The paper examines whether it is better to separate a model/sample to account for differences between large and small firms when developing a predictive tool to identify which companies are likely to become distressed, and therefore at risk to the supply chain. Previous research has scaled data...