Many financial institution planning discussions over the past year have focused on FASB’s decision to move to the CECL standard as a replacement for the current accounting methodology of allowance for loan and lease losses (ALLL) and other than temporary impairment (OTTI). Entities filing with the SEC will be required to adopt CECL beginning in fiscal years following December 15, 2019. Public business entities (PBEs) will be required to adopt the standard in fiscal years beginning after December 15, 2020. Currently, non-public entities and credit unions do not need to implement CECL until the fiscal year following December 15, 2021. While SEC filers may already have their CECL solution determined, other depositories, specifically non-PBEs, may still be determining the steps needed to prepare for the new CECL standard.

1. Determine your data collection process
Data collection is one of the most important – and resource-constraining – tasks needed to develop a well-informed loss model. Creating an efficient data collection process can require significant human hours and potentially burdensome costs. Given the time and costs needed to create a data management system, it should be a given priority. An appropriate process should allow for data users to retrieve critical elements needed to analyze the probability a single loan will default and the resulting losses, if any, that may occur.

This requires the institution to pull updated credit fields, such as credit score, LTV ratio, back-end DTI ratio, debt service coverage ratio, etc. Ideally, details regarding type of collateral should be incorporated into any loan tape. Loans secured by real estate should have identifiable traits like property type and occupancy status flagged. C&I loans secured by inventory should identify the type of collateral. The feasibility to collect appropriate fields can play a significant role in the adopted loss model.

2. Consider the models available
Once an institution has developed the plans for data collection and has a reasonable estimate of the type and number of fields that can be gathered, the CECL model should be considered. Given the number of options for projecting CECL losses, the selection of the model should be less about cost and more determined by balance sheet complexity and intended use. FASB’s approved models provide institutions the flexibility to choose from simplistic to more complex processes.

Financial depositories with simplistic balance sheets or that do not have the ability to capture significant data should consider the simpler, and less costly, CECL models. The potential precision that comes along with a more complex model will be lost on those without appropriate data. Balance sheet complexity is often defined by an institution with esoteric loan products that do not fit the mold of the generic financial depository. Real estate loans for non-traditional single-family residences, single-purpose CRE, or C&I secured by collateral that does not have a readily available market value would require a more complex process to appropriately estimate losses.

3. Make decisions about intended use 
Your institution’s intended use should be another consideration in addition to the availability of appropriate data. For institutions looking at CECL as merely a check-the-box regulatory process, the institution should appropriately balance model complexity with auditor expectations. Some institutions may view CECL as just a procedure that will be required by auditors; however, it can be an opportunity to consistently assess balance sheet risk. Those looking at CECL simply as the former would not derive much benefit from a model that provides significant precision and is highly-tuned to market expectations. In such cases, it would may be more appropriate to verify auditor expectations and choose a model that is less cost prohibitive.

If your institution is looking to acquire more value out of the required process, a more complex approach would be the better choice. CECL is an opportunity to analyze the credit position of the balance sheet. With loss projections informed, an institution will be able to review concentration risks along with assuring that pricing is appropriate given credit costs. CECL can provide the start to a more extensive stress test on the balance sheet to highlight more nuanced risks that might not be discovered with a more simplified loss model.

The sooner you start planning, the better
Credit unions and other non-public business entities still have time to prepare for CECL. This extra time, however, should not be wasted as data initiatives may take significant effort and resources to accomplish. Given the importance of data availability in loss modeling, it can drive the decision regarding which type of model should be adopted. While the best resource for non-public business entities to use when deciding CECL models may be the auditor, remember that CECL can be more than a method to project ALLL, and such a process can greatly enhance the institution’s risk management.

Thomas Griswold, CFA

Director, Advisory Services at ALM First