Effects of debt collection practices on loss given default

In this article, we propose an LGD model that is solely based on legal and internal debt collection actions. Our model is supported by empirical tests in which it performs better than a usual firm specific model. This result is noteworthy when we recall that the model has only binary variables that indicate whether an action was taken. Our model can be applied to update the LGD of distressed firms in a timely manner reflecting the actions taken during the debt collection period. It also can be used to assess the effect of a recovery action and to determine whether to apply an action to certain types of debt.

► We propose an LGD model based on legal and internal debt collection actions. ► Our model performs better than a firm specific model in both in- and out-of-samples. ► Combining two models further improves explanatory power and prediction accuracy. ► LGD of distressed debts can be continuously updated reflecting the actions applied. ► Cost and benefit of a recovery action can be assessed via our model.