Corporate Failure Prediction Modeling: Distorted by Business Groups' Internal Capital Markets?
研究发现，在破产预测模型中同时纳入子公司和集团层面的变量，能显著提升预测准确性；集团支持会提高子公司的生存概率，尤其是核心业务子公司。
Abstract: Most models in the bankruptcy prediction literature implicitly assume companies are stand‐alone entities. However, in view of the importance of business groups in Continental Europe, ignoring group ties may have a negative impact on predictive reliability. We find that models encompassing both bankruptcy variables defined at subsidiary level and at group level have a substantially better fit and classification performance. Furthermore we find that the group's support causes improved survival chances for subsidiaries, especially when these subsidiaries belong to the group's core business. Overall our results are consistent with existing theoretical and empirical findings from the internal capital markets literature.