A Bivariate High-Frequency-Based Volatility Model for Optimal Futures Hedging

摘要

This study examines the usefulness of high-frequency data for estimating hedge ratios for different hedging horizons. By jointly modeling the returns and conditional expectation of the covariation, the multivariate high-frequency-based volatility (HEAVY) model generates spot-futures distributions over longer horizons. Using the data on international equity index futures, performance comparisons between HEAVY and generalized autoregressive conditional heteroskedasticity (GARCH) hedge ratios indicate that HEAVY hedge ratios perform more effectively than GARCHhedge ratios at shorter hedging horizons. This implies that the distinct properties of short-time response and short-run momentum effects revealed in the HEAVY model are vital for hedge ratio estimation. (C) 2017 Wiley Periodicals, Inc.