A radial basis function approach to pricing and hedging options incorporating transaction costs

Nonparametric methods of pricing options have been available for some time, providing a viable alternative to traditional parametric methods. A general class of methods known as learning networks has been making significant inroads in option pricing literature. This thesis will adopt McLoone's hybrid linear/nonlinear training algorithm in developing RBF network models for the purpose of pricing and hedging options. An empirical study was first conducted to determine the suitability of our RBF network models in recovering simulated Black-Scholes option prices. We then assessed their ability to replicate options without transaction costs using some predefined backward induction performance measures.In addition, we further implemented our RBF network models on the new self-financing hedging strategy developed by Lai & Lim (2004), which is based on minimizing the expected cumulative hedging error and additional transaction rebalancing costs. Leland's strategy of discrete, regular revision time replications via a readjusted delta was used to gauge the effectiveness of this new strategy using both RBF network and Black-Scholes deltas.

Author dc:creator

• TING JEUM NGIT

• Learning Networks, Radial Basis Function, Optimization, Option Hedging and Pricing, Transaction Costs