Application Of Time Series Analysis In Carbon Market Prediction And Analysis

With the rapid development of science and technology, time series analysis has already been more and more popular, especially when the Professor Robert Engle and Professor Clive Granger obtained the Nobel prize in economics in 2003. Time series analysis method is mainly based on the knowledge of mathematical statistics, it is very common that time series used in the system description, analysis, predict the future decision-making and control aspects.As everyone knows that, the emission of carbon dioxide have caused very serious pollution to the environment of the world, and it will become worse and worse. The carbon trade is based on a international law, it is aim at the reduction of greenhouse gas emission. The prediction and research of carbon trading market has become more and more important for global social life.This article will predict and analyze the carbon price in the CER market, compare with the international carbon trading market, propose the advice to the carbon market in China. The article mainly used the methods、technology and software of time series analysis. Specifically, ARIMA-GARCH and BP、RBF model are used. This paper attempts to explain that it is feasible that making a forecast of carbon price by using neural network models. We use EVIEWS and MATLAB to do the carbon price prediction and simulation respectively. At the end of this article, we put forward that China should operate the carbon trading market in a Chinese characteristics way,after learning the lessons from European carbon trading market.The main innovative achievements of this article are:(1) The basic framework of time series analysis problems was detailed deducted and elaborated, it needs fully consideration of the characteristics of time series data when we build a basic model of time series.(2) The model of neural network and its optimization based on time series analysis have been put forward, BP and RBF neural network are especially introduced.(3) We got the effective forecast of short-term carbon price by modeling on the latest data of carbon price. We also analyzed the advantages and disadvantages of the models, including the ARIMA-GARCH model and BP、RBF models, the two neural network models.(4) We put forward the prospect of carbon market in China, compared the European carbon market with the actual situation of China.The research in this paper shows the important applications of time series analysis in carbon price time series. It has great value for the construction of the carbon market in China by comparing to the European carbon market.