Do Changes in Reserves Proxy Well for Official Intervention?

The study of foreign exchange intervention has been slowed by the traditional reluctance of central banks to release intervention data to researchers. To circumvent the lack of actual intervention data, researchers often have used publicly available foreign exchange reserves data to proxy for the confidential intervention data. Little research has been done, however, to compare the characteristics of reserves and intervention. In this article, the author addresses that issue by using time-series techniques and measures of correlation to compare American, Swiss, and German monthly intervention and reserves series. Although the raw correlations are modest, ranging from about 0.12 to 0.42, some simple adjustments for seasonality and ERM realignments can increase the correlations in the Swiss and German data. More sophisticated adjustment techniques would be difficult and time-consuming.