Testing the Expectile Market Timing Ability of Mutual Fund Managers

Summary This study examines the market timing abilities of US mutual fund managers using an expectile‐based predictive regression model with ARMA‐GARCH covariates, effectively capturing both lower and upper tail market risks. To evaluate timing skills across different expectile levels, we introduce a nonparametric profile empirical likelihood (PEL) test, whose test statistic follows a chi‐square distribution under the null hypothesis. Simulation results confirm the accuracy of the PEL test in terms of empirical size and power across various settings. An empirical analysis utilizing CRSP data reveals significant variation in market timing abilities across expectile levels, with stronger skills observed at the tail level compared with the mean level (). Furthermore, the interplay between market timing and stock‐picking skills appears to be context‐dependent. These findings highlight the importance of assessing timing abilities across multiple expectile levels to capture diverse market dynamics.