The Federal Reserve determined that the 31 banks subject to this year's annual stress test have sufficient capital to withstand a significant recession. Under the severely adverse scenario, the Fed projects that the banks' aggregate common equity Tier 1 (CET1) capital ratio would decline by 280 basis points (bps), to 9.9% (before rising to 10.4% in the final quarter of the stress test) from 12.7% (see table 1). While this marks a greater decline than last year's 250 bps, it is within the range of recent stress tests.

Table 1

We estimate that the stress capital buffer (SCB) increased for 12 of the 23 banks with U.S.-domiciled parents included in the stress test. These 12 banks will have higher minimum capital ratios starting in the third quarter of 2024 (Oct. 1). The Fed usually responds to a bank's request for reconsideration of its stress test results within 30 days, at which point it will publish the finalized results for all banks.

Chart 1

We believe the final Basel III endgame proposal will play a bigger role than the stress test results in how much capital banks need to hold. Based on statements from regulators, we expect the final Basel III rules to be less stringent than the original proposal, which could prompt most banks to increase their shareholder payouts this year.

Share repurchases have been relatively low the last two years, probably partly in anticipation of tighter capital rules and partly amid concerns about market conditions, the economy, and unrealized losses. We estimate that the share repurchases of the 23 banks with U.S.-based parents subject to the Fed's stress test this year equated to about a third of their earnings in aggregate last year, little changed from 2022 and materially down from 2021 and the years in the lead-up to the pandemic.

Roughly half of those 23 banks announced increases in their dividends following the stress test results, including all eight global systemically important banks (GSIBs). Other banks indicated they would maintain their dividends or made no announcement.

That said, banks with smaller buffers over their requirements are less likely to significantly increase repurchases. We expect the Category III and IV banks with significant unrealized losses in their available-for-sale (AFS) securities portfolios will likely be parsimonious with repurchases.

In their July 2023 proposal for the final Basel III standards, regulators proposed to disallow those banks from excluding the effect of those unrealized losses in their regulatory capital. (Category I and II banks already must count unrealized losses on AFS in their regulatory capital.)

How We Consider Stress Test Results In Ratings

The stress test results typically don't result in any material surprises that directly lead us to change our credit ratings. Still, we do consider the results as important inputs into our rating assessments, for example, when we forecast banks' capital levels and S&P Global Ratings risk-adjusted capital (RAC) ratios.

We may also compare the reported SCBs and loan loss rates to the assumptions in our rating analysis. For example, large differences between a bank's internal stress test results and the Fed's results may inform our view of risk. We also compare the Fed's assessment of loss rates with our own analysis of a bank's risk appetite and underwriting standards.

This year, overall, we believe the relative strengths and weaknesses of these banks from the test, including loan loss rates, roughly aligned with the assumptions we factor into our ratings. Notably, we recently revised our outlook on American Express Co.--a perennial outperformer on the Fed's stress test--to positive from stable (see "American Express Co. Outlook Revised To Positive From Stable On Consistent Strong Results; 'BBB+/A-2' Ratings Affirmed," June 13, 2024).

Many Banks' SCBs Will Increase

We expect the SCBs to increase for many banks with a U.S.-domiciled parent that participated in this year's stress test (see table 2).

Table 2

Still To Come

We believe the final Basel III endgame proposal--rather than the stress test--will have the biggest impact on how much capital banks need to hold. While the proposal from July 2023 indicated that capital requirements could rise substantially for large banks, the Fed's Vice Chair for Supervision Michael Barr recently said he expects "broad and material changes" to the proposal based on comments regulators received.

The Fed also didn't incorporate the current expected credit loss (CECL) model into its calculation of the allowance for credit losses in the 2024 stress test. Doing so could meaningfully increase provisions (to build a bigger allowance for stress conditions) and thus could raise SCBs. This may be included in future stress tests.

Appendix

Severity of this year's test results versus last

Table 3

Although the severity of this year's stress test was similar to last year's, this year's results reflected higher losses. The Fed says this is largely because bank balance sheets are now somewhat riskier. Specifically, banks' credit card balances are significantly higher than last year, with more delinquencies, and the aggregate credit card loan loss rate remains elevated at 17.6% in 2024, compared to 17.4% last year.

In addition, banks' corporate credit portfolios have become riskier, partly due to banks doing more internal downgrades of these loans, which resulted in higher projected corporate losses in the Fed's stress test (the aggregate commercial and industrial loan loss rate was 8.1% in 2024, from 6.7% last year).

Banks' preprovision net revenue (PPNR), which helps offset projected losses, was lower this year due to weaker noninterest revenue, largely reflecting lower investment banking and mortgage banking results.

Earnings capacity and loan loss rates are key determinants for SCBs

A bank's ability to deliver PPNR under the stress scenario to help offset credit and other losses helps to determine its SCB. More specifically, a bank with an elevated loss rate--either due to a concentration in riskier loans (i.e., credit cards) or riskier underwriting standards--and weak PPNR would likely have a higher SCB.

That said, other factors such as changes in other comprehensive income, a bank's dividend rate, also play a part in determining a bank's SCB. In addition, PPNR and loss rates are for the full nine quarters of the stress, while the SCB is determined at the level of a bank's minimum capital ratio during the nine quarters.

Chart 2

Results from the Fed's exploratory scenarios

The Fed's stress test scenario this year didn't explicitly incorporate a bank's ability to withstand significant increases in funding costs or large deposit outflows, which were big concerns in last year's regional bank turmoil. However, this year the Fed also ran an "exploratory analysis" including two funding stresses (for all banks tested) and two trading book stresses (for the largest, most complex banks).

The exploratory analysis is a distinct exercise from the stress test and explores additional hypothetical risks to the broader banking system. That said, it has no direct impact on SCBs or capital requirements. The Fed does not publish individual bank results under the exploratory analysis, only aggregate results.

The funding stresses included:

• A rapid repricing of a large proportion of deposits under a severe global recession, combined with high and persistent inflation and rising interest rates; and
• A rapid repricing of a large proportion of deposits under a moderate global recession, combined with increasing inflation and rising interest rates.

The funding stress analysis also assumes:

• Banks must convert 20% of their non-interest-bearing deposits into time deposits, and
• The shift occurs all at once, at the beginning of the projection horizon.

The Fed believes that in aggregate both scenarios would be manageable for the largest U.S. banks.

In the severe recession scenario, the aggregate CET1 ratio across the 31 banks falls by 2.7 percentage points to a minimum of 10.0%. Under the moderate stagflation scenario, the aggregate CET1 ratio falls by 1.1 percentage points to a minimum of 11.6%. The differences in capital declines mainly owe to different projections of PPNR due to higher funding costs, a change in the value of fixed-income securities, and higher loan loss provisions due to the weaker economic conditions.

Chart 3

The two other exploratory scenarios for the GSIBs were:

• A market shock featuring a sudden dislocation to financial markets stemming from expectations of reduced global economic activity and higher inflation, and
• A market shock featuring a sudden dislocation to financial markets stemming from expectations of severe recessions in the U.S. and other countries.

In addition, in both scenarios, the five hedge funds that represent the largest hedge fund counterparty exposures for each bank fail.

The results of the market shock elements suggest that, in aggregate, the largest banks' losses under these shocks would total 1.0%-1.2% of their risk-weighted assets, which suggests that such a shock would be manageable for the largest and most complex banks.

Chart 4

Ranking the banks by their stress test results

We have ranked the banks that participated in the stress test by the decline in their capital ratios (see table 4). We looked at a bank's starting CET1 ratio and its minimum CET1 ratio during the nine quarters of the stress test, starting at the end of fourth-quarter 2023. The smaller the decline from the beginning ratio to the minimum ratio, the stronger the bank's capital and earnings position is relative to its risk, according to the results of the Fed's stress test.

Of note, the following banks had minimum capital ratios below 7%:

• BMO Financial Corp.,
• HSBC North America Holdings Inc., and
• Citizens Financial Group Inc.

Table 4

Loan loss rates by bank

The Fed projected in this year's stress test that the tested banks, in aggregate, would experience $685 billion in losses on loans and other positions over the nine quarters of the projection horizon, mostly from loan losses.

Smaller sources of losses in the stress test were trading and counterparty credit, additional items such as loans booked under the fair-value option, and securities. For amortized cost loans, the projected aggregate losses amounted to a 7.1% loss rate, higher than last year's test (which involved fewer banks).

However, eight banks' loss rates fell relative to last year, even in the severely adverse scenario (see table 5).

Table 5

Commercial real estate

This year's stress test results showed that the biggest banks are resilient with respect to commercial real estate (CRE; see table 6). We believe CRE will likely continue to see pressure in the coming years (see "Some U.S. Regional Banks Could Face Higher Risk If Commercial Real Estate Asset Quality Worsens, March 26, 2024).

In the severely adverse scenario, the Fed assumed a 40% decline in CRE prices, concentrated in properties that are most at risk of a sustained drop in income and asset values (e.g., office). The aggregate loss rate of 8.8% in this year's stress test was the same as the prior year. Notably, we believe that many smaller banks--those not included in the stress test--are more exposed to CRE and could therefore be hit harder by CRE price deterioration than the larger banks.

Table 6

Related Research

• Some U.S. Regional Banks Could Face Higher Risk If Commercial Real Estate Asset Quality Worsens, March 26, 2024
• How The U.S. Proposes To Implement Basel III Capital Rules And The Impact On U.S. Bank Capital Ratios, Jan. 11, 2024
• Banks Held Up Well In The Fed's Stress Test, But They Likely Won't Ramp Up Share Buybacks June 2023, July 7, 2023
• After A Tougher Fed Stress Test, Capital Requirements Will Rise For Some U.S. Banks, June 28, 2022
• Fed Stress Test Results Give Banks Room To Release The Capital They Built Up During The Pandemic, June 28, 2021
• The Fed's Stress Test Results Open The Door For U.S. Banks To Increase Capital Returns In 2021, Dec. 22, 2020