Distinction between Intra-day and Overnight Stock Return Distribution and its Influence on Option Pricing

This research investigates the importance of including the overnight behavior of the stock price in an option pricing model. During the trading hours stock prices process follows a stochastic model, which allows for stochastic volatility and random jumps. We model the overnight behavior by a single jump, which is independent of the intra-day component. In our research we consider Heston, Bates and Variance Gamma models, extend them by introducing an overnight jump and compare their performance. We _nd that introducing an overnight component in already existing option pricing models leads to option prices closer to the ones observed on the market. However the overnight jump alone is not able to capture the stock price behavior, random jumps should also be included. We _nd that the most successful of the considered option pricing model uses both a random jump during the day and scaled-t distributed overnight jumps, where the improvement from a normally distributed overnight jumps is signi_cant