参数法、半参数法和非参数法计算我国铜期货市场VaR之比较  

The Comparison of Chinese Copper Futures Market VaR Estimation Using Parametric Method , Semiparametric Method and Nonparametric Method

作　　者：刘向丽[1,2] 成思危[1] 汪寿阳[3] 洪永淼 

机构地区：[1]中国科学院研究生院管理学院,北京100190 [2]北京信息工程学院基础部,北京100101 [3]中国科学院数学与系统科学研究院,北京100190 [4]美国康乃尔大学经济学系与统计科学系,纽约14850

出　　处：《管理评论》2008年第6期3-8,共6页Management Review

基　　金：国家自然科学基金委员会优秀创新研究群体基金(70300501)

摘　　要：本文分别采用参数法、半参数法与非参数法来估计铜期货市场的风险值。由于利好与利空信息对期货市场波动性的影响是不对称的,因此本文参数法采用非对称的EGARCH与TGARCH模型来估计铜期货市场的条件VaR,半参数法采用前四阶矩方法,非参数法采用核函数的方法。一般金融市场只考虑下跌风险,而期货市场由于存在做空机制,所以本文亦提出期货市场上涨风险的概念,并分别估计了三种方法下的下跌VaR与上涨VaR,用返回检验对每种方法的有效性进行了评价。实证研究表明:不论是下跌VaR还是上涨VaR,非参数法对数据的拟和要远远优于另外两种方法,具有很强的实际意义。This paper adopts parametric method, semiparametric method and nonparametric method to estimate the VaR of the futures market. In view that good news and bad news have different impact on the market volatility, we adopt EGARCH and TGARCH model in parametric method and adopt the first four order moments in semiparametric method. The nonparametric method is based on kernel function. The general financial markets only consider the downside VaR. This paper also introduces a new conception of the upside VaR for the first time and estimates both the downside VaR and the upside VaR with the above three methods. We adopt backtest to evaluate the power of each method. Our findings indicate that no matter it is downside VaR or upside VaR, the nonparametric method gains an advantage over the other two methods. So the nonparametric method has powerful meaning in reality.

关 键 词：期货市场 VAR GARCH 核函数 返回检验 

分 类 号：F724.5[经济管理—产业经济] F224