EXPAR-GARCH model for Forecasting Financial Time-series Data

In this paper, a novel model which is a combined Exponential autoregressive (EXPAR) and Generalized autoregressive conditional heteroscedastic (GARCH) model has been developed for capturing the characteristic features of financial time-series data. The method for estimation of parameters using EXPAR-GARCH model has been proposed. Further, the developed methodology has also been illustrated using financial time-series data sets that have both cyclical and volatile characteristics. It is seen that the combined model performs exceedingly well, for goodness-of-fit and forecast performance, compared to the individual models.