Option valuation using Garch-type models : Empirical evidence using USD/INR data

The aim of this paper1 is to evaluate the option pricing performance of three GARCHtype models against the Black-Scholes model. The standard GARCH model and the asymmetric GJR-GARCH and EGARCH models are employed with the Gaussian innovation distribution. The models are assessed on two sample sets of USD/INR currency European call options data obtained from the National Stock Exchange of India. A Monte Carlo simulation approach is implemented on the estimated GARCH models for the valuation of European call options. An accuracy analysis of the option prices is then performed across different moneyness and maturity using relative loss functions. The three GARCH models performed better than the Black-Scholes model in all samples of USD/INR for all moneyness and maturity. From an overall standpoint, the GJR-GARCH model displayed the best option pricing accuracy.