Empirical Performance Of Alternative Option Pricing Models-Based On The Hong Kong Market

Over the past three decades, the basic assumptions behind BS Model deviated from the Market quickly with the consistent changing and developing of financial markets, which significantly influenced the pricing precision of BS Model. In order to price options more precisely, numerous option pricing models are derived based on relaxing the BS assumptions about the stochastic process of interest rate and underlying asset price. The main aspect of improvement for BS model is the stochastic process of underlying asset price, which can be divided into first-order improvement and second-order improvement. The first-order improvement is: (1) adding the stochastic volatility into the stochastic process of the underlying asset price, and (2) adding the random jump into the stochastic process of the underlying asset price; the second-order improvement is: (1) adding random jump into the stochastic volatility process, (2) assuming the random jump follows exponential distribution instead of normal distribution, (3) assuming a nonzero correlation coefficient between the jump parts of the stochastic volatility process and the stochastic underlying asset price process .Based on the option and future data form Hong Kong Security Exchange, this paper investigates the pricing efficiency of BS model and the alternative option pricing models which give the first-order improvement for BS assumption on stochastic process of the underlying asset price, including the SV (Stochastic Volatility) model from Heston (1993) and the SVJ (Stochastic Volatility and Random Jump) model form Bates (1996), and the effect on the pricing precision of BS model from the two first-order improvements. Through the positive investigation, we intent to find the best option pricing model for Hong Kong market and the most effective first-order improvement for BS model.The results are: (1) the SVJ model has the best pricing precision, followed by the SV model, and the BS model ranks the last; (2) adding the stochastic volatility into the stochastic process of the underlying asset price can significantly improve the pricing precision of BS model, especially for short term—out of the money options, however, including the random jump can only improve the pricing ability of SV model marginally.