中文摘要：

利用高维Archimedean Copula模型对合成CDO进行定价，在传统简单Archimedean Copula的基础上，基于三种不同的方式，引入多个参数，从而解决作为市场基准的Gaussian Copula模型下存在相关性微笑的问题。对于特殊的大样本同质资产组合，违约损失分布可以直接从违约概率得到。而对于一般性的资产组合，可以得到损失的特征函数，从而通过快速Fourier变换，计算出违约的分布。最后，给出了数值计算结果。

英文摘要：

Models with high-dimensional Archimedean Copula were used to price synthetic CDO. Based on traditional Archimedean Copula, multiparameter Archimedean Copula was derived with three different methods. Through these methods, correlation skews problem in benchmark Gaussian Copula model can be solved. For large homogeneous portfolios, loss distribution can be directly got from default probability. Some numerical results are listed in the last part.