改进的pair-copula参数估计方法在经济研究中的应用

Improved Pair Copulas Estimation with Applications in Economics

对于pair-copula中的参数估计,大多假设copula函数的参数和条件变量独立,将参数简化成一个不依赖于条件变量的常数。本文假设copula函数的参数和条件变量不独立,该参数是以条件变量为自变量的一元函数。应用该方法实证分析了"克强指数"三个指标铁路货运量、工业用电量和贷款发放量的对数增长率之间的关系,研究发现该方法优于简化的pair-copula参数估计,并且得出在固定铁路货运量不变时,工业用电量和银行贷款发放量成负相关关系,且这种负相关性随铁路货运量增加而减弱。

For the estimation of parameters in pair-copula, the parameters are often assumed to be independent of the conditioning variables and are simplified as constant. In this paper, assuming that the parameters and conditioning variables are not independent, and one parameter is an unary function of conditioning variable. Apply this method to analyze the relationship between the logarithmic growth of the three indexes of ＂Li Keqiang＇s Index＂, such as volume of railway freight, industrial power eonsmnption, and loan amount. The study found that this method is better than the simplifying assumption. And if the volume of railway freight is fixed, the industrial power consumption and loan amount are negatively correlated. The negative correlation weakens as the volume of railway freight increases.