Credit Risk & Underwriting
Default Probability in Shipping Credit Models
Summary
Probability of default (PD) expresses the likelihood that an obligor meets a specified default definition over a stated horizon.
Why this matters in ship finance
Explicit PD estimation enables portfolio comparison and expected-loss analysis, but shipping data and regime shifts can make estimates fragile.
The concept
Models may use borrower financial ratios, payment behaviour, fleet characteristics, freight exposure and governance; for regulated banks, prudential definitions and calibration requirements are relevant. PD is different from loss given default.
How it is used in practice
Validate event definitions, observation windows, missing data, out-of-time performance and calibration across shipping-cycle regimes.
Practical issues
A contractual covenant breach is not automatically a prudential default or IFRS 9 impairment event. Do not state model scores as verified probabilities without validation.
How ShipFinance.ai uses this concept
Present calibrated PD ranges, explanatory variables and transparent model limitations.
Key takeaways
- Define default before estimating PD
- Rare events demand cautious validation
- PD and recovery are different risk dimensions
Primary legal, rule and scholarly sources
- Schinas & Kewitsch (2015), Loans and Risk Management Considerations, in BK3 - scholarly chapter
- Schinas, Grau & Johns (eds.), The HSBA Handbook on Ship Finance (Springer, 2015) - scholarly reference, not legislation
- Regulation (EU) No 575/2013 - Capital Requirements Regulation (CRR; lenders within scope)
- IFRS 9 - financial instruments and expected credit losses (where IFRS applies)