Don’t miss out on news you won’t find anywhere else — get The Memo US and The Memo Europe in your inbox every two weeks.
Some of the US's biggest systemically important institutions have incorporated the benefits of significant risk transfers (SRTs) into the data submitted for regulatory stress tests from the Fed, while their smaller peers like Ally Financial, US Bancorp and Huntington Bancshares have not, according to sources.
US banks with over $250bn in assets are subject to annual supervisory tests by the Fed, and banks with assets between $100bn and $250bn undergo periodic stress tests under Dodd-Frank.
This year the Fed published its results for the 22 largest US banks it tested, some of which have SRTs.
The largest five to six banks are more likely to have the time and resources to incorporate its SRTs into its stress test. Some of the largest US banks have managed to do so, 9fin understands.
Smaller banks (of the ones tested) are likely to have relatively lower stress testing surcharges to begin with, lowering the incremental benefit of adding SRT to its framework.
Regardless of their size banks do not have a direct influence or understanding on what the Fed includes in its stress test framework, and whether it will incorporate the benefits of the SRTs submitted into its stress test results.
Ally and Huntington were not in the “large” 22 US banks this year, but still large enough to be stress tested and have SRTs, did not have their SRTs incorporated into their stress test framework, 9fin understands. These banks were counted in the top 31 “large” bank tests last year.
The Federal Reserve does stress tests every year to ensure that large banks are sufficiently capitalized and able to lend even in a severe recession.
However, the framework for doing the stress tests varies across institutions and has been criticized as opaque. An area where it notably varies is the incorporation of SRT transactions into its results.
The Fed collects information and data for the stress tests from bank’s FR Y-14 reports, which are a set of capital assessments and stress testing reports submitted by banks. These reports include an annual report with quantitative projections of the balance sheet performance and qualitative information on methodologies, quarterly reports with granular data on varies asset classes and capital components and monthly reports with detailed loan-level data.
In its stress test results, the Fed noted: “At times, Federal Reserve staff engage staff from banks subject to stress test to better understand the submitted bank data and make adjustments, if needed.”
The Fed mentions in its stress test results that it considers "synthetic securitizations" and that it incorporates “a richer dataset and considered this type of credit protection in modeling fair value-option/held-for sale loan losses.” It added this provision in 2024.
Credit risk mitigants are not automatically recognized in the projected loan losses on a portfolio. Banks must prove to regulators that some of those projected losses can be reduced by the presence of an SRT.
It has been debated whether SRTs would be considered in the reporting depending on their structure, whether they have a CUSIP ID (as some credit-linked notes do) and whether the SRT was done available for sale loans versus held for investment loans.
The testing frameworks reportedly vary across institutions and remain opaque. “I have one client who is allowed to fully incorporate the effects of their SRT into their stress test, and another one who isn't, and that's because the Fed deals on an individual basis with each stress testing institution,” said Matt Bisanz speaking at a DealCatalyst’s Significant Risk Transfer conference last week.
“Theoretically, it makes sense that the existence of an SRT trade would mitigate the ‘trough’ used to calculate the stress capital buffer over the nine quarter stress test horizon,” said a lawyer. “That said, I know there have been persistent challenges with squeezing an SRT trade into the FR Y-14 reporting data that the Fed uses to calibrate the stress tests for the banks.”
In April, the Fed proposed a rule to average the stress test results of two consecutive years and improve transparency over the testing framework. Proponents say it will help reduce year-over-year volatility in bank capital requirements. Critics have pushed back on giving the banks access to details about the stress test ahead of time saying it could give banks a chance to game the results to lower their capital requirements.
“The proposal will ... improve the design and related assumptions for certain models including credit losses, operational risk and securities,” said Michelle Bowman, Vice Chair for Supervision at the Federal Reserve, speaking at the Federal Reserve stress testing research conference in Boston today.
Market participants hope this will, among other things, give a clearer integration of SRT into the modeling of credit losses.
The finalized stress testing proposal is expected to be released this month or the following, 9fin understands.
Industry groups including the Bank Policy Institute and the American Bankers’ Association filed a lawsuit at the end of last year against the Fed, citing the “flawed” stress testing framework and are asking for more transparency over the stress testing process.
“The current opaque regime, combined with the lack of clear standards for the global market shock and the operational risk charge, continues to produce capital charges that are inaccurate, volatile and excessive,” said Greg Baer, president and CEO of the Bank Policy Institute, at the time of filing the lawsuit.
All the top US banks passed their stress tests at the end of June with flying colors: “All 22 banks tested remained above their minimum CET1 capital requirements during the stress scenario, after absorbing total projected hypothetical losses of more than $550bn,” said the Fed at the time.
Banks have a stress capital buffer requirement stemming from the stress test and can ask that the requirement be reconsidered based on an independent assessment of the bank’s loan portfolio. The lowest stress capital buffer requirement banks can have is 2.5%.
According to the large bank stress test results, Ally and US Bank have a stress capital buffer requirement of 2.6% and Huntington has one of 2.5%, meaning they are already essentially at the floor and thus would have little room to improve.
Goldman Sachs has a stress capital buffer requirement of 3.4% and Citigroup has one of 3.6%.
Banks can request to have their stress capital buffers reviewed and potentially lowered.
On 30 September, the Fed lowered Morgan Stanley’s stress capital buffer requirement for this year from 5.1% to 4.3%, at the bank’s request and following a subsequent analysis of the bank’s fair value option loan portfolio.
Last year, the Fed lowered Goldman’s stress capital buffer requirement from 6.4% to 6.2%.
Ally, Citi and US Bancorp declined to comment. Huntington and Goldman Sachs did not reply to request for comment.






