[Quant Note] REGULATORY REQUIREMENTS: GOVERNANCE AND MODELING OF CREDIT RISK PARAMETERS

2020 – June 25th

Author : Quant Practice – Quanteam

Today, the governance of credit risk models and the estimation of credit risk parameters (Probability of Default, Loss Given Default and Expected Loss Best Estimate) are subject to an ever growing supervision from the banking supervisory authorities. This is why we will analyze in this note the regulatory requirements in terms of governance and estimation of credit risk parameters.

The note is mainly based on the EBA guidelines (EBA/GL/2017/161 and EBA/GL/2019/032) and on the TRIM guide (Targeted Review of Internal Models).

These guidelines, coming into effect in January 2021, detail the requirements on the estimation of credit risk parameters. The TRIM guide explains how the supervisory authorities interpret the European regulations on the internal models of risk assessment (credit, market and counterparty) and on the model’s governance. The TRIM exercise took place from 2015 to 2019 and aimed to:

– ensure the credibility of internal models and confirm their adequacy to the calculation of regulatory own funds;

– ensure the regulatory compliance of the models and reduce the variability of the RWA (Risk-Weighted Assets) due to inappropriate
modeling.

The aim of this note is to analyze the regulatory requirements on the credit risk in order to show the fundamental notions of the modeling to which the supervisors are now very attentive.

 

¹ EBA Consultation paper on Guidelines on PD estimation LGD estimation and the treatment of defaulted exposures, 14 November 2016.

2 EBA Guidelines for the estimation of LGD appropriate for an economic downturn (‘Downturn LGD estimation’), 6 March 2019.

Download the White Paper

Quant Note #1 - REGULATORY REQUIREMENTS: GOVERNANCE AND MODELING OF CREDIT RISK PARAMETERS