Joint Agency Policy Statement: Supervisory Policy Statement Concerning a Supervisory Framework for Measuring and Assessing Banks' Interest Rate Risk Exposure

Federal RegisterAug 2, 1995

Ask Donna

What actually matters in this document.

Text

[[Page 39495]]

DEPARTMENT OF THE TREASURY

Office of the Comptroller of the Currency

12 CFR Part 3

[Docket No. 95-17]

FEDERAL RESERVE SYSTEM

12 CFR Part 208

[Docket No. R-0802]

FEDERAL DEPOSIT INSURANCE CORPORATION

12 CFR Part 325

Joint Agency Policy Statement: Supervisory Policy Statement

Concerning a Supervisory Framework for Measuring and Assessing Banks'

Interest Rate Risk Exposure

AGENCIES: Office of the Comptroller of the Currency (OCC), Treasury;

Board of Governors of the Federal Reserve System (Board); and Federal

Deposit Insurance Corporation (FDIC).

ACTION: Policy statement; request for comment.

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

SUMMARY: The OCC, the Board, and the FDIC (collectively, ``the

agencies'') seek comment on a proposed interagency Supervisory Policy

to establish a uniform supervisory framework for measuring banks'

interest rate risk (IRR) exposures. The proposed policy establishes a

framework that the agencies would use to measure and monitor the level

of IRR at individual banks. The measurement process proposed and

described in this policy statement is intended to facilitate the

agencies' assessment of a bank's IRR exposure and its capital adequacy.

The results of the supervisory and internal models would be one factor

used by the agencies in their assessments' of a bank's capital adequacy

for IRR. Other factors that the agencies will consider include the

quality of the bank's IRR risk management process, the overall

financial condition of the bank, and the level of other risks at the

bank for which capital is needed. Pursuant to the final rule banks may

be required to hold additional capital.

The proposed supervisory framework provides measures of the change

in a bank's economic value for a given change in interest rates using a

supervisory model. The framework also considers the results of a bank's

internal model results when that model provides a measure of the change

in a bank's economic value. Banks not specifically exempted from

detailed IRR reporting would submit new IRR Call Report schedules

indicating the maturity, repricing, or price sensitivity of their

various on- and off-balance sheet instruments. A bank also would have

the option of reporting its internal model estimates of the price

sensitivity of its major portfolios and its economic value.

Concurrent with the publication of this proposed Supervisory Policy

statement, the agencies have issued a final rule that amends their

capital guidelines for IRR. Those amendments indicate that the agencies

will consider in their evaluation of a bank's capital adequacy, the

exposure of a bank's capital and economic value to changes in interest

rates. The amendments are in response to section 305 of the FDIC

Improvement Act of 1991 (FDICIA) which requires the agencies to amend

their risk-based capital standards to take adequate account of interest

rate risk.

As noted in the discussion of the final rule on IRR, the agencies

intend, at a subsequent date, to incorporate explicit minimum

requirements for IRR into their risk-based capital standards. The

agencies anticipate that the measurement framework described in this

proposed policy, will be the basis for such a capital requirement.

Toward that end, the agencies intend to work with the industry to

evaluate the reliability and accuracy of the results from the

supervisory model and bank internal models. Any explicit minimum

capital charge would be implemented through the agencies' rulemaking

process and would provide the opportunity for public comment before a

final rule is adopted.

DATES: Comments must be received by October 2, 1995.

ADDRESSES: Interested parties are invited to submit written comments to

any or all of the agencies. All comments will be shared among the

agencies.

OCC: Written comments should be submitted to Docket No. 95-17,

Communications Division, Ninth Floor, Office of the Comptroller of the

Currency, 250 E Street, S.W., Washington, D.C. 20219, Attention: Karen

Carter. Comments will be available for inspection and photocopying at

that address.

Board of Governors: Comments, which should refer to Docket No. R-

0802, may be mailed to Mr. William Wiles, Secretary, Board of Governors

of the Federal Reserve System, 20th and Constitution Avenue, N.W.,

Washington, D.C. 20551. Comments addressed to Mr. Wiles may also be

delivered to the Board's mail room between 8:45 a.m. and 5:15 p.m. and

to the security control room outside of those hours. Both the mail room

and control room are accessible from the courtyard entrance on 20th

Street between Constitution Avenue and C Street, N.W. Comments may be

inspected in Room B-1122 between 9:00 a.m. and 5:00 p.m., except as

provided in 261.8 of the Board's ``Rules Regarding Availability of

Information,'' 12 CFR 261.8.

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 5:00 p.m. [FAX

number (202) 898-3838; Internet address: comments @ fdic.gov]. Comments

will be available for inspection and photocopying in Room 7118, 550

17th Street, N.W., Washington, D.C. 20429, between 9:00 a.m. and 4:30

p.m. on business days.

FOR FURTHER INFORMATION CONTACT:

OCC: Christina Benson, Capital Markets Specialist, or Lisa

Lintecum, National Bank Examiner (202/874-5070), Office of the Chief

National Bank Examiner; Michael Carhill, Financial Economist, Risk

Analysis Division (202/874-5700); and Ronald Shimabukuro, Senior

Attorney, Bank Operations and Assets Division (202/874-4460), Office of

the Comptroller of the Currency, 250 E Street, S.W., Washington, D.C.

20219.

Board of Governors: James Houpt, Assistant Director (202/452-3358),

William F. Treacy, Supervisory Financial Analyst (202/452-3859),

Division of Banking Supervision and Regulation; Gregory Baer, Managing

Senior Counsel (202/452-3236), Legal Division, Board of Governors of

the Federal Reserve System. For the hearing impaired only,

Telecommunication Device for the Deaf (TDD), Dorothea Thompson (202/

452-3544), Board of Governors of the Federal Reserve System, 20th and C

Streets, N.W., Washington, D.C. 20551.

FDIC: William A. Stark, Assistant Director (202/898-6972) or

Phillip J. Bond, Senior Capital Markets Specialist (202/898-3519),

Division of Supervision, Federal Deposit Insurance Corporation, 550

17th Street, N.W., Washington, D.C. 20429.

SUPPLEMENTARY INFORMATION:

I. Introduction

Interest rate risk is the risk that changes in market interest

rates will have an adverse effect on a bank's

[[Page 39496]]

earnings and its underlying economic value. Changes in interest rates

affect a bank's reported earnings by changing its net interest income

and the level of other interest-sensitive income and operating

expenses. The underlying economic value of the bank's assets,

liabilities, and off-balance sheet instruments also is affected by

changes in interest rates. These changes occur because the present

value of future cash flows and in some cases, the cash flows

themselves, are affected when interest rates change. The combined

effects of the changes in these present values reflect the change in

the bank's underlying economic value.

Interest rate risk is inherent in the role of banks as financial

intermediaries. However, a bank that has an excessive level of interest

rate risk can face diminished future earnings, impaired liquidity and

capital positions, and, ultimately, may jeopardize its solvency.

The agencies believe that safety and soundness requires effective

management and measurement of interest rate risk, and each agency has

provided supervisory guidance to banks and examiners on this subject.

In addition, the agencies believe that a bank's capital adequacy should

be assessed in the context of the risks it faces, including interest

rate risk. Section 305 of FDICIA Pub. L. 102-242 (12 U.S.C. 1828 note),

on which a final rule is being issued at the same time as this

statement, specifically requires the agencies to take account of

interest rate risk in assessing capital adequacy. Both of these aspects

of interest rate risk depend on, among other things, a meaningful

measurement of the bank's risk exposure.

The agencies believe that a bank should have an IRR measurement

system that is commensurate with the nature and scope of its IRR

exposures. Among the difficulties in performing a supervisory

evaluation of interest rate risk, however, is that measurement systems

and management philosophies can differ significantly from one bank to

another. As a result, although two banks may each be well-managed,

their measured exposure may not be directly comparable. This difficulty

has been magnified by the rapid pace of change in financial markets and

instruments themselves.

In implementing Section 305 of FDICIA, and in light of the rapid

evolution in financial instruments and practices, the agencies believe

there is a need for a more formal supervisory assessment of banks'

interest rate risk exposures. To support that effort, the agencies

propose a measurement framework that includes a supervisory measurement

system (``supervisory model'') that will, on a standardized basis,

measure the risk of all banks not exempted from reporting additional

information on their IRR exposures. In addition, banks will be

encouraged to report, through a voluntary and confidential supplemental

Call Report schedule, the results of their internal IRR measurement

systems. These measured results would then serve as an additional

source of information for an examiner's assessment of the bank's risk

management and capital adequacy. The results also would provide

information on industry trends and patterns that will better inform

both present and future supervisory efforts related to interest rate

risk.

The measurement framework described in this policy statement

focuses on the exposure to a bank's underlying economic value from

movements in market interest rates. The exposure to a bank's economic

value, as used in this policy statement, is defined as the change in

the present value of its assets, minus the change in the present value

of its liabilities, plus the change in the present value of its off-

balance sheet interest-rate positions. The agencies haven chosen this

focus because they believe that changes in a bank's economic value best

reflect the potential impact of embedded options and the potential

exposure that the bank's current business activities pose to the bank's

future earnings stream, and hence, its ability to sustain adequate

capital levels. Changes in economic value measure the effect that a

change in interest rates will have on the value of all of the future

cash flows generated by a bank's current financial positions, not just

those cash flows which affect earnings over the few months or quarters.

Thus, changes in economic value provide a more comprehensive measure of

risk than measures which focus solely on the exposure to a bank's near-

term earnings. It is for this reason that the agencies have amended

their capital standards to identify explicitly a bank's exposure to

declines in economic value from changes in interest rates as an

important factor to consider in evaluating a bank's capital adequacy.

II. Summary of Approach

In assessing the sensitivity of a bank's economic value to changes

in interest rates, the agencies are proposing to use the results of a

supervisory model and, for those electing to provide such analysis, the

results of banks' own internal models. These assessments will rely on

data reported in regulatory Call Reports. Recognizing that the burden

for reporting IRR exposures would fall most heavily on smaller

organizations with limited resources, the policy statement makes

provisions for smaller, well-managed institutions that are less likely

to be significantly exposed to IRR to be exempt from additional

reporting. As described in further detail in the policy statement, the

agencies propose that banks with (i) assets under $300 million, (ii)

composite supervisory CAMEL ratings of 1 or 2 and, (iii) moderate or

low holdings of assets with intermediate and long term maturity or

repricing characteristics, be exempted from expanded reporting

requirements for IRR.

Banks that are not specifically exempted by the proposed policy

statement will submit additional Call Report information on the

repricing and maturity of their portfolios. The proposed supervisory

model applies a series of IRR risk-weights to a bank's reported

repricing and maturity balances. These weights estimate the price

sensitivity of a bank's reported balances to a 200 basis point increase

and decrease in interest rates. The summation of these balances, along

with certain price sensitivity information that a bank may be required

to self-report, results in a net risk-weighted exposure for the bank.

That exposure represents the estimated change in the bank's economic

value to the specified rate change.

The proposed supervisory model represents a refinement of the model

presented in the September 1993 notice of proposed rulemaking

(September NPR) [58 FR 48206, September 14, 1993]. The September NPR

solicited comments on a framework for measuring banks' exposure to IRR

for capital purposes pursuant to Section 305 of FDICIA. The final rule

for Section 305 does not incorporate an explicit measurement framework

for IRR into the agencies' risk-based capital standards. The agencies

have concluded that it is appropriate to first collect industry data

and evaluate the performance of the measurement framework before

explicitly incorporating the results of that framework into their risk-

based capital standards. The data collected by the agencies will assist

current supervisory efforts and will facilitate the development of a

measurement framework that could be explicitly incorporated into

capital standards in the future. This proposed policy statement would

implement that supervisory measurement framework. The proposed

framework is broadly consistent with the one discussed in the September

NPR. The agencies, however, have made several refinements to the

[[Page 39497]]

supervisory model to improve its accuracy while still endeavoring to

limit the burden of the expanded reporting and maintain model

transparency. The refinements to the September NPR model include:

(1) Separate risk-weights and reporting for residential

adjustable-rate mortgages;

(2) Separate risk-weights and reporting for residential fixed-

rate mortgages and all other amortizing assets;

(3) Self-reporting by banks of price sensitivities of

instruments with complex and/or non-standardized cash flow

characteristics such as structured notes, collateralized mortgage

obligations (CMOs), and mortgage servicing rights;

(4) Supplemental reporting for banks with concentrations in

adjustable- and fixed-rate mortgage loans.

(5) Greater flexibility in reporting deposits without stated

maturity or repricing dates;

(6) Separate reporting and treatment in the baseline schedule

for residential mortgage loans which are held by the bank for sale

and delivery to a secondary market participant under terms of a

binding commitment.

A summary of the public comments and agency analysis that led to

these refinements are included in section IV of this document and the

refinements themselves are described in detail in the policy statement

and accompanying reporting instructions.

For a bank choosing also to report the results of its internal IRR

model, the agencies are proposing to collect the dollar change in value

of the bank's major portfolios and the net change in the bank's

economic value using the same rate scenario incorporated in the

supervisory model. To the extent specific details concerning a bank's

financial instruments are incorporated in an internal model with

adequate integrity and reasonable assumptions, those results should

provide the agencies with an improved understanding of a bank's IRR

profile. For a bank reporting internal model results, an examiner would

have the benefit of weighing the results of both measures in assessing

a bank's overall IRR exposure for capital adequacy purposes. Moreover,

comparisons between the results of the supervisory model and internal

models are expected to aid the agencies in determining what, if any,

refinements should be made to the proposed measurement framework before

incorporating it into a minimum capital charge for IRR.

III. CDFI Section 335 Considerations

On September 23, 1994 the Reigle Community Development and

Regulatory Improvement Act of 1994 (``CDFI'') (Pub. L. 103-325) was

enacted. Section 335 of CDFI amended section 305 of FDICIA by

instructing the agencies to be sure that steps taken to implement

Section 305 ``take into account the size and activities of the

institutions and do not cause undue reporting burdens.'' The agencies

believe that the Congressional mandate to avoid undue reporting burdens

is also applicable and desirable for purposes of implementing the

proposed policy statement. Consequently, as already noted, the agencies

have formulated a reporting exemption test that takes into account the

size and activities of an institution. In addition, the reporting

requirements for the supervisory model also considers the nature and

scope of a bank's activities. Banks holding certain types of financial

instruments that often have complex or nonstandardized cash flow

characteristics will be expected to have the ability to calculate on

their own, or obtain from reliable sources, estimates of those

instruments' market value sensitivity. Banks with holdings of fixed-

and adjustable-rate residential mortgage loans and securities that

exceed certain levels would be required to report additional

information on those portfolios to better assess the embedded option

risks associated with those products.

IV. September 1993 Notice of Proposed Rulemaking

A. Description of September NPR

In September 1993, the Banking Agencies issued a notice of proposed

rulemaking (September NPR) [58 FR 48206, September 14, 1993] that

solicited comments on a framework for measuring a bank's IRR exposure

and determining the amount of capital the bank needed for IRR.

The framework outlined in the September NPR incorporated the use of

a three-level measurement process to evaluate banks' IRR exposures. The

first measure was a quantitative screen, based on existing Call Report

information, that exempted potential low risk banks from additional

reporting requirements. The exemption screen used two criteria: (1) The

amount of a bank's off-balance-sheet interest rate contracts in

relation to its total assets; and (2) the relation between a bank's

fixed- and floating-rate loans and securities that mature or reprice

beyond five years and its total capital.

Banks not meeting the proposed exemption test were required to

calculate their economic exposure by either: (1) A supervisory model

that measured the change in the economic value of bank for a specified

change in interest rates; or (2) the bank's own IRR model, provided

that the model was deemed adequate by examiners for the nature and

scope of the bank's activities and that it measured the bank's economic

exposure using the interest rate scenarios specified by the agencies.

B. Comments on the September NPR Measurement Framework

The agencies collectively received a total of 133 comments on the

September NPR. The majority of commenters were banks. Thrift, trade

associations, bank consultants, and other government-sponsored agencies

and regulators also commented. The majority of commenters responded

favorably to modifications that the agencies made from an earlier,

advanced notice of proposed rulemaking (ANPR) [57 FR 35507, August 10,

1992]. In particular, most commenters expressed strong support for

using the results of a bank's own IRR model to determine its level of

exposure and corresponding need for capital. Commenters noted the

potential inaccuracies of standardized regulatory models as one reason

for allowing the use of internal models. Internal models, they

believed, would better capture the unique characteristics of individual

bank portfolios. Many commenters also stated that permitting the use of

internal models would provide banks with incentives to improve their

internal risk measurement systems.

Many commenters raised concerns about various elements of the

measurement framework outlined in the September NPR. Most commenters

believed that the proposed treatment of non-maturity deposits

understated their effective maturity. Others questioned the accuracy of

the proposed supervisory model and the appropriateness of the proposed

exemption test criteria.

C. Agencies' Responses to Comments

The agencies have carefully considered the concerns raised by

commenters regarding the structure and elements of the proposed

measurement framework and the accuracy of the proposed supervisory

model. Although the agencies have decided to retain many of the

principles and structures outlined in the September NPR framework, the

agencies are also proposing several modifications and refinements to

that framework. These modifications include changes to the proposed

exemption criteria, the structure of the supervisory model, and the

treatment of certain types of assets and non-maturity deposits. These

modifications are discussed in greater detail in the sections that

follow.

[[Page 39498]]

1. Exemption Criteria

The September NPR included criteria that would exempt a bank from

additional measurement and reporting requirements. The proposal set

forth the following two criteria that a bank would have to meet to

qualify for an exemption:

(1) The total notional principal amount of all of the bank's off-

balance-sheet interest rate contracts must not exceed 10 percent of its

assets; and

(2) 15 percent of the sum of the bank's fixed- and floating-rate

loans and securities that mature or reprice beyond 5 years must be less

than 30 percent of its total capital.

There was general support among commenters for some type of

exemption. The majority of commenters addressing this issue, however,

voiced concerns with the proposed test. Many commenters believed that a

10 percent threshold for off-balance sheet contracts would discourage

the use of such instruments in managing and reducing IRR exposures.

Commenters also expressed concerns that the maturity test, incorporated

in the second criterion, used contractual maturities rather than

expected average lives and would overstate the risk associated with

amortizing loans and securities, such as mortgage-related products.

Several commenters suggested modifying the criterion to use bank

management's estimates of average lives, rather than contractual

maturities.

Several commenters questioned whether the proposed exemption

criteria provided sufficient safeguards against exempting banks that

may pose significant risks to the Bank Insurance Fund due to their

potential IRR exposures. A few commenters noted the potential for

material intermediate-term maturity (e.g., 1- to 5-years) mismatches. A

minority of commenters question the need for, or efficacy of, any

exemption test.

The agencies continue to believe that an exemption is desirable and

that section 335 of CDFI Bill reinforces the need to consider ways of

minimizing burdens associated with this policy statement. The agencies

also believe that there is a need to ensure sufficient safeguards

against exempting banks that may pose significant systemic risks or

costs to the Bank Insurance Fund. Consequently, the agencies propose to

modify the exemption test to focus on three considerations: the size of

the bank; the quality of its overall condition and management, as

measured by its composite CAMEL rating; and the level of its potential

repricing exposure as measured by its intermediate and longer-term

assets. Specifically, to be exempted, a bank would have to meet the all

of the following three conditions:

(1) The bank must have total assets of less than $300 million; and

(2) Have a ``1'' or ``2'' composite CAMEL 1 rating from its

primary supervisor; and

\1\ CAMEL refers to the Uniform Financial Institution's Rating

System that the agencies have adopted. Each bank is assigned a

uniform composite rating based on an evaluation of pertinent

financial and operational standards, criteria and principles. This

overall rating is expressed through use of a numerical scale of

``1'' through ``5'' with ``1'' indicating the highest rating and

``5'' the lowest. The composite rating assess five key performance

dimensions that are commonly identified by the acronym ``CAMEL'':

Capital adequacy, Asset quality, Management, Earnings and Liquidity.

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

(3) The sum of:

(a) 30 percent of its loans and securities with contractual

maturity or repricing dates between one and five years, and

(b) 100 percent of its loans and securities with contractual

maturity or repricing dates beyond five years must be less than 30

percent of the bank's total assets.

Banks that meet this proposed exemption test could elect to submit the

proposed IRR Call Report schedules on a voluntary basis. The agencies

encourage such voluntary reporting.

The exemption test does not alleviate the need for an exempted bank

to employ sound IRR measurement and management practices and to have

sufficient capital for its risk exposure. Exempted banks will continue

to be subject to safety and soundness IRR examinations that the

agencies may conduct. As a result of such examinations, a bank that is

exempt from this policy statement may be directed by their primary

supervisor to improve its IRR measurement and management practices, or

to hold additional capital for IRR. In addition, the agencies would

retain the right to require any bank to comply with the provisions of

this policy statement and any subsequent rulemakings regarding IRR.

2. Interest Rate Scenarios

The September NPR outlined a number of factors that should be

considered in selecting an appropriate interest scenario for measuring

banks' IRR exposures and evaluating capital adequacy. These factors

included:

(1) The time horizon over which banks and supervisors could

reasonably be expected to identify risk and implement mitigating

responses;

(2) The likelihood of occurrence, as reflected by historical rate

volatility; and

(3) The appropriate historical sample period used to determine the

likelihood of a given rate movement.

The agencies sought comment on several alternative methodologies

for developing appropriate interest rate scenarios, including both

parallel and non-parallel changes in interest rates. Among the non-

parallel methods, the interest rate scenario could be based upon

observed nominal changes in interest rates, or upon observed

proportional changes in interest rates. As an alternative, the agencies

also sought comment on using a simple parallel shift in interest rates

across the entire maturity spectrum (``parallel rate shocks'').

The agencies received a range of comments on the selection and

determination of the appropriate interest rate scenarios. Commenters

were divided on whether a short or long historical sample was most

appropriate for determining the potential range of interest rate

movements. Those favoring a shorter sample period believed such a

period best reflected current and likely probabilities of rate changes.

Others favored a longer sample period, primarily to minimize the impact

of any one rate cycle. Opinions were also divided on whether a monthly,

quarterly, or annual time horizon was most appropriate for analyzing

potential rate scenarios. The majority of commenters favored either a

monthly or quarterly horizon, on the grounds that such time frames

represented the time bank management would need to implement risk

mitigating actions in response to an adverse movement in interest

rates. Others, however, disagreed and favored the use of an annual time

horizon.

Commenters also expressed diverse views on whether the proposed

rate scenarios should be based on nominal or proportional changes in

historical rates, or on the basis of a simple parallel rate shock. A

majority of commenters argued against the use of parallel rate shocks,

on the grounds that such scenarios were not realistic of probable

future interest rate changes. Of these commenters, most favored

scenarios that would be based on proportional rate changes, such that

the size of the rate change used to measure exposures would depend

upon, and vary with, the current level of market interest rates. Other

commenters, however, favored the use of parallel rate shocks, primarily

on the grounds of simplicity and ease of understanding.

The agencies propose to use a simple 200 basis point, instantaneous

parallel upward and downward shift in interest rates for measuring and

evaluating

[[Page 39499]]

banks' exposures for purposes of assessing capital adequacy. The

agencies believe that such rate movements are realistically

conservative given the movements in interest rates experienced in 1994.

They also believe that such rate scenarios are sufficiently transparent

and easy to understand that they can be easily incorporated into either

a bank's own IRR model or the supervisory model. The scenarios are

incorporated into the proposed supervisory model via the proposed risk-

weights that are applied to a bank's reported maturity and repricing

balances.

The agencies stress that their adoption of these rate scenarios

does not replace the need for a bank to evaluate its IRR exposure over

a wider range of possible rate changes for its own risk management

purposes. Such rate changes may include non-parallel yield curve shifts

and gradual, as well as immediate, rate changes. To ensure greater

consistency, however, in the agencies' assessments of banks' exposures

and their need for capital, banks are encouraged to include the

proposed instantaneous and parallel 200 basis point rate scenarios into

their internal IRR measurement processes.

3. Structure of Supervisory Model

The supervisory model in the September NPR grouped assets,

liabilities, and off-balance-sheet positions by various categories,

based on their general cash flow and product characteristics. Each

category and time band was assigned risk-weights corresponding to a

rising rate scenario and a declining-rate scenario. The risk-weights

were constructed by the agencies, using hypothetical market instruments

that were representative of the category being measured. For amortizing

instruments, the risk-weights incorporated assumptions about

prepayments.

A number of commenters expressed concerns regarding the accuracy of

the model proposed in the September NPR. Frequently cited concerns

included: the use of hypothetical, rather than bank-specific,

instruments to derive risk weights; the level of data aggregation; the

use of standardized prepayment assumptions; and the treatment of

interest rate protection agreements (caps and floors). A number of

commenters voiced concerns about the treatment of residential mortgage-

related products. In general, these commenters believed that additional

detail on mortgage holdings, such as coupon information on fixed- rate

mortgages, and more explicit information on periodic and lifetime

interest caps for adjustable-rate products, would improve the model's

accuracy.

The agencies sought comment in the September NPR on whether

commercial banks with portfolios that are similar to thrift should be

required to use the Net Portfolio Value model used by the Office of

Thrift Supervision (OTS) for federally-supervised thrift institutions.

Most commenters believed that such a requirement would impose

substantially greater reporting burdens without necessarily improving

the accuracy of the measure and might create incentives for banks to

substitute such a model for the judgment of bank management. A minority

of commenters disagreed and stated that the approach and data used by

the OTS were superior and more accurate than what the banking agencies

had proposed.

The agencies have carefully considered commenters' concerns about

the proposed supervisory model's accuracy. The agencies believe it is

critical to have a supervisory model that can identify banks with

significant IRR exposures. They also are attentive to the risk that

model measurement errors could lead to undesirable incentives or

incorrect assessments regarding the risk and complexity of products,

activities, or banks. At the same time, the agencies recognize the need

to balance the desire for increased accuracy against the potential

costs of greater reporting detail and model complexity. The agencies

are particularly concerned that the supervisory model retain sufficient

transparency so that bankers can understand its methodology and

anticipate and compute their bank's measured exposure and that it not

replace the role or need for sound internal interest rate risk

management systems.

The agencies intend to make five modifications to the structure of

the supervisory model to improve its accuracy and which are described

below. The first four changes modify the basic supervisory model

outlined in the September NPR. This revised basic model will be the

baseline model for non-exempted banks. The last modification creates

supplemental modules for banks that have concentrations in residential

mortgage-related instruments. The agencies are mindful that the

supplemental schedules will impose additional reporting requirements

for some banks. Nonetheless, the agencies are concerned that the

baseline model may not be sufficiently accurate to capture the risk at

banks with significant holdings of mortgage loans or mortgage pass-

through securities, and therefore propose to require additional

reporting for those banks. A detailed description of the model, the

risk weights, and information requirements are discussed in the policy

statement. Schedule 1, provided in the attached policy statement,

illustrates the type of information that will be used in the baseline

supervisory model, while Schedules 2-4 illustrate the information used

for the supplemental modules.

a. Adjustable-rate residential mortgages. The first modification

that the agencies have made is to treat adjustable-rate residential

mortgage loans and securities (ARMs) separately from fixed-rate

residential mortgage assets. As modified, information on ARMs will be

reported by a bank on the basis of the reset frequency of the ARM's

pricing index, rather than by the ARM's next date to repricing. In

addition, a bank will report ARMs that are currently within 200 basis

points of their lifetime cap separately from those ARMs that are

further away from their lifetime caps. The agencies believe that this

stratification of ARM products will provide a better reflection of

their potential price sensitivity to changes in market interest rates

than the treatment described in the September NPR.

b. Fixed-rate residential mortgages and other amortizing assets.

The second modification the agencies made is to treat fixed-rate

residential mortgage assets separately from other amortizing assets. In

the September NPR, these assets had been combined into a single

category. As a result of this combination, the same prepayment

assumptions were applied to all amortizing assests. By separating these

two categories, the agencies propose to apply different prepayment

assumptions to the two categories.

c. Self-reporting of market value sensitivities. The third

modification will require a bank that holds certain types of financial

instruments to provide in its Call Report submissions, estimates of

changes in market value sensitivities of those instruments for the

specified 200 basis point interest rate scenarios. These estimates may

be obtained from the bank's own internal risk measurement systems or

from reliable third-party sources, provided that the bank knows,

understands, and documents the assumptions underlying those estimates.

All estimates and supporting documentation will be subject to examiner

review. The September NPR used this approach for certain mortgage

derivatives securities. The agencies propose to extend this treatment

to other products. The products for which banks would be required to

self-report market value sensitivities generally have complex options

or cash flow

[[Page 39500]]

characteristics. These characteristics make it difficult to adequately

measure these products in a standardized model without collecting

detailed transaction-oriented data.

Self-reporting of market value sensitivities generally would be

required for the following products or portfolios:

(1) All mortgage-backed derivative securities that meet the

FFIEC's definition of ``high-risk.'' 2

\2\ Effective February 10, 1992, the agencies and the Office of

Thrift Supervision adopted revised supervisory policies on

securities activities that were developed under the auspices of the

FFIEC. The revised policies established a framework for identifying

``high-risk mortgage derivative products.''

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

(2) All structured notes, as defined in the Call Report

instructions;

(3) Non-high-risk mortgage derivative securities when those

holdings represent 10 percent or more of a bank's assets.

(4) Mortgage servicing rights that are capitalized and reported

on the bank's balance sheet;

(5) Off-balance-sheet interest rate options, caps, and floors,

including interest rate swaps with embedded option characteristics.

The agencies believe that given the potential price sensitivity of

these products or portfolios to interest rate changes, it is reasonable

to expect banks to be able to calculate or obtain reliable estimates of

their market value sensitivities. Industry comments on the availability

of such information are especially welcomed.

d. Trading account portfolios. The agencies also propose to change

the manner in which trading account positions are treated in the

supervisory model. These changes are in response to commenters concerns

regarding the burden associated with distributing trading positions

into the maturity ladder and applying a 200 basis point rate shock to

those positions.

As modified, banks will be asked to self report the change in the

economic value of all of their trading account positions for a 100

basis point parallel increase or decrease in interest rates. This rate

change, smaller than the 200 basis point change used for the rest of

the bank's holdings, reflects the shorter holding period typical for

trading account positions. It also is similar to the 100 basis point

scenario used by the Basle Committee on Banking Supervision (Basle

Committee) in its April 1995 proposal on capital requirements for the

market risks of traded debt securities.\3\

\3\ The Basle Committee on Banking Supervision is a committee of

banking supervisory authorities which was established by the

central-bank Governors of the Group of Ten countries in 1975.

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

The agencies believe the self-reporting treatment for trading

accounts is consistent with supervisory guidance issued by each of the

agencies that directs banks with significant trading activities to have

internal risk measurement and limit systems commensurate with the size

and complexity of their activities.

As previously noted, the Basle Committee has recently released for

comment a proposal to incorporate the market risks of trading

activities into the Basle Accord risk-based capital standards.\4\ The

agencies published in the Federal Register on July 25, 1995 (60 FR

38082) a notice of proposed rulemaking on the Basle market risk

proposal. If the agencies adopt a final rule to implement the Basle

market risk proposal for banks with a large concentration of trading

activities, the agencies anticipate that modifications to this policy

statement will be required to ensure that IRR exposures arising from

those activities are not ``double-counted.'' One approach that the

agencies are considering is to exclude trading activities from this

proposed policy statement and IRR measure for those banks that are

subject to the market risk proposal. If such an approach is adopted,

those banks would be exempted from having to report the changes in the

market value of their trading portfolios for the IRR measure. If,

however, a bank's trading portfolio offsets the exposure from other

components of the bank's balance sheet, this treatment would overstate

the bank's total IRR exposure.

\4\ The Committee's proposal is described in a consultative

paper, entitled ``Planned Supplement to the Capital Accord to

Incorporate Market Risks,'' issued in Basle, Switzerland on April

12, 1995. Copies of that paper may be obtained by contacting: The

OCC's Communications Division, Ninth Floor, Office of the

Comptroller of the Currency, 250 E Street, S.W., Washington, D.C.

20219. A copy of the paper also is available at the FDIC Reading

Room, 550 North 17th Street, NW, Washington, D.C.

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

e. Supplemental modules. The final modification made by the

agencies to the supervisory model structure is the development of

supplemental modules for fixed-rate and adjustable-rate residential

mortgage loans and pass-through securities. A bank whose holdings of

these products exceeds certain threshold levels will be required to

report additional information on those holdings in their Call Report

submissions. The agencies will apply expanded tables of risk-weights to

those portfolios. The supplemental module for fixed-rate residential

mortgages requires a bank to stratify its balances into eight coupon

ranges. The agencies have developed separate risk-weights for each

coupon range which reflect the differences in expected prepayment

speeds that are associated with the underlying coupon rates. To develop

these risk-weights, the agencies have used the September 30, 1994

pricing tables generated by the Office of Thrift Supervision's Net

Portfolio Value Model.\5\ The agencies will apply this supplemental

module and associated risk-weights when a bank's holdings of fixed-rate

residential mortgage loans and pass-through securities represent 20

percent or more of its total assets. Schedule 2 in the attached policy

statement illustrates the information that will be used in the

supplemental module for fixed-rate residential mortgages. This expanded

module will be optional for a bank whose holdings of these instruments

are less than 20 percent of its assets.

\5\ Appendix 4 of the policy statement provides a description of

the derivation of the risk-weights for the baseline supervisory

model and supplemental modules.

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

Two levels of supplemental modules have been developed by the

agencies for adjustable-rate residential mortgages. The first level,

illustrated by Schedule 3 in the attached policy statement, requires

information on ARMs to be stratified by reset frequency (as in the

baseline model), periodic caps, and the ARMs' distances from lifetime

caps. This module will be used by the agencies when a bank's ARM

holdings are greater than 10 but less than 25 percent of its assets.

The second level, illustrated by Schedule 4 in the attached policy

statement, requires that ARM balances be further stratified by the

underlying rate index of the ARM. This module will apply to banks whose

holdings equal or exceed 25 percent of their total assets. The agencies

have developed risk-weights that correspond with each various reset

frequency, lifetime cap, periodic cap, and, index combination, again

using pricing tables generated from the OTS Net Portfolio Value Model.

The agencies are mindful that many commenters to the September NPR

raised concerns about tradeoffs between attempts to improve the

supervisory model accuracy and associated reporting burdens, especially

with regards to the use of the OTS model. Nonetheless, the agencies

believe the distribution of coupons for fixed-rate mortgage portfolios

and the interaction of the parameters illustrated in Schedules 3 and 4

significantly affect the price sensitivity of mortgage loans and

securities. The agencies believe that by explicitly considering these

parameters, the supplemental modules will enhance the accuracy of the

supervisory model. The agencies believe that this increased accuracy is

[[Page 39501]]

warranted due to the increased holdings of mortgage products among

commercial and savings banks. They also note the flexibility that many

banks exercise in their ability to tailor the various pricing

combinations of their ARM products. As banks expand their activities in

these products, the agencies are particularly concerned that banks not

ignore the potential impact and interaction of these pricing

parameters.

Draft instructions for completing the supplemental modules and a

technical description of the risk-weights used in the modules are

provided in the appendices 2 and 4 to the proposed policy statement.

4. Non-maturity deposit assumptions. The September NPR established

limits on the maximum maturities that a bank could attribute to its

non-maturity deposits when measuring its IRR exposures for capital

adequacy. Non-maturity deposits were defined to be those instruments

without a specific maturity or repricing date and included demand

deposits (DDA), negotiable order of withdrawal (NOW), savings, and

money market deposit (MMDA) accounts. In the September NPR, banks were

subject to the following constraints in distributing these deposits

across time bands:

(1) A bank could distribute its DDA and MMDA accounts across any

of the first three time bands, with a maximum of 40 percent of those

balances in the 1 to 3 year time band;

(2) A bank could distribute its savings and NOW account balances

across any of the first four time bands, with a maximum of 40

percent of the total of those balances in the 3 to 5 year time band.

The treatment of non-maturity deposits was one of the most

commented upon aspects of the September NPR. Most commenters stated

that the proposed treatment could, in many cases, understate the

effective maturity of these deposits and urged the agencies to adopt a

more flexible approach or extend the permissible maturities. Commenters

expressed concern that the adoption of the proposed rules could lead to

incorrect assessments of risk exposures or inappropriate incentives to

shorten asset maturities.

The agencies recognize that the treatment of non-maturity deposits

will be, for many banks, the single most important assumption in

measuring their IRR exposures. The agencies also agree that many banks

historically have been able to exercise considerable flexibility in the

timing and magnitude of pricing changes for these accounts. It is for

this reason that the agencies had proposed to allow banks some

flexibility in the treatment of these deposits. Nonetheless, the

agencies believe that there are risks associated with assuming that a

bank has sufficient flexibility in its management of these deposits so

as to offset any IRR position it may have. While these deposits can, in

many circumstances, help to mitigate a bank's IRR exposure, historical

experience suggests that an institution can incur significant levels of

IRR though it may have sizeable holdings of non-maturity deposits. The

agencies also are concerned that increased competitive pressures and

changing customer demographics may, over time, make these deposits more

rate sensitive or prone to migration into other investment vehicles.

Given these considerations, the agencies believe it is appropriate

to extend, but not eliminate, the maximum permissible maturities for

non-maturity deposits. Within these maturity ranges, a bank would have

the flexibility to distribute its balances based on its own assumptions

and experience. The agencies will expect that bank management will be

able to document to examiners the rationale for the treatment they have

chosen.

In addition to extending permissible maturities, the agencies

believe that demand deposit balances held by businesses should be

treated differently than demand balances held by other entities. In

particular, the agencies believe that a shorter maturity is appropriate

for commercial demand deposit accounts since many of these accounts are

in the form of compensating balances.\6\ The implicit earnings from

these compensating balances are often used to offset service charges

incurred by the customer, and the level of these implicit earnings

attributed to the deposits is generally dependent upon the level of

current market rates. As such, these balances behave very much like

interest-sensitive balances. As market rates increase, the level of

balances drops due to a higher earnings credit, while as rates decline,

the level of balances will generally increase.

\6\ For purposes of this policy statement, the term

``commercial'' is used to mean ``nonpersonal'' as that term is

defined under the Board of Governor of the Federal Reserve System's

Regulation D dealing with reserve requirements.

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

The agencies propose to extend the range of permissible maturities

for non-maturity deposits by revising the distribution rules for those

deposits. As proposed, a bank may distribute its deposits across time

bands according to its individual assumptions and experience, subject

to the following constraints:

(1) Commercial Demand Deposits: A bank would report 50 percent

of it's commercial demand deposits in the 0-3 month time band. The

remaining balances may be distributed across the first four time

bands, with a maximum of 20 percent of total balances in the 3-5

year time band.

(2) Retail DDA, Savings, and NOW Accounts: A bank may distribute

the balances in these accounts across any of the first five time

bands, with a maximum of 20 percent in the 5-10 year time band and

no more than 40 percent combined in the 3-5 and 5-10 year bands.

(3) MMDA Accounts: A bank may distribute the balances in these

accounts across any of the first three time bands, with a maximum of

50 percent in the 1-3 year band.

Table A summarizes the distribution that would result if a bank

reported its balances so as to maximize its allowable maturities.

Table A.--Maturity Distribution Limits for Non-Maturity Deposits

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

0-3 months 3-12 months 1-3 years 3-5 years 5-10 years

(percent) (percent) (percent) (percent) (percent)

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

Commercial DDA................................. 50 0 30 20 ...........

Retail DDA..................................... 0 0 60 20 20

MMDA........................................... 0 50 50 ........... ...........

Savings........................................ 0 0 60 20 20

NOW............................................ 0 0 60 20 20

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

[[Page 39502]]

The agencies believe that these maturity limits provide appropriate

guidelines for the purpose of standardized IRR measurement across the

banking industry. These limits are not intended to replace the need for

banks to evaluate and consider the sensitivity of their individual

deposit bases when managing their IRR exposures. Examiners will

consider a bank's assessment of its deposit base and how those

assessments may differ from those used in the standardized supervisory

model during the examination process when evaluating a bank's capital

adequacy for IRR. The agencies do not propose to require banks to

incorporate these assumptions into their internal IRR models when

submitting internal model results to the agencies. Rather, through the

examination process, examiners will consider whether the treatment used

in the bank's model is appropriate, based on the analysis the bank

provides.

5. Use of a Bank's Internal IRR Model

The September NPR permitted a bank to use the results of its

internal IRR model, as an alternative to the supervisory model, when

assessing its need for capital for IRR, provided that its model was

deemed adequate by the appropriate supervisor. Most commenters

expressed strong support for using the results of a bank's internal

model and believed that such a model would provide a more accurate

assessment of risk than the proposed supervisory model.

The proposed policy statement provides for the consideration of a

bank's internal model results in the assessment of that bank's level of

IRR exposure and its need for capital. The results and quality of a

bank's IRR measurement process will be one factor that examiners will

consider in assessing a bank's need for capital. Among the factors that

an examiner will consider when evaluating the quality of a bank's

internal model is whether the risk profile it generates is an adequate

measure of the bank's risk position, taking account of the types of

instruments held or offered by the bank, the integrity and completeness

of the data used in the model, and whether the assumptions and

relationships underlying the model are reasonable. When assessing the

exposure of a bank's economic value to changes in interest rates,

examiners generally will place greater reliance on the results of a

bank's internal model, rather than the supervisory model, provided that

the bank's own model:

(1) Measures IRR from an economic perspective, as defined in this

proposal;

(2) Uses the proposed supervisory scenario of an instantaneous and

parallel 200 basis point movement in interest rates; and

(3) Is deemed by the examiner to provide a more accurate assessment

of the bank's IRR risk profile than the supervisory model and meets the

criteria discussed in Section VII of the proposed policy statement.

Reacting to the September NPR, some commenters requested the

agencies to provide more explicit guidelines on the criteria that

examiners will use to evaluate the adequacy of a bank's model. Other

commenters cautioned the agencies against creating checklists of

acceptable assumptions or measurement techniques. Such lists, they

believed, would be incomplete given the diverse nature of banks and

would stifle innovation in both risk measurement and product

development. Some commenters also expressed concern that the

assumptions and results of the supervisory model would be used as an

explicit benchmark against which internal models would be judged and

compared. These commenters were concerned that examiners would require

the bank to conduct detailed and ongoing reconciliations between the

bank's internal model and the supervisory model results. Such

requirements, they believed, imposed unnecessary burdens and lessened

the incentives for banks to use their own IRR models. Commenters

raising these concerns generally urged the agencies to refrain from

imposing supervisory model assumptions on bank models and from

requiring banks that have their own internal model to report the

information required for the supervisory model.

A key issue for the agencies, and one reason for delaying the

implementation of explicit minimum capital standards for IRR, is the

degree of specification the agencies need to establish when internal

models are used for assessing regulatory capital adequacy. The agencies

are aware that there are a variety of measurement systems and

assumptions in use by the industry to measure exposures. While such

variation may be appropriate given the diverse nature of commercial

banks, it may lead to different assessments of risk and hence, capital

requirements, for institutions that have similar risk profiles. More

explicit guidance from the agencies on acceptable techniques and

assumptions could help to lessen this variation and the risk that

different amounts of capital may be required for banks with similar

portfolios. Such guidance also would help reduce inconsistencies among

examiners and agencies in evaluating internal models. Efforts to devise

more explicit guidance could, however, result in standards which are

inappropriate for some institutions and may impede the industry's

continued innovation of more sophisticated risk measurement techniques.

The agencies welcome industry comments and suggestions on criteria and

standards that they should establish for accepting internal model

results.

With regard to reporting, the agencies propose that internal model

results be reported on voluntary basis in a supplemental Call Report

schedule like that portrayed in Schedule A. In response to the concerns

of many commenters, the agencies propose that such reporting be on a

confidential basis. Although many commenters to the September NPR

requested that banks submitting internal model results not be required

to also report the data required for the supervisory model, the

agencies propose the data for the supervisory model be collected from

all non-exempt banks. While recognizing the reporting burden that this

imposes, the agencies believe that collecting data for both internal

and the proposed supervisory model results will be important for

effective supervision. Moreover, such data also will help the agencies

evaluate the use of both the supervisory model and internal models as

the basis for ultimately establishing minimum capital charges for IRR.

By monitoring the maturity and repricing data collected for the

supervisory model, the agencies will be able to assess whether

supervisory and internal models results capture major shifts in

portfolio compositions. Such monitoring may help identify key model

assumptions that should be highlighted for examiner review and common

strengths or weaknesses of internal measures when compared to the

supervisory model. This information will help the agencies to provide

better guidance to examiners and bankers on acceptable risk measurement

techniques. It will also assist the agencies in determining what, if

any, improvements could be made to the proposed supervisory model

before explicit minimum capital charges are implemented.

V. Reporting Requirements

The implementation of this policy statement relies on changes to

the Call Report. The examples of Call Report schedules shown in this

proposal and the accompanying draft reporting instructions for those

schedules are provided to assist the reader in analyzing the full

implications of the proposal. Once comments are received on the

measurement framework and any

[[Page 39503]]

modifications that the agencies believe are appropriate are made, the

proposed Call Report schedules would also be amended to reflect those

changes. At that time, the Call Report schedules would be submitted to

FFIEC's Reports Task Force for inclusion in the comment document for

March 1996 Call Report changes. The FFIEC will submit any Call Report

changes to OMB for review as required under the Paperwork Reduction Act

44 U.S.C. 3501. Opportunity for public comment is always provided in

relation to such a submission. Nevertheless, the agencies invite

comments regarding the paperwork implications of this proposed policy

statement, and will carefully consider any comments received in the

development of the policy, as well as in recommending to the FFIEC

proposed revisions to the Call Report.

VI. Implementation Schedule

The agencies propose to require any additional reporting by non-

exempt banks beginning with the March 1996 Call Reports. Full

implementation of this policy statement for assessing the adequacy of

bank capital would be effective December 31, 1996.

VII. Requests for Comments

Comments are requested on all aspects of the proposed policy

statement, including the suggested implementation schedule. The

agencies particularly request comments on the following issues:

1. Exemption for Small Banks

The agencies propose to exempt certain small banks from the

proposed policy statement and associated reporting requirements in

order to lessen regulatory burdens on small, well-managed banks. The

criteria for exemption considers the size of the bank, its overall

CAMEL rating and the proportion of assets in intermediate and longer-

term maturities.

a. Are the three criteria used for the exemption appropriate and

reasonable?

b. Does the use of a bank's confidential CAMEL rating as one of the

exemption criteria raise concerns that it may allow public users of

Call Reports to discern a bank's CAMEL rating?

c. Does the proposed exemption criteria provide adequate safeguards

against exempting banks that pose significant risks to the deposit

insurance fund due to IRR?

2. Baseline Supervisory Model

The agencies are proposing that all non-exempted banks provide

information for a baseline supervisory model, the results of which,

would be one factor that an examiner would use to assess a bank's level

of IRR exposure and its need for capital. The baseline model uses seven

time bands and applies a series of risk-weights to a bank's reported

repricing and maturities balances in each of those time bands. For

certain types of instruments or activities, a bank would be required to

provide their own estimate of the change in value (self-report) of the

instruments or activities for the specified interest rate scenario.

a. Does the proposed baseline supervisory model provide a

reasonable basis for measuring a bank's IRR exposure? If not, what

changes should be made to the model?

b. Are the amount and type of data proposed to be collected for the

model appropriate and reasonable? If not, what changes could be made

either to improve the usefulness of the data collected and/or reduce

the burden of the proposal?

c. Do banks have the ability to calculate or obtain reasonable

estimates of changes in market values for the items where self-

reporting would be required? If not, how should such items be

incorporated into the model? What factors should examiners consider in

reviewing and assessing the reliability of bank's self-reported

estimates?

d. Are the risk-weights proposed for the baseline model appropriate

for an immediate and parallel 200 basis change in interest rates?

e. What portion, if any, of the proposed Call Report interest rate

risk data and output from the proposed supervisory measurement system

should be made available to the public through Call Report disclosures

and the Uniform Bank Performance Report?

3. Treatment of Non-Maturity Deposits

The agencies propose limits on how a bank could distribute deposits

without specified maturities (DDA, NOW, MMDA and savings) among the

time bands for the supervisory model. In setting these limits, the

agencies propose to treat commercial DDA balances separately from other

DDA balances. As proposed, these limits only apply to the standardized

supervisory model. The proposal would give an examiner the latitude to

use a bank's own non-maturity deposit assumptions when evaluating the

bank's capital adequacy for IRR provided that the bank can demonstrate

and support those assumptions.

a. Is it appropriate to treat commercial DDA balances separately

from other DDA balances?

b. Are the proposed maturity limits reasonable for a standardized

reporting and measurement framework?

c. Is it appropriate to give examiners latitude to use a bank's own

non-maturity deposit assumptions? If so, should the agencies specify

minimum standards of analysis that will be acceptable for banks that

wish to use their own assumptions? What types of analyses or factors

should be incorporated into such standards?

4. Supplemental Modules for Mortgage Holdings

The agencies have proposed supplemental reporting and expanded

risk-weight tables that would apply to banks that have concentrations

in either fixed- or adjustable-rate residential mortgage products.

These supplemental modules are designed to improve the supervisory

model's accuracy by incorporating more fully, the parameters which may

affect a mortgage's price sensitivity. The agencies propose to derive

the risk-weights for the supplemental modules from pricing tables

generated by the OTS's Net Portfolio Value Model (OTS model).

a. Is the information that would be collected for the supplemental

modules appropriate and meaningful? If not, what changes should be

made?

b. Are the thresholds proposed for requiring a bank to use the

supplemental modules appropriate? If not, what threshold would be

appropriate?

c. Do the supplemental modules and risk-weights sufficiently

address concerns about the supervisory model's accuracy for banks with

significant holdings of residential mortgage products? Will their use

lessen the possibility of different regulatory treatment for

institutions subject to the OTS model and those subject to this policy

statement?

d. Will the use of the supplemental modules and the associated

risk-weights used in those modules provide appropriate incentives for

bank decision-making? Will their use discourage the development of a

bank's own measurement capabilities?

e. Is the OTS model a reasonable source for developing the risk-

weights used in this module? If not, are there other sources that would

be more better?

f. The agencies believe the supplemental schedules related to

mortgages are necessary because the price sensitivity of these products

may vary substantially depending upon their coupon and cap

characteristics. Are the proposed supplemental schedules appropriate

and is the level of precision sought by the agencies reasonable?

[[Page 39504]]

5. Frequency of Updating Risk-Weights

In the interest of minimizing regulatory burden and providing

greater transparency and certainty for the supervisory model, the

agencies propose to update the risk weights for the baseline and

supplemental schedules only in the event of a significant movement in

market rates or other market factors that materially change the

accuracy of the derived price sensitivities and associated risk

weights. The OTS, in contrast, recalculates the price sensitivities for

its model each quarter in order to achieve the precision it believes

necessary to distinguish among different coupon rates of mortgage and

other products.

a. Does the agencies' intention to limit the updating of risk-

weights represent an appropriate balance among the objectives of

minimizing regulatory burden, providing transparency and certainty, and

providing sufficient measurement accuracy? If not, what other

approaches would be appropriate?

b. Does this limitation on updating risk weights materially reduce

the benefits and accuracy that the supplemental schedules for mortgages

are designed to provide?

c. The supplemental reporting schedule for fixed-rate mortgages

proposes to collect balance information by set coupon ranges. An

alternative that the agencies have considered is to collect balances on

the basis of their distance from prevailing current market coupons.

Such a treatment would allow the risk weight applied to any given

mortgage coupon to vary as its spread to current mortgage rates varies.

Would such a treatment be an improvement over the approach currently

proposed by the agencies? What, if any, difficulties would be

encountered in reporting balances on the basis of their spread to

current mortgage coupons?

6. Use of Carrying Values

In the interest of simplicity, the agencies propose to apply the

risk weights, including those derived from the OTS price sensitivities,

to the carrying value of a bank's instruments. To the extent that the

carrying and market values differ, this introduces an error in the

estimated price sensitivity of an instrument. The price sensitivity of

instruments whose market values exceed their carrying values will be

understated whereas the price sensitivity of instruments whose market

values are below carrying values will be overstated.

a. Is the use of carrying values an appropriate simplification and

does the use of carrying values for both assets and liabilities

sufficiently mitigate the materiality of such errors? If not, what

other approach(es) would be appropriate?

7. Use of Internal Models

a. Does the proposed policy statement provide appropriate

incentives for the use of banks' internal models and for banks to

enhance their internal risk measurement systems?

b. Are the criteria described for assessing a bank's internal model

appropriate? What other factors or criteria should examiners consider

in assessing and reviewing a bank's internal model results?

c. Should the agencies provide additional guidelines on acceptable

parameters, assumptions, and methodologies for internal models? What

types of guidance would be most useful?

d. Is the proposed voluntary schedule for reporting internal model

results appropriate? Are there sufficient incentives for banks to

provide this information on a voluntary basis?

8. Treatment of Trading Account

The agencies propose that banks ``self-report'' the change in value

of their trading account activities for a 100 basis point change in

interest rates. The agencies also are considering whether trading

account activities should be excluded from this policy statement and

IRR measure if a bank is subject to the market risk capital

requirements as proposed by the Basle Committee.

a. Is the 100 basis point interest rate scenario that the agencies

propose to use when measuring the IRR exposure in a bank's trading

portfolio appropriate? If not, what scenario would be appropriate?

b. What modifications, if any, should be made to this proposal for

banks that may be subject to the Basle Committee's proposed capital

standards for market risk in trading activities? What, if any,

operational problems would be created if such banks were simply

exempted from including and reporting their trading activities for

purposes of this policy statement? What, if any, competitive issues

would such a treatment present?

The text of the proposed policy statement follows. The first two

appendices to the proposed policy statement provide proposed reporting

schedules and accompanying instructions for those schedules that are

under consideration by the agencies as part of this proposed policy

statement. The third appendix provides the risk weights that would be

used in the proposed supervisory model. The fourth appendix provides

technical descriptions of the derivation of the model's risk weights

and the supplemental modules for residential mortgage-related products.

Proposed Policy Statement

I. Purpose

This supervisory policy statement is adopted by 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), collectively, the ``agencies.'' The statement

establishes a supervisory framework that the agencies will use to

assess and measure the interest rate risk (IRR) exposures of insured

commercial and FDIC supervised savings banks. The results of this

measurement framework will be used by the agencies in their evaluation

of a bank's IRR exposure and whether it needs capital for IRR. Each

agency has additional guidance and policies on the measurement and

management of IRR. Those policies and guidelines set forth each

agency's expectations regarding safe and sound banking practices for

IRR management. This policy statement does not replace or supersede

those issuances. The adoption of this policy statement by the agencies

does not replace the agencies' expectations that all insured depository

institutions have internal IRR measurement and management processes

that are commensurate with the nature and level of their IRR exposures.

II. Background

Interest rate risk is the adverse effect that changes in market

interest rates have on a bank's earnings and its underlying economic

value. Changes in interest rates affect a bank's earnings by changing

its net interest income and the level of other interest-sensitive

income and operating expenses. The underlying economic value of the

bank's assets, liabilities, and off-balance sheet instruments also are

affected by changes in interest rates. These changes occur because the

present value of future cash flows and in some cases, the cash flows

themselves, change when interest rates change. The combined effects of

the changes in these present values reflect the change in the bank's

underlying economic value.

Interest rate risk is inherent in the role of banks as financial

intermediaries. Interest rate risk, however, introduces volatility to

bank earnings and to the economic value of the bank. A bank that has an

excessive level of IRR can diminish its future earnings, impair its

[[Page 39505]]

liquidity and capital positions, and, ultimately, jeopardize its

solvency.

The agencies believe that safety and soundness requires effective

management and measurement of IRR, and each agency has provided

supervisory guidance to banks and examiners on this subject. In

addition, the agencies believe that a bank's capital adequacy should be

assessed in the context of the risks it faces, including interest rate

risk. Both of these aspects of IRR depend, among other things, on a

meaningful measurement of the bank's risk exposure.