Entropy and credit risk in highly correlated markets

TL;DR

This paper compares two corporate default models by measuring divergence in their predictions across different market conditions, revealing increased inconsistency during financial instability, especially in highly correlated markets.

Contribution

It introduces a method to quantify differences between default models using divergence measures and highlights how market size and correlation affect model predictions during instability.

Findings

Divergence increases in highly correlated, volatile, large markets.

Divergence is minimal in small, low-correlation markets with high leverage.

Models become more inconsistent during financial crises.

Abstract

We compare two models of corporate default by calculating the Jeffreys-Kullback-Leibler divergence between their predicted default probabilities when asset correlations are either high or low. Our main results show that the divergence between the two models increases in highly correlated, volatile, and large markets, but that it is closer to zero in small markets, when asset correlations are low and firms are highly leveraged. These findings suggest that during periods of financial instability the single-and multi-factor models of corporate default will generate increasingly inconsistent predictions.