quarterly bulletin autumn 2002 - bank of england · quarterly bulletin—autumn 2002 markets and...

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 [email protected]

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


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.


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


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


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


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.


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


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.


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


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.


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


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


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


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.


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


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


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


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.


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


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.


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


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


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


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.


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


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.’


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


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.


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)


282


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)


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


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)


c) DDooeess PPaarrlliiaammeenntt hhaavvee tthhee ppoowweerr ttoo vveettoo aappppooiinnttmmeennttss ttoo tthhee ccoommmmiitttteeee??

Yes No



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


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


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


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


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


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.


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


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.


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


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.


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


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


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


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).


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


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.


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


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


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.


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


TThhoommaass,, RR ((11999977bb)), ‘The demand for M4: a sectoral analysis—Part 2—The corporate sector’, Bank of England


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


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).


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


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.


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


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.


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


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


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.


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


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.


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


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.


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


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.


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


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.


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


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.


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


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.


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


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


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


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


The balance-sheet information content of UK company

profit warnings


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 [email protected]

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 [email protected]

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

quarterly bulletin autumn 2002 - bank of england · quarterly bulletin—autumn 2002 markets and...

Documents