An Empirical Study on Option Pricing Model Based on Stochastic Volatility

The prices obtained from the standard Black-Scholes model differ significantly from observed prices.These systematic valuation errors are documented in a stylized fact,the smile effect.These empirical biases reflect the fact that in reality volatility is not constant,but time-depending.Researchers create many models to correct the volatility smile.This paper examines the performance of two common extensions of the Black-Scholes framework,namely a GARCH and a stochastic volatility(Heston) option pricing model.The models are calibrated to Chinese option prices.When we analyze the fit to observed prices,both models underestimate the market price,and GARCH clearly dominates both Heston and the benchmark Black Scholes model.