A Goodness-of-fit Test for Copulas Based on Martingale Transformation
A Goodness-of-fit Test for Copulas Based on Martingale Transformation
复制标题
基于鞅变换的Copula拟合优度检验
DOI：
10.1016/j.jeconom.2019.08.007
复制
发表时间：
2020
影响因子：
6.3
通讯作者：
Xu Zheng
中科院分区：
文献类型：
--
作者：
Xiaohui Lu;Xu Zheng
本文对具有动态边际分布的Copula过程，如Gestival过程和阿尔马过程，提出了一种渐近分布无关检验。检验是基于参数估计的边际分布的经验copula过程。应用Khmaladze（1982，1988，1993）鞅变换方法，变换后的经验过程收敛于一个标准高斯过程，从而得到的检验统计量是渐近分布自由的. Monte Carlo模拟表明，该检验在有限样本下表现良好。实证应用测试欧元/美元和英镑/美元汇率之间的Copula。
This paper proposes an asymptotically distribution-free test for copulas with dynamic marginal distributions, such as GARCH and ARMA processes. The test is based on the empirical copula process with parametrically estimated marginal distributions. By applying the Khmaladze (1982, 1988, 1993) martingale transformation method, the transformed empirical process converges to a standard Gaussian process, so the resulting test statistics are asymptotically distribution-free. Monte Carlo simulations show that the test performs well in finite samples. An empirical application to test copulas between EUR/USD and GBP/USD exchange rates is provided.