IDIOSYNCRATIC RISK IN TRADING AND NON-TRADING HOURS AND THE MOMENTUM EFFECT

Abstract

In line with the notion that idiosyncratic volatility (IVol) acts as an arbitrage cost, prior empirical papers that studied the relationship between close-to-close IVol and momentum mispricing found larger momentum profits amongst high IVol stocks. I find that by decomposing close-to-close IVol into day (open to close) and night (close to open) components, I am able to more precisely determine if a stock has high or low IVol by studying their return patterns during the day and night separately. Using this approach, I found amplified risk-adjusted momentum return differences between high IVol stocks and low IVol stocks – about 10 times larger than prior empirical methodologies, which do not use the same form of IVol decomposition. Furthermore, I also find during rare episodic ‘momentum crashes’, stocks with high day and night IVol experienced larger negative returns compared to their low IVol counterparts.