Performance of stochastic option pricing models and construction of volatility smiles for option pricing in an emerging derivatives market

Abstract

This thesis provides the necessary methodology for the introduction of a fully functioning derivatives market in Vietnam where both European and exotic style derivatives can be traded, satisfying the needs of various potential market participants. The thesis has two objectives. The first objective is carrying out a comparison of performance between four stochastic option pricing models (Heston, Heston++, Bates and Heston-Hull-White), based on pricing a cross-section of stock options across various industries. The second is proposing a method of constructing implied volatility (IV) smiles for stock options in a new or illiquid option market (Vietnam in this case), using data from an existing market to do so (the US in this case). For the first objective, the results show that the Heston model performs the best in in-sample pricing, as well as capturing the characteristics of the market, while Heston ++ performs the best in out-of-sample pricing and hedging. For the second objective, the three proposed methods for constructing IV smiles, namely correlation, K-nearest neighbor (KNN) and weighted KNN, perform reasonably well, with weighted KNN considered the best among them.