FRBSF Economic Letter
2004-26; September 17, 2004
Supervising Interest Rate Risk Management
Over the past 20 years, financial institutions have
made significant efforts to establish and improve their procedures
for interest
rate risk management, including using economic models of interest
rates and related models of credit risk (Lopez 2001a, b). At the
same time, bank supervisors worldwide, including the Federal Reserve,
have been expanding their knowledge and oversight of interest rate
risk management techniques. For example, U.S. bank supervisors
recently issued supervisory guidance on sound risk management practices
regarding the valuation of mortgage servicing rights (Board of
Governors 2003).
The centerpiece of these international supervisory efforts is
the revised Basel Capital Accord, which was released in June 2004
and
is to be fully implemented by year-end 2007. This Economic
Letter
reviews the Accord's stated principles on interest rate risk. In
brief, the principles strongly support the idea that banks' internal
risk assessments should, whenever possible, form the basis for
supervisory oversight of their interest rate risk profiles. The
principles suggest supervisory guidelines for assessing the adequacy
of interest rate risk management systems, such as focusing on banks'
internal control functions and stress-testing results.
Components
of interest rate risk
Interest rate risk (IRR) is defined as the
change in a bank's portfolio value due to interest rate fluctuations.
Taking on IRR is a key
part of what banks do; but taking on excessive IRR could threaten
a bank's earnings and its capital base, raising concerns for bank
supervisors. In practice, IRR management systems have been developed
to measure and control such risk exposures, both in the trading
book (i.e., assets that are relatively liquid and regularly traded)
and in the banking book (i.e., assets, such as loans, that are
much less actively traded).
IRR can be roughly decomposed into four
categories: repricing risk, yield curve risk, basis risk, and optionality
(see Basel Committee
on Banking Supervision (BCBS) 2003). Repricing risk refers to fluctuations
in interest rate levels that have differing impacts on bank assets
and liabilities; for example, a portfolio of long-term, fixed-rate
loans funded with short-term deposits (i.e., a case of duration
mismatch) could significantly decrease in value when rates increase,
since the loan payments are fixed (and funding costs have increased).
Yield curve risk refers to changes in portfolio values caused by
unanticipated shifts in the slope and shape of the yield curve;
for example, short-term rates might rise faster than long-term
rates, clearly affecting the profitability of funding long-term
loans with short-term deposits. Basis risk refers to the imperfect
correlation between index rates across different interest rate
markets for similar maturities; for example, a bank funding loans
whose payments are based on U.S. Treasury rates with deposits based
on Libor rates is exposed to the risk of unexpected changes in
the spread between these index rates. Finally, optionality refers
to risks arising from interest rate options embedded in a bank
assets, liabilities, and off-balance-sheet positions. Such options
can be explicitly purchased from established markets for interest
rate derivatives or included as a term within a loan contract,
such as the prepayment option included in residential mortgages.
IRR
management
Banks have access to a wide array of financial tools
for managing their IRR, such as standard asset-liability management
procedures
and interest rate derivatives. Banks commonly use one of two approaches
when assessing aggregate IRR exposures across their various business
lines and portfolios—the traditional earnings approach and the
more challenging economic value approach. The earnings approach
focuses on how interest rate changes affect a bank's overall earnings,
which are typically measured as net interest income (the difference
between total interest income and total interest expenses). Broader
measures that include non-interest income, such as revenue from
mortgage servicing activities, and expenses have become common,
however. The main point of this approach is to examine earnings
sensitivity to interest rate fluctuations of different sizes.
The
economic value approach takes a broader perspective on IRR management
by focusing on how interest rate changes affect total
expected net cash flows from all of a bank's operations. Thus,
this approach examines expected cash flows from assets minus expected
payments on liabilities plus the expected net cash flows from off-balance-sheet
positions, such as fees charged for borrower credit lines. This
approach is more challenging to conduct since, at a minimum, it
requires collecting and aggregating more data; at the same time,
it provides greater insight into a bank's aggregate IRR exposure.
In
addition to such aggregate IRR management approaches, banks use
more focused IRR measurement techniques for derivatives and
other instruments with especially complex risk profiles, such as
mortgage-backed securities. While the aggregate approaches typically
involve making judgmental adjustments to interest rates and tracking
their impact across the bank, the focused techniques explicitly
use mathematical models of interest rate dynamics for various index
rates and their yield curves. For example, many possible future
interest rate paths are generated and used to examine the potential
effects of interest rate changes on portfolio values, investment
returns, and cash flows from different assets. Since the models
can examine the components of interest rate risk separately, risk
managers use them to gauge and control their portfolios' exposures
to a broader range of interest rate fluctuations. In theory, the
more sophisticated IRR management techniques could be applied to
the bank as a whole. Important developments in this direction have
been made, but several important challenges still remain, especially
in aggregating IRR exposures across business lines.
A key advantage
of these mathematical IRR management techniques is that they provide
a consistent framework for analyzing a wide
variety of possible interest rate scenarios. For example, banks
can consider multiple scenarios accounting for changes in the general
level of interest rates and changes in the relationships among
interest rates. However, since models are just simplifications
of actual phenomena, prudent IRR management requires considering
extreme scenarios that might not be within a given model's structure.
This practice is commonly called stress-testing, since the underlying
model and IRR management system are "stressed" by examining
uncommon, although not implausible, scenarios. Common stress scenarios
include abrupt changes in the general level of interest rates (i.e.,
repricing risk), changes in the relationships among key market
rates (i.e., basis risk), changes in the slope and shape of the
yield curve (i.e., yield curve risk), changes in the liquidity
of key financial markets, and changes in the volatility of market
rates. Optionality risks typically are affected by all of these
scenarios.
Supervisory guidelines
As part of its ongoing efforts to address
international bank supervisory issues and to support the revised
Basel Capital Accord, the BCBS
recently issued a summary paper regarding general principles on
IRR management. The principles were intended to be used in the
supervisory evaluation of the adequacy and effectiveness of bank
IRR management systems and in developing supervisory responses
to these systems. The principles are based on the current IRR management
practices of large international banks and are intended for IRR
exposures arising from trading and book activities.
The principles
advocate that banks have in place comprehensive management systems
that measure and control IRR exposures effectively.
The systems must be subject to appropriate board of directors and
senior management oversight. Specifically with respect to supervisors,
the principles advocate that banks' own IRR management systems
should, whenever possible, form the basis of supervisors' measurement
of and response to their interest rate sensitivity. The BCBS principles
can be grouped into four categories: IRR management oversight issues,
issues related to adequate bank policies and procedures, issues
specific to IRR monitoring and control, and specific supervisory
issues.
With respect to management oversight issues, the principles
state that a bank's board of directors should approve IRR strategies
and policies and ensure that senior management effectively monitors,
communicates, and controls these risks. Furthermore, risk managers
within the IRR management system must be independent from the risk-taking
functions of the bank in order to avoid potential conflicts of
interest. Risk managers also should be able to report IRR exposures
directly to senior management and the board of directors.
Senior
management must ensure that a bank's IRR policies and procedures
are clearly defined and consistent with the nature and complexity
of the bank's activities. For example, senior management could
articulate its risk tolerance, both for the bank as a whole and
for the disaggregated business units, by crafting policy statements
identifying specific interest rate instruments and activities that
are permissible. When proposing new interest rate products or activities,
management should work to identify the inherent risks clearly and
ensure that adequate procedures and controls are in place before
introducing them.
With respect to IRR monitoring and control issues,
banks must capture all material IRR exposures, whether in their
trading or banking
books, within their management systems. Operating limits and related
practices for keeping IRR exposures within levels consistent with
internal policies must be clearly established and enforced. Furthermore,
all IRR modeling assumptions and parameters must be well documented
and updated with reasonable frequency. Stress-testing should be
regularly used to assess the bank's interest rate sensitivity and
examine the appropriateness of key modeling assumptions. Stress-test
results must be considered when establishing and reviewing IRR
policies and procedures. A bank must have adequate information
systems for reporting accurate IRR exposure information on a timely
basis to its board of directors and senior management. Finally,
effective IRR management systems require regular evaluations by
independent auditors, whether internal or external.
With respect
to supervisory issues, the BCBS principles address four main concerns.
First, since banks' own systems are to form
the basis of supervisory oversight of IRR management, supervisors
should receive sufficient and timely information with which to
evaluate bank's IRR systems. For example, supervisors should have
ready access to information on the range of maturities and currencies
in bank portfolios, including off-balance-sheet items. Information
contained in internal management reports, such as earnings and
economic value estimates, and the results of stress tests would
also be useful. Second, banks should disclose publicly information
on their aggregate IRR exposures and their policies for managing
them. The BCBS has issued recommendations for the public disclosure
of information on IRR as part of the overall review of the Basel
Accord (Lopez 2003).
Third, to facilitate supervisory monitoring
of IRR exposures across institutions, banks should try to use standardized
rate changes
to provide the results of their internal measurement systems, expressed
in terms of changes to economic value. According to the BCBS guidelines,
these rate changes should in principle be determined by banks but
based on the recommended criteria. For example, for IRR exposures
in G-10 currencies, banks should consider either a parallel rate
change of ±200 basis points or the changes implied by the
1st and 99th percentiles of historically observed interest rate
changes over at least five years. Fourth, senior management and
boards of directors should periodically review both the design
and the results of their stress tests. Supervisors will continue
to expect institutions to examine multiple scenarios in evaluating
the appropriate level of their IRR exposures.
If supervisors determine that a bank's management system does
not capture its IRR exposures fully, the bank would be required
to
bring its system up to the appropriate supervisory standards.
If supervisors determine that a bank is not holding sufficient
capital
for its level of IRR exposure, especially in the banking book,
remedial action should be considered, requiring the bank to reduce
its risk or to set aside additional capital or a combination
of the two, depending on the situation.
Conclusion
In support of the revised Basel Accord, the BCBS has
issued several guidelines regarding IRR management for both bankers
and bank
supervisors. The BCBS is aware that banks' IRR management
techniques continue
to evolve, so certain details of their guidelines will need
to be updated. However, the principle that banks' own assessments
of their IRR exposures should form the basis of supervisory
oversight
is a defining characteristic of future supervisory efforts.
Jose A. Lopez
Senior Economist References
[URLs accessed September 2004.]
Basel Committee on Banking Supervision. 2004. "Principles
for the Management and Supervision of Interest Rate Risk."
http://www.bis.org/publ/bcbs108.pdf
Board of Governors of the Federal Reserve
System. 2003. "Risk
Management and Valuation of Mortgage Servicing Assets Arising from
Mortgage Banking Activities." Supervisory Letter SR03-4.
http://www.federalreserve.gov/boarddocs/SRLETTERS/2003/sr0304.htm
Lopez,
J.A. 2003. "Disclosure as a Supervisory Tool: Pillar
3 of Basel II." FRBSF Economic Letter 2003-22 (August 1).
http://www.frbsf.org/publications/economics/letter/2003/el2003-22.html
Lopez,
J.A. 2001a. "Financial Instruments for Mitigating Credit
Risk." FRBSF Economic Letter 2001-34 (November 23).
http://www.frbsf.org/publications/economics/letter/2001/el2001-34.html
Lopez,
J.A. 2001b. "Modeling Credit Risk for Commercial Loans." FRBSF
Economic Letter 2001-12 (April 27).
http://www.frbsf.org/publications/economics/letter/2001/el2001-12.html
|
