Expected Call Option Returns: A Reconciliation of Stock and Index Options and the Significance of Holding Period

Coval and Shumway (2001) show that under certain simple assumptions the expected returns to call options should be higher than the expected returns to their underlying assets, and increasing in the strike price. The authors find empirical support for these predictions by studying index option returns over the January 1990 to October 1995 period. Ni (2008), however, demonstrates that this does not hold empirically for stock options, where returns are relatively low and decreasing in the strike price. The author finds some evidence that a preference for idiosyncratic skewness in returns might explain this result. In this paper I show that the different results documented in these two papers are related to the differences in methods used to determine the "moneyness" of the options and the holding periods used to calculate the returns. Furthermore, when the bid-ask spread is included in the calculation of holding period returns a consistent pattern of low returns which are decreasing in the strike price emerges for both stock and index Calls.