Are Jumps in Stock Returns Diversifiable? Evidence and Implications for Option Pricing

This paper studies the diversifiability of jumps in stock returns. It presents a multivariate time-series model of the stochastic process for an index and its component stocks that explicitly admits discrete common jumps. Maximum likelihood estimation for such a model is developed and applied to the daily Major Market Index and its component stocks for the period 1985 through 1990. The paper finds that Poisson-distributed jumps observed from both the index and its component stocks constitute nondiversifiable risk, implying that the standard assumption in option pricing that these jumps are not priced may be invalid.