The aggregate degree of investor diversification can make conventional aggregate idiosyncratic risk measures very noisy proxies for the priced aggregate undiversified idiosyncratic risk. We postulate that the noisy measures are responsible for the less robust and thus controversial evidence on the relation between idiosyncratic risk and expected future market return. We propose a simple method of noise reduction to re-examine the predictive relation. The simple method is to include two different noisy measures (dual-predictor) in the predictive regression. Noise effects tend to cancel out due to a high correlation in the noise components of the two noisy measures. We find that the dual-predictor alone predicts future market return with an adjusted R2 of more than 4% and produces large economic gains in trading strategies based on out-of-sample forecasts. The predictability of market return by the dual-predictor is robust to the inclusion of other known predictors and to the choices of sample period and market index.
Keywords: Expected Return, Idiosyncratic Risk, Predictability, Undiversified Risk.
JEL Classification: G12; G14; G17
학회소식
학회소식
[2010년 제 4차] When Does Idiosyncratic Risk Really Matter?
작성자 : 관리자
조회수 : 933
게시일 :
2010-12-09