“Why do Firms with High Idiosyncratic Volatility and High Trading Volume Volatility have Low Return?” by Chuan Yang HWANG
Chuan Yang HWANG
Nanyang Technological University
Thomas J. George
University of Houston
Ang, Hodrick, Xing and Zhang (2006) documented a puzzling negative relationship between return and idiosyncratic volatility (hereafter referred to as the AHXZ result). Since then, Bali and Cakici (2006) have suggested that the AHXZ result is not robust and is absent when equally-weighted portfolios are considered. However, we show that once we exclude the tax-loss selling effect in January and exclude penny stocks from the sample, the AHXZ result is very significant in equally-weighted portfolios and is robust to idiosyncratic volatility measures of different data frequency and to the returns measured in different holding periods. We hypothesize that the AHXZ result is caused by investors’ mispricing of high idiosyncratic volatility stocks due to the difficulty of estimating the fundamental of such stocks and a lack of analyst coverage. All of the testable implications from this hypothesis are supported: 1) Low returns of the high idiosyncratic volatility stocks only exist in firms with low analyst coverage, 2) Earning announcement returns are negative for stocks with high volatility and are more negative when the analyst coverage is low, 3) Accounting returns of high volatility stocks are on average lower than those of low volatility stocks, especially for low analyst coverage firms. Finally, we show that our hypothesis also explains a related puzzle documented by Chordia, Subrahmanyam and Anshuman (2001) that stocks of high trading volume volatility earn low returns.