Model Risk Management

Our model validation and backtesting services support model risk management and compliance with the requirements arising out of Recommendations W and R of the Polish Financial Supervision Authority (KNF). Our support ensures the performance of periodic, annual validations and monitoring of models that are necessary to maintain or raise the high quality of the models used. In validating or backtesting the parameters of an ECL model, we also apply IFRS 9 and verify whether the standard’s guidelines are met as part of the ECL model.

An independent validation of a material model performed by an outside expert will also ensure that the guidelines of Recommendation W are fully complied with in terms of rotation/change of model validator and smoothness of transition, and will provide a different quality perspective of the tested models.

An audit of the model risk management process may be a comprehensive audit of the entire process or may constitute support for the internal audit function in selected elements of the process and in terms of quantity, and in particular in the measurement and assessment of model risk during the audit.

Validation of ECL model parameters in accordance with the IFRS 9 portfolio approach or selected elements thereof, including the Forward Looking Information approach:

Validation of PD, LGD, CCF, EAD, SICR and ECL parameters (qualitative and quantitative analyses) in the following areas:

Data quality analysis
Model operating efficiency assessment
Model algorithm form analysis
Implementation correctness analysis
Documentation quality
Model risk level assessment in accordance with Recommendation W

We can also perform a validation of credit risk scoring and rating models covering the same scope as for IFRS 9 models.

Validation of stable deposit base estimation models (core deposit models):

Data quality analysis
Model operating efficiency assessment
Model algorithm form analysis
Implementation correctness analysis
Documentation quality
Model risk level assessment in accordance with Recommendation W
Alternative model proposal

Backtesting of IFRS 9/scoring/rating models:

Data quality analysis
Population/segment/rating stability assessment
Segmentation adequacy
Backtest of model results

Backtesting of stable deposit base estimation models (core deposit models):

Data quality analysis
Deposit category/segment split adequacy and stability assessment
Backtest of model results

Audit of the model risk management process and verification of compliance with Recommendation W requirements

Verification of model risk management policies and procedures and their compliance with Recommendation W, analysis of the process and its elements
Concept analysis of methodology adopted to assess model materiality, exposure to risk and level of risk of material models and aggregate risk of models, verification of timeliness and compliance with materiality and risk level assessment procedures
Verification of model register and journals
Analysis of reporting process and model risk reports for adequacy of management information and timeliness of reporting to bank organs
Analysis of controls in the process of model risk management
Validation process analysis