[Federal Register Volume 61, Number 46 (Thursday, March 7, 1996)]
[Proposed Rules]
[Pages 9114-9119]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: 96-5235]



 ========================================================================
 Proposed Rules
                                                 Federal Register
 ________________________________________________________________________
 
 This section of the FEDERAL REGISTER contains notices to the public of 
 the proposed issuance of rules and regulations. The purpose of these 
 notices is to give interested persons an opportunity to participate in 
 the rule making prior to the adoption of the final rules.
 
 ========================================================================
 

  Federal Register / Vol. 61, No. 46 / Thursday, March 7, 1996 / 
Proposed Rules  

[[Page 9114]]



DEPARTMENT OF THE TREASURY

Office of the Comptroller of the Currency

12 CFR Part 3

[Docket No. 96-05]
RIN 1557-AB14

FEDERAL RESERVE SYSTEM

12 CFR Parts 208 and 225

[Regulations H and Y; Docket No. R-0884]

FEDERAL DEPOSIT INSURANCE CORPORATION

12 CFR Part 325

RIN 3064-AB72


Risk-Based Capital Standards; Market Risk; Internal Models 
Backtesting

AGENCIES: Office of the Comptroller of the Currency, Treasury; Board of 
Governors of the Federal Reserve System; and Federal Deposit Insurance 
Corporation.

ACTION: Joint notice of proposed rulemaking.

-----------------------------------------------------------------------

SUMMARY: The Office of the Comptroller of the Currency (OCC), the Board 
of Governors of the Federal Reserve System (Board), and the Federal 
Deposit Insurance Corporation (FDIC) (Agencies) are proposing to amend 
their July 25, 1995, proposal to incorporate a measure for market risk 
into their respective risk-based capital standards. The proposed 
amendment would provide additional guidance to an institution about how 
the multiplication factor used to calculate capital requirements for 
market risk under the internal models approach would be adjusted if 
comparisons of its internal model's previous estimates with actual 
trading results indicate that the internal model is inaccurate. The 
proposed amendment would increase the market risk capital charge for an 
institution with an inaccurate model.

DATES: Comments must be received on or before April 5, 1996.

ADDRESSES: Comments should be directed to:
    OCC: Comments may be submitted to Docket No. 96-05, Communications 
Division, Third Floor, Office of the Comptroller of the Currency, 250 E 
Street, S.W., Washington, D.C., 20219. Comments will be available for 
inspection and photocopying at that address. In addition, comments may 
be sent by facsimile transmission to FAX number (202) 874-5274, or by 
electronic mail to [email protected].
    Board: Comments directed to the Board should refer to Docket No. R-
0884 and may be mailed to William W. Wiles, Secretary, Board of 
Governors of the Federal Reserve System, 20th Street and Constitution 
Avenue, N.W., Washington, D.C., 20551. Comments may also be delivered 
to Room B-2222 of the Eccles Building between 8:45 a.m. and 5:15 p.m. 
weekdays, or to the guard station in the Eccles Building courtyard on 
20th Street, N.W., (between Constitution Avenue and C Street) at any 
time. Comments may be inspected in Room MP-500 of the Martin Building 
between 9 a.m. and 5 p.m. weekdays, except as provided in 12 CFR 261.8 
of the Board's rules regarding availability of information.
    FDIC: Written comments should be sent to Jerry L. Langley, 
Executive Secretary, Attention: Room F--402, Federal Deposit Insurance 
Corporation, 550 17th Street N.W., Washington, D.C. 20429. Comments may 
be hand delivered to Room F--402, 1776 F Street N.W., Washington, D.C. 
20429 on business days between 8:30 a.m. and 5 p.m. (Fax number (202) 
898-3838; Internet address: [email protected]). Comments will be 
available for inspection and photocopying in Room 7118, 550 17th 
Street, N.W., Washington, D.C. 20429, between 9 a.m. and 4:30 p.m. on 
business days.

FOR FURTHER INFORMATION CONTACT:
    OCC: Margot Schwadron, Financial Analyst, or Christina Benson, 
Capital Markets Specialist (202/874-5070), Office of the Chief National 
Bank Examiner. For legal issues, Ronald Shimabukuro, Senior Attorney, 
or Andrew Gutierrez, Attorney (202/874-5090), Legislative and 
Regulatory Activities Division.
    Board: Roger Cole, Deputy Associate Director (202/452-2618), James 
Houpt, Assistant Director (202/452-3358), Barbara Bouchard, Supervisory 
Financial Analyst (202/452-3072), Division of Banking Supervision and 
Regulation; or Stephanie Martin, Senior Attorney (202/452-3198), Legal 
Division. For the Hearing impaired only, Telecommunication Device for 
the Deaf, Dorothea Thompson (202/452-3544).
    FDIC: William A. Stark, Assistant Director, (202/898-6972), Miguel 
D. Browne, Deputy Assistant Director, (202/898-6789), or Kenton Fox, 
Senior Capital Markets Specialist, (202/898-7119), Division of 
Supervision; Jamey Basham, Counsel, (202/898-7265) Legal Division, 
FDIC, 550 17th Street N.W., Washington, D.C. 20429.

SUPPLEMENTARY INFORMATION:

Background

    The Agencies' risk-based capital standards are based upon 
principles contained in the agreement on International Convergence of 
Capital Measurement and Capital Standards (Accord) issued in July 1988. 
The Accord, proposed by the Basle Committee on Banking Supervision 
(Committee) and endorsed by the central bank governors of the Group of 
Ten (G-10) countries,1 assesses an institution's capital adequacy 
by weighting its assets and off-balance-sheet exposures on the basis of 
credit risk. In April 1995, the Committee issued a consultative 
proposal to supplement the Accord to cover market risk, specifically 
market risk in foreign exchange and commodity activities and in debt 
and equity instruments held in trading portfolios, in addition to 
credit risk.2 On July 25, 1995, the Board, the OCC, and the FDIC 
issued a joint proposal to amend their respective risk-based capital 
standards in accordance

[[Page 9115]]

with the consultative proposal (60 FR 38082) (July 1995 proposal). 
Under the July 1995 proposal, an institution with relatively large 
trading activities would calculate a capital charge for market risk 
using either its own internal value-at-risk (VAR) 3 model 
(internal models approach) or, alternatively, risk measurement 
techniques that were developed by the Committee (standardized 
approach). The institution would integrate the market risk capital 
charge into its risk-based capital ratios.
---------------------------------------------------------------------------

    \1\ The Committee is composed of representatives of the central 
banks and supervisory authorities from the G-10 countries (Belgium, 
Canada, France, Germany, Italy, Japan, Netherlands, Sweden, 
Switzerland, the United Kingdom, and the United States) and 
Luxembourg. The Agencies each adopted risk-based capital standards 
implementing the Accord in 1989.
    \2\ The Committee's document is entitled ``Proposal to issue a 
Supplement to the Basle Capital Accord to cover market risk.'' On 
December 11, 1995, the G-10 Governors endorsed a final supplement to 
the Accord incorporating a measure for market risk, subject to the 
completion of rulemaking procedures in countries that require such 
action. The final supplement is entitled ``Amendment to the Capital 
Accord to incorporate market risks.'' The proposal and the final 
supplement are available through the Board's and the OCC's Freedom 
of Information Office and the FDIC's Reading Room.
    \3\ Generally, the VAR is an estimate of the maximum amount that 
could be lost on a set of positions due to general market movements 
over a given holding period, measured with a specified confidence 
level.
---------------------------------------------------------------------------

    Under the internal models approach, an institution would calculate 
a VAR amount using its internal model, subject to certain qualitative 
and quantitative regulatory parameters. The institution's capital 
charge for market risk would equal the greater of (1) its previous 
day's VAR amount (calculated based upon a 99 percent confidence level 
and a ten-day holding period); or (2) an average of the daily VAR 
amounts over the preceding 60 business days multiplied by a minimum 
multiplication factor of three.
    The July 1995 proposal also provides that the Agencies could adjust 
the multiplication factor to increase an institution's capital 
requirement based on an assessment of the quality and historical 
accuracy of the institution's risk management system. One of the 
proposal's qualitative criteria, which supervisors would use to 
evaluate the quality and accuracy of a risk management system, is that 
an institution would have to conduct regular backtesting. Backtesting 
involves comparing the VAR amounts generated by the institution's 
internal model against its actual daily profits and losses (outcomes).

Supervisory Framework for the Use of Backtesting

    Since issuing its consultative proposal, the Committee developed a 
framework that more explicitly incorporates backtesting into the 
internal models approach and directly links backtesting results to 
required capital levels.4 This framework recognizes that 
backtesting can be useful in evaluating the accuracy of an 
institution's internal model, and also acknowledges that even accurate 
models (i.e., models whose true coverage level is 99 percent) can 
perform poorly under certain conditions.
---------------------------------------------------------------------------

    \4\ The Committee sets out this framework in a document entitled 
``Supervisory framework for the use of `backtesting' in conjunction 
with the internal models approach to market risk capital 
requirements,'' which accompanies the document entitled ``Amendment 
to the Capital Accord to incorporate market risks,'' supra note 2.
---------------------------------------------------------------------------

    The Agencies agree with the Committee that backtesting can be a 
useful tool in evaluating the performance of an institution's internal 
model but recognize that backtesting techniques are still evolving and 
that they differ among institutions. The Agencies believe that the 
framework for backtesting developed by the Committee adequately 
recognizes the limitations of backtesting, while providing incentives 
for institutions to improve the efficiency of their internal models. 
The Agencies, therefore, are proposing to amend their July 1995 
proposal to incorporate a backtesting framework similar to the one 
endorsed by the G-10 Governors, as described later in the supplementary 
information.
    Under the supervisory framework for backtesting, an institution 
must compare its internal model's daily VAR amount with the following 
day's trading outcome. The institution must use the daily VAR amount 
generated for internal risk measurement purposes, not the daily VAR 
amount generated for supervisory capital purposes. Moreover, when 
making this comparison, the institution must first adjust the VAR 
amount, if necessary, to correspond to an assumed one-day holding 
period and a 99 percent confidence level.
    An institution must count the number of times that the magnitude of 
trading losses on a single day, if any, exceeds the corresponding day's 
adjusted VAR amount during the most recent 250 business days 
(approximately one year) to determine the number of exceptions. The 
number of exceptions, in turn, will determine whether and how much an 
institution must adjust the multiplication factor it would use when 
calculating capital requirements for market risk. However, if the 
institution demonstrates to its supervisor's satisfaction that an 
exception resulted from an accurate model affected by unusual events, 
the supervisor may allow the institution to disregard that exception.
    The Agencies recognize that there may be several explanations for 
exceptions. For example, an exception may result when an institution's 
internal model does not capture the risk of certain positions or when 
model volatilities or correlations are not calculated correctly. This 
type of exception reflects a problem with the basic integrity of the 
model. In other cases, the model may not measure market risk with 
sufficient precision, implying the need to refine the model. Other 
types of exceptions, on the other hand, may occur occasionally even 
with accurate models, such as exceptions resulting from unexpected 
market volatility or large intra-day changes in the institution's 
portfolio.
    Backtesting results also could prompt the supervisor to require 
improvements in an institution's risk measurement and management 
systems or additional capital for market risk. When considering 
supervisory responses, the Agencies would take into account the extent 
to which trading losses exceed the VAR amounts, since exceptions that 
greatly exceed VAR amounts are of greater concern than are exceptions 
that exceed them only slightly. The Agencies also could consider, for 
example, other statistical test results provided by the institution, 
documented explanations for individual exceptions, and the 
institution's compliance with applicable qualitative and quantitative 
internal model standards. The first backtesting for regulatory capital 
purposes is scheduled to begin in January 1999, using VAR amounts and 
trading outcomes beginning in January 1998.

Framework for Interpreting Backtesting Results

    This framework attempts to balance the possibility that an accurate 
risk model would be determined inaccurate (Type I error) and the 
possibility that an inaccurate model would be determined accurate (Type 
II error). Consequently, it divides the number of possible exceptions 
into three zones:
    (1) The green zone (four or fewer exceptions)--Backtest results do 
not themselves suggest a problem with the quality or accuracy of the 
institution's internal model. In these cases, backtest results are 
viewed as acceptable, given the supervisors' concerns of committing a 
Type I error. Within this zone, there is no presumed increase to an 
institution's multiplication factor.
    (2) The yellow zone (five through nine exceptions)--Backtest 
results raise questions about a model's accuracy, but could be 
consistent with either an accurate or inaccurate model. If the number 
of exceptions places an institution into the yellow zone, then it must 
adjust its multiplication factor. Because a larger number of exceptions 
carries a stronger presumption that the model is inaccurate, the 
adjustment to an institution's multiplication factor increases with the 
number of exceptions. Accordingly, the institution would adjust its 
multiplication factor by the amount corresponding to the number of 
exceptions as shown in Table 1.
    (3) The red zone (ten or more exceptions)--Backtest results 
indicate a

[[Page 9116]]

problem with the institution's internal model, and the probability that 
the model is accurate is remote. Unless the high number of exceptions 
is attributed to a regime shift involving dramatic changes in financial 
market conditions that result in a number of exceptions for the same 
reason in a short period of time, the institution must increase its 
multiplication factor from three to four, and improve its risk 
measurement and management system.
    The presumed adjustments to an institution's multiplication factor 
based on the number of exceptions follow:

   Table 1--Adjustment in Multiplication Factor From Results of Backtesting Based on 250 Trading Outcomes \1\   
----------------------------------------------------------------------------------------------------------------
                                                                                                      Cumulative
                                                                                      Adjustment to  probability
                      Zone                                No. of exceptions          multiplication      (in    
                                                                                         factor        percent) 
----------------------------------------------------------------------------------------------------------------
Green Zone......................................  4 or fewer.......................         0.00         89.22  
                                                  5................................         0.40         95.88  
                                                  6................................         0.50         98.63  
Yellow Zone.....................................  7................................         0.65         99.60  
                                                  8................................         0.75         99.89  
                                                  9................................         0.85         99.97  
Red Zone........................................  10 or more.......................         1.00         99.99  
----------------------------------------------------------------------------------------------------------------
\1\ The zones are defined according to the cumulative probability of obtaining up to a given number of          
  exceptions in a sample of 250 independent observations when the true level of coverage is 99 percent. The     
  yellow zone begins where the cumulative probability equals or exceeds 95 percent, and the red zone begins     
  where the cumulative probability equals or exceeds 99.99 percent.                                             

    The Agencies urge institutions to continue working on improving the 
accuracy of backtests that use actual trading outcomes and to develop 
the capability to perform backtests based on the hypothetical changes 
in portfolio value that would occur if there were no intra-holding 
period changes (e.g., from fee income or intra-holding period changes 
in portfolio composition).

Questions on Which the Agencies Specifically Request Comment

    1. Some industry participants have argued that VAR measures cannot 
be compared against actual trading outcomes because the actual outcomes 
will be contaminated by intra-day trading and the inclusion of fee 
income booked in connection with the sale of new products. The results 
of intra-day trading, they believe, will tend to increase the 
volatility of trading outcomes while the inclusion of fee income may 
mask problems with the internal model. Others have argued that the 
actual trading outcomes experienced by the bank are the most important 
and relevant figures for risk management and backtesting purposes.
    What are the merits and problems associated with performing 
backtesting on the basis of hypothetical outcomes (e.g., the changes in 
portfolio values that would occur if end-of-day positions remained 
unchanged with no intra-day trading or fee income)?
    What are the merits and problems associated with performing 
backtesting on the basis of actual trading profits and losses?
    2. What, if any, operational problems may institutions encounter in 
implementing the proposed backtesting framework? What changes, if any, 
should the Agencies consider to alleviate those problems?
    3. What type of events or regime shifts might generate exceptions 
that the Agencies should view as not warranting an increase in an 
institution's multiplication factor? How should the Agencies factor in 
or exclude the effects of regime shifts from subsequent backtesting 
exercises?
    4. The adjustments to the multiplication factor set forth in Table 
1 of the proposal are based on the number of exceptions in a sample of 
250 independent observations. Should the Agencies permit institutions 
to use other sample sizes and, if so, what degree of flexibility should 
be provided?
    5. The Agencies recognize that an institution may utilize different 
parameters (e.g., historical observation period) for the VAR model that 
it employs for its own risk management purposes than for the VAR model 
that determines its market risk capital requirements (as specified in 
the July 1995 proposal). Should the adjustment to an institution's 
multiplication factor be determined using trading outcomes backtested 
against the institution's VAR amounts generated for internal risk 
management purposes or against the VAR amounts generated for market 
risk capital requirements? Should the Agencies permit an institution to 
choose? Should backtesting be required against both sets of VAR 
amounts?

Regulatory Flexibility Act Analysis

OCC Regulatory Flexibility Act Analysis

    Pursuant to section 605(b) of the Regulatory Flexibility Act, the 
Comptroller of the Currency certifies that this proposal would not have 
a significant impact on a substantial number of small business entities 
in accord with the spirit and purposes of the Regulatory Flexibility 
Act (5 U.S.C. 601 et seq.). Accordingly, a regulatory flexibility 
analysis is not required. The impact of this proposal on banks 
regardless of size is expected to be minimal. Further, this proposal 
generally would apply to larger banks with significant trading 
activities and would cover only trading activities and foreign exchange 
and commodity positions throughout the bank.

Board Regulatory Flexibility Act Analysis

    Pursuant to section 605(b) of the Regulatory Flexibility Act, the 
Board does not believe this proposal would have a significant impact on 
a substantial number of small business entities in accord with the 
spirit and purposes of the Regulatory Flexibility Act (5 U.S.C. 601 et 
seq.). Accordingly, a regulatory flexibility analysis is not required. 
In addition, because the risk-based capital standards generally do not 
apply to bank holding companies with consolidated assets of less than 
$150 million, this proposal would not affect such companies.

FDIC Regulatory Flexibility Act Analysis

    Pursuant to section 605(b) of the Regulatory Flexibility Act (Pub. 
L. 96-354, 5 U.S.C. 601 et seq.), it is certified that the proposal 
would not have a significant impact on a substantial number of small 
entities.

Paperwork Reduction Act

    The Agencies have determined that this proposal would not increase 
the regulatory paperwork burden of banking organizations pursuant to 
the provisions

[[Page 9117]]

of the Paperwork Reduction Act (44 U.S.C. 3501 et seq.).

OCC Executive Order 12866 Determination

    The OCC has determined that this proposal is not a significant 
regulatory action under Executive Order 12866.

OCC Unfunded Mandates Reform Act of 1995 Determination

    The OCC has determined that this proposal would not result in 
expenditures by state, local, and tribal governments, or by the private 
sector, of $100 million or more in any one year. Accordingly, a 
budgetary impact statement is not required under section 202 of the 
Unfunded Mandates Reform Act of 1995.

List of Subjects

12 CFR Part 3

    Administrative practice and procedure, Capital, National banks, 
Reporting and recordkeeping requirements, Risk.

12 CFR Part 208

    Accounting, Agriculture, Banks, banking, Confidential business 
information, Crime, Currency, Federal Reserve System, Mortgages, 
Reporting and recordkeeping requirements, Securities.

12 CFR Part 225

    Administrative practice and procedure, Banks, banking, Federal 
Reserve System, Holding companies, Reporting and recordkeeping 
requirements, Securities.

12 CFR Part 325

    Administrative practice and procedure, Banks, banking, Capital 
adequacy, Reporting and recordkeeping requirements, Savings 
associations, State non-member banks.

Authority and Issuance

Office of the Comptroller of the Currency

12 CFR CHAPTER I

    For the reasons set out in the preamble, part 3 of title 12 of 
chapter I of the Code of Federal Regulations, as proposed to be amended 
at 60 FR 38082, is further proposed to be amended as follows:

PART 3--MINIMUM CAPITAL RATIOS; ISSUANCE OF DIRECTIVES

    1. The authority citation for part 3 continues to read as follows:

    Authority: 12 U.S.C. 93a, 161, 1818, 1828(n), 1828 note, 1831n 
note, 3907, and 3909.

    2. Appendix B to part 3 as proposed to be added at 60 FR 38095 
would be amended by revising paragraph (a)(2) of section 4 and by 
adding a new paragraph (d) to section 5 to read as follows:

Appendix B to Part 3--Market Risk

* * * * *

Section 4. Market Risk Exposure

* * * * *
    (a) * * *
    (2) The average of the daily value-at-risk amounts for each of 
the preceding 60 business days times a multiplication factor of 
three, except as provided in section 5(d).
* * * * *

Section 5. Qualifying Internal Market Risk Model

* * * * *
    (d) Backtesting. A bank using an internal market risk model 
shall conduct backtesting as follows:
    (1) The bank shall conduct backtesting quarterly;
    (2) For each backtesting, the bank shall compare the previous 
250 business days' trading outcomes with the corresponding daily 
value-at-risk measurements generated for its internal risk 
measurement purposes, calibrated to a one-day holding period and a 
99 percent confidence level;
    (3) The bank shall consider each business day for which the 
trading loss, if any, exceeds the daily value-at-risk measurement as 
an exception; however, the OCC may allow the bank to disregard an 
exception if it determines that the exception does not reflect an 
inaccurate model; and
    (4) Depending on the number of exceptions, a bank shall adjust 
the multiplication factor of three described in section 4(a)(2) of 
this appendix B by the corresponding amount indicated in Section 
5(d)(4) Table, and shall use the adjusted multiplication factor when 
determining its market risk capital requirements until it obtains 
the next quarter's backtesting results, unless the OCC determines 
that a different adjustment or other action is appropriate:

Section 5(d)(4) Table.--Adjustment to Multiplication Factor From Results
              of Backtesting Based on 250 Trading Outcomes              
------------------------------------------------------------------------
                                                           Adjustment to
                    No. of exceptions                     multiplication
                                                              factor    
------------------------------------------------------------------------
4 or fewer..............................................         0.00   
5.......................................................         0.40   
6.......................................................         0.50   
7.......................................................         0.65   
8.......................................................         0.75   
9.......................................................         0.85   
10 or more..............................................         1.00   
------------------------------------------------------------------------

* * * * *
    Dated: February 26, 1996.
Eugene A. Ludwig,
Comptroller of the Currency.

Federal Reserve Board

12 CFR CHAPTER II

    For the reasons set forth in the preamble, parts 208 and 225 of 
title 12 of chapter II of the Code of Federal Regulations, as proposed 
to be amended at 60 FR 38082 (July 25, 1995) are further proposed to be 
amended as follows:

PART 208--MEMBERSHIP OF STATE BANKING INSTITUTIONS IN THE FEDERAL 
RESERVE SYSTEM (REGULATION H)

    1. The authority citation for part 208 continues to read as 
follows:

    Authority: 12 U.S.C. 36, 248(a), 248(c), 321-338a, 371d, 461, 
481-486, 601, 611, 1814, 1823(j), 1828(o), 1831o, 1831p-1, 3105, 
3310, 3331-3351, and 3906-3909; 15 U.S.C. 78b, 78l(b), 78l(g), 
78l(i), 78o-4(c)(5), 78q, 78q-1, and 78w; 31 U.S.C. 5318; 42 U.S.C. 
4012a, 4104a, 4104b, 4106, and 4128.

    2. In appendix E to part 208 as proposed to be added at 60 FR 
38103, section III.B. would be amended by revising paragraph 2.a. and 
adding a new paragraph 3 to read as follows:

Appendix E to Part 208--Capital Adequacy Guidelines for State Member 
Banks: Market Risk Measure

* * * * *

III. The Internal Models Approach

* * * * *
    B. * * *
    2. * * *
    a. A bank must have a risk control unit that is independent from 
its business trading units and reports directly to senior management 
of the bank. The unit must be responsible for designing and 
implementing the bank's risk management system and analyzing daily 
reports on the output of the bank's risk measurement model in the 
context of trading limits. The unit must conduct regular backtesting 
13 and adjust its multiplication factor, if appropriate, in 
accordance with section III.B.3. of this appendix E.
* * * * *
    c. * * *
    3. In addition to any backtesting the bank may conduct as part 
of its internal risk management system, the bank must conduct, for 
regulatory capital purposes, backtesting that meets the following 
criteria:
    a. The backtesting must be conducted quarterly, using the most 
recent 250 trading days' outcomes and VAR measures, which encompass 
approximately twelve months. The VAR measures must be calibrated to 
a one-day holding period and a 99 percent confidence level.
    b. The bank should identify the number of exceptions (that is, 
cases where the

[[Page 9118]]

magnitude of the daily trading loss, if any, exceeds the previous 
day's VAR measure) to determine its appropriate zone and level 
within a zone, as set forth in Table A of section III.B.3.c. of this 
appendix E.
    c. A bank should adjust its multiplication factor by the amount 
indicated in Table A of this paragraph c., unless the Federal 
Reserve determines that a different adjustment or other action is 
appropriate:

     Table A.--Adjustment to Multiplication Factor from Results of Backtesting Based on 250 Trading Outcomes    
----------------------------------------------------------------------------------------------------------------
                                                                                                      Cumulative
                                                                                      Adjustment to       1     
                      Zone                            Level (No. of exceptions)      multiplication  probability
                                                                                         factor          (in    
-------------------------------------------------------------------------------------------------------percent)-
Green Zone......................................  4 or fewer.......................         0.00         89.22  
                                                  5................................         0.40         95.88  
                                                  6................................         0.50         98.63  
Yellow Zone.....................................  7................................         0.65         99.60  
                                                  8................................         0.75         99.89  
                                                  9................................         0.85         99.97  
Red Zone........................................  10 or more.......................         1.00         99.99  
----------------------------------------------------------------------------------------------------------------
\1\ The zones are defined according to the cumulative probability of obtaining up to a given number of          
  exceptions in a sample of 250 independent observations when the true coverage level is 99 percent. The yellow 
  zone begins where cumulative probability equals or exceeds 95 percent, and the red zone begins where the      
  cumulative probability equals or exceeds 99.99 percent.                                                       

* * * * *

PART 225--BANK HOLDING COMPANIES AND CHANGE IN BANK CONTROL 
(REGULATION Y)

    1. The authority citation for part 225 continues to read as 
follows:

    Authority: 12 U.S.C. 1817(j)(13), 1818, 1828(o), 1831i, 1831p-1, 
1843(c)(8), 1844(b), 1972(1), 3106, 3108, 3310, 3331-3351, 3907, and 
3909.

    2. In appendix E to part 225 as proposed to be added at 60 FR 
38116, section III.B. would be amended by revising paragraph 2.a. and 
adding a new paragraph 3 to read as follows:

Appendix E to Part 225--Capital Adequacy Guidelines for Bank Holding 
Companies: Market Risk Measure

* * * * *

III. The Internal Models Approach

* * * * *
    B. * * *
    2. * * *
    a. A institution must have a risk control unit that is 
independent from its business trading units and reports directly to 
senior management of the bank holding company. The unit must be 
responsible for designing and implementing the institution's risk 
management system and analyzing daily reports on the output of the 
institution's risk measurement model in the context of trading 
limits. The unit must conduct regular backtesting 13 and adjust 
its multiplication factor, if appropriate, in accordance with 
section III.B.3. of this appendix E.
* * * * *
    c. * * *
    3. In addition to any backtesting the bank holding company may 
conduct as part of its internal risk management system, the bank 
holding company must conduct, for regulatory capital purposes, 
backtesting that meets the following criteria:
    a. The backtesting must be conducted quarterly, using the most 
recent 250 trading days' outcomes and VAR measures, which encompass 
approximately twelve months. The VAR measures must be calibrated to 
a one-day holding period and a 99 percent confidence level.
    b. The bank holding company should identify the number of 
exceptions (that is, cases where the magnitude of the daily trading 
loss, if any, exceeds the previous day's VAR measure) to determine 
its appropriate zone and level within a zone, as set forth in Table 
A of section III.B.3.c. of this appendix E.
    c. An institution should adjust its multiplication factor by the 
amount indicated in Table A of this paragraph c., unless the Federal 
Reserve determines that a different adjustment or other action is 
appropriate:

     Table A.--Adjustment to Multiplication Factor From Results of Backtesting Based on 250 Trading Outcomes    
----------------------------------------------------------------------------------------------------------------
                                                                                    Adjustment to    Cumulative 
                      Zone                           Level (No. of exceptions)     multiplication       \1\     
                                                                                       factor       probability 
----------------------------------------------------------------------------------------------------(in percent)
Green Zone.....................................  4 or fewer......................           0.00          89.22 
                                                 5...............................           0.40          95.88 
                                                 6...............................           0.50          98.63 
Yellow Zone....................................  7...............................           0.65          99.60 
                                                 8...............................           0.75          99.89 
                                                 9...............................           0.85          99.97 
Red Zone.......................................  10 or more......................           1.00         99.99  
----------------------------------------------------------------------------------------------------------------
\1\ The zones are defined according to the cumulative probability of obtaining up to a given number of          
  exceptions in a sample of 250 independent observations when the true coverage level is 99 percent. The yellow 
  zone begins where cumulative probability equals or exceeds 95 percent, and the red zone begins where the      
  cumulative probability equals or exceeds 99.99 percent.                                                       


[[Page 9119]]


* * * * *
    By order of the Board of Governors of the Federal Reserve 
System, February 9, 1996.
---------------------------------------------------------------------------

    \13\ Back-testing includes ex post comparisons of the risk 
measures generated by the model against the actual daily changes in 
portfolio value.
---------------------------------------------------------------------------

William W. Wiles,
Secetary of the Board.

Federal Deposit Insurance Corporation

12 CFR CHAPTER III

    For the reasons set forth in the preamble, part 325 of title 12 of 
chapter III of the Code of Federal Regulations, as proposed to be 
amended at 60 FR 38082 (July 25, 1995), is further proposed to be 
amended as follows:

PART 325--CAPITAL MAINTENANCE

    1. The authority citation for part 325 continues to read as 
follows:

    Authority: 12 U.S.C. 1815(a), 1815(b), 1816, 1818(a), 1818(b), 
1818(c), 1818(t), 1819(Tenth), 1828(c), 1828(d), 1828(i), 1828(n), 
1828(o), 1831o, 3907, 3909, 4808; Pub. L. 102-233, 105 Stat. 1761, 
1789, 1790 (12 U.S.C. 1831n note); Pub. L. 102-242, 105 Stat. 2236, 
2355, 2386 (12 U.S.C. 1828 note).

    2. In appendix C to part 325 as proposed to be added at 60 FR 
38129, section III.B.2. introductory text and section III.B.2.a. would 
be revised and section III.B.3. would be added to read as follows:

Appendix C to Part 325--Risk-Based Capital for State Non-Member Banks: 
Market Risk

* * * * *

III. The Internal Models Approach

* * * * *
    B. * * *
    1. * * *
    2. A bank must meet the following minimum qualitative criteria 
before using its internal model to measure its exposure to market 
risk.\13\
    a. A bank must have a risk control unit that is independent from 
its business trading units and reports directly to senior management 
of the bank. The unit must be responsible for designing and 
implementing the bank's risk management system and analyzing daily 
reports on the output of the bank's risk measurement model in the 
context of trading limits. The unit must conduct regular backtesting 
\14\ and adjust its multiplication factor, if appropriate, in 
accordance with section III.B.3. of this appendix C.
* * * * *
    3. In addition to any backtesting the bank may conduct as part 
of its internal risk management system, the bank must conduct, for 
regulatory capital purposes, backtesting that meets the following 
criteria:
    a. The backtesting must be conducted quarterly, using the most 
recent 250 trading days' outcomes and VAR measures, which encompass 
approximately twelve months. The VAR measures must be calibrated to 
a one-day holding period and a 99 percent confidence level.
    b. The bank should identify the number of exceptions (that is, 
cases where the magnitude of the daily trading loss, if any, exceeds 
the previous day's VAR measure) to determine its appropriate zone 
and level within a zone, as set forth in Table A of section 
III.B.3.c. of this appendix C.
    c. A bank should adjust its multiplication factor by the amount 
indicated in Table A, unless the FDIC determines that a different 
adjustment or other action is appropriate.

     Table A.--Adjustment to Multiplication Factor From Results of Backtesting Based on 250 Trading Outcomes    
----------------------------------------------------------------------------------------------------------------
                                                                                    Adjustment to  Cumulative\1\
                      Zone                           Level  No. of exceptions)     multiplication   probability 
                                                                                       factor       (in percent)
----------------------------------------------------------------------------------------------------------------
Green Zone.....................................  4 or fewer......................           0.00          89.22 
                                                 5...............................           0.40          95.88 
                                                 6...............................           0.50          98.63 
Yellow Zone....................................  7...............................           0.65          99.60 
                                                 8...............................           0.75          99.89 
                                                 9...............................           0.85          99.97 
Red Zone.......................................  10 or more......................           1.00         99.99  
----------------------------------------------------------------------------------------------------------------
\1\ The zones are defined according to the cumulative probability of obtaining up to a given number of          
  exceptions in a sample of 250 independent observations when the true coverage level is 99 percent. The yellow 
  zone begins where cumulative probability equals or exceeds 95 percent, and the red zone begins where the      
  cumulative probability equals or exceeds 99.99 percent.                                                       

* * * * *
    By order of the Board of Directors.
---------------------------------------------------------------------------

    \13\ If the FDIC is not satisfied with the extent to which a 
bank meets these criteria, the FDIC may adjust the multiplication 
factor used to calculate market risk capital requirements or 
otherwise increase capital requirements.
    \14\ Back-testing includes ex post comparisons of the risk 
measures generated by the model against the actual daily changes in 
portfolio value.

    Dated at Washington, D.C., this 27th day of February 1996.
Jerry L. Langley,
Executive Secretary.
[FR Doc. 96-5235 Filed 3-6-96; 8:45 am]
BILLING CODE 4810-33-P (\1/3\), 6210-01-P (\1/3\), 6714-01-P (\1/3\)