Abstract

This paper examines the influence of the survivorship bias on performance persistence in Canadian mutual funds. Our sample covers the period of January 1986 till December 1999. Spreads of the survivorship bias on mutual fund returns are gauged by comparing the difference between the sample of surviving funds and the sample of surviving and defunct funds. The comparisons are conducted first only on equity funds, and later on funds in all categories. Contingency tables are used to address the question of performance persistence. Cross Product Ratios (CPR) are obtained for all funds, active and inactive, on an annual basis. Probit models are used to explore the odds of and factors that contribute to the disappearance of funds. Major findings of this study are as follows. The effects of the survivorship bias on Canadian mutual funds are nontrivial. Persistence of fund performance has been found, while reversals are also observed. The persistence is correlated across managers; this may be due to certain common factors. An examination of fund disappearance in the probit models indicates that funds' return, size, and expense ratios are significant predictors of fund's attrition, while the optional sales charges, whether a fund is affiliated with an insurance company, and how long the fund has been in existence are also significant other factors. These results are consistent with those reported for the US mutual fund industry.