基于得分驱动模型的广东和湖北两省碳排放权交易市场波动特征研究

Research on Fluctuation Characteristics of Carbon Emissions Trading Markets in Guangdong and Hubei Provinces Based on Score-Driven Model

现有关于碳排放权价格波动特征方面的研究多采用GARCH族模型,但是常见GARCH模型中方差是关于过去收益率平方的线性模型,这使得方差对于极端数值往往响应过度且这样效应存续时间较长.针对这一问题,Harvey和Harvey&Sucarrat提出了得分驱动的模型构建思路.采用得分驱动的模型构建思路,利用Beta-skew t-EGARCH模型对广东和湖北两省碳排放权交易市场波动进行研究.使用单组分、双组分Beta-skew t-EGARCH模型和常用EGARCH模型对两组碳排放权收益率序列进行拟合,极大似然估计值和BIC值均表明单组分Beta-skew t-EGARCH模型的拟合效果更好.

Existing studies on the price fluctuation characteristics of carbon emission rights mostly use the GARCH family model,but the variance in the common GARCH model is a linear model about the square of the past return,which makes the variance often over-response to extreme values and this effect lasts for a long time.In response to this problem,Harvey and Harvey&Sucarrat proposed a score-driven model construction idea.This paper adopts the idea of scoring-driven model construction,and uses the Beta-skew t-EGARCH model to study the fluctuation of carbon emission trading market in Guangdong and Hubei provinces.The single-component,two-component Beta-skew t-EGARCH model and the commonly used EGARCH model are used to fit the carbon emission right yield series.The maximum likelihood estimates and BIC values both indicate the single-component Beta-skew t-EGARCH The fitting effect of the model is better.