Request a Brochure

Learn, explore and unleash your inner chef.

For more information, contact:
Yiota Andreou

What are the benefits of using IRB models as risk management tool?

Ever since the introduction of Basel II, IRB models already contributed  a lot in many banks to enhance credit risk management. Obviously, the use of IRB models in risk management is already well spread amongst banks. Developing and maintaining IRB models demands a substantial effort and resource contribution, it is hence key to leverage as much as possible on this investment. In many banks, overall quality of IRB models is excellent and frequently monitored, although I do acknowledge that some concerns are valid and the ECB launched the TRIM inspections for good reasons. Most of the concerns are rather based on the calibration of the models from a prudential point of view (model uncertainty, data governance) and are not undermining the ranking power of the IRB models. By ranking power I mean the ability of the model to discriminate between high risk and low risk clients. If a model with a good ranking power is available, then it seems logical from an efficiency and consistency point of view, to start from this model and to use it for multiple reasons: IRB capital calculation, IFRS ECL calculation, acquisition strategy, pricing and risk management. The underlying challenge for all the mentioned domains is by and large the same: discriminating good clients from bad clients by ranking them based on creditworthiness. Given this common objective, it is – certainly for a retail bank – quite natural to start from the same ingredients. It goes without saying that different calibrations might be needed to translate the common ranking model into an IRB PD, a PiT PD, a pricing parameter.

What are the challenges for using internal models? 

From a model management point of view, the two most important challenges are data management and data governance on the one hand and model change management on the other hand. Data governance and data management is key and becoming even more important. This probably doesn’t need further explanation. From a model change management point of view, the most important challenge in my opinion is the different time horizons for model changes. It is clear that for acquisition or for loan pricing model changes should be implemented as soon as possible; whereas on the other end of the spectrum it typically takes 12-18 months (if not more) to have an IRB model change approved by the regulator.

Could you elaborate on the implications of IFRS9 for financial stability? 

It is clear that IFRS9 will increase the coverage rates of the banks. The change from an incurred loss model to a forward looking expected loss model is in my opinion a very good evolution and the right way forward. It allows banks to properly take provisions or book impairments for anticipated credit losses rather than having to wait for impairment triggers. However, it comes with a downside as well. The IFRS9 LECL requirements will create more P&L volatility. It is key that the wanted volatility, ie linked to a real deterioration of creditworthiness, is preserved, but the unwanted volatility is minimized. The staging and the multiple economic scenario approach are still judgemental areas and might result in some unwanted volatility. It is hard to see to which extent those two features will indeed contribute to comparable and transparent results. Probably, a kind of convergence is needed in 2018-2019.

What would you like to achieve by attending the 3rd Edition IRB Models, the Standardised Approach for Credit Risk, and Capital Floors Conference?

Conferences are generally very useful to keep in touch with the industry, exchange ideas with peers and get updates on the hot topics. So, I am looking forward to interesting presentations and active discussions especially on the link between IRB models and IFRS 9. 

Ahead of the 3rd Edition IRB Models, the Standardised Approach for Credit Risk, and Capital Floors Conference, we spoke with Pieter Desmedt, Head of Retail Credit Risk, AXA Bank Belgium, about the benefits of using IRB models as a risk management tool.

Practical Insights From:
Addiko Bank
AXA Bank Belgium
Banco de Portugal
Bank of Ireland
Danske Bank
European Banking Authority
FBN Bank
Global Credit Data
Jyske Bank
Nova KBM
Oesterreichische Nationalbank
Prudential Regulatory Authority
Shawbrook Bank
UIster Bank
UK Finance


About the Conference:

This marcus evans event will enable banks to discuss the latest updates from the Basel committee and how banks are developing or amending their credit risk models in order to comply with the requirements, as well as looking in detail at both the benefits and the issues with the use of standardised models, and the lessons that can be learned from the ECB’s TRIM exercise. The event will also look at how a bank’s credit risk modelling can be streamlined by looking at IRB models, IFRS9 credit risk modelling and the ICAAP in an integrated approach, in addition to discussing remaining key issues before the deadline for the IFRS9 implementation in January. By discussing the issues in the credit risk modelling area today, individual firms, as well as the industry as a whole, will be able to proceed in the knowledge that everyone is moving forward together.
3rd Edition IRB Models, the Standardised Approach for Credit Risk, and Capital Floors Conference will take place from the 27th until the 28th of November 2017 in London, UK.

Copyright © 2017 Marcus Evans. All rights reserved.

Previous Attendees Include: 

Alpha Bank
Bank of Ireland
BNP Paribas
Credit Suisse
Danske Bank
Metro Bank
Morgan Stanley
Nedbank Sydbank

About the speaker:

Pieter Desmedt is currently retail credit risk manager in Axa Bank Belgium, the banking subsidiary of the global insurance group Axa. In this role he is leading the credit risk modelling team managing the IRB models and he is responsible for the impairment measurement under IFRS9. Before moving to Axa Bank Belgium in Februari 2016, Pieter worked 13 years for KBC Group of which 7 years in the credit risk management department overseeing a number of quantitative credit risk management activities like stress testing, economic capital and credit portfolio modelling. From mid 2012 until January 2016, he assumed management positions in KBC Consumer Finance. Pieter Desmedt holds a PhD in pure mathematics from the University of Leuven (Belgium).

The benefits of using IRB models as a risk management tool


An interview with Pieter Desmedt, Head of Retail Credit Risk, AXA Bank Belgium

Pieter Desmedt, Head of Retail Credit Risk, AXA Bank Belgium

Fix the following errors: