Do High-Frequency Volatility Methods Improve the Accuracies of Risk Forecasts? Evidence from Stock Indexes and Portfolio

• DOI: 10.1142/s0219477521500322
• Publication Date: 2020-12-31T07:19:03Z
• Authors: Siqi Xu; Kun Yang; Yifeng Zhang; Bo Li
• Keywords: Downside risk; Stock (firearms); Realized variance

Though the high-frequency volatility approaches are increasingly introduced to forecast financial risk in recent years, whether they can improve accuracies of forecasts remains controversial. This paper compares forecasting abilities four pairs low- and models, by calculating evaluating downside upside value-at-risk expected shortfall stock indexes portfolio. The empirical results show that, first, all models well filter serial dependence extremes, conditional standard deviation obtained from GARCH model performs best filtering dependence. Secondly, backtesting index portfolio consistent. More specifically, traditional low-frequency produce more accurate most cases, whereas methods also manifest some advantages extreme forecasting.