Asset/Liability Model Validation and Verification
If your institution currently uses either a bond dealer sponsored or outsourced ALM solution, R2Metrics can perform various services to improve the model’s accuracy and ensure regulatory compliance of model assumptions.
Level 1 Validation
- Validation of the Asset/Liability report information by comparing category balances for the ALM report to the Financial Institution’s general ledger.
- Review the chart of accounts and underlying integrity of data used in the primary report.
- A comparison of bond portfolio market value sensitivity estimates to ALM report forecasts.
- A review of all important assumptions underlying model results – including loan betas and applicable prepayment speeds, as well as deposit betas, decay rates, and early redemption methodology. Compare current funding composition to historical periods of 2006/2007 to estimate migration from lower yielding non-maturity deposits into higher yielding CD’s in the event that interest rates increase.
- Back testing of the Asset/Liability report by comparing projections of asset yields and liability costs to model forecasts from one year ago.
- Verification of the Asset/Liability report’s compliance with the Financial Institution’s Asset/Liability Management Policy and review of the Policy to ensure that it is in compliance with the latest regulatory statements.
- Highlight potential modeling issues that may potentially impact overall model accuracy.
Level 2 Validation
- Verify accuracy of the ALM report by running parallel R2Metrics ALM model and comparing asset yield, liability cost and net interest margin sensitivity forecasts under different interest rate shock scenarios, as well as changes in market value of equity. R2Metrics will provide a summary opinion letter with supporting tables detailing modeling discrepancies as well as the likely causes of these discrepancies and suggestions for corrective action.
Assumption Stress Testing
- Once the base case ALM report is complete, R2Metrics will rerun the model in order to stress test key assumptions. Key assumptions include but are not limited to loan CPRs, loan and deposit betas, CD early redemption penalties, non-maturity deposit decay rates, and flattening the yield curve.