PASSING GRADES AREN´T ENOUGH ON BANKS´ STRESS TEST / THE WALL STREET JOURNAL

Passing Grades Aren’t Enough on Banks’ Stress Test

Banks again proved their soundness in the Federal Reserve’s annual exam, but higher capital requirements may still squeeze them

By Telis Demos

Citigroup was among the largest global banks that saw bigger capital drawdowns versus last year’s stress test. / PHOTO: VICTOR J. BLUE/BLOOMBERG NEWS

Big banks don’t appear to be in any danger of going bust, according to the Federal Reserve’s latest stress test. 

But that doesn’t mean shareholders have nothing to worry about.

For investors, what often matters most about stress testing isn’t the big picture, but rather the fine details. 

Seemingly small moves in a bank’s capital requirements can have a big impact on their ability to return capital and push returns on equity higher. 

And this test showed that for some of the biggest banks, there are no easy answers right now.

In this year’s test, the examined big banks’ aggregate key capital ratio fell by as much as 2.7 percentage points in the most stressful economic scenario. 

Last year, that drop was 2.4 percentage points, and it was 2.6 percentage points in the special December 2020 pandemic test. 

The size of each bank’s dip helps determine their “stress capital buffer” requirement. 

Among the very largest global banks, Bank of America, Citigroup and JPMorgan Chase saw bigger capital drawdowns versus last year’s test, while Morgan Stanley and Goldman Sachs had slight decreases versus last year, causing their shares to rally sharply on Friday. 

Wells Fargo’s didn’t change.

What changed for banks wasn’t always what you might expect, like the rate of loan losses. 

Projected loan losses as a percentage of average balances actually fell on average across Bank of America, Citigroup and JPMorgan Chase versus the prior year’s test. 

However, their loan-loss provisions rose, partly reflecting the fact that banks have broadly released substantial reserves since the last test. 

That gave them a smaller cushion when it came to the Fed’s projection of provisioning.

Banks’ stressed capital drawdowns, plus four quarters of planned dividends, determine the buffers. 

Wolfe Research analysts estimate that Bank of America, Citigroup and JPMorgan Chase might see their minimums increase by roughly a percentage point. 

But banks can’t fully lay out their buffers until Monday evening, when they can also announce their capital-return plans.

These potential stress-test capital minimum increases come on top of already-rising global systemic risk buffers for the very biggest banks, reflecting their growth in size since the pandemic. 

To get ahead of these rising minimums, banks can raise equity, retain more earnings, or even deliberately shrink their balance sheets. 

All of that could be rough for shareholders.

There are some offsets to capital-ratio pressures. 

Big banks already anticipate some possible shrinking as fast-moving deposits come off their books in search of higher rates elsewhere. 

Banks can also possibly just run very close to, or even below, their minimum requirements.

Unlike under the prior rules banks don’t have to get their capital plans preapproved. 

Under the new regime, payouts become restricted on a sliding scale as banks dip further below their regulatory buffer requirements. 

Since realistically no bank will be eager to curtail dividends, what comes into focus is the degree of planned share buybacks.

An investor might decide they are comfortable with a bank running very close to or even below their minimum, arguing it won’t interfere too much with payouts and preferring it to the alternatives. 

But beyond a bank’s ability to just survive a recessionary scenario, there is the question of how much it can lean into potential opportunities, like by expanding its balance sheet to soak up lending or trading flow.

This leaves investors in a tough spot: Smaller payouts mean less money in shareholders’ pockets today in an uncertain market, but they also might preserve longer-term potential. 

It is a multiple-choice question where none of the answers might feel totally right.