TSK Fuzzy Inference System Based GARCH Model for Forecasting Exchange Rate Volatility

This paper applies TSK fuzzy inference system to the GARCH model for predicting the conditional volatility of foreign exchange rates returns. Out-of-sample forecast results of using TSK-based GARCH model are compared with that of an ANN-based and a SVM-based GARCH models, respectively. The empirical study shows that for the RMSE, MAE and Mincer-Zarnowitz regression test, the TSK-based GARCH model outperforms the ANN-based and SVM-based GARCH models. Therefore, TSK-based GARCH model is expected to be important in developing the novel strategies for volatility trading and advanced risk management.