[House Hearing, 118 Congress]
[From the U.S. Government Publishing Office]
STRESS TESTING: WHAT'S INSIDE
THE BLACK BOX?
=======================================================================
HEARING
before the
SUBCOMMITTEE ON FINANCIAL INSTITUTIONS AND MONETARY POLICY
of the
COMMITTEE ON FINANCIAL SERVICES
U.S. HOUSE OF REPRESENTATIVES
ONE HUNDRED EIGHTEENTH CONGRESS
SECOND SESSION
__________
JUNE 26, 2024
__________
Serial No. 118-98
Printed for the use of the Committee on Financial Services
[GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]
www.govinfo.gov
_______
U.S. GOVERNMENT PUBLISHING OFFICE
56-624 PDF WASHINGTON : 2026
HOUSE COMMITTEE ON FINANCIAL SERVICES
PATRICK McHENRY, North Carolina, Chairman
FRENCH HILL, Arkansas, Vice MAXINE WATERS, California, Ranking
Chairman Member
FRANK D. LUCAS, Oklahoma SYLVIA R. GARCIA, Texas, Vice
PETE SESSIONS, Texas Ranking Member
BILL POSEY, Florida NYDIA M. VELAZQUEZ, New York
BLAINE LUETKEMEYER, Missouri BRAD SHERMAN, California
BILL HUIZENGA, Michigan GREGORY W. MEEKS, New York
ANN WAGNER, Missouri DAVID SCOTT, Georgia
ANDY BARR, Kentucky STEPHEN F. LYNCH, Massachusetts
ROGER WILLIAMS, Texas AL GREEN, Texas
TOM EMMER, Minnesota EMANUEL CLEAVER, Missouri
BARRY LOUDERMILK, Georgia JAMES A. HIMES, Connecticut
ALEXANDER X. MOONEY, West Virginia BILL FOSTER, Illinois
WARREN DAVIDSON, Ohio JOYCE BEATTY, Ohio
JOHN W. ROSE, Tennessee JUAN VARGAS, California
BRYAN STEIL, Wisconsin JOSH GOTTHEIMER, New Jersey
WILLIAM R. TIMMONS, IV, South VICENTE GONZALEZ, Texas
Carolina SEAN CASTEN, Illinois
RALPH NORMAN, South Carolina AYANNA PRESSLEY, Massachusetts
DANIEL MEUSER, Pennsylvania STEVEN HORSFORD, Nevada
SCOTT FITZGERALD, Wisconsin RASHIDA TLAIB, Michigan
ANDREW R. GARBARINO, New York RITCHIE TORRES, New York
YOUNG KIM, California NIKEMA WILLIAMS, Georgia
BYRON DONALDS, Florida WILEY NICKEL, North Carolina
MIKE FLOOD, Nebraska BRITTANY PETTERSEN, Colorado
MICHAEL LAWLER, New York
ZACHARY NUNN, Iowa
MONICA DE LA CRUZ, Texas
ERIN HOUCHIN, Indiana
ANDREW OGLES, Tennessee
Matthew Hoffman, Staff Director
------
SUBCOMMITTEE ON FINANCIAL INSTITUTIONS AND MONETARY POLICY
ANDY BARR, Kentucky, Chairman
BARRY LOUDERMILK, Georgia, Vice BILL FOSTER, Illinois, Ranking
Chairman Member
BILL POSEY, Florida AYANNA PRESSLEY, Massachusetts,
BLAINE LUETKEMEYER, Missouri Vice Ranking Member
ROGER WILLIAMS, Texas NYDIA M. VELAZQUEZ, New York
JOHN W. ROSE, Tennessee BRAD SHERMAN, California,
WILLIAM R. TIMMONS, IV, South GREGORY W. MEEKS, New York
Carolina DAVID SCOTT, Georgia
RALPH NORMAN, South Carolina AL GREEN, Texas
SCOTT FITZGERALD, Wisconsin JOYCE BEATTY, Ohio
YOUNG KIM, California JUAN VARGAS, California
BYRON DONALDS, Florida SEAN CASTEN, Illinois
MONICA DE LA CRUZ, Texas
ANDREW OGLES, Tennessee
C O N T E N T S
----------
Wednesday, June 26, 2024
OPENING STATEMENTS
Page
Hon. Andy Barr, Chairman of the Subcommittee on Financial
Institutions and Monetary Policy, a U.S. Representative from
Kentucky....................................................... 1
Hon. Bill Foster, Ranking Member of the Subcommittee on Financial
Institutions and Monetary Policy, a U.S. Representative from
Illinois....................................................... 3
WITNESSES
Mr. Sean Campbell, Chief Economist & Head of Policy Research,
Financial Services Forum....................................... 4
Prepared Statement........................................... 7
Mr. Francisco Covas, Executive Vice President & Head of Research,
Bank Policy Institute.......................................... 35
Prepared Statement........................................... 37
Mr. Jonathan Gould, Partner, Jones Day........................... 58
Prepared Statement........................................... 60
Mr. Greg Feldberg, Research Director, Yale Program on Financial
Stability, and Research Scholar and Lecturer, Yale School of
Management..................................................... 64
Prepared Statement........................................... 66
APPENDIX
ADDITIONAL MATERIAL SUBMITTED FOR THE RECORD
Hon. Andy Barr:
Securities Industry and Financial Markets (SIFMA)............ 100
RESPONSES TO QUESTIONS FOR THE RECORD
Written responses to questions for the record from Mr. Francisco
Covas..........................................................
Representative Maxine Waters................................. 107
LEGISLATION
H.R. ------, the "Fair Audits and Inspections for Regulators'
Exams Act"..................................................... 108
H.R. 8591, the "Federal Reserve Financial Accountability and
Transparency Act".............................................. 121
STRESS TESTING: WHAT'S INSIDE
THE BLACK BOX?
----------
Wednesday, June 26, 2024
U.S. House of Representatives,
Subcommittee on Financial Institutions
and Monetary Policy,
Committee on Financial Services,
Washington, DC.
The subcommittee met, pursuant to notice, at 2:04 p.m., in
room 2128, Rayburn House Office Building, Hon. Andy Barr
[chairman of the subcommittee] presiding.
Present: Representatives Barr, Posey, Luetkemeyer, Williams
of Texas, Loudermilk, Rose, Fitzgerald, Kim, De La Cruz,
Foster, Sherman, Scott, Green, Beatty, Vargas, Casten, and
Pressley.
Chairman Barr. The committee will now come to order.
Without objection, the chair is authorized to declare a
recess of the committee at any time.
This hearing is titled, ``Stress Testing: What's Inside the
Black Box?''
Without objection, all members will have 5 legislative days
within which to submit extraneous materials to the chair for
inclusion in the record.
I now recognize myself for 5 minutes to give an opening
statement.
HON. ANDY BARR, CHAIRMAN OF THE SUBCOMMITTEE ON FINANCIAL
INSTITUTIONS AND MONETARY POLICY, A U.S. REPRESENTATIVE FROM
KENTUCKY
When done right, stress testing can be beneficial to banks,
consumers, and the financial system. Stress tests demonstrate
to everyone that the U.S.' largest banks are well capitalized,
resilient, and capable of functioning and lending throughout
the economic cycle but for all these benefits to flow from
stress tests, the tests must be credible, with transparency and
accountability.
Instead of running stress tests in an open and accountable
manner subject to public scrutiny, the Federal Reserve cloaks
the stress test under a veil of secrecy. This is no legal basis
for the secrecy, and the perceived benefits of secrecy are
illusory at best.
No one disputes that the stress tests are used in binding
capital requirements, including the stress capital buffer, and
can be used to restrict banks from paying dividends. The stress
tests, including the scenarios and models, must be subject to
the notice and comment rulemaking process.
I am certain that we will hear from our colleagues on the
other side of the aisle that key aspects of stress tests must
remain secret and not subject to public comment. We will
probably hear that banks will simply, ``game the tests,'' and
that there is not enough time to take public comment on stress
test scenarios.
Even if these arguments were legally tenable, which they
are not, they still do not carry water. As we will hear from
our witnesses, it is not clear that the banks could easily game
the stress test, even if they wanted to and the Fed already has
plenty of traditional supervisory tools to address gaming. More
importantly, if disclosing models resulted in better stress
performance without improved risk profiles, this suggests there
are material weaknesses in the models themselves rather than a
problem with public disclosure.
It is critical that the Fed gets stress testing right and
with growing politicization of the Federal banking agencies, it
is possible that rogue officials pursuing political agendas
could also game the tests by manipulating models and
assumptions to obtain desired results for heightened capital
requirements or reduce dividends.
The stress tests have real world impacts: Banks make
capital allocations based on how they think they will perform
on stress tests. This means that banks are making decisions
about how to lend based in part on how they think the Fed's
models will work, and that has massive ramifications for the
real economy. Because banks are only making guesses, they must
be more conservative with how they allocate their capital to
ensure that they can easily pass the stress tests, regardless
of the underlying models and assumptions used. Again, this has
real-world impact on the families, farmers and small businesses
who rely on banks to access credit.
Moreover, the stress tests impacts how much dividends banks
can pay. It is important that banks can return their profits to
shareholders as otherwise investors will not choose to invest
in banks in the future.
This hearing builds on the work that our committee has been
doing in this Congress to improve the stress testing regime.
This committee recently reported out my bill, the Bank
Resilience and Regulatory Improvement Act, which would require
that the Fed disclose the underlying models and assumptions
used in the stress tests, and require that the Fed take public
comment on each year's stress testing scenarios.
Of course, there is nothing keeping the Fed from fixing
these problems on their own, but due to their continued
unwillingness to do so, I applaud my colleagues for advancing
this important piece of legislation.
One final note, we have to also recognize and remember that
stress testing and the way in which stress tests are conducted
interface with the existing regulatory regimes that are out
there. As the banking agencies are reviewing the proposal on
Basel III Endgame, we must keep in mind how that can work
cumulatively with any changes to stress tests, also the long-
term debt rule. These requirements can pile up and the way in
which they work and are interchangeable is very, very
important, and so there is interplay among these regulatory and
administrative decisions.
So I look forward to our witnesses and their testimony and
shedding at least a glimmer of light on these opaque processes.
With that, I now recognize the ranking member of the
Subcommittee on Financial Institutions and Monetary Policy, the
gentleman from Illinois, Dr. Foster, for 4 minutes for an
opening statement.
HON. BILL FOSTER, RANKING MEMBER OF THE SUBCOMMITTEE ON
FINANCIAL INSTITUTIONS AND MONETARY POLICY, A U.S.
REPRESENTATIVE FROM ILLINOIS
Mr. Foster. Thank you, Chairman Barr.
Thank you to our witnesses for joining us, particularly
those of you who spent part of your life just trying to
understand what befell our economy in 2008 and how to keep it
from happening again.
In 2008 during the crisis, large U.S. banks faced a crisis
of confidence and its economic conditions continued to
deteriorate, and the question came up just how much worse
things could get. The outlook was dim at best, and action was
needed to assure investors and the public that U.S. banks were
capable of withstanding continued negative conditions in our
economy.
In 2009, regulators conducted simultaneous stress tests of
the 19 large U.S. banks to project their ability to continue
should leading economic indicators continue to go downhill. By
publishing the results of the stress tests and requiring banks
to correct capital deficiencies, these stress tests were a big
part of what restored confidence in large U.S. banks that they
would be able to continue serving Americans if the crisis were
to continue to worsen.
This process is viewed as, almost universally, as a
success. Led to the creation of the permanent stress testing
regime that we now know today, with Congress requiring regular
internal and supervisory stress testing of large banks as part
of the Dodd-Frank Act, but as stress testing has evolved,
industry participants have petitioned for greater transparency
and predictability of stress tests from regulators, which has
its pros and cons, which I hope we are going to have a
thoughtful discussion of this in this hearing.
Under the last administration, the Federal Reserve began
publishing significantly more details regarding the design of
stress test models and their methodology, including projected
loss rates for certain types of loans. Despite the added
transparency, industry participants and others would continue
to advocate for closer to total transparency of stress test
scenarios and models, both afterwards and even prior to the
stress tests.
So academics and former Federal Reserve officials rightly
cautioned against this approach of total transparency, raising
concerns that too much transparency and stress testing would
allow financial institutions to game the stress test in various
ways.
When I first heard about the concept of stress test back in
2009, I thought that was a good idea and was impressed at how
it worked, but I was also warned at the time that it is not a
panacea. History gives us examples of the weaknesses of fully
transparent stress tests. One relevant to this discussion is,
in 2002, the Office of Financial Housing Enterprise Oversight
administered what they called stress tests to Fannie Mae and
Freddie Mac, which they passed and which were not really viewed
as a success in terms of diagnosing real problems they had.
The agency did disclose all the comprehensive details about
their stress testing model, including the scenarios, empirical
specifications, and parameter estimates for public notice and
comment. I do not think anyone or very few people appreciated
not only the fragility of asset valuations there but the
correlations in those asset valuations. No one could
anticipate, for example, that you would have a nationwide drop
in housing prices and it was not reflected in the risks that
the market priced into various complex financial devices that
were put as a layer on top of housing finance.
So complete transparency and limitations on the agency's
ability to update the model to reflect the changing economic
conditions reduced the efficiencies of those tests, and we have
to avoid that mistake here. In retrospect, the academic
community has noted that these factors provide a very poor
window into enterprises' true risk exposures.
Like I said, I hope for a really thoughtful discussion of
the pros and cons of different kinds of transparency here.
I yield back.
Chairman Barr. The gentleman yields back. Thank you.
Today we welcome the testimony of Sean Campbell. Mr.
Campbell is the chief economist and head of policy research at
the Financial Services Forum.
Francisco Covas. Mr. Covas is executive vice president and
head of research at the Bank Policy Institute.
Jonathan Gould. Mr. Gould is a partner at the Jones Day law
firm.
Greg Feldberg. Mr. Feldberg is a research director and
research scholar and lecturer at Yale University.
We thank each of you for taking time to be here today. Each
of you will be recognized for 5 minutes to give an oral
presentation of your testimony.
Without objection, each of your written statements will be
made part of the record.
Mr. Campbell, you are now recognized for 5 minutes to give
your oral remarks.
STATEMENTS OF SEAN CAMPBELL, CHIEF ECONOMIST & HEAD OF POLICY
RESEARCH, FINANCIAL SERVICES FORUM
Mr. Campbell. Chairman Barr, Ranking Member Foster, and
members of the subcommittee, my name is Sean Campbell, and I am
the chief economist at the Financial Services Forum, which
represents the eight globally systemically important banks, or
GSIBs, headquartered in the United States. I am a Ph.D.
economist with nearly 14 years of experience at the Federal
Reserve, where I was deeply involved in regulatory policies and
stress testing development critical to the stability of the
U.S. financial system, particularly following the 2008
financial crisis. The Forum appreciates the opportunity to
testify today.
Over the last 15 years, U.S. GSIBs have demonstrated their
strength and resilience through rigorous and dynamic annual
stress tests, holding significantly more capital compared to
stress losses even under severe economic scenarios.
Additionally, U.S. GSIBs demonstrated their strength during
recent real-life stress tests, supporting the economy during
the coronavirus disease 2019 (COVID-19) pandemic and fostering
financial stability during recent regional bank failures.
An open and transparent regulatory process is a
longstanding and well-accepted principle of good governance
that provides the public with important information while also
helping to ensure the public accountability of regulators. The
stress testing framework, however, lacks transparency compared
to other regulatory measures, like the risk-based capital
requirements that are currently undergoing a public and
transparent notice and comment process.
Stress test opacity limits public understanding and hinders
the broader discussion necessary to improve the stress testing
process and mitigate any negative side effects for the broader
economy.
Transparency is essential to promoting an effective and
efficient banking system. Importantly, available research
indicates that stress tests have clear impacts on credit
allocation in the U.S. economy and, in some instances, can lead
to reduced lending in areas, such as small business and
healthcare.
As is the case for all other risk-based capital
requirements, banks need to understand the regulatory rules to
which they are subject to make informed decisions. Stress
testing opacity forces banks to make decisions based on
incomplete data and information that hinders banks core
function of stewarding financial resources to their most
productive and efficient use.
Greater transparency and stress testing would improve bank
risk management and capital planning, while allowing for more
informed assessments by people across the economy of its impact
on credit availability and facilitating greater economic
growth.
To be sure, the Federal Reserve does provide some limited
disclosures to the public. The provided disclosures, however,
are generally broad and high level, without providing the type
of pertinent information that would admit a serious and
critical assessment of the stress tests. As an example, in
describing its model for net income, the Federal Reserve
discloses that--and here I am quoting from the Federal
Reserve's public website--the specific macroeconomic variables
that enter each regression model differ across equations and
are chosen based on statistical predictive power and economic
theory.
To analogize, this is a bit like asking your grandmother
for her secret chocolate cake recipe and being told, oh, dear,
it is so simple, just mix some flower, eggs, butter, and other
flavorings together and there you have it.
Stress testing is the only major prudential requirement
that does not provide the public with an open and transparent
process. This lack of transparency is largely a vestige of the
rapid approach that was taken to executing the initial round of
stress tests in the midst of the financial crisis. Then, you
need to quickly assure the public of bank strength and
financial stability arguably outweighed the need for
transparency. Fifteen years on, there is no compelling public
policy rationale to maintain a largely opaque approach to
stress testing.
Today's approach should be broadly consistent with the
public notice and comment standards applied to the rest of the
bank regulatory framework, which includes Basel III, the GSIB
capital surcharge, and liquidity standards. Critically
important is the fact that the results of the stress tests now
feed into the annual stress capital buffers for each firm,
functioning as a de facto regulatory requirement by directly
augmenting required capital amounts and, in turn, the
allocation of capital to individuals, companies, and important
sectors of the economy.
In conclusion, greater transparency in stress testing is
essential to achieving a more effective, efficient, and
resilient financial system. Remedies that should be part of a
broader public policy discussion should include, one, more
formal engagement with the public on setting hypothetical
scenarios so regulators can gain insight to salient risks from
a diverse set of viewpoints; two, requiring that stress test
models be subject to the usual notice and comment process that
applies broadly to other prudential requirements; and three,
requiring that the stress testing process be subject to a
rigorous external audit that would make its findings known to
the board and to the public to instill further confidence in
the validity and robustness of the stress testing regime.
Overall, aligning stress tests with the public process
afforded other bank regulatory standards would help improve
public trust, promote economic stability, and ensure that banks
can support the economy to their fullest potential.
Thank you, and I appreciate your questions.
[The prepared statement of Mr. Campbell follows:]
[GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]
Chairman Barr. The gentleman's time has expired.
Mr. Covas is now recognized for 5 minutes.
STATEMENT OF FRANCISCO COVAS, EXECUTIVE VICE PRESIDENT & HEAD
OF RESEARCH, BANK POLICY INSTITUTE
Mr. Covas. Chairman Barr, Ranking Member Foster, and
subcommittee members, thank you for the opportunity to testify
today.
My name is Francisco Covas, head of research at the Bank
Policy Institute. We are a research and advocacy group
supported by banks with over $100 billion in assets. Our
membership spans a full range of banks under the Federal
Reserve's stress tests, so we are keen to testify today.
The stress testing regime that underpins the capital
framework for large banks consistently demonstrate that U.S.
banks are well capitalized and can continue to support the
economy during economic downturns. Since 2020, stress tests
results have been used to set the stress capital charge,
effectively part of larger banks' minimum capital requirements.
We support using a stress capital buffer in overall capital
requirements. However, the lack of transparency in the Federal
Reserve stress testing regime has significant economic costs
that can adversely affect U.S. financial system and the broader
economy. The opacity and inaccuracy of the Fed's models create
uncertainty for banks about the level of capital they are
required to hold, leading to excess ``uncertainty buffers''
that reduce credit availability and decrease market liquidity.
In my written testimony, I provide several examples that
highlight the inaccuracies and the volatility inherent in
supervisory stress testing models used to determine the stress
capital buffer.
Year after year, we observe significant fluctuations in the
stress capital buffer that often fail to reflect equitable
changes in banks' risk profiles. This disconnect can be
attributed to various factors.
First, the lack of granularity in the Fed's revenue
projections, which are prone to inaccuracies and volatility.
Second, the Fed's models consistently underestimate trading
revenues during times of higher market volatility, a finding
that runs counter to historical experience. Third, operational
risk loss models have proven to be inaccurate, with projected
losses increasing in the post-pandemic period despite the
absence of the corresponding rates in real-world losses.
Fourth, long lost models also tend to overestimate losses
compared to banks' internal projections, likely due to the
Fed's inability to properly account for bank-specific factors.
Fifth, compounding these issues is a flawed reconsideration
process for appealing stress test results, which suffers from a
lack of independence and transparency.
It is also important to have notice and common on the
scenarios used to calculate the stress capital buffer. Our
analysis reveals that the 2024 stress scenarios divert
significantly from historical experience and the Fed's own
guidance. Several macroeconomic variables in these scenarios
are dramatically more severe than those observed in previous
U.S. recessions.
Consider this example: In just the first quarter of the
stress planning period, the severely adverse scenario in the
year's stress test assumes a decline in the house price index
exceeding 15 percent, and this is not annualized. This
projection far exceeds what we observed during the 2007, 2009
global financial crisis. Such an extreme scenario implies
unprecedented shifts in other macroeconomic variables,
including sharper initial increases in employment rate and
steeper declines in short-term rates. The result, financial
institutions face considerably higher projected losses and
lower revenues compared to prior severe recessions.
Moreover, the calibration of the global markets are a
component that applies to the largest banks and has become less
transparent over time, so it is not clear what scenario the
global markets (GMS) currently represents.
In summary, increased transparency in the Fed's stress
testing regime is necessary to reduce uncertainty and promote a
more efficient financial system. U.S. banks are very well
capitalized, but the current stress capital buffer framework
suffers from excessive volatility, inaccurate projections, and
a flawed reconsideration process, leading to less lending and
increased market liquidity.
Moreover, the stress analysis violate the Fed's own
guidance, and that is significantly more severe than historical
recessions. Allowing public comment on scenarios and
supervisory models would enable extensive review by experts,
academics, and banks. This would increase transparency in
scenario design and improve model accuracy, fostering a more
effective financial system that better serves U.S. economy's
needs.
Thank you again for the opportunity to testify. I look
forward to your questions.
[The prepared statement of Mr. Covas follows:]
[GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]
Chairman Barr. Thank you.
Now, Mr. Gould, you are now recognized for 5 minutes.
STATEMENT OF JONATHAN GOULD, PARTNER, JONES DAY
Mr. Gould. Chairman Barr, Ranking Member Foster, and
members of the subcommittee, thank you for the opportunity to
discuss stress testing.
My testimony is my own. I am speaking today solely in my
personal capacity. I am not speaking on behalf of any clients
or my law firm.
Stress testing has transformed bank capital from a
backward-looking static concept into a forward-looking dynamic
requirement but its current integration into the overall
capital framework rests uneasily with the requirements of
administrative law and agency overemphasis on regulatory
capital as the primary supervisory tool risks marginalizing the
important role of examination in ensuring the safety and
soundness of banks.
The modern form of supervisory stress testing traces its
roots to the 2008 financial crisis. To restore confidence in
the health of the largest bank holding companies, the Federal
Government designed and administered the Supervisory Capital
Assessment Program, or SCAP. Under this program, regulators
assessed whether banks would remain well-capitalized in an
adverse macroeconomic scenario. The SCAP was widely viewed as a
success, but like many crisis-era actions, legal authorities
and process were not scrutinized too closely given the
exigencies of the moment.
Congress subsequently endorsed and ratified the concept of
supervisory stress testing in the Dodd-Frank Act the following
year but like other extraordinary programs borne of the crisis,
supervisory stress testing has evolved into something more. As
the Fed embarked on the creation of its Comprehensive Capital
Adequacy Review, or CCAR, it melded regulatory capital, stress
testing, and capital planning into an omnibus framework, part
supervisory exercise and part regulatory requirement, making
liberal use of guidance to knit the new edifice together.
Supervisory stress testing became a cornerstone of a new
approach to regulation and supervision of the Nation's largest
financial institutions. In and of itself, supervisory stress
testing would be merely hortatory and, at critical times,
illuminating for regulators, markets, and the banks themselves,
but under the Fed's hybrid framework, supervisory stress
testing is also used to limit capital distributions and has now
become the means for setting binding capital requirements for
large banks through the introduction of the stress capital
buffer.
The Fed's continuing failure to seek public comment on its
stress models and scenarios and, in the case of the former,
even disclose them at all raises significant concerns under
basic principles of administrative law. First, the
Administrative Procedure Act requires agencies to pursue
legislative rules through notice and comment rulemaking but the
Fed has not sought public comment on its models and scenarios.
In the case of the stressed capital buffer and planned capital
distributions, its related rulemakings are, at best,
incomplete, omitting the most critical components. These
omissions are particularly egregious since they not only inform
but dictate what is arguably the most important regulatory
requirement on banks--their capital levels. Nor is this an area
of legal ambiguity or first impression. Courts have repeatedly
held that models treated by agencies as binding must go through
public comment.
Second, because the Fed's models and assumptions are the
basis for its actions and it has not disclosed its models, it
is unable to explain its decisionmaking process. As a result,
its actions appear arbitrary and capricious.
At its core, the Fed has created a framework that allows it
discretion to set capital levels for banks through a scenario
choice and model design with limited accountability. Its only
constraint is the models themselves, but the public lacks
visibility into these models as well as their design, back-
testing and revision. Put another way, the Fed's capital stress
buffer is like a highway with an unposted speed limit that
changes on a regular basis, and one only learns the speed limit
after being pulled over. Whatever the value of stress testing,
its utility is no substitute for its legality.
Supervisory stress testing and the capital edifice it
supports are the paradigmatic post-2008 reforms. This new
approach to regulation and supervision is too often marked by
its complexity, opacity, and only episodic adherence to basic
principles of administrative law. As Federal banking agencies
appear poised to double down in these and other areas, Congress
should ensure that agencies abide by applicable legal
requirements and assess the effectiveness and costs of current
post-2008 regulatory reforms before expanding them.
Thank you again for the opportunity to testify. I look
forward to your questions.
[The prepared statement of Mr. Gould follows:]
[GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]
Chairman Barr. Thank you for your testimony.
Now, Mr. Feldberg, you are now recognized for 5 minutes.
STATEMENT OF GREG FELDBERG, RESEARCH DIRECTOR, YALE PROGRAM ON
FINANCIAL STABILITY, AND RESEARCH SCHOLAR AND LECTURER, YALE
SCHOOL OF MANAGEMENT
Mr. Feldberg. Chairman Barr, Ranking Member Foster, members
of the subcommittee, thank you for the opportunity to testify.
Stress tests have played an important role in our financial
history. Along with two of my fellow panelists today, I was
involved in the first stress test on the staff of the Federal
Reserve, and I think we are all very proud of how that turned
out in 2009.
I would like to make four points about the state of stress
testing today. First, supervisory stress tests should be
countercyclical. Stress tests should help supervisors keep
their guard up during those long periods of financial calm when
the possibility of a banking crisis and its associated costs
appear remote. In good times, stress tests should be tougher,
and we should resist the inevitable pressures to reduce the
rigor of the process.
Second, stress tests should use multiple scenarios.
Financial crises tend to come from unexpected places. The Bank
of England and European Central Bank have long used
``exploratory'' scenarios every other year. They are not
binding, they simply help the understanding of regulators and
banking system about the risks out there. For the first time
this year, the Federal Reserve has introduced nonbinding
exploratory scenarios, and I welcome that development.
In short, the U.S. innovated the use of supervisory stress
tests during the crisis in 2008. Early efforts in other
jurisdictions did not measure up as they learned in other
jurisdictions to ours, but in later years the innovation has
been elsewhere in other countries.
Third, stress tests are part of a broader regulatory and
supervisory toolkit. Some have argued that the Fed's stress
test should have included Silicon Valley Bank, the first large
bank to fail last year in the rising interest rate scenario. I
am not sure that would have worked. It is possible that rising
rates could have boosted net interest income in the test,
making up for the decline in asset values.
The International Monetary Fund actually made that very
point in 2020 in commenting to the U.S. about its financial
stability system. The International Monetary Fund (IMF)
recommended instead that the U.S. implement an existing Basel
standard for interest rate risk management with quantitative
thresholds, as many other countries have done. I made that
argument last year in a blog.
There is no doubt that Silicon Valley would have failed on
the Basel interest rate risk test.
The lesson is that supervisors need many tools in their
toolkit, a lesson was, again--that was again driven home
recently when supervisors rejected several big banks'
resolution plans.
I guess the main topic of this hearing, and fourth, that
transparency can be a double-edged sword. As I noted in a paper
in 2019, the goal should be to reveal just enough about the
Fed's methods to help banks develop their models and manage
their risks but not so much that it becomes a predictable
compliance exercise, which was what it really would become if
grandma showed you a recipe like Sean said.
I will make two observations about that tradeoff. First, we
are already revealing a lot to the regulated industry, which
may allow them to merely optimize the test. After what you have
heard today, you may be surprised at how much we reveal in the
U.S. I refer everyone on the committee to review the Fed's
public methodology and scenario documents. You might be
surprised at the level of detail after what you have heard
today. You will see 50 pages describing the models, including
descriptions of inputs and equations, 26 pages on loss rates. I
had a colleague on my team look at other jurisdictions to see
what they revealed. The U.S. was far ahead of other countries,
except Bank of England, to some extent and he could not find
another country that revealed its loss rates.
So you might wonder what it is. We would like to be more--
we are hearing we would like to be more transparent about from
the folks here today.
Sean Campbell, his testimony, his written testimony was
very helpful in understanding what it is we are being asked of
and his testimony is fairly long. I think the main thing he is
asking for is, gosh, we do get a lot of information about the
models, we do get a lot of information about loss rates, but we
are missing the parameters. Just tell us the parameters and we
will really be able to make grandma's cake. I think at some
point that becomes no longer a supervisory test but something
that lacks any kind of impact.
We have a good example of what happens when you put the
parameters of a stress test out for public comment like what is
being proposed today. The supervisor of the government-
sponsored enterprises, Freddie Mac and Freddie--Fannie Mae and
Freddie Mac, were subject to congressionally mandated stress
tests from 2002 to 2008, which today did not detect any risks.
In fact, they showed strong compliance. Why? It turns out the
supervisor, Office of Federal Housing Enterprise Oversight
(OFHEO), never tweaked its models despite growing obvious risk
in the mortgage market. One paper found that the requirements
to publish all stress scenarios, empirical specifications, and
parameter estimates in the Federal Register meant that would
have been administratively cumbersome to make any material
changes to the underlying models.
In short, supervisory stress tests are very important for
bank risk management. I am concerned, as are many others, that
the tests have become too routinized and that further
disclosures from regulators will simply make them more so.
[The prepared statement of Mr. Feldberg follows:]
[GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]
Chairman Barr. Thank you for your testimony.
We will now turn to member questions, and the chair now
recognizes himself for 5 minutes for questioning.
We will start with Mr. Gould. Will you please explain why
it is legally problematic for the Fed to keep secret the stress
test models and methodologies?
Mr. Gould. Thank you, Mr. Chairman.
Yes, so because the models and scenarios that the Fed is
using, and at least with respect to the models not disclosing,
are not put out for public comment, and because the Fed relies
upon those models and scenarios to, essentially, calculate the
stress capital buffer, which is a binding constraint on banks'
capital, those models and scenarios are, essentially,
legislative rules, and under the Administrative Procedure Act
they have to go through public notice and comment rulemaking.
Chairman Barr. Are there also legal problems with the fact
that the Fed does not give the public any opportunity to
comment on the scenarios used in the tests?
Mr. Gould. Well, yes, public comment is a requirement under
the Administrative Procedure Act (APA) for legislative rules.
So yes, that is a problem as well.
Chairman Barr. Could public comment improve the--not only
the process but the tests themselves?
Mr. Gould. Yes. So setting aside the legal requirement that
they have to do this from a policy standpoint, yes, it could
inform it. In fact, that is the entire rationale behind public
comment, generally. It is to inform the agency promulgating the
rule on how to make the rule better.
Chairman Barr. Mr. Covas, a number of commenters, including
you, have noted the overlap between the stress testing and the
Basel III Endgame proposal, as I suggested in my opening
comments. Can you talk about how supervisory stress tests
already account for market risk and operational risk?
Mr. Covas. Thank you for the question.
Yes, the Fed projects operational risk losses over the nine
quarters of the stress horizon. That threat reduces revenues
under stress for the banks subject to the stress test, so
operational risk is already accounted.
For market risk, the Fed, as for the largest banks,
subjects the largest banks to global market shock, for which
assumes an instantaneous shock to banks' market positions and
result in significant losses.
So operational risk losses account for about 25 percent of
losses of all the banks, such as the stress tests. Market risk
losses accounts for roughly 20 percent for market risk losses
for the largest banks in the stress test.
Chairman Barr. We are told that the Fed is undertaking
broad and material changes to the proposal, but if you--if the
Basel III Endgame proposal were finalized as it has been
proposed, do you think that there would be a double count of
these risks? Meaning that banks will be required to hold more
capital than is necessary to address their market and
operational risks.
Mr. Covas. So yes. The Basel III proposal basically doubles
capital requirements for operational risk and material increase
capital requirements for market risk. The way the banks
allocate capital across business lines, they look at the entire
capital requirement both based on the Basel framework as well
as the result of the stress tests. There are significant
overlaps because both of the stress losses--operational risk
losses, for example, they are stress losses and the stress
tests, by definition, also under the new Basel framework, they
also are considered to be stress losses. They are all
calibrated back to the global financial crisis, mortgage losses
that the banks experience, and they effectively are double
counted in a sense that they are requirements to hold capital
for exactly the same type of risks.
Chairman Barr. So I think that is very important that the
Fed then take into account, as they undertake these broader
material changes to Basel III, that they take into account what
the stress tests actually require as well so that there is not
double counting.
Mr. Campbell, finally, there were numerous concerns
expressed on the Basel III Endgame proposal. Fed Chair Powell
and Vice Chair Barr have both stated that these comments have
been helpful in ensuring proper consideration of the costs and
benefits and the expected decrease in availability of credit.
It seems to me that the static capital requirements have
benefited greatly from transparency and public engagement. At
the end of the day, stress tests are essentially dynamic
capital requirements, as Mr. Gould pointed out.
Is there any reason to think that stress tests would not
similarly benefit from transparency and public engagement?
Mr. Campbell. No. I think, in fact, indeed there is every
reason to believe that the stress tests would benefit greatly
from that kind of public engagement. As evidenced by the Basel
III Endgame process, there has been a voluminous amount of
comments on those rules and that proposal. Most of the
commenters, quite frankly, have been from nonbanks. As you
indicated, the Federal Reserve key senior policymakers have
indicated that those comments have been broad and helpful and
that they will result in broad and material changes. So it is
puzzling, quite frankly, to understand why the Federal Reserve
would not want to avail itself of the same benefit in the
context----
Chairman Barr. Finally, Mr. Campbell, can you address this
accusation of gaming with transparency?
Mr. Campbell. Sure. So I think gaming is a misnomer, and I
think--the issue is that banks use information in the
regulatory structure and the regulatory architecture to engage
in prudent risk management. They look at risk weights, they
look at use of capital surcharges. All of the facets of the
regulatory--bank regulatory framework are available to them.
They can look at those things. They are publicly available
because they are in the notice and comment process, and they
use that information to engage in prudent risk management.
That is no different for stress testing than it would be in
the way that firms look at risk weights in the context of Basel
III Endgame. They all feed into the total overall capital
requirement that applies to the firm and they use that
information.
Chairman Barr. Thank you. My time has expired.
The chair now recognizes the ranking member, Dr. Foster,
for 5 minutes.
Mr. Foster. Thank you, Mr. Chair.
One of the defining characteristics of the 2008 crisis was
the unanticipated level of interconnection and contagion. It is
hard for me to see how pure capital requirements can pick that
up as well as a stress test. You know, if you think of the
question of American International Group (AIG) when it came out
was no one anticipated just how interconnected AIG was and what
a body blow it would be to the entire international banking
system if AIG went bust. That became a huge issue when it, in
fact, did go bust.
Mr. Feldberg, can you talk a little bit about how a stress
test might be able to go after a contagion and interconnect in
this in a way that just pure regulatory capital requirements
might not?
Mr. Feldberg. Sure. I think the basic innovation of the
stress test is the introduction of a severe shock that is
imposed on the entire sector, the entire financial sector, to
the extent that this--the shock affects intermediaries and you
are supposed to be catching the impact that would have on
banks. I am not particularly convinced that the AIG scenario
would have been caught by the stress test the way they are
currently designed.
After the global financial crisis, there were many
responses by the regulatory community to try to address the AIG
incident and to make sure that does not happen again, ranging
from the way banks manage their risks to the way credit risk
transfer and structure credit products are handled in the first
place. I think the development of the credit risk transfer
market in recent years suggests that, that some of those risks
are rising again, and that would be a useful thing to add to
the stress test. I am not sure if it was included this year and
I hope they include it next year.
Mr. Foster. I am a little--it appears that there is a
little talking past each other. What exactly is meant by
increased transparency? You know, so radical transparency means
that the Fed just publishes all the Python source code for all
its models for all of the--which could be done, but that is
only half the problem. The public is going to understand what
is involved. They will also have access to--need access to the
granular information that you get from banks which seems would
have huge problems. I doubt that, Mr. Campbell, your members
would be enthusiastic about revealing that to the public.
How do you--it would be--so when you are talking about
letting the public see this, the public seems really to be the
analyst inside the giant banks rather than the general public.
They will be able to see how the details of the model applies
to all the private information they hold. So how do we deal
with this? If you have a lot more transparency preventing the
public from reverse engineering, many of the positions and
private information of the large banks and--or, generally, how
do you deal with that tradeoff?
Mr. Campbell. So----
Mr. Foster. Well, first off, I want to compliment you on
your--it took me two glasses of wine to get through your
written testimony, and it was excellent, very thoughtful. Thank
you.
Mr. Campbell. Great. Now, next time I will aim for three
glasses, but thanks very much.
To address your question, first of all, I think I would say
that when we talk about the public--so clearly that includes
the banks, so I want to be absolutely transparent about that.
So the public includes the banks but the public also includes
not general members of the public like people like my dad, but
people who are, say, economics professors at Stanford
University, people who are risk management experts inside the
industry, and there are a variety of people in the public that
have the wherewithal and the ability to make sense of and
understand better disclosures of who are not, essentially, John
Q. Public, but who are experts in the field. So, we should be
thinking about those people.
Then with respect to your point about the information, so
the one thing that I would point out is that the nature of the
data the Federal Reserve collects from the banks in the context
of the stress testing process is publicly available through the
Y-14 forms. Those forms are on the Federal Reserve's public
website. The data is not filled in, but the public has a window
into exactly what data is being collected by the Federal
Reserve. It has access to the instructions----
Mr. Foster. Yes. For example, Mr. Covas was complaining
about what appeared to be sort of random jumps in the stress
capital buffer. Those could be due to real differences. For
example, there is an article in The Wall Street Journal about
the synthetic risk transfers. Even though, instead of taking it
out of the banking system, it would boomerang back into the
banking system.
So that is the sort of thing that should cause a big--if a
bank gets involved in that, there should be a big variation in
the capital stress buffer if they decide to get involved in
that sort of event. The problem is that you are going to have
to reveal a lot of details about what the banks' positions
really are that caused it to have a big variation on the
capital stress buffer.
Mr. Covas. Very quickly. I mean, I do not think--so a lot
of the issues arise on the revenue side. For example, in terms
of the volatility you cited, the banks would like to know or
the public would like to know just which variables are being
used, the sensitivities and the coefficients. A lot of the
issues are due to the lack of data that the Fed is using, and I
think the banks have much more better data and more experience,
and that could be dialog similar to Basel Endgame where the Fed
receives comments from the industry and from subject matter
experts that leads to improvement of the models.
Mr. Foster. I think the gavel is coming down here, but
thank you.
Chairman Barr. Thank you.
The gentleman from Florida, Mr. Posey, is now recognized.
Mr. Posey. Thank you very much, Chairman Barr.
Mr. Covas, would you please explain how the lack of
transparency in Fed stress models resulted in uncertainty and
impairs the ability of banks to do capital management planning?
Mr. Covas. Sure. Thank you for your question.
So stress test results, the results can change every year,
and that the fact it is a capital requirement, it is not
uncommon to see increases in capital requirements in excess of
1 percentage point, and the banks have 3 months to comply with
the new requirement.
So just to give you a contrast, under Basel, the average
estimate by the agencies was about 2 percentage points
increasing capital requirements and the banks have 3 years to
comply, plus they are subject to notice and comment. So the
banks effectively have 3 months to comply with the new
requirement, which if they--no bank wants to be inside the
buffer when the requirement goes live. Usually the results come
at the end of June, go live October 1. So they have to make
significant adjustments to their portfolios if they want to
meet the new requirements.
So ex ante, in face of this uncertainty, the banks just
decide to underinvest in activities that tend to be--that are
more cyclically sensitive; namely, small businesses and lending
to households with less than pristine credit scores. They tend
to be the exposures that they tend to economize and that is how
it reduces lending and economic activity.
Mr. Posey. Okay. You have written on the impact of stress
tests on capital requirements and accessibility of credit for
small businesses. Could you summarize that research for me?
Mr. Covas. Sure. So it is basically the fact that small
businesses are a highly cyclical sensitive activity. When you
have, in the Fed, analysis assuming unemployment rate goes from
4 percent to 10 percent, the risk of lending to those borrowers
increase significantly and the banks just have to allocate much
more capital. So over time, the banks have tried--they want to
economize capital and assess the return on the investment
versus the cost of making such investments. Over time, what we
have seen in the data is that the banks, subject to the stress
tests, have moved away from these more cyclically sensitive
activities like small business lending.
Mr. Posey. Thank you.
Mr. Campbell, you did research suggesting the Federal
Reserve's stress models overestimate Gross Domestic Product
(GDP) declines and stress test losses. Could you please explain
how the overestimates capital requirements and increases the
cost of capital to banks?
Mr. Campbell. Sure. Again, briefly, you are correct that we
have done some research at the Financial Services Forum showing
that the decline in GDP that you find in the stress tests is,
larger than historical experience. So during the financial
crisis, GDP fell about--really GDP fell about 4 percent. In the
most recent stress tests it fell by 8 percent, and it kind of
bumps around in that general region. So it is effectively
overstated relative to what we have seen in other severe
historical recessions.
It is also overstated relative to other empirical sort of
regularities that relate GDP declines to unemployment. The
obvious impact there is that, although we do not know exactly
how the stress testing models are put together, real GDP
declines are a driver of losses in all the stress testing
models. So whether we are talking about auto loan losses,
mortgage losses, business loan losses, a decline in GDP is
going to result in those loan losses.
If you ramp up the declining GDP, you are going to ramp up
the loan losses. That is going to make those assets look
riskier, maybe riskier than they actually are, given the
overstatement and the decline of GDP, and that is going to
result in increased capital requirements which has the natural
effect of raising the cost of funding and then incentivizing a
pull back in lending to those sectors into those markets.
Mr. Posey. Thank you very much, Mr. Campbell.
Mr. Chairman, I yield the balance of my time.
Chairman Barr. The gentleman yields.
Will Mr. Campbell please pull the microphone a little
closer to you for future testimony. Thanks.
The gentleman from California, Mr. Sherman, is now
recognized.
Mr. Sherman. It is not surprising in a capitalist country
that we have fights over the allocation of capital and there is
a fight between blue collar families operating small
businesses, pizzerias, plumbing supply operations, and Wall
Street, that prefers billion and multibillion dollar
transactions. All the decisionmakers are grad school educated,
people wearing ties like me and you.
I will not say that the decisionmakers and the regulators
hate the blue collar families, they just do not feel
comfortable with them. So when we have to make a decision as to
where the money goes, it does not go in $250,000 loans to
pizzerias. It goes in multibillion dollar transactions on Wall
Street, and I see these stress tests as pushing us further in
that direction.
You can see from a bank examiner's standpoint, Silicon
Valley Bank buys $50 billion worth of treasuries, oh, that
looks wonderful. Wall Street understands that. All the files
are perfect. We kind of know what the credit rating is of the
borrower. If you actually look at the file for a $250,000 loan,
it is a bit messy, some things are handwritten. It would not
get you an A at any major business school. So this capital is
shown because of--this fight--is shown by Basel III proposals
that give only a 65 percent rate rather than a 100 percent rate
when the borrower is a publicly traded company.
The greatest problem we have is the fight between interest
rate risk and credit risk. The pizzeria loan is nothing but
credit risk. The Treasury investment is nothing but interest
rate risk. So we have a system that focuses on the credit risk
and deliberately ignores the credit risk--the interest rate
risk. They do it deliberately and they do it by what I call the
stupid depositor model, which is that if interest rates go up,
your whole portfolio of treasuries goes down in value, but the
bank makes money because it has lots of stupid depositors who
will leave their money in non-interest bearing accounts and not
take it out and not put it in money markets yielding 5 percent.
You know who turned out to be stupid? Not the depositors at
Silicon Valley Bank; the bankers at Silicon Valley Bank and the
regulators of Silicon Valley Bank.
So I want to ask each of our witnesses, are you aware of
any stress test that focused on the risk of interest rates
going up, a pretty obvious risk, because 5 years ago, interest
rates were 0.0. They had only one way to go, and did that, a
test, a stress test for interest rates going up and depositors
not being stupid. Any of our witnesses aware of any such stress
test?
I do not see--oh.
Mr. Feldberg. So the interest rate risk, the Basel standard
that I mentioned, is essentially a stress of a portfolio
against changes in interest rate risk and interest rates.
Mr. Sherman. That is proposed for the future. We have had
stress tests for a while.
Mr. Feldberg. No, other--other countries been doing it for
20 years almost.
Mr. Sherman. Has the United States, as our regulators, done
a stress test looking at the very stress that was pretty
obvious that interest rates would go up having been----
Mr. Feldberg. No, that is exactly right, they did not do
that.
Mr. Sherman. Then even worse for Silicon Valley Bank, et
cetera, two worst things: One is they did not even have to face
a stress test because they were below the limit and the limit
should have applied. Then second, to illustrate how it is not
the depositors that are stupid, Silicon Valley Bank had, in
fact, an insurance policy against interest rates going up and
their portfolio bonds going down. You know what they did with
that insurance policy? They sold it to generate money. They
could generate bonuses for their executives.
I realize I have used a lot of my time. I will just say
that I do not see there is much excuse for the regulators not
to publish everything after the test. You do not give away the
questions before the test but after the test. This is a
democracy, things are supposed to be public.
I yield back.
Chairman Barr. The gentleman yields.
The gentleman from Missouri, Mr. Luetkemeyer, is now
recognized.
Mr. Luetkemeyer. Thank you, Mr. Chairman.
I thank our witnesses for being here today, an interesting
discussion today.
What is the purpose of the stress tests? Mr. Campbell, what
do you think the purpose of a stress test should be?
Mr. Campbell. I think the purpose of the stress test should
be to assess the capital adequacy of the Nation's largest
banks.
Mr. Luetkemeyer. Okay. If you are going to assess that,
would you not want to know how you are going to be assessed, so
that you can manage your bank accordingly? I mean, that would
seem to me what I would want to know.
Mr. Campbell. Yes, absolutely.
Mr. Luetkemeyer. So what we have here right now is a game
of gotcha, if we do not have the Fed be able to give us the
information that is going to be on the stress test so that we
can change our model. To me, the stress test should be the
Fed's own thought process of what the next crisis is going to
be, what the next concern should be with regards to managing
banks and our economy coming down the road.
Mr. Sherman made a great point here a minute ago with
regards to the interest rate risk that got Silicon Valley and
several other banks. It was attested for. Is that correct?
Mr. Campbell. I am sorry, I did not quite hear you.
Mr. Luetkemeyer. Okay. The stress test that Mr. Sherman was
talking about with regards to interest rate risk, that was not
on the stress test.
Mr. Campbell. Correct.
Mr. Luetkemeyer. I think it is being incorporated now, if I
understand it correctly. So they missed the future problem that
could have--that they should have been able to see, as you
pointed out. Everybody should know if they are interest rate's
at zero, down the road you are going to see them go up.
To me this stress test should be something that would help
banks manage their future problems according to what the Fed
sees those problems to be. So the discussion then is, why do we
not want to know those things ahead of time so that we can
actually manage our banks? You want to wait a year before you
get your stress test, before you can actually change your
model, so that you can minimize your risk? Does that make sense
to you?
Mr. Campbell. No, not generally.
Mr. Luetkemeyer. Mr. Gould, what do you think of my analogy
so far in my discussion, am I on point or am I off on this?
Mr. Gould. I think you are on point, Congressman. I would
just note that I think the underlying analogy to the concept of
gaming is fundamentally off, right. This is not a situation
where testing the bank's knowledge, and so we do not want to go
into the answer set. Capital--this is a compliance obligation
and we are setting capital requirements. Inherent in setting
capital requirements is establishing incentives. Some of the
very incentives that Congressman Sherman was mentioning have
altered the nature of the banking industry and the clients and
customers it serves. So I think it is very important that we
know these in advance.
Mr. Luetkemeyer. So to me the argument about gaming the
system is false. I think we need to be showing how this test
could actually help banks manage their banks better. That is
what it should be all about at the end of the day. Have a
better banking system and a more sound economy and everything
is going to run much better, but anyway.
Mr. Gould, you went into a lengthy discussion, I know Mr.
Covas also had some comments with regards to the Administrative
Procedures Act and the fact the Fed does not adhere to that. So
if they do not adhere to administer--that is a law, by the way.
If they are not adhering to that, are they in violation of the
law?
Mr. Gould. Congressman, that is up for--that is for a court
to decide. Certainly, I have not heard a convincing defense as
to why they are not providing a notice in common on these
models.
Mr. Luetkemeyer. You are an attorney, are you not?
Mr. Gould. Yes, sir.
Mr. Luetkemeyer. Do you advise banks, with your practice?
Mr. Gould. I do.
Mr. Luetkemeyer. Do you advise the banks to go after the
Fed and sue them if they are going to try and impart this thing
on to them without going through the proper procedures and it
is in violation--it is a violation of law?
Mr. Gould. Look, I think that is a individual determination
that banks have to weigh pros and cons because, as you know,
banks have a lot of----
Mr. Luetkemeyer. Okay. Well, you were advising the American
Bankers Association or the--which represents mostly big banks,
would you advise them to take this to the court?
Mr. Gould. Congressman, I would certainly advise them to
consider it strongly.
Mr. Luetkemeyer. Okay. Very good.
Question number three: One of the things that concerns me
is that we have a situation here with this proposal that the
Fed is going through with stress testing and it seems very
inconsistent when you look at Current Expected Credit Losses
(CECL). Here they asked the banks to be able to do their own
modeling with regards to the local economy, with regards to how
they should determine loan loss. You have one bank on one side
of the street, one bank on the other side of the street come up
with completely different models and that it could be okay.
So now we are talking about stress tests, it goes across
everybody, and they are sitting here with no ability to have an
impact. To me, whenever you have people--the reason for comment
period, by the way, is to have people who actually know what
they are doing in the business to be able to tell you whether
it is going to work or not.
Mr. Chairman, I will yield back. The balance of my time is
gone.
Chairman Barr. The gentleman yields.
The gentleman from Georgia, Mr. Scott, is now recognized.
Mr. Scott. Thank you, Mr. Chairman.
You know, we worked on this. I felt this was a real
critical part of Dodd-Frank. Many of us on this committee were
around back then and put a lot of work into it.
Mr. Feldberg, there is one piece that we have been leaving
out here. There are two pieces to this, and we talked about the
banks, and that is the nonbanks. I want you to kind of delve
into that a moment. Because oftentimes we miss the real deal in
solving a problem, much like we did with Dodd-Frank, and we had
to call attention, I know I did with Barney Frank, and say we
cannot solve this with just bailing out the banks. We have to
do something else for those people who are losing their jobs,
who are losing their homes, and we came up with the hardest
hit.
Tell us about the nonbanks. How can shock scenarios and
hypothetical models be designed to best analyze the
interconnections between the banks and the nonbanks?
Mr. Feldberg. That is a very good question. Thank you.
Other countries have taken a much more aggressive approach
on this. The bank of England is currently doing an exploratory
analysis----
Mr. Scott. You said the Bank of England?
Mr. Feldberg. Yes, of banks and nonbanks. The U.S. has not
done this yet. I think part of the--another aspect of Dodd-
Frank was designating systemic firms that were not paying for
supervision, which might have led to stress testing of those
entities but we have not attempted that yet. There are models
around the world, other Central Banks, the IMF, that conducts
stress tests on the nonbanking sector, and that is something
that the U.S. ought to be doing too. It is harder when you do
not have supervisory regulatory authority over an entity to
involve them in a stress test.
Mr. Scott. Do you feel it is possible to build a process
through which banks could replace some Fed models with their
own model, subject to careful consideration and Fed approval?
Mr. Feldberg. So I would not recommend replacing the Fed
stress test, but the banks that do have their own models, and I
think the existence of both Fed models and the banks' own
models is a good thing to have both of those checks over the
risk management of the bank.
Mr. Scott. Yes. Now, how can Federal regulators safeguard,
search a process against poor quality of risk management or
without sacrificing the Fed's goal of consistency across the
institutions subject to stress testing, and that is
particularly for the most poorly managed firms.
Mr. Feldberg. So there used to be a qualitative objection
that the Fed raised in the stress testing process, which has
been taken out in recent years but the Federal Reserve
supervisory process should allow for what you just said quite
extensively. If the Fed is unhappy with an outlying bank's risk
mismanagement practices, the Fed's supervisory authorities
should be quite extensive for addressing that through the
traditional tools.
Mr. Scott. Do you feel that we are on the right course in
this? Are there any other things that we on this committee can
do to make sure that never again will we have the bank
failures, the collapse of our financial system? That is the
whole purpose of the work we did with Dodd-Frank. Is there any
more we need to do, particularly in this area that you and Yale
University are doing such great work on?
Mr. Feldberg. There is certainly more that could be done
with non-banks. I mentioned interest-rate risk management. I
think part of it is keeping track of financial stability
risks----
Mr. Scott. You said interest rates----
Mr. Feldberg. Risk management, oversight----
Mr. Scott. Good. All right.
Mr. Feldberg. Then holding hearings like this to identify
financial stability risks that might be overseen--or, might be
missed by the regulators, like Congressman Foster mentioned,
systemic risks----
Mr. Scott. Thank you very much, Professor.
Chairman Barr. The gentleman's time has expired.
The gentleman from Texas, the chairman of the Small
Business Committee, Mr. Williams.
Mr. Williams of Texas. Thank you, Chairman.
Thank you all for being here today, and I am from Texas.
Now, the Federal Reserve stress-testing scenarios can play
a key role in enhancing the stability of the financial system.
Annual stress-test scenarios can increase the public's
confidence in banks' ability to handle economic headwinds and
maintains confidence in the overall U.S. banking system.
However, the Federal Reserve continues to limit
transparency by failing to disclose their stress-test models--
we have talked about that--while also denying stakeholders and
the general public the opportunity to provide feedback on
stress-test scenarios.
So greater transparency and accountability in stress tests
provides all stakeholders, big and small, with the opportunity
to identify flaws and suggest improvements to make the
scenarios more useful and the outcomes more accurate.
So, Mr. Gould, can you please elaborate on the benefits of
greater transparency in stress testing and why a proper notice-
and-comment period is important?
Mr. Gould. Yes, Congressman.
First of all, as a matter of administrative law, if this is
a legislative rule, which it certainly appears to be, then the
models and scenarios themselves must be disclosed and they must
be subject to public comment.
From a policy standpoint, public comment is really
important because it actually, ideally, makes the ultimate work
product better, number one. Number two, it also, I think,
increases credibility in the agency that is promulgating the
legislative rule. Three, it serves as an element of
accountability for that agency as well.
Mr. Williams of Texas. Okay.
The Fed's stress-testing regime is a black box comprised of
thousands of pages of complex data, formulas, and models. This
complexity and lack of transparency makes it difficult for
outsiders to understand how these tests work or how they can
predict their own outcomes. So, consequently, both banks and
the public often struggle to grasp the full implications of the
results.
Now, the complexity of the process can lead to uncertainty
and mistrust among stakeholders who rely on these tests to
ensure financial stability.
Mr. Covas, can you discuss how the secrecy and lack of
information on the stress-test scenarios makes it harder to
evaluate the analytical effectiveness and accuracy of the test
outcomes?
Mr. Covas. Thank you for your question.
SThe Fed models, because they are based on data that is not
very granular, the results change significantly from 1 year to
the next, just due to changes in banks' balance sheets and the
fact that the scenarios themselves are also changing in ways
that are unpredictable.
Now, the results--for example, what happened during the
post-pandemic, the Fed increased significantly its balance
sheet. By buying Treasury securities by the value--the amount
of $1 trillion, banks' balance sheets increased significantly
with the reserve balances. According to the Federal Reserve
models, the higher is your balance sheet, in particular by
having--the higher are your expenses. That increases banks'
capital requirements in the stress test, because they are
holding more reserve balances, which is at zero to no
interest--no interest expense.
As a result, the banks need to price in the increase in
capital requirements in the loans they are making to households
and businesses, and as a result, that reduces credit
availability and ultimately impacts--negatively impacts
economic growth.
Mr. Williams of Texas. Okay. Thank you.
Over the last year, I have talked at length about the
dangers of increasing capital requirements on banks and the
impact that these changes can have on businesses' ability to
access reliable credit.
Now, a proposal like Basel III Endgame forced banks to hold
more capital, which in turn increases overall borrowing costs
and reduces availability of credit for businesses and
consumers, in many cases, right down Main Street.
So this is why I am extremely concerned with the lack of
transparency surrounding stress tests, because stress tests can
be used to set binding capital requirements for banks. We must
ensure transparency and public engagement on stress testing to
reduce any redundancies which the capital requirements require.
Mr. Campbell, quickly, what are the real-world impacts
stress tests have on banks' capital requirements and how does
this impact the cost of credit and access to capital for
businesses like mine?
Mr. Campbell. Sure. In the last couple of years, the
Federal Reserve has undergone actually a massive project to
formally include the stress-testing process into the formal
bank regulatory capital framework. Today, stress tests are run,
and those result in what has already been discussed today, is
the stress capital buffer. The stress capital buffer is indeed
a regular element of a firm's required bank capital
requirement.
So, after running the stress test, if their stress capital
buffer goes up by, say, a full percentage point, that increases
the amount of capital that they need to maintain against all of
the assets on their balance sheet. As they need to raise more
capital on their balance sheet to support the assets on their
balance sheet, that increases the cost of funding, because bank
equity capital is the most expensive form of finance. So, as
they raise more capital, that raises the cost of funding.
As that raises the cost of a bank's funding, for any
business--it does not matter if you are a bank or a car dealer
or a bakery--when the cost of your inputs goes up, you turn
around and you have to charge the customer more. So,
ultimately, that is going to impact people who are borrowing
from banks, and that is going to increase the interest rates
that they get charged on loans.
Mr. Williams of Texas. I do not raise the cost of my cars.
Chairman Barr. The gentleman's time has expired.
The gentleman from Texas--the other gentleman from Texas,
Mr. Green.
Mr. Green. Thank you, Mr. Chairman.
I thank the ranking member.
I thank the witnesses for appearing.
I would like for Mr. Feldberg to respond, please.
Sir, if we make the banks aware of what we are testing,
would that encourage them to be prepared for things that are
not being tested, such that they would not take into
consideration some of the many other concerns that may hurt us
when we have a downturn?
Mr. Feldberg. Sorry. Could you rephrase that?
Mr. Green. Yes. I am simply saying that if the banks are
aware, if they know what the testing is all about----
Mr. Feldberg. Uh-huh.
Mr. Green [continuing]. I am confident they would be
preparing for this.
Mr. Feldberg. Uh-huh.
Mr. Green. What about the many other things that they might
avoid or not prepare themselves for because they focused only
on what they are concerned with in terms of passing a test?
Mr. Feldberg. Uh-huh.
Mr. Green. Your response, please.
Mr. Feldberg. Yes, no, I think that is a fair point.
I think some disclosure--the Fed discloses quite a lot, as
I pointed out, and I really think--I really want to emphasize
that point. The Fed discloses a lot about the stress-test
methodology, more than any other country. To some extent, that
is a good thing. I think the argument I was making was that, at
some point, you are providing so much data that all they are
doing is optimizing to the test, and it is----
Mr. Green. Exactly.
Mr. Feldberg [continuing]. and it is diluting the point of
the exercise.
Former Vice Chair of Supervision Dan Tarullo, who was the
leader who led the stress-test process through much of the
first 5 or 10 years that it was run, has expressed concern
about the excessive disclosures that are already being made, to
some extent, and that could be made under what has been
contemplated. He has made the argument that if you are really
going to go this far and, what I would call, just make it like
the OFHEO process that had such dire results with the
Government-Sponsored Enterprises (GSEs), you might want to just
consider something else.
It was kind of shocking to read Dan Tarullo's paper in
Brookings just a month or two ago, where he said at this point,
maybe we should just take away the connection of the stress
test to capital. Because if we are going to go ahead and make
it this predictable and easy to game, then really maybe we just
need another tool. It was very distressing to see him say that.
He laid out what it would look like if we were going to
decouple it from capital.
Mr. Green. If we should decouple, what is that other tool?
Mr. Feldberg. So, I mean--so, in his view, if you were
going to, say--and I think it should be pointed out that other
countries that do stress testing do not tie it as directly to
capital as we do. So, essentially, it would just be looking
like other countries. It would be an exploratory scenario that
would help the regulators--inform the regulators on what the
capital charges would be for individual banks.
That, of course, the way it works in other countries, would
probably raise even more objections as far as transparency. I
mean, if you look at Bank of England, any country, they conduct
the stress test, and they say, this informs our individual
capital charge for that bank. So, essentially, that is more of
a black box than we are talking about here. So I do not think
the banks are going to want to see that alternative.
Whatever you do, Tarullo made the point that if you are
going to get rid of the stress test and its impact on bank
capital, you are going to have to raise the basic level of
capital before the stress test, just because you have gotten
rid of all that sensitivity and analysis and its role in the
capital process.
Mr. Green. Thank you, Mr. Chairman. I yield back the
balance of the time.
Chairman Barr. The gentleman yields.
The gentleman from Tennessee, Mr. Rose, is recognized.
Mr. Rose. Thank you, Chairman Barr, and thank you to our
witnesses, and also Ranking Member Lynch for holding the
hearing today.
Mr. Covas, in your testimony, you highlight that the
reconsideration process of stress tests are inadequate. Could
you--or, is inadequate. Could you elaborate as to why you
believe that?
Mr. Covas. Thank you for your question. Yes.
Since the stress capital buffer framework was put in place
in 2020, we have had eight reconsideration requests, for which
the Fed has effectively denied those requests.
However we do not know--to the best of my knowledge, there
has been no change in adjustments in the stress capital buffer
of those banks, and there is no explanation of how those
reconsideration requests are currently being addressed in terms
of change in the models and any timelines of how they are
addressing the shortcomings of the models that have been
pointed out by the firms.
Mr. Rose. So it is truly a black box out of which the----
Mr. Covas. Yes. It is very difficult to understand how the
results change from 1 year to next, despite some of the
disclosures. I have been following the stress tests for some
time. It is always--whenever we get the results to compare
against our expectations, it is very difficult to know exactly
how the results change from 1 year to the next.
Mr. Rose. Has there ever been a situation where the Federal
Reserve has admitted fault and revised the stress capital
buffer for a financial institution that appealed?
Mr. Covas. To the best of my knowledge, no.
Mr. Rose. Okay.
How would establishing a truly independent review mechanism
with clear procedures and timelines solve the problems with the
reconsideration process, in your opinion?
Mr. Covas. So, just forcing--adding much more transparency
around the process is going to force the Federal Reserve to
just disclose a lot more information for the public and for the
firms themselves that can analyze.
For example, notice-and-comment on the models would address
many of the questions the banks have in terms of the behavior
of the models. If they still disagree of the relationships,
they can, again, through the notice-and-comments process, try
to improve the models.
At the same time, this does not--is going to obligate the
Fed to actually follow the comments provided by the industry.
The Fed could always deny and not follow those.
So I do think it is important to have a third-party
validator that just reviews the interchange of information
between the commenters and the Fed and provides
recommendations. You know, a good example would be the
Government Accountability Office (GAO) to perform that role.
Mr. Rose. As a recovering lawyer, the failure to solicit
public comment on stress models and scenarios appears to
violate administrative law. Your thoughts on that?
Mr. Covas. So, coming from an economist, the Stress Capital
Buffer (SCB) is effectively a capital requirement, no different
from the Basel capital requirements that we have been
discussing over the past year. So they are rules that affect
banks' decisions in terms of capital allocation. So, yes, I do
think they should be subject to the same requirements as the
Basel Endgame proposal, for example.
Mr. Rose. Mr. Gould, I know that you share these concerns
as well. Have the courts reached any conclusions as to whether
models treated by agencies as binding must go through public
comment?
Mr. Gould. Yes, they have, and the courts have concluded
that models that are treated as binding by the agencies are
subject to the public notice-and-comment.
Mr. Rose. You think that would be appropriate here?
Mr. Gould. It would seem to me. Again, I have not heard
from the Fed any justification for an exception to the
legislative rule process.
Mr. Rose. You have also described the action of the Federal
Reserve as appearing arbitrary and capricious. Can you explain
why the Federal Reserve might appear that way due to its
failure to disclose its models?
Mr. Gould. Yes. I think this could come up, too, in the,
kind of, reconsideration process. Because the Fed has not
provided transparency around its models, it is very hard for
them to fully explain how they arrived at the outputs to those
models.
Mr. Rose. Thank you. I appreciate that.
As a former bank board member, I have witnessed firsthand
how regulation can be burdensome and ineffective.
Mr. Campbell, do you think that stress tests are useful for
banks and supervisors?
Mr. Campbell. I think they have the potential and the scope
to be useful for banks and supervisors. I think the current
implementation and the lack of transparency limits their
usefulness considerably, and improved transparency would
improve the extent to which they can be useful to banks and
supervisors.
Mr. Rose. What are some realistic changes that would
provide immediate benefits to both the banks and supervisors,
in your view?
Mr. Campbell. As we have already discussed today, greater
transparency on the implementation and the specification of the
stress-test models themselves, as well as greater transparency
on the scenario generation process, I think, would help to
foster greater understanding amongst banks, amongst the public,
and amongst regulators as to exactly how the models work, how
they are risk-ranking different activities that are sitting on
banks' balance sheets, and would help improve capital
management decisions and understanding of the banking sector
more broadly.
Mr. Rose. Thank you.
I see my time has expired. Thank you, witnesses, and I
yield back.
Chairman Barr. The gentleman yields.
The gentleman from California, Mr. Vargas, is recognized.
Mr. Vargas. Thank you very much, Mr. Chairman.
First of all, I would like to say that I think this has
been an excellent hearing. I am not exactly sure what side I
fall on now. I am a little more agnostic, I think, to this
issue. I do see both sides, to be frank.
I mean, I took a lot of tests in my life. If I had the
questions prior to the test, I think I would have done better
on some. I do agree with Mr. Sherman; after the test is over, I
do not understand why it is not disclosed completely. I do not
understand that aspect of it, to be frank. So I am not sure.
Whether it conforms to the APA or any other rules, I am not
sure. For years here, I kept hearing how the Consumer Financial
Protection Bureau (CFPB), their funding source, did not conform
to the law. Of course, the Supreme Court came down quite
clearly and said, of course it does, it is right down the
middle, it is no big deal. So I am not sure here either if it
does or does not. I know you have some stronger opinions here.
I do want to--and I always ask this question, so I am going
to ask it of you. I think it would be unfair if I did not. How
many of you believe in climate change, that the climate is
changing?
Okay, all of you. I appreciate that. In a reasonable group
like this, I would think that we would come up with that
reasonable answer, and we did.
I do want to ask about that, because I do think that is a
risk that we are not looking at strongly enough and
transparently enough, frankly. I know that the Securities and
Exchange Commission (SEC) now has taken a look at this, and I
know the Fed is looking at it. They have conducted a pilot
project on this program.
I wanted to tell you, I believe climate change is real, and
we have seen things in my own district--we just--I represent
San Diego--we had a 1,000-year event. You know, we had one of
those 50 years ago--not even 50 years--20 years ago, and it is
another 1,000-year event. It is not a 1,000-year event if it
happens every 20 years.
So obviously the climate is changing. I want to know if
these stress tests that we have are, in fact, taking that into
account. It is hard to know if we do not know what they are--
would someone like to comment on that?
Mr. Feldberg. I----
Mr. Vargas. Yes, go ahead, sir.
Mr. Feldberg. I mean, I think the long-term changes in
climate are just way beyond the timeframe of the typical stress
test. Even having said that, the climate stress scenario that
the Federal Reserve just conducted, I think, was extremely
useful in understanding these risks and some of those risks
really should show up in the next--in the planning period for a
stress test when----
Mr. Vargas. Well, I do not know that they are way beyond
the time--I mean, you see all the things that are happening
now. In fact----
Mr. Feldberg. So----
Mr. Vargas [continuing]. examples, you have data gaps right
now in real estate exposures. You have----
Mr. Feldberg. That is right.
Mr. Vargas [continuing]. insurance and infrastructure
exposures. The banks' financing, I am not sure that it is not
within the scope and timing of these tests.
Mr. Feldberg. So----
Mr. Vargas. Can you comment on that?
Mr. Feldberg. So, I mean, I think, to the extent that a
bank is experiencing--has real estate on the coast and is
experiencing borrowers whose insurance rates are going up and
that is affecting their credit risk, we should not even be
talking about climate scenarios; that should already be in
their credit risk analysis. I mean, there should not be
anything controversial about that.
Mr. Vargas. I do not think there should be.
Would anyone disagree with that?
Mr. Campbell. Yes.
Mr. Vargas. Would anyone disagree with that, that it
already should be in the models?
Mr. Campbell. I think, to the extent that climate risk is
approximate in first order of risk on bank balance sheets, then
it should be included in the models.
I think there is an open question as to whether or not
climate risk is a first-order imminent risk in the banking
sector that would have material consequences to the banking
sector. I think the available evidence and research suggests
that it is not. As to Greg's point, especially over the time
horizons over which we think capital is a useful economic
resource, it is not clear that climate risk is a first-order
issue.
To the extent that it is an important economic risk, then
it should be captured in any risk model.
Mr. Vargas. Anyone disagree with that?
I mean, that seems reasonable to me. I mean, I would
disagree that it is not. I think it is but reasonable people
can disagree. That is why it should be tested, it should be
looked at, the information, the data should be there.
Mr. Feldberg. I mean, I think--I mean, the insurability of
coastal areas is already in question. I mean, this is happening
now.
Mr. Vargas. Uh-huh.
Mr. Feldberg. That part is happening now.
The harder question, which the Fed also tested for, is the
transition risk and that is the risk, very simply, of what
happens when policy really finally transitions, if it does.
Mr. Vargas. Uh-huh.
Mr. Feldberg. So they tested net-zero 2050. The idea is,
somehow the global community is going to agree that we need to
do whatever it takes to get to net-zero by 2050, and what does
that mean for the financial system? That is not going to----
Mr. Vargas. Right. My time is about----
Mr. Feldberg. That is not going to happen in the next 24
months, but it is a serious risk----
Mr. Vargas. Yes.
Mr. Feldberg [continuing]. that bankers should be
considering in----
Mr. Vargas. I have 10 seconds, so I am just going to say, I
agree with what has been said here. I think it has been a very
good hearing, and I appreciate it. I think it is important that
this information, the stress-test information, be disclosed at
some point.
Thank you.
Chairman Barr. The gentleman from Illinois, Mr. Casten, is
recognized.
Mr. Casten. I did not expect to go so soon, but thank you.
Great minds are thinking alike with Mr. Vargas here. I want
to follow up on this climate-risk thread. I guess I want to--I
guess I am going to start with you, Mr. Campbell, since your
clients are the GSIBs who are well-exposed to these stress
tests.
Broadly speaking--and tell me if I am missing anything--if
I am a GSIB, I have a risk on my balance sheet, I can either--
to the extent I want to hold it and it is insurable, I can
insure it and I can offload that risk onto an insurance
company; to the extent I have a buyer and I can offload it, I
can sell it or I suppose I can write it off.
Is there any other--if I do not want to hold the risk, do I
have any other options, or have I accurately captured all the
choices I have?
Mr. Campbell. If you do not want to hold it?
Mr. Vargas. If I do not want to hold the risk, I can either
insure it, sell it, or write it off.
Mr. Campbell. Some combination of that sounds right to me,
yes.
Mr. Casten. Okay.
I asked the question because the characterization that
climate risks are not near-term risks, I think, sits a little
awkwardly with the reality, that we have actually seen lots of
data that the more likely you are to hold a mortgage in a
flood-prone area, the more likely you are to sell off that
mortgage to Fannie and Freddie, which is path one. We have seen
a whole bunch of areas of, can we get insurance? We have seen
Fannie and Freddie move to, you know, are the reinsurance
markets there, which we can insure it? Then, of course, we have
these reports and stories about people in Texas and Florida who
are now just sitting on uninsured homes where they are
essentially writing it off, right?
So the risk is there. I think the question is not so much,
do the banks perceive the risk, as much as, are the banks able
to offload that risk into some other part of the financial
system?
So, if we agree that they have the risk, then--I guess I
will pivot to you, Mr. Feldberg--do you think that FSOC has the
ability, authority, talent to identify where that risk is
pooling in our financial system?
Mr. Feldberg. So I think the FSOC certainly has the
responsibility to do it. A lot of what you are talking about
means the insurance sector, which is still State-regulated in
the U.S. The State regulators sit on FSOC. So, to that extent,
the FSOC could raise the issue with insurance.
It really is an issue that has to be looked at in terms of
banks and insurance companies at the same time----
Mr. Casten. I mean, I get concerned----
Mr. Feldberg [continuing]. because of the interconnections
between them on this issue.
Mr. Casten. My wife is in the insurance industry, and she
has no obligation to provide insurance in every State in the
country.
Mr. Feldberg. That is right.
Mr. Casten. Right? So I get nervous that does not eliminate
the risk, and we are seeing it pool up in those other places.
Mr. Feldberg. Uh-huh.
Mr. Casten. I raise all that because I think this question
about transparency--I share my Jesuit friend's agnosticism. I
have never met a buyer who does not want transparency, and I
have never met a seller who wants transparency, right?
So, if we were to go and identify all of these risks--and
let us say we are sitting here and we say, there are risks in
this society, and, by the way, maybe there are some of these
climate risks that are held by the GSIBs, which is not
insurable, which--the buyer is sufficiently informed that they
do not want to buy it at a price you are willing to sell----
Mr. Feldberg. Uh-huh.
Mr. Casten [continuing]. should we publicize that? Are we
on board for transparency in that case or does that risk
jeopardizing the solvency of the GSIBs because they now are
sitting on risk rather than being able to offload that into
some other part of the economy that is not subject to stress
tests?
Does anybody have any--what should we do in that case?
Should we make this information public, or should we punt it to
a future Congress?
Mr. Feldberg. That is a good question. I think we have been
talking about the disclosure of the inputs into the process,
and now you are bringing up the disclosure of the outputs of
the process, which--there are also tradeoffs, like I mentioned
in my paper, there.
You certainly want the disclosures about stress-test
results to be informative to the public but you also do not
want them to raise undue concern about individual institutions.
It is part of the supervisory process, which has been a very
confidential process since it was started 150 years ago.
Certainly, information like that, you might not want that to be
public.
Mr. Casten. Yes. Well--and if any of you have thoughts to
continue that conversation.
There was a study out of Harvard that I think was in a
recent Senate hearing, that they found that real estate in the
U.S. may be overvalued by as much as $237 billion because there
are sort of second-tier rating agencies that are providing
insurance for second-tier banks that are therefore able to
offload this onto Fannie and Freddie.
I get it, right? Like, if there is a sucker who is willing
to buy something from me, I will sell. All of us, whether as
regulators or legislators, are ultimately going to have to fix
that.
I think--I will close where I started. These climate risks
are big, and they are real, and they are now but we need to do
stress testing of the whole financial system, not just of the
six biggest banks.
I yield back.
Chairman Barr. The gentleman yields.
The gentlewoman from California, Mrs. Kim.
Mrs. Kim. Thank you, Chairman Barr.
I want to thank the witnesses for joining us today.
I am all for ensuring the stability of our financial
system. However, when we talk to small businesses and
households, we should also be frank and transparent about how
increasing capital requirements and making models in scenarios
less predictable can reduce lending and access to capital.
First question to you, Mr. Campbell. If the Fed decides to
broaden the scope of stress-test scenarios without making them
more transparent, can you elaborate on how this decision could
change stress capital buffer requirements?
Mr. Campbell. Sure. So you have in mind the idea that the
Federal Reserve would use more and different kinds of scenarios
rather than a single scenario, just to make sure I understand
the question?
Mrs. Kim. To me, it seems that it could make them less
predictable. So, if----
Mr. Campbell. Yep.
Mrs. Kim [continuing]. multiple stress-test scenarios make
capital requirements less predictable----
Mr. Campbell. Yes.
Mrs. Kim [continuing]. how can that influence lending to
small businesses and families?
I ask this because my fear is that, with high interest
rates, mortgages and other financial tools, that will increase
in cost due to unpredictability of stress tests.
Mr. Covas, do you have anything to add to this question?
Mr. Covas. Sure. I would just add that, if you are
increasing scope regarding stress-test results, the banks will
want to hold more capital as a precautionary buffer. As I said
earlier, no bank wants to be below their minimums by the time
the requirements go live. So, yes, that will reduce credit
availability and provision of capital to markets.
Mrs. Kim. Thank you.
We understand that Basel III Endgame proposal would
increase the capital banks must hold with respect to market
risk and this is the same risk that is being stress-tested in
the global market shock component. I would say that this sounds
like double counting of these risks to me.
I know Chairman Barr touched on the double-counting aspect.
So if you could tweak either the market risk component of the
Basel III proposal or the global market shock component, which
one of the two components do you find less necessary and would
you eliminate one or both?
Mr. Covas. I mean, I can answer.
Mrs. Kim. Mr. Covas, uh-huh?
Mr. Covas. I mean, we were disappointed in terms of the
Basel proposal. You want to look holistically about the
requirements and look at the requirements at the product level
and see which ones are in excess. I think you want to address
both of the requirements, potentially.
So the current--the existing market risk framework has some
flaws for which the Fed, through the stress test, has
enumerated some of those flaws through the global market shock.
The new proposal fixes some of those flaws.
So I do think you ought to go back to the global market
shock and make some adjustments so that ultimately you are not
overcapitalizing for market risk.
Mrs. Kim. Mr. Campbell, could you add to that, please?
Mr. Campbell. I mean, I would broadly agree with that
sentiment.
I think the most important thing to ensure is that the
global market shock and the treatment of market risk assets in
the context of Basel III Endgame work in a conceptually
consistent and logical manner and that they do not, basically,
work at cross-purposes with each other. I think that is part of
what is going on in the current regime.
Mrs. Kim. Well, let me move on to another matter. It has
been suggested that being more transparent with models and
scenarios could lead to gaming and that it could be like the
teacher revealing the questions to the students before the
exam.
Mr. Covas, do you think the Federal Reserve Board revealing
methodologies and increasing public engagement on the stress-
test framework would lead to banks gaming the stress test?
Mr. Covas. As we discussed in the hearing, no. I mean,
again, this is--the stress tests are a capital rule. The banks
need to allocate the portfolio allocation of capital through
the different products they offer. They need to know what the
capital requirement is.
You know, banks change their portfolios driven by business
decisions and market demand, not because they want to game or
reduce the capital charge in the stress tests.
Mrs. Kim. Uh-huh.
Mr. Covas. The models, if they are accurate--the banks have
no incentive to game accurate models. If they are inaccurate,
they should be put out for notice and comment to fix the
models.
Finally, the Fed already has processes in place to look at
banks' exposures through the global market shock. Again, these
allow--they have been doing that for the last at least 3 or 4
years, where they look carefully at banks' exposures and if
they think the banks are--some of the exposures are just to
reduce the charge in the stress test, they are disallowed. So I
think----
Mrs. Kim. Mr. Campbell, on the same topic, can you explain
why it is difficult to game stress tests when you are talking
about hundreds of billions in the balance sheet of a bank?
Mr. Campbell. Sure.
So I think, again, this notion of gaming the stress test is
largely a misnomer. What banks are doing is, they are using the
information that they have access to in the risk weights, in
the context of things like the GSIB capital surcharge, and, in
principle, in the context of the stress test, to assess the
riskiness of the assets on their balance sheet and then make
funding decisions on a relevant basis in concert with the other
information that they bring to the table in the context of
their own risk management decisions. The better information
that they have access to, the more complete that information,
the better risk management decision that they can make.
To the extent--as Francisco and others have pointed out, to
the extent that the Federal Reserve is concerned about some
sort of gaming activities or something else, they have a
variety of remedies at their disposal that they can use through
the supervisory process----
Chairman Barr. The----
Mr. Campbell [continuing]. to deal with that.
Mrs. Kim. Thank you.
Mr. Campbell. They do not need to limit the transparency in
the stress-testing regime to deal with that issue.
Mrs. Kim. I yield back my time.
Chairman Barr. The gentlewoman's time has expired.
The gentlelady from Massachusetts, Ms. Pressley, is now
recognized. This will be the last questioner of the hearing.
Ms. Pressley. Thank you, Mr. Chair.
Thank you to our witnesses for joining us today.
The 2008 financial crisis exposed the weaknesses and
vulnerabilities in our financial system, causing a recession
that robbed an entire generation of economic opportunity and
many are still reeling from these effects.
The modern stress-testing regime that regulators use today
was born out of that crisis and is now a staple of our
regulatory framework and oversight of banks.
My Republican colleagues, however, want to go beyond the
rollback of Dodd-Frank in 2018 that directly led to the
collapse of Silicon Valley Bank (SVB) and other banks last
year. They want to continue down a path of deregulation that
will undoubtedly increase systemic risk in our financial system
and increase economic harm to our constituents. Make it make
sense.
Mr. Feldberg, how did regular stress testing restore
confidence in the banking sector following the 2008 financial
crisis?
Mr. Feldberg. So, in the middle of the financial crisis,
there was widespread concern about the capital of all the
largest banks in the U.S., and I think there was just a lack of
belief in regulatory capital information that was out there.
There was a lack of confidence in the measure of capital.
So the point of the stress test was to throw a very
difficult scenario at the banks and to ask the question, what
would happen if this thing got even worse than it is? I think
it was seen as a very credible test. The important thing,
really, was that it be seen as credible by the markets.
The results showed some banks passed, some banks failed.
Banks went ahead and raised capital in an environment in which
confidence was improving. I guess there was some luck that they
were able to raise capital and only a small number of banks
ended up really failing the test.
It was a--it was a remarkable opportunity to address the
crisis and get it going----
Ms. Pressley. Thank you.
Mr. Feldberg [continuing]. get things going again.
Ms. Pressley. Thank you.
Again, deregulation is a failed approach. Yet bank
lobbyists are continuing to push for less stringent oversight.
My colleagues across the aisle are working with big banks to
publicly reveal the mechanics behind stress testing. This is
not a pro-transparency, good-governance argument. They simply
want to help banks game the system and circumvent oversight.
Mr. Feldberg, in 2019, the Fed provided details to banks
about how they conduct stress tests, and this diminished the
value of those tests.
If the Fed reveals even more information--just building
upon what we have been talking about today; please expound--
would that not allow banks to game the tests, undermining the
very purpose of examining how resilient and prepared banks are
for unexpected shocks to the financial system?
Mr. Feldberg. The banks spend a great deal of money trying
to reverse-engineer the stress tests, and do not think they do
that for any other reason than trying to game the process.
They can already do a petty good job of reverse-engineering
because they already get more information about the stress-test
models than any other country gives their banks.
If they got what is being asked for today, it would
emasculate the process, and we would have something similar to
what OFHEO had with the GSEs--an ineffective process that does
not keep up with changes in the financial system.
Ms. Pressley. Thank you.
Mr. Feldberg. I would like to--I mean, just--an argument
that came up earlier was, why do we not reveal these things
after the test? I would discourage that idea from going much
further.
As I keep on saying, we actually do disclose a great deal
already, so the difference in what we would be disclosing after
and before the release of the results is not very much, but it
is that one little secret sauce that prevents the banks from
essentially just replicating the models and baking grandma's
cake, and then you just--you do not have a test anymore.
Ms. Pressley. Thank you.
When implementing stress tests, regulators should be
prioritizing what is best for consumers and the stability of
our financial system, not the desires of greedy bank CEOs.
If Republicans and big banks want greater transparency,
then they should practice what they preach. I have a long list
of areas where banks should be more transparent: the use of AI
in decisionmaking, their investments in mass incarceration, the
funding of fossil-fuel projects that kill our environment, the
denials of mortgages and business loans to disproportionately
Black and Brown communities, and so much more.
The Fed should not give away the answer key to stress
testing. After witnessing some of the largest bank failures in
our history last year, our response cannot be further
deregulation that weakens our banking system. We need
meaningful and stringent stress tests for a safer and more
stable financial system.
Thank you. I yield.
Chairman Barr. The gentlelady yields.
I would like to thank our witnesses for their testimony
today.
As a matter of a concluding observation, I would just say
that I thought we had great testimony, great points made by all
of our witnesses. My takeaway is that we can have more
transparency in stress testing, we can reduce the opacity. We
can do that certainly on the input side of things to provide
the accountability, to improve the process, to improve the
actual stress testing itself and we do not have to do it in a
way that would create the risks that Mr. Feldberg pointed out
on the outputs.
I think he makes a good point, that the goal should be to
reveal just enough about banks to help market participants
evaluate risk but not so much to undermine confidence. I think
improving transparency at the front end does not necessarily
have to lead to undermining confidence on the back end, and
results of stress tests can be disclosed to the supervised
institution so that they can evaluate those risks without
undermining confidence.
So, with that, I appreciate the testimony. I think this was
a great hearing.
Without objection, all members will have 5 legislative days
within which to submit additional written questions for the
witnesses to the chair, which will be forwarded to the
witnesses for their response. I ask our witnesses to please
respond as promptly as you are able.
[The information referred to can be found in the appendix.]
Chairman Barr. This hearing is now adjourned.
[Whereupon, at 3:41 p.m., the subcommittee was adjourned.]
A P P E N D I X
[GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]
[all]