Volatility Modeling

This paper attempts to compare the forecasting performance of the ARIMA model and hybrid ARMA-GARCH Models by using daily data of the Iran’s exchange rate against the U.S. Dollar (IRR/USD) for the period of 20 March 2014 to 20 June 2015. The period of 20 March 2014 to 19 April 2015 was used to build the model while remaining data were used to do out of sample forecasting and check the forecasting ability of the model. All the data were collected from central bank of Iran. First of all, the stationary of the exchange rate series is examined using unit root test which showed the series as non stationary. To make the exchange rate series stationary, the exchange rates are transformed to exchange rate returns. By using Box-Jenkins method, the appropriate ARIMA model was obtained and for capturing volatilities of returns series, some hybrid models such as: ARIMA-GARCH, ARIMA-IGARCH, ARIMA-GJR and ARIMA-EGARCH have been estimated. The results indicate that in terms of the lowest RMSE, MAE and TIC criteria, the best model is ARIMA((7,2),(12)) –EGARCH(2,1). This model captures the volatility and leverage effect in the exchange rate returns and its forecasting performance is better than others.