buying it, a couple of tears later, at 147. For his part, Stan says that by "floating" he means that a country will not defend a particular exchange rate. That's a broad (and in my view entirely correct) view of "floating". The test, I suppose, would be to ask whether Singapore might be the model for some of these countries - with the exchange rate quite heavily managed through intervention and interest rate actions, but with the authorities ready to step well back when shocks arrive or when fundamentals change. In doing so they are ready to re-group the defence lines further back (when they can be more sure the rate has gone "far enough"), and provide strong support at these levels¹.

The second issue is how to limit and cope better with the inevitable volatility. The first task might be to see what can be done to smooth capital flows. It is misguided to think about capital controls simply in terms of outflows (although there may be a place for these in a crisis, in the form of debt stand-stills). The problem came about because the inflows were too large. It was too easy for unsophisticated firms to borrow in foreign currency (most notably the Bangkok International Banking Facility). Prudential regulators took no account of the great credit risk which domestic banks were running when they made loans denominated in foreign exchange (and even when they made loans in local currency to firms which had taken on a large foreign exchange exposure)². In future, they need to find a way to build this risk into their regulations. Publicly available credit registers of foreign borrowing might help to make borrowers and lenders more conscious of the risks. There may well be a place for a Chilean-type withholding tax on inflows, tailored to discourage short-term flows. Of course these will not work perfectly, but the idea is just to throw a bit of sand in the wheels.

The point here is not just to discourage some flows (although in my view it would be no great loss if short-term flows were inhibited), but to make the banking system much less vulnerable if the exchange rate does move a lot. It was the linkage between foreign exchange crisis and banking crisis that proved so catastrophic. I'm unimpressed by those who argue that the crisis would have been avoided if depreciations had been allowed or encouraged earlier (i.e. somehow engineered in the period when the capital inflows were putting upward pressure on exchange rates). This leaves unanswered the question: how to avoid systemic problems in the banking system as the exchange rate fell, administering huge losses to those who had (over-)borrowed in foreign exchange. The best that could be argued along these lines is to say that if the crisis had somehow been precipitated earlier, it might have been less damaging. A small crash before the car built up to full speed does seem preferable, but hardly ideal.

Let me sum up. The Asian Crisis has left us just about unanimous in our view that one point on the spectrum of exchange rate regimes - fixed but adjustable - is not tenable. But it would be too glib to argue that some other point on the spectrum offers an easy answer. For my part, I accept that there are good reasons for the widespread "fear of floating". The approach suggested here is four-fold. First, as always, careful macro policies which are alert and ready to respond to signs of asset price bubbles and other signs of euphoria (even former central bankers never forget that it is their duty to take away the punch bowl just when the party is getting going!). Second, to see what can be done to reduce capital-flow reversals by limiting and smoothing inflows (in the same way that the sudden and damaging stopping in an auto accident can be reduced by prior action - not going so fast). Third, to accept some occasional exchange rate intervention to limit misalignment. And fourth, to work (mainly through prudential regulations) to make the banks more robust and resilient, particularly against exchange rate movements. Such intrusive prudential supervision can be well justified by pointing to the huge and damaging externalities of the crisis. The authorities should not be intimidated by the free-market vigilantes - whether simplistic model-builders from academia or self-serving self-appointed spokesmen for financial markets - who argued so vociferously before the crisis in defence of inaction (and who were often found among those arguing for government action after the crisis).

My guess is that there would not be big differences with Stan on the majority of these issues (although he would no doubt find more clear, subtle and diplomatic language to make some of the points). What strikes me in looking back over the period is how much we all learned. The Fund's learning process meant that Brazil was better handled than, say, Indonesia. The Fund seems better placed to handle future crises, and is doing much to make them less frequent. This progress is in no small measure a reflection of the intellectual power, energy and open-mindedness that Stan brought to his role at the Fund. Today's talk reminds us of these extraordinary talents.

¹ This is, I might note in passing, very different from the kind of Mexican intervention that Stan quotes with approval in his excellent Robbins Lectures - the attempt to smooth out short-term volatility. This seems to me to be a rather pointless objective - forward cover against these sort of minor short-term swings is cheap and readily available. The real issue for policy makers is not short-term volatility, but misalignment - big changes which last long enough to affect resource allocation. If there is a purpose in intervention, it is to try to fill in the troughs and lop off the peaks of misalignment (not, of course, to try to change the fundamental underlying rate).

² I might note in passing that it is simply a misunderstanding of the macro-economics to argue that these borrowers should have hedged their exposure. Individually they could do this, but collectively a country can't hedge its foreign exchange exposure without finding foreigners who are prepared to take a corresponding local-currency exposure, and these are limited. The effect of general hedging would be to nullify the capital inflow, as counterparties to the hedges offset their exposure by buying foreign currency assets.