Probability Distribution in Volatility and Value-at-Risk Forecasting for Cryptocurrency

This study investigates the role of probability distribution in forecasting volatility and Value-at-Risk (VaR). We use the Realized GARCH model and high-frequency data from the cryptocurrency market and show that the role of probability distribution varies across different situations. A skewed-t distribution contributes to better performance in forecasting the VaR; however, a skewed and heavy-tailed distribution does not improve the accuracy of volatility forecasting. For predictive models, the Realized GARCH model improves forecasting accuracy and provides economic gains for investors