Unaccounted model risk for Basel IRB models deemed acceptable by conventional validation criteria

Henry Penikas

Risk Management, 2023, vol. 25, issue 4, No 3, 25 pages

Abstract: Abstract Validation procedures are designed to prevent IRB models with model risk from being used in day-to-day business operations. The models that do pass are expected to entail negligible model risk. However, no one has studied the magnitude of such risk. Moreover, most of the prior literature criticises the IRB models either for missing certain features (such as concentration) or for being overconservative due to reliance on the 99.9% confidence level. By simulating around 700 scenarios for the hypothetical loan book, we are able to demonstrate the materiality of the model risk of the accepted IRB models. In addition, we explain that the accuracy of the conservative margin adjusting model calibration cannot offset the underlying model risk when the discriminatory power is not perfect (which is almost always the case). Ultimately, we arrive at suggested risk-weight (RW) mark-ups which enable the revealed and previously unaccounted for IRB model risk to be captured.

Keywords: Validation; IRB; Basel II; AUROC; CLAR; Loss shortfall; Brier skill score; Traffic light approach (search for similar items in EconPapers)
JEL-codes: C52 G28 G32 (search for similar items in EconPapers)
Date: 2023
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
http://link.springer.com/10.1057/s41283-023-00129-x Abstract (text/html)
Access to the full text of the articles in this series is restricted.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:pal:risman:v:25:y:2023:i:4:d:10.1057_s41283-023-00129-x

Ordering information: This journal article can be ordered from
https://www.palgrave.com/gp/journal/41283

DOI: 10.1057/s41283-023-00129-x

Access Statistics for this article

Risk Management is currently edited by Igor Loncarski

More articles in Risk Management from Palgrave Macmillan
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().