CECL will strengthen, not hinder, financial system
Since the financial crisis struck in full force a decade ago, the nation’s financial system has been substantially reformed. Most of the reforms were fought hard by financial institutions, and while some of the reforms probably went too far and were counterproductive, for the most part they succeeded. The system is on much sounder financial ground.
Complaints are mounting again, this time over the next big reform facing the financial system — the adoption of current expected credit loss accounting. CECL, as it is called, is a sea change in how financial institutions account for losses on their loans and securities. The Financial Accounting Standards Board is requiring adoption of CECL by 2020. The complaints are misplaced: CECL will result in an even stronger financial system and economy.
Under current incurred loss accounting standards, financial institutions add to their loan losses only when they are certain of a loss on their loans and securities. This is typically 90 days after a borrower’s last loan payment. Under CECL, financial institutions must add to their loan losses when they make the loan, since there is an estimable probability that a loan will default at some point in its life. That probability can vary depending on the quality of the loan and the expected economic conditions during the loan’s life. If the economy is expected to be strong, then the probability will be lower, and vice versa.
The economic logic of moving from incurred loss accounting to CECL is that CECL is less procyclical. That is, in a recession when unemployment is rising quickly and borrowers fall short on their loan payments, banks must start adding more to their loss reserves, hurting profitability and capital. Banks have no choice but to tighten their underwriting standards, curtailing the availability of credit and adding to the economy’s woes.
This is precisely what happened a decade ago. Loan losses ballooned when the great recession struck, undermining bank capital and precipitating a credit crunch. This is a significant part of why the economic recovery from the downturn was so painfully slow.
Under CECL, banks would have increased their loan loss allowances during the housing bubble as underwriting dramatically eroded and subprime mortgage lending was surging. Under any expected future economic conditions, there would be more mortgage defaults. Loan losses under CECL would have risen, hurting profits and capital and forcing banks to become more cautious in their lending. There may have been a housing bubble anyway, but it certainly would not have been as big, and the subsequent bursting of the bubble not as catastrophic.
To test this logic, Moody’s simulated how most banks would likely implement CECL for a portfolio of Fannie Mae and Freddie Mac residential mortgage loans at different times leading up to, during, and after the crisis. We used our baseline forecasts for the economy at these times and our current model of the economy to produce upside and downside scenarios around this baseline. Mortgage defaults are modeled based on borrowers’ credit scores, loan-to-value and debt-to-income ratios.
No, we did not predict the great recession, although we gave plenty of warning of a high risk of a serious financial event and recession. Predicting downturns is all but impossible, and it would be a mistake to make any changes to accounting rules on the premise that economists will predict future recessions even roughly accurately. They won’t.
Despite this, our simulations show that CECL accounting results in meaningfully less procyclicality in loan losses, capital levels, underwriting, lending activity and the economy’s performance than incurred loss accounting. If CECL had been in place, say, at the end of 2004, the housing bubble would not have grown as large, and the housing bust would not have been as cataclysmic. And if CECL had been in place at the end of 2009 as the crisis was winding down, the subsequent credit crunch would have been less severe.
Banks are highly skeptical of these arguments. As with other regulatory changes implemented over the past decade, banks strongly object to this accounting change. The American Bankers Association has called for a delay in CECL’s implementation, and the Bank Policy Institute, a trade group for major banks, has been highly critical. To be sure, they have good reason to be nervous. The change is big and it impacts bottom lines. A change in loan losses directly affects the returns banks report to investors each quarter.
Moreover, banks reasonably expect that if required to reserve more upfront when making a loan, they should be given some capital relief by regulators. That is, if banks sock away more at the start of a loan in case it goes bad in the future, then their capital standards should reflect it. Bank regulators have acknowledged this, and seem set to make the appropriate adjustments.
Banks also make a good case for their being allowed to account for the future interest income on a loan at origination. The interest rate on the loan reflects the loan’s credit risk — the greater the risk the higher the loan rate and interest income. But if banks count the interest income, they should also count the interest expense to fund the loan. This gets very tricky to account for, and thus the reluctance of the Financial Accounting Standards Board to go down this road.
Financial institutions in much of the rest of the world adopted the international analogue of CECL this year to little fanfare. Other than a bit of agita at troubled Italian banks and Canadian banks with large mortgage portfolios, there was barely a ripple. The international standard differs from CECL, but not by enough to dilute the message that the new standard should be reasonably gracefully adopted. Highlighting the countercyclical aspects of the new accounting, a large British bank recently added to reserves in anticipation of fallout from a potential hard Brexit.
The U.S. banking system is as strong as it has ever been and as profitable. This, despite the substantial reforms imposed on the system since the crisis. Fears that these reforms would diminish the system have been misplaced. As will the fears over the fast-approaching adoption of CECL.