Stability of mutual fund systematic risk statistics

Empirical data for 85 mutual funds are used to test the intertemporal stability of their systematic risk statistics. Reasons why the portfolios may be nonstationary are suggested. A random coefficient model developed by Theil [37] is employed to test for the stability of each fund's beta. The data suggest that some funds do exhibit a beta that is best described as being a random coefficient. However, the percentage of funds exhibiting this characteristic was not statistically different from the percentage of randomly created portfolios that exhibited a random beta coefficient. The findings of this study support the statistical models employed in two other recent studies [18,21] to test for the stability of beta. Yet, for mutual funds that do exhibit a random beta coefficient, the partitioning of the total risk of the portfolio return into systematic and unsystematic risk is no longer valid for explaining the total risk.