Development a New Hybrid Modeling Approach for European Option Pricing

One of the important risk management tools in financial markets, is financial derivatives. In this study, we investigate the problem of European option pricing. The most key input to option pricing models is volatility. For accurate modeling of volatility, we use three famous GARCH type models including GARCH, EGARCH, GJR-GARCH. With using the results of the best GARCH-type model, we develop two nonparametric models based on Neural Networks and Neuro-Fuzzy Networks to price call options for S&P 500 index. We compare the obtained results with those of Black-Scholes model and show that the Black-Scholes model in not appropriate for at-the-money options. Furthermore, by comparing the Neural Network and Neuro-Fuzzy approaches with Black-Scholes model, we observe that the accuracy of non-parametric models are better than the Black-Scholes model.