Post on 02-Oct-2020
transcript
Bank of England Volume 42 Number 3
QuarterlyBulletinAutumn 2002
Bank of England Quarterly BulletinAutumn 2002
Summary 247
Recent economic and financial developments
Markets and operations 249
Research and analysis
Committees versus individuals: an experimental analysis of monetary policy decision-making 262
Parliamentary scrutiny of central banks in the United Kingdom and overseas 274
Ageing and the UK economy 285
The balance-sheet information content of UK company profit warnings 292
Money and credit in an inflation-targeting regime 299
Summaries of recent Bank of England working papersFinancial liberalisation and consumers’ expenditure: ‘FLIB’ re-examined 308Soft liquidity constraints and precautionary saving 309The implications of an ageing population for the UK economy 310On gross worker flows in the United Kingdom: evidence from the Labour Force Survey 311Regulatory and ‘economic’ solvency standards for internationally active banks 312Factor utilisation and productivity estimates for the United Kingdom 313Productivity versus welfare: or, GDP versus Weitzman’s NDP 314Understanding UK inflation: the role of openness 315Committees versus individuals: an experimental analysis of monetary policydecision-making 316The role of corporate balance sheets and bank lending policies in a financialaccelerator framework 317
Printed by Park Communications© Bank of England 2002
ISSN 0005-5166
Report
International Financial Architecture: the Central Bank Governors’Symposium 2002 318
Speeches
The monetary policy dilemma in the context of the internationalenvironmentSpeech by the Governor given at the Lord Mayor’s Banquet for Bankers and Merchants of the City of London at the Mansion House on 26 June 2002 326
Monetary policy issues: past, present, futureSpeech by Stephen Nickell, member of the Bank’s Monetary Policy Committee, delivered at a lunch organised by Business Link and the Coventry and Warwickshire Chamber of Commerce inLeamington Spa on 19 June 2002 329
The contents page, with links to the articles in PDF, is available at
www.bankofengland.co.uk/qbcontents/index.html
Authors of articles can be contacted at forename.surname@bankofengland.co.uk
The speeches contained in the Bulletin can be found at
www.bankofengland.co.uk/speech/index.html
Volume 42 Number 3
247
Quarterly Bulletin—Autumn 2002
Markets and operations(pages 249–61)
This article reviews developments in international and domestic financial markets,
drawing on information from the Bank of England’s market contacts, and describes
the Bank’s market operations in the period 17 May 2002 to 23 August 2002.
Research and analysis(pages 262–317)
Research work published by the Bank is intended to contribute to debate, and
does not necessarily reflect the views of the Bank or of MPC members.
Committees versus individuals: an experimental analysis of monetary policydecision-making (by Clare Lombardelli and James Talbot of the Bank’s Monetary
Assessment and Strategy Division and James Proudman of the Bank’s Conjunctural
Assessment and Projections Division). This article reports the results of an
experimental analysis of monetary policy decision-making under uncertainty. The
experiment used a large sample of economically literate undergraduate and
postgraduate students from the London School of Economics to play a simple
monetary policy game, both as individuals and in committees of five players. The
result—that groups made better decisions than individuals—accords with a previous
study in the United States with Princeton University students. The experiment also
attempted to establish why group decision-making is superior: although some of the
improvement was related to committees using majority voting when making decisions,
there was a significant additional committee benefit associated with members being
able to observe each other’s voting behaviour.
Parliamentary scrutiny of central banks in the United Kingdom and overseas(by Jonathan Lepper, formerly of the Secretariat to the House of Commons Treasury
Committee and Gabriel Sterne of the Bank’s International Economic Analysis
Division). This article reviews the parliamentary scrutiny of central banks in
14 countries using the results from a new survey. There is wide variation in the
nature of parliamentary scrutiny within the sample. There is no firm evidence in
these data, however, to suggest that particular types of framework are associated with
different overall levels of parliamentary scrutiny. The Bank of Japan, Bank of England,
European Central Bank (ECB) and Federal Reserve each make higher-than-average
appearances before their respective parliaments to discuss monetary policy issues,
and the technical support provided to the relevant committees is relatively high in the
US Congress and in the European Parliament. The level of scrutiny can be
circumstance specific, and some inflation-targeting frameworks have defined specific
conditions that would trigger scrutiny and the form it would take.
Ageing and the UK economy (by Garry Young of the Bank’s Domestic Finance
Division). This article argues that overall living standards in the United Kingdom are
set to double over the next 50 years alongside a sharp increase in the proportion of
people over retirement age. While there are clear risks to this outlook, these would
be present even without demographic change. Nevertheless an ageing population
does appear to increase the risks to the financial welfare of individuals, especially in
their old age. If people living longer do not save more when they are working, then
either they have to consume less in their old age or work for longer than would have
248
BBaannkk ooff EEnnggllaanndd QQuuaarrtteerrllyy BBuulllleettiinn:: Autumn 2002
been the case had greater provision been made for retirement. This risk is
heightened by general uncertainty about asset returns which becomes more
important as the number of people reliant on private pensions increases.
The balance-sheet information content of UK company profit warnings (by Allan
Kearns and John Whitley of the Bank’s Domestic Finance Division). This article looks
at the information content of profit warnings issued by UK private non-financial
companies over the period 1997–2001 in relation to measures of their profitability
and balance-sheet strength. It finds that profit warnings are associated with a
persistent fall in profit margins and that this decline in margins is larger than for
companies who do not issue warnings. The article also finds that profit warnings
contain incremental information for other balance-sheet variables: those firms who
issue warnings are also more likely to see their gearing levels rise, and investment and
dividends fall, than other firms whose profit margins also fall but who do not issue a
warning.
Money and credit in an inflation-targeting regime (by Andrew Hauser and Andrew
Brigden of the Bank’s Monetary Assessment and Strategy Division). This article is
one of a series on the UK monetary policy process. It discusses how the assessment
of money and credit data fits into the Bank’s quarterly forecast round.
International Financial Architecture: the Central Bank Governors’ Symposium 2002The Central Bank Governors’ Symposium 2002 examined the architecture of the
world’s financial system. Horst Koehler, Managing Director of the IMF, and the Bank
of England’s two Deputy Governors at the time, David Clementi and Mervyn King,
gave the main addresses. This article summarises what they said. It also gives a
precis of eight background papers provided for the occasion. Taken together, these
eleven contributions explore general aspects of the international financial
architecture, as well as discussing how financial crises may be contained or prevented,
and best resolved when they do occur.
Report(pages 318–25)
249
Sterling asset prices(1)
Sterling fixed interest markets
The Bank of England’s Monetary Policy Committee
(MPC) left the official repo rate unchanged at 4% during
the period, but forward interest rates, as implied by
short sterling contracts and by gilts, fell significantly
(Charts 1 and 2). As of 23 August, the December short
sterling contract implied a rate of 4.01%, down from
5.22% on 17 May, and, according to market participants,
consistent with a central expectation that the official
repo rate would remain at 4% until the end of the year.
Reuters’ polls of economists’ forecasts also showed a fall
in interest rate expectations for the end of 2002
(Chart 3); the poll conducted over 27–29 August(2)
indicated a mean expectation of 3.98%, compared with
Markets and operations
This article reviews developments in sterling fixed income and foreign exchange markets since theSummer Quarterly Bulletin.
" Sterling interest rates have fallen at all maturities, against a background of lower equity prices. " Gilts were included in London Clearing House’s RepoClear service. " On 9 September, Continuous Linked Settlement for foreign exchange was introduced, greatly
reducing settlement risk.
Chart 1Sterling three-month Libor, and expectations from futures contracts
18 July 2002
23 August 2002
20 June 2002
17 May 2002
4.00
4.25
4.50
4.75
5.00
5.25
5.50
5.75
6.00
Three-month£ Libor
Bank of England’srepo rate
Per cent
Expiry dates of contracts
0.00Mar. Sept. Mar. Sept. Mar. Sept. Mar. Sept. Mar. Sept.
2001 02 03 04 05
3.75
(1) The period under review is 17 May (the data cut-off for the previous Quarterly Bulletin) to 23 August.(2) Shortly after the end of the period under review.
Chart 2Three-month forward gilt yields(a)
(a) Gilt yields are derived using the Bank’s VRP curve. For further details see Anderson, N and Sleath, J (1999), ‘New estimates of the real and nominal yield curves’, Bank of England Quarterly Bulletin, November, pages 384–96.
4.0
4.5
5.0
5.5
6.0Per cent
17 May 2002
23 August 2002
0.02004 06 08 10 12 14 16 18 20 22 24
3.5
Chart 3Expectations of economists for the Bank’s repo rate at end-2002
Source: Reuters.
0
10
20
30
40
50
60
70
80
3.50 3.75 4.00 4.25 4.50 4.75 5.00 5.25
Expected repo rate in per cent
Frequency (per cent)
28–29 May27–29 August
250
BBaannkk ooff EEnnggllaanndd QQuuaarrtteerrllyy BBuulllleettiinn:: Autumn 2002
4.63% for the poll conducted over 28–29 May.
Short-term interest rate uncertainty in the United
Kingdom, as inferred from options prices, increased from
mid-July, but fell somewhat towards the end of the
period (Chart 4).
The nominal yields of conventional gilts fell by more
than the real yields of index-linked gilts. In
consequence, implied breakeven inflation rates—the
difference between nominal yields and real yields—also
fell (Chart 5). But market contacts did not suggest that
changes in mean inflation expectations were a
significant factor in explaining the fall in breakeven
inflation rates. Real gilt yields rose during the middle of
the period, perhaps partly in response to several
index-linked corporate issues.
In the period of sharp equity market declines until
late July, movements in market interest rates followed
equity indices closely(1) (Chart 6), reflecting assessments
of the implications for aggregate demand and hence
monetary policy (see the August Inflation Report).
Between 17 May and 30 July, for each 1% fall in
the FTSE 100 the rate implied by the June 2003
short sterling contract fell, on average, by 3.3 basis
points. Among economic data and surveys,
weaker-than-expected RPIX data for May and June also
led to falls in rates implied by futures contracts, as did
the MPC’s decision not to change the Bank of England’s
official rate in June (Table A). Short-term forward
interest rates fell by more than those at longer
maturities (Chart 7).
Chart 4Sterling interest rate uncertainty at various horizons(a)
Sources: LIFFE and Bank of England.
(a) Implied standard deviations of short sterling futures contracts.
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
J F M A M J J A S O N D J F M A M J J A
2001 02
Twelve-month
Percentage points
17 May 2002
Three-month
Six-month
Chart 5International ten-year breakeven inflation rates(a)
(a) Breakeven inflation rates are calculated as the difference between the yield of a conventional bond and the yield of an index-linked government bond with a maturity of approximately ten years. Indexation is based on the following: RPI for the United Kingdom, CPI excluding tobacco for France, HICP excluding tobacco for French index-linked bonds indexed to euro-area inflation, and the CPI Urban index for the United States.
Chart 6FTSE 100 and the rate implied by the June 2003 short sterling contract
Sources: Bloomberg and Bank of England.
0.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
2.6
2.8
3.0
United StatesFrance/ Euro HICP
France
United Kingdom
Per cent
J F M A M J J A S O N D J F M A M J J A2001 02
17 May 2002
1.0
4.25
4.50
4.75
5.00
5.25
5.50
5.75
Regression uses data from 17 May 2002 to 30 July 2002
30 July
Per cent
0.003600 3800 4000 4200 4400 4600 4800 5000 5200 5400
FTSE 100 (index)
4.00
23 August
17 May
Table AMarket interest rate reactions to some economic newsand official publications(a)
Expected Actual Intraday Daily change change (basis (basis points) (b) points) (c)
MPC decision (6/6) n.a. n.a. -6 -6UK industrial production
(m-o-m) (11/6) 0.50% 1.10% 4 5US advance retail sales (m-o-m) (13/6)-0.30% -0.90% -3 -5US Michigan confidence survey
(preliminary) (14/6) 96.50 90.80 -6 -5RPIX (18/6) 2.00% 1.80% -3 -7RPIX (16/7) 1.70% 1.50% -11 -8US GDP (31/7) 2.30% 1.10% -4 -7US ISM manufacturing (1/8) 55.00 50.50 -3 -7UK industrial production (m-o-m)
(5/8) -0.70% -4.30% -4 -11Inflation Report (7/8) n.a. n.a. -1 -9FOMC announcement (13/8) n.a. n.a. -5 -2SEC deadline (14/8) n.a. n.a. 5 14
n.a. = not available.
Source: Bloomberg.
(a) Reactions in rates implied by short sterling futures contracts (September 2002 contract up to 19 June, subsequently December 2002 contract).
(b) Change in rates implied by short sterling from 15 minutes before to 15 minutes after the economic news release, publication of document or start of speech, or for overnight news from closing price to 30 minutes after start of trading the following day.
(c) For overnight news, from closing price on the day of the news to closing price the following day.
(1) In price terms, bonds and equities were highly negatively correlated.
Markets and operations
251
Short-run market interest rates falling in response to
sharply weaker equity indices also characterised euro
and US dollar markets (Charts 8 and 9). The close
comovement of the major international equity indices
(Table B) suggests that they were driven by common
factors—the most obvious immediate trigger being
revelations of accounting irregularities at US companies,
notably WorldCom in May. Investors appear to have
reassessed the reliability of reported earnings often used
as the basis for equity valuations, and perhaps also the
effectiveness of incentives (such as share options) and
controls (such as external audit) designed to reconcile
the interests of corporate managers with those of
shareholders.
These concerns appeared to prompt a more widespread
reappraisal of equity valuations internationally. The
share price rises of the late 1990s had left conventional
valuation measures, such as price-earnings ratios, well
above their historical averages, even after the correction
from the peaks of 2000. As a result of the falls in equity
markets, price-earnings ratios moved closer to their
long-run averages (Chart 10). Prices of equity index
options suggested, however, that market participants’
perceptions of short-term equity market risk increased.
In late July, implied volatilities rose to levels not reached
since the Long-Term Capital Management (LTCM) crisis
of 1998.
Chart 7Cumulative changes in interest rate expectations(a)
(a) ‘Short sterling’ is the three-month Libor implied by December 2002 short sterling contract. Other rates are three-month forward rates, using the Bank’s VRP curve.
160
140
120
100
80
60
40
20
0
20
M J J A
Short sterling
10 years
25 years
2 years
2002
–
+
Basis points
Chart 8International equity indices
Source: Thomson Financial Datastream.
60
70
80
90
100
110
J F M A M J J A
Percentage change 17 May 2002 to 23 August 2002
FTSE All-Share -16%S&P 500 -15%Euro Stoxx -20%Topix -13% 4 Jan. 2002 = 100
FTSEAll-Share
Topix
S&P 500
Euro Stoxx
2002
17 May 2002
Chart 9Cumulative changes in short-term interest rateexpectations(a)
Sources: Bloomberg and Reuters.
(a) As indicated by changes in interest rates implied by futures contracts maturing in December 2002.
160
140
120
100
80
60
40
20
0
20
M J J A
Basis points
2002
Short sterling
Eurodollar
Euribor
–
+
Table BWeekly correlations of changes in international equity indices
Since 1992 2002 17 May to 23 August
All-Share/S&P 500 0.62 0.78 0.80All-Share/Euro Stoxx 0.79 0.87 0.89S&P 500/Euro Stoxx 0.67 0.82 0.79
Chart 10FTSE All-Share and S&P 500 price-earnings ratios(a)
Source: Thomson Financial Datastream.
(a) ‘Earnings’ are those reported over the past year.
0
5
10
15
20
25
30
35
1973 75 77 79 81 83 85 87 89 91 93 95 97 99 2001
S&P 500
FTSE All-Share
S&P 500 average(1973 to present)
FTSE All-Share average
(1973 to present)
Ratio
252
BBaannkk ooff EEnnggllaanndd QQuuaarrtteerrllyy BBuulllleettiinn:: Autumn 2002
Trading in equity and bond markets seems, nevertheless,
to have remained orderly—even during the sharp equity
market falls in the first half of July. There was some talk
of insurance companies selling equities and equity
futures to limit losses in case of further price falls. But
contacts suggested that selling of equities by a number
of UK insurers had occurred steadily over a longer
period in order to reduce the proportion of their
portfolios invested in equities, often in favour of
corporate bonds. Daily turnover of September
FTSE 100 futures increased threefold in the first few
weeks of July, more than in the underlying equity market.
Market participants reported investor purchases of
short-maturity gilts and Treasury bills as ‘safe-haven’
securities in the face of the sharp falls in equity markets.
For example, the 61/2% Treasury 2003 and 8% Treasury
2003 gilts briefly traded at low yields relative to general
collateral (GC) repo rates. However, the spread of
unsecured interbank rates over gilt repo rates did not
widen significantly (Chart 11).
Normalised implied volatilities of ten-year options on
ten-year sterling interest rate swaps (swaptions(1)) rose in
June and July (Chart 12). Long-maturity sterling
swaptions have in the past been used by UK insurance
companies to hedge their exposure to interest rate risk
from having issued guaranteed annuity products.(2)
The increase in long-maturity swaption volatilities in
June and July may partly have reflected actual, or
expected future, buying of long-maturity swaptions by
some UK insurance companies following falls in equity
markets and long-term interest rates. But by the end of
the review period these implied volatilities had eased
back.
From the end of July equity markets were more stable for
a while, but money market interest rates continued to
fall in early August, as market participants interpreted
economic data in the United States and Europe as
indicating that global economic recovery would be
slower than previously expected. Weaker-than-expected
industrial production data for June in the United
Kingdom and GDP data in the United States, including
downward revisions to 2001 GDP data, contributed to
falls in implied rates (Table A).
Sterling forward interest rates at medium and long
maturities fell during the first half of August (Chart 7),
in line with US and European markets. Contacts
suggested that this might have reflected UK institutional
investors extending the maturity of their bond holdings.
Following the passing of the US Securities and Exchange
Commission’s (SEC) 14 August deadline for companies to
certify the accuracy of their financial statements,
equities rose sharply, as did short-term interest rate
expectations. That most chief executives and chief
financial officers of large US companies attested to the
accuracy of their financial statements without further
significant revelations may to some degree have reduced
concerns about the integrity of reported earnings. As
equity indices increased, implied volatilities fell back
somewhat from their late-July highs.
Chart 11Spread of three and six-month sterling interbank rates over GC repo rates(a)
(a) Interbank is the offer rate, GC repo is the bid rate; five-day moving averages.
6
8
10
12
14
16
18
20
0J F M A M J J A S O N D J F M A M J J A
2001 02
Basis points
Six months
Three months
17 May 2002
4
(1) Normalised implied volatilities are the product of implied volatilities of the swaption and the forward swap rateunderlying the swaption. See Financial Stability Review, June 2002, page 24 for a description of swaptions.
(2) See Financial Stability Review, December 2001, pages 152–54.
Chart 12Normalised implied volatilities(a) of ten-year/ten-yearswaptions
Source: Deutsche Bank.
(a) Implied volatilities multiplied by the forward swap rate underlying the swaption.
50
60
70
80
90
100
110Basis points
0J F M A M J J A S O N D J F M A M J J A
2001 02
Sterling
Dollar
Euro
17 May 2002
40
Markets and operations
253
In summary, equities were highly volatile during the
period, both day-to-day and intraday, and, for a period
when the official rate did not change, money market
interest rates were also relatively volatile (Chart 13).
Crucially, though, notwithstanding heightened
uncertainty, market conditions were orderly and there
was no generalised or abrupt ‘flight to quality’ as seen
for example during Autumn 1998 when LTCM failed.
While spreads between sterling swap rates and gilt
yields widened during the review period, they widened
much more sharply during the second half of 1998
(Chart 14).
In the sterling market, the yield spread of the bank
liability curve over gilts widened at all maturities over
the review period, and by most at longer maturities
(Chart 15).(1) Late in the period, the spread became
wide enough to prompt supranational issuance of
fixed-rate sterling debt. Such issuers usually swap their
liabilities back to floating rate, receiving sterling fixed in
a swap. Reflecting this, the issuance triggered a slight
narrowing in the swap spread.
Spreads of sterling corporate bond yields over swap rates
also widened, particularly for sub investment-grade
bonds (Chart 16). For investment-grade bonds, spreads
widened by most on BBB and A-rated bonds,
of which a large proportion was issued by UK
non-financial companies (Chart 17). Spreads on
sterling corporate bonds issued by media and
financial (including insurance) companies widened
most over the period.(2) In contrast, spreads on
mortgage-backed and other asset-backed securities
narrowed slightly.(3)
Issuance in the sterling-denominated non-government
bond market was about £18.5 billion in 2002 Q2
(Table C), compared with about £17 billion in 2002 Q1.
New issues were predominantly long-maturity fixed-rate
bonds. A large proportion was rated AAA, with about
60% backed by mortgages or other assets. Some issues
by UK non-financial companies also carried an
(1) The bank liability curve is a yield curve derived from interbank money market interest rates and interest rate swaps.For more information, see Brooke, M, Cooper, N and Scholtes, C (2000), Bank of England Quarterly Bulletin,November, pages 392–402.
(2) Based on Merrill Lynch Global Index System indices.(3) Based on Merrill Lynch Global Index System indices.
Chart 13Historical standard deviations of FTSE 100 index andimplied rates from short sterling futures contracts(a)
Sources: Bloomberg, LIFFE and Bank of England.
(a) For FTSE 100, calculated as the rolling 60-day standard deviation (annualised) of logarithmic returns. For short sterling, calculated as the rolling 60-day standard deviation (annualised) of percentage change in yield as implied by the mean of the probability density function six months ahead, as derived from options.
0
5
10
15
20
25
30
35
40
45
J A J O J A J O J A J O J A J O J A J
1998 99 2000 01 02
FTSE 100
Short sterling
Per cent
Chart 14Ten-year swap spreads(a)
Source: Bloomberg.
(a) Five-day moving average of the difference between ten-year swap rates and ten-year government bond yields.
20
0
20
40
60
80
100
120
140
J A J O J A J O J A J O J A J O J A J
Basis points
1998 02
United Kingdom
United States
Euro area
17 May 2002
99 2000 01
–
+
Chart 15Spread of bank liability curve over GC repo/gilt curve zero coupon yields(a)
(a) GC repo/gilt yields using the Bank’s VRP curve (see Chart 2). For BLC curve, see footnote 1 on this page.
0.0
0.1
0.2
0.3
0.4
0.5
0.6
2004 06 08 10 12 14 16 18 20 22 24 26
17 May 2002
23 August 2002
Maturity
Percentage points
254
BBaannkk ooff EEnnggllaanndd QQuuaarrtteerrllyy BBuulllleettiinn:: Autumn 2002
AAA-rating as a result of credit enhancement in the form
of a guarantee (or ‘wrap’) from a monoline insurance
company.(1) Total issuance by UK non-financial
companies was little changed on the previous quarter,
while issuance by overseas companies was higher. The
widening of corporate spreads may have deterred
issuance towards the end of the review period,
particularly for lower-rated issuers: issuance between
the start of August and 23 August was lower than the
average for the equivalent period in 1999, 2000 and
2001.
Sterling exchange rates
Between 17 May and 23 August, sterling appreciated by
4.1% against the dollar and depreciated by 1.4% against
the euro and by 1.0% against the yen (Chart 18).
Sterling’s effective exchange rate index (ERI) ended the
period slightly lower, down 0.4%.
Over the review period, the change in the dollar-sterling
exchange rate was broadly consistent with relative
movements in interest rates, but the change in the
euro-sterling exchange rate was less consistent. Table D
Chart 16Sterling corporate bond spreads by credit rating
Source: Merrill Lynch.
Chart 17Composition of Merrill Lynch sterling investment-grade corporate bond indices(a)
Source: Merrill Lynch.
(a) As per August member lists.
0
500
1,000
1,500
2,000
AAA AA A BBB BB B C0
50
100
150
200
Change (left-hand scale)
17 May 2002 (right-hand scale)
23 Aug. 2002 (right-hand scale)
Basis pointsBasis points
Investment grade Sub investment grade
0
10
20
30
40
50
60
70
AAA AA A BBB
UK financial UK securitisedUK non-financial Overseas financialOverseas securitised Overseas non-financial
Per cent
Rating
Table CSterling bond issuance in 2002 Q2DDMMOO ggiilltt aauuccttiioonnss (£ millions)
CCoonnvveennttiioonnaall Date Amount issued Stock29.05.02 2,250 5% Treasury Stock 202525.06.02 3,000 5% Treasury Stock 2008
IInnddeexx--lliinnkkeedd Date Amount issued Stock24.04.02 425 21/2% Index-linked Stock 2020
CCoorrppoorraattee iissssuuaannccee Amount (£ billions)By credit rating:
Number BBB andof issues Total (a) AAA AA A lower
Fixed-rate issuesUK corporates 17 4.1 1.1 0.0 1.4 1.7UK financials 12 1.7 0.4 0.4 0.9 0.1Supranationals 1 0.2 0.2 0.0 0.0 0.0Overseas borrowers 19 6.5 1.4 1.9 2.6 0.7TToottaall (a) 44 99 1122..66 33..11 22..33 44..88 22..44
FRNsUK corporates 2 0.7 0.0 0.0 0.0 0.7UK financials 35 4.5 3.4 0.1 0.9 0.2Supranationals 0 0.0 0.0 0.0 0.0 0.0Overseas borrowers 9 0.8 0.2 0.5 0.2 0.0TToottaall (a) 44 66 66..00 33..55 00..55 11..11 00..99
Sources: Bank of England, Debt Management Office, Moody’s and Standard and Poor’s.
(a) Totals may not sum exactly due to rounding.
(1) For more information, see the box on monoline bond insurers in Rule, D (2001), ‘Risk transfer between banks,insurance companies and capital markets’, Financial Stability Review, December, pages 137–59.
Markets and operations
255
illustrates a decomposition of exchange rate movements
according to the uncovered interest rate parity
condition, which seeks to identify the role of interest
rate news in explaining exchange rate moves.(1) Interest
rate news here is measured as the cumulative expected
return on a ten-year government bond over a ten-year
horizon. In the United States, this measure fell by
almost 4 percentage points more than in the United
Kingdom (11.0 versus 7.3 percentage points
respectively), broadly consistent with the direction and
magnitude of the change in the dollar-sterling exchange
rate. But in the euro area, it fell by only 0.3 percentage
points less than in the United Kingdom (7.0 versus
7.3 percentage points). While this difference was
consistent with the direction of the euro-sterling
exchange rate movement, it was not large enough to
explain its size.
The change in the dollar-sterling exchange rate also
appears to have been broadly consistent with changes in
relative economic growth forecasts, but the change in
the euro-sterling exchange rate was not. Between May
and August, Consensus growth forecasts for the United
Kingdom for 2002 were scaled down by 0.1 percentage
points, compared with 0.5 percentage points for the
United States and 0.2 percentage points for the euro
area.
Several other factors also influenced sterling’s value
against other currencies. The depreciation of the
sterling ERI from 17 May until the end of June
(Chart 18) may have been partly attributable to a
relatively high level of actual and potential merger and
acquisition activity by UK companies abroad. Market
contacts also reported some speculative EMU
convergence trades during this period, particularly short
positions against the Swedish krona, putting pressure on
sterling.
The sterling ERI appreciated throughout July. This was
primarily accounted for by a change in the value of the
dollar, which depreciated against all major currencies.
Market contacts suggested that this in part reflected
renewed concerns about the sustainability of the US
current account deficit and the cross-border capital
flows required to finance it. The dollar’s depreciation
appeared to be linked to falls in equity markets
(Chart 19), and perhaps therefore to increased doubts
about whether US assets would continue to deliver
relatively higher returns.
In the second half of July, sterling appreciated against all
major currencies, but particularly the euro. Market
contacts mentioned as factors the perception of a more
positive economic outlook for the United Kingdom
compared with the countries of the euro area, and the
positive impact on sterling from the unwinding of EMU
Chart 18Sterling exchange rates(a)
95
100
105
110
M
17 May 2002 = 100
£/$
£ ERI
£/€
£/Y
JJ A
2002
(a) A number above 100 indicates sterling appreciation.
Chart 19US dollar and US equity indices
50
60
70
80
90
100
110
J F M A M J J A
4 Jan. 2002 = 100
$/€
2002
Dow Jones
S&P 500
Nasdaq
Percentage change 17 May 2002 to 23 August 2002
Dow Jones -14%S&P 500 -15%Nasdaq -21%$/€ -5%
17 May 2002
(1) The method of decomposing the uncovered interest parity condition to assess the impact of interest rate news on the exchange rate is explained in Brigden, A, Martin, B and Salmon, C (1997), ‘Decomposing exchange rate movements according to the uncovered interest rate parity condition’, Bank of England Quarterly Bulletin, November,pages 377–89.
Table DExchange rate movements and interest rate news: 17 May to 23 August(a)
Percentage points
Sterling ERI Euro-sterling Dollar-sterling Dollar-euro
[A] Actual change -0.37 -1.41 4.12 5.61[B] Interest rate news -0.05 -0.29 3.67 3.96
of which [C] domestic -7.28 -7.28 -7.28 -6.99[D] foreign 7.23 6.99 10.95 10.95
(a) [B] = [C] + [D]. Interest rate calculations use the Bank’s VRP curve. For details, see Chart 2.
256
BBaannkk ooff EEnnggllaanndd QQuuaarrtteerrllyy BBuulllleettiinn:: Autumn 2002
convergence trades, particularly of short positions
against the Swedish krona. In addition, market contacts
reported increased demand for sterling by UK
corporates engaged in active hedging of overseas
earnings.
During August, sterling initially depreciated against
all three major currencies, leading to a fall in the
sterling ERI. This may in part have been attributable
to a strengthening of the US dollar, but also to
weaker-than-expected UK macroeconomic news, such as
industrial production and consumer confidence.
Sterling subsequently rose against both the euro and the
yen, but fell against the dollar, so that the sterling ERI
changed little over the remainder of the period.
On previous occasions when the dollar has depreciated
against the euro, so has sterling. Over this period,
sterling’s depreciation against the euro was broadly
consistent with the historical correlation between the
euro-sterling and euro-dollar exchange rates. Options
prices can give an indication of how closely correlated
the sterling and euro exchange rates are expected to be.
The implied correlation between sterling and the euro
(based on exchange rate movements against the dollar)
rose slightly over the review period at both the
one-month and one-year maturity (Chart 20). The
one-month implied correlation coefficient rose to 0.80
from 0.78, while the one-year implied correlation
coefficient rose to 0.80 from 0.79. Since mid-2000, this
implied correlation had steadily increased, at both
maturities. On 16 July 2002, both measures reached
their highest level since the creation of the single
currency in 1999.
In contrast, the one-month implied correlation
coefficient of sterling with the dollar (based on
exchange rate movements against the euro) fell to 0.59
from 0.77 over the period from 17 May to 23 August,
indicating that market participants expected that
sterling would be less correlated with the dollar in future
(Chart 21).
As in other asset markets, uncertainty increased in
foreign exchange markets, as measured by implied
volatilities derived from options prices (Chart 22). In
April 2002, actual and implied one-month volatilities for
an average of the five most traded currency pairs against
the US dollar(1) fell to their lowest levels since
May 1998 and November 1996, respectively. But
between May and August, actual and implied one-month
volatilities rose by 3.2 and 2.1 percentage points
respectively. The one-month implied volatility for US
Chart 20Implied correlation between the euro and sterling(versus the dollar)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
J A J O J A J O J A J O J A J1999 2000 01 02
One month
One year
Correlation coefficient
Chart 21One-month implied correlations
0.6
0.7
0.8
0.9
1.0
0.0J F M A M J J A
2002
Euro and sterling (versus dollar)
Sterling and dollar (versus euro)
Correlation coefficient17 May 2002
0.5
Chart 22One-month implied and actual exchange rate volatility(a)
0
2
4
6
8
10
12
14
16
18
1995 96 97 98 99 2000 01 02
Per cent
Actual
Implied
(a) For an average of the five most traded currency pairs against the US dollar.
(1) As reported in the Bank for International Settlements’ (BIS) Triennial Central Bank Survey (April 2001), the five mosttraded currency pairs by turnover against the US dollar are the euro, the yen, sterling, the Swiss franc and the Canadiandollar. For further analysis, see the box on ‘Exchange rate volatility’ (2002), Bank of England Quarterly Bulletin,Summer, pages 142–43.
Markets and operations
257
dollar-sterling increased by 2.3 percentage points. In
contrast, the implied one-year volatility of the
euro-sterling exchange rate was unchanged, and the
actual one-year volatility for euro-sterling fell slightly, by
0.4 percentage points (Chart 23).
Against the background of the usual summer lull in
activity in the foreign exchange market, the increase in
implied volatility may have reflected a reduction in
risk-taking. Especially in the latter half of the period,
the most conspicuous actors in the market were thought
to have very short investment horizons, and market
contacts frequently ascribed sharp intraday movements
to model-based traders. Increased uncertainty about the
global growth outlook led many medium and
longer-term speculators to withdraw from the market,
contributing to a relative lack of liquidity in some
currency pairs at times. Against this, trading volumes
have increased over the course of the year, with high
volumes perceived to have been traded on high-volatility
days.
Developments in the structure of sterlingmarkets
The past few months have seen two significant
developments in the sterling market infrastructure, as
well as further developments in instruments and trading
patterns.
Continuous Linked Settlement (CLS)
The Continuous Linked Settlement Bank (CLSB) began
live operations on 9 September, settling foreign
exchange transactions between seven major currencies,
including sterling. The other currencies included from
the start are the Australian dollar, Canadian dollar, euro,
Japanese yen, Swiss franc and US dollar, with more likely
to be added in due course.
The intraday principal exposures entailed in foreign
exchange settlement were highlighted in 1974 by the
failure of Bankhaus Herstatt. The official sector’s
response was to strengthen bank supervision against
internationally agreed standards promulgated by the
Basel Committee on Banking Supervision, which was
established in 1974. During the 1980s and into the
1990s, payment system reforms focused primarily on
strengthening domestic wholesale payments mechanisms
through the introduction of real-time gross settlement
(RTGS). As those agendas made progress, in the
mid-1990s attention returned to curing the remaining
‘Herstatt risk’ problem. In a 1996 report prepared by
the G10 Committee on Payment and Settlement
Systems,(1) central banks set out a remedial strategy, a
key component of which was that private-sector groups
should provide risk-reducing multi-currency settlement
services. CLS has been the main industry response. It is
designed to enable settlement banks to eliminate foreign
exchange settlement risk by settling bought and sold
currencies on a ‘payment-versus-payment’ basis.
CLSB settles foreign exchange transactions in a five-hour
window. It holds accounts with the respective central
banks and uses their RTGS payments systems to make
and receive payments. Settlement members submit
trades to CLSB, and by 6.30 am Central European Time
(CET) are told the net amounts they are due that day to
receive (for currencies in which they are long overall)
and pay for (for currencies in which they are short
overall). Settlement members pay in the net amounts
they owe between 7.00 CET and 12.00 CET, subject to a
schedule set by CLSB, with minimum amounts required
to be paid in by specific times. During this period CLSB
attempts to settle trades individually—this can occur
only if both settlement members have sufficient funds in
their respective accounts to do this. If not, the trade is
sent to the back of a queue and CLSB attempts to settle
the next trade. Each trade is checked until all are
settled and all long balances have been paid out (by
12.00 CET); if funds are insufficient, CLSB cannot settle
the trade. There are a number of safeguards in place in
case a bank fails to pay in the funds it owes, but
ultimately CLSB will eliminate settlement risk only for
those trades it has been able to settle and not
necessarily for all those trades that have been submitted
to it.
Chart 23One-year euro-sterling exchange rate volatility
5
7
9
11
13
15
0J A J O J A J O J A J O J A J
1999 2000 01 02
Per cent
Implied
Actual
3
(1) See http://www.bis.org/publ/cpss17.htm
258
BBaannkk ooff EEnnggllaanndd QQuuaarrtteerrllyy BBuulllleettiinn:: Autumn 2002
At least initially, some foreign exchange transactions are
being settled outside and some through CLSB. As a
result, sterling settlement members could potentially
face imbalances between CLSB pay-in obligations in a
particular currency and receipts relating to transactions
settled outside CLS. The pay-in window for sterling is in
the morning UK time, and CHAPS banks providing
sterling banking services to settlement members may use
their access to intraday liquidity from the Bank of
England against eligible RTGS collateral in order to
bridge any such intraday sterling mismatches. The Bank
of England is monitoring the pattern of demand for such
liquidity.
It is difficult to assess what impact CLS will have
eventually on the broad structure of the foreign
exchange market. At present, 66 shareholder banks own
CLSB. Many will be settlement members and will seek to
sign up non-shareholder banks in order to offer them
third-party settlement services within CLS, so the total
number of banks eventually using CLS, both directly and
indirectly, is potentially large. This could bring cheaper
settlement costs for all market participants. Market
anecdote from some participants suggests that a
differential pricing structure could develop between
trades settled in CLS (and thus not subject to settlement
risk) and those settled outside CLS. There has also been
discussion of the possibility that higher fixed and lower
marginal costs of foreign exchange settlement could lead
to a concentration of business into a smaller number of
global players, with other banks using their pricing and
settlement services for their own clients. The Bank will
keep any such behavioural effects under review.
LCH RepoClear for gilts and SwapClear
On 5 August, the London Clearing House (LCH) added
gilts to its RepoClear service, under which it acts as
central counterparty (CCP) for bond repo transactions
and also for outright purchases.(1) This has the following
effects:
" Balance-sheet netting. When a trade is registered
with LCH, the existing bilateral agreement is
replaced by two new agreements between LCH and
the two banks. As a result, exposures are netted
multilaterally.
" This type of netting also reduces the number of
deliveries, as participants have a single settlement
per security with LCH, rather than with many other
market participants. LCH estimates an average
daily netting efficiency of the order of 65% for
RepoClear.
" Likewise, usage of bilateral credit lines is reduced.
Rather than having many exposures to each other,
participants have margined exposure to LCH.
" Because exposures are to LCH, rather than to other
market participants, anonymous trading is
facilitated.
" LCH provides the option of Straight-Through
Processing (STP), which greatly reduces the need
for ticket writing and paperwork. This, together
with the single counterparty and standardised
contract terms, helps reduce operational risk in
repo.
In general, provided that a CCP is well constructed, with
highly professional and effective risk management, it can
improve the management of some risks within a market,
and make the functioning of the market more resilient
during a crisis.(2)
In the first 14 days of operation, daily volumes averaged
£5.6 billion split between 120 tickets, with an average
maturity of 7.6 days. Automatic trading systems and
voice brokers both had significant proportions of the
business, with a greater proportion of business executed
via voice brokers than in other European government
repo markets cleared through LCH. Money market
participants had already reported that the liquidity at
the short end of the sterling cash yield curve had
deepened following greater use of automatic trading
systems ahead of the introduction of central
counterparty settlement; this may help to explain the
increase in turnover in the gilt repo market in 2002 Q2
shown in Table F. Some say that the benefits outlined
above will enable them to do greater amounts of gilt
repo business, especially at calendar quarter ends. But it
is too soon to judge the significance of any increase in
market depth and liquidity.
In June LCH expanded to 30 years the maturity of
sterling (along with US dollar, euro and yen) interest rate
(1) For more information on RepoClear, see LCH’s web site: http://www.lch.co.uk/RepoClear/BusinessBenefits.htm and http://www.lch.co.uk/press_releases/05082002.htm
(2) See Hills, R and Rule, D (1999), ‘Counterparty credit risk in wholesale payment and settlement systems’, FinancialStability Review, November, pages 98–114.
Markets and operations
259
swaps for which it will act as CCP through SwapClear.(1)
Previously the limit had been ten years. Reflecting
SwapClear’s standardised processes, around 80% of swap
trades cleared by LCH are now confirmed between
counterparties (and cleared) within a day of the trade,
which represents a significant improvement on previous
industry practice.
Money market funds
In addition to these developments in sterling markets,
contacts have commented on the growth of money
market funds. As elsewhere, they invest in money market
assets such as CDs, Treasury bills, repo and commercial
paper. In contrast with the United States, where funds
are also popular with retail investors, sterling money
market funds typically cater for institutional clients, for
example companies, pension funds and local authorities.
They offer such customers an alternative to keeping their
cash balances in demand deposit accounts, the return
on which is typically related to the overnight rate.
The United States has a large domestic money market
mutual fund sector, with assets of $2,238 billion in
2001.(2) Its growth was stimulated in the 1970s as retail
savers switched some of their deposits from banks to
money market mutual funds, as market interest rates rose
above maximum interest rates on time deposits imposed
by Regulation Q.(3) Restrictions on the payment of
interest on companies’ current (checking) accounts may
also have contributed. In contrast to France, where
there are currently no interest-bearing sight deposits, no
such distortion is present in the United Kingdom, so it is
not clear how much sterling money market funds will
grow in future.
Members of the UK-based Institutional Money Market
Funds Association, whose funds are rated AAA, had
assets under management of $22.8 billion in sterling,
$15.5 billion in euros and $65.1 billion in US dollars, as
at 16 August 2002.(4) For comparison, £606 billion was
outstanding in the sterling money market in the United
Kingdom as of end-June 2002 (Table E).
Sterling money markets
More generally, amounts outstanding in the sterling
money markets rose by £30 billion to £606 billion in
2002 Q2, having risen by £35 billion in the previous
quarter (Table E). Within the total, interbank deposits
increased sharply, by around £40 billion, compared with
an increase of under £20 billion over the previous
twelve months. In part, the increase reflected intragroup
activity following a group restructuring. Market contacts
suggest, however, that the increase might also have
reflected increased precautionary investment in
short-term money market assets, including cash
deposits. Non-bank financial institutions’ (such as
pension funds, insurance companies and securities
dealers) sterling deposits with banks in the United
Kingdom increased by £3.6 billion over the quarter, and
by almost £10 billion in June, compared with a fall over
the year to March.
Data collected by the Financial Services Authority for
the sterling stock liquidity regime (SLR), which ensures
that the major UK-incorporated banks match an element
of their potential outflows of sterling liabilities with
holdings of liquid assets,(5) indicate that over the three
months to mid-July, these banks’ net wholesale funds
becoming due over the next five days fell significantly.
Table ESterling money marketsAmounts outstanding: £ billions
Interbank CDs Gilt Stock Eligible Commercial Other TToottaall(a) (a) repo (b) lending (b) bills (a) paper (a) (c)
2000 Q1 156 132 100 51 14 15 6 447744Q2 159 135 124 54 12 16 7 550077Q3 162 125 127 53 12 16 7 550022Q4 151 130 128 62 11 18 9 550099
2001 Q1 171 141 126 67 13 19 7 554444Q2 177 131 128 67 12 22 6 554433Q3 187 134 144 52 11 21 6 555555Q4 185 131 130 48 11 20 16 554411
2002 Q1 190 139 134 66 11 22 14 557766Q2 229 130 144 46 11 26 20 660066
(a) Reporting dates are end-quarters.(b) Reporting dates are end-February for Q1, end-May for Q2, end-August for Q3, end-November for Q4.(c) Including Treasury bills, sell/buy-backs and local authority bills.
(1) See Financial Stability Review, June 2002, page 97.(2) Figure taken from ‘Money Fund Report’, produced by iMoneyNet Inc.(3) Regulation Q was issued by the Federal Reserve. See Eatwell, J et al (1992), The New Palgrave Dictionary of Economics,
Macmillan.(4) Figures taken from ‘European Money Fund Report’, produced by iMoneyNet Inc.(5) See also Chaplin, G, Emblow, A and Michael, I (2000), ‘Banking system liquidity: developments and issues’, Financial
Stability Review, December, pages 93–112, and the Financial Stability Review, June 2002, pages 86–87.
260
BBaannkk ooff EEnnggllaanndd QQuuaarrtteerrllyy BBuulllleettiinn:: Autumn 2002
Market contacts suggested that this partly reflected an
increase in short-term (overnight to five days) interbank
lending, which is netted against short-term wholesale
borrowing for SLR purposes; this would be consistent
with the increase in interbank deposits shown in
Table E. The banks’ apparent preference for lending
funds short term in the interbank market, rather than
holding longer-term money market assets, may have
reflected a reported reluctance to take positions beyond
very short maturities, given the relatively flat money
market yield curve and perceived interest rate
uncertainty.
Major UK-owned banks’ holdings of CDs and bank bills
also fell over the three months to mid-July, and the value
of CDs outstanding in the market as a whole declined
significantly in the three months to end-June (Table E).
To the extent that large UK banks received inflows of
deposits from institutional investors, this may have
reduced their need for CD issuance.
Bank of England official operations
Over the review period, spreads of one-month CD,
interbank and general collateral repo rates averaged 14,
7 and 18 basis points below the Bank’s repo rate
respectively, compared with 10, 9 and 20 basis points
over the year to 17 May. Overnight cash rates almost
entirely remained within the range determined by the
Bank’s collateralised overnight lending and deposit
facilities. The average spread between the Sterling
Overnight Index Average (SONIA) and the Bank’s
repo rate was minus 10 basis points in May, minus
48 basis points in June, minus 25 basis points in
July and plus 16 basis points from 1 to 23 August
(Chart 24).
Volatility, as measured by the standard deviation of the
daily changes in two-week interbank interest rates over a
one-month window, remained broadly constant
throughout 2002, at around 10 basis points.
Open market operations (OMOs)
The stock of money market refinancing held on the
Bank’s balance sheet (comprising the short-term assets
acquired via the Bank’s open market operations) was
slightly higher than in the previous three-month
period (Chart 25). This reflected an increase in
the note circulation, partly as a result of increased
demand associated with the Jubilee weekend and the
World Cup.
The effect of the small increase in the stock of
refinancing was offset by a fall in the rate of turnover of
the stock, leaving the average daily shortage broadly
unchanged. Daily money market shortages averaged
£2.59 billion between May and July, compared with
£2.53 billion during the previous three-month period
(Table G). During May, June and July, counterparties
Chart 24Spread of SONIA, two-week and one-month interbank rates over the Bank’s repo rate
1.5
1.0
0.5
0.0
0.5
1.0
1.5
J A S O N D J F M A M J J A
2001 02
Percentage points
One-month interbank
SONIA
Two-week interbank
–
+
Table FTurnover of money market instrumentsAverage daily amount, £ billions
2001 2002Q1 Q2 Q3 Q4 Q1 Q2
Short sterling futures (a) 60.0 66.0 71.5 69.6 74.1 69.9Gilt repo (b) 15.7 17.9 18.2 20.0 21.3 25.1Interbank (overnight) 10.3 11.1 9.3 10.8 12.4 12.4CDs, bank bills andTreasury bills 11.8 12.4 11.4 11.7 10.5 11.1
Sources: CrestCo, LIFFE, Wholesale Markets Brokers’ Association and Bank of England.
(a) Sum of all 20 contracts extant, converted to equivalent nominal amount.(b) Quarters are to end-February (Q1), end-May (Q2), end-August (Q3) and
end-November (Q4).
Chart 25Stock of money market refinancing and daily shortages
0
2
4
6
8
10
12
14
16
18
20
Q2 Q4 Q2 Q4 Q2 Q4 Q2 Q4 Q2 Q4 Q20.0
0.5
1.0
1.5
2.0
2.5
3.0
Average stock of foreign exchange swaps (left-hand scale)
Average stock of money market refinancing (left-hand scale)
Average daily money market shortage (right-hand scale)
£ billions£ billions
1997 98 99 2000 01 02
Markets and operations
261
refinanced 78% of the daily money market shortages in
the 9.45 am and 2.30 pm rounds of operations (which
largely have a two-week maturity) at the official repo
rate, and 22% in the late rounds of operations, on an
overnight basis and at a spread over the official repo
rate (Chart 26).
Counterparties made use of the Bank’s deposit
facility on three days during the review period. In
order to leave the market square by close of business,
the Bank accordingly increased the amount of
refinancing available to settlement banks at the
4.20 pm late repo facility by the size of these deposits.
On each occasion, the settlement banks borrowed the
full amount of refinancing available. The deposit
facility provides a floor to the interbank overnight rate,
and consequently other short-dated market interest
rates.
Gilts accounted for around £11.5 billion (or 62%) of the
stock of collateral taken by the Bank in its official money
market operations during May, June and July (Chart 27).
Euro-denominated eligible securities(1) (issued by EEA
governments and supranational bodies) accounted for
around £4 billion (or 23%) of the collateral, the same
absolute level as in the previous three-month period.
The increase in the use of bills as OMO collateral
towards the end of the period may partly have reflected
increased Treasury bill issuance by the Debt
Management Office: the stock of Treasury bills
increased from about £8 billion at end-April to about
£13 billion at end-July.
Bank of England euro issues
The Bank of England continued to hold regular monthly
auctions of euro bills during the period. Each month
€900 million of bills were auctioned, comprising
€600 million of three-month and €300 million of
six-month Bank of England euro bills. The stock of euro
bills outstanding on 23 August was €3.6 billion. The
auctions continued to be oversubscribed, with the issues
being covered an average of 6.5 times the amount on
offer; bids were accepted at average yields of between
Euribor minus 8 and 12 basis points.
The Bank of England did not issue any euro notes
during the period under review.
Chart 27Instruments used as OMO collateral
0
2
4
6
8
10
12
14
16
18
20
22
J A J O J A J O J A J
£ billions
BillsEuro-denominated securitiesGilts
2000 01 02
Chart 26Refinancing provided in the Bank’s open marketoperations
0
10
20
30
40
50
60
70
80
90
100
J A J O J A J O J A J
Average overnight refinancing as a percentage of the daily shortages
Average 2.30 pm refinancing as a percentage of the daily shortagesAverage 9.45 am refinancing as a percentage of the daily shortages
2000
Per cent
01 02
Table GAverage daily money market shortages£ billions
1998 Year 1.421999 Year 1.202000 Year 2.022001 Year 2.482002 Q1 2.51
April 2.17May 3.28June 1.92July 2.46
(1) A list of eligible securities is available on the Bank’s web site:www.bankofengland.co.uk/markets/money/eligiblesecurities.htm
262
Introduction
On 6 May 1997, the Monetary Policy Committee (MPC)
of the Bank of England was established and granted
operational independence in setting short-term interest
rates to achieve the government’s inflation target of
2.5%. This new framework replaced the previous system
of a single individual—the Chancellor of the
Exchequer—deciding on the appropriate level of UK
base rates.
Why delegate monetary policy to a committee? The
academic argument for central bank independence is
well established (see, for example, Barro and Gordon
(1983)). And in practice, there is strong evidence from
across the world to suggest that committees are the
preferred arrangement for setting monetary policy by
central banks. For instance, a wide-ranging survey
undertaken by Fry, Julius, Mahadeva, Roger and Sterne
(2000) finds that 79 central banks out of a sample of 88
use some form of committee structure when setting
monetary policy. By weight of numbers, it appears to be
accepted that setting interest rates by committee is
superior. And the intuitive argument that committees
make better decisions than individuals—because they
allow decision-makers to pool judgment—also seems
plausible.
The theoretical economics literature has less to say
about the consequences of delegating responsibility to a
committee: the hypothesis that groups make better
monetary policy decisions is difficult to test, due to a
lack of comparable empirical data. This problem
motivated Blinder and Morgan (2000) to adopt a
different approach: carrying out a ‘laboratory
experiment’ on a large sample of Princeton University
students to test whether groups do indeed make
monetary policy decisions differently.
In an experiment, the researcher can isolate the relative
performance of individual and group behaviour,
controlling for differences in the abilities, incentives and
preferences of the decision-makers, and of the
environment in which they work. The main drawback is
that it is artificial—it is not possible to replicate exactly
the complexities of real-world policy-making in the
context of a simple experiment. But the results may still
be informative when thinking about the arrangements
for monetary policy making.
Although experimental techniques are relatively new to
monetary economics, they are well established in other
branches of economics such as asset pricing, game
theory and decision-making under uncertainty.(2) In
addition, psychologists have studied group behaviour for
Committees versus individuals: an experimental analysisof monetary policy decision-making
This article reports the results of an experimental analysis of monetary policy decision-making underuncertainty. The experiment used a large sample of economically literate undergraduate andpostgraduate students from the London School of Economics to play a simple monetary policy game,both as individuals and in committees of five players. The result—that groups made better decisionsthan individuals—accords with a previous study in the United States with Princeton Universitystudents. The experiment also attempted to establish why group decision-making is superior: althoughsome of the improvement was related to committees using majority voting when making decisions, therewas a significant additional committee benefit associated with members being able to observe eachother’s voting behaviour.
(1) The authors would like to thank the London School of Economics for its help and assistance in this project, inparticular Richard Jackman, Paul Jackson and Gill Wedlake, but also all the LSE students who took part. A longerversion of this paper is published on the same day as this Bulletin in the Bank of England Working Paper series, no. 165.
(2) See Davis and Holt (1993) and Kagel and Roth (1995) for excellent surveys of this literature.
By Clare Lombardelli and James Talbot of the Bank’s Monetary Assessment and Strategy Divisionand James Proudman of the Bank’s Conjunctural Assessment and Projections Division.(1)
Committees versus individuals: an experimental analysis of monetary policy decision-making
263
many years, and a series of experiments—for example,
Hall (1971), Janis (1972) and Myers (1982)—have
shown that group decisions are rarely equal to the sum
of their parts. Group performance depends on the
nature of the interaction between members and the task
in hand, but the consensus view seems to be that for
complex tasks, decisions taken by committee should be
at least as good as the average of the individuals that
comprise it.
This hypothesis was supported by the results of Blinder
and Morgan (2000). In their experiment, groups made
substantially better decisions on average than
individuals. And, just as in real life, there were also
disagreements between committee members over interest
rate decisions. But, contrary to their expectations,
groups did not make decisions more slowly than
individuals.
Examining whether groups make better decisions than
individuals is the main focus of this article. It describes
a new experiment with students from the London School
of Economics, which explored in more detail why groups
are superior. One explanation is that majority voting
helps to eliminate the poor decisions of a minority of
members. But this experiment provided evidence that
committees do more than just this, allowing members to
pool information and—through communicating with
each other—learn more about the game they are
playing. And it explicitly tested whether the ability to
exchange information through discussion improved
performance.
Such a finding would not be surprising if players came to
the experiment with different views about the nature of
the (unknown) model of the economy. So the
experiment tried to examine such differences of opinion
by means of a questionnaire designed to help establish
these prior beliefs. Asking participants to fill in the
questionnaire again at the end indicated how much they
learned about the underlying model during the
experiment.
The rest of this article is organised as follows. The first
section describes the economic model used and the
structure of the experiment; the second section
discusses the results; finally, the article concludes by
trying to draw some inferences from our work for the
design of monetary frameworks in the real world.
Experiment outline
(i) The model
Participants were asked to act as monetary policy makers
by attempting to ‘control’ a simple macroeconomic
model that was subject to randomly generated shocks in
each period, as well as a structural shock that occurred
at some point during the game. The model used in the
experiment (see Appendix 1 for further details) has two
equations—a Phillips curve and an IS curve—and is of a
type that is widely used for policy analysis in modern
macroeconomics (see, for example, Fuhrer and Moore
(1995)). Although the model is stylised, where possible
it was calibrated with a view to matching UK
macroeconomic data (see Bank of England (1999, 2000)
for more details of the calibration of such models).
Players were asked to choose the path for the short-term
interest rate after observing the response of the
endogenous variables—output and inflation—in the
previous period. The model has an ‘optimal policy rule’
(see Appendix 1) that provides a useful benchmark
against which to compare individual and group
decisions.
(ii) Modelling prior beliefs
An intriguing feature of Blinder and Morgan’s (2000)
results was that committee members frequently
disagreed about their decisions, despite having identical
incentives and information. But even without observing
such differences in voting—whether experimentally, or
in real life—it seems entirely plausible that committee
members can think differently about how to respond to
the same economic news.
This should be especially true of a committee where
members have diverse backgrounds and beliefs. At the
beginning of the experiment, players filled in a
questionnaire that attempted to reveal these prior
beliefs.(1) The questionnaire was designed so that
answers could be directly compared with the parameters
of the model and the coefficients of the optimal rule.
During the experiment, players should learn about the
structure of the economy—just like real-world
policy-makers—by observing the response of inflation
and output to changes in interest rates, updating their
prior beliefs, and changing their perception of the
(unknown) actual model accordingly. Participants
(1) See Appendix 2 for a copy of the questionnaire.
264
BBaannkk ooff EEnnggllaanndd QQuuaarrtteerrllyy BBuulllleettiinn:: Autumn 2002
revealed some of what they learned by completing the
same questionnaire again at the end of the experiment.
(iii) Information flows and incentives for players
Players received a clear mandate at the beginning of the
experiment: their objective was to maximise a ‘score’
function that penalised deviations of output and
inflation from their ‘target’ values. The participants
knew that at the end of the game they would be paid in
pounds according to the following formula:
Payoff = 10 + Average score/10
where the score was averaged over the 16 rounds of the
game. The maximum payoff was £20; and the minimum,
£10. In practice, most students earned around
£15–£16.
As in real life, the participants did not know with
certainty the exact structure of the economy they were
attempting to analyse. The only information given to
participants about the model was that it was linear and
broadly characterised the structure of the UK economy.
They were also told that the economy was subject to
random shocks in each period, and that a structural
change occurred at some point during each game. The
challenge for players was to extract the signal from the
noise and change their behaviour accordingly in order
to maximise their score.
(iv) Outline of the experiment
The participants played the game under a number of
different decision-making structures. The sequencing of
the experiment is summarised in Table A below. But first
it is perhaps helpful to define some terminology.
A period refers to a unit of time corresponding to one
interest rate decision and an observation of output and
inflation. The players were also given a score in each
period. Each round consisted of ten periods. At the end
of each round, individuals were given a final score
(corresponding to the average score over its ten
constituent periods), the game was reset to its initial
state and the next round (ten-period game) would begin.
There were four rounds in a stage. Stage 1
corresponded to four individual rounds (numbered
1–4). In Stages 2 and 3 (rounds 5–8 and rounds 9–12
respectively) individuals set interest rates together in
committees of five players. Some committees were
allowed to discuss their decisions in Stage 2 while others
were not. Those arrangements were reversed in Stage 3.
Stage 4 (rounds 13–16) consisted of a further four
individual rounds, with participants playing separate
games.
Participants were allocated into groups of five. They
were given a set of instructions and asked to fill in
the questionnaire. Players had about ten minutes to
practise on their own with the actual version of the
game used in the experiment before starting to play ‘for
real’. In each period participants had to decide what
interest rate to set in response to developments in the
‘economy’.
In the first stage, the participants acted as individual
policy-makers, playing separate games on separate
computers for four rounds. Beginning with round 1, the
game started at period 1, where participants decided
on the appropriate level for the interest rate after
observing the initial values of inflation and output with
a one-period lag.(1) This vote was entered into the
computer and the game proceeded to period 2. The
computer displayed output and inflation outturns for
period 1, along with the score for that round and the
interest rate decision. The same process was repeated
until the game reached period 10. At this point, players
were told their average score for round 1, the game was
reset, and play continued, for a further three rounds.
The committee phase was played in two stages (Stages 2
and 3 in Table A above). Stage 2 began at round 5. The
five players observed the same information in each
period—the level of output and inflation of the previous
period(s) as well as the history of interest rates and
scores—and entered their votes while sitting at separate
computers. But this time, in each period, the computer
Table AThe structure of the monetary policy experimentRead instructions sheet
Fill in ‘Priors Questionnaire’
Practice rounds No score recorded
Stage 1: Rounds 1–4 Played as individuals
Stage 2: Rounds 5–8 Played as a group (i) No discussion(ii) With discussion
Stage 3: Rounds 9–12 Played as a group (i) With discussion(ii) No discussion
Stage 4: Rounds 13–16 Played as individuals
Fill in ‘Priors Questionnaire’
Students are paid according to their average score across the four stages
(1) Inflation and output would always be close to their target values initially.
Committees versus individuals: an experimental analysis of monetary policy decision-making
265
selected, and then set, the median vote of the group (as
a proxy for a majority-voting rule). Participants
observed this committee decision, as well as the
response of output and inflation to it. They also saw the
(unattributed) votes of their fellow committee members
and overall score for the period and the round so far.
Again each round lasted for ten periods. Stage 2
finished in round 8.
The committees were divided into two sets. For one set,
discussion among members of the committee was not
allowed in stage 2. For the other set, discussion was
permitted. In stage 3 the organisation of stage 2 was
reversed. The ordering of the discussion and no
discussion games was organised in this way to control
for learning.
Stage 4 (rounds 13–16) served as another control
mechanism, to ensure that the comparison between
individual and committee play was not affected by the
fact that participants had had four (or more) individual
rounds to learn before entering the committee stage. By
returning to individual play at the end of the
experiment, it was possible to verify that the
improvement in scores during the committee stages
(rounds 5–12) was not just an extension of the learning
trend observed in rounds 1–4.
(v) The data
The experiment was conducted on ten evenings between
12 November and 11 December 2001 at the London
School of Economics. Participation in the experiment
was voluntary with 170 students taking part in 34
independent experiments: that is to say 34 committees
with 16 score observations for each.(1)
Chart 1 shows a breakdown of the participants by
course studied: half of the students were postgraduate
economists. And although a small minority (5%) was not
currently studying an economics-related discipline, all
students had taken at least one undergraduate-level
economics course.
Results
The main focus of the experiment was to provide
evidence on the differences between group and
individual policy-making and to offer some insight into
explaining these differences. But indirectly, the results
also allow some analysis of what players learned
about both the model and how to play the game over
time.
(i) Learning about the model
Players’ answers to the initial questionnaire gave some
insight into their prior beliefs about the structure of the
economy. All answers to the questionnaire were in
numeric form, and each question was related to either
the parameters of the model, or the associated optimal
rule (see Appendix 1 for details).
Participants also filled in the same questionnaire again
at the end of the experiment. One test of learning is
therefore the extent of convergence in these views
towards the actual parameter values over the course of
the game.(2) To this end, a useful statistic is the mean
squared error (MSE). The MSE is calculated as the
average of the squared errors made by each player when
responding to each question.
Over all players and questions, the total MSE statistic
decreased; from 0.17 in the initial questionnaire to 0.15
at the end of the experiment. This fall is significant at
the 1% level—suggesting that players’ responses were
closer to the actual parameters of the model at the end
of the experiment. The standard deviation of responses
to the questionnaire also narrowed significantly (at the
1% level) from 1.59 to 1.45, suggesting some
convergence of views among players.
Can we decompose this improvement further? Chart 2
shows the change in MSE for individual questions: the
5%
50%
31%
Postgraduate economicsPostgraduate non-economicsUndergraduate economics (final year)Undergraduate economics (second year)
14%
Chart 1A breakdown of players by course studied
(1) A further 15 students participated in an alternative version of the experiment described below.(2) See Lombardelli, Proudman and Talbot (2002) for a discussion of how the responses to the questionnaire can be
compared with the parameters of the model and the optimal rule.
266
BBaannkk ooff EEnnggllaanndd QQuuaarrtteerrllyy BBuulllleettiinn:: Autumn 2002
dashed lines represent the reduction in error required
for a significant improvement (at the 5% level) in
response to each of the questions. This implies that
participants learned most about the lags in the
transmission mechanism of monetary policy (Q2) and
the weight they should attach to deviations of output
from trend in their ‘rule’ (Q3). The change in response
to the other questions was more mixed. Participants did
less well at working out the parameters of the model
(Q4–8)—particularly how much impact interest rate
changes have on output (Q5) and the long-run impact of
output on inflation (Q8). But each game may have been
too short to learn much about these aspects—especially
the long-run neutrality property of the model. There
was also a fall in the MSE of responses to the question
on how cautious monetary policy makers should be
when setting interest rates (Q1), but this was not
significant.
(ii) Learning about the game
The results of the questionnaire provide tentative
evidence of learning about certain aspects of the model
and the nature of the optimal rule, but does this mean
that players actually became better at playing the game
over time?
Chart 3 below shows a summary of the mean scores
attained by the 34 committees over time. This is broken
down into the first set of individual play (rounds 1–4),
committee play (rounds 5–12) and then individual play
for a second time (rounds 13–16). For the individual
rounds, the ‘committee’ score is taken to be the mean of
the scores across the five individuals playing separately.
For the committee rounds, this statistic is the mean
score from committee decisions.
There are three striking features of the data:
(i) the significant upward trend in the results over
time—indicative of the learning that occurred
during the game;
(ii) the large rise in scores when players moved to
committee decision-making in round 5; and
(iii) the large downward move in scores when
participants returned to playing as individuals in
round 13.
The dispersion of scores in any given round—measured
by the standard deviation across players—more than
halved during the game from 76 in round 1 to 35 in
round 16. This suggests that the worst players learned
most about the game: those who performed poorly in
the first rounds got disproportionately better.
Chart 4 shows that it was not just the worst players who
learned during the course of the experiment. We can
rank the five players in each committee by their initial
performance (in rounds 1–4), and calculate how much
they improved by the final round. Although the worst
players learned most, and only the worst two players in
each committee made a significant improvement (again
the dashed lines represent 5% significance levels), even
the best players improved somewhat by the end of the
game.
(iii) Group versus individual performance
There was strong evidence that decisions taken by
committees were superior to those of individuals.
0.08
0.06
0.04
0.02
0.00
0.02
0.04
0.06
0.08
0.10
Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8
Lower mean squared error in final questionnaire
Reduction in meansquared error (b)
+
–
Chart 2Reduction in mean squared error of responses between the initial and the final questionnaire(a)
(a) See Appendix 2 for details of the questionnaire.(b) Dashed lines indicate significance at the 5% level.
0
10
20
30
40
50
60
70
80
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Individual games
Committee games
Individual games
Round
Average score
Score
Chart 3Average committee scores over time
Committees versus individuals: an experimental analysis of monetary policy decision-making
267
The average committee score of 68 (over rounds 5–12
in Chart 3 above) was nearly two-thirds better than
the average score of 41 for the individual rounds
(rounds 1–4 and 13–16 in Chart 3). And this increase
in score is significant at the 1% level.
To give some idea of the scale of the improvement, the
average score of someone following the optimal rule (see
Appendix 1) would be 85, much higher than the best
individual player’s score (71), but only slightly better
than the best committee (83). On average, moving from
individual decision-making to a committee structure
closed nearly two-thirds of the ‘policy gap’.
What explains this improvement in committee
performance? There are (at least) two distinct,
competing hypotheses that can be used to explain why
committee decisions are superior to those of the
individuals that comprise it:
Hypothesis 1: A committee with ‘majority’ voting can
neutralise the impact of some members playing badly in
any given game.
Hypothesis 2: Committees allow members to improve
performance by sharing information and learning from
each other.
Chart 5 shows a visual representation of the
contribution of these two hypotheses to the
improvement of committees over individuals. The
blue line represents the average—over the 34
independent groups of five players—of the median
player’s score. The red line is simply the mean score
across all players in each committee.(1) Line C is the
mean score over all the committee rounds and line D is
the mean score over rounds 13–16 for the median
players in each of those rounds. The overall
improvement in performance—generated by setting
interest rates by committee—is therefore measured as
the distance between C and A: the difference between
the average score in the final individual round and the
committee rounds.
The chart decomposes this improvement into two
distinct components. The difference between the
score of the mean and median player in the individual
rounds (the distance B–A in Chart 5) represents the
adverse effect of a minority of poor performers on
the mean individual score. This is therefore the
extent of improvement under Hypothesis 1 described
above. And this portion of the difference in means is
significant at the 1% level. So Hypothesis 1 cannot be
rejected.
The remainder should represent the contribution of
Hypothesis 2: C–B (the portion of the committee
improvement not explained by the move to majority
voting). This difference is also significant at the 1%
level, so Hypothesis 2 cannot be rejected either.
Another striking feature of both these results and those
of Blinder and Morgan (2000) was the significant
decline in scores as participants move back to individual
play, in this case at the end of round 12. By definition,
this component of the committee improvement
(represented by distance C–D in Chart 5) cannot be
0
10
20
30
40
50
60
70
1st 2nd 3rd 4th 5th
Improvement in scores betweenrounds 1–4 and 16 (a)
Players ranked in order of performance in rounds 1–4
Chart 4Improvements in scores for players ranked by initial performance
(a) Dashed lines indicate significance at the 5% level.
0
10
20
30
40
50
60
70
80
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Score
DC
A
B
Round
Mean scores
Median scores
Individual games
Committee games
Individual games
0
Chart 5Mean and median scores for committee members
(1) Note that the mean score in the committee rounds is the score of the committee’s interest rate decision.
268
BBaannkk ooff EEnnggllaanndd QQuuaarrtteerrllyy BBuulllleettiinn:: Autumn 2002
associated with learning about the game over time,
because players know at least as much about the model
in round 13 as they did before. So it therefore seems
likely that this residual effect stems from the ability of
committees to pool judgment, expertise and skill. This
fall in scores is significant at the 1% level too: in other
words, there is ‘something special’ about committees in
addition to their ability to aid learning and to strip out
the effects of ‘bad’ play.
Further evidence that a committee is more than just the
sum of its parts is shown by asking whether the
performance in the committee stages was better than
the mean score of the best individual in each committee
when playing alone. The mean committee score (68)
was somewhat higher than the mean score of the
best individual (65) (this difference is significant at
the 10% level), providing evidence that committees did
more than just replicate the behaviour of their best
individual.
(iv) What makes a good committee?
If committees improve decision-making by exploiting
their members’ ability to pool information and
knowledge and to learn from each other, communication
must be key. As discussed earlier, the experiment
included two different ways of organising committee
decision-making: one where participants were
allowed to discuss their views and another where
no verbal communication was allowed. Perhaps
the most surprising result was that the ability to
discuss did not significantly improve committee
performance.
This result was in contrast to earlier trials of the game
on Bank staff. So, in addition to the main experiment
described above, a further small sample of students was
asked to play a different version of the game as a
robustness check. This variant was designed so as to
raise the implicit benefit of discussion: committee
members were told—with a lag of up to two periods—
that a shock had occurred, and the length of this
information lag was allowed to vary across players. The
ability to discuss was therefore more valuable because
committee members with more timely information could
share this with others more quickly by verbal
communication. The average score of discussion
committees was higher than for non-discussion in this
version of the game, although the small sample size—
three committees—meant that the significance of this
improvement could not be tested.
Another hypothesis is that people can communicate in
different ways. And the benefits of different forms of
communication are likely to depend on the nature of the
game, as well as the individuals taking part. There are
many games—for example snooker or chess—that may
be easier to learn by watching, rather than through
discussion. But for the main version of the game, and
for this set of students, discussion did not provide more
information than could be acquired by observing others’
votes.
There is also some evidence from the psychology
literature that discussion may not always enhance group
performance. The idea of ‘group polarisation’—as
proposed by Myers (1982)—suggests that discussion
tends to polarise any initial tendency within the group.
This is because people have an innate desire to compare
themselves favourably with each other, and so take
increasingly extreme positions in favour of the initial
group proposition. One way around this problem is to
ensure that a frank and open exchange of views takes
place at the beginning of the discussion—as outlined in
an earlier study by Hall (1971) who showed that groups
who established a common consensus quickly were often
less effective.
So if discussion did not help committees to improve
their scores in our experiment, what sort of behaviour
does? Lombardelli, Proudman and Talbot (2002)
explored this question in more detail, using an
econometric analysis to model scores over time
and across committees. After controlling for
committee-specific features—such as the innate ability
of participants to play the game—the model captured
the upward trend in scores over time, and the rise in
scores during the committee stages. Committee scores
were positively related to the period in which the
structural shock occurred in each round. Intuitively, the
earlier in the game the structural shock took place, the
more difficult the economy was to control over the
remainder of the game—particularly if it took some time
for the player to recognise that such a shock had
occurred. Higher interest rate activism—as measured
by the standard deviation of the interest rate in each
ten-period game—was associated with lower scores for
both individuals and committees.
But on the whole, the econometric analysis reinforced
the results presented above: that committees performed
significantly better than individuals, and that there was
some evidence of participants learning about the game
over time.
Committees versus individuals: an experimental analysis of monetary policy decision-making
269
Conclusions
This article discusses an experimental analysis of
monetary policy decision-making. Although such a
stylised experiment can never hope to capture fully the
complexity of the decision problem faced by real-world
policy-makers, the results provide evidence that the
decisions made by committees were superior to those of
a single individual. And there is also evidence to suggest
that committee performance was, on average, better than
the performance of the best individual. This suggests
that the real-world preference for setting interest rates
by committee is justified.
The experiment also tried to examine why committee
decisions were superior to those of individuals. A
significant portion of the improvement could be
attributed to the process of majority voting. But there
was also evidence that there is something ‘special’ about
committees beyond their ability to strip out the effect of
bad play. The ability of committees to allow the pooling
of judgment and information (in whatever form) means
that a group can be more than just the sum of its parts.
Perhaps surprisingly, committees who were able to
discuss their decisions did not perform better than
those where discussion was not allowed. It seems that,
in the experiment, it was possible to glean the same
amount of information by observing the votes of other
committee members. But, as noted above, real-world
policy-making is undoubtedly a more complex affair. The
Monetary Policy Committee takes into account a much
wider range of data and information than just lagged
inflation and output when making its monthly interest
rate decision.
What this simple experiment has shown is that it is not
enough simply to take a majority decision among
fixed views that have been reached in isolation. The
pooling of knowledge among committee members—in
whatever form—is one important reason why group
decision-making is superior. And that reflects one
feature of the practical operation of the Monetary
Policy Committee—the exchange of views among
the group helps to determine the votes of each
individual.
270
BBaannkk ooff EEnnggllaanndd QQuuaarrtteerrllyy BBuulllleettiinn:: Autumn 2002
Appendix 1The model and a derivation of the ‘optimal rule’
Although participants were not provided with the underlying equations the model can be described by the following
equations:
yt – y* = 0.8(yt–1 – y*) – 0.5(Rt – pt – r*) + g– + ht ((11))
pt = 0.7pt–1 + 0.3pt–2 + 0.2(yt – y*) + nt ((22))
where: yt is log output, y* is the log of the natural rate of output (calibrated arbitrarily to 5), pt is inflation, Rt is the
nominal interest rate and r* is the neutral real interest rate (calibrated to 3% per year). g– is a permanent shock, ht and
nt are shocks corresponding to a random draw from a normal distribution ~ N(0, 0.01) in each period.
Equation ((11)) is an ‘IS curve’. The current deviation of output from its natural rate (yt – y*) is a function of its
one-period lag, and the deviation of the real interest rate from its neutral level in the current period (Rt – pt – r*).
Equation ((22)) is a ‘Phillips curve’. Inflation is a function of lagged values of itself and the current deviation of output
from its natural rate. The coefficients on lagged inflation sum to one, reflecting the fact that although a short-run
trade-off between output and inflation may exist, the Phillips curve is vertical in the long run.
Assuming that players attempt to maximise their score (St) in each period of the game, the decision problem of each
player can be written as:
MaxEt–1{St} s.t. ((11)) and ((22)) where: St = 100 – 40|yt – y*| – 40|pt – p*| ((33))rt
where p* is the inflation target, calibrated to 2.5%.
Approximating ((33)) as a linear quadratic, we derive the optimal rule by substituting in the constraints ((11)) and ((22)) and
differentiating with respect to rt to give:
rt = 1.6yt–1 + 0.27pt–1 + 0.115pt–2 + 2g– ((44))
Obviously, the distribution of g– is also unknown to participants in the experiment, so ((44)) is the ‘certainty equivalent
optimal rule’. Svensson and Woodford (2000) note that—under the assumption that the loss function is quadratic—
the optimal policy rule under partial information is the same as its full-information counterpart. We use this optimal
rule to calibrate the ‘ideal’ responses to the questionnaire and to conduct simulations of the model—see Lombardelli,
Proudman and Talbot (2002) for further details.
Committees versus individuals: an experimental analysis of monetary policy decision-making
271
Appendix 2Prior beliefs questionnaire
Date: Group:
PPlleeaassee ssppeenndd aa ffeeww mmiinnuutteess ffiill lliinngg iinn tthhiiss qquueessttiioonnnnaaiirree,, ccoonncceennttrraattiinngg iinn ppaarrttiiccuullaarr oonn tthhee
qquueessttiioonnss iinn iittaalliiccss.. IItt ddooeessnn’’tt mmaatttteerr iiff yyoouu aarree nnoott ffaammiilliiaarr wwiitthh tthhee jjaarrggoonn iinn bbrraacckkeettss:: tthhiiss iiss
mmeerreellyy ttoo hheellpp uuss ccaalliibbrraattee yyoouurr rreessppoonnssee..
BBEE AASS HHOONNEESSTT AASS YYOOUU CCAANN——TTHHEERREE AARREE NNOO RRIIGGHHTT OORR WWRROONNGG AANNSSWWEERRSS!!
What is your player number?
1) To what extent should monetary policy makers respond cautiously to shocks (ie if their interest rate reaction
function includes the following expression it = ait–1 + ...., what weight should they place on a)?
Not at all cautiously (ie a = 0) Very cautiously (ie a = 1)
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
2) After how many quarters is the maximum impact of monetary policy on inflation felt?
0 1 2 3 4 5 6 7 8 9 10
3) What relative weight should monetary policy makers place on smoothing output compared with controlling
inflation (ie if their reaction function includes the following expression it = a(yt – Y) + (1 – a)(pt –p*) + ...., what weight
should they place on a)?
None (ie a = 0) All (ie a = 1)
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
4) To what extent are shocks to output persistent (ie if the expression for output included the following term
yt = ayt–1 + ...., what weight do you think a would take)?
Not at all persistent (ie a = 0) Completely persistent (ie a = 1)
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
272
BBaannkk ooff EEnnggllaanndd QQuuaarrtteerrllyy BBuulllleettiinn:: Autumn 2002
5) How sensitive is output to changes in interest rates (ie if the expression for output included the following term
yt = ait + ...., what weight do you think a would take)?
Not at all sensitive (ie a = 0) Very sensitive (ie a = 1)
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
6) To what extent are shocks to inflation persistent (ie if the expression for inflation included the following term
pt = apt–1 + ...., what weight do you think a would take)?
Not at all persistent (ie a = 0) Completely persistent (ie a = 1)
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
7) To what extent is inflation sensitive to deviations of output from trend in the short run (ie if the expression for
inflation included the following term pt = a(yt–1 – Y) + ...., what weight do you think a would take)?
Not at all sensitive (ie a = 0) Highly sensitive (ie a = 1)
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
8) To what extent is inflation sensitive to deviations of output from trend in the long run?
Not at all sensitive (ie a = 0) Highly sensitive (ie a = 1)
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
9) What course are you studying?
………………………………………………………………………………………………………………….
10) Are you….
Undergraduate: 2nd year Undergraduate: 3rd year Graduate student
Committees versus individuals: an experimental analysis of monetary policy decision-making
273
References
BBaannkk ooff EEnnggllaanndd ((11999999)), Economic models at the Bank of England.
BBaannkk ooff EEnnggllaanndd ((22000000)), Economic models at the Bank of England, September 2000 update.
BBaarrrroo,, RR JJ aanndd GGoorrddoonn,, DD BB ((11998833)), ‘Rules, discretion and reputation in a model of monetary policy’, Journal of
Monetary Economics, Vol. 12, No. 1, pages 101–21.
BBlliinnddeerr,, AA SS aanndd MMoorrggaann,, JJ ((22000000)), ‘Are two heads better than one: an experimental analysis of group vs
individual decision making’, NBER Working Paper, No. 7909, September.
DDaavviiss,, DD DD aanndd HHoolltt,, CC AA ((11999933)), Experimental economics, Princeton University Press.
FFrryy,, MM,, JJuulliiuuss,, DD,, MMaahhaaddeevvaa,, LL,, RRooggeerr,, SS aanndd SStteerrnnee,, GG ((22000000)), ‘Key issues in the choice of monetary policy
framework’, in Mahadeva, L and Sterne, G (eds), Monetary frameworks in a global context, Routledge.
FFuuhhrreerr,, JJ CC aanndd MMoooorree,, GG RR ((11999955)), ‘Inflation persistence’, Quarterly Journal of Economics, Vol. 110, No. 1,
pages 127–59.
HHaallll ,, JJ ((11997711)), ‘Decisions, decisions, decisions’, Psychology Today, November.
JJaanniiss,, II LL ((11997722), Victims of groupthink, Boston: Houghton Mifflin.
KKaaggeell ,, JJ HH aanndd RRootthh,, AA EE ((11999955)), The handbook of experimental economics, Princeton University Press.
LLoommbbaarrddeellllii ,, CC,, PPrroouuddmmaann,, AA JJ aanndd TTaallbboott,, JJ II ((22000022)), ‘Committees versus individuals: an experimental
analysis of monetary policy decision-making’, Bank of England Working Paper no. 165.
MMyyeerrss,, DD GG ((11998822)), ‘Polarizing effects of social comparison’, in Bransdatter, H, Davis, J H and Stocker-Kreichgauer, G
(eds), Group decision-making, New York: Academic Press.
SSvveennssssoonn,, LL EE OO aanndd WWooooddffoorrdd,, MM ((22000000)), ‘Indicator variables for optimal policy’, NBER Working Paper,
No. 7953, October.
274
Introduction
A number of countries have considered how to achieve
an appropriate level of scrutiny of the conduct of
monetary policy within a framework of central bank
independence. There are a variety of ways in which such
scrutiny can be exercised (for example through the press
or by the legislature). In a parliamentary democracy—
where it is for the parliament to hold the executive to
account—an important method will be through the
appearance of central bankers in front of parliament or
its representatives.
Parliaments may call central bank officials to account for
their monetary policy actions at regular calendar
intervals. In some countries, predetermined
appearances may be supplemented by additional
appearances should these be warranted by economic
conditions. And in several inflation-targeting
frameworks, there exist predefined conditions under
which central banks account for their actions.
Using results from a specially constructed survey (see
Annex 1), this article examines in detail the
parliamentary scrutiny of central banks. It quantifies (i)
how frequently parliamentary committees call central
bank officials in front of them; (ii) how often these
appearances are to discuss monetary policy; and (iii) the
level of technical support provided to the parliamentary
committee in advance of each hearing. And it asks more
qualitative questions on (iv) how the number of
appearances of central bank officials before parliament
is decided; (v) who is responsible for appointing the
policy-making board of the central bank; and (vi) who in
the central bank is responsible for monetary policy. The
article also uses information on recognised procedures
that are undertaken when a target is missed, using data
from Fry et al (2000).
Issues in defining and measuringaccountability and parliamentary scrutiny
The concepts of transparency and accountability have
become a focus among policy-makers and academic
researchers in recent years. One aspect of
accountability is the formal duty to justify what has been
done. In its Codes of Good Practice on Transparency
(Section IV), the IMF (2000) argues that ‘Officials of the
central bank should be available to appear before a
designated public authority to report on the conduct of
monetary policy, explain the policy objective(s) of their
institution, describe their performance in achieving
Parliamentary scrutiny of central banks in the UnitedKingdom and overseas
This article reviews the parliamentary(3) scrutiny of central banks in 14 countries using the results from anew survey. There is wide variation in the nature of parliamentary scrutiny within the sample. There isno firm evidence in these data, however, to suggest that particular types of framework are associatedwith different overall levels of parliamentary scrutiny. The Bank of Japan, Bank of England, EuropeanCentral Bank (ECB) and Federal Reserve each make higher-than-average appearances before theirrespective parliaments to discuss monetary policy issues, and the technical support provided to therelevant committees is relatively high in the US Congress and in the European Parliament. The level ofscrutiny can be circumstance specific, and some inflation-targeting frameworks have defined specificconditions that would trigger scrutiny and the form it would take.
By Jonathan Lepper, formerly of the Secretariat to the House of Commons Treasury Committee(1)
and Gabriel Sterne of the Bank’s International Economic Analysis Division.(2)
(1) Currently Economic Adviser, HM Treasury.(2) The authors are very grateful to all those who completed the questionnaires and checked the compiled responses for
their countries. These included the staff of the various central banks, parliaments and British Embassies in thecountries surveyed. The views in this article and any mistakes in the data reported are, however, solely theresponsibility of the authors and not those of the Bank of England, HM Treasury, or any of the survey respondents.
(3) The term ‘parliament’ is used throughout the paper as a generic reference to the law-making assembly of a country.
Parliamentary scrutiny of central banks in the United Kingdom and overseas
275
their objective(s), and as appropriate, exchange views on
the state of the economy and the financial system.’ This
approach is in line with the aims set out by the UK
[House of Commons] Treasury Committee, which, in its
1997 report on the ‘Accountability of the Bank of
England’, examined how it might best hold the MPC to
account and concluded that:
‘… by bringing information into the public domain we
can help clarify the thinking and actions of those
responsible for the formulation and delivery of monetary
policy and the rigorous scrutiny of the basis for policy
decisions will enhance the credibility and effectiveness
of the monetary framework as a whole.’
Parliamentary scrutiny is an important aspect of central
bank accountability. There is no single definition of
parliamentary accountability of central banks, although
a number of authors have defined and measured aspects
of accountability. Briault, Haldane and King (1996)
suggest that both legal aspects of accountability and
more subtle forms of accountability or transparency may
be important. In a similar vein, De Haan, Amtenbrink
and Eijffinger (1999) and De Haan and Eijffinger (2000)
define central bank accountability to have three main
features: the explicit definition and ranking of
objectives; the transparency of monetary policy; and
who bears final responsibility for monetary policy. In
presenting a comprehensive index of transparency in
nine central banks, Eijffinger and Geraats (2002) discuss
a sub-index of political transparency that includes
measures consistent with broadly accepted notions of
accountability, whereas Fry et al (2000) provide a
number of measures of parliamentary accountability.
Some of these studies have analysed whether the central
bank is subject to monitoring by parliament, though
none in such detail as presented here.
Survey method
The analysis in this article is based on a small survey of
14 countries. We sent a short questionnaire(1) to each
country asking for details about the number of
parliamentary hearings held with central bank officials
in the year to May 2001, the proportion of these
hearings related to monetary policy, the proportion
attended by the head of the central bank, and whether
the parliamentary committee conducting the hearings
had the power to veto appointments to the monetary
policy board. The questionnaire also requested details
on the number of technical staff available to the
parliamentary committee and whether they required
additional advice from outside experts. In addition,
the questionnaire asked for supplementary details
about the method of appointment of the policy-setting
committee in each central bank. We also report
results on measures of scrutiny that may be triggered if
the central bank misses its target. Some information
was taken from results published in Fry et al (2000) and
has been revised and extended in this survey. In
aggregate, the present results provide more detail than
previously, though we recognise that it is impossible
to measure the precise quality of parliamentary
scrutiny with such summary information. We
subsequently asked each central bank to check an
earlier draft of the article for factual accuracy and
general comments.
The sample chosen includes various monetary
frameworks. Monetary policy instruments are set
independently of government in all countries in our
sample. Government(2) is to varying degrees involved in
setting numerical targets for inflation in all frameworks,
except in the United States and Japan, while in the euro
area, the Treaty establishing the Community specifies a
mandate for price stability and the ECB has quantified
the objective. The ECB is accountable to the European
Parliament for monetary policy actions affecting all
twelve member countries, including France, Germany
and Italy, so the inclusion of these countries in the
sample is not intended to represent a direct comparison
in terms of the overall level of parliamentary scrutiny of
monetary policy in these countries.
Results
(i) Quantity of parliamentary appearances
In the majority of countries surveyed a minimum
number of appearances before parliament is either laid
out in statute or determined by a formal agreement
between the parliament and the central bank. For
example, Article 40 of the European Parliament rules of
procedure states that the ECB President shall appear at
least four times a year. In most of these cases,
parliament also has the option of holding additional
hearings. For instance, Article 113 of the Treaty
stipulates that ‘the President of the ECB and the other
members of the Executive Board may, at the request of
(1) See Annex 1 for full questionnaire.(2) Government is taken to be the executive policy-making body of a state, parliament is the legislative authority.
276
BBaannkk ooff EEnnggllaanndd QQuuaarrtteerrllyy BBuulllleettiinn:: Autumn 2002
the European Parliament or on their own initiative, be
heard by the competent committees of the European
Parliament.’
In four countries (Canada, New Zealand, Norway and the
United Kingdom), there is no statutory number of
appearances and the decision on the number rests with
parliament. In the United Kingdom, the Governor and
members of the Monetary Policy Committee (MPC) are
regularly invited to appear in front of Select Committees
of both the lower and upper Houses of Parliament to
discuss monetary policy issues. Appearances in front of
the House of Commons Treasury Committee usually
follow publication of the February, May and November
Inflation Report, although the Treasury Committee
reserves the right to call the Governor more often should
economic conditions warrant it. In Japan there is no
statutory minimum and the number of appearances is
decided through co-ordination between the Diet and the
Bank of Japan dependent upon the economic and
financial conditions at the time. Some appearances in
the Diet, however, may be relatively short.
Table A shows the number of parliamentary hearings
attended by central bank officials in the year to
May 2001 and the percentage of those hearings
attended by the head of the central bank. Central bank
officials from Israel, the Czech Republic, Japan, the
United States and the United Kingdom attended the
highest number of parliamentary hearings.
There was considerable diversity in the number of
parliamentary appearances by central bank officials in
the year to May 2001. Such diversity may reflect
variation across countries in the requirements and
preferences with respect to accounting for monetary
policy actions, but also that some central banks are more
likely to make appearances not directly related to
monetary policy, and that in some countries there is
more than one chamber of parliament before which the
central bank appears. The number ranged from between
51 and 100 in Israel, to zero in Germany, where the
Finance Committee of the Bundestag does not monitor
the Bundesbank and has a limited role in holding it to
account. The Bundesbank may, however, decide to
appear on a voluntary basis before parliament (or
relevant committees) and has done so in the past. In
Norway constitutional custom has focused on the
Minister as the responsible official to Parliament, not the
head of public bodies under a Minister’s domain. In the
year to May 2001 the Governor of the Norges Bank
attended one parliamentary hearing to discuss a non
monetary policy related matter.
The total number of appearances tends to be higher, and
the proportion attended by the head of the central bank
lower, in countries where there are appearances during
which the focus is not directly related to monetary
policy. In Israel, the central bank is an official economic
adviser to the government and its staff appear before
Parliament to discuss this advice. Relatively few of these
appearances are by the Governor. In the Czech
Republic, the year to May 2001 was atypical with a large
number of hearings being held with officials to discuss
work on amendments of the Act on the Czech National
Bank. The Governor of the Czech National Bank attends
all the hearings on monetary policy. Likewise, the
Governor of the Reserve Bank of Australia attends all the
hearings on monetary policy matters, but here too the
year to May 2001 was atypical with a number of more
specialised issues being considered which the Governor
did not attend. In the United States, Federal Reserve
officials appear before Congress to discuss a wide range
of economic and financial issues in addition to monetary
policy. In the United Kingdom, Bank of England officials
also appeared in front of Parliamentary Committees to
discuss European Monetary Union, globalisation and
cash and debt management.
The Governors of the central banks in Canada and
New Zealand attended all the meetings. The Bank of
Canada endeavours to appear four times a year, twice
before a House Committee and twice before the Senate
Committee. In addition, special interest items may also
cause one or other of the committees to invite the Bank
to appear. The President of the ECB attends a quarterly
dialogue at the European Parliament. Of the remainder
Table ANumber of parliamentary appearances in the year toMay 2001
Numbers of parliamentary Percentage of appearances appearances (a) by by head of central bankcentral bankers
Australia 4 50Canada 1–5 100Czech Republic 21–30 1–20Euro area 9 66France 6–10 81–100Germany 0 n/aIsrael 51–100 1–20Italy 1–5 61–80Japan 34 81–100Korea 1–5 81–100New Zealand 4 100Norway 1 100United Kingdom 14 41–60United States 18 41–60
n/a = not applicable.
(a) Questionnaire asked respondents to tick boxes indicative of the range of appearances (eg 1–5). Exact number of appearances is included when supplied.
Parliamentary scrutiny of central banks in the United Kingdom and overseas
277
of the scheduled hearings one is attended by the
vice-president to present the annual report and one by
ECB board members.
In Israel, the Governor appears before the Knesset about
five times a year, on a range of matters relating to the
activity of the Bank, mostly before the Finance
Committee. In the United States Chairman Greenspan
attended about half of the testimonies given by Fed
officials to Congress, and in the United Kingdom the
Governor of the Bank of England attended 50% of all
the parliamentary hearings conducted with Bank
officials. (A number of the parliamentary hearings were
with the House of Lords where Monetary Policy
Committee members attended without the Governor.)
The ECB is accountable to the European Parliament for
monetary policy actions affecting the euro area,
including France, Germany and Italy. Nevertheless, the
President of the Banque de France is also called to give
evidence on monetary policy to the French Parliament
where he takes collective responsibility for the actions of
the ECB.
In the United Kingdom, Canada, Japan and the United
States the number of hearings may be larger due to
requirements to appear before both houses of their
respective parliaments.(1) In the United States the
prepared testimony may be the same to both the Senate
and the House of Representatives, though responses to
questioning may, of course, be different. There are four
annual ‘official’ monetary policy testimonies (two in each
chamber) and in addition there are usually a few each
year on the macroeconomy as background for
congressional consideration of the budget.
Chart 1 shows the number of parliamentary hearings
conducted in total or in part on monetary policy issues.
It shows a ranking of countries similar to that in Table A,
with the number of appearances ranging from 34 in
Japan to zero in Germany, Italy and Norway. Taken at
face value the chart perhaps overstates the difference
between Japan and other countries. The number of
annual parliamentary hearings conducted solely on
monetary policy issues is only four, which are the
biannual hearings in two houses of the Diet. Some other
hearings before Diet committees deal not only with
monetary policy actions, but also a broad range of other
themes that falls in their jurisdiction. The duration of
appearances by Bank of Japan officials to account for
monetary policy actions may be relatively short,
occasionally lasting only 15 minutes. The US Congress
holds fewer hearings on monetary policy than the
European, French and UK parliaments, both in absolute
terms and in proportion to the total number of hearings.
There is no significant statistical relationship between
the number of parliamentary appearances and central
bank independence in this sample.(2) The Bundesbank
has been widely cited as a central bank whose high
degree of independence was coupled with a low level of
parliamentary scrutiny; yet there is no evidence in this
sample that such a negative relationship holds more
widely.
(ii) Scrutiny that depends upon prespecified circumstances
Certain events might trigger pre-ordained actions
whereby the central bank is required to explain its
policies. The results are shown in Table B, using
updated information that was originally collected by
Fry et al (2000).
(1) The European Parliament, Israel, New Zealand, Norway and South Korea have a unicameral parliamentary system.Australia, Canada, the Czech Republic, France, Germany, Italy, Japan, the United Kingdom and the United States have abicameral parliamentary system. Central banks operating in the context of a bicameral system may be called to appearbefore both houses of parliament. Even in circumstances where prepared testimony is the same, however, thequestions asked to central bank officials will be different, and this justifies including each appearance as distinct inTable A.
(2) In regressions of the number of appearances on a constant and a measure of independence taken from the Fry et al(2000) survey and of the log of the number of appearances on this measure of independence there was no evidencefor a significant relationship, even at the 20% level.
0
5
10
15
20
25
30
35Number in year to May 2001
Jap
an
Eu
ro a
rea
Isra
el
Fran
ce
Un
ited
Kin
gd
om
Un
ited
Sta
tes
New
Zeal
and
Ko
rea
Au
stra
lia
Cze
ch R
ep
ub
lic
Can
ada
No
rway
Germ
any
Ital
y
Chart 1Parliamentary appearances to discuss monetary policy issues
Note: Some observations are based on mid-points of ranges in questionnaire responses.
278
BBaannkk ooff EEnnggllaanndd QQuuaarrtteerrllyy BBuulllleettiinn:: Autumn 2002
The table illustrates that a number of central banks are
required or have themselves committed to provide
detailed explanations when and if a target is missed.
Sometimes these explanations may be provided in
existing publications (the Czech Republic, Norway and
the United States). In New Zealand special measures
may be initiated by the government. In other cases
additional (open) letters may be required (the United
Kingdom and, although not in our sample, Sweden). Of
the countries in our sample the procedures are not a
legal requirement in any country. In the United
Kingdom the initial remit set for the MPC requires the
Governor of the Bank of England to write an open letter
to the Chancellor whenever inflation deviates by more
than 1 percentage point from its target. The remit is
not, however, in the Act of Parliament providing the legal
basis for independence.
(iii) Level of support for parliamentary committees
An assessment of the degree of parliamentary scrutiny
may be enhanced by considering not only the number of
hearings conducted but also the effectiveness of each
hearing.(1) The survey asked about the number of
parliamentary analytical research staff supporting each
committee. Nine countries provided details of staffing
arrangements for parliamentary committees. As Chart 2
Table BAre there procedures when a target (or numerical objective) is missed? Are they a legal requirement?(i)
Established Legal requirement Details
Australia No No
Canada Yes No The Renewal of the Inflation-Control Target, May 2001, available at http://www.bankofcanada.ca/en/press/pr01-9.htm states that ‘If CPI inflation persistently deviates from the 2 per cent target midpoint, the Bank will give special attention in its Monetary Policy Reports or Updates to explaining why inflation has deviated to such an extent from the target midpoint, what steps (if any) are being taken to ensure that inflation moves back to this midpoint, and when inflation is expected to return to the midpoint.’
Czech Republic Yes No Explanations as to why the target is missed are presented in the Inflation Report. The Board’s discussion of explanations is presented in the minutes which are included in the Inflation Report.
Euro area Yes No The ECB has provided a numerical quantification of its primary objective of price stability. In its Monthly Bulletins and at appearances before the European Parliament, the ECB reports about its monetary policy and whether it has achieved its objective, and if not, why this has been the case.
France (ii) No No
Germany (ii) No No
Israel No No
Italy (ii) No No
Japan No No There is no published numerical policy target or objective.
Korea No No
New Zealand Yes No When in 1995 and 1996 the inflation target was missed, the Minister of Finance wrote to the non-executive Directors of the Bank asking for their opinion on the Governor’s performance.
Norway Yes No When the inflation target was adopted in March 2001 the Norges Bank undertook to provide an assessment in its annual report to the government if there were significant deviations between the actual price inflation and the target. Particular emphasis would be placed on deviations outside the plus or minus one percentage point range. The Ministry of Finance stated in a White Paper in March 2001 that other circumstances might necessitate such an assessment to the government on occasions other than the annual report.
United Kingdom Yes No Under the initial remit set for the MPC the Governor is required to send an open letter to the Chancellor of the Exchequer following a Monetary Policy Committee meeting and referring as necessary to the Inflation Report.
United States No No The semi-annual reports to Congress were initiated in 1978 by the Humphrey-Hawkins Act but the Federal Reserve is no longer required to explain any deviations from the intermediate monetary targets in its semi-annual report to Congress. The Act now requires that ‘the Chairman of the Board shall appear before the Congress at semi-annual hearings, as specified in paragraph (2), regarding (A) the efforts, activities, objectives and plans of the Board and the Federal Open Market Committee with respect to the conduct of monetary policy; and (B) economic developments and prospects for the future described in the report required in subsection (b) of this section.’
Source: Fry et al (2000), extended and updated in the current survey.
(i) Most central banks in the sample may, to some extent, explain misses to any target or numerical objective in standard bulletins and parliamentary appearances. The extent to which suchprocedures may be characterised as ‘established’ was assessed by each central bank. The authors recognise the possibility of subjectivity in responses.
(ii) The French, German and Italian central banks are not responsible for the conduct of monetary policy in the euro area.
0
2
4
6
8
10
12
14
16
Un
ited
Sta
tes
Eu
ro a
rea
Fran
ce
Can
ada
Un
ited
Kin
gd
om
Au
stra
lia
Ko
rea
New
Zeal
and
No
rway
Chart 2Number of parliamentary committee technical research staff
Notes: In Canada, the staff can vary considerably because the committees consider many issues in addition to monetary policy and may add staff for certain items.The US number is approximate. Committees and legislatures employ large numbers of staff, but the precise number with specific economic expertise is difficult to estimate precisely.
(1) A concern raised by Svensson (2001) is that ‘There [is] a range of experience and monetary policy expertise amongstmembers [of the Finance and Expenditure Select Committee of the New Zealand Parliament], which may act to reducethe effectiveness of the monitoring function.’
Parliamentary scrutiny of central banks in the United Kingdom and overseas
279
shows, apart from the United States, which has between
10 and 20 technical research staff, the level of internal
technical support offered to parliamentary committees is
relatively limited in the countries for which information
is available.
To complement the internal staffing the parliamentary
committees in Australia, the European Parliament and
the United Kingdom also receive additional briefing
from panels of outside monetary policy experts. In the
United States, congressional committees frequently call
experts to provide both written and oral testimony. Of
the nine countries that replied only in Norway did the
parliamentary committee receive no technical support.(1)
(iv) Monetary policy decision-makers and their appointment
The nature of central bank independence and scrutiny
may in some circumstances be affected by the degree to
which the executive branch of government is involved in
the appointment of the members of the monetary policy
making board. Table C summarises where the
responsibility for setting monetary policy and
appointments to the central bank rests.
The executive branch of government(2) is involved in the
appointment of monetary policy decision-makers in all
countries surveyed, apart from Canada, where the
central bank board chooses the head of the Bank of
Canada, although the Minister of Finance must approve
the board’s decision. In Japan and the United States the
Diet and Congress have statutory powers to veto the
appointments. In the case of the ECB Article 112(b) of
the Treaty Establishing the European Community states
that the European Parliament needs to be consulted
before the appointment of ECB executive board
members by Heads of State and Government. Other
countries’ parliaments do not possess such vetoes.
New Zealand and Israel are, in practice, the only
countries in the sample where decisions on monetary
policy rest solely with the head of the central bank
rather than with a committee. Nevertheless it is the
head of the central bank who is most frequently (in the
case of Japan and Korea exclusively) called before
parliament to be held to account for monetary policy
actions.
Conclusions
In this small sample of 14 central banks there is
considerable diversity in the number of appearances by
central bankers before parliament to discuss monetary
policy issues. Bank of Japan officials appear around
30 times each year—albeit relatively briefly on
average—and Bank of England and ECB officials also
make higher-than-average appearances before
parliament. In contrast, officials from both the
Bundesbank (even before European Monetary Union)
and the Norges Bank have not appeared before
parliament to discuss monetary policy.
There is no firm evidence in these data to suggest that
particular types of framework are associated with
different overall levels of parliamentary scrutiny. Neither
is there significant evidence of a correlation between the
degree of independence of central banks in the sample,
and the number of appearances related to monetary
policy. The nature of parliamentary scrutiny of monetary
policy may, however, vary according to framework type.
Some inflation-targeting frameworks have defined
ex ante both the specific circumstances in which
scrutiny will be triggered (when the target is missed by
more than a particular amount), and the form it would
take.
The survey provides detailed information about the
nature of parliamentary scrutiny of monetary policy.
In-house technical support offered to parliamentary
committees is usually limited to five members of staff or
(1) This is unsurprising since the Norges Bank, as well as the Bundesbank and Banca d’Italia, did not appear beforeparliament to account for monetary policy in the year in question.
(2) In the Czech Republic the President is responsible for appointing the policy board.
Table CResponsibility for central bank appointments
Monetary policy Policy board Parliamentary veto set by: appointed by: on appointment to
monetary policy boards:
Australia CB Com Exec NoCanada CB Com Board of CB NoCzech Republic CB Com Pres NoEuro area CB Com Heads of Govt NoIsrael Head of CB Exec NoJapan CB Com Exec YesKorea CB Com Exec NoNew Zealand Head of CB Exec NoNorway CB Com Exec NoUnited Kingdom CB Com Exec NoUnited States CB Com Exec Yes
CB = Central Bank; Com = Committee; Exec = Executive branch of government;Pres = president
Notes: According to the Bank of Canada Act, the Governor is responsible for monetary policy.Since 1994, however, the Governor has made decisions through the GoverningCouncil—the group that in addition to the Governor, consists of the Senior DeputyGovernor and the Deputy Governors (currently four). According to the Act the BankBoard of Directors appoints the Senior Deputy Governor and Deputy Governors.Executive Board Members of the ECB are appointed by the Heads of State andGovernment. Governors of euro-area national central banks, who automatically becomemembers of the Governing Council of the ECB, are appointed by their respectivenational authorities. In the United Kingdom, two Executive Directors of the nine-member Monetary Policy Committee are appointed by the Governor afterconsultation with the Chancellor. The three Governors are crown appointments. Theother appointments are made by the Chancellor.
280
BBaannkk ooff EEnnggllaanndd QQuuaarrtteerrllyy BBuulllleettiinn:: Autumn 2002
fewer (although a number of committees supplement
in-house support with additional advice from outside
experts). Parliamentary hearings on monetary policy are
in the main held with the head of the central bank, even
when monetary policy decisions are made on a
committee basis.
The survey also asks about parliamentary scrutiny of
appointments to monetary policy making committees.
We find that government is almost universally involved in
executive appointments in the sample. Generally,
however, there is no requirement for a parliamentary
check on this appointments procedure.
Parliamentary scrutiny of central banks in the United Kingdom and overseas
281
Annex 1Parliamentary scrutiny questionnaire
This questionnaire is part of a pilot study making a global comparison of the level of parliamentary scrutiny of central
banks. If you have any questions regarding the completion of the questionnaire please see the contact details at the
end of the questionnaire.
Please mark (eg ring or underline) the appropriate answer or answers where applicable. Any additional information
you may wish to provide can be given at the end of each question.
We intend to make the results available to you for comment by the end of July.
Name of Central Bank: ………………………………………….
QQuueessttiioonn 11
a) HHooww mmaannyy sseeppaarraattee ttiimmeess ddiidd cceennttrraall bbaannkk ooffffiicciiaallss aappppeeaarr bbeeffoorree PPaarrlliiaammeenntt oorr iittss
rreepprreesseennttaattiivveess iinn tthhee llaasstt yyeeaarr?? (Please count joint appearances by two or more officials at the same
hearing as a single appearance)
none 1–5 6–10 11–20 21–30 31–40 41–50 51–100 100+
b) WWhhaatt ppeerrcceennttaaggee ooff tthheessee aappppeeaarraanncceess wweerree bbyy tthhee CChhaaiirrmmaann//GGoovveerrnnoorr ooff tthhee cceennttrraall bbaannkk??
(either alone or accompanied by other central bank officials)
1–20 21–40 41–60 61–80 81–100
c) WWhhaatt ppeerrcceennttaaggee ooff tthhee ttoottaall nnuummbbeerr ooff hheeaarriinnggss wwaass,, iinn tthhee mmaaiinn,, rreellaatteedd ttoo mmoonneettaarryy ppoolliiccyy
ccoonncceerrnnss??
1–20 21–40 41–60 61–80 81–100
d) IIss tthhiiss nnuummbbeerr ooff hheeaarriinnggss,, ddiivviissiioonn ooff ssuubbjjeeccttss,, aanndd ppeerrcceennttaaggee ooff aappppeeaarraanncceess,, rreepprreesseennttaattiivvee
ooff aa ttyyppiiccaall yyeeaarr?? (If not, please explain, eg if it varies owing to the state of the economy?)
AAddddiittiioonnaall ccoommmmeennttss::
QQuueessttiioonn 22
a) HHooww mmaannyy mmeemmbbeerrss ooff ssttaaffff wwoorrkk ffuullll ttiimmee ffoorr tthhee ppaarrlliiaammeennttaarryy ccoommmmiitttteeee((ss)) rreessppoonnssiibbllee ffoorr
hhoollddiinngg tthhee cceennttrraall bbaannkk ttoo aaccccoouunntt??
1–5 6–10 10–25 25–50 50+
b) WWhhaatt ppeerrcceennttaaggee ooff tthhiiss ssttaaffff pprroovviiddeess rreesseeaarrcchh//aannaallyyttiiccaall ssuuppppoorrtt??
1–20 21–40 41–60 61–80 81–100
c) DDooeess aannyybbooddyy aaddvviissee tthhee ccoommmmiitttteeee((ss)) oonn aa ppaarrtt--ttiimmee bbaassiiss,, iiff ssoo wwhhoo?? (eg in the UK parliament a
panel of expert economists briefs the Treasury Committee in advance of hearings on monetary policy)
AAddddiittiioonnaall ccoommmmeennttss::
282
BBaannkk ooff EEnnggllaanndd QQuuaarrtteerrllyy BBuulllleettiinn:: Autumn 2002
QQuueessttiioonn 33
a) HHooww iiss tthhee nnuummbbeerr ooff aappppeeaarraanncceess bbeeffoorree PPaarrlliiaammeenntt,, mmaaddee bbyy cceennttrraall bbaannkk ooffffiicciiaallss,, ddeecciiddeedd??
By statute (ie set out in law) By the parliamentary committee
Both Other (please specify)
AAddddiittiioonnaall ccoommmmeennttss::
QQuueessttiioonn 44
a) WWhhoo iiss rreessppoonnssiibbllee ffoorr aappppooiinnttiinngg tthhee CChhaaiirrmmaann//GGoovveerrnnoorr ooff tthhee cceennttrraall bbaannkk??
Central bank board Parliament Council of ministers
Prime Minister/Finance Minister Other (please specify)
b) DDooeess PPaarrlliiaammeenntt hhaavvee tthhee ppoowweerr ttoo vveettoo tthhee aappppooiinnttmmeenntt ooff tthhee CChhaaiirrmmaann//GGoovveerrnnoorr ooff tthhee
cceennttrraall bbaannkk??
Yes No
AAddddiittiioonnaall ccoommmmeennttss::
QQuueessttiioonn 55
a) WWhhoo iinn tthhee cceennttrraall bbaannkk iiss rreessppoonnssiibbllee ffoorr ddeetteerrmmiinniinngg mmoonneettaarryy ppoolliiccyy??
Chairman/Governor Committee of officials including chairman
If the answer to question 5a is ‘Chairman/Governor’ then you have completed the questionnaire. See end of
page 4 for details on how to return the questionnaire.
If the answer to question 5a is ‘Committee of officials including chairman’ please go on to question 5b.
Questions 5b and 5c concern the membership of the monetary policy making committee other than the
chairman.
b) WWhhoo iiss rreessppoonnssiibbllee ffoorr aappppooiinnttiinngg tthhee mmeemmbbeerrss ooff tthhee ccoommmmiitttteeee??
Central bank board Parliament Council of ministers
Prime Minister/Finance Minister Other (please specify)
AAddddiittiioonnaall ccoommmmeennttss::
c) DDooeess PPaarrlliiaammeenntt hhaavvee tthhee ppoowweerr ttoo vveettoo aappppooiinnttmmeennttss ttoo tthhee ccoommmmiitttteeee??
Yes No
AAddddiittiioonnaall ccoommmmeennttss::
Parliamentary scrutiny of central banks in the United Kingdom and overseas
283
Annex 2
To obtain a more precise dataset answers to the following questions would be a helpful addition to the original
questionnaire:
1. What is the principal objective of the parliamentary committee when it holds hearings with central bank officials?
2. How many members of parliament make up this committee?
● On average how many attend meetings with central bank officials?
● How many members of the parliamentary committee have any formal economic expertise (eg academic
training)?
3. [[IIff lleessss tthhaann 2200]] Exactly how many parliamentary hearings did central bank officials attend in the last year?
● How many of these appearances were made by the head of the central bank?
4. [[IIff lleessss tthhaann 2200]] Exactly how many hearings on monetary policy did central bank officials attend in the last
year?
● How many of these were appearances made by the head of the central bank?
5. How often does the parliamentary committee request that the central bank provide written evidence on the
conduct of monetary policy?
6. [[IIff lleessss tthhaann 1100]] Exactly how many members of staff work full time for the parliamentary committee
responsible for holding the central bank to account?
● How many of these members of staff provide analytical/technical support?
284
BBaannkk ooff EEnnggllaanndd QQuuaarrtteerrllyy BBuulllleettiinn:: Autumn 2002
References
BBrriiaauulltt,, CC,, HHaallddaannee,, AA aanndd KKiinngg,, MM ((11999966)), ‘Independence and accountability’, Bank of England Working Paper
no. 49.
DDee HHaaaann,, JJ ,, AAmmtteennbbrriinnkk,, FF aanndd EEiijjffffiinnggeerr,, SS CC WW ((11999999)), ‘Accountability of central banks: aspects and
quantification’, Banca Nationale del lavoro Quarterly Review 209, pages 169–93.
DDee HHaaaann,, JJ aanndd EEiijjffffiinnggeerr,, SS ((22000000)), ‘The democratic accountability of the European Central Bank: a comment
on two fairy-tales’, The Journal of Common Market Studies, September.
EEiijjffffiinnggeerr,, SS CC WW aanndd GGeerraaaattss,, PP ((22000022)), ‘How transparent are central banks?’, Cambridge University, mimeo.
FFrryy,, MM,, JJuulliiuuss,, DD,, MMaahhaaddeevvaa,, LL,, RRooggeerr,, SS aanndd SStteerrnnee,, GG ((22000000)), ‘Key issues in the choice of monetary policy
framework’, in Mahadeva, L and Sterne, G (eds), Monetary frameworks in a global context, Routledge, London.
IIMMFF ((22000000)), Supporting document to the code of good practices on transparency in monetary and financial policies:
Washington DC, 2000, www.imf.org/external/np/mae/mft/sup/index.htm
SSvveennssssoonn,, LL ((22000011)), Independent review of the operation of monetary policy in New Zealand: report to the
Minister of Finance, www.princeton.edu/~svensson/NZ/RevNZMP.htm
TTrreeaassuurryy CCoommmmiitttteeee ((ooff UUKK HHoouussee ooff CCoommmmoonnss)) ((11999977)), ‘Accountability of the Bank of England’, First Report
of Session 1997–98, HC 282, London. The Stationery Office Ltd.
285
In common with most OECD countries, the average age
of the UK population is expected to rise in the current
century, reflecting the maturing of the baby boom
generation, lower fertility rates and increased longevity.
On current trends, the average will rise from 38.6 years
in 1998 to 44 years by 2040 and the number of people
over 75 will increase from about 4.4 million in 2000 to
8.3 million in 2040.(2)
The economic impact of changes in the age structure of
the population is likely to be widespread, depending on
how people react to the welcome prospect of living
longer. The potential effect on saving, the allocation of
funds around the financial system and the risks that this
entails are of direct relevance to the Bank of England,
since they affect its core purposes.
This article provides a preliminary assessment of the
effects of ageing on UK economic growth and the living
standards of different age groups within the population,
summarising work done within the Bank in co-operation
with the Financial Services Authority (FSA). It begins by
describing how average living standards in the United
Kingdom might develop over the course of this century,
taking account of anticipated demographic changes. It
then goes on to discuss the sensitivity of this outlook to
the way in which the overall level of saving in the
economy might change as the population ages, taking
account of the interaction between the level of saving,
national income and the rate of return on assets. It is
shown that demographic change might have a
differential impact, benefiting some generations and not
others. This arises partly from changes in the rate of
return on assets. The article goes on to review some of
the available evidence on the link between the rate of
return and ageing, emphasising the risks inherent in
asset returns. It concludes by summarising some of the
key issues arising from this discussion and identifying
where the main vulnerabilities lie.
Living standards and demographic change inthe United Kingdom
Chart 1 illustrates the anticipated extent of demographic
change in the United Kingdom in the coming 60 or so
years. It shows that, taking 60 as the retirement age, the
number of working-age people per pensioner is due to
fall from around three now to about two in 30 years’ time
and then to stabilise around that ratio.
At a simple level, such change affects living standards
because of increasing ‘dependence’; a rise in the
number of people with a claim to the country’s resources
relative to those involved in producing them. But the
level of living standards is also affected by the amount of
productive capital available, as well as the effectiveness
with which labour is used. It is possible that the impact
on living standards of increased dependence will be
offset by changes in saving, by longer working lives or by
increased productivity. Table A shows how trends in the
Ageing and the UK economy(1)
This article argues that overall living standards in the United Kingdom are set to double over the next50 years alongside a sharp increase in the proportion of people over retirement age. While there areclear risks to this outlook, these would be present even without demographic change. Nevertheless anageing population does appear to increase the risks to the financial welfare of individuals, especially intheir old age. If people living longer do not save more when they are working, then either they have toconsume less in their old age or work for longer than would have been the case had greater provisionbeen made for retirement. This risk is heightened by general uncertainty about asset returns whichbecomes more important as the number of people reliant on private pensions increases.
(1) A more technical version of this paper is available as a Bank working paper (Young (2002)). This work has been usedas background to an FSA thematic review of ‘The implications of an ageing population for the FSA’. The outcome ofthat review was published on 20 May in ‘Financing the future: mind the gap!’.
(2) Government Actuary’s Department 1998 population projections.
By Garry Young of the Bank’s Domestic Finance Division.
286
BBaannkk ooff EEnnggllaanndd QQuuaarrtteerrllyy BBuulllleettiinn:: Autumn 2002
population in different age groups over recent and
future ten-year periods translate into aggregate output
and output per head of population under specific
cautious assumptions about capital accumulation,
labour participation rates and technological progress.
Here capital is cautiously assumed to grow at 10% per
ten-year period, less than half the growth rate seen in
the past three decades, while underlying labour
productivity grows at 1.75% per year, broadly in line
with the average over the past 40 years and consistent
with the assumptions underlying the Government’s
long-term fiscal projections (HM Treasury (2002)).
Participation rates are assumed not to change and are
set at their recent levels.
Under these assumptions, living standards, as
represented by output per head of the population, are
projected to grow more slowly than in the post-war years.
Nevertheless, growth is fast enough that by 2061–68 the
level of living standards is still two and a half times as
great as in the 1991–2000 period. The projected broad
increase in living standards is consistent with similar
projections made for other countries. Indeed, in one of
the earliest studies of the economic effects of ageing,
Cutler, Poterba, Sheiner and Summers (1990) find that,
while increasing dependence reduces living standards in
the long run (relative to levels without a change in
dependence), this would be fully reversed by only a
0.15 percentage point a year increase in productivity
growth.
This comforting conclusion is dependent on a number
of uncertain factors and would be adversely affected by
sharp falls in either productivity growth or the rate of
capital accumulation. While either is possible and
therefore a source of general uncertainty, their possible
link to demographic change needs to be clarified. There
is very little theoretical argument or empirical evidence
to link productivity growth to demographic change
directly, apart from the effect on average productivity as
large cohorts move through different stages of the
productivity lifecycle.(1) There is, however, a relationship
between demographic change and productivity through
the effect on national saving and hence capital
accumulation.(2) In small open economies changes in
national saving are as likely to be reflected in foreign as
in domestic investment. So when assessing the likely
future evolution of living standards it would also be
important to take account of the build-up of claims on
foreign countries.
Chart 1Over 60-year-olds as a percentage of 15 to 60-year-olds
0
15
30
45
60
75
1961 71 81 91 2001 11 21 31 41 51 61
Per cent
Source: Government Actuary’s Department.
Table ADemographic trends and living standards
Population of age group Effective labour Capital stock Output Output per head Growth in output(millions) supply (millions, (£ billions, (£ billions, of population per head
2000 equivalent) 1995 prices) 1995 prices) (£ thousand per head,0–14 15–29 30–44 45–59 60–74 75+ 1995 prices) Annualised
(per cent)
1961–70 12.8 11.2 10.3 10.4 7.3 2.4 15.8 843 371 6.81971–80 12.9 12.2 10.2 9.9 8.1 2.5 19.0 1,198 477 8.6 2.41981–90 11.0 13.3 11.5 9.3 8.1 3.7 24.1 1,451 578 10.2 1.71991–2000 11.3 12.1 12.9 10.4 7.9 4.2 30.1 1,795 731 12.4 2.02001–10 10.9 11.6 13.5 11.9 8.3 4.6 37.3 1,974 868 14.3 1.42011–20 10.6 11.8 12.0 13.3 9.9 5.0 43.9 2,172 997 15.9 1.12021–30 10.6 11.2 12.4 12.4 11.3 6.3 51.3 2,389 1,141 17.8 1.12031–40 10.4 11.2 12.0 11.9 11.8 7.6 59.6 2,628 1,301 20.1 1.22041–50 10.2 11.1 11.6 12.1 10.9 8.8 69.9 2,891 1,492 23.1 1.42051–60 10.1 10.8 11.7 11.6 11.0 8.4 81.8 3,180 1,709 26.8 1.52061–68 10.0 10.7 11.5 11.5 10.8 8.3 94.5 3,498 1,940 30.9 1.5
Notes to table: The ‘effective’ labour supply is constructed assuming participation rates of 0.75, 0.85, 0.70 and 0.1 for 15–29, 30–44, 45–59 and 60–74 year-old age groups respectively. Theeffectiveness of a unit of labour (normalised at one per employee in 2000) is assumed to grow at 1.75% per year. 30–44 year-olds are assumed to be 40% more productive than others. Output (GDPat constant 1995 market prices) and the capital stock (in constant 1995 prices) are averaged over each ten-year period. The future capital stock is cautiously assumed to grow at 10% per ten-yearperiod. Future output is generated by a Cobb-Douglas production function with capital share of 0.35. Sources: Population projections from Government Actuary’s Department, capital stock and output from Office for National Statistics.
(1) Cutler et al (1990) is one of the few papers in the literature that try to link productivity growth to demographic change.(2) This could also affect the rate of technological change if technological improvements are embodied in capital
investment.
Ageing and the UK economy
287
While overall savings levels cannot be identified simply
with the saving of individuals, the link between
aggregate saving and ageing is usually approached by
considering saving over the individual lifecycle.(1) If,
as seems logical, people tend to save most in their
middle age and dissave in their old age, then aggregate
saving might be expected to increase when the
proportion of middle-aged people in the population
increases and decline when the proportion of old people
increases.
In the most readily available data, the observed pattern
of saving across different age groups does not match up
easily with the predictions of lifecycle theory. Chart 2
shows measured saving rates by age group in the United
Kingdom in 1974 and 1995 based on data from the
Family Expenditure Survey (FES) presented in Banks and
Rohwedder (2000). Saving as defined here represents
the accumulation of financial assets and does not take
account of the accumulation of housing assets or any
pension fund built up by employer contributions. It also
fails to take account of wealth accumulated through
capital gains on assets.
One of the striking features of this chart is the high
median rate of saving by the retired at a time of life
when they might be expected to be running down assets.
This is the so-called ‘retirement savings puzzle’ analysed
by Banks, Blundell and Tanner (1998). They show that
this cannot be accounted for by mortality risk, the
removal of work-related costs or demographic factors,
although it may reflect differential mortality given that
those with the highest pension incomes live longest. It
might also be accounted for by noting the complexities
in measuring pensioner income. As Miles (1999) has
pointed out, household surveys like the FES measure
pensioner income incorrectly, because some of the
receipts classified as income are depleting the pension
fund of which the pensioner is a member and should
properly be treated as dissaving.
Hussain (1998) adjusts the saving rate of pensioners for
this form of mismeasurement and suggests that the ‘true’
saving rate of pensioners, taking account of the
depletion of pension funds, is minus 8% of disposable
income. From this he predicts a decline in the personal
saving rate from a peak of around 12% in 2005 to a low
of around 9% by 2040 as a consequence of demographic
change. This can have a relatively large impact because
in a closed economy a lower rate of saving reduces
capital accumulation, the capital stock and hence output
and subsequent saving. Young (2002) shows illustrative
projections of output per head in a baseline case where
the ratio of investment to output is fixed at recent levels
and in an alternative case where aggregate investment
responds to exogenously determined cohort-specific
saving rates adjusted in line with Hussain’s estimates.
Living standards grow more slowly in the latter case,
so that by 2060 they are about 10% lower than they
would be in the fixed investment rate case. While
this difference is substantial, in both cases living
standards in the future are projected to be substantially
higher than they are now, in line with the estimates in
Table A.
Aside from the problem in estimating age group specific
saving rates, there are a number of other difficulties with
the forgoing illustration. In particular, it cannot be
assumed that the age-specific saving rates will remain
constant. As Chart 2 illustrates, the saving rates of
different age groups have changed over time, reflecting
different aggregate influences on saving, as well as
factors specific to particular cohorts. It is also likely
that demographic change will have a number of effects
on saving rates and welfare at particular points in the
lifecycle, depending on what is causing the demographic
shift. Moreover, the approach adopted so far has
ignored many of the complex interactions that can only
be readily allowed for in a more general setting. In
particular, a general equilibrium approach would also
enable an analysis of the impact that demographic
change could have on asset prices.
Chart 2Saving rates by age group
0.05
0.00
0.05
0.10
0.15
0.20
0.25
0.30
20–24
25–29
30–3435–39
40–44
45–49
50–54
55–59
60–6465–69
70–7475–79
80–8485–89
As a proportion of household income
+
_
1974
1995
(1) For the United Kingdom, household saving in 2000 amounted to 31/2% of GDP, corporate saving to 9% of GDP andgovernment saving to 31/2% of GDP.
288
BBaannkk ooff EEnnggllaanndd QQuuaarrtteerrllyy BBuulllleettiinn:: Autumn 2002
Assessing the impact of demographic change
In Young (2002), a simplified dynamic general
equilibrium model is used to assess the impact of
demographic change. The model outlines the possible
effects of ageing on the supply of labour and capital, the
key factors in determining the amount of resources that
the economy can produce. It also shows, when markets
clear, how such changes affect real wages and real
interest rates, the rewards to labour and capital that
determine the living standards of different groups within
the population. The model thus abstracts from many
important features of the actual economy in order to
focus on the essential details. In particular, it focuses on
a situation where pensions are provided solely from
private saving. It also ignores inflation and the
possibility of capital flows between different countries,
so that real interest rates are determined by purely
domestic factors.
Within the model, the impact of ageing on saving and
hence capital accumulation depends on the behaviour of
individual households and how they respond to
changing economic incentives. One possibility is that
saving is chosen to spread spending evenly over the
maximum possible lifetime of each household taking
account of expected future incomes. The difficulty with
this approach to modelling saving is that it assumes very
high powers of calculation on the part of individual
households and is not necessarily consistent with
empirical evidence on household saving behaviour. In
order to assess the extent to which the analysis is robust
to different assumptions about household behaviour a
second case is considered where households are
assumed to follow simple ‘rules of thumb’. In this case
they spend a fixed proportion of their current resources
throughout their lives and do not take account of the
fact that they would be better off by altering the amount
they save as economic conditions change. Importantly,
there is no change in saving when the expected period
of retirement lengthens.
Within this framework, the welfare implications of an
ageing population are shown to depend upon the type of
demographic shock that brings it about. Furthermore,
some types of demographic shocks have opposing effects
on the welfare of different generations. Consider first
the effect of a baby boom. This is a temporary
demographic shock with no impact on the length of life
of individuals. It reduces the average living standards of
the baby boom generation while improving the living
standards of their parents and children. This arises
simply from the fact that the baby boom generation are
effectively more plentiful and this reduces the value of
their labour when they are working and the value of
their capital when retired.
By contrast, the impact of greater longevity is different
since it affects the average length of life of all
individuals. It has an adverse impact on the
consumption of all generations. This follows from the
assumption that people do not extend their working lives
when life expectancy rises. With longer life spans but no
change to labour supply, households have to spread their
resources over a longer period and hence must consume
less. Relaxing the assumption of fixed labour supply
would change the results and tend to reduce the impact
of increased longevity on consumption.
The impact of reduced fertility, a permanent
demographic shock that has no impact on the length of
life of individuals, is different yet again. It has little
effect on individual welfare in the long run, although it
does improve the reward to labour relative to that of
capital by reducing the number of people of working age
relative to those living off savings. This has an adverse
impact on those who are old when the change in fertility
occurs as they lose from lower real interest rates without
benefiting from having higher real wages when they are
young.
In some of the cases considered, optimal consumption in
retirement is reduced relative to what it would have been
without a demographic shock. These effects are
compounded when households determine their
consumption by following rules of thumb, since in this
case they do not make the necessary adjustment to their
consumption when they are young.
The model also reveals some useful predictions about
medium to longer-term real interest rates and hence
asset prices. The implications of demographic shocks
for real interest rates are dependent on the effect on
capital accumulation and the type of household
behaviour assumed. In the case of a baby boom and
lower fertility, real interest rates move in a qualitatively
similar way in both the optimising and rule-of-thumb
cases, although the magnitude of the effects is different.
When households are optimisers the effects are generally
small, as saving responds to changes in real interest rates
and so dampens their movement. In the case of
increased longevity, the two assumptions about
household behaviour give different predictions about
the direction of change in real interest rates, reflecting
Ageing and the UK economy
289
different responses in saving. When households are
optimisers, aggregate saving rises in response to
increased survival rates and this depresses real interest
rates mildly. But when households make no provision for
increased survival, aggregate saving falls, raising
equilibrium real interest rates. In all of the cases
considered, the impact on real interest rates is
quantitatively small (less than 1 percentage point),
especially when households are optimisers. This is
consistent with the wider academic literature. For
example, in his analysis of the effect of ageing on the
UK economy, Miles (1999) shows the real interest rate
falling by 0.4 percentage points over the 30 years from
the late 1990s.
Of course, the assumption that medium to long-term real
interest rates are determined within any national
economy is inconsistent with high levels of capital
mobility within the international economy. When it is
mobile, capital will tend to flow to countries where the
rate of return is highest, equalising risk-adjusted rates of
return where mobility is perfect. For small open
economies, purely domestic demographic shocks would
have no effect on the domestic rate of return and
countries would export capital when the domestic
saving rate was high and import it when it was low. In
these circumstances, the model would be a useful
description of how the global economy might respond to
ageing.
Brooks (2000) outlines the implications of population
ageing using a calibrated model of the world economy.
His simulations show that there will be a turning point
in regional saving-investment balances between 2010
and 2030 when the European Union and North America
will experience a substantial decline in savings relative
to investment as their populations age rapidly. This
shift will be financed by capital flows from less
developed regions that are projected to become capital
exporters.
Empirical evidence on asset returns anddemographic structure
As noted above, calibrated theoretical models generally
show only modest effects of demographic change on real
interest rates and, by implication, asset prices. But this
prediction is sensitive to the precise specification of the
model. Furthermore, this contrasts sharply with some
popular claims, especially in the United States, that the
increase in asset prices in the 1990s was partly caused
by the movement of the baby boom generation into the
high-saving part of its lifecycle. There are similar
predictions of an asset price ‘meltdown’ when the baby
boomers attempt to sell their assets on retirement,
although, as Poterba (2001) has noted, this is difficult to
reconcile with the view that any such effect should
already be priced into asset prices determined in
forward-looking markets.
These claims can be assessed by examining whether
asset price movements have been linked to shifts in the
demographic structure that have occurred in the past.
Mankiw and Weil (1989) analysed the relationship
between house prices and the age structure of the
US population. They forecast that reduced housing
demand would result from ageing of the US population
after 1990 and this would lead to house prices lower
than ‘any time in recent years’. Of course, house prices
did not fall as predicted over the 1990s. This does not
refute the thrust of the Mankiw-Weil analysis since other
factors have undoubtedly changed so as to offset the
impact of demographic changes, but it does emphasise
the need for caution in making predictions about asset
prices without acknowledging the wider uncertainty that
exists.
Similar trends in the house-buying population were
suggested as a cause of the lacklustre state of the UK
housing market in the mid-1990s (Wallace (2001)), but
the subsequent housing recovery again suggests that
demographic trends are not the only cause of house
price increases.
In a wide-ranging survey, Poterba (2001) questions
whether it is possible to test for low-frequency patterns
in asset prices: ‘There is one Baby Boom shock in the
post-war US demographic experience, and as the Baby
Boom cohort has approached fifty, real stock market
wealth has risen rapidly. This is consistent with some
variants of the demographic demand hypothesis.
Whether fifty years of prices and returns on this
experience represent one observation, or fifty, is however
an open question’. Despite this caveat, Poterba goes on
to analyse the empirical evidence. He concludes that ‘it
is difficult to find a robust relationship between asset
returns on stocks, bonds, bills and the age structure of
the US population over the last seventy years’.
This negative result is consistent with the small effects
on asset returns from demographic change generated by
the theoretical models and suggests that it cannot be
isolated in the data because of other influences.
290
BBaannkk ooff EEnnggllaanndd QQuuaarrtteerrllyy BBuulllleettiinn:: Autumn 2002
Chart 3 shows the ex post annual average rate of return
on equity in the United Kingdom on investments held
for 20 years at a time with dividends re-invested. The
well-known volatility in asset returns shown in Chart 3
draws attention to some of the key risks to which
investors reliant on private saving are exposed. In
particular, some savers will reach retirement age, having
invested their savings at high rates of return, while
others will be much less fortunate. Using these figures, a
20-year investment in 1954 in UK equities with
dividends re-invested would have barely grown at all,
whereas the same investment made in 1974 would have
grown 13-fold. This simple illustration demonstrates the
vulnerability of generations of private savers to the risk
inherent in financial assets. It probably poses a much
greater threat to the living standards of future
pensioners than any impact resulting from demographic
change.
Conclusion
Under relatively cautious assumptions about
technological progress and capital accumulation,
aggregate living standards could still double over the
next 50 years, despite the projected marked increase in
the proportion of people over retirement age. Even after
allowing for the possibly depressing effect of an ageing
population on saving rates and hence on capital
accumulation, average material living standards should
still be significantly higher. Theoretical models and
empirical research suggest that demographic change
tends to have little effect on asset prices, which is in any
case dwarfed by their usual volatility.
Alongside this picture of improving living standards, the
risks to individual welfare may have increased as a result
of demographic change. This has occurred in three
main ways. First, there has been a shift throughout the
world from public to private provision for old age,
increasing the proportion of people exposed to asset
price fluctuations. Second, the size of the group
exposed to such risks is growing larger as a direct result
of ageing. Third, any adverse financial effects of greater
longevity are most likely to be felt in old age. This third
effect arises since people living longer have to spread
their lifetime incomes over more years of life, implying a
need for more saving when working. If this does not
occur, then either consumption has to be lower in old
age or people have to work for longer than would have
been the case had proper provision been made for
retirement. A recent report by the Financial Services
Authority (2002) considers in more detail some of the
key risks associated with demographic change in the
United Kingdom and outlines some of their implications
for consumers, financial firms and the authorities.
Chart 3Long-run asset returns(a)
4
2
0
2
4
6
8
10
12
14
16
Average
Per cent
1918
+
_
25 30 35 40 45 50 55 60 65 70 75 80 85 90 95
(a) Average annual return on investments in UK equity made 20 years earlier with dividends re-invested.
Sources: Barclays Capital, Bank of England calculations.
Ageing and the UK economy
291
References
BBaannkkss,, JJ ,, BBlluunnddeellll ,, RR aanndd TTaannnneerr,, SS ((11999988)), ‘Is there a retirement savings puzzle?’, American Economic Review,
Vol. 88, pages 769–88.
BBaannkkss,, JJ aanndd RRoohhwweeddddeerr,, SS ((22000000)), ‘Life-cycle saving patterns and pension arrangements in the UK’, Institute for
Fiscal Studies, mimeo.
BBrrooookkss,, RR ((22000000)), ‘Population ageing and global capital flows in a parallel universe’, International Monetary Fund
Working Paper 00/151.
CCuuttlleerr,, DD,, PPootteerrbbaa,, JJ ,, SShheeiinneerr,, LL aanndd SSuummmmeerrss,, LL ((11999900)), ‘An ageing society: opportunity or challenge?’,
Brookings Papers on Economic Activity, 1990:1, pages 1–73.
FFiinnaanncciiaall SSeerrvviicceess AAuutthhoorriittyy ((22000022)), ‘Financing the future: mind the gap!’, May.
HHMM TTrreeaassuurryy ((22000022)), ‘Budget 2002: The strength to make long-term decisions’, HC 592.
HHuussssaaiinn,, II ((11999988)), ‘Ageing populations, pensions and capital markets’, Financial Services Authority, mimeo.
MMaannkkiiww,, NN GG aanndd WWeeiill ,, DD NN ((11998899)), ‘The baby boom, the baby bust, and the housing market’, Regional Science
and Urban Economics, pages 235–58.
MMiilleess,, DD ((11999999)), ‘Modelling the impact of demographic change upon the economy’, Economic Journal, Vol. 109,
pages 1–36.
PPootteerrbbaa,, JJ MM ((22000011)), ‘Demographic structure and asset returns’, Review of Economics and Statistics, Vol. 83,
pages 565–84.
WWaallllaaccee,, PP ((22000011)), Agequake, Nicholas Brearley Publishing.
YYoouunngg,, GG ((22000022)), ‘The implications of an ageing population for the UK economy’, Bank of England Working
Paper no. 159.
292
Introduction
Profitability is a key indicator of corporate health and
profit warnings indicate unexpected developments that
may imply lower profitability and increased financial
fragility for the firms issuing these warnings. They can
therefore be a useful leading indicator, especially as they
are mandatory and oblige companies to reveal
immediately any change in prospects that might have a
bearing on their share price.
This article relates profit warnings for a sample of UK
quoted non-financial companies that have issued profit
warnings between 1997 and 2001 to their profitability
and balance-sheet strength (gearing and liquidity),
before and after the warnings. Previous authors have
focused on the impact of profit warnings on the share
prices of issuing companies. For example, Clare (2001)
estimates, for a subset of UK firms issuing profit
warnings, that the average share price reduction relative
to the FTSE 100 can be as much as 13% on the day a
warning is issued.
Profit warnings data are an indicator of unexpected
adverse shocks directly affecting the financial position
of companies. This contrasts with other indirect
indicators that embody the revisions to expectations of
agents outside the company; for example, changes in
ratings reflect revised expectations by rating agencies
about the financial viability of companies. The results of
this paper support the view that profit warnings are
associated with a (persistent) fall in profit margins for
the majority of firms who issue a warning. Moreover, the
incidence and size of the fall in profit margins is greater
than for firms not issuing warnings. Although it may not
be surprising that profit warnings are associated with
lower profitability, previous work has not identified the
degree of persistence of the lower profitability, nor its
extent. And the analysis does suggest that profit
warnings do not merely represent previous overly
optimistic expectations of profitability, with no
necessary implications for actual profit levels; on the
contrary, they do appear to contain forward-looking
information about actual profit levels.
These results can be interpreted in the wider context of
examining how firms respond to unexpected financial
shocks or financial pressure. Studies of corporate
behaviour suggest that firms adjust to exogenous shocks
or financial pressure on their cash flows by cutting
investment, dividends or employment.(1) The research
finds that firms who issue warnings are also more likely
to see their gearing levels rise and dividends and
investment fall (relative to the prewarning position, and
to other firms whose profit margins fall but who do not
issue a warning).
The article begins with a discussion of the data and
some descriptive statistics on profit warnings, as well as
the research method used. It then quantifies the impact
of profit warnings on profit margins and examines the
impact of any additional information from profit
warnings on other balance-sheet variables; namely the
gearing and liquidity levels and discretionary
expenditures of issuing firms.
The balance-sheet information content of UK companyprofit warnings
This article looks at the information content of profit warnings issued by UK private non-financialcompanies over the period 1997–2001 in relation to measures of their profitability and balance-sheetstrength. It finds that profit warnings are associated with a persistent fall in profit margins and thatthis decline in margins is larger than for companies who do not issue warnings. The article also findsthat profit warnings contain incremental information for other balance-sheet variables: those firms whoissue warnings are also more likely to see their gearing levels rise, and investment and dividends fall,than other firms whose profit margins also fall but who do not issue a warning.
By Allan Kearns and John Whitley of the Bank’s Domestic Finance Division.
(1) See, Fazzari et al (1988), Bernanke et al (1996), Nickell and Nicolitsas (1999), Benito and Young (2001).
The balance-sheet information content of UK company profit warnings
293
Data and method
Trading statements must be issued by listed companies
in the event that they become aware of developments in
their financial situation or business that are significant
enough to move the company’s share price
substantially.(1) There are many events that typically give
rise to trading statements; the listing rules suggest that
dividend announcements, board appointments or
departures, profit warnings, share dealings by directors
or substantial shareholders, acquisitions and disposals,
annual and interim results and any offers of securities
should all be announced in a trading statement.
As used in this article ‘profit warning’ is a negative
trading statement. Firms issue profit warning statements
when they have material reason to project their
profitability to be lower than previously expected; by the
company itself, by its shareholders or by research
analysts. In general, a company must issue a profit
warning as soon as possible after it has become aware of
new circumstances that have caused it to revise down its
expected future profitability. A short delay between the
discovery of this new information and a firm’s profit
warning is allowed only in exceptional circumstances
where the company needs time to clarify the situation.
Even in these circumstances a company has an
obligation to issue a holding statement outlining the
subject matter of its investigations. Together these rules
imply that profit warnings should indicate promptly
revised expectations of future profitability by the issuing
company. To the extent that other market participants
believe the company’s trading statement, we might also
expect these participants to revise their expectations for
the same company.
This article looks at warnings issued between 1997 and
2001. The UK definition of a profit warning was
unchanged over this period. However, UK listing rules
are different from those in operation in other
jurisdictions.(2)
A database of warnings issued by all listed firms on the
London Stock Exchange (LSE) since January 1997 is
maintained by the Bank of England. The data are
updated on a monthly basis by using the key word
search facility in Reuters Business Briefing.(3) The
database consists of 1,323 warnings issued by 1,047
firms over this period.(4) The number of profit warnings
in 2001 was higher than that recorded in any of the
previous four years and more than double the number
recorded in 2000 (see Chart 1).
This database has been matched with a database of UK
company accounts in order to examine the impact of
warnings on company-account variables for over 700 of
these firms.(5) The set of financial accounts covering the
year in which a profit warning was issued has been
identified for each firm. Combining the data on profit
warnings with company accounts data suggests that
profit warnings are spread across the whole distribution
of profitability (the shares of profit warnings broadly
match the percentiles of the values of profit margins (see
Chart 2)). So profit warnings are not confined to
low-profitability companies.
The method adopted in this paper is to pool the
observations on firms issuing profit warnings across all
(1) Paragraph 9.2 in the listing rules that govern firms listed on the London Stock Exchange states that ‘[a] company mustnotify the Company Announcements Office without delay of all relevant information which is not public knowledgeconcerning a change in (a) the company’s financial condition, (b) the performance of its business or (c) the company’sexpectation of its performance; which, if made public, would likely lead to substantial movement in the price of itslisted securities.’ A copy of the listing rules published by the Financial Services Authority can be found atwww.fsa.gov.uk/ukla/2_listinginfo.html
(2) For example, the ‘Fair Disclosure’ rules in force in the United States until October 2000 did not require full public andimmediate disclosure of material information. These rules were tightened somewhat with effect from October 2000,see www.sec.gov/rules/final/33-7881.htm
(3) The key words ‘profit warning’ and various combinations of ‘profit’ and ‘warn’ are entered into the search facility. Thesearch results are examined to confirm that they are profit-warning statements issued by UK quoted companies.
(4) This is not the only source of profit-warnings data. For example, Ernst & Young also compiles a database from profitwarnings as reported in the financial press, see www.ey.com/global/content.nsf/UK/Profit_Warnings. The two series arealmost identical.
(5) Less than 50% of 2001 year-end accounts was available when this analysis was undertaken. This limits the possibledegree of matching of accounts information with profit warnings for that year.
0
100
200
300
400
500
600
1997 98 99 2000 01
Number of warnings
Chart 1Annual number of profit warnings
294
BBaannkk ooff EEnnggllaanndd QQuuaarrtteerrllyy BBuulllleettiinn:: Autumn 2002
years available (1997–2001), comparing the balance
sheets of companies at the financial year-end
immediately before and after the warning. Multiple
warnings issued by the same firm within the same
financial year are counted as one observation. The
change in balance-sheet variables for firms issuing profit
warnings is benchmarked against a control group of
firms who did not issue a profit warning throughout the
period 1997–2001. There are approximately 3,000
observations in this control group.
Firms may issue profit warnings as a result of a change in
macroeconomic conditions; because of factors affecting
all firms in a specific sector; or following changes
specific to the company. Macroeconomic conditions are
common to all firms, and may be a major source of
changes in the total number of profit warnings.
However, overall macroeconomic conditions were
relatively stable during the period considered, albeit
with some marked weakening through 2001.
Furthermore, macroeconomic conditions are unlikely to
explain the differences in profitability changes between
those firms who issue profit warnings and those who do
not. To identify the role of sectoral shocks a comparison
would need to be made between firms making profit
warnings in a particular sector, and those in the same
sector not issuing a warning. That is not attempted
here.
Profitability
Throughout this article, the operating profit margin is
used as the measure of profitability. Profit margins are
not the only indicator of profitability, and others might
be used, such as the rate of return on capital. The
margin measure makes comparisons within the sample
easier. Further, use of the rate of return would introduce
the problem of trying to compare new-economy firms
who may have little tangible capital stock (and hence
capital for accounting purposes) with old-economy firms
with lots of capital stock.
Although profit warnings are issued merely when profits
are expected to be lower than previously envisaged, in
practice, they are associated with falling profit margins
for the majority of issuing firms (see Chart 3). Over
three quarters of the firms who issued a profit warning
in the sample (just over 500 firms) experienced a fall in
profit margins between the set of financial accounts
immediately preceding the warning and the subsequent
year-end set of accounts. Of the remaining quarter that
saw profit margins rise, some experienced lower growth
in margins than in the previous year, but the majority
(rather surprisingly) experienced an increase in the
growth rate of their margins. Firms who did not issue a
profit warning were less likely to experience a fall in
profit margins, and more likely to record an increase in
the growth of profit margins, than firms issuing profit
warnings (see Chart 4).
The median deterioration in profit margins (measured as
the proportional change) for the profit-warning cohort
was larger than for the no-warning control group (see
Table A). Indeed, the median no-warning control group
0
10
20
30
40
50
60
70
80
90
100
1997 98 99 2000 01
>90th75–90th50–75th
25–50th10–25th<10th
Per cent
Chart 2Share of profit warnings in firms of varyingprofitability(a)
(a) Each firm issuing a profit warning is assigned to a category based on its relativeprofitability in the year prior to the warning. The categories are defined by reference to the values of profit margins at various percentiles (ie firms above the 90th percentile or between the 75th and 90th percentile).
5%
77%
18%
OPM falls
OPM rises but growth rate falls
OPM rises at a faster rate
Chart 3Change in profitability for firms issuing profitwarnings(a)
(a) OPM is operating profit margin. The change is calculated between the set of financial accounts immediately pre and post-warning. The change in margins for the no-warning group is calculated over the equivalent set of financial accounts.
The balance-sheet information content of UK company profit warnings
295
experienced no deterioration in profit margins. Not only
do companies that issue profit warnings face a higher
likelihood of a fall in profit margins after the warning
than other firms, but they are also likely to experience a
similar fall (ie they are more concentrated around the
mean fall in margins). Around one in ten firms who
issued a profit warning moved from profit-making to
loss-making after the warning, around twice as many as
for companies that did not issue a warning (5%). Given
that the distribution of profitability is not of a standard
form (for example, normal), measures of statistical
significance for these results cannot be easily computed.
But the differences between the profit-warning and
non-warning groups appear to be indicative. These
differences do not seem to be related to the size of the
firm. Although larger firms have a greater propensity to
issue profit warnings, comparing changes in profit
margins across firms of similar size still shows that
firms who have issued warnings experience greater
reductions in margins than non-warning firms, whether
for all firms, or only those experiencing falling profit
margins.
So far, it has been shown that companies that issue
warnings are more likely than other companies to
experience a deterioration in profit margins. It is
possible that this is only a temporary effect, and that
profit margins soon return to the prewarning level.
Benito (2001) shows that there is a relatively high
degree of persistence of corporate profitability. The
analysis in this article suggests that this deterioration is
more persistent for companies issuing a profit warning
than for companies that did not issue a warning and yet
also experienced a fall in profit margins (see Chart 5).
After two years, around 80% of firms who had issued a
profit warning had margins still below prewarning levels,
compared with around 70% of the non-warning firms
who had lower profitability. There is some evidence that
profitability had not returned to prewarning levels even
after four years, but this applies only to firms issuing
warnings up to 1999 (ie halfway through the sample
period).
An alternative way of looking at persistence of
profitability is to use a transition matrix that shows the
proportion of companies that move from one quintile of
the distribution of profitability to another over a period
of a year. This is shown as Table B, averaged over
1997–2001, both for firms who issued a profit warning
and firms who did not. The diagonal elements give the
proportions that remain in the same profitability ranking
(quintile) between one year and the next—a measure of
persistence.(1)
35%
47%
18%
OPM falls
OPM rises but growth rate falls
OPM rises at a faster rate
Chart 4Change in profitability for firms not issuing profitwarnings(a)
(a) OPM is operating profit margin. The change is calculated between the set of financial accounts immediately pre and post-warning. The change in margins for the no-warning group is calculated over the equivalent set of financial accounts.
Table ASize of change in profitability Per cent
Median proportional change in profit margins: (a)
Profit-warning cohort No-warning control group
All firms -20.7 0.0
(a) Between the set of financial year-end accounts immediately before and after the warning.
0
10
20
30
40
50
60
70
80
90
100
+1 year +2 year +3 year +4 year
Profit warning
No profit warning Percentage
Chart 5Non-recovery rates of firms with falling profit margins(a)
(a) The non-recovery rate is the share of all firms who experience falling profit margins and do not see their margins recover to prewarning levels for 1, 2, 3 and 4 years.
(1) Some 738 firms issued warnings in the sample and so each row in the transition matrix represents around 150 firms.There are some 3,000-plus firms in the control group and so around 600 firms in each row.
296
BBaannkk ooff EEnnggllaanndd QQuuaarrtteerrllyy BBuulllleettiinn:: Autumn 2002
Consistent with the findings of Benito (2001) there is a
high degree of persistence for both groups, but firms
who issue warnings are more likely than other firms to
see their relative profitability position change in the
year following the profit warning. Tests described in
Proudman et al (1998, pages 49–50) show that these
differences between the two groups are statistically
significant.(1) Taken together, these two sets of findings
on persistence show that companies issuing a warning
are more likely to experience a change in profitability
and that the fall in profitability is then more persistent
than for other firms.
Balance-sheet changes
The impact of profit warnings on a company’s balance
sheet and profit and loss account might not be
confined to profit margins alone. Under the listing rules
for the dissemination of price-sensitive information, a
quoted company is obliged to notify the market as soon
as it anticipates lower profitability than previously
envisaged. The immediacy with which these disclosures
have to be made reduces the scope for firms to have
revised their balance sheets before the warning and
increases the probability of finding ‘knock-on’ revisions
to other balance-sheet variables in response to the
warning.
Firms might react initially to falling profitability by
reducing cash holdings because cash holdings are the
cheapest alternative source of income compared with
debt or equity (Myers and Majluf (1984)). Otherwise
companies might compensate for falling internal
funds by increasing external finance (ie issuing new
equity, increasing debt finance) and/or reducing
discretionary expenditures, such as those related to
employment, dividends or investment (see, for example,
Fazzari et al (1988), Bond and Meghir (1994),
Bernanke et al (1996), Benito and Young (2001) and
Nickell and Nicolitsas (1999)). This applies to all
firms who experience falling profitability, whether
following a profit warning or not. By comparing
the different responses of companies who have and
have not issued profit warnings to a fall in profitability
we can assess whether profit warnings have
incremental information for other balance-sheet
variables.
There is some evidence for a greater propensity for firms
who have experienced falling margins following a
warning to have subsequently increased gearing,
reduced investment rates and dividend payout ratios, in
comparison with firms who experienced falling margins
without issuing a warning. But there is little difference
in the propensity of both groups to experience a fall in
liquidity (Table C).(2) Profit-warning firms therefore
seem to become more financially frail following the
warning than do firms who also experience a fall in
margins but who do not issue a warning.
Conclusions
Previous studies have concentrated on the relationship
between profit warnings and share price changes. The
analysis described in this paper is new in that it
compares changes in profitability and other
balance-sheet indicators between firms who have issued
profits warnings and a control group of firms who have
not issued warnings over the same period. There are
three key findings.
First, over three quarters of firms experience falling
profit margins in the financial year in which they issue a
Table BA transition matrix of profitabilityProfit-warning firms No profit warning firms
Group 1 2 3 4 5 Group 1 2 3 4 5after/ after/before before
1 63 27 3 5 2 1 71 20 3 3 32 25 60 14 1 0 2 13 59 25 2 13 8 35 51 6 0 3 4 14 58 22 24 6 11 33 45 5 4 4 3 11 67 155 3 3 7 26 61 5 2 2 2 11 83
Notes: Groups 1–5 correspond to firms below the 20th percentile of profit margins, betweenthe 20th and 40th percentile, between the 40th and 60th percentiles, between the60th and 80th percentiles, and above the 80th percentile in each year respectively.Each row sums to 100% and individual entries describe the proportionate chance ofstaying in the same group both before and after the warning.
(1) Table B also implies that firms in the least profitable quintile have less persistence in profitability than firms in thehigher-profitability groups, irrespective of whether they issue warnings or not. This is also consistent with Benito (2001) who concludes that such companies are able to recover from periods of relatively poor performancemore rapidly than linear models of profit persistence would predict, conditional on their survival.
(2) The distribution of all these variables is non-normal and measures of statistical significance for these results cannoteasily be computed.
Table CIncremental information from firms with falling profitmargins
Issued a profit warning (a) Did not issue a warning (a)
Gearing rises 59.2 51.5Liquidity falls 56.2 56.4Investment rate falls 60.8 51.6Dividend payout rate falls 63.5 43.1
(a) 77% of firms who issued a warning also experienced a fall in profit margins; 47% of firmswho did not issue a warning experienced a fall in margins.
The balance-sheet information content of UK company profit warnings
297
warning. Fewer than 20% of firms issue a profit warning
and then experience increased growth in profit margins
in the same financial year.
Second, there is evidence that the decline in profitability
is not temporary, once it has occurred. Some 80% of
profit-warning firms in the sample did not see their
profit margins recover to prewarning levels within at
least two years. This persistence is more marked than for
firms who did not issue a warning.
Third, profit warnings are associated not only with
falling profitability, but also with adverse
developments in other balance-sheet variables—
increased gearing, reduced investment and
reduced dividend payout ratios—for greater
proportions of profit-warning firms than of non profit
warning firms. Profit warnings therefore appear to
have some incremental information for overall
balance-sheet health beyond that implied by falling
profitability.
Data appendix
Capital gearing at replacement costGross debt divided by capital stock measured at replacement cost. Gross debt (Datastream item (DS) 1301). Capital
stock is measured on a replacement cost basis. The raw data provide cost of plant and machinery (DS328), buildings
(DS327) and other assets (DS329) separately in gross historic cost terms.
Replacement cost capital stock, K, is estimated using the perpetual inventory method formula:
Kit+1 = (1 + Pt) Kit (1 – d) + IJit
where Pt is the inflation rate for the particular asset type in year t; d is the rate of depreciation of the asset. This is
assumed to be equal to 0.025 for buildings, 0.08 for plant and machinery and 0.05 for other assets; IJ is investment in
a particular asset. Total stocks and work in progress (DS364) are then added for each company in a particular year to
obtain the company capital stock in that year. For the company’s first observation, the replacement cost is assumed
equal to the gross historic cost.
DividendsOrdinary dividends net of advance corporation tax (DS187). Dividend-sales payout ratio is DS187 divided by sales
(DS104).
Investment rateInvestment is the proportionate change in the capital stock measured at replacement cost.
LiquidityRatio of cash and equivalents (DS375) to current liabilities (DS389).
Profit marginsEarnings before interest and tax divided by sales. This is calculated as pre-tax profit (DS137) plus net interest paid
(DS157-DS143), divided by sales (DS104).
298
BBaannkk ooff EEnnggllaanndd QQuuaarrtteerrllyy BBuulllleettiinn:: Autumn 2002
References
BBeenniittoo,, AA ((22000011)), ‘‘Oscillate wildly’: asymmetries and persistence in company-level profitability’, Bank of England
Working Paper no. 128.
BBeenniittoo,, AA aanndd YYoouunngg,, GG ((22000011)), ‘Hard Times or Great Expectations?: Dividend omissions and dividend cuts by
UK firms’, Bank of England Working Paper no. 147.
BBeerrnnaannkkee,, BB,, GGeerrttlleerr,, MM aanndd GGiillcchhrriisstt,, SS ((11999966)), ‘The financial accelerator and the flight to quality’, Review of
Economics and Statistics, Vol. 78, pages 1–15.
BBoonndd,, SS aanndd MMeegghhiirr,, CC ((11999944)), ‘Dynamic investment models and the firm’s financial policy’, Review of Economic
Studies, Vol. 61, pages 197–222.
CCllaarree,, AA ((22000011)), ‘The information in UK company profit warnings’, Bank of England Quarterly Bulletin, Spring,
pages 104–09.
FFaazzzzaarrii ,, SS,, HHuubbbbaarrdd,, RR aanndd PPeetteerrsseenn,, BB ((11998888)), ‘Financing constraints and corporate investment’, Brookings
Papers on Economic Activity, Vol. 1, pages 141–203.
MMyyeerrss,, SS aanndd MMaajjlluuff ,, NN ((11998844)), ‘Corporate financing and investment decisions when firms have information that
investors do not have’, Journal of Financial Economics, Vol. 13, pages 187–221.
NNiicckkeellll ,, SS aanndd NNiiccoolliittssaass,, DD ((11999999)), ‘How does financial pressure affect firms?’, European Economic Review,
Vol. 43, pages 1,453–56.
PPrroouuddmmaann,, JJ ,, RReeddddiinngg,, SS aanndd BBiiaanncchhii,, MM ((11999988)), ‘Is international openness associated with faster economic
growth’, in Proudman, J and Redding, S (eds), Openness and growth, pages 33–59, Bank of England.
299
Introduction
Inflation is ultimately a monetary phenomenon. Over
long periods of time, and across many different
countries, there has been a very close relationship
between the average annual rate of growth of the money
stock, and the average annual rate of increase of prices
(see Chart 1). This suggests that the monetary
aggregates should play an important part in the
Monetary Policy Committee’s assessment of the outlook
for inflation. However, extracting information from these
data is not a mechanical process. Over policy-relevant
time horizons, the monetary aggregates will be
influenced by many factors, such as cyclical shifts in the
demand for money and credit, and innovations in
financial structure, products and regulation. That
explains why, in the short term, the correlation between
monetary growth and inflation is much less marked (see
Chart 2), and why, soon after its inception, the MPC
took the decision not to reinstate the monitoring ranges
for money that had been in place under the previous
Government.(3)
It is important to understand and quantify the factors
affecting monetary variables, and to draw out the key
implications for the inflation outlook. In carrying out
this task, the Bank of England can draw on a substantial
quantity of published research in monetary economics,
together with a wide variety of other tools for economic,
Money and credit in an inflation-targeting regime(1)
This article is one of a series on the UK monetary policy process.(2) It discusses how the assessment ofmoney and credit data fits into the Bank’s quarterly forecast round.
(1) An earlier version of this paper was presented at a workshop ‘Monetary analysis: tools and applications’ held at theEuropean Central Bank in Frankfurt on 20–21 November 2000. It was subsequently published in Klöckers andWilleke (2001).
(2) See also Bean and Jenkinson (2001) and Clews (2002).(3) These charts are taken from King (2002). A full account of the decision not to reinstate monitoring ranges is given in
the box on pages 8–9 of the November 1997 Inflation Report.
By Andrew Hauser and Andrew Brigden of the Bank’s Monetary Assessment and Strategy Division.
Chart 1Average annual inflation and broad money growth rates over the past 20 years(a)
0
20
40
60
80
100
0 20 40 60 80 100
Money growth, per cent
Inflation, per cent
Chart 2Average annual inflation and broad money growth rates over the past 2 years(a)
0
20
40
60
80
100
0 20 40 60 80 100Money growth, per cent
Inflation, per cent
(a) Based on data up to 1999 for 116 different countries. For presentational purposescountries with average inflation or broad money growth rates exceeding 100% have been excluded.
(a) Based on data up to 1999 for 116 different countries. For presentational purposescountries with average inflation or broad money growth rates exceeding 100% have been excluded.
300
BBaannkk ooff EEnnggllaanndd QQuuaarrtteerrllyy BBuulllleettiinn:: Autumn 2002
institutional and statistical analysis.(1) These tools
are used in many different ways to brief the MPC.
This article focuses on how formal theoretical and
empirical models are used to provide a quantitative
evaluation of the information in money and credit
aggregates as an input to the MPC’s quarterly inflation
forecast.
Money, credit and the Bank’s suite of models
Bean and Jenkinson (2001) discuss how and why MPC
members use a ‘suite’ of economic models in forming
their overall forecast judgments. One element of this
suite is a macroeconometric model (MM), which
includes equations for each of the key behavioural
relationships in the UK economy. That model is under
continual development; the most recent published
version was in September 2000.(2) But it is clear that,
given substantial uncertainty over the true structure of
the economy, and the need in any model-building
exercise to focus on some economic interactions at the
expense of others, no single model is likely to be able to
encompass all possible factors bearing on the inflation
outlook. The MM is therefore supplemented by a suite
of auxiliary models, including a number of models based
on money and credit data, designed to provide answers
to the types of question not readily addressed within the
MM, to analyse specific policy issues, or to act as a
potential check on the output of the MM.
The MM has three central properties desirable of any
model used for providing monetary policy advice. First,
the long-run behaviour of real variables, such as GDP
and employment, is independent of the level of prices.
Second, there is no long-run trade-off between inflation
and output. And third, both the level of prices, and the
rate of inflation, depend ultimately on monetary policy.
However, in common with nearly all large-scale
macroeconometric models used in other central banks
and research institutions, money in the MM reacts
passively to changes in measures of economic activity,
wealth and the rate of interest. Active roles for money
and credit in the transmission mechanism are less
well-developed. Consequently, if the inflation
projections relied exclusively on the MM, then important
factors affecting the outlook for inflation might be
overlooked. It is therefore important that the MPC has
access to alternative models capable of capturing and
quantifying any incremental information in money and
credit. Research at the Bank of England over the past
few years has generated a rich set of empirical and
theoretical models suitable for this purpose.(3) Bank
staff use these models to provide the MPC with
an updated analysis of the potential information in
money and credit at an early stage in each forecast
round.
Aims and objectives of the monetaryassessment process
Implicit in the suite-of-models approach is a recognition
that views about the role played by different variables,
including (alternative definitions of) money and credit,
may vary across academics, informed commentators and
policy-makers. A central aim of the monetary assessment
process is to reflect this range of views by offering
policy-makers a menu of alternative projections and
simulations, depending on the extent to which monetary
variables are thought to play an active role in the
transmission mechanism.
Monetary variables may be helpful to policy-makers in a
number of different ways.
11 MMoonneeyy aanndd ccrreeddiitt aass sshhoorrtt--rruunn iinnddiiccaattoorr
vvaarriiaabblleess.. Monetary statistics are available more
rapidly than most other economic data, and are
usually drawn from a complete population, making
them less vulnerable to sampling errors.
Consequently, they might assist policy-makers by
providing an early, independent read on economic
events. Statistical evidence on the value of this
information is given in Astley and Haldane (1995).
22 MMoonneeyy aanndd ccrreeddiitt aass aa ssoouurrccee ooff
iinnccrreemmeennttaall iinnffoorrmmaattiioonn oonn tthhee
ttrraannssmmiissssiioonn ooff sshhoocckkss.. Data on money and
credit can provide policy-makers with incremental
information on the transmission of shocks through
the economy, at least at certain points in the
economic cycle. This sort of information can have
a bearing on economic developments over a
number of years. It is one area where monetary
models are able to complement the output of the
MM, and help to illuminate key issues in the
preparation of the inflation projection. Taking this
information on board reduces the probability of
making policy mistakes.
(1) A comprehensive list of references is provided at the end of this paper.(2) Bank of England (2000).(3) See, for example, Nelson (2002) for a recent study of the possible incremental information contained in M0, also
known as base money.
Money and credit in an inflation-targeting regime
301
33 MMoonneeyy aanndd ccrreeddiitt aass iinnddiiccaattoorrss ooff tthhee
uunnddeerrllyyiinngg ssttaannccee ooff mmoonneettaarryy ppoolliiccyy.. The
monetary aggregates might provide incremental
information on the tightness or looseness of policy
over and above other measures, such as the
short-term interest rate and the amount of spare
capacity in the economy. Though related to the
second view, in the sense that money contains
information incremental to other variables, the
distinguishing characteristic of this view is the
explicit link to the underlying stance of policy itself.
Money and credit as short-run indicatorvariables
Money and credit aggregates can be informative about
the short-run outlook for activity and inflation. Bank
staff use models based on household and corporate data
to draw conclusions about the near-term outlook for
consumption and investment, and models based on
whole-economy aggregates to generate projections for
aggregate demand. The Bank of England has for some
time now looked particularly closely at sectoral money
and credit data, reflecting a belief that sectoral
relationships are likely to be more stable than those at
an aggregate level.(1)
The short-run outlook for consumption
Most of these indicator-variable projections are based on
a modelling philosophy set out in Thomas (1996, 1997a,
1997b) and Janssen (1996). The first stage is to estimate
a system of equations for households’ money holdings,
consumption, income, wealth, inflation and interest
rates. From this system, two long-run relationships are
identified, which—after testing and imposing a set of
theory-based restrictions—are interpreted as long-run
equilibrium consumption and long-run equilibrium
money holdings. The table shows examples of the sorts
of equations that have been obtained, taken from
Thomas (1997a). Households’ long-run equilibrium
money holdings depend positively on income and
wealth, with a joint coefficient of unity, positively on the
relative interest rate on deposits, and negatively on
inflation (which measures the relative return between
real and financial assets). Consumption depends
positively on income and wealth—again with a joint unit
coefficient—negatively on a measure of the real interest
rate, and negatively on the change in unemployment, a
term intended to capture the effects of precautionary
saving.
If households’ actual money holdings were always in
long-run equilibrium, as represented by the equation for
m* in the table, and if all of the right-hand side variables
in the equation were measured without error in real
time, then money would contain no incremental
information about the economy. In practice, however,
none of these conditions are likely to hold. Even if
money holdings were entirely demand determined,
monetary data are available earlier, and with less chance
of error, than those on incomes, prices and wealth, and
so are helpful to policy-makers as short-run indicator
variables.
But households’ money holdings may not always be in
long-run equilibrium. If adjusting portfolios is costly, or
yields are sluggish to adjust, households may be willing
to accept higher or lower money balances for a short
period of time, as a temporary abode of purchasing
power.(2) Long-run equilibrium is then restored only
gradually, as households attempt to eliminate any
excess holdings through purchases of goods, and
financial and physical assets. In this scenario, money
also plays an incremental role in the transmission of
shocks.
These considerations suggest that there may be useful
information in both money growth, and in any gap
between actual money holdings and estimated
equilibrium holdings. To capture both of these potential
effects econometrically, the estimated long-run
relationships are embedded in separate equations for
the quarterly growth rates of money and credit.
Consistent with the discussion above, residuals from the
long-run money demand function are found to help
explain the path of consumption. So the resulting
system expresses consumption growth as a function of
both monetary growth and ‘money gaps’. For given
assumptions about the exogenous variables, the model
Long-run money demand and long-run consumptionequationsHouseholds’ long-run equilibrium money holdings (m*)
m*t = 0.5yt + 0.5wt + 0.44(idt – it) – 1.6pt
Households’ long-run equilibrium consumption (c*)
c*t = 0.9yt + 0.1wt – 0.64(it – pt) – 1.2Dut
m = log real household sector M4 y = log real household sector disposable income w = log real gross household sector wealth id = weighted average interest rate on household sector M4 i = three-month Treasury bill rate p = annualised rate of increase of consumption deflator c = log real consumption expenditure u = unemployment rate
(1) For further details on this see, for example, Thomas (1997a).(2) This is sometimes referred to as the buffer stock theory of money.
302
BBaannkk ooff EEnnggllaanndd QQuuaarrtteerrllyy BBuulllleettiinn:: Autumn 2002
can therefore be used to generate ‘money-based’
projections for future consumption growth.
Even within the class of monetary models, there is
debate about which monetary aggregate best captures
the concept of money likely to be most closely associated
with activity. So, as part of the monetary assessment
process, forecasts are presented from models using both
household M4 and household Divisia money, together
with output from a model based on narrow money.(1)
These projections have been helpful in highlighting
influences on consumption which may have been missed
by other forecasting approaches that omit an explicit
role for money.
The short-run outlook for investment
A similar exercise is carried out for the corporate sector.
The model used is based on Brigden and Mizen (1999),
and is functionally similar to the consumption model
described above. The main difference is that, in
addition to equations for corporate money demand and
investment expenditure, the model also includes a
long-run relationship describing firms’ bank borrowing.
The rationale for this is that firms have greater capacity
to manage their net liquidity position than households,
so looking at only one side of their balance sheets could
give a particularly misleading picture. The model relates
investment growth to both change and disequilibrium
terms in firms’ money and credit. This type of research
can help to enrich the analysis of particular forecast
issues, even where the projections from the auxiliary
models themselves are not given a central role in the
MPC’s own assessment.
Recent years have seen significant changes in the
structure of corporate finance in the United Kingdom. A
stronger government fiscal position and lower inflation,
together with increased internationalisation and
innovation in capital markets, have led firms
progressively to shift more of their liabilities from bank
to market-based finance (see Chart 3). As a result, some
of the traditional relationships between bank deposits,
credit and investment have become less reliable
over time. The links between firms’ overall financial
position, activity and inflation remain a key focus of
policy-makers’ concerns, however. Recently, Bank staff
have constructed a ‘financial accelerator’ model of the
corporate sector, which is discussed later in this article.
The short-run outlook for aggregate nominal activity
Bank staff also use systems of equations based on
whole-economy (rather than sectoral) money aggregates
to derive short-run forecasts for nominal GDP, inflation
and real activity. Again, a range of projections are
presented from models based on alternative monetary
aggregates: narrow money (M0), retail money (M2) and
broad money (M4). All of these potentially contain
information about the stance of monetary policy.
The reason for producing models based on both M2 and
M4 relates mainly to the behaviour of non-bank
financial firms, or ‘OFCs’, the money holdings of which
are captured in M4, but have little impact on M2.
Recent years have seen sharp growth in OFCs’ balance
sheets, associated with the structural changes to the
corporate sector discussed above, the growth in asset
prices, the increased use of securitisations of retail
portfolios, and so on. This growth has also been
associated with significant fluctuations in OFCs’ money
holdings, and therefore in the rate of growth of M4 (see
Chart 4). Some commentators believe that the money
holdings of at least part of the OFC sector may have a
causal impact on inflation via their effects on asset
prices. Others prefer to see OFCs’ money as primarily
reflecting volatile financial activity, or the precise way in
which financing arrangements are constructed, with
little implication for the prices of goods and services.(2)
In the absence of conclusive evidence it is important
that policy-makers have access to projections consistent
with either interpretation. Bank staff therefore present
forecasts from both M2 and M4 models.
Chart 3Stock of PNFCs’ borrowing from banks and buildingsocieties as a fraction of their total liabilities
6
7
8
9
10
11
12
13
14
15
1976 80 84 88 92 96 2000
Per cent
0
(1) Divisia is a measure of money that weights together different components—including notes and coin, and sight andtime deposits—according to their relative liquidity. The construction of Divisia money is discussed in Fisher, Hudsonand Pradhan (1993).
(2) These issues have been regularly discussed in the Bank’s Inflation Report. See, for instance, the box on pages 6–7 ofthe May 1998 Report. For a recent empirical study of OFCs’ money and credit holdings see Chrystal and Mizen (2001).
Money and credit in an inflation-targeting regime
303
Money and credit as a source of incrementalinformation on the transmission of shocks
A second strand of the regular assessment of money and
credit data focuses on their role in the transmission
mechanism. Much of this involves a careful assessment
of prevailing credit conditions. A key insight from the
‘credit channel’ literature is that—contrary to the
predictions of more traditional economic models—
credit may play a separate role in the propagation of
economic shocks.(1) This issue is likely to be most acute
at times of potential financial instability, such as during
the events related to Russia and to Long-Term Capital
Management (LTCM) in 1998, when policy-makers spent
considerable time assessing the implications for the
financial health of the UK financial and private
sectors.(2) But it is not necessarily limited to such
periods. As the recent financial accelerator literature
stresses, if the effective cost of capital depends not just
on observed rates but also on agents’ financial positions,
credit conditions could also contain incremental
information relevant to an assessment of the true
cyclical position of the economy.(3)
At present, much of the assessment in this area is based
on careful data analysis, backed by a substantial pack of
charts and tables summarising the latest developments
in credit conditions. The analysis includes a review of
households’ and firms’ financial positions—including an
assessment of gearing levels—a review of non-price
credit effects and an appraisal of the private sector’s
borrowing capacity. A simple example of the type of data
on which this analysis is based is given in Chart 5,
which compares households’ debt to wealth and debt to
income ratios, two alternative measures of household
gearing. In preparing this work for the MPC, Bank staff
working in Monetary Analysis are able to draw heavily on
assessments prepared by colleagues in the Financial
Stability area.(4)
Increasingly, however, this analysis is also being
informed by the type of more formal theoretical and
empirical work discussed elsewhere in this article. The
Bank has developed a calibrated financial accelerator
model of the UK corporate sector. This model,
described in Hall (2001), links firms’ spending
behaviour to their overall financial position via an
external finance premium.(5) It offers a potentially rich
range of insights into the role of credit in the
transmission mechanism, the effects of financial
innovation, and the impact of asset price movements. A
similar model has recently been developed for the
household sector.(6) And it is hoped that in the future
these theoretical models will be supplemented by more
detailed econometric studies based on micro data sets.
An attractive aspect of the calibrated theoretical models
is that they provide a quantitative link between measures
of financial health and expenditures. They also allow
(1) A summary of key aspects of the credit channel literature is given in Bernanke and Gertler (1995).(2) See, for instance, Section 1 of the November 1998 Inflation Report.(3) For a discussion of the policy implications of a credit channel, and other financial frictions, see Bean, Larsen and
Nikolov (2002).(4) A full overview of these issues is given in the Financial Stability Review (published twice a year), which is also available
on the Bank’s web site.(5) More specifically, the external finance premium (the difference between the cost of borrowing money and the cost of
using one’s own funds) varies inversely with the size of the firm’s balance sheet.(6) See Aoki, Proudman and Vlieghe (2001).
Chart 4Growth rates of the monetary aggregates
0
2
4
6
8
10
12
14
1992 94 96 98 2000 02
Percentage changes on a year earlier
M4
M0
M2
Chart 5Household sector gearing ratios
15
17
19
21
23
25
1987 89 91 93 95 97 99 2001
80
90
100
110
120
130Per centPer cent
Debt to net wealth (a)
Debt to income (b)
00
(a) Ratio of households’ total financial liabilities to the sum of households’ net financial and physical assets.
(b) Ratio of households’ total financial liabilities to households’ annualised gross disposable income.
304
BBaannkk ooff EEnnggllaanndd QQuuaarrtteerrllyy BBuulllleettiinn:: Autumn 2002
staff to perform a wide variety of simulations, based on
alternative assumptions about possible future shocks or
changes to the structure of the economy. An example of
this type of analysis is given in Chart 6, which shows
how the response of investment in the corporate sector
model to a monetary policy shock varies under
alternative assumptions about gearing levels. Because of
its relative simplicity, the model does not capture every
aspect of the data—such as the speed of adjustment in
response to shocks—particularly well. But it can, for
example, be used to provide an estimate of the
incremental effect of higher gearing on the level of
investment under alternative scenarios.
Money and credit as indicators of theunderlying stance of monetary policy
The final element of the monetary assessment process is
based on the view that money and credit data might
contain information about the underlying monetary
stance.
Each quarter, Bank staff review the output of a number
of simple policy ‘rules’.(1) One of these rules—the Taylor
rule—relates the short-term interest rate to the
deviation of inflation from target in the previous period,
and to an estimate of the output gap in the previous
period.(2) Another rule, the Brainard rule,(3) suggests a
way for policy-makers to behave when they are uncertain
about some of the key parameters in the economy.(4) A
third rule, the McCallum rule, effectively puts more
weight on signals from the monetary data.(5)
The McCallum rule is of the following form:
Mt = x* – vt-1 + 0.5(x* – xt-1) ((11))
where M is the rule’s prescription for narrow money
growth, x is actual nominal income growth, x* is nominal
income growth in steady state,(6) and v is trend velocity(7)
growth (a four-quarter moving average of narrow money
velocity growth). The rule implies that narrow money
growth should be lowered whenever nominal income
growth is above its steady-state value. The velocity term
in the rule also allows for changes in the rule’s
prescriptions due to shifts in money demand,(8) which
should be accommodated by the monetary authority.(9)
The staff use the output of each rule as an illustrative
benchmark. Whenever the actual policy rate differs
substantially from that suggested by either the Taylor or
the Brainard rule, or the growth rate of money differs
substantially from that suggested by the McCallum rule,
it is necessary for staff to consider why.
In addition to these measures, the MPC is provided with
an analysis based on a small monetary model, known as a
structural vector autoregression (SVAR), of the UK
economy, developed by Dhar, Pain and Thomas (2000).
An important lesson drawn from experience with earlier
models is that, in general, there is no single
deterministic link between money, activity and inflation.
In practice, all three are determined simultaneously, so
the precise relationships between them will depend on
the underlying shocks affecting the economy at each
point in time. Policy-makers should therefore beware of
drawing firm inferences about future inflationary
pressures solely on the basis of past correlations
Chart 6Response of investment to a positive monetary policy shock of 100 basis points(a)
-6
-5
-4
-3
-2
-1
0
0 1 2 3 4 5 6 7 8 9 10 11 12
Percentage deviation from steady-state level
Quarters after shock
Low corporate gearing
High corporate gearing
(a) Based on a simulation of the model described in Hall (2001).
(1) Some of these simple rules are described in more detail in Stuart (1996).(2) Taylor’s original rule for the United States is given in Taylor (1993).(3) Named after Brainard (1967).(4) The Bank’s Brainard rule is derived from a small four-equation model for output, inflation, the short-term interest rate
and the sterling effective exchange rate. It is described in Martin and Salmon (1999).(5) McCallum’s original rule for the United States is given in McCallum (1988).(6) This is based on an inflation target of 2.5% and assumed trend real GDP growth.(7) The velocity of circulation of money is the speed with which money circulates throughout the economy over a given
time period. Velocity is given by nominal GDP divided by the money stock.(8) Inflation in monetary models is driven by excess money supply. So shifts in money demand should be met by an
increase in the money supply in order to maintain equilibrium in the money market and keep inflation stable.(9) Persistent differences between the McCallum rule’s prescriptions and actual money base growth in the absence of
inflationary pressure could also indicate trend velocity shifts or changes in the economy’s trend growth rate.
Money and credit in an inflation-targeting regime
305
between simple monetary statistics and other economic
variables. Instead, a view must be taken on the likely
structural shocks driving the economy. SVARs represent
one way of attempting to identify these shocks.
The monetary SVAR model is estimated as follows.
Starting from a system of eight equations for real money,
income, inflation and a set of asset prices, four long-run
relationships are identified, including a money demand
equation, a term structure relationship, and an
asset-pricing function. Theory-based restrictions are
then used to identify four permanent and four
temporary shocks. These include both temporary and
permanent monetary policy shocks, and two types of
velocity shock, one reflecting changes in the provision of
credit by the banking system, and one reflecting changes
in liquidity preference.
In the model, different shocks have different
implications for the observed correlations between
money, measures of economic activity, and prices. This
has long been understood from the monetary assessment
side. But the formalisation and quantification of this
idea represents an important advance from this line of
research. The model also provides Bank staff with a
highly flexible tool for use in longer-term policy analysis
and advice.
Summary
The inflation projection published in the quarterly
Inflation Report is drawn up by, and owned by, the MPC.
In producing its forecast, assisted by Bank staff, the
Committee is able to draw on the analysis of the
monetary models, other parts of the suite of models, or
any other sources. In principle, the MPC might use
output from the monetary models in several ways. First,
the indicator-variable projections might be used to help
form a judgment about the evolution of different
components of expenditure during the preceding
quarter,(1) improving the data set for the very near-term
outlook. Second, the analysis on money and credit in
the transmission mechanism might be used to adjust the
output of some of the equations in the MM or elsewhere
where the money or credit data are thought to contain
incremental information. Third, the longer-run
projections and measures of the monetary stance
provide a potential cross-check on the provisional
outputs of the forecast. And, fourth, any or all of this
material could help the MPC to assess the risks around
possible central projections, reflected in the inflation
and GDP fan charts.
The monetary models, like other parts of the Bank’s
suite of models, are subject to continual review. In
providing a menu of alternative projections, it is
recognised that different policy-makers may have
different interpretations of, or put different weights
on, developments in the money and credit data. Some
of the models described here may also be thought to
have more relevance at some points in the cycle, or in
certain market conditions, than others. There is
therefore no mechanical link between the material
provided, and the MPC’s final projections for GDP and
inflation. Nevertheless, it is important that Committee
members are made aware of, and are able to draw on, the
best possible technical representations of these
alternative paradigms when taking monetary policy
decisions.
(1) Official estimates of many economic time series will not be available for either the current or the preceding quarter atthe time each Inflation Report is published.
306
BBaannkk ooff EEnnggllaanndd QQuuaarrtteerrllyy BBuulllleettiinn:: Autumn 2002
References
AAookkii ,, KK,, PPrroouuddmmaann,, JJ aanndd VVlliieegghhee,, GG ((22000011)), ‘Why house prices matter’, Bank of England Quarterly Bulletin,
Winter, pages 460–68.
AAssttlleeyy,, MM aanndd HHaallddaannee,, AA GG ((11999955)), ‘Money as an indicator’, Bank of England Working Paper no. 35.
BBaannkk ooff EEnnggllaanndd ((22000000)), Economic models at the Bank of England, September 2000 update.
BBeeaann,, CC aanndd JJeennkkiinnssoonn,, NN ((22000011)), ‘The formulation of monetary policy at the Bank of England’, Bank of England
Quarterly Bulletin, Winter, pages 434–41.
BBeeaann,, CC,, LLaarrsseenn,, JJ aanndd NNiikkoolloovv,, KK ((22000022)), ‘Financial frictions and the monetary transmission mechanism:
theory, evidence and policy implications’, European Central Bank Working Paper no. 113.
BBeerrnnaannkkee,, BB aanndd GGeerrttlleerr,, MM ((11999955)), ‘Inside the black box: the credit channel of monetary policy transmission’,
Journal of Economic Perspectives, Vol. 9 pages 27–48.
BBrraaiinnaarrdd,, WW ((11996677)), ‘Uncertainty and the effectiveness of policy’, American Economic Review, Vol. 57 pages 2–19.
BBrriiggddeenn,, AA aanndd MMiizzeenn,, PP ((11999999)), ‘Money, credit and investment in the UK corporate sector’, Bank of England
Working Paper no. 100.
CChhrryyssttaall ,, KK AA aanndd MMiizzeenn,, PP ((22000011)), ‘Other financial corporations: Cinderella or ugly sister of empirical monetary
economics?’, Bank of England Working Paper no. 151.
CClleewwss,, RR ((22000022)), ‘Asset prices and inflation’, Bank of England Quarterly Bulletin, Summer, pages 178–85.
DDhhaarr,, SS KK,, PPaaiinn,, DD aanndd TThhoommaass,, RR ((22000000)), ‘A small structural empirical model of the UK monetary transmission
mechanism’, Bank of England Working Paper no. 113.
FFiisshheerr,, PP GG,, HHuuddssoonn,, SS aanndd PPrraaddhhaann,, MM ((11999933)), ‘Divisia indices for money: an appraisal of theory and practice’,
Bank of England Working Paper no. 9.
HHaallll ,, SS ((22000011)), ‘Financial accelerator effects in UK business cycles’, Bank of England Working Paper no. 150.
JJaannsssseenn,, NN ((11999966)), ‘The demand for Divisia money by the personal sector and by industrial and commercial
companies’, Bank of England Quarterly Bulletin, Vol. 36, pages 405–11.
KKiinngg,, MM ((22000022)), ‘No money, no inflation—the role of money in the economy’, Bank of England Quarterly Bulletin,
Summer, pages 162–77.
KKllööcckkeerrss,, HH--JJ aanndd WWiilllleekkee,, CC ((22000011)), Monetary analysis: tools and applications, Frankfurt: European Central
Bank.
MMaarrttiinn,, BB aanndd SSaallmmoonn,, CC ((11999999)), ‘Should uncertain monetary policy-makers do less?’, Bank of England Working
Paper no. 99.
MMccCCaalllluumm,, BB ((11998888)), ‘Robustness properties of a rule for monetary policy’, Carnegie-Rochester Conference Series
on Public Policy, Vol. 29, pages 53–84.
Money and credit in an inflation-targeting regime
307
NNeellssoonn,, EE ((22000022)), ‘Direct effects of base money on aggregate demand: theory and evidence’, Journal of Monetary
Economics, Vol. 49, pages 687–708.
SSttuuaarrtt,, AA ((11999966)), ‘Simple monetary policy rules’, Bank of England Quarterly Bulletin, Vol. 36, No. 3, pages 281–87.
TTaayylloorr,, JJ ((11999933)), ‘Discretion versus policy rules in practice’, Carnegie-Rochester Conference Series on Public Policy,
Vol. 39, pages 195–214.
TThhoommaass,, RR ((11999966)), ‘Understanding broad money’, Bank of England Quarterly Bulletin, Vol. 36, pages 163–79.
TThhoommaass,, RR ((11999977aa)), ‘The demand for M4: a sectoral analysis—Part 1—The personal sector’, Bank of England
Working Paper no. 61.
TThhoommaass,, RR ((11999977bb)), ‘The demand for M4: a sectoral analysis—Part 2—The corporate sector’, Bank of England
Working Paper no. 62.
308
The failure of consumption models to predict the fall in
the United Kingdom’s savings ratio in the late 1980s and
its rise in the early 1990s led some economists to look at
models of forward-looking consumers who may be
unable to borrow. A theoretical weakness in these early
papers is that they assumed that the proportion of
liquidity-constrained individuals does not change. In
the United Kingdom, increased competition in the
lending market in the 1980s eased restrictions on
borrowers, and is likely to have reduced the number of
credit-constrained consumers. So models that assume
that the proportion of constrained individuals remains
constant through time may not match UK experience.
To address this shortcoming, some economists have
specified forward-looking consumption functions that
assume that the proportion of credit-constrained
consumers is inversely related to proxies for financial
liberalisation. Their results suggest that this method is
able to explain UK consumption data.
This paper examines whether recent UK consumption
behaviour can indeed be explained using this method.
To this effect we construct a proxy for financial
liberalisation (FLIB). FLIB is defined as the sum of the
constant and the residuals in a regression of the loan to
value ratio on the house price to income ratio, the
nominal post-tax mortgage rate and a two-year moving
average of the post-tax mortgage rate. We then
re-examine a forward-looking consumption model which
uses FLIB as a variable identifying the proportion of
liquidity-constrained individuals. We find that this
implementation of the model, which inevitably embodies
joint hypotheses about consumption behaviour and
about the measurement of financial liberalisation, is not
robust and does not give a plausible picture of the
number of people who were liquidity constrained in the
1990s.
We argue that one possible explanation for these results
is that the liberalisation proxy is unable to depict
accurately the consequences of UK financial
deregulation in the 1990s. FLIB’s behaviour in the
1990s suggests that all the liberalisation that occurred
in the 1980s was reversed the following decade, which
seems implausible. This in turn means that, by assuming
that the proportion of constrained agents in the
economy is a function of FLIB, the consumption model
examined in this paper does not derive a plausible
measure of this key variable. We argue that the mapping
from the FLIB index of liberalisation to the proportion of
constrained consumers is somewhat arbitrary and that
some of the assumptions made to derive a functional
form for an estimatable UK consumption function might
account for the failures encountered in this paper.
Finally, we attempt to explain the behaviour of FLIB in
the 1990s. We identify the sharp reduction in the
nominal interest rate as the main factor accounting for
FLIB’s reversal over the 1990s. We also argue that
lending institutions may have changed the emphasis of
their lending criteria towards loan to income ratios after
liberalisation.
Financial liberalisation and consumers’ expenditure: ‘FLIB’re-examinedWorking Paper no. 157
Emilio Fernandez-Corugedo and Simon Price
309
Miles Kimball defines a precautionary motive as ‘any
aspect of an agent’s preferences which causes a risk to
affect decisions other than the decision of how
strenuously to avoid the risk itself and risks correlated
with it (which is governed by risk aversion). A
precautionary motive leads an agent to respond to a risk
by making adjustments that will help to reduce the
expected cost of the risk’. Thus, precautionary saving
arises when forward-looking consumers accumulate
wealth today for the purpose of reducing the impact of
future uncertainty on future consumption decisions.
Liquidity constraints arise when consumers have
difficulties to obtain credit. More specifically, soft
liquidity constraints represent the situation where
consumers are able to borrow, but incur penalties which
increase with the amount borrowed. Hard liquidity
constraints refer to the unavailability of credit
altogether.
The modern consumption literature has examined the
problem of how much to consume and save each period
under two polar scenarios. One scenario considers
perfect capital markets where no barriers to borrowing
exist and where interest rates are the same for savers and
borrowers. The other scenario assumes that consumers
are not able to borrow at all. Both scenarios, however,
do not seem to match what is commonly observed in
developed economies: consumers often borrow and face
interest rates that are higher for debt than for saving.
The theoretical implications for consumption arising
from the two polar cases are summarised by Carroll and
Kimball in two papers that provide the conditions under
which the introduction of uncertainty and liquidity
constraints leads to precautionary saving, and analyse
how precautionary saving. Technically, these conditions
require the interaction of risk (either to labour income
or to the rate of return) with liquidity constraints and/or
with certain functional forms for the utility function.
The literature finds at least three important implications
of the inability to borrow (hard constraints) for
consumption. First, hard constraints increase
precautionary saving around levels of wealth where the
constraints bind. Second, if consumers face the
possibility of becoming constrained at any point in the
future, they will behave as if they were constrained
today, even in the absence of a current liquidity
constraint. Finally, the introduction of further
borrowing constraints does not necessarily lead to an
increase in precautionary saving.
This paper considers the implications for consumption
behaviour when households are allowed to borrow, but
face penalties that increase with the amount
borrowed. The introduction of this type of constraint
does not lead to consumers behaving very differently
from consumers who face hard constraints. A soft
constraint increases precautionary saving and affects
prior periods, although the introduction of further soft
constraints can lead to lower precautionary saving.
However, a new result is that the amount of
precautionary saving is reduced when hard
constraints are relaxed and become soft. The intuition
behind this result is simple: when consumers cannot
borrow, they must have savings to avoid shocks that
could leave them with low levels of income. A
relaxation of the borrowing constraint means that
consumers do not need to have these (high) savings to
avoid adverse shocks to income. More technically, the
paper shows the effects that soft liquidity constraints
have on the value, marginal value and consumption
functions in a dynamic programme. The introduction of
a soft constraint makes consumers more averse to risk
(since the value function becomes more concave) and
also more prudent (since the marginal value function
becomes more convex). An implication is that the
resulting consumption function becomes concave with
respect to wealth.
Soft liquidity constraints and precautionary savingWorking Paper no. 158
Emilio Fernandez-Corugedo
310
This paper discusses the impact of demographic change on the
UK economy, looking at effects on GDP growth and GDP per
head, saving and capital investment, interest rates, asset prices
and the distribution of national income. It also considers the
risks associated with demographic change. A key finding,
widely supported in the academic literature, is that even under
relatively cautious assumptions about technological progress
and capital accumulation, aggregate living standards (as
measured by GDP per head) are set to double over the next
50 years. While there are clear risks to this aggregate outlook,
these would be present even without demographic change.
The impact of ageing on the rate of saving and capital
accumulation is one of the key uncertainties surrounding any
projection of long-term growth. The paper analyses this in the
context of a model where people are reliant on their own
saving for their retirement income and considers three
different types of demographic shocks: a baby boom, an
increase in longevity and a decline in fertility. The overlapping
generations model used for this purpose makes it possible to
assess the impact of these shocks on the welfare of different
generations under different assumptions about household
behaviour. It finds that a baby boom has an adverse effect on
the baby boom generation for the obvious reason that when
they are of working age their abundance drives down wages
and when they are of retirement age the abundance of their
saving drives down the rate of return. The impact of a baby
boom on other generations is largely beneficial. Increases in
longevity, not accompanied by changes in labour supply, have a
detrimental effect on annual consumption per head for the
obvious reason that people have more years over which to
spread their consumption. Changes in fertility appear to have
very little effect on individual consumption per head, although
they clearly affect aggregate quantities because of changes in
the number of people.
An important conclusion of these models is that while
individual consumption over the life cycle may not be strongly
affected by demographic change, there can be large effects at
particular parts of the life cycle when individuals do not
attempt to spread their consumption evenly. For example, the
analysis of greater longevity suggests that this might reduce
individual life-time consumption by about 2% if the change
is spread evenly over time. But if individuals follow
rule-of-thumb behaviour prior to retirement and do not
accumulate enough assets, the reduction in their life-time
consumption will be concentrated into the years when they are
old. This is particularly important at the current juncture
since many people in prime saving age will observe their own
pensioner parents living longer without any obvious adverse
effect on their consumption. This could be misunderstood as
suggesting that their own saving for retirement is adequate.
Yet the formal model suggests that the early generations to
benefit from greater longevity do not have to reduce their
consumption since the capital accumulated by previous
generations is not affected by their longevity. But their
children will receive smaller bequests. Moreover, the current
generation of pensioners has benefited from extraordinarily
high asset returns which are unlikely to be repeated.
The implications of this analysis for interest rates are modest.
This is consistent with other research which suggests that the
effect of demographic change on asset prices more generally is
likely to be small. This leads on to the second conclusion of
this paper, that the risks to the living standards of individuals
and individual cohorts are large. While the impact of
demographic change on asset prices is small, the historical
volatility of asset prices and rates of return is significant. This
is unlikely to be affected by demographic change, but it means
that those relying on financial market returns for their
retirement income could be much less lucky than those who
enjoyed the high returns of the 1980s and 1990s.
Moreover, the projected increase in the number of people in
this position raises the risks of large numbers suffering the
effects of financial shocks, as well as the risks to
macroeconomic and financial stability. Recent experience
with endowment mortgages emphasises that the returns on
long-term investments can turn out to be substantially
different to expectations. In a similar way, a period of very low
rates of return on capital would leave people with much lower
pension entitlements than had been anticipated. This can
occur even when overall asset returns have been strong if
investors have poorly diversified portfolios, but the adverse
effect of it occurring for a substantial group of savers could be
severe. Such an outcome would have macroeconomic
repercussions if lower expenditure by the retired was
intensified by lower spending by those of working age who
become concerned about their own retirement income. It
would have systemic implications if lower asset returns meant
that debts could not be paid.
Given the lack of financial sophistication of many households,
there is a clear educational role for financial regulators in
informing people of the risks they face and what action they
might take.
The implications of an ageing population for the UKeconomyWorking Paper no. 159
Garry Young
311
Empirical studies of worker flows in the United States and
Europe have found that these flows are large when compared
with the change in the stocks of employment and
non-employment and have a distinct cyclical pattern. In the
United Kingdom, studies of this kind have been hampered by
limitations in the available data. In this paper we make use of
newly released longitudinal data from the Labour Force Survey
(LFS) to document the size and cyclical patterns of the gross
worker flows in the United Kingdom.
The motivation for considering gross worker flows is a simple
one: to uncover what lies behind the headline levels of—and
changes in—key statistics such as employment and
unemployment. In particular, data on gross worker flows allow
us to observe two features of these flows: their magnitude and
cyclical properties. The magnitude of worker flows may allow
us to gauge the flexibility of an economy, as the rate at which
workers flow from less efficient plants to more efficient ones
will affect how quickly an economy responds to economic
shocks. And the cyclical properties of the gross flows allow us
to uncover how labour demand is met over the business cycle.
In short, the availability of data on gross worker flows allows us
to go behind the aggregate stock data to examine the nature of
labour market dynamics.
Data on gross flows may be affected by measurement biases to
a greater extent than the levels data. In particular, sample
attrition and response error may cause errors in estimating the
flows. We test this by looking at the number of ‘inconsistent’
transitions. In the LFS, individuals in employment and
unemployment are asked not only about their current state,
but also how long they have been in that state. If the duration
contradicts the transition, then the transition is ‘inconsistent’.
We observe a significant level of inconsistent transitions, but
suspect that most of the error occurs because individuals are
unclear as to their exact duration in any state rather than
about their current state. To the extent that these transitions
are not genuine, they will lead to overestimation of the gross
flows.
Over the past five years, the stock of unemployed fell by an
average of 40,000 per quarter. Given an average stock of
1.9 million, this may seem to suggest that the market for
labour can be characterised as fairly static. Yet such a
conclusion would be wrong. We find that, over the same
period, almost three-quarters of a million people entered
unemployment in a quarter, with numbers drawn equally from
employment and inactivity. Similarly, almost one million
people start a new job each quarter after previously being
unemployed or inactive.
Theoretical models of labour market flows generate predictions
about the cyclical pattern of flows and associated hazard rates
(the chances of making a transition from a given labour market
state to another). These predictions can be tested using the
LFS longitudinal data. In particular, we examine the cyclicality
of both the gross flows and the associated hazard rates in the
United Kingdom using a variety of data and techniques. We
find that:
1. Flows from employment to unemployment are
countercyclical, as is the hazard rate. The reverse flow,
from unemployment to employment, is also
countercyclical—while its associated hazard is strongly
procyclical.
2. Flows from employment to inactivity tend to be
procyclical and there is no clear pattern to the
associated hazard rate. Flows from unemployment to
inactivity appear to be countercyclical.
3. Flows and hazards from inactivity are imprecisely
measured, and we cannot be confident of any statement
on their cyclical characteristics.
4. Flows of workers moving from one job to another,
without a recorded period of unemployment or
inactivity, are strongly procyclical.
These findings are broadly consistent with similar results for
the United States and Europe.
In addition, we are also able to measure the incidence of
job-to-job flows. Little is known about these flows in the
United Kingdom and previous research has tended to focus on
the prevalence of on-the-job search without knowing whether
that search was successful. We show that 2.9% of those in
employment change employer in an average quarter. This
represents a movement of three-quarters of a million workers.
Unsurprisingly, the probability of making such a move is much
higher for those who are engaged in on-the-job search. Such
movements tend to occur much more frequently for workers
with short tenure in their initial job. This is consistent with
findings in the literature suggesting that individuals search on
the job when they are in poor matches. As tenure lengthens
and job-specific human capital is acquired, the incentive to
move jobs falls.
On gross worker flows in the United Kingdom: evidencefrom the Labour Force SurveyWorking Paper no. 160
Brian Bell and James Smith
312
The Basel Committee is currently engaged in designing anew Accord on the capital adequacy of internationallyactive banks that will supersede the original 1988Accord. Under the old Accord internationally activebanks in G10 countries are required to hold broad andnarrow capital that is no less than 8% and 4%respectively of their risk-weighted assets. Whileexposures to banks, sovereigns and mortgage assets aretreated differently, the bulk of banks’ private sectorassets are subject to the same capital charges no matterhow risky they are. In framing the new Accord, the BaselCommittee intends to make capital charges on differentexposures much more risk sensitive.
A major question confronting regulators is how high theoverall average level of capital charges for representativebanks should be. The 8% in the original Basel Accordwas chosen on the basis that this was the minimum levelof capital observed among banks that were perceived tobe following best industry practice. When capitalcharges are being redesigned as part of the preparationof the new Accord, it is natural to consider what currentlevels of regulatory capital imply for financial stabilityand to what extent these are a binding constraint onbanks.
This paper employs a standard credit risk model toinvestigate the survival probabilities implied by thecurrent minimum level of narrow regulatory capital(subordinated debt included in the wider definitiondoes not affect survival probabilities). In particular, forcorporate loan portfolios with representative qualitydistributions, we calculate the likelihood that the banksholding the portfolio would survive over a one-yearhorizon if their capital were at the regulatory minimum.The model employed is a simplified version of the widelyused CreditMetrics approach, developed by JP Morgan.
We then compare the survival probabilities implied byminimum levels of regulatory capital with those levels of‘economic’ capital that internationally active banksactually hold. We do this in two ways. First, we againemploy CreditMetrics calculations. Second, we examinethe historical survival rates associated with the ratingsthat banks receive from the main rating agencies. Thislatter examination is complicated by the fact that, inmany cases, banks’ agency ratings are boosted by marketexpectations that the authorities will provide support ifbanks experience difficulties. To get around this
problem, we calculate, using an econometric model,what ratings banks would have if they were not expectedto be able to obtain support.
We conclude from our calculations that the one-yearsurvival probability or solvency standard implied by thecurrent Basel Accord minimum capital levels is between99.0% and 99.9% depending on the quality of thecorporate loan book used. Our investigation of thesolvency standard implicit in the ‘economic’ capitallevels that internationally active banks actually holdsuggests that it is substantially higher than 99.9%.
Having investigated the relation between the survivalrates implied by regulatory minimum capital levels andby the capital levels that banks actually hold, we ask afurther question: why do banks make such apparentlyconservative capital decisions, selecting economiccapital that significantly exceeds the regulatoryminimum? It is difficult to answer this questionconclusively but we argue that the evidence is at leastconsistent with one explanation, namely that banks areobliged to maintain higher capital levels in order toobtain access to certain wholesale markets, most notablythe swap market, participation in which is a prerequisitefor operating a modern large-scale, internationally activebank. To make the case that market discipline of thiskind is an influence on banks’ choices, we show that thevolume of banks’ swap liabilities, conditioning on banksize, is significantly correlated with the bank’s creditrating. Large international banks wanting to deal insignificant swap volumes appear to have to maintainhigh ratings.
The main implication of our analysis is that maintaining minimum regulatory capital levels in thenew Basel Accord at levels similar to those that applyunder the 1988 Accord would not act as a majorconstraint on most internationally active banks, sincethey already operate on higher ‘economic’ solvencystandards than those implicit in the Basel regulatoryminimum. While different reasons might be adduced for why banks adopt a relatively conservative approachin their capital-setting decisions, one possibility thatseems consistent with data on swap market volumes isthat the need to maintain access to certain wholesalemarkets, which is crucial to operating a large bank,necessitates a fairly stringent ‘economic’ solvencystandard.
Regulatory and ‘economic’ solvency standards forinternationally active banksWorking Paper no. 161
Patricia Jackson, William Perruadin and Victoria Saporta
313
Capacity utilisation—or time-varying factor input
utilisation—is a key component of the supply side of the
economy and is often thought to provide information
regarding the build-up of inflationary pressures.
Though difficult to measure, capacity utilisation may
also account for much of the variation in aggregate
output. So it may provide useful insights into the
characterisation of the business cycle.
This paper evaluates the implications of a general
equilibrium model of time-varying factor utilisation
under the assumption of factor hoarding. We assume
that production relies on labour and capital services.
The contribution from labour depends not only on the
number of hours people work, but also on the
productive effort they exert during those hours.
Similarly, the contribution from capital takes into
account not only the number of machines in the factory,
but also the intensity at which they operate. There are
costs associated with utilising factors intensively:
workers suffer disutility, machines wear out more quickly.
But since firms must choose in advance how much
capital stock and employment to rent, they may
under-utilise these inputs in equilibrium. Machines may
be left idle; workers may spend time sweeping the
factory floor. We find that firms initially respond to
unanticipated shocks by altering factor utilisation rates.
In subsequent periods, firms adjust their physical stock
of capital and employment. As a result, utilisation rates
are a leading indicator of firms’ hiring of both capital
and labour.
We then use the model to derive estimates of capital
utilisation and labour effort for the United Kingdom. By
explicitly accounting for variations in factor utilisation,
these help to estimate total factor productivity (TFP)—
that portion of output growth not due to growth in
capital or labour more accurately.
Our estimate of capital utilisation for the United
Kingdom matches survey-based measures of capacity
utilisation quite closely, supporting the view that these
measures accurately reflect the degree to which firms are
utilising their existing capital stock. All measures
indicate that capital utilisation rose during the 1990s—
though it has recently fallen back somewhat—reflecting
a declining capital-to-output ratio over the period. The
predicted positive and leading relationship between
capital utilisation and investment in the model in turn
indicates the potential usefulness of surveys for
forecasting investment.
Movements in total hours worked drive our estimate of
labour effort. Given the costs to adjusting employment,
this is quite intuitive. When a boom is in its initial
stages, firms demand an increase in effort in order to
generate labour services. Only after a time can firms
satisfy their demand for increased labour services by
increasing total hours worked, with effort slowly
returning to normal levels. Our estimated series for
labour effort shows a decline after the mid-1990s. This
decline is a reflection of the sharp increase in total
hours worked over that period. Contrary to theoretical
predictions, however, our effort series is only weakly
correlated with both a manufacturing-based measure of
labour effort and average hours worked.
Our estimate of TFP is found to be less cyclical than the
traditional measure, the Solow residual. Nevertheless, a
weighted average of capital utilisation and labour
effort—which we call aggregate factor utilisation—is
not closely related to the Solow residual. This suggests
that measures that conflate both capacity utilisation and
temporary fluctuations in TFP (as the Solow residual
does) may be misleading indicators of excess demand
pressure.
Rather, our measure of aggregate factor utilisation is
more correlated with detrended labour productivity. In
some ways this is not surprising: if capital and labour are
slow to adjust, then much of the variation in factor
inputs—and hence output—over the business cycle
must come from utilisation and effort. This supports the
view that labour hoarding is responsible for much of the
cyclicality in measured labour productivity. In fact,
labour productivity, when calculated as output per unit
of effective labour input, is much less cyclical than a
simple measure of output per hour.
Factor utilisation and productivity estimates for the United KingdomWorking Paper no. 162
Jens Larsen, Katharine Neiss and Fergal Shortall
314
Traditionally, productivity at the aggregate level has
been measured using GDP, ie the measure of output is
gross of depreciation. But suppose that the composition
of the capital stock is shifting towards assets with
shorter lives, so that the average depreciation rate is
rising. This suggests that some part of what GDP
measures as an increase in output may be illusory: some
of the extra output is needed just to maintain the capital
stock at its existing level. This question is given a
sharper focus by the experience of the United States in
the 1990s, where the growth rates of labour productivity
and total factor productivity (TFP) rose, while investment
shifted towards short-lived ICT assets. This raises the
possibility that the US productivity improvement might
be just a statistical illusion.
This paper has a theoretical and an empirical part. In
the theoretical part, I compare measures of productivity
with measures of welfare. I conclude that while GDP is
satisfactory as a measure of output, it is outclassed as a
measure of welfare by what I call Weitzman’s NDP
(WNDP). This is nominal net domestic product
(consumption plus net investment) deflated by the price
index for consumption. I extend the theory behind
WNDP to the case where TFP growth can vary across
sectors. This is the empirically relevant case for
analysing recent US experience.
The aggregate TFP growth rate is the rate at which the
GDP frontier is shifting out over time. This can be
decomposed into a weighted average of the TFP growth
rates in the various industries. Analogously, we can
define the rate at which the WNDP frontier is shifting
out over time. I call this the growth rate of total factor
welfare (TFW). Like aggregate TFP growth, TFW growth
can also be decomposed into a weighted average of TFP
growth rates in the various industries, but the weights
are not the same as for the GDP frontier. Hence the
growth of welfare over time can be analysed using the
same tools as have been developed for the analysis of the
growth of output.
In the empirical part of the paper, I apply some of these
ideas to the experience of the United States in the
1990s. In principle, one might expect WNDP to have
grown more slowly than GDP over this period, for several
reasons. First, the weight on consumption is higher in
WNDP (or in NDP) than in GDP and consumption has
been growing more slowly than investment. Second, the
relative price of investment goods has been falling and
this reduces WNDP growth. Third, one might have
expected depreciation to have risen as a proportion of
GDP, thus raising the share of consumption in WNDP
still further.
In practice, WNDP has grown a bit more slowly than
GDP. But the gap between the two growth rates was
actually somewhat larger in the period 1973–90 than it
was post 1990. And the acceleration of WNDP post
1995 was equal to that of GDP. The explanation is
twofold. The ratio of depreciation to GDP has in fact
been stable, despite the growing importance of
short-lived assets. And net investment has grown more
rapidly than gross investment. The growth rates of TFP
and of TFW in the US non-farm business sector are also
compared and found to be similar in the 1990s.
Moreover, they display an almost identical increase after
1995.
GDP is a measure of output, not of welfare. So even if
GDP had grown significantly faster than WNDP, this
would not by itself suggest measurement error. In fact,
the two have grown at similar rates in the 1990s and
accelerated by the same amount. So it seems that, in
practice, GDP has provided as reliable a measure of the
improvement in US living standards over this period as
WNDP, even though WNDP is conceptually superior as a
welfare measure.
Productivity versus welfare: or, GDP versus Weitzman’sNDPWorking Paper no. 163
Nicholas Oulton
315
This paper re-examines inflation dynamics in the United
Kingdom. Our main motivation is the recent low
inflation, low unemployment era in the United States,
the United Kingdom and the euro area. This has led to
overpredictions of inflation using standard
specifications of traditional Phillips curves. This has
been a major motivation for a ‘New Phillips Curve’
approach, which has had success for the United States
and the euro area. The reason the United Kingdom is
an interesting case to study is that it is a far more open
economy than the United States or the euro area.
In this paper we analyse whether the openness can
explain the overprediction problem in traditional
Phillips curve estimates and whether it affects the
performance of ‘New Phillips Curve’ estimates. The
paper is divided into two parts. First, we document the
overprediction problem for the United Kingdom and try
to solve it in a traditional Phillips curve framework. We
introduce external shocks from two sources: terms of
trade shocks and domestically generated inflation (DGI).
We find that external shocks do not fully solve the
overprediction problem within this framework. We
further argue that there is a more general
misspecification problem with traditional Phillips curve
estimates, due to the presence of regime changes and
structural change in the UK economy.
Second, we look at ‘New Phillips Curve’ estimates. They
do not perform particularly well: real marginal cost is
not significant in our baseline specification. Further
investigation suggests the relationship between marginal
cost and inflation broke down around the mid-1980s.
When we use a labour-share measure adjusted for the
public sector, real marginal cost becomes significant, but
the goodness of fit of the model—based on fundamental
inflation—is still very poor.
Next, we extend our ‘New Phillips Curve’ model to allow
for open-economy influences. In particular, we take into
account imported intermediate goods. When we allow
for imported intermediate goods the relationship
between inflation and marginal cost improves
significantly. Fundamental inflation performs better
than previously, but still has a tendency to underpredict
and then overpredict inflation: something also present
in the traditional Phillips curve estimates.
Finally, we decompose the open-economy measure of
marginal cost to learn more about its driving forces. We
find that a wage mark-up component is important and
highly countercyclical. We also find that relative price
movements, of taxes relative to overall prices and of
imported intermediate goods relative to wages, have
been a negative influence on marginal costs over the
1990s. Understanding likely future developments in
these relative prices could contribute to the assessment
of prospects for marginal costs and the pressures on
inflation.
Time-varying desired mark-ups may, in part, explain
why the open-economy New Phillips Curve still
underpredicts and then overpredicts inflation. In the
models considered in this paper, the desired mark-up is
assumed to be constant. This is important to the extent
that the desired mark-up varies cyclically and can be
influenced by external factors. For example, recently
there has been much speculation that the high level of
sterling has forced manufacturers to cut their margins
on exported goods. This is equivalent to a fall in the
desired mark-up and will have a negative impact on
inflation in the GDP deflator. This idea fits well in a
customer market model. In a customer market model,
firms are assumed to be monopolistically competitive,
and set their own mark-up, taking the mark-up of other
firms as given. However, there is a dynamic element to
the firm’s problem in that higher relative prices reduce
market share. In addition, some consumers are assumed
to pay a cost when switching from one firm to another.
This kind of model provides a justification for firms to
allow the desired mark-up to vary, in the short term, in
order to stop the long-term loss to profitability of losing
customers. It may also be a key factor that exporters
take into account, by allowing margins to vary in
reaction to changes in exchange rates, rather than the
foreign price of the exported good. Recent high levels of
sterling may have reduced the desired mark-up and thus
potentially explain the overprediction of actual inflation
by fundamental inflation. We plan to look at this in
greater detail in future work.
Understanding UK inflation: the role of opennessWorking Paper no. 164
Ravi Balakrishnan and J David López-Salido
316
Evidence from around the world suggests that the
majority of central banks take monetary policy decisions
by committee rather than through a single individual.
Despite this observation, there is little direct empirical
or theoretical evidence on the relative merits of
monetary policy decision-making by committees versus
individuals. Recent work by Blinder and Morgan has
sought to shed light on this question by taking an
experimental approach, the main result being that the
decisions of committees were superior to those of
individuals. Although the results of our paper support
this conclusion, we attempt to extend their work by
examining and testing several hypotheses as to why this
improvement might come about.
To this end, we asked a large sample of economically
literate undergraduate and postgraduate students from
the London School of Economics to play a simple
monetary policy game. Participants acted as monetary
policy makers, setting interest rates to ‘control’ a simple
macroeconomic model calibrated to match UK data and
subject to an unknown combination of shocks. Each
participant acted as both individual decision-maker and
as part of a committee of five players. All players faced
an identical incentive structure: performance was
judged according to a score function that penalises
deviations of output and inflation from their target
values; and they were paid according to their
performance.
Just like actual policy-makers, participants in our
experiment were forced to make decisions in an
uncertain world, while observing only the evolution of
the endogenous variables over time. As in real life, these
monetary policy makers did not know with certainty the
exact structure of the economy they were attempting to
analyse. To the extent that players came to the
experiment with different prior beliefs about the
structure of the model, they may have responded
differently to the same set of shocks. So we modelled
these differences of opinion by asking participants to fill
in a questionnaire that attempted to reveal these prior
beliefs. By asking players to fill in the same
questionnaire at the end of the game we were able to
discern some evidence of players learning about the
underlying model of the economy over time. And for the
‘worst’ players, their improvement in scores over time
was positively and significantly related to the extent of
their learning about the underlying model.
Like Blinder and Morgan, we found that committees
performed significantly better than the individuals who
composed them. There are several competing
hypotheses as to why. Our results suggest two reasons
why committees make better decisions. First, collective
decision-making appears to give more weight to the
better and less weight to the worse committee
members—as judged by their scores when playing the
game as individuals—than would be implied by taking
the mean of their individual performance. But we find
evidence that committees do more than this, enabling
all members to improve their performance by sharing
information and learning from each other. For example,
the performance of the committee was on average
better than that of its ‘best’ policy-maker when playing
alone.
In our experiment, we also explicitly tested whether the
ability to discuss a decision drives the observed
improvement in performance. In practice, this did not
appear to be the case: in our simple monetary policy
game, participants were able to share enough
information by simply observing each other’s behaviour.
But we were able to illustrate how the relative
importance of different types of communication
depends upon the nature of the decision problem in a
variant of the game in which we slightly altered the
structure so as to raise the relative importance of
discussion. When we did so, committees that discuss
performed better.
Committees versus individuals: an experimental analysisof monetary policy decision-makingWorking Paper no. 165
Clare Lombardelli, James Proudman and James Talbot
317
Recent financial crises have illustrated that the financial
positions of borrowers and lenders—financial stability
considerations—can influence the way in which official
interest rate changes affect spending and inflation—
monetary stability considerations. A substantial
academic literature has developed considering potential
macroeconomic impacts of financing decisions by
borrowers and lenders. Among these so-called ‘credit
channel’ models, the recent financial accelerator
approach of Bernanke, Gertler and Gilchrist seems
particularly suited to an analysis of how corporate sector
balance sheets and the behaviour of banks can affect
the monetary transmission mechanism.
In credit channel models, firms often find it more costly
to finance investment projects with external funds rather
than with internally generated resources. This ‘external
finance premium’ may arise because lenders face costs
from observing and/or controlling the risks involved in
supplying funds to borrowers. These agency costs, and
the external finance premium, may vary with borrowers’
financial health. For example, the stake of a borrower in
an investment project (measured by the degree to which
it is able to finance a project using internal funds) may
provide a signal of the unobserved risk of lending. It
may also affect the borrower’s incentive to act diligently
and to report project outcomes truthfully.
The financial accelerator model used in this paper
embeds a similar imperfect information problem in the
supply of external finance in a standard macroeconomic
framework. Our paper examines how a range of
interesting financial scenarios can arise out of this
model and in turn, how these scenarios affect the
dynamic response of the model economy to alternative
shocks (for example monetary news or productivity
shocks). These scenarios are defined in terms of
steady-state credit spreads, bank lending policies and
corporate financial health. The main objective is to
examine how the strength of the monetary transmission
mechanism might vary across such scenarios.
Our simulations of the model show how balance sheet
positions of the financial and non-financial corporate
sectors can affect the monetary transmission
mechanism. We show that in certain financial scenarios
the financial accelerator mechanism is very potent,
whereas in others it has little incremental impact. This
implies that, for a given shock in the model economy,
monetary policy can be less or more proactive,
respectively.
In addition, the model simulation results suggest that
certain parameters may merit particular attention. For
example, the sensitivity of bank lending to news about
corporate financial health has an especially marked
impact on the model’s dynamics. And as illustrated in
previous work, leverage also plays an important role in
amplifying and propagating shocks. But we also show
that the strength of the financial accelerator cannot be
attributed to a single variable. For example, we observe
that the financial accelerator can be weak, both when
leverage is low and banks are relatively restrictive in
their lending, and when leverage is high and banks are
very accommodative.
These theoretical results are consistent with real-world
experience that bank and non-financial corporate
balance sheets can, at times, have a marked impact on
the effectiveness of monetary policy. But while the
specific model used in this paper provides an attractive
analytical framework for thinking about potential
qualitative effects of changes in financial conditions on
real variables, we think its quantitative results need to be
interpreted with caution, as in all calibrated simulated
models. Moreover, although the model can be used to
analyse certain important interactions between financial
imperfections and the monetary transmission
mechanism, it leaves out several features that one might
want to incorporate in a more general model of financial
stability. For example, the model has a relatively
restricted financial structure with a focus on debt
finance. Financial institutions are sparsely modelled,
with limited potential for effects from the bank lending
channel. That suggests that further work to develop
quantitative models that incorporate these features may
provide further insights into interactions between
monetary and financial stability.
The role of corporate balance sheets and bank lendingpolicies in a financial accelerator frameworkWorking Paper no. 166
Simon Hall and Anne Vila Wetherilt
318
The Symposium
The Central Bank Governors’ Symposium 2002 took
place at the Bank of England on Friday 5 July. The
Governor opened the symposium. There were three
main speakers: Dr Horst Koehler, the Managing Director
of the International Monetary Fund, and the Bank of
England’s two Deputy Governors at the time,
David Clementi and Mervyn King. The subject this year
was ‘International Financial Architecture’. Their
speeches, and the set of background papers assembled
for the event, explored various aspects of this topic. This
summary has grouped that material into three sets. The
first set includes overviews from the IMF and the BIS, a
salutary lesson from history, and a sketch of some
germane international economic issues. The second set
concentrates on how financial crises may best be
contained and prevented. And the third looks at how
they should be resolved. Horst Koehler’s paper led the
first set, David Clementi’s the second and Mervyn King’s
the third.
The Governors of the Reserve Banks of India and South
Africa, and the Prezes of the National Bank of Poland,
acted as discussants. They provided comments on the
three speeches and the background papers. A lively
debate ensued. Dr Koehler gave robust, frank and
detailed answers to numerous questions from the
assembled Governors and heads of delegation. Fifty
central banks were represented at the symposium, which
had been preceded, as usual, by a high-level workshop
organised by the Bank of England’s Centre for Central
Banking Studies. This year the workshop was devoted to
the challenges faced by central banks, and the role
played by the Centre’s training and collaborative
research in helping to meet them.
Overview contributions
Horst Koehler began his address by outlining how the
IMF had amended the architecture of the world’s
financial system in the five years since the onset of the
Asian crisis. The IMF was promoting transparency in
member countries, and was now publishing most
country documents. Surveillance and assistance for
preventing crises, and loan facilities for resolving them,
were being sharpened. Conditionality was being
streamlined, and members urged to assume fuller
responsibility for and ownership of programmes of
reform. With other bodies, the IMF had established new
codes and standards and Financial Sector Assessment
Programmes (FSAPs). And measures against money
laundering and the finance of terrorism were being
strengthened.
The future stability and growth of the world economy,
Dr Koehler argued, depended above all on four factors.
These were medium-term budgetary balance in the
United States; much needed structural reform to
support faster expansion in Europe; deregulation,
restructuring and swifter disposal of non-performing
loans in Japan; and stronger institutions and greater
income equality in Latin America. Fighting world
poverty underscored the need for more and freer trade,
not less. Safeguarding the world’s financial system
International Financial Architecture: the Central BankGovernors’ Symposium 2002
The Central Bank Governors' Symposium 2002 examined the architecture of the world's financialsystem. Horst Koehler, Managing Director of the IMF, and the Bank of England's two Deputy Governorsat the time, David Clementi and Mervyn King, gave the main addresses. This article summarises whatthey said. It also gives a precis of eight background papers provided for the occasion. Taken together,these eleven contributions explore general aspects of the international financial architecture, as well asdiscussing how financial crises may be contained or prevented, and best resolved when they do occur.
(1) I should like to thank Sir Edward George, Bill Allen, Charlie Bean, Alastair Clark, David Clementi, Andrew Crockett,Andy Haldane, Andrew Hauser, Simon Hayes, Mervyn King, Mark Kruger, and Hyun Shin very warmly for illuminatingdiscussions and helpful comments, but to exculpate them from any errors.
By Peter Sinclair,(1) Director, Centre for Central Banking Studies.
International Financial Architecture
319
required stronger accounting and disclosure procedures
in the United States and elsewhere. Successful crisis
resolution needed debtors and private creditors to
assume full responsibility for risks, as well as improved
techniques for judging debt sustainability, clearer limits
to Fund lending, and new devices to facilitate an orderly
restructuring of debt when crises hit. Horst Koehler
ended by thanking central banks for participating in
FSAPs, adopting codes and standards, and maintaining
the prudent policies on which the success of the IMF’s
work so much depended.
For Andrew Crockett (General Manager of the Bank for
International Settlements), ‘international financial
architecture’ had two aspects: rules governing all
economic and financial contacts between countries, and
the institutions where those rules were set, monitored
and applied. Noted achievements since 1945 had been
the multilateral approach, and the great liberalisation of
trade. Changes included the spread of floating exchange
rates, the replacement of administrative controls by
market-driven processes for balance of payments
adjustment and liquidity, and the growing involvement of
more national authorities, and the private sector, in the
work of the IMF and the World Bank.
The division of tasks between national and international
bodies needed some rethinking, Andrew Crockett
advised. Financial supervision was a national
responsibility and should be recognised as such.
Policy-makers needed to look carefully at the degree to
which private markets acted efficiently and without
undue turbulence. Sound public finances and
transparency were critical for preventing and containing
financial crises, which, in 18 recent cases, cost victim
countries an average of 22% of a year’s GDP. To stop
repetition, financial markets needed to become deeper,
broader and more resilient; accounting standards less
diverse; corporate governance and insolvency
procedures improved; law enforcement fairer and more
coherent; and payment systems strengthened.
Resolving crises called, in his view, for creditors’
governments not to finance private creditors’ withdrawal
from troubled debtors. Private creditors must be fully
involved in crisis resolution, and discussing the form this
should take was a priority. The Financial Stability Forum
had a key coordination role for national regulators, but
should not replace them. Another priority was the
creation of an institutional mechanism for involving the
private sector in the work of the IMF.
What lessons did history provide about financial crises?
The usual view was that crises often began with a weak
bank, and the panic that triggered a run on it and
maybe other institutions. But crises could develop in
other ways, and in circumstances where all parties
regarded their positions as absolutely safe. Isabel
Schnabel (University of Mannheim) and Hyun Song Shin
(London School of Economics) explored the details of
the North European financial crisis of 1763, centred in
Amsterdam, Hamburg and Berlin, and the disturbing
resemblances it bore to the collapse of Long-Term
Capital Management 235 years later.
Events in 1763 provided a salutary reminder, they
argued, of some valuable principles not well or widely
understood. One was that a connected lending chain
could not eliminate all risks, even if every participant in
it deemed herself to be fully hedged. There were
unsuspected knock-on effects from the failure of one
party to meet obligations; credit and counterparty risks
were correlated, and increasingly so in the event of
trouble. Too much pressure on one link could set off
balance sheet contagion, and cause all the other links to
snap too. Today’s parallel upward trends in many agents’
gross assets and liabilities, coupled with direct and
indirect counterparty risks, off balance sheet items and
the proliferation of potentially perilous derivatives, made
the task of quantifying true exposure an extremely
challenging one.
The problem in 1763 had been compounded by a second
phenomenon, clearly as visible then as it would be today.
This was liquidity risk. If one market participant tried to
meet obligations by selling a seemingly liquid asset they
all held—in 1763 this was grain—its price could tumble,
suddenly and unexpectedly lowering the net worth of
others in the chain. Dangerous and pervasive as the
events of 1763 had been, they did not constitute a
Pareto deterioration; Amsterdam’s woes were
instrumental in securing the growth of London as the
world’s leading financial centre.
The dynamics of inflation and debt in an open economy
were the first topic explored by Peter Sinclair (Director
of the Centre for Central Banking Studies at the Bank of
England). Under foresight, and if undisturbed, these two
variables should drift in the same direction towards
long-run values pinned down by fiscal policy parameters,
growth, real interest rates, and deficit-financing patterns.
He adapted a paper of Fry and Sinclair (2002) on this to
allow for an open economy with a freely floating
320
BBaannkk ooff EEnnggllaanndd QQuuaarrtteerrllyy BBuulllleettiinn:: Autumn 2002
exchange rate.(1) Revised expectations about
fundamentals caused the price level and the exchange
rate to jump up or down. The time profiles of inflation
and the exchange rate depended on bond maturity,
currency denomination and indexing (if any). Ambiguity
about future fiscal policy was inflationary in its own
right. Changes in expectations surrounding this,
coupled with any official resistance to allowing exchange
rates to respond to them, could be the surest recipe for a
financial crisis later on. Argentina’s recent crisis,
described most recently by Powell (2002), was but the
latest example of this.
Sinclair went on to argue that there were valuable gains
from opening up capital trade across national borders,
gains similar in character to those from constructing
domestic financial markets. He concluded with a
suggestion for modifying models of consumption and
capital accumulation that enabled one to pin down
plausible trajectories for countries’ net claims on each
other, avoiding the unpalatable conclusions of some
conventional models.
Contributions on crises and how to preventand contain them
David Clementi began by emphasising how costly
financial crises were, in terms of lost output, added
government debt and maybe subsequent monetary
disorder. His main focus was on crisis-prevention
measures, and recent progress made.
He advanced three propositions. The first was that
recent crises underscored the vital need for sound
economic and financial management, and appropriate
policy design. Proposition two was a plea for following
up vulnerability assessments of countries’ finances with a
continuing process of monitoring and managing macro
financial stability threats. The third proposition centred
on the crucial need for the successful engagement of
private sector investors with crisis prevention initiatives.
Central to proposition one, David Clementi argued, were
the twelve standards that the Financial Stability Forum
had identified as paramount, including, in particular,
those relating to transparency in both policy and data.
More transparent credible policy implied lower risk
premia in the markets. Accountability made for better
policy. And good and timely data provision reinforced
both. Constructing and publishing more comprehensive
balance sheets for nations and for the main sectors
within them—households, non-financial companies,
financial institutions and government—was an
important priority.
For David Clementi, one lesson of Enron’s collapse
was that even the most advanced countries must look
closely for possible holes in their defences against
financial stability threats. The type of control system
appropriate for some countries may not always be best
for others.
Turning to his second proposition, David Clementi
highlighted the benefits of FSAPs and Reports on the
Observance of Standards and Codes (ROSCs). The
United Kingdom, he noted, was currently undergoing an
FSAP of its own. But there was no unique blueprint.
Financial authorities needed to monitor financial
stability proactively, he argued, and to assign a high
priority to improving their mechanisms for safeguarding
financial stability in the face of financial innovation and
structural change.
Third, he argued that the full benefits of
crisis-prevention policy could be obtained only if
best-practice economic management by policy-makers
were met by best-practice risk assessment by the private
sector. Policy-makers could help through appropriate
regulatory design. But the private sector needed to
make better use of information on macro financial
positions if a strong link between economic and
financial management and the cost of capital were to be
ensured.
Peter Sinclair, in his summary background paper,
maintained that abolishing financial crises was like
abolishing sin: a worthy principle, perhaps, but also a
utopian dream. Much like crime, pollution, and
unemployment, financial crises were to some degree
inevitable. And like them, good policy should seek to
limit their gravity and frequency. Economists could even
think of an optimal incidence of financial crises: costly
as they were, the expected marginal cost of truly
eliminating them, even if it were feasible, could be
enormous, far above the marginal benefit. International
financial crises could probably be stopped, for example,
by a simple expedient. Suspending currency
convertibility and banning all international capital
movements, on pain of death for convicted offenders,
would surely achieve this. But the costs in terms of
(1) The official IMF classification of countries according to exchange rate system has recently been questioned by Reinhartand Rogoff (2002).
International Financial Architecture
321
misallocated world capital, and cross-country
risk-trading forgone, stressed for example by Sinclair and
Shu (2001), would be simply prohibitive.
A debtor typically had not just one creditor, but several.
This was true of individual borrowers, and truer still of
sovereign states. When trouble arose, creditors needed
to be induced to act in concert. This was no easy task.
Delay and discord made speedy renegotiation
impossible. Workouts became disorderly. Financial
assets were liquidated too early. The costs of this were
deadweight. Prasanna Gai’s (Bank of England and
Australian National University) paper explored these
ideas.(1)
Creditors could be brought to coordinate in several ways.
Prasanna Gai mentioned country clubs, swap
arrangements, liquidity management and payments
standstills as four practical devices to achieve this.
What role did official intervention play? For him, the
IMF had two functions: whistleblower, and fireman. It
could stop play if it thought the borrower was cheating,
falsely claiming insolvency. That was strategic default.
As fireman, the IMF could rescue the project from some
of the damage capital flight brought through official
finance.
The net benefit of having the IMF play these roles
turned on the accuracy with which it could perceive the
borrower’s true financial position, the cost of creditor
non-coordination (the damage from premature flight),
and the IMF’s ability to limit that damage. Often net
benefits were positive. But they need not be,
particularly if its monitoring ability was low when its
efficacy in limiting the cost of creditor flight was high,
triggering moral hazard problems. So high-quality
official monitoring was crucial. That called for timely,
accurate and comprehensive data about the borrower’s
position. Gai concluded with some interesting
observations about the role of private sector finance,
and what the paper might mean for East Asia.
Liz Dixon, Andy Haldane and Simon Hayes (Bank of
England) stressed the fact that the nature of financial
crises had changed recently. In the 1970s and 1980s, it
had been current account balance of payments deficits
that had often set them off. By contrast, 1990s’ crises
had begun on the capital account, often deepening as a
result of maturity and currency mismatches in the victim
countries’ national balance sheets.
The main lesson to be drawn from such crises was the
need for balance sheet monitoring and management
before they broke out. More and better data, enhanced
surveillance and laying down guidelines for managing
balance sheets for the authorities and the banks were all
needed, and important steps had been taken, through
FSAPs and ROSCs for example, to provide them. The
next stage, the authors argued, was to progress from
macro prudential analysis to comprehensive
macroeconomic analysis. Looking at national balance
sheets was valuable, but identifying the balance sheets of
component aggregate units, such as the private
non-bank sector, was an urgent need. So too was
scrutiny of off balance sheet transactions. And were
more and better data enough, they asked? Greater
transparency may alter behaviour, and the genesis of
crises. But lessons based on behaviour under one
regime may tell us little about what happens when that
regime changed.
All were agreed that macro prudential surveillance was
crucial to crisis prevention. But which data did this
require? What analytical methods were needed to carry
out this work effectively? These key questions were
addressed by Philip Davis (Brunel University).
Davis pleaded for a selective synthesis of the theories of
financial instability. Experience taught us, he
maintained, that ingredients in this synthesis should
include information asymmetries, disaster myopia on the
part of banks and regulators, recognising the differences
between risk and uncertainty, asset bubbles and risk
mispricing, together with herding, bank runs, currency
mismatch and industrial organisation issues.
For Philip Davis, key data were for flow of funds;
financial prices and spreads; monetary, banking,
external financing and macroeconomic statistics; and
(1) The paper starts with a summary of a recent two-period model due to Chui, Gai and Haldane (2002), whichencompasses two kinds of financial crisis. One arises when a borrower’s (random) supply capacity—productivity—turns out so low that he is insolvent. The second occurs above this level, where it is low enough for unco-ordinatedcreditors to withdraw funds (the loan contract permits this) if enough of them are too fearful. Capital flight damagesthe investment project that the loans finance. In the second case, any crisis is based on beliefs. One key feature inthat case is that a creditor who quits makes it more likely that others would want to pull out too. And unless the modelis tight, anything can happen in this region. Morris and Shin (2000), in a similar context, paint each creditor ashaving a different but imperfect signal in period one about what period two supply will be—and this can remove themultiplicity of solutions down to one. Chui et al (2002) show that the Morris-Shin unique solution applies in theirmodel too.
322
BBaannkk ooff EEnnggllaanndd QQuuaarrtteerrllyy BBuulllleettiinn:: Autumn 2002
qualitative information on such phenomena as
regulation, innovation and risk correlations.
Contributions on how to resolve crises
Resolving and preventing financial crises were
inseparable issues, Mervyn King argued. The right
approach to crisis resolution had to balance two
objectives: not just minimising the costs of a crisis
when it happened, but also minimising the likelihood of
future crises later on. Proposals about crisis resolution
needed to be assessed from both standpoints. The
second objective called for a proper alignment of
incentives for all parties—borrowers, creditors and the
official sector.
Anne Krueger’s call for a Sovereign Debt Restructuring
Mechanism (SDRM), and recent proposals for expanding
the use of collective action clauses in international bond
contracts were valuable, Mervyn King maintained. But
they did little to meet the second aim of keeping future
crises to a minimum.
Prospects of official finance could deter borrowers
sensing trouble from starting to restructure debt.
Instead, it may tempt them, and their creditors, to
temporise and even gamble by augmenting it.
Exceptional access to large official loans had become
more common, and unpredictable in scale.
So presumptive (though not rigid) limits on official
financing were desirable, and so too was the judicious
use of payments suspensions and standstills when crises
struck. Borrowers in difficulty should face a clear set of
options, and encouragement to seek early, market-based
solutions to payments problems. Official finance should
be a backstop, not a first resort: official lending into
arrears could depend on an orderly renegotiation of
debt, with payments temporarily suspended. These
principles needed to become operational, and without
delay.
More clarity about IMF financing decisions and
rescheduling procedures should stabilise capital flows,
trim risk premia and thus serve to lower, not raise, the
cost of capital to emerging countries.
To Mervyn King, the private sector’s initial responses to
these ideas had been encouraging. They represented in
his view a healthy evolution of the world’s financial
system from which all players could benefit.
Andy Haldane (Bank of England) and Mark Kruger’s
(Bank of Canada) paper was also devoted to these issues.
Their argument was that the IMF’s response to most past
crises—bridging finance, often heavily conditional on
reform, and in the hope of generating private sector
capital inflows—was not immune to criticism. Creditors’
and debtors’ incentives might have been affected
perversely. There were costly uncertainties about the
scale of help; and future crises could become likelier
when everyone understood that such help may be very
generous after the event.
The alternative framework they proposed aimed to align
incentives for all parties affected by an international
financial crisis, and to stipulate a clear sequence of
actions for the players. Limited official finance for a
crisis victim country should come first, they argued:
enough to avoid excessively rapid and painful
adjustments, but limited to circumvent the moral hazard
difficulties that arose with any form of insurance that
paid the insured to act differently.
The IMF’s normal official lending limits should apply in
crises, Haldane and Kruger advised. Clarity here should
help by reducing risk premia for emerging countries; by
dissuading private creditors from playing a waiting game
at times of strain and gambling on a breach of those
limits; and by deterring them to some degree from
excessive lending in the first place.
The second element in the Haldane-Kruger picture
concerned the parties themselves. The debtor should be
free to choose (after consulting creditors) from a menu
of possibilities. These included bond exchange, debt
rollover, and temporary standstills. Standstills on
overseas debt service and repayment could help to
coordinate creditors, align the parties’ incentives, and
buy time and order for restructuring and reform. The
guidelines the authors suggested for standstills included
transparency, equal creditor treatment, new lending
seniority, a time limit, and rules about debt reductions
in the event of illiquidity (rephasing debt at unchanged
present value) and insolvency (cutting back to a
sustainable level).
While urging respect for normal limits on official
financing, they recognised the case for flexibility in
extremis. But write-downs for insolvent defaulters, and
general expectations of official lending ceilings, should
in the long run curb future overlending, so often a key
ingredient in the crisis.
International Financial Architecture
323
In a related paper, Andy Haldane explored and compared
two proposals for reinforcing the international financial
architecture. The first was Krueger’s (2002) call for a
Sovereign Debt Restructuring Mechanism (SDRM), or
international bankruptcy court embracing some of the
features of Chapter 11 in the US system. The second
concerned the suggestions advanced in the
Haldane-Kruger (HK) paper summarised above. What
the two share was their emphasis on standstills. But
while HK recommended a non-statutory solution device
for both illiquidity and insolvency crises, the SDRM was
statutory and confined to insolvency.
Pronouncing insolvency involved, inter alia, guesses
about the future course of the debtor country’s GDP and
interest rates. Andy Haldane argued that there were
bound to be more countries in a ‘grey zone’ between
clear illiquidity and clear insolvency, than there are
cases of the last kind. As SDRM gave interim official
finance, the borrower’s returns had a floor, so that
his losses were limited in bad outcomes, thus tempting
him to gamble. The official community, whatever it
might say and want ex ante, tended to forbear, and could
end up inducing and indeed financing private sector
outflows.
These drawbacks Haldane saw in the SDRM were
compensated by four strengths. Two—protecting
creditors’ interests and granting fresh loans seniority—
were common to the HK proposal. Two were not. These
were the risks of creditors suing or disrupting debt
restructuring agreements. He argued that law suits,
while not uncommon (over a quarter of sovereign
defaults since 1975 had provoked them), succeeded only
rarely. And hold-ups had been attempted by only tiny
minorities of creditors. The SDRM was a long-term
proposal that would require lengthy and widespread
legislation, but could be seen as complementary to the
simpler recommendations of HK, which could be
implemented sooner, and would also embrace liquidity
or grey-zone crises.
Some concluding remarks(1)
In his background summary paper, Peter Sinclair’s
discussion of the international financial architecture
explored the analogies of buildings and plumbing. He
posed a number of questions, and then distilled answers
from the papers presented. If it were a building, he
asked how deep and strong should its foundations be?
How secure against nature, and man’s misbehaviour?
Should it be a fortress, to control ingress and maintain
privacy, or a translucent glasshouse, its inner workings
plain to any outsider? Should its flooring be
partitioned? Would watchtowers be needed? What
signs of attack should the sentries look out for?
How free should financial actors be to move around
their own national stage, or those of other countries?
Should we be conducting our affairs on one global stage
in the long run? Should the building be wired to record
all participants’ transactions? Should it be a
panopticon, with camera monitors in every nook? How
far could we trust market systems to translate
individuals’ actions taken in informed self-interest into
efficient aggregate outcomes? Should the private sector
build part of the edifice, and how could we best involve
it in its operation? Or should this be left to national or
supranational authorities? And if a global architect
offered plans, exactly who would debate, amend and
ratify them, and where, when and how? If the
supranational authority were to be the partner of
national authorities, what precise form should that
partnership take?
Controls, rules, oversight and risk-proofing were all
costly, directly and indirectly. Their benefits, however
large, were presumably subject to diminishing returns at
the margin. So where would the ideal balance be
struck? What would be the lowest-cost method of
constructing the building, for any given size and level of
complexity?
The building was already there. Would it have to be
rebuilt? Should its architecture be uniform? Or would
pluralism be permissible or apt? Would the same
standards, rules or ratio formulae for all financial
institutions be right for all countries? Would it just be
banks we had to encompass? Should there be a
state-of-the-art building for the richest countries’
financial systems, with more latitude and simplicity for
others? And on what time scale?
Might a rigid structure for international finance impede
innovation and growth? Could it accentuate cyclical
instabilities? Would it just be shelter against the worst
shocks that nature could throw that constituted the core
(1) The author wishes to emphasise that this is his personal selection of conclusions to emerge from the backgroundpapers prepared for the Symposium, and the three main addresses given there. They should not be seen asconclusions agreed by participants at the Symposium, nor as conclusions the Bank of England drew from it.
324
BBaannkk ooff EEnnggllaanndd QQuuaarrtteerrllyy BBuulllleettiinn:: Autumn 2002
function of the building? Would reducing the frequency
and gravity of shocks be the right objective? Would it
be wiser to accept that some shocks were unavoidable
(or best not resisted), and aim, instead, to allot
irreducible risks to those best placed by circumstance or
temperament to bear them—and resolve the effects of
bad shocks to achieve an ideal trade-off between fairness
and proper incentives for future actions?
Maybe it was humbler features of the financial edifice
with which central banks were most concerned.
Plumbing could be a better analogy. Would it be the
task of central financiers to provide safe conduits for
liquidity, without which trading and production could
dry up? Would international financial architecture
embrace irrigating and protecting trades (in goods, risks
and assets) within national borders, and not just across
them? And should good plumbing include a proper
mechanism for the swift write-off and disposal of
defunct assets in the banking system? What tolerances
to risk and stress should the plumbing exhibit? Must
there always be enough liquidity everywhere, or could
that promote risky behaviour or cause inflationary
floods? How much should be spent on removing
dangerous impurities? How thoroughly and often
should the plumbing be inspected and renewed?
The main conclusions offered by the Symposium papers
suggested those questions might be answered as follows.
If international financial architecture were an edifice,
that it should contain much glass, as well as fora
where the private sector, national authorities and
supranational bodies could meet and work in harmony.
There should also be watchhouses for phenomena and
data. The main existing central building, the IMF,
certainly did not need replacing, although its links to
the rest of the structure might require some
strengthening, in addition to the recent changes in
train. Much important work had already been done;
but questionnaires and site checks needed to be
followed by a continuing process of repair and renewal.
Rather too many parts of the building remained
unconnected to the main concourse and to each other,
to the detriment of their inhabitants. A structural
survey would reveal that the interface with government
budgets was a key weakness in some areas. Using large
official financing to rescue public sector overborrowers
and private sector overlenders from their past errors was
not a buttress, but a threat, to the fabric in the future.
Standstills in emergencies would offer much more
promise. And the building’s future safety was perhaps
guaranteed best by adopting rules that no longer
rewarded misbehaviour or excessive risk taking.
International Financial Architecture
325
References
CChhuuii,, MM,, GGaaii,, PP aanndd HHaallddaannee,, AA ((22000022)), ‘Sovereign liquidity crises: analytics and implications for public policy’,
Journal of Banking and Finance, forthcoming.
FFrryy,, MM aanndd SSiinnccllaaiirr,, PP ((22000022)), ‘Inflation, debt, fiscal policy and ambiguity’, International Journal of Finance and
Economics, forthcoming.
KKrruueeggeerr,, AA ((22000022)), ‘A new approach to sovereign debt restructuring’, IMF, mimeo.
MMoorrrriiss,, RR aanndd SShhiinn,, HH SS ((22000000)), ‘Rethinking multiple equilibria in macroeconomic modelling’, NBER
Macroeconomics Annual, MIT Press.
PPoowweellll ,, AA ((22000022)), ‘The Argentine crisis: bad luck, bad economics, bad politics, bad advice?’, paper presented at the
CCBS, May 2002.
RReeiinnhhaarrtt,, CC aanndd RRooggooffff ,, KK ((22000022)), ‘The modern history of exchange rate arrangements’, paper presented to the
CCBS Conference on International Capital Movements, June 2002.
SSiinnccllaaiirr,, PP aanndd SShhuu,, CC ((22000011)), ‘International capital movements and the international dimension to financial
crises, in Brealey, R et al (eds), Financial stability and central banks, Routledge.
326
On this great annual occasion last year I spoke about the
problems that confronted us in keeping the UK economy
on course in the face of the cold winds blowing from
abroad. I drew attention to the risk that, in seeking to
avoid being sucked down into the Charybdis of adverse
external influences, by lowering interest rates in this
country, we could find ourselves thrown onto the Scylla
of excessive domestic exuberance.
In the event, the cruel terrorist attacks on New York
and Washington on 11 September had a damaging
short-term impact on economic conditions everywhere.
We were drawn irresistibly towards Charybdis despite the
resilience of domestic consumer demand, which we
sought to sustain by steering towards Scylla, with further
reductions in interest rates. The economy as a whole
became becalmed over the winter.
That, of course, was disappointing after 37 successive
quarters of relatively steady growth. But the economy as
a whole still managed to grow in the year to the first
quarter—by 1% on the present data—which was
somewhat faster than in a number of other G7
countries. The labour market held up well, with the
number of people in employment (on the LFS measure)
rising by 184,000 in the year to April, to an all-time
high; and while the number of unemployed people on
the same measure rose somewhat over the year, it fell on
the claimant count basis, to 945,000 in May, a 26-year
low. And inflation, though rather more volatile from
month to month, averaged 2.2% over the past year, very
close to our 21/2% target, although it fell to 1.8% in May.
For only the third time in the past nine years the rate of
growth in the year to the first quarter fell below the rate
of inflation.
So we have much to be thankful for, despite the hostile
external environment we have had to contend with. We
can also be grateful that the difficult international
environment has not seriously undermined the stability
of the global financial system—though that’s another
story.
The key question now is where are we going from here.
Well, the relatively good news is that the external
economic storms seem to be beginning to abate.
The United States in particular, which experienced
negative growth in the middle of last year, saw a
surprisingly strong recovery on the back of a reversal of
falling stocks, over the winter; and while this may not
have continued on the same scale through the spring,
consumer spending has remained encouragingly
resilient. The uncertainty looking forward relates
primarily to the prospects for a recovery of investment
spending as we move into the second half of the year.
Most economic analysts—including ourselves in the
Bank—are reasonably optimistic, pointing in particular
to the continuing underlying strength of US productivity
growth despite the economic slowdown and evidence
that demand for computer hardware has picked up. The
consensus economic forecast is for overall growth in the
United States—perhaps after something of a lull in the
The monetary policy dilemma in the context of theinternational environment
In his annual speech(1) at the Mansion House, the GGoovveerrnnoorr describes the United Kingdom’s relativelystrong economic performance over the past year, despite the difficult international environment. Lookingforward, the GGoovveerrnnoorr points to signs of recovery in the United States, the eurozone and Japan,suggesting that the pressure on the United Kingdom’s externally-exposed sectors should lessen,particularly if sterling’s recent overall depreciation were maintained. He notes that domestic demandgrowth will need to moderate to accommodate the expected increase in external demand. But theprospect for the United Kingdom is for a return to better-balanced growth, with inflation remaining closeto target.
(1) Given at the Lord Mayor’s Banquet for Bankers and Merchants of the City of London at the Mansion House on 26 June 2002. This speech can be found on the Bank’s web site at www.bankofengland.co.uk/speeches/speech174.htm
The monetary policy dilemma in the context of the international environment
327
second quarter—to pick up steadily to around trend—to
3% or perhaps somewhat more—through the second
half of the year into next. But it has to be said that many
US businesses themselves seem less convinced about
future earnings prospects; and financial markets, too,
remain uncertain, notably about equity valuations—
partly as a result of the recent spate of corporate
governance and accounting failures.
The outcome in the United States, of course, is
fundamentally important to the prospects elsewhere.
But on the reasonably optimistic consensus view for the
United States, the outlook is for recovery to around
trend growth in the eurozone and for modest positive
growth even in Japan. And there are encouraging signs
of stronger growth elsewhere in Asia—though parts of
South America have problems of their own.
If global economic recovery seems likely to provide a
more hospitable international environment for our
own economy, so too do recent developments in
foreign exchange markets. Last year I pointed out
that sterling’s exchange rate was at a 15-year low
against the dollar, but close to its peak against the
euro. In overall terms sterling’s effective exchange rate
index against other currencies generally had been
relatively stable—at around 105 plus or minus 5%—for
most of the past two to three years. That pattern of
exchange rates made life particularly difficult for the
euro-exposed sectors or our economy, and, given that
the eurozone represents over 50% of our external trade,
contributed significantly to the imbalance within our
economy.
Happily, from our point of view, and indeed in the
context of the global external imbalance, we have
recently seen a significant strengthening of the euro
against the dollar, and to a lesser extent against sterling,
which will help to ease some of the earlier tensions.
Sterling’s exchange rate has recently moved to a two-year
high against the dollar and nearly a three-year low
against the euro. Sterling has consequently also
weakened in overall effective terms, from around 107
some months ago to currently about 103—which is still
within the earlier range.
It is frankly anyone’s guess whether the recent pattern of
exchange rate movements will be extended, but what has
happened so far, taken together with the improved
prospect for global demand, suggests that the adverse
external factors weighing down on the UK economy over
the past two years or so should now diminish. That
offers the prospect of both stronger and more balanced
demand and overall output growth. Indeed, we are
already seeing—in surveys, but also now in actual data
for the past month or so—increasing signs of a recovery
in overall output growth, including a recovery in the
manufacturing sector, which was the hardest hit by the
global environment last year.
But our policy dilemma has not simply gone away. In
fact, trying to weigh up the prospects for the various
different components of demand, in order to decide
what we need to do with interest rates to keep overall
aggregate demand growing broadly in line with the
underlying supply-side capacity of the economy to meet
that demand, is a perpetual dilemma.
In the present context, what that means is that, in order
to accommodate the expected improvement in external
demand, we are likely to need a moderation in the recent
strength of domestic demand growth—particularly the
consumer spending growth, which has sustained the
overall economy through the period of global
weakness—if we are to avoid a build-up of inflationary
pressure further ahead.
The necessary moderation of consumer demand growth
could come about of its own accord, if, for example,
consumers become more reluctant to take on additional
debt. It is perhaps not surprising that, as they have
become more acclimatised to a low inflation/low
nominal interest rate environment, households have
been prepared to incur more debt relative to incomes,
and the debt-to-income ratio has indeed risen to an
all-time high over recent years. But that adjustment
cannot go on indefinitely and will at some point come to
an end.
In the meantime, many commentators have recently
focused on accelerating house price inflation—which is
closely related to the build-up of debt—as worrying
evidence of a bubble that will eventually burst. And we,
on the MPC, agree that the current rate of house price
increases is unsustainable, although it is less obvious
that that necessarily applies also to the present level of
house prices, though of course it may. In any event,
while, of course, we pay close attention to what is
happening to house prices, as we do to other asset
prices, our principal concern, in this part of the
equation, is with consumer demand as a whole—and
that depends on a range of other factors, not just on
house prices.
328
BBaannkk ooff EEnnggllaanndd QQuuaarrtteerrllyy BBuulllleettiinn:: Autumn 2002
What is clear is that, if external demand does improve as
we expect, and if consumer demand growth does not
moderate to accommodate it, then we will need at some
point to raise interest rates to bring that moderation
about. It is important that both borrowers and lenders
should understand and take account of that possibility
in their decision-making. But you will note the
qualifications. It is not a threat—if anything it is a
promise. It’s what our mandate from the Chancellor
requires us to do for the very good reason that low,
stable inflation is both a necessary condition for and a
reflection of overall macroeconomic stability over the
medium and longer term, which is clearly what we all
want to see.
Given the qualifications, it is not a promise on which we
will necessarily deliver immediately. We start from a
position in which inflation is somewhat below our
symmetrical target, and there is little evidence at this
point of significant inflationary pressure; wage
pressures, in particular, remain relatively benign—
despite the robust labour market—and it is crucial that
that should continue. That gives us a certain amount of
time to assess the unfolding evidence on the evolution of
both the external situation and the strength of overall
domestic demand, though one can legitimately debate
just how much risk we can afford to take by waiting. But
if and when we do decide to act, you should see that as a
sign of the strength of the economy rather than as
evidence of weakness. Indeed, my message to you this
evening is that, after a difficult passage over the past
twelve months, we can reasonably now look forward to
more favourable offshore winds in the period ahead,
opening the way to somewhat stronger and better
balanced overall output growth, but with inflation
remaining close to target. Few things are certain in
life—the distance between Scylla and Charybdis has not
widened, but it is a more comforting prospect than I was
able to offer you last year.
My Lord Mayor, I’m not sure how much comfort that
message will bring to the Merchants and Bankers of the
City of London here this evening. But they will certainly
have enjoyed, as I have, your very generous hospitality,
which has been in the very best tradition of the
occasion. And on behalf of all your guests I thank you—
and the Lady Mayoress—for that. I thank you, too, for
your sterling efforts—throughout your mayoralty—in
support of the City, in its civic affairs, in its business
activity, and in its increasing engagement with our
neighbouring communities. And I thank you not least
for the role that you have played—on behalf of all of
us—in helping to celebrate so splendidly Her Majesty’s
Golden Jubilee.
329
Introduction
Every so often, a particular issue excites those who are
interested in monetary policy. For some period, the
issue preoccupies analysts, then it tends to fade as it is
replaced by something more exciting. In fact, given the
slow-moving nature of the economy, issues tend to
remain important for long periods, sometimes many
years. For example, over two years ago, during my
confirmation hearings at the Treasury Select Committee,
there was a lengthy discussion of the strength of sterling
and when it was going to fall.
The advantage of this is that one can write about this
year’s hot topics and also last year’s hot topics without
wasting the reader’s time. So here I look at some
current issues, such as consumer debt and house prices,
and some of last year’s topics, such as the New Economy.
In fact, in what follows, I cover a wide range of issues, all
of which are still live. First, I consider the question, has
the MPC exhibited a deflationary bias? The answer is
no. Second, I discuss the New Economy and discuss the
question, is the United Kingdom going to experience a
surge in trend productivity growth(3) in the near future?
The answer is no. Third, I analyse the ‘imbalances’
which currently afflict the UK economy. In particular, I
focus on consumption growth, debt and house prices;
domestic demand growth and the exchange rate; weak
manufacturing and strong services. Broadly, I conclude
that unsustainable imbalances do not require any
special response from the MPC over and above its
watching brief on inflationary pressures looking forward.
Finally, I look at the current prospects for monetary
policy and explain why, when the MPC central projection
for inflation rises above target at the two-year forecast
horizon, this should not automatically mean that
interest rates have to rise immediately in order to keep
inflation close to target.
Has the MPC exhibited deflationary bias?
It is often noted by commentators that, aside from the
odd month, RPIX inflation has been below the 2.5%
target since 1999 Q2. Periodically this has led to
accusations that monetary policy has been too tight and
that the MPC has had a deflationary bias.(4) Of course,
policy being too tight, ex post, does not necessarily
imply a deflationary bias. The decisions may be spot on,
but subsequent inflation reducing shocks can easily
produce undershooting over quite long periods.
To investigate the hypothesis that the MPC has been
biased towards deflation, we report (in Table A,
column 2) the path of RPIX inflation had the short-run
Monetary policy issues: past, present, future
In this speech,(1) Professor Stephen Nickell(2) considers four issues. First, has the MPC exhibited adeflationary bias? The answer is no. Second, is the United Kingdom going to experience a surge intrend productivity growth in the near future for New Economy reasons? The answer is no. Third, thereis an analysis of the ‘imbalances’ that currently afflict the UK economy. The broad conclusion is thatunsustainable imbalances do not require any special response from the MPC over and above its watchingbrief on inflationary pressures looking forward. Fourth, there is a discussion of the current prospect formonetary policy. This explains why, when the MPC central projection for inflation rises above target atthe two-year forecast horizon, this should not automatically imply a rise in interest rates.
(1) Delivered as a speech at a lunch organised by Business Link and the Coventry and Warwickshire Chamber ofCommerce in Leamington Spa on 19 June 2002. This speech can be found on the Bank’s web site atwww.bankofengland.co.uk/speeches/speech173.pdf
(2) Member of the Bank of England’s Monetary Policy Committee and School Professor of Economics, London School ofEconomics. Opinions expressed here are entirely personal and do not reflect the views of any official body. I am mostgrateful to Kate Barker, Nicoletta Batini, Charlie Bean, Ian Bond, Jo Cutler, Nick Davey, Jenni Greenslade, Brian Jackson, Mervyn King and Kenny Turnbull for help with the production of this paper.
(3) The emphasis is on trend growth here. There will be a surge in productivity growth as we emerge from the recessionfor purely cyclical reasons.
(4) Most recently this issue arose in the press following the presentation of Wallis (2001) at the Royal Economic SocietyConference in March 2002 and the publication of Wadhwani (2002). Wallis himself does not, in fact, argue thatmonetary policy has had a deflationary bias although he does remark that ‘excessive concern with upside risks was notjustified over the period considered’.
330
BBaannkk ooff EEnnggllaanndd QQuuaarrtteerrllyy BBuulllleettiinn:: Autumn 2002
interest rate been set at 1/4 percentage point lower than
its actual value since 1997 Q3.(1) That is, we suppose
that, in 1997 Q3, the MPC reduced the repo rate by 1/4 percentage point on top of any other changes it
actually made and that all subsequent MPC decisions on
rate changes remained unchanged. To see what is going
on, it is convenient to divide the period from 1999 Q2
into two sub-periods, namely 1999 Q2 to 2001 Q1 and
2001 Q2 to 2002 Q1.
First, we consider the inflation outturns for the two-year
sub-period 1999 Q2 to 2001 Q1. Over this time, the
average inflation undershoot each quarter was
0.39 percentage points and, had the MPC set the repo
rate at a level 1/4 percentage point lower, the average
inflation undershoot would have been 0.18 percentage
points. Furthermore, throughout the two-year period,
the RPIX inflation rate would still have remained below
target in every quarter. So over this period, we have
prima facie evidence of deflationary bias.
Assuming that the full impact of interest rate changes
takes about two years to come through, the question we
must now investigate is why the decisions taken by the
MPC in 1997 Q2 to 1999 Q1 generated a repo rate
which was probably 1/4 percentage point too high in the
light of the inflation outcome two years later. Before
looking at specific reasons, it is worth noting, first, that
to discover, ex post, that an economic policy was a tiny
fraction away from what would have been optimal with
20/20 hindsight, is hardly a major criticism. Second, it
should be borne in mind that throughout the decision
period, 1997 Q2 to 1999 Q1, inflation was actually
above target in every quarter.
So, over the period 1997 Q2 to 1999 Q1, was the MPC
exhibiting a small, perhaps even understandable,(2)
amount of deflationary bias, or was the subsequent
undershoot the consequence of inflation-reducing
shocks which could not reasonably have been forecast?
Two factors were important over this period. More or
less throughout, the exchange rate forecasts seriously
underpredicted the outcome. Like most forecasters, the
MPC uses an element of the ‘uncovered interest parity’
theory(3) when trying to guess where the exchange rate is
going. The idea here is that if domestic interest rates
are higher than foreign interest rates, investors must
think that the domestic exchange rate is going to fall,
otherwise prospective returns on domestic assets
would completely dominate those on foreign assets,
so why would anyone hold foreign assets? Since over
the period, UK interest rates did, indeed, tend to be
higher than average foreign rates, the MPC, along
with other forecasters, predicted that the sterling
exchange rate would tend to fall over the two-year
forecast horizon. In fact, the UK exchange rate was
trending upwards for much of the period from 1997 Q2
to 2000 Q2, an upward move which continued to
surprise (see Chart 1).
And these surprise upward moves in the rate exerted
downward pressure on import price inflation and hence
on RPIX inflation throughout the relevant period.
Table ARetail price inflation (per cent per annum): 1998–2002
RPIX RPIX RPIY1/4 point off interest rates since 1997 Q3
1998 2 2.94 2.95 2.253 2.55 2.57 2.114 2.53 2.56 1.83
1999 1 2.53 2.59 1.832 2.30 2.39 1.623 2.17 2.29 1.424 2.16 2.31 1.67
2000 1 2.09 2.28 1.932 2.07 2.29 1.723 2.13 2.38 1.844 2.11 2.40 1.77
2001 1 1.87 2.20 1.582 2.26 2.63 2.633 2.38 2.78 2.814 1.95 2.38 2.41
2002 1 2.37 2.83 2.73
Sources: RPIX and RPIY provided by ONS. Column 2 is based on a simulation using the Bank of England Medium Term Model.
Chart 1Sterling effective exchange rate
Source: ONS.
(1) This is based on a simulation of the Bank of England Macroeconomic Model (see Bank of England (2000)).(2) Understandable because inflation was above target throughout the period and the MPC was new and needed to build
up credibility.(3) Actually, MPC exchange rate forecasts are an average of the path based on uncovered interest parity and an
unchanged level. Ex post, neither work particularly well, but then neither does any other known forecasting method.
Monetary policy issues: past, present, future
331
The second factor was the rate of wage inflation. Over at
least some of the relevant period, earnings forecasts
tended to be too high, partly because earnings data were
inaccurate on the upside and partly because
improvements in the workings of the labour market were
not fully apparent in the 1997/98 period.
I think it may readily be argued that movements in both
the exchange rate and wage inflation over the relevant
period could not reasonably have been forecast given
the available information, and both these movements
were systematically favourable for inflation. In the light
of this, I think it is hard to make the accusation of
systematic deflationary bias stick over the decision
period 1997 Q2 to 1999 Q1. So what about the
decision period (ie, 1999 Q2 to 2000 Q1) for the RPIX
inflation outcome in 2001 Q2 to 2002 Q1?
Over this latter period, the RPIX inflation undershoot of
the target was, on average, 0.26 percentage points each
quarter. However, as we can see from Table A, column 2,
had the repo rate been 1/4 percentage point lower since
1997 Q3, there would have been an average quarterly
overshoot of 0.16 percentage points. Although
0.16 percentage points is smaller than 0.26 percentage
points it would be pedantic to argue that there was even
prima facie evidence of deflationary bias over the
year-long decision period from 1999 Q2. It is, however,
true that food price inflation during 2001 was
surprisingly high, at least in part because of the bad
weather in the preceding winter and the foot-and-mouth
crisis. On the other hand, the fact that average excise
taxes were not uprated in line with inflation in the 2001
budget systematically reduced RPI inflation during the
following year. This at least partly contributed to the
upward surge in RPIY inflation during the period
2001 Q2 to 2002 Q1 seen in Table A, column 3.(1) On
balance, therefore, it seems hard to convict the MPC of
deflationary bias during this later period. Of course, the
consequences of the MPC decisions taken since
2000 Q3 have yet to work through fully, so we must wait
a little before judgment on that period can be passed.
The New Economy
By 2000, it had become clear that in the second half of
the 1990s, labour productivity growth in the United
States had been greater than in most major OECD
countries for the first time since the Second World War.
Until 1995, labour productivity growth in the OECD was
broadly consistent with the notion that the United
States was the country on the technological frontier and
the other countries were slowly and fitfully catching up.
From 1995, however, the United States appeared to be
pulling away, with trend productivity growth having risen
by around 1 percentage point per annum, a very
surprising event. Something new appeared to be
happening, so not surprisingly the phrase ‘New
Economy’ was coined.
Huge amounts of research have been devoted to
understanding what is going on. In the United States, to
explain the surge in growth. In Europe, to explain the
absence of such a surge.
The key issue for our purposes lies in the answer to the
questions: (i) Can we expect a surge in UK labour
productivity growth? And (ii) If so, when? This issue is
vital for monetary policy, because if an increase in trend
labour productivity growth is expected, then potential
supply growth can also be expected to increase. This
implies that a higher level of aggregate demand growth
is consistent with stable inflation and monetary policy
might have to be looser than would otherwise be the
case. The correct monetary policy response would, of
course, depend on the extent to which the expected
increase in trend productivity growth, of itself, generated
increases in consumption and investment.
What happened in the United States?
It is clear from the data that the productivity surge in
the United States from 1995 was intimately connected to
information technology (IT). This is starkly
demonstrated in Table B, taken from Stiroh (2001),
which shows how the increase in US labour productivity
growth after 1995 was concentrated more or less
Table BAnnual labour productivity growth (per cent) in theUnited States, 1987–99Industries 1987–95 1995–99 Share of total output
in per cent (1999)
IT-producing 8.24 11.90 5.3IT-intensive 1.24 2.61 47.3Other 0.98 1.11 47.4
Source: Stiroh (2001). The productivity numbers are employment-weighted averages acrossindustries. The IT-producing sectors are industrial machinery and equipment, andelectronic and other electric equipment.The IT-intensive industries are those whose 1995 IT capital shares were above the1995 median. The main sectors here are telecom, wholesale trade, retail trade,finance and insurance (not real estate), business services and health services. Thesesix sectors make up 77% of the total output of the IT-intensive industries.
(1) RPIY inflation is RPIX inflation with taxes excluded. Part of the surge in RPIY inflation is also due to base effects, inother words, it would have risen even if excise taxes had been uprated in line with inflation in the 2001 budgetbecause they were increased above inflation in the previous budget.
332
BBaannkk ooff EEnnggllaanndd QQuuaarrtteerrllyy BBuulllleettiinn:: Autumn 2002
exclusively in the small IT-producing sector and among
the IT-intensive users, most of which are in the service
sector. The remaining half of the economy has seen no
significant rise in productivity growth over the relevant
period. There are three significant ways in which IT
generates improved productivity growth. First, simply
accumulating IT capital faster raises productivity growth
almost mechanically, so long as workers equipped with
more IT capital become more productive. Second, the
rate of production of the scientific knowledge involved in
producing IT equipment has risen. This is exemplified
by Moore’s law, according to which the processing power
embodied in chips doubles every 18 months. This
generates the productivity surge in the IT-producing
sector as well as rapidly falling IT equipment prices.
Finally, firms learn how to use IT equipment more
effectively, recognising that they have to make
complementary organisational investments (see
Brynjolfsson and Hitt (2000), for example). The
contributions of these three different factors to the total
surge in productivity growth are roughly 40%, 25%, 35%
respectively.(1)
So what has been going on in the United States is fairly
clear. There has been a very high rate of productivity
growth in IT production for some time now and this
became even more rapid in the late 1990s. As a
consequence, the price of IT equipment has been falling
fast and this encouraged a very rapid rate of
accumulation of IT equipment by many firms outside the
IT production sector, particularly in the large service
sectors of the economy. These firms have been able to
make very effective use of this IT equipment, often by
comprehensive reorganisation of their operations, and
this has led to a marked improvement in productivity
growth in these sectors. These improvements are driven,
fundamentally, by the decline in the price of IT
equipment and this can be expected to continue for
some time. However, simply possessing lots of new IT
equipment is not enough. Productivity improvements
also depend on the ability of management to make good
use of it via best practice methods. Many US companies
have been able to do this, partly because they are driven
into it by the very high levels of competition in many
sectors of the US economy. Finally, it is worth remarking
that, despite the recent downturn, US productivity
growth looks as if it will continue at high levels for some
time yet.
Is this surge in productivity growth going to happen inthe United Kingdom?
By 1999, it had become clear to all that the New
Economy was well under way in the United States. This
led many to believe that trend productivity growth was
going to rise in the United Kingdom. For example, a
Goldman Sachs paper (Davies, Brookes and Williams
(2000)) cautiously predicted that UK productivity
growth in 2000–05 would be 0.8 percentage points
per annum higher than in 1995–2000, a very substantial
improvement. But before looking to the future, we must
first look at what was going on in the United Kingdom
while US productivity growth was taking off.
Not very much, at least according to the official
data. On more or less every measure, average UK
productivity growth in the second half of the 1990s was
below that in the first half and below the long-run
average. To be a bit more specific, between 1995 and
1999, US labour productivity in manufacturing rose
by 24 percentage points relative to that in the
United Kingdom and in market services, it rose by
13 percentage points (O’Mahony (2002), Table 5). It is
worth recording that similar numbers apply also to the
European Union relative to the United States, although
the gap is smaller in manufacturing and larger in market
services.
So the situation is one where UK productivity
performance in the second half of the 1990s was
exceptionally modest, with labour productivity
starting in 1995 at a substantially lower level than
in the United States in all sectors, followed by a
marked widening of this gap in the subsequent five
years. So what factors might help us to understand
this picture, particularly as firms in the United Kingdom
had just as much access to rapidly cheapening IT
equipment as those in the United States? Consider first
the role of direct inputs into production. As in the
United States, investment in IT equipment in the United
Kingdom rose substantially in the second half of the
1990s, as it did in telecoms equipment. Yet the
productivity growth of IT users was actually lower in the
second half of the 1990s in the United Kingdom than
over the previous 15 years (see UBS (2002), Table 6).
This is the key factor, because, as we saw in Table B, it is
the IT users who have driven the improvement in the
United States.
(1) There is a lot of disagreement about these numbers. Looking at Jorgenson and Stiroh (2000), Oliner and Sichel(2000), Whelan (2000), US Council of Economic Advisors (2001), Gordon (2000), the respective percentages are inthe ranges 30%–50%, 12%–40%, 0%–63%. The contribution of the third factor is a particularly contentious issue.
Monetary policy issues: past, present, future
333
It has been argued that some of this stark contrast is
simply down to measurement, with US output and price
data taking better account of quality change than the
corresponding UK data. However, as UBS (2002)
indicates, even taking account of this, the contrast
between the performances of the United Kingdom
and the United States remains. A further argument
points out that the unemployed who obtained work in
large numbers in the United Kingdom from 1994 were
less skilled than the average employee and this held
back productivity growth in the later 1990s.
Unfortunately, this argument applies equally in the
United States over the same period. Furthermore, the
evidence presented in Crafts and O’Mahony (2000)
suggests that labour force skill differences did not
play any significant role in the widening gap
between the United Kingdom and the United States
in the second half of the 1990s. So we are left with
the simple fact that firms in the United States made
much better use of their new IT equipment than firms
in the United Kingdom over this period. The question
thus remains, what are the underlying forces at
work here and are they going to moderate in the near
future?
Underlying forces holding the United Kingdom back andprospects for the future
It has long been known that, while the United Kingdom
has a very strong academic science base, it has been
weak at translating this strength into innovation and
industrial performance, with the notable exception of
the pharmaceuticals sector. Over the past 20 years, the
share of both publicly and privately funded R&D
expenditure in GDP has actually fallen in the United
Kingdom, whereas the opposite has happened in the
United States. R&D expenditure is particularly
important in this regard, because not only does it
generate innovations, but it also helps firms absorb the
innovations of others (see Griffith et al (1999)), a factor
which is particularly relevant here.
This is just a symptom, however. Underlying this are the
following basic factors. First, as is clear from the analysis
presented in McKinsey Global Institute (1998), Nickell
and Van Reenen (2002) and Baily (2001), the
competitive pressures on US firms are, on average,
greater than those exerted on firms in the United
Kingdom. These pressures are fundamental in
encouraging the use of new technologies and more
generally in forcing firms to find ways to improve their
operations. Second, in the United Kingdom, general
management skills are not as highly valued in the market
place as skills in finance, accountancy and consultancy,
so the brightest graduates tend to go into the latter
areas. Furthermore, because a high proportion of UK
companies is not operating at the frontier of best
practice, the majority of managers learn the job outside
a best-practice environment. This, of itself, inhibits the
absorption of innovations. Finally, while the UK
education system is good for those at the upper end of
the ability range, the structure in place for post-school
vocational education is weak. This leads to a noticeable
shortfall in technician skills which holds back the
absorption of new technologies.
Are any of these structural factors changing so that UK
firms will start making significantly better use of IT in
the near future? Starting with competition, many will
find it surprising that there is a shortage of competition
in the United Kingdom. Surely after 20 years of
privatisation, deregulation and globalisation,
competition is very fierce. However, it should not be
forgotten that we have had 20 years of ‘restructuring’ in
many sectors, much of which has had the effect of
sustaining and even concentrating market power. So net
rates of return on capital in private non-financial
corporations have been higher over the past five years
than over most of the previous 25 (see Chart 3 below,
page 339), hardly a sign of significantly greater
competition which might have been expected to cut
profitability.(1) This has been reinforced by the fact that,
(1) In standard models, increasing competition leads to lower rates of return on capital. For those who like formalanalysis, if a firm has a Cobb-Douglas technology and faces a downward-sloping demand curve, maximising profitwould imply
where
P is the output price, Y is output, K is capital, L is employment, W is the wage, C is the cost of capital, (1 – a),a arethe Cobb-Douglas exponents and h is the demand elasticity. These equations imply that profits,
and that . But
where PI is the price of investment goods, r – P.I /PI is the ‘own’ real interest rate and V is the rate of depreciation. So
the return on capital,
which is decreasing in the standard measure of competition, h.
334
BBaannkk ooff EEnnggllaanndd QQuuaarrtteerrllyy BBuulllleettiinn:: Autumn 2002
by comparison with the United States, UK competition
law has been relatively feeble. However, the UK
anti-trust system was significantly strengthened from
1 March 2000, when the 1998 Competition Act came
into force. This, along with further prospective
tightening as a result of future planned changes in
competition law, should raise the level of competitive
intensity in the United Kingdom in the future.
A second recent change that could have some
longer-term impact on productivity growth is the
introduction of R&D subsidies. The available evidence
suggests that these will raise business spending on R&D
as a proportion of GDP and that this will have a positive
impact on productivity over the longer term. Of the
other issues mentioned above, the only one which looks
as if it might improve any time soon is the state of
post-school vocational education. Some policy effort is
currently being devoted to this area, although it would
be a mistake to be too optimistic about the outcome.
The history of policy in vocational education over the
past 20 years reveals just how difficult it is to make
significant improvements.
To summarise, the United Kingdom has not seen a surge
in IT-generated productivity growth since 1995. This has
made the productivity gap, which was wide in 1995,
considerably wider by the present time. This period
contrasts with the period from 1980 to 1995 when
productivity growth in the United Kingdom was higher
than in the United States and there was some closing of
the productivity gap (see O’Mahony and de Boer (2002),
Table 2). The persistence of a wide productivity gap and
the ability of the US economy to make better use of new
IT equipment are due to structural problems in the UK
economy. These include low levels of R&D spending and
innovation, low levels of competitive pressure on firms,
weakness in general management skills, with the labour
market valuations placed on finance and consultancy
skills being much higher than on general management,
and, finally, weakness in post-school vocational
education. Some of these problems are the subject of
systematic attention from policy-makers but it would be
unrealistic to expect any dramatic changes in the near
future. As a consequence, there is no justification for
setting monetary policy in the expectation that the UK
economy will experience a surge in trend productivity
growth in the near future. It may happen, but it would
be unwise to bet on it.(1)
Imbalances
The UK economy is currently suffering from a severe
case of ‘imbalances’, mostly of the ‘unsustainable’ variety.
Of course, since the economy is rarely spot on its
balanced growth path, the situation at any moment is
almost always unsustainable. So is the present pattern of
‘unsustainable imbalances’ unduly worrying? Before
discussing the nature of the current imbalances, why
should imbalances in general concern the MPC? There
are two sorts of reasons. First, as the economy moves
back towards a more balanced situation this process may
involve increasing inflationary or deflationary pressures,
which would require action on the part of the MPC. If
the economy starts from a more unbalanced position,
these pressures may well be greater. It is good to be
prepared for such possibilities, although it may well be
that no action is required until the pressures seem likely
to materialise. Second, certain types of imbalance may
resolve themselves very rapidly (eg, the bursting of an
asset price bubble), which generates a high level of
output and inflation volatility, making it harder for the
MPC to hit the inflation target. The tricky question for
the MPC is what action to take, if any, prior to the rapid
resolution of the imbalance when it is uncertain when,
or indeed if, any such rapid resolution will occur.
In the light of this general discussion, what are the
specific imbalances currently facing us? There are three
which we shall discuss.
(i) Consumption growth has been higher than GDP
growth in all bar two quarters since 1998 Q1. In
the more recent past, this consumption growth has
been backed by very rapid growth in household
debt, much of it secured on a housing stock that is
rapidly rising in value. Household debt to income
ratios are at record levels.
(ii) Related to this is the fact that annual domestic
demand growth has been higher than annual GDP
growth for the past five years. Over the same
period, the balance of payments deficit has been
rising steadily as a proportion of GDP.
(1) Lest it be thought that there is any contradiction between this statement and the much discussed rise in trend outputgrowth assumed by the Treasury in the Budget forecast, there is not. The Treasury assumes no rise in trendproductivity growth, looking forward. The rise in trend output growth comes from the assumed rise in employmentgrowth arising from the projected increase in the growth in the population of working age provided by theGovernment Actuary’s Department. Note that trend output growth is the sum of trend labour productivity growth andtrend employment growth.
Monetary policy issues: past, present, future
335
(iii) Manufacturing growth is much lower than growth
in the service sector and the rate of return on
capital in manufacturing is at a low level, both
absolutely, and relative to that in services.
Consumption, debt and house prices
Associated with high levels of consumption growth have
been two related factors, both of which have caused
widespread concern, namely very high levels of growth of
household debt and of house prices. We consider each
in turn.
(i) Debt
The level of household debt relative to post-tax income
in the United Kingdom is currently 118%, a record level,
having grown by over 15 percentage points since 1997.
Such record levels are by no means unusual in the sense
that debt to income ratios in many other countries are
of the same order of magnitude,(1) but if this level is
worrying, then this fact hardly makes it less worrying.
Furthermore, the debt to income ratio is still rising.
This increasing debt is being used both to fund property
purchase and to help finance a high level of
consumption growth, which has recently been driven by
growth in durables consumption.
Why might households wish to hold higher levels of debt
today than in the past? One good reason is that we now
live in a period of low inflation and hence of low
nominal interest rates. Here is a simple example.
Suppose that inflation is 12% and interest rates are 15%.
Consider a household taking out a substantial debt of
four times its annual disposable income, repayable over
25 years. Then in the first year, they would have to pay
over 60% of post-tax income in interest and repayments.
Typically lenders would not allow a loan of this size
because of the enormous size of the payments required
relative to income. Of course, after ten years, annual
disposable income would have risen by more than a
multiple of three at 12% inflation and the costs of
interest plus repayment would be relatively modest. By
the end of the loan, after 25 years, the repayment and
interest costs would be negligible as disposable income
would have risen by 17 times! But, because borrowers
cannot cope with the early years of the loan, they
probably would neither want nor be allowed to take on
such a debt.(2)
However, if inflation is 2.5% and interest rates are 5.5%,
so the real interest rate is the same, this front-end
loading problem no longer applies. In the first year, a
loan of the same size would require a payment of around
one quarter of annual disposable income, a situation
which prudent lenders and borrowers might be prepared
to contemplate. After ten years, the payments on the
loan would have diminished very little as a proportion of
disposable income (by around 20%), so the prudent
borrowers would have to recognise they were in for a
long haul. Nevertheless, it is easy to see that in a low
inflation environment, households might well wish to,
and be allowed to, take on more debt relative to income,
despite unchanged real interest rates. So it is no
surprise that in the high-inflation period from 1974 to
1980 the average loan to income ratios for first-time
house buyers were less than 2, whereas in the late 1990s
they were in excess of 2.2 according to data from the
Survey of Mortgage Lenders. Furthermore, the rules
governing income multiples that mortgage lenders will
contemplate are significantly more generous today than
in the high-inflation 1970s.
All this both enables us to understand why households
might wish to borrow more in a low-inflation
environment and why this may be a prudent course.
This is reflected that the relatively low levels of
household ‘income gearing’ we see today, despite record
levels of debt (see Chart 2). At some point, of course,
the move to higher ‘equilibrium’ levels of debt will be
complete and consumption growth will slow down as a
consequence.(3) Such a shift would not be of undue
concern, indeed the MPC has some slowing of
consumption growth in its forecast. The only real
problem is the uncertainty about when it will occur, but
this is a run-of-the-mill monetary policy problem, not
one which deserves the limelight.
But perhaps consumers are taking on all this debt, not
as a result of prudent decision-making on their part and
on the part of lenders, but as a consequence of
over-optimism. For example, households might take on
higher levels of debt because they expect their real
(1) For example, in Germany, the United States and Australia.(2) Things are a bit more complex than this because it may be argued that, in a period of high inflation, borrowers and
lenders would be happy to keep increasing the size of any loan as inflation continues to boost disposable income. Inpractice, however, the transaction costs associated with this process were so large during the high-inflation period inthe United Kingdom that such increases were relatively infrequent.
(3) Formally, this process might be modelled by supposing that households faced a relaxation in their liquidityconstraints, with the precise point in time at which this relaxation happened differing across households. This wouldlead to a higher rate of aggregate consumption growth for some period.
336
BBaannkk ooff EEnnggllaanndd QQuuaarrtteerrllyy BBuulllleettiinn:: Autumn 2002
incomes to grow more rapidly than normal in the future.
This may have arisen because, for a variety of reasons,
real disposable income has been particularly high over
the past two years. Given the prospects for productivity
growth, which we have already discussed, such a
projection would be mistaken. If such an error were
being made, once it became apparent, debt growth and
hence consumption growth would tend to slow.
However, it seems unlikely that this would generate a
sudden collapse in consumption. Much more likely is
the sort of gradual slowdown we are expecting to happen
anyway.
So is there any reason why we should be particularly
worried about high and rising levels of debt? Of course,
one obvious point is that the higher the level of debt,
the more responsive is the absolute level of household
interest charges to changes in interest rates and this may
lead consumption to be more sensitive to monetary
policy shifts. This is something that the MPC must take
into account when setting rates, but it need not cause us
undue concern. Of course, when interest rates rose from
8% to 15% during the late 1980s, the high levels of debt
which had been accumulated at that time
unquestionably made the subsequent recession
substantially worse. However, the lesson to be drawn is
not that high levels of debt are bad, per se, but that we
should not allow the inflationary position to become so
bad that such drastic interest rate rises become
necessary to get inflation under control. Of course, one
of the other consequences of the dramatic interest rate
rises in the late 1980s was a collapse in the housing
market. House prices are again approaching levels,
relative to earnings, that ruled at that time,(1) so is this a
problem?
(ii) House prices
House prices have grown by between 10% and 20% in
the past year, depending on which index is used. This is
a simple consequence of high demand (low mortgage
rates, high rates of population growth, the attractions of
buy-to-let relative to equity investments) meeting low
supply (lowest rate of new house building since the War).
Housing wealth serves as collateral for borrowing and
has a direct impact on consumption. If housing wealth
grows faster, consumption and aggregate demand grow
faster and, looking forward, inflation rates will be higher.
So the rate of growth of housing wealth is a significant
factor in the deliberations of the MPC. This is
straightforward. But should the MPC worry about house
price inflation over and above its direct implications for
future inflation?
It has been argued that there are particular dangers if a
house price bubble develops. Consider the following
scenario. In the rapidly expanding buy-to-let market
(now around 5% of housing transactions) a symptom of a
bubble would be if individuals invested in buy-to-let
property when expected rental incomes were not enough
to cover running costs plus mortgage payments, relying
on rapid capital gains to make it a worthwhile
investment. The bubble then bursts when house price
growth starts to slow for other reasons. The buy-to-let
investors start to suffer from cash-flow problems and the
slow growth in the value of their (housing) collateral
limits borrowing opportunities, so they dump their
properties on the market and house prices fall. This
would affect regular owner-occupiers although, given low
interest rates, they would have no difficulty in servicing
their debts even if they had negative equity. Of course,
negative equity could cause further problems if, for
example, lenders insisted on higher interest rates
because of the reduced collateral, or even insisted on
some loan repayment. However, it is unlikely that a fall
in house prices would cause the problems of the early
1990s, when interest rates were 15%, but it may still be
something to be avoided.
5
7
9
11
13
15
17
1987 89 91 93 95 97 99 2001
Regular payments (a) (excluding unsecured principal repayments)
Regular payments (a)
Interest payments only
Percentage of annual personal disposable income
0
Chart 2Measures of UK household income gearing(a)
Sources: ONS, Financial Research Survey and Bank of England.
(a) ‘Regular payments’ should be considered as non-discretionary payments made as they include estimates of all interest payments, regular mortgage principal repayments and unsecured loan principal repayments. They do not include principal repayments on credit cards as these data are unavailable.Information on unsecured loan principal repayments is not available prior to 1997. Regular mortgage principal repayments are estimated prior to 1998.
(1) In 2002 Q1, the house price to earnings ratio was between 8.1% (ODPM) and 20.8% (Halifax) below the 1989 peak.The house price to personal disposable income ratio was between 25.3% (ODPM) and 36.5% (Halifax) below the 1989peak.
Monetary policy issues: past, present, future
337
So where does the MPC enter this story? The boom and
bust in the housing market does not cause any obvious
problems for the average level of inflation, but it might
increase volatility, which would make it harder for the
MPC to keep close to the target. So could the MPC do
anything about this scenario if it thought it was
developing? In order to answer this question, we have to
step back a little and consider how the MPC operates.
By and large, it is sensible to think of the MPC as each
month assessing where RPIX inflation is going over the
next couple of years, and if it is expected to go above
target, to raise rates and if it is expected to go below
target, to lower rates. There may be a tendency for
commentators to focus on the two-year horizon, by
which time the bulk of the current changes in rates will
have worked through, but the relationship between the
central projection of inflation relative to target at this
two-year horizon and the current decision is not
mechanical. For example, if the central projection is
below target for most of the time, only rising to the
target at the end of the two-year horizon, it may not be
necessary to raise interest rates immediately in order to
hit the target. The target could still be attained by
raising rates later and this might be desirable if, for
example, this path of rates generates a less volatile path
of domestic demand in the meantime, which would
enable the MPC to keep inflation closer to target further
out. Furthermore, in this situation, raising rates later
keeps inflation closer to target ahead of the two-year
horizon.
How does this potential flexibility relate to a housing
market bubble? If the MPC was certain that a housing
market bubble was developing, it has been suggested
that it could, perhaps, generate a smoother path of
output and inflation by having a higher interest rate
than would otherwise be necessary in the absence of the
nascent bubble, in order to prevent the bubble actually
taking off. The suggestion, therefore, is that the MPC
could do better by responding to asset price bubbles
over and above the response that would occur in the
normal way of things, because asset prices affect
demand and inflation prospects.
The problem with this proposal is that, if it is to work
out for the best, bubbles or speculative booms in asset
prices must be rapidly identifiable and readily
distinguishable, for example, from rapid movements in
asset prices generated by warranted changes in
expectations about fundamentals. The use of monetary
policy to interfere with the latter kind of changes in
asset prices would not improve inflation performance
and would distort the allocation of resources for
investment. So it is not to be recommended. Generally
speaking, it is not possible, ex ante, to identify
bubbles or speculative booms with any certainty, so
the use of monetary policy to ‘nip them in the bud’
is not normally a feasible strategy. We simply do not
have enough information to operate this kind of
sophisticated policy in a reliable fashion.(1) As a
general rule it is better simply to focus on the
longer-term consequences of asset price movements
for inflation and leave it at that. Of course, if house
prices really take off, MPC forecasts of inflation would
also rise and the MPC would tend to raise rates. So
this strategy is not a recipe for complacency, merely a
common-sense framework for dealing with the
problem.
Domestic demand growth and balance of paymentsdeficits
Related to the high level of consumption growth in
recent years has been the high rate of domestic demand
growth overall (consumption plus investment plus
government spending), which has also been outstripping
GDP growth for at least five years. Despite favourable
movements in the terms of trade over this period, we
have been spending more than we produce, with the
gap being filled by the rest of the world, who have
been supplying us with more than we have been
supplying them with. So we have had a persistent trade
deficit over this period, which is now running at over 2%
of GDP.
The first question to ask is whether this is sustainable?
For example, if UK citizens have lots of overseas assets,
generating high levels of interest and dividends and
these substantially exceed the interest and dividends
paid out to foreign holders of UK assets, then we could
go on with a trade deficit forever. So to help consider
sustainability, we add these income and transfer flows to
the trade deficit, the result being the current account
deficit. On average over the past three years this is only
a shade lower than the trade deficit, being just under
2% of GDP. If correct, this would imply that foreigners
were acquiring UK assets at a considerably faster rate
than UK citizens were acquiring foreign assets. Is this
an unsustainable situation?
(1) Not to mention the fact that a substantial rise in rates might be necessary to choke off a speculative boom and thatsuch a rise would create further instability.
338
BBaannkk ooff EEnnggllaanndd QQuuaarrtteerrllyy BBuulllleettiinn:: Autumn 2002
The answer is not clear cut. First, it is worth noting that,
even before domestic demand started growing faster
than GDP, we regularly had a current account deficit.
For example, from 1970 to 1994, our deficit apparently
averaged 1% of GDP over the whole period. Second, if
some countries are in deficit, others must be in surplus.
The United States has an enormous deficit on its balance
of payments (around $400 billion per year). So where
are the big surplus countries? There are two. Japan
is a major surplus country but the biggest by far is a
place called Discrepancy, which has had an annual
surplus of around five times the UK deficit over each
of the past three years. Much of this simply reflects
mismeasurement, which systematically overstates deficits
and understates surpluses. One reason for this is that it
is much harder to keep track of dividend and interest
receipts, which often accrue to individuals, than of
dividend and interest payouts, which are generally
undertaken by companies or official bodies. The upshot
is that the measured UK balance of payments deficit may
overstate the true figure by an unknown amount. A third
point germane to sustainability is the fact that the
balance of payments measures omit capital gains and
losses on asset holdings. There is some evidence to
suggest that capital gains on UK holdings of foreign
assets exceed those on foreign holdings of domestic
assets. This would increase the sustainability associated
with any measured UK deficit.(1)
The upshot of this discussion is that measured payments
deficits of the current size can be sustained for
considerable periods without significant adjustments
being required. Furthermore, as the United States and
Europe recover, their demand growth will rise relative to
that in the United Kingdom (which has had a less severe
slowdown). As a consequence, the UK payments deficit
should fall back, particularly if consumption growth
slows for reasons discussed in the previous section. The
fundamental question is the extent to which, during the
process where domestic consumption growth slows and
world demand growth rises, the sterling exchange rate
falls, thereby generating additional inflationary pressure.
Some might argue that a fall in the exchange rate is
necessary to help shift demand and supply patterns
towards the international sector and assist in
‘rebalancing’ the economy. This may or may not be true,
but the key issue for the MPC is whether we should
adjust interest rates today in order to forestall the
potential inflationary consequences of a possible
significant exchange rate fall in the future.
Turning to the issue of exchange rate movements, I said
during my confirmation hearings at the Treasury Select
Committee that any forecasts I make of the exchange
rate are not worth the paper they are written on. I see
no reason to change this view, although I should add
that the evidence suggests that modest current account
deficits can be sustained for many years and are of no
great value for forecasting exchange rate movements in
the medium term. Of course, when the MPC presents its
forecasts in the Inflation Report, we have to make some
assumption about future movements in the exchange
rate. As we have already noted, the MPC’s current
convention is to include in its forecast an element based
on uncovered interest parity. Given the international
pattern of interest rates, this generates a very modest
decline in the effective sterling exchange rate of around
21/2% over the two-year forecast horizon. This
forecasting convention has nothing to do with the issues
currently being discussed. The important question here
is, should we move interest rates up today in order to
forestall the potential inflationary consequences of a
significant exchange rate fall that might come about as
UK domestic demand growth slows and international
demand growth speeds up? In my view, the answer is
simply no. If, and when, such a fall in the exchange rate
comes about, then is the time to make appropriate
adjustments, if any, in interest rates. To act
pre-emptively on this front is difficult because of the
huge uncertainties involved in forecasting exchange rate
movements and not really necessary because the lags to
demand from interest rates and exchange rates are of the
same order of magnitude.(2)
The imbalance between manufacturing and services
The third imbalance we shall discuss is that associated
with the recent rapid decline in the manufacturing
sector and the contrasting continuing strength of the
services sector. This pattern is directly associated with
the fact that the manufacturing sector is more open to
international competition than the service sector and so
the manufacturing sector has been strongly hit by the
weakness of the world economy in the recent past and by
the strength of sterling since 1997. Again, this might
(1) For example, once direct investment is measured at market value as opposed to the book value used in the standardstatistics, the overall UK external asset position swings from one of net liabilities to one of net assets. In particular, ifthe method of adjusting direct investment due to Cliff Pratten is used, the deterioration in the UK external net assetposition since 1996 is eliminated. For details, see Senior and Westwood (2001), in particular the box on page 390.
(2) The argument here is that the first-round, price-level effects from sudden exchange rate moves should beaccommodated, with monetary policy only acting on the potential second-round effects.
Monetary policy issues: past, present, future
339
suggest that if the gap between the growth rates of the
two sectors is to close, not only does the world economy
have to recover, which it is expected to do, but the
sterling effective exchange rate has to fall.
In order to analyse this question, it is helpful to look at
the broader historical context. First, the manufacturing
sector in all developed countries has been contracting
relative to the service sector for at least 30 years. In
other words, some degree of ‘imbalance’ is the normal
state of affairs. Second, reflecting this, we find that
employment in manufacturing in the United Kingdom
has been falling at an average rate of around 125,000
per year since it reached its peak in 1966. So an average
of over 2,000 manufacturing jobs have been lost in the
United Kingdom in every week for over 35 years. Since
overall employment has risen over the same period, this
loss has been more than compensated by the jobs
gained in other sectors.
This continuing shift in activity out of manufacturing
and into services has been driven by the fact that the
returns on capital employed in manufacturing have been
substantially below those in other sectors since at least
1970, as we can see from Chart 3. Looking more closely
at the 1990s, we find that, when sterling was weak in the
period from 1992 to 1996, rates of return in
manufacturing rose to their highest level in the past
30 years. Even when sterling strengthened markedly up
to 2000, the rate of return remained in excess of 8%,
far higher than the average in the 1970s and 1980s.
Only when world demand shrank dramatically in 2001
did manufacturing returns fall back towards the
previous average. So even if sterling remains strong, we
can expect manufacturing returns to revert to a
historically relatively high level when world demand
growth recovers.
Overall, the history of sectoral employment shares and
rates of return suggests that a return to a normal state of
imbalance might come about without any substantial
moves in the exchange rate.(1) This does not mean that
they won’t happen. All that is being said here is that a
substantial fall in sterling is not inevitable because of
the manufacturing/services imbalance. While this
imbalance has devastating implications for those
individuals and regions on the losing side, the
consequences for monetary policy, as opposed to social
security policy or regional and industrial policy, are
relatively slight.
Final reflections on imbalances
We have discussed the usual three suspects, first
consumption growth, debt and house prices, second
domestic demand growth and the exchange rate and
finally, weak manufacturing and strong services. The
following points have emerged.
First, there are good reasons why households will
wish to, and be allowed to, take on higher levels of debt
in an era of low inflation and low interest rates. At some
stage, this move to a higher level of debt can be
expected to come to an end of its own accord with
consumption growth slowing as a consequence. There is
no good reason why monetary policy should respond
specifically to the current levels of debt over and above
the response required if the consequent high levels of
consumption growth lead to excessive inflationary
pressure.
Second, while the current rate of house price inflation
clearly cannot continue indefinitely, there is no reason
for a special increase in interest rates specifically to
‘cool down the overheating housing market’. If the
impact of rising housing wealth on demand is such as to
lead us to expect inflation to move above target, then the
MPC must act. Generally speaking, it simply does not
have enough information to start following a policy of
pricking asset price bubbles, en passant.
Third, domestic demand growth has outstripped the
growth of domestic output for some years and this has
(1) Of course, the relevant exchange rate here is the real exchange rate. However, since 1993, real and nominal exchangerates have not diverged greatly because inflation rates in the major trading nations of the OECD have been very closeby recent historical standards.
0
2
4
6
8
10
12
14
16
18
Services
Manufacturing
Total
1970 75 80 85 90 95 2000
Chart 3Net rates of return
Source: ONS.
340
BBaannkk ooff EEnnggllaanndd QQuuaarrtteerrllyy BBuulllleettiinn:: Autumn 2002
resulted in a current account deficit of around 2% of
GDP. Recovery in the world economy may reduce this
deficit, but questions remain concerning the
sustainability of this situation. It is arguable that the
required slowdown in domestic demand may be
associated with a significant and permanent fall in the
effective exchange rate. While this may happen at some
stage,(1) it is by no means certain. Furthermore, there is
no reason for the MPC to react to the possible
inflationary consequences of such a fall until it actually
comes about.
Fourth, the dramatic contrast between a growing service
sector and a contracting manufacturing sector is, in
part, the consequence of the recent weakness in the
international economy and the strength of sterling and,
in part, a reflection of the longer-term decline of UK
manufacturing which started in the mid-1960s. Over
the past 35 years, an average of 125,000 manufacturing
jobs have disappeared every year, with at least as many
being gained in other sectors. When the world economy
recovers, the gap between the growth rates of
manufacturing and services will narrow; we should not
expect it to disappear, but merely to revert to
longer-term trends. Monetary policy has no important
role in this process. Overall therefore, unsustainable
imbalances do not normally require any special response
from the MPC over and above its watching brief on
inflationary pressures looking forward.
The Budget and other near-term challenges formonetary policy
The general thinking of the MPC on the outlook for the
future is set out in the latest Inflation Report
(May 2002) so I shall only emphasise two crucial issues.
First, the recovery in the rest of the world, and the
recovery in the United Kingdom are both looking a little
fragile. In the United States, there is no strong evidence
of any recovery in investment which would underpin a
strong economy going forward. In the euro area, the
weak recovery is almost entirely dependent on buoyant
net trade and in the United Kingdom, while the surveys
look good, we are still waiting for real action outside the
retail sector.
Overlaying this is the Budget. The key points on
spending were additional expenditure on health from
2003/04 and tax credits to improve work incentives and
provide further support for those on low incomes.
These changes are paid for by 1 percentage point
increases in the rate of both employees’ and employers’
National Insurance contributions (NICs) taking effect
from April 2003. The implications of these changes on
future demand growth and inflation are fraught with
uncertainties. The demand effects depend crucially on
four factors. First, how much of the increased health
expenditure is spent on higher pay for existing workers
in the health sector? Second, what proportion of the
additional health spending goes on imports relative to
the import content of the private expenditure that is
displaced because of the tax increases. This is an issue
because the average import content of private
expenditure is greater than the average import content
of public expenditure. Third, how much of the tax credit
income gets spent relative to the reduction in spending
generated by the tax increases that pay for it? This is an
issue because the recipients of the tax credits, on
average, spend a higher proportion of their income than
the average tax payer. Fourth, how much is private
consumption moderated in advance of the increase in
employee NICs? The MPC has made judgments on all
these questions based on all the information they could
muster. But the level of uncertainty here is very great.
Then there is the direct impact on inflation of increases
in NICs. Increases in employees’ NICs will affect
inflation if the employees succeed in obtaining
additional pay rises to compensate. Increases in
employers’ NICs inevitably raise labour costs in the first
instance and the implications for inflation then depend
on the extent to which these increases are passed on to
prices, are absorbed by reduced margins or additional
productivity increases or are compensated by employees
accepting lower pay increases. Again, the MPC has made
judgments based on past behaviour and other
information. And again, the level of uncertainty is very
great.
Further work is under way at the Bank to try and reduce
the range of uncertainty, but in the meantime, the
results of the MPC’s judgments, as reported in the
Inflation Report, generate a central projection for a
steady recovery in the United Kingdom. GDP growth is
above trend by 2003 and inflation below target until the
very end of the forecast horizon, by which time it is
rising quite rapidly. A mechanical rule that translates
the central projection of inflation at the two-year
horizon directly into an immediate interest rate
adjustment suggests a rise in interest rates at the May
meeting. The fact that this did not happen has been
(1) Indeed, it may already be happening.
Monetary policy issues: past, present, future
341
taken by some as an indication that the MPC has taken
its eyes off the inflation target.
This is simply not the case. The central projection has
inflation below target until the last quarter of the
forecast horizon. While it may take two years for a rise
in interest rates to have its maximum impact on
inflation, it will start having an impact well before the
two years are up. This immediately implies that a rise
in interest rates some time later will still affect
demand early enough to lower inflation to the target
level at the end of the two-year period. Given the
current fragility of the recovery in the United Kingdom,
some delay in raising rates seems to be an eminently
sensible strategy.
Summary and conclusions
Today, I have covered four issues that loom large when
considering UK monetary policy, past, present and to
come. First, I ask the question, has the MPC
demonstrated a bias towards deflation? The answer is
that it is difficult to convict the MPC of such a bias,
despite the chronic undershooting of the 2.5% inflation
target since 1999 Q2. To get this answer, I consider
what would have happened had interest rates been 1/4 percentage point lower from 1997 Q3. Over the
two-year period from 1999 Q2 to 2001 Q1, inflation
would still have undershot the target in every quarter.
Nevertheless, I absolve the MPC of deflationary bias over
the relevant decision period 1997 Q2 to 1999 Q1,
essentially because, during this period, there was a
tendency to underpredict the sterling effective exchange
rate and to overpredict wage inflation. Both these
mispredictions could not reasonably have been avoided,
given the information available at the time.
During the year from 2001 Q2 to 2002 Q1, had interest
rates been 1/4 percentage point lower since 1997 Q3,
inflation would have overshot the target most of the
time, so during the relevant decision period from
1999 Q2, there is no evidence of deflationary bias. Of
course, for decisions taken since mid-2000, we have yet
to see the full consequences. Overall, it is hard to
convict the MPC of deflationary bias and, as a general
point, the fact that MPC decisions are found to be only
a fraction away from optimal with the benefit of 20/20
hindsight looks like something to celebrate rather than
carp about.
The second question discussed is whether or not we can
expect a surge in productivity growth in the United
Kingdom because of the New Economy. I argue that the
answer is probably not. Since 1995, unlike in the United
States, the United Kingdom has not seen a boom in
IT-generated productivity growth despite a lot of IT
investment. This contrasts with the previous 15 years
when productivity growth in the United Kingdom was
higher than in the United States. However, there
remained a large productivity gap which has been
widening since 1995 as the US has made better use of
new IT equipment. So why is the United Kingdom
unlikely to start catching up in the near future?
Basically, because of a number of structural problems.
Relative to the United States, the United Kingdom has
lower levels of R&D spending and innovation, lower
levels of competitive pressure on firms and weaknesses
in both general management and post-school vocational
education. While some of these problems are the
subject of systematic attention from policy-makers it will
take some time before any results will start to show
through. In the meantime, there is no justification for
setting UK monetary policy in the expectation of a surge
in trend productivity growth.
The third issue I have dealt with is that of ‘imbalances’.
I have discussed three aspects of the imbalances
problem: (i) consumption growth, debt and house
prices; (ii) domestic demand growth and the exchange
rate; and (iii) weak manufacturing and strong services.
My conclusions are as follows. Consumers are
reasonable in taking on higher levels of debt because we
now have low inflation, and low general levels of interest
rates. At some point, the move to higher levels of debt
will come to an end of its own accord, with consumption
growth slowing as a consequence. The problem for the
MPC is that it doesn’t know when. However, there is no
strong argument for monetary policy to respond
specifically to the current levels of debt over and above
the response required if the consequent high levels of
consumption growth lead to excessive inflationary
pressure.
Turning to house prices, there is no reason for a special
increase in interest rates specifically to ‘cool down the
housing market’. The MPC should act if and when the
impact of rising housing wealth on demand is such as to
push forecast inflation above target. Generally speaking,
it does not have enough information to follow a policy of
pricking asset price bubbles, en passant. On domestic
demand growth, the fact that this has been stronger
than output growth has put pressure on the balance of
payments. Recovery in the world economy will reduce
342
BBaannkk ooff EEnnggllaanndd QQuuaarrtteerrllyy BBuulllleettiinn:: Autumn 2002
the deficit, as will the prospective slowdown in
consumption growth when it eventually occurs. This
process may or may not be associated with a fall in
sterling. Whether sterling falls or not, there is no reason
for the MPC to react to any possible inflationary
consequences of such a fall until it actually happens.
Finally, on ‘imbalances’, there is the dramatic contrast
between expanding services and contracting
manufacturing. The gap in growth rates between the two
has been particularly wide because of the strength of
sterling and the weakness of the world economy.
However, some gap between the two is the normal state
of affairs. Since the mid-1960s, an average of 125,000
manufacturing jobs have disappeared every year.
Furthermore, over the whole period, rates of return on
capital in manufacturing have been far below the
average rates of return in other sectors. So when the
world economy recovers, the gap will narrow but not
disappear. Furthermore, monetary policy has no
important role in this process, which is far more the
concern of regional and industrial policy and social
security policy.
The final issue I have discussed is the immediate
challenge facing monetary policy. The consequences of
recent changes in the macroeconomy and the Budget
have led the MPC to produce a central projection for a
strong recovery in the United Kingdom with GDP
growth above trend by 2003 and inflation below target
until the very end of the forecast horizon when it moves
above target and is rising quite rapidly. A mechanical
rule that translates the central projection of inflation at
the two-year horizon directly into an immediate interest
rate adjustment would have suggested a rise in rates in
May. But such a mechanical rule is not the best way of
keeping inflation close to target. While it may take two
years for a rise in interest rates to have its full impact,
the impact starts well before then. So in a situation
where inflation only rises above target at the end of the
two-year horizon, a rise in interest rates further down
the line would still affect demand early enough to lower
inflation to the target level at the end of the two-year
period and beyond. And given the current fragile nature
of the recovery in the United Kingdom, some delay in
raising rates seems to be the correct strategy for keeping
inflation as close as possible to target over the long haul.
Monetary policy issues: past, present, future
343
References
BBaaiillyy,, MM NN ((22000011)), ‘Macroeconomic implications of the New Economy’, Economic Policy for the Information
Economy, symposium sponsored by the Federal Reserve Bank of Kansas City, Jackson Hole, Wyoming
(30 August–1 September 2001).
BBaannkk ooff EEnnggllaanndd ((22000000)), Economic Models at the Bank of England, September 2000 update.
BBrryynnjjoollffssssoonn,, EE aanndd HHiitttt,, LL ((22000000)), ‘Beyond computation: information technology, organizational transformation
and business practices’, Journal of Economic Perspectives, Vol. 14 (4), Fall, pages 23–48.
CCoouunncciill ooff EEccoonnoommiicc AAddvviisseerrss ((22000011)), ‘Annual Report’, Economic Report of the President, Washington D.C.,
January.
CCrraaffttss,, NN aanndd OO’’MMaahhoonnyy,, MM ((22000000)), A perspective on UK productivity performance, London School of
Economics, mimeo.
DDaavviieess,, GG,, BBrrooookkeess,, MM aanndd WWiilllliiaammss,, NN ((22000000)), ‘Technology, the Internet and the New Global Economy’,
Goldman Sachs Global Economics Paper No. 39, Goldman Sachs, March.
GGoorrddoonn,, RR JJ ((22000000)), ‘Does the ‘New Economy’ measure up to the great inventions of the past?’, Journal of Economic
Perspectives, Vol. 14 (4), Fall, pages 49–74.
GGrriiffffiitthh,, RR,, RReeddddiinngg,, SS aanndd VVaann RReeeenneenn,, JJ ((11999999)), Mapping the two faces of R&D: productivity growth in a
panel of OECD manufacturing industries, Institute for Fiscal Studies, mimeo.
JJoorrggeennssoonn,, DD aanndd SSttiirroohh,, KK ((22000000)), ‘Raising the speed limit: US economic growth in the information age’,
Brookings Papers on Economic Activity, (1), pages 125–211.
MMccKKiinnsseeyy GGlloobbaall IInnssttiittuuttee ((11999988)), Driving productivity and growth in the UK economy, Washington D.C.
NNiicckkeellll ,, SS aanndd VVaann RReeeenneenn,, JJ ((22000022)), ‘Country studies: the United Kingdom’, Technological Innovation and
Economic Performance, edited by Steil, B, Victor, D G and Nelson, R R, Princeton University Press.
OOlliinneerr,, SS DD aanndd SSiicchheell ,, DD EE ((22000000)), ‘The resurgence of growth in the late 1990s: is information technology the
story?’, Journal of Economic Perspectives, Vol. 14 (4), Fall, pages 3–22.
OO’’MMaahhoonnyy,, MM ((22000022)), ‘Productivity and convergence in the EU’, National Institute Economic Review, Vol. 180, April,
pages 72–82.
OO’’MMaahhoonnyy,, MM aanndd ddee BBooeerr,, WW ((22000022)), ‘Britain’s relative productivity performance: has anything changed?’,
National Institute Economic Review, Vol. 179, January, pages 38–43.
SSeenniioorr,, SS aanndd WWeessttwwoooodd,, RR ((22000011)), ‘The external balance sheet of the United Kingdom: implications for
financial stability?’, Bank of England Quarterly Bulletin, Winter, pages 388–405.
SSttiirroohh,, KK JJ ((22000011)), ‘Information technology and the US productivity revival: what do the industry data say?’,
Federal Reserve Bank of New York Staff Reports, No. 115, January.
UUBBSS ((22000022)), ‘European potential: summary’, UBS Global Asset Management, Research Paper, 12 April.
344
BBaannkk ooff EEnnggllaanndd QQuuaarrtteerrllyy BBuulllleettiinn:: Autumn 2002
WWaaddhhwwaannii,, SS BB ((22000022)), ‘The MPC: some further challenges’, speech delivered at the National Institute of Economic
and Social Research, 16 May.
WWaalllliiss,, KK FF ((22000011)), Chi-squared tests of interval and density forecasts, and the Bank of England’s fan charts,
Department of Economics, Warwick University.
WWhheellaann,, KK ((22000000)), Computers, obsolescence and productivity, Board of Governors of the Federal Reserve System,
Finance and Economics Discussion Series, No. 2000–06, January.
Speeches made by Bank personnel since publication of the previous Bulletin are listed below.
MMaannssiioonn HHoouussee ssppeeeecchh..Speech by The Rt Hon Sir Edward George, Governor, at the Lord Mayor’s Banquet for Bankers and Merchants of the City of London,Mansion House on 26 June 2002.www.bankofengland.co.uk/speeches/speech174.htm Reproduced on pages 326–28 of this Bulletin.
MMoonneettaarryy ppoolliiccyy iissssuueess:: ppaasstt,, pprreesseenntt,, ffuuttuurree..Speech by Stephen Nickell at a lunch organised by Business Link and the Coventry and Warwickshire Chamber of Commerce,Leamington Spa on 19 June 2002.www.bankofengland.co.uk/speeches/speech173.pdf Reproduced on pages 329–44 of this Bulletin.
Bank of England speeches
Contents of recent Quarterly Bulletins
The articles and speeches which have been published recently in the Quarterly Bulletin are listed below. Articles from
November 1998 onwards are available on the Bank’s web site at www.bankofengland.co.uk/qbcontents/index.html
Articles and speeches (indicated S)
November 1999
Sterling market liquidity over the Y2K period
Public sector debt: end March 1999
The external balance sheet of the United Kingdom:
recent developments
News and the sterling markets
New estimates of the UK real and nominal yield curves
Government debt structure and monetary conditions
Challenges for monetary policy: new and old (S)
Sterling’s puzzling behaviour (S)
Monetary policy and asset prices (S)
Interest rates and the UK economy—a policy for all
seasons (S)
February 2000
Sterling wholesale markets: developments in 1999
Recent developments in extracting information from
options markets
Stock prices, stock indexes and index funds
Private equity: implications for financial efficiency and
stability
Back to the future of low global inflation (S)
British unemployment and monetary policy (S)
Before the Millennium: from the City of London (S)
May 2000
A comparison of long bond yields in the United
Kingdom, the United States, and Germany
Money, lending and spending: a study of the UK
non-financial corporate sector and households
Monetary policy and the euro (S)
The new economy and the old monetary economics (S)
The impact of the Internet on UK inflation (S)
Monetary policy and the supply side (S)
August 2000
Public sector debt: end-March 2000
Age structure and the UK unemployment rate
Financial market reactions to interest rate
announcements and macroeconomic data releases
Common message standards for electronic commerce in
wholesale financial markets
The environment for monetary policy (S)
Monetary union and economic growth (S)
August 2000 (continued)
The exchange rate and the MPC: what can we do? (S)
The work of the Monetary Policy Committee (S)
November 2000
The external balance sheet of the United Kingdom:
implications for financial stability?
Economic models at the Bank of England
International financial crises and public policy: some
welfare analysis
Central banks and financial stability
Inferring market interest rate expectations from money
market rates
Central bank independence (S)
Britain and the euro (S)
Monetary challenges in a ‘New Economy’ (S)
Spring 2001
Sterling wholesale markets: developments in 2000
The Kohn report on MPC procedures
Bank capital standards: the new Basel Accord
The financing of technology-based small firms: a review
of the literature
Measuring interest accruals on tradable debt securities
in economic and financial statistics
Saving, wealth and consumption
Mortgage equity withdrawal and consumption
The information in UK company profit warnings
Interpreting movements in high-yield corporate bond
market spreads
International and domestic uncertainties (S)
Current threats to global financial stability—a European
view (S)
Summer 2001
The Bank of England inflation attitudes survey
The London Foreign Exchange Joint Standing
Committee: a review of 2000
Over-the-counter interest rate options
Explaining the difference between the growth of M4
deposits and M4 lending: implications of recent
developments in public finances
Using surveys of investment intentions
Summer 2001 (continued)
Can differences in industrial structure explain
divergencies in regional economic growth?
Has there been a structural improvement in US
productivity?
International efforts to improve the functioning of the
global economy (S)
Monetary stability as a foundation for sustained
growth (S)
The ‘new economy’: myths and realities (S)
The impact of the US slowdown on the UK economy (S)
Autumn 2001
Public attitudes about inflation: a comparative analysis
Measuring capital services in the United Kingdom
Capital flows and exchange rates
Balancing domestic and external demand (S)
The international financial system: a new
partnership (S)
‘Hanes Dwy Ddinas’ or ‘A Tale of Two Cities’ (S)
Has UK labour market performance changed? (S)
Some reflections on the MPC (S)
Winter 2001
The external balance sheet of the United Kingdom:
implications for financial stability
Public sector debt: end-March 2001
The foreign exchange and over-the-counter derivatives
markets in the United Kingdom
The Bank’s contacts with the money, repo and stock
lending markets
The formulation of monetary policy at the Bank of
England
Credit channel effects in the monetary transmission
mechanism
Financial effects on corporate investment in UK business
cycles
Why house prices matter
The prospects for the UK and world economies (S)
Maintaining financial stability in a rapidly changing
world: some threats and opportunities (S)
Monetary policy: addressing the uncertainties (S)
Economic imbalances and UK monetary policy (S)
Do we have a new economy? (S)
Spring 2002
The London Foreign Exchange Joint Standing
Committee: a review of 2001
Spring 2002 (continued)
Provision of finance to smaller quoted companies: some
evidence from survey responses and liaison meetings
Explaining trends in UK business investment
Building a real-time database for GDP(E)
Electronic trading in wholesale financial markets: its
wider impact and policy issues
Analysts’ earnings forecasts and equity valuations
On market-based measures of inflation expectations
Equity wealth and consumption—the experience of
Germany, France and Italy in an international context
Monetary policy, the global economy and prospects for
the United Kingdom (S)
Three questions and a forecast (S)
Twenty-first century markets (S)
The stock market, capacity uncertainties and the outlook
for UK inflation (S)
Summer 2002
Public attitudes to inflation
The Bank of England’s operations in the sterling money
markets
No money, no inflation—the role of money in the
economy
Asset prices and inflation
Durables and the recent strength of household spending
Working time in the United Kingdom: evidence from the
Labour Force Survey
Why are UK imports so cyclical?
Monetary challenges (S)
The Monetary Policy Committee: five years on (S)
Household indebtedness, the exchange rate and risks to
the UK economy (S)
Autumn 2002
Committees versus individuals: an experimental analysis
of monetary policy decision-making
Parliamentary scrutiny of central banks in the United
Kingdom and overseas
Ageing and the UK economy
The balance-sheet information content of UK company
profit warnings
Money and credit in an inflation-targeting regime
International Financial Architecture: the Central Bank
Governors’ Symposium 2002
The monetary policy dilemma in the context of the
international environment (S)
Monetary policy issues: past, present, future (S)
Bank of England publications
Working papers
Working papers are free of charge; a complete list is available from the address below. An up-to-date list of workingpapers is also maintained on the Bank of England’s web site at www.bankofengland.co.uk/wp/index.html, whereabstracts of all papers may be found. Papers published since January 1997 are available in full, in PDF.
No. Title Author
125 Assessing the impact of macroeconomic news announcements on securities prices under Andrew Claredifferent monetary policy regimes (February 2001) Roger Courtenay
126 New estimates of the UK real and nominal yield curves (March 2001) Nicola AndersonJohn Sleath
127 Sticky prices and volatile output (April 2001) Martin EllisonAndrew Scott
128 ‘Oscillate Wildly’: asymmetries and persistence in company-level profitability Andrew Benito(April 2001)
129 Investment-specific technological progress in the United Kingdom (April 2001) Hasan BakhshiJens Larsen
130 The real interest rate gap as an inflation indicator (April 2001) Katharine S NeissEdward Nelson
131 The structure of credit risk: spread volatility and ratings transitions (May 2001) Rudiger KieselWilliam PerraudinAlex Taylor
132 Ratings versus equity-based credit risk modelling: an empirical analysis (May 2001) Pamela NickellWilliam PerraudinSimone Varotto
133 Stability of ratings transitions (May 2001) Pamela NickellWilliam PerraudinSimone Varotto
134 Consumption, money and lending: a joint model for the UK household sector K Alec Chrystal(May 2001) Paul Mizen
135 Hybrid inflation and price level targeting (May 2001) Nicoletta BatiniAnthony Yates
136 Crisis costs and debtor discipline: the efficacy of public policy in sovereign debt Prasanna Gaicrises (May 2001) Simon Hayes
Hyun Song Shin
137 Leading indicator information in UK equity prices: an assessment of economic tracking Simon Hayesportfolios (May 2001)
138 PPP and the real exchange rate–real interest rate differential puzzle revisited: evidence Georgios E Chortareasfrom non-stationary panel data (June 2001) Rebecca L Driver
139 The United Kingdom’s small banks’ crisis of the early 1990s: what were the leading Andrew Loganindicators of failure? (July 2001)
140 ICT and productivity growth in the United Kingdom (July 2001) Nicholas Oulton
141 The fallacy of the fiscal theory of the price level, again (July 2001) Willem H Buiter
142 Band-pass filtering, cointegration, and business cycle analysis (September 2001) Luca Benati
143 Does it pay to be transparent? International evidence from central bank forecasts Georgios Chortareas(November 2001) David Stasavage
Gabriel Sterne
The Bank of England publishes information on all aspects of its work in many formats. Listed below are some of themain Bank of England publications. For a full list, please refer to our web site www.bankofengland.co.uk/publications
144 Costs of banking system instability: some empirical evidence (November 2001) Glenn HoggarthRicardo ReisVictoria Saporta
145 Skill imbalances in the UK labour market: 1979–99 (December 2001) Pablo Burriel-LlombartJonathan Thomas
146 Indicators of fragility in the UK corporate sector (December 2001) Gertjan W Vlieghe
147 Hard Times or Great Expectations?: Dividend omissions and dividend cuts by UK firms Andrew Benito(December 2001) Garry Young
148 UK inflation in the 1970s and 1980s: the role of output gap mismeasurement Edward Nelson(December 2001) Kalin Nikolov
149 Monetary policy rules for an open economy (December 2001) Nicoletta BatiniRichard HarrisonStephen P Millard
150 Financial accelerator effects in UK business cycles (December 2001) Simon Hall
151 Other financial corporations: Cinderella or ugly sister of empirical monetary economics? K Alec Chrystal(December 2001) Paul Mizen
152 How uncertain are the welfare costs of inflation? (February 2002) Hasan BakhshiBen MartinTony Yates
153 Do changes in structural factors explain movements in the equilibrium rate of Vincenzo Cassinounemployment? (April 2002) Richard Thornton
154 A monetary model of factor utilisation (April 2002) Katharine S NeissEvi Pappa
155 Monetary policy and stagflation in the UK (May 2002) Edward NelsonKalin Nikolov
156 Equilibrium exchange rates and supply-side performance (June 2002) Gianluca BenignoChristoph Thoenissen
157 Financial liberalisation and consumers’ expenditure: ‘FLIB’ re-examined (July 2002) Emilio Fernandez-CorugedoSimon Price
158 Soft liquidity constraints and precautionary saving (July 2002) Emilio Fernandez-Corugedo
159 The implications of an ageing population for the UK economy (July 2002) Garry Young
160 On gross worker flows in the United Kingdom: evidence from the Labour Force Survey Brian Bell(July 2002) James Smith
161 Regulatory and ‘economic’ solvency standards for internationally active banks Patricia Jackson(August 2002) William Perraudin
Victoria Saporta
162 Factor utilisation and productivity estimates for the United Kingdom (August 2002) Jens LarsenKatharine NeissFergal Shortall
163 Productivity versus welfare: or, GDP versus Weitzman’s NDP (August 2002) Nicholas Oulton
164 Understanding UK inflation: the role of openness (September 2002) Ravi BalakrishnanJ David López-Salido
165 Committees versus individuals: an experimental analysis of monetary policy Clare Lombardellidecision-making (September 2002) James Proudman
James Talbot
166 The role of corporate balance sheets and bank lending policies in a financial accelerator Simon Hallframework (September 2002) Anne Vila Wetherilt
External MPC Unit discussion papers
The MPC Unit discussion paper series reports on research carried out by, or under supervision of, the external members of the Monetary Policy Committee. Papers are available from the Bank’s web site atwww.bankofengland.co.uk/mpc/extmpcpaper0000n.pdf (where n refers to the paper number). The following papershave been published recently.
No. Title Author
5 Monetary policy for an open economy: an alternative framework with optimising agents Bennett T McCallumand sticky prices (October 2001) Edward Nelson
6 The lag from monetary policy actions to inflation: Friedman revisited (October 2001) Nicoletta BatiniEdward Nelson
7 The future of macroeconomic policy in the European Union (February 2002) Christopher Allsopp
8 Too much too soon: instability and indeterminacy with forward-looking rules Nicoletta Batini(March 2002) Joseph Pearlman
9 The pricing behaviour of UK firms (April 2002) Nicoletta BatiniBrian JacksonStephen Nickell
Monetary and Financial Statistics
Monetary and Financial Statistics (Bankstats) contains detailed information on money and lending, monetary andfinancial institutions’ balance sheets, analyses of bank deposits and lending, international business of banks, publicsector debt, money markets, issues of securities and short-term paper, interest and exchange rates, explanatory notes totables, and occasional related articles. Bankstats is published quarterly in paper form, priced at £60 per annum in theUnited Kingdom (four issues). It is also available monthly free of charge from the Bank’s web site at:www.bankofengland.co.uk/mfsd/latest.htm
Further details are available from: Daxa Khilosia, Monetary and Financial Statistics Division, Bank of England:telephone 020 7601 5353; fax 020 7601 3208; e-mail daxa.khilosia@bankofengland.co.uk
The following articles have been published in recent issues of Monetary and Financial Statistics. They may also befound on the Bank of England web site at www.bankofengland.co.uk/mfsd/article
Title Author Month of issue Page numbers
Economic activity of bank holding companies Michelle Rowe July 2002 3–5
Development of euro business of banks in the Richard Walton July 2002 1–2European Union
A work programme in financial statistics— Ben Norman April 2002 5–7April 2002 update
Prices indices: a report on a meeting of the Darran Tucker April 2002 1–4Financial Statistics and Business Statistics Users’ (Office for NationalGroups Statistics)
Financial Stability Review
The Financial Stability Review is published twice a year, in June and December. Its purpose is to encourage informeddebate on financial stability; survey potential risks to financial stability; and analyse ways to promote and maintain astable financial system. The Bank of England intends this publication to be read by those who are responsible for, orhave interest in, maintaining and promoting financial stability at a national or international level. It is of especialinterest to policy-makers in the United Kingdom and abroad; international financial institutions; academics;journalists; market infrastructure providers; and financial market participants. It is available from Financial StabilityReview, Bank of England HO-3, Threadneedle Street, London, EC2R 8AH.
Practical issues arising from the euro
This is a series of booklets providing a London perspective on the development of euro-denominated financial marketsand the supporting financial infrastructure, and describing the planning and preparation for possible future UK entry.Recent editions have focused on the completion of the transition from the former national currencies to the euro inearly 2002, and the lessons that may be drawn from it. Copies are available from Public Enquiries Group, Bank ofEngland, Threadneedle Street, London, EC2R 8AH.
Economic models at the Bank of England
The Economic models at the Bank of England book, published in April 1999, contains details of the economicmodelling tools that help the Monetary Policy Committee in its work. The price of the book is £10.00. An update waspublished in September 2000 and is available free of charge.
Quarterly Bulletin
The Quarterly Bulletin provides regular commentary on market developments and UK monetary policy operations. Italso contains research and analysis and reports on a wide range of topical economic and financial issues, both domesticand international.
Back issues of the Quarterly Bulletin from 1981 are available for sale. Summary pages of the Bulletin from February 1994, giving a brief description of each of the articles, are available on the Bank’s web site atwww.bankofengland.co.uk/bulletin/index.html
The Bulletin is also available from ProQuest Information and Learning: enquiries from customers in Japan and Northand South America should be addressed to ProQuest Information and Learning, 300 North Zeeb Road, Ann Arbor, Michigan 48106, United States of America; customers from all other countries should apply to The Quorum, Barnwell Road, Cambridge, CB5 8SW, telephone 01223 215512.
An index of the Quarterly Bulletin is also available to customers free of charge. It is produced annually, and listsalphabetically terms used in the Bulletin and articles written by named authors.
Bound volumes of the Quarterly Bulletin for the period 1960–85 (in reprint form for the period 1960–85) can beobtained from Schmidt Periodicals GmbH, Ortsteil Dettendorf, D-83075 Bad Feilnbach, Germany, at a price of €105per volume or €2,510 per set.
Inflation Report
The Bank’s quarterly Inflation Report sets out the detailed economic analysis and inflation projections on which theBank’s Monetary Policy Committee bases its interest rate decisions, and presents an assessment of the prospects for UKinflation over the following two years.
The Report starts with an overview of economic developments; this is followed by six sections:
● analysis of money, credit and financial market data, including the exchange rate;● analysis of demand and output;● analysis of the labour market;● analysis of costs and prices;● summary of monetary policy during the quarter; and● assessment of the medium-term inflation prospects and risks.
The minutes of the meetings of the Bank’s Monetary Policy Committee (previously published as part of the InflationReport) now appear as a separate publication on the same day as the Report.
Publication dates
From 2002, copies of the Quarterly Bulletin and Inflation Report can be bought separately, or as a combined packagefor a discounted rate. Current prices are shown overleaf. Publication dates for 2002 are as follows:
Quarterly Bulletin Inflation Report
Spring 18 March February 13 FebruarySummer 18 June May 15 MayAutumn 23 September August 7 AugustWinter 16 December November 13 November
Copies of the Quarterly Bulletin and Inflation Report can be bought separately, or as a ccoommbbiinneedd package for adiscounted rate. Subscriptions for a full year are also available at a discount. The prices are set out below:
Destination 2002 2001
Quarterly Bulletin Quarterly Bulletin Inflation Report Quarterly Bulletin Inflation Reportand Inflation only only and Inflation onlyReport package Report package
Annual Single Annual Single Annual Single Annual Single Annual Single
United Kingdom,by first-class mail (1) £27.00 £7.50 £21.00 £6.00 £10.50 £3.00 £40.00 £10.00 £12.00 £3.00
Academics, UUKK oonnllyy £18.00 £5.00 £14.00 £4.00 £7.00 £2.00 £27.00 £6.75 £8.00 £2.00Students, UUKK oonnllyy £9.00 £2.50 £7.00 £2.00 £3.50 £1.00 £14.00 £3.50 £4.50 £1.50
European countriesincluding the Republic ofIreland, by letter service £33.00 £9.00 £25.00 £7.00 £13.00 £4.00 £48.00 £12.00 £14.00 £3.50
Countries outside Europe:Surface mail £33.00 £9.00 £25.00 £7.00 £13.00 £4.00 £48.00 £12.00 £14.00 £3.50
Air mail: Zone 1 (2) £43.00 £12.00 £34.00 £9.00 £17.00 £5.00 £64.00 £16.00 £21.00 £5.25
Zone 2 (3) £45.00 £12.00 £35.00 £9.00 £18.00 £5.00 £66.00 £16.50 £22.00 £5.50
(1) Subscribers who wish to collect their copy(ies) of the Bulletin and/or Inflation Report may make arrangements to do so by writing to the address given below. Copies will be available to personal callers at the Bank from 10.30 am on the day of issue and from 8.30 am on the following day.
(2) All countries other than those in Zone 2.(3) Australasia, Japan, People’s Republic of China, the Philippines and Korea.
Readers who wish to become rreegguullaarr ssuubbssccrriibbeerrss, or who wish to purchase single copies, should send to the Bank,at the address given below, the appropriate remittance, payable to the Bank of England, together with full addressdetails, including the name or position of recipients in companies or institutions. If you wish to pay by VViissaa,,MMaasstteerrccaarrdd,, SSwwiittcchh oorr DDeellttaa, please telephone 020 7601 4030. Existing subscribers will be invited to renew theirsubscriptions automatically. Copies can also be obtained over the counter at the Bank’s front entrance.
The ccoonncceessssiioonnaarryy rraatteess for the Quarterly Bulletin and Inflation Report are noted above in italics. AAccaaddeemmiiccss aattUUKK iinnssttiittuuttiioonnss of further and higher education are entitled to a concessionary rate. They should apply on theirinstitution’s notepaper, giving details of their current post. SSttuuddeennttss aanndd sseeccoonnddaarryy sscchhoooollss iinn tthhee UUnniitteeddKKiinnggddoomm are also entitled to a concessionary rate. Requests for concessionary copies should be accompanied by anexplanatory letter; students should provide details of their course and the institution at which they are studying.
These publications are available from Publications Group, Bank of England, Threadneedle Street, London, EC2R 8AH; telephone 020 7601 4030; fax 020 7601 3298; e-mail mapublications@bankofengland.co.uk
General enquiries about the Bank of England should be made to 020 7601 4444.The Bank of England’s web site is at: www.bankofengland.co.uk
Issued by the Bank of England Publications Group.
Quarterly Bulletin and Inflation Report subscription details