CMStatistics 2022: Start Registration
View Submission - CFE
Title: Intraday foreign exchange rate volatility forecasting: Univariate and multilevel functional GARCH models Authors:  Fearghal Kearney - Queens University Belfast (United Kingdom)
Han Lin Shang - Australian National University (Australia)
Yuqian Zhao - University of Kent (United Kingdom) [presenting]
Abstract: The aim is to predict conditional intraday volatility in foreign exchange (FX) markets using functional Generalized AutoRegressive Conditional Heteroscedasticity (GARCH) models. We contribute to the existing functional GARCH-type models by accounting for the stylised features of long-range conditional heteroscedasticity and cross-dependence in major FX currencies through estimating the models with long-range dependent and multi-level functional principal component basis functions, as well as incorporating the intraday bid-ask spread microstructure information. Overall, we demonstrate the statistical and economic superiority of appropriately modelling FX volatility using various functional GARCH-based models. Remarkably, we find that taking account of cross-dependency dynamics between the major currencies can significantly improve intraday conditional volatility forecasting in the FX market. Intraday risk management benefits and inter-daily asset allocation applications are presented to highlight the practical benefits of our proposed approaches.