Within days before leaving for a family vacation out west this summer, we found out certain areas were closed due to flooding and were probably going to be closed for the season. Fortunately, the west side of Yellowstone National Park opened up the day we arrived, but Yellowstone’s popular north side attraction where bears were most commonly seen remained closed. However, on our drive away from Yellowstone, we spotted a bear, the highlight of our trip!
It was not only about being grateful for family time out west, but it was also about seeing nature’s beauty and wildlife and looking beyond what was clearly visible, contributing to making our vacation more enriching. Like the enriching experience we had out west, our ALM consultants don’t just brush the surface in our independent model validation engagements. We do a deeper look in to understanding your Bank’s ALM process and procedures and specific model assumptions.
Model risk is inherent in ALM models. Regulatory guidance for model risk management was released over a decade ago (2011) and expectations continue to grow. ALM model inputs, particularly the integrity of the data and the model assumptions, are what result in output for the ALCO’s informed decision-making process. So understanding the impact of reported interest rate risk from the Bank’s underlying data and assumptions is an important step of the Bank’s overall ALM process.
As model risk is ongoing, we have been observing some banks transitioning to more robust ALM models and outsourced servicing. Understanding the intricacies of your Bank’s ALM model and model assumption inputs, will help to give you confidence in your ALM model and servicer selection. Economic uncertainty continues and largely why generating alternative scenarios for changing market interest rates and sensitivity testing of key model assumptions play a pivotal role in the Bank’s financial success. Key model assumptions alluded to are loan and amortizing security prepayments, security call optionality, and non-maturity deposit rate sensitivities (betas) and decay lives.
Here are some best practice considerations in model assumption analyses:
- Ensure you have the proper chart of account breakout in the ALM model based on the attributes of balance sheet accounts (i.e. fixed- versus variable-rate, maturity/amortization terms, pricing strategies, growth, etc.) to sufficiently capture the repricing opportunities for model simulations.
- Spend time analyzing model assumptions for significant balance accounts as smaller balance accounts will likely have an immaterial impact interest rate risk results.
- Determine whether accounts analyzed in bank-specific assumption studies should be segregated further to better capture customers’ behaviors.For instance, segregate by account type versus general ledger account.
- Remove outliers from historical assumption studies based on quantitative data and apply qualitative factors as appropriate.
- Generate before and after analyses from key model assumption changes.This will help show the interest rate risk impact from each assumption change.
- Ensure model assumptions and changes impacting interest rate risk results, along with assumption methodologies, are well-documented for ALCO and Board meeting discussions and approval.
We have a team of ALM experts here at Wipfli. We are more than your independent auditor. We are your trusted advisor.