A Letter to the Editor

By Mr. de Beaufort Wijnholds
Executive Director
International Monetary Fund

The Economist
October 14, 2000

Sir,

While you are right in observing that the most discussed policy matters at the IMF/World Bank meeting in Prague were the interventions in the euro and the oil market (September 30), you seem to side with the skeptics, especially about euro intervention, citing the `hard' truth that foreign -exchange support purchases succeed only in rare circumstances. Unfortunately you do not adequately address the crucial question whether the coordinated intervention of September 22 actually constitutes such a rare occasion. Consider that the jump in the rate of the euro (initially from 83 to 90 dollarcents, then settling around 88) was quite large if - as you guesstimate - the intervention amounted to only $3 to 5 billion; that the G7 got its act together quickly (a pretty rare occurrence); that the effect of the intervention has now lasted for about two weeks and that the euro rate was not affected by the negative vote on Danish participation in EMU.

You state that none of the conditions for making interventions a success have been met. While you agree - with everybody else - that the euro has become clearly undervalued, you deny that any speculative activity has been involved. How can you possibly know this? To most practitioners (I am a former central banker) it would appear that the accelerated undershooting of the euro that was taking place prior to September 22 was precisely fueled by position-taking on the assumption that the monetary authorities had no stomach for coordinated intervention (too close to the US elections).

Intervention is not for the faint-hearted and, yes, the participants have taken risks in the process, but precisely because the risk of doing nothing was considered greater: another indication that this has been indeed one of those rare circumstances where intervention is called for. Are huge amounts needed to make intervention work? It depends on what your objective is. Stabilizing a clearly undervalued rate requires considerably less ammunition than trying to substantially reverse a steep depreciation.

It is also time to put to rest the popular, but fallacious, view that since central bank reserves equal only a fraction of the daily turnover on foreign-exchange markets, intervention can be merely a drop in the bucket. The fact is that interventions to be effective do not have to match the total - or even a significant part - of all transactions conducted on the foreign-exchange market, but only the ex ante imbalances in these transactions. In other words, intervention in order to be effective, needs to be at least equal in size to the extent to which the demand for the undervalued currency falls short ex ante of the supply for it. This ex ante imbalance tends to be only a modest fraction of total market turnover. Unfortunately there are no usable estimates of the size of such imbalances, which we can only determine ex post, so that intervention comes down to Fingerspitzengefühl, as they say in Frankfurt.