Risk Forecasting Simulation Models

The Client

Global Financial Institution

The Challenge

A “top tier” financial institution had a large and diverse portfolio of loans on their books.  They wanted to value and assess the risk of their loan portfolios on a monthly basis using numerous economic scenarios.  In addition, the institution required these forecasting models for their CCAR (Comprehensive Capital Analysis & Review) submission to the Federal Reserve.  The forecasting models used a very advanced simulation based approach that required enormous amounts of computational power.  The computational power requirements of the model were causing numerous performance issues that prevented the institution from using these advanced models – the models would literally take weeks to run.

The Solution

  • Delivered an optimized solution utilizing Cognilytics Decision Analyzer and SAP HANA
  • Addressed major performance problems and demonstrated over 500x speedup on model execution compared to existing implementation
  • Instituted model management, repository, and auditing controls in order to satisfy regulatory requirements

The Approach

  • Implemented advanced simulation based credit risk models utilizing Decision Analyzer, Hadoop, and SAP HANA
  • Developed a centralized model management and control environment
  • Built advanced self-service business intelligence and risk reporting platform utilizing SAP HANA, an in-memory database platform