Estimating Short-term Default Probabilities Conditional to Economic Conditions: Applications of Regularisation Approach and Economic Adjustment Coefficients
Data publikacji: 20 cze 2025
Zakres stron: 178 - 197
Otrzymano: 14 gru 2024
Przyjęty: 31 sie 2024
DOI: https://doi.org/10.2478/bsrj-2025-0009
Słowa kluczowe
© 2025 Siti Aisyah Mustafa et al., published by Sciendo
This work is licensed under the Creative Commons Attribution 4.0 International License.
Background
Corporate bonds are crucial for corporations as they provide a flexible and often less costly alternative to equity financing. However, rising corporate debt levels, along with rating downgrades and economic uncertainty, can cause corporations to face financial distress, exacerbating the probability of default.
Objectives
The purpose of this paper is to estimate bond default probabilities conditional on fluctuations in economic growth over short-term frequencies using inputs from rating transitions.
Methods/Approach
The estimation is based on a Markov chain framework and the incorporation of economic growth by utilizing specifications of the economic adjustment coefficient. Further, quasi-optimisation of the roots matrix is utilized to extend the model within a quarterly domain.
Results
Economic growth (proxied by GDP) carries little informational content on the future default probabilities. Non-investment grade ratings depict higher default probability, while investment-grade ratings yield default propensity of less than 1.1% in the next quarters and exhibit higher distance between default probabilities by tenor points and neighbouring states as the time horizon lengthens.
Conclusions
First, practitioners can measure forward-looking bond exposure across different tenure buckets using the estimation approach developed in this study. Second, by considering historical fluctuations in the economic cycle as an additional factor for estimating future default probability, this study informs financial market regulators by providing entities with an alternative reference point to their in-house generated models, helping them meet regulatory requirements.