My dissertation examines the effect of arbitrage risk on a large set of anomalies in the cross-section of stock returns. I find that long-short anomaly returns are concentrated among stocks with the most arbitrage risk, as proxied by idiosyncratic volatility. Volatility deters arbitrage activity against mispricing, and this effect is asymmetric between the long and short ends of each strategy. Greater profitability among high IVOL stocks can be attributed primarily to the extreme overpricing of the short leg in each anomaly. A composite strategy across all individual anomalies yields 1.3% alpha per month, and conditioning on the level of idiosyncratic volatility increases this alpha to 2.9%. An interaction strategy that buys the low-IVOL long leg and sells the high-IVOL short leg of each anomaly generates on average 2.5 times standard anomaly returns. Furthermore, these returns are higher following periods of high sentiment, when overpricing is likely to be more prevalent. These findings are robust to controls for size and liquidity, alternative estimates of volatility, and different portfolio construction schemes. In addition to the cross-sectional effects of IVOL, an aggregate IVOL measure across stocks also has time-series predictability for average anomaly returns.
展开▼
