A tug of war of forecasting the US stock market volatility: Oil futures overnight versus intraday information. (31st August 2022)

• Record Type: Journal Article
• Authors: Ma, Feng; Wahab, M. I. M.; Chevallier, Julien; Li, Ziyang
• Is Part Of: Journal of forecasting. Volume 42:Number 1(2023)
• Page Start: 60
• Page End: 75
• Publication Date: 2022-08-31
• Subjects: oil future overnight RV -- oil futures intraday RV -- signed intraday returns -- the US stock market volatility; Forecasting -- Periodicals; Forecasting -- Mathematical models -- Periodicals; 003.2
• DOI: 10.1002/for.2903
• Languages: English
• ISSNs: 0277-6693

Abstract

Abstract: This study is the first to examine the impacts of overnight and intraday oil futures cross‐market information on predicting the US stock market volatility the high‐frequency data. In‐sample estimations present that high overnight oil futures RV can lead to high RV of the S&P 500. Moreover, negative overnight returns are more powerful than positive components, implying the existence of the leverage effect. From statistical and economic perspectives, out‐of‐sample results indicate that the decompositions of overnight oil futures and intraday RVs, based on signed intraday returns, can significantly increase the models' predictive ability. Finally, when considering the US stock market overnight effect, the decompositions are still useful to predict volatility, especially during high US stock market fluctuations and high and low EPU states.