ESTIMATING THE EFFECTS OF INFORMATION SURPRISES AND TRADING ON STOCK RETURNS USING A MIXED JUMP-DIFFUSION MODEL

I present a methodology that uses the mixed jump-diffusion model for s tock returns to estimate the separate effects of information surprises and strategic trading around corporate events. Using simulation techn iques, I show that for events with multiple announcements spread over a long time, the estimators derived from the mixed jump-diffusion mode l are more powerful compared to the traditional cumulative abnormal re turn estimators. The new methodology is used to study the separate eff ects of information surprises and strategic trading associated with bl ockholdings and subsequent targeted repurchases. I find that for more than 93 percent of the firms in our sample the mixed jump-diffusion mo del in statistically superior to the pure diffusion model in describin g stock returns. More important, I find a statistically significant ne gative effect due to trading, while the average effect around announce ments is statistically insignificant. In contrast, the standard cumula tive abnormal return is not statistically different from zero.