The role of trading volume in volatility forecasting

Abstract

Current models of volatility generally either use historical returns or option implied volatility to generate forecasts. Motivated by recent findings in the strand of literature focusing on volume-return (price) volatility relationships, this paper proposes the introduction of trading volume into various ARCH frameworks to improve forecasts. In particular, ex-ante evidence indicates that the incorporation of option implied volatility and trading volume into an EGARCH model leads to outperformance over other alternate forecast approaches. Noticeably, abnormal returns obtained from trading simulation underscores the improvement in forecast accuracy to be economically significant. These results remain robust to different measures of volatility and volume and offers scope for investors to more accurately predict volatility in the future. © 2010.