Analysis Of Influencing Factors And Forecasting To Carbon Emission Quota Prices

In recent years,China has achieved rapid development in the field of carbon finance,which has played a huge role in promoting energy conservation and emission reduction in various industries.The significant progress of carbon finance has a positive impact on China’s economic transformation and the development of low-carbon industries,giving China a strong driving force for energy conservation and emission reduction technology.Economic transformation,low-carbon economic development,and low-carbon technological progress are crucial for the high-quality development of China’s economy.Not only is it a leap in economic level,but it will also greatly improve people’s living standards and conditions.The carbon emissions trading market defines property rights and introduces market mechanisms to reduce carbon emissions levels,while also reducing social regulatory costs.Therefore,it is necessary to promote the construction of China’s carbon emissions trading market.In this context,economic transformation,the incubation and growth of low-carbon industries and technologies will form a positive correlation with the development of China’s carbon emissions trading market.Through the construction of a carbon emission trading market,China can shift from an unreasonable input-output situation of high energy consumption,high pollution,and low returns to a low-carbon and cost-effective economy,promoting high-quality development of the Chinese economy.Based on the panel data of five regional carbon emission trading points from January 1,2020 to June 30,2021,this paper studies the impact of various factors on the closing price of carbon emission quotas,and on this basis,predicts the closing price of the representative carbon emission quotas in Hubei Province.This thesis first screens 19 variables at six levels: traditional energy,macroeconomics,foreign factors,environmental factors,external shocks,new energy and new economy by using adaptive Lasso,smooth shear absolute smoothing algorithm(SCAD)and minimum maximum concavity penalty algorithm(MCP).On this basis,we set up a long panel to analyze the data of the emission trading centers in Beijing,Shanghai,Guangzhou,Hubei and Shenzhen.For the long panel,we used the fixed-effect model,PCSE model and comprehensive FGLS model of SCC command regression to compare and analyze the estimation results of the panel.In the prediction of carbon trading prices,the traditional time series method ARIMA-GARCH model,the machine learning method LSTM model and XGBOOST model were used to predict the price of Hubei Province.The main conclusions of the thesis through empirical evidence are:First,the adaptive Lasso algorithm,SCAD algorithm,and MCP algorithm are used for variable screening,and variables with coefficients that are not zero under different methods are selected as reserved variables,ensuring good results in subsequent influencing factor analysis and price prediction.As for the data from the five regional trading centers,the factors that affect the closing price of carbon emission quota trading are the thermal coal index,Shanghai Composite Index,digital economy index,air quality index,green economy index,temperature,purchasing manager’s comprehensive PMI,and CPI.Second,the LSDV test first indicates that there is an individual effect;Secondly,among the eight variables,the thermal coal index,digital economy index,air quality index,green economy index,temperature and PMI have a positive change relationship with the closing price of carbon emission quota trading,while the closing price of the Shanghai Composite Index and CPI and carbon emission quota trading have a negative change relationship,and the influence path of these factors on the closing price of carbon emission quota trading is different.Third,in the prediction of closing price,through the comparison of RMSE and MAE two evaluation indicators,LSTM is better than XGBOOST and ARIMA-GARCH models in prediction,and the prediction effect of XGBOOST model is better than ARIMA-GARCH model.The LSTM model and XGBOOST model have advantages over the traditional time series model ARIMA-GARCH model in the face of strong fluctuations in data and processing and predicting extreme values.By analyzing of the influencing factors and prediction to the closing price of carbon emission allowances,this thesis provides a theoretical reference for stabilizing the price of China’s carbon emission trading market and improving the carbon emission trading mechanism.