CMStatistics 2022: Start Registration
View Submission - CFE
A1897
Title: Forecasting the conditional Covariance: Using daily, intraday returns, or both Authors:  Yongdeng Xu - Cardiff University (United Kingdom) [presenting]
Abstract: The intraday returns are used to construct the ``realized'' covariance, which provides a more precise measurement of covariance than the daily returns. When working with high-frequency data from markets that operate during a reduced time, an approach is needed to correct the missing overnight information. Typically, overnight covariance is added to the realized covariance, or the realized covariance is rescaled to the daily covariance. A recently developed high frequency-based multivariate GARCH (HF-MGARCH) model makes use of both daily and intraday returns, and provides more accurate forecasts of daily covariance. We compare the forecasting performance daily realized covariance models with the HF-MGARCH models. Statistically, we find that HF-MGARCH models perform better than the realized covariance models augmented with overnight returns. The daily rescaled realized covariance performs almost as well as the HF-MGARCH model. Economically, we find that a risk-averse investor would be willing to pay 20 to 80 basis points per year to capture the observed gains in portfolio performance by switching from intraday realized covariance to HF-MGARCH modelling of the daily conditional covariance.