On the Idiosyncratic Volatility and Heteroscedasticity in Stock Return Data

Lee, Woongki

doi:2018.47.6.947

Korean J Financ Stud > Volume 47(6); 2018 > Article

Korean Journal of Financial Studies 2018;47(6):947-975.
DOI: https://doi.org/10.26845/KJFS.2018.12.47.6.947 Published online December 31, 2018.

On the Idiosyncratic Volatility and Heteroscedasticity in Stock Return Data

Woongki Lee

고유변동성과 수익률 이분산성에 관한 논고

이웅기

고려대학교

Abstract

I analyze the idiosyncratic volatility (IVol) effects on expected returns over the period from April 1992 to March 2017 in the Korean stock market. The IVols are measured as in Ang, Hodrick, Xing, and Zhang (2006). Simulation results show the majority of individual return volatilities comes from non-systematic risks, implying the total volatilities and IVols are highly correlated and there must exist heteroscedasticity in stock returns. I propose a measure of bias in statistical power due to the heteroscedasticity. By using this measure, I find that equally weighted portfolios distort IVol effects most severely as compared to the portfolios from value-weight or median scheme. The unbiased IVol effects are more significant than the results from the previous research.

Key Words: 고유변동성,이분산성,비대칭성,통계적 검정력 편향,분위수 회귀분석,Idiosyncratic Volatility,Heteroscedasticity,Asymmetry,Bias in Statistical Power,Quantile Regression