Research On The Risk Spillover Effect Of The Oil Market On The Carbon Market Based On Copula-CoVaR

In the context of global energy conservation and emission reduction,countries around the world have taken the carbon market as an important way of energy conservation and emission reduction to control carbon dioxide emissions.The carbon market regards carbon emission rights as a commodity that can be bought and sold openly in the market.It aims to control the total carbon emissions of the global contracting countries by limiting the carbon emission allowances of each country in the contract,so that the carbon emission rights can be traded between countries or enterprises with insufficient and excessive demand.Carbon market has both similarities and differences with the traditional financial market.On the one hand,common factors influencing the financial market will have an impact on the carbon market price;on the other hand,specific factors such as the price of fossil energy represented by oil will also affect the carbon market price.Because the combustion of oil is the main cause of the rise in carbon dioxide emissions,there is an inseparable relationship between the price of oil and the price of carbon emissions trading.The price of oil will affect the demand for oil by enterprises,and thus affect the amount of carbon dioxide emissions,so that the demand for carbon emission rights will change,and it will eventually be reflected in changes in the carbon market price,this leads to the risk loss events occurring in the oil market will quickly spread to the carbon market,that is,in theory,the oil market will have a certain risk spillover effect on China’s carbon market.At the same time,under the background of economic globalization China’s foreign oil dependency is more and more big,the domestic oil price is in line with the international market,the domestic oil price will be affected by the international oil price,so that the oil market at home and abroad on the risk of loss event will influence the development of China’s carbon market,and China’s carbon market risk of overflow.Therefore,how to measure the risk spillover effect of domestic and foreign oil market on carbon market has become an urgent task for the healthy development of carbon market.Because the yield data of financial markets usually has a characteristic of thick peaks and tails.The traditional empirical distribution method used to estimate the edge distribution can often fit most of the observations,but the fitting effect on the tail values is very poor.The extreme tail risk is the most noteworthy in the risk measurement,and there is a nonlinear correlation between most financial markets.Therefore,in the process of risk measurement,on the one hand,it is necessary to select an edge distribution model that can describe the extreme risks of financial markets,and on the other hand,it is necessary to introduce a new theoretical method to describe the interdependence between markets.Therefore,based on the consideration of extreme market risk conditions,this paper combines the extreme value theory that fits the marginal distribution of financial sequences with the Copula function describing the dependence relationship,and applies the CoVaR method to the oil market risk spillover effect on the carbon market In the research,the risk spillover effect was directly converted into a specific value,and the EVT-Copula-CoVaR model was constructed to quantitatively analyze the direction and intensity of the oil market risk spillover to the carbon market.On the one hand,this article compares the risk spillover effects of oil markets on different carbon markets and explores the differences in risk spillover effects between different carbon pilots,on the other hand,compare the risk spillover effects of domestic and foreign oil markets on the same carbon market,and clarify the differences in the direction and intensity of risk spillover effects between different oil market and carbon market combinations.The main conclusions are:From the calculation results of risk spillover effects:First,when the oil market is in a risk condition,the CoVaR value of the carbon market yield sequence is greater than the corresponding VaR value,which leads to a risk spillover effect is all positive.That is,the risk events of the domestic and foreign oil markets have a positive spillover effect on each pilot carbon market.When a risk event occurs in the oil market,the risk of the carbon market will increase accordingly;Second,comparing different carbon markets,it can be found that under the same confidence level,the oil market has the largest spillover effect on the Hubei carbon market,followed by the Guangdong carbon market,and the third and fourth are the Shenzhen carbon market and the Beijing carbon market,respectively.The risk spillover effect on the Shanghai carbon market is minimal.The reason may be that Hubei carbon market has the largest trading volume and turnover,so it is more susceptible to the influence of the oil market;Third,comparing the spillover effect of the domestic and foreign oil markets on the carbon market,we can find that the spillover effect of the foreign oil market on the carbon market is greater than the spillover effect of the domestic oil market on the same carbon market.As a result,the average daily oil import volume has increased year by year and surpassed the United States,thus becoming the world’s largest oil importer,marking the continuous increase of domestic oil dependence on foreign countries,so the carbon market is relatively more affected by the international oil market;Fourth,the higher the confidence level,the more obvious the risk spillover effect.From the validity of the risk measurement model:The number of failure days for the VaR predictions of domestic and foreign oil markets,five carbon markets,and CoVaR predictions for ten market combinations are all within the critical range of the failure frequency test,that is,the measures of the VaR values of the seven markets and the CoVaR values of the ten market combinations are valid,indicating that the EVT-Copula-CoVaR model constructed in this paper and the Monte Carlo simulation method used in risk measurement have certain accuracy.