Financial Institutions (FIs) have accounted for loan losses the same way for 40 years by using the “incurred loss” model. In the near future (2019 or 2020), FIs will change their loan loss accounting methodology to an “expected loss” model.
The incurred loss model allowed an FI to write-off a loan as it became known to the institution that the loan would not pay off. An FI also relied on historical information about a portfolio of loans to apply reserves against loan losses. Under the Current Expected Credit Loss Model (CECL), the financial institution will need to understand historical, current, and future conditions when accounting for loan losses. In addition to this, the rule requires an FI to preemptively reserve money against a loan, or portfolio of loans.
Since the idea of CECL was originally proposed by the Financial Accounting Standards Board (FASB), there’s been numerous articles written about it. Discussions with clients, and articles written by bankers suggest that most financial institutions are against the idea of changing the current accounting process. However, regulators support the switch believing that the accounting change will prevent many of the loan loss problems encountered during the downturn. As someone who’s worked in this industry for many years, I’m skeptical that this accounting change will reduce the risks of a catastrophic event to the extent the FASB, and others, believe it will.
When the concept was first introduced, industry analysts suggested that upon implementation, equity capital would take a huge hit. Estimates at the time suggested that capital ratios at many institutions would be reduced by 50 percent or more, as the result of the large amount of money booked against loan loss reserves. However, more recent articles such as this one from S&P Global Market Intelligence, suggest that improvements in the economy place the capital hits in the 20 to 30 percent range. Still, this is a huge number for many institutions to try and swallow.
To help with this reduction, the Federal Reserve and other regulatory agencies proposed a three-year phasing-in approach to help mitigate the capital impairment expected with CECL implementation. All of the industry regulatory groups are in favor of this extension, and as of the writing of this article, the Fed is still in the “waiting for comments” period. Once the comments are evaluated, the Fed should announce its final decisions on this potential extension.
The change will likely impact how FIs approach lending in the future. For instance, some institutions may look at multi-year credit lines and move those to shorter terms, such as revolving on an annual basis. Loans less than one year in duration may be easier to measure and reserve for. FIs may also re-evaluate certain loan portfolio products if they determine the risk of loss is higher. This could force some borrowers to seek other funding sources that may be more expensive in the long run.
Discussions with clients suggest that some FIs are considering changing some of their portfolios. Clients that have extensive credit lines that extend out several years are considering changing those to one year renewals to make it easier to manage their loss exposures. Still other clients indicated that CECL will not alter their lending behavior other than perhaps creating a bit more vigilance around exposure limits. And a few clients indicated that they will not make changes until after they see impacts to their loss reserves based on the new calculations.
A safer financial services industry is good for the economy as a whole, however, reserving for potential losses earlier may not prevent a catastrophic meltdown, it will depend on scenario building and understanding the best way to manage this in the long term. Most FIs are not happy with this accounting change, but it will inevitably occur with numerous rules and regulations that are unpopular amongst financial institutions.
Time will tell whether this change will do more harm than good. As with any change in process or regulation, there may be unintended consequences. Unfortunately, there is no mathematical, or other model to predict what these may be.