【摘要】 Logistic模型是研究违约率的主流方法之一,但目前的研究未能对最优样本配比与分界点这两个基本问题给予足够的重视。本文采用模拟分析方法设计了样本配比—临界点的情景,通过实证比较的方法得出1∶3的样本配比与0.647的临界点比较适合我国的情况,而常用的1∶1的样本配比可能并不适用。更多还原

【Abstract】 The literatures about logistic default risk modeling by now have not paid enough attention to fundamental questions,that is,what is the optimal pairing structure of the training sample and what is the optimal cutoff point?This paper does some primary research into these questions.We first design typical scenarios of different sample pairing structure and cutoff point combination.Then comparing logistic model estimation and prediction efficiency in these different scenarios,we concludes in the end that 1∶3 pairing of default companies and healthy companies and cutoff value of 0.647 fit with China’s data set,and widely used 1∶1 structure may be not suitable.更多还原