Objective and CRISIL's solution

• A leading global bank engaged CRISIL to evaluate, test and document more than 50 interest rates and FX pricing models. The goal was to prepare the models for validation in line with the internally approved SR 11-07 compliant template
• Models: Linear, Xccy, VoI and Curve, Callable, Quanto 
• Platforms/Tools: P&L Analysis Tools, Model Monitoring Framework, Python-based Unt Tests, Regression Testing Suite and Stability Testing 
• CRISIL GR&A set up a global, multi-geography 14-member team in a short time and used a phased approach to understand the deliverables and deliver complete, validation-ready and SR 11-07 compliant model documentation
• In the first phase, a three-member team was assembled at the client location in London, and the team spent six weeks working closely with the bank’s Quants team in understanding their analytics library, approach to documentation and aligning expectations on intensity of testing
• The team members then dispersed to different geographies (India, London, and New York) and set up larger teams in respective geographies, thereby creating a scalable operating model 
• The phased approach ensured that all learnings were transferred to other members within the team, and CRISIL was able to quickly scale up and transfer quantitative understanding of various interest rates models
• CRISIL team also added additional value by developing re-usable scripts and tools to assist with backtesting, stability testing and benchmarking of models

Client impact

• CRISIL team was able to deliver more than 56 models in six months, which was 10-15% more than promised. The client was able to meet deadline committed to regulators on preparing model documents for validation
  

Questions

Looking for high-end research and risk services? Reach out to us at: