Forecasting exchange rate volatility and hedging

The managment of risks arising from exchange rate volatility has become a very important concept in today's conditions. In order to ensure full risk management, it is important to dominate the market conditions and to interpret the economic indicators correctly. On the other hand, to make predictions about exchange rate volatility can protect against exchange rate risks. For this aim, the volatility estimation in exchange rates should be estimated with the right estimation models. In this study, exchange rate data are used for weekly $/TL and €/TL rates between 01.01.2010 and 28.12.2018. Then ARCH, GARCH, E-GARCH, T-GARCH and A-PARCH models were estimated. In order to find the best prediction model, both dynamic analysis RMSE (Root Mean Square Error) and Akaike Information Criteria were used in the comparison between models. As a result of the study, the best model which minimizes the numerical data is found to be dynamic E-GARCH (1,1) for dollar rate and dynamic GARCH (1,1) model for euro rate according to RMSE criterion. In comparison with the Akaike criterion, T-GARCH for dollar rate and A-PARCH model for euro rate were the best estimation model. However, contrary to expectations, no leverage effect was observed in the prediction models. In other words, positive shocks cause more volatility. This study is a first in the literature in terms of conducting the exchange rate crisis that started in 2018 and making the comparison between models for this data set with two methods. Volatility, ARCH Model, GARCH Model, Financial Derivatives