Abstract

My paper investigates in which periods the idiosyncratic volatility anomaly is observable, and the trend in recent years. It uses a graphical methodology that allows the reader to assess the effects of different starting and ending months. Plots for the value-weighted portfolio show that near the end of the sample period, the Ang et al. (2006) anomaly either attenuates or disappears. Consistent with Bali and Cakici (2008), the effect is weaker and insignificant for the equal-weighted portfolio. Using 5F and 6F benchmark return models shows similar results that differ quantitatively, but not qualitatively.