<DOC>
[108 Senate Hearings]
[From the U.S. Government Printing Office via GPO Access]
[DOCID: f:94514.wais]
S. Hrg. 108-495
REVIEW OF THE NEW
BASEL CAPITAL ACCORD
=======================================================================
HEARING
before the
COMMITTEE ON
BANKING,HOUSING,AND URBAN AFFAIRS
UNITED STATES SENATE
ONE HUNDRED EIGHTH CONGRESS
FIRST SESSION
ON
THE NEW BASEL CAPITAL ACCORD PROPOSAL, WHICH WILL ADDRESS AN AREA WHICH
IS IMPORTANT TO THE SAFE AND SOUND FUNCTION OF OUR BANKING SYSTEM
__________
JUNE 18, 2003
__________
Printed for the use of the Committee on Banking, Housing, and Urban
Affairs
U.S. GOVERNMENT PRINTING OFFICE
94-514 WASHINGTON : DC
____________________________________________________________________________
For Sale by the Superintendent of Documents, U.S. Government Printing Office
Internet: bookstore.gpo.gov Phone: toll free (866) 512-1800; (202) 512ÿ091800
Fax: (202) 512ÿ092250 Mail: Stop SSOP, Washington, DC 20402ÿ090001
COMMITTEE ON BANKING, HOUSING, AND URBAN AFFAIRS
RICHARD C. SHELBY, Alabama, Chairman
ROBERT F. BENNETT, Utah PAUL S. SARBANES, Maryland
WAYNE ALLARD, Colorado CHRISTOPHER J. DODD, Connecticut
MICHAEL B. ENZI, Wyoming TIM JOHNSON, South Dakota
CHUCK HAGEL, Nebraska JACK REED, Rhode Island
RICK SANTORUM, Pennsylvania CHARLES E. SCHUMER, New York
JIM BUNNING, Kentucky EVAN BAYH, Indiana
MIKE CRAPO, Idaho ZELL MILLER, Georgia
JOHN E. SUNUNU, New Hampshire THOMAS R. CARPER, Delaware
ELIZABETH DOLE, North Carolina DEBBIE STABENOW, Michigan
LINCOLN D. CHAFEE, Rhode Island JON S. CORZINE, New Jersey
Kathleen L. Casey, Staff Director and Counsel
Steven B. Harris, Democratic Staff Director and Chief Counsel
Mark F. Oesterle, Counsel
Martin J. Gruenberg, Democratic Senior Counsel
Vincent Meehan, Democratic Counsel
Joseph R. Kolinski, Chief Clerk and Computer Systems Administrator
George E. Whittle, Editor
(ii)
C O N T E N T S
----------
WEDNESDAY, JUNE 18, 2003
Page
Opening statement of Chairman Shelby............................. 1
Opening statements, comments, or prepared statements of:
Senator Sarbanes............................................. 2
Senator Allard............................................... 3
Prepared statement....................................... 53
Senator Reed................................................. 3
Senator Sununu............................................... 3
Senator Corzine.............................................. 21
WITNESSES
Roger W. Ferguson, Jr., Vice Chairman, Board of Governors of the
Federal Reserve System......................................... 3
Prepared statement........................................... 53
John D. Hawke, Jr., Comptroller of the Currency, U.S. Department
of the Treasury................................................ 7
Prepared statement........................................... 71
Donald E. Powell, Chairman, Federal Deposit Insurance Corporation 10
Prepared statement........................................... 83
James E. Gilleran, Director, Office of Thrift Supervision........ 11
Prepared statement........................................... 86
Maurice H. Hartigan, II, President and CEO, RMA--The Risk
Management Association......................................... 33
Prepared statement........................................... 90
Micah S. Green, President, The Bond Market Association........... 35
Prepared statement........................................... 92
Edward I. Altman, Max L. Heine Professor of Finance, Leonard N.
Stern School of Business, New York University.................. 36
Prepared statement........................................... 96
Karen Shaw Petrou, Managing Partner, Federal Financial Analytics,
Inc............................................................ 39
Prepared statement........................................... 98
D. Wilson Ervin, Managing Director, Credit Suisse First Boston,
on Behalf of the Financial Services Roundtable................. 41
Prepared statement........................................... 103
Kevin M. Blakely, Executive Vice President and Chief Risk
Officer, KeyCorp............................................... 44
Prepared statement........................................... 112
(iii)
REVIEW OF THE NEW BASEL CAPITAL ACCORD
----------
WEDNESDAY, JUNE 18, 2003
U.S. Senate,
Committee on Banking, Housing, and Urban Affairs,
Washington, DC.
The Committee met at 10:05 a.m. in room SD-538 of the
Dirksen Senate Office Building, Senator Richard C. Shelby
(Chairman of the Committee) presiding.
OPENING STATEMENT OF CHAIRMAN RICHARD C. SHELBY
Chairman Shelby. The hearing will come to order.
I want to first thank the witnesses for coming here this
morning. The purpose of the hearing today is to consider the
New Basel Capital Accord Proposal. This important proposal
addresses an area that is extremely important to the safe and
sound function of our banking system. With only a quick glance
at the economic history books, one can readily determine that
thinly capitalized banks pose huge risks to depositors, the
banking system, and to the overall economy, and perhaps to the
taxpayers.
To protect against such risks, we have employed minimum
capital requirements as a means to ensure that banks possess
the financial integrity necessary to carry on banking
activities. What this really means is that bank capital, the
bank owner's money, is ``on the line'' with the other bank
resources used to conduct business. Thus, bank losses translate
to bank owners' losses.
By ensuring the sharing of losses amongst bank depositors,
creditors, and owners, capital requirements properly align the
interests of these groups, and alignment of these interests is
crucial. Capital is a very valuable thing. There is tremendous
competition for it. Those who provide it expect something in
return for it. Those who obtain it must protect it and must
make sure it produces. Thus, with their own capital in the
breach, banks have developed very sophisticated risk
identification, analysis, and management tools to achieve these
ends. Ultimately, the combination of capital requirements and
risk management techniques have served us well.
Today, we are considering proposed changes to the current
capital regime, changes which could have very serious effects
on the amount of risk-based capital banks are required to hold,
on the risk management techniques they employ, and even on the
domestic and international competitive landscapes.
Because of the significant nature of these issues, I
believe this Committee has a responsibility to closely
scrutinize the proposal and, at a minimum, become aware of its
ramifications so that we can draw our own conclusions regarding
its merit.
I thank you for being here today, and I look forward to
your testimony a little later.
Senator Sarbanes.
STATEMENT OF SENATOR PAUL S. SARBANES
Senator Sarbanes. Thank you very much, Mr. Chairman.
First, I want to welcome this distinguished panel of
regulators, our line of defense, perhaps, for the Banking
Committee this morning. And I want to commend Chairman Shelby
for holding this oversight hearing on the Basel Capital Accord
which is now being worked on.
The New Basel Accord is a highly complex proposal with
potentially significant consequences for both the United States
and the international financial system, and I think it is
appropriate for the Committee to review its likely impact, in
fact, before the agreement is actually concluded and the
process of implementing it begins. The whole dynamic changes at
that point, and if there are concerns about it and what its
implications are, I think it is important that they be heard
earlier rather than later. Otherwise, we run the risk of
examining what has been agreed upon, which is different from
examining something that has not yet been agreed upon.
The first Basel Accord, concluded in 1988 and fully
implemented by 1992, was an effort to establish international
standards for the measurement of bank capital in order to bring
about greater uniformity of regulation and reduce risk in the
international financial system. I think it is generally
acknowledged to have been a significant step forward, but there
now appears to be agreement that the system for measuring risk
under Basel I may be inadequate today, particularly for large,
complex financial institutions. In fact, former Federal Reserve
Board Governor Laurence Meyer observed, ``Large, complex
banking organizations now routinely structure their portfolios
in ways that arbitrage around the current capital standard.
These banks can often lower their capital requirements with
little, if any, reduction in their actual risk taking. As a
result, reported capital ratios may, and often do, overstate a
bank's true financial strength.'' That is former Fed Board
Governor Laurence Meyer.
The proposed Basel II Accord is an effort to capture in a
more sophisticated way the financial risks undertaken by
banking institutions and to assign capital requirements
appropriate to those risks. There appears to be agreement about
the broad goals of the Basel II Accord, but press accounts
report there are significant differences of view among the bank
regulators about the agreement itself and, indeed, within the
American banking industry.
Most of the discussion is focused on whether the proposed
Accord would raise or reduce capital; whether it would place
one set of U.S. banking institutions at a competitive
disadvantage to another; or whether U.S. institutions generally
would be placed at a disadvantage relative to foreign
institutions; and, finally, whether the agreement is too
complex and difficult to implement.
These are very important questions. It seems to me
appropriate to have a public discussion about them. We need to
have some sense of the domestic and international impact of the
Accord. One of the key questions that will need to be addressed
today is whether the regulators believe they have a sufficient
grasp on the impact of the proposed agreement to conclude it
internationally and implement it domestically. I must say,
having sat on this Committee for a number of years, I would
have concerns about any agreement that would significantly
reduce the capital held by large financial institutions in
terms of the safety and soundness of the system.
Mr. Chairman, I look forward to the testimony.
Chairman Shelby. Thank you.
Senator Allard.
COMMENTS OF SENATOR WAYNE ALLARD
Senator Allard. Mr. Chairman, I would like to hear from our
panel, and to expedite that, I would like to just submit my
comments for the record.
Chairman Shelby. Without objection, it will be entered into
the record.
Senator Allard. And I thank them personally for being here,
and I look forward to their testimony.
Chairman Shelby. Senator Reed.
COMMENTS OF SENATOR JACK REED
Senator Reed. Mr. Chairman, I think Senator Allard set the
right direction. And I look forward to hearing the testimony.
Chairman Shelby. Senator Sununu.
COMMENTS OF SENATOR JOHN E. SUNUNU
Senator Sununu. Thank you, Mr. Chairman. I have no opening
statement. I am looking forward to hearing from the witnesses.
Chairman Shelby. Our first panel today, we have with us the
Honorable Roger W. Ferguson, Jr., who is the Vice Chairman of
the Board of Governors of the Federal Reserve System. He is no
stranger to this Committee. Welcome, Governor. The Honorable
John Hawke, Comptroller of the Currency, who has also spent a
lot of time with us. The Honorable Donald Powell, Chairman of
the FDIC, who has been here many times. And Honorable James
Gilleran, Director of the Office of Thrift Supervision.
Gentlemen, we welcome all of you. All of your statements
will be made part of the record in their entirety. You proceed
as you wish. Governor Ferguson, we will start with you.
STATEMENT OF ROGER W. FERGUSON, JR.
VICE CHAIRMAN, BOARD OF GOVERNORS
OF THE FEDERAL RESERVE SYSTEM
Mr. Ferguson. Thank you, Mr. Chairman, and also Senator
Sarbanes and Members of the Committee. It is certainly a
pleasure to be here before you today on behalf of the Board of
Governors to discuss Basel II, which, as you have already
indicated, is the evolving New Capital Accord for
internationally active banks. I appreciate the fact that the
full statement will be made part of the record.
Under Basel II, most of the banks in this country will
remain under the current capital regime. The operations of the
vast majority of U.S. banks do not require the full panoply of
sophisticated risk management techniques involved in the
advanced versions of Basel II. And the simpler versions of the
new accord will not provide our banks, with their current
supervisory and disclosure requirements, much additional
benefit.
We have an entirely different view of our largest and most
complicated banking organizations, especially those with
significant
operations abroad. A very important objective of the proposed
Basel II is to continue to promote consistency of capital
requirements for banks that compete directly in global markets.
And the supervisors of the industrial nations have agreed that
banks operating across national boundaries should be under
common capital standards to assure a competitive balance, and
that standard will soon be
Basel II.
Supervisors also want to encourage the largest banking
organizations in the world to continue to incorporate into
their operations the most sophisticated techniques for both the
measurement and the management of risk. Substantial difficulty
at any one of those entities could have significant effects on
global financial markets. In our view, prudential supervisors
and central bankers would be remiss if we did not address the
evolving complexity of our largest banks and ensure that modern
techniques were being used to manage their risks.
In the United States, the supervisory agencies will be
proposing that to meet these objectives, banks with large
foreign exposures and/or banks that are large and complex
should be in the set of core banks that would be required to
adopt Basel II. And the U.S. supervisors have concluded that
these banks should be required to adopt the advanced versions
of Basel II, the so-called Advanced Internal Ratings Based, or
A-IRB, approach for measuring credit risk and the Advanced
Measurement Approaches, or AMA, for measuring operational risk.
These approaches, which are described in more detail in my
written statement, are the most consistent with best practice
risk measurement and risk management.
Ten U.S. banks meet the proposed criteria to be core banks
and, thus, would be required under our proposal to adopt the A-
IRB and the AMA to measure their credit and operational risks,
respectively. We would also permit any bank that meets the
infrastructure requirements of the A-IRB and the AMA--that is,
the ability to quantify and develop the necessary risk
parameters on credit exposures and also to develop measurement
systems for operational risk exposures--to choose Basel II.
We anticipate that about 10 or so large banks now outside
of that core group that I have discussed would choose to adopt
Basel II in the near term. Thus, in total, we expect about 20
banks to adopt the advanced versions of Basel II before or
shortly after implementation date.
Now, let me turn to three issues that some have raised,
and, in fact, these reflect some of the comments that you two
raised, Mr. Chairman and Senator Sarbanes, in your opening
comments.
The first is competitive equity. While this concern takes
several forms, the most frequently voiced is the view that
competitive imbalance might result from the so-called
``bifurcated rules'' requiring Basel II for large banks while
applying the current capital rules for all other U.S. banks.
The fear in this regard is that banks that remain under the
current capital rules, with capital charges that are not as
risk-sensitive as those in Basel II, might be at a competitive
disadvantage compared to the Basel II banks that would get
lower capital charges on less risky assets.
We take this concern seriously and will be exploring it
through the upcoming Advanced Notice of Proposed Rulemaking.
But without prejudging the issue, there are some reasons to
believe that little, if any, competitive disadvantage will be
brought to those banks remaining under the current capital
regime.
The basic question here is the role of minimum regulatory
capital requirements in the determination of the price and the
availability of credit. Our understanding of bank pricing is
that it starts with the capital allocations that the banks
themselves make internally within their own organizations, then
factors in explicit recognition of the riskiness of credit, and
is then further adjusted on the basis of market conditions and
local competition from bank and nonbank sources. In some
markets, some banks will be relatively passive price takers. In
either case, regulatory capital, which is what Basel II deals
with, in particular, regulatory minimum capital, is mostly
irrelevant in the pricing decision and, therefore, unlikely to
cause competitive disparities.
Moreover, most banks, and especially the smaller ones,
today hold capital far in excess of the regulatory minimums for
a variety of reasons. Thus, changes in their own or their
rival's minimum regulatory capital due to Basel II generally
would not have much effect on the level of capital that they
choose to hold and would, therefore, not necessarily affect
internal capital allocations, which are the allocations that
drive, in part, pricing decisions.
Finally, the banks that most frequently express a fear of
being disadvantaged by a bifurcated regulatory regime have for
years faced capital arbitrage from larger rivals, who are able
to reduce their capital charges by securitizing loans for which
the regulatory charge was too high relative to either the
market or economic capital charge. The A-IRB approach would
provide, in effect, risk-sensitive capital charges for lower-
risk assets that are similar to what the larger banks have for
years already obtained through capital arbitrage. In short,
competitive realities between banks might not change in many
markets in which minimum regulatory capital charges would
become more explicitly risk sensitive.
Let me repeat that I do not mean to dismiss competitive
equity concerns at all. Indeed, I hope that the comments on the
Advanced Notice of Proposed Rulemaking, the ANPR, might bring
forth insights and analyses that respond directly to the
issues, particularly the observations I have just made. But, I
really must say that, we need to see reasoned analysis and not
just assertions.
A second area of concern is the proposed Pillar 1 treatment
of operational risk. Operational risk refers to losses from
failures of systems, controls, or people. Capital charges for
such risks have been implicit under Basel I for the last 15
years. These risks will, for the first time, be explicitly
subject to capital charges under the Basel II proposal.
Operational disruptions have caused banks to suffer huge
losses and, in some cases, failure both here and abroad. In an
increasingly technologically driven banking system, operational
risks have become an even larger share of total risk. Frankly,
at some banks, they are probably the dominant risk. To avoid
addressing them would be imprudent and would leave a
considerable gap in our regulatory system.
The AMA for determining capital charges on operational risk
is a principles-based approach that would obligate banks to
evaluate their own operational risks in a structured but
flexible way. Importantly, a bank could reduce its operational
risk charge by adopting procedures, systems, and controls that
reduce its risk or by shifting the risk to others through
measures such as insurance.
Some banks for which operational risk is the dominant risk
oppose an explicit capital charge and would prefer that
operational risk be handled case-by-case through the
supervisory review of buffer capital rather than be subject to
an explicit regulatory capital charge. The Federal Reserve
believes that would be a mistake because it would greatly
reduce the transparency of risk and capital that is such an
important part of Basel II, and it would make it very difficult
to treat risks comparably across banks.
The third concern I would like to discuss is the fear that
the combination of credit and operational risk capital charges
for those U.S. banks that are under Basel II would decline too
much for prudent supervisory purposes. Speaking for the Federal
Reserve Board, let me underline that we could not support a
final Basel II that we felt caused capital to decline to unsafe
and unsound levels at the largest banks. There will be several
stages before final implementation at which resulting capital
levels can and will be evaluated. At any of those stages, if
the evidence suggested that capital were declining too much,
the Federal Reserve Board would insist that Basel II be
adjusted or recalibrated, regardless of the difficulties with
bankers here or abroad or with supervisors in other countries.
But let us keep this in mind. Supervisors can maintain the same
level of average capital in the banking industry, either by
requiring each bank to maintain its current Basel I capital
level or by recognizing that there will be divergent levels
among banks dictated by different risk profiles.
To go through the process of devising a more risk-sensitive
capital framework, just to end, bank-by-bank with the same
Basel I
results, I think would be pointless. There will be some greater
dispersion and greater dispersion in required capital ratios,
if that
reflects underlying risk, is an objective and not a problem to
be
overcome.
Of course, I should add that capital ratios are not the
sole consideration. The improved risk measurement and
management and its integration into the supervisory system
under Basel II are also critical to ensure the safety and
soundness of the banking system.
We are now in the middle of the comment period for the
third Basel Consultative Paper, and next month we will begin
our comment period on the agencies' Advanced Notice of Proposed
Rulemaking. The comments on the domestic rulemaking as well as
on the third Consultative Paper will be critical in developing
the negotiating position of the U.S. agencies and highlighting
the need for any potential modifications in the proposal. The
U.S. agencies are committed to careful and considered review of
the comments received. The record already underlines that
comments and dialogue with bankers has had a substantive impact
on the Basel II proposal, and that will continue. But at this
stage of the proposal, comments that are based on evidence and
analysis are most likely to be effective.
In conclusion, the Basel II framework is the product of
extensive multiyear dialogues with the banking industry
regarding evolving ``best practice'' risk management techniques
in every significant area of banking activity. Accordingly, by
aligning supervision and regulation with these techniques, it
provides a great step forward in protecting our financial
system and that of other nations to the benefit of our
citizens.
We now face three choices: we can reject Basel II, we can
delay Basel II as an indirect way of sidetracking it, or we can
continue the domestic and international process using the
public comment and implementation process to make whatever
changes are necessary to make Basel II work effectively and
efficiently. The first two options require staying with Basel
I, which is not a viable option for our largest banks. The
third option recognizes that an international capital framework
is in our self-interest since our institutions are the major
beneficiary of a sound international financial system. The Fed
strongly supports that third option.
I will be happy to respond to your questions.
Chairman Shelby. Mr. Hawke.
STATEMENT OF JOHN D. HAWKE, JR.
COMPTROLLER OF THE CURRENCY
U.S. DEPARTMENT OF THE TREASURY
Mr. Hawke. Chairman Shelby, Senator Sarbanes, and Members
of the Committee, thank you for inviting the Office of the
Comptroller of the Currency to participate in this important
hearing.
I want to assure the Committee that the OCC, which has the
sole statutory responsibility for promulgating capital
regulations for national banks, will not sign off on a final
Basel II framework for U.S. banks until we have determined,
through our domestic rulemaking process, that any changes to
our domestic capital regulations are practical, effective, and
in the best interests of the U.S. public and our banking
system.
My written testimony provides a detailed discussion of the
background and content of Basel II, and I appreciate, Mr.
Chairman, that it will be included in the record in its
entirety. It discusses the important issues with which this
Committee is properly concerned. I would like to use this time
before the Committee today to make four important points that
may help to put today's testimony in proper focus.
First, all of the U.S. banking agencies share a concern
about the potential effect of Basel II on the capital levels of
large U.S. banks. Our banking system has performed remarkably
well in difficult economic conditions in recent years. I
believe that is due in substantial part to the strong capital
position our banks have maintained. While a more risk-sensitive
system of capital calculation might be expected to have the
effect of reducing the capital of some banks, we would not be
comfortable if the consequence of Basel II were to bring about
very large decreases in required minimum capital levels. By the
same token, if Basel II were to threaten significant increases
in the capital of some banks, it could undermine support for
the proposal and might threaten the competitiveness of those
banks. As things stand today, we simply do not have
sufficiently reliable information on the effect of these
proposals on individual institutions or on the banking industry
as a whole. Before we can make a valid assessment of whether
the results are appropriate and acceptable, we have to know, to
a much greater degree of reliability than we now have, just
what the results of Basel II will be.
The OCC believes that significant additional quantitative
impact analysis will be necessary. Ideally, this should take
the form of another study by the Basel Committee itself. But
even if the Basel Committee does not undertake such an
additional study, I believe that it is absolutely essential
that the U.S. agencies make such an assessment prior to the
adoption of final implementing regulations. I strongly believe
that we cannot responsibly adopt final rules implementing Basel
II until we have not only determined with a high degree of
reliability what the impact will be on the capital of our
banks, but have made the judgment that the impact is acceptable
and conducive to the maintenance of a safe and sound banking
system in the United States.
I believe all of the U.S. banking agencies share that
objective, and we expect to work closely together to resolve
any open issues.
Second, some have perceived there to be significant
differences among the U.S. banking agencies in their approach
to Basel II and have suggested that some external mechanism is
needed to resolve such differences. I do not think that is a
correct conclusion.
On the contrary, I believe the agencies have worked
exceedingly well together on this project for the past 4 years
and will continue to do so. To be sure, we have not always
agreed on every one of the multitude of complex issues that
Basel II has presented, but that is no more than one would
reasonably expect when a group of experts have brought their
individual perspectives to bear on difficult issues. Where
there have been differences, we have worked our way through
them in a highly professional and collaborative manner.
The Advance Notice of Proposed Rulemaking (ANPR) for
implementation of Basel II in the United States that the
agencies will soon jointly issue is another example of a highly
collegial and collaborative process. Our staffs have been
laboring together diligently to get us prepared for this first
round of rulemaking. In
addition, we are now in the final stages of internal review on
draft interagency guidance that we will jointly issue
concurrently with the ANPR to clarify and elaborate on our
expectations for those of our banks that will be subject to
Basel II, and that guidance has been developed in a process in
which every agency had substantial input. While reaching
agreement on some of the proposed requirements was no small
feat, I believe that every agency will concur with the outcome.
Considerable consultation and deliberation still lie ahead
before we can even consider final adoption of the implementing
regulations. But I have every confidence that the agencies will
continue to approach the issues in the same constructive and
cooperative spirit that has prevailed up to now.
Third, as I said earlier, I believe we are all committed to
a process that has real integrity to it. The current Basel
Committee timeline presents a daunting challenge to both the
U.S. banking agencies and to the banking industry. While it is
clearly necessary to address the acknowledged deficiencies in
the current Basel Accord, the banking agencies must better
understand the full range and scale of likely consequences
before finalizing any proposal. We have identified in our
written testimony the milestones that the agencies must meet
under the current Basel II timeline. They include: Basel
Committee consideration of comments received by it
on its latest Consultative Paper; the issuance of an ANPR and
draft supervisory guidance in the United States with a 90-day
period for comments, which we expect to issue in mid-July; full
consideration of those comments; the issuance of a definitive
paper by the Basel Committee; the drafting and issuance for
comment in the United States of a proposed regulation
implementing the final Basel paper; the conduct of a further
quantitative impact study, as I have just mentioned;
consideration of the comments received on the NPR; and,
finally, the issuance of a definitive U.S. implementing
regulation.
Each of these steps is critical in a prudential
consideration of Basel II in the United States, and the
agencies will be working closely together at every step. I
anticipate that we will also be working in close communication
with committees of the Congress.
If we find that our current target implementation of
January 1, 2007, is simply not doable--and my personal opinion
is that realization of that target may be very difficult--we
will take more time. But it is too early to draw that
conclusion yet. The important point is that we will take great
care not to let the time frame shape the debate. Equally
important is that the time frame will be secondary to our
responsibility to fully consider all comments received during
our notice and comment process. If we determine through this
process that changes to the proposal are necessary, we will
make those views known to the Basel Committee, and we will not
implement proposed revisions until those changes are made.
Finally, some have viewed the New Basel II approach as
leaving it up to the banks to determine their own minimum
capital, or, as some have said, putting the fox in charge of
the chicken coop. This is categorically not the case. While a
bank's internal models and risk assessment systems will be the
starting point for the calculation of capital, bank supervisors
will be heavily involved at every stage of the process. We will
publish extensive guidance and standards that the banks will
have to observe. We will not only validate the models and
systems but will assure that they are being applied with
integrity.
In my view, the bank supervisory system that we have in the
United States is unsurpassed anywhere in the world in both its
quality and in the intensity with which it is applied, and we
are not going to allow Basel II to change that. In fact, if we
do not believe at the end of the day that Basel II will enhance
the quality and effectiveness of our supervision, we should
have serious reservations about proceeding in this direction.
Moreover, while Basel II has largely been designed by
economists and mathematicians, and while these so-called
``quants'' will play an important role in our oversight of the
implementation of Basel II, the role of our traditional bank
examiners will continue to be of enormous importance. Such
values as asset quality, credit culture, managerial competence,
and the adequacy of internal controls cannot be determined by
mathematical models or formulas. Nor can many of the risks that
banks face be properly evaluated except by the application of
seasoned and expert judgment. I can assure you that those
national banks covered by Basel II will continue to be closely
monitored and supervised by highly qualified and experienced
national bank examiners who will continue to have a full-time,
on-site presence. The new process will not replace them. It
will simply give them even better tools to assess the true
nature and measure of the risks confronting the banks for which
they are responsible.
I am pleased to have had this opportunity to provide our
views, Mr. Chairman, on this important initiative, and I would
be happy to answer any questions.
Chairman Shelby. Mr. Powell.
STATEMENT OF DONALD E. POWELL
CHAIRMAN, FEDERAL DEPOSIT INSURANCE CORPORATION
Chairman Powell. Thank you, Mr. Chairman and Members of the
Committee. I appreciate the Committee's interest in the New
Basel Capital Accord.
I believe that Basel II ranks among the most important
pieces of proposed banking regulation in our Nation's history.
The FDIC supports the goal of lining up capital regulation with
the economic substance of risks that banks take. Basel II
encourages a disciplined approach to risk management, and it
addresses important weaknesses in our current capital rules. We
applaud the intense and prolonged efforts that have been made
to address these important issues. We are approaching a
crossroads where judgments will need to be made on some
critical issues. We have an interagency process and a public
comment period to help reach those judgments, and I am
confident that our process will result in an appropriate
outcome. My written testimony provides a broad overview of some
of the critical judgments that will need to be made before the
agencies commit to adopt Basel II in the United States.
The first key issue is capital adequacy. The Basel II
formulas allow, at least in principle, for significant capital
reductions. The proposals issued by the Basel Committee specify
that after a phase-in period, there would be no floor on the
level of risk-based
capital that banks would be required to hold. The level of
risk-based capital that banks actually hold would depend upon
their own internal estimates of risk--validated by their
supervisors--and on the demands of the marketplace. It is very
difficult to predict the ultimate effect of Basel II on overall
bank capital, and we do know that the formulas are forceful
tools for affecting risk-based capital requirements.
There is no question that Basel formulas would help the
regulators differentiate risk. The formulas cannot stand on
their own. Banks face other risk besides credit risk and
operational risk. Lending behavior can change over time,
causing losses to escalate in activities perceived as low risk.
The fact is that no one knows what the future holds. For these
and other reasons, the FDIC believes that Basel II must be
supplemented by the continued application of existing
regulatory minimum leverage capital and prompt corrective
action requirements. I am very gratified at the support our
fellow bank regulators have expressed for this conclusion.
We also understand that a leverage ratio alone cannot
provide protection without the support of sound risk-based
capital rules. It will be necessary to better understand the
impact of the proposals on the capital required for specific
activities.
Maintaining capital adequacy under Basel II would be an
ongoing task. Validating banks' internal risk estimates would
be a challenge. Doing so consistently across agencies would be
a greater challenge for which an interagency process would be
needed. The other key issue is competitive equity. Basel II has
been expected to provide some degree of regulatory capital
relief. The banks that stand to be directly affected by Basel
II have expressed strong support for such capital relief. They
have expressed concerns where they believe Basel II capital was
too high. The key policy question is: What economic benefits
and costs would come with changes in regulatory capital
requirements? Would the economic benefit of lower-risk-based
capital requirements for large banks enhance their competitive
posture or accelerate industry consolidation?
We recognize there are differences of opinion about the
importance of competitive equity issues. That is why we need to
pay close attention to the comments we receive on this issue.
The agencies received a number of comments on both sides of
this issue at a recent industry outreach meeting and this
dialogue will continue.
In short, the ingredients of the success of Basel II
continue to be: Appropriate minimum capital standards; a
consistent approach to validating banks' risk estimates; an
adequate vetting of competitive issues; and, time to address
these and other policy issues as we finalize our views on this
Accord.
We will continue to work closely with our fellow regulators
to work through these important issues and reach the right
conclusions. We are committed to evaluating the costs and
benefits of the Basel II proposals and their impact on the U.S.
banking industry and the safety and soundness of the financial
system.
Thank you for the opportunity to present the views of the
FDIC.
Chairman Shelby. Mr. Gilleran.
STATEMENT OF JAMES E. GILLERAN
DIRECTOR, OFFICE OF THRIFT SUPERVISION
Mr. Gilleran. Chairman Shelby, Ranking Member Sarbanes, and
Members of the Committee.
My fellow regulators have outlined very well, I believe,
all of the risks that are inherent in Basel II, so just a
couple of personal comments. I think that Basel II has moved
the ball along very well in trying to think through the
question of risk versus capital, and I think it has been a real
contribution for all of us who are dealing with capital all the
time.
When you look at Basel I, you look at an idea that is very
simple. Everyone understands it. It is applied to all, and it
has yielded what everybody concludes is capital which seems to
be sufficient. In fact, we have come off of just 2 years of
almost the best results the financial services industry has
ever had, and this year looks very well, so the capital levels
do not seem to have impeded the results in the industry.
Basel II, while having a step forward in thought process,
concludes with a very complicated, detailed system that applies
to only a few, and everybody believes will result in lower
capital. Therefore, I think that your interest in this subject,
and I believe the work that has been done by my fellow
regulators, must continue so that we are all very sure that
what we are doing in this next step will be safe and sound.
Thank you very much.
Chairman Shelby. Thank you.
Governor Ferguson, first I am going to read an excerpt from
the June 4, 2003 Wall Street Journal citing a recent Mercer
Oliver Wyman study.
``The new accord will also change the amount of regulatory
capital required from banks for different types of lending,
meaning that certain types of products such as mortgages,
lending to large corporations, and leasing will receive less
back-up capital and could potentially be more profitable. Other
types of lending, such as project financing and lending to
small businesses will require more capital and so could become
less profitable. Oliver Wyman expects this will lead banks to
move away from products they cannot thrive in.''
Governor, looking at the results of this study as reported
in The Wall Street Journal on June 4, I would conclude that the
new accord will result in some winners and losers here, so to
speak, in different sectors of the economy. In other words,
this proposal could produce significant shifts in the economy.
Inasmuch, it could be said that we are not merely talking about
reform of banking regulation alone, we are talking about in a
sense, some people believe, a proposal that involves policy
decisions that will have significant macroeconomic effects.
What is your take on that?
Mr. Ferguson. I would not reach that conclusion based on my
understanding of Basel II.
Chairman Shelby. Do you believe that there will be winners
and losers? There are always winners and----
Mr. Ferguson. By definition we are changing, if this goes
through, we are changing the status quo, and there will be
those who like that and those who do not. I do not think of any
of them as winners and losers, because what we are trying to do
under Basel II is to have regulatory capital that better
reflects risk, and I see no losers in that process.
Chairman Shelby. Do you see small business as a possible
loser? This is what had turned this economy for years is the
hiring of probably 75 to 80 percent of our people. And if they
are not in the play like they have been, this could have a very
significant impact on our economy in the future.
Mr. Ferguson. I do not think small businesses will be a
loser in the sense you are talking about.
Chairman Shelby. Why?
Mr. Ferguson. The reason is as follows. First, I am not
sure that Oliver Wyman had it just right on what is going to
happen in terms of risk-based capital for different portfolios.
I would set that to one side. These things are still being
calibrated, et cetera.
Second, and more importantly, what we have seen is that
small businesses have found a very large number of sources of
capital or lending, some from large banks, some from small
banks. None of that is being driven, I believe, by regulatory
minimum capital. I think it is driven by a broader range of
issues and regulatory minimum is relatively unimportant in that
panoply of things that drive the decision of banks to lend to
small businesses.
Chairman Shelby. Governor, do you have some studies that
you can share with the Committee that would back up what you
are saying here with some analysis, not just your opinion but
some other analysis?
Mr. Ferguson. That is a fair question. I do not have
studies at this stage. What we will be doing though in the ANPR
is asking this question to get a greater sense of the
information that you have just talked about. In fact, broadly
speaking, the entire issue with respect to competition and
different portfolios will be heavily examined under the ANPR,
which will give us a better fact base, because I am looking at
the same issue you are. As I said in my opening remarks, I do
not think, based on my analysis and based on conversation with
bankers, that anything other than economic capital and market
structure are the kinds of things that drive these decisions.
It is not, based on my current analysis, the regulatory minimum
that determines pricing or availability.
I would also observe that the thing that has been driving
small business lending has been frankly more a question of
technology, and to some degree, pricing. As I have been on the
Board and observed what is happening over the past many years--
--
Chairman Shelby. In what way? Could you explain?
Mr. Ferguson. Yes. I will explain both. What has happened
in small business is that it has historically been, for many,
many years the purview of local bankers, and those community
bankers still have a very strong role that they are playing
with respect to lending to small businesses.
Chairman Shelby. They are the small business lender.
Mr. Ferguson. Absolutely. That linkage still seems to hold
true.
Chairman Shelby. But will it hold true in the future, like
my colleague--I know my time is nearly up and we will have
another round, but Senator Sarbanes raised that question
earlier, alluded to that question.
Mr. Ferguson. Right, he alluded to it.
Chairman Shelby. It is very important.
Mr. Ferguson. It is very important. I believe----
Chairman Shelby. These are not accords that we should even
look at in a cursory manner, as the Comptroller of the Currency
said. The Basel II Accord has deep and broad ramifications.
Mr. Ferguson. That is exactly why we are here today, and
that is why we commend you for holding this hearing.
I think it will continue to hold true going forward,
because I do not believe that banks choose to get into or out
of a business based on regulatory minimum capital. Indeed, we
are trying to structure regulatory minimum capital so that it
does not impact the strategic decisions of a bank to get into
or out of a business.
Chairman Shelby. My time is up, but I want to get this in,
and I will turn to Senator Sarbanes. Governor, who is
responsible here? In other words, this is a very basic
question, but I think it is an important practical concern, at
least of mine, perhaps other Members. I would like to know who
is taking the sum total of this proposal, has digested it, and
has a firm grasp of its entirety? I mean this is far reaching.
Who has?
Mr. Ferguson. Personally, I have been working with this for
a year and a half. I have been heavily briefed.
Chairman Shelby. I know you have. We have talked about it.
Mr. Ferguson. I have worked through lots of the details. I
believe I have a good grasp of the vast majority of the issues.
There are a few issues here that I would want to understand a
little better. I think in each one of our agencies, each would
give perhaps a different answer.
Chairman Shelby. We have some regulators here, the
Comptroller, the FDIC Chairman, OTS Director, that voice other
concerns, and they should. I would like to know who is
ultimately
responsible for this and who is accountable for a success or
failure. Is it Chairman Greenspan? Is it the Fed? Is it the
Comptroller? Is it the FDIC Chairman? Is it the OTS Director?
Mr. Ferguson. You have before you a panel of the regulators
of the major depository financial institutions. I think we have
a collective responsibility, as both Comptroller Hawke and
Chairman Powell have indicated, to work very closely together.
Yes, these are tough issues and there are places where by
definition----
Chairman Shelby. Profound issues, are they not?
Mr. Ferguson. Absolutely profound. There are places where
we will initially disagree. We will come to a middle ground
that we all agree on, that we think is in the best interest of
the United States, and we have, I believe, a collective
responsibility to you and to the country to understand these
implications of these profound changes, and to give you our
best judgment on the impact that they will have. That is one of
the reasons why I support what the Comptroller said with
respect to the need to have ongoing quantitative impact studies
so that we have a factual evaluation of the impact that this is
likely to have in the U.S. banking industry.
Chairman Shelby. We have this group, an ad hoc group of
regulators, who have convened in Switzerland over the last few
years. They put this together. Ultimately, I believe that this
Committee, the Committee of jurisdiction in the Senate; as
Senator Sarbanes said, capital is very important. A lot of us
have been on this Committee a long time, where we have visited
the taxpayer on things where there was not sufficient capital.
This is going to lower some capital standards. It is going to
concern me and other Members.
Mr. Ferguson. As I have said, I would not jump to
conclusions yet that it is going to lower capital overall in
the industry. I think it will create greater dispersion of
capital because there will be those banks that need to have
higher regulatory minimum capital, and there will be some that
will need to have lower regulatory minimum capital. When this
Basel Committee got started in this process, the goal was to
keep capital at about the same level as it is now, and thus
far, first indications from early quantitative studies suggest
that indeed if it lowers capital, it will be a very small
amount, maybe about 6 percent or so. But as I said in my
opening remarks, and I think all of us would say the same, if
we believe, based on looking at the quantitative impact studies
that we do going forward, that capital is lowered to levels
that are unsafe and unsound, then we will have to go back and
renegotiate, we will recalibrate, and I am sure your Committee
would want to hear that we will do that. You have heard my
commitment that the Fed believes that, and I think that is
consistently shared across all the regulators before you.
Chairman Shelby. Senator Sarbanes.
Senator Sarbanes. Thank you, Mr. Chairman.
Am I correct that the schedule for concluding the
international agreement in Basel is the end of this year?
Mr. Ferguson. That is the current schedule.
Senator Sarbanes. Six months away?
Mr. Ferguson. That is the current schedule, yes.
Senator Sarbanes. We have the extent of concern or
question, if you do not want to label it disagreement, amongst
the legislators as reflected in their statements this morning,
and we are only 6 months out; is that right?
Mr. Ferguson. If I can continue to respond for myself, and
allow them to speak. I would say, what you have heard from the
regulators I would not describe as disagreement. I think we
understand what the issues are.
Senator Sarbanes. You may disagree on what should be done
about them.
Mr. Ferguson. I think we will reach an agreement on what
should be done. This is certainly profound without question,
but I think we will continue as we have to date to have an
honest discussion among ourselves of what the options are and
to develop a middle ground. I share some of the nature of your
concern, and I share what Comptroller Hawke had to say. We have
to give ourselves sufficient time in the comment period before
we go back and do the final negotiations.
Senator Sarbanes. Let me go to some of the basic concepts.
The Economist, on March 29 of this year, in an article said,
amongst other things, the following: ``American regulators
intend to apply the new rules to fewer than a dozen other
banks. This is a choker for European regulators who see Basel
II like Basel I, as a global standard to be applied to all
banks.''
That raises the first question. What is the rationale for
applying Basel II to a limited number of banks?
Mr. Ferguson. You are addressing that question to me,
Senator?
Senator Sarbanes. I do not know. Anyone on the panel who
wants to answer it.
Mr. Hawke. Let me give my colleague a rest.
Mr. Ferguson. Thank you.
[Laughter.]
Mr. Hawke. I think the basic premise, Senator Sarbanes, in
Basel I, as in Basel II, was that these accords were intended
to apply to internationally active banks in order to achieve
competitive equality on the international scene. The Basel
Committee has played a valuable role in trying to establish a
framework for capital that can be adopted by any banking system
anyplace in the world.
Senator Sarbanes. Did Basel I apply to all banks or a
limited number of banks?
Mr. Hawke. We voluntarily decided to apply Basel I to all
of our banks.
Senator Sarbanes. To all banks.
Mr. Hawke. But the premise in Basel I, as in Basel II, was
that it was intended to be applied to internationally active
banks. And the 10 or 12 banks that we intend to apply it to
constitute probably 95 percent or more of the international
exposures of U.S. banks. So, by selecting that group, we have,
I think, kept faith with the premise of the Basel Committee. We
do not see any useful purpose to be served in applying Basel II
to the thousands of smaller banks that we have in the United
States which are already better capitalized than their larger
counterparts, and, I think, significantly more intensively
regulated than comparably sized banks anyplace else in the
world.
Senator Sarbanes. I gather one concern is that this will
now enable their larger counterparts to have less capital vis-
a-vis the smaller banks. Who determines which banks Basel II
will apply, to whom it will apply; who determines that?
Mr. Hawke. We have, through a joint interagency process,
established guidelines that look both at the size of the bank
and the extent of its international exposures, and it was
through that process that we jointly selected the standard that
resulted in 10 or 12 of our largest banks being included in the
mandatory category.
Senator Sarbanes. Is it also the case that another 10 or 12
banks can voluntarily subject to the Basel Accord?
Mr. Hawke. We did not include any numerical limit. Any bank
that can establish that it has the capacity to implement the
systems that would be required can apply for regulatory
approval to come under Basel II.
Senator Sarbanes. So, you can calculate what advantage it
will give you and decide to have that apply to you; is that
correct?
Mr. Hawke. That could be the case, yes.
Senator Sarbanes. Well, what kind of regulatory system is
that? Let me ask this question. Is it the case that the models
to determine the risk will be internal models of the banks
themselves?
Mr. Hawke. As I mentioned in my statement, Senator
Sarbanes, the banks will develop the models, but under very
carefully defined guidelines that we will set forth for them,
and under intense validation procedures that we will follow,
and under continuing supervision by us with respect to the
application of those models.
Senator Sarbanes. But they will develop the models
themselves.
Mr. Hawke. The models would be developed by the banks
subject to our validation and continuing oversight.
Senator Sarbanes. That is a pretty tricky thing, is it not?
Mr. Hawke. There are people who have expressed reservations
about that, the fox-guarding-the-chicken-coop syndrome, but I
do not think that is the case. First, as I said in my
statement, the basic nature of bank supervision is not going to
change. We are going to have full-time, on-site examiners in
all of these large banks, just as we do today. Second, a new
breed of people will be coming into the banks--mathematicians
and economists who helped construct these rules--and they will
be validating the models and helping our examiners to oversee
the integrity of the application of the models. While
intuitively many of us have had exactly that concern, I do not
think this amounts to turning capital calculation over to the
banks themselves.
Senator Sarbanes. Another concern I have heard is the
uncertainty of the impact of the proposed agreement on major
U.S. financial institutions. Let me leave aside for a moment
now the bifurcation, right, that it is only going to apply to a
limited number of financial institutions, with others able to
choose to have it apply if they decide it is to their
advantage. I understand that three tests have been done to
measure the impact of the new agreement on the largest U.S.
financial institutions, and that no clear picture of the impact
of the agreement on the capital held by the largest U.S. banks
has emerged. In fact, and I want to just verify this factually,
that the latest test shows the capital of some U.S.
institutions rising by 40 percent and the capital at other U.S.
institutions declining by 35 percent. Is that correct?
Mr. Hawke. I think that is an accurate reflection of what
the so-called QIS-3 reported, but I think QIS-3 was severely
flawed. That is why we are insisting on a subsequent
quantitative impact study that is overseen by the regulators
and carefully monitored. That range would give me enormous
concern.
Senator Sarbanes. I would hope so.
Mr. Hawke. At both ends of the spectrum.
Mr. Ferguson. May I add one other point here, sir, on that?
I think you should also recognize that what you are talking
about I think is just the credit risk component of capital.
There is also an operational risk component as well to this
whole accord, which in most cases I think would end up reducing
some of that range that you just talked about. Again, you also
want us to ask the question of to what degree, if this is
calibrated correctly, are we reflecting underlying risk? I
agree with Comptroller Hawke that obviously going forward we
will need to have more of these studies to try to continue to
get a better handle on these issues that you are raising.
Senator Sarbanes. My time is up. Before I close, Mr. Powell
and Mr. Gilleran, you may want to add some observations on this
discussion we have just been having here with Mr. Hawke and Mr.
Ferguson.
Mr. Powell. I think, Senator, your comments call to
attention a strong view that the FDIC has, and that is, for
minimum regulatory capital. That is very important to us at the
FDIC. While I can understand that there are those in the
marketplace who believe this is an outmoded ideal, I think it
has served us well in good times and bad times over the years.
I would also make the comment that economic capital and
market capital for the last 13 years has exceeded regulatory
capital. While these models are wonderful, I think the
discussion that we have had here shows some of the
inconsistency in them. Now, I
believe in them, and we support that model, so we support that
capital should be based upon risk, but we strongly support also
minimum capital.
Senator Sarbanes. Mr. Gilleran.
Mr. Gilleran. Senator, I think that we also have probably
misjudged the number of financial institutions that will be
making application to have Basel II apply to them, and even
though initially it is believed that just a few large
organizations will apply because of the complexity and the need
for computer programming and for a history to be able to
demonstrate that the programs work, I believe that there are
many consulting firms queuing up out there to be able to
provide the data and the support to much smaller institutions
so they can make application to use Basel II concepts. So, I
believe the regulators will have many more organizations' plans
to look at than some expect.
Chairman Shelby. Senator Allard.
Senator Allard. Thank you, Mr. Chairman.
Basel II's third pillar has a requirement that banks not
only calculate the risk positions in capital requirements, but
it also ensures that the calculations are disclosed for review
by the markets. I personally like to have disclosure for the
markets because I think the more informed the consumers are,
the better. I am curious if you have had any concerns brought
to your attention about those disclosure requirements? I know
that in a competitive environment banks sometimes have certain
information that is proprietary in nature that they do not want
their competitors to know. Have you received any comments in
that regard?
Mr. Ferguson. We have been working this for many years, and
in fact this pillar you are referring to, Pillar 3, has evolved
over time. It is clearly not intended to have any disclosure of
competitively proprietary information. It is intended to have
disclosure of information that the market should know. This
disclosure, by the way, is part of the answer to some of the
questions that have been raised before by Senator Sarbanes,
because it is not just simply our judgment and validation of
the inputs, but the market will be able to look at the inputs
to some degree and make some comparisons under Pillar 3. So, we
worked very hard to make sure that the disclosure regime under
Pillar 3 avoids the kind of competitive issues that you have
talked about, while being sufficient to give the market the
information that it needs to help us do the work that we need
to do with respect to validations.
Mr. Hawke. If I could just add one quick point to that,
Senator Allard. One important function of Pillar 3, in addition
to providing the market with information, is to provide a basis
for supervisors and banks in different countries to help assure
that Basel II is being applied even-handedly by supervisors in
different countries. That kind of transparency is very
important to the integrity of the
process.
Senator Allard. I would think that would be very important
in that regard.
Let's say we have passed what you have recommended, and the
regulations have become effective and everybody agrees on them.
How do you make the transition from a Basel I to a Basel II
bank? That is a hypothetical, but I think it is something that
would very possibly happen in the future with the consolidation
of the banks. Two or three banks could consolidate, and
suddenly they are big enough to compete in the international
market. How do you transition them into that arena once you
have standards in place? What are your plans for doing that?
Maybe the Fed is making appropriate plans, or the other members
of the panel have some thoughts on that. I am curious as to how
you see that happening.
Mr. Hawke. Beyond that group of what we call the mandatory
Basel banks, the 10 or 12 that we will explicitly subject to
Basel, there may be 20 or more other banks that initially or
over time will come to us and try to demonstrate that they have
the systems capacity and the sophistication to opt in to Basel
II. It will start with a judgment on the part of the bank as to
whether they think it is to their advantage, either in terms of
market perception or for other reasons, to come under the Basel
II regime. We will have to determine that they are capable of
developing the systems that are going to be necessary and that
they have the kind of risk management and risk measurement
capacity that will be necessary to deal with the Basel II
requirements. We will make that judgment on a bank-by-bank
basis.
Senator Allard. Mr. Gilleran.
Mr. Gilleran. Senator, if a bank is rated under Basel II
and it has been deemed to have excess capital then, then it has
to decide what to do with it. Do you either pay it out to your
shareholders as a dividend? Do you take on additional risk in
your portfolio to utilize the capital, or do you acquire other
financial institutions?
One of the things that may come out of the application of
Basel II to only larger banks is the fact that it might provide
a number of larger banks with excess capital to do many
acquisitions within the industry, and therefore, one of the
questions long range is whether or not there is a roll-up in
the financial services industry even greater than there has
been. This, Senator Shelby, I think has impact on small
business lending because if you then have a roll-up of
community banks, then you would probably have an impact on
loans to small businesses. So that is an impact.
On your question on disclosure, Senator Allard, the problem
that I have is that the information in connection with Basel II
provides a tremendous amount of information. I am unconvinced
yet though that all of the disclosures will be really
communicating the real true risk in the institution, and I
believe that is something that we have to be careful about
going forward, whether or not a lot of data is there, but not
enough disclosure of the real risks.
Senator Allard. I see my time is up.
Chairman Shelby. You go ahead, proceed.
Senator Allard. Well, just one other thought, and it was
brought up by Mr. Hawke in his response. Basel II will require
a significant investment in systems and training, particularly
in examiners. All of you have admitted that Basel II is much
more complicated than Basel I. What difficulties do you see in
both getting your examiners trained and acquisition of
equipment, if any?
Mr. Hawke. That process has been ongoing now for quite a
while. Our conventional safety and soundness examiners have
been involved in the process. They are intimately involved,
both within our agency and on an interagency basis, with the
implementation process. So there is a great deal of ground
laying that is going on. We are attempting to expand our staff
of quants, if you will, to deal with Basel issues as well. We
have all invested a lot in the preparations for Basel II.
Senator Allard. Thank you, Mr. Chairman.
Chairman Shelby. Senator Reed.
Senator Reed. Thank you very much, Mr. Chairman, and thank
you, gentleman.
Let me raise a series of questions. First, one of the key
aspects of the Basel II Accord is operational risk and it
raises two questions in my mind. First, whether a quantitative
approach, a certain level of capital, is the appropriate
response or a qualitative approach of actually evaluating the
systems that the institution has and making adjustments or
directions to the system is the best approach. Mr. Ferguson,
Mr. Hawke, might you comment on that, Mr. Powell and Mr.
Gilleran.
Mr. Ferguson. Well, I will start. We believe, and we have
seen some recent evidence, that it is indeed more possible now
than it was a few years ago to get a more quantified approach
to operational risk. We had at the Federal Reserve Bank of New
York, a week ago more or less, a conference that brought in
banks from around the world, each one of which disclosed for
the first time the approaches they have used, and we were
actually quite impressed to see the degree to which progress
has been made. What banks are doing these days--this is again,
leading edge banks, not all banks but leading edge banks--are
using both their internal operational loss information,
external databases, scenario analysis, information from some
insurance companies, for example, to come up with a very solid
quantified approach to operational risk. I can name a few
names. We have put the presentations on the website. So, I
think indeed our degree of confidence in the ability to do more
quantified approaches to operational risk has gone up, and it
is a sign of how quickly things have changed in the industry.
Obviously, one of the things that is done in that regard is to
think about the impact of procedures, systems, backup sites, et
cetera, on the probability of an operational problem having an
impact on the ongoing day-to-day operations of the bank. So it
does involve some management judgment as well.
What Basel II calls for is that the approaches to doing the
internal assessments of operational risk be things that we can
replicate, that they be systematic and not purely judgmental,
that they be based on data and facts that we can also evaluate.
The good news is, as I have said, things have moved very much
in that direction, and we have a pretty high degree--I think
all of us, though it might vary somewhat--a reasonably high
degree of confidence that indeed systems have moved in the
right direction to make this measure of operational risk more
quantifiable than it was, let us say, 5 years ago.
Senator Reed. Mr. Hawke.
Mr. Hawke. Senator, I have participated in the Basel
Committee for 4\1/2\ years. Throughout that entire period I
argued strenuously that operational risk should be left to
supervisory assessment under Pillar 2, a qualitative
assessment, because if you believe that operational risk
inheres in the strength of an institution's internal controls,
that is inherently a qualitative type of judgment. We, working
together with the Fed and the other agencies, pursued the
development of the Advanced Measurement Approach (AMA), which
is a Pillar 1 approach, to the calculation of operational risk,
and it has a very high degree of supervisory judgment involved
in that process.
I think whether we put operational risk under Pillar 1 or
Pillar 2 is not a consequential issue any more. I think that
even if it were under Pillar 2, we would still have to have a
framework for determining what kind of operational risk charge
we would apply. I think the AMA approach that we have developed
is flexible. It has a lot of supervisory discretion and
evaluation built into it, and I think that is where we need to
continue to work on perfecting it.
Senator Reed. Mr. Powell.
Mr. Powell. I just say supervisory oversight--I think the
Comptroller said it well--is the key to this whole operational
issue.
Mr. Gilleran. I believe that the Comptroller is correct,
that it really does not matter whether it is in Pillar 1 or
Pillar 2, as long as it is recognized. The interesting thing
about Basel I is that Basel I implicitly recognizes operational
risk, and that the Basel I system has worked pretty well, and
has recognized that risk because we have gone through several
years of tough economic times, and yet the banking system has
held up. Therefore the implicit recognition has been a valid
method.
Basel II does require this explicit recognition, which is a
whole different way of going about it, much more intellectual,
but it remains to be seen whether it is better.
Senator Reed. Thank you.
Chairman Shelby. Senator Corzine.
STATEMENT OF SENATOR JON S. CORZINE
Senator Corzine. Thank you very much, Mr. Chairman. I
welcome the witnesses.
I will just say from my own personal experience that this
is one of the most difficult topics that needs to be dealt
with, and the complexity is almost mind-boggling. I want to
understand what the process of the oversight of the self-
derived models are. How is that foreseen to be certain? Are you
going to certify different outside consultants or is it only
going to be an internally driven responsibility of the
regulators to look at the models that the organizations are
putting together, and how do you see that process?
The second question is, I think one of the most talked
about and thought about risks are derivative risks. Is there
any description on how that process, that you could share with
us, that is tractable, that would be considered? And if we have
time, I would love to hear what you think the international
implications of this are, competitive implications for
Americans, financial system on a relative basis, because there
are some elements in here on assessment of risk with regard
toward legal exposures and other things that are pretty
significant and I would like to understand how those fit.
But I will go back to the complexity issue. How are we
going to assess the processes in a way that it is fair,
consistent, in a way that people will feel like the system is
not rigged for those that brightest quants?
Mr. Hawke. Let me take the first question, Senator. First
of all, I can assure you we will not be certifying consultants.
That is not our intention at all. The validation process starts
with the Basel documents, which have very detailed standards
for risk measurement and risk management systems built into
them, and for the kinds of models that will pass muster.
Senator Corzine. How are those going to be tested?
Mr. Hawke. The second step will be the validation of those
models by our examiners and experts, who will come into the
banks and review the models, review their compliance with the
Basel standards and test them out. There will be an ongoing
process beyond that where our examiners and experts will look
at the application of the models to see how the models are
actually working in practice. This is going to be an ongoing
process that starts with the articulation of detailed standards
by the Basel Committee itself.
Let me defer to Governor Ferguson on the derivatives
question.
Mr. Ferguson. You asked a question on derivatives. You will
see, when we come up to the ANPR, that we are going to ask some
questions in that regard. The reality is, I think you will hear
later, to broaden your question a bit beyond derivatives, to
other products that are much more technically intensive, that
drive securitization, for example. We have a number of very
specific questions in the ANPR to try to get to some of the
issues that you are talking about. The reality is that this is
one of the areas that has been most complicated. We are
continuing to work on it. You will hear from some people on the
second panel, a range of issues and concerns. So at this stage
I would say it is very much still being focused on, because it
is, as you point out, very complicated.
Senator Corzine. So with regard to those specific areas, it
is still a work in progress.
Mr. Ferguson. It is still a bit of a work in progress. We
have a proposal, but all of these things are a work in
progress, and we are looking for comment. We have gotten
feedback from a number of the institutions. We have gotten
feedback from the Bond Market Association, for example, with
specific questions that we want to ask them to try to get some
better input into these matters, because it is something that
we still have to work through. This is, by the way, an area in
which the United States is the world leader, as you know. We
want to make sure that as we go down this path we do not
undercut our ability to continue to innovate in this regard,
and that we have a much better handle on the risk profiles that
are starting to emerge.
Mr. Hawke. If I could just add, Senator, we are about to
put out in mid-July, jointly, substantial supervisory guidance
on this whole internal ratings-based approach, and standards
for the models and our expectations for management capacity to
deal with those models. That is going out for comment in a
month or so. Right now it runs about 130 pages.
Senator Corzine. Is operational risk included in your
derivative formulation of measurement of risk?
Mr. Ferguson. It is not at this point. I think this is an
interesting question that we need to raise. We have thought of
derivative risk and securitization risk more down the path of
credit risk. Obviously, there are some associated legal
ramifications that might emerge that would fall into
operational risk, but at this stage we are thinking of them
separately. I think again it is something we can question. But
the reality, as you may have sensed, with respect to
operational risk in the areas of systems, regulations, et
cetera, is that it is already sufficiently complicated. We have
not then thrown in the derivative issues for a reason that I
think you would sympathize with and understand.
Mr. Powell. I was just going to comment on your earlier
question about validating the banks' risk models. I think it is
very important for consistency that that process should be
interagency, like the same approach that we have with the
Shared National Credits, because there obviously could be some
potential inconsistencies, so it should be an interagency team
effort.
Senator Corzine. And is there every intent on having that?
Mr. Powell. I think I have shared this with all of the
gentleman here at the table--there is a spirit of cooperation
here, and the intent would be that there is consistency, yes.
Mr. Gilleran. I think we have demonstrated over the last
several years the very good cooperative effort through the
FFIEC Act and through other regulatory issues that we are
capable of cooperatively dealing in most important areas. I
think that will happen.
I think the question you raise on the international impact,
I think goes to the heart of whether or not the international
regulation of banks is the same, and the United States clearly,
in my view, is the leader in bank regulation. I think that on
an ongoing basis, whether or not every country is looking at it
the same way that we are, as an important issue.
Senator Corzine. Certainly is in the long run a competitive
issue. Capital is important actually on how you run your
business. So if somebody gets a leg up, whether it is smaller
banks versus larger banks or Swiss banks versus United States
banks, whoever the primary supervisor is, how they look at
those various models can end up potentially leading to
disparities in how capital is applied to the various risks that
are being taken. I am all for risk-based capital standards. It
is just one heck of a complex issue.
Mr. Ferguson. May I respond to the international question?
We are well aware of the need to keep the regulatory community
looking at these issues pretty much in the same way. The Basel
Committee has put together a group called the Accord
Implementation Group that brings together the regulators to
share best practice and findings and try to create that sense
of commonality across borders. By definition, no one can
guarantee that any process is going to be foolproof, but we
have not left this issue unexamined. We have created this
entire process to get to some of the issues that you have
raised, Senator.
Senator Corzine. Thank you.
Chairman Shelby. Thank you, Senator.
The part of the proposal that will be applied to our banks,
as I understand it, relies a great deal on banks' internal
models. Do the regulators have the personnel with the requisite
expertise to conduct a thorough review of these systems? How
can we know that questionable practices or assumptions will be
caught by the regulators before the models are in full
operation? It is my understanding that modeling generally
involves the use and analysis of historical value. Please help
us understand the value provided by the use of modeling.
Governor?
Mr. Ferguson. I will start. I think we should take a step
back here because I think there is a point that is being
missed. The concepts in Basel II are not things that the
regulators thought of independent of the market. I would say,
in fact, just the opposite. What Basel II is trying to do at
base is catch up with where leading edge banks already are in
the way they run their banks. Not all banks are there. But we
are trying to reinforce the decisions that banks have already
made and to encourage others to adopt some of these techniques.
So this is not brand new terra incognita to the banks. They are
not starting at ground zero.
The second point I make is on models. We have chosen not to
have capital under Basel II that is purely model driven. It is
true that the banks would provide some inputs into the formulas
that we have derived for determining capital. So you should
not, Senator, have the impression that somehow or another there
is going to be a model that chunks away that no one understands
and then a number----
Chairman Shelby. There is a model here. Excuse me. There is
a model here though, is it not?
Mr. Ferguson. There are approaches and models the banks use
to determine some of the inputs, but the regulatory formulas
take those inputs and then determine what the capital is going
to be.
Now, your question with respect to historical data. It is
absolutely true that in order to evaluate your sense of a risk
on probability of default or a loss given default, one of the
inputs will be historical experience. That works extremely well
in some portfolios for sure. There may be others in which it
does not work as well. Banks will add to that important
judgments, for example, stress testing. One of the things the
regulators will look at is whether or not indeed a model or an
approach to estimating these inputs looks across an entire
cycle, so that banks are not just picking up the good times,
but also, frankly, the bad times.
Chairman Shelby. That is where capital is imported though.
Mr. Ferguson. One of the reasons that one wants to have
stress testing and a cross-cycle input is to get capital that
deals with both the good times and the bad times. I think you
should be aware that indeed one of the things we would be
looking at is whether or not the data that are being used cut
across a broad swath of time and not a short period that may or
may not be representative, and also whether or not there is
stress testing, for example, to figure out in the more extreme
scenario what kind of loss a bank is likely to experience, what
kind of exposures it is likely to have, et cetera.
You asked an important question that Comptroller Hawke has
answered, and I would endorse what he said. The regulators will
have to make investments in training and staff. I would add two
things that he did not say. One is that this entire process, at
least from the standpoint of my agency, has allowed us already
to start to build skills, just in the interaction over the last
4 years to do this. And also--again this is the Fed maybe more
than the other agencies--we are blessed by already having,
because of the nature of what we do, a number of economists
already trained in these areas. Some of them have moved over to
be the heads of supervision or senior members in supervision.
We still have some way to go. The banks are not at ground zero.
Frankly, nor are we. By definition, I do not want to
overestimate or underplay the amount of investment that we are
going to have to undertake.
Chairman Shelby. Mr. Hawke, you are the Comptroller. I
would like your views on this. This is important.
Mr. Hawke. Let me say I envy their unlimited budget.
[Laughter.]
Chairman Shelby. At the Federal Reserve.
Mr. Hawke. At the Federal Reserve.
I think it is important to recognize that the use of models
is not something new or something that is going to be initiated
by this Basel process. Models have been used for years. Since
1997, banks have been using models to measure value-at-risk or
market risk, as part of the Basel process. I think we at the
OCC have been in the forefront of using economists in the
examination process, working with banks and their models. The
use of models as measurement tools in the supervision process
is not brand new with Basel II.
Chairman Shelby. But you cannot just use the model. You
have got to have other means too, have you not, other than this
model?
Mr. Hawke. Oh, absolutely. As I said before, the role for
conventional bank examiners will be undiminished. I view Basel
II, if it really works, as providing our examiners with better
tools to do the job that they already do.
Chairman Shelby. We were told that the Long-Term Capital
Management used models developed by Nobel prize winning
economists and their models failed. So as a comptroller, any
model, you will have to be looking at it closely to see if it
does the job that you want, right?
Mr. Hawke. I can assure you, we will not hire any Nobel
prize winners at the OCC.
[Laughter.]
Mr. Hawke. But I think it is also important to recognize
that the models in Long-Term Capital Management did not have
the kind of supervisory triangulation that we are going to
bring to bear in this process.
Chairman Shelby. Mr. Powell.
Mr. Powell. Senator, I think you bring a valid point. I
think models are important. I think models are necessary, and I
agree with what my fellow regulators have said. Models have
been around for some time, and I think they are useful tools.
But I am reminded of the old statement, junk-in, junk-out. I
think models are wonderful, but the input data is extremely
important. That is where the supervision, and where the
oversight of the regulation will be critically important--to
look at those estimates going forward. But I think models are
very useful.
Chairman Shelby. Mr. Gilleran.
Mr. Gilleran. Regulators have faced sophistication all
along the last decades because of data processing, and we are
up to the task I think from a technical point of view, and we
will hire others if we need them. But what regulators bring to
the table is a gimlet eye, and it is that gimlet eye that is a
very important one in the process, and I think that will
continue to be the case.
Chairman Shelby. From what I understand, this proposal
involves taking the internal management practices of the banks,
and saying such practices will now largely inform the
regulatory standards. Typically--and I will direct this to you,
Mr. Hawke--is it not Congress or at least the regulators who
develop the policies which banks then implement? Does this
proposal turn this equation around?
Mr. Hawke. No, I do not think so, Mr. Chairman. We
implement the basic policies that Congress sets. Congress has
given us at the OCC the task of determining the rules under
which the capital for our banks will be determined and
evaluating capital adequacy on an ongoing basis. We are not
turning over the policymaking function to the banks. The rules
will be set in great detail, and there will be very careful
oversight.
Chairman Shelby. It can probably be said that there are
some considerable differences between the markets of all the
countries that are participating in the New Basel II Accord.
What considerations were made regarding these differences? In
this country our banks use very sophisticated methods of
securitization. What impact will the proposal have on
securitization in the United States?
Mr. Hawke. That has been a great concern of mine, not only
because U.S. banks do far more in the way of securitization
than European banks, but also because national banks in
particular are very heavy into the securitization markets. So,
we have tried to make sure that the rules that emanate from
Basel II on securitization do not discriminate against U.S.
banks.
There are other aspects of this whole structure that
concern me in terms of international comparability and that
relate to the nature of supervision. Our large banks, including
the 10 to 12 large banks that will be covered by Basel II, have
full-time, on-site resident examiners that are there day in and
day out. Some of our counterparts on the Basel Committee
examine their banks once every 2 years or once every 5 years.
There is the potential for disparity here in application just
because of the disparity in the nature of the supervisory
functions from country to country.
Chairman Shelby. Can we, just for the record, definitively
establish the number and perhaps even the specific entities
which will be required to comply with this proposal? Is the
number you cited in your testimony, up to 20 banks definitive?
I recall that Basel I was intended to apply only to
internationally active banks, but now covers all U.S. banks.
Do you want to take that, Governor Ferguson?
Mr. Ferguson. Well, I will start and my colleagues will
chime in and correct me if I am wrong.
The plan, as we have indicated, is to have about 10 banks
that will be mandatory banks. We think another 10 or so will
want to opt in early on in this process for a variety of
reasons. The question you ask puts us into a broader time frame
though. The reality is, not in the short term but I would
expect over the intermediate term, if indeed Basel II does what
it is intended to do, which is to create more risk-sensitive
capital which gives a better signal to management of what is
going on, more banks will find it in their interest to opt in.
That will also occur as the cost of building some of these
systems and hiring some of these people comes down. So over
time, not in the short term but over the intermediate term,
almost regardless of what we as regulators do, I would expect,
if this is indeed beneficial in helping banks to run themselves
by giving them a better sense of risk, that we will have more
banks gradually opt in.
We will also, I think, as banks grow, find that we as
regulators want more banks to come in because they will reach a
larger scale of domestic operations or international
operations, but I do not think that is going to be a short-term
issue. I think it is more of an intermediate-term issue.
Chairman Shelby. Mr. Hawke.
Mr. Hawke. Mr. Chairman, the standard by which we would
determine which banks are and which are not mandatory is one of
the issues that will be addressed in the rulemaking. The
standard will take into account both size and international
exposure. Right now, our best estimate is that there will be
about 10 banks that will be included. Those 10 banks probably
account for, as I said, 95 percent of the foreign exposures of
U.S. banks, and if another 10 banks were to opt in, we expect
that we would cover probably 99 percent of the foreign
exposures of U.S. banks.
Chairman Shelby. Mr. Powell.
Mr. Powell. Senator, may I say one thing about the comment
on the role of Congress and the role of the regulator? Basel
does not do away with prompt corrective action that Congress
implemented during the crisis, and that still will be part of
it.
Chairman Shelby. Mr. Gilleran.
Mr. Gilleran. I must say too that Basel is not going to
take away the regulatory responsibility to be able to step up
and tell any bank that it feels it is operating in an unsafe
and unsound manner to have more capital. So therefore, this is
a system for measurement, but it does not take away our
regulatory powers or responsibilities.
Chairman Shelby. Senator Sarbanes.
Senator Sarbanes. That is interesting. I would like to
follow up on that comment of Mr. Gilleran right there.
Does that mean that if an agreement is concluded, our own
regulators could then impose subsequently a higher capital
requirement on our banks? Is that the point you were making?
Mr. Gilleran. That would be my conclusion if a bank adopts
it, but if we feel it has been adopted incorrectly or
inappropriately, or we feel our estimation of the risks are
such that more capital is required, then it is our
responsibility to ask for it and to get it.
Senator Sarbanes. Is that right?
Mr. Ferguson. Yes, that is right. You also should
recognize, as Chairman Powell has indicated, we already have
some areas of
difference from foreign countries. We have prompt corrective
action, for example. We have the leverage ratio. We have the
various legal requirements to become a financial holding
company which require that subsidiary banks be well
capitalized. There is also the market discipline. So there are
a number of different tools that we have in the United States,
some that are legislative, some that come from us as
regulators, that are likely to increase the amount of capital
we have far above or somewhat above the regulatory minimums. I
should also mention the fact that banks themselves, because
they recognize the cyclical ups and downs, already hold, and
will continue to hold, a cushion of capital above the
regulatory minimum that we are talking about here.
Senator Sarbanes. Now I want to address what is required by
the regulatory regime. If our regulators determine that what
Basel II produces is inadequate, to what extent are they
constrained by the agreement?
Mr. Ferguson. Should I respond to that? The agreement
already has a number of places that have what is called
national discretion, where we can accept or reject some
approaches. We also clearly have said over time, and will
continue to say, that as we get a better sense of the impact of
this on capital and as best practice changes over time, we
will, if necessary, go back and we will renegotiate. We
obviously, by definition, have the right and the authority to
make, if you will, a step back from the agreement and to make
some unilateral changes if we have to. That would not be the
better approach. The better approach, for a variety of reasons,
is to make changes in the international context because we want
to make sure that we have some control over the way our banks
that are operating overseas are treated, and therefore having
an international agreement is the better approach. But we have
not given up national approaches here if we find that we are
having a great deal of difficulty with things that we think are
important.
Senator Sarbanes. All of this argues for making sure that
we get it right in the first place, does it not?
Mr. Ferguson. I think it argues for two things. One,
obviously, making sure we get it right in the first place. Two,
reinforcing the message that I have just given you and that we
have also given to our negotiating partners, that as we see
things evolve and change, we have the authority and we reserve
the right to go back and renegotiate in places where we have to
renegotiate, but obviously, we want to do the best we can to
get it right the first time.
But we should also be mindful, Senator, that we, I think,
all believe that Basel I is not sending the correct signals to
us, to the market, or to the banks about the capital that is
being required. While we want to get this right the first time,
recognize that delay is not necessarily holding onto a safe and
sound system. Basel I is outdated for our largest institutions,
and so it is important for us to continue to work for a common
process, but to do it with a certain resolve so that we are not
leaving large institutions on a capital framework that is not
risk-sensitive and probably not fully
reflective of what is going on in those banks.
Senator Sarbanes. I guess my concern is that you seem to be
very close to the concluding date, but not yet to have either
worked out a number of problems that everyone concedes are
problems, or to have developed, at least on our side, a full-
scale consensus as to what should be done.
Mr. Hawke. Let me say, Senator Sarbanes, that there is a
lot of process to come here. We will be going out with an
Advance Notice of Proposed Rulemaking, which will be the first
occasion where we have officially solicited comments from the
entire U.S. public, not just the banks that would be affected,
on Basel II. The Basel Committee's schedule includes trying to
finalize the Basel product by the end of this year. I think
that may be too tight a schedule, given the need for us to
analyze the comments that come in in response to the ANPR and
to feed that back into the Basel process. Following the Basel
Committee's decision, we are going to have another quantitative
impact study, plus the potential for an economic analysis that
might be dictated by an Executive Order that applies to
rulemakings that have significant economic impact. We will have
the NPR process, which will be another opportunity for public
comment. It will not be until all of those things have run
their course that we will be in a position to consider whether
we are going to adopt the Basel proposal finally. So there is a
lot of process, a lot of opportunity for input into this before
the end of the day.
Mr. Ferguson. May I add two other points to that? The final
implementation for this is expected to be the end of 2006, very
beginning of 2007. Before that we will have a year of parallel
running between Basel II and Basel I, which will give us
another chance to see if we are comfortable. Then in the first
2 years there are capital floors that are put in to make sure
that if there are going to be reductions, we understand the
sources of them and that we are comfortable with them.
In some sense, Senator, what we are looking at ultimately,
before we have an unfettered Basel II, is many years from now,
not next year. So, 2004 will be when we hope we get the basics
all run. We still have room to make adjustments.
Senator Sarbanes. Once you close the agreement you are in a
different framework than before you close the agreement. I made
that point earlier, and I just want to repeat it.
Mr. Chairman, could I ask one final question?
Chairman Shelby. You may proceed.
Senator Sarbanes. In her written testimony, which will be
on the next panel, Karen Shaw Petrou, who has appeared before
this Committee a number of times and given us some very
perceptive and helpful testimony, poses a question, ``Whether
the complexities in Basel II's advanced models are so daunting
that supervisors at home and abroad will not be able to ensure
that banks actually comply with the new capital rules.''
Before I put out my question I will just tell you a little
story. John Biggs of TIAA-CREF testified before our Committee
when we were working on the corporate and accountants
responsibility legislation, and he told a story of one of his
analysts who came in to see him, his financial analyst, who
told him that he just could not see where Enron was making all
this money, that he had been over their statements again and
again, and he just could not figure it all out. And Biggs says
to him, ``Well, if you cannot figure it out, we had better sell
it.'' And they sold it. This was a couple of years before Enron
took a nosedive and went into bankruptcy.
So my question to you is whether you have the necessary
expertise and resources to implement Basel II and effectively
supervise the banks' internal risk measurements. I mean,
everyone says right from the beginning, this is extremely
complicated and complex. Only the big banks can handle it, and
so forth and so on. That obviously raises a concern about
whether the system can be gamed. Where are we on that rather
important question?
Mr. Hawke. I think that is a very important question, and I
think it does present us with some challenges in terms of
training and retention of people. I disagree with the
suggestion that we are not up to the task of evaluating the
models that our banks are going to be using. As I mentioned
earlier, Senator Sarbanes----
Senator Sarbanes. Who will evaluate the models that other
banks are using internationally, our competitors?
Mr. Hawke. Their supervisors will be evaluating those
models, and that is certainly a cause for concern because the
intensity of the supervisory process in some of the Basel
countries is quite different from ours. One of the purposes
that the disclosure pillar, Pillar 3, is supposed to serve in
this process is to provide as much transparency as we can with
respect to this process, so that there can be some cross-
checking from country to country, so that we can look at what
is going on in other countries and determine the integrity with
which the process is being applied.
Mr. Powell. Senator, I hope this is the last time I will
say this, but I think again you are raising lots of issues on
the complexity and the validation issues and the competition.
Again, I would just say that that is another reason for
regulatory minimum capital.
Mr. Ferguson. May I throw in a third if you will? A couple
of points I want to make. You are also going to hear on the
second panel from one of the bankers here who has a technical
orientation by his background, who if I recall his testimony,
talks a little bit about a set of models called VAR models,
value-at-risk models, which 10 years ago were brand new and
went through exactly the same degree of concern and
uncertainty, and today we discover are well-established. We
understand them. So, yes, this is in some sense relatively new,
and I will come back to that point in a minute. But these
things do evolve our own understanding and that of the banks
does evolve.
The second point to make is in some sense we have no choice
here. We always can choose what we do, but what this is doing,
again, is reflecting what many leading edge banks are already
doing. It is reflecting what some of our regulators, some of
our supervisors already have to come to grips with. It is
making those changes more visible. I agree with you it is
certainly important. We should not go into this being naive. We
should not go into this thinking that it is easy, but we should
not refuse to go into it because it is new, because indeed it
is already happening in many of our leading edge banks. I will
not underestimate the importance of the complexity and the
degree of concern and the caution that you are raising,
Senator, because they are extremely well thought out,
obviously. But I do not want to leave the impression that we
are stepping to terra completely incognita. We are, I think as
regulators, doing what many of the leading edge banks are
doing, we are reflecting where the well-run have already gone.
You talked about the complexity allowing some gaming of
Basel II. The reality is that one of the reasons we are going
down this path is that--it is a quote that you heard from Larry
Meyer--there is, if you will, gaming--I know ``gaming'' sounds
much too pejorative--there is capital arbitrage already going
on. One of the things we are trying to do is confront that
directly by getting the capital to reflect the risk because we
now have a system in which capital may well not be reflecting
risk. Yes, this is complex for sure, but we are not going to be
naive about going into this. We are not running into it with
our eyes closed, if you will. But we also know that in some
ways we are doing only what the market has already started to
do.
Mr. Gilleran. May I speak on this?
Senator Sarbanes. Yes, sir.
Mr. Gilleran. Senator, I believe that we will have the
people necessary to do probably the finest regulatory job that
can be done in connection with Basel II.
In answer to your question, any time that you would make a
system variable based upon the institution's own determinations
of risk, that you are going to have a system where mistakes
will be made, either intentionally or unintentionally, that has
to be dealt with in the regulatory scheme. So any time that you
would come up with a system like this, you are going to have to
take into consideration that there will be mistakes made.
Senator Sarbanes. Thank you.
Chairman Shelby. Senator Allard.
Senator Allard. I just have one question regarding the
origin of needing to update Basel I. Did the Basel II proposal
come about because we had financial institutions in this
country concerned about the competitive environment, or did it
arise from the regulatory community's concern with safety and
soundness issues? How is it that the need to change the
original Accord came up and was brought to your attention?
Mr. Hawke. I think a little bit of both. There was a
growing recognition, as Governor Ferguson has said, and as
Larry Meyer said in the quote that was read earlier, that the
old system was too coarse, that in defining several risk
buckets, it was not really accurately reflecting risk. That was
becoming evident as time went on.
Senator Allard. That was a concern of yours.
Mr. Hawke. Yes, Senator. And there was also a concern that
we needed a better system of oversight internationally, that
Basel I was not being applied in an even-handed way
internationally.
Senator Allard. Regulators internationally share some of
those concerns?
Mr. Hawke. Yes.
Mr. Ferguson. I would like to add a third reason, which is
a sense from the leadership of our largest banks. They were
observing that the way they were managing and thinking about
internal risk was diverging from what they had to do for
capital purposes. While by definition those things may not
always be perfectly aligned, I think all of us would recognize
that there is probably a benefit to having banks building their
basic regulatory capital under our guidance, supervision,
validation, et cetera, in a way that is at least consistent
with the way that they manage themselves because they should be
getting the same signals from their own economic capital models
and judgments, and also from what they are doing with respect
to regulatory capital, what the minimum regulatory capital is
telling them.
Senator Allard. Are these large banks concerned about not
being able to compete because of the capitalization
requirements that we have here in the U.S. as compared to other
countries?
Mr. Hawke. Well, I have not heard that. I think that, as
Governor Ferguson said earlier, regulatory capital requirements
generally, frequently lag behind a bank's economic capital.
Banks are maintaining higher regulatory capital than the
regulatory minimums that are required.
Senator Allard. In this country?
Mr. Hawke. In this country.
Senator Allard. How do U.S. banks view that requirement in
regards to competing internationally with other banks that have
not had a base in other countries?
Mr. Ferguson. I do not think they have seen that recently
as a disadvantage. To be fair, when the Basel process started
back in 1988, there was some concern that perhaps some banks in
some nations were letting their regulatory capital slip to
levels that were really unreasonably low. I have not heard
recently the kind of concern you are talking about, and the
reason is that the banks that we are talking about, the largest
internationally active banks around the world, compete with
each other, but they also go to the markets for funding. The
markets really are, in addition to their internal needs,
demanding a certain level of capital. That is, for reasons
Comptroller Hawke indicated and other reasons, higher often
than the regulatory minimum. So, I had not heard from our banks
a sense that they were at a competitive disadvantage because of
regulatory differences of this type. That had been true 20
years ago more or less, but that is not one of the incentives
that is driving this round of discussion with respect to Basel
II.
Mr. Hawke. Senator Allard, if I could just take a second to
go back to your earlier point. Under Basel I with the very few
risk buckets that cover all bank assets, it became clear that
within a particular risk bucket, there could be included assets
of widely different risk characteristics, and yet they had the
same capital charge. One of the things that did was to
encourage banks to invest in the riskier assets because there
was no differentiation in the capital charge for those assets.
So the current approach is to try to more closely match the
capital allocation to the risk of a particular asset.
Senator Allard. I see.
Thank you, Mr. Chairman.
Chairman Shelby. Thank you.
Obviously, we have, and you have, a lot of unanswered
questions. We expect answers here on the Committee, and we feel
that is our obligation before things get finalized. In going
forward, I hope that you will provide some of the means to the
Committee that we will be kept abreast of changes in the
proposal. All the Members have raised serious questions here
today, and we do not have all the answers yet, and I do not
think you do.
But we are very interested in this. We appreciate, Governor
Ferguson, you representing the Fed today, spent your morning
here with us; Comptroller of the Currency, John Hawke; Don
Powell, Chairman of the FDIC; James Gilleran, Director of the
Office of Thrift Supervision. It has been a long morning. Thank
you for your participation again before the Banking Committee.
Mr. Powell. Thank you.
Mr. Hawke. Thank you, Senator.
Mr. Gilleran. Thank you.
Mr. Ferguson. Thank you.
Chairman Shelby. We will now, although we are in a late
morning, going to call the second panel up, Mr. Maurice
Hartigan, President and CEO, the Risk Management Association;
Mr. Micah Green, President of the Bond Market Association;
Professor Edward Altman, Max L. Heine Professor of Finance,
Stern School of Business, New York University; Ms. Karen Shaw
Petrou, Managing Partner, Federal Financial Analytics; Mr.
Wilson Ervin, Managing Director, Strategic Risk Management,
Credit Suisse First Boston, and he will be testifying on behalf
of the Financial Services Roundtable; and Mr. Kevin Blakely,
Executive Vice President and Chief Risk Officer, Key
Corporation.
All of your written testimony will be made part of the
record. As I said, you have been here all morning. You have
heard all the regulators. We would like to move this hearing on
as quickly as possible. All of your written testimony will be
made part of the record in its entirety.
Mr. Hartigan, we will start with you. If you can sum up
your testimony as brief as you can. Thank you so much.
STATEMENT OF MAURICE H. HARTIGAN, II
PRESIDENT AND CEO
RMA--THE RISK MANAGEMENT ASSOCIATION
Mr. Hartigan. Good morning, Mr. Chairman, Senator Sarbanes,
and Members of the Committee. Thank you for inviting me to
appear before the Committee.
I am the President and CEO of RMA, the Risk Management
Association. RMA is a member-driven professional association
whose sole purpose is to advance the use of sound risk
principles in the financial services industry.
The main point I want to make to you today is that the New
Basel Accord will be a step forward for the United States and
world banking industries, provided it is modified as it is
being finalized and provided it is implemented flexibly. It
will be a step forward because it is directionally correct in
improving the risk sensitivity of regulatory minimum capital
adequacy standards. But it must be modified to ensure that it
is not too conservative, that these are truly minimum and not
maximum capital standards, and to ensure that it is not too
prescriptive.
The 1988 Accord relied solely on a regulatory minimum
capital standard. In contrast, the new Accord will be grounded
on three principles or ``pillars'' as they are called: Capital
requirements, enhanced supervision, and greater disclosure.
This alone represents a significant improvement.
Nonetheless, we have specific concerns in this area. Pillar
1, which deals with the capital standard itself, must contain
assurances that Basel will evolve toward a full models-based
approach for credit risk, and it must avoid arbitrary
specificity. Pillar 2, which deals with the implementation of
the standard through the process of supervision, must allow
regulators enough discretion to accommodate the diversity of
best practices in risk management today. Pillar 3, which
requires increased disclosure in order to provide greater
market discipline, must ensure that comparability is meaningful
across the varying international accounting regimes.
Our research to date suggests that the new Accord, as
proposed in the Third Consultative Paper, will require more
overall capital than many banks' internal risk-rating systems
require today, even though for some banks and some portfolios,
the new overall requirement will be somewhat less than under
the old Accord. This will often be inappropriate.
The new Accord should represent a true minimum capital
requirement. For well-run banks in normal times, this implies
that regulatory capital levels should be set below a bank's
economic capital based on best practice internal risk
measurement procedures.
Given the newness of the fields of study surrounding credit
and operational risk management, it is natural that regulators
should be prone to conservatism. But too much capital is just
as bad as too little capital. Too much capital will drive down
the risk-adjusted rates of return on a particular business line
and cause bankers to lend less than they otherwise would and
should. This is not a good thing either, for the shareholders
of the bank, the loan customers, or the general economy.
The next point I wish to make is that the process to reform
the 1988 Accord has had a positive impact on the development of
risk measurement and management procedures in the financial
services industry. Moreover, the dialogue between the industry
and its regulators surrounding Basel reform, while not without
frustration on both sides, has been useful and productive.
While outstanding issues clearly remain, some quite
significant, continued discussion with the industry is ongoing,
and I would expect this to be the case throughout the reform
process, and into the implementation stage as well. Indeed, it
may not be possible to resolve a number of specific issues
without an active two-way dialogue between regulators and the
industry as the implementation process takes place. Further
discussion can only help promote innovation and investment in
best practices throughout the industry.
It is for this reason that the reform process must
continue. However, it must be framed as a work in progress.
There cannot be a prescribed end state for sound risk
management practices. Otherwise, the ink on the new Accord
would not be dry before it became obsolete, much like the 1988
Capital Accord.
The quantitative analysis supporting sound credit risk
measurement and management are still evolving. Many of these
emerging practices were born out of the last economic downturn.
The resilience of the financial services industry over the past
3 years should not go without comment. Many have credited the
industry success to the better risk management practices
established over the past decade. I clearly agree.
Furthermore, in our own review of Basel II, we find that
some of the new requirements are written in a very prescriptive
fashion that does not lend itself to allowing individual banks
to employ a diversity of best practices. Without such diversity
we cannot have continued evolution of best practices, and
without evolution we could not have had the improvements in
risk measurement that have occurred over the past decade.
In the interest of time, I redirect your attention to my
writ-
ten testimony for a fuller treatment of two more technical
points
which are of great importance, in essence, that the internal
risk ratings-based approach must be followed with a full
internal models approach to capital; and second, capital is
required only for unexpected loss known as UL, not for expected
loss, as the current version of the Accord argues. RMA has
additional technical concerns specific to the Third
Consultative Paper which we will address in the formal
response.
To conclude, I would like to reiterate RMA's belief that
the reform process has helped advance the practice of sound
risk measurement and management within the industry. RMA is
hopeful that the New Capital Accord can be structured to
encourage and enhance continued industry innovation and that it
will recognize the benefit that diversity of practice within
the industry provides.
Thank you, and I would be happy to answer any questions
that you may have.
Chairman Shelby. Thank you.
Mr. Green.
STATEMENT OF MICAH S. GREEN, PRESIDENT
THE BOND MARKET ASSOCIATION
Mr. Green. Thank you, Chairman Shelby and Members of the
Committee, for the opportunity to testify today on the Basel
Committee on Banking Supervision's proposed New Capital
Accords, or Basel II. My name is Micah Green, and I am
President of the Bond Market Association, which represents
securities firms and banks active in the United States and
global bond markets. Association member firms account for at
least 95 percent of all bond market activity in the United
States, in addition to much of the bond underwriting and
trading in the rest of the world. Together with our affiliates,
the American and European Securitization Forums, we represent a
majoirity of the participants in the growing securi-
tization markets in the United States and Europe. The following
comments focus only on those issues related to Basel II that
are most important to our membership.
First let me say the association supports the Basel
Committee's overall goal of rationalizing the current risk-
based capital regime and aligning regulatory capital
requirements more closely with actual credit risk. We are
grateful to the Federal Reserve Board and the other banking
regulators, in particular, Vice Chairman Roger Ferguson, for
working with us to address the issues presented by the proposed
capital accord revisions that are important to our membership.
We are still concerned, however, that if not amended, Basel II
will diminish the economic benefits derived from large and
growing sectors of the capital markets, benefits which accrue
to consumers, as well as businesses.
I will first make one general comment on the direction of
Basel II and then focus on the two areas most important to us:
Securitization and repurchase, or repos, and securities lending
transactions.
With regard to Basel II broadly, we believe it is important
that this agreement not be viewed as the last word on
regulatory capital. Risk management techniques are continually
evolving, and the financial markets need a regulatory capital
accord that evolves with them. Basel II must, therefore, be
crafted in a way that ensures that it can better adapt to
changing market products and developments. Ultimately, the
global financial community will need to move toward a broader
reliance on internal risk models to determine appropriate
capital levels.
Securitization is the process of converting illiquid
financial assets, like loans and other receivables, into
securities, which can then be traded in the capital markets. It
is a large and growing market with tremendous economic benefits
for consumers and businesses. Securitization lowers borrowing
costs for consumers and others, improves risk management, and
draws new sources of capital to the lending markets. Consumers
benefit from these efficiencies with lower interest rates and
lower prices.
Just to summarize the size of the securitization market, in
the United States over the last 5 years it has increased 5-
fold, to $2.7 trillion. In Europe, it has increased 20-fold
over that same period of time, to a much lower level overall
but still $151 billion of outstanding securities. And in Asia,
where the marketplace is just
getting started in the last 7 years, the marketplace has
increased 510-fold, to a level right now of $51 billion, and it
is anticipated it will grow quite rapidly.
Financial institutions participate in securitization as
issuers and investors and as part of the risk management
functions. For securitization generally under Basel II, the
proposed risk weights for securitization positions held by
banks are too high in light of the actual credit risk presented
by these products. The proposed rules use unrealistically
conservative assumptions that, cumulatively, would require
financial institutions to set aside excessive levels of
capital. And considering who ultimately benefits from a vibrant
securitization market, consumers of homes, car buyers, or other
people who need capital, this is very important.
Repo and securities lending transactions, although little
known outside the wholesale financial markets, are vital to our
capital markets' liquidity and efficiency. Repo and securities
lending transactions allow market participants to finance and
hedge trading positions safely, cheaply, and efficiently. It is
utilized by elements of the Government. The Federal Reserve, in
fact, uses this marketplace to implement its monetary policy.
Basel II may require banks to take capital charges inconsistent
with the actual level of risks present in repo and securities
lending transactions. Financial institutions should have
greater flexibility to employ supervisory-approved internal
risk models created to assess counterparty risk in order to
accurately reflect risks present in these transactions.
We agree completely that the current regulatory scheme for
bank capital, Basel I, needs significant revision. The current
regulations are outdated and inflexible, as you have heard
before. Updating the regime can produce significant benefits,
including the promotion of fair global competition, incentives
for better internal risk management, and an economically
efficient allocation of capital. Getting it wrong, however, and
implementing capital regulations which do not reflect modern
practices or true credit risks on balance sheets will diminish
or eliminate the market efficiencies.
The Basel Committee is on the right track. We need them to
improve it even more, and we look forward to working with them
in this process.
Thank you again, Mr. Chairman.
Chairman Shelby. Professor Altman.
STATEMENT OF EDWARD I. ALTMAN
MAX L. HEINE PROFESSOR OF FINANCE
LEONARD N. STERN SCHOOL OF BUSINESS
NEW YORK UNIVERSITY
Dr. Altman. Thank you very much, Senator Shelby, for
inviting me here today.
I have followed the Basel II's consultative papers since
the first one was issued in June 1999. A colleague and I have
written several commentaries to them, particularly with respect
to Pillar 1, the capital adequacy based on specific risk
characteristics of bank counterparties. Our major comments were
that capital requirements related to expected and unexpected
losses from corporate and other loans should be based on actual
historical experience of the Loss Given Default from the
corporate bond and bank loan markets. The original 1999
suggestions bore absolutely no resemblance to real-world
experience. However, they have made significant modifications
since 1999, and I am pleased to say that the
current revision does a much better job of relating the
requirements to default experience, although in my opinion
still too little capital is being required for the most risky
categories, and probably too much capital for the least risky
categories.
A problem with the suggested regulations, however, is the
complexity in determining capital requirements and the somewhat
arbitrary choice of modifications to the standardized scale due
to such items as the size of the counterparty and the
existence, or not, of collateral on the loan or the bond. For
example, Senator Shelby, you mentioned before small and medium-
sized enterprises as a very important part of our system.
Chairman Shelby. Do you agree with that?
Mr. Altman. Absolutely.
Chairman Shelby. Okay.
Mr. Altman. Absolutely. The Basel Accord under the current
recommendations will give lower capital requirements for
comparable risk levels for as much as 25 to 50 percent less
capital for SME's than larger counterparties. The argument that
the correlation of default rates among these small
counterparties is lower than for larger corporations may be
valid, but I have seen little evidence that the haircut for
SME's for these loans should be as much as 50 percent. In my
opinion, this was a concession to those national banking
systems of the world whereby SME's are the vast majority of
borrowers; hence, lower capital requirements for banks in those
countries. It is also true, however, that SME's make up the
vast majority of loan assets of smaller banks in the United
States and the same lower capital requirements would hold for
U.S. SME's and the banks that make these loans, close to all
but 100 of our Nation's 8,000 banks. But, as I will now
discuss, almost all of U.S. banks will not be required to
follow the recommendations of Basel II. So the reduced capital
requirements for SME's will not be relevant and the old Basel I
8-percent rule will probably still be in effect for all but the
very largest U.S. banks. In other words, SME's in other
countries will be advantaged vis-a-vis the United States.
As I indicated above and as you probably all are aware by
now, and you heard the regulators this morning, the central
banks of the world and other bank regulatory bodies set
national regulatory policy based on Basel's recommendations,
but they do not have to accept what Basel suggests. Indeed, it
came as an enormous surprise to some observers, including this
writer, that only the largest 10 U.S. banks, and perhaps the
next 10 to 20 banks in terms of asset size and other
requirements, would be required to conform and the next 10 to
20 having the option, to opt into Basel II, and for all other
banks, Basel I still remains. In other words, the ``bad wine''
that Basel I might have been drinking in 1988 is still going to
be drunk in the coming years by all but 10, maybe 20, banks in
the United States.
While it is true that as much as two-thirds of all bank
assets are held by the top 30 U.S. banks and more than 95
percent of the foreign exposures of these banks will be covered
under Basel II, it is likely that all the rest of our banks,
almost 8,000, will not be asked to conform and will probably
not do so for many of the reasons you heard this morning: Basel
II is too complex and too costly; the U.S. banking system is
presently more than adequately capitalized; the added Basel II
capital required for operating risk is based on highly
arbitrary and extremely difficult-to-measure variables; and,
finally, the Federal Reserve System's and other regulatory
bodies' maximum leverage ratios and prompt corrective action
have worked very well. In other words, if it ain't broke, don't
fix it.
I believe that the choice of only the 10 largest commercial
banks to conform to Basel II and the IRB approaches is
unfortunate and should be reconsidered. Notwithstanding, the
recent consolidation movement of many of our largest and most
sophisticated banks, the possible exemption of number 11 to
number 30, including such large banks as HSBC Bank, Citibank
[West], Bank of New York, Key Bank, which you will hear later
from, State Street Bank, number 11 to 15, and the very likely
exemption of number 31 to 50, including such seemingly large
banks as Charter One, Am South, Union Bank of California,
Mellon Bank, and Northern Trust, to name just a few, seems
arbitrary and belittles the possible sophistication and
motivation of these banks which would be substan-
tial institutions in most other countries of the world. For
example, the 50th largest bank in the United States in terms of
assets, Compass Bank, $24 billion in assets, or in terms of
deposits Mellon Bank, would be huge institutions in most
countries. They do not have to conform.
The choice of a round number like 10 would seem to be
insensitive to world opinion, as well as to the risk management
motivation. Speaking from an economic standpoint, rather than a
political one, I would prefer to see either no banks be
required or some exemption level whereby the costs/benefits to
our banking system would be more rationally presented and
defended, not just the fact that they are internationally
active or not. Certainly, a number like the top 50 to 100 banks
would be much more in line with the number of banks conforming
in other countries.
Our largest banks are probably relatively happy to conform
to Basel II even with its complexity and added costs to develop
the systems and models we have been talking about. Some of
those models, by the way, I helped develop a long time ago, and
I am very happy to see that they are getting the play that they
are. The reason is that they expect total capital required for
credit assets will be less than what is required under the
current regime. So we may have a new regulatory regime where
everyone, large and small banks, as well as our bank
regulators, are relatively pleased with the changes
recommended. That does not necessarily mean that it is good
legislation.
I have always felt that despite its problems with
complexity, too low capital requirements for risky
counterparties, and the difficulty in managing against
operating risks, Basel II had one extremely
important by-product, at least one: The motivation for banks to
develop or improve upon their existing credit-scoring models
and systems to reduce total losses from nonperforming and
eventually charged-off loans. These systems can be used to rate
and set capital for all bank customers rather than using a
``one-size-fits-all,'' the 8-percent rule which is now in
effect. I have observed the enormous strides achieved by banks
throughout the world--mostly outside the United States, I might
add--including ones of all size and location, as to developing
risk management systems and training of personnel to prepare
for Basel II. Indeed, from what I can surmise, banks in most
countries, especially in the European Union, will all have to
adhere to Basel II's Standardized, Foundation, or Advanced IRB
approaches. Granted that regulators in these countries will
need to sanction far fewer banks than U.S. regulators would
have to do if all banks are mandated to conform, it must have
come as a surprise, perhaps even a resentful shock, that the
vast majority of U.S. banks will not adhere to Basel II. This
is especially true since the United States and its
representatives to the BIS were the early champions of the need
to change the way banks allocate capital for credit risk of
their clients.
What is disappointing to me is that the Fed's decision to
exempt all but the largest banks from building and implementing
IRB approaches, et cetera, for risk management systems will
demotivate the rest of the banks to do so. Under Basel I, they
did not have to do so. They are not going to have to do so in
the future, although some will opt to do so, as the regulators
said. I am very sensitive to the problems in Basel II and what
the regulators are faced with. But, in conclusion, what I would
like to say is that our decision to exempt smaller banks from
Basel II may backfire if many of the world's smaller, or even
larger, banks in other countries decide also to opt out of the
system because of what we do. This may cause an international
problem in banking, which can affect us due to a contagion
effect.
I have other comments on the procyclicality issue, but in
the interest of time, I would like to just conclude and thank
you again. I am certainly happy to answer any questions.
Chairman Shelby. Thank you.
Ms. Petrou.
STATEMENT OF KAREN SHAW PETROU
MANAGING PARTNER
FEDERAL FINANCIAL ANALYTICS, INC.
Ms. Petrou. Thank you very much, Mr. Chairman, Senator
Sarbanes, and the Committee as a whole. I am Karen Shaw Petrou,
Managing Partner of Federal Financial Analytics, which is a
firm that advises on the strategic impact of U.S. legislative,
regulatory, and policy events like the Basel Accord. I also
serve as Executive Director of a group called the Financial
Guardian Group, which represents those U.S. banks most
particularly concerned with the proposed new capital charge for
operational risk-based capital.
I thank Senator Sarbanes for mentioning that I have been
here before. Through the 1980's, as a matter of fact, when this
Committee spent a tremendous amount of time and ultimately had
to allocate an awful lot of taxpayer money to rescue the FSLIC
and address the savings and loan crisis. My firm then also
advised a national commission on the causes of the S&L crisis
chartered by Congress in 1989, and we concluded that the
predicate cause of the savings and loan debacle was the failure
in the early 1980's by the thrift regulators to set appropriate
regulatory capital. I am a strong believer in regulatory
capital because, at its most simple level, it means that
shareholders put their money first. Their money is up before
the deposit insurance fund, before the lender of last resort.
It is a critical discipline.
I must respectfully disagree with Mr. Ferguson in terms of
the importance of regulatory capital in driving decisionmaking.
It is a profound driver of profit decisions at the most senior
level of every financial services firm, bank and nonbank. And
as a result, it can have major policy impact.
I would point to, for example, the question of Fannie Mae
and Freddie Mac, an issue now before this Committee as a result
of recent events. Congress and the markets have allowed Fannie
and Freddie to operate with regulatory capital somewhere
between a half or 20 percent of that which would be required on
the same assets if they were held by an insured depository. If
you want to look and ask yourself why have these two GSE's
doubled in size every 5 years to now hold $3.3 trillion in
obligations, are they smarter than everyone else? Maybe. Do
they have advantages as GSE's? For sure. But, fundamentally,
they can run a lot bigger and maintain tremendous profitability
on much smaller amounts of regulatory capital. So these issues
are very important, and they will have profound implications
for the financial services industry.
As Chairman Powell said, Basel II, for all its hundreds of
pages and formulas, is a major strategic driver of the
competitive direction of our financial services industry and
its ability to serve a core customer base in a safe and sound
fashion going forward.
I would just like to emphasize four points in the interest
of time this morning.
First, I would suggest that, as Basel II concludes,
hopefully quickly, and the U.S. regulators wrestle with their
own implementing rules, we focus on first things first. Five
years ago, Basel was set in motion to deal with the regulatory
arbitrage issue, that is, banks holding high-risk assets when
their regulatory capital ratios were too low, and low-risk ones
left the banking system because the economic capital,
regulatory capital numbers simply did not match. That is
arbitrage, politely, or gamesmanship, not so politely. And it
is risky.
There is a lot in Basel II on which all agree. Some of it
is in the more simple standardized models, and, sure, they are
not perfect. But they are a lot better than what we have right
now, and that part of Basel should move forward quickly because
existing ongoing regulatory arbitrage undermines our banking
system.
I was surprised to see the regulators post on their website
a couple of weeks ago the goals of Basel II, and suddenly they
now are improving internal risk management, promoting market
discipline, and, most mysterious of all, imposing a new
operational risk-based capital charge. Two of those three goals
make sense, but the underlying purpose of Basel II regulatory
arbitrage termination has disappeared, and I would urge
refocusing on that.
Second, I do not believe that one-size-fits-all capital
works. We have imposed the leveraged capital requirement. It is
the most primitive of all of those on the table here in the
United States because of concern about other sources of risk.
But it is a piece of the puzzle that drives low-risk assets out
of the banking system. Again, our minimum capital requirements
are leverage standards. The way risk-based capital works now is
one of the reasons why in recent years the highest-quality,
lowest-risk mortgage assets have left our banks and savings
associations for the secondary market and highest-risk,
subprime loans are now being held by banks. We need a capital
system that promotes safety and soundness not through an
arbitrary cushion of capital but, rather, a system that rewards
good risk-taking and imposes higher capital when higher risks
are taken.
The operational risk-based capital proposal, in my opinion,
has a perverse incentive. It will encourage risk-taking not
reduce it. And in the wake of September 11, we have all learned
how really serious operational risk can be. And, most
importantly, in those tragic days we also learned the value of
effective operational risk mitigation, contingency planning,
back-up facilities, and in the days thereafter, insurance.
The crude capital charge proposed in Basel II will promote
operational risk-taking because banks will have to hold an
arbitrary amount of capital based on gross income, regardless
of the way in which they invest in operational risk management
or mitigation. I know the United States regulators have
decided, rightly, that that approach is so flawed that we
should not impose it here. But maintaining the operational risk
capital charge in Basel means that it will still guide major EU
and Japanese banks, and we cannot wall ourselves off from their
operational risk. We need a good regulatory and supervisory
system that rewards appropriate operational risk management and
mitigation that can and should be done under good supervision.
Finally, I think the challenge of Basel II is just that:
Good supervision. We have it here. It does not exist uniformly
elsewhere. Pillar 1 in the Basel new paper is about 200-some-
odd pages, not counting footnotes. Pillar 2, which is supposed
to improve supervision, is about 20 pages. I would rather see
them reversed and put our efforts into ensuring that here and
abroad, when banks threaten their safety and soundness or pose
large economic risks, supervisors intervene quickly and
meaningfully and, if necessary, shut the banks down. That does
not exist elsewhere, and I think that is a fundamental
challenge still left on the table.
Thank you very much.
Chairman Shelby. Mr. Ervin.
STATEMENT OF D. WILSON ERVIN
MANAGING DIRECTOR, STRATEGIC RISK MANAGEMENT
CREDIT SUISSE FIRST BOSTON
ON BEHALF OF THE
FINANCIAL SERVICES ROUNDTABLE
Mr. Ervin. Good afternoon, and thank you for inviting me
here today. My name is Wilson Ervin. I am presenting testimony
today on behalf of Credit Suisse First Boston and on behalf of
our trade group, the Financial Services Roundtable. CSFB is a
major participant in global capital markets employing
approximately 20,000 people, mostly here in the United States.
We are regulated as a U.S. broker-dealer, a U.S. financial
holding company, and also as a Swiss bank. I head CSFB's risk
management function. My job is to assess the risks of my bank
and protect our capital, a goal that is similar to many of the
goals of bank supervisors and at the heart of the Basel
reforms.
We agree with the importance of bringing the current regime
up-to-date and fully support the objectives of Basel II. The
regulators who have worked on Basel II have addressed a great
many challenging issues with stamina and sophistication. I
would like to thank Governor Ferguson, Comptroller Hawke, and
FDIC Chairman Powell very much for their openness and
willingness to listen during this process.
Yet, while there is much to admire in the new rules, there
are also many elements that still raise serious concern. On
balance, we believe the advantages of the new rules now
outweigh the drawbacks, but this balance remains close. This is
a frustrating outcome for an initiative with so much potential.
Today, I would like to highlight four macro issues that we
believe are particularly important: Number one, the current
proposal is unnecessarily complex and costly and suffers from
an excessive reliance on detailed, prescriptive formulae.
Number two, procy-
clicality. The new accord could reduce liquidity in the credit
markets during economic downturns and potentially deepen
economic recessions. Number three, the operational risk charge
is highly controversial. And, number four, the disclosure
requirements require burdensome additional paperwork, raising
costs but adding little information of value to users.
The first topic I would like to address is the high cost
and prescriptive nature of the new accord. Conceptually, the
Committee has attempted to capture current industry best
practice and then boil it down into a fixed formula. These new
rules, while well-intentioned, will add burdensome
qualification, testing, and reporting
requirements and will be inconsistent with changing market
reality and evolving best practice. The cost of implementing
these sys-
tems will be very high. We estimate approximately $70 to $100
million in start-up costs for our firm. This increased cost
could run into the billions when added up across the whole
banking sector globally and could tilt the playing field
between banks and non-
banks very significantly.
The compliance costs of this accord will also be material
and will be driven by how these rules are enforced. This is
particularly true for international banks, who are regulated by
multiple supervisors and subject to the risk of conflicting
interpretations. This raises the risk of getting caught in a
Catch-22 between regulators. While this problem does exist
today to some extent, it will be a lot more important given the
number of complex rules under Basel II. The Fed has recently
indicated a constructive approach here which we hope represents
the start of a broader international effort to resolve
this problem.
CSFB and other Roundtable members are also concerned about
the cumulative effect of the numerous conservative choices that
are built into the fabric of Basel II. A good example can be
found in the rules for securitizations, which are a common
method for financing housing and credit card loans in the
United States. The rules proposed for this area include complex
formulae and flowcharts with tough controls over business
judgment. The combined effect of each of these individual items
adds up to regulatory capital requirements that can depart
significantly from the true economic capital needs that Basel
II is aiming to emulate.
The rules seem to be drafted primarily to avoid any
possible circumvention, but are likely to deter good financial
transactions as well. We believe the securitization rules will
tend to raise capital requirements in markets where this
technology is most advanced, notably here in the United States.
Our suggested response to the problems of prescriptiveness
and high cost is for the Basel Committee to place a much
greater emphasis on a principles-based approach. Whereas Pillar
1 currently sets out capital calculations in a detailed,
prescriptive way, the approach of Pillar 2 is to force the
development of better internal models, based on evolving best
practice, and then to scrutinize those results through the exam
process.
The new rules will change the capital requirements for bank
lending which can have consequential effects on the state of
the economy more broadly. Based on a review of the last 20
years of credit cycles, our calculations indicate that the new
rules will require much more bank capital during an economic
downturn when compared to the current system. My personal
estimate is that my bank would have cut back its lending in
these circumstances by perhaps 20 percent if the Basel II rules
were in place during the recent recession. If all banks cut
back in lending at the same time, as they will tend to do under
a common regulatory regime, the potential adverse effect on the
real economy could lengthen and deepen an economic slump.
The proposed quantification of operational risk, the risk
of breakdowns in systems and people, is highly controversial.
All of the Roundtable's member companies agree that evaluating
and controlling this risk is important and should be required.
However, many banks, including CSFB, believe the current
approach is deeply flawed because of the difficulties of
measuring and predicting this type of risk using a quantitative
model. Other banks take an opposite view and believe that the
operational risk rules will lead to improved risk governance
and better transparency and are, therefore, appropriately
placed in the Pillar 1 category.
One of the strengths of the Basel II proposals is that they
go beyond capital calculations in Pillar 1. They look to
improve market discipline via greater disclosure. While we
appreciate that the Pillar 3 disclosure requirements have been
reduced, they continue to be burdensome and potentially
confusing. While we certainly support transparency, as Chairman
Greenspan said recently, there is a large difference between
more disclosure and more transparency.
In sum, much hard work has been put into Basel II, but much
also remains ahead. In the pressure to finalize and implement
the accord, we hope enough time will be provided for everyone--
banks and supervisors alike--to consider the implications of
this new regime. Streamlining the current proposal will require
strong discipline in the final round of drafting and a return
to the original philosophy underlying the project. I believe
very much can be accomplished if we increase the emphasis on
best practice prin-
ciples rather than rigid formula, and if we increase the weight
of
Pillar 2.
Pillars 2 and 3 have real people behind them: Regulators
and the market. People can adapt to changes and new markets
more easily than a rulebook can. This also puts the burden back
where it should be--on the shoulders of bank management--to
demonstrate to regulators and the public that they are doing a
good job. That is in the spirit of the Sarbanes-Oxley reforms,
and I think it is a smart and durable way to improve
discipline.
Thank you.
Chairman Shelby. Mr. Blakely.
STATEMENT OF KEVIN M. BLAKELY
EXECUTIVE VICE PRESIDENT AND CHIEF RISK OFFICER
KEYCORP
Mr. Blakely. Thank you, Mr. Chairman, Senator Sarbanes, and
other Members of the Committee. I am here today on behalf of
KeyCorp, the 11th largest banking company in the United States
KeyCorp has total assets of approximately $85 billion and spans
the northern half of the United States from Maine to Alaska.
While the vast majority of our business is domestically based,
we do have a modest amount of international business activity.
KeyCorp is not one of the companies included in the
definition of ``the top 10 most internationally active
institutions.'' Accordingly, under the present regulatory
guidance, we will not be required to comply with Basel II when
it becomes effective in 2006. Nonetheless, it is our intent to
qualify as an advanced model institution. We simply believe
that it is good banking practice to develop the risk management
tools that are the foundation of Basel II. If that qualifies us
as an advanced model company under the new accord, so much the
better.
KeyCorp believes that Basel I is broken and that a new
accord needs to be implemented. Basel II is a major step
forward, and we applaud its approach. It is not perfect now,
nor will it be perfect when implemented, nor perfect 10 years
after implementation. Regardless, it is light years ahead of
Basel I, as well as any other proposal we have seen to date.
We acknowledge that it is complex, but banking is a complex
business. A simple solution to complex issues is probably not
the right medicine. As an industry, we should not shy away from
the remedy simply because it is complex. We should work
collectively with the regulators to find the right solution,
not the easy one.
We have our doubts as to the high cost figures attributed
to Basel II. Our own experience to date has proved to the
contrary. We believe that many of the Basel II costs are simply
expenditures we should otherwise be making as a matter of sound
banking practice. Good risk management costs money, but it is
intended to help avoid even bigger costs that arise from bad
risk management.
We do not believe the adoption of Basel II will trap the
financial services industry in a time warp. It will not stifle
the creation of new risk management tools, as some have
alleged. Banks will continue to develop better methods of
managing risks regardless of what Basel II requires.
We believe there is substantial merit to including as much
as we can in Pillar 1 versus Pillar 2. One of the greatest
benefits that Basel II promises is that it will utilize the
invisible hand of the market to discipline wayward
institutions. In order to do that, investors must have adequate
information to compare the risk of one institution against
another on an apples-to-apples comparison basis. Pillar 1 is
the best vehicle for ensuring that banks report on a consistent
basis.
Mr. Chairman, KeyCorp appreciates the opportunity to share
our views on Basel II. We want to make sure our industry
operates within a safe and sound environment. We know that this
is the goal of the Committee, as well as our friends in the
regulatory world. While Basel II is far from perfect, it
certainly moves us further down the path.
Chairman Shelby. Thank you.
I think everyone and I hope everyone recognizes that risk
management practices must constantly improve. You are in the
business. That said, I am not sure there is complete agreement
as to what actually leads to improved risk management
practices. Is it market forces? Is it the use of detailed and
specific capital rules such as those that have been proposed?
What kind of impact can using these kinds of rules have on the
development of risk management practices in the future? Mr.
Ervin.
Mr. Ervin. I think all those things are critically
important. It is not just regulatory capital calculations that
drive innovations in the market and drive innovations in best
practice, although I think they can be helpful.
Our concern here is that the complexity prescriptiveness of
what they are trying to build in Pillar 1 may actually restrict
that process, may restrict the evolution of market best
practice and may trap us--I disagree with Mr. Blakely on this--
may trap us to some extent----
Chairman Shelby. You are trapped in the model, aren't you?
Mr. Ervin. To some extent we are trapped in models today.
We are living in an advanced world, and I do not think we can
turn back the clock. The question is: Do we have the right
market forces and the right ways to keep up with how the world
is evolving? Or do we end up having to run two sets of books,
one for a potentially outdated regulatory regime and another
one for the world as it has come to be evolving?
Chairman Shelby. Ms. Petrou, the proposal, as I understand
it, requires banks to set aside capital for operational risk,
assuming that the potential for these types can be quantified.
Is it really possible to measure these risks? And don't we call
some of those risks, you know, acts of God for a reason? You
have heard that all your life. Do you want to comment?
Ms. Petrou. I do not think it is possible at this point to
quantify operational risk or to measure it in any way on which
anybody agrees. In fact, Basel's own committee, the Risk
Management Group, recently after reviewing the data that was
gathered in an effort to come up with Basel II, said that the
data needed to be used with caution. And another Basel
Committee has said that it does not believe that a quantifiable
operational risk capital charge is in sight.
I believe the proposed ops risk capital charge now is, in
essence, a way to top off the credit risk charge to protect
against drops in credit risk capital. But when low-risk books
of credit risk are there, then I think the capital should drop,
and we should not be fudging the books, as the operational risk
charge would do.
Mr. Ervin. Senator Shelby, if I might expand on that just
for a few moments?
Chairman Shelby. Yes, you go ahead.
Mr. Ervin. I am a quantitative person by nature. I came up
through that part of the bank, and I would love to be able to--
--
Chairman Shelby. You are still there, too.
[Laughter.]
Mr. Ervin. But I would love to crack the operational risk
management using quantitative people, hire a bunch of my people
in my department and have the problem solved. I am just not
convinced we can do that today. And I am concerned that we are
building the system still on models that have not been
validated, still have not been proven.
Chairman Shelby. This is a work in progress in a sense?
Mr. Ervin. Absolutely.
Chairman Shelby. Professor.
Mr. Altman. If I might piggyback on both comments.
Chairman Shelby. Go ahead.
Mr. Altman. I completely agree with them. I think it was a
major mistake to add operating risk to Basel II. I mean, it
almost was tacked on as an afterthought rather than the main
problem, which was credit risk, which was what Basel II was
supposed to handle. And then everyone plowed around as to how
to measure this, and I agree completely with both Karen Petrou
and Mr. Ervin that this is whistling in the wind. Most of the
banks have no idea, and the regulators have less, about
operating risk. I think it is plug figure, and it is
unfortunate because it detracts from all of Basel II. Nobody
likes it. I do not know anybody I have ever met who likes the
operating risk part of Basel II. And yet it is in there. I
think we should have the guts to say it does not belong in
there and get rid of it.
With respect to the first question you asked, it might add,
one of my comments earlier was if we do not require some opting
in by all but the largest banks, we are really going to
demotivate risk management at these other institutions. They
are going to throw up their hands and say it is too complex, I
do not need to change what I have been doing all this time. And
just the fact that we have not had any major bank failures
lately--we have short memories. Things back in the 1980's were
not so good. If we have a lot of stress to the system, I do
believe that prompt corrective action on the part of our
regulators, which is not part of Basel II, is a great thing we
have, and Basel II does not have it. And I would like to see
that be put in as well.
Mr. Hartigan. Mr. Chairman.
Chairman Shelby. Sorry, Mr. Hartigan.
Mr. Hartigan. If I could just respond on operational risk,
it is a nettlesome issue. It will continue to be. It is
something which is not science.
Chairman Shelby. While you are on that, would you please
touch on the new capital charge for operational risk? That is
all involved here.
Mr. Hartigan. As the Comptroller of the Currency said, it
really matters not whether it is 1 or 2. I think what we must
do, what the industry must do, is continue to observe and
analyze the data which surrounds operational risk, be more
comfortable with it, acknowledge it more, and acknowledge that
it does have a role in the capital charge.
Chairman Shelby. Do any of you have any concerns about
competitive issues associated with the new proposal? We know
what the proposal basically is.
Ms. Petrou. I would just like to say it has some unique
U.S. issues. We have talked about international
competitiveness. But it is also very important to get the rules
right here because some big financial services firms and some
little ones that elect to go in will be in, and some big ones
will be out by virtue of their charter choice. And some of the
ones that might think they are in can change that charter
choice and take them out.
In the EU, the proposal is to apply Basel to all financial
services firms, but under U.S. law it can only apply to
financial holding companies.
If the regulatory capital incentives are not better aligned
with the economic ones, some significant advantages to perhaps
non-
bank institutions will result. And that will push assets out of
our good supervisory framework. So, I think that is a very
troublesome issue that needs careful attention.
Chairman Shelby. Professor Altman, I want to direct this
question to you and also to Mr. Ervin. Could you elaborate on
your statements regarding the procyclical aspects of the
proposal? The proposal could increase booms and busts during
the economic cycle, some people believe.
Mr. Altman. Yes. I skipped over it in the interest of time.
Thank you for giving me the opportunity.
Procyclicality is considered by some as a very serious
issue. Other people, including most of the regulators in the
United States, seem to think that it is not that serious of an
issue. I personally think that the New Basel II has all the
ingredients to increase procyclicality, perhaps dramatically.
It exists already. Banks cut back in difficult times, either
through credit rationing or----
Chairman Shelby. What could that do to the economy?
Mr. Altman. That is what I am getting to.
Chairman Shelby. Okay.
Mr. Altman. In times of stress, if the banks are motivated
through Basel II higher capital requirements on losses and
downgrades, then they will lend even less. That is exactly the
time that we need it. One of the reasons we have had so many
bankruptcies and defaults in the last couple of years is that
banks were much too liberal in 1993 through 1998, and that is
procyclicality. They were fat and happy. That is good. But they
made too many bad loans as a result of being that. So that is
the procyclicality.
I recommend that we consider a smoothing of the capital
requirements, that more capital be required in good times and
less capital in bad times. You have got to prompt corrective
action anyway to move into the banks if they have got too
little capital. So exercise that, but reduce that
procyclicality by much more proactive action on the part of the
regulators.
Chairman Shelby. The bottom line is capital is important.
Let's face it.
Mr. Altman. Absolutely.
Chairman Shelby. Mr. Ervin, do you have a comment?
Mr. Ervin. I would support what Professor Altman said. I do
believe this could be potentially quite important. When we have
looked at our own bank and looked at our own behavior, we think
this could significantly affect the number of loans we choose
to make. And if you pull liquidity out of the market in the
tough times, that will have an effect in the economy. I am not
smart enough to know exactly what effect that will have, but we
do believe that could be a potentially substantial impact, and
it is something that the U.S. regulators in particular need to
be aware of and have contingency plans to make sure we do not
fall into that trap.
Chairman Shelby. Mr. Blakely.
Mr. Blakely. Mr. Chairman, if I may, we disagree on the
issue of procyclicality because the whole concept of
procyclicality depends on the banks' operating at the absolute
bare minimum level of capital. I do not think any bank worth
its salt is going to be doing that. Most banks will maintain a
buffer zone of capital above the bare minimum that should----
Chairman Shelby. Well, banks get nervous during stress
time.
Mr. Blakely. Banks do get nervous, but banks will not stop
lending during a recessionary environment. Lending is the
business that most banks are in. It is our lifeline of revenue.
Chairman Shelby. Excuse me a minute. They might not quit
lending, but they curtail some lending or tighten up on credit.
We know that. And that can exacerbate an economic downturn, a
cyclical thing, as Dr. Altman references. Is that correct?
Mr. Altman. That is correct.
Mr. Blakely. I will not deny that banks do curtail in some
areas of lending. But, again, banks do need to continue lending
in order to keep their revenue stream going. Also during a
recessionary environment, you are afforded the opportunity to
get better underwriting standards and to get better pricing.
There is an old saying that says that the best of loans are
made in the worst of times and the worst of loans are made in
the best of times. And that is true right now.
We do not stop lending. We may become a little bit more
cautious, and as long as we have that buffer zone of capital,
we will continue to lend when opportunities present themselves.
Chairman Shelby. Thank you very much, Senator Sarbanes, for
your indulgence.
Senator Sarbanes. I will be very brief because it is
getting late, and this panel has been with us quite a while.
I am a little concerned or perhaps even confused by some of
what I have been hearing from this panel. First of all, is
there anyone on the panel who disagrees with the statements
made by the previous panel that Basel I is deficient and does
not constitute a framework within which we should continue to
function? That is what I understood the previous panel to say.
If you disagree with that, you should correct me. But is there
anyone at the table who disagrees with that.
Mr. Hartigan. No.
Mr. Green. No.
Mr. Altman. No.
Ms. Petrou. No.
Mr. Ervin. No.
Mr. Blakely. No.
Senator Sarbanes. I think the reporter should indicate
that----
[Laughter.]
All right. Now, my next question is: Is there anyone at the
table who thinks the Basel II process should be terminated?
Mr. Hartigan. No.
Mr. Green. No
Mr. Altman. No.
Ms. Petrou. No.
Mr. Ervin. No.
Mr. Blakely. No.
Senator Sarbanes. The question then is: How do we carry
forward in the Basel II process to take into account some, if
not all, of the concerns, to review the concerns that have been
enunciated, and to try to weigh them and see what can be done?
Would that be a fair statement?
Mr. Hartigan. Could I start? I think that many of these
issues that we have discussed today really have to continue to
be broken down in a hierarchy of needs, and that some of the
issues in the implementation are scalable. I think that the
whole system of capital based on risk, and appropriate capital,
is where the industry is moving and, indeed, where we are being
encouraged to move, where the industry has been encouraged to
move by the regulators.
I think that the way to do it, Senator Sarbanes, is to
continue to test the observations and to make recommendations
based on valid tests. It may be that we do not meet the
timetable that is in place now. Timetables can be moved. But at
the end of the day, the system will be better off for a capital
system based on risk.
Mr. Green. Senator Sarbanes, I would say that this
Committee deserves a good deal of credit for holding this
hearing in a timely way before the comment period on the
Consultative Paper 3 is complete. The fact is, what you heard
here is unanimity in favor of the Basel process and support for
what the regulators are trying to get done generally. And yet
we all have issues that we would like to raise consistent with
our support. By holding the hearing, you have advanced the
dialogue of those constructive points of criticism.
We came here today to put on your radar screen and to
reiterate on the radar screens of the Basel Committee the
effects on the securitization market and the repurchase
agreement and securities lending market, and yet not be
critical of the Basel process. It is not even a fine line. We
can be wholly supportive of the Basel process and point out
things that still need to be dealt with before it is finalized
and fully implemented. So we thank you for what you have done.
Senator Sarbanes. Does anyone else want to add anything on
that point?
Mr. Blakely. I would just like to say that since the
initial draft of Basel II came out, there has been tremendous
progress that has been made on it, and that has come about
through work between the banking institutions themselves, as
well as the regulators trying to come together on common
ground.
As I look back retrospectively and look at the progress
that has been made, it is nothing short of phenomenal, and I
think by the time Basel II eventually arrives, it will be
better, much better than it is now. We do all agree that Basel
II is the right direction to go. We just do not necessarily
agree on certain aspects of it. But we will eventually get
there.
Senator Sarbanes. Well, Mr. Chairman, I--sorry, yes.
Mr. Ervin. If I may just expand on that, I would agree that
it has improved a lot, and I do agree that it is also through
hearings like this one that the public airing and the public
dialogue can be improved, and that has had a major impact on
the quality of the work recently released in CP-3.
I do think there is significantly more work to do in
balancing between not sticking with an outmoded regime or going
to Basel II as it exists today. I think as Professor Altman
says, we need to have the guts to take out the parts that do
not seem to be working, do not seem to be getting consensus,
and focus on the parts that work, get that implemented, and
take a more evolutionary approach. I think that will be a smart
way to move forward here.
Senator Sarbanes. Well, it is an interesting process, and I
just wanted to clarify that. I once was the Executive Director
of the Charter Revision Commission in Baltimore, and we had to
take our proposed charter to the City Council. The director of
finance, who was a very influential figure in the city
government, came to testify. He was in support of what we were
trying to do. But then he got in front of the City Council, and
he had a number of pinpointed criticisms to make of our work.
And by the end of his first session with the City Council, they
were convinced he was against the charter changes. We had to go
back, and he had to make very clear that he was basically in
favor of the charter changes, and he was just trying to improve
it here and there. And I wanted to be clear on that.
This Economist article I quoted earlier, illustrates the
difficult problem of this international relationship, said,
``For the past 5 years, the world's financial regulators have
been working on a new set of rules for bank capital called
Basel II. Getting this far has taken a lot of sweat and horse
trading. American bankers and regulators have been at the
forefront. American financial institutions have debated the
rule changes as keenly as anybody. Imagine, therefore, the
consternation of other committee members on learning how
America plans to treat the new rulebook,'' and then they went
on.
You know, we have gone far enough down this path that we
have, in a sense, that problem on our hands. On the other hand,
I do not think that should lead us not to insist on shaping
this thing in a way that people can look at it and say it did
not make sense. And I do think the regulators need, one, to
work to get themselves in alignment and, two, to continue to
interact with the private sector to address some of these
concerns so we develop a broader and deeper consensus about
what we are seeking to do. And in that respect, I think, Mr.
Chairman, that it was a very constructive contribution for you
to schedule this hearing.
Chairman Shelby. Thank you.
A couple of quick observations. Mr. Green, has there been
any work done on the effects of Basel II on securitization that
considers the downstream effects on consumer financing?
Mr. Green. Yes, that is our principal level of concern. As
they have currently defined regulatory requirements, we believe
that they are significantly too severe. And we think it could
have an effect on the ability of financial institutions who are
very active participants in that securitization market, not
just of originators of loans and----
Chairman Shelby. And have an effect on our economy, right?
Mr. Green. Absolutely, Mr. Chairman. All this high finance
and complexity really does boil down to how does it affect real
people. And the securitization market is one of those examples
as to how real people would be affected, whether you are buying
a home, refinancing a home, buying a car, have credit card
debt, or just simply need capital. That $2.7 trillion, $151
billion, and $51 billion in the United States, Europe, and Asia
of a market----
Chairman Shelby. That has to be addressed, doesn't it?
Mr. Green. Absolutely.
Chairman Shelby. If the Fed does not pull operational risk,
should we address it here in Congress, Dr. Altman? Should we
address operational risk here?
Mr. Altman. I am not sure of the proper forum for that. I
am not sure also that capital will--more capital will obviate
the possibility of massive losses due to fraud or some other
types of activity which is under operational risk. You put
another 0.6 percent of capital for operating risk, but then you
have got a rogue trader or some fraud action, and you have got
an institution being brought down, regardless of that other 0.6
percent.
Chairman Shelby. It has happened.
Mr. Altman. Yes, exactly, and it is going to happen again.
It has to happen again because that is the nature of fraud. I
do not know when it is going to happen, and I do not think this
extra 0.6 percent capital is going to do anything with respect
to operating risk.
Chairman Shelby. Ms. Petrou, you know a lot about capital.
What do you think? Do you agree with him or disagree?
Ms. Petrou. Yes, I do.
Chairman Shelby. You agree?
Ms. Petrou. Yes, I think it is a meaningless charge, and to
your question of where Congress should go, I think what you and
on the House side as well, these hearings are focusing
attention back on the policy issues that Basel raises. When you
have rules that are hundreds of pages long with the kinds of
formulas as Mr. Ervin has said, all of us at this table and
everywhere, we get into these ``how do we do it'' debates, and
the ``should we do it'' issues get lost. And I think you are
helping everyone focus back on the ``should it be done'' issue,
which is of paramount importance.
Chairman Shelby. Well, Senator Sarbanes and other Members
of the Committee have thought this is a very important hearing.
I know it is very technical in nature and probably boring to a
lot of people, but it will have a tremendous effect on our
economy down the road in our banking system.
I want to thank all of you for being here and being
patient, waiting through the first panel, but when we have
these types of panels, yours and the other one before you, it
is hard to leave.
Thank you very much.
The hearing is adjourned.
[Whereupon, at 12:59 p.m., the hearing was adjourned.]
[Prepared statements supplied for the record follow:]
PREPARED STATEMENT OF SENATOR WAYNE ALLARD
I would like to thank Chairman Shelby for holding this important
hearing today to discuss the changing supervisory needs of some of the
world's largest and most complex internationally active banks. The
nature and activities of these banking institutions have evolved since
the inception of the Basel I Capital Accords in 1988. It is necessary
that the standards and requirements of internationally active U.S.
banking organizations are subject to the appropriate standards and
requirements to ensure that they remain competitive, healthy, and well-
capitalized.
The varying capital requirements for internationally active banks
made the 1988 Basel I Accord necessary to address the problem of
competitive inequality by establishing uniform capital requirements. In
the last decade or so, securitization and the use of derivatives have
prompted the Basel Committee to reconsider the relevance of the 1988
Accord standards. Since 1999, the Committee has been working to develop
a more risk sensitive capital adequacy framework to replace Basel I.
For most banks in the United States, Basel I is more than adequate
in establishing a capital framework. For a few large, complex, and
internationally active banks, the framework needs adjustment in order
to maintain the health of banking organizations in the United States. I
would like to thank the regulators for their ongoing work and attention
to this critical issue. I also thank all of our witnesses for appearing
before the Committee today. I look forward to your testimony.
----------
PREPARED STATEMENT OF ROGER W. FERGUSON, JR.
Vice Chairman
Board of Governors of the Federal Reserve System
June 18, 2003
Chairman Shelby, Senator Sarbanes, Members of the Committee, it is
a pleasure to appear before you this morning on behalf of the Board of
Governors to discuss Basel II, the evolving New Capital Accord for
internationally active banking organizations. After 5 years of
discussion, the proposal is entering its final stage of public comment
and review, although there still remains additional steps to the
process.
Why Is a New Capital Standard Necessary?
The banking supervisors in this country believe that Basel I, the
current capital regime adopted in 1988, must be replaced for the
largest, most complex banks for three major reasons: (1) Basel I has
serious shortcomings as it applies to these large entities, (2) the art
of risk management has evolved at the largest banks, and (3) the
banking system has become increasingly concentrated.
Shortcomings in Basel I
Basel I was a major step forward in capital regulation. For most
banks in this country Basel I, as we in the United States have
augmented it, is now--and for the foreseeable future will be--more than
adequate as a capital framework. However, for the small number of
large, complex, internationally active banking organizations, Basel I
has serious shortcomings which are becoming more evident with time.
Developing a replacement to apply to these banking organizations is
imperative.
Basel I is too simplistic to address the activities of our most
complex banking institutions. The framework has only four risk
categories, and most loans receive the same regulatory capital charge
even though loans made by banks encompass the whole spectrum of credit
quality. The limited differentiation among the degrees of risk means
that the calculated capital ratios are too often uninformative and
might well provide misleading information for banks with risky or
problem credits or, for that matter, with portfolios dominated by very
safe loans.
Moreover, the limited number of risk categories creates incentives
for banks to game the system through capital arbitrage. Capital
arbitrage is the avoidance of certain minimum capital charges through
the sale or securitization of bank assets for which the capital
requirement that the market would impose is less than the current
regulatory capital charge. For example, credit card loans and
residential mortgages are securitized in volume, rather than held on
banks' balance sheets, because the market requires less capital, in the
form of bank credit enhancements, than Basel I requires in capital
charges. This behavior by banks is perfectly understandable, even
desirable in terms of economic efficiency. But it means that banks that
engage in such arbitrage retain the higher-risk assets for which the
regulatory capital charge--calibrated to assets of average quality--is
on average too low.
To be sure, through the examination process supervisors are still
able to evaluate the true risk position of the bank, but the regulatory
minimum capital ratios of the larger banks are becoming less and less
meaningful, a trend that will only accelerate. Not only are creditors,
counterparties, and investors less able to evaluate the capital
strength of individual banks from what are supposed to be risk-based
capital ratios, but also regulations and statutory requirements tied to
capital ratios have less meaning as well. Basel I capital ratios
neither adequately reflect risk nor measure bank strength at the larger
banks.
The Evolving State of the Art
Risk measurement and management have improved significantly beyond
the state of the art of 15 years ago, when Basel I was developed. Banks
themselves have created some of the new techniques to improve their
risk management and internal economic capital measures in order to be
more effective competitors and to control and manage their credit
losses. But clearly banks can go considerably further. One objective of
Basel II is to speed adoption of these new techniques and to promote
the further evolution of risk measurement and management by harnessing
them to the regulatory process.
Increased Heterogeneity and Concentration in Banking
Market pressures have led to consolidation in banking around the
world. Our own banking system has not been immune; it, too, has become
increasingly concentrated with a small number of very large banks
operating across a wide range of product and geographic markets. The
operations of these large banks are tremendously complex and
sophisticated, and they have markedly different product mixes. At the
same time, significant weakness in one of these entities has the
potential for severely adverse macroeconomic consequences. Although
their insured liabilities have been declining over time as a share of
their total funding, these organizations, with their scale and role in
payment and settlement systems and in derivatives markets, have
presented the authorities with an increasing moral hazard. It is
imperative that the regulatory framework should encourage these banks
to adopt the best possible risk-measurement and management techniques
while allowing for the considerable differences in their business
strategies. Basel II presents an opportunity for supervisors to
encourage these and other large banks to push their management frontier
forward.
Basel II
The proposed substitute for the current capital accord, Basel II,
is more complex than its predecessor for very good reasons. First, the
assessment of risk in an environment of a growing number of instruments
and strategies with subtle differences in risk-reward characteristics
is inevitably complicated
Second, the Basel II reform has several objectives: U.S.
supervisors are trying to improve risk measurement and management both
domestically and internationally; to link to the extent that we can the
amount of required capital to the amount of risk taken; to further
focus the supervisor-bank dialog on the measurement and management of
risk and the risk-capital nexus; and to make all of this transparent to
the counterparties that ultimately fund--and hence share--these risk
positions.
To achieve all these objectives, the framework for Basel II
contains three elements, called Pillars 1, 2, and 3. The most important
pillar, Pillar 1, consists of minimum capital requirements--that is,
the rules by which a bank calculates its capital ratio and by which its
supervisor assesses whether it is in compliance with the
minimum capital threshold. As under Basel I, a bank's risk-based
capital ratio under Basel II would have a numerator representing the
capital available to the bank and a denominator that would be a measure
of the risks faced by the bank, referred to as ``risk-weighted
assets.'' The definition of regulatory capital in the form of equity,
reserves, and subordinated debt and the minimum required ratio, 8
percent, are not changing. What would be different is the definition of
risk-weighted assets, that is, the methods used to measure the
``riskiness'' of the loans and investments held by the bank. It is this
modified definition of risk-weighted assets, its greater risk-
sensitivity, that is the hallmark of Basel II. The modified definition
of risk-weighted assets would also include an explicit, rather than
implicit, treatment of ``operational risk.''
Pillar 2 addresses supervisory oversight; it encompasses the
concept that well-managed banks should seek to go beyond simple
compliance with minimum capital requirements and perform for themselves
a comprehensive assessment of whether they have sufficient capital to
support their risks. In addition, on the basis of their knowledge of
industry practices at a range of institutions, supervisors should
provide constructive feedback to bank management on these internal
assessments.
Finally, Pillar 3 seeks to complement these activities with
stronger market discipline by requiring banks publicly to disclose key
measures related to their risk and capital positions. The concept of
these three mutually reinforcing pillars has been central to the Basel
II effort.
Scope of Application in the United States
The U.S. supervisory agencies will propose that most banking
organizations in this country remain under the existing Basel I-type
capital rules and would continue to have no explicit capital charge for
operational risk. Earlier I emphasized that Basel I had outlived its
usefulness for the larger banking organizations. How then did we
conclude that most of our banks should remain under rules based on the
old accord?
Banks Remaining Under Current Capital Rules
To begin with, most of our banks have relatively straightforward
balance sheets and do not yet need the full panoply of sophisticated
risk-management techniques required under the advanced versions of
Basel II. In addition, for various reasons, most of our banks now hold
considerable capital in excess of regulatory minimums: More than 93
percent have risk-weighted capital ratios in excess of 10 percent--an
attained ratio that is 25 percent above the current regulatory minimum.
No additional capital would likely have to be held if these
institutions were required to adopt Basel II.
Moreover, U.S. banks have long been subject to comprehensive and
thorough supervision that is much less common in most other countries
planning to implement Basel II. Indeed, U.S. supervisors will continue
to be interested in reviewing and understanding the risk-measurement
and management processes of all banks. Our banks also disclose
considerable information through regulatory reports and under
accounting rules and requirements of the Securities and Exchange
Commission; they already provide significant disclosure--consistent
with Pillar 3 of Basel II.
Thus, when we balanced the costs of imposing a new capital regime
on thousands of our banks against the benefits--slightly more risk
sensitivity of capital requirements under, say, the standardized
version of Basel II for credit risk, and somewhat more disclosure--it
did not seem worthwhile to require most of our banks to take that step.
Countries with an institutional structure different from ours might
clearly find universal application of Basel II to benefit their banking
system, but we do not think that imposing Basel II on most of our banks
is either necessary or practical.
Banks Moving to Basel II
We have an entirely different view for our largest and most
complicated bank-
ing organizations, especially those with significant operations abroad.
Among the most important objectives of both Basel I and the proposed
Basel II is to promote competitive consistency of capital requirements
for banks that compete directly in global markets.
Another important objective has been to encourage the largest
banking organizations of the world to continue to incorporate into
their operations the most sophisticated techniques for the measurement
and management of risk. As I have noted, these entities use financial
instruments and procedures that are not adequately captured by the
Basel I paradigm. They have already begun to use--or have the
capability to adopt--the techniques of modern finance to measure and
manage their exposures; and because substantial difficulty at one of the largest banking organizations could have significant effects on global
financial markets, all of the largest banks should be using these
procedures. In our view, prudential supervisors and central bankers
would be remiss if we did not address the evolving complexity of our
largest banks and ensure that modern techniques were being used to
manage their risks. The U.S. supervisors have concluded that the
advanced versions of Basel II--the Advanced Internal Ratings-Based (A-
IRB) approach for measuring credit risk and the Advanced Measurement
Approaches (AMA) for measuring operational risk--are best suited to
achieve this last objective.
Under the A-IRB approach, a banking organization would have to
estimate, for each credit exposure, the probability that the borrower
will default, the likely size of the loss that will be incurred in the
event of default: And--where the lender has an undrawn line of credit
or loan commitment to the borrower--an estimate of what the amount
borrowed is likely to be at the time a default occurs. These three key
inputs--probability of default (PD), loss given default (LGD), and
exposure at default (EAD)--are inputs that would be used in formulas
provided by supervisors to determine the minimum required capital for a
given portfolio of exposure. While the organization would estimate
these key inputs, the estimates would have to be rigorously based on
empirical information, using procedures and controls validated by its
supervisor, and the results would have to accurately measure risk.
Those banks that are required, or choose, to adopt the A-IRB
approach to measuring credit risk, would also be required to hold
capital for operational risk, using a procedure known as the Advanced
Management Approach (AMA) to establish the size of that charge. Under
the AMA, banks themselves would bear the primary responsibility for
developing their own methodology for assessing their own operational
risk capital requirement. To be sure, supervisors would require that
the procedures used are comprehensive, systematic, and consistent with
certain broad outlines, and must review and validate each bank's
process. In this way, a bank's ``op risk'' capital charge would reflect
its own environment and controls. Importantly, the size of the charge
could be reduced by actions that the bank takes to mitigate operational
risk. This provides an important incentive for the bank to take actions
to limit their potential losses from operational problems.
Determining Basel II Banks
To promote a more level global playing field, the banking agencies
in the United States will be proposing in the forthcoming Advance
Notice of Proposed Rulemaking (ANPR) that those U.S. banking
organizations with foreign exposure above a specified amount would be
in the core set of banks that would be required to adopt the advanced
versions of Basel II. To improve risk management at those organizations
whose disruption would have the largest effect on the global economy,
we would also require the same of banks whose scale exceeds a specified
amount. That is, banks meeting either the foreign exposure criterion or
the asset size criterion would be required to adopt the advanced
versions of Basel II, although most banks meeting one criterion also
meet the other.
Ten U.S. banks meet the proposed criteria to be core banks and thus
would be required, under our proposal, to adopt A-IRB and AMA to
measure their credit and operational risks, respectively. As they grow,
other banks could very well meet the criteria and thus shift into the
core group in the years ahead. We would also permit any bank that meets
the infrastructure requirements of A-IRB and AMA--the ability to
quantify and develop the necessary risk parameters on credit exposures
and develop measurement systems for operational risk exposures--to
choose Basel II. Banks that choose to use A-IRB and AMA would need to
consider several factors, including the benefits of Basel II relative
to its costs, the nature of their operations, the capital impact, and
the message they want to send their counterparties about their risk-
management techniques. We anticipate that after conducting such a
review, about 10 or so large banks now outside the core group would
choose to adopt Basel II in the near-term. Thus we expect about 20
banks to adopt the advanced version of Basel II before or shortly after
the initial implementation date.
Over time, other large banks, perhaps responding to market pressure
and facing declining costs and wider understanding of the technology,
may also choose this capital regime, but we do not think that the cost-
benefit assessment would induce smaller banks to do so for a very long
time. Our discussions with the rating agencies confirm they do not
expect that regional banks would find adoption of Basel II to be cost
effective in the initial implementation period. Preliminary surveys of
the views of bank equity security analysts indicate that they are more
focused on the disclosure aspects of Basel II rather than on the scope
of application. To be clear, supervisors have no intention of
pressuring any of the banks outside the core group to adopt Basel II.
The 10 core banks that would be required to adopt Basel II,
together with the approximately 10 self-selecting banks that we
anticipate would adopt it before or shortly after the initial
implementation date, today account for 99 percent of the foreign assets
and two-thirds of all the assets of domestic U.S. banking
organizations, a rate of coverage demonstrating the importance of these
entities to the United States and global banking and financial markets.
These data also underscore our commitment to international competitive
equity and the adoption of best-practice policies at the organizations
critical to our financial stability while minimizing cost and
disruption at our purely domestic, less-complicated organizations.
Issues
Bankers have identified three key areas of concern: Cost,
competitive equity, and Pillar 1 treatment of operational risk.
Cost
Implementing A-IRB and AMA in this country is going to be expensive
for the small number of banks for which it will be required, for other
banks choosing it, and for the supervisors. For the banks, the greatest
expense would be establishing the mechanisms necessary for a bank to
evaluate and control its risk exposures more formally. The A-IRB
approach would not eliminate losses: Banks are in the business of
taking risk, and where there are risks, there will be losses. But we
believe that the better risk-management that is required for the A-IRB
and AMA would better align risk and return and thereby provide benefits
to bank stakeholders and the economy. And, more risk-sensitive capital
requirements would assist in ensuring that banks would have sufficient
capital to absorb losses when they do occur. The cost-benefit ratio
looks right to the supervisors.
This ratio is further enhanced because attributing to Basel II all
the costs associated with the adoption of modern, formal risk-
management systems is a logical fallacy. The large banks that would be
required, or that would choose, to adopt A-IRB and AMA must compete for
funding in a global marketplace and thus already have adopted many of
these processes and would continue to develop them even without Basel
II. The new accord may well appropriately speed up the adoption
process, but overall, the costs of adopting these processes are being
forced on these banks not by Basel II but by the requirements of doing
business in an increasingly complex financial environment. In any
event, the ANPR will include questions designed to quantify the cost of
implementing Basel II.
Competitive Equity
A second key concern is competitive equity. Some are concerned that
the U.S. supervisors would be more stringent in their application of
Basel II rules than other countries and would thereby place U.S. banks
at a competitive disadvantage. To address this concern, the Basel
Agreement establishes an Accord Implementation Group (AIG), made up of
senior supervisors from each Basel member country, which has already
begun to meet. It is the AIG's task to work out common standards and
procedures and act as a forum in which conflicts can be addressed. No
doubt some differences in application would be unavoidable across
banking systems with different institutional and supervisory
structures, but all of the supervisors, and certainly the Federal
Reserve, would remain alert to this issue and work to minimize it. I
also emphasize that, as is the case today, U.S. bank subsidiaries of
foreign banks would be operating under U.S. rules, just as foreign bank
subsidiaries of U.S. banks would be operating under host-country rules.
Another issue relates to the concern among U.S. Basel II banks of
the potential competitive edge that might be given to any bank that
would have its capital requirements lowered by more than that of
another Basel II bank. The essence of Basel II is that it is designed
to link the capital requirement to the risk of the exposures of each
individual bank. A bank that holds mainly lower-risk assets, such as
high-quality residential mortgages, would have no advantage over a
rival that held mainly lower-quality, and therefore riskier, commercial
loans just because the former had lower required capital charges. The
capital requirements should be a function of risk taken, and, under
Basel II, if the two banks had very similar loans, they both should
have a very similar required capital charge. For this reason,
competitive equity among Basel II banks in this country should not be a
genuine issue because capital should reflect risk taken. Under the
current capital regime, banks with different risk profiles have the
same capital requirements, creating now a competitive inequity for the
banks that have chosen lower-risk profiles.
The most frequently voiced concern about possible competitive
imbalance reflects the ``bifurcated'' rules implicit in the U.S.
supervisors' proposed scope of application: That is, requiring Basel II
through A-IRB and AMA for a small number of large banks while requiring
the current capital rules for all other U.S. banks. The stated concern
of some observers is that the banks that remained under the current
capital rules, with capital charges that are not as risk sensitive,
would be at a competitive disadvantage compared to Basel II banks that
would get lower capital charges on less-risky assets. The same credit
exposure might have a lower regulatory minimum capital charge at a
Basel II bank than at a Basel I bank. Of course, Basel II banks would
have higher capital charges on higher-risk assets and the cost of
adopting a new infrastructure, neither of which Basel I banks would
have. And any bank that might feel threatened could adopt Basel II if
they would make the investment required to reach the qualifying
criteria.
But a concern remains about competitive equity in our proposed
scope of application, one that could present some difficult trade-offs
if the competitive issue is real and significant. On the one hand is
the pressing need to reform the capital system for the largest banks
and the practical arguments for retaining the present system for most
U.S. banks. Against that is the concern that there might be an
unintended consequence of disadvantaging those banks that would remain
on the current capital regime.
We take the latter concern seriously and will be exploring it
through the ANPR. But, without prejudging the issue, there are reasons
to believe that little if any competitive disadvantage would be brought
to those banks remaining under the current capital regime.
The basic question is the role of minimum regulatory capital
requirements in the determination of the price and availability of
credit. Economic analysis suggests that regulatory capital should be
considerably less important than the capital allocations that banks
make internally within their organization, so-called economic capital.
Our understanding of bank pricing is that it starts with economic
capital and the explicit recognition of the riskiness of the credit and
is then adjusted on the basis of market conditions and local
competition from bank and nonbank sources. In some markets, some banks
will be relatively passive price takers. In either case, regulatory
capital is mostly irrelevant in the pricing decision, and therefore
unlikely to cause competitive disparities.
Moreover, most banks, and especially the smaller ones, hold capital
far in excess of regulatory minimums for various reasons. Thus, changes
in their own or their rivals' minimum regulatory capital generally
would not have much effect on the level of capital they choose to hold
and would therefore not necessarily affect internal capital allocations
for pricing purposes.
In addition, the banks that most frequently express a fear of being
disadvantaged by a bifurcated regulatory regime have for years faced
capital arbitrage from larger rivals who were able to reduce their
capital charges by securitizing loans for which the regulatory charge
was too high relative to the market or economic capital charge. The
more risk-sensitive A-IRB in fact would reduce the regulatory capital
charge in just those areas where capital requirements are too high
under the current regime. In those areas, capital arbitrage has already
reduced the regulatory capital charge. The A-IRB would provide, in
effect, risk-sensitive capital charges for lower-risk assets that are
similar to what the larger banks have for years already obtained
through capital arbitrage. In short, competitive realities between
banks might not change in many markets in which minimum regulatory
capital charges would become more explicitly risk sensitive.
Concerns have also been raised about the effect of Basel II capital
requirements on the competitive relationships between depository
institutions and their nondepository rivals. Of course, the argument
that economic capital is the driving force in pricing applies in this
case, too. Its role is only reinforced by the fact that the cost of
capital and funding is less at insured depositories than at their
nondepository rivals because of the safety net. Insured deposits and
access to the Federal
Reserve discount window (and Federal Home Loan Bank advances) let
insured depositories operate with far less capital or collateralization
than the market would otherwise require of them and far less than it
does require of nondepository rivals. Again, Basel II would not change
those market realities.
Let me repeat that I do not mean to dismiss competitive equity
concerns. Indeed, I hope that the comments on the ANPR bring forth
insights and analyses that respond directly to the issues, particularly
the observations I have just made. But, I must say, we need to see
reasoned analysis and not assertions.
Operational Risk
The third key area of concern is the proposed Pillar 1 treatment of
operational risk. Operational risk refers to losses from failures of
systems, controls, or people and will, for the first time, be
explicitly subject to capital charges under the Basel II proposal.
Neither operational risk nor capital to offset it are new concepts.
Supervisors have been expecting banks to manage operational risk for
some time, and banks have been holding capital against it. Under Basel
I both operational and credit risks have been implicitly covered in one
measure of risk and one capital charge. But Basel II, by designing a
risk-based system for credit and operational risk, separates the two
risks and would require capital to be held for each separately.
Operational disruptions have caused banks to suffer huge losses
and, in some cases, failure here and abroad. At times they have
dominated the business news and even the front pages. Appendix 1 to
this statement lists the 10 largest such events of recent years. In an
increasingly technology-driven banking system, operational risks have
become an even larger share of total risk; at some banks they are the
dominant risk. To avoid addressing them would be imprudent and would
leave a considerable gap in our regulatory system.
A capital charge to cover operational risk would no more eliminate
operational risk than a capital charge for credit risk eliminates
credit risk. For both risks, capital is a measure of a bank's ability
to absorb losses and survive without endangering the banking and
financial system. The AMA for determining capital charges on
operational risk is a principles-based approach that would obligate
banks to evaluate their own operational risks in a structured but
flexible way. Importantly, a bank could reduce its operational-risk
charge by adopting procedures, systems, and controls that reduce its
risk or by shifting the risk to others through measures such as
insurance. This approach parallels that for credit risk, in which
capital charges can be reduced by shifting to less-risky exposures or
by making use of risk-mitigation techniques such as collateral or
guarantees.
Some banks for which operational risk is the dominant risk oppose
an explicit capital charge on operational risk. Some of these
organizations tend to have little credit exposure and hence very small
required capital under the current regime, but would have significant
required capital charges should operational risk be explicitly treated
under Pillar 1 of Basel II. Such banks, and also some whose principal
risks are credit-related, would prefer that operational risk be handled
case by case through the supervisory review of buffer capital under
Pillar 2 of the Basel proposal rather than be subject to an explicit
regulatory capital charge under Pillar 1. The Federal Reserve believes
that would be a mistake because it would greatly reduce the
transparency of risk and capital that is such an important part of
Basel II and would make it very difficult to treat risks comparably
across banks because Pillar 2 is judgmentally based.
Most of the banks to which Basel II would apply in the United
States are well along in developing their AMA-based capital charge and
believe that the process has already induced them to adopt risk-
reducing innovations. Presentations at a conference held late last
month illustrated the significant advances in operational-risk
quantification being made by most internationally active banks. The
presentations were made by representatives from most of the major banks
in Europe, Asia, and North America, and many presenters
enthusiastically supported the use of AMA-type techniques to
incorporate operational risk in their formal modeling of economic
capital. Many banks also acknowledged the important role played by the
Basel process in encouraging them to develop improved operational risk
management.\1\
---------------------------------------------------------------------------
\1\ Papers from that conference are available at http://
www.newyorkfed.org/pihome/news/speeches/2003/con052903.html.
---------------------------------------------------------------------------
Overall Capital and An Evolving Basel II
Before I move on to other issues, I would like to address the
concern that the combination of credit and operational risk capital
charges for those United States banks that are under Basel II would
decline too much for prudent supervisory purposes. Speaking for the
Federal Reserve Board, let me underline that we could not support a
final Basel II that we felt caused capital to decline to unsafe and
unsound levels at the largest banks. That is why we anticipate that the
United States authorities would conduct a Quantitative Impact Study
(QIS) in 2004 to supplement the one conducted late last year; I
anticipate at least one or two more before final implementation. It is
also why CP-3 calls for 1 year of parallel (Basel I and II) capital
calculation and a 2-year phase-in with capital floors set at 90 and 80
percent, respectively, of the Basel I levels before full Basel II
implementation. At any of those stages, if the evidence suggested that
capital were declining too much the Federal Reserve Board would insist
that Basel II be adjusted or recalibrated, regardless of the
difficulties with bankers here and abroad or with supervisors in other
countries. This is the stated position of the Board and our supervisors
and has not changed during the process.
Of course, capital ratios are not the sole consideration. The
improved risk measurement and management, and its integration into the
supervisory system, under Basel II, are also critical to ensuring the
safety and soundness of the banking system. When coupled with the
special U.S. features, such as prompt corrective action, minimum
leverage ratios, statutory provisions that make capital a prerequisite
to exercising additional powers, and market demands for buffer capital,
some modest reduction in the minimum regulatory capital for sound,
well-managed banks could be tolerable. I note that banks with lower
risk profiles, as a matter of sound public policy, should have lower
capital than banks with higher-risk profiles. Greater dispersion in
required capital ratios, if reflective of underlying risk, is an
objective, not a problem to be overcome.
I should also underline that Basel II is designed to adapt to
changing technology and procedures. I fully expect that in the years
ahead banks and supervisors will develop better ways of estimating risk
parameters as well as better functions that convert those parameters to
capital requirements. When they do, these changes could be substituted
directly into the Basel II framework, portfolio by portfolio if
necessary. Basel II would not lock risk management into any particular
structure; rather Basel II could evolve as best practice evolves and,
as it were, be evergreen.
The Schedule and Transparency
I would like to say a few words about the schedule. In a few weeks,
the agencies will be publishing their joint ANPR for a 90-day comment
period, and will also issue early drafts of related supervisory
guidance so that banks can have a fuller understanding of supervisory
expectations and more carefully begin their planning process. The
comments on the domestic rulemaking as well as on CP-3 will be critical
in developing the negotiating position of the U.S. agencies, and
highlighting the need for any potential modifications in the proposal.
The U.S. agencies are committed to careful and considered review of the
comments received.
When the comments on CP-3 and the ANPR have been received, the
agencies will review them and meet to discuss whether changes are
required in the Basel II proposal. In November, we are scheduled to
meet in Basel to negotiate our remaining differences. I fear this part
of the schedule may be too tight because it may not provide U.S.
negotiators with sufficient time to digest the comments on the ANPR and
develop a national position to present to our negotiating partners.
There may well be some slippage from the November target, but this
slippage in the schedule is unlikely to be very great.
In any event, implementation in this country of the final agreement
on Basel II would require a Notice of Proposed Rulemaking (NPR) in 2004
and a review of comments followed by a final rule before the end of
2004. On a parallel track, core banks and potential opt in banks in the
United States will be having preliminary discussions with their
relevant supervisors in 2003 and 2004 to develop a work plan and
schedule. As I noted, we intend to conduct more Quantitative Impact
Studies, starting in 2004, so we can be more certain of the impact of
the proposed changes on individual banks and the banking system. As it
stands now, core and opt in banks will be asked by the fall of 2004 to
develop an action plan leading up to final implementation.
Implementation by the end of 2006 would be desirable, but each bank's
plan will be based on a joint assessment by the individual bank and its
relevant supervisors of a realistic schedule; for some banks the
adoption date may be beyond the end of 2006 because of the complexity
of the required changes in systems. It is our preference to have an
institution ``do it right'' rather than ``do it quickly''. We do not
plan to force any bank into a regime for which it is not ready, but
supervisors do expect a formal plan and a reasonable implementation
date. At any time during that period, we can slow down the schedule or
revise the rules if there is a good reason to do so.
The development of Basel II has been highly transparent from the
beginning and will remain so. All of the consultative papers over the
past 5 years have been supported by a large number of public papers and
documents to provide background on the concepts, framework, and
options. After each previous consultative paper, extensive public
comment has been followed by significant refinement and improvement of
the proposal.
During the past 5 years, a number of meetings with bankers have
been held in Basel and in other nations, including the United States.
Over the past 18 months, I have chaired a series of meetings with
bankers, often jointly with Comptroller Hawke. More than 20 U.S. banks
late last year joined 365 others around the world in the third
Quantitative Impact Survey (QIS-3), which was intended to estimate the
effects of Basel II on their operations. The banking agencies last
month held three regional meetings with the bankers that would not be
required to adopt Basel II but might have an interest in choosing to
adopt the A-IRB approach and the AMA. Our purpose was to ensure that
these banks understand the proposal and the options it provides
them.\2\ As I noted, in about 1 month the banking agencies in this
country hope to release an ANPR that will outline and seek comment on
specific proposals for the application of Basel II in this country. In
the past week or so we have also released two White Papers to help
commenters frame their views on commercial real estate and the capital
implications of recognizing certain guarantees. These, too, are
available at our web site.
---------------------------------------------------------------------------
\2\ The documents used in these presentations are available at the
Board's web site, http://www.Federalreserve.gov/banknreg.htm (Documents
Relating to U.S. Implementation of Basel II).
---------------------------------------------------------------------------
This dialog with bankers has had a substantive impact on the Basel
II proposal. I have attached to my statement a comparison of some of
the major provisions of Basel II as proposed in each of the three
consultative documents published by the Basel Committee on Bank
Supervision (Appendix 2). As you can see, commenters have significantly
influenced the shape and detail of the proposal. For example, comments
about the earlier proposed crude formulas for addressing operational
risk led to a change in the way capital for operational risk may be
calculated; banks may now use their own methods for assessing this form
of risk, as long as these methods are sufficiently comprehensive and
systematic and meet a set of principles-based qualifying criteria. That
is the AMA. The mechanism for establishing capital for credit risk has
also evolved significantly since the first Consultative Paper on the
basis of industry comments and suggestions; as a result, a large number
of exposure types are now treated separately. Similarly, disclosure
rules have been simplified and streamlined in response to industry
concerns.
At this stage of the proposal, comments that are based on evidence
and analysis are most likely to be effective. Perhaps an example of the
importance of supporting evidence in causing a change in positions
might be useful. As some Members of this Committee may know, the
Federal Reserve had concluded earlier, on the basis of both supervisory
judgment and the available evidence, that the risk associated with
commercial real estate loans on certain existing or completed property
required a capital charge higher than the capital charge on other
commercial real estate and on commercial and industrial loans. In
recent weeks, however, our analysis of additional data suggested that
the evidence was contradictory. With such inconsistent empirical
evidence, we concluded that, despite our supervisory judgment on the
potential risk of these exposures, we could not support requiring a
higher minimum capital charge on commercial real estate loans on any
existing or completed property, and we will not do so.
In the same vein, we also remain open minded about proposals that
simplify the proposal but attain its objective. Both the modifications
of the proposals in CP-3 and the changes in U.S. supervisory views, as
evidenced by the commercial real estate proposal, testify to the
willingness of the agencies, even at this late stage of the process, to
entertain new ideas and to change previous views when warranted.
Summary
The existing capital regime must be replaced for the large,
internationally active banks whose operations have outgrown the simple
paradigm of Basel I and whose scale requires improved risk-management
and supervisory techniques to minimize the risk of disruptions to world
financial markets. Fortunately, the state of the art of risk
measurement and risk management has improved dramatically since the
first capital accord was adopted, and the new techniques are the basis
for the proposed new accord. In my judgment, we have no alternative but
to adopt, as soon as practical, these approaches for the supervision of
our larger banks.
The Basel II framework is the product of extensive multiyear
dialogs with the banking industry regarding evolving best practice
risk-management techniques in every significant area of banking
activity. Accordingly, by aligning supervision and regulation with
these techniques, it provides a great step forward in protecting our
financial system and that of other nations to the benefit of our
citizens. Basel II will provide strong incentives for banks to continue
improving their internal risk-management capabilities as well as the
tools for supervisors to focus on emerging problems and issues more
rapidly than ever before.
I am pleased to appear before you today to report on this effort as
it nears completion. Open discussion of complex issues has been at the
heart of the Basel II development process from the outset and will
continue to characterize it as Basel II evolves further.
<GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT>
PREPARED STATEMENT OF JOHN D. HAWKE, JR.
Comptroller of the Currency
U.S. Department of the Treasury
June 18, 2003
Introduction
Chairman Shelby, Senator Sarbanes, and Members of the Committee,
thank you for inviting the Office of the Comptroller of the Currency
(OCC) to participate in this hearing on proposed revisions to the 1988
Capital Accord developed by the Basel Committee on Banking Supervision
(Basel Committee). I welcome the efforts of the Committee to focus
attention on these critical issues. The health of the U.S. commercial
banking system is a critical element to a strong economy. Thus, it is
essential that any regulatory changes that might affect the condition
and competitiveness of our banking system be fully understood and
carefully evaluated by the banking industry, the U.S. Congress and the
American public.
The 1988 Accord, referred to as Basel I, established the framework
for the risk-based capital adequacy standards applicable to
internationally active commercial banks in all of the G-10 countries,
and it has been adopted by most other banking authorities around the
world. U.S. banking and thrift agencies have applied the 1988 framework
to all U.S. insured depository institutions.
By the late 1990's, it became evident that Basel I had become
outdated. The increased scope and complexity of the banking activities
of our largest banking institutions over the last decade and the
unintended consequences of various provisions of the regulations,
severely undercut the utility of the Capital Accord. Basel I simply
does not provide a meaningful measure of the risks faced by large,
internationally active banks or the capital they should hold against
those risks.
Consequently, over the past several years, the Basel Committee has
been developing a more detailed and risk sensitive capital adequacy
framework to replace Basel I. The Committee's first draft document,
Consultative Paper No. 1 (CP-1), was issued in June 1999. It laid the
groundwork for the new capital adequacy framework (Basel II), but
provided few details. The Committee provided additional details on the
specifics of Basel II in its January 2001 issuance of Consultative
Paper No. 2 (CP-2). Although more detailed, CP-2 still left a number of
key issues unaddressed and unresolved. The Committee's most recent
paper, Consultative Paper No. 3 (CP-3), which I will discuss today, was
issued on April 29 of this year.
As work on these consultative papers has progressed, the Basel
Committee also has attempted to gauge the impact of its proposals on
the required capital levels of banking institutions through a series of
quantitative impact studies. In May, the Committee published the
results of the most recent assessment, the third quantitative impact
study (QIS-3). While the Committee concluded that the results were
generally in line with the objectives of Basel II, the QIS-3 data still
do not provide a sufficiently reliable estimate of the likely
regulatory capital requirements for banks subject to Basel II. More
work in this area is clearly warranted and I will discuss this later in
my testimony.
The Basel Committee has outlined an aggressive timeline for the
remaining actions leading to the adoption of Basel II. As a
consequence, the U.S. banking agencies, the agencies responsible for
the maintenance of capital adequacy standards for U.S. financial
institutions, are faced with a daunting task. While we will work
earnestly in this effort, the timeline should be seen as a means to an
end, not an end in itself. As will be highlighted in my testimony,
basic principles of safety and soundness demand that the banking
agencies have a more complete understanding of the consequences of this
proposal on the overall capital levels of affected institutions, the
competitive effects on our financial system, and associated compliance
costs and burdens before moving forward to finalize this proposal.
Our current primary focus in this effort is the development of U.S.
implementing regulations and policies. As I will discuss later, the OCC
and the other U.S. banking agencies will soon issue for comment
proposed revisions to U.S. risk-based capital regulations to reflect
the primary components of Basel II. Let me be absolutely clear about
the integrity of this rulemaking process--the OCC, which has the sole
statutory responsibility for promulgating capital regulations for
national banks, will not begin implementing a final Basel II framework
until we have conducted whatever cost-benefit and impact analyses that
are required, and fully considered all comments received during our
notice and comment process--as we would with any
domestic rulemaking. If we determine through this process that changes
to the proposal are necessary, we will not implement proposed revisions
until appropriate changes are made. We made this point quite clearly to
our Basel Committee colleagues before we agreed to go forward with CP-
3. Indeed, many of them will also have to go through their own internal
domestic processes before they can adopt the Basel II framework.
Current Basel Proposal
The Basel Committee deserves considerable credit for its
articulation of Basel II in CP-3. The proposal is still exceedingly
complex, but CP-3 is a clearer presentation of inherently difficult
material than its predecessors. This is an important step, since
regardless of the complexity of the proposal, it is important the
industry and other interested parties have a clear understanding of the
proposed Accord.
The attachment to this written statement provides a summary of the
substantive provisions contained in CP-3. As before, this iteration of
the proposed new Accord has three mutually reinforcing ``pillars'' that
comprise the framework for assessing bank capital adequacy. The first
pillar of the new Accord is the minimum regulatory capital requirement.
The Pillar 1 capital requirement includes a credit risk charge,
measured by either a standardized approach or one of the new internal
ratings-based (IRB) approaches (foundation or advanced), an operational
risk charge, and a market risk charge. Again, the attached document
provides a more detailed description of the various components of the
Pillar 1 charge.
Pillar 2 addresses supervisory review. It is, ``intended to ensure
not only that banks have adequate capital to support all the risks in
their business, but also to encourage banks to develop and use better
risk management techniques in monitoring and managing these risks.''
This pillar encourages supervisors to assess banks' internal approaches
to capital allocation and internal assessments of capital adequacy,
and, subject to national discretion, provides an opportunity for the
supervisor to indicate where such approaches do not appear sufficient.
Pillar 2 should also be seen as a way to focus supervisors on other
means of addressing risks in a bank's portfolio, such as improving
overall risk management techniques and internal controls.
The third pillar recognizes that market discipline has the
potential to reinforce capital regulation and other supervisory efforts
to ensure the safety and soundness of the banking system. Thus, the
Committee is proposing a wide range of disclosure initiatives, which
are designed to make the risk and capital positions of a bank more
transparent. As a bank begins to use the more advanced methodologies,
such as the Advanced IRB approach, the new Accord will require a
significant increase in the level of disclosure. In essence, the
tradeoff for greater reliance on a bank's own assessment of the
building blocks of capital adequacy is greater transparency.
U.S. Implementation Actions
It is important to recognize that the Basel Accord is not self-
executing in the U.S. Even when adopted by the Basel Committee, Basel
II will not apply to U.S. institutions unless and until the U.S.
banking agencies adopt regulations to implement it. In accordance with
the Administrative Procedure Act, 5 U.S.C. 551, et seq., the U.S.
banking agencies must publish notice and seek comment from all
interested persons on any such proposal, and must fully consider those
comments, before adopting a new capital regulation in final form.
Obviously, the OCC and the other Federal banking agencies intend to
comply fully with these requirements. The importance of this rulemaking
makes this comment process particularly critical to our success. Thus,
we welcome this process as a means for positive contribution to this
deliberative effort. We believe that the solicitation and assessment of
comments is a critical step in determining the feasibility,
effectiveness, and expected consequences of Basel II and related
domestic capital regulations.
Next month, the U.S. banking agencies expect to jointly issue an
Advance Notice of Proposed Rulemaking (ANPR) soliciting comment on
proposed revisions to the existing domestic capital adequacy
regulations that would implement Basel II. The ANPR will be largely
based on CP-3, and will provide a description of proposed
revisions to current capital regulations, while seeking comment on
outstanding or contentious issues associated with the proposal. The
ANPR will also request information on the cost of implementing the
proposal, and will seek comment on the competitive implications in both
domestic and international markets for banks of all sizes. In
conjunction with the ANPR, the banking agencies will also issue for
comment draft supervisory guidance articulating general supervisory
expectations for banks seeking to implement Basel II-compliant
methodologies for the Advanced Measurement Approach (AMA) to
operational risk and Advanced IRB for corporate credits. Recognizing
that CP-3 is a complex document, we understand the importance of
providing U.S. banks an opportunity to review and comment on U.S.
implementing documents as soon as practicable. By describing these
concepts within the context of our existing regulatory and supervisory
regime, the ANPR and draft guidance will provide a meaningful forum for
a full discussion of Basel II.
After assessing comments generated during the ANPR process, the
U.S. banking agencies will consider a complete cost analysis in
accordance with applicable rulemaking requirements, including the
standards of Executive Order 12866, discussed below, and will develop
specific regulatory language for a joint Notice of Proposed Rulemaking
(NPR). Again, the banking industry and other interested parties will
have an opportunity to comment on this fully articulated proposal
before any revisions to our capital regulations are finalized.
Let me now focus on two important, unique features of the U.S.
regulatory capital regime that will be highlighted in the ANPR and
NPR--the scope of application of Basel II and the content and structure
of the proposed revisions to the capital adequacy regulations. First,
the United States expects to set forth in the ANPR proposed criteria
for identifying which banks in the United States will be subject to the
new Accord. Despite language in the 1988 Capital Accord that permitted
a more limited application, U.S. banking and thrift agencies applied
the Basel framework to all U.S. insured depository institutions. As we
will highlight in the forthcoming ANPR, the U.S. agencies have
determined to apply Basel II concepts more narrowly. Specifically,
consistent with the focus of the Basel Capital Accord on banks that
compete in the global marketplace, we will propose applying Basel II
concepts on a mandatory basis only to large, internationally active
institutions that compete on a significant global basis with other
financial service providers. Other institutions will have the
opportunity to voluntarily opt into the Basel framework upon
application to, and approval by, their primary Federal supervisor.
Preliminary analysis by the U.S. agencies suggests that under the
narrow approach we are proposing, there are currently fewer than a
dozen U.S. banks that would be mandatorily subject to Basel II-based
regulatory capital requirements. Of course, the approach of requiring
only a small population of banks to comply with Basel II will be
subject to notice and comment in the ANPR and will be definitively
resolved only after the U.S. rulemaking process has been completed.
Second, in developing revisions to existing capital adequacy
regulations, U.S. banking agencies recognize that the revised
regulation, and interagency implementation policies, need not follow
the literal structure and language of Basel II. While consistent with
the objectives, general principles and core elements of the revised
Basel Accord, the language, structure, and degree of detail of U.S.
implementing documents may be very different from Basel II. These
implementation differences are reflective of the particular statutory,
regulatory and accounting structures and practices in place in the
United States. It is important to note that U.S. implementation actions
do not contemplate changes to many fundamental aspects of our
regulatory/supervisory process, including a focus on regular on-site
supervision, our prompt corrective action rules, and our minimum
leverage ratio for capital adequacy. As described more fully in the
attachment, the U.S. agencies will propose for notice and comment a
Basel II-based regime incorporating only the Advanced IRB approach for
credit risk, the AMA for operational risk, and the internal models
approach for market risk.
We are also very cognizant that in connection with this, or any
rulemaking, existing requirements may compel preparation of detailed
analysis of the costs, benefits, and other effects of our regulations,
depending on threshold determinations of whether the rulemaking in
question triggers the substantive requirements of particular statutes
or Executive Orders. Relevant requirements are set forth in the
Regulatory Flexibility Act (RFA), the Unfunded Mandates Reform Act of
1995 (UMRA) and Executive Order 12866 (E.O. 12866). Issuance of the
ANPR will help us identify and determine costs, benefits, and other
effects of the proposed rulemaking, for purposes of complying with
these requirements.
Timing
As I noted early on in my testimony, the Basel Committee timeline
presents a daunting task to both the U.S. banking agencies and the
banking industry. While it is clearly necessary to move forward in
addressing the acknowledged deficiencies in the current Basel Capital
Accord, the banking agencies must better understand the full range and
scale of likely consequences before finalizing any proposal. The list
provided below identifies the milestones the OCC must meet under the
current Basel II timeline. Each step is critical in a prudential
consideration of Basel II in the United States:
<bullet> Consideration of comments received by the Basel Committee on
CP-3. The comment period on this document concludes on July 31.
<bullet> Finalization, issuance, and consideration of comments on the
U.S. ANPR. Based on the current estimates, the notice and comment
period will run from July to
October.
<bullet> Finalization, issuance, and consideration of comments on
supervisory guidance on Corporate IRB and AMA methodologies. Based
on current estimates, the notice and comment period will run from
July to October.
<bullet> Development, issuance, and consideration of comments on
supervisory guidance on other substantive aspects of Basel II-based
regulations, especially including retail IRB. Based on current
estimates, the agencies hope to commence solicitation of comment on
this guidance by year-end 2003.
<bullet> Participation in the Basel Committee's consideration of Basel
II. Under the current timeline, the Committee is to consider
approval of Basel II in December of this year.
<bullet> Development, issuance, and analysis of results of additional
agency efforts to evaluate the prospective effects of Basel II
implementation. E.O. 12866 may compel the OCC and OTS to undertake
such analysis prior to the issuance of an NPR. Even without regard
to this requirement, however, it is essential that we have a
reliable estimate of the impact of Basel II on the capital and
competitive position of U.S. banks.
<bullet> Development, issuance, and consideration of comments on the
U.S. NPR. This document would only be issued after the Basel
Committee finalizes its consideration of Basel II. If the existing
timeline is maintained, solicitation of comment on the NPR would
commence no earlier than the first quarter of 2004.
<bullet> Development and issuance of a U.S. final rule and supervisory
guidance. Again, assuming the present timeline is maintained, our
best estimate for the issue date of a final rule implementing Basel
II is the third or fourth quarter of 2004.
<bullet> Completion of all necessary supervision-related steps to
implement Basel II-based regulations in advance of the presently
proposed December 2006 effective date. Most significantly, the
agencies need to determine whether each bank subject to Basel II-
based regulations has appropriate systems and procedures in place
to qualify for using the A-IRB and AMA.
Status of Basel Proposal--Outstanding Issues
In commencing an objective assessment of the status of Basel II, it
is important to reiterate and reaffirm the commendable work of the
Basel Committee, and in particular, the strong and intelligent
leadership of its former Chairman, William McDonough. The OCC firmly
supports the objectives of Basel II. These objectives constitute a
sound conceptual basis for the development of a new regulatory capital
regime and should continue to serve as a useful benchmark to gauge our
progress in this effort. Nonetheless, much of that conceptual basis has
not been tested in practice in any manner approaching the magnitude of
Basel II. We continue to be concerned about the potential for
unintended or unanticipated consequences of the Basel II proposals.
Implementation Challenges
At its foundation, the Basel II proposals permit qualifying
institutions to calculate their minimum risk-based capital requirements
by reference to their own internal systems and methodologies. While it
is the hallmark of Basel II, a greater alignment of internal risk
assessment with minimum regulatory capital derived through internal
models represents a radical departure from our existing regulatory
capital framework. As we will highlight in the ANPR and accompanying
guidance, this reliance on internal risk assessment systems mandates
changes in the way we structure our capital regulations and, in certain
important respects, how we conduct our supervisory activities. The
fundamental question for the banking agencies in assessing Basel II is
the issue the OCC has previously identified--whether the regime will
work in practice, as well as theory, as the basis for a regulatory
capital regime.
For bank supervisors and other external stakeholders to be in a
position to rely on a bank's internal process in the establishment of
regulatory capital requirements, there must be a high degree of
confidence that regulators can establish and enforce appropriate risk
measurement and management standards consistently across the banks
subject to a Basel II-based regime. The challenge for supervisors is to
create a verifiably accurate system that appropriately balances the
need for flexibility, to promote continued improvement in risk
management practices, with the need for objective standards, to ensure
consistency in application across institutions and supervisors, both
foreign and domestic.
The capital rule we implement must respect the evolutionary nature
of risk management. As regulators, we must acknowledge that we are
still in the relatively early days of model-based credit and
operational risk measurement and management. We must recognize the
inevitability of further innovation and improvements in this area. This
respect for the evolutionary nature of this discipline must then be
reconciled with the need for objective standards to ensure consistency
in application. Much of the detail and complexity within Basel II
derives from the need to establish more objective expectations for bank
rating systems, control mechanisms, audit processes, data systems, and
other internal determinations of risk by individual banks. In many
cases, this has led to the establishment of supervisory standards in
areas previously left to management discretion or supervisory judgment.
Not surprisingly, the regulatory community has struggled with the
establishment of these standards. Failing to achieve the proper balance
for these often conflicting objectives while moving forward with the
radically different Basel II-based regime can have dramatic
consequences. If our regulation and supervisory process is overly
flexible, bank internal calculations of capital adequacy may prove
insufficient, noncomparable, or both. If we err on the other extreme,
we establish an excessively
prescriptive supervisory regime that stifles innovation, imposes undue
regulatory burden, and inappropriately narrows the role of judgment.
This need to carefully balance dramatically opposed objectives,
together with the significant uncertainties that still exist about the
practical feasibility of these proposed changes to the Capital Accord,
raise doubts about the achievability of the timeframe established by
the Basel Committee.
Competitive Equality
A stated goal of the Basel Committee in developing Basel II was
that, ``the Accord should continue to enhance competitive equality.''
Realistically, we are not yet in a position to assess definitively the
full range of consequences from the implementation of Basel II,
including its effect on competitive equality in the global financial
marketplace. There are risks that Basel II may create or exacerbate
relative advantages between domestic banks and foreign banks; between
banks and nonbanks; and between large domestic banks and mid-size/small
domestic banks. It is imperative that the U.S. banking agencies remain
sensitive to these concerns and assess, to the extent possible, any
unintended consequences resulting from the implementation of Basel II.
One of the primary objectives of the Basel Committee itself is the
reduction of gaps and differences in international supervisory coverage
by national supervisory agencies, especially as it relates to large
internationally active banks that compete on a significant global basis
with other financial service providers. This principle of competitive
equality and a level playing field for international banks is an
admirable one, and an appropriate goal of the Committee's efforts. Yet,
the very complexity of the rules themselves calls this objective into
question. Bank supervision varies significantly from one country to
another in approach, intrusiveness, and quality. Is it realistic to
think that an enormously complex set of rules will be applied in an
evenhanded way across such a broad spectrum of supervisory regimes? For
example, the OCC has as many as 30 to 40 full-time resident examiners
in our largest banks. They are intimately involved as supervisors in
assessing the banks' operations and judging the banks' compliance with
a myriad of laws, rules, and guidelines. Some other countries may send
examiners in once a year to a comparably sized institution, or may
examine such an institution thoroughly only every 5 years, or may put
heavy reliance on the oversight of outside auditors.\1\
---------------------------------------------------------------------------
\1\ See, Daniel E. Nolle, ``Bank Supervision in the United States
and the G-10: Implications for Basel II,'' RMA Journal, June 2003.
---------------------------------------------------------------------------
It is fair to ask, I think, in which type of supervisory regime
detailed, prescriptive capital rules are more likely to be robustly and
reliably enforced. The Basel Committee has not undertaken to set
standards of supervision for member countries. Yet the attainment of
competitive equity among internationally active banks is a bedrock
principle of Basel II. Can we really achieve competitive equality
without addressing disparities in supervision, particularly when we are
operating on the assumption that the complex new rules we are writing will be applied in an evenhanded way throughout the world?
Another principle source of competition for many banks is not other
insured depository institutions, but nonbanks. This situation is
especially pronounced in businesses such as asset management and
payments processing. As you are aware, however, regulations implementing
Basel II-based concepts in the United States will apply only to insured depository institutions and their holding companies. While differences
in regulatory requirements for banks and nonbanks exist today, many institutions have voiced concern that implementation of Basel II may
unduly exacerbate the current differences. These concerns have been
mainly focused on the effects on competition from the application of the operational risk proposal and the enhanced disclosures required under
Pillar 3.
Finally, there is concern about the potential effect of Basel II on
the competitive balance between large and small banks. As implemented
in the United States, Basel II would result in a bifurcated regulatory
capital regime, with large banks subject to Basel II-based requirements
and small and mid-sized banks subject to the current capital regime.
This structure is premised on the belief that, to the extent possible,
regulations should reflect the size, structure, complexity, and risk
profile of banking institutions. The Basel II framework was developed
to address the unique risks of large internationally active
institutions. Mandatory application of such a framework to small banks,
with its associated costs, was deemed inappropriate. In fact, the
banking agencies sought comment from the banking industry, especially
smaller institutions, on the development of a simplified capital framework
specifically for noncomplex institutions.\2\ Industry comments were
overwhelming negative on the proposal--most institutions felt that the
cost of adopting a new regulatory capital regime outweighed any
potential benefits. Accordingly, the banking agencies tabled the
proposal.
---------------------------------------------------------------------------
\2\See Advance Notice of Proposed Rulemaking, Simplified Capital
Framework for Non-Complex Institutions, 65 FR 66193 (November 3, 2000).
---------------------------------------------------------------------------
With that said, the banking agencies need to continue to assess the
competitive effects of a bifurcated regulatory capital regime, and it
is one of the areas on which we will seek guidance in our ANPR. There
are several concerns in this regard. First, banks using a Basel II-
based regime may have a lower minimum capital requirement, allowing
those banks to grow and compete more aggressively with smaller banks
for both assets and liabilities. To be sure, banks subject to the New
Basel II requirements will incur very significant systems and
compliance costs in preparing for the new regime. These concerns are
discussed in more detail in the ``Calibration'' section below.
Moreover, banks using a Basel II-based regime may have significantly
higher or lower marginal regulatory capital charges than non-Basel
banks for some types of loan products, resulting in potential pricing
differentials. While Basel II might enable larger banks to compete more
effectively for high quality credits, it could also result in larger
concentrations of lower quality credits in smaller institutions.
Finally, the potential implications on industry consolidation are
simply not known. The banking agencies must continue to assess this
situation and, if warranted, take steps to mitigate adverse effects on
the competitive balance between large and small banks. We would be
seriously concerned if, as an unintended consequence of the
implementation of Basel II, we significantly alter the structure of
banking in the United States.
Calibration
The first objective of the Basel Committee in embarking on the
Basel II effort was to calibrate minimum capital requirements to bring
about a level of capital in the industry that, on average, is
approximately equal to the global requirements of the present Basel
Accord. That calibration was to be designed to provide an incentive to
banks to develop and maintain sophisticated and risk-sensitive internal
ratings-based systems.
In order to gauge its success in meeting that objective, the Basel
Committee attempted to measure the impact of its proposals on the
required capital levels of banking institutions through several
quantitative impact studies. On May 5, 2003, the Committee published an
overview of the results of its most recent assessment, the third
quantitative impact study (QIS-3). On the basis of QIS-3 results, the
Committee concluded that the aggregate results were generally in line
with the objectives established for Basel II.
Unfortunately, the QIS-3 data do not provide a reliable estimate of
the likely regulatory capital requirements for banks subject to Basel
II. And banks encountered several practical impediments to providing
accurate estimates of the effect of the proposals on their measured
ratios; thus, the estimated risk-based capital ratios were subject to a
substantial margin of error. For example, in many cases, existing bank
systems were not able to produce the data requirements necessary for
inputs required by the new Accord. In some areas, the QIS-3
instructions were not sufficiently clear or were misinterpreted, and in
other cases, the proposals were still in flux as banks were completing
the survey. Most important, QIS-3 was completed without the rigorous
supervisory validation and oversight that would occur when the proposal
actually takes effect.
A key concern is that focusing on the overall results of the QIS-3
exercise masks the wide dispersion of results for individual
institutions. In the U.S., measured against current risk-weighted
assets, the use of advanced approaches yielded results that ranged from
a decrease in regulatory capital requirements of 36 percent to an
increase of 43 percent. Similarly broad dispersions are found in a
great many of the underlying components that make up the total capital
requirement. While some dispersion of results in a truly more risk-
sensitive framework would be expected, we are not convinced that the
wide ranges indicated by QIS-3 can be explained by relative differences
in risk among institutions; it appears that comparability of QIS-3
results among different institutions may be severely lacking.
Finally, the quantitative studies that have been done to date have
been based on unilateral inputs from the participating banks. We and
other supervisors have had only very limited ability to review the
veracity of the results. I want to be clear that we have no reason to
believe that U.S. banks did not make every effort to provide results as
accurate as possible given the constraints they were operating under.
Nonetheless, it is certainly conceivable--I would say highly likely--
that the results might change significantly, and not necessarily in any
particular direction, when all the intricacies of real-world implementation come into play. It seems fair to assume that banks will have fewer
incentives to take conservative stances and greater incentives to exploit
any loopholes or gray areas in the final rules; the extent to which these effects might be offset, or exceeded by, greater supervisory oversight
is unknown.
Notwithstanding the significant uncertainties noted above, it
presently appears that the required capital levels of some U.S.
institutions could drop significantly, even taking into account the
temporary minimum floor capital requirements, discussed in the
attachment. The OCC does not believe that some reduction in minimum
regulatory capital requirements for certain institutions is, in and of
itself, an adverse feature of Basel II. Such a result is only
acceptable, however, if the reduction is based on a regulatory capital
regime that appropriately reflects the degree of risk in that bank's
positions and activities. Given the fact that relevant bank systems and
procedures are still in development, the OCC is not yet in a position
to make that determination as it relates to Basel II. As such, the OCC
is not yet comfortable allowing national banks to materially lower
their current capital levels simply on the basis of the output of the
currently proposed Basel II framework.
The OCC expects that an additional quantitative study will be
necessary after the Basel Committee's work on Basel II is completed.
Ideally, this should take the form of another global study by the Basel
Committee itself--that is, a QIS-4. However, even if the Basel
Committee does not undertake such a study, I believe that it is
absolutely essential that the U.S. agencies do so prior to the adoption
of final implementing regulations. I strongly believe that we cannot
responsibly adopt final rules implementing Basel II until we have not
only determined with a high degree of reliability what the impact will
be on the capital of our banks, but we have also made the judgment that
the impact is acceptable and conducive to the maintenance of a safe and
sound banking system in the United States.
Conclusion
As I have indicated, the OCC firmly supports the objectives of
Basel II--a more risk-sensitive and accurate capital regime. However,
in light of the issues that have been identified with the current
iteration of Basel II, the U.S. banking agencies must now determine how
best to proceed on this critically important issue. I believe the
following are essential elements in the agencies' consideration of
Basel II implementation within the United States.
First, the agencies need to move forward with the solicitation of
comments on a Basel II-related ANPR and associated guidance. That is
the most effective mechanism to have full and complete consideration of
the proposal from all interested parties. The solicitation of comments on a proposed regulatory and supervisory structure for Basel II implementation
will also permit supervisors to tangibly assess the feasibility of the proposal.
Second, the agencies need to undertake additional steps to evaluate
the costs, benefits, and other effects of the proposal before moving
forward with any final regulatory action. Frankly, we simply need
additional information to reasonably address the numerous issues,
concerns, and uncertainties associated with Basel II implementation. We
must better understand the likely consequences of this proposal on
overall capital levels of affected institutions, the competitive
effects on our financial
system, and associated compliance costs and burdens. In determining the
appropriate additional steps, the agencies should consider the
obligations imposed under E.O. 12866, the other statutory requirements
for consideration of costs and impact, lessons learned from QIS-3, and
perhaps, a U.S. version of QIS-4.
Third, as I have consistently reiterated, if we determine through
this process that changes to the Basel II proposal are necessary, the
U.S. agencies must pursue those changes, both domestically and in the
Basel Committee. In this regard, the U.S. agencies should not foreclose
consideration of alternative proposals that address the acknowledged
deficiencies of the 1988 Accord but that do not constitute such a
radical departure from our existing regulatory capital framework.
Fourth, the overarching consideration for supervisors in moving
forward on Basel II is the need to act in accordance with our primary
mission--to ensure the continued maintenance of a robust and safe and
sound banking system. We need to incent banks to continue to better
measure and manage the full panoply of risks they face and to make use
of new and evolving risk management practices. We must also
ensure that prudential consideration of safety and soundness principles
remain
paramount.
As I said in the beginning of my statement, the OCC, the agency to
which Congress has committed the authority to define capital
requirements for national banks, will not sign off on implementation of
a final Basel II framework until we have fully considered all comments
received during our notice and comment process. Given the importance of
this proposal, the significant issues that remain unresolved, and the
prospect that whatever emerges from this process is likely to govern
the financial landscape for years to come, we need to take whatever
time is necessary to develop and implement a revised risk-based capital
regime that achieves the stated objectives of the Basel Committee in
both theory as well as practice.
I am pleased to have had this opportunity to provide our views on
this important initiative, and I would be happy to answer any questions
you may have.
* * *
Attachment
Summary of Basel II: The Proposed New Accord Office of the Comptroller
of the Currency
The Basel Committee (the Committee) has been developing the new
Accord over the past 5 years. During that time, three full-scale
consultative papers (June 1999, January 2001, and April 2003) and
numerous working papers supporting various
elements of the new Accord have been released to the industry for
comment. This summary is intended to convey a general idea of the
structure and substance of the proposed new Accord, and does not
attempt to provide a complete analysis. It is based on the most recent
publications from the Basel Committee, notably the New Basel Capital
Accord (Consultative Document) which is out for comment until July 31;
the document can be found on the Committee's website at http://
www.bis.org/bcbs/index.htm.
The new Accord will include menus of approaches for measuring the
capital required for credit risk, market risk, and operational risk.
For credit risk and operational risk, each of the proposed approaches
is described briefly below; capital charges for market risk are
unchanged in the new Accord and are not discussed here. Some of the
approaches described are unlikely to be implemented in the United
States and have been noted as such. Moreover, based on preliminary
analysis by the U.S. agencies, currently there are less than a dozen
U.S. banks that would be mandatorily subject to Basel-based regulatory
capital requirements. While other banks would be permitted to opt in to
the Basel rules (subject to meeting prudential qualification
requirements), the U.S. capital rules will remain in place for the vast
majority of U.S. banks that either are not required to or do not opt to
apply the Basel II framework. Of course, any issues regarding U.S.
implementation of the new Accord will be definitively resolved only
after the U.S. rulemaking process has been completed.
The current structure of the Accord has been influenced by the
results of several quantitative impact studies (QIS), the most recent
of which was completed in December 2002. Approximately 20 U.S. banks
participated in the QIS exercise in
December and the results have been factored into the most recent
version of the Accord. Changes were made in several areas including the
treatment of retail credits, specialized lending, securitization, and
operational risk.
General Structure of the Proposed New Accord
The new Accord has three mutually reinforcing ``pillars'' that make
up the framework for assessing capital adequacy in a bank. The first
pillar of the new Accord is the minimum regulatory capital charge. In
order to calculate the capital charge under Pillar 1, banks will have
to determine the individual charges for credit, market, and operational
risk. The new Accord offers a series of options for calculating credit
and operational risk. Market risk will remain unchanged from a 1996
amendment to the Accord. The new options for credit and operational
risk were designed to be available to a wide range of banks, from
relatively simple to very complex. For credit risk, the Pillar 1
capital requirement includes both the standardized approach, updated
since the 1988 Accord, and the new Internal Ratings-Based (IRB)
approaches (foundation and advanced). Pillar 1 has been the focal point
of much of the discussion and comment from the industry on the new
Accord.
Pillar 2 covers supervisory review and banks' obligation to hold
sufficient capital vis-a-vis their risk profile. The pillar is,
``intended to ensure not only that banks have adequate capital to
support all the risks in their business, but also to encourage banks to
develop and use better risk management techniques in monitoring and
managing these risks.'' This pillar encourages supervisors to assess
banks' internal approaches to capital allocation and internal
assessments of capital adequacy. It provides an opportunity for the
supervisor to indicate where such approaches do not appear sufficient.
Pillar 2 is also a way to focus supervisors on other means of
addressing risks in bank's portfolio, such as improving risk management
techniques and internal controls.
The third pillar recognizes that market discipline has the
potential to reinforce capital regulation and other supervisory efforts
to ensure the safety and soundness of the banking system. Thus, the new
Accord proposes a wide range of disclosure initiatives, which are
designed to make the risk and capital positions of a bank more
transparent. As a bank begins to use the more advanced methodologies
for market and operational risk, the new Accord will require a
significant increase in the level of disclosure. In essence, the
tradeoff for greater reliance on a bank's own assessment of capital
adequacy is greater transparency. This pillar has been subject to
numerous changes as the Committee has worked to balance the need for
robust
disclosure with a recognition of the proprietary and confidential
nature of some of the information.
Capital for Credit Risk
Under Basel II, banks must select one of three approaches to
determine their capital for credit risk. The three approaches, from
simplest to most complex are: The standardized approach, the foundation
IRB, and the advanced IRB.
Standardized Approach
The 1988 Accord introduced the standardized risk-bucketing approach
for setting the minimum regulatory capital requirement, which is still
used in the United States today. The approach has been subject to
criticism that it lacks sufficient risk sensitivity. The revised
standardized approach under Basel II enhances the 1988 Accord by
providing greater, though still limited, risk sensitivity.
Key changes to create a more risk-sensitive framework include the
refinement and addition of risk buckets, the introduction of external
credit ratings, and a wider recognition of credit risk mitigation
techniques. Risk weights are still determined by category of the
borrower--sovereign, bank, or corporate--but within each of these
categories changes have been made to make the capital more reflective
of the riskiness of the asset category. For example, the risk weight on
mortgage loans has decreased from 50 percent to 35 percent and the risk
weight on certain retail credits has moved from 100 percent to 75
percent. Risk weights for externally rated corporate credits, currently
100 percent, will range from 20 percent to 150 percent. Sovereign risk
weights are no longer dependent upon whether a country is a member of
the Organization for Economic Cooperation and Development (OECD), but
rather on the external rating identified for the country.
The standardized approach is not likely to be implemented in the
United States. U.S. supervisors believe that credit risk measured under
the standardized approach of Basel II would generally not be
appreciably different than that measured under current rules for most
U.S. banks, and the marginal changes in capital requirements would not
justify the cost of implementation.
Internal Ratings-Based Approach (Foundation and Advanced)
The IRB approach represents a fundamental shift in the Committee's
thinking on regulatory capital. It builds on internal credit risk
rating practices used by some institutions to estimate the amount of
capital they believe necessary to support their economic risks. In
recent years, as a result of technological and financial innovations
and the growth of the securities markets, leading banking institutions
throughout the world have improved their measurement and management of
credit risks. These developments have encouraged the supervisory
authorities to devote greater attention to introducing more risk-
sensitive regulatory capital requirements, particularly for large,
complex banking organizations.
Banks must meet an extensive set of eligibility standards or
``qualifying criteria'' in order to use the IRB approach. Because the
requirements include both qualitative and quantitative measures,
national supervisors will need to evaluate compliance with them to
determine which banks may apply the new framework. The requirements
vary by both the type of exposure and whether the bank intends to use
the simpler foundation IRB framework or the more advanced IRB
framework. The requirements are extensive and cover a number of
different areas, including rating system design, risk rating system
operations, corporate governance, and validation of internal estimates.
A brief sample of actual criteria include:
<bullet> The board of directors and senior management have a
responsibility to oversee all material aspects of the IRB
framework, including rating and probability of default (PD)
estimation processes, frequency and content of risk rating
management reports, documentation of risk rating determinations,
and evaluation of control functions.
<bullet> A 1-year PD estimate for each grade must be provided as a
minimum input.
<bullet> Banks must collect and store historical data on borrower
defaults, rating decisions, rating histories, rating migration,
information used to assign ratings, PD estimate histories, key
borrower characteristics, and facility information.
As mentioned above, the requirements that a bank must meet are
partially dependent upon which of the two IRB approaches a bank will
use. The first methodology, called the foundation approach, requires
fewer direct inputs by banks and provides several supervisory
parameters that, in many cases, carry over from those proposed for the
standardized approach. For a variety of reasons, the United States does
not plan to introduce the foundation approach in its regulations. The
second approach, the advanced IRB approach, allows banks much greater
use of their internal assessments in calculating the regulatory capital
requirements. This flexibility is subject to the constraints of
prudential regulation, current banking practices and capabilities, and
the need for sufficiently compatible standards among countries to
maintain competitive equality among banks worldwide.
There are four key inputs that are needed under IRB, for both the
foundation and advanced approaches. The first element is the PD of a
borrower; the bank is required to provide the PD in both the foundation
and the advanced approaches. The second input is the estimate of loss
severity, known as the loss given default (LGD). The final two elements
are the amount at risk in the event of default or exposure at default
(EAD) and the facility's remaining maturity (M). LGD, EAD, and M are
provided by supervisors in the foundation approach, but must be
provided by banks operating under the advanced approach (subject to
supervisory review and validation). For each exposure, the risk weight
is a function of PD, LGD, and EAD.
The IRB approach envisions internal rating systems that are two-
dimensional. One dimension focuses on the borrower's financial capacity
and PD estimates that quantify the likelihood of default by the
borrower, independent of the structure of the facility. The other
dimension takes into account transaction-specific factors such as
terms, structure, and collateral. These characteristics would determine
the second dimension, that is, the LGD. Implicit in this treatment is
the assumption that when a borrower defaults on one obligation, it will
generally default on all its obligations. (This assumption is relaxed
with the IRB treatment of retail portfolios.)
Calculating the capital charge under the IRB approach involves
several steps. The first of these steps is the breakdown of the bank's
portfolio into five categories: Corporate (including commercial real
estate), retail, bank, sovereign, and equity. The IRB rules differ to
varying degrees across these portfolios. As a result, the IRB capital
charge is calculated by category, with the PD, LGD, and EAD inputs
potentially differing across these categories. Supervisory approval is
needed before banks can use the IRB approach for any of the five
categories. The minimum requirements described above were written to
apply across these five types of exposures.
Another important step is the determination by the bank of the PD's
for its loan grading categories. The PD of an exposure is the 1-year PD
associated with the borrower grade, subject to a floor of 0.03 percent
(excluding sovereigns). The determination of PD's for borrowers
supported by guarantees or credit derivatives is more
complex. Banks under the advanced approach would use their internal
assessments of the degree of risk transfer within supervisory defined
parameters, while those under the foundation approach would use the
framework set forth in the new credit risk mitigation provisions.
Overall, the PD must be, ``grounded in historical experience and
empirical evidence,'' while being ``forward looking'' and
``conservative.'' A reference definition of default has been developed
for use in PD estimation and internal data collection of realized
defaults.
Once the PD has been established, banks must then establish the
dimensions of LGD based on collateral and M. Under the foundation
approach, M is assumed to be 2.5 years. There are several options that
may be selected for the advanced approach, but in general, M is defined
as the greater of 1 year or the remaining effective maturity in years.
After the bank determines the PD's and LGD's for all applicable
exposures, these combinations can be mapped into regulatory risk
weights. The risk weights, which are calibrated to include coverage for
both expected and unexpected losses, are expressed as a continuous
function. The minimum capital charge is then determined by multiplying
the risk weight by the amount expected to be outstanding at the time of
default (EAD), and by 8 percent.
A final step in this process involves the ongoing review by the
supervisors of the systems used to develop the IRB capital charge.
Periodically, supervisors will need to validate these systems and
review the internal controls that provide the foundation for the IRB
approach. In addition, supervisors will also have to consider, under
Pillar 2, whether the amount of capital generated by the IRB approach
is commensurate with the bank's risk profile.
Implementation of the IRB Approach
In addition to the requirement that a bank meet the qualifying or
eligibility criteria, the new Accord requires that banks using the IRB
approach run parallel
systems for 1 year before implementation. This means that a bank
planning to implement the IRB approach in December 2006, will actually
have to begin calculating results as of December 2005, while continuing
to run its current systems.
Adjustments to the Capital Charge for Credit Risk
There are additional considerations that banks may have to factor
in when determining the capital charge for credit risk. These
additional considerations will further adjust required capital, outside
of the requirements of the different approaches to credit risk. The two
primary adjustments that might be made to the credit risk charge are
for credit risk mitigation and asset securitization.
Credit Risk Mitigation
The new Accord provides a measure of capital relief for certain
qualifying risk-mitigating techniques used by banks. However, it is
important to note that most of the credit risk mitigation proposals in
the new Accord are only directly relevant to the standardized or
foundation IRB approaches, which are not likely to be used in the
United States. In the advanced IRB approach, credit risk mitigation
must meet certain qualitative requirements, such as legal certainty. In
addition, specific proposals related to maturity mismatches and
backtesting requirements of certain model results are applicable to the
Advanced IRB approach. Otherwise, it is assumed that any credit risk
mitigation efforts will be factored into the PD's and LGD's assigned by
the bank.
With that caveat in mind, the section on credit risk mitigation in
the new Accord attempts to provide some rough approximations of the
risk reduction attributable to various forms of collateralized credit
exposures, guarantees, credit deriva-
tives, and on-balance sheet netting arrangements. The Committee has
proposed a conceptual approach to these risk mitigation techniques
that, while recognizing their risk reduction benefits, attempts to
capture the additional risks posed by such transactions.
The credit risk mitigation proposal provides both a simple and a
comprehensive approach to dealing with collateral. The proposal expands
the range of eligible collateral from that recognized in Basel I. It
also discusses the appropriate treatment for maturity mismatches
between the credit risk mitigant and the underlying credit exposure.
The proposal introduces ``haircuts,'' which the bank may estimate, to
cover the market price and foreign exchange volatility that may be
inherent in collateral. The proposal allows banks to greatly reduce the
capital requirements for exposures with large amounts of high quality
collateral. There are strict quantitative and qualitative factors that
must be met in order for a bank to be permitted to use its own haircut
estimates. The proposal encourages the use of credit risk mitigation by
expanding the type of collateral, guarantors, and transaction
structures that are recognized for capital reduction. Different types
of credit risk mitigation techniques pose different levels of
additional risk; the proposal incorporates flexibility that recognizes
these differences and adjusts the capital treatment accordingly.
Asset Securitization
Asset securitization is clearly an important issue in the United
States, as the securitization market is significantly greater than the
securitization market of any other Basel-member country. The Committee
believes that it is important to construct a more comprehensive
framework to better reflect the risks inherent in the many forms of
asset securitizations, including traditional and synthetic forms.
The securitization framework in the New Basel Accord applies
generally when there is a transaction that involves the stratification
or tranching of credit risk. The Committee has developed securitization
approaches for both standardized and IRB banks. The level of complexity
is significantly higher for IRB banks. The framework tries to focus on
the economic substance of the transaction, rather than on its
legal form.
Under the proposal for the treatment of securitizations by
standardized banks, the capital charge is generally determined by
multiplying the amount of the securiti-
zation exposure by the risk weight mapped to the long- and short-term
rating categories. Off-balance sheet exposures are subject to a
conversion factor before the
appropriate risk weight is applied. The proposal does allow for some
recognition of credit risk mitigants provided on securitization
exposures, but that recognition is permitted only when the bank meets a
series of stringent criteria.
Banks that adopt the IRB approach for credit risk are generally
required to use one of two methods for determining capital requirements
for securitization exposures. One method is the Supervisory Formula
Approach (SFA), under which capital is calculated through the use of
five bank-supplied inputs: The IRB capital charge on the underlying
securitized exposures (as if held directly on the bank's balance
sheet); the tranche's credit enhancement level and thickness; the
pool's effective number of loans; and the pool's exposure weighted
average loss given default (LGD). The second method is known as the
Ratings-Based Approach (RBA). Under this approach, capital is
determined by multiplying the amount of the exposure by the appropriate
asset-backed security risk weights, which depend on external rating
grades, short- or long-term. Granularity of the pool and the level of
seniority of the position are also considered.
The securitization proposal is one of the newest pieces of the
Accord and its potential impact on the industry is still being
assessed. In the December 2002 QIS exercise, banks were asked for the
first time to provide data on the relative impact of the proposals. The
QIS results did not provide entirely reliable results. However, the
Committee has responded to some of the concerns raised during the QIS
process by making changes to the securitization framework. One key
change was the introduction of a simpler approach for liquidity
facilities.
Operational Risk
One of the most significant changes in the new Accord is the
proposal for an operational risk charge. It is expected to represent,
on average, 10-15 percent of the total minimum regulatory capital
charge. The framework is based upon the following operational risk
definition: The risk of loss resulting from inadequate or failed
internal processes, people, and systems or from external events. This
includes legal risk, but excludes strategic and reputational risks.
The Committee has proposed three approaches to calculate the
operational risk charge, which represent a continuum of increasing
sophistication and risk sensitivity. The Basic Indicator Approach (BIA)
is the simplest of the three approaches; the capital charge is
determined by taking an alpha factor decided by the Committee and
multiplying it by an indicator, gross income. The next approach is
known as the Standardized Approach and is similar to the BIA, but
breaks out gross income into business lines. The Committee has
introduced an Alternative Standardized Approach to address some of the
concerns raised by the results of the December 2002 QIS exercise; this
is not a separate approach, but rather a modification to the
Standardized Approach. Because there is no compelling link between
these measures and the level of operational risk, the United States
does not plan to utilize the BIA or the Standardized Approach
(including the Alternative Standardized Approach) to determine the
capital charge for operational risk.
The Committee has made the most significant changes to the advanced
approach since it was originally introduced in January 2001. At that
time, the Committee envisaged a single, very prescriptive advanced
approach for operational risk, similar to credit risk. However, after
numerous comments from the industry, the Committee made substantive
changes in the proposal to reflect the evolutionary nature of the
operational risk framework. The Committee recognized that, unlike
credit risk, there are very little data and no internal systems
specifically designed to target operational risk; instead, banks and
supervisors rely primarily on internal controls to deal with a myriad
of banking risks that cannot be as readily quantified as credit and
market risks.
The Committee considered the comments and analyzed the state of the
art of operational risk and developed what is known as the Advanced
Measurement Approaches (AMA). Rather than prescribing one methodology,
the AMA will allow banks the option of designing the operational risk
measurement framework that best suits their institution, subject to
some broad criteria. The criteria will be the key to achieving a
certain level of consistency and comparability among institutions, as
well as providing a margin of comfort to supervisors who must assess
these differing systems. The criteria currently identified in the new
Accord include the need for internal and external data, scenario
analysis, and consideration of business environment and internal
control factors. Banks may also, under the AMA, consider the impact of
risk mitigation (such as insurance), again subject to certain criteria
set to ensure that the risk mitigants act as an effective capital-
replacement tool.
Temporary Capital Floors
Two floors that have been established for the Basel II framework.
In the first year of implementation, an institution's required minimum
level of regulatory risk-based capital cannot be less than 90 percent
of the minimum level of capital that would be required under the
Agencies' general risk-based capital rules. In the following year, an
institution's minimum level of regulatory risk-based capital cannot be
less than 80 percent of the minimum amount required under the Agencies'
general risk-based capital rules.
----------
PREPARED STATEMENT OF DONALD E. POWELL
Chairman, Federal Deposit Insurance Corporation
June 18, 2003
Thank you, Mr. Chairman and Members of the Committee. I welcome the
opportunity to testify on behalf of the Federal Deposit Insurance
Corporation on the New Basel Capital Accord (Basel II). The proposals
contained in the Third Consultative Paper (CP-3) recently published by
the Basel Committee on Banking Supervision, if adopted in the United
States, would easily rank among the most important pieces of banking
regulation in our Nation's history.
Introduction
Basel II would change bank capital regulation in the United States
in at least three important ways. First, rather than emphasizing simple
preset minimum numerical capital ratios, Basel II would allow
qualifying banks to use their own internal risk estimates as inputs to
regulator-supplied formulas with the supervisors
providing oversight and evaluation of the banks' ability to measure
risk. Second, the new framework would formally adopt a ``bifurcated''
capital system in the United States: One set of rules for the large,
complex, and internationally active institutions, and another set for
the balance of banks in the country. A third key change is that the
total minimum regulatory capital charge under the new framework will
include an explicit charge for operational risk. For those large
institutions that qualify, the new framework may lead to reduced credit
risk capital requirements for certain asset classes with additional capital held based on a flexible operational risk charge.
The FDIC supports the overall goal of Basel II, which is to create
regulatory capital standards that are more sensitive to the economic
substance of risks taken by these large banks, to limit their
opportunities for regulatory capital arbitrage and to encourage sound
risk management.
Over the years that Basel II has been under development, the Basel
Committee and the U.S. Federal supervisors have reached out to the
industry and the public for comment on how to more closely align the
proposed new framework with the ways that large banks measure risk.
There have been quantitative impact studies to assess the potential
impact on capital levels. We have been engaged in roundtables and
discussions. Over this time, various aspects of the new framework have
been refined and changed. Today, these refinements are reflected in CP-
3, which the Basel Committee recently released for additional comment.
The work in this country continues. The agencies intend to issue an
Advance Notice of Proposed Rulemaking (ANPR) that will suggest how CP-3
will be proposed for adoption in the United States and will seek
additional comments on all facets of Basel II. As in the past, it can
be anticipated that further changes to the framework may be required.
The FDIC is committed to an interagency process to achieve the overall
goals of Basel II and to fully understand its possible impact on bank
capital levels and competitiveness.
The goal of more closely tying regulatory capital to banks' own
internal assessment of risk is a good one. This goal is reached in part
by using regulatory capital formulas that are based on ways of
measuring credit risk and allocating internal capital that, to some
degree, are already in place in large banks. The term ``economic
capital'' is often used to refer to the amount of capital that should
be allocated to an activity according to the results of a numerical
loss analysis. Banks use models based on historical data and economic
analysis to estimate future losses and the amount of income, reserves,
and capital needed to ensure their portfolios conform to management's
target level of risk.
These calculations produce different results for different bank
activities. For example, the measured risk on residential mortgages
might be much less than the measured risk on construction loans. The
bank might use the economic capital measures to compute its risk-
adjusted returns on the two activities and to assist its pricing
decisions. This is a disciplined approach to risk management, and Basel
II establishes firm expectations for banks to be rigorous in this
respect. Basel II expands these risk management expectations beyond the
area of credit risk and into the realm of operational risk.
Tying capital requirements closer to risk and increasing the
incentives for disciplined risk management have the potential to
improve the safety and soundness of the U.S. financial system. The FDIC
supports enhancing the incentives for the largest banks in the United
States to strengthen risk management processes. Tying regulatory
capital closer to risk would reduce the incentives for banks to make
uneconomic decisions designed to reduce regulatory capital.
At the same time, the domestic impact of Basel II has not been
determined. Given current analysis, it seems likely Basel II will
confer some degree of regulatory capital benefits on the limited number
of banks that qualify, in exchange for their substantial investments in systems and infrastructure intended to improve risk management. The
critical issue for the safety and soundness of our financial system is
whether the improvements in risk management systems, and the resulting bank risk profiles, would justify the level of capital reductions that banks
might ultimately realize.
It is virtually impossible to quantify at this time the potential
changes in capital under Basel II. Basel II proposes floors by which
risk-based capital would be allowed to decline by at most 10 percent
the first year of implementation, and at most 20 percent the second
year. After the second year, Basel II does not impose a floor on the
minimum risk-based capital requirement. A quantitative study conducted
in the fall of 2002 showed a wide range of changes in capital
requirements for 19 large U.S. banks under the Advanced Internal
Ratings-Based (A-IRB) approach, with an average reduction in capital
requirements for credit risk of 17 percent. In this study, the
reduction in capital was offset by the operational risk capital charge,
which was substantial. However, the amount of this operational risk
charge was by necessity estimated using an approach that will not be
used in the United States.
The agencies understand that the results to date of the impact
studies do not provide a full picture of the possible impact of Basel
II. There are many moving parts to the proposal and the banks'
participation in the study was on a best efforts basis. Moreover, in
the United States, leverage ratio floors and the demands of the
marketplace would act as a constraint on the potential reduction in
actual capital.
Still, these initial estimated results show that the Basel II
formulas are potent instruments for affecting risk-based capital
requirements in the United States. This is a matter of great interest
to the FDIC and we are committed to working with the other banking
agencies as we move forward to more accurately assess the impact of the
proposed new standards.
A significant business challenge for the banking and thrift
agencies would be how to achieve interagency consistency in the
application of these complex rules. Required capital charges will
depend heavily on the ongoing judgments of banks and regulators about a
variety of specific risks.
In addition to understanding the impact of Basel II on capital
levels, we must also understand the significance of mandating two tiers
of regulatory capital standards--a bifurcated framework that will offer
competitors different regulatory capital charges for similar assets.
The critical issues in terms of the competitive playing field are
whether the direct competitors of a core group of about 10 large banks
would feel forced to opt in to the new framework for competitive
reasons, and whether banks in the tier below those able to opt in would
be at substantial competitive risk.
To resolve these fundamental issues satisfactorily, much hard work
remains. Given the magnitude of the issues, we must proceed carefully.
Capital Adequacy
The U.S. banking system has weathered the last 10 years better than
the banking systems of some other countries for a number of reasons.
One significant reason
is strong capital levels. Bank capital is subject to Federal
legislation and regulation because of its critical importance to the
health and well-being of the U.S. financial system. An adequate capital
cushion enhances banks' financial flexibility and their ability to
withstand periods of adversity. As insurer, the FDIC has a vital stake
in the adequacy of bank capital--as do our fellow regulators and all
U.S. taxpayers. Congress recognized this important principle when it
established the Prompt Corrective Action (PCA) requirements in the
Federal Deposit Insurance Corporation Improvement Act. A critical
aspect of the existing PCA regulations is the minimum
leverage capital requirement. To be considered well-capitalized, a bank
must have a ratio of Tier 1 capital-to-total assets (the leverage
ratio) of at least 5 percent. Banks with leverage ratios under 4
percent are considered undercapitalized. The agencies agree that
maintaining the minimum regulatory capital standards as
reflected in the current PCA legislation and existing implementing
regulations is very important.
Capital is not the only thing needed for safety-and-soundness. The
strong risk management that Basel II promotes is also essential. There
is no denying that banks with good risk management and a lower-risk
profile should be able to operate with somewhat less capital than more
risky banks. But there is also no denying that when the unexpected
happens, the hard-earned benefits of risk management can evaporate
overnight without adequate capital.
The sophistication of the measurement of economic capital can make
it easy to lose sight of the fact that, in reality, no one knows the
range of potential future losses for a given activity, or the
associated probabilities. Certain risk management practitioners express
great faith in the calculation of economic capital, and believe that
the regulatory capital standard should in all instances be less than
the economic capital amount. The idea behind this philosophy is that
banks tend to be forced out of low-risk activities where regulatory
capital requirements exceed economic capital requirements. It is this
belief that gives us concern about a clash of expectations about Basel
II between a number of prominent risk management practitioners on the
one hand, and the FDIC and our fellow bank regulatory agencies on the
other.
As the regulators move forward to finalize our views on Basel II,
we need to proceed cautiously. Where a proposal seems to run counter to
established U.S. supervisory practice, we need to ask whether the
established practice should be reexamined in light of the proposed new
rules, or whether the new rules need to be reexamined for U.S.
purposes.
Basel II is the object of intense scrutiny and comment. Changes
have been and will be suggested by banks in many areas, including the
treatment of commercial real estate, credit cards (and the related
issue of future margin income), mortgages, securitizations, and capital
recognition of certain risk-mitigating activities. The potential for
many moving parts could make it difficult to evaluate the capital
impact or the competitive impact of Basel II. Yet, we believe that we
must achieve a better understanding of these issues before the bank
regulatory agencies commit the United States to the new framework.
Interagency Consistency
Basel II would provide banks and supervisors some flexibility to
determine what capital would be held on an ongoing basis. The degree of
conservatism to apply to a particular situation would often be a
judgment call. Is the loss given default on a secured commercial loan
likely to be 20 percent or 40 percent? Capital for that loan would
double, or be cut in half, depending on the answer--and the answer
could well depend on a mix of historical data, the specific
underwriting methods used by individual banks and the specific
analytical techniques banks use to make their case. Supervisors would
need to validate--uniformly and consistently across banks--the answers
to such questions. In this new framework, regulators must be prepared
to challenge the modeled outputs of sophisticated risk measurement
systems of the largest U.S. financial institutions, a difficult and
demanding task. It will require courage and discipline to respond to
this new challenge.
Much progress has been made by the regulators and the industry in
deciding how this validation might be done. Interagency guidelines are
being drafted and implementation approaches are being discussed. The
FDIC has an active interest in the development of a sound approach to
ensure the consistent and uniform review of bank risk measurement
systems under Basel II.
A Level Playing Field
Capitalism, with its inevitable winners and losers, is about
competition. It is the job of the regulators to make certain that the
competition is fair. In our capitalist system, one of the key functions
of regulation is to ensure the rules do not display favoritism and that
the competitive struggle is carried out on equal terms. We need to
evaluate Basel II against this standard before committing to implement
it in the United States.
The proposed agreement raises several very important questions. The
fundamental question is what are the economic benefits of the
regulatory capital relief some banks might realize under Basel II?
Conversely, what are the costs of additional capital they might be
required to hold for certain activities? Would small or mid-sized
regional banks, unable to qualify for the new framework, become
acquisition targets of Basel II banks whose reduced capital has boosted
their returns on equity? Would a large credit card bank that must hold
capital for unused credit card lines be at a disadvantage to a non-
Basel bank that faces no such requirements? Would a securitizing
regional bank that is forced to deduct most of its retained interests
from capital be at a disadvantage to a Basel bank whose deductions from
capital would now be capped? What would be the ramifications of
significantly reduced capital requirements for Basel banks on specific
assets held by banks of all sizes, such as mortgage-backed securities
issued by the Federal Government sponsored enterprises?
The Basel II formulas are designed to work for large diversified
portfolios, and the capital requirements they produce might be too low
for most small banks. The Basel framework also requires significant
systems investments at a level likely beyond the reach of--and not
essential for--small institutions. Therefore, it is not practical to
think that any competitive concerns that may exist could be resolved
simply by allowing all banks access to the Basel framework.
To a large extent, the banking system in the United States is
already a two-tier system, with large financial institutions possessing
the vast majority of U.S. bank assets. Still, we must evaluate
thoroughly whether Basel II will unnecessarily disturb this current,
albeit divided, field of competition. Even though the industry may
already be divided between the large and complex and the small and less
complex, banking supervisors must understand fully whether Basel II
adds significant additional competitive pressures or would trigger
additional industry consolidation. The ANPR will seek input from all
interested parties, including banks that believe they will be
competitively harmed if they cannot embrace the Basel II framework.
Conclusion
An Advance Notice of Proposed Rulemaking will be issued this summer
and will reflect the United States banking and thrift agencies' views
on how Basel II would be adopted in the United States. More
importantly, it will present issues and concerns, and raise questions
to the industry and the public. The comments will pro-
vide invaluable insight to many of the key concerns being raised by the
agencies and by Congress.
Given the importance of these issues, it is vital that we treat the
implementation of Basel II in the United States as we would any other
proposed regulation--with a dose of skepticism, a willingness to
entertain the discussion of options, and a commitment to fully explore
potential costs and benefits before reaching a final decision. We need
to listen carefully to comments that will be received in the rulemaking
process to ensure we address these threshold issues.
It also is important that the financial services industry, the
Congress, and the banking agencies have a full opportunity to review
the response to the ANPR and achieve a better understanding of the
impact of this proposed agreement before we commit the United States to
the Basel II approach. The FDIC has no interest in delaying the
agreement and its implementation beyond what is necessary to address
the issues we have raised and to understand the impact of this new
system of capital regulation.
I have full confidence that this interagency process will work and
will arrive at an appropriate outcome. The FDIC will continue to remain
fully involved in this process and will work to ensure that the goals
of Basel II and of Congress are being met as the process moves forward.
Thank you for the opportunity to present the views of the FDIC.
----------
PREPARED STATEMENT OF JAMES E. GILLERAN
Director, Office of Thrift Supervision
June 18, 2003
Introduction
Good morning, Chairman Shelby, Senator Sarbanes, and Members of the
Committee. Thank you for the opportunity to discuss the proposed
revisions to the 1988 Capital Accord (Basel I) developed by the Basel
Committee on Banking Supervision (BSC). Although the Office of Thrift
Supervision (OTS) has been involved in the Basel process for some time,
we have only recently attempted to engage ourselves in the process
internationally. While we are very supportive of the Basel process,
there are numerous policy implications involved in the recently
proposed international capital standards for banking organizations in
the United States. These include issues that we all must strive to
understand and address. I welcome your efforts to highlight these pending
and important changes.
The proposed change in capital standards currently under
consideration arises from a third consultative paper, CP-3, recently
issued for public comment by the BSC. CP-3 is expected to result in the
New Basel Capital Accord, or Basel II. Basel II will directly affect the largest and most internationally active banking organizations around the world, including approximately 10 banking organizations in the United States. Basel II may also significantly impact, albeit indirectly, all other banking organizations around the world, including roughly 9,500 institutions in the United States. These institutions include large, medium, and small banks and thrifts that operate nationally, regionally, and at the community level, many of which compete domestically with our largest internationally active banking organizations.
Development of Basel II
Basel I, signed in 1988, addressed only the largest,
internationally active banks in G-10 countries and encouraged countries
outside the G-10 to adopt the framework for their banks that were
operating internationally. The underlying principles of Basel I,
however, were intended to apply to all banking organizations of any
size and activity. Thus, while OTS did not sign Basel I, we applied it
along with the other Federal banking agencies. Since Basel I, the four
banking agencies have developed risk-based capital standards consistent
with its underlying principles, but with modifications intended to
enhance risk sensitivity.
In connection with our involvement and experience with Basel I, OTS
has been monitoring for many years the work leading up to Basel II.
Because of the potential impact of Basel II on the institutions we
regulate, we stepped up our involvement in the Basel process. In
anticipation of the domestic application of Basel II, OTS is
participating fully in preparation of an interagency Advanced Notice of
Proposed Rulemaking (ANPR), with accompanying supervisory guidance, to
be published in the Federal Register in the near future. The initiative
will trigger the official kick-off of the national debate on the
subject of new international capital standards, but, as you are aware,
many of the issues raised by Basel II have already attracted
significant attention. While OTS has not been directly involved in the
international deliberations to date, our role on the domestic front--
particularly in the mortgage markets--provides us a unique and useful
perspective for this discussion.
In Basel I, the BSC identified two fundamental objectives at the
heart of its work on regulatory convergence. As the Committee stated,
first, ``the new framework should serve to strengthen the soundness and
stability of the international banking system; and [second,] the
framework should be fair and have a high degree of consistency in its
application to banks in different countries with a view to diminishing
an existing source of competitive inequality among international
banks.'' Although the BSC developed a far more detailed and risk-
sensitive capital adequacy framework in Basel II than in the original
Accord, it does not stray from the objectives set 15 years earlier. In
fact, the BSC expanded upon these objectives as a guide to its efforts
in producing the current proposal. In particular, the Committee
observed that Basel II should:
<bullet> Continue to promote safety and soundness and at least maintain
the current overall level of capital in the system.
<bullet> Continue to enhance competitive equality.
<bullet> Establish a more comprehensive approach to address risk.
<bullet> Contain approaches to capital adequacy that are appropriately
sensitive to risk.
<bullet> Focus on internationally active banks, although its underlying
principles should be suitable for application to all banking
organizations.
While the objectives for Basel II set forth by the BSC are
important to ensure consistency and competitiveness among
internationally active banking organizations, the impact of the
proposed changes may affect many other banking entities domestically.
It is important to encourage a thorough discussion among the
regulators, Congress, and the thousands of banking organizations in the
United States that may be affected, directly or indirectly, by Basel
II. Hearings such as this and the upcoming ANPR will help stimulate
this debate.
Overview of Basel II
Basel II contains three ``pillars'' that are intended to be
mutually reinforcing. Pillar 1 is a minimum regulatory capital
requirement; Pillar 2 addresses supervisory review; and Pillar 3 is
intended to promote risk and capital transparency. Briefly, a
description of these is as follows:
<bullet> Pillar 1 includes a credit risk component that is measured by
either a standardized approach or one of two internal ratings-based
approaches. The two ratings-based approaches or models are the
Advanced Internal Ratings-Based (A-IRB)
approach and the ``Foundation'' approach. Pillar 1 also includes an
operational risk component that has several optional approaches.
The centerpiece of the operational risk component of Pillar 1 also
permits use of an internal model, the Advanced Measurement Approach
(AMA).
<bullet> Pillar 2 is viewed by the BSC as a way for the banking
supervisors to attain better overall risk management and internal
controls at the banking organizations we regulate.
<bullet> Pillar 3 includes a wide range of disclosure initiatives
designed to make the risk and capital positions of banking
organizations more transparent.
Issues for Consideration
As I noted at the outset, OTS has only recently sought to be
involved internationally in the Basel process. While we are supportive
of this process and encouraged by the work completed so far, both
domestically and internationally, there are a number of issues that we
have considered regarding the application of Basel II in the United
States. In the following discussion, I highlight some of these issues.
Competitive Equality
Regardless of how we strive to explain Basel II, the extraordinary
technical detail at its core is substantial. Our banking organizations
will need to master the complexity of Basel II to provide effective
feedback during the upcoming ANPR comment process on the balance of its
burdens and benefits. As we proceed, we need their input to weigh
changes to our existing capital rules, and to assure ourselves that our
actions do not significantly alter the competitive landscape for all
U.S. banking entities. We want to assure that U.S. banking
organizations remain healthy, competitive, and well-capitalized.
The key principle underlying Basel II, and the basis for the
advancement from Basel I, is greater risk sensitivity. This principle
has as much meaning for a small community banking organization as it
does for a large internationally active institution. The challenge lies
in how to address this issue simultaneously for both types of banking
organizations, especially considering that under the proposed scope of
application in the United States, all but the few largest banking
organizations will not be ``Basel II banks.'' A significant issue in
this debate is whether we maintain consistent capital standards for all
banking organizations for lending activities that have the same risk
characteristics.
From our standpoint, maintaining competitive equality for community
banks is important, particularly as our economy is showing encouraging
signs of improvement. Community banking organizations play a
significant role in small business lending, which feeds new job
creation. ``Community banks are one of the key sources of credit and
other financial services to small businesses--the most prolific job
creating sector of our economy. Small businesses employ 60 percent of
the Nation's workforce and have created two-thirds of all the net new
jobs since 1970.'' \1\
---------------------------------------------------------------------------
\1\ Statement of Paul G. Merski, Chief Economist and Director of
Federal Tax Policy, Independent Community Bankers of America, before
the House Small Business Committee, March 1, 2002.
---------------------------------------------------------------------------
Another aspect of this issue that we must consider is the extent to
which we alter our existing capital rules, applicable to all banks, to
accommodate changes proposed by Basel II. For example, under Basel I,
the blunt-edged risk-based capital requirement for 1-4 family
residential mortgages (a 50 percent risk-weight, or 4 percent capital
requirement) is not commensurate with the historical risk associated
with residential mortgage lending in the United States. For residential
mortgage loans with relatively low loan-to-value ratios, a
substantially lower risk--weight is more reflective of loss experience.
By contrast, the Federal banking agencies have concluded that for some
concentrations of subprime loans, a significantly higher risk weight
than 100 percent--and therefore, a capital requirement higher than 8
percent might be more appropriate. While Basel II is intended to
enhance the risk sensitivity of our capital rules, it is important that
the proposed changes are truly
reflective of actual risk, as measured over an appropriate historical
timeframe.
Supervisory Effectiveness
Another important issue is the potential impact of Basel II on our
supervisory effectiveness. The U.S. bank regulatory system is
considered to be among the most comprehensive and admired in the world.
Capital requirements are only part of our multifaceted supervisory
response to ensure safety and soundness. Our supervisory system is
grounded in a regular program of on-site examinations complemented by
comprehensive and frequent reporting and off-site monitoring--a level
of supervisory review that may be unparalleled.
As we move forward with a relatively dramatic approach that places
a tremendous emphasis on capital, we must be careful not to minimize or
diminish the other supervisory tools and regulatory judgment that is
integral to our supervisory system. In particular, we should focus on
how Basel II fits within and improves our system, and how to strike the
right balance between capital rules and effective supervisory
oversight. In the end, sound regulatory judgment is the key to our
supervisory effectiveness and cannot be compromised.
Accountability in a Ratings-Based Capital Model
A corollary to this issue is the role of examiners and our
examination process in evaluating ratings-based models dictated in a
Basel II supervisory world. The application of Basel II in the United
States will include complex mathematical formulas and models used to
measure regulatory capital levels for our largest financial
institutions. While prior regulatory approval is required to use the
models, once obtained, an institution would effectively set its own
capital requirements. This would be based largely on inputs derived
from credit assessments from the institution's own credit risk and
operational risk models.
The accuracy and consistency of ratings is extremely important in
any ratings-based system. Numerous subjective decisions are made daily
by bank personnel regarding model inputs. These inputs involve
judgments made on items such as rating a loan's probability of default,
an estimate of loss given default, and the probability of a major loss
arising from an institution's operational risk. It is important to keep
in mind that these are human inputs, and are not infallible. Of
particular concern is how to account for the subjectivity of the
``human factor'' as we implement and apply Basel II.
Equally important is that we take the steps necessary to support
and train our examiners who will be expected to review the many
subjective decisions made under, and evaluate the mathematical models
of, Basel II. We must also consider how the Basel II models and
mathematical formulas reconcile with our existing rules, such as with
our asset risk classification and prompt corrective action rules. This
includes whether any of our existing rules, in addition to risk-based
capital, would have to be changed to accommodate Basel II.
Operational Risk
Another important issue is the operational risk capital charge in
Basel II. The concerns include the difficulty of trying to measure
something that cannot be readily modeled. Currently, the ability to
measure and quantify operational risk is less advanced than the
measurement and quantification of credit risk. In addition, the
boundaries between credit risk and operational risk are not always
clear. Another question is whether operational risk should receive a
more qualitative Pillar 2 supervisory review as opposed to the
quantitative Pillar 1 approach proposed in Basel II. This question is
significant because assessment of operational risk inherently involves
human judgment, which lies more squarely within Pillar 2.
There are also questions about the availability of good data to
measure operational risk. Under the AMA model of Pillar 1, the most
sophisticated institutions would use available external data to measure
risk and compute their own capital charge. While data may be readily
available for ordinary risk events that can be budgeted, truly high-
risk loss events occur infrequently. We must consider how to proceed
where there is a lack of readily available data for precisely the type
of risk for which capital may be most relevant to a particular institution
or group of institutions.
We will also want to consider the positive effect that an
institution's internal systems and controls have on operational risk
exposure. In computing their operational risk capital charge, it is
important to understand whether and how different institutions would
allocate capital appropriately for weaknesses in their internal systems
and controls, as well as the disincentives in doing so. This is
important to ensure both consistency and accuracy in the operational
risk capital charge.
Conclusion
Thank you, Chairman Shelby, Senator Sarbanes, and Members of the
Committee for the opportunity to testify on Basel II. As you are aware,
Basel II raises very significant issues not only for our very largest
banking organizations, but potentially for all our insured
institutions. I urge all of the Members of the Committee to remain
involved in this process going forward.
PREPARED STATEMENT OF MAURICE H. HARTIGAN, II
President and CEO
RMA--The Risk Management Association
June 18, 2003
Good morning, Mr. Chairman and Members of the Committee. Thank you
for inviting me to appear before the Committee to discuss the important
work under way to reform the 1988 Capital Accord, sometimes known as
the Basel Accord. My name is Maurice Hartigan and I am the President
and CEO of RMA--the Risk Management Association. RMA is a member-driven
professional association whose sole purpose is to advance the use of
sound risk principles in the financial services industry. RMA promotes
an enterprise-wide approach to risk management that focuses on credit
risk, market risk, and operational risk.\1\
---------------------------------------------------------------------------
\1\ Headquartered in Philadelphia, RMA has 3,000 institutional
members that include banks of all sizes as well as nonbank
institutions. They are represented in the Association by 16,000
commercial loan, credit, and risk management professionals in the 50
States, Puerto Rico, Canada, and numerous foreign cities, including
Hong Kong, Singapore, and London. RMA was founded in 1914 and formerly
known as Robert Morris Associates.
---------------------------------------------------------------------------
RMA has been actively involved in the reform of the 1988 Accord. In
1999, we formed the RMA Capital Working Group, consisting of the chief
economic capital officers of major banking institutions in North
America. Our group conducted research to demonstrate how banks use
their internal risk rating systems to assign economic capital. The RMA
Capital Working Group has produced a substantial body of research, and
has commented extensively on different drafts of the new Accord. It is
currently formulating comments to the most recent Basel Committee
Draft, the third consultative paper. This group also plans to comment
on the forthcoming interagency advanced notice of proposed rulemaking
that will deal with the U.S. implementation of the new Accord.
The main point I want to make to you today is that the New Basel
Accord will be a step forward for the U.S. and world banking
industries, provided it is modified as it is being finalized and
provided it is implemented flexibly. It will be a step forward because
it is directionally correct in improving the risk sensitivity of
regulatory minimum capital adequacy standards. But it must be modified
to ensure that it is not too conservative--that these are truly minimum
and not maximum capital standards--and to ensure that it is not too
prescriptive.
The purpose of the New Basel Accord is to make capital regulation
truly risk sensitive. The 1988 Accord was called the Risk-Based Capital
Accord, but it was that in name only. The new accord is designed to be
much more risk sensitive. It will require additional capital for
activities that are more risky and less capital for those that are not.
The 1988 Accord relied solely on a regulatory minimum capital standard.
In contrast, the new accord will be grounded on three principles or
``pillars'' as they are called: (1) Capital requirements, (2) enhanced
supervision, and (3) greater disclosure. This alone represents a
significant improvement.
Nonetheless, we have specific concerns in this area. Pillar 1,
which deals with the capital standard itself, must contain assurances
that Basel will evolve toward a full models-based approach for credit
risk, and it must avoid arbitrary specificity. Pillar 2, which deals
with the implementation of the standard through the process of
supervision, must allow regulators enough discretion to accommodate the
diversity of best practices in risk management today. Pillar 3, which
requires increased disclosure in order to provide greater market
discipline, must ensure that comparability is meaningful across the
varying international accounting regimes.
Our research to date suggests that the new accord, as proposed in
the third consultative paper, will require more overall capital than
many banks' internal risk rating systems require today, even though for
some banks and some portfolios, the new overall requirement will be
somewhat less than under the old accord.\2\ This will often be
inappropriate.
---------------------------------------------------------------------------
\2\ All of RMA's research and our formal responses to the
Consultatives Papers issued by the Basel Committee are available on our
Web site at www.rmahq.org. RMA's Securities Lending Committee has also
responded to the proposed treatment of securities lending activities,
and the work of that Committee is available to the public on our Web
site as well. Institutions participating in the research are listed on
the Web site and may hold views different from those expressed in this
testimony.
---------------------------------------------------------------------------
The new accord should represent a true minimum capital requirement.
For well run banks in normal times this implies that regulatory capital
levels should be set below a bank's economic capital based on best-
practice internal risk measurement procedures.
RMA and many others within the industry have long argued that
regulatory capital requirements should be more closely aligned with an
institution's own internal risk rating systems. Best-practice
institutions today assign internal capital to their portfolios and
measure performance on a risk-adjusted basis. Doing so enables them to
better price for risk and maximize shareholder value. Thus, good
business practices are consistent with the economic capital principles
underlying the proposed new accord.
The old Capital Accord requires best-practice institutions to
maintain two completely separate capital regimes: An internal system
that mirrors their true risk
profile, and a regulatory capital system that is a simple, flat capital
charge. Advanced-practice institutions do not manage risk based on the
current regulatory capital requirements. It would not be in their
shareholders' or their customers' best
interests to do so. This fact has certainly not gone unnoticed by the
regulators. Indeed, that is why reform of the 1988 Accord is under way.
For best-practice institutions, the possibility to align internal
capital estimation processes and regulatory capital procedures
represents a significant and meaningful improvement over the current
system. Turning this promising possibility into reality is not an easy
task, however. And that is why we are here before you today for a
review of the New Basel Accord.
The process to reform the 1988 Capital Accord has had a positive
impact on the development of risk measurement and management procedures
in the financial services industry. Moreover, the dialog between the
industry and its regulators surrounding Basel reform, while not without
frustration on both sides, has been useful and productive. While
outstanding issues clearly remain, some quite significant, continued
discussion with the industry is ongoing, and I would expect that to be
the case throughout the reform process and into the implementation
stage as well. Indeed, it may not be possible to resolve a number of
specific issues without an active two-way dialog between regulators and
the industry as the implementation process takes place.
Further discussion can only help promote innovation and investment
in best practices throughout the industry. It is for this reason that
the reform process must
continue. However, it must be framed as a work in progress. There
cannot be a prescribed ``end state'' for sound risk management
practices. Otherwise, the ink on the new accord would not be dry before
it became obsolete. This is much like the 1988 Capital Accord.
The quantitative analytics supporting sound credit risk measurement
and management are still evolving. Many of these emerging practices
were born out of the last economic downturn. The resilience of the
financial services industry over the past 3 years should not go without
comment. Many have credited the industry's success to the better risk
management practices established over the past decade. I would have to
agree.
One way to look at the new accord is that it is aimed at bringing
capital adequacy standards for credit and operational risk closer to
those for market risk. For some time, market risk has had a well-
established language among practitioners, strong analytics, and a
robust disclosure framework to support it. It is for this reason that
amendments to the 1988 Capital Accord were adopted in 1995 to
acknowledge the industry's advancement in the field.
Credit and operational risk management are still evolving to catch
up with market risk management. The practice of credit risk measurement
and management will no doubt benefit greatly over the next 2 years as
new data become available to populate quantitative credit risk modeling
systems. Operational risk measurement is a younger field, and it is
making strides on the back of our achievements in credit and market
risk.
Given the newness of the fields of study surrounding credit and
operational risk management, it is natural that regulators should be
prone to conservatism. But too much capital is just as bad as too
little capital. Too much capital will drive down the risk-adjusted
rates of return on a particular business line and cause bankers to lend
less than they otherwise would and should. This is not a good thing for
the shareholders of the bank, the loan customers of the bank, or the
general economy.
Furthermore, in our own review of Basel II, we find that some of
the new requirements are written in a very prescriptive fashion that
does not lend itself to allowing individual banks to employ a diversity
of best practices. Without such diversity we cannot have continued
evolution of best practices, and without evolution we could not have
had the improvements in risk measurement that have occurred over the
past decade.
I would now like to touch on two areas, which are somewhat more
technical in nature, about which we have great concern at present.
Foremost is the adoption by Basel of the same credit risk model as used
by advanced banks. A key parameter of these models--the degree to which
loan losses are correlated--is set by Basel, not by the empirical
research of best-practice banks. In some cases, such as certain retail
loan products, this critical parameter has been set too high by Basel,
causing the regulatory capital minimums to be too high. This is why RMA
has consistently stated in all our papers to the Basel Committee that,
``we believe strongly that the Internal Ratings-Based (IRB) approach
must be followed with a full internal models approach to capital.''
Second, RMA also has repeatedly argued that the Basel definition of
capital should be changed to conform to the definition used by the
industry. Indeed, Basel II will run into problems to the extent that
the Basel view of capital differs substantially from the view of
economic capital held by the industry. In the industry view, economic
capital is required only for unexpected loss (known as UL). The Basel
Committee has proposed that both UL and expected loss (known as EL) be
included in bank capital. RMA disagrees. For purposes of estimating
economic capital and capital adequacy, EL is covered by earnings
(spread and fees, net of expenses), and we believe that it is double
counting to include expected losses in capital. Indeed, if EL is
included in bank regulatory capital, it will clearly disadvantage banks
with their nonbank competitors.
RMA has additional technical concerns specific to the Third
Consultative Paper that we will address in our formal response.
To conclude, I would like to reiterate RMA's belief that the reform
process has helped advance the practice of sound risk measurement and
management within the industry. RMA is hopeful that the New Capital
Accord can be structured to encourage and enhance continued industry
innovation and that it will recognize the benefit that diversity of
practice within the industry provides.
Much good work has been done in conjunction with the new accord. It
has helped foster valuable research that has contributed to industry
innovation. It has also focused the industry and its regulators on the
need for additional research. Data limitations remain in a number of
key areas, and this is likely to be the case for some time. Again, this
only reinforces the fact that development of the new accord must be an
ongoing process.
Regulatory capital standards must evolve over time as practices
within the industry evolve. Otherwise, the industry and its regulators
will continue to face the same limitations embedded in the current
accord.
The only way for this goal to be achieved is to allow for the
development, over time, of a full Internal Models-Based approach to
bank capital. The proposals contained within the third consultative
paper, subject to the specific concerns we will be addressing shortly,
can represent a necessary start to this process. Thank you, and I would
be happy to answer any questions that you might have.
----------
PREPARED STATEMENT OF MICAH S. GREEN
President, The Bond Market Association
June 18, 2003
On the Basel II Capital Accord
The Bond Market Association is grateful for the opportunity to
testify on the Basel Committee on Banking Supervision's proposed New
Capital Accords, or Basel II. The Bond Market Association represents
securities firms and banks that underwrite, distribute and trade debt
securities domestically and internationally. Association member firms
account for in excess of 95 percent of all primary issuance and
secondary market activity in the U.S. debt capital markets. Through our
affiliate American and European Securitization Forums, we represent a
majority of the participants in the growing securitization markets in
the United States and Europe. The following comments focus on only
those issues related to Basel II that are most important to our
membership.
TBMA Supports the Goals of Basel II
The Association supports the Basel Committee's overall goal of
rationalizing the current risk-based capital regime, and aligning
regulatory capital requirements more closely with actual credit risk.
This goal is critically important to the global financial market, in
which capital flows are increasingly mobile and interdependent. Also,
we are grateful to the Federal Reserve Board and other U.S. bank
regulatory agencies for working with us to address the issues presented
by the proposed capital accord revisions that affect the domestic bond
market. While some of our concerns expressed previously were addressed
in the Basel Committee's third consultative paper (CP-3) on Basel II,
critical issues still remain.
The Basel Committee has an important role in promoting a prudential
but efficient allocation of capital throughout the banking system. An
updated regulatory capital regime can produce significant benefits,
including the promotion of fair global competition, the creation of
incentives for better internal risk management, and an economically
efficient allocation of capital to its most productive uses.
Although we support the direction and goals embodied in Basel II,
the revised Accord should not be viewed as the last word on regulatory
capital. In attempting to promulgate a universal rules-based system
that applies the same basic capital requirements to all regulated
financial institutions, Basel II--like its predecessor--is overly rigid
and prescriptive in certain critical respects. However, no such ``one-
size-fits-all'' regulatory capital regime can fully accommodate the
unique needs of these diverse institutions, or flexibly respond to
rapid changes in the financial markets in which they operate, without
suffering from this basic limitation. To overcome this deficiency, the
global financial community will need to move toward a broader reliance
on internal risk models, with supervisory review and approval, to
determine appropriate regulatory capital levels, and we encourage
financial market regulators to continue moving in this direction.
In the meantime, our comments focus on aspects of the proposed
Accord that we believe will, at least in the short-term, facilitate the
goal of aligning regulatory capital requirements more closely with
actual credit risk.
The Association has principally focused on two areas of the
proposed Basel Accord that significantly affect the bond markets--
securitizations and collateralized transactions, including securities
repurchase (repo) and securities lending arrangements. By creating more
risk-sensitive capital standards in these areas, Basel II can ensure
these transactions continue to serve as useful funding, liquidity, and
risk management tools.
Securitizations allow banks and other entities to obtain efficient
funding and to remove certain risks from their balance sheet so they
can be borne by other parties who desire such an exposure. Repo and
securities lending transactions also aid institutions in managing risk
by allowing them to readily obtain securities in order to meet delivery
obligations and to hedge exposures arising from separate transactions.
Setting regulatory capital charges too high for these increasingly
important and widely used arrangements threatens to distort economic
decisionmaking on the part of a financial institution. This has the
potential of eroding the significant benefits that consumers and
businesses alike realize from securitization and collateralized
transactions.
Background on the Securitization and the Repo and Securities Lending
Markets
Market Size
The past several years have seen phenomenal global growth of the
securitization market. Since 1995, the United States, European, and
Asian markets combined have grown from $497 billion to $2.9 trillion.
The U.S. market by itself has accounted for about 95 percent of that
volume.
The repo market has also shown steady growth over the same period.
Approximately $1.7 trillion in repo and securities lending transactions
were outstanding on average in 1996 and today an average $3.7 trillion
are outstanding. Hundreds of billions of dollars in repo transactions
are conducted daily to fund the positions of bond market participants
and allow the Federal Reserve Board to conduct open market operations.
Benefits of Securitization and Securities Lending and Repo Agreements
Securitization offers numerous benefits to consumers, investors,
regulators, corporations, and financial institutions.
Securitization has developed as a large market that provides an
efficient funding mechanism for originators of receivables, loans,
bonds, mortgages, and other financial assets. Securitization performs a
crucial role for the entire U.S. economy by providing liquidity to
nearly all major sectors including the residential and commercial real
estate industry, the automobile industry, the consumer credit industry,
the leasing industry, and the bank commercial lending and corporate
credit markets. In addition, securitization has provided a means for
banks to effectively disperse the risk of various positions they hold
throughout the broader financial market.
Securitization provides low-cost financing for banks and other
companies, lowers borrowing costs for consumers and homebuyers, adds
liquidity to banks' balance sheets, provides for efficient bank balance
sheet and capital management, and draws nontraditional sources of
capital to the consumer and corporate lending markets. The efficiencies
introduced by securitization are passed on to consumers and businesses
in the form of more widely available credit, lower interest rates, and
lower prices.
Securities Lending and Repurchase Transactions
Securities lending and repo transactions are integral to
maintaining liquidity in the capital markets. They are a secure and
flexible method of obtaining funding and securities for market
participants. For example, a market participant may purchase securities
which are then sold in a repo transaction, with an agreement to
repurchase such securities sometime in the future. The repo seller can
use the proceeds of this transaction to fund their initial purchase.
The repo buyer is able to invest funds for short periods in a safe and
liquid product. By providing a ready source of funding, repos and
securities lending transactions are critical to maintaining liquidity
in the bond markets. In the Treasury markets in particular, this
liquidity ensures that the Treasury's borrowing costs are kept low. In
short, America's capital markets operate as efficiently as they do
because wholesale market participants can use repos and securities
lending transaction to finance and hedge positions. The liquidity and
efficiency provided by the repo market lowers financing costs for the
Federal Government, homebuyers, corporations, and consumers.
Basel II's Impact on Securitization and the Repo and Securities Lending
Market
The Association applauds the goal of the Basel Accord to allow
financial institutions the ability to more closely tailor risk-based
capital requirements to the actual amount of risk present in financial
transactions. The proposed Accord, however, does not currently meet
this goal because under the proposal, institutions would be required to
maintain a higher level of capital than is warranted by the practical
risk of their positions. We have summarized below some of our principal
concerns in connection with the proposed capital treatment of
securitization exposures and repo and securities lending transactions.
The Association is continuing to develop additional quantitative and
analytical arguments to support these points, which will be submitted
prior to the July 31 comment deadline in response to the CP-3. The
Association will share our comments with Committee Members at that
time.
Securitization
The Association is troubled by the treatment in Basel II of certain
securitization products and positions. We are especially concerned that
if Basel II is not amended, the onerous capital charges imposed on
banks will discourage them from engaging in securitization
transactions. As a result, the benefits conveyed by a robust and
efficient securitization market would be diminished or lost.
Securitization Risk Weights Are Too High
The floor capital charge is too high for many types of
securitization positions, given their actual risk profile.
Subinvestment grade positions in particular attract too high a capital
charge under the proposals, given the actual credit risk they present.
Many of the key assumptions underlying securitization formulas and risk
weights are too conservative, and lack a proper theoretical or
empirical foundation.
By setting the floor requirements at a higher level than the actual
risk of a position, Basel II reduces incentives for banks to
participate in securitizations. This would lower incentives to conduct
transactions that actually lessen a bank's risk exposure and that allow
banks effectively to disseminate the risk of a particular transaction
throughout the marketplace.
Conservative Rules Result in Inordinately High Charges
In establishing rules governing the manner in which regulatory
capital computations are to be made, Basel II defaults to the
conservative alternative so often that--cumulatively--these rules
result in an inappropriately high capital charge for securitizations.
For example, given the general ability under Basel II to rely upon
qualified external ratings to determine regulatory capital
requirements, we believe that originators of securitized assets should
be able to use such ratings to determine risk weights, even if this
produces a lower capital charge than if the assets had not been
securitized. Originators do not have this ability under the proposal as
drafted. There are numerous other examples of excessively conservative
rules that--in the aggregate--produce unduly high capital charges for
securitizations.
Synthetic Securitizations Should Not Be Discriminated Against
Higher capital charges should not be levied against synthetic
securitizations, in comparison to traditional asset securitizations.
(Synthetic securitizations involve the bundling and securitization of
credit exposures, rather than the underlying financial assets.)
Synthetic securitizations are increasingly used by financial
institutions to manage their balance sheets, and provide additional
options and flexibility for risk management. Since the risk profile of
a synthetic asset is the same as for a cash asset, the risk-based
capital treatment should be equivalent. However, this would not be the
outcome under the proposals as currently drafted and, in several
respects, synthetic securitization positions attract inordinately high
capital charges.
Limited Credit Risk Inherent in Liquidity Facilities Should be
Recognized
In a number of important respects the Basel II proposals would
require financial institutions to hold disproportionately high levels
of capital against liquidity facilities they provide in connection with
securitizations. Such liquidity facilities are
extended by financial institutions to a variety of securitization
issuance vehicles, including but not limited to asset-backed commercial
paper conduits. Through the securitization market, these conduits
provide competitive short-term financing for a wide range of asset
originators. The performance history of liquidity facilities in this
context demonstrates that the likelihood of draws are extremely low,
and the incidence of credit losses negligible.
We believe that internal modeling is the most appropriate method
for determining regulatory capital for liquidity facilities. The key
operational requirement for liquidity facilities is that there be an
asset quality test that adjusts dynamically to preclude funding of
defaulted assets. Such a dynamic test is one that is built into
liquidity facilities that have been in the market for many years. This
has led to historical performance data showing the relatively low risk
of draws and of losses on such draws.
Under Basel II, if a liquidity position is not rated, we believe
that a bank should be able to look through to the risk weight assigned
to the underlying transaction that the liquidity supports if that
underlying transaction has been externally rated. Given that the
underlying transaction reflects the ultimate risk of a liquidity
position, we see no reason not to permit the reliance on the rating of
that transaction if a liquidity position itself is not rated.
Securities Lending and Repurchase Transactions
The Association is concerned that Basel II, as proposed, falls
short with regard to recognizing modern risk-management techniques as
they relate to secured transactions such as securities lending and
repurchase transactions. By failing to account for methods widely used
to mitigate risk exposure, capital charges for banks would not reflect
true balance sheet risk. The undue capital charges would ultimately
result in less efficient and more costly markets.
Encourage the Use of Cross-Product Netting as a Risk Management
Technique
The Association believes that the manner in which risk-based
capital requirements for repo and securities lending transactions are
calculated should be revisited along with the treatment of similar
collateralized transactions. The Association strongly believes that
transactions which present similar risks--and mitigate against similar
risks--as repo and securities lending transactions should be treated in
the same way for risk-based capital purposes. Many financial
institutions currently manage risks for all collateralized transactions
in a uniform manner.
After conforming the manner in which risk is calculated for repo
and securities lending transactions and other collateralized
transactions, the Basel Accord should take the next logical step and
allow for recognition of the netting of exposures across such
transactions. Currently, the Basel Accord contemplates netting only
between repo and securities lending transactions. It is widely
recognized that netting exposures across different transactions helps
financial institutions reduce their exposure to the risks such
transactions present. Providing incentives in the Basel Accord through
broader recognition of cross-product netting will provide added
incentives for financial institutions to implement this risk-reducing
practice.
Encourage the Use of Internal Risk Models
It is the Association's view that allowing financial institutions
to utilize internal risk models--as Basel II would--to determine
counterparty risk for collateralized transactions is a step in the
right direction. Basel II should not, however, dictate rigid rules as
to what models financial institutions must utilize in determining risk.
The Accord should allow financial institutions to utilize their own
risk models subject to the review and approval of national supervisors
under Pillar 2 of the Basel Accord. Otherwise, financial institutions
would likely devote resources to creating a model that may not
accurately capture the risks present in collateralized transactions. In
addition, the Association believes the Accord should not set out a
rigid backtesting regime for such models. (In this case, backtesting
refers to evaluating the performance of a model based on historical
data.) In any event, the backtesting regime currently set out in the
Basel Accord risks dissuading financial institutions from improving
upon their existing risk management practices through the use of
internal risk models by risking the imposition of significantly
increased capital charges. As currently contemplated, should the
results of the backtesting regime generate a number of mismatches or
``exceptions'' between estimated and actual data, an institution's
risk-based capital charge would be significantly increased. Such
backtesting regime--and its potentially punitive results--do not have
any commercially reasonable basis in relation to the repo and
securities lending markets.
Conclusion
The Association supports the overall goal of the Basel Committee to
align capital requirements for financial institutions more closely to
actual credit risk. While the revised Accord has the potential to move
regulatory capital requirements in the right direction, the Association
continues to have fundamental concerns with the proposal that must be
addressed to uphold the Basel Committee's stated goals
without causing economic distortions in the securitization, repo, and
securities lending markets.
The Association looks forward to continuing its dialog with the
Federal Reserve Board and other U.S. regulators on the issues we have
addressed above. We plan to offer formal comments on the third
consultative paper this summer, and when the Board issues its advanced
notice of proposed rulemaking describing the U.S. implementation of
Basel II, the Association will provide further input.
----------
PREPARED STATEMENT OF EDWARD I. ALTMAN, PH.D.
Max L. Heine Professor of Finance
Leonard N. Stern School of Business
New York University
June 18, 2003
Thank you for inviting me to the Senate hearings on the B.I.S.
recommended regulations on Capital Allocations for Bank Credit and
Other Assets--the so-called ``Basel II'' Accord. I have followed Basel
II's consultative papers since the first one was issued in June 1999.
We have submitted two formal commentaries to the Basel Commission on
Bank Supervision, primarily related to the first of the so-called
pillars of the new recommendations--capital adequacy based on the
specific risk characteristics of bank counterparties. Our major
comments were that the capital requirements related to expected and
unexpected losses from corporate and other loans should be based on the
actual historical experience of Loss Given Default (LGD) from the
corporate bond and bank loan markets. The original 1999 suggestions
bore little resemblance to actual performance and we pointed this out
fairly precisely. I am pleased to note that the latest version of Basel
II's capital requirements based on the riskiness of bank portfolios
does a much better job of relating the requirements to default
experience, although too little capital is still being required for the
most risky categories.
A problem with the suggested regulations, however, is the
complexity in determining capital requirements and the somewhat
arbitrary choice of modifications to the standardized scale due to such
items as the size of the counterparty and the existence, or not, of
collateral on the loan/bond. For example, small and medium-sized
enterprises (SME's) are given lower capital requirements for comparable
risk levels of as much as 25-50 percent less capital. The argument that
the correlation of default rates amongst these small counterparties is
lower than for larger corporations may be valid, but I have seen little
evidence that the ``haircut'' for these loans should be as much as 50
percent. In my opinion, this was a concession to those national banking
systems of the world whereby SME's are the vast majority of borrowers--
hence lower capital requirements for banks in those countries. It is
also true that SME's make up the vast majority of loan assets of the
smaller banks in the United States and the same lower capital
requirements would hold for U.S. SME's and the banks that make these
loans--close to all but 100 of our Nation's 8,000 banks. But, as I will
now discuss, almost all of U.S. banks will not be required to follow
the recommendations of Basel II, so the reduced capital requirements on
SME's will not be relevant and the old Basel I's 8 percent rate will
probably still be in effect for all except the very largest U.S. banks.
United States vs. Rest of the World and Basel II
As I indicated above, and as you are probably all aware of, Basel
II's recommendations on credit risk and operating risk dimensions of
bank activity are just that--recommendations. The Central Banks of the
world, and other bank regulatory bodies, who set national bank
regulatory policy, may or may not choose to conform to all or any parts
of Basel II's recommendations. Indeed, it came as an enormous surprise
to some observers, including this writer, that only the largest 10 U.S.
banks, and perhaps the next 10-20 banks in terms of asset size, would
be required (top 10) or will have the option (next 10-20) to follow the
advanced Internal Rate-Based (IRB) version of Basel II's Accord with
respect to specifying the LGD dimensions of their portfolios and hence,
set capital requirements based on portfolios risk characteristics.
While it is true that as much as two-thirds of all bank assets are held
by the top 30 U.S. banks and more than 95 percent of foreign bank
assets operating in the United States will be covered by Basel II's
most sophisticated guidelines, it is likely that all the rest of our
banks (almost 8,000 smaller banks) will not be asked to conform and
will probably not do so for the following reasons:
(1) Basel II is too complex and costly to introduce and conform
with.
(2) The U.S. banking system is presently more than adequately
capitalized and the recent decade's experience of very low numbers of
bank failures makes change unnecessary.
(3) The added Basel II capital required for operating risk is based
on highly arbitrary and extremely difficult to measure variables.
(4) The Federal Reserve System's, and other bank regulatory
agencies, policy of ``prompt corrective action,'' and maximum leveraged
ratios, when bank capital falls below a certain specified level has
worked very well in the United States and is not specified as part of
Basel II--even in Pillar 2's regulatory oversight.
In other words, ``if it ain't broke, don't fix it!''
I believe that the choice of only the largest 10 commercial banks
to conform to Basel II, and the IRB approaches, is unfortunate and
should be reconsidered. Not withstanding the recent consolidation
movement of many of our largest and most sophisticated banks, the
possible exemption of #11 to #30 (including HSBC Bank USA, Citibank
[West], Bank of New York, Key Bank, and State Street, #11-15) and the
very likely exemption of #31 to #50 up to 8,000 (including such
seemingly large banks as Charter One, Am South, Union Bank of
California, Mellon Bank, and Northern Trust, to name just a few, seems
arbitrary and belittles the possible sophistication and motivation of
these banks which would be substantial institutions in most other
countries of the world. For example, the 50th largest bank in the U.S.
in terms of assets (Compass Bank with $24.3 billion) or in terms of
deposits (Mellon Bank with $15.2 billion) would be huge institutions in
most countries.
The choice of a round number, like ten, would seem to be
insensitive to world opinion as well as to risk management motivation.
Speaking from an economic standpoint, rather than a political one, I
would prefer to see either no banks be required to conform or some
exemption level whereby the costs/benefits to our bank-
ing system would be more rationally presented and defended. Certainly,
a number like the top 50-100 banks would be much more in line with the
number of banks conforming in other countries. This would help ensure a
``level playing field'' amongst banks.
Our largest banks are probably relatively happy to conform to Basel
II even with its complexity and added costs to develop information and
credit scoring systems to conform to the requirements of the advanced
Internal Rating-Based (IRB) systems mandated under Basel II. The reason
is that they expect that the total capital required for credit assets
will be less than what is required under the current regime (which will
continue until 2007). So, we may have a new regulatory regime where
everyone--large and small banks, as well as out bank regulators--are
relatively pleased with the changes recommended under Basel II.
A Related Disappointing Result of U.S. Policy
I have always felt that despite its problems with: (1) complexity,
(2) too low capital requirements on the more risky counterparty assets
and (3) the difficulty of managing against operations risks, Basel II
had one extremely important by-product--the motivation for banks to
develop or improve upon their existing credit scoring models and
systems to reduce total losses from nonperforming and eventually
charged-off loans. These systems can be used to rate and set capital
for all bank customers rather than setting a ``one-size-fits-all'' (8
percent) requirement on them. I have observed the enormous strides
achieved by banks throughout the world, including ones of all size and
location, as to developing risk management systems and training of
personnel to prepare for Basel II. Indeed, from what I can surmise,
banks in most countries, especially in the European Union, will all
adhere to Basel II's Standardized, Foundation or Advance IRB
approaches. Granted that regulators in these countries will need to
sanction far fewer banks than U.S. regulators would have to do if all
banks are mandated to conform, it must have come as a surprise, perhaps
even a resentful shock, that the vast majority of U.S. banks will not
adhere to Basel II. This is especially true since the United States and
its representatives to the B.I.S. were early champions of the need to
change the way banks allocate capital for credit risk of their clients.
What is disappointing to me is that the Fed's decision to exempt
all but the largest banks from building and implementing an IRB
approach of some level of increased sophistication will reduce the
motivation for most banks to move to a more risk sensitive lending
policy. I recommend that our Federal regulators require some level of
added due diligence on the part of banks with respect to economic
capital decisionmaking and the use of credit scoring or rating systems
even if they are not absolutely required under the old (and continuing)
Basel Accords. One way to accomplish this on a cost effective basis is
for smaller banks to combine resources (data and money) to accomplish
these goals. Our decision to exempt smaller banks from Basel II may
backfire if many of the world's smaller, or even larger, banks from
developing and developed countries, also opt out of the process,
leading to greater instability in these banking systems and perhaps to
ours through contagion.
A Note on Procyclicality
One of the likely by-products of Basel II and its reliance on
systems that require a careful assessment of credit ratings and loss
given default reserves is the possible procyclical impact. That is, any
system which requires more capital when defaults increase and banks'
portfolios become more risky, such as what is likely to occur during
periods of economic stress, will motivate banks to provide even less
credit, for example, ration credit or a credit-crunch, thereby
exacerbating economic downturns. The opposite will likely occur during
periods of above average economic growth, thereby causing too much easy
credit and subsequent higher levels of defaults and charge offs than
would have been the case under the old system. Now I am aware that this
problem called procyclicality already exists due to banks and other
lending and capital providers having ``short memories'' of the last
period of economic stress. Indeed, the procyclical problem resulting
from the benign credit cycle of the mid-1990's (1993-1998) helped to
cause the enormous level of defaults in 2000-2002. And, our research
shows that when bond and loan defaults increase, we can expect a
coincident reduction in recovery rates. Hence, the LGD result will be
even greater due to the negative correlation between probabilities of
default and recoveries given default. I would be surprised if bank
regulators, and the banks themselves, have considered this double
negative effect in times of economic stress. Fortunately, our banking
system was very well capitalized prior to these problems and seems to
have weathered the avalanche of large firm bankruptcies (77 in 2001/
2002 with liabilities greater than $1 billion) without too much stress.
Despite out seeming capital adequate condition and the fact that a
great deal of procyclical behavior (for example, herding, over
compensation for short-term loan losses) can be expected from current
bank regulatory guidelines, I suggest that the Fed consider a more
smoothed capital allocation system to even out the normal fluctuations
in bank reserves, capital allocations, and lending behavior. This would
require more capital set aside in good times and less during periods of
stress.
Conclusion
Thank you for inviting me to attend today's hearings and express my
views. On balance, Basel II has many positive recommendations but still
may prove to be inadequate to overcome strong systemic problems that
normally could be mitigated by a well-capitalized and prudent
regulatory oversight policy. I look forward to observing the results of
Basel II on a worldwide level as well as special concern for the U.S.
banking system. For your information, I have provided my bio-sketch as
an attachment to this document.
----------
PREPARED STATEMENT OF KAREN SHAW PETROU
Managing Partner
Federal Financial Analytics, Inc.
June 18, 2003
It is an honor to appear today before this Committee to discuss the
potential ramifications of the international risk-based capital rules
under consideration in Basel for U.S. financial institutions and--even
more important--for the economy that depends upon them. I am the
Managing Partner of Federal Financial Analytics, a consulting firm that
advises on U.S. legislative, regulatory, and policy issues affecting
strategic planning. In this capacity, we advise a variety of companies
on the implications of specific sections of the Basel proposal. We also
advise the Financial Guardian Group, which represents those U.S. banks
most concerned with the proposed operational risk-based capital charge.
In my testimony I will focus on the most recent version of the
Basel rules--the third consultative paper or CP-3, as well as on the
advance notice of proposed rulemaking (ANPR) on which the U.S.
regulators are now working. Although the
effective date for the new version of the international capital rules--
Basel II--is December 31, 2006, its actual impact will be felt far more
quickly. Indeed some markets have already begun to change in
anticipation of the Basel standards. As the final shape of the rules
becomes more clear, financial markets--and the larger economy--will
change more noticeably. Congress' review of the rules thus comes in a
timely fashion that ensures any policy concerns posed by the rules can
be addressed well in advance.
Much in CP-3 is very worthwhile. Overall, Basel II is a worthy and
overdue effort to fix the problems in Basel I that have created all too
many opportunities for banks to ``arbitrage'' the capital rules. When
regulatory capital diverges from the ``economic'' capital dictated by
the markets, banks change their portfolio, pricing, and risk decisions.
This has profound impact on overall franchise value and on key lines of
business, as well as affecting the cost and availability of credit to
consumers and companies across the country and around the world.
When regulatory capital is too low, banks can take undue risk--a
problem observed since Basel I went into effect that needs a quick
remedy. When regulatory capital is too high, banks cannot compete
against nonbanks, and assets flee the banking system with possible
adverse consequences for overall market stability.
However, Basel II now has gone far from its initial clear goal of
ending regulatory arbitrage. In fact, the most recent statement of the
purposes of Basel II--posed 2 weeks ago by U.S. regulators--no longer
even mentions this. Now, the goals of Basel II are said to be:
Improvements to internal risk management and capital allocation,
enhanced market discipline and--of all things--a new capital charge for
operational risk. I shall have more to say about the operational risk
capital charge later, but suffice it to say that this proposal worsens
the relationship between regulatory capital and risk--absolutely the
reverse of where Basel II initially intended to go. If Basel II cannot
be brought back to its initial and important purpose, then the U.S.
capital rules alone should change to do so.
Based on our review of the third consultative paper and recent
statements from U.S. regulators about our own implementing rules:
<bullet> There is a ``first-things-first'' solution that fixes the
Basel II's complexity problem. Much in the proposal can be quickly
implemented at reasonable cost for all banks and savings
associations here and abroad. U.S. regulators should act now on
those sections of Basel II on which they can agree unanimously, and
defer other sections until they can do so.
<bullet> A ``one-size-fits-all'' approach won't work in the United
States Capital should go up or down with risk, not be squeezed into
the current requirements that were originally set with scant regard
for actual credit losses. Unique factors in the U.S. market make it
especially important that bank capital appropriately reflect risk.
<bullet> The operational risk-based capital section of Basel II remains
deeply flawed and should be dropped. Regulatory capital for
operational risk will increase risk, not reduce it, and strong
supervision with enforced standards is the right way to address
operational risk.
<bullet> Simple capital rules are essential for effective supervision.
Agencies here and elsewhere cannot administer over-sophisticated
rules. Further, laboring to do so will divert resources from
emerging risks that often prove the undoing of individual
institutions or serious risks to the financial system as a whole.
Capital is not the only driver of safety and soundness. Banks have
collapsed in the past and will fail in the future even as they hold
more than the minimum amount of regulatory capital.
Economists and financial analysts have spent literally thousands of
hours working to revise the risk-based capital standards that govern
interationally active banks around the world and all insured
depositories in the United States. This effort is a very important one,
as problems in Basel I have led to undue risk-taking and other concerns
that warrant immediate attention. However, in the 5 years in which
Basel II has been crafted, more and more attention has been devoted to
the increasingly complex models that attempt to anticipate expected and
unexpected losses in every line of business, under every scenario in
each country for all time. The defense of this effort is that financial
markets are now complex, so capital must be too. However, the universe
is very complex, yet Einstein found a very simple formula that helped
to explain it. Complexity is a weakness, not a strength, and Basel II
should be difficult only when absolutely necessary to capture subtle
risks with potentially severe consequences.
Basel II rightly rests on three pillars: improved regulatory
capital standards, better supervision and more disclosure. If Pillars 2
and 3 work well, then Pillar 1--the capital standards--need not be as
formulaic and far-reaching as currently proposed because supervisors
will have ample tools to tailor regulatory capital to individual
circumstances and markets will know when this is not being done.
One reason regulators rely so much on regulatory capital is the
lack of effective supervision in many major financial markets. Here,
though, supervisors have ample authority to discipline banks for
problems that have nothing to do with capital standards. Companies
must, for example, be ``well managed,'' as well as ``well capitalized''
to be financial holding companies and enjoy the privileges provided in
Gramm-Leach-Bliley. Further, supervisors measure banks on a ``CAMELS''
scale in which capital--the C--is just one of a range of factors--all
weighted equally--on which critical enforcement actions hinge. The
other factors are asset quality, management, earnings, liquidity and
sensitivity (to various risk factors). If non-U.S. regulators adopted a
similarly wide-reaching supervisory regime--and backed it up with
meaningful sanctions such as those deployed here--then much of the
complexity in Basel II could fade away and the rules could focus on
ending major sources of regulatory arbitrage that, on the one hand,
threaten safety and soundness and, on the other, unnecessarily
undermine bank profitability.
First Things First
Despite the intention of having a balanced international regulatory
framework that emphasizes more than just regulatory capital, the vast
majority of staff time has been spent on the regulatory capital
charges. U.S. regulators, I think, could have done much for the global
financial system and avoided many of the pitfalls in Basel II if more
attention had been paid to exporting our strict supervisory standards
and their effective enforcement. Japan, in particular, would benefit
greatly from this--it is a clear case in which nominal adherence to
regulatory capital has done nothing to prevent a grave banking crisis
with serious macroeconomic impact.
As Basel II advanced and the capital models grew ever more complex,
U.S. regulators rightly became increasingly concerned about how this
would work in our unique banking system. In sharp contrast to Japan and
the European Union, we here have thousands of banks and savings
associations; foreign banking systems are far more concentrated into a
few nationwide banks. Regulators also--rightly--became concerned about
several pieces of the simpler sections of Basel II that resulted from
complex multilateral negotiations in which the end goal was often
obscured. This is particularly true with the more simple versions of
the operational risk-based capital standards, which are not only
flawed, but could also actually increase--not reduce--banking risk.
Further, even the relatively simple sections of Basel II grew ever more
complex as negotiators sought to solve each problem and individual
national political objectives as the rules worked their way along over
the years.
Based on these fears--some of them quite right--U.S. regulators
have come up with a solution--mostly wrong. They now plan to impose
only the most complex versions of Basel II and then to do so only for
the Nation's largest banks. This may limit the pain, but it also
undermines the gains close at hand in Basel II. Where Basel II drops
regulatory capital--which it does dramatically in traditional lines of
business like mortgages and small-business lending--banks left out of
Basel II, which will still be required to comply with Basel I, will be
at a serious competitive disadvantage to big ones in it. Where the
models are overly complex or--worse--wrong, the fact that only big
banks must comply with them does nothing to redress the adverse impact
they might have.
Further, writing off the most flawed sections of Basel II in the
United States does nothing to address potential serious consequences in
the global economy. U.S. banks--especially large ones--compete head-on
with non-U.S. banks here and abroad. If differences in the simpler
parts of Basel II--called the standardized approaches--give non-U.S.
banks an excuse to rely on overlax rules and inadequate enforcement,
then the major strength U.S. banks now have in the international
financial services market will be undermined. Worse still, major
financial services firms could operate under capital rules that do not
actually address real risk.
For all its flaws, much in the standardized proposal for credit
risk reflects broad agreement on improvements to Basel I. U.S.
regulators should turn this into clear language and propose it for
smaller banks, while refining the advanced models and offering them to
large ones. Where no agreement is in sight--on asset securitization,
for example--regulators should act now on those areas where broad
consensus exists and defer the others until it emerges or regulators
are sure--absolutely sure--they are right and the industry is wrong.
U.S. regulators now diverge on many key areas of Basel II, and they
should act in unison on issues where they intend to contradict the best
evidence and advocacy the industry can muster.
One Size Won't Fit All
As U.S. regulators have turned their attention from the
international negotiations to implementation of Basel II at home, a
major dispute has arisen over whether to follow the new rules where
they lead. Under Basel II, capital could go down below current levels,
especially for very large banks with major retail or mortgage
operations. It is for that reason that the operational risk-based
capital proposal has been superimposed on the credit risk reforms Basel
II initially sought. It is also the
reason why some U.S. regulators are now reasserting the importance of
the most primitive of all capital charges--a simple leverage one--on
banks and their parent holding companies. ``Topping off'' the right
amount of credit risk capital with the operational charge and a
surcharge for ``leverage'' will so undermine Basel II--especially in
light of its high implementation cost--as to raise serious questions
about whether the entire exercise is worthwhile.
I shall have more to say about the operational risk charge below.
With it, Basel II should not be implemented at all. Without it, a sound
regulatory capital scheme is in sight.
The leverage rule is a unique U.S. capital standard, and it is one
that should be dropped as Basel II comes into force. Indeed, it is one
that should have been dropped years ago. The leverage standard is a
simple ratio of capital to on-balance sheet assets calculated without
regard to risk. Under the leverage standard, a bank holds the same
amount of capital if its book of business is solid gold or unsecured
credit card loans to dubious borrowers. It is a capital standard that
could not be more crude, but U.S. regulators clung to it in 1988
because they weren't sure they trusted Basel I. They wanted some form
of insurance because they knew then--as now--that credit risk rules did
not capture interest-rate risk. You will recall that this latter risk
was the predicate cause of the collapse of our savings and loans--which
cost taxpayers more than $250 billion and kept this Committee extremely
busy for over a decade.
So, ironically, Basel II still doesn't address interest-rate risk
(IRR). Although the regulators think they know enough about operational
risk to put it in the Pillar 1 mandated capital standards, they have
decided to leave IRR in Pillar 2. In 1988, regulators were right about
the problems measuring IRR; now, they are not. Markets price trillions
of dollars of IRR each year in a fashion that Fed Chairman Greenspan
has rightly praised.
Why then keep the leverage rule? U.S. agencies appear to be
clinging to it because they are afraid to follow Basel II's models
where they lead. In some cases, the advanced models propose massive
drops in regulatory capital. This is particularly true in mortgages and
small-business loans--key lines of business for smaller banks that will
face major competitive problems if big banks get to drop regulatory
capital under Basel II while they are kept in the cold of Basel I. Of
course, in other cases, Basel II will dramatically raise capital--for
high-risk loans and certain equity holdings, for example. To adopt
Basel II when it goes up and block it when it goes down is to create a
regulatory capital regime that leaves arbitrage largely in place--again
profoundly undermining why all this started in the first place.
The best way to protect the deposit insurance funds from risk and
small banks from competitive harm is to introduce Basel II's most
advanced model-driven sections in an incremental way that--
essentially--hedges the model-builders' bets. How to do this? Despite
the complexity of the advanced internal ratings-based approach to
credit risk, it can be introduced in a remarkably easy way. Upon
conclusion of Basel II's comment period and a review of all the
analyses of the sophisticated models, regulators should make up their
minds about the ``right'' amount of credit risk-based capital for
specific assets. Where they cannot agree, as noted, they should defer
action. Where they can, they should implement Basel II--but only in a
phased-in fashion. If, for example, the ``right'' amount of capital is
a dramatic drop, then set a schedule in which capital slides down year
after year across the board for all banks that qualify to use the
advanced models. Where it goes up a lot, capital should similarly be
phased in.
This incremental approach has two advantages. First, as noted, it
hedges the regulators' bet on the skills of their model builders and
the ability of supervisors to handle the complex new rules (on which
more below). Second, it addresses concern that Basel II will exacerbate
booms and worsen busts--``procyclicality'' in Basel speak. To be sure,
phasing in Basel over time doesn't eliminate procyclicality, but it
ensures that regulators are certain of their capital models when these
come into full force, while giving them time also to assess the value
of stress testing and other measures now under consideration.
Eliminate the Pillar 1 Operational Risk Capital Charge
Basel II's pending proposal and, we are told, the draft U.S.
implementing rules will include a new regulatory capital charge for
operational risk. Operational risk is that resulting from human or
systems failures, natural disasters, and even terrorist attack. There
is, though, no accepted definition of operational risk for supervisory
purposes--for example, does it include reputation risk? Basel II says
no--for now--but this risk has frequently proven the most serious of
all in a business fundamentally founded on investor and depositor
confidence. What about events like September 11--catastrophic
operational risk? Basel II now has them in--although they were out at
the end of last year--but who knows how to measure the likelihood of
another attack and then to decide just how much capital is enough and
whether capital the right antidote?
It is particularly hard to understand why Basel II has a specific
capital charge for this risk when one notes that many of its own
documents agree that it cannot be well defined. The Basel Risk
Management Group, for example, said its own data need to be used with
``caution'' and that a specific capital charge cannot now be based on
them. A major Basel Committee on global financial safety also concluded
earlier this year that there is now no way to determine a quantitative
regulatory capital charge.
Another major unanswered question: Is any amount of capital enough
against catastrophic risk? I do not think so, and indeed imposing an
operational risk-based capital (ORBC) requirement will create a serious
and perverse incentive for banks to skimp on the forms of operational
risk management and mitigation that proved their worth in the most
recent and terrible case of catastrophic operational risk, the attack
on the World Trade Center. What worked after the terrorist attack--
apart from undaunted heroism--were the backup systems and contingency
plans that well-prepared financial firms had put in place. What worked
in the terrible days thereafter--other than sheer courage and
determination--was insurance. In Basel-speak, these are operational
risk management and mitigation. Both are costly--indeed, the back up
systems, which U.S. regulators have mandated since September 11, are
very much so. Imposing a simple, arbitrary charge against operational
risk will lead many banks to rely on this, not proven ways to protect
themselves, their customers and the financial system more generally.
Basel II now includes three variations on a regulatory ORBC
requirement. Two of these--the ``basic indicator'' and ``standardized''
ones--rely on a simple percentage of gross income to calculate ORBC.
This is, quite simply, nonsensical. There is no correlation between
income and risk. In fact, operational risk also runs counter to gross
income because banks that spend more on risk management and mitigation
have less profits. Banks that generally have trouble making money also
tend to be riskier--again, an inverse correlation between gross income
and operational risk, not the positive, linear one on which Basel II
relies.
U.S. regulators have, apparently, realized that the two simple
approaches to operational risk in Basel II do not work. As a result,
they are planning only to impose the ``advanced measurement approach''
(AMA) here. This will, though, leave the other two methods in place in
the EU and Japan, creating a perverse incentive for big banks there to
run undue amounts of operational risk. We cannot wall ourselves off
from the problems this will create, and U.S. regulators should thus
push hard for meaningful supervisory standards for operational risk
that bind all financial services firms, not compromise on a deeply
flawed regulatory capital model.
Further, the AMA does not solve the fundamental problems with an
ORBC charge in Pillar 1. Some of these are unique to U.S. banks--which
must compete with major nonbanks in lines of business like asset
management and payments processing. Basel II in the United States will
not cover nonbanks. Specialized banks will thus face major competitive
pressures that may force them to review whether continuing to remain a
bank is worthwhile. ORBC not only creates incentives for increased
operational risk, but it also may create one for nonbank charters. This
would drive assets outside our sound, proven system of bank
supervision.
The pending ORBC charge will also put U.S. banks at a competitive
disadvantage against EU and Japanese ones because ``legal risk''
results in a regulatory capital charge. Our legal system is unique--no
other nation has our plaintiffs' bar or our extensive array of laws
designed to protect consumers, prevent discrimination, and promote
workplace safety. There is no evidence that any of this legal risk has
ever caused any U.S. bank to fail, and current law already requires
reserves for material legal risk (and these must also be disclosed).
The AMA epitomizes the problems in Basel II where reliance is
placed on unproven models over which U.S. regulators rightly do not
agree. Acceptance of these models now puts banks at undue and
unnecessary risk--risk far better addressed through effective
supervision with meaningful enforcement.
Can Supervisors Supervise Under Basel II?
Finally, I would like to turn to the question of whether the
complexities in Basel II's advanced models are so daunting that
supervisors at home and abroad will not be able to ensure that banks
actually comply with the new capital rules. This is a major concern,
and one the regulators are already trying to address through a major
Basel Committee focused on supervisory implementation. In the United
States, the agencies now think the best way to handle the complexity
problem is to make Basel II apply only to the biggest banks, whose
examiners tend to be those most familiar with complex financial
arrangements. However, as noted, applying Basel II here only to the
biggest banks will create a range of competitive and safety problems,
while leaving the supervisory capability question largely unresolved.
A recent survey of the cost of Basel implementation for the larger
banks expected to use the advanced models indicates that it will reach
$200 million per bank. One has to ask how it can cost so much for banks
and not pose a comparable burden on supervisors who must assess these
elaborate models. In point of fact, the rules must be as costly for the
supervisors as for the supervised or undue reliance will be placed on
untested models. If supervisors instead rely on ``benchmarks'' they
will in effect superimpose standardized credit and operational
standards that obviate the flexibility hoped for from the advanced
approaches.
These problems are not addressed by the proposed qualifying
conditions for use of the advanced models--more board and senior
management involvement, for example--because none of the proposed
standards addresses the fundamental problem posed by complexity, let
alone how top management can divert resources from their many other
pressing investor protection and safety-and-soundness responsibilities.
The right solution to the supervisory resource problem is the same
as the right solution to the other challenges posed by Basel II: Impose
a uniform system of improved rules across the board and then change
them gradually over time as the rules are tested and we all learn how
to work under them. Back up the more sophisticated models with
meaningful supervision that binds banks in the EU and Japan, not just
United States banks and give investors simple, clear disclosures to
help them understand just how much capital banks have and whether the
supervisors are concerned about it.
----------
PREPARED STATEMENT OF D. WILSON ERVIN
Managing Director
Credit Suisse First Boston
on Behalf of the Financial Services Roundtable
June 18, 2003
Introduction
Good morning Mr. Chairman. I want to thank you for holding these
hearings today and inviting me to appear before the Committee. My name
is Wilson Ervin and I am a Managing Director of Credit Suisse First
Boston (CSFB).\1\ I head our Strategic Risk Management (or SRM)
department and also chair its risk committee. I am presenting testimony
today on behalf of CSFB and on behalf of our trade group, the Financial
Services Roundtable.\2\ CSFB employs approximately 20,000 people,
primarily in the United States, and is a major participant in the
capital markets. It ranks among the top firms in raising money for
companies around the world and is a leading underwriter of mortgage and
credit card financing. The firm is also among the largest managers of
funds invested in private companies.
---------------------------------------------------------------------------
\1\ Credit Suisse First Boston (CSFB) is a U.S. financial holding
company and leading global investment bank serving institutional,
corporate, government, and high net worth clients. CSFB's businesses
include securities underwriting, sales and trading, investment banking,
private equity, financial advisory services, investment research,
venture capital, and asset management. CSFB operates in more than 89
locations across more than 37 countries on six continents. The Firm is
a business unit of Zurich-based Credit Suisse Group, a leading global
financial services company.
\2\ The Financial Services Roundtable is a national association
representing 100 of the largest integrated financial services companies
in the United States providing banking, insurance, securities, and
investment products and services to American consumers.
---------------------------------------------------------------------------
My department is responsible for assessing the risk profile of CSFB
on a global basis and for recommending corrective action where
appropriate to protect our capital. This objective is very similar to
many of the goals of bank supervisors, including the drafters of the
proposed Basel Accord--to deter very large losses and protect bank
solvency.
The Basel II Capital proposals have been the topic of intense
discussion and debate in the financial and regulatory community for the
past several years. The
industry supports the objectives of the Basel process: To better align
regulatory capital to underlying economic risks, promote better risk
management, and foster international consistency in regulatory
standards. The proposed Accord is not a minor refinement to the bank
regulatory process, but is, instead, a wholesale reform of bank
regulation--a regime that covers roughly $2 trillion of capital and is
a key economic engine for most developed markets. The impacts of these
seemingly technical discussions will affect banks, the markets, and the
economy in a deep way, and we would be wise to consider the effects
carefully before implementation.
Before I start, I would like to note that I have personally
developed tremendous respect for the diligence and stamina of the
regulators who have worked on Basel II. They have had to address a
great many complex and challenging issues, and have been tenacious in
trying to develop a ``best practice'' solution for each. Balancing all
of this and applying it to very different financial markets around the
world--with political sensitivities in each--does not make this an easy
job. I wish to express appreciation for the efforts of Federal Reserve
Board Vice Chairman Roger Ferguson, who has met with CSFB and
Roundtable member companies several times in the past few weeks to
listen to our concerns on the proposed Accord. Comptroller Hawke and
FDIC Chairman Powell have also had open doors for discussion throughout
the long process of developing the new Accord. We look forward to
continuing this dialog as Basel II moves closer toward formal adoption
and throughout the implementation period.
CSFB and the Roundtable have worked hard to be constructive
commentators on the new rules, particularly in respect to practical
implementation issues. The recent revision of the proposals--called CP-
3--included significant improvements, and demonstrated a willingness by
regulators to address specific issues raised by industry and academic
critics. Just last week, the Federal Reserve announced that, in
implementing Basel II in the United States, the regulators propose to
reduce the capital charges on many types of commercial real estate
loans, in response to comments and new data from the banking industry.
We support the direction in which the Accord has been moving recently,
and appreciate the regulators' willingness to reexamine earlier
conclusions and consider further changes.
However, in spite of the hard work of the Basel Committee and
industry, we believe substantial areas for improvement still remain.
Basel II has considerable momentum, and most people in the industry
believe it will likely be implemented in the relatively near future. On
balance, we believe that the advantages of the reform now outweigh the
drawbacks, although that balance remains close, and in several areas,
open issues remain. This is a frustrating outcome for an initiative
with so much potential. We hope these hearings will help illuminate
some of the important remaining issues that need to be addressed, so
that the Basel II reforms can live up to their original, very worthy
goals.
Today, without getting too involved in the technical details of the
Accord, I would like to highlight four ``macro'' issues which we
believe are particularly important:
1. The current Basel proposal is unnecessarily complex and costly,
and suffers from an excessive reliance on detailed, prescriptive rules.
Under the rubric of comparability, these international rules could
bring a more formulaic, inflexible style of regulation to the United
States, which currently enjoys a much better balance between black-
letter rules and supervisory consultations.
2. The new Accord and its sensitivity to credit ratings could
reduce liquidity in the credit markets during economic downturns,
potentially extending or deepening economic recessions
(procyclicality).
3. The operational risk capital charge proposed by the Basel
Committee remains highly controversial. Some Roundtable members support
the proposed Pillar 1 operational risk charge; others believe
operational risk should be addressed through Pillar 2 supervisory
reviews instead.
4. The disclosures required under Pillar 3 of the new Accord are
likely to add perhaps 20 pages of highly technical data to bank
reporting requirements, raising costs and adding little information of
value to the reader. While we appreciate that the Pillar 3 disclosure
requirements have been reduced, they continue to be burdensome and
potentially confusing.
Prescriptiveness, Cost, and Adaptability
The first topic I would like to address is the overall cost and
prescriptive tone of the new capital rules, and the effect this will
have on whether the rules remain relevant over time. The new rules
shift the regulatory regime toward a highly complex, formula-based
system, and will diminish the important role that is currently played
by human judgment. Implementation of these rules will be high cost, but
not highly cost effective. Moreover, we believe the very complexity of
the new rules and the delicate political balance represented in them
will make it challenging to update the rules over time.
Most of this prescriptiveness is to be found in Pillar 1, which
describes the ``recipe'' for calculating capital requirements. The most
recent draft of the Pillar 1 calculations ran to nearly 200 pages,
roughly 5 times the length of the original Basel Accord (not including
technical papers and additional guidance that is expected to be
issued). This is a common result from this kind of process. Once you
start developing a system that attempts to capture the complexity of
the real world in a series of mathematical rules, it is very hard to
stop halfway. One issue or another will always be of major concern for
some institution or country. Many of the Pillar 1 rules reflect a
political compromise as much as the results of a scientific approach to
risk management. The result is a very elaborate system that tries to
address all circumstances by being ever more complex, and currently
staggers under its own weight. The Basel Committee has done a
commendable job in streamlining the earlier drafts in CP-3--the earlier
drafts of Pillar 1 rules were even longer--but this remains a
fundamental issue.
Perhaps the underlying issue in this respect is the prescriptive
nature of the new Accord. Conceptually, the Committee has attempted to
capture current industry best practices and boil them down into fixed
formulae, adding burdensome qualification, testing, and reporting
requirements.\3\ These new regulatory requirements, while well-
intentioned, will be unduly burdensome and inconsistent with changing
market reality and evolving best practice.\4\ It is our recommendation
that the Committee establish some basic requirements largely around the
key input parameters and exposure calculations and publish best
practices that provide guidance to banks and
supervisors rather than a rigid rulebook.
---------------------------------------------------------------------------
\3\ One editorial recently described this approach as ``prescribing
and proscribing in equal measure . . . a monster that cannot clear the
first hurdle: flexibility.'' Risk Magazine, editorial page, June 2003
edition.
\4\ For example, the eligibility requirements for institutions to
qualify to use the Accord's advanced IRB methods for credit risk
capital charges are too detailed and burdensome. In general, we believe
that these eligibility requirements should be scaled back and replaced
with more general guidance.
A specific example is the testing requirements for credit exposure
in repurchase agreements, an area with historically very low losses. To
its credit, the Basel Committee permits the use of internal market risk
models to estimate potential collateral shortfalls under stress, which
is in line with modern practice. However, the Committee requires
substantial additional testing to use this technique, even though this
calculation is based on the same model that governs overall market
risk, which is a much bigger risk and already subject to comprehensive
regulatory oversight.
---------------------------------------------------------------------------
CSFB and other Roundtable members are also concerned about the
cumulative
effect of numerous conservative choices and assumptions that are built
into this complex fabric. Each of these can be debated separately, and
many are extremely technical. But the combined effect of each of these
individual items adds up to regulatory capital requirements that can
depart significantly from the true economic capital needs that Basel II
was aiming to emulate.\5\
---------------------------------------------------------------------------
\5\ To mention a few examples:
(i) The Accord significantly overstates the credit risk capital
charges for exposures hedged by guarantees and credit derivatives, by
failing to recognize the much lower risk of joint defaults by debtors
and guarantors and by applying overly conservative rules on maturity
mismatches.
(ii) The proposed Accord requires capital against Expected Losses,
even though these losses are already covered by loan loss reserves, and
Future Margin Income is generally recognized only for credit card
exposures.
---------------------------------------------------------------------------
Home/Host Country Issues
The complexity of the new rules poses particular challenges for an
international bank that is regulated by supervisors in multiple
countries. CSFB, for example, will be required to implement Basel II as
both a Swiss bank and a U.S. financial holding company. Our
implementation will be governed primarily by the Swiss Federal Banking
Commission, in conjunction with the Federal Reserve in the United
States and the Financial Services Authority in the United Kingdom, and
also by other regulators around the world.
Most international banks face a similar set of interlocking
regulation in which both home and host countries interpret and enforce
rules. This can give rise to conflicts, even under an international
standard like the Basel Accord. At times, we have been given
conflicting requirements by home and host regulators under Basel I,
making compliance an impossible ``Catch-22''. While we have been able
to resolve these issues to date, the potential tension between ``home
and host'' regulators will become a bigger issue given the much wider
and more detailed Basel II regime. If each country decides to require
its own local rules and local data for each of the many calculations
required under Basel II, the compliance burden will go from bad to
worse. The Basel Committee has formed an Accord Implementation Group to
deal with cross-border implementation issues, but experience shows that
some differences between multiple supervisors are inevitable.
We are pleased to note that, in a speech last week, Vice Chairman
Ferguson indicated that the U.S. banking regulators expect to accept
the Basel II approaches and calculations followed by a bank's home
country supervisors, when evaluating an international bank with U.S.
branches and for purposes of eligibility of Gramm-Leach-Bliley Act
financial holding company status. This is reassuring to hear. We hope
that other host countries adopt similar policies that defer to home
country regulators, and that similar issues related to subsidiary banks
also are addressed. We believe that stronger proposals should be
developed to resolve home/host country conflicts in a timely and more
predictable manner.
Securitization
A germane example of Basel II's complexity and excess
prescriptiveness is its proposal for asset securitization. Asset
securitizations are a cornerstone of how the U.S. markets finance
residential mortgages, consumer credit card balances, automobile loans,
and other receivables. The draft rules here are daunting, potentially
quite burdensome, and often difficult to interpret. The result is that
only a few experts in each area are likely to understand this and other
specialized rules of the Accord. Yet, the interpretation of these
experts on some technical points can have enormous impact on the
capital calculation.
These rules are written to deter possible arbitrages in the new
rules, but risk throwing the ``baby out with the bath water.'' The
industry and regulatory communities generally agree on the objective
that capital should be similar before and after securitization, since
the total economic risk is unchanged. However, apportioning the risks
properly among the different securities poses a difficult challenge for
any set of static rules. The Basel Committee's current proposal under
CP-3 takes a conservative approach to this problem, focusing on
avoiding improper capital arbitrage by building a technically complex
system with a ``belts-and-suspenders'' philosophy. Unfortunately, this
approach can also interfere with legitimate transactions and could
undermine a widely accepted risk management tool used by many United
States institutions.
Several problems remain that should be reviewed by the regulators.
First, the mere act of securitization and distribution will tend to
increase the capital charge assigned to the same pool of assets.\6\
This increased capital is an important issue for the U.S. markets in
particular, as foreign markets are much less reliant on securitization
technology. This could raise costs for funding U.S. consumer loans and
other asset classes where securitization techniques are important.
Foreign regulators have much less at stake in their local markets.
---------------------------------------------------------------------------
\6\ For example, the originating bank is charged the full risk of
the pool if it retains a sufficiently large position in the junior
securities. A second bank that purchases the senior securities also
will be charged significant capital, meaning that the capital required
of the banking system will be higher than if the assets had simply been
held on an institution's balance sheet directly.
---------------------------------------------------------------------------
Second, the calculations are subject to difficult interpretations,
which can give rise to ``cliff edge'' uncertainties, where capital
charges can change by a factor of 10 or more depending on whether a
particular instrument can be fit into a specific regulatory box. For
example, a credit line provided to support a credit card or receivables
facility might attract a risk weighting of 100 percent if the bank can
satisfy a number of technical tests about the structure of the credit
facility.\7\ However, this charge can skyrocket to 1250 percent (that
is, an outright deduction from capital) if a bank cannot meet one of
these compliance requirements. This is a conservative approach \8\ that
will certainly help deter arbitrage, but it may also deter good
finance. It also will tend to restrict the evolution of new markets and
new securities, since these future instruments might not fit easily
into today's compartments. As with other areas of the Accord, we
believe that moving to a more principles-based system that leaves more
discretion to banks--subject to thorough supervisory oversight--will
provide a more durable and flexible solution for the long term. It will
be important to incorporate these changes into both the final text and
in the practical implementation of the rules.\9\
---------------------------------------------------------------------------
\7\ In particular, questions remain regarding the proposed
treatment of liquidity facilities for asset-backed commercial paper
programs, which would face capital charges that seem disproportionately
high relative to the level of risk.
\8\ Some also have argued that the risk weights on such securitized
assets are too high as a more general matter. Similarly rated corporate
loans often attract a much lower capital charge.
\9\ For example, banks that qualify for the Advanced IRB approach
should be allowed to use internal ratings to determine risk weights,
which is not allowed for securitizations under CP-3. Ratings based on
rating agency methodologies or reasonably equivalent approaches, for
example, should provide supervisors sufficient comfort that a market
test has been met. Liquidity facilities and credit enhancements for
asset-backed commercial paper conduits are prime examples where this
approach could be easily adopted.
---------------------------------------------------------------------------
Cost
The monetary cost of complying with the Basel II rules will be
significant. For Credit Suisse Group, our holding company, we estimate
that our initial costs will be $70mm to $100mm just to implement the
system, plus substantial ongoing costs. Multiply that by thousands of
banks globally and this will amount to many billions of dollars of
additional costs. Some of these costs will be passed on to consumers
and corporations, and some of these costs may force banks to exit
certain activities leaving these markets to unregulated entities.
A major driver of the cost / benefit ratio of the new rules will
depend on how they are applied. For example, there are more than 50
specific requirements that must each be met to use the so-called IRB
advanced credit system. If each of them is interpreted and tested to
rigorous audit standards, there will be enormous costs in compliance
though the relevance to better risk management will be small. I would
note that implementation costs will be substantial for regulators as
well as for the banking community.
Even more important, perhaps, than the direct monetary costs, are
the indirect costs. These will depend on whether the new rules support
the real risk management needs of the business, or whether they become
an extra bureaucratic burden or even a diversion. CSFB's internal
assessment is that most of the additional resources required will not
be in the risk control departments. Instead, most of these new
resources will be needed in the areas of financial reporting and IT
support systems, in order to generate the volume of data and reports
that Basel II requires to a reliable, audit quality standard. While
further systems development provide some important benefits, this
result suggests that the gains in risk management quality from the new
proposal are likely to be relatively modest.
Adaptability
The proposed Basel rules are based on the financial markets as they
work today, but are so complex and heavily negotiated that they will be
difficult to update over time. Indeed, some commentators have suggested
that the Accord will be outdated by the time of implementation.\10\
---------------------------------------------------------------------------
\10\ See Risk Magazine, footnote 3 above.
---------------------------------------------------------------------------
The draft Accord also requires banks to use the Basel II processes
in their internal management in many areas, regardless of whether they
remain relevant for business practices. If bank management is required
to compute and manage by the Basel II rules anyway, further
improvements in internal practice could be seen as both costly and
irrelevant. As a result, the Basel Accord could actually slow the
progress of better private sector risk management techniques.
Proposal
Our suggested response to the problems of prescriptiveness and high
cost is for the Basel Committee to place a much greater emphasis on the
principles-based approach that underlies the ``Pillar 2'' section of
the proposed Accord.\11\ Whereas Pillar 1 sets out regulatory capital
calculations in a detailed, prescriptive way, the approach of Pillar 2
is to force firms to develop their own internal models, based on
evolving best-practice, and then to scrutinize the results through the
examination process and regulatory guidance. This ``principles-based''
approach, subject to some reasonable benchmarks and guidelines to
maintain consistency, has some important natural advantages compared to
the complex ``black-letter'' style rules currently prescribed by
regulators under Pillar 1. Pillar 2 encourages banks and regulators to
work together over time to improve risk management practice, rather
than forcing compliance with a potentially dated rulebook. That
approach permits steady, evolutionary improvement and should therefore
be more durable and relevant than Pillar 1 rules that are designed with
today's markets in mind.
---------------------------------------------------------------------------
\11\ Our Pillar 2 comments here are strictly focused on the credit
risk capital charges. As noted later in this testimony, Roundtable
members have differing views on whether any operational risk charge
should be addressed under Pillar 1 or Pillar 2.
---------------------------------------------------------------------------
Addressing this issue will not be simple in the short time left
before the rules are finalized. If these rules are all applied as black
letter law and interpreted strictly, the new rules will be both costly
and--since the risk management advances that lead in part to Basel II
will not end in 2003--potentially irrelevant to ongoing best practice.
We encourage an approach that emphasizes principles and simplicity as
the rules are finalized, and a less onerous ``trust but verify''
approach to compliance. Specifically, we would support adding
statements to the Accord to emphasize that compliance with the rules
will be based not on ``box checking'' but with the spirit of the rules,
based on economic content.
Impact on Competition
We believe that the cost and complex rules of Pillar 1 will have
significant impacts on competition, and could tilt the current playing
field significantly in various markets. This will be particularly
important in the United States, where nonbank competitors like
investment banks, finance companies, and insurance companies represent
a large part of the financial system. The Basel rules do not apply to
them. If the costs of Basel II are high, banks will earn a lower return
on capital, will grow more slowly and may lose market share. There may
even be some incentives to exit businesses or to de-bank altogether. We
believe that the Basel Committee needs to do significantly more work in
assessing the competitive impact of the rules across the financial
marketplace.
Procyclicality
The new rules will change how banks calculate and manage their
capital and the amount of business they choose to do. If banks all act
in concert--as they will tend to do under a common regulatory regime--
this can significantly increase or decrease liquidity in the credit
markets and ultimately affect the real economy. We have analyzed this
effect over the last 20 years of credit cycles. Our calculations
suggest that the impact on required bank capital will be substantial.
In particular, the New Basel II calculations could require much more
bank capital during economic recessions than the current system. The
process by which these rules could widen economic swings is called
``procyclicality.'' \12\ This is, in effect, an implicit change in
macroeconomic policy and it would be wise to consider that carefully.
---------------------------------------------------------------------------
\12\ CSFB has taken particular interest in this issue among
Roundtable members. See American Banker, ``Basel Capital Accord Must
Leave Some Room for Human Judgment'' by Wilson Ervin and Joseph Seidel,
August 30, 2002, and Risk Magazine, ``Procyclicality in the New Basel
Accord'' by Wilson Ervin and Tom Wilde, October 2001.
---------------------------------------------------------------------------
As a practical example, consider the credit environment of the last
2 years. We have seen a huge number of credit rating downgrades, which
have increased the real risk of bank portfolios. The current system is
relatively indifferent to this change in terms of required regulatory
capital, but the proposed system will require significantly more
capital when companies are downgraded. Banks will have to choose
between raising more capital during recessions or reducing the amount
of lending that they do.
Some regulators have suggested that the fear of a capital shortfall
will change banks' risk assessment and lending behavior so that this
issue will disappear. Implicitly, they suggest that bank's risk
assessment will improve so much that mistakes will be a thing of the
past. While it would be wonderful if banks could always foresee the
future, I do not think that is realistic. Bank management already makes
risk assessment a top priority--it is perhaps the core judgment that
determines whether a bank thrives or fails. Unfortunately, economies
are likely to remain cyclical and predictions about the future will
inevitably turn out to include their share of mistakes.
Cutting lending during a downturn is probably smart, if your
perspective is focused solely on bank solvency. However, it raises
significant issues for the wider economy. My personal estimate is that
my bank would have cut back its lending by perhaps an additional 20
percent to 30 percent if the Basel II rules were in place during 2002.
If all banks cut back at the same time, the potential adverse impact on
the real economy could lengthen and deepen the recession. We are
currently working through an economic slowdown; it is difficult to
think that adding pressure on bank capital during this period would be
helpful to economic recovery. In fact, it defeats part of the reason
for regulating banks in the first place--in order to have a stable
supply of capital to support the underlying economy. We need to be
particularly careful here because the new system is imposed across the
whole banking system and everyone will have to operate at the same time
on the same rules. Herd behavior can make smaller problems into bigger
ones.
The regulatory community has acknowledged this as a potentially
serious issue, but we believe that further attention is warranted,
because the consequences of getting this wrong are potentially quite
important to the broader economy. When the first quantitative proposals
in January 2001 revealed a significant potential problem, the
regulators did react with a revised and somewhat ``flatter'' risk-
weight curve.\13\ However, while this reduces the scale of the issue
somewhat, it does not grasp the nettle.
---------------------------------------------------------------------------
\13\ Basel Committee on Banking Supervision, Working Paper
``Potential Modifications to the Committee's Proposals'', Bank for
International Settlements, November 2001.
---------------------------------------------------------------------------
The current Pillar 2 proposals include a credit risk ``stress
test'' which is directly linked to possible additional capital
requirements.\14\ The exact design of this test remains unclear but the
language suggests it amounts to an extra layer of buffer capital so
that banks will not need to dig into their core capital in tough times.
In effect, this is like creating a second fire department, because you
want to always keep the first fire department in reserve. Creating two
fire departments or requiring two pools of capital is unnecessarily
expensive and doesn't seem to address the fundamental issue. That issue
is that a risk sensitive system will inevitably lead to varying capital
requirements through time, and that is a result that will require
explicit management and thoughtful preparation. As with other areas of
the Basel Accord, adding some flexibility to the rules is the simplest
and most practical way of preventing these inevitable stresses from
building up into major crises.
---------------------------------------------------------------------------
\14\ Basel Committee on Banking Supervision, CP-3, April 2003,
Paragraph 397-399 and 724.
---------------------------------------------------------------------------
At a minimum, we suggest that the Basel Committee add to the
proposed Accord an explicit acknowledgment that capital levels may
fluctuate, and that Pillar 2 reviews and stress tests not become one-
way ratchets that only increase regulatory capital requirements. If a
stress test is to work properly, then when tough times arrive, banks
should be permitted to live within their plans, and regulators should
resist the temptation to continue to require the same untouched capital
cushion. Otherwise, Basel II's stress test will not in fact reduce
procyclicality, but will simply amount to an unpublished higher minimum
capital standard.
Operational Risk
In addition to reforming capital charges for credit risk, Basel II
establishes a new capital charge for operational risk--the risk of
breakdowns in systems and people. This is the most controversial
element of the proposed Accord.
Financial Services Roundtable Comment
It is important to distinguish between the concepts of managing
operational risk and imposing a separate, quantitative capital
requirement for it. All of the Roundtable's member companies agree that
evaluating and controlling operational risk is important and should be
required as a supervisory and business matter. Roundtable members do
not agree on whether or how operational risk should be reflected in
regulatory capital calculations. Many companies believe operational
risk can best be addressed through case-by-case supervisory reviews
under Pillar 2; others favor a quantitative and a publicly disclosed
capital charge under Pillar 1.
In several forums, the Roundtable itself has opposed a separate
capital charge for operational risk and has argued for handling the
issue through supervisory reviews under Pillar 2, much as interest rate
and liquidity risk are handled. The Roundtable's senior management has
expressed its concerns directly with Federal Reserve Board Chairman
Greenspan and Vice Chairman Ferguson. Many Roundtable member companies
strongly oppose any Pillar 1 operational risk capital charge. However,
several Roundtable member companies just as firmly support Basel II's
proposed Pillar 1 approach, following the development of the Accord's
``Advanced Measurement Approach'' (AMA), which gives banks flexibility
to use their own internal methods for determining the regulatory
capital needed for operational risk. Institutions that support a Pillar
1 operational risk charge believe it would improve transparency and
comparability and bring regulatory capital requirements into closer
alignment with the ``economic capital'' determinations used in these
banks' internal management decisions. These institutions contend that
any approach other than an explicit Pillar 1 charge for operational
risk would impede progress toward a level playing field, by affecting
the process of calibrating regulatory capital minimums. That is, these
members believe that if an operational risk charge were not included in
Pillar 1, the resulting capital charges on credit risk and market risk
would remain higher to compensate, making it more difficult for
international banks to compete with institutions that are not covered
by the new Accord.
The Roundtable continues to have concerns about the proposed
operational risk capital charge, as well as several technical questions
about its implementation. One problem that all of our members agree
upon is that the proposed Accord fails to give enough recognition to
the benefits of insurance in mitigating operational risk.
CSFB Comment
CSFB is concerned about the attempt to base an operational risk
capital charge on new, unproven models, and believes this approach is
problematic and possibly even counter-productive. We agree that
operational risk is a critical risk to manage, and we set aside
significant capital to cover potential surprises in our internal
capital allocation process. However, we do not believe that operational
risk can be modeled in the quantitative way proposed under the Basel II
rules. Many efforts to measure operational risk have been proposed,
often focusing on limited areas (for example, operations processing
losses) that happen to be susceptible to statistical techniques. But
these methods are not generally relevant to major risks, such as fraud,
a changing legal environment or a major disaster, which are the risks
that require capital. Operational risk capital is primarily to insure
against the risk of being fundamentally surprised by a major event, but
it is difficult to predict and measure what you do not expect.
Basel II and other regulatory initiatives will push banks to devote
significant resources toward operational risk systems and loss
databases, but I personally feel that these resources could be better
utilized elsewhere. Basel II's Advanced Measurement Approach to
operational risk requires banks to attempt to verify their models
statistically. Many are working hard on this, but we have yet to see
any model that has actually been verified in a robust way. In fact, by
emphasizing quantitative numbers for operational risk, we may be
creating a real danger--creating a false sense of security that we have
measured operational risk and hence controlled it. I am a model-
oriented, technical person by training, but I do not want to rely on a
model that is built on speculative assumptions.
It is encouraging that the Basel Committee has sent signals
suggesting an increased degree of flexibility in operational risk
calculations. I am hopeful that this will bear out through the
implementation. But we will still have a long way to go, and I am
concerned that there will be a tendency to revert to prescriptive and
unscientific requirements as regulators develop specific rules for
approving operational risk models.
Pillar 3--Disclosure Rules
One of the strengths of the Basel II proposals is that they look
beyond just calculating and maintaining capital levels. In designing
Basel II, regulators realized that capital requirements--the so-called
``Pillar 1''--could never ensure the safety and soundness of the
banking system alone. They understood that ultimately it is more
important to encourage constructive relationships between financial
institutions, their supervisors and the market to produce good risk
management. This reasoning, which has the strong support of the banking
industry, has lead to the creation of the two qualitative Pillars of
the Basel Accord. Pillar 2 deals with the supervisory review process
and, in particular, regulatory oversight of banks' internal economic
risk assessments. Pillar 3 seeks to enhance market discipline through
increased public disclosure requirements.
The concepts behind the proposed rules for Pillar 2 and 3 are well
accepted by the industry and regulators alike. However, many of the
detailed proposals in the Pillar 3 market disclosures section are cause
for concern in the industry. Unfortunately, the development of Pillar 3
is an area where consultation between the industry and the regulators
came late in the process. Although CP-3 has improved the situation
somewhat, we believe the proposals still are overly prescriptive,
burdensome and subject to misinterpretation. The Pillar 3 requirements
also reflect a somewhat narrow view of risk, focusing exclusively on a
specific regulatory view of risk capital.
We currently publish approximately 20 pages of risk information in
our annual report, and we support transparency and disclosure as very
worthwhile goals. The Pillar 3 proposals would add a large mass of
additional disclosure which is highly technical in nature and which we
believe will be of little benefit to the reader. Indeed, few people are
able to digest all of the information that is already presented on
risks, but now this information could be lost in a deeper, more
technical pile of data. The additional requirements proposed under
Pillar 3 are more likely to confuse than illuminate.
As Chairman Greenspan has recently remarked, transparency is not
the same as disclosure: ``Transparency challenges market participants
not only to provide information, but also to place that information in
a context that makes it meaningful.'' \15\ In this, we believe the
prescriptive, volume oriented focus of Pillar 3 falls short.
---------------------------------------------------------------------------
\15\ Remarks by Federal Reserve Board Chairman Alan Greenspan,
Corporate Governance, at the 2003 Conference on Bank Structure and
Competition, May 8, 2003.
---------------------------------------------------------------------------
Of particular concern are the numerous required disclosures that
relate directly to the capital calculations performed within Pillar 1.
Instead of disclosing measures of risk used in internal risk management
systems, these disclosures mandate an
explicit regulatory capital view of risk. In the most complex areas,
such as asset securitization, these disclosures will surely be
mystifying to all but the most expert audiences.
Moreover, given the likely longevity of the Basel II accord (the
current accord is in its 14th year), there is a need to ensure risk
management practice is able to
mature beyond the concepts now embedded in the Basel II proposals. Just
as the market has moved beyond the current accord, there will
inevitably come a time when some Pillar 1 calculations are no longer
regarded as good measures of risk for all products. In that case, it
must be possible for banks to alter disclosures to represent emerging
best practices. Under Pillar 3 as currently proposed, banks will likely
find themselves constrained to disclosing risks under a system that is
no longer wholly relevant.
In designing the details of Pillar 3, the Basel Committee has
placed too much emphasis on quantity, rather than quality, of disclosure. It is emphasizing consistency by prescription instead of consensus. In
contrast, the demands of the market have produced broadly comparable
and largely voluntary disclosures of market risk by banks. This is an
example of how Pillar 3 should work. It would be more effective if
Pillar 3 established a general set of principles, and then allowed the
discipline of the market to produce continuous improvement in risk
disclosure. This would produce information that the market actually
desires, rather than seeking to impose today's ideas on future market
participants by fiat.
Summary
We are at an important crossroads in the reform effort. A lot of
good hard work on designing the framework and gaining political
consensus has been accomplished. We have a high regard for the efforts
of the Basel Committee and the regulators who have worked so hard to
capture the best current practices in risk assessment. CSFB and the
Roundtable have tried to contribute to the specifics of those
discussions in a constructive manner. We believe that the current
proposal should be streamlined significantly, reducing the level of
prescriptiveness and cost, so that the advantages of this project are
not tarnished by its current shortcomings.
Simplifying the massive weight of detailed rules in Pillar 1 will
require continued discipline in the final round of drafting. It will
also require a new emphasis on the ``spirit'' of the rules, both as the
rules are finalized and when they move to the implementation phase with
national regulators. If, instead, these rules are written and
interpreted as black-letter regulations, set at a highly technical
audit standard, the cost of overall implementation will be high. Such
an approach would mean the calculations could also become increasingly
outdated and less relevant to risk management best practice over time.
We can hope that all national regulators will avoid this pitfall, but
international banks will tend to be driven by the standards set by the
strictest and most literal of their major regulators.
Much hard work has been put into Basel II, but much also remains
ahead. The timetable for implementation is challenging, particularly
since the Accord's require a minimum of 3 years of data for the
advanced calculations--meaning that banks will need to revise systems
to begin collecting the new information by early next year. In the
pressure to finalize and implement the Accord, we hope that enough time
will be provided for everyone--banks and supervisors alike--to digest
and think about the implications of the new regime, and to develop
appropriate transition rules.
As a final comment, I believe that much more can be accomplished by
increasing the emphasis on the concepts of Pillar 2 and Pillar 3, and a
focus on the principles of evolving best practice rather than fixed
formulae. This approach would not only help address ``prescriptiveness,
cost and adaptability'', but could also help address the issues of
operational risk and procyclicality. Pillars 2 and 3 have real people
on the other side--regulators and the market. Human judgment can adapt
to changes and new markets more easily than a rulebook can. This
approach, properly applied, also puts the burden back where it should
be--on the shoulders of bank management to demonstrate to the
regulators and the public that they are doing a good job. That is in
the spirit of the Sarbanes-Oxley reforms, and I think it is a smart,
durable way to improve discipline and maintain best practice standards.
Finally, it should also make the new system more responsive to
change and therefore more relevant over time. Without adjustments to
make Basel II more flexible and to allow it to evolve over time, I am
afraid we might have to start work on a Basel 3 before the ink is dry
on the current effort.
Thank you.
----------
PREPARED STATEMENT OF KEVIN M. BLAKELY
Executive Vice President and Chief Risk Officer, KeyCorp
June 18, 2003
Introduction
Thank you, Mr. Chairman. I am here today on behalf of KeyCorp, the
11th largest banking company in the United States. KeyCorp has total
assets of approximately $85 billion, and spans the northern half of the
United States from Maine to Alaska. While the vast majority of our
business is domestically based, we do have a modest level of
international business activity.
KeyCorp is not one of the institutions included in the definition
of ``top 10 most internationally active institutions.'' Accordingly,
under the present regulatory guidance, we will not be required to
comply with Basel II when it becomes effective in 2006. Nonetheless, it
is our intent to qualify as an advanced practice institution. We simply
believe that it is good banking practice to develop the risk management
tools that are the foundation of Basel II: If that qualifies us as an
advanced practice company under the new accord, so much the better.
I believe my testimony today provides a rather unique perspective
on the issue of whether or not Basel II is good for the banking
industry. For the first 17 years of my professional career I was a bank
regulator with the Office of the Comptroller of the Currency (OCC).
Much of my time with the OCC was spent dealing with problem and failing
institutions. During my last several years with the OCC, I was Deputy
Comptroller for Special Supervision. That is a nice way of saying I was
responsible for the department that dealt with severely troubled and
failing financial institutions.
My tenure in the Special Supervision Department ran from 1986
through 1990, a time when a significant number of banks failed in the
United States. I was able to see first hand the myriad of reasons that
caused banks to get into trouble. Not the least of these was the
inability to appropriately identify and manage their risks.
I left the OCC in 1990 to join the deeply troubled Ameritrust
Corporation in Cleveland, Ohio. Ameritrust was a $12 billion company
that had encountered difficulties arising from its loan portfolio. I
was part of the new management team focused on turning the company
around. Over an 18-month period, Ameritrust lurched from one crisis to
another, but we eventually were able to stabilize the company. During
the interim period I lived, first hand, through the effects of a firm
that had little in the way of risk management practices and tools.
My experience with the OCC's failing banks division and the
Ameritrust debacle convinced me that there had to be a better way of
managing risk in the banking
industry.
In 1992, Ameritrust was acquired by Society Corporation, the
precursor of today's KeyCorp. I was placed in the position Executive
Vice President of Credit Policy and Risk Management. In this capacity,
I was given the opportunity to explore and
experiment with new risk management tools that were beginning to bud in
the industry. I was encouraged to do so by our CEO who expressed a
desire to have a system whereby he could understand the totality of
risk that our company faced on a daily basis.
Our CEO envisioned a process that could tell him how much aggregate
risk the company was taking, including the risks that emanated from our
credit, market, and operational activities. He wanted a system that
could allow us to increase, decrease, or maintain our risk position as
circumstances warranted. Neither of us realized it at the time, but he
was describing a process that is today commonly called ``enterprise-
wide risk management.''
In 1993, I commenced the first step of his vision by installing a
Value-at-Risk (VAR) system in our company's trading floor. VAR was a
highly complex model
designed to measure risk in the bond, equity, and foreign exchange
trading we undertook on a daily basis. Due to the complexity of a VAR
model, I had to engage several Ph.D.'s to help us implement it. During
the course of their engagement, I happened to mention my frustration in
finding an enterprise wide system that could aggregate the risk of each
of our banking activities. One of the Ph.D.'s suggested that I look
into the concept of economic capital allocation, now commonly known as
``risk-based capital.''
Once I investigated the premise of risk-based capital allocation, I
concluded I had discovered a powerful risk management tool.
Implementing such a model at KeyCorp would enable us to allocate
capital to our lines of business based on the amount of risk they took.
Each line of business would be charged for the amount of credit risk,
market risk, and operational risk they encountered. Using the aggregate
of that capital charge as the denominator, and the revenue they
generated as the numerator, we could determine which lines of business
were getting appropriately paid for the risk they took. For the first
time, we would be able to put all our lines of business on an apples-
to-apples comparison basis. Hence, the ability to know our level of
risk and whether or not we would be paid for the risk being taken.
Further, we would be able to aggregate the total amount of capital
being allocated to all our lines of business to understand the totality
of risk our company was taking. It was the enterprise-wide solution we
had been looking for.
KeyCorp commenced building an economic capital allocation program
in the mid-1990's because we firmly believed that it was the right
thing to do. It has taken us nearly a decade to build it, and we are
still not finished with it. Nonetheless, even after nearly 10 years we
remain convinced that it is the best way to run our company. No
regulator has told us that we must do this.
We are pleased to note that this powerful risk management tool,
economic capital allocation, is now the underlying driver of Basel II.
Our company was highly critical of the initial version of Basel II and
publicly stated as much. We felt that it failed to address the
sophistication and complexity that our industry routinely operated in.
We felt it was inadequate and little better than the original Basel I.
Put simply, it did not adequately address risk sensitivity. However,
over the next several years we were pleasantly surprised to see how
Basel II became a much better document. The regulators working on the
new accord have been genuinely receptive to hearing the concerns that
KeyCorp and others have raised. We haven't always gotten our way, but
at least we have been heard.
We believe that Basel II is now on the right track. Financial
institutions will need to develop more sophisticated risk management
tools to support the risk-based capital premise upon which it is built.
This is a good thing. In today's world of complex financial markets,
tools such as value-at-risk, two-dimensional loan grading systems,
enterprise data warehouses, and operational loss databases are not a
luxury; they are a necessity. In order to understand their risk
positions, banks should be calculating risk-based capital and using
these tools to do so. While models are no substitute for human
judgment, they certainly create a more informed human with whom to make
the decision.
One of the benefits we see in the Basel II proposal is that we will
finally be free to price our products and services commensurate with
the risk they entail. As previously mentioned, Basel I provides very
little in the way of risk sensitivity. One of the perversities of this
shortcoming is that it has driven high quality borrowers away from the
banking industry. These clients can access providers of credit not
subject to the costly level of capital that banks are currently
required to hold. In essence, banks are forced to overprice for this
business, and they lose it to other cheaper, nonregulated providers.
Conversely, Basel I's simplistic 8 percent capital requirement has
allowed banks to hold less capital than they should against borrowers
that are high risk. This has resulted in banks underpricing such
credit. It should be no surprise, then, that Basel I has chased high-
quality credits away from banks, while attracting low-quality credits
to them.
If banks are allowed to calculate the proper level of capital to be
held based on a realistic stratification of credit risk, this serious
problem will largely disappear. This is one of the tenets that Basel II
is based upon: You hold the level of capital necessary to support the
risk, and price for it accordingly.
I would now like to address some of the criticisms that have been
leveled against Basel II. These would include its cost, complexity,
inflexibility, and propensity to foster procyclicality. I would also
like to provide a few comments on the merits of Basel II's Pillar 1
versus Pillar 2.
Cost
Much has been said about the cost of building the models necessary
to comply with Basel II. At KeyCorp, we wonder how anyone can afford
not to build them. We, ourselves, have painfully learned the cost of
not having them. In 1996, our risk-based capital process was still in
its embryonic stage: In truth it did not begin to take hold until 2000.
In 1996, we were still calculating profitability measures
utilizing the primitive 8 percent capital standard stipulated by Basel
I. On this basis, one of our loan portfolios, leveraged lending, was
producing an eye-popping return on equity close to 30 percent. As a
consequence, we unfortunately pursued expansion of leveraged lending
over the next several years. At the end of 1998, the quality of this
portfolio began to collapse and we have written-off many millions of
dollars since.
We have looked retrospectively on our experience with this
portfolio. We believe if we had had our risk-based capital model in
place (the kind proposed by Basel II) our anticipated return would have
been in the single digit range. Such knowledge would have caused us to
avoid this particular lending activity and to seek other opportunities
that offered better risk/reward ratios.
Through this experience, we have learned an important lesson from
which others can benefit. The entire cost of the nearly 10-year effort
to implement our economic capital model (the same kind proposed by
Basel II) pales in comparison to the cost of not having it in place.
We have read that others estimate the cost of compliance with Basel
II to be staggeringly high. We are not convinced this is the case, and
it certainly has not been so at KeyCorp. Yes, we have spent multiple
millions of dollars over the years investing in risk management tools
and models, but we have done so because we believe those tools are
necessary to conduct our business in a safe and sound manner. Frankly,
they will also make us a better competitor. The more we understand our
risk, the better we will be at managing and pricing for it.
Some have criticized the cost of auditing and back-testing the
accuracy of the models that Basel II is based upon. We view such
activities as nothing more than good common sense. Auditing and back-
testing of outputs is critical to ensuring that the model is producing
reasonable numbers. Auditing/back-testing serve as the tuning devices
necessary to modify the models' calculations. For example, auditing and
back-testing of VAR models is an accepted practice in the industry now:
Everyone knows their benefit. We view auditing/back-testing as
necessary investments needed to create a better model. Better models
create better understanding of risk and the ability to better manage
it. Better management of risk results in lower losses to banks. We
believe the cost of auditing/back-testing is inconsequential compared
to the losses that can occur due to inferior risk management processes.
Before one accepts the large figures attributed to Basel II
compliance, one must subtract the costs of building the risk management
systems that a good financial institution would invest in, regardless.
We do not believe the gap between the two is significant.
Complexity
We cannot deny that Basel II is a complex document. It is. Yet, it
needs to be. Banking is a complex business that needs complex solutions
to the issues it faces. We should not run from complexity but instead
be willing to face it and manage our way through it.
I have previously mentioned that KeyCorp installed a VAR system for
its trading floors in the early 1990's. At that time, many were saying
VAR systems were exceedingly complex, expensive, and too mathematically driven. Yet, today VAR systems are widely recognized as the standard by
which to manage risk in their trading books. VAR is a superior risk
management tool that never would have come to be had the financial
services industry been intimidated by its complexity. I reiterate: When
VAR first surfaced, it was accused of being too complex, costly, and
mathematically driven, the same crimes Basel II stands accused of
today. Yet, VAR has become the industry standard.
Inflexibility
Some fear Basel II will trap the industry with year 2000 era risk
management tools and stifle creation of new ones. We believe this
concern is overstated. The 1988 Basel Accord was a woefully inadequate
document from the start. Its simplistic approach mandated a specific
capital level and made no provisions to the contrary. Yet, over the
past 15 years, the financial services industry has continued to develop
new risk management tools never envisioned by the 1988 Accord. These
would include: VAR models, two-dimensional loan grading systems,
economic capital models, and enterprise-wide data warehouses. The fact
that such tools were not contemplated by Basel I did not interfere with
the industry's pursuit of them. We anticipate a similar situation with
Basel II--banks will continue to pursue a better risk management
mousetrap. We will acknowledge, however, that regulators must be
willing to consider the new tools as they are developed, and work with
the industry to accommodate them as their effectiveness is
demonstrated.
Procyclicality
We have frequently heard that regulators are concerned that Basel
II might allow substantial capital to escape from the banking system.
We believe the whole premise of procyclicality is evidence that such
concerns may be overstated. Basel II capital levels represent the
minimum level of capital that an institution is to hold. The premise of
procyclicality assumes that banks operate at or near the minimum
capital level. We believe it is highly unlikely that any banking
company worth its salt will allow their capital to sink to the lowest
acceptable level.
Some argue that under Basel II, economic downturns will cause
financial institutions to become more reluctant to lend when liquidity
is most needed. Banks would be placed in a position of making a
difficult choice: Immediately raise new capital or stop lending. In
truth, there is a third choice that most banks will probably follow:
Retain a buffer level of capital to accommodate cyclical changes in
risk that everyone knows will inevitably occur.
We believe that even in times of economic stress, banks genuinely
desire to make new loans to drive their own revenue streams. Our
current economic situation is a prime example: Banks are anxious to
lend money. The demand is not there.
Pillar 1 versus Pillar 2
One of the basic principles of Basel II is to make risk transparent
so that it is comparable from one institution to another. Pillar 1
encourages a formulaic based system that will enable this to occur.
Consistency of methodology is critical to empower investors,
regulators, and depositors with the information they need to gauge the
risk of the institution with whom they are dealing. Without Pillar 1's
consistency of approach, a Tower of Babel syndrome can occur.
Pillar 2 relies more on flexible judgment as to how much capital is
warranted at an institution. We acknowledge and accept that regulators
must have the flexibility to invoke their authority to ignore the
results of Pillar 1 when circumstances so dictate. However, completely
abandoning Pillar 1 in favor of Pillar 2 yanks any comparability
benefit away from investors and depositors. The invisible hand of the
market will be impeded in its ability to quickly discipline a wayward
institution.
For example, much has been said about the need to place operational
risk under a Pillar 2 approach. In this regard, the individual
regulator that happened to be examining a particular bank would largely
determine the adequacy of capital held for its operational risk. This
lends itself to varying assessments, interpretations, methodologies,
and enforcements. An investor attempting to compare the level of
capital held for operational risk at multiple banks must assume that
different examiners will utilize the exact same thinking in their
operational risk assessments. That simply doesn't happen. A more
formulaic approach, where all banks are using the same scorecard, lends
itself much more to consistent comparability.
The mere presence of a Basel II draft has caused many in the
industry to start contemplating new ways of tracking operational risk.
This would include KeyCorp. We have commenced building an operational
risk database that will give us better information regarding the
source, size, and amount of operational losses. This database will
ultimately serve as the system that feeds our operational risk model.
We believe it can be supplemented by exchanging information on
operational risk losses with other financial institutions. This will
help us build the critical mass necessary to create reliable,
predictive loss forecasting models. I will readily admit that we have a
way to go in this particular area, but the presence of Basel II over
our heads encouraged KeyCorp and others in the industry to get moving
on building the databases sooner.
Conclusion
In conclusion, KeyCorp believes that Basel I is hopelessly broken
and that a new accord needs to be implemented. Basel II is a major step
forward and we applaud its approach. It is not perfect now, nor will it
be perfect when implemented, nor perfect 10 years after implementation.
Regardless, it is light years ahead of Basel I, as well as any other
proposal we have seen to date. It should be supported.
We acknowledge it is complex, but banking is a complex business. A
simple solution to complex issues is probably not the right medicine.
As an industry, we should not shy away from the remedy simply because
it is complex. Instead, we should work collectively with the regulators
to find the right solution, not the easy one.
We have our doubts as to the high cost figures attributed to Basel
II. Our own experience to date has proven to the contrary. Further, we
believe many Basel II costs are simply expenditures we should otherwise
be making as a matter of sound banking practice. Good risk management
costs money, but it is intended to help avoid even bigger costs that
arise from bad risk management.
We do not believe adoption of Basel II will trap the financial
services industry in a time warp. Banks will continue to develop better
methods of managing risk regardless of what Basel II requires. However,
regulators must be open and responsive as these new tools are
developed.
We believe there is substantial merit to including as much as we
can in Pillar 1 versus Pillar 2. One of the greatest benefits that
Basel II promises is that it will utilize the invisible hand of the
market to discipline wayward institutions. In order to do that,
investors must have adequate information to compare the risk of one
institution against another on an apples-to-apples basis. Pillar 1 is
the best vehicle for ensuring that banks report on a consistent basis.
Mr. Chairman, KeyCorp appreciates the opportunity to share our
views on Basel II. We want to make sure our industry operates within a
safe and sound environment. We know this is the goal of the Committee
as well as our friends in the regulatory world. While Basel II is far
from perfect, it certainly moves us further down the path.