via e-mail
November 4, 2003
Office of the Comptroller of the Currency
250 E Street, SW
Public Information Room
Mailstop 1-5
Washington, DC 20219
Ms. Jennifer J. Johnson, Secretary
Board of Governors, Federal Reserve System
20th Street and Constitution Avenue, NW
Washington, DC 20551
Robert E. Feldman
Executive Secretary,
Federal Deposit Insurance Corporation
550 17th Street, NW
Washington, DC 20429
Attention: Comments
Regulation Comments
Chief Counsel’s Office
Office of Thrift Supervision
1700 G Street, NW
Washington, DC 20552
Ladies and Gentlemen:
Juniper Financial Corp. and its wholly
owned subsidiary Juniper Bank (hereinafter collectively referred to as
“Juniper”) appreciate the opportunity to comment on the Advance Notice
of Proposed Rulemaking (“ANPR”) regarding the “Risk-Based Capital
Guidelines; Implementation of New Basel Capital Accord”.
Juniper Bank is a partnership focused
issuer of credit cards, with over $1.2 billion in managed credit card
receivables, approximately $500 million of on book assets, and
approximately 700,000 credit card accounts. Founded in 2001 it is one of
the fastest growing credit card issuers in the United States; yet it is
also one of the smallest banks issuing credit cards nationwide. Juniper
is an 89% owned subsidiary of the Canadian Imperial Bank of Commerce, a
United States Financial Holding Company, with approximately $285 billion
(Canadian) of on book assets and $688 billion (Canadian) of managed
assets.
At the outset, Juniper would like to
emphasize that it supports the Basel Committee on Banking Supervision’s
(“Committee”) goal of more precisely assessing regulatory capital
requirements in relation to risk. Any process that more accurately
aligns regulatory capital to risk improves the safety and soundness of
the banking system. At the same time, Juniper believes that any new
capital regime should be employed in a manner that does not create
competitive inequities or undue regulatory burden--especially if the new
regime does not more accurately align regulatory capital to risk.
Banking organizations should not be accorded a competitive handicap vis
a vis their competitors because they are forced to comply with the
Advanced Internal Rating Based (“IRB”) approach to credit risk in the
New Accord or because they continue to apply the current general risk
based capital rules. Moreover, any new regulatory capital regime should
be structured so that banks that are engaged in certain lines of
business, such as credit cards, are not unduly disadvantaged vis a vis
their non-bank competitors or vis a vis banks engaged in other lines of
banking business. Unfortunately, we at Juniper are concerned that the
New Basel Capital Accord (“New Accord”) might do just that.
Juniper believes that, as presently
constituted, the New Accord would require all credit card issuers to
hold more regulatory capital against their credit card assets than
presently required. This would cause banking organizations that issue
credit cards to be competitively disadvantaged vis a vis their non
banking competitors and could result in banking organizations shifting
investments away from credit cards to other banking product lines that
require less regulatory capital. Juniper itself would be negatively
impacted, either as a subsidiary of a large internationally active bank
that will be required to comply with the New Accord, or as a small
independent issuer of credit cards in the United States. Moreover,
Juniper believes much of this imbalance is caused by a misunderstanding
of the risks posed by credit cards and that a few relatively simple
modifications to the New Accord would go a long way in mitigating these
competitive inequities and would not cause incremental risk to the
safety and soundness of credit card banking.
Juniper notes that on October 11, 2003
the Committee issued a news release which declared that it has
“identified opportunities to improve the framework” of the New Accord.
Juniper further notes that the proposed areas identified for improvement
include the proposed allocation of regulatory capital for expected
losses, as opposed to unexpected losses, and the regulatory capital
treatment of “credit card commitments” and securitized assets. Juniper
strongly supports the direction in which the Committee seems to be
heading. It is consistent with many of the points we make below. At the
same time, since we have not seen the Committee’s actual proposed
amendments to the New Accord, we will comment on the ANPR as it is
presently drafted. Hopefully our comments will assist the Agencies in
their ongoing deliberations with the Committee regarding the drafting of
actual amendments.
Juniper has the following recommendations
regarding the New Accord:
Summary
1) Regulatory capital should not be
allocated for Expected Losses (“EL”), especially for Qualified Retail
Exposures (“QREs”), due to fact that the EL is incorporated into the
interest rate and pricing spread of QREs; 2) In any event, the proposed
future margin income (“FMI”) offset to EL should be increased from 75%
to 100% as consistent with economic pricing reality; 3) Banks should not
be required to incorporate undrawn lines of credit into their
calculation of the estimate of default (“EAD”) or loss given default (“LGD”);
4) Just as important, the requirement to hold capital against undrawn
lines of credit card accounts that have been securitized should be
eliminated; 5) Banks should not need to deduct dollar for dollar capital
from retained positions in securitizations between zero and KIRB if the
retained position has been rated by an independent debt rating agency;
6) The asset value correlation (“AVC”) to probability of default (“PD”)
does not reflect economic reality, 7) The dollar for dollar reduction to
Tier 1 Capital for capitalized FMI should be modified to apply only to
amounts of capitalized FMI greater than 25% of Tier 1 Capital; 8)
Operational risk regulatory capital considerations should incent
investment in risk mitigation and contingency planning; 9) While greater
transparency is a laudable concept, care must be taken to ensure that
Pillar 3 does not require the public disclosure of proprietary and
competitive information; 10) An additional overarching concern is that
the New Accord is incredibly complex and costly to implement.
1. Regulatory
Capital should not be allocated for EL.
As stated in the ANPR (p. 23), the EL is
incorporated into the interest rate and pricing spreads of all retail
banking products. It is simply one of the costs that retail lenders take
into consideration when pricing their products (along with costs for
acquiring the account, costs for servicing the account, etc.) with the
expectation that the price will more than cover all expected costs
(i.e., the profit margin). Assuming the retail lender is appropriately
pricing its product, the FMI associated with that product should be more
than sufficient to cover EL. Requiring regulatory capital to be set
aside for EL adversely impacts products with higher ELs (and higher
pricing to cover those higher ELs) – without any showing that it is
needed to cover increased risk. Credit cards especially would be
adversely impacted as they generally have higher ELs than most other
banking products; they also are generally priced higher to absorb the
higher ELs.
While it is entirely appropriate and
prudent to require regulatory capital to cover unexpected losses (“UL”),
given the fact that EL is baked into the price of retail banking
products, regulatory capital should not be assessed for EL. We note that
the Committee, in its October 11th communiqué, seems to agree with this
proposition. We also note, however, that the Committee seems to have
substituted reserves for regulatory capital and to have required that
reserves must equal EL. We are concerned that this also might constitute
a “one size fits all” approach that is not appropriate for all banking
products and that regulators may require credit card issuers to beef up
reserves to a greater degree than for other banking products. This could
create many of the same issues that would be created by requiring that
regulatory capital be allocated for EL. We at Juniper submit that it
would be more appropriate for the supervisory function to review, as
they do today, the retail bank’s estimate of EL, provisioning policies
and the adequacy of reserves as part of the supervisory review. If they
find that the estimate of EL, provisioning policies or reserves are
inadequate, they can require additional reserves or additional
regulatory capital as they do now. If the retail bank does a good job of
estimating EL (versus actual performance) and prices its products
accordingly and adequately provides for reserves, it should not be
penalized simply because the amount of EL is high.
2. The FMI Offset to EL
for QREs should be increased from 75% to 100%.
As a partial acknowledgement of the above
argument, the ANPR proposes that the total capital held against EL be
reduced by 75% of eligible FMI. We note that the proposed 75% offset is
itself reduced from a proposed 90% in the Quantitative Impact Study. At
the very least, should regulatory capital be allocated for EL, Juniper
believes that the appropriate offset number should be 100%.
As stated previously, banks price their
products by incorporating EL into their pricing. Not allowing a 100% FMI
offset clearly results in a regulatory capital charge that is higher for
higher priced products than for lower priced products regardless of any
increase in actual risk to the banking system. Moreover, the proposed
definition of eligible FMI for QREs is limited to the amount of income
the QRE accounts can be expected to generate over the next 12 months.
Anticipated income for new accounts can not be included. In order to use
FMI as an offset, the banking organization must be able to support their
estimate of eligible FMI on the basis of historical data. These appear
to be conservative and appropriate limitations. These limitations
further underscore the proposition that FMI should be allowed to offset
up to 100% of EL since FMI is conservatively defined. If the realistic
amount of FMI more than covers EL over the next year – it should be
allowed to be a total offset to EL-- again, with the caveat that the
supervisory process has the authority to disallow any portion of FMI
that the banking organization’s primary regulator believes does not
comport with safe and sound banking practice.
3. Banking
organizations should not be required to incorporate undrawn lines of
credit into
their calculations of the EAD or LGD
There should be no requirement to hold
regulatory capital against undrawn lines of credit where the undrawn
line can be terminated at will. On page 40 of the ANPR, it is
acknowledged that there is a substantial difference between credit card
undrawn lines of credit and undrawn wholesale lines of credit – “not
only in degree but also in kind.” The big difference is that banking
organizations have much more control over the credit risk imposed by
undrawn credit card lines than undrawn corporate lines. Unlike a
committed corporate line of credit, there are no limits to a credit card
lender’s ability to reduce its line exposure. Credit card banks can and
do actively manage their credit risk exposure by constantly reviewing
credit card accounts and their associated credit lines and by reducing
the lines on those accounts that are identified as high risk. It is one
of their primary methods of managing credit risk exposure and is
substantially different from corporate lines of credit. Requiring
regulatory capital to be set aside for undrawn corporate lines of credit
in the same manner as undrawn credit card lines is tantamount to
favoring corporate over credit card lending and creates competitive
inequality without consideration of the amount of actual credit risk
involved. It is our hope that the Committee’s reference in its October
11th communiqué to revisiting provisions regarding “credit card
commitments” represents an undertaking to address this issue.
4. The Requirement to hold
capital against undrawn lines of credit card accounts that have
been securitized should be eliminated.
The proposal that banking organizations
be required to hold capital against the full amount of undrawn lines on
accounts that have been securitized (ANPR, p. 41) is inconsistent with
the inherent risk posed by these lines. As stated above credit card
lines on credit card accounts are uncommitted lines; credit card issuers
can and do reduce and/or terminate unused lines at will in order to
manage their credit card risk. Just as important, in a typical
securitization, both drawn balances and undrawn balances are
securitized; third party investors are obligated to purchase at par
newly originated receivables on securitized accounts in order to
maintain their investor interest during the securitization’s revolving
period. Investors are required to purchase the receivables on a pro-rata
basis from all accounts in the Master Trust, including those in high
risk or high EL segments. There should not be a requirement to hold
regulatory capital against those undrawn lines of credit that may or may
not be accessed, when at the time those lines are accessed, the newly
created receivables are securitized. These are receivables that never
make it to the banking organization’s books – holding regulatory capital
against them makes no sense. At the very least, the requirement to
assess regulatory capital against undrawn lines of credit card accounts
should be limited to those securitizations that do not sell new loan
originations.
5. Banks should not
be required to deduct dollar for dollar capital from all retained
positions in securitizations between zero
and KIRB if the retained position has been
rated by a rating agency.
The requirement to deduct dollar for
dollar capital for all retained positions in securitizations between
zero and KIRB (ANPR, p. 78) does not provide equitable relief if the
retained position is rated by a independent debt rating agency. The
purpose of the external rating is to evaluate the level of retained
risk. The requirement to deduct dollar for dollar capital for retained
positions is not consistent with the risk the retained position actually
poses since it gives no credit to the rating assigned by an independent
agency. Moreover, this proposed dollar for dollar treatment is
inconsistent with the risk weighted capital approach required for
investors in securitizations. It also is inconsistent with the
standardized approach to calculating regulatory capital. We suggest that
the appropriate position would be to risk weight the retained position
based on the rating assigned by the debt rating agency – thereby
aligning the regulatory capital requirement with the risk the retained
position actually poses.
6. The proposed AVC to PD
does not reflect economic reality.
The ANPR assumes that AVC for QREs
declines as PD rises; that higher credit quality borrowers are more
likely to experience simultaneous defaults (on a proportionate basis)
than pools of lower quality borrowers because higher wealth individuals
are more sensitive to macroeconomic events (see ANPR p.44). This does
not comport with the experience of our industry. Historically, the
credit card industry has not shown a higher correlation of losses from
lower risk individuals during macroeconomic shocks – and any correlation
that might exist is significantly lower than the threshold used in the
PD calculation – that the AVC curve should be significantly flatter. The
result of the proposed curve is that companies originating less risky
accounts are being penalized with a higher correlation factor than is
warranted and ultimately a higher capital requirement.
7. The dollar for dollar
reduction to Tier 1 capital for certain securitization exposures should
be modified.
The proposed requirement that banking
organizations deduct capitalized FMI from Tier 1 capital should be
limited only to amounts of capitalized FMI greater than 25% of Tier 1
Capital. Other capital deductions should be deducted from total capital,
not 50% from Tier 1 and 50% from Tier 2. Both recommendations are
consistent with current FFIEC guidelines.
8. Operational Risk Rules
should incent the establishment of preventative controls.
Juniper agrees that a systemic and
rigorous analysis of Operational Risk is to be encouraged – especially
if it leads to taking measures to mitigate or reduce operational risk.
To the extent that the New Accord’s emphasis on operation risk achieves
that goal, it is to be lauded. Juniper believes that the most effective
way to manage operational risk is the establishment and maintenance of a
strong control and compliance environment; yet the New Accord gives no
weight or incentive for preventative controls. Banks are given no credit
for the investment in risk management and contingency planning. At the
very least, more weight must be given to the establishment of a strong
control and compliance environment and some sort of reduced capital
requirement be accorded for investments which directly lead to reduced
operational risk. This is best done through the supervisory process.
9. Pillar 3 should be
reworded to ensure it does not require public disclosure of proprietary
and confidential information.
Increased transparency is a laudable
concept and ensuring public disclosure as to how an institution
calculates its regulatory capital requirements could enhance market
discipline. However, any disclosure requirements contained in Pillar 3
need to be balanced against the concern of regulatory burden, complexity
and more importantly, the need to protect proprietary and confidential
information. Mandating disclosures beyond those currently mandated by
debt rating agencies, accounting and securities authorities could result
in the disclosure of information that non-regulated competitors are not
required to disclose. It will also add significant costs. Moreover, a
banking organization’s disclosure of various components of credit risk
(type of credit exposure, geographic distribution of loans, etc., see
ANPR p. 100) could result in the disclosure of highly confidential and
proprietary information that could be accessed and used by a banking
organization’s competitors. Significantly, this would impact relatively
small competitors with a limited product set such as Juniper to a
greater degree than it would larger institutions (with a more diverse
set of product lines) by revealing far more competitive information
about its sole product line – credit cards.
We recommend that the agencies work
closely with accounting and securities authorities to implement a
cohesive and consistent disclosure scheme.
10. The New Accord is
incredibly complex and will impose substantial regulatory burden
and implementation costs on those
banking organizations adopting it; time will be
needed to comply with it.
An overarching concern regarding the New
Accord is that it is too complex and the corresponding increase the
regulatory burden imposed on those attempting to comply with it is too
great. It has been estimated that the costs to even small banks for
complying with the New Accord will be at least $10 million* (it might
not be that much for Juniper, but it will be significant). Moreover,
given the overall complexity of the New Accord, banking organizations
are going to need to develop the systems, infrastructure and expertise
to support the New Accord. Time will be needed for implementation if
implementation is to be done well – at a minimum four years (regulators
themselves will require time to develop and coordinate their approach
for all institutions). Juniper appreciates the recent six month
extension for drafting the rule; it proposes that at least another year
extension be provided for implementation and compliance.
Conclusion – The net effect of the
above is that as presently constituted, the New Accord could cause
significant competitive harm to banks (versus their unregulated
competitors) and particularly to US banks focused on issuing credit
cards. Hopefully, the October 11th communiqué signals a willingness to
address the concerns. However, the New Accord as presently drafted,
would require credit card issuers to increase the amount of regulatory
capital to be held against credit card receivables; favor other forms of
retail lending over credit card lending, and would impose enormous
compliance costs and burdens. Juniper as a subsidiary of a large
international bank will be required to comply. Moreover, even were
Juniper to be spun off from CIBC, (as an independent relatively small
bank) Basel II would adversely impact Juniper. The compliance costs and
extra regulatory capital requirements could reduce its valuation in an
IPO. Even were it not required to adopt the Advance IRB approach (“AIA”)
to calculating regulatory capital and the Advance Measurement Approach
(“AMA”) to calculating operational risk, it is likely that its
supervisory regulators would impose additional regulatory capital
requirements on Juniper so as not to give it a competitive advantage
versus its larger credit card competitors who would be required to adopt
the Advanced IRB and AMA approaches to calculating regulatory capital.
While maybe not inevitable, there would clearly be some pressure to
place all credit card issuers on a so called “even playing field” and it
is certain that the pressure would be to require all issuers to increase
their levels of regulatory capital, not decrease them. Yet, unlike its
larger competitors, Juniper is not large enough to absorb easily
additional regulatory and compliance costs. Juniper and other small
issuers would clearly be more adversely impacted by any requirement to
increase regulatory capital than its competitors. Care should be taken
to ensure smaller credit card issuers are not unduly implaced.
Accordingly, we urge the Agencies to
consider seriously these suggestions and recommendations set forth
above. They would go a long way in ameliorating any competitive
inequities that the new Accord, as presently drafted, might cause.
Again, thank you for the opportunity to submit our comments to the ADPR.
Should anyone desire, we at Juniper would be delighted to discuss our
concerns further.
Sincerely,
Clinton W. Walker
Juniper Bank
Wilmington, DE
* Petrou, Karen Shaw, “Policy Issues
in Complex Proposals Warrant Congressional Scrutiny”, Testimony before
the Domestic and International Monetary Policy, Trade and Technology
Subcommittee on Financial Services, U.S. House of Representatives,
February 27, 2003.
| 
