An Option Pricing Model Using High Frequency Data

Abstract We propose a European call option evaluation framework accommodating GARCH-M model and its extension to handle irregularly spaced high-frequency data. The framework takes Bayesian approach to derive the predictive distribution for option prices and their volatilities. These predictive distributions vary as time approaches to the expiry data and provide credibility intervals to evaluate the option market. In empirical study, we illustrate the application using KOSPI200 and its options. Our approach results well-suited in the simulation based option pricing and explains a behavior of option prices close to the expiry.