The EUR/PLN exchange rate is of great importance to Polish exporting industries as well as to investors in the Polish region. As a consequence, there is a particular interest in managing foreign exchange (FX) risks due to short and long FX exposures. By implementing a variety of GARCH models under different return distributions, we forecast the exchange rate volatility of daily returns of EUR/PLN
... [Show full abstract] exchange rates. Confirming findings of recent literature, we ap-ply Fractionally Integrated GARCH (FIGARCH) and Asymmetric Power ARCH (APARCH) which reveal a statistically significant long memory effect and an asymmetry in volatility in both time series. These characteristics implicate some challenges in volatility fore-casting and Value-at-Risk (VaR) predictions in short and long trading positions. Therefore, we combine these two effects in the Fractionally Integrated Asymmetric Power ARCH (FIAPARCH) modeling framework which yields the best goodness-of-fit of all aforementioned models. The FIAPARCH outperforms all other models in regard to the applied loss functions and is found to provide the best VaR prediction results. Our findings contribute to research on volatility of financial instruments in Poland and expand the findings of existing literature, raise awareness of combined effects to practitioners, and may give insights to Polish financial regulators, as well.