Nonparametric Tests of Alternative Option Pricing Models Using All Reported Trades and Quotes on the 30 Most Active CBOE Option Classes from August 23, 1976 through August 31, 1978

Authors

  • MARK RUBINSTEIN

    Search for more papers by this author
    • Graduate School of Business Administration, University of California at Berkeley. The author would like to thank the Chicago Board Options Exchange and Interactive Data Corporation for supplying the data necessary to undertake this study, and the Institute for Quantitative Research in Finance for funding. The original version of this paper contained only an empirical analysis of the statistical significance of deviations of market prices from Black-Scholes values. Special thanks are due to Kamal Duggirala who, while he was a student at Berkeley, independently replicated the results of the original version of the paper and went on to provide the supplemental analysis of economic significance.


ABSTRACT

The tests reported here differ in several ways from those of most other papers testing option pricing models: an extremely large sample of observations of both trades and bid-ask quotes is examined, careful consideration is given to discarding misleading records, nonparametric rather than parametric statistical tests are used, reported results are not sensitive to measurement of stock volatility, special care is taken to incorporate the effects of dividends and early exercise, a simple method is developed to test several option pricing formulas simultaneously, and the statistical significance and consistency across subsamples of the most important reported results are unusually high. The three key results are: (1) short-maturity out-of-the-money calls are priced significantly higher relative to other calls than the Black-Scholes model would predict, (2) striking price biases relative to the Black-Scholes model are also statistically significant but have reversed themselves after long periods of time, and (3) no single option pricing model currently developed seems likely to explain this reversal.

Ancillary