A pending new requirement for how banks should write down the value of troubled loans is providing a ringside seat for those in capital markets who want to understand how or why accounting and auditing are becoming more difficult by the day.

Historical numbers fail to satisfy demands for more forward-looking information. Prescriptive rules are being replaced by more principles-based requirements that can be adapted to different business environments. That means estimates and forecasts are becoming more common, but they are clashing with demands for more precision. And those battles on the front lines are getting more heated.

In fact, it nearly came to blows recently, when members of the Financial Accounting Standards Board met with representatives of the Independent Community Bankers of America and the American Bankers Association to try to bridge the gap on their different views of how the new rule would play out in practice.

“If you move forward with this model as it is now, all these banks will be out of business,” said James Kendrick, vice president of accounting and capital policy for the ICBA, referring to thousands of community banks that will be affected by the new rule. “The entire rural area of America—it won’t be underbanked. It will be unbanked. You do not listen. You do not understand.”

FASB member Larry Smith rebuked Kendrick: “You really are not trying to understand what we require,” he said. “All you have to do to implement our model is consider historical losses, which you say you want to use as your starting point. Then look forward for a reasonable period of time that you think you can feel good about your predictions. Tweak that historical rate by X or Y, then go forward.”

“If you move forward with this model as it is now, all these banks will be out of business. The entire rural area of America -- it won’t be underbanked. It will be unbanked.”

James Kendrick, VP of Accounting & Capital Policy, ICBA

Current accounting rules require banks to recognize losses on their debt securities or loan portfolios only when a loss actually occurs, even if the bank knows things are heading south. The financial crisis, which some attribute to the current accounting model that prevents any visibility into signs of trouble on the horizon, inspired FASB to make some changes.

FASB developed the “current expected credit loss” model, or CECL model, to factor more forward-looking information into the allowances banks would record for loan losses. The new approach would require entities to record on their balance sheets the net carrying amount of their financial assets minus an allowance for credit losses.

That forward-looking concept is not a concern for banks, even smaller community banks that are up in arms over the CECL model. “We are in agreement on this point,” said Tim Zimmerman, president and CEO of Standard Bank in Monroeville, Pa., near Pittsburgh. “I want to be able to anticipate. We’re OK with the rule being changed to allow us to anticipate. What we don’t want is a real prescriptive method that this is how you have to do it and this is the only thing that will be acceptable.”

Joanne Wakim, chief accountant for the Federal Reserve, tried to assuage Zimmerman’s fears. “The Fed feels the same way in terms of being committed to CECL being operationalized in a way that is simple and practical for community bankers.” The Fed is messaging to its examiners: “You should expect to see CECL implemented differently at different banks,” she said.

Greg Ohlendorf, president & CEO of First Community Bank and Trust, isn’t convinced it will work in practice as FASB and banking regulators say they envision. A regulator examining big banks with sophisticated data-driven modeling is sure to push that expectation onto smaller banks, he said. And auditors will press for heavy documentation of assumptions and estimates. “If you haven’t been in a community bank auditor exam recently, when you get into that room, the party changes,” he said. “I don’t want to be the guy in here just with my hair on fire. ... I know what happens when it leaves the printing press and goes into the field.”


Below is a summary of the Fed’s view of the CECL Model and Supervisory Approach.
Key Elements of the Proposed Standard
• Allowances are to be based on CECL model
– CECL is applicable to loans and debt instruments held at amortized cost, as well as receivables, lease receivables, and loan commitments
• Expected credit losses are defined in the exposure draft as “current estimate of all contractual cash flows not expected to be collected.”
– No triggers, no thresholds
• Quicker recognition of losses is an expected outcome
– Changes in allowance balances reflect changes in credit quality and flow through bank earnings
Measurement of Expected Credit Losses
A current estimate of all contractual cash flows not expected to be collected should incorporate:
–  Internally and externally available information
–  Information about past events, current conditions, and reasonable and supportable forecasts
–  Quantitative and qualitative factors specific to borrowers and the economic environment, including underwriting standards
·       Choice of methods include:
–  Loss-rate methods
–  Probability of default methods
–  Discounted cash flow methods
–  Roll-rate methods
–  Provision matrix method using loss factors
·       Any reasonable approach may be used, provided it reflects that some risk of default, however small, always exists and that zero allowance for loan and lease losses would be rare.
·       Entities should leverage current internal credit risk management approach and systems to measure expected credit loss.
Supervisory Approach
Institutions Are Encourage to:
–  Become familiar with the proposed changes
–  Involve all relevant business lines in preparation for the implementation of the CECL model
–  Discuss the proposed CECL model with industry peers and external auditors
–  Begin identifying and collecting actual loss data required for the implementation of the CECL model
Source: Federal Reserve Bank of St. Louis

FASB has not yet issued the final standard on credit impairment that will roll out the new CECL model, but the board currently expects issuance in the second quarter of 2016. The board determined late in 2015 that the standard will take effect in 2019, a year after an already-issued standard on how to recognize and measure financial instruments takes effect.

FASB has formed a Transition Resource Group that has already met to begin considering and help address some of the challenges banks will face in adopting the new approach. A similar TRG has addressed some 90 separate questions with the implementation of the revenue recognition standard, leading to a handful of changes to the final standard even before it takes effect. FASB member Tom Linsmeier said he anticipates the challenges now raised by community banks will be vetted through that TRG process.

After the dust settled on the heated dispute, Mike Gullette, vice president of accounting and financial management for the ABA, said he was disappointed in the discussion. “I think there is still no consensus on the level of work or complexity of work that needs to be done considering today’s audit environment,” he said. “The audit practice has changed a lot in the last three years. Companies may still feel a need to bring in models merely because they’re going to be asked to provide a lot of documentation to support assumptions.”

Sydney Garmong, a partner with audit firm Crowe Horwath, said it’s no secret that regulators are already concerned about the quality of audits under the current model. Inspection reports from the Public Company Accounting Oversight Board are full of findings that auditors failed to adequately audit inputs to allowances for loan losses. “It’s challenging today,” she said. “When you add the additional subjectivity into that with the new model, it’s simply going to make it harder.”

Regulatory and audit response to the new standard are legitimate concerns, says Mike Lundberg, a partner and director of financial institution services at audit firm RSM. “There is a clear commitment from regulators to make this scalable to community banks and credit unions, but there is a lot of uncertainty around what that scalability looks like,” he says.

Where banks follow a less complex equation to arrive at their expected loan loss, they generally will arrive at a bigger number that needs to be kept in reserve, Lundberg says. That worries banks as well. “There’s a belief that smaller institutions could be disadvantaged by a less complex model because it may result in higher reserves and more volatility in reserves,” says Faye Miller, another partner at RSM.

In addition to pressure from the banking community, FASB has also heard from more than 60 members of Congress who are calling on the board to revisit the model. Bankers are asking FASB to re-expose the standard for more public comment and add more examples to the standard to illustrate a simpler approach for simpler business models.

Raul Gupta, a partner at Grant Thornton who recently finished a fellowship at FASB where he worked on the impairment standard, says FASB is in a tough spot. “Many times the board has seen in the past when they give examples people read those examples as bright line rules,” he says.

Kendrick is hopeful FASB and its constituents can keep the dialogue going. “The concerns here are pretty extreme,” he says. “FASB has been open to listening to our bankers but I don’t know if they will be making any accommodations.”