成交量信息有助于预测碳价格波动吗——来自中国碳市场的经验证据  

Can Trading Volume Be Helpful to Predict Carbon Price Volatility?——Evidence from Carbon Markets in China

作　　者：杜坤海 王鹏 Du Kunhai;Wang Peng

机构地区：[1]西南财经大学中国金融研究中心

出　　处：《财经科学》2020年第1期42-54,共13页Finance & Economics

基　　金：中央高校基本科研业务费专项资金资助项目“宏观金融安全分析及预警”(JBK1805003)

摘　　要：成交量信息是否有助于预测资产价格波动?学术界目前存在两种截然相反的观点。本文以湖北、深圳、广东、北京等国内代表性碳市场为例,通过将成交量信息纳入二阶矩和高阶矩波动模型之中,对上述问题进行了实证研究。通过采用样本外递归预测法与更为严谨和稳健的SPA预测能力检验法,本文研究发现:(1)成交量信息对中国碳市场的波动过程普遍具有显著影响;(2)除方差外,碳收益的偏度和峰度也具有显著的波动聚集性和杠杆效应特征;(3)总体上讲,含成交量信息的波动模型比不含成交量信息的波动模型具有更优的预测能力,且高阶矩波动模型预测能力要优于二阶矩波动模型。最后,初步提出将该类模型运用于碳市场风险管理的思想。In academic circles,there are two opposite views about whether the trading volume information can help predict the volatility of asset prices.By taking four representative carbon markets in China as example,the paper made a empirical research with various GARCH models aiming at the above questions.The empirical results show that,the variance,skewness,kurtosis of carbon allowance return in China have notable time-varying characteristics,volatility clustering characteristics and leverage effect,and the volume information have significant and different effects on volatility of most carbon allowance return.Generally speaking,GARCH models with volume information have more outstanding forecasting performance,and higher moments volatility models have more outstanding forecasting performance.Finally,we suppose that the associated research between this model and risk management in carbon trading market would be significant.

关 键 词：碳市场 成交量 波动预测 SPA检验 

分 类 号：F832.5[经济管理—金融学]