In the wake of the financial crisis, the two Titans that create accounting standards tried to hammer out agreement over how to account for impairment of loans. These bodies are the Financial Accounting Standards Board (FASB) and the International Accounting Standards Board (IASB).
“FASB and IASB share the same goal but unfortunately were not able to agree on the same standard,” said Emil Lopez, Director of Risk Measurement at Moody’s Analytics. FASB calls their impairment standard “current expected credit losses” (CECL) whereas IASB deals with impairment and expected losses in their International Financial Reporting Standards as IFRS 9.
Lopez was the third of three panellists at the webinar “The Long Road to CECL” sponsored by the Global Association of Risk Professionals on September 8, 2016. He described the main differences of the impairment frameworks proposed by the two organizations.
Both standards now refer to “credit loss estimates that reflect historical, current, and forward-looking information,” he said. They both require “reasonable and supportable forecast.”
However, the horizon used for estimating loss is different. The CECL looks at the entire “life of loan” whereas IFRS 9 has a 12-month forward look for performing loans. The net result is that CECL will estimate greater losses, since it’s over a longer time span, and will be less volatile, since loans will have more time in which to turn around.
Another difference in this clash of titans is that IFRS 9 makes use of unbiased, probability-weighted scenarios. It does not require allowance for purchased credit-impaired (PCI) assets, which in the case of CECL, is done using a gross-up method.
The timelines are different. “IFRS 9 is coming on line much sooner than CECL,” he noted, with an implementation deadline of 2018 compared to 2020-2021 for CECL.
Lopez summarized methods for estimating expected credit losses: discounted cash flow (DCF); loss rate; roll rate; vintage analysis; and the probability of default (PD).
The probability of default (PD) method is very common for commercial firms, and is considered “best practice” since it can be quite granular, if data are available to support it. It requires determination of loss given default (LGD).
“CECL may change the role of qualitative overlays,” said Lopez. The old way of calculating the allowance for loan and lease losses (ALLL) did not take into account forward-looking information about losses.
“There is no doubt that the qualitative overlay will be an important tool,” he said.
Since the Dodd-Frank Act Stress Test (DFAST) and Comprehensive Capital Analysis and Review (CCAR) regulations also require forecasts, there might be some means of incorporating linking quantitatively the forward-looking information used by all three determinations.
However, Lopez advised to proceed with caution. Stress models are optimized for stress scenarios. Any bias in stress testing would have to be modified for CECL. “A separate model shocks the ordinary model; for example, how a recession affects the PDs,” he said, referring to the probability of default, “or it might shock the underlying drivers such as changes to leverage or profitability.”
Credit quality of receivables and allowances should be broken down by class and segment; by age; and also by vintage. “Vintage is not a new concept,” he said, but making the information available to investors is. ª