On Threshold Quantile Regression with Heteroskedasticity: Stock Return-Volume Relations

• Activity Number: 419
• Type: Contributed
• Date/Time: Tuesday, August 11, 2015 : 2:00 PM to 3:50 PM
• Sponsor: Business and Economic Statistics Section
• Author(s): Cathy W.S. Chen* and Mike K.P. So and Thomas C. Chiang
• Companies: Feng Chia University and The Hong Kong University of Science and Technology and Drexel University
• Keywords: Asymmetric Laplace distribution ; Quantile regression ; Volume Asymmetric ; HP-filter ; Market beta ; MCMC

Abstract

This paper presents a CAPM-based threshold quantile regression model with GARCH specification to examine relations between stock excess returns and "abnormal trading volume." An adaptive Bayesian Markov chain Monte Carlo scheme, exploiting the link between the quantile loss function and the asymmetric-Laplace distribution, is employed for estimation and inference, simultaneously estimating and accounting for nonlinear heteroskedasticity plus unknown threshold limits and delay lags. Six daily Dow Jones Industrial stocks are considered, the proposed model captures asymmetric risk through market beta and volume coefficient that change discretely between regimes that are driven by market information and various quantile levels. This study finds significantly negative effects of abnormal volume on stock excess return under low quantile levels, nevertheless there are significantly positive effects under high quantile levels. The evidence indicates that each market beta varies with different quantile levels, capturing different states of market conditions.