By Teri Grams, Abrigo Advisor
As we get closer to implementation of the new current expected credit loss accounting standard, or CECL, many of us are looking for as much information as we can to help in understanding everything that should be considered in putting together our CECL models.
CECL is best thought of as part of a holistic view that focuses on Enterprise Risk Management and not just credit risk alone. It is no longer advisable to operate in silos as you might have done in the past when credit loss reserves were calculated by the Credit Officer for the institution, while Asset/Liability forecasts were in the domain of the CFO. With the need to look at a more forecasted approach to determine credit reserves, it makes sense to leverage some of the assumptions that have been used in the past for developing loan assumptions in your ALM models. These assumptions include loan growth, the average life of loans, which takes into account prepayment assumptions, and loan loss forecasts.
To start, it is important to take into account from a capital plan perspective, what is it you want to achieve as an organization? The capital plan sets the direction for the entire organization to ensure all areas of the institution are on the same path and have realistic goals set. Once those goals are set, you will know what you need to achieve in asset growth and earnings growth in order to meet your capital goals. All areas must work together in order to achieve those goals. For example, if you are trying to maintain your capital level and want to grow assets, you will need a certain level of return on assets, or ROA. How loan loss reserves are determined needs to play into these goals. If you are forecasting loan growth in a pool with potentially high credit losses, how does that impact earnings? Setting loan pricing in a way that is attractive to your best credit borrowers may keep you from increasing loan rates, but on the other side of the balance sheet you may have to increase rates on deposits to keep funding these loans. This can quickly lead to earnings compression if not managed properly.
Having the proper information available in order to appropriately set loan prices by breaking out loan categories into pools with similar characteristics and with similar loss rates will benefit A/L models and CECL models. Breaking out loan details in a way that provides the same level of characteristics in both models that match will make it easier to use assumptions from one model to the other. One example is fine-tuning prepayment speeds to get appropriate cash flows in your A/L model and then using the same prepayment analysis for the same pool of loans in your CECL model. Prepayment analysis would go a long way to helping provide valuable cash flow data that impacts your A/L model and would also be helpful in calculations for credit losses. This information will help determine how different pools have typically been paying down and what losses might have incurred during the life of the loan.
Developing a usable CECL model should be at the top of your list of to-dos even if you are not scheduled to implement for a few years yet. Being prepared ahead of time and making sure you have the right data and developing or modifying your current credit loss models will go a long way toward making CECL implementation much easier.