Predicting Financial Volatility: High-Frequency Time-Series Forecasts Vis-a-Vis Implied Volatility
25 Pages Posted: 1 Mar 2002
Date Written: February 21, 2002
Recent evidence suggests option implied volatility provides better forecasts of financial volatility than time-series models based on historical daily returns. In particular it is found that daily GARCH forecasts have no or little incremental information over that already contained in implied volatilities. In this study both the measurement and the forecasting of financial volatility is improved using high-frequency data and the latest proposed model for volatility, a long memory model. The results indicate that volatility forecasts based on historical intraday returns do provide good volatility forecasts that can compete with implied volatility and sometimes even outperform implied volatility.
Keywords: Implied volatility, long memory, high-frequency data
JEL Classification: G14
Suggested Citation: Suggested Citation