Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
11023361 | Journal of Empirical Finance | 2018 | 26 Pages |
Abstract
In this paper, we develop a nonparametric methodology for estimating and testing time-varying fund alphas and betas as well as their long-run counterparts (i.e., their time-series averages). Traditional linear factor model arises as a special case without time variation in coefficients. Monte Carlo simulation evidence suggests that our methodology performs well in finite samples. Applying our methodology to U.S. mutual funds and hedge funds, we find most fund alphas decrease with time. Combining our methodology with the bootstrap method which controls for 'luck', positive long-run alphas of mutual funds but hedge funds disappear, while negative long-run alphas of both mutual and hedge funds remain. We further check the robustness of our results by altering benchmarks, fund skill indicators and samples.
Keywords
Related Topics
Social Sciences and Humanities
Economics, Econometrics and Finance
Economics and Econometrics
Authors
Biqing Cai, Tingting Cheng, Cheng Yan,