Conditional Diseconomies of Scale in the Actively-Managed Mutual Fund Industry: What Do the Outliers in the Data Tell Us?

Recent research suggests that improper identification of outliers can lead to distorted inference. We investigate this issue by examining the role that multivariate outliers play in research outcomes using the Chen et al. (2004) study. We find that the documented negative relation between scale and return performance in the actively managed mutual fund industry is an artifact of extreme observations. A manual examination of the most influential observations with verifications against outside sources shows that these outliers are largely bad data. Removing the errors reduces the point estimates on the effect of fund size, rendering it economically and statistically insignificant. Further analysis employing regressions that mitigate outlier-induced bias and extending the sample through 2014 confirm our findings. Our evidence contributes to the recent research on the importance of outlier identification in finance research.