Support Vector Regression Model of Currency Options Pricing with Stochastic Volatility Models and Forward Exchange Rate

Support Vector Regression (SVR) is a new general learning method, proposed in the statistical Learning Theory. In this paper, we construct a new nonparametric currency options pricing model with SVR approach. We integrate stochastic volatility models (SV) into SVR to upgrade the forecasting ability of the price of currency options. And we use the forward exchange rate as the input variable of SVR, as the forward exchange rate takes the interest rates of the pair of currencies into account. The inputs of SVR will include the moneyness (Spot rate/strike price), forward exchange rate, the volatility of the spot rate, domestic risk free simple interest rate, and the time to maturity. Analytical results reveals that new model provides greater predictability than traditional approaches such as GK model and artificial neural network option pricing model on the same data sets.