Empirical performance of alternative option pricing models with stochastic volatility and leverage effects

In this study, we develop a unified framework to analyze the asset price processes underlying option prices, and test a variety of model specifications using the S&P 500 index options. Specifically, we classify option pricing models, the stochastic volatility and leverage effects of which are generated by three channels: via the diffusion, via jumps, or via both. The empirical results from our specification analysis show that the performance of option pricing models can be improved significantly by generating stochastic return volatilities with two factor processes and with two sources of leverage effects that come separately from the jump and diffusion components.