Supporting ALM Prepayment Forecast Assumptions

Many institutions struggle with inconsistencies between ALM and CECL prepayment rates, often due to reliance on generic third-party data. This can lead to inaccurate risk assessments and regulatory challenges.

The best CECL analytics are built with

ARCSys offers a solution:

We utilize your institution’s actual loan-level prepayment history to generate highly accurate and customized prepayment forecasts.
We can customize your forecasts based on any loan types, not just CECL pools, as we calculate prepayments on every loan.
We analyze historical data to identify statistically significant relationships between prepayment rates and key interest rate factors.
Our forecasts seamlessly integrate with your CECL modeling requirements, ensuring consistency and improving the accuracy of your allowance for credit loss.

Our ALM Prepayment Forecast Assumptions Support Includes:

An initial and annual prepayment analysis to determine which interest rates shown have the strongest relationship to the prepayment rates for the portfolio.
A detailed report that includes each interest rate’s relationship with each pool’s prepayment rates, graphs of the forecasted prepayment rates for each pool, and documentation describing the statistical models used to find the relationships.
A monthly or quarterly prepayment forecast of each pool or loan type requested, utilizing the baseline forecast (static or dynamic) and a scenario plus and minus 1-3% for each statistically significant interest rate factor for each pool or loan type.
A detailed analysis of all modeling and statistical results.

ARCSys will collaborate with your institution to find statistically significant and directionally consistent relationships between interest rates and prepayment rates. If there is a particular interest rate you prefer, ARCSys will check to see how strong the relationship is between the selected interest rate and your prepayment rate. We will also investigate different periods to identify the strongest relationship that is also directionally consistent.