碳排放交易市场的建立,是一个基于经济学理论来解决气候变暖问题的具有价值的途径,其目的是发展低碳经济。在欧盟排放交易体系一级市场上,以欧盟排放配额(European Union Allowances, EUA)作为主要交易标的物的碳排放权交易市场已经成为一个重要的新兴贸易市场。随着碳排放权交易市场的不断发展,该市场的资本化程度逐渐深化,其金融属性也日益显著,并逐步融入到国际资本市场体系之中。与其它资本市场相类似,碳排放权交易市场之间也存在着复杂的非线性相关关系,而Copula函数可以用来捕捉这种相依结构特征。因此,文章选取欧盟排放配额(EUA)期货的日价格时间序列数据,首先假设新息序列服从学生t分布,运用ARMA-GARCH模型对经调整的对数收益率序列进行过滤,采用极大似然方法估计模型的参数,并得到残差序列,同时将其标准化而得到标准化残差;然后,将Kendall s tau秩相关系数作为权重,采用最大生成树算法(maximum spanning tree algorithm)的序贯Copula选择方法构建合适的规则藤Copula模型,并运用基于序贯的极大似然方法估计规则藤Copula模型,以描述碳排放权交易市场之间复杂的相依结构特征。研究结果发现:在无条件下,t-copula函数可以较好地捕捉碳排放权市场之间的相依关系,说明市场存在明显的对称尾部;在Dec10EUA、Dec12EUA、Dec13EUA 市场相依结构固定下, Dec11EUA 与 Dec14EUA 市场之间的相依结构可以采用Gaussian copula函数来描述,而在Dec10EUA、Dec13EUA市场相依结构确定不变情形下,Dec12EUA与Dec14EUA市场之间的相依结构则适合采用Frank copula函数来捕捉,说明这些市场之间并没有出现尾部特征。进一步地,文章分别选择White信息矩阵等式拟合优度检验和基于概率积分转换(probability integral transform,PIT)与经验Copula过程(empirical copula process,ECP)混合方法的拟合优度检验,并基于Bootstrap方法,以Cramer von Mises(CvM)检验统计量作为度量测度,来对模型进行拟合优度的检验。研究发现,构建的规则藤Copula模型能够较好地捕捉碳排放权市场之间的相依结构。这一研究结果,为准确探讨碳排放权交易市场之间、碳排放权交易市场与其它资本市场之间套期保值策略提供了一定的参考意义,也有利于提高碳排放权市场产品定价的准确度。
英文摘要:
Formation of the carbon emission trading market is a valuable approach to deal with the global warming problems based on economic theory, which aims at the development of a low-carbon economy. In the primary market of the European Union emission trading system, the carbon emission trading market becomes an important emerging market where the European Union Allowance (EUA) is taken as the main object of the market transaction. With the development of the carbon emission trading market, its capitalization gradually deepens, the financial properties significantly increase and the market becomes integrated into the system of the international capital markets. Due to its similarity to other capital markets, a complex non-linear correlation exists in the carbon emission trading market, and the copula functions can be used to capture the characteristics of the dependence structure. Therefore, the paper chose the data on daily price series of EUA futures, assuming the series of innovations follows the Student s t-distribution, filtered the adjusted log-returns of EUA futures using the ARMA-GARCH model, and estimated the parameters in the model, obtained the series of the residuals and standardized them. Then, it took the coefficients of the Kendall s tau as the weights of the trees in the vine structure, constructed a feasible regular vine copula model by using the sequential selection approach based on the maximum spanning tree algorithm, and estimated the parameters by applying the sequential maximum likelihood method to describe the characteristics of the dependence structure of the carbon emission trading market. It is shown that the t-copula function can capture the dependence of the markets in the unconditional context, indicating that the carbon emission trading market has significant symmetric tails. Besides, the Gaussian copula function and Frank copula function describe the dependence of Dec11EUA and Dec14EUA conditional on Dec10EUA, Dec12EUA and Dec13EUA, and the dependence of Dec12EUA and Dec14EUA conditional on Dec10EUA and Dec13EUA indicating that there are no tails in the market. Furthermore, the paper tested the performance of the modeled regular vine copula framework by using the Goodness-of-fit tests of the White s information matrix equation and the combination of probability integral transform (PIT) approach and empirical copula process (ECP), which is based on the bootstrap method and Cramer von Mises (CvM) test statistics. The results show that the modeled regular vine copula framework performs well to describe the nonlinear dependence structure of the carbon emission market. This result provides some references to the discussion of the hedging strategy within the carbon emission trading market and with other capital markets. It is also beneficial to improve the accuracy of the risk management and pricing strategies in the carbon emission market.