Rolling Connectedness Of Crude Oil Futures And EPU With Chinese Commodity Market In Time-frequency Domain

For the past few years,commodity markets have attracted extensive attention from investors and researchers due to the frequent and violent fluctuations of commodity prices and increasing demand for commodities from investors,with much investment capital flowing into commodity markets.Currently,with the in-depth development of global financial integration and the continuous advancement of industrialization in emerging economies,the dependence of Chinese commodities on global related industries,especially the crude oil futures,keeps high growth.Against a backdrop of increased uncertainty,the effect of economic policy uncertainty on domestic financial asset investment and other aspects cannot be ignored.In view of this,this paper develops rolling window wavelet vector autoregression and connectedness networks to evaluate the time-varying characteristics of the connectedness of crude oil futures,economic policy uncertainty and Chinese commodity markets,in order to provide some reference and asset allocation for policy makers and investors.In order to study the time-frequency rolling connectedness of crude oil futures,EPU and Chinese commodity markets,this paper selects the weekly data from August20,2004 to December 31,2021,and combines wavelet vector autoregression model for empirical research.The wavelet transform method can analyze the connectedness of crude oil futures,EPU and Chinese commodity markets under different investment periods.At the same time,rolling windows analysis can provide time-varying connectedness studies under different market conditions.This paper firstly expounds the impact mechanism of crude oil futures and EPU on commodity markets,then constructs a corresponding empirical model based on wavelet transform and vector autoregression theory,then conducts descriptive statistical analysis and granger causality test on data samples.In the empirical research section,we examine connectedness across different scales and investigate dynamic connectedness and timevarying effects based on rolling windows.Additionally,we choose different rolling windows for robustness checks.Finally,policy suggestions are put forward in order to provide policy and investment suggestions for domestic policy makers and domestic and foreign investors according to the empirical results.The empirical results show that: first,the impact of crude oil futures and EPU on commodities varies across different time scales,and the long-term is higher than the short-term.Second,oil futures and EPU are positive in the directional connectedness network,and they affect financial markets across different scales primarily as information contributors.Third,there are significant differences in the impact of EPU and crude oil futures on commodity markets under different market conditions.Particularly during the crisis,aggregate connectedness across time scales remained high,and the contribution of EPU and crude oil futures to the commodity futures sector increased significantly.Fourth,the impact of EPU and crude oil futures on different commodity industries is heterogeneous.Oil’s contribution to energy commodities is higher than other commodities,while economic policy uncertainty affects grains more than other commodity sectors.Finally,we found that the degree of connectedness is sensitive to the size of the window.The shorter the window period,the higher the total connectedness,but the stronger the connectedness fluctuation;and the longer the selected window,the lower the connectedness value,but the weaker the connectedness fluctuation.