Asset/liability management: Will your model pass regulatory scrutiny
The emphasis from various regulatory agencies on asset/liability management (ALM) continues to increase as interest rates trend upward. Are you prepared for this scrutiny?
While all banks are required to obtain an annual, independent validation of their interest rate risk model, there are several steps you can take to expedite the process and confirm your ALM model is on track to pass regulatory scrutiny.
All banks are required to obtain an annual, independent validation of their interest rate risk model.
- Document and maintain support logic for your ALM model assumptions, including a historical look back. You should document all logic used to derive key assumptions for your ALM model, including prepayment rates, beta rates, and deposit decay rates. It’s also important to make sure your model captures loan floors. This includes the potential delay in interest income in rising rate scenarios and the minimum expected interest income in declining rate scenarios.
- Conduct periodic modeling. If key assumptions prove to be different than anticipated, stress testing will indicate the impact to bank earnings under various scenarios. Additionally, it’s important to model for changing yield curves since the flattening, steepening, or twisting of the interest rate yield curve may have an impact on interest rate risk results.
You should also consider surge deposits in periodic modeling. A large increase in non-maturity deposit balances as a percent of total deposits, and what may happen if customers move out of these product types, is a major concern of regulators.
- Communicate regularly with your asset-liability committee (ALCO) and board of directors. To support appropriate oversight and reporting requirements, share your process and review your model logic with your ALCO and board of directors. This should include discussion of key modeling assumptions and a review of your model results compared to policy guidelines. Include interest rate risk results for both one- and two-year scenarios. Risks may be more evident in year two than in year one. Be sure to maintain documentation in all committee/board minutes.
- Perform an annual back testing of ALM model results. The comparison of model projections to actual results will help support the accuracy of your model.
Regulatory agencies continue to monitor ALM very closely. Making it a priority now will save you headaches later on.