A1448
Title: Forecasting exchange rate realized volatility: An amalgamation approach
Authors: Ekaterini Panopoulou - University of Essex (United Kingdom) [presenting]
Antonis Alexandridis - University of Macedonia (Greece)
Ioannis Souropanis - Kent Business School (Greece)
Abstract: The importance of realized volatility (RV) forecasting in exchange rates has both practical and academic merit. The aim is to provide a comprehensive analysis of the forecasting ability of financial and macroeconomic variables for future exchange rate realized volatility. Seven widely traded currencies are employed against the USD and linear models, and a variety of machine learning is examined, including dimensionality reduction and forecast combination approaches, along with creating a grand forecast (amalgamation approach) from these approaches. The findings highlight the predictive power of the amalgamation approach. Furthermore, forecasts on the separate frequencies of RV using wavelet analysis are generated, in order to extract frequency-related information and timing effects in the performance of the methods are examined. Overall, strong evidence that macroeconomic and financial predictors should be taken into consideration in RV forecasting is provided, along with the essential information that wavelet decomposition may provide.