A Simple Measure of Default-Risk Based on Endogenous Credit-Risk Models

Abstract

We introduce a new distance-to-default (DD) measure based on observable covariates, allowing us to bypass any model-based inference (e.g., Merton, 1974), that works well. It is based on the following result: The default event defined by endogenous credit-risk models, a sufficiently low asset value, is also described by a low ratio of the equity price to the firm's negative net cash flow. With a logistic function and observable covariates---equity price, equity volatility, and the firm's negative net cash flow---we fit one-year default probabilities (DP). For a single firm (Leland and Toft, 1996), fitting is almost perfect. For a set of different asset-volatility firms, including the new DD measure, which depends on the previous covariates, reduces fitting errors to a few basis points. The new DD measure picks up the asset-volatility effect in the DP---lining up well with the unobservable (Leland-Toft) asset-value based DD.

Keywords: Default and credit risk; Distance to default; Equity prices; Debt service and negative earnings; Equity volatility.

JEL Classification: G12, G21, G22, G33