ESTIMATION OF TIME-VARYING SYSTEMATIC RISK AND PERFORMANCE FOR MUTUAL FUND PORTFOLIOS: AN APPLICATION OF SWITCHING REGRESSION

There is substantial risk level nonstationarity in mutual fund portfolios. The Sharpe-Lintner-Mossin (SLM) model exhibited more nonstationarity than the Black capital asset pricing model. A bias is found in favor of high risk mutual fund portfolios. The hypothesis is rejected that the switching regression estimates of mutual fund performance are unbiased risk-adjusted measures. Separable performance measures are needed for the division of responsibility and allocation of resources among 2 tasks if the nonstationarity is due to the timing activities of fund managers. Opportunities for investor loss and gain are greater for a fund that does not deviate from its target level if the fund engages in timing activites.