Comparison of ARCH models in forecasting volatility on the EUR/USD currency market

Abstract

In this paper the generalized autoregressive conditional heteroscedastic models were applied for modeling volatility of the exchange rate of EUR / USD for daily observations using dataset of period starting 1 January 2010 to 30 December 2016. The paper analyzes both asymmetric and symmetric models that found numerous facts about exchange rate returns such as volatility clustering and leverage effect. The performance of GARCH and GARCH-M models as well EGARCH, GJR-GARCH and APARCH (models with different residual distributions were analyzed to a given dataset. The best models for forecasting volatility of EUR/USD exchange rates are APARCH, GJR-GARCH and EGARCH model with Student’s t-distribution.