[Senate Hearing 113-551]
[From the U.S. Government Publishing Office]





                                                        S. Hrg. 113-551


                      EXAMINING THE GAO REPORT ON
     EXPECTATIONS OF GOVERNMENT SUPPORT FOR BANK HOLDING COMPANIES

=======================================================================

                                HEARING

                               before the

                            SUBCOMMITTEE ON
             FINANCIAL INSTITUTIONS AND CONSUMER PROTECTION

                                 of the

                              COMMITTEE ON
                   BANKING,HOUSING,AND URBAN AFFAIRS
                          UNITED STATES SENATE

                    ONE HUNDRED THIRTEENTH CONGRESS

                             SECOND SESSION

                                   ON

      EXAMINING WHETHER THE ERA OF TOO BIG TO FAIL IS FINALLY OVER

                               __________

                             JULY 31, 2014

                               __________

  Printed for the use of the Committee on Banking, Housing, and Urban 
                                Affairs


                 Available at: http: //www.fdsys.gov /
                 
                 
                 
                                     ______

                         U.S. GOVERNMENT PUBLISHING OFFICE 

93-180 PDF                     WASHINGTON : 2015 
-----------------------------------------------------------------------
  For sale by the Superintendent of Documents, U.S. Government Publishing 
  Office Internet: bookstore.gpo.gov Phone: toll free (866) 512-1800; 
         DC area (202) 512-1800 Fax: (202) 512-2104 Mail: Stop IDCC, 
                          Washington, DC 20402-0001
                                          
                 


            COMMITTEE ON BANKING, HOUSING, AND URBAN AFFAIRS

                  TIM JOHNSON, South Dakota, Chairman

JACK REED, Rhode Island              MIKE CRAPO, Idaho
CHARLES E. SCHUMER, New York         RICHARD C. SHELBY, Alabama
ROBERT MENENDEZ, New Jersey          BOB CORKER, Tennessee
SHERROD BROWN, Ohio                  DAVID VITTER, Louisiana
JON TESTER, Montana                  MIKE JOHANNS, Nebraska
MARK R. WARNER, Virginia             PATRICK J. TOOMEY, Pennsylvania
JEFF MERKLEY, Oregon                 MARK KIRK, Illinois
KAY HAGAN, North Carolina            JERRY MORAN, Kansas
JOE MANCHIN III, West Virginia       TOM COBURN, Oklahoma
ELIZABETH WARREN, Massachusetts      DEAN HELLER, Nevada
HEIDI HEITKAMP, North Dakota

                       Charles Yi, Staff Director

                Gregg Richard, Republican Staff Director

                       Dawn Ratliff, Chief Clerk

                      Shelvin Simmons, IT Director

                          Jim Crowell, Editor

                                 ______

     Subcommittee on Financial Institutions and Consumer Protection

                     SHERROD BROWN, Ohio, Chairman

       PATRICK J. TOOMEY, Pennsylvania, Ranking Republican Member

JACK REED, Rhode Island              RICHARD C. SHELBY, Alabama
CHARLES E. SCHUMER, New York         DAVID VITTER, Louisiana
ROBERT MENENDEZ, New Jersey          MIKE JOHANNS, Nebraska
JON TESTER, Montana                  JERRY MORAN, Kansas
JEFF MERKLEY, Oregon                 DEAN HELLER, Nevada
KAY HAGAN, North Carolina            BOB CORKER, Tennessee
ELIZABETH WARREN, Massachusetts

               Graham Steele, Subcommittee Staff Director

        Geoffrey Okamoto, Republican Subcommittee Staff Director

                  Megan Cheney, Legislative Assistant

                                  (ii)

                            C O N T E N T S

                              ----------                              

                        THURSDAY, JULY 31, 2014

                                                                   Page

Opening statement of Chairman Brown..............................     1

Opening statements, comments, or prepared statements of:
    Senator Toomey...............................................     3
    Senator Vitter...............................................     4

                               WITNESSES

Lawrance Evans, Director, Financial Markets and Community 
  Investment, Government Accountability Office...................     6
    Prepared statement...........................................    34
    Responses to written questions of:
        Chairman Brown...........................................   198
Deniz Anginer, Assistant Professor of Finance, Pamplin School of 
  Business, Virginia Tech........................................     8
    Prepared statement...........................................    50
Edward Kane, Professor of Finance, Boston College................    10
    Prepared statement...........................................   124
Anat Admati, George G.C. Parker Professor of Finance and 
  Economics, Graduate School of Business, Stanford University....    13
    Prepared statement...........................................   130
Douglas Holtz-Eakin, President, American Action Forum............    15
    Prepared statement...........................................   195
    Responses to written questions of:
        Chairman Brown...........................................   199

              Additional Material Supplied for the Record

Charts submitted by Chairman Sherrod Brown.......................   201
Report submitted by the Government Accountability Office.........   202
Statement submitted by the American Bankers Association..........   296
Comments regarding the ``GAO Report on Large Bank Holding 
  Companies'', by Allan H. Meltzer...............................   303

                                 (iii)

 
EXAMINING THE GAO REPORT ON EXPECTATIONS OF GOVERNMENT SUPPORT FOR BANK 
                           HOLDING COMPANIES

                              ----------                              


                        THURSDAY, JULY 31, 2014

                                       U.S. Senate,
       Subcommittee on Financial Institutions and Consumer 
                                                 Protection
           Committee on Banking, Housing, and Urban Affairs
                                                    Washington, DC.
    The Subcommittee met at 2:02 p.m., in room 538, Dirksen 
Senate Office Building, Senator Sherrod Brown, Chairman of the 
Subcommittee, presiding.

          OPENING STATEMENT OF CHAIRMAN SHERROD BROWN

    Chairman Brown. The Subcommittee will come to order.
    Thank you, Senator Toomey, for working with us on so many 
issues, including your cooperation in this hearing. Thank you.
    Thank you, Senator Vitter, for joining us.
    There will be other Members here, too.
    And thanks to the witnesses whom I will introduce in a 
moment.
    Too big to fail is the Government policy that ensures that 
certain financial institutions cannot be allowed to fail 
because their failures would cause too much damage to our 
Nation's financial system and our Nation's economy.
    This is the Subcommittee's sixth hearing since 2011 on the 
issue of too-big-to-fail banks. We are here to ask again 
whether too big to fail is finally over.
    Some Wall Street institutions and their paid consultants 
and lobbyists argue that no bank is too big to fail.
    Upton Sinclair, the American writer, once said, ``It is 
difficult to get a man to understand something, when his salary 
depends on his not understanding it.''
    But most objective observers agree with Federal Reserve 
Chair Janet Yellen who said that ``Our work is not finished.'' 
When the four largest U.S. banks are 25 percent larger today 
than they were in 2007, before the implosion of the economy, it 
is hard to disagree with her statement.
    Yesterday, Christy Romero, the Special Inspector General 
for TARP, released a report in which she concluded the six 
largest U.S. bank holding companies ``remain interconnected to 
each other in 2013 as they were in 2008.'' She agreed we have 
more to do.
    Too big to fail distorts incentives and encourages 
excessive growth and leverage and complexity.
    Today we are here to examine one effect of too-big-to-fail 
policies--the financial benefits enjoyed by the largest banks. 
The GAO report shows that megabanks' funding advantage varies 
based upon how one measures it. Unfortunately, this estimate 
contains many limitations and is clouded by both extraordinary 
interest rate policies and a number of subsidies that are 
difficult to quantify.
    But this report has some valuable lessons about too big to 
fail and shows that right now the subsidy may have been reduced 
but could be about 50 basis points. GAO also says the subsidy 
may have been reduced not because we ended too big to fail but 
because of, their words, ``improvements in banks' holding 
companies' financial conditions.''
    Unfortunately, market perceptions of financial conditions 
change quickly. For example, banks' credit default swap spreads 
increased suddenly and not until well into 2008.
    Secretary Lew told this Committee last month that the only 
real moment when you know for sure is when there is a crisis.
    We rarely have the foresight to know when a financial 
crisis is about to happen. GAO's report, though, gives us a 
glimpse of the next crisis. It estimates that the funding 
advantage would return to levels similar to those in 2008 as 
investors and depositors flock--and that word, flock, is 
exactly right--as depositors and investors flock to the 
megabanks because they believe the U.S. Government will rescue 
them.
    In another crisis, the biggest banks' advantages 
potentially rise to as much as 500 basis points. Taking an 
estimate from one of the most conservative models of this 
scenario, the 6 largest banks would get an annual subsidy of 
about $13 billion in all of their liabilities. Using some of 
the higher estimates, it could be 10 times that.
    This suggests that under Secretary Lew's test, the largest 
banks are still too big to fail; taxpayers are still supplying 
them with an implicit guarantee.
    Whether you view the Government support as a form of 
catastrophic insurance or a stock option or a nontransparent 
contingent liability for the Federal budget, we know that 
taxpayers really, in the end, never receive full value for it.
    The first GAO report issued in November, when Dr. Evans 
joined us before, showed that Government programs underprice 
support during a crisis.
    As a second report shows, we have not taken the necessary 
steps to ensure that we will not have to prop up the largest 
banks again.
    Unless you think that we can eliminate financial crises 
forever, the GAO report is another reminder we have more work 
to do to eliminate too-big-to-fail policies and the advantages 
and the distortions which they create.
    When we think about the costs and benefits of too-big-to-
fail policies, industry wants us to think only about their 
costs.
    Steel companies dump waste into a river. They then argue it 
will be costly to clean it up, but it has a higher human cost 
to the miners and the children who get sick from the pollution. 
It passes more health care costs onto our society. The same 
with damage done by policies on too big to fail.
    Those who believe in a society with rules understand that 
auto safety might cost car companies to install seatbelts and 
airbags, but those protections save lives.
    And financial rules might cost bank executives a little 
smaller bonuses, somewhat smaller dividends perhaps, but they 
will help prevent a repeat of what we had 5 years ago with 
millions of foreclosures and millions of lost jobs.
    Senator Toomey, thank you.

             STATEMENT OF SENATOR PATRICK J. TOOMEY

    Senator Toomey. Thank you very much, Chairman Brown, and 
thanks for having this hearing.
    I want to also thank Senator Vitter for requesting the GAO 
report and thank the GAO for their contribution to this 
discussion.
    It seems to me the GAO report on whether or not big banks 
get a subsidy is inconclusive, but in any case the real work 
that we ought to be focused on, the real issue here, is the 
extent to which Dodd-Frank has codified too big to fail.
    The way I see it, what Dodd-Frank does is it looks at 
banks, designates them as SIFIs and then attempts to 
micromanage them so massively and so completely that failure 
is, theoretically, not possible. I think that is the basic 
mentality of Dodd-Frank.
    Well, there are some problems with this approach.
    One is we have institutions designated as SIFIs who 
absolutely are not systemically important, and we have had 
hearings where we have had that discussion. There is no 
question in my mind about that.
    We also have the massive direct and indirect costs of 
complying with the overregulation and the micromanagement, 
which I would suggest goes well beyond a slightly diminished 
dividend to shareholders. It means there is less credit 
available, and the credit that is available is available at 
higher prices. So that is a problem for our entire economy.
    I would also suggest that the premise that regulators, as 
long as they have enough power, will make it impossible for an 
institution to fail; that is based on the mistaken notion that 
these regulators are omniscient, or have greater wisdom or 
intuition than they did before 2008, because there was no 
absence of regulators at the time that the financial crisis 
hit.
    Finally, I would simply argue that failure has to be an 
option. In a market-based economy, in a capitalist system, in a 
free society, you have to be free to fail. And that is without 
a taxpayer bailout.
    And so that is what we ought to be looking to achieve--a 
system where we can have a failure that is not catastrophic and 
that does not involve taxpayers being forced to make a bailout.
    So my suggestion is that what we ought to do--and I am open 
and looking forward to having your support on my legislation, 
Mr. Chairman--is let's repeal Title II. To the extent that 
anybody thinks there is a subsidy or there is a codification of 
too big to fail, it must reside in Title II, which is where the 
orderly liquidation authority is.
    And rather than have this subjective process, which is what 
Title II is, where there is no option for restructuring, there 
is no certainty about creditor rights and there is an explicit 
mechanism for taxpayers to come in and be forced to make a 
bailout, rather than all of that, why not do this in 
bankruptcy, which is what bankruptcy is supposed to be all 
about?
    If you have a properly designed bankruptcy code, which I 
think does require some modifications, creditors would be on 
the hook for losses, not taxpayers; you could allow for either 
a liquidation or a reorganization, whichever makes sense; 
creditors of equal standing would be treated equally rather 
than the subjective treatment that Title II of Dodd-Frank 
contemplates; and you have a bridge bank mechanism that would 
allow for a resolution to occur without systemic problems.
    So I think that too big to fail is a real concern. I think 
the real solution is to go to the heart of where the problem 
is. The problem is in Dodd-Frank, and so I hope we will be able 
at some point to address that.
    Again, I thank you for having this hearing, Mr. Chairman.
    Chairman Brown. Thanks, Senator Toomey.
    Senator Warren? No opening statement.
    Senator Vitter.

               STATEMENT OF SENATOR DAVID VITTER

    Senator Vitter. Thank you, Chairman Brown, for holding this 
hearing today and for partnering with me in requesting the 
study from the GAO that was released today on the expectation 
of Government support for megabanks-large bank holding 
companies.
    Chairman Brown and I started our work together based on a 
shared belief that Dodd-Frank had not ended too big to fail. We 
began by writing the regulators in 2012, urging them to use 
their statutory authority to deal with the too-big-to-fail 
problem and end bailouts once and for all. Unfortunately, they 
were not as aggressive as they needed to be.
    We then introduced our legislation requiring prior capital 
standards, and in January 2013, we asked the GAO to study and 
report to Congress on the perks of a megabank being considered 
too big to fail.
    I guess I disagree somewhat with my colleague, Senator 
Toomey, that this report is completely inconclusive. I think it 
is very helpful, and I think it moves the debate significantly.
    Not long ago, a lot of folks led by the megabanks were 
denying any funding advantage, any too-big-to-fail subsidy. Now 
I think that debate is over. Everyone agrees it exists, and we 
are debating how big it is and for what reasons it is here or 
here or wherever.
    I think that is a significant shift in the debate and we 
are approaching a consensus on this. From this report, I think 
it supports that consensus.
    Let me just point to a couple of quotes.
    ``Remaining market assumptions about Government support can 
give rise to advantages for the largest bank holding companies 
in three broad categories to the extent these assumptions 
affect decisions by investors, counterparties and customers of 
these firms. Those categories are funding costs, financial 
contracts that reference ratings and an ability to attract 
customers.''
    And then in addition, out of this GAO report, it says that 
if today's megabanks had the same environment and credit risk 
as in 2008 the market would expect bailouts and their funding 
advantage would increase to between 102 and 495 basis points. 
So that is significant.
    Finally, let me point to one other key metric that I am 
very concerned about. It is outside of this report, but it is a 
key metric that is unassailable, and that is the accelerating 
pace since the crisis of consolidation.
    The megabanks have gotten much bigger. Smaller community 
banks have gotten far fewer in number. That was a preexisting 
trend, but that trend has been put on steroids through the 
crisis and Dodd-Frank. That trend is unassailable, beyond 
debate.
    And I think that is very worrisome for our banking system 
and sadly ironic, given that on the private side--Government 
was certainly responsible for the crisis in terms of many 
policies, but on the private side it was very large 
institutions, not smaller community banks. They have 
essentially benefited in terms of where they are in the market, 
benefited through the crisis and, in light, benefited by our 
response in terms of Dodd-Frank.
    Those smaller institutions that had nothing to do with the 
crisis have receded and in a much more vulnerable position than 
they were before the crisis and the legislative response.
    I look forward to our witnesses today and to an informed 
discussion based on what they have to say.
    Thank you.
    Chairman Brown. Thank you, Senator Vitter. Thank you for 
your work on requesting this report and as Senator Toomey said, 
too.
    Dr. Lawrance Evans is a Director in the Financial Markets 
team at the GAO, where he has also served as an economist and 
led engagements in the GAO's International Affairs and Trade 
Team. Prior to his service there, Dr. Evans was a research 
fellow at Amherst College and a research assistant at the 
Center for Economic Policy and Analysis.
    Welcome back, Dr. Evans.
    Mr. Evans. Thank you.
    Chairman Brown. Dr. Deniz Anginer is an assistant professor 
of finance at Virginia Tech's Pamplin School of Business and is 
currently on leave from the position as a financial economist 
with the Development Research Group at the World Bank. Prior to 
joining the World Bank, Dr. Anginer worked as a risk and 
finance consultant at Oliver Wyman.
    Dr. Edward Kane is a research professor of finance at 
Boston College. In addition to his more than 50-year career in 
banking and economic teaching and research, Professor Kane 
served as a consultant to private companies and Government 
agencies including the Office of the Comptroller of the 
Currency, the General Accounting Office and the Federal Reserve 
Banks of Boston and Chicago. He currently serves as a research 
associate at the National Bureau of Economic Research and is a 
member of the Shadow Financial Regulatory Committee.
    Welcome.
    Dr. Anat Admati is the George Parker Professor of Finance 
and Economics at Stanford University's Graduate School of 
Business. Professor Admati spent more than 30 years studying 
and teaching finance and economics and has written extensively 
about the function and regulation of financial markets.
    Thank you, Dr. Admati, for being here.
    Douglas Holtz-Eakin is President of the American Action 
Forum. Dr. Eakin was a member of the Financial Crisis Inquiry 
Commission and served as Director of the Congressional Budget 
Office and Chief Economist of the President's Council of 
Economic Advisors. In addition to his appointments across 
multiple Administrations, Dr. Holtz-Eakin was a senior fellow 
at the Peterson Institute for International Economics, Director 
of the Maurice Greenberg Center for Geoeconomic Studies and the 
Volcker Chair of International Economics at the Council on 
Foreign Relations.
    Welcome back to one of your many times to testify, Dr. 
Holtz-Eakin.
    Dr. Evans, if you would begin.

 STATEMENT OF LAWRANCE EVANS, DIRECTOR, FINANCIAL MARKETS AND 
     COMMUNITY INVESTMENT, GOVERNMENT ACCOUNTABILITY OFFICE

    Mr. Evans. Thank you. Chairman Brown, Ranking Member 
Toomey, and Members of the Subcommittee, it is my pleasure to 
appear before you today to discuss the existence of any funding 
advantages the largest banks may have received due to perceived 
Government support.
    The report we released today reflects GAO's extensive work 
on this issue. When I appeared before the Subcommittee in 
January, I noted that the question of whether or not banks 
receive benefits because of investor expectations of loss 
protection was largely an empirical one. As a result, my 
remarks today will focus on our original quantitative analysis.
    It is important to note up front that this was a difficult 
task, particularly because measuring investor perceptions is 
complicated. The very nature of our exercise necessitated a 
number of methodological choices and raised a number of issues 
over which reasonable people may disagree.
    Consumers of this research should know that precise and 
firm conclusions likely reflect key assumptions and confident 
decisions on difficult methodological issues that would benefit 
from professional skepticism and full disclosure. As a result, 
our report carries a heavy dose of caution and nuance and 
reflects the uncertainty underpinning our modeling effort.
    There are many funding sources one could analyze. We 
selected senior unsecured bonds.
    Bond yield spreads are a direct measure of what actual 
investors charge banks to borrow money in the market and are 
sensitive to credit risk and, thus, investor expectations. 
Senior unsecured debt is among the most important sources of 
nondeposit funding and intended to absorb losses under FDIC's 
resolution authority.
    While there are many approaches to examining potential 
funding cost advantages, we chose an econometric methodology. 
This allows us to examine the relationship between size and 
bond funding costs and take into account other factors that 
might drive funding cost differences, like credit risk and bond 
liquidity.
    Specifically, our model allows the funding cost-risk 
relationship to be influenced by bank holding company size and 
for that influence to change from year to year.
    Our time period covers 2006 to 2013, which captures many 
important changes in the regulatory environment and any impact 
this may have had on market expectations.
    Because uncertainty is inherent in modeling, it is best 
practice to analyze and report its effects. We directly 
incorporate sensitivity analysis into our findings by 
estimating 42 models for each year. The multiple model 
specifications reflect the various defensible ways to 
incorporate credit risk, bond liquidity and size into the 
analysis.
    For example, there is no agreement in the literature on 
which institutions may be considered too big to fail. 
Therefore, we capture systemic importance using different 
measures, including total assets and indicators for banks 
designated as G-SIBs and those designated as SIFIs by Dodd-
Frank.
    Our analysis suggests a funding cost advantage for large 
banks during the financial crisis but provides mixed evidence 
of such advantages in recent years. For example, most models in 
2013 suggest that funding cost advantages have declined or 
reversed, but there were a few models that found lower funding 
costs for large banks.
    Because we are agnostic about which models are the right 
ones, GAO remains cautious against strong conclusions. The 
weight of the evidence, though, suggests progress has been 
made, but it is still too soon to declare victory as funding 
cost advantages might resurface should crisis conditions 
reemerge.
    This is an important possibility to consider because 
changes over time in our estimates of the funding cost-size 
relationship may reflect changes in investors' belief about, 
one, the likelihood that a bank will fail, two, the likelihood 
that it will be rescued by the Government if it fails, and 
three, the size of the losses the Government may impose on 
investors if it rescues the bank. However, we cannot precisely 
identify the influence of each of these factors.
    In a hypothetical scenario, when we assume credit risk 
returned to financial crisis levels in 2013, most of our models 
suggest, again, lower funding costs for larger banks.
    The hypothetical scenarios we examined did not provide the 
same overwhelming type of evidence of advantages we found in 
2008 and 2009, and Dodd-Frank and other financial reforms could 
make financial crisis risk scenarios less likely.
    But the outcome of these scenarios suggests that in 
addition to changes in expectations of Government support some 
of what we see is likely related to balance sheet repair and, 
therefore, a lower probability of failure.
    To be frank, it may take another crisis to truly test the 
effectiveness of financial reforms.
    We detail important limitations associated with our 
econometric analysis in the written statement and full report. 
Suffice to say here that our work is not perfect, and we have 
not exhausted the many ways one might investigate funding cost 
advantages. Users of the report should give our cautionary 
notes serious attention before moving from our findings to 
public policy.
    Now we understand that stakeholders are invested on both 
sides of this issue. We encourage interested parties to base 
their examination of our work on an objective and thorough 
reading of the actual report.
    Shortly, you will hear from Professor Kane, who likely only 
had a limited time to review our report. Professor Kane's 
written testimony contains a number of statements that are 
either inaccurate, mischaracterize our methodology or result 
from the application of inappropriate criteria to assess the 
validity of our study.
    GAO will welcome an opportunity to respond as we believe we 
have made an important contribution to the literature and the 
public policy debate.
    Chairman Brown, Ranking Member Toomey, and Members of the 
Subcommittee, this concludes my prepared statement. I look 
forward to any questions you might have.
    Chairman Brown. Thank you, Dr. Evans.
    Dr. Anginer.

  STATEMENT OF DENIZ ANGINER, ASSISTANT PROFESSOR OF FINANCE, 
           PAMPLIN SCHOOL OF BUSINESS, VIRGINIA TECH

    Mr. Anginer. Mr. Chairman and the distinguished Members of 
the Subcommittee, thank you for convening today's hearing and 
inviting me to testify.
    My name is Deniz Anginer. I am an assistant professor at 
Pamplin Business School and Virginia Tech.
    Along with my colleagues, Viral Acharya and Joe Warburton, 
I have examined market expectations of implicit Government 
guarantees to so-called too-big-to-fail institutions. Most of 
my testimony is based on this research.
    The too-big-to-fail doctrine holds that the Government will 
not allow large financial institutions to fail if their failure 
would cause significant disruption to the financial system and 
to economic activity.
    In our research, we find that large financial institutions 
and their investors expect the Government to back the debts of 
these institutions should they encounter financial difficulty. 
These expectations of Government support are embedded in the 
price of bonds issued by major financial firms, allowing them 
to borrow at lower rates.
    Expectation of Government support by the market also 
results in a distortion in how risk is reflected in the debt 
prices of large financial institutions.
    An explicit Government guarantee dulls market discipline by 
reducing investors' incentives to monitor and to price the 
risk-taking of large financial firms.
    In our analysis, we show that while a positive relationship 
exists between risk and cost of debt for medium- and small-
sized institutions, this relationship is 75 percent weaker for 
the largest institutions. Changes in leverage and capital 
ratios are likewise less sensitive to changes in risk for these 
large financial firms.
    Because they pay a lower price for risk than other 
financial institutions, the perceived guarantee provides too-
big-to-fail firms with a funding advantage. We estimate a 
funding cost advantage of approximately 30 basis points over 
the years 1990 to 2012, peaking at more than 100 basis points 
in 2009.
    The total value of the subsidy amounted to about $30 
billion per year on average over the same time period, topping 
at $150 billion in 2009.
    We also examined nonfinancial firms. If bond investors 
believe that all of the largest firms, both financial and 
nonfinancial, are too big to fail, then large nonfinancial 
firms should enjoy subsidies similar to that of large financial 
firms. However, we find that this is not the case, suggesting 
that the difference is likely due to an expectation of an 
implicit Government guarantee.
    Compared to the GAO study, we find lower implicit subsidy 
values for the years 2007 to 2011 and slightly higher numbers 
in 2012. We have not examined the year 2013, the year in which 
the GAO finds the greatest decline.
    Although most of the attention will be paid to the analyses 
that try to quantify the dollar values of the subsidy and its 
changes over time, it is important to note that it is very 
difficult to directly relate these changes to the introduction 
of Dodd-Frank and other regulations.
    It is very hard to separate out changes in the probability 
of large financial firms experiencing distress from the 
probability that they will be bailed out. As the GAO report 
points out, this is especially true as the risk premium in the 
market has declined in recent years and large financial firms 
have seen significant improvements in their balance sheets and 
capital ratios, reducing their probability of experiencing 
financial distress.
    Although it is very difficult to establish a direct link 
between regulations and changes in subsidy over time, examining 
these changes in subsidy using alternative methods over a short 
time window would be more helpful in analyzing the impact of 
Dodd-Frank and other regulations.
    In our study, we examined changes in risk sensitivities of 
cost of debt after the introduction of Dodd-Frank. We examined 
changes in subsidies accruing to large financial firms compared 
to nonfinancial firms. We also examined the cost of implicitly 
guaranteed debt to explicitly guaranteed debt issued by the 
same firm under the FDIC's Temporary Liquidity Guarantee 
Program.
    Using these alternative approaches, we find that Dodd-Frank 
did not significantly alter investors' expectations that the 
Government will bail out too-big-to-fail financial firms should 
they falter.
    Despite its no-bailout pledge, Dodd-Frank leaves open many 
avenues for future rescues. For instance, the Federal Reserve 
can offer a broad-based lending facility to a group of 
financial institutions in order to provide a disguised bailout 
to the industry or a single firm.
    In addition, Congress can sidestep Dodd-Frank by amending 
or repealing it or by allowing regulators to interpret their 
authority in ways that protect creditors and support large 
institutions.
    Finally, it is also important to note that the analysis 
conducted by us and the GAO only measured the direct subsidy 
that may accrue to too-big-to-fail firms. There may be other 
indirect effects such as misallocation of capital or excessive 
and correlated risk-taking to exploit the implicit guarantees 
that are not captured by the analysis.
    To conclude, Governments are generally not required to make 
any apparent financial commitment or outlay or request funds 
from the legislatures or taxpayers when they implicitly 
guarantee too-big-to-fail institutions. Implicit guarantees 
lack the transparency and accountability that accompany 
explicit policy decisions.
    Taxpayers' interests could be better served by estimating 
on an ongoing basis, both in good times and in bad times, the 
accumulated value of the subsidy. Public accounting of 
accumulated too-big-to-fail costs might restrain those 
Government actions and policies that encourage too-big-to-fail 
expectations.
    Thank you.
    Chairman Brown. Thank you, Dr. Anginer.
    Dr. Kane, welcome.

 STATEMENT OF EDWARD KANE, PROFESSOR OF FINANCE, BOSTON COLLEGE

    Mr. Kane. Thank you, Mr. Chairman, for inviting me to 
testify today. I want to congratulate you and the rest of the 
Committee for continuing to battle against too big to fail in 
the face industry efforts to tell us it has gone away. Finally, 
I also want to thank Mr. Evans for making you more eager to 
hear what I have to say.
    [Laughter.]
    Mr. Kane. What I have to say is that GAO bungled the 
assignment you gave it. The GAO goes wrong at the outset in how 
it defines too big to fail.
    The definition of too big to fail offered in the report's 
first sentence is incomplete. It describes too big to fail 
(TBTF) as an active policy of intervention when the most 
important part of TBTF is a passive policy of forbearance, 
which allows institutions that are insolvent to continue to 
roll over, and even expand, their debt.
    Deeply insolvent banks are what I term zombie institutions. 
They can only prevail because they are backed by the black 
magic of Government implicit guarantees.
    The GAO also misunderstands the character of the funding 
advantages that your Committee asked them to study. The GAO 
treats these guarantees as if they are merely a form of bond 
insurance on outstanding bonds.
    The character of too-big-to-fail guarantees is richer than 
insurance on outstanding bonds because, as long as regulators 
forbear from resolving its insolvency, a truly too-big-to-fail 
firm can extract further guarantees by issuing endless amounts 
of additional debt.
    So what is funding cost? Funding cost is the cost of the 
funding mix. Being too big to fail lowers both the cost of 
debt, which GAO studied, and the cost of equity, which it did 
not.
    too big to fail guarantees lower the risk that flows 
through to holders of both kinds of securities. It chops off 
their losses at a certain point and directs the flow of further 
losses to taxpayers.
    This means that, period by period, the incremental 
reduction in interest payments on outstanding bonds, deposits 
and repos is only part of the subsidy that the stockholders 
enjoy. The missing part is the increase in stock prices that 
comes from having investors discount the firm's current and 
future cash-flows at an artificially low rate of return on 
equity.
    Limitless guarantees shift the risk of the deepest possible 
losses away from creditors and stockholders. It is as if the 
profit flow move through a pipeline with a Y in it. Once a TBTF 
becomes insolvent, further losses go to the taxpayers until the 
economy recovers.
    So the issue is not whether things are better today. The 
economy is better; so the banks will be better.
    The issue is whether we continue to encourage them to take 
on too much tail risk.
    The value of these incentives is the greatest part of what 
the GAO missed. We must recognize that guarantee contracts have 
two components. The first allows the guarantee party to put 
responsibility for covering losses that exceed the value of the 
assets of the bank holding company to the guarantor. No 
guarantor wants to expose itself to unlimited losses on this 
put.
    For this reason, all guarantee contracts incorporate a 
stop-loss provision that gives the guarantor a call on the 
assets of the firm. Ordinarily, the stop-loss kicks in just as 
insolvency is approached or breached.
    In the FDIC Improvement Act of 1991, efforts to exercise 
the Government's call is termed prompt corrective action. We 
did not see prompt corrective action in 2008 for TBTF 
institutes.
    By definition, the Government's right to take over a firm's 
assets will never be exercised in a financial institution that 
is truly too big to fail. Nonexercise means that the Government 
has effectively ceded the value of its loss-stopping rights to 
the too-big-to-fail BHC's stockholders. The value that 
forbearance gives away is what the GAO's measure ignores.
    I offer a picture, Figure 2, in my testimony that graphs 
the behavior of AIG's stock price before, during and after the 
2008 crisis. The only time AIG's stock price approached zero--
and it did so twice--was when the notion of a Government 
takeover was seriously under discussion so that the probability 
of stockholders' continued rescue was falling. As soon as this 
course of action was tabled, the stock price surged again 
because, TBTF policies were turning the stop-loss back to the 
stockholders.
    Also, the designation of systemically important financial 
institutions is really not a binary condition; that is, it is 
wrong to say that a BHC either is TBTF or it is not. TBTF does 
not start at a particular size; it lies on a continuum and is 
influenced by several variables. Any firm's access to Senators 
and Congresspersons grows with its geographic footprint--this 
is part of the problem caused by the ongoing consolidation 
among the biggest BHCs--and with the number of employees that 
can be persuaded to contribute to reelection campaigns.
    To do a proper investigation, one cannot just look at bond 
markets. One should be looking also the stock market.
    I present in Figure 3 some work that Hovakimian, Luc 
Laeven, and I have done to estimate average dividend that 
taxpayers ought to have been paid by large banking firms from 
1974 to 2010. We can see the cyclical pattern that we have been 
talking about. But we can also see secular learning about and 
growing exploration of the value of these guarantees.
    Prior to each new recession, at the peak, more benefit has 
been extracted by too-big-to-fail institutions. I fear what is 
apt to happen in the next crisis.
    In deciding to ignore studies that use the contingent-claim 
approach to evaluate TBTF subsidies, the GAO fell into the trap 
of thinking of bailout expenditures as either loans or 
insurance. It is important that we understand the difference 
between guarantees, insurance, and loan contracts.
    An insurance company does not double and redouble the 
coverage of drivers it knows to be reckless.
    Similarly, lifelines provided to an underwater firm cannot 
be thought of as low-interest loans. Loans are just not 
available to firms that are in dire straits.
    The ability to extract implicit guarantees on new debt and 
the hugely below-market character of bailout programs means the 
repayment of funds that were actually advanced--i.e., just the 
funds that were actually advanced--does not show that a bailout 
program is a good deal for taxpayers.
    I believe that the politicians who have made that claim are 
embarassing everyone in Government. They are causing the 
public, who does understand this, to lose confidence in the 
policymaking.
    So what should we do to sanction reckless pursuit of TBTF 
subsidies? That is the second part of my statement: How can we 
sanction the exploitation of too-big-to-fail quantities?
    I have stressed that in principle the risks in backstopping 
these firms cannot be calculated and priced in the 
straightforward way that the risks of bonds or insurance 
contracts can.
    Now I want to convince you to characterize bailout support 
as equity funding, as loss-absorbing equity funding, provided 
to a zombie firm when no one else will give it a nickel. We 
have got to see that managers who adopt risk management 
strategies that willfully conceal and abuse taxpayers' equity 
stake are sanctioned explicitly by corporate and criminal law 
rather than excused by insurance law as inevitable moral 
hazard.
    I believe the way we frame problems is critically important 
in making policy.
    If we think of bailout support as a loan, if recipients pay 
it back, it is a good loan.
    If we think of it as insurance, we would suppose that 
actuaries have been able to somehow figure out the risks and 
that the Government should be able to price and control its 
exposure to moral hazard.
    I am saying----
    Chairman Brown. Please wrap up.
    Mr. Kane. OK. Recklessly pursuing tail risk is an ethical 
violation.
    Regulatory capture has actually infiltrated the 
bureaucratic system that is supposed to limit risk-taking and 
sewn loopholes into the rules. Capital requirements in 
particular have gone very, very awry.
    I believe that genuine reform would compel the Department 
of Justice to prosecute megabank holding companies that engaged 
in easy-to-document securities fraud. There is value in 
documenting the violations and prosecuting these crimes in open 
court.
    I know that Senator Warren has been pushing a bill that 
would make settlement deals much more transparent. We need to 
underscore how managers benefit when the fines fall only on the 
shareholders.
    But BHC managers that have committed theft by safety net 
from taxpayers are individuals. If we do not set up sanctions 
that punish individuals, we are going to get even more theft in 
the future.
    Chairman Brown. OK. I am going to cut you off. Thank you, 
Dr. Kane.
    Dr. Admati, thank you.

   STATEMENT OF ANAT ADMATI, GEORGE G.C. PARKER PROFESSOR OF 
 FINANCE AND ECONOMICS, GRADUATE SCHOOL OF BUSINESS, STANFORD 
                           UNIVERSITY

    Ms. Admati. Chairman Brown, Ranking Member Toomey, and 
Subcommittee Members, I am very grateful for the opportunity to 
speak to you today.
    too big to fail is primarily about the collateral damage 
from the failure of some very large and complex companies.
    The global financial system is highly interconnected, 
opaque and fragile, like a set of dominos near one another. The 
systemic dominos are particularly large and central. If one of 
them fails, or if people fear that it might be unable to 
fulfill some of its many promises, the system may collapse. And 
the same is true is multiple banks are in trouble at the same 
time even if they are not so big.
    The trends are not encouraging on this despite the Dodd-
Frank Act, and I can elaborate on that later if you would like.
    In this scenario, the Government and the Fed must choose 
whether to let the natural process of failure play out or to 
intervene and, if so, how.
    When Lehman Brothers, a medium-sized investment bank that 
did not take deposits, filed for bankruptcy almost 6 years, the 
experience was traumatic. Massive interventions supported 
numerous other institutions. Some of the largest ones were 
among big-time recipients.
    Here are some questions:
    Is this fragile system the best system we can have?
    Must we live with it like we must live with the risk of an 
earthquake?
    Was the financial crisis the 100-year flood, and it is not 
cost-beneficial to build an expensive dam just for the rare 
event?
    Would making the system safer entail sacrificing its 
benefits?
    Is the problem just that they are too big and, therefore, 
breaking them up would solve the problems?
    Is finding a way to make them fail the solution?
    The answer to all these questions is no.
    We first must diagnose the problems and see which ones are 
most solvable and at reasonable cost. This does not happen, 
unfortunately. We are still living in a sick system, and we 
keep missing the most effective and straightforward medicines, 
and we can do something about this.
    Let me say, briefly, the failed scenarios are all bad, and 
even changing the bankruptcy law will not help that. Bankruptcy 
certainly, as it is, is not going to work without collateral 
damage. So we can commit to something, but we have to bear the 
results of what we decide to do.
    I hope the Fed does not need too many iterations with its 
living wills to admit this obviously reality. I am glad Senator 
Warren pushed Chair Yellen about this recently.
    I have further questions to ask, and I did in my statement.
    Title I of Dodd-Frank gives the Fed perfect authority to 
act not only once it admits that bankruptcy is not a viable 
option without harm but, right now, while it is still 
iterating. Instead, the Fed is failing the public.
    Title II, orderly liquidation authority, is not really 
liquidation. It can maintain the firm's infected intents, too.
    The FDIC is doing the best it can. This option is better 
than bankruptcy right now, certainly, but Title II is also not 
the solution of too big to fail, and it is especially not a 
solution to the inefficiencies and distortions in the system.
    Harm starts much before the fail is reached. In fact, the 
system is inefficient and distorted every single day, and much 
more can be done to improve it.
    This hearing is about implicit supports, and we heard 
something about it. I have a lot to say on that, and in fact, 
some of the documents that I submitted elaborate a little bit 
more on some of what Dr. Kane said over here and more about 
what these insidious subsidies really are.
    Implicit supports that the banks do not pay for do create a 
subsidy. Measuring the size of this subsidy is really 
complicated. A lot of assumptions are made, and often bad 
assumptions are made.
    But it is also besides the point is what I what to say. 
Even if you charge the companies every penny of the monetary 
cost of this subsidy, even if you did, and that would be 
difficult, this is a very bad system.
    I cite some papers in my statement that the GAO report and 
industry studies do not cite.
    And the bottom line is very clear; the subsidies are real, 
and they are very large.
    The main problem with the guarantees is that they reinforce 
an already distorted set of incentives and exacerbate the 
inefficiencies of the system. They create perverse incentives, 
and they enable excessive growth and other bad decisions. They 
intensify the conflict of interest between the banks and the 
rest of society.
    We are not getting what we want for these subsidies. We are 
maintaining a sick system.
    The key to the fragility of this system is the fragility of 
the institutions in it. This is not about the risk they take or 
about micromanaging them. This is about how they fund their 
investment.
    This is the most immediately fixable disease; banks simply 
use too little equity and too much debt. The tough regulations 
allow them to fund with 95 percent debt. This is unheard of 
anywhere without regulation, and there is nothing good about 
it.
    Citigroup would have satisfied these requirements in 
December 2007. These requirements are outrageously 
inappropriate.
    At the current levels, this level of indebtedness magnifies 
the risk in the system and creates fragility without doing 
anything useful for us at all. It works only for those in the 
system. The rest of us got talked into allowing it and living 
with the bad consequences.
    If you care about access to credit, we are not getting it. 
We are getting distorted credit. We get too little for some 
things that we want funded and too much elsewhere; that is 
wasteful.
    The key reason is that the lenders themselves have too much 
debt. They behave like distressed or zombie borrowers all the 
time.
    The people in the system are compensated to take risk, and 
the regulations are so bad; they distort the incentives further 
through a use of a risk weights that favor some investments 
over others and are manipulatable and distorted.
    We want risk to be taken with the right funding so that 
when things go wrong and do not work out there is no collateral 
damage.
    You may also believe that regulations are automatically bad 
and costly, but that is false. Some regulations are essential 
and good. Forcing banks to use more equity and less debt, if 
done effectively, brings only benefit.
    It acts to correct the distortions. It is a correction to 
what otherwise is crony capitalism. This is an area where the 
markets are not working, and only regulations or laws can 
correct this.
    We are doing a little bit of it, and they will tell you 
that we are doing a lot more. But we are doing it bad and 
insufficiently and ineffectively.
    The analogy I give is having speed limits that are entirely 
inappropriate. The fail is like ambulances and hospitals. We 
are not going to allow trucks to drive at 95 miles an hour 
through a residential neighborhood and take the chance that 
they implode and they burn the engine meanwhile. This is not a 
safe speed, and it is not hard to see that.
    Similarly, no corporation must live on 95 percent debt. 
Nobody does except in banking. Nobody can live like the banks 
except for the way they are allowed to and get away with. And 
that is without regulating it.
    I urge you to engage on these issues and to do what you can 
to improve this part of the regulation. The public should own 
this place, and the public is not served. Flawed claims that 
seem to have outside impact on this important debate are having 
that impact and should not.
    I look forward to a further discussion.
    Chairman Brown. Thank you, Dr. Admati.
    Dr. Holtz-Eakin.

 STATEMENT OF DOUGLAS HOLTZ-EAKIN, PRESIDENT, AMERICAN ACTION 
                             FORUM

    Mr. Holtz-Eakin. Thank you, Chairman Brown, Ranking Member 
Toomey, and Members of the Committee, for the privilege of 
being here today.
    I am going to make four brief points in my oral remarks, 
each of which is elaborated on in the written testimony that I 
submitted.
    Point number one is that any expectation of support for a 
bank holding company is, at its root, the result of 
discretionary policy actions taken, and the problem begins with 
policymakers and ends with policymakers. It is not something 
that the banks did. It is not something that the creditors did.
    It is something that the policymakers, in a discretionary 
fashion, did. And creditors took it into account, charged too 
little in their loans to banks. Banks correctly responded to 
those incentives and expanded their portfolios inappropriately.
    It is a very simple identification of the problem. It is a 
policymaking problem.
    The second point is that the history of such interventions 
in a discretionary fashion is very erratic. It has consisted of 
interventions on behalf of large firms and on behalf of some 
very small firms. It has been interventions on behalf of 
financial firms and nonfinancial firms. Even in the most recent 
crisis, we saw both large and small banks receive some 
intervention, and we saw auto companies receive intervention.
    The nature of the intervention has changed. In some cases, 
it is to preserve the financial stake of bondholders. In other 
cases, they have been wiped out.
    And so it was unsurprising to me, having written my 
testimony before I read the GAO report, that when I read the 
GAO report they said there was a wide range of expectations 
among market participants and monitors about the nature of a 
potential Government intervention. Given the history, that 
range of expectations is utterly unsurprising to me and 
something the GAO should have found.
    The third point is that econometric attempts to pull out of 
differences in bond yields anything like quantifiable too-big-
to-fail subsidy is a really elusive quest. It is well 
established that there are borrowing differentials between 
large and small entities in lots of industry, and indeed, 
financial services does not stand out as an especially large 
differential in the data.
    There are good reasons why markets might reward diversified 
firms, firms that have greater liquidity, and reward them even 
disproportionately in a crisis when liquidity and being able to 
move your financial assets is especially important.
    As a result, taking apart the differences in those bond 
yields is going to be highly sensitive to the nature of the 
specification, the nature of the estimation.
    And I think what you see in the GAO report is a stark 
tribute to exactly that--that we are going to get a different 
answer depending on how we do this analysis and you are 
unlikely to find a single sign or a single number on which to 
hang the analysis.
    And the last point is that given that this is an issue that 
comes from policymaker interventions it is hardly surprising 
that any too-big-to-fail expectation would change over time. We 
have seen radically important changes in the policymaking 
environment since the most recent financial crisis--Dodd-Frank, 
the Federal Reserve's activities, you know, look across all the 
regulators, the existence of the FSOC, the activities of the 
FSOC.
    One would expect this to change. Indeed, the GAO had found 
that it had diminished somewhat.
    If you want to eliminate it, you have to eliminate the 
problem, which is discretion.
    And in that regard I am quite sympathetic with the notion 
of Senator Toomey and his coauthors, that what you want to do 
is eliminate the discretion, put a fixed set of rules in the 
hands of a bankruptcy judge, design the rules so that you can 
deal with the admitted problem that illiquid financial 
institutions can become insolvent in a rapid fashion. So you 
have to build the system around that.
    But the solution to this is not to be found in changing the 
behavior of the creditors or the banks. It is changing the 
opportunity for behavior on the part of the policymakers.
    And I think the bankruptcy route and things like that are 
the most promising way for eliminating this problem in the 
future.
    Thank you, and I look forward to the chance to answer your 
questions.
    Chairman Brown. Thank you, Dr. Holtz-Eakin.
    I want to start by observing that the GAO used three 
industry-funded studies to design this report, and the GAO 
arranged meetings with corporate treasurers of companies 
suggested exclusively--I believe exclusively--by the U.S. 
Chamber of Commerce, the same organization that sent a letter 
to the Subcommittee, sort of extolling the virtues of the 
largest banks in the country. So, just that observation.
    I want to start with Dr. Anginer and then work my way 
across the table, about the study.
    The GAO said the subsidy may have been reduced because of 
``improvements in banks' holding companies' financial 
conditions.''
    One of the witnesses at our last hearing said that banks 
are in the business of taking prudent risks.
    Tom Hoenig estimates that the 8 U.S. globally, systemically 
important banks, the G-SIBs, have a 6.5 percent leverage ratio 
under U.S. accounting rules and a 4.62 percent ratio under 
international accounting rules. By comparison, the 10 largest 
banks in this country, under $1 billion, have a 9.25 percent 
leverage ratio under both measures.
    So my question is--and starting with you, Mr. Anginer, and 
working across--how do we encourage institutions to engage in 
prudent risks useful to our economy, prudent risks without 
incenting the kind of reckless behavior that leads to bailouts?
    If you would answer that, and then Dr. Kane and Dr. Admati 
and Dr. Holtz-Eakin.
    Mr. Anginer. I think market discipline is very important. 
So the type of policies that encourage greater market 
discipline would be very useful in that regard.
    One of the things that we show in our analysis is that one 
of the effects of this too-big-to-fail doctrine is that it 
dulls market discipline. It reduces investors' incentives to 
monitor the risk-taking of these financial institutions.
    So, to the extent that these policies reduce those 
incentives, you would expect greater risk-taking and not only 
greater risk-taking but a certain type of risk-taking as well.
    So, if you are a bank and you fail and none of the other 
banks are failing, you will most likely not get a bailout. If 
you fail when everybody else is failing, then you will get a 
bailout.
    So this sort of incentivizes banks to take on a correlated 
risk, or similar risk, to all the other banks. This is also 
another important aspect that is often overlooked.
    And what we showed in our analysis is that this is usually 
penalized in other industries. So, if you are a company and you 
take on risk similar to other companies in the industry, the 
market will actually penalize you whereas in the financial 
sector we do not see this; they actually get a benefit.
    So market discipline is very important toward this type of 
perverse incentives.
    Chairman Brown. Dr. Kane.
    Mr. Kane. What I see is a failure of the rule of law. 
Corporate and property law needs to recognize that the taxpayer 
is given an equity stake in these firms. And by thinking of the 
safety net as insurance, we excuse it as moral hazard and say 
it is regulators' problem to stop it, even though they do not 
have the tools to stop it.
    Individual managers are paid to take potentially ruinous 
tail risks and rewarded and promoted on the basis of the value 
they extract by taking tail risks.
    Congress needs to enact offsetting personal and corporate 
penalties for willful efforts to pursue risks that recklessly 
abuse taxpayers' equity stake and visit problems on 
nonfinancial and household sectors in the economy. Corporate 
penalties could include forced sales of some or all kinds of 
business.
    I think it is useful to think of the taxpayers' stake in 
each too-big-to-fail firm as if it were a trust fund and 
conceive of Government officials as fiduciaries responsible for 
managing that fund.
    The reforms I propose seek to give regulators and managers 
and directors of too-big-to-fail firms an explicit and codified 
fiduciary duty to measure, disclose and service taxpayers' 
stakeholding fairly and competently.
    We have to rework bureaucratic and private incentives to 
make this happen. That means regulatory agency and corporate 
mission statements must explicitly define, embrace and enforce 
fiduciary duties of loyalty, competence, and care to 
taxpayers--to taxpayers--who are implicit shareholders and are 
being treated unfairly relative to the explicit shareholders.
    Chairman Brown. Dr. Admati.
    Ms. Admati. Well, corporate governance is broken in banks, 
mostly because most of the money they invest is creditors' 
money, but the creditors are not really there in the 
governance.
    And what I am proposing and what is being proposed before 
is very complicated to do, but the first thing we have to do is 
bring the risk back to the investors. And the investors that 
should take the risk are the shareholders, and there should be 
more of their money.
    It is as if you increase their liability, but instead, you 
just have more of them with limited liability because that is 
what corporations are.
    So partnerships have more liability. Banks should have more 
responsibility and liability on their balance sheets. And the 
way to do that is not through triggers and through imposing it 
on creditors because of the collateral damage of doing that.
    That is why the natural place is the way other companies do 
it, and the way they do it in other companies is because the 
creditors start telling them that they will not stand for it 
anymore unless they write incredibly punishing covenants and 
conditions and increase the costs.
    So if you have more shareholders, then you get more 
discipline from them.
    If banks were sent to the equity markets, they would be 
told by equity investors that they are too opaque. That is what 
they are saying. Listen to equity investors say that they 
cannot understand the disclosures, that they are uninvestable.
    And if they do not give them a high price for their equity, 
then we need to know that and we need to know why. And that is 
the price of their worth because maybe their balance sheets are 
too inflated.
    So the point is if they cannot raise equity then--at any 
price, then they fail a market stress test, and that is a 
stress test much better than the stress test that the Fed 
wants.
    Chairman Brown. Dr. Holtz-Eakin. Mr. Holtz-Eakin. I will 
echo at the outset some of those remarks in that you want 
creditors doing their job, which is monitoring the risks that 
are embedded in the activities of a bank.
    And in addition to that you want the shareholders keenly 
aware that their money is at risk and that if, in fact, the 
bank is not well-run they will lose it and that it will go 
away.
    If they do that, if the external forces convey appropriate 
appreciation of the risks, something very important will 
happen.
    One of the things I learned on the Financial Crisis 
Commission is that many of these institutions had remarkably 
bad internal risk assessment during the course of this period. 
I was stunned, but it was really very surprising.
    If the outside creditors and shareholders are telling you, 
do not worry about risk, you do not devote anything in your 
corporate management to controlling those risks. And it conveys 
all through these organizations.
    So we would get better risk management, not through the 
regulatory approach, but just from the outside incentives 
permeating the culture of the firm.
    Now, to do that, you have to take off the table the 
assistance, or expectation of assistance, by the Government in 
bad times. And that is the hardest thing because, honestly, 
when things go bad policymakers become very risk-averse, 
infinitely risk-averse.
    And, you know, they have asked many times, is there any 
chance this could happen?
    You say, well, there is a two-tenths of a billionth of a 
probability that something very bad is going to happen.
    And they say, OK, well, then let's pull whatever leverage 
we have got.
    You have to take those levers away to get this to work, and 
that should be the focus of the Committee--finding ways to 
establish rules that will reorganize or liquidate these 
entities. Then the outside forces will do their job.
    Chairman Brown. Thank you.
    Senator Toomey.
    Senator Toomey. Thank you, Mr. Chairman.
    Senator Moran. Would the Senator yield just for a second?
    Senator Toomey. I would be happy to yield to the Senator 
from Kansas.
    Senator Moran. Mr. Chairman I am leaving to speak on the 
floor. I appreciate the Senator from Pennsylvania allowing me 
just a minute of his time, and I hope to return to ask 
substantive questions.
    But it is troublesome to me that in reading the testimony 
and hearing the testimony of the witnesses at least some of 
them appear to have known what was in the GAO report before the 
GAO report became public.
    And I tried recently, yesterday, to see if I could get a 
copy of the GAO report so I could be more intelligent in my 
asking questions, without success. I do not know whether 
something happened here that is inappropriate, but it does seem 
to me that there may be a double standard in who has 
information about their testimony and who did not.
    Senator Toomey. Thank you.
    Dr. Holtz-Eakin, I want to follow up.
    I think you make a very, very important point here about 
where the expectations for bailout come from, which one of the 
things you have stressed is previous policy decisions, right?
    Mr. Holtz-Eakin. Yes.
    Senator Toomey. So if, in fact, the Government comes along 
and bails out institutions, then there is obviously some level 
of expectation that the Government will do this again in the 
future.
    But would you agree that in Title II of Dodd-Frank there is 
an explicit mechanism by which the regulators are empowered at 
their discretion to tap into taxpayer funding and execute a 
bailout of sorts.
    And so it would seem to me that the codification of that 
mechanism also contributes perhaps very significantly to the 
expectation that it might, in fact, get used.
    Could you comment on that?
    Mr. Holtz-Eakin. Yes, I am quite sympathetic to that point 
of view.
    I mean, you can think of this in many ways as codifying 
what has been our experience with Fannie Mae and Freddie Mac, 
who we put on taxpayer funds, kept alive for a long time, have 
not fundamentally resolved in any way, and now you hear lots of 
agitation for them to be simply returned to the private sector.
    There is no discipline then, right? They get bailed out, 
and they survive.
    Senator Toomey. So is it your view that if we repeal Title 
II, which is the bailout mechanism of Dodd-Frank, and we made 
the reforms necessary for there to be a credible bankruptcy 
resolution that would actually be orderly and which would 
follow the ordinary rules of bankruptcy to the extent that we 
can--and I think you need to make some modifications to the 
current code to get there--would that at least diminish the 
likely expectations that there would be a taxpayer-funded 
bailout?
    Mr. Holtz-Eakin. I think so, yes. My reading of the broad 
amount of research--and it is not specifically the GAO report--
is that we have already seen a diminishment. That expectation 
has dropped dramatically, postcrisis.
    This would further diminish it. So I do not think it would 
impose a great shock on financial markets.
    I do think it would be a step toward sounder policymaking. 
It is a sensible thing to do.
    Senator Toomey. Would there be--if that were to be the 
case, if we had that policy change and the expectations of a 
bailout correspondingly diminished, would you be able to 
comment on any benefits for taxpayers, for markets, for 
allocation of credit? Are there other benefits that come about?
    Mr. Holtz-Eakin. Certainly, you would have less of a need 
for an elaborate supervisory regime, which is quite costly and 
raises the cost of credit and diminishes access to credit for 
consumers. I think that would be a tremendous benefit.
    You would have the benefit of a set of rules that all 
market participants understood rather than guesses at the 
discretionary actions of policymakers in a crisis.
    And all of those things produce better outcomes.
    Senator Toomey. Dr. Kane, if I understood your comment, I 
thought at one point you said that your recommendation is for 
the Government to impose penalties on the people who would take 
excessive risks because the taxpayers are the involuntary 
shareholders, effectively.
    Mr. Kane. Absolutely.
    Senator Toomey. Well, that, of course, supposes that the 
Government knows what is excessive, knows what the penalty 
ought to be for what somebody decides is excessive.
    And I wonder if a better approach is simply to make sure 
that the taxpayer is not an involuntary shareholder in the 
institution in the first place so that creditors can enforce 
behavior by virtue of their decisions on pricing and allocation 
of credit.
    Mr. Kane. Well, I think the best is often an enemy of the 
good; that is, we would like to make sure that authorities did 
not make bad policy, but the stresses of a crisis make bailouts 
so much easier than resolutions.
    We see bailout behavior all around the world. Along with 
some colleagues at the World Bank and the IMF, I have studied 
what happened to the safety net in 196 countries during this 
last crisis.
    In country after country, if they did not have deposit 
insurance, they created it. If they had deposit insurance, they 
extended the coverage and even guaranteed other kinds of debt 
of TBTF organizations. It is just so much easier to extend 
guarantees than running the risk of disorganizing the system in 
crisis circumstances.
    To get them to do hard things, we have to change the 
incentives of the bankers who play this risky game, and not 
just blame the authorities.
    What is the purpose of Section II? It responds to the 
excuse for inaction that authorities gave. This excuse was: 
``Oh, we did not have the powers to resolve TBTF firms.''
    So DFA gave regulators more powers, but power is not the 
true problem.
    The problem is incentives. Top officials are risking their 
careers and reputations, if they try to allocate losses. It 
resembles the issues in exercising tough love. It is very hard 
for parents to show tough love when their child is caught, say, 
smoking marijuana or worse. Instead, they hire a lawyer to save 
them from the consequences of their bad behavior.
    On the other hand, the goal is not to send hordes of 
managers to prison. What we want to do is to build a system 
where the people who have been abusing the safety net have 
better incentives in the future.
    Senator Toomey. I think Dr. Admati has something she would 
like to add to this.
    Ms. Admati. All the benefits that were just mentioned here 
to having a bankruptcy code--which, by the way, in the 
modification of the bankruptcy code that you would need to have 
all the big institutions fail under bankruptcy, and good luck 
to all of us in that scenario--you would also need some kind of 
backstop.
    You yourself said you were going to need some liquidity 
support.
    It would look--and I heard the FDIC discuss this with 
bankruptcy experts. It would look very much like Title II. So 
where bankruptcy would change, it would change in the direction 
of where Title II is right now. I think the distinctions there 
are very minor.
    The point is that all the benefits to credit and less 
micromanaging are going to happen with more equity. Then there 
would be more money if banks retain their earnings, more money 
to make loans and better incentives.
    And so there is nothing about correcting that distortion 
that goes against anything that we want in this system. 
Nothing.
    Senator Toomey. Well, I see my time is expired, but Mr. 
Chairman----
    Chairman Brown. If you want to continue with Dr. Anginer, 
if you want to.
    Senator Toomey. Fine.
    Mr. Anginer. Just on the point of whether Dodd-Frank might 
have codified too big to fail, actually, there is some 
empirical support for that. We looked at risk sensitivity of 
cost of debt for these large financial institutions that are 
deemed to be too big to fail by Dodd-Frank, and there is 
actually a decline in risk sensitivity.
    So the cost of debt has become less sensitive to risk after 
the introduction, again, suggesting that this too big to fail 
might have been codified.
    Senator Toomey. Thank you, Mr. Chairman.
    Chairman Brown. Senator Reed.
    Senator Reed. Well, thank you very much.
    This has been a very informative panel. Thank you all.
    Dr. Admati, I think you made the point, and you made 
several good points, that the size issue is sort of irrelevant 
in many respects, that there is no perfect size, right size, 
that it goes really more to capital that the company has, the 
equity and the leverage ratios.
    But a lot of times we have to take very simple approaches, 
and there has been a lot of discussion about not letting 
institutions grow beyond $250 billion. That seems, to me, to be 
treating a symptom and not effective.
    Ms. Admati. I am glad you asked about that because about 
the size, the size is very large, and you have to wonder why. 
My answer to that is because the funding is too easy.
    And we saw conglomerates break up on their own weight, on 
their own inefficiency.
    If the funding gets straightened out, then we might begin 
to see it. For example, if it was beginning, if you were 
beginning to pass more and more of the down side to the 
shareholders, they might begin to see how inefficient these 
corporations are, and we might see a natural breakups that just 
make more nimble, better pieces of it.
    Yes, I think that some of the crude measures of size at 
some point by Government, et cetera, at some point, it really 
becomes just outsized by any measure at all. And so I can see 
that some of the crude rules that say, OK, enough is enough for 
size.
    These corporations are the largest in the economy by asset 
size. The amount that is controlled, sprawling across the 
globe, is unfathomable. Every study of governance shows a total 
breakdown, and all the repeated scandals are reinforcing that. 
This is a reckless industry.
    Senator Reed. But we could sort of exclude the larger ones 
with the hope that that would solve the problem, but that would 
not necessarily help those that sort of get underneath the 
limit if they are insufficiently capitalized or overleveraged.
    Ms. Admati. Well, you can have a system with tiny, little 
pieces all failing at the same time, all interconnected, and it 
is not--that is why just size is not going to do it.
    Senator Reed. I understand.
    And you raised some other issues. It is not just size. It 
is two others, which makes this a very challenging problem.
    It is management--the inability, because they are so large, 
to effectively manage even if the intention is to manage it 
well.
    And then, the interconnection to other institutions that 
might be poorly managed or led, or that take business 
approaches that compel the others to follow.
    And let me add a third dimension, which is we are in a 
global system, and what we do is affected by what other 
countries do.
    And, in fact, frankly, you have heard it before. One of the 
reasons these banks are so big is because they are global and 
their global competitors are just as big.
    And, oh, by the way and you might comment on this--they 
have a too-big-to-fail regime overseas, which is a competitive 
advantage that they have.
    Can you comment on that?
    Ms. Admati. Yes. And, again, I am very thankful for the 
question.
    That other countries might do worse than us is not a reason 
to follow them.
    So a huge problem in Europe, in fact, which is misdiagnosed 
because the Germans like to protect their banks, and so do the 
French, is that they have a sicker banking system than we do, 
and very bloated.
    And so, yes, we are not here to support a particular 
industry on our back. We are not allowing pollution when other 
countries allow it, and we should not allow excessive risk even 
if other countries allow it.
    We should protect our citizens. The foreign banks that are 
here, we control what they can do here.
    We should worry about what we can worry about and lead the 
world as opposed to follow to the lowest common denominator.
    Senator Reed. But let me ask Dr. Kane, too, to comment 
because Dr. Kane suggested--and I think your testimony together 
sort of is sympathetic--in that there are two paths.
    One is regulatory forbearance--of not being tough on 
capital, on equity, on management skills and things like that, 
and the other is the implicit subsidy for debt and for equity. 
And those are the two paths.
    For the first path there has to be very active regulation 
by the Federal Reserve particularly because I must say I do not 
think we are in a position to, on a daily basis, pass 
legislation to fix this.
    And then the second path is a much more structural path 
about just the public perception--equity perceptions.
    So I have just been very general.
    Dr. Kane, do you want to comment on this discussion we have 
had?
    Mr. Kane. Yes.
    First, when you ask why are U.S. firms so active globally, 
I would point out two disturbing phenomena.
    In the futures markets, we see parent firms ``de-
guaranteeing'' their foreign susidiaries stripping formal 
guarantees from foreign affiliates. This lets the parent search 
out markets that can least well discipline their risk-taking. 
This is exposing the U.S. safety net because distressed subs 
are going to be supported by the parent in a pinch and by U.S. 
taxpayers in the next crisis.
    Second, we are observing what is called ``inversion,'' 
where, say, a company like Morgan Chase could merge with a 
small foreign firm, move the headquarters abroad and not pay 
corporate taxes here. In fact, the CEO claimed the right to do 
this.
    To me, this behavior is an ethical scandal. I am apalled by 
the ethical environment of an industry that would thumb its 
nose at the Government that rescued it a mere 5 years ago.
    I mean, again, in ordinary human relations we would sock 
somebody like that in the nose.
    You know, it just----
    Senator Reed. You must be from Boston.
    Mr. Kane. Well, I spent a lot of time there.
    No, I am from Northeast Washington, DC. We did a lot of 
that, here, too.
    [Laughter.]
    Senator Reed. I just--may I have one?
    Chairman Brown. Sure, go ahead.
    Senator Reed. I think this has been terribly useful. And I 
will ask, at least, are we better off than we were in 2008 with 
some of the provisions we have adopted in Dodd-Frank like those 
on derivatives, and the fact that we have actually put emphasis 
on higher capital levels?
    I think there is a sense like nothing has changed at all, 
and my sense is some things have changed, maybe not enough, but 
some things have changed.
    Mr. Kane. Again, yes, some things have changed, but cross-
country swaps regulation is currently a complete disaster.
    In the U.S., we have the SEC and the CFTC trying to divide 
responsibility. Swaps and other derivatives are very protean 
types of instruments. There are futures now on swaps trading on 
exchanges. There is a problem with the capitalization of 
central clearing parties. They have taken on much more risk 
than before, but none of them has increased its reserve funds 
to support that, and no one is compelling them to do so.
    So one could say we are not in the same position in the 
sense that the details are different, but U.S. taxpayers are 
still very exposed. The next crisis threatens to be worse than 
the last one.
    Ms. Admati. I would say that some of the things that were 
done in Dodd-Frank are essential, and they gave authority that 
can be used well.
    The problem is the implementation. A few of the things that 
got written were written in such a way that makes 
implementation very difficult, like Volcker. And so in the end 
of all of that, you are not sure whether the thousand pages of 
rules are really going to do as much as was hoped. So that is 
one.
    And then Title VII is very important. So this is critical 
because the modernization act was an absolute disaster, and we 
know that.
    So the ability to regulate derivatives is there. However, 
CFTC is not getting budgets to do it. And Gensler did a mighty 
job of it except, you know, could not.
    So in the end the lesson that was learned was that the ones 
that did the bailouts extolled them, and the people learned 
that there will be more bailouts.
    So whether we can try to commit to not do them, it is very 
difficult. We should try to prevent it.
    Why are we having ambulances when we have speed limits so 
dangerous? That is the question we have to first ask.
    Senator Reed. Thank you.
    Thank you, Mr. Chairman. You are very kind. Thank you.
    Chairman Brown. Thanks, Senator Reed.
    Senator Vitter.
    Senator Vitter. Thank you, again, to all of our panelists.
    Let me start with sort of the biggest bottom-line question. 
Do any of you think that too big to fail as policy and market 
perception has ended, and if so, why?
    Anybody?
    Ms. Admati. The answer is no.
    Senator Vitter. OK. Do you----
    Mr. Kane. No, for me, too.
    Senator Vitter. OK.
    Mr. Holtz-Eakin. Greatly diminished, however. I mean, not 
gone but greatly diminished.
    Mr. Evans. And, clearly, we show that there is a variation 
over time across models, and it may depend on credit risk 
conditions. So we would say, no.
    Senator Vitter. OK. And do any of you think that firms 
perceived as too big to fail do not have a market advantage 
because of that?
    Mr. Kane. No.
    Ms. Admati. They do have a funding advantage.
    Mr. Holtz-Eakin. That is not so obvious, I would say, at 
this point, honestly.
    There would be a funding advantage, but these are also the 
same entities that are subject to an enhanced supervisory 
regime. They have a much bigger regulatory cost. They have 
capital charges. And on balance it is not obvious they have a 
competitive advantage against smaller banks.
    Senator Vitter. OK.
    Mr. Anginer. And just to clarify, I mean, losses to the 
taxpayer will depend on two things. One is the probability that 
these institutions will fail, as Dr. Evans pointed out, as well 
as the likelihood that they will be bailed out.
    Senator Vitter. Right.
    Mr. Anginer. So some of the regulations actually increase 
some of the--made the banks in a much better position than they 
were five, 6 years ago.
    So the likelihood of failure has come down, but a 
likelihood of a bailout still remains. It is good to make that 
distinction.
    Senator Vitter. So my second question, does anyone else 
think--Dr. Holtz-Eakin says it is questionable.
    Does anyone else think that a too big to fail does not have 
a market advantage as a result?
    Mr. Kane. Could I add something? This increased regulatory 
burden is scalable, so that it has actually helped big banks. 
They can assign someone to fill out all the forms, while 
managers of very small banks find this burdening. Smaller banks 
are going to have to get bigger in order to spread the costs of 
filling out the paperwork.
    Senator Vitter. Right. Well, I would certainly agree with 
that, and that sort of goes to Dr. Holtz-Eakin's point.
    I think the overall new regulatory environment is an 
advantage for big firms, not small firms, even though the 
burden is bigger for big firms.
    But, anyway, I want to use my time efficiently.
    Mr. Evans, do you have any response to the criticism that 
you all studied too-big-to-fail funding advantage in an overly 
narrow way by looking at bond debt?
    Dr. Kane talked about the entire equity side. In addition, 
a lot of people think that much of the advantage is in short-
term funding of money market liabilities, which you did not 
look at. Do you have a response to that broad set of 
criticisms?
    Mr. Evans. Right. Certainly, we did not exhaust the various 
ways institutions might benefit from perceptions of being too 
big to fail, but we used the dominant methodology in the 
literature, which is to study one particular slice of the 
liability stack.
    We think bonds are extremely important. You can learn a lot 
about bonds. It tells you a lot about what happens below it.
    So, if we are talking about trends--and, remember, we did 
not try to quantify a subsidy. We are talking about what 
happens over time. And if it happens in bonds, it is quite 
likely that in places where there is a lower priority you are 
seeing it as well.
    Now Dr. Kane, when he says equity, he is really referring 
to some of these option pricing modeling approaches, which are 
highly theoretical. In fact, if I know leverage, if I know 
volatility, I can mechanically produce a result, and there is 
no room for investor expectation.
    So our model was strongly informed by Professor Anginer's 
work----
    Senator Vitter. OK.
    Mr. Evans. ----and an independent review by some highly 
respected scholars.
    Senator Vitter. OK. And, finally, Dr. Admati, in your 
testimony you particularly focus on the significance and 
perhaps the potency of capital as a tool to dramatically lessen 
risk. Why is that, and can you expand on that a little bit?
    Ms. Admati. Well, equity is the most natural loss absorber. 
Equity gets the upside. Why should anybody but equity bear more 
of the downside?
    The fact of the matter is equity bears a downside for most 
corporations, first and foremost, and there is no corporation 
that has so much debt.
    And the banks do not have to have so much debt. They are 
not in the right range of equity levels that are reasonable for 
corporations. Without regulation, the markets tell them.
    And this is related to the point about what do supervisors 
need to do. The way to think about that, in my view, is not 
that the supervisors and the regulators are like equity 
holders. They take the place of the breakdown of credit markets 
for these precisely because the banks' ability to borrow comes 
with not enough strings attached that usual creditors do. And 
that includes insolvency.
    Most companies could not live like the banks.
    If you erase the labels from the banks and you gave them--
of course, you would have to erase a few zeroes so they are not 
recognizable as such, and gave their balance sheets to the 
banks, they would--with the disclosures that they have, they 
would not be able to borrow.
    My hypothetical funding costs for them? Very large, if they 
really lived on their own.
    In a hypothetical that they really are on their own, if we 
surely believe that with all the covenants and the 
prioritization that you have to think about, creditors will not 
come.
    What happens in the banks in reality--and it is easy to 
understand. Start with deposits. Start especially with 
deposits. Depositors are the most passive creditors. They do 
not even know they are creditors. They do not even think of 
themselves as creditors. That is where the problem starts.
    A bank CEO could say, I have a lot of deposits; therefore, 
I do not have a lot of debt.
    Right there is the problem. The fact that the bank CEO can 
momentarily forget that he owes the deposit shows you the 
problem right there.
    From that time on--and deposits are unsecured--they can go 
and borrow more with the assets they buy with deposits, and the 
creditors will let them do it. And then the next creditor will 
let them do it.
    Companies cannot live like that in the real world, and the 
banks should not be allowed to.
    So what the regulators are doing is only coming in instead 
of the usual credit or covenants that would normally happen to 
a company that look like this. They should not live like this.
    That is what we have to do. The missing piece of it is that 
there is no credit discipline, and there will not be, and it is 
not the most effective way to get the discipline.
    First, push the rest of the balance sheet. Then there is 
obviously corporate governance that has to care about risk 
management.
    First, it has to be their own money.
    Senator Vitter. Thank you.
    Chairman Brown. Thank you, Senator Vitter, and thank you 
for your assistance on this.
    Mr. Kane. May I just say one thing?
    Chairman Brown. Dr. Kane, certainly.
    Mr. Kane. Mr. Evans said that the GAO used the ``dominant 
methodology.''
    There are at least two broad methodologies. I believe the 
other one is actually dominant because we must be concerned 
about the taxpayer put--the ability to put losses to the 
taxpayer when they exceed stockholder equity. Equity is the 
natural loss-absorbing mechanism. To not look at taxpayer 
exposure as an option is conceptually inferior.
    The GAO methodology has a lot of people working on it who 
do not thoroughly understand option pricing.
    Chairman Brown. Dr. Evans, if you would like to speak to 
that.
    Mr. Evans. Just to quickly say, you would use that approach 
if you thought you were sure a subsidy exists and you just 
wanted to know how large it is.
    I mean, Merton, of Black-Scholes and Merton, used it in a 
deposit insurance context where it is most appropriate.
    Again, this is a highly theoretical model and makes some 
strong and extreme assumptions, and again, there is no room for 
investor expectations.
    Mr. Kane. Well, that is not true. It is changes in 
expectations that cause movements in the stock price that 
others use to pull estimates of the option out.
    So I am saying you do not quite understand the mechanism 
and are downplaying the role of assumptions in other economic 
research.
    Chairman Brown. Dr. Anginer, conclude on that discussion, 
referee for a moment, and then let me move on.
    Mr. Anginer. Sure, just a point on the methodology. It 
reminds me of the old joke about looking for a lost key under a 
lightpost because that is where the light is, not where the key 
is lost.
    And the reason we study bonds is that is where the data are 
available and makes the analysis much easier.
    But having said that, it is likely that we are 
underestimating risk for a number of reasons.
    One is that we are using equity prices.
    Also, bonds, publicly traded bonds, are the first ones that 
are going to get hit on the balance sheet structure, on the 
debt structure, when there is a failure.
    So there is good reason to think that most likely we are 
underestimating; so these subsidy numbers would be larger.
    Chairman Brown. OK. Thank you.
    Dr. Admati, you said that we should compare banks' funding 
with and without Government support rather than compare bank by 
larger banks and smaller banks, if I recall.
    One way to do this is to incorporate ratings uplift that 
big banks get as a result of that implied support, actually, 
that regional community banks do not get.
    The GAO report does not do that. They say it is an indirect 
measure of the funding advantage.
    How important--and feel free, all of you, to comment on 
this, too, but it is directed first to you, Dr. Admati.
    How important is including or excluding this uplift in 
measuring financial benefits for the too-big-to-fail 
institutions?
    Ms. Admati. Well, I think the credibility of credit rating 
agencies, of course, is somewhat diminished after the crisis. 
So we should be always skeptical of what they say.
    However, they are capturing something real, which is that 
investors perceive--and the credit rating agencies know--that 
in the hypothetical if the banks are really on their own their 
funding costs would be a whole lot higher.
    And it has not just the interest rate. It has the whole 
conditions that come with it.
    Most of what scares companies from becoming highly indebted 
is that the creditors will write such restrictive conditions 
that will not allow them to move.
    And the banks get credit under incredibly comfortable, easy 
burdens that are different from other companies.
    Credit uplifts are trying to get at the way to do this, 
which is not to compare to other companies but to compare the 
company as it is to the hypothetical company without support.
    That is what we are after, conceptually. The problem is it 
is a hypothetical that is counterfactual, that we do not have.
    That is why I said in my hypothetical, with really no--with 
knowing that the depositor can come in the bank, that other 
short-term creditors can come and dilute you as a junior 
creditor, the banks will have a very hard time borrowing. And 
some of the small banks have a hard time and, in fact, do not 
have bonds. So, already, you are biasing the whole discussion.
    The point is the banks are funding at an unreal world that 
they feel entitled to, that is completely outside the normal 
markets.
    So I think the credit rating--it is too bad that it was not 
included. Some of the studies that include it are informative, 
I would say.
    Chairman Brown. Dr. Holtz-Eakin, do you have something on 
that?
    I would just note--and someone can correct me if I am 
wrong--recently, most, if not all, the major credit rating 
agencies have removed the credit uplifts. So you could not do 
that now. If you care about too big to fail now, in 2013 and 
'14, it is gone from that perspective.
    Chairman Brown. Dr. Anginer and then Dr. Evans.
    Mr. Anginer. So just on that, we actually examined the 
effect of these ratings on the pricing of debt.
    So some rating agencies issue two types of ratings. One is 
called a standalone rating, just incorporating the risk of an 
institution on its own, without any Government support, and 
another rating that incorporates the Government support.
    So, if you look at how these two types of ratings are 
priced, it is really the Government support that is being 
priced, not the standalone rest.
    So it is true that some of the rating agencies have 
recently downgraded some of the Government support, but we do 
not know what the pricing implications of those are.
    Chairman Brown. Dr. Evans.
    Mr. Evans. And I would say rating agencies are a less 
direct measure. In fact, they are indirect. If you can go 
straight to the market and study actual investors that 
participate in the market, you are going to get a much cleaner 
answer.
    Now we have interviewed a number of investment firms. Some 
of them do their own credit rating analysis. So it is not 
necessarily the case that you can go from credit ratings to 
actual bond spreads.
    Chairman Brown. OK. Dr. Kane, and then I want to move on to 
a last question, and then we will wrap up.
    Mr. Kane. I do not see why you do not use all the data 
possible. Why would you look in just one place when you lost 
your keys in your house?
    Chairman Brown. Fair enough question.
    Mr. Evans. And that would apply to the option pricing 
approach, too, which is, again, highly theoretical.
    And if the option pricing model is a hammer and the world 
is a nail, you will always look there.
    But I will point out, in Dr. Kane's written testimony, 
Figure 3--which, again, you would expect to see this trend 
because volatility is lower, leverage is lower--looks a lot 
like the graphics from GAO's report.
    Chairman Brown. OK. Dr. Anginer touched on an issue. I want 
to ask one question, and all of you feel free to weigh in here.
    A lot of the talk is about whether too big to fail 
overfocuses on what would happen if one institution got into 
trouble and needed to be resolved.
    When you look at the last terrible financial crisis and you 
look to the future, it seems more likely--I mean it does not 
seem so likely that one large institution will fail, and only 
one. Large, universal banks, by and large, conduct the same 
activities, have the same kind of portfolios, making it 
unlikely that there will only be one getting into trouble at a 
time.
    So two questions, and I guess I will start with you, Dr. 
Anginer, since you touched on it earlier, and I just want you 
to expand.
    How does the universal nature of the largest banks affect 
too big to fail, initially?
    And then what sort of risk-taking--because, ultimately, 
this hearing is about what too big to fail leads to, and that 
is incenting risk-taking among these large banks and paying 
little price for it, except the price the public pays.
    What sort of risk-taking does that sort of universal nature 
of the largest banks incent, and does it make it more likely we 
have a systemic event rather than just isolated and just an 
isolated failure?
    So, Dr. Anginer and then anybody else that wants to weigh 
in.
    Mr. Anginer. Sure. I think large financial institutions are 
incentivized to take these type of correlated risks.
    As I mentioned before, again, if you are a large financial 
institution and you fail when nobody else is failing, you are 
less likely to get bailed out. This incentivizes you to take 
on--do activities that others are doing.
    And just to----
    Chairman Brown. So you are saying that if you are a CEO of 
a large bank and you understand what you just said, that means 
that you are going to want to act like the others. You want to 
mimic the other banks; they want to mimic you.
    Mr. Anginer. Exactly, exactly.
    Chairman Brown. OK, for sort of safety in numbers. So you 
are all bailed out because one of you would not be----
    Mr. Anginer. Exactly.
    Chairman Brown. ----if it were solely one. OK.
    Mr. Anginer. Exactly. And we actually do see this in the 
data as well.
    So, if you are in another industry and you do what 
everybody else is doing, taking on similar risk, the investors 
in the market actually penalize you.
    Why? Because they want to be diversified. They do not want 
their company to fail when everybody else is failing because 
they are going to suffer greater losses.
    We see the opposite effect in the financial sector. They 
actually get a benefit when they take on similar risk.
    So, again, that is because we have these perverse 
incentives, that because of this too-many-to-fail effect, this 
universal banking model that you mentioned. And it leads to all 
sorts of perverse incentives, actually increasing systemic 
fragility.
    Chairman Brown. Dr. Admati, next.
    Ms. Admati. Yes. I did not talk too much about the report 
and did not get into the details, but my commentary goes to a 
number of other issues that did not come up here.
    And I just want to say--and it is related to your 
question--this is something incredibly interconnected. That 
means the bailouts are interconnected. When AIG is bailed out, 
the banks are bailed out. When you bail out Greece, the German 
banks get bailed out. That is why you cannot even look exactly 
at one company and isolate these costs.
    It is much more complicated because it is not just that 
they do the same thing; it is that most of their activities are 
with each other. Actually, a small fraction of them come out to 
the rest of the economy. There is just so much intersystem 
activity in those balance sheets.
    And so the bailout issues are very, very complicated. In 
terms of measuring the subsidy, a lot of issues come.
    I want to say as well that I mentioned at least one op-ed, 
and it was specifically about this study here, that says that 
that study is underestimated.
    And I mentioned studies that are not cited in the GAO 
report that go to the return on the equity, not options, but a 
study by Kelly, et al., that you did not use, and Lustig and 
all of those that were not there, some of which was not 
options. And it showed directly that the banks benefit from all 
kinds of--in all kinds of ways.
    So the subsidy is underrated.
    On volatility is low now, credit ratings took the uplift, 
all of those things--I want to remind everybody; 2006 was a 
great year. Volatility was very, very low. Everybody was making 
record profits. OK.
    So let's just remember the good times can stop very 
quickly, and the euphoria of those good times and the low 
volatilities of these good times can change dramatically. 
Within less than a year or two, or starting mid-2007, 
volatilities have shifted from 10 percent on the VIX sort of 
index to 70 percent.
    And I was teaching through that time, so I could see it.
    Chairman Brown. Thank you.
    Dr. Holtz-Eakin.
    Mr. Holtz-Eakin. So there is an important difference 
between interconnectedness and suffering a common shock. And 
what we saw in the last crisis was a worldwide credit and 
housing bubble that, when it broke, was a common shock across 
all financial institutions.
    Institutions that rely on short-term funding and are 
heavily leveraged are the recipe for financial problems in 
those circumstances, and it is appropriate then to worry about 
an enhanced supervisory regime for those big institutions. And 
we have that now, and I think that is a sensible thing.
    And you should not somehow believe that they are all going 
to fall apart in the same instance. Indeed, I believe if you go 
back and just reread the history of the evolution of the 2007-
2008 crisis, everyone's remembrance of those awful days when 
Lehman went down, AIG, Fannie, Freddie, in close order. But the 
lead-up time was well over a year before that, a year and a 
half.
    I mean, we had early indicators of problems and stresses 
and policies that accommodated them, and the importance of the 
policy being better could have cut that off at the beginning.
    Now we are going to debate forever whether we should have 
let Bear Stearns go down along with Lehman or saved them both, 
but saving one and not the other was a disaster. That is 
inconsistent discretionary policymaking, and that is the 
problem.
    Ms. Admati. I would characterize the problem differently. 
The problem is one of forbearance and of not prompt corrective 
action. Fail is not a good option. That is the difficulty here.
    You can put all your eggs in the fail option, but we do not 
have to. That is why we have supervision that would interfere 
before.
    And I agree; the signs were there. But the fail allowed 
dividends out--dividends that would then have to be plugged in, 
or were plugged in, by top money that was actually debt and not 
equity, that they did not lend, that they just wanted to return 
so they could pay bonuses again.
    Chairman Brown. Dr. Kane.
    Mr. Kane. I would just like to emphasize the political side 
of being interconnected and being subject to common shocks. It 
is very hard in our system to say we are going to close, say, 
Bank of America and not Citi. That choice would just bring a 
tremendous amount of problems.
    Mr. Holtz-Eakin. That is why you should not give them a 
choice.
    Mr. Kane. Well, the----
    Ms. Admati. Well, some were not that big, by the way.
    Mr. Kane. Yeah, but my point is that if most BHCs are 
failing whatever test you run, or if they all are subject to a 
common shock, the situation requires choices to be made. The 
Government will not have the people in place to sort things 
out.
    Even with the slower processing of insolvency in 2008-2009, 
the FDIC brought people out of retirement to come back to work 
to help them. They did not have enough experienced personnel.
    Chairman Brown. Thank you all. This was a very helpful 
discussion. Thank you.
    And you saw a lot of interest from my colleagues. Some, I 
assume, will have questions. Some Members of the Subcommittee 
perhaps or the full Committee, too, may have questions. If you 
would get answers to those within a week, it would be very 
helpful.
    Special thanks to Graham Steele and Megan Cheney in my 
office and Travis Johnson in Senator Vitter's office for the 
work they have done for a year leading up to this.
    We have done a lot of Subcommittee hearings in this 
Subcommittee. This is one of the most important because it is 
the one that has such effect on our financial system. And the 
input from all of you was quite valuable.
    So the Subcommittee is adjourned and thank you.
    [Whereupon, at 3:38 p.m., the hearing was adjourned.]
    [Prepared statements, responses to written questions, and 
additional material supplied for the record follow:]
                  PREPARED STATEMENT OF LAWRANCE EVANS
   Director, Financial Markets and Community Investment, Government 
                         Accountability Office
                             July 31, 2014
































                  PREPARED STATEMENT OF DENIZ ANGINER
 Assistant Professor of Finance, Pamplin School of Business, Virginia 
                                  Tech
                             July 31, 2014
    Mr. Chairman and the distinguished Members of the Subcommittee, 
thank you for convening today's hearing, and inviting me to testify. My 
name is Deniz Anginer. I am an assistant Professor at the Pamplin 
Business School at Virginia Tech. Along with my colleagues, Viral 
Acharya and Joe Warburton, I have examined market expectations of 
implicit Government guarantees to so called ``too-big-to-fail'' 
institutions. \1\ Most of my testimony is based on this research.
---------------------------------------------------------------------------
     \1\ Viral V. Acharya, Deniz Anginer, and A. Joseph Warburton, 
``The End of Market Discipline? Investor Expectations of Implicit 
Government Guarantees'' (available at: http://ssrn.com/
abstract=1961656).
---------------------------------------------------------------------------
    The too-big-to-fail (TBTF) doctrine holds that the Government will 
not allow large financial institutions to fail if their failure would 
cause significant disruption to the financial system and economic 
activity. In our research, we find that large financial institutions 
and their investors expect the Government to back the debts of these 
institutions should they encounter financial difficulty. These 
expectations of Government support are embedded in the prices of bonds 
issued by major financial institutions, allowing them to borrow at 
lower rates.
    Expectation of Government support by the market also results in a 
distortion in how risk is reflected in the debt prices of large 
financial institutions. An implicit Government guarantee dulls market 
discipline by reducing investors' incentives to monitor and price the 
risk taking of large financial institutions. In our analyses, we show 
that while a positive relationship exists between risk and cost of debt 
for medium- and small-sized institutions, this relationship is 75 
percent weaker for the largest institutions. Changes in leverage and 
capital ratios are, likewise, less sensitive to changes in risk for 
these large institutions. \2\
---------------------------------------------------------------------------
     \2\ Acharya, Anginer, and Warburton (2014).
---------------------------------------------------------------------------
    Because they pay a lower price for risk than other financial 
institutions, the perceived guarantee provides TBTF institutions with a 
funding advantage. We find that the implicit subsidy has provided these 
institutions an average funding cost advantage of approximately 30 
basis points per year over the 1990-2012 period, peaking at more than 
100 basis points in 2009. The total value of the subsidy amounted to 
about $30 billion per year on average over the 1990-2012 period, 
topping $150 billion in 2009. We have also examined nonfinancial firms. 
If bond investors believe that all of the largest firms (both financial 
and nonfinancial) are too big to fail, then large nonfinancial firms 
should enjoy a size subsidy similar to that of large financial 
institutions. However, we find this is not the case.
    Compared to the GAO study, we find lower implicit subsidy values 
for the years 2007 to 2011 and slightly higher numbers in 2012. We have 
not examined 2013, the year in which the GAO finds the greatest 
decline. Although most of the attention will be paid to the analyses 
that try to quantify the dollar values of the subsidy and its changes 
over time, it is important to note that it is very difficult to 
directly relate these changes to the introduction of Dodd-Frank and 
other regulations.
    It is very hard to separate out changes in probabilities of large 
financial institutions experiencing distress from the probability that 
they will be bailed out. As the GAO report points out, this is 
especially true as the risk premium has declined in recent years and 
the large financial institutions have seen significant improvements in 
their balance sheets and capital ratios reducing their probability of 
experiencing financial distress.
    Although it is very difficult to establish a direct link between 
regulations and changes in subsidy over time, examining these changes 
using alternative methods over a short time window would be more 
helpful in analyzing the impact of Dodd-Frank and other regulations. 
For instance, in our study we examined changes in risk sensitivities of 
cost of debt after the introduction of Dodd-Frank. We examined changes 
in subsidies accruing to large financial firms compared to nonfinancial 
firms. We also examined the cost of implicitly guaranteed debt to 
explicitly guaranteed debt issued by the same firm under FDIC's 
Temporary Liquidity Guarantee Program.
    Using these alternative approaches, we find that Dodd-Frank did not 
significantly alter investors' expectations that the Government will 
bail out TBTF financial institutions should they falter. Despite its 
no-bailout pledge, Dodd-Frank leaves open many avenues for future TBTF 
rescues. For instance, the Federal Reserve can offer a broad-based 
lending facility to a group of financial institutions in order to 
provide a disguised bailout to the industry or a single firm. In 
addition, Congress can sidestep Dodd-Frank by amending or repealing it 
or by allowing regulators to interpret their authority in ways that 
protect creditors and support large institutions. \3\ As former Kansas 
City Fed President, Thomas Hoenig, noted: ``The final decision on 
solvency is not market driven but rests with different regulatory 
agencies and finally with the Secretary of the Treasury, which will 
bring political considerations into what should be a financial 
determination.''
---------------------------------------------------------------------------
     \3\ See, e.g., David Skeel, ``The New Financial Deal: 
Understanding the Dodd-Frank Act and Its (Unintended) Consequences'' 
(2011); Arthur E. Wilmarth, ``The Dodd-Frank Act: A Flawed and 
Inadequate Response to the Too-Big-To-Fail Problem'', 89 Oregon Law 
Review 951 (2011); Standard & Poor's, ``The U.S. Government Says 
Support for Banks Will Be Different `Next Time'--But Will It?'', (July 
12, 2011). One aspect of the recent regulations, that our analyses 
suggest had a mild impact in reducing market expectations of support, 
is the release of the specifics of the FDIC's Single Point-of-Entry 
approach to resolving financial institutions under OLA. This is 
consistent with the GAO interviews with large investors who point to 
the SPOE approach as affecting their expectations of future Government 
support.
---------------------------------------------------------------------------
    Finally, it is also important to note that the analyses conducted 
by us and the GAO only measure the direct subsidy that may accrue to 
TBTF institutions. There may be other indirect effects such as 
misallocation of capital or excessive and correlated risk-taking (to 
exploit the implicit guarantee) that are not captured by the analyses.
    Governments are generally not required to make any apparent 
financial commitment or outlay, or request funds from legislatures or 
taxpayers, when they implicitly guarantee TBTF institutions. Implicit 
guarantees lack the transparency and accountability that accompany 
explicit policy decisions. Taxpayer interests could be better served, 
in both good times and bad, by estimating on an ongoing basis the 
accumulated value of this subsidy. Public accounting of accumulated 
TBTF costs might restrain those Government actions and policies that 
encourage TBTF expectations.
    Thank you for your time.
References
Acharya, V., D. Anginer, and A.J. Warburton (2014) ``The End of Market 
    Discipline? Investor Expectations of Implicit State Guarantees''. 
    NYU Working Paper.
Skeel, David, (2011), The New Financial Deal: Understanding the Dodd-
    Frank Act and Its (Unintended) Consequences (Hoboken, NJ: John 
    Wiley).
Standard & Poor's, (2011), ``The U.S. Government Says Support for Banks 
    Will Be Different `Next Time'--But Will It?'' (July 12).
Wilmarth, Arthur E., (2011), ``The Dodd-Frank Act: A Flawed and 
    Inadequate Response to the Too-Big-To-Fail Problem'', Oregon Law 
    Review 89, 951-1057.
Appendix




                   PREPARED STATEMENT OF EDWARD KANE
                  Professor of Finance, Boston College
                             July 31, 2014
    I want to begin by thanking Chairman Brown for inviting me to 
testify today and to congratulate him and the Subcommittee for 
continuing to battle against the pernicious and unfair advantages that 
panic-driven crisis-management policies confer on mega-institutions, 
not only in this country but in financial-center countries around the 
world. The claim that the Dodd-Frank Act of 2010 or Basel III can end 
these advantages is a dangerous pipe dream. There will always be 
institutions that regulators will--especially in crisis circumstances--
find macroeconomically, politically, and administratively too difficult 
to fail and unwind. The existence of a powerful propensity to rescue 
such too-big-to-fail (TBTF) firms is the central lesson taught both by 
the S&L mess and by the Great Financial Crisis.
The GAO Has Bungled Its Assignment
    The GAO goes wrong at the outset. The definition of TBTF offered in 
the Report's first sentence (lines 9-10) is incomplete. It describes 
TBTF as an active policy of ``intervention'' without confronting the 
more dangerous additional role played by passive capital forbearance.
    The title of this hearing focuses on ``funding advantages'' that 
TBTF BHCs receive from expectations of unlimited Government support. 
The GAO's estimated 42 statistical models each year seek to explain in 
a robust manner only how the interest spreads between bonds issued by 
large BHCs and comparable Treasuries relate to BHC size and credit 
risk. This conception of TBTF subsidies treats TBTF guarantees as if 
they were merely a form of bond insurance and builds in an additional 
downward bias by not using volume-based proxies for the extent to which 
after-issue trading in individual BHC bonds is less liquid than in 
Treasuries.
    But even if they were modeled perfectly, spreads on outstanding 
bonds capture only part of the impact of TBTF guarantees. TBTF 
guarantees are different from bond insurance because, as long as 
regulators forbear from resolving a BHC's insolvency, a truly TBTF firm 
can extract further guarantees by issuing endless amounts of additional 
debt.
Funding Cost Is More Than Debt Costs
    A BHC's ``funding cost'' is the cost of its ``funding mix.'' Being 
TBTF lowers both the cost of debt and the cost of equity. This is 
because TBTF guarantees lower the risk that flows through to the 
holders of both kinds of contracts. The lower discount rate on TBTF 
equity means that, period by period, a TBTF institution's incremental 
reduction in interest payments on outstanding bonds, deposits, and 
repos is only part of the subsidy its stockholders enjoy. The other 
part is the increase in its stock price that comes from having 
investors discount all of the firm's current and future cash flows at 
an artificially low risk-adjusted cost of equity. This intangible 
benefit generates capital gains for stockholders and shows up in the 
ratio of TBTF firms' stock price to book value. Other things equal 
(including the threat of closure), a TBTF firm's price-to-book ratio 
increases with firm size. For four quarters in 2012-2013, Figure 1 
compares the behavior of this ratio for banks in different size ranges. 
The comparisons show that on average this ratio increases with size in 
all four quarters.
    I hope that contemplating the following numerical example can drive 
home the need to account for the equity-funding component of annual and 
capitalized TBTF subsidies. Let us suppose a TBTF institution is 
projected to earn $12 billion a year forever and that $2 billion of its 
earnings comes from the reduction in its cost of debt. By hypothesis, 
market participants recognize that TBTF guarantees shift a range of the 
deepest possible losses away from creditors and stockholders to 
taxpayers. If authorities were expected to take over the firm and pay 
off guaranteed creditors just as it became insolvent, the debt 
component would be the whole story. But because authorities are 
expected to leave the stock in play come hell or high water, TBTF 
policies give comfort to shareholders, too. This comfort lowers the 
risk class of the stock, so that the warranted return on equity falls.
    Let us assume that the opportunity cost of equity would be 12 
percent without the TBTF guarantee, but--in the presence of the contra-
liability provided by the unlimited guarantee--this cost falls to 10 
percent. Then, the capitalized subsidy built into the stock price would 
be not $16.7 billion ($2 billion/.12) or even $20 billion (=$2 
billion/.10), but $36.3 billion. The capitalized subsidy is the 
difference between the $83.3 billion stock-market value of the 
unguaranteed firm (=$10 billion/.12) and the $120 billion ($12 
billion/.10) in value that develops under TBTF guarantees. The annual 
subsidy that would deserve to be passed through the Federal budget 
would be $4.4 billion: the $2 billion in interest saving plus another 
$2.4 billion (.02 x $120 billion) subsidy on the firm's equity funding. 
So, for this hypothetical BHC, the annual subsidy to equity would prove 
roughly the same size as the subsidy to debt.
    The warranted rate of return on the stock of deeply 
undercapitalized firms like Citi and B of A would have been sky high 
and their stock would have been declared worthless long ago if market 
participants were not convinced that authorities are afraid to force 
them to resolve their weaknesses. Had these BHCs' assets and 
liabilities been transferred to bridge institutions or put into 
resolution, losses that contractually deserved to be incurred by 
uninsured creditors and postcrisis increases in the TBTF stock prices 
would have accrued to taxpayers.
    A simpler way to see what the GAO has missed is to think carefully 
about the structure of guarantee contracts. An external guarantee 
allows the guaranteed party to put responsibility for covering debts 
that exceed the value of BHC assets to the guarantor. No guarantor 
wants to expose itself to unlimited losses on this put. For this 
reason, all guarantee contracts incorporate a stop-loss provision that 
gives the guarantor a call on the guaranteed party's assets. 
Ordinarily, this right kicks in just the insolvency threshold is 
breached. In the FDIC Improvement Act of 1991, efforts to exercise this 
call are termed ``prompt corrective action'' (PCA).
    By definition, the Government's right to take over the firm's 
assets will never be exercised in a financial organization is truly 
TBTF. This means that the Government has effectively ceded the value of 
its loss-stopping rights to TBTF stockholders. The value of this 
giveaway is what the GAO's measure ignores.
    I can clarify this further by examining Figure 2. This figure 
graphs the behavior of AIG's stock price before, during, and after the 
2008 crisis. The only times AIG's stock price approached zero was when 
a Government takeover of the firm was being actively discussed. Each 
time that this possibility was tabled, trading picked up and the stock 
price soared as new stockholders tried to share in the value of the 
unexercised call.
GAO Neglect of Differences in Political Clout
    Postcrisis reforms seek to classify particular firms as either 
systemically important financial institutions (SIFIs) or not. But TBTF 
status is not a binary condition and does not start at a particular 
size. A firm's access to Senators and Congresspersons grows steadily 
with its geographic footprint and with the number of employees that can 
be persuaded to contribute to reelection campaigns. TBTF BHCs give 
heavily to candidates in both political parties as Ferguson, Jorgenson, 
and Chen (2013) have documented. Holding size constant, the more 
organizationally complex and politically influential an institution 
becomes, the better the chance that Government examiners will find it 
difficult to observe its exposure to tail risk and to discipline such 
risk adequately.
Need To Bring in the Behavior of Stock Market Prices
    To capture the full extent of TBTF subsidies, it is critical to 
make use of stock-market data. Figure 3 of my presentation tracks 
annualized estimates that Armen Hovakimian, Luc Laeven, and I (2012) 
have made of the average dividend that taxpayers ought to have been 
paid on their stake in large BHCs. This Figure plots the mean value of 
the credit support in annualized basis points per dollar of assets 
supplied to large banking organizations, quarter by quarter between 
1974 and 2010. The surge in the third quarter of 2008 is remarkable, as 
is its steady fallback afterwards.
    Regulators and policymakers persistently misframe bailout 
expenditures as either loans or insurance. This false characterization 
helps TBTF firms and their creditors to steal wealth from taxpayers. An 
insurance company does not double and redouble its coverage of drivers 
it knows to be reckless. Similarly, lifelines provided to an underwater 
firm should not be thought of as low-interest loans. Loans are simply 
not available to openly insolvent firms from conventional sources. The 
ability to extract implicit guarantees on new debt and the hugely 
below-market character of bailout programs means that the repayment of 
funds that were actually advanced does not show that a bailout program 
is a good deal for taxpayers.
    Bailout funding can more accurately be described as unbalanced 
equity investments whose substantial downside deserves to carry at 
least a 15 percent to 20 percent contractual return. The Government's 
bailout deals compare very unfavorably with the deal Warren Buffet 
negotiated in rescuing Goldman-Sachs. Buffet's deal carried a running 
yield of 10 percent and included warrants that gave him a substantial 
claim on Goldman's future profits. Government credit support 
transferred or ``put'' to taxpayers the bill for past and interim 
losses at numerous insolvent or nearly insolvent TBTF firms. 
Authorities chose this path without weighing the full range of out-of-
pocket and implicit costs of their rescue programs against the costs 
and benefits of alternative programs such as prepackaged bankruptcy or 
temporary nationalization and without documenting differences in the 
way each deal would distribute benefits and costs across the populace 
(see Bair, 2012).
    In my opinion, it is shameful for Government officials to imply 
that TBTF bailouts were good deals for taxpayers. On balance, the 
bailouts transferred wealth and economic opportunity from ordinary 
taxpayers to much higher-income stakeholders in TBTF firms. Ordinary 
citizens understand that this is unfair and officials that deny the 
unfairness undermine confidence in the integrity of economic 
policymaking going forward.
How To Sanction the Pursuit of TBTF Subsidies
    I hope my testimony convinces you that, in principle, the risks in 
backstopping TBTF firms cannot be calculated and priced in the 
straightforward ways that the risks of a bond or insurance contract 
can. Taxpayer guarantees to TBTF creditors provide unlimited loss-
absorbing equity funding to zombie firms at a time when no sensible 
private party would even advance them a dime.
    I want to convince you further that interpreting bailout support as 
equity funding implies that managers who adopt risk-management 
strategies that willfully conceal and abuse taxpayers' equity stake 
should be sanctioned explicitly by corporate and criminal law rather 
than excused by insurance law as inevitable moral hazard.
    I find it disgraceful that corporate law legitimizes managerial 
efforts to exploit taxpayers' equity position. The norm of maximizing 
stockholder value is inappropriate for TBTF firms. In TBTF 
institutions, this norm leaves taxpayers' unbooked equity stake 
inferior to that of ordinary shareholders in five ways:

  1.  Taxpayers cannot trade their positions away.

  2.  Downside liability is not contractually limited, but upside gain 
        is.

  3.  Taxpayer Positions carry no procedural or disclosure safeguards.

  4.  Taxpayer positions are not recognized legally as an ``equitable 
        interest.'' (This means TBTF firms may exploit them without 
        fear of lawsuits.)

  5.  TBTF Managers can and do abuse taxpayers by blocking or delaying 
        recovery and resolution.
The Problem of Regulatory Capture
    In and out of crisis, taxpayer interests are poorly represented by 
regulators because politicians and regulators have kept themselves less 
than fully accountable for the costs of bailouts and have 
simultaneously pursued conflicting political and bureaucratic goals. 
Over the years, the financial industry has infiltrated the bureaucratic 
system that ought to monitor and regulate aggressive risk-taking and 
woven huge loopholes into the fabric of capital requirements that--then 
and now--are supposed to keep financial instability in check. The 
industry's capture of the regulatory system is politically very well-
defended, because the subsidies are in part shifted forward to 
creditors and to customers in various industries (e.g., in realty and 
construction).
    Capture can be demonstrated in at least four complementary ways: 
(1) by enumerating the problems that the Dodd-Frank Act set aside (such 
as how to define systemic risk operationally or how to resolve the 
Fannie and Freddie mess); (2) by examining the many loose ends left in 
the Act's efforts to handle regulation-induced innovation (especially 
in swaps) and to deal with institutions that have made or are making 
themselves too large, too complex, and too well-connected politically 
and bureaucratically to be closed and unwound; (3) by noting that 
crisis-management policies have helped the largest BHCs to become even 
larger; and (4) by recognizing that postcrisis reforms continue to 
feature loophole-ridden measures of accounting capital as the 
cornerstone of financial-stability policy.
Why Capital Requirements Can't Adequately Protect Taxpayers From BHC 
        Shareholders
    Besides setting minimums that are far too low, gaping imperfections 
exist in weighing risks and measuring capital that open and solidify 
avoidance opportunities (see Admati and Hellwig, 2013). Actual and 
potential zombie institutions can use accounting tricks, organizational 
complexity, and innovative instruments to hide risk exposures and 
accumulate losses until their insolvency becomes so immense that they 
can panic regulators and command life support from them.
    The Basel control framework (see Basel Committee on Banking 
Supervision, 2013) is built on the fiction that all or most SIFIs can 
be persuaded to forgo individually profitable credit business for the 
greater good. This seems awfully naive (see Schelling, ``Strategy of 
Conflict''). The naivete lies in a set of unrealistic assumptions about 
the regulatory game: (1) that accounting ratios are difficult to 
misrepresent; (2) that supervisors are hard to mislead; (3) that 
bankers dutifully accept statutory burdens rather than work 
aggressively to adjust their risk profile to neutralize the net effect 
of capital restrictions on SIFI profits and market capitalization; and 
(4) that meritorious commitments to protect unsophisticated depositors 
and to keep systemically important markets and institutions from 
breaking down in difficult circumstances do not provide convenient 
cover that tempts officials to obligate taxpayer funds over-generously 
and without revealing the full picture of fiscal and incentive effects.
    Capital requirements are merely restraints. Improved capital 
requirements increase the difficulty of extracting TBTF subsidies, but 
they do not reduce the legitimacy of adopting strategies that willfully 
pursue this goal. To do this, I propose that Congress declare that 
taxpayers have an equitable interest in any institution that can be 
shown to extract a subsidy from the safety net. In common law, an 
``equitable interest'' is understood as a balance-sheet position that 
gives its owner a right to compensation from damages. I believe that we 
should conceive of this compensation as the dividend taxpayers would be 
paid on their implicit equity stake in any accounting period if 
information asymmetries did not exist. The net value of taxpayers' 
stake in a TBTF firm increases with the extent to which creditors and 
stockholders are confident that they can hide tail risks and, if 
ruinous losses emerge, scare authorities into funding the losses 
without extracting due compensation.
    Genuine reform would compel the DOJ to prosecute megabank holding 
companies that engaged in easy-to-document securities fraud. Numerous 
representations and warranties can be shown to be deliberately 
deceptive and designed to benefit individual firms at the expense of 
the rest of us. As legal persons and convicted felons, guilty BHCs 
could be forced to break themselves up. Subsidiaries of felonious 
companies could lose the right to take insured deposits or act as 
broker-dealer firms and futures merchants. The beauty of such penalties 
is that managements and not Governments would have to design the 
breakup plan.
    Living wills, enhanced resolution authority, clawbacks of 
undeserved executive compensation, and an Office of Financial Research 
are potentially useful tools. But the failure to prosecute any TBTF 
firm or top manager in open court for criminal securities fraud tells 
us how easy it is to collect fines (because they are paid by 
stockholders) and how hard it can be for regulators to discipline 
individual managers of influential and interconnected BHCs. For top 
management, corporate-level fines are a nondeterrent slap on the wrist. 
Moreover, only a portion of most fines compensate the taxpayer by 
flowing through to the Treasury. Sad to say, most of these criticisms 
apply to the reform programs that are unfolding in the European Union 
as well.
The Problem of Fairness
    Fairness is the heart of the Rule of Law. Whether or not enhanced 
resolution or contracts with bail-in provisions can be enforced in 
difficult circumstances, Corporate and/or Property Law needs: (1) to 
recognize that regulators' demonstrated propensity to bail out 
creditors and shareholders in TBTF firms (e.g., in AIG, Fannie, and 
Freddie) assigns taxpayers' a disadvantaged equity position in each 
TBTF firm, and (2) to enact personal and corporate penalties for 
willful efforts to pursue risks that abuse taxpayers' equity stake and 
pervert the pattern of real investment. Corporate penalties could 
include forced sales of some or all lines of business.
    It is useful to think of taxpayers' stake in each TBTF firm as if 
it were a trust fund and conceive of Government officials as 
fiduciaries responsible for managing that fund. The purpose of the 
reforms I propose is to give regulators, along with managers and 
directors of TBTF firms, an explicit and codified fiduciary duty to 
measure, disclose, and service taxpayers' stake-holding fairly. To 
overcome short-term benefits from ducking their implicit fiduciary 
responsibilities, regulators, managers, and board members need to face 
stricter legal liability for neglecting or incompetently performing 
these fiduciary duties.
    Governments must rework bureaucratic and private incentives to 
focus reporting responsibilities for regulators and institutions on 
uncovering the value of safety-net support. Regulatory-agency and 
corporate mission statements must explicitly define, embrace, and 
enforce fiduciary duties of loyalty, competence and care to taxpayers 
in operational and accountable ways. Otherwise, it is unreasonable to 
hope that managers will--or that regulatory staff can--contain systemic 
risk during future rounds of boom and bust.
    The report the GAO released today (General Accountability Office, 
2014) is a small step in this direction. The downside of the report is 
that TBTF firms are going to trumpet GAO's low-ball and conceptually 
deficient measurement of the subsidy as if it were gospel.
    To support a culture of fiduciary duty, I have long maintained that 
we need to strengthen training and recruitment procedures for high-
ranking regulators. If it were up to me, I would establish the 
equivalent of a military academy for financial regulators and train 
cadets from around the world. The curriculum would not just teach 
cadets how to calculate, aggregate, and monitor the costs of safety-net 
support in individual institutions and countries. The core of the 
curriculum would be to drill students in the duties they will owe the 
citizenry and to instruct them in how to confront and overcome the 
nasty political pressures that elite institutions exert when and as 
they become increasingly undercapitalized.
    Politically, a financial crisis is a struggle by financial firms 
whose assets have collapsed in value to offload the bulk of their 
losses onto creditors, customers, and taxpayers. In the early months of 
the 2008 crisis, Fed and Treasury officials assisted economically 
insolvent zombie institutions (such as Bear Stearns and AIG) to book 
new risks and to transfer their losses onto the Government's balance 
sheet. Authorities did this by mischaracterizing the causes of these 
institutions' distress as a shortage of market liquidity and helping 
insolvent firms to expand and roll over their otherwise unattractive 
debt. Far from assisting zombie institutions to address their 
insolvency, unwisely targeted and inadequately monitored Government 
credit support encouraged troubled firms not only to hold, but even to 
redouble the kinds of go-for-broke gambles that pushed them into 
insolvency in the first place.
    Indiscriminately bailing out giant firms was a mistake that has 
hampered, rather than promoted economic recovery. Similarly, prolonged 
uncertainty about the future of Fannie and Freddie continues to disrupt 
housing-finance activity. Blanket bailouts evoke gambles for 
resurrection among zombie and near-zombie beneficiary firms like AIG, 
while uncertainty about who will finally bear the extravagant costs of 
these programs dampens spending plans in every sector. These problems 
divert and restrain the flows of credit and real investment necessary 
to trigger and sustain a healthy economic recovery.
References
Admati, A., and Hellwig, M. (2013). ``The Bankers' New Clothes: What's 
    Wrong With Banking and What To Do About It?'', Princeton University 
    Press, Princeton.
Bair, S. (2012). ``Bull by the Horns: Fighting To Save Main Street From 
    Wall Street and Wall Street From Itself'', Free Press, New York.
Basel Committee on Banking Supervision (2014). ``Supervisory Guidelines 
    for Identifying and Dealing With Weak Banks: Consultative 
    Document'', Banking for International Settlements, Basel.
Ferguson, T., Jorgenson, P., and Chen, J., (2013). ``Party Competition 
    and Industrial Structure in the 2012 Elections: Who's Really 
    Driving the Taxi to the Dark Side?'', International Journal of 
    Political Economy, 42 (Summer), 3-41.
Hovakimian, A., Kane, E.J., and Laeven, L.A. (2012). ``Variation in 
    Systemic Risk at U.S. Banks During 1974-2010'', available at SSRN: 
    http://ssrn.com/abstract=2031798 or http://dx.doi.org/10.2139/
    ssrn.2031798 (May 29).
United States Government Accountability Office, July 2014. ``Bank 
    Holding Companies: Expectations of Government Support'', (GAO-14-
    621).
    
    
    
    
                   PREPARED STATEMENT OF ANAT ADMATI
George G.C. Parker Professor of Finance and Economics, Graduate School 
                    of Business, Stanford University
                             July 31, 2014
    Chairman Brown, Ranking Member Toomey, and Members of this 
Subcommittee, I commend you for holding this hearing and am grateful 
for the opportunity to speak to you. I am a Professor of Finance and 
Economics at Stanford Graduate School of Business and my recent 
research and writings have focused on issues immediately relevant to 
today's hearing.
    Recent experiences have helped foster the expectations of 
Government support mentioned in the title of this hearing. Since 2008, 
the Treasury, the Federal Reserve and the FDIC provided through various 
programs massive and unprecedented support to the financial system. The 
largest bank holding companies, to varying degrees, have had access to 
hundreds of billions, even trillions of dollars in relatively cheap 
loans and guarantees, and they benefited from bailouts of their 
counterparties such as AIG. For some, e.g., Citigroup, the support was 
critical. \1\
---------------------------------------------------------------------------
     \1\ The banks and the Federal Reserve tried to keep information 
about the extent of Fed loans hidden. The information was released 
after Bloomberg fought in court. See Phil Kuntz and Bob Ivry, ``Fed 
Once-Secret Loan Crisis Data Compiled by Bloomberg Released to 
Public'', Bloomberg, December 22, 2011. Citigroup is discussed further 
below.
---------------------------------------------------------------------------
    Trillions of U.S. taxpayer funds were put at risk. The supports 
prevented the collapse of the system and helped many financial 
institutions avoid default, bankruptcy, or resolution in which their 
shareholders would be wiped out and at least some of their creditors 
would suffer losses. Yet, the programs did little to solve the housing 
crisis, failed to improve business lending meaningfully, and at times 
were excessively generous and inefficient. \2\
---------------------------------------------------------------------------
     \2\ Cole (2012) shows that TARP did not help improve business 
lending, which is not surprising since the programs did not reduce the 
institutions' indebtedness and the resulting debt overhangs (see Admati 
and Hellwig, 2013a, chapter 3, and Admati et al., 2014). Barofsky 
(2012) and Bair (2012) describe the bailouts programs. Additional 
references in the notes to chapter 9 of Admati and Hellwig (2013a), 
whose text is attached to this testimony.
---------------------------------------------------------------------------
    Implicit guarantees for which banks do not pay create a subsidy, 
essentially free insurance for their debts, or at least a partial 
insurance that lowers the likelihood of losses in some scenarios. 
Because such subsidies are implicit and invisible, determining their 
value with any precision is difficult; there is no market in which the 
implicit guarantees are being valued (although some have tried to use 
credit insurance contracts to try to estimate their value). Any 
estimate depends on many variables that change over time, and 
estimation requires making many assumptions; such assumptions might or 
might not be true in reality. In fact, many of the variables that 
affect the size of the subsidy vary across different institution in 
complex ways. Moreover, actions by the institutions, by investors, and 
by regulators also have important impacts. Later in this document I 
will have additional comments on measuring the subsidies.
    When implicit-guarantee subsidies are provided to institutions that 
have significant discretion about their investments and the risks they 
take, the results can be perverse. Policymakers may hope that the 
subsidies are passed on to specific investments or people, but the 
institutions, as they benefit from the guarantees, may well have 
incentives to make different investments altogether.
    For example, guarantees may be provided in the hope that the banks 
will make certain loans, when in fact, given their compensation 
structures and the flawed regulations we have in place (e.g., the use 
of risk weights), the banks may only make the loan if it is very safe 
or if it is guaranteed by the Government. Instead, banks may prefer to 
invest in derivatives markets with more upside.
    The institutions benefiting most from the subsidies often deny the 
existence of any benefit and claim that they are happy to give up the 
implicit subsidies. ``Please,'' they may say, ``let banks fail when 
they should.'' \3\ The difficulty is that letting systemic 
institutions--or many institutions at the same time--fail may be 
disruptive and entail enormous collateral damage. Even if the direct 
costs of the failure are covered, the disruptions and inefficiencies 
that result are costly for the economy and the harm is borne by 
innocent citizens. As I will explain below, we do not have workable 
options for the failure of systemic institutions; moreover, the harm 
from their distress and even from the fear of their failure creates 
instabilities.
---------------------------------------------------------------------------
     \3\ For example, in his letter to shareholders in April 2011, 
Jamie Dimon, CEO of JPMorgan Chase, denies his bank benefits from 
implicit subsidies and suggested that the industry pay any expenses 
associated with the failure of ``dumb banks.'' For a response to this 
letter, see Anat Admati, ``An Open Letter to JPMorgan Chase Board of 
Directors'', reprinted in Huffington Post, June 14, 2011). This letter, 
which was sent to at least one person within the bank, did not receive 
any acknowledgment and did not appear to affect the banks' strategy. 
Misleading comments by bank executives and bank lobbyists as well as 
others are discussed in Admati and Hellwig (2013a) and in a number of 
short pieces, some of which are cited later in this document.
---------------------------------------------------------------------------
    Financial crises are sometimes portrayed as if they were 
unpreventable natural disasters, implying that bailouts are similar to 
emergency aid after an earthquake. This narrative is misleading. The 
crisis of 2007-2009 was an implosion of a system that had become too 
fragile, reckless, and distorted. Regulatory failures, including flawed 
and ineffectively enforced regulations, must take much of the blame for 
the excessive fragility and the buildup of risk. These failures can be 
corrected, and regulators have authority to do so under current laws, 
but, remarkably, obvious lessons have not been learned, and not enough 
has been done to make the system as safe as it can and should be. Some 
counterproductive laws have also remained in place. \4\
---------------------------------------------------------------------------
     \4\ I am referring, for example, to the distortive corporate tax 
code that penalizes equity funding and encourages borrowing, which can 
become excessive, and to the sweeping exemptions of repos and 
derivatives from stay in bankruptcy, which has likely enabled and 
encourage excessive growth in these markets. These issues will come up 
briefly below and they are discussed in Admati and Hellwig (2013a, 
chapters 5, 9, and 10).
---------------------------------------------------------------------------
    The situation in banking is disturbingly similar to allowing heavy 
trucks with dangerous cargo to drive recklessly at 95 miles per hour in 
residential neighborhoods. If drivers get a bonus for reaching the 
destination quickly, and face little risk of injury or death even in an 
explosion (imagine that they have a special protective mechanism), they 
will drive recklessly and endanger innocent citizens. Authorities can 
send firefighters to put out fires and medics to treat the injured if 
an explosion occurs, but the public would wonder why truck companies 
reward reckless driving and, most importantly, why a safer speed limit 
was not set and enforced to prevent harm.
    Similar questions must be asked about the failure of financial 
regulation. We should have a financial system that supports the economy 
as efficiently and consistently as possible without major distortions. 
The system we have instead is too dangerous, exposing the public to 
unnecessary risk and distorting the economy. Much can be done--even 
within existing laws--to improve this situation.
    This Committee has an important role in helping bring about 
beneficial changes. In the rest of this document, I will elaborate on 
the above statements, diagnose the key problems, and outline some 
recommendations. Additional materials are attached and referenced; I 
will be happy to provide more at your request.
Can/Will Large Bank Holding Companies Ever ``Fail'' and if so, How?
    The Dodd-Frank Act (DFA) intended, among other things, to eliminate 
bailouts. Yet virtually everyone involved in the financial system--even 
if some would not admit it--expects that the Government, possibly 
through the Federal Reserve and FDIC, will again provide supports to 
large bank holding companies and other institutions considered 
``systemic'' if authorities fear that the failure of these institutions 
would cause significant harm to the economy. If many small institutions 
become distressed at the same time, they too may be supported.
    This assessment is based on the realities of today's system and the 
state of its regulation. \5\ Whereas regulators receive significant 
authority under DFA (some of which they had all along), the 
implementation of the law has been messy and uneven. Some of the most 
critical rules are insufficient and flawed; others appear wasteful, too 
costly relative to the benefit they provide.
---------------------------------------------------------------------------
     \5\ The dynamics of contagion are explained in Admati and Hellwig 
(2013a, chapter 5), White (2014), and testimony of James Thomson before 
this Subcommittee on July 16, 2014.
---------------------------------------------------------------------------
    Policymakers who were involved in the bailouts extol the virtues of 
their actions while appearing willfully blind to their failure to 
reduce the fragility of the system before the crisis and to learn the 
lessons since. If anything, investors may reasonably expect that 
supports would be forthcoming for fear of another ``Lehman moment'' 
even with the alternative to bankruptcy offered through the new and 
still untested resolution authority by FDIC.
    The DFA titles most relevant for this discussion are Titles 1, 2, 
and 7. I'd like to focus my discussion mainly on Titles 1 and 2, 
although Title 7, which deals with derivatives markets, is also 
critical. The still-too-opaque markets in derivatives allow banks to 
hide enormous amount of risk from investors and regulators. Ineffective 
implementation of Title 7 and poor disclosures can undermine Titles 1 
and 2 and the objective of having a healthier financial system.
    Stating the obvious (but see more below for nuances), a business 
``fails'' when it does not fulfill its debt commitments or is feared to 
be unable to pay the debts. For ``normal'' companies in the U.S., 
failure involves filing for bankruptcy or liquidation under Chapter 11 
or Chapter 7.
    Title 1 of DFA requires, among other things, that large bank 
holding companies submit ``living wills'' to regulators. These 
documents are meant to play out a scenario in which the holding company 
goes through bankruptcy process, presumably under Chapter 11. In her 
testimony before your Committee on July 15, 2014, Fed Chair Janet 
Yellen was asked some pointed questions about the living-wills process 
by Senator Elizabeth Warren. The exchange brings out some key issues. 
According to Chair Yellen, the largest bank holding companies have by 
now submitted three rounds of living-wills documents, and received 
feedback on the first set of submissions. The parts of these documents 
that are made public provide little information, often less than is 
included in standard financial statements. The full submission, 
according to Chair Yellen, goes into tens of thousands of pages.
    Senator Warren asked Chair Yellen a critical question: ``Can you 
honestly say that JPMorgan could be resolved in a rapid and orderly 
fashion as described in its plans with no threats to the economy and no 
need for a taxpayer bailout?'' The Senator pointed out that JPMorgan 
Chase has $2.5 trillion in assets and 3,391 subsidiaries, compared to 
Lehmann Brothers, which had $639 billion in assets and 209 subsidiaries 
prior to its failure.
    The Lehman Brothers bankruptcy, filed on September 15, 2008, caused 
severe disruption and damage to the global financial system. In its 
immediate aftermath, stock prices imploded, investors withdrew from 
money market funds, money market funds refused to renew their loans to 
banks, and banks stopped lending to each other. Banks furiously tried 
to sell assets, which further depressed prices. Within 2 weeks, many 
banks faced the prospect of default.
    To prevent a complete meltdown of the system, Governments and 
central banks all over the world provided massive supports to financial 
institutions. These interventions stopped the decline, but the downturn 
in economic activity was still the sharpest since the Great Depression. 
Anton Valukas, the lawyer appointed by the bankruptcy court to 
investigate Lehman Brothers, put it succinctly: ``Everybody got hurt. 
The entire economy has suffered from the fall of Lehman Brothers . . . 
the whole world.'' Within 21 months, American households lost $17 
trillion; reported unemployment hit 10.1 percent at its peak in 2009. 
\6\
---------------------------------------------------------------------------
     \6\ These two paragraphs are adapted from Admati and Hellwig 
(2013a, p. 11), and the crisis is described in some detail in chapters 
5 and 9 (the latter is attached to this testimony). Mr. Valukas made 
the statement here quoted in an interview on CBS 60 Minutes, aired 
April 22, 2012. The last fact is included in the 2011 report of the 
Financial Crisis Inquiry Commission (p. 390).
---------------------------------------------------------------------------
    Chair Yellen stated that Title 1 of DFA only requires the Fed to 
give feedback to the companies about their plans. She referred to an 
``iterative process'' of submission and feedback. Title 1 apparently 
does not require that regulators give a pass/fail grade to the living 
wills nor to determine definitively whether bankruptcy is a viable 
option. However, the title definitely authorizes regulators to take a 
number of strong actions if they find that bankruptcy would entail too 
much collateral damage. Such actions include increasing capital 
(equity) requirements, requiring that structures be simplified and 
assets sold (potentially ``breaking up'' the banks), etc.
    The U.S. bankruptcy code to which Lehman Brothers was subjected has 
not changed since 2008. Other countries have different processes, which 
Lehman Brothers' foreign subsidiaries must follow. The tens of 
thousands of pages of living wills JPMorgan Chase has submitted to 
regulators might be of some use should it file for bankruptcy, at least 
under U.S. law (although they may well be dated by the relevant time, 
because banks' counterparties and businesses can change in a matter of 
days or months). But the process will not be much faster and simpler 
than Lehman Brothers bankruptcy. Moreover, should the numerous 
counterparties of JPMorgan Chase become concerned that bankruptcy might 
be forthcoming, runs and disruptions similar to those observed in 2007-
08 when Bear Stearns and Lehman Brothers became distressed will likely 
start significantly before any filing.
    It defies credibility to suggest that, at the current speed of the 
``iterative process'' that Chair Yellen described regarding the living 
wills, and without major changes to their structure and funding mix, 
enormously large and complex institutions like JPMorgan Chase will be 
able to go through bankruptcy without major harmful effects. Yet, 
regulators may continue to ``iterate'' and fail to use their authority 
to act even knowing that bankruptcy is not viable, refusing to admit to 
and deal with this reality. I doubt this situation was the intent of 
Title 1.
    DFA authors, perhaps mindful after the Lehman Brothers experience 
that bankruptcy may not be a realistic option for large financial 
institutions, included an alternative mechanism in Title 2, which gives 
the FDIC authority to deal with the failure of any institution deemed 
``systemic'' through a so-called Orderly Liquidation Authority (which 
actually doesn't intend to liquidate the company). The FDIC has engaged 
in the last few years in a serious effort to make its plans for this 
process credible, focusing on an approach called Single Point of Entry 
(SPOE).
    SPOE represents an important and useful development, but, as 
bankruptcy expert David Skeel (2014, p. 3) assesses, ``the technique 
also has important vulnerabilities, and some of the claims made on its 
behalf are quite exaggerated.'' Among them, SPOE does not work for 
institutions that are active globally and that have systemically 
important operations in several countries, unless all the countries 
that are involved agree to such an approach. A recent coordination 
effort between U.S. and UK may allow for SPOE of U.S. authorities in 
U.S. holdings companies without intervention of UK authorities in UK 
subsidiaries, so the problem of UK authorities entering a Lehman 
Brothers subsidiary and finding that there is no cash to keep 
systemically important functions going might not arise.
    However, the U.S.-UK coordination is the only attempt of this sort, 
and it does not seem to be fully symmetric. If Barclays or Deutsche 
Bank were to run into trouble, U.S. authorities would probably not be 
willing to accept SPOE resolution by the domestic authorities of these 
banks, but instead would intervene directly in the holdings companies 
that organize these banks' U.S. activities. Multiple-entry resolution, 
however, destroys operational procedures that have been managed in 
integrated fashion across jurisdictions, for cash management, as in the 
case of Lehman Brothers, or, even more importantly, the joint use of 
Information Technology systems.
    From the perspective of the different countries involved, single-
entry resolution would involve significant conflicts of interest. If 
U.S. authorities had been in charge of Lehman Brothers, London, as well 
as the parent, would they have paid proper attention to London-specific 
concerns, including the systemically important market-making activities 
of Lehman Brothers in London? Alternatively, is it acceptable for U.S. 
authorities to follow the procedure suggested in the living will of 
Deutsche Bank, which argues that damage from resolution would be 
minimized if U.S. authorities were willing to trust the German 
authorities (Bafin, the supervisor, and FMSA, the resolution 
authority)? In a cross-border setting, SPOE resolution leaves too much 
room for the authority in charge to shift losses to other countries and 
it is therefore hardly workable. \7\
---------------------------------------------------------------------------
     \7\ Even the Nordic countries have not been able to agree on an 
SPOE procedure for Nordea.
---------------------------------------------------------------------------
    Even if we had SPOE resolution for globally systemically important 
banks, some of these banks would most likely be ``too big to fail.'' 
Procedures would be lengthy and cumbersome and, meanwhile, there might 
be substantial systemic fallout. Regulators would then be reluctant to 
use the procedure if multiple financial institutions face default at 
the same time, or if resolution would expose problems at one or more 
subsidiaries. In sum, Title 2 is useful, but it is certainly not a 
silver bullet for addressing the ``too-big-to-fail'' problem and it 
does not eliminate expectations of support for large bank holding 
companies. Moreover and importantly, even under the best scenario, 
using Title 2 resolution would be costly and entail collateral damage 
and, as in the case of bankruptcy, the distress of the corporation, and 
the fear or anticipation that Title 2 resolution might be invoked by 
its counterparties would likely already cause harm. \8\
---------------------------------------------------------------------------
     \8\ See also White (2014) for a discussion of the issues regarding 
``fail'' scenarios in ``too big to fail.''
---------------------------------------------------------------------------
    The living wills requirements and Title 2 of DFA try to make 
palatable the notion that, like other companies, financial institutions 
structured as limited liability corporations should fail if they take 
risk and become unable to pay their debts, thus wiping out their 
shareholders and imposing losses on their creditors through an orderly 
legal process. In a vibrant market economy, innovations involve risk, 
and failures should be tolerated.
    For normal companies, bankruptcy typically follows an actual or 
imminent default. Restructuring debts may allow the company to continue 
operating. Bankruptcy laws try to control the actions of managers and 
shareholders in insolvent companies, who have incentive to benefit 
themselves at the expense of creditors by taking out cash or gambling 
for survival. Since such problems and the legal and other costs of 
bankruptcy are anticipated by creditors, the terms of the debt claims, 
including both the interest rate and the conditions the contract puts 
on the borrower, are set by prudent lenders to compensate for the 
losses in the event of default and bankruptcy, and to control 
borrowers' actions that go against the lenders' interests.
    A source of great inefficiency in banking is that banking 
institutions can persist in a state of distress or even insolvency 
without their creditors becoming alarmed and without the institution 
experiencing the difficulty of most distressed borrowers to raise funds 
and continue operating. One reason for this anomaly is that banks' 
creditors include depositors, who are insured and dispersed. Depositors 
are particularly passive in their role as lenders to the banks (a 
status most of them do not quite realize they have) and do not behave 
as normal creditors with standard debt contracts. Depositors rely on 
insurance and regulators to protect them.
    Banks can use depositors' funds to invest in various loans and 
other assets that can sometimes be used as collateral and enable the 
bank to borrow even more under attractive terms. Creditors whose debts 
are secured by collateral care less than unsecured creditors about the 
borrower's solvency. Lending to financial institutions through so-
called repurchase agreements (repos) is even safer than secured 
lending, because, under safe harbor laws from 2005, repos, as well as 
derivatives, are exempted from the normal stay in bankruptcy. \9\
---------------------------------------------------------------------------
     \9\ Skeel and Jackson (2012), and Mark Roe, another bankruptcy 
expert, (see, e.g., ``Reforming Repo Rules'', Project Syndicate, 
December 21, 2011), call for reexamining these exemptions. Skeel (2014) 
also warns with regard to Title 2 resolution that ``it reinforces 
problematic incentives for financial institutions to rely on short-term 
financing.''
---------------------------------------------------------------------------
    For bank holding companies considered too big to fail, even 
unsecured bond holders feel reasonably sure they will be paid in full. 
In the financial crisis the creditors of numerous banking institutions, 
including those whose claims had counted as ``regulatory capital'' and 
were meant to absorb losses, were paid in full even as the institutions 
received large amounts of bailout funds and other supports. As 
discussed above, even today, and despite DFA, it is quite possible and 
even likely that the creditors of one of the largest bank holding 
companies will be paid in full even if the institution is insolvent.
    As long as creditors are paid and do not constrain the borrowing 
bank much, it can continue operating. In that case, only regulators are 
in a position to intervene even as highly distressed or insolvent 
borrowers, including banks, are extremely inefficient and their 
decisions are distorted by conflicts of interest with creditor. In 
fact, I will argue below that by most standards, the banks are 
permanently in a state of financial distress, yet they manage to get 
away with it.
Essential, Yet Flawed and Insufficient Regulation
    In addition to the living-wills requirement, Title 1 of DFA 
authorizes the Federal Reserve, in collaboration with other regulators, 
to design prudential regulations meant to maintain the safety and 
soundness of the system. The Fed is charged with regulating bank 
holding companies as well as all institutions declared systemic by the 
Financial Stability Oversight Committee.
    As discussed above, the scenarios that involve default and failure 
of systemic institutions are complicated, disruptive, and harmful. 
There are no good options. It thus appears particularly important to 
try to prevent reaching these failure situations through prudent 
supervision and regulations. Most important among those safety measures 
are capital requirements meant to control the funding mix of these 
companies, including to ensure that they fund their investments by 
appropriate amount equity--money from owners and shareholders--so that 
they can continue making loans and investments and still pay their 
debts even if they incur losses. (Note: the jargon that refers to 
capital as something banks ``hold'' or ``set aside'' is confusing, 
suggesting that capital represents idle funds like cash reserves that 
banks cannot use, which is false. \10\)
---------------------------------------------------------------------------
     \10\ On this insidious confusion, see Admati and Hellwig (2013a, 
chapters 1 and 6), Admati et al. (2013, section 3.1), Claims 1 and 2 
Admati and Hellwig (2014), which is attached to this testimony, and my 
Tedx Stanford talk http://www.youtube.com/
watch?v=s_I4vx7gHPQ&feature=youtu.be&a.
---------------------------------------------------------------------------
    According to its financial statements, on December 31, 2007, the 
largest bank holding company at the time, Citigroup, reported that its 
shareholder equity or net worth (the difference between its reported 
assets and liabilities) was 5.2 percent of its total assets. 
Citigroup's assets were valued at almost $2.2 trillion. As Lawrence 
White from New York University Stern School notes, however, this 
information does not capture some important facts. He writes (White, 
2014, p. 7, footnotes omitted): ``Citigroup is best understood as a 
(roughly) $1.2 trillion depository institution, on top of which was a 
(roughly) $1 trillion holding company (including its nondepository 
subsidiaries). The holding company's net worth was smaller than the 
depository's net worth; in essence, if the net worth of the depository 
(i.e., the capital of the depository, which also counted as an asset 
for the holding company) was ignored, the holding company was 
insolvent.''
    Citigroup proceeded to collapse at the end of 2008 and needed a 
series of bailouts and massive other supports. Remarkably, the 
Government injected of $25 billion of TARP funds into Citigroup on 
October, 8, 2008, and, even with the market value Citigroup stock 
falling below $25 billion in November, the company was offered tens of 
billions in additional bailouts and hundreds of billions in cheap loans 
and guarantees from the Fed. (Citigroup, according to Arthur Wilmarth 
from George Washington University Law School is ``a case study in 
managerial and regulatory failure.'' \11\)
---------------------------------------------------------------------------
     \11\ See Wilmarth (2014). Bair (2012) and Barofsky (2012) include 
vivid descriptions of the bailouts.
---------------------------------------------------------------------------
    Indeed, regulators often show forbearance and allow insolvent banks 
to persist and even hide their losses. Insolvent institutions are 
highly dysfunctional and harm the economy. They do not make new loans 
and may become reckless, gambling for survival or looting the 
institutions. Recklessness was pervasive in the Savings and Loan Crisis 
of the 1980, and the dysfunctionality of weak banks is evident in 
Europe in recent years. Yet, when banks are supported, their 
indebtedness is often maintained because the supports are given in the 
form of more loans. \12\ Solvent corporations can in fact raise equity 
at some price, although their managers and shareholders are unlikely to 
do so voluntarily. Creditors or regulators can bring about reduction of 
indebtedness through covenants or regulation. \13\
---------------------------------------------------------------------------
     \12\ Onaran (2011) argues that both Citigroup and Bank of America 
were insolvent or ``zombies'' even in 2010. Admati and Hellwig (2013a, 
chapters 3, 4, and 11) emphasize the harm of allowing weak banks to 
persist.
     \13\ Admati et al. (2014) discusses in detail how borrower-
creditor conflicts affect funding decisions in highly indebted 
corporations, and the analysis is particularly applicable to banks.
---------------------------------------------------------------------------
    A glaring failure of regulatory reform efforts across the globe 
(not just in the U.S., indeed, the situation is worse in Europe) is 
that, even as the largest global financial institutions have grown ever 
bigger, more complex, more connected and more dangerous, they continue 
to be allowed to operate with dangerously high levels of indebtedness 
and much too little equity, and to hide too much risk in opaque markets 
and off their balance sheets.
    The minimal requirements agreed upon in Basel III allow equity to 
be as low as 3 percent of the total assets. Even with the harsher U.S. 
requirements, 95 percent of the total assets of the largest bank 
holding companies can be funded with debt. Note that this requirement 
would have been satisfied by Citigroup in December, 2007. Capital 
regulations also rely on an enormously complex and manipulable system 
of risk weights that distorts banks' decisions and exacerbates the 
fragility of the system, among other things making business lending 
relatively unattractive.
    Bankers and regulators claim that the new capital regulations are 
tough when in fact these reforms amount to a tweak and they have no 
valid justification. In the speeding analogy, the reforms are analogous 
to reducing the speed limit for loaded trucks from 90 miles per hour to 
85 miles per hour in residential neighborhoods, with police unable to 
measure the actual speed. The claims made to justify the regulation or 
to fight higher equity requirements are fraught with flaws that range 
from false statements to misleading claims that divert the discussion. 
These statements are discussed in details in many of my writings, with 
colleagues, over the last 4 years; a small sample of which is attached 
to this testimony. \14\
---------------------------------------------------------------------------
     \14\ See Admati (2014), Admati and Hellwig (2013a, 2014), and 
Admati et al. (2013, 2014). Admati et al. (2013) was first posted in 
August, 2010. These and additional references are available at http://
bankersnewclothes.com/: and (for more academic writing) http://
www.gsb.stanford.edu/news/research/admati.etal.html.
---------------------------------------------------------------------------
    A key observation for understanding corporate funding decisions is 
that heavy borrowing creates strong conflicts of interest between 
borrowers and lenders and potentially distorts the investments and 
funding decisions made by borrowers once debt is in place. Overhanging 
debts create inefficiencies when borrowers--or managers in an indebted 
corporation acting in the interests of shareholders--make decisions in 
their own interest and do not take into account the impact of their 
actions on creditors or third parties. For example, borrowers may 
underinvest in worthy projects if they expect the returns to accrue in 
part to their creditors or they may make excessively risky investments 
if they expect the downside of the risks to be borne by creditors, or 
by deposit insurance institutions and taxpayers. \15\
---------------------------------------------------------------------------
     \15\ As discussed in Admati and Hellwig (2013a, chapter 3), the 
effects of overhanging debt can be seen in the case of homeowners who 
would not invest in the house if its value is low relative to the 
mortgage, or who might take a second mortgage even as this may put the 
lender of their first mortgage at risk.
---------------------------------------------------------------------------
    As a result of these distortions and other costs associated with 
distress or bankruptcy, heavy borrowing can actually reduce the total 
value of a firm (i.e., the sum of the values of all claims, including 
debt and equity). Borrower-creditor conflicts also create an 
``addiction'' to debt on the part of heavy borrowers, biasing 
subsequent funding decisions towards more debt and away from equity 
that makes existing creditors safer. \16\ As mentioned above, the 
conflicts are particularly intense when corporations are in a state of 
distress or insolvency, which for most corporations are rare but which 
in fact are considered normal in banking.
---------------------------------------------------------------------------
     \16\ This phenomenon is explores in details in Admati et al. 
(2014), which is highly relevant to understanding the rationale for 
leverage regulation. See also Admati et al. (2013) and Admati (2014).
---------------------------------------------------------------------------
    Without any regulation of their funding, and despite a (distortive) 
tax code that subsidizes borrowing and penalizes the use of equity, 
most corporations do not borrow heavily. \17\ Even those who tend to 
use more debt, including private equity firms or Real Estate Investment 
Trusts, rarely have less than 30 percent equity in their funding mix. 
As discussed above, prudent creditors write restrictive covenants that 
constrain dividend payouts and other decisions by the borrower, and 
adjust the cost of borrowing to reflect anticipated legal costs and 
delays should the borrower go into bankruptcy, as well as the 
possibility that the borrower would take additional debt that might 
dilute their claims.
---------------------------------------------------------------------------
     \17\ White (2014) provides some comparisons based on book value of 
equity. The comparisons of banks and nonbanks on the basis of market 
value are starker. The latter have on average 60 percent or more equity 
relative to total assets.
---------------------------------------------------------------------------
    Banks, however, can persist in distress because they do not 
experience the ``dark side of borrowing,'' including the increased 
costs and harsh terms that naturally prevent other corporations from 
heavy borrowing. Although they use a lot of debt, much of this debt 
comes with fewer strings attached than those other borrowers face (and, 
indeed, the terms the banks often place those to whom they lend). 
Deposit insurance and implicit guarantees lighten the burden of debt, 
allowing banks to continue to borrow and take risks without much effect 
on the terms of their debts. Supports and guarantees enable, encourage, 
and feed this addiction to debt. \18\
---------------------------------------------------------------------------
     \18\ Some claim that debt disciplines managers. In banking, this 
idea is a myth, as discussed in Admati et al. (2013, section 5), Admati 
and Hellwig (2013b) and Admati and Hellwig (2014, Claim 22), attached.
---------------------------------------------------------------------------
    Guarantees can also exacerbate the inefficiencies and distortions 
in banks' investment decisions. If you could use borrowed money in a 
casino, keep the winnings and continue to borrow when you lose, you 
would certainly love gambling even if the odds were significantly 
against you. Chapter 9 of Admati and Hellwig (2013a), whose text is 
attached to this testimony, provides an accessible explanation.
    The fact that banks choose to rely so much on debt does not mean 
that their indebtedness levels are essential or efficient. These levels 
are the result of a failure of internal governance and a failure of 
normal credit markets to constrain the love of borrowing by banks and 
bankers. Compensation structures that reward return on equity (ROE), 
which are pervasive in banking, effectively pay bankers to gamble at 
the expense of creditors or taxpayers who are exposed to greater risks. 
Even shareholders may be exposed to risks for which they are not 
properly compensated. \19\ Few benefit while the rest are harmed by 
this situation. When markets fail, effective laws and regulations must 
correct the distortions. Otherwise laissez faire can become crony 
capitalism.
---------------------------------------------------------------------------
     \19\ This is explained in detail in chapter 8 of Admati and 
Hellwig (2013a) and in many other writings. See Claim 8 in Admati and 
Hellwig (2014), attached.
---------------------------------------------------------------------------
    The idea of finding ways for banks to fail, discussed above, is 
obviously meant to bring back market discipline into banking. However, 
given the collateral damage from the failure of one or more 
institutions, and the fact that disruptions and harm start even before 
an actual default, the primary focus should be on prevention. Much more 
can be done on this front. There is simply no justification for the 
current inefficient levels of indebtedness in banking. Reducing it will 
achieve major benefits for society at virtually no relevant costs.
    The inefficiencies of heavy borrowing in banking also distort the 
provision of credit in the economy. Making loans is a critical 
contribution banks can make to the economy. \20\ Heavily indebted 
banks, however, may make too few worthy (but relatively ``boring'') 
business loans that don't have much upside, while at the same time 
making too many risky loans, including credit card loans, which may 
lead others to borrow too much and suffer the consequences. The 
distortions create cycles of booms, busts and crises. Regulations based 
on risk weights exacerbate these distortions.
---------------------------------------------------------------------------
     \20\ Despite the emphasis often placed on banks as sources of 
credit for firms, lending is actually a small part of what the largest 
bank holding companies do (see Admati and Hellwig (2013a, chapter 6). 
On the evolution of business of banking in the U.S., see Omarova 
(2013).
---------------------------------------------------------------------------
    It is possible and highly beneficial to transition to a system in 
which banks use significantly more equity, thus reducing the likelihood 
of costly failures or bailouts and at the same time permitting banks to 
invests more efficiently on behalf of all its investors, thus 
supporting the economy better and with fewer distortions.
    Whereas many extol the importance of increasing equity 
requirements, the status quo seems to be the benchmark against which 
changes are measured. This benchmark is entirely inappropriate. Banks 
are as fragile as they are only because those who make decisions in the 
banks benefit from the status quo and they have so far gotten away with 
maintaining it, even after the most recent crisis.
    Requiring that banks use more equity is not a silver bullet, and 
much depends on the details of the regulations and its implementation 
and enforcement, but effective regulation of banks' indebtedness can 
make other, more costly, regulations less important or necessary. \21\ 
Liquidity breakdowns are less likely if banks can trust each other to 
be solvent, and the liquidity offered by deposits and other short term 
debt by banks would only be enhanced if banks have more equity. \22\
---------------------------------------------------------------------------
     \21\ In chapter 11 of Admati and Hellwig (2013a) we outline 
briefly how better regulation can be designed and how to transition to 
a better system.
     \22\ These issues are discussed in detail in Admati and Hellwig 
(2013a, chapter 10).
---------------------------------------------------------------------------
    Existing laws still allow regulators to revise capital regulation. 
Title 1, specifically in the context of the living wills requirements, 
allows significant increases in equity requirements for institutions 
deemed systemic, if regulators admit that bankruptcy is not a viable 
option.
Comments on Measuring the Value of the Implicit Subsidies
    As discussed at the start of this document, it is very difficult to 
measure the value of the implicit subsidy associated with guarantees. 
Because there are no markets for these guarantees, assumptions must be 
made about the underlying forces and the data being used. One can also 
try to focus on the cost to taxpayers or in terms of benefits to 
banking institutions who receive the subsidies. In fact, these two need 
not be the same because of the collateral impact of the banks' choices 
of investment and funding, and especially of their distress and failure 
scenarios.
    In assessing the costs to taxpayers, it is important to realize 
that expenses for supporting financial institutions in a systemic 
crisis occur at the every moment when the macro-economy is doing 
poorly, the country's fiscal situation is very tight and money is 
sorely needed in many places. \23\ Similarly, in assessing the benefits 
to banks, it is important to realize that Government guarantees are 
most useful in times of crisis, when private protection schemes are 
breaking down and the very survival of the institution is at stake. As 
discussed above, banks' decisions about lending and investments are 
most distorted at that time, and bailouts that do not reduce 
indebtedness and thus do not alleviate banks' distress may keep banks 
going but be unhelpful to the rest of the economy. (Ineffective banking 
regulations have caused much harm in Europe in recent years; many 
problems can be traced to a weak and bloated banking system and the 
politics of banking.)
---------------------------------------------------------------------------
     \23\ For example, in the Swedish crisis of 1992-1994, Government 
support for the banks necessitated cutbacks in other Government 
spending, which greatly contributed to the sharp economic recession. 
Citizens in Ireland and Iceland are still suffering.
---------------------------------------------------------------------------
    With these caveats, I will make a few observations about attempts 
to estimate the size of the subsidy, but I do not wish to focus on this 
technical issue. As I will argue below, the size of the subsidy does 
not actually matter much to the policy recommendation.

  1.  There is compelling evidence that the Government provided a 
        sector-wide collective bailout guarantees to the financial 
        sector in 2007-2009. \24\
---------------------------------------------------------------------------
     \24\ In one example, Kelly, et al. (2014) document the fact that 
during the recent financial crisis, (out-of-the-money) index put 
options that provide protection against large drops in the value of the 
entire financial sector were surprisingly cheap compared to the 
individual options of the financial institutions that are part of this 
index. This finding is consistent with the notion that the Government 
will not tolerate large equity losses for the financial sector as a 
whole. As a result, the market underprices the cost of insurance 
against these sector-wide losses for financials.

  2.  The value of the subsidy, if thought of as the amount the banks 
        would have to pay to receive perpetual (even partial) insurance 
        for their debts in the private markets, is sensitive to many 
        variables and can change dramatically over time depending on 
        the level of uncertainty, the state of the local and global 
        economy, and various fragilities in the financial system. The 
        value is highest when uncertainty is large and when the economy 
        and/or the financial sector are weak, and especially in a 
        crisis. Boom times, however, when the value of the subsidy 
        might be thought low, can quickly turn to bust. For example, 
        uncertainty indicators were low in 2006 and through summer 2007 
---------------------------------------------------------------------------
        only to explode in late 2008 and 2009.)

  3.  When focusing on the funding costs of the institutions, 
        particularly their borrowing costs, the relevant thought 
        experiment in trying to assess the value of the implicit 
        subsidies to the institutions who receive them from an ex ante 
        perspective, i.e., when institutions fund their investments in 
        light of the expectations of support, is to consider how 
        institutions would have fared in the hypothetical scenario in 
        which they tried to raise funding, such as unsecured, junior 
        debt, without any chance of a guarantee, and specifically in a 
        world in which the full costs of any failure, including 
        bankruptcy costs and the distortions of distress and 
        insolvency, would fall on shareholders and creditors. This 
        counterfactual scenario cannot be observed, thus comparison 
        requires many assumptions. One approach is to use credit 
        ratings uplifts. The approach makes sense if the uplifts 
        actually capture the true distinctions in the context of an 
        individual institution and specific bond issuance.

  4.  None of the approaches takes into account the extreme opacity of 
        the large banking institutions' and the difficulty in assessing 
        their risks, including those lurking off their balance sheets 
        and in derivatives markets. \25\ Many banks use derivatives to 
        get certain risks off their balance sheets. But then the 
        counterparties on these derivatives might fail. If the 
        counterparties have many parallel positions, as was the case 
        when AIG wrote credit default swaps for $500 billion on 
        mortgage-backed securities, CDOs, and the like, the risk that 
        the counterparty might fail is correlated with the underlying 
        risk, i.e., the attempt to hedge risks through derivatives may 
        end up being ineffective. In the case of AIG, fear of systemic 
        fallout from such a failure was a major reason for the bailout.
---------------------------------------------------------------------------
     \25\ On the poor disclosures of the banks and investors' inability 
to assess their risk, see for example Jesse Eisinger and Frank Partnoy, 
``What's Inside America's Banks'', The Atlantic, January 2, 2013.

  5.  Correlations of risks, i.e., the risk that the same event affects 
        multiple institutions, are notoriously difficult to measure. 
        This is especially true of the correlations among the risks 
        against which derivative contracts are written and the default 
        risks on these contracts. If these correlations are improperly 
        measured, however, credit ratings and credit ratings uplifts 
        are unlikely to be reliable. If these correlations are 
        neglected, as has been the case in the past (for example the 
        possibility that housing price declines will affect numerous 
        mortgages at the same time), the estimates of the total risk in 
        banks' assets are likely to be too low, and so are all 
        estimated of the value of Government guarantees protecting 
---------------------------------------------------------------------------
        against such risks.

  6.  In this context, it is also important to appreciate the role 
        played by Government guarantees for counterparties of banking 
        institutions. In a financial system with a complex network of 
        inter-institution contracts, the individual institution 
        benefits not only from Government guarantees protecting its own 
        creditors but also from Government guarantees protecting the 
        counterparties of those in which it invested. For example, the 
        AIG bailout benefited many counterparties of AIG, not the least 
        of these being the many banks that had purchased credit 
        insurance from AIG. The benefit of such protection for AIG to, 
        say, Goldman Sachs, however, cannot be assessed merely by 
        looking at data for Goldman Sachs and relating the interest 
        Goldman Sachs must pay to the risks they are taking. The 
        embeddedness of their activities in a system to which the 
        Government provides comprehensive support can hardly be 
        gathered from data about individual institutions.

  7.  Even a resolution process such as under Title 2 of DFA may offer 
        guarantees to some of the institutions' debt in order to avoid 
        disruptions or runs, which would transfer some downside risk to 
        the Government at least temporarily. \26\
---------------------------------------------------------------------------
     \26\ DFA directs the FDIC to cover any shortfall by charging the 
surviving institutions, but doing so might be difficult if they too are 
experiencing losses.

  8.  Being able to borrow at below-market rates relative to the risk 
        taken with the investments provides a subsidy that affects the 
        institutions' stock price and can favorably affect the terms at 
        which the institution can raise equity. When an insolvent 
        institution is given supports and does not fail, its 
        shareholders are not wiped out. Other things equal, therefore, 
        a systemic institution's stock price is higher in reality than 
        in the hypothetical without support. Indeed, raising equity has 
        been surprisingly cheap for the largest U.S. banks over the 
        past four decades, but expensive for the smallest banks, 
        because large bank stocks are priced under the assumption that 
        they are relative safe while the stocks of small banks are not, 
        despite the fact that large banks tend to be more heavily 
        indebted. \27\ The fact that guarantee become an asset, and the 
        fact that commonly used assumptions about the risks banks are 
        subject to may well be inappropriate, may lead the value of the 
        subsidies in some studies to be under-estimated. \28\
---------------------------------------------------------------------------
     \27\ Gandhi and Lustig (2014) find that over the past four decades 
the stock returns realized on the largest U.S. commercial banks, after 
adjusting for risk differences, are abnormally low compared to the 
stock returns on the smallest U.S. commercial banks. These differences 
are large (around 6 percent per year). The authors also provide 
evidence that large bank stocks are significantly less exposed to 
losses during recessions and financial crises, even though these large 
banks are typically much more heavily indebted. These findings are 
consistent with the notion that Government guarantees are perceived by 
investors to protect shareholders in large banks, but not in small 
banks, in financial disasters.
     \28\ See, for example, Stefan Nagel, ``Too Big To Fail Is Bigger 
Than You Think'', Bloomberg, March 2, 2014.

  9.  Comparisons between the interest charged on debt of large and 
        small banks may not be informative because the large banks may 
        well have significant risks that are harder to assess due to 
        their more opaque disclosures. As mentioned earlier, this 
        applies particularly to banks heavily involved in derivatives 
        trading. The larger banks also tend to have more complex 
        structures, more lines of business, and more off-balance sheet 
        exposures than small banks. These factors would affect funding 
        costs in the hypothetical scenario without support and thus the 
        comparison between large and small banks, and they might not be 
        sufficiently observable to correct for. Similar considerations 
        apply to comparisons of large banks with other large 
        corporations, whose disclosures, and business models are often 
---------------------------------------------------------------------------
        simpler and less opaque.

    The challenges in measuring how the banking industry as a whole, 
and especially the largest institutions, benefit from the possibility 
of future support do not change my bottom line, that the subsidy is 
perverse and insidious, rewarding and encouraging recklessness and 
excessive use of debt which endangers the public while allowing banks 
to make investments of many kinds to maximize their own profits that 
may not always benefit society.
    Because the public pays for any subsidy, and the result of implicit 
supports is a dangerous and distorted system, these subsidies are, on 
net, enormously costly for society. Even if banks were to pay in full 
for the guarantees, at least collectively--similar to how deposit 
insurance works--the impact of the implicit support is harmful and 
distortive. The same institutions whose failure would cause significant 
collateral damage--individually and when they fail at the same time--
have incentives to borrow too much, take too much risk, and become more 
highly interconnected, so as to increase the likelihood of Government 
support. In responding to these incentives, they can put us at yet more 
harm, unless these incentives are countered effectively by regulations. 
\29\
---------------------------------------------------------------------------
     \29\ See, for example, Brandao et al. (2013) for evidence on 
excessive risk taking as a result of expectations and support. Section 
5 in Admati et al. (2014) which discusses the why the leverage ratchet 
effect (addiction to borrowing by heavy borrowers) is particularly 
relevant in banking and exacerbated by guarantees, and this effect 
exacerbates other distortions. Admati and Hellwig (2013a, chapter 9) 
provide additional references. See also Anat Admati, ``Bank Immensity 
Undermines Responsibility'', New York Times Room for Debate, May 16, 
2014.
---------------------------------------------------------------------------
    Among the perverse consequences of implicit guarantees is that they 
encourage and enable the largest institutions to grow even to 
inefficiently large sizes. There is no valid evidence of true scale 
economies for banks as they grow to trillions in assets. Such sizes are 
unseen in the rest of the economy. \30\ Indeed, the problem of ``empire 
building'' by managers to benefit themselves appears particularly 
severe in banking. \31\ The largest institutions seem to suffer from 
serious governance and control problems, as evidenced by repeated 
scandals and fines. \32\ However, because the status of being too big 
to fail confers significant benefits and better access to funding, the 
largest institutions are unlikely to shrink naturally (as conglomerates 
often do).
---------------------------------------------------------------------------
     \30\ Davis and Tracey (2014) use estimates of the subsidies based 
on credit rating uplifts and argue that, once the effect of subsidies 
is controlled for, the largest institutions are ``too large to be 
efficient.''
     \31\ For example, Mayo (2011) describes excessive growth that 
appears inefficient, for example in Citigroup. A recent book (Fraser, 
2014) describes the recklessness of the Royal Bank of Scotland and its 
CEO, which led to its spectacular failure and bailout by UK taxpayers.
     \32\ For example, the report by the Senate Committee on 
Investigation chaired by Senator Carl Levin on ``London Whale'' 
scandal, entitled ``JPMorgan Chase Whale Trades: A Case History of 
Derivatives Risks and Abuses'', reveals serious control problems in our 
largest banks. Suspicion of fraud and other evasion of laws and 
regulations appear routinely in the press.
---------------------------------------------------------------------------
    These perverse effects undermine any notion of market discipline 
and they breed recklessness, even lawlessness, on the part of those 
within the largest institutions who benefit the most from the 
guarantees and subsidies, whose compensation reward gambling, and who 
rarely pay a personal price when charges for wrongdoings, including 
crimes, are settled by authorities or when excessive risks that harm 
the public, and even the shareholders of the corporations, are taken. 
Both corporate governance and regulations appear to fail. It is 
essential to take steps to counter these perverse incentives of the 
implicit subsidies and reduce their impact.
    Fortunately, there is a straightforward and cost-effective way to 
do just that while reaping other critical benefits; that is to reduce 
banks' excessive use of debt and requiring significantly more equity 
than banks are currently required to have. \33\ There is no reason for 
banks to live so dangerously. Importantly, aside from possibly losing 
subsidies associated with borrowing, the overall funding costs of banks 
would not increase if they use more equity and less debt. \34\ Since 
subsidies come from public funds, reducing them does not represent a 
social cost.
---------------------------------------------------------------------------
     \33\ Additional benefits are outlined in Admati et al. (2013, 
section 2) and Admati (2014).
     \34\ This is explained in details in Admati et al. (2013, see 
especially section 4); see chapter 9 of Admati and Hellwig (2013a) 
Claim 11 in Admati and Hellwig (2014), both in attached documents. 
Taxes are public funds, and the tax impact of higher equity 
requirements can easily be neutralized, as explained in Admati et al., 
(2013, section 4.1).
---------------------------------------------------------------------------
    Encouraging and subsidizing banks to fund themselves with as much 
debt as is currently allowed (up to 95 percent for the large bank 
holding companies) as perverse as encouraging and subsidizing reckless 
speed for trucks or rewarding the captains of large oil tankers to go 
ever closer to the coast. More equity would force banks to stand more 
on their own when they take risk, rather than shift some of the risk 
and cost of bearing it to others. Shareholders who benefit from the 
upside, and not creditors or taxpayers, should be the ones to bear the 
downside.
    Whatever else is done to reform the financial system so it works 
better for the rest of the economy, bringing banks' indebtedness to 
more reasonable levels appears enormously cost-beneficial. With the 
perverse incentives banks have, and their ability to get away with 
harmful actions, many of the problems will not be corrected by markets. 
Making the system safer requires focused and effectively enforced 
regulation. If the size of individual banks, or of the banking 
industry, shrinks as a result, the resulting size would likely be more 
appropriate. The size and structure of firms and industries should be 
determined by undistorted markets, but the markets we have are entirely 
distorted. Bloated and inefficient, the financial industry may be able 
to attract talented workforce that may be more productive elsewhere in 
the economy. This system works for few and harms all the rest. When 
regulations fail to correct such distortions and harm, the public pays 
the price. Because the issues are misunderstood and the harm from 
excessive risk in finance, unlike that from exploding trucks, is 
abstract, the public may not fully realize the situation, particularly 
with the extent of lobbying by the industry.
Summary: If not Now, When?\35\
    In March, 2013, the Senate voted unanimously to approve an 
amendment proposed by Senators Brown and Vitter to eliminate the too-
big-to-fail subsidies. As discussed above, among the many benefits of 
forcing the large banks to use more equity and less debt is that any 
subsidy they benefit from is immediately reduced. This benefit is 
obtained without having to break up the banks, and is realized in 
addition to all the other benefits of preventing their failure and 
reducing the distortions in their lending.
---------------------------------------------------------------------------
     \35\ This is the title of chapter 11 in Admati and Hellwig 
(2013a), whose epigraph is ``time has a trick of getting rotten before 
it gets ripe.'' For an excerpt, see Anat Admati and Martin Hellwig, 
``Must Financial Reform Await Another Crisis?'' Bloomberg View, 
February 6, 2013.
---------------------------------------------------------------------------
    The focus on making the failure option palatable is as misguided as 
a focus on preparing ambulances for a possible explosion while police 
allows loaded trucks to drive at 95 miles an hour in residential 
neighborhoods. Whoever pays for the ambulances, explosions harm 
innocent people. Requiring that banks fund themselves so that those who 
benefit from the upside of risk bear more of its downside brings about 
more safety and corrects distortions.
    In the exchange on July 15, 2014, between Senator Warren and Chair 
Yellen referred to earlier, Senator Warren pointed out that under Title 
1 of DFA, the Fed has authority to break up the largest bank holding 
companies if it finds that bankruptcy is not a viable option if they 
fail. The Fed certainly has authority to ban dividends and other 
payouts to shareholders until banks are better prepared to absorb 
losses from risks they take without failing or becoming distressed.
    As it goes through the ``iterative process'' of the living wills, 
and while it is not ready to assert that the failure of the largest 
bank holding companies will not harm the economy, the Fed must act 
prudently and protect the public. Corporations routinely retain their 
profits to fund investments, and banks should do the same. Retained 
profits would enable banks to make more worthy loans, and may increase 
their incentives to actually make them. The profits from any 
investments belong to shareholders as long as debt is paid. \36\
---------------------------------------------------------------------------
     \36\ Warren Buffett's company Berkshire Hathaway, for example, 
rarely makes payouts to its shareholders, continuing to invest on their 
behalf and retained earnings are considered first in the ``pecking 
order'' of funding. See Admati et al. (2014), for example.
---------------------------------------------------------------------------
    Not only do banks have access to their own profits to become more 
resilient, they can sell shares to investors at appropriate prices. 
Other companies may be forced by debt covenants or prohibitive 
borrowing costs to raise equity when they are distressed. For banks, 
action must come from regulators. Banks unable to raise equity at any 
price fail a basic market ``stress tests'' and might be too opaque or 
not viable without subsidies. Such banks are unhealthy and must be 
dealt with promptly.
    The Fed justifies allowing banks to make payouts to their 
shareholders on the basis of ``stress tests.'' This methodology uses 
models to predict regulatory capital levels that mean little in actual 
distress and especially in a crisis. The models are incapable of 
predicting the within-system dynamics that might follow adverse 
scenarios because the Fed does not have sufficient information on the 
many layers of interconnectedness that go beyond single counterparty 
exposures. Trusting models that should not be trusted has contributed 
to the causes of the financial crisis. The lesson from the failures of 
these models must be learned, particularly when there is no scarcity of 
equity just for banks, and no justification for allowing them to live 
as dangerously as they do. \37\
---------------------------------------------------------------------------
     \37\ See Claims 13-14 in Admati and Hellwig (2014), attached, for 
a brief discussion.
---------------------------------------------------------------------------
    If banks deny that they benefit from implicit subsidies, moreover, 
they cannot at the same time complain that their funding costs would 
increase significantly if they must use more equity. \38\ The fact that 
banks are anxious to make payouts to their shareholders rather than use 
their profits for making worthy loans, even at their very low equity 
levels, calls into question their motives and exposes the disconnect 
between claims that higher equity requirements would prevent lending 
and making payout to shareholders instead of using the funds to make 
loans.
---------------------------------------------------------------------------
     \38\ In that case, the only private cost is that banks might have 
to pay more corporate taxes, but, as explained in Admati et al. (2013, 
section 4.1), this is not a social cost, and the effect can anyway be 
neutralized.
---------------------------------------------------------------------------
    It is baffling that the Fed finds it appropriate, before it can 
assert that the largest bank holding companies would not harm the 
economy if they fail, to allow these institutions to make payouts to 
shareholders that deplete their most reliable loss-absorbing capacity, 
namely their equity. \39\ A significant increase in equity requirements 
must be considered the most cost-effective way to make it less likely 
that we face difficult choices when institutions become weak, as well 
as to reduce the fragility of the system and many distortions. The Fed 
has the responsibility and the ability to protect the public, yet as a 
regulator, it has failed the public. On behalf of the public, I hope 
you will take my comments into consideration and implore it to do 
better. \40\
---------------------------------------------------------------------------
     \39\ I have written many commentaries on this issue, see Anat 
Admati, ``Dividends Can Wait Until the Banks Are Stronger'', Financial 
Times, January 19, 2011, ``Only Recapitalized Banks Should Pay 
Dividends'', a letter signed by 16 academics, Financial Times, February 
15, 2011, Anat Admati, ``Fed Runs Scared With Boost to Bank 
Dividends'', Bloomberg View, February 24, 2011, and ``Why the Bank 
Dividends Are a Bad Idea'', Reuters, March 14, 2012. Admati and Hellwig 
(2011, chapter 11) provide a more detailed explanation.
     \40\ Other claims are made in response to such recommendations, 
such as concerns about the so-called shadow banking system or about the 
competitiveness of our banks. These concerns are invalid excuses, as 
explained in Admati and Hellwig (2013a, chapters 12 and 13) and Claims 
26-28 in Admati and Hellwig (2014), attached.
---------------------------------------------------------------------------
References\41\
Admati, Anat R. (2014), ``The Compelling Case for Stronger and More 
    Effective Leverage Regulation in Banking'', Journal of Legal 
    Studies, forthcoming.
---------------------------------------------------------------------------
     \41\ Most of the references (at least in working paper form) are 
available online. My own academic papers and other writings on the 
topic are posted at http://www.gsb.stanford.edu/news/research/
admati.etal.html.
---------------------------------------------------------------------------
Admati, Anat R., Peter M., DeMarzo, Martin F. Hellwig, and Paul 
    Pfleiderer (2013), ``Fallacies, Irrelevant Facts, and Myths in the 
    Discussion of Capital Regulation: Why Bank Equity Is Not Socially 
    Expensive'', Working paper; and (2014), ``The Leverage Ratchet 
    Effect'', Working paper.
Admati, Anat R., and Martin F. Hellwig (2013a), ``The Bankers' New 
    Clothes: What's Wrong With Banking and What To Do about It'', 
    Princeton University Press. (Excerpts and links available at http:/
    /bankersnewclothes.com/.)
Admati, Anat R., and Martin F. Hellwig (2013b), ``Does Debt Discipline 
    Bankers? An Academic Myth About Bank Indebtedness'', Working paper 
    (based on a chapter omitted from Admati and Hellwig (2013a).
Admati, Anat R., and Martin F. Hellwig (2014), ``The Parade of Bankers' 
    New Clothes Continues: 28 Flawed Claims Debunked'', available at 
    http://bankersnewclothes.com/wp-content/uploads/2014/07/Parade-
    continues-July-2014.pdf.
Bair, Sheila (2012), ``Bull by the Horns: Fighting To Save Main Street 
    From Wall Street and Wall Street from Itself'', Free Press.
Barofsky, Neil (2012), ``Bailout: An Inside Account of How Washington 
    Abandoned Main Street While Rescuing Wall Street'', Free Press.
Brandao, Luis Marques, Ricardo Correa, and Horacio Sapriza (2013), 
    ``International Evidence on Government Support and Risk Taking in 
    the Banking Sector'', IMF Working Paper WP/13/94.
Cole, Rebel A. (2013), ``How Did the Financial Crisis Affect Business 
    Lending in the U.S.?'' Working paper.
Davies, Richard, and Belinda Tracy (2014), ``Too Big To Be Efficient? 
    The Impact of Implicit Subsidies on Estimates of Scale Economies in 
    Banking'', Journal of Money, Credit and Banking, 219-253. (Working 
    paper version dated 2012 available online.)
Fraser, Ian (2014), Shredded: Inside RBS, the Bank That Broke Britain, 
    Birlinn; ``Too Big To Be Efficient? The Impact of Implicit 
    Subsidies on Estimates of Scale Economies in Banking'', Journal of 
    Money, Credit and Banking.
Gandhi, Priyank, and Hanno Lustig (2014), ``Size Anomalies in U.S. Bank 
    Stock Returns'', Journal of Finance, Forthcoming. (Working paper 
    dated 2012 available on SSRN.com.)
Kelly, Bryan, Hanno Lustig, and Stijn Van Niewerburgh (2012), ``Too-
    Systemic-To-Fail: What Option Markets Imply About Sector-Wide 
    Government Guarantees'', Working paper.
Mayo, Mike (2011), ``Exile on Wall Street: One Analyst's Fight To Save 
    the Big Banks From Themselves'', John Wiley and Sons.
Omarova, Saula T. (2013), ``The Merchants of Wall Street: Banking, 
    Commerce, and Commodities'', Minnesota Law Review, 2-78.
Onaran, Yalman (2011), ``Zombie Banks: How Broken Banks and Debtor 
    Nations Are Crippling the Global Economy'', Bloomberg.
Skeel, David A., Jr. (2014), ``Single Point of Entry and the Bankruptcy 
    Alternative'', Working paper.
Skeel, David A., Jr., and Thomas H. Jackson (2012), ``Transaction 
    Consistency and the New Finance in Bankruptcy'', Columbia Law 
    Review, 152-202.
White, Lawrence J. (2014), ``The Basics of `Too Big To Fail' '', 
    Forthcoming in Paul H. Schultz, ed. Dodd-Frank and the Future of 
    Finance, MIT Press.
    
    
    
    
               PREPARED STATEMENT OF DOUGLAS HOLTZ-EAKIN
                    President, American Action Forum
                             July 31, 2014
    Chairman Brown, Ranking Member Toomey, and Members of the 
Subcommittee, I am grateful for the privilege of appearing today. In my 
brief testimony today I would like to make four main points:

    Any expectations of Government support for bank holding 
        companies is at root a problem created by policymakers' 
        discretionary actions;

    The history of Federal Government assistance is not a 
        pattern of consistent intervention on behalf of large firms, 
        but rather an erratic and unpredictable series of interventions 
        on behalf of firms large, small, financial, and nonfinancial;

    Attempts to measure any ``implicit too-big-to-fail (TBTF) 
        subsidy'' is an elusive quest due to the many confounding 
        factors; and

    Any market TBTF expectation is hardly fixed, but is 
        necessarily a changing reality.

    To start, I should stipulate that I do not seek in this testimony 
to specifically criticize or address the Government Accountability 
Office report on Government support for bank holding companies, nor the 
several other reports from other institutions on the same topic. Let me 
also stipulate that no firm (financial or otherwise) should ever 
benefit from an unfair advantage owing to policy-induced bias. Herein I 
only hope to provide the Committee with a brief conceptual discussion 
of some of the issues surrounding the question too big to fail (TBTF) 
and implicit subsidies. \1\
---------------------------------------------------------------------------
     \1\ A longer discussion of these issues can be found in Satya 
Thallam, ``Reconsidering Too Big To Fail'', American Action Forum, 
Research, March 12, 2014, http://americanactionforum.org/research/
reconsidering-too-big-to-fail.
---------------------------------------------------------------------------
The Policymaker-Creditor Nexus
    What is too big to fail? \2\ It is not a market failure, like an 
externality; it ``is a rational market response to expectations set by 
Government policy.'' \3\ The proximate beneficiaries of any perceived 
bailout expectations (the banks) benefit passively--the ultimate source 
of any implicit subsidy exists at the nexus of the banks' creditors and 
expectations imputed from policymaker choices. As Minneapolis Fed 
President Narayana Kocherlakota put it, the proper conception of too 
big to fail ``emphasizes the role of creditor beliefs . . . The beliefs 
of other parties are much less relevant.'' \4\
---------------------------------------------------------------------------
     \2\ Here I use the phrase ``too big to fail'' or ``TBTF''.
     \3\ Supra, n. 1.
     \4\ Narayana Kocherlakota, ``Too-Big-To-Fail: The Role of 
Metrics'', Speech delivered at ``Quantifying the `Too Big to Fail' 
Subsidy Workshop'', Federal Reserve Bank of Minneapolis, Minneapolis, 
Minnesota, November 18, 2013, http://www.minneapolisfed.org/
news_events/pres/speech_display.cfm?id=5203.
---------------------------------------------------------------------------
    A creditor's belief that an institution will receive Government 
support will be rooted in an expectation that policymakers will take 
extraordinary steps to prevent contagion from one firm's failure to 
spread to others. Financial interdependencies may be the transmission 
mechanism for shocks to spread throughout a system, but policymakers 
make the ultimate decision to intervene, creating the ex ante 
expectation in the first place. Thus policymakers attempting to 
eliminate any implicit TBTF subsidy will need to look to themselves--or 
more specifically they will need to consider the rules and regulations 
which open the door to future intervention, or even lead creditors to 
believe intervention is forthcoming.
Unpredictability
    The Federal Government has a dubious history of intervening in 
times of economic distress to save certain firms or otherwise mitigate 
their losses. Unfortunately for analysts and policymakers seeking to 
determine the financial effects of these interventions, this history is 
inconsistent, not hewing to any rule or regularity. In the most recent 
financial crisis, the Federal Government's response swung from pillar 
to post, intervening (Bear Stearns), then not (Lehman Brothers), 
intermittently providing assistance to investment banks, banks large 
and small (TARP), investment funds, and automakers (GM and Chrysler). 
If we go further back, we see intervention on behalf of a large, 
conventional commercial bank (Continental Illinois), a not particularly 
large or major money center institution (Long Term Capital Management), 
savings and loans, airlines, and even a city (New York City). \5\ It's 
the very breadth and variety of these interventions (not to mention the 
extreme infrequency relative to the gross number of large firm failures 
during the same period) that should lead one to be skeptical of claims 
purporting a robust relationship between certain firms' insolvency and 
Government rescue.
---------------------------------------------------------------------------
     \5\ Jesse Nankin and Krista Kjellman Schmidt, ``History of U.S. 
Gov't Bailouts'', ProPublica April 15, 2009, http://www.propublica.org/
special/government-bailouts.
---------------------------------------------------------------------------
    And yet, even this incomplete list understates the highly 
heterogeneous nature of interventions. Loans, loan guarantees, capital 
infusions, stock purchases and warrants, direct transfers--all 
interventions are not born the same, and more to the point, have 
differing effects on different parties.
    The larger point is that if, for example, investors in a bank 
holding company's bonds are pricing in a discount (lower yield) owing 
to some probability of a bailout conditional on insolvency, we must 
presume those investors have determined the likelihood not only that 
policymakers will in fact intervene, but that they have also correctly 
identified the firm that will receive assistance, and that the 
intervention will benefit them as opposed to shareholders, executives, 
employees, or even other classes of debtholders. Indeed, as we saw in 
the Chrysler bailout, some bondholders were in fact made worse off. \6\
---------------------------------------------------------------------------
     \6\ Todd Zywicki, ``The Auto Bailout and the Rule of Law'', 
National Journal, No. 7, Spring 2011, http://www.nationalaffairs.com/
publications/detail/the-auto-bailout-and-the-rule-of-law.
---------------------------------------------------------------------------
Confounding Factors
    One prevailing line of thinking points to the fact that large 
financial institutions can borrow more cheaply relative to smaller 
ones, and thus this differential is evidence of TBTF. But there are 
many factors that affect the funding costs of various institutions. 
This large-small differential in fact exists across most industries, 
with the banking industry somewhere near the middle. \7\ Differences in 
the liquidity of debt, risk diversification, information limitations, 
and other factors may explain much or all of the differential. That 
said, there still might be a part of the differential that cannot be 
explained by size-dependent factors--a TBTF subsidy may still be 
embedded. But any attempt to quantify the TBTF subsidy using cost of 
funding will need to successfully separate out the non-TBTF factors, 
which is exceedingly difficult and perhaps even impossible.
---------------------------------------------------------------------------
     \7\ Randall Kroszner, ``A Review of Bank Funding Differentials'', 
Presented at ``Too Big to Fail and Its Implications on Bank Funding 
Costs'', NYU School of Business, October 8, 2013, http://
www.stern.nyu.edu/cons/groups/content/documents/webasset/
con_044532.pdf.
---------------------------------------------------------------------------
    If a TBTF subsidy does exist, it stands to reason that it exists on 
a continuum, rather than simply as a binary condition between those 
firms that are TBTF and those that are absolutely not. Thus properly 
determining the subsidy portion of the differential is beside the point 
if one cannot properly identify the two categories of institutions.
Changing Expectations
    The yield spread between firms may be ever changing; indeed, it has 
at times even become negative. \8\ As it changes, one must conclude 
either that: (1) the TBTF subsidy is in fact changing and transferring 
among institutions over time; \9\ (2) the yield spread attributable to 
TBTF is being swamped by other effects; or (3) the yield spread is not 
a reliable measure of the TBTF subsidy.
---------------------------------------------------------------------------
     \8\ Steve Strongin, et al., ``Measuring the TBTF Effect on Bond 
Pricing'', Goldman Sachs Global Markets Institute, May 2013, http://
www.goldmansachs.com/our-thinking/public-policy/regulatory-reform/
measuring-tbtf-doc.pdf. See also Kroszner 2013, supra.
     \9\ Kenichi Ueda and Beatrice Weder di Mauro, ``Quantifying 
Structural Subsidy Values for Systemically Important Financial 
Institutions'', IMF Working Paper 12/128, May 2012, http://www.imf.org/
external/pubs/ft/wp/2012/wp12128.pdf.
---------------------------------------------------------------------------
    Consider what bank investors and creditors must have thought about 
the likelihood of rescue following the collapse of Bear Stearns as 
compared to after the collapse of Lehman Brothers. In just 1 year, 
real-time policy choices must have drastically changed the implied TBTF 
subsidy. Thus the outcome of any TBTF study will be directly affected 
by the window of time chosen to examine. But the larger point is that 
any policy chosen now or in the near future as a TBTF corrective may be 
(in the best case scenario) appropriately targeted for some fixed state 
of the world, but cannot easily adjust to changing conditions. In the 
extreme, would such a policy corrective (such as a tax) become a refund 
if and when the TBTF subsidy reverses?
Final Thoughts
    Two wrongs do not make a right. Even if we presume the existence of 
a consistent and significant TBTF subsidy, one must consider the net 
effect of applying another distortion on top of the first. That is, 
proposed solutions such as a bank tax or a financial transactions tax 
applied to TBTF institutions are attempting to counteract a distorting 
dynamic created by policymaker expectations and creditors' response 
with a punitive measure which works along a somewhat different channel.
    Discretion is the handmaiden of bailouts. Time consistency in 
policymaking is an age-old problem and is not limited to financial 
crises. \10\ Congress should focus its energy on those mechanisms 
which: (1) make bank failures easier and predictable; and (2) limit 
policy choices even in a time of crisis.
---------------------------------------------------------------------------
     \10\ See Finn Kydland and Edward Prescott, ``Rules Rather Than 
Discretion: The Inconsistency of Optimal Plans'', Journal of Political 
Economy, Vol. 85, Iss. 3, pp. 473-492, June 1977. The difficulty of 
time-consistent plans are illustrated by a recent report which stated: 
``If the only choices are between bailout and fire-sale liquidations or 
value-destroying reorganizations that can result in a contagious panic 
and collapse of the financial system, responsible policymakers 
typically choose bailout as the lesser of two evils.'' John Bovenzi, 
Randall Guynn, and Thomas H. Jackson, ``Too Big to Fail: The Path to a 
Solution'', Report of the Failure Resolution Task Force of the 
Financial Regulatory Reform Initiative, Bipartisan Policy Center, May 
2013, p. 43.
---------------------------------------------------------------------------
    One particularly promising avenue in this regard is to replace 
Title II of the Dodd-Frank Act with a bankruptcy process for banks. 
This would place decisions in the hands of a court, and not either an 
agency or the Congress. In the process it would limit discretion and 
clarify the outlook for creditors.
    The Dodd-Frank Act happened. Whether one considers the Dodd-Frank 
Act a positive or negative change to financial regulation, there is 
little argument that it has a significant effect on financial 
institutions. This includes numerous new requirements and restrictions 
on the industry, many of them directed specifically at the largest bank 
holdings companies. The upshot is that any perceived advantages must be 
considered on net with any of these new costs.
    Thank you and I look forward to answering your questions.
        RESPONSES TO WRITTEN QUESTIONS OF CHAIRMAN BROWN
                      FROM LAWRENCE EVANS

Q.1. GAO used wide-ranging data and a number of models to 
estimate the various funding cost differences in its study. 
Will GAO make the full documentation of its data, coding, 
methodology, etc., available to third parties--either for free 
or at cost--so that independent experts can examine the data a 
processes to make their own evaluations and draw their own 
conclusions?

A.1. While GAO is not subject to the Freedom of Information 
Act, its disclosure policy follows the spirit of the act 
consistent with GAO's duties and responsibilities to the 
Congress. Upon written request, GAO may provide GAO records 
associated with this engagement to parties wishing to replicate 
our work. However, please be advised that certain exemptions to 
disclosure apply and GAO does not release certain information 
including proprietary and trade secret data. See 4 CFR 81.5 
and 4 CFR 81.6. \1\ Interested parties should submit their 
request in writing to GAO's Chief Quality Officer. The request 
may be emailed to [email protected], faxed to (202) 512-
5806, or mailed by traditional mail.
---------------------------------------------------------------------------
     \1\ For information concerning fees and charges, please see 4 CFR 
81.7.

Q.2. Has GAO released any information--such as a list of 
meetings, conference calls, and other conversations--regarding 
---------------------------------------------------------------------------
the parties with whom you consulted in preparing this report?

A.2. No, GAO has not released indentifying information on the 
parties interviewed for the report. \2\ The Objectives Scope 
and Methodology section of the report discloses only that we 
conducted interviews with representatives from credit rating 
agencies, investment firms, and corporations that are customers 
of banks, bank holding companies of various sizes, bank 
industry associations, public interest groups, academics, and 
other experts.
---------------------------------------------------------------------------
     \2\ As per our protocols, for any ongoing work--except for 
classified work and investigations--GAO will disclose, if asked (e.g., 
by Members, congressional staff, agencies, or the press) the source of 
the request and the project's objectives, scope, and methodology. 
Additionally, all congressional offices have, through the Senate and 
House intranet connections to GAO, access to the background and key 
research questions for active GAO assignments, except for those cases 
where the reporting of such work would result in disclosing classified 
or other sensitive information. The information we volunteer would not 
include the detailed information on the parties we selected for 
interview.

Q.3. Would you be willing to provide such information, 
---------------------------------------------------------------------------
including the party initiating the contact?

A.3. Yes, GAO will grant Members, upon their written request, 
access to the available information at GAO offices or will 
provide copies of such lists. The Objectives, Scope, and 
Methodology Section of the report details the criteria we used 
to select parties to interview.

Additional Information To Correct the Record

    During the hearing Chairman Brown stated,

        GAO used three industry-funded studies to design this 
        report . . .

    For the record, this is inaccurate. To inform our 
econometric approach and understand the breadth of results and 
methodological approaches, we reviewed 16 studies--1 of which 
was conducted by researchers at a large bank holding company 
and two others that were sponsored by a trade group 
representing large commercial banks. Taking into consideration 
the strengths and limitations of different methodologies, we 
developed our own econometric approach. We then selected three 
experts with relevant expertise to review our methodology and 
assess its strengths and limitations. These experts reviewed 
our approach before we implemented it and provided comments. In 
many instances, we made changes or additions to our models to 
address their comments, and in other instances, we disclosed 
additional limitations of the models. Before selecting these 
experts, we reviewed potential sources of conflicts of 
interest, and we determined that the experts we selected did 
not have any material conflicts of interest for the purpose of 
reviewing our work. Note well, the GAO approach was influenced 
most significantly by the research conducted by Dr. and his 
colleagues.
    During the hearing Chairman Brown stated

         . . . and the GAO arranged meetings with corporate 
        treasurers of companies suggested exclusively--I 
        believe exclusively--by the U.S. Chamber of Commerce . 
        . .

    GAO--not the Chamber of Commerce--indentified firms we 
planned to contact. At GAO's request, the Chamber informed its 
members that GAO was reaching out to corporate customers of 
banks. GAO then contacted and interviewed four of the member 
firms that expressed interest and met our criteria. We 
subsequently selected two additional firms to achieve further 
diversity across industry sectors. Note, GAO selected U.S. 
corporations from different industry sectors and with a range 
of banking needs. These corporate treasurers provided a diverse 
set of views on the issue.
                                ------                                


        RESPONSES TO WRITTEN QUESTIONS OF CHAIRMAN BROWN
                    FROM DOUGLAS HOLTZ-EAKIN

Q.1. During the hearing, I asked all of the witnesses about the 
ratings ``uplift'' given to banks by ratings agencies as a 
result of perceived Government support. In response, you said 
that ``recently most if not all the major credit agencies have 
removed the credit uplifts . . . If you care about `too big to 
fail' now and 2013, '14, it's gone from that perspective.''
    It is true that Moody's recently removed the uplift for the 
holding company debt of the eight U.S. G-SIBs, while 
maintaining its assumption of support for bank-level senior 
debt. However, S&P released a report on August 4th--after the 
Subcommittee hearing--titled ``U.S. Banks: Government Support 
Is Fading But Not Gone--Yet.'' In it, they maintain a 1-2 notch 
uplift for the U.S. G-SIBs. In March, Fitch Ratings also 
released support ratings for some U.S. G-SIBs that ``reflect 
Fitch's expectation that there remains an extremely high 
probability of support from the U.S. Government (rated `AAA', 
Rating Outlook Stable) if required. This expectation reflects 
the U.S.'s extremely high ability to support its banks 
especially given its strong financial flexibility, though 
propensity is becoming less certain.''
    Would you care to amend or revise your response to the 
question in any way?

A.1. Thank you for bringing these more recent data to my 
attention. The Moody's report indicates that the basic trend 
toward no Government support remains, but that the pace is 
slower than I had estimated and markets may not as yet have 
fully reflected these developments as I had anticipated.
              Additional Material Supplied for the Record
               CHARTS SUBMITTED BY CHAIRMAN SHERROD BROWN
               



        REPORT SUBMITTED BY THE GOVERNMENT ACCOUNTABILITY OFFICE




COMMENTS REGARDING THE ``GAO REPORT ON LARGE BANK HOLDING COMPANIES'', 
                          BY ALLAN H. MELTZER