Exchange rate movements and central bank intervention

Insook Kim, Purdue University

Abstract

This thesis examines the behavior of exchange rates and the central bank intervention policy for Germany, Japan and the United Kingdom during the managed floating exchange rates, 1973-1989. In the first two chapters, we show the evidence of leaning-against-the-wind policy by estimating intervention reaction functions and presenting the evidence that, although central bank intervention in dollars appears unprofitable over some periods in the 1970s, the results are more positive if one examines the entire period up to 1989 in terms of the profits of the central bank. In Chapter 3, we modify the Corrado and Taylor (1986) model and assume two stochastic models in which the pure random walk model is a special case: In the first model, Case I, we introduce a first-order moving-average process and in the second model, Case 2, a first-order autoregressive process. The results show that, except in Corrado and Taylor's special case, there is always a leaning-against-the-wind policy that yields positive expected profits to the central bank. The variance of the changes in the exchange rates is reduced through a leaning-against-the-wind policy in both cases. In Chapter 4, we examine empirically whether or not exchange rates follow a random walk process. From regressions of the exchange rate on the exchange rate one day earlier, we find that they follow a random walk, and if there is a mean reverting process, it is very slow. Second, a regression of the daily change in the exchange rates on the change lagged one day indicated that there is a small but significant tendency for the change to continue in the same direction in the next day. Third, a variance-ratio test calls for a rejection of random walk hypothesis because the variance of changes over longer intervals is larger than expected if exchange rates followed a random walk. Fourth, a variance ratio test on monthly exchange rate data supports the claim by Huizinga (1987) that there is a mean-reverting tendency after 3-4 years.

Degree

Ph.D.

Advisors

Carlson, Purdue University.

Subject Area

Economics

COinS