Accounting Under FASB's Credit Loss Model: Too Much of an Audit Challenge?


When the Financial Accounting Standards Board began to flesh out what became a fresh acronym in the halls of U.S. accounting's chief rulemakers – FASB's "current expected credit loss," or CECL, model for reporting of loan and other credit losses—the board and its staff gave reasons for discarding an impairment approach that had been painstakingly crafted in meeting after meeting with the International Accounting Standards Board.

That decision to table the jointly drafted credit loss accounting model was not easy, FASB members acknowledged.

Board members such as Daryl Buck and a staff accountant stressed the need to devise an impairment blueprint that is "understandable, operable and auditable." Those three words have become almost a mantra in the project, one of the most important rule-making efforts on the board's agenda today.

The so-called "three-bucket" impairment model worked out by the Norwalk, Conn., board and the London-based IASB had led to "significant concerns" among constituents that it would not be "understandable, operable or auditable," as noted in the issue of the occasional publication "FASB in Focus" posted Dec. 20, when the U.S. panel released its revised proposal on impairment (09 APPR 13, 1/4/13).

However, in early 2013, two accountants who are well known in rulemaking circles—Ed Trott and Jack Ciesielski—have raised questions about whether the FASB's proposed expected credit losses model would present difficult auditing challenges. Trott is a former member of FASB and a retired partner at KPMG, with more than 30 years under his belt at the firm. Ciesielski, who has served on FASB's Emerging Issues Task Force and various advisory panels, is a Baltimore security analyst and publisher of The Analyst's Accounting Observer.

During a panel discussion on auditing at a Feb. 13 corporate governance workshop of the Practising Law Institute in New York, Ciesielski had a fleeting exchange on the impairment topic with James Kroeker, a Deloitte & Touche partner fresh from a stint as chief accountant at the Securities and Exchange Commission. Ciesielski suggested that the two talk off-line about how to audit the FASB's current expect credit loss model for impairment.

In the Feb. 11 edition of the AAO, Ciesielski, president of R.G. Associates, Inc., offered detailed analysis of the CECL model. In that analysis, he addressed provisions on how a credit loss allowance is estimated. That could take into account "information about past events," as Ciesielski wrote, as well as "reasonable and supportable forecasts and their implications for expected credit losses," to quote the FASB proposal of last December.

Managers thus can depart from using information about current conditions and "exercise a greater degree of judgment about the impact of future economic conditions" in figuring the allowance for credit losses, the Baltimore analyst-accountant suggested.

"Slippery Thing to Audit."

"This has particularly negative implications for auditors, because it will be a very slippery thing to audit," Ciesielski continued. "What may be `a reasonable and supportable forecast' by management should not always be acceptable to auditors. As subjective as forecasts can be, it may all come down to a particular world view held by managers and auditors—which may or may not coincide."

The robustness of prescribed disclosures would hold the key to avoiding potential `cookie jar' reserving and other earnings management issues that can arise when such judgments are called for in such accounting prescriptions, Ciesielski suggested.

Fair Value Versus `ECL.'

For his part, Ed Trott studied the FASB proposal on impairment and concluded that it came up short in a comparison with more of a fair value-based recognition and measurement model. The U.S. board, which retreated from a fair value-based proposal on classifying and measuring instruments that it floated in 2010, does not show any indication of going back to such an approach.

During his almost eight years on FASB, Trott figuratively made a journey, from being very skeptical about fair value for financial instruments to looking much more favorably on that basis of accounting.

In his analysis, contained in a Jan. 15 comment letter to the board where he sat for almost eight years, until mid-2007, Trott wrote about "verifiability." He tied that to the auditability issue with regard to what he called the expected credit loss, or ECL, model.

"I do not believe entity-specific ECL is as auditable as a [fair value] measurement," Trott wrote. "I question how an auditor with financial assets of the same credit quality and subject to the same future events will be able to determine that the ECL of the individual clients are in accordance with GAAP if the entity-specific forecasts of the future are significantly different."

Trott added: "How will the auditor determine that the significantly different forecasts are both based on relevant financial information about past events, including historical loss experience with similar assets, current conditions, and reasonable and supportable forecasts that affect the expected collectability of the assets `remaining cash flows'?"

"Auditing FV measurements is a challenge," Trott wrote. "However, such measurements are made based on much more observable information than an entity-specific forecast of the future."

At FASB, a spokeswoman declined direct comment March 21 when asked to respond to the auditability questions raised by Ciesielski and Trott. She said the board is in the listening phase in the due process on the Dec. 2012 proposal and is receiving comment letters on it. She suggested that such issues or related ones might be addressed in an FAQ on the impairment model that was to be posted at FASB's Website about a week from the date of the brief interview.

By Steve Burkholder
Bloomberg BNA staff correspondent