We propose a novel framework for credit risk modeling, where default or failure information and rating or expert information are jointly incorporated in the model. These sources of information are modeled as response variables in a multivariate ordinal regression model estimated by a composite likelihood procedure. The proposed framework provides probabilities of default conditional on the rating information observed at the beginning of a predetermined period and is able to account for missing failure or credit rating information. Our approach is, to the best of our knowledge, the first that consistently combines failure-prediction models, where default indicators are used as responses, with so-called shadow rating models, where the responses are estimates of default probabilities usually derived from the leading credit rating agencies. In our empirical analysis we apply the proposed framework to a data set of US firms over the period from 1985 to 2014. Different sets of financial ratios constructed from financial statements and market information are selected as bankruptcy predictors in line with the standard literature in failure-prediction modeling. We find that the joint model of failures and credit ratings outperforms state-of-the-art failure prediction models and shadow rating approaches in terms of prediction accuracy and discriminatory power.

中文翻译：

---

失败和信用评级的联合模型

我们提出了一种新的信用风险建模框架，其中违约或失败信息以及评级或专家信息被共同纳入模型中。这些信息源被建模为多元序数回归模型中的响应变量，该模型由复合似然程序估计。提议的框架根据在预定时期开始时观察到的评级信息提供违约概率，并能够解释缺失的失败或信用评级信息。据我们所知，我们的方法是第一个始终将故障预测模型（默认指标用作响应）与所谓的影子评级模型（其中响应是通常来自领先的违约概率的估计）的方法。信用评级机构。在我们的实证分析中，我们将提出的框架应用于 1985 年至 2014 年期间的美国公司数据集。根据失败的标准文献，选择从财务报表和市场信息构建的不同财务比率集作为破产预测因子-预测建模。我们发现，故障和信用评级的联合模型在预测准确性和辨别力方面优于最先进的故障预测模型和影子评级方法。