Floating exchange rates allow a country to achieve external balance while maintaining control over its money supply and monetary policy.

But floating exchange rates are also highly variable, more variable than we expected when many countries shifted to floating rates in 1973.

Governments that have chosen floating exchange rates worry about the large amount of variability, and nearly all manage the float to some extent. Some governments manage their floating rates closely. If the floating exchange rate is heavily managed, then it behaves more like a fixed exchange rate, and the analysis of the previous chapter is relevant. Other governments, including the governments of most major countries that have chosen floating rates, use management selectively. Occasionally the government intervenes in the foreign exchange market.

Is selective or occasional intervention effective in influencing exchange rates? Thirty years ago the conventional wisdom was clear. If the intervention is not sterilized, then it can be effective.

However, it is effective not because it is intervention but rather because it changes the money supply. Unsterilized intervention is simply another way to implement a change in the domestic money supply and monetary policy. By changing the money supply, it can have a substantial effect on the exchange rate. If the intervention is sterilized, the conventional wisdom was that it would not be effective in changing the exchange rate, at least not much or for long. Yet interventions by the U.S., Japanese, and British monetary authorities, among others, are fully sterilized.

The conventional wisdom was based on studies that showed little effect of sterilized intervention.

It was also based on the relatively small sizes of interventions. In a market where total daily trading was hundreds of billions of dollars, interventions that typically were less than \($1\) billion seemed too small to have much impact.

More recent studies have challenged this conventional wisdom. How might sterilized intervention be effective, even though it does not change the domestic money supply and is relatively small? The most likely way is by changing the exchange-rate expectations of international financial investors and speculators.

Intervention can act as a signal from the monetary authorities that they are not happy with the current level or trend of the exchange rate.

The authorities show that they are willing to do something (intervention) now and they signal that they may be willing to do something more in the future. For instance, the authorities may be willing to change monetary policy and interest rates in the future if the path for the exchange rate remains unacceptable. Sterilized interventions then can be a type of news that influences expectations. If international investors take the signal seriously, they adjust their exchange-rate expectations. Changed expectations alter international capital flows, changing the exchange rate in the direction desired by the authorities.

For instance, in 1985 the major governments announced in the Plaza Agreement that they were committed to reducing the exchange-rate value of the dollar. They intervened to sell dollars.

International investors shifted to expecting the dollar to depreciate by more than they had previously thought, and the exchange-rate value of the dollar declined rapidly.

Recent studies indicate that sterilized intervention can be effective. One indirect measure of effectiveness is whether the monetary authority makes a profit or a loss on its interventions over time. If the authority buys a currency and succeeds in driving its value up (and sells to drive the value down), the authority makes a profit

(buy low, sell high). An early study found that central banks generally incurred losses on their intervention in the 1970s. However, a study of U.S. intervention during the 1980s found that the U.S. monetary authority made a profit of over

\($12\) billion on its dollar–DM interventions and a profit of over \($4\) billion on its dollar–yen interventions during this period. Another study concluded that interventions during the mid- and late 1980s significantly affected exchange-rate expectations in the direction intended.

Recently, a number of monetary authorities began releasing data on their daily intervention activities, information that had previously been kept secret. With these data the quality of statistical studies has improved dramatically. The typical findings are as follows:

• Intervention is usually effective in the short period of two weeks or less, in reversing the direction of the trend of the exchange rate, or at least in reducing the speed of the trend (when the authority is leaning against the wind).

• Larger interventions are more successful.

• Coordinated intervention, in which two or more monetary authorities intervene jointly to try to influence an exchange rate, is much more powerful than an intervention by one country of the same total size.

• The effectiveness of the intervention often diminishes after this two-week period, and often there is no discernible effect one month later.

Here is one example. After the introduction of the euro in 1999, the European Central Bank (ECB)

intervened to influence the exchange-rate value of the euro on only four days, in the fall of 2000, after the euro’s exchange-rate value had declined substantially during 1999–2000. On September 22 a coordinated intervention by the ECB and the monetary authorities of the United States, Japan, Canada, and Britain bought several billions of euros. The initial impact was to increase the euro’s value by over 4 percent, though about half of the effect was lost during the remainder of the day.

The intervention was successful in reversing the direction of the trend for about 5 days, and after 15 days the euro’s decline was less than what it would have been if the previous trend had simply continued. Still, the euro’s decline continued, and the ECB intervened again on November 3, 6, and 9.

The same effects of reversing the trend for about 5 days and smoothing the decline for about 15 days occurred. These interventions may have had some lasting impact because at about this time the value of the euro stopped its downward trend and began to vary around a trend that was essentially flat.

Here’s another example. The Japanese government has been quite concerned about the variability of the yen’s value. From early 1991 to 1995, the yen appreciated from 140 per dollar to 80.

Then the yen depreciated to 146 in 1998, appreciated to about 100 in 2000, and then depreciated to about 125 by April 2001.

Japan’s monetary authority bought dollars (and sold yen) on 168 days during this time period, for total purchases of over \($200\) billion. It sold dollars (and bought yen) on 33 days, a total of

\($37\) billion.

Ito (2003) reckons that these interventions were highly profitable for the Japanese government.

It bought dollars at an average price of 104 and sold dollars at an average price of 130.

It realized capital gains of \($8\) billion, earned an interest differential on its holdings of the dollars it bought of about \($31\) billion, and had an unrealized capital gain at the end of the period of \($29\) billion. In fact, the Japanese government seemed to have an unstated target exchange rate of about 125 per dollar. Sales of dollars were at rates above 125, and purchases below.

By combining days of intervention that are close together, Fatum and Hutchison (2006) identify 43 separate instances of intervention, 29 of buying dollars during 1993–1996 and 1999–2000 and 14 of selling dollars during 1991–1992 and 1997–1998. In 34 of these, Japan’s monetary authority was clearly leaning against the wind—

intervening against the trend during the previous two days. Of these, 24 reversed the trend for the two days after the intervention, and in another 5 the trend rate of exchange-rate change was lowered. The major failures occurred in the first half of 1995, when continual interventions could not prevent the yen from appreciating from 100 to the then-amazing level of 80 yen per dollar.

However, even when the Japanese authorities succeeded during this short two-day window, there was generally little effect on the exchange rate a month after the intervention.

Coordinated interventions with the U.S.
monetary authority (which occurred on 23 days during the decade) were much more powerful than interventions only by Japan. Ito estimates that, for 1996–2001, independent intervention of \($2\) billion by the Japanese authority moved the yen value by 0.2 percent on average. Coordinated intervention of \($1\) billion by each of the two authorities moved the yen value by about 5.0 percent.
After intervening rather little during 2001 and 2002, the Japanese government engaged in massive interventions during 2003 and the first quarter of 2004 to attempt to prevent appreciation of the yen. The government intervened on about 40 percent of these days and purchased a total of about \($315\) billion. Nonetheless, the yen did appreciate from 120 per dollar at the end of 2002 to 104 per dollar at the end of March 2004, although the appreciation might have been larger if there had been no intervention. The Japanese government then abruptly ended all intervention after March 2004, and the yen value was fairly steady for several years after the cessation.
Our beliefs about the effectiveness of sterilized intervention by the major countries is now cautious and nuanced. If the time frame is a few days, intervention seems often to be successful. If the time frame is one month or more, it usually seems to be unsuccessful. Still, there are times, such as late 2000 for the euro, when the monetary authorities believe that the market has pushed an exchange rate far from its fundamental value.
As Michael Mussa, then chief economist for the International Monetary Fund, said a few days before the September 2000 euro intervention, “Circumstances for intervention are very rare, but they do arise. One has to ask, if not now, when?”
DISCUSSION QUESTION On March 11, 2011, a devastating earthquake and tsunami hit Japan. Surprisingly, during the next week, the Japanese yen appreciated by 5 percent against the U.S. dollar. Should the Japanese central bank have intervened in the foreign exchange market? If so, how should it have intervened? If not, why not?