Overnight Trading And The Forecasting Of Financial Asset Volatility

In the context of global economic and financial integration,financial asset volatility prediction is of great significance to financial market risk management.Although digital transactions make it convenient for investors to conduct cross-border transactions,the existence of time differences will still affect investors’ transactions,which also highlights the importance of overnight information.This paper takes the all-weather trading of U.S.light crude oil(WTI)futures as an example to explore the impact of night-time trading fluctuations,which are ignored in existing research,on the prediction of intraday price fluctuations of financial assets.This paper uses a heterogeneous autoregressive(HAR)model to empirically analyze the intraday volatility modeling of WTI crude oil,and studies whether the addition of price fluctuation information in night trading can improve the prediction performance of intraday price fluctuations of US crude oil futures.And the research in this paper finds that the night-time fluctuations of the crude oil market can effectively improve the prediction accuracy of intra-sample fluctuations of WTI crude oil futures.And in out-of-sample forecasting,whether daily,monthly or weekly night trading information indicators are added separately,or all three are added at the same time,the new HAR extended model can effectively improve the prediction accuracy of intraday price fluctuations of crude oil futures.In addition,this paper uses the recursive window method and the HAR-CJ model to test the robustness,and the above conclusions are still robust.Based on the intraday 5-minute highfrequency trading data of the crude oil futures market,this paper provides rigorous evidence for the impact of night trading fluctuations on the intraday price fluctuations of financial market assets,and explores the spillover effect and forecasting effect of night trading on intraday fluctuations.The main innovations of this paper are as follows:(1)A new indicator,overnight information,was found for intraday price volatility prediction in financial markets.In the study of crude oil futures market,scholars have added various predictive indicators to the HAR-RV benchmark model to study whether these indicators can improve the predictive ability of the model.In this paper,we add the indicator of overnight information to the benchmark model,and conduct in-sample prediction,out-of-sample prediction and robustness test respectively,and analyze and confirm the improvement effect of this indicator on in-sample and out-of-sample prediction accuracy.(2)This paper provides new evidence on the importance of overnight information in international financial product transactions in the context of economic globalization.Many scholars have introduced overnight information as an explanatory variable into the volatility model to study its impact on the forecast of asset price volatility in various financial markets.In the context of globalized financial transactions,the nighttime transactions of international financial assets mainly come from other countries.And we also find that the night-time transaction fluctuations of financial assets can effectively predict the intraday fluctuations of the crude oil market,which proves the importance of nighttime transaction information in cross-border transactions.(3)The spillover relationship between overnight volatility and intraday volatility is decomposed by time-frequency,and the dynamic spillover effect under different time-frequency is discussed.And the method of Baruník 和 K?ehlík(2018)is used to decompose the volatility of crude oil futures by time-frequency,which deepens the understanding of the existing research on the volatility spillover effect of financial assets overnight and intraday.