Independent
Validation
Comprehensive assurance of Model ecosystem in line with best practices.
Overview
Our ‘Independent Model Validation Index’ methodology is designed to support the local regulators in benchmarking the performance of models (all types of portfolios) of different banks in its jurisdiction during a year or over the years. We believe ‘Independent Validation’ is important for meaningful implementation of regulatory reforms and to increase the resilience & quality of risk models, as these models underpin the provisions (IFRS 9) & regulatory capital (Basel II/III), at the core of banking and regulatory reforms.
It has become even more important post implementation of IFRS 9 as provisions calculated by the internal models have direct impact on the profitability of the financial institution. To err on either side may increase the volatility of the earnings of the financial institution.
Our independent validation services are designed to consistently assess the performance of internal models to provide assurance to our clients about their efficacy and limitations. We have developed an integrated framework which assesses the model performance on both qualitative and quantitative parameters.
We use our proprietary model validation system designed to handle large volumes of data and ability to generate detailed model validation reports inclusive of data cleaning and preparation modules. Tool helps us to scale our validation services without compromising on the quality of the outcome.
Our validation index, which is constituted of quantitative and qualitative validation, is designed to monitor the performance of the internal models over the years using a consistent framework. This will help clients to understand the improvement in model performance over the years.
Robustness of ‘Quantitative Validation’ of the model is dependent on the availability of covariate & default data.
Even where sufficient data is available and standard model validation can be done with relative ease, the key challenge is to assess the reliability of the quantitative measures i.e. how does qualitative components of the overarching model ecosystem impact the performance of the model for it to be suitable for use in decision making process, such as processes, systems, governance and availability of adequately skilled resources for development/modification & maintenance of the models.
‘Validation Score Index’ is designed to incorporate the assessment of these integral qualitative components to arrive at a comprehensive ‘Validation Score’, which can be used within the client to benchmark different models/scorecards or across the clients by the regulators to benchmark the similar models of different banks.