What Keeps Community Banks up at Night
In October 2017, the Conference of State Bank Supervisors (CSBS) and the Federal Reserve jointly published “Community Banking in the 21st Century/2017,” profiling results of a recent national survey of more than 600 community banks, conducted by banking commissioners in 37 states. The community banks identified the considerable technology challenges underlying the industry’s future as one of the top issues they face. Survey findings revealed that respondents expressed concern over increasing competition in small business lending both from entirely new fintech entrants and from credit unions and larger banks.
When it came to regulatory relief, respondents were divided in whether the recent growth of compliance expenditures might soon subside. While implications of FASB’s radical change in credit loss accounting did not appear as a topic in the Introduction or the survey questions, it did appear as matter of concern in the notes of more than 50% of the 30 state commissioners who conducted follow-up interviews with survey participants.
Community Banks Express Concern Over CECL Burden in Recent Survey
In expressing their concerns about CECL, the community banks have much in common with larger banks. The scale and frequency of measurement that CECL will require -- all loans, large and small, commercial and otherwise, in most cases on a quarterly basis -- have led the consulting community to stress the need for comprehensive and early automation of all suitable CECL-related exercises. That means quantitative credit evaluation wherever it can take place.
CECL & The Model Question: Why You Should be Using Term PD
Larger banks are nearly unanimous in using Term PDs on their C&I books, although not always exclusively.
- Estimated Credit Loss = Probability of Default x Loss Given Default;
- or, ECL = PD x LGD . *
Considered to be the most robust approach available – Probability of Default best permits banks of whatever size – banks that can arrive confidently at PDs -- to incorporate the changing risk profiles of their individual corporate obligors over time. Such banks much prefer this approach to the simple discounted cash flow (DCF) method as it is commonly deployed among small and medium-sized banks.
In simple DCF exercises the lender typically applies the contractual rate of the loan continually, although it may be subject to increase in the face of marked deterioration in the obligor’s circumstances. Such exercises are therefore static in almost all cases when the policy goal of CECL is the dynamic expression of the lender’s and the industry’s up-to-date vulnerabilities. However, DCF can be made dynamic if the lender adopts a rigorous discount rate regime that is fully reflective of present conditions in the C&I book, a regime informed by point-in-time calibrations of PD.
*Here LGD is a net dollar amount. The formula can also be expressed as: ECL = PD x LGD x EAD, where LGD is a rate of loss and EAD (Exposure at Default) is the gross dollar amount of the loan.
The Challenge for Arriving at Probability of Default
Most community banks have limited model inventory for arriving at Probability of Default. On the other hand, most community banks have no trouble in arriving at the second factor, LGD, which will be a matter of discrete evaluation of the combination of business and business-owner collateral. The third factor, EAD, will be self-evident. Therefore, if community banks determine that well-informed PDs from a third party are available on affordable terms and meet CECL’s “reasonable and supportable” threshold, then they can consider adopting the PD x LGD x EAD formula as either a champion or a challenger model.
Automation is Key for Tackling CECL’s Massive Scale
The sheer volume of data ingesting – the heavy lifting that CECL will require -- is sure to put a large strain on any bank’s resources. It will be impossible to keep up with recurring CECL reporting obligations without leveraging technology aggressively. That said, many community bankers see technology as an obligatory and expensive challenge. One way to lighten the CECL workload is through API (application programming interface) facilitation, providing real-time access to critical data. It is inexpensive and serves the overall goal of automation well. RapidRatings is accustomed to ingesting data by a variety of means, so that it can serve almost any depository surveilling or originating commercial portfolios.
CECL calls for frequent reporting on loan pools grouped according to the lender’s choice of similar risk characteristics: vintage, tenor, industry, geography, creditworthiness, pricing, etc. FASB intends pool reporting to serve the many lenders who will find it much easier to make the required judgments of “more than insignificant” loan deterioration on a top-down basis. RapidRatings plans to deliver its ratings production in the subscriber’s choice of presentation – by pool or by individual company or both. However, any pool presentation of ours will always be a bottom-up exercise – rating the component companies individually before rolling them up into their pools. The approach will assure a precision that is unavailable from top-down generalizations.
As Survey respondents make clear, community banks will need all the help they can get to maintain leadership in small business loans in coming years. With careful CECL planning and leveraging the right tools, however, it doesn’t need to be another major compliance burden on community banks.