via e-mail
November 3, 2003
Robert E. Feldman, Executive Secretary
Attention: Comments
Federal Deposit Insurance Corporation
550 17th Street N.W.
Washington, D.C. 20429
Comments@FDIC.gov
Re: Risk-Based Capital Guidelines; Implementation of New Basel
Capital Accord 68 FR 45900 (August 4, 2003)
Dear Mr. Feldman:
On behalf of People’s Bank (“People’s”), I am pleased to comment on
the Advance Notice of Proposed Rulemaking (ANPR) concerning
implementation of the new Basel Capital Accord (the “New Accord” or
“Basel II”) that is being developed by the Basel Committee on Banking
Supervision at the Bank for International Settlements. People’s is a
state-chartered savings bank, headquartered in Bridgeport, CT; it has
managed assets of $12 billion and 150 branches across the state of
Connecticut.
Position Summary
People’s Bank supports the efforts of the Basel Committee and the
U.S. regulatory agencies to more closely link minimum capital
requirements with an institution’s risk profile. However, we have a
number of substantive and inter-related concerns regarding both the New
Accord itself and the proposed manner of its implementation in the
United States. These include:
1. The undue emphasis in the New Accord on complex and theoretical
models;
2. The potential competitive implications of the proposed two-tiered
regulatory capital framework, wherein some number of financial
institutions will be competing pursuant to one set of capital rules and
a significantly greater number of institutions will be governed under
another set of capital rules;
3. The “all-or-nothing” nature of the proposed implementation in the
United States; and
4. The potential that in the future market and/or regulatory forces
will compel smaller and less complex institutions to implement costly
systems and processes that are of only limited benefit.
As a result, we urge that the agencies consider alternative
approaches that might address potential competitive dislocations without
abandoning the underlying goals of the New Accord.
Emphasis on Complex Models
We are well aware of the value of sophisticated models in managing
risk as well as other aspects of our business. When developed and
utilized in an appropriate manner, models can add great value to an
institution’s efforts to manage risk, and regulators are to be commended
for their efforts to encourage the adoption of best practices by all
institutions.
However, we also agree with the testimony of FDIC Chairman Powell
that “in reality, no one knows the range of potential future losses for
a given activity, or the associated probabilities”. We believe that
effective management and supervision of capital requires only a
reasonable estimation of risk, rather than the expense and illusion of
attempting to achieve precision. While highly sophisticated models based
on granular data are indisputably of value, it is important to bear in
mind that they are but one tool within the risk management toolkit.
We believe that the proposal places an emphasis on sophisticated
quantification that is out of proportion to the impact of such
quantification on either estimating or reducing the actual risk profile
of an organization. While the additional data is undoubtedly of some
value, it is simply not believable that an institution that makes the
very significant investment required to develop that data is
significantly less risky than one with an identical profile but more
rudimentary calculations.
This is true in terms of both operational risk and credit risk. While
quantification of operational risks is of value to the effective
management of those risks, it is clearly secondary to effective
oversight and management practices and a corporate culture attuned to
understanding and mitigating risk. However, the effect of the proposal
would suggest that an institution that undertakes the level of
quantification required under the Advanced Management Approach requires
significantly less capital than a similar institution that maintains an
effective operational risk management program but has not made the same
investment in quantification of operational risk. While convenient from
a capital calculation standpoint, this assertion is theoretically
unsound.
The same holds true for the treatment of credit risk. Under the
proposal, required capital will not reflect even broadly the risk level
associated with the underlying portfolios (i.e. it will continue to be
measured using Basel I standards) unless the institution invests in the
development and maintenance of granular data that is arguably of only
incremental value in determining appropriate levels of required capital.
The illusion of precision created by the use of more sophisticated
models can suggest that less data-intensive models are significantly
less valid when in fact the actual difference in validity may in fact be
much less significant.
In addition, we are concerned about the potential for significant,
and potentially systemic, disruption in cases in which models have been
inappropriately developed or utilized. Recent history is replete with
instances in which the utilization of sophisticated and opaque models
have resulted in significant disruption to the economy and the financial
system.
These concerns are magnified in this instance for two reasons. First,
we are troubled by the potential for either intentional or unintentional
manipulation of model results. The typically wide range of “reasonable”
assumptions, coupled with the incentive to use those assumptions which
produce the lowest possible capital requirements, make this approach
problematic. Second, we are concerned that it may not be possible for
regulators to provide effective oversight given the complexity of the
models, the number and nature of the required assumptions, and the
incentives cited above on the part of those they are examining.
In summary, we are concerned that the heavy emphasis on investments
in quantification that are only affordable for the largest institutions
not only puts smaller and less complex institutions at a disadvantage,
but is out of proportion to the incremental value of the more
sophisticated approaches and may in fact introduce new systemic risks to
the banking system.
Competitive Implications of a Two-Tiered Framework
While we appreciate the efforts of the agencies to spare smaller
institutions from the expense and effort of adopting the New Accord, we
believe that a two-tiered framework is inherently risky, based primarily
on the high potential for unintended consequences.
In particular, we believe that the proposed two-tiered system has the
potential to incent non-Basel II institutions to increase their
concentrations of riskier assets (against which they will be required to
hold less capital than their New Accord counterparts) and to decrease
their concentrations of less risky assets (against which they will be
required to hold relatively more capital).
We take issue with the argument that changes in regulatory capital
for large banks will not impact smaller banks. Capital is a fundamental
element of financial management, and is actively measured and managed by
all institutions. It is a key focus of investors, rating agencies, and
other external parties. While it may not in all cases be explicitly
linked to the pricing of each individual transaction, it certainly is a
critical consideration in the analyses and decisions that govern an
institution’s appetite for a particular product line. To the extent that
larger institutions receive favored treatment in this regard, it will be
reflected in their financial performance and will inevitably influence
investor sentiment.
Even if one were to assume that the proposed two-tiered approach
would not directly influence the decision-making process at non-Basel II
institutions, it is highly likely to result in indirect impacts. For
instance, as Basel II institutions adjust their pricing and their
appetite for certain types of assets based on the revised capital
requirements, the competitive impacts will be felt by the smaller
institutions with which they compete. This raises the possibility, for
example, that smaller institutions may find themselves unable to compete
with pricing offered by Basel II institutions on residential mortgages
but increasingly competitive in more risky product categories.
We must admit that, to our knowledge, there is no empirical evidence
to suggest that a two-tiered regulatory capital framework will harm
smaller institutions. However, it is equally true that there is no
empirical evidence that this will not be the case, and, if it were to
happen, no reason to believe that action could be taken in a manner
timely enough to mitigate the negative impacts. To the extent that the
United States is committed to maintaining a robust banking system which
includes numerous competitors of varying size and geography, we believe
it is incumbent on the agencies to proactively ensure, rather than
simply assume, that the proposal will not result in competitive
inequities.
“All or Nothing” Implementation
The proposed implementation requires that, in order to realize the
potential benefits of a more risk-sensitive regulatory capital regime,
institutions must implement the most advanced versions of the New Accord
– the Advanced Internal Ratings Based (A-IRB) approach for credit risk
and the Advanced Management Approach (AMA) for operational risk. We
believe that this approach is unnecessarily draconian in that it makes
it all but impossible for smaller institutions to realize even the most
obvious benefits.
A good example of our concern relates to Residential Mortgages. This
asset type is popular with relatively small and less complex
institutions, and there is ample evidence to suggest that, under the
current regulatory capital requirements, it is assessed an
inappropriately high capital charge. The New Accord quite rightly
enables institutions with portfolios of high asset quality mortgages to
reduce their regulatory capital on those assets, as well as recognize
the benefits of private mortgage insurance where appropriate.
Unfortunately, under the proposal this adjustment will only be
available to Basel II institutions, putting this benefit out-of-reach
for all but the largest institutions.
One can debate whether or not an institution should be required to
meet the extensive and, in our view, overly complex A-IRB requirements
specific to residential mortgages in order to enjoy reduced regulatory
capital on those assets. However, the proposal requires that, in
addition to meeting those requirements specific to residential
mortgages, institutions would have to meet costly requirements specific
to all other asset classes as well as significant operational risk
quantification requirements in order to enjoy more appropriate treatment
of its mortgage portfolio. In our view, this is akin to failing a
student in history because the student did not sign up for an advanced
biology class. If prime quality residential mortgage assets are
materially less risky than is reflected under the current regulatory
capital requirements, the more accurate treatment should be more
reasonably available to all institutions and not solely to those that
make large and unrelated expenditures.
The residential mortgage example is but one particularly stark
example of concerns related to this issue. Clearly, a very large
institution focused on residential mortgage lending could potentially
achieve competitive advantage relative to non-Basel II banks based on
its ability to implement the New Accord. But more generally, we are
concerned that the potential for competitive dislocation is present in
any situation in which there is a yawning gap in the treatment of an
identical asset pool under two separate capital regimes.
Future Developments
As mentioned above, People’s Bank is appreciative of the effort by
the agencies to spare smaller institutions from the very large resource
commitment required to implement the New Accord. However, we are
concerned that this benefit may prove only temporary as implementation
of the New Accord inevitably raises expectations on the part of
investors, rating agencies, and even the agencies themselves.
We are fully supportive of efforts by the agencies to incent
institutions to improve their risk management capabilities. Replacing
the current regulatory capital regime with a more risk-sensitive one is
consistent with that goal. However, as cited above, we believe that some
of the complex and costly requirements of the New Accord are out of
proportion to the resulting risk management benefits, particularly in
the case of smaller institutions. Based on this reasoning, the decision
to exempt smaller and less complex institutions from Basel II is a wise
one.
However, as part of this process the agencies should consider not
only the immediate issues but also the likely future implications of
this decision. It is inevitable that as investors, analysts, rating
agencies and others become accustomed to the volume of data that will
become available from Basel II institutions, they will expect to see the
same level of data from other institutions. When it is unavailable, it
is likely that their opinion of an institution will be diminished,
consciously and/or sub-consciously. The implicit regulatory “seal of
approval” that comes with adoption of the New Accord (particularly in
terms of the supervisory oversight of risk management processes) will
inevitably influence third parties as they evaluate the management and
performance of the institutions they assess. In addition, it is likely
that secondary market practices will evolve to require data available
from Basel II banks, and institutions which are unable to provide it
will be unable to participate on equal terms with those that can.
This would be a defensible and perhaps even desirable outcome if in
fact the existence of the added data actually provided those third
parties with greater insights or more assurance regarding the risk
sensitivity of a particular institution. However, we would argue for the
reasons discussed above that, particularly in the case of smaller and
less complex institutions, adoption of the New Accord will be more
reflective of an institution’s size than its level of risk. To the
extent that this is not understood or accepted by third parties, the
potential exists for disparate treatment in the public markets between
core banks and non-Basel II institutions. We fear that smaller
institutions may eventually have little choice but to opt in, at great
expense and with less ability to spread implementation and ongoing costs
over a large enterprise.
We believe firmly that the agencies should strive to ensure that
smaller and less complex institutions are not disadvantaged in any way
by responsible business decisions to avoid the monetary and other costs
of opting in to the New Accord and that the proposal does not
satisfactorily address this concern.
Recommendation
The development and implementation of a new regulatory capital regime
is undeniably complex and demanding. The agencies must take into account
a wide range of perspectives in seeking to ensure that regulatory
capital is applied based on the relative risk levels of various
institutions. In particular, the agencies need to ensure that large and
internationally active U.S. institutions are treated fairly relative to
their global competitors. At the same time, the agencies need to ensure
that smaller U.S. institutions are not unfairly disadvantaged relative
to larger domestic or global competitors. Our impression is that more
time and effort has been expended to date on understanding and
addressing the New Accord’s impact on larger institutions than on
smaller competitors. While that focus was necessary to bring the process
to this point, we believe that it is now appropriate to expend
additional time and energy on addressing issues of concern to smaller
institutions.
We believe that the issues discussed above can best be addressed
through the development of a new regulatory capital regime that would
apply to all institutions regardless of size, and which would address
the most glaring deficiencies of Basel I without requiring the somewhat
overwhelming complexity of Basel II.
This approach would primarily involve adding more “buckets” to Basel
I which would enable differentiated treatment of more finite asset pools
based not only on the product type, but also on elements such as
loan-to-value ratio, credit scores, etc. Given the fact that capital
requirements can be effectively determined using reasonable
approximations rather than theoretically precise determinations of risk,
this approach would succeed in replacing Basel I with a more
risk-sensitive approach without introducing the cost, complexity or
systemic risk of the more arcane approach.
In terms of operational risk, we are supportive of the agencies’
efforts to heighten awareness of the importance of an effective
operational risk management program. As such, we suggest that the
agencies provide banks with capital-related incentives for the
implementation and management of such a program, with quantification of
operational risks a somewhat secondary element. For instance,
institutions might receive capital relief if, under pillar two, they are
in compliance with AMA level operational risk management activities, a
slightly lesser incentive if they maintain an effective program that
lacks the more arcane elements of quantification but satisfies examiners
that the institution’s operational risks are effectively identified,
assessed, and managed, and somewhat higher required capital if they do
not have an effective program in place.
We recognize that the agencies may feel it necessary to require the
A-IRB and AMA approaches as outlined in the proposal for larger and
internationally active institutions for reasons of global
competitiveness and cooperation. To the extent this is the case, we
still believe that the introduction of a new regulatory capital regime
as described above would be effective in addressing the competitive
concerns of smaller and less complex institutions. Such an approach
could dramatically reduce the gaps in capital treatment for institutions
with similar levels of risk at reasonable cost to the institutions and
the agencies that oversee them.
Thank you for the opportunity to comment on this important matter. If
you or any member of your staff have any questions concerning the
matters discussed in this letter, please do not hesitate to contact me
at (203) 338-4585, or via e-mail at Bill.Kosturko@peoples.com or Ken
Weinstein at (203) 338-3052, or via e-mail at Ken.Weinstein@peoples.com.
Very truly yours,
William T. Kosturko
Executive Vice President/General Counsel
Legal and Government Affairs
People's Bank
Bridgeport Center
850 Main Street
Bridgeport, CT 06604-4913
Cc: The Honorable Christopher J. Dodd, United States Senate
The Honorable Christopher H. Shays, United States House of
Representatives
