Empirical competitiveness of deterministic option pricing models: Evidences from the recent waves of financial upheavals in India

Abstract

It is acknowledged by the leading researchers that the fundamental force behind the emergence of advanced option pricing models was the result of abundant empirical research analysis, leading to the fact that the asset return distribution is non-log-normal. It obviously resulted in strong moneyness and maturity pricing biases of Black-Scholes (BS). The concept defined the assertive distributional assumptions for Nifty (benchmark index of India), because the stated view supported the idea and its implications majorly focused on establishing the well-protected alternative model. The core concept of almost all the option pricing models was substantially indicating towards the flexible distributional structure. Such behaviour not only correlated underlying stock returns and its volatility but also subdued the level of skewness and kurtosis. Therefore, this article endeavours to investigate empirically the out-of-sample moneyness-maturity forecasting performance of deterministic volatility option pricing models during the recent waves of economic imbalance. It forecasts S&P CNX Nifty 50 index options of India against significant BS option pricing model and also attempts to find out forecasting performances of a family of deterministic volatility option pricing models, namely, Practitioner Black-Scholes (PBS), Gram-Charlier (GC) and Constant Elasticity of Variance (CEV). The accuracy of models' prices testify to the relative market price of Nifty index options using error metrics. For the purpose of this research, the prices are compared analytically through continuously updating the parameters of the models using cross-sectional option data, almost on a daily basis. The underlying focus of the article emphasizes on how to investigate the parameters of various models, pertaining specifically to a period of turbulence in the Indian economy. Eventually, while giving verdict to the results deduced, we measure out that the PBS model has smaller out-of-sample valuation errors in pricing Nifty Index options than the CEV and GC. However, it remains challenging to root out price bias completely out of all models we have framed out and applied. © 2013 Macmillan Publishers Ltd.