Model Risk Management

The risk management landscape has evolved considerably over the years due to an increasing focus on regulatory requirements, especially after the financial crisis of 2008. Risk management (as a function) has since become an integral part of financial decision-making, in contrary to the earlier siloed approach. With the extended use of analytics to measure and monitor risks, use of statistical models has become crucial, leading to model risk management. So risk practitioners now need to ensure the efficacy of risk models while addressing growing regulatory compliance requirements of conducting regular model validations and reviews, and meeting use-test requirements.

The WNS Model Risk Management Centre of Excellence (CoE) helps clients with every aspect of the modeling ecosystem by addressing the imperatives of Basel guidelines, CCAR and D-FAST requirements, along with the local regulatory needs. The CoE leverages WNS’ proven capabilities as a BPM leader in the areas of analytics, technology, process re-engineering and project management to provide end-to-end solutions in banking, financial risk modeling, and model risk management. Our teams integrate with clients’ broader corporate information network to deliver on the critical mandates. We help clients ‘free up’ scarce onshore resources to focus on core business activities while engaging our offshore team to assist the client with other tactical tasks, including model development and performance analysis, model remediation and recalibration, and others

Services

Model Development:

• Development and testing of advanced statistical / econometric models and machine learning algorithms
• Development of benchmark models to test efficacy of existing models

Model Performance Analysis:

• Model validation services to validate the five central pillars of a model - theory, assumptions, data, algorithm and output
• Model stress testing in accordance with regulatory guidance / risk appetite

Model Remediation and Recalibration:

• Regular audits of model deficiencies based on performance reviews
• Model recalibration

Documentation:

• Reviews of model documentation
• Documentation of model components, assumptions and others in accordance with best practices

Automation: Automation of routine model review and validation tasks, such as

• Parameter stability
• Back testing
• Sensitivity analysis