🤖 AI Summary
This paper addresses the insufficient accuracy of intraday foreign exchange (FX) volatility curve forecasting, which hampers real-time risk management. To this end, we propose a novel functional GARCH-X model—the first to apply functional time series modeling to intraday FX volatility prediction. The model explicitly captures both cross-currency协同 volatility dynamics and the time-varying impact of microstructure variables (e.g., bid-ask spreads). By integrating cross-asset dependence structures with intraday high-frequency information, it significantly improves volatility curve forecast accuracy and enables construction of intraday Value-at-Risk (VaR) curves. Empirical results demonstrate that the generated VaR curves effectively mitigate extreme losses and enhance trading strategy robustness, delivering substantial economic value.
📝 Abstract
This paper seeks to analyse and predict conditional intraday volatility curves in FX markets using functional Generalised AutoRegressive Conditional Heteroscedasticity (GARCH) models. Remarkably, taking account of cross-dependency dynamics between the major currencies significantly improves intraday conditional volatility forecasting. Additionally, incorporating intraday bid-ask spread using a functional GARCH-X model further enhances predictability. The precise volatility forecasts motivate the construction of intraday Value-at-Risk (VaR). An intraday risk management application highlights that predicted intraday VaR curves can help mitigate dramatic losses in intraday trading strategies, showcasing their practical economic benefits.
