Do Arbitrageurs Really Avoid High Idiosyncratic Risk Stocks?

Abstract

It is widely believed that stocks with high idiosyncratic risk exhibit stronger anomalies because arbitrageurs avoid holding these stocks due to diversification concerns, allowing deviations of prices from fundamental values. In this paper we test this proposition using hedge fund holding data. Controlling for stock size, we find that hedge funds allocate on average more capital towards holding high idiosyncratic stocks than they do towards low idiosyncratic risk stocks. Contrary to the prediction that diversification concerns prevent arbitrageurs from holding high idiosyncratic risk stocks, we find that the effect is stronger for small hedge funds and for less diversified hedge funds, and does not vary with hedge fund leverage. We also find that hedge fund trades in high idiosyncratic risk stocks earn significantly higher abnormal returns than trades in low idiosyncratic risk stocks. We propose that these results can be reconciled by high idiosyncratic risk being a proxy for information noise about the fundamental value. Consistent with this idea, we find that hedge funds require stronger mispricing signals from highly idiosyncratic stocks before trading.

Keywords: limits-to-arbitrage, arbitrageurs, hedge funds, mispricing, diversification, idiosyncratic risk, idiosyncratic volatility, volatility, anomalies, anomaly, frictions, event study, noise, analysts, dividends, signal-to-noise, fundamental value, mispricing, trading,signal

JEL Classification: G11, G12, G14, G23