Evaluating mutual fund performance: an application of minimum convex input requirement set approach

Abstract

This paper has employed Chang's (Eur. J. Oper. Res. 115 (1999) 497–506) minimum convex input requirement set (MCIRS) approach to evaluate the performance of US mutual funds. Unlike the traditional methods, the MCIRS approach does not need to assume a particular functional form for the return generating process. The empirical results show that maximum capital gain and growth funds have done worse than growth and income funds, actively managers underperform a passive investment strategy, low risk funds outperform high risk funds, and no load funds outperform load funds. The paper also finds that funds with low beta and small assets have operated more efficiently. Scope and purpose Most of empirical studies of mutual fund performance use some asset pricing model (e.g., the capital asset pricing model) that has a specific functional form. The derived empirical results are controversial since the validity of asset pricing models is questionable. This paper has employed Chang's (1999) minimum convex input requirement set (MCIRS) approach to evaluate the performance of US mutual funds. The MCIRS approach can assign an efficiency score for each mutual fund. Unlike the traditional asset pricing methods, the MCIRS approach does not need to assume a particular functional form for the return generating process. The empirical results show that maximum capital gain and growth funds have done worse than growth and income funds, actively managers underperform a passive investment strategy (e.g., index funds), low risk funds outperform high risk funds, and no load funds outperform load funds. The paper also finds that funds with low beta (i.e., risk) and small assets have operated more efficiently.