The Washington PostDemocracy Dies in Darkness

Banks’ Early Caution Could Pay Off in Late 2023

JPMorgan Chase & Co. headquarters in New York, US, on Wednesday, Jan. 18, 2023. JPMorgan Chase & Co., the biggest US bank, said this year’s net interest income will be lower than analysts expected as the economy shows signs of slippage. (Photographer: Bloomberg/Bloomberg)

As corporate America’s fourth-quarter financial results continue to roll in, investors are searching those reports for clues to how a possible recession will affect profits in 2023. For one industry at least — financial services — they may need to look back further than that.

Regulatory and accounting changes in recent years mean that the rising unemployment rate anticipated by the Federal Reserve has already been incorporated in the earnings that banks and other financial companies reported for last year, not just in their forecasts for 2023. And while the Fed’s estimates require the financial industry to be cautious for now, they might also end up providing a boost to the economy later this year if that big jump in the unemployment rate doesn’t materialize.

The accounting change is known as CECL, or current expected credit losses, which was a response to shortcomings during the 2008 financial crisis and went into effect for public companies at the end of 2019. Prior to the financial crisis, banks set aside money to cover losses on their assets based on how their portfolio was actually performing. Rising delinquencies and charge-offs would hurt earnings. 

The problem that CECL seeks to address is when that real-time data understates future losses based on an unfolding or anticipated negative economic shock. Under CECL, banks and other companies with loans or debt holdings need to set aside money for losses based on a view of the future, even before the negative scenario has materialized. If it’s sunny today but there’s a hurricane forecast for next week, expected losses from the hurricane need to be taken today.

The future unemployment rate is a key input in any financial company’s scenario analysis. The question is what the unemployment forecast should be for 2023 in order to set aside money today for anticipated losses stemming from a higher rate. An added wrinkle is that other post-2008 regulatory changes require banks and other publicly traded financial services companies to pass annual Federal Reserve stress tests to get permission from the Fed to return capital to shareholders in the form of stock buybacks and dividends.

So it’s perhaps not surprising that the Fed’s forecast for the economy helps establish the base case for the financial services industry. There is no point, and arguably no benefit, in having an aggressively optimistic view on the economy if you ultimately need the Fed to sign off on your capital-return plans.

We’ve seen this so far across multiple financial companies in their fourth-quarter earnings reports. JPMorgan Chase & Co. set aside an additional $1.4 billion — or roughly 50 cents a share —  “driven by updates to the firm’s macroeconomic outlook, which now reflects a mild recession in the central case,” according to the company’s chief financial officer. Its forecast has the unemployment rate rising to 4.9%, slightly higher than the 4.6% projected by the Federal Reserve at its December meeting.

Two other financial services companies that depend even more on consumer credit for their profits reported similar unemployment forecasts for 2023. Discover Financial Services projected its net charge-off rate would more than double in 2023 to more than 3.9%. As its CFO explained, that’s in part due to its assumption that the unemployment rate would rise this year to between 4.5% and 6%. Ally Financial Inc., which provides auto financing for consumers, projected the unemployment rate would approach 5% by the end of 2023, and because of that it expects to be building its capital buffer this year rather than buying back shares.

What’s noteworthy is that even though these companies set aside reserves in the fourth quarter based on similar unemployment forecasts, the markets didn’t seem to really buy it. Discover’s stock, despite the elevated charge-off forecast, closed higher on Friday than it did the day before it reported earnings. Ally’s stock surged 20% the day after its earnings report.

That shows that even as the Fed forecasts an end-of-year unemployment rate that’s 1.1 percentage points higher than current levels, and even though banks have set aside money to cover losses for that scenario, markets are skeptical anything like that will actually materialize.

Should markets end up being correct, the likely outcome is that financial companies will have to release some of the reserves that they set aside as losses this past quarter, counting them as profit later this year. That’s one of the dynamics with CECL — sometimes pessimistic forecasts don’t come true, and it means banks get to report profit when expected loan losses don’t end up happening.

That boost to capital levels could then be lent out, invested or used as buybacks to fuel growth for the companies and the overall economy as a soft landing is unfolding. The longer the labor market remains resilient, the greater the chance that banks have excess capital to put to work in the back half of the year.

More From Other Writers at Bloomberg Opinion:

• No, Banks Aren’t Stronger Than They Need to Be: Editorial

• Strange Times for JPMorgan, Other Big US Banks: Paul J. Davies

• Don’t Get Disoriented by Recession-Talk Fatigue: Jonathan Levin

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

Conor Sen is a Bloomberg Opinion columnist. He is founder of Peachtree Creek Investments.

More stories like this are available on

©2023 Bloomberg L.P.