Several people have attempted to evaluate the performance of mutual funds. Treynor [17] and Sharpe [15] have developed performance measures which make it possible to establish relative rankings for such funds. Treynor and Mazuy [18] have devised a statistical test for determining whether mutual funds successfully anticipate major fluctuations in the stock market. Jensen [7] has provided an absolute measure of performance which can be used to determine whether mutual funds earn higher or lower returns than those expected for the level of risk associated with their portfolios. McDonald [11] has employed the measures of performance developed by Sharpe, Treynor, and Jensen to evaluate the objectives, risk, and return of mutual funds in the period 1960–1969. Although these studies have examined mutual fund performance, none has employed an analytical framework for dealing explicitly with the nonstationarity which is likely to exist in the risk-return relationships for such funds [13].