Written evidence submitted by the Bank
of England
TOWARDS A MACRO-PRUDENTIAL INSTRUMENT
At its hearing on 26 February, the Treasury
Committee requested from the Bank of England a paper on counter-cyclical
macro-prudential instruments (MPI).
The rationale for an MPI is well illustrated
by the events over the past decade. In the UK and internationally,
monetary policy was aimed at stabilising the overall level of
inflation. It achieved that in part through balancing aggregate
demand and aggregate supply in the economy. At the same time,
financial regulation focussed on the conduct and resilience of
individual institutions. For much of this decade that approach
appeared to work well, with demand and inflation stable. There
were also very few failures of financial institutions.
But no instrument or institution was charged
explicitly with controlling overall financial conditions, except
insofar as this affected inflation, aggregate demand or individual
institutions. In that environment, bank balance sheets grew unchecked.
Between 2000 and 2007, they roughly trebled in size. Latent
vulnerabilities built-up within the financial system. The credit
crisis of the past 18 months has exposed those vulnerabilities,
with highly adverse consequences for both the financial system
and the real economy.
With hindsight, there was a gap between the
macro-economic and the micro-prudential arms of policy. The growth
of the financial sector might have been moderated, and the subsequent
crisis made less painful, had there been an instrument filling
the gap between the macro-economic and the micro-prudential—a
macro-prudential policy instrument. If implemented correctly,
this might have resulted in a more stable path for both the real
economy and the financial sector.
This is easier said than done. Implementation
challenges for an MPI are considerable. This paper provides a
preliminary assessment of some of the key operational issues involved
in the design of an MPI. These include:
— Key design features of an MPI (objectives,
instruments);
— Potential objectives of an MPI;
— Potential instruments for implementing
an MPI;
— Practical problems in implementing an
MPI.
DESIGNING A
MACRO-PRUDENTIAL
INSTRUMENT
From the second World War up until the early
1980s, various quantitative restrictions were placed on UK commercial
banks in an attempt to stabilise their balance sheets and thereby
the real economy. This historical experience provides lessons
for the design of an MPI. MPIs are intended to curb cyclical variations
in credit provision. But they may also lower average amounts of
credit being provided to certain classes of borrower, thereby
potentially constraining growth.
The sub-prime experience in the US is salutary.
With hindsight, this is seen as a period of laxity in credit provision,
with large costs for the US and global economies in general and
for US home-owners in particular. But it is important not to forget
that, ahead of crisis, the relaxation of credit constraints for
large cohorts of previously credit-constrained US households was
seen as a success story. In designing an MPI, the key is to find
a balance between these factors.
At a high level, the key design features of
an MPI are threefold:
— Objectives: What is the MPI seeking
to achieve?
— Instruments: How is the MPI calibrated
to achieve these objectives?
— Institutions: Who is charged with
implementing the MPI?
These decisions follow a natural sequence. For
example, without first defining the underlying objectives of an
MPI, it is difficult to determine which instruments are most appropriate
and how they should be adjusted. And decisions on the objectives
and instruments of an MPI should logically precede decisions on
who should operate it.
Over the past couple of months, there have been
several reports from the official sector internationally and from
the academic community on the design of an MPI. Several international
committees are also engaged in work programmes. Annex A lists
some of those reports and committees. In general, these reports
have tended to focus on the "who" more than the "what"
and the "how".
A better starting point is "what"
and "how". But the choice of objectives and instruments
raises difficult analytical and practical issues and potential
trade-offs. The remainder of this paper focuses on those operational
issues.
OBJECTIVES OF
MACRO-PRUDENTIAL
POLICY
The current debate on MPIs has illustrated that,
as yet, consensus on objectives has not been reached. To illustrate,
at one end of the spectrum are a set of proposals which are essentially
about making banks more resilient against cyclical variations
in the economy. They are, first and foremost, about ensuring the
safety and soundness of banks and their depositors and creditors.
These measures have been the focus of official sector reports
to date.
One example of such a policy is "dynamic
provisioning". This is a set of rules which aim to ensure
banks set aside sufficient reserves for a cyclical downturn, providing
an additional cushion for banks and thereby better enabling them
to maintain lending during a recession.[1]
This regime has operated in Spain for a number of years. Leverage
ratios—a measure of banks' assets relative to their equity—are
a second potential counter-cyclical measure.
At the other end of the spectrum are proposals
which are essentially about dampening the growth in credit. Policies
which adjust regulatory ratios, or margin requirements, in response
to excessive credit growth or asset price inflation would fall
into this category. These measures have been discussed in a number
of academic reports. They go beyond dynamic provisioning, which
did not appear to constrain credit growth in Spain much over recent
years.
At present, both sets of policies are being
captured under the "macro-prudential" umbrella. But
these measures could operate in different ways, potentially requiring
different instruments and placing different informational demands
on the authorities. They may also have potentially different implications
for the behaviour of the financial sector and the real economy.
The current UK conjuncture provides a good illustration
of the potentially different objectives an MPI could serve. If
the only concern at the present time was protecting banks from
the downturn, the authorities would be raising capital requirements
to provide an extra buffer. But on broader macroeconomic grounds,
there is a case for actually lowering capital ratios, so giving
banks extra flexibility to lend. Before hardwiring either approach
into the design of an MPI, it will be critical to understand and
evaluate these different approaches and their consequences.
INSTRUMENTS OF
MACRO-PRUDENTIAL
POLICY
There are several aspects to this, including:
(a) Which instrument?
In theory, an MPI could be used to exercise
control over almost any aspect of banks' balance sheets—for
example, capital or debt on the liabilities side, or lending on
the assets side. This control could also be exercised using either
prices or quantities. Annex B sets out some options and provides
some examples.
Choosing between these instruments involves
trade-offs. It involves balancing the desire to exercise leverage
over credit supply decisions on the one hand, and the desire to
minimise effects on the commercial decision-making of financial
institutions on the other. For example, adjusting regulatory capital
ratios would be one means of operating an MPI. This would cause
less interference in banks' decision-making. At the same time,
its impact on banks' lending choices would be indirect and thereby
uncertain in extent. For example, it is unclear whether lowering
capital ratios for banks at present would encourage them to lend.
Instruments that act directly on the assets
side of banks' balance sheets—for example, direct lending
controls or prescribed loan-to-value ratios—would strike
a different balance. They would, on the face of it, score better
in terms of their influence on credit supply decisions. That is
why, for example, the UK authorities have during this year used
lending agreements to support the economy. But this would come
at the expense of greater impact on, and hence potential distortion
to, commercial banks' decision-making.
This is a second area where further analysis
of the operational choices, and the tradeoffs they present, would
be essential before putting an MPI into practice.
(b) Single v multiple and rules v discretion?
Other dimensions to instrument choice include
whether there should be one instrument or many, and whether that
instrument should operate according to a predefined rule or be
discretionary. Earlier UK experience is revealing here. Multiplying
the number of restrictions on banks' balance sheets was rarely
beneficial. It added complexity and thus distortion without any
correspondingly greater degree of control. As in a monetary policy
context, this suggests there should be a strong preference for
simple, targeted measures wherever possible and we should aim
to avoid a proliferation of instruments.
On rules versus discretion, a case can be made
analytically for an MPI having rule-like features. Rules can reinforce
the credibility of a regime, by acting as a bulwark against forbearance—for
example, offsetting the inevitable incentive to avoid raising
required capital ratios when a credit boom was in full swing.
Rules also increase clarity and hence policy transparency.
Against that, the inflexibility of fixed rules
can be a constraint in some circumstances—for example, if
underlying behaviour in the economy is changing. That is one of
the key lessons from history. When implementing restrictions on
banks' balance sheets, whether for prudential or macroeconomic
purposes, Goodhart's Law (that historical relationships are apt
to change after a policy is implemented) has been an ever-present
problem.
This may point towards an MPI needing to operate
within a framework of "constrained discretion", combining
some rule-like features with some discretion. By analogy, this
is now widely accepted internationally as the optimal framework
for the implementation of monetary policy.
PRACTICAL IMPLEMENTATION
There are a large number of potential practical
problems in implementing an MPI. These arise almost irrespective
of the precise operational model. Many of these practical issues
hinge critically on the choice of end-objective. They include:
— Institutional scope of regulation:
Historical experience suggests that there will inevitably be strong
incentives to avoid regulatory rules: for example, Regulation
Q in the US stimulated the euro-dollar market in London; the 1970s
"Corset" in the UK encouraged disintermediation, for
example through the acceptances market; and, more recently, Basel
I stimulated growth in the shadow banking system. At a minimum,
this calls for a degree of flexibility when determining the appropriate
institutional scope of an MPI. More broadly, however, this underscores
the importance of determining the appropriate objective of an
MPI. If the aim is to protect depositors, this suggests a focus
on deposit-taking institutions. If the focus is on credit supply,
this may speak to a potentially different set of institutions.
— International scope of regulation:
An important additional complication is the treatment of cross-border
banks. Without consistent application of the regime internationally,
there would be strong avoidance incentives—for example,
by booking business in countries where the macro-prudential regime
was looser or non-existent. Consider a London branch of a Swiss
bank lending to a US firm expanding its operations in Germany.
Who should operate the lever, to which entity should it apply
and calibrated to whose credit cycle? There is no easy answer
to those questions. The answers are once again importantly influenced
by the objective of an MPI. If the objective is resilience of
the financial system, then conditions in Switzerland and the UK
become central; if the objective is stabilisation of the credit
cycle, then credit conditions in the US and Germany become a crucial
determinant.
— Consistency with other policy instruments:
A successful MPI will support the other arms of policy—macro-economic
policy and micro-prudential policy. This suggests there needs
to be consistency between these arms of policy. Objectives are
again key. Narrower macro-prudential tools call for consistency
with micro-prudential instruments; broader macro-prudential tools
for consistency with monetary policy.
— Bank-specific v system-wide calibration:
Should macro-prudential tools be calibrated to individual firms'
own balance sheets positions or to the balance sheet of the system
as a whole? A case can be made for either and the case rests,
once again, on end-objectives. The greater the orientation of
an MPI towards the resilience of the banks, the stronger the case
for calibrating to individual institutions' balance sheets. The
greater the orientation towards dampening the credit cycle, the
more important becomes the need to calibrate interventions according
to system-wide financial conditions and behaviour.
NEXT STEPS
There is now a clear consensus in favour of
a counter-cyclical MPI. That is considerable progress and it is
important that this opportunity to reform the financial system
is pursued. At the same time, if an MPI is to be implemented,
it is crucial that it is robust and credible. That will require
a considerable programme of work to tackle the operational issues
raised above. It will also require a broader consideration of
other tools to increase systemic resilience.
The experience with monetary regimes suggests
that process cannot and should not be rushed. There is time to
consider carefully the design of new instruments. Nor should the
process be conducted piecemeal, with initiative layered on initiative
without a clear sense of direction. The law of unintended consequences
applies forcefully when introducing new policy instruments. The
aim should be to deliver a macro-prudential regime which both
matters and which lasts.
9 April 2009
Annex A
SELECTED REPORTS AND COMMITTEES ON MACRO-PRUDENTIAL
INSTRUMENTS
TABLE: SELECTED
REPORTS AND
COMMITTEES ON
MACROPRUDENTIAL INSTRUMENTS
Official sector reports
IMF: "The Perimeter of Financial Regulation"
(2009) available at: www.imf.org/extemal/pubs/ft/spn/2009/spn0907.pdf
IMF: "Lessons of the Financial Crisis for
Future Regulation of Financial Institutions and Markets, and for
Liquidity Management" (2009) available at: www.imf.org/external/np/pp/eng/2009/020409.pdf
IMF: "Lessons of the Global Crisis for
Macroeconomic Policy" (2009) available at: http://www.imf.org/external/np/pp/eng/2009/021909.pdf
G30: "Financial Reform: A Framework for
Financial Stability" G30 Report (chaired by Paul Volcker)
(2009) available at: www.group30.org/pubs/recommendations.pdf
FSF: "Report of the Financial Stability
Forum on Addressing Procyclicality in the Financial System"
(2009) available at: http://www.fsforum.org/
FSF: "Report of the Financial Stability
Forum on Enhancing Market and Institutional Resilience" (2008)
available at: http://www.fsforum.org/list/fsf_publications/tid_110/index.htm
Basel Committee, Speech by Chairman Wellink:
"Basel Committee initiatives in response to the financial
crisis" (2009) available at: www.bis.org/review/r090330a.pdf
EU: "Report by the High-Level Group on
Financial Supervision in the EU" (chaired by Jacques de Larosiere)
(2009) available at: ec.europa.eu/commission_barroso/president/pdf/statement_20090225_en.pdf
FSA: Turner Review and DP 09/ 02: "A Regulatory
Response to the Global Banking Crisis" (2009) available at:
www.fsa.gov.uk/pubs/discussion/dp09_02.pdf
US: The Department of the Treasury Blueprint
for a Modernized Financial Regulatory Structure ("Paulson
plan", 2008) available at: www.treas.gov/press/releases/reports/Blueprint.pdf
US: Framework For Regulatory Reform (Geithner's
"outline for regulatory reform", 2009) available at:
http://www.ustreas.gov/press/releases/tg72.htm
The Tripartite Review Preliminary report (2009)
by James Sassoon available at: https://www.tripartitereview.co.uk/
Academic reports
"The Fundamental Principles of Financial
Regulation" (2009) by Brunnermeier, Crockett, Goodhart, Persaud
& Shin available at: http://www.voxeu.org/index.php?q=node/2796
"Restoring financial stability: how to
repair a failed system" (2008) edited by Viral Acharya and
Matthew Richardson. See: http://whitepapers.stern.nyu.edu/home.html
"New Ideas for the London Summit"
(Chatham House and Atlantic Council) (2009) available at: www.chathamhouse.org.
uk/publications/papers/download/-/id/727/file/13733_r0409_g20.
pdf
International Committees with work underway
Financial Stability Forum (Working Group on
Market and Institutional Resilience)
Basel Committee on Banking Supervision
EU Economic and Financial Committee (Working
Group on Procyclicality)
European Banking Committee/Committee of European
Banking Supervisors (Joint working group on supplementary measures)
Committee of European Banking Supervisors (Expert
Group on Prudential Requirements and Working Group on Cyclically)
Annex B
EXAMPLES OF BANK BALANCE SHEET INSTRUMENTS
Side of the balance sheet affected
| First variable affected | Policy tool
| Comments |
| Liability side | Quantities
| Capital Requirements (floor) | These requirements could take many different forms. For example, a discretionary counter-cyclical buffer, or non distributable cyclical reserves ("dynamic provisioning").
|
| | Core funding requirement (floor)
| Liquidity regulation could impose constraints on the extent to which banks can use less stable sources of funding to grow rapidly. The FSA's proposed "core funding ratio" is an example of this.
|
| | Margining requirements (floor)
| Broad-based collateral arrangements or a margin-setting authority could enforce margining rules.
|
| | Controls on the growth of banks' IBELs [interest-bearing eligible liabilities] (ceiling)
| For example, the "Corset" (used in the UK during the 1970s). This scheme penalised banks whose IBELs grew faster than the prescribed rate.
|
| Prices | Deposit rate ceiling (ceiling)
| A deposit rate ceiling could be used to constrain banks' ability to expand rapidly, funded by high-paying retail deposits.
|
| Asset side | Quantities |
Lending controls (ceiling) | Direct controls on the quantity of bank lending were in place prior to the introduction of Competition and Credit Control in 1971. Between 1965 and 1971, ceilings were used to target a specific rate of lending growth.
|
| | Loan-to-value/Loan to-income ratios (ceiling)
| This approach was used in Hong Kong in the 1990s. The HKMA had a recommended maximum LTV ratio in 1991 of 70% for property lending.
|
| | Cash reserve requirements (floor)
| Requirements to hold government bonds or cash reserves deposited at the Bank.
|
| Prices | Loan rate control (floor or ceiling)
| If used as a ceiling, loan rate control would choke off lending, because it would hamper banks' ability to lend to higher risk customers. If used as a floor, control of the loan rate could act directly on the demand for credit.
|
| |
| |
1
See, for example, Box 6: Countercyclical measures, Bank of England
Financial Stability Report, October 2008. Back
| 
