mpc.theyserveforyou.com

mpc:

MINUTES OF THE MONETARY POLICY COMMITTEE MEETING HELD ON3-4 NOVEMBER 1999

1 Before completing their November inflation and output growth projections and turning to their immediate policy decision, the Committee discussed demand and output; the labour market; prices and costs; money, credit and asset prices; and the world economic outlook. Prior to the meeting, the Committee was briefed by Treasury officials on the Chancellor's latest projections for economic activity and the public finances.

Demand and output

2 Output was estimated to have increased by 0.9% between Q2 and Q3, significantly higher than projected in the Committee's August Inflation Report; and growth in the first half of the year had been revised upwards, so that the level of activity was also higher than expected. The quarterly rate of manufacturing output growth had been 1% in Q3. The recovery seemed to be spreading both regionally and, with some exceptions such as agriculture, sectorally. The twelve-month growth rate of final domestic demand in Q2, the latest period for which data were available, had been 4.5%, well above trend. Strong domestic demand had to some extent been offset in the first part of the year by destocking and, over a longer period, by weak net trade, reflecting the rise in sterling's exchange rate and shocks to world economic growth. Aggregate demand had therefore grown more slowly than final domestic demand, and this had contained pressures on the economy's supply capacity. More recently, however, net trade had been stronger than expected, and in Q2 had even made a positive contribution to GDP growth. Against this background, the Committee discussed whether the growth in demand was sustainable, and if not whether it would slow spontaneously, ie without a further policy tightening. 3 Within final domestic demand, consumption growth had been especially buoyant. A number of possible special factors were identified. First, strong growth in real earnings, and so real personal disposable income, might have reflected not just tight labour market conditions, but also inflation outturns that had been lower than expected when pay bargains were struck. Other things being equal, any such effect was unlikely to continue as inflation expectations had come down over the past few years, and were now broadly in line with the 2½% target. However, the Committee had built into its latest central projection for inflation a number of particular downward influences on prices - such as

mpc: 2 utility price cuts and the effects of increased competition on retailers' margins - which would tend to reduce RPIX inflation, and it was unclear to what extent these were incorporated into wage bargainers' inflation expectations. 4 Second, the rapid growth in consumption in Q1 and Q2 partly reflected recovery from the slowdown last autumn, when confidence had been hit by the shocks to the world economy and international financial markets. Third, consumer spending growth earlier in the year had partly been accounted for by an earlier-than-usual spike in car sales, reflecting changes in the timing of new registrations. That was expected to unwind in the second half of the year. Retail sales growth appeared, however, to have strengthened during the year according to data for Q3 and some surveys, so that overall consumption growth might not slow much. Moreover, strong output growth in Q3, evidence of strong tax receipts (including corporate taxes), and the relative strength of household money and credit growth and of the housing market suggested that strong final domestic demand growth might have continued beyond the first half of the year. 5 There were, though, some possible indications that consumption growth had steadied. Indicators of retail sales in Q4 appeared firm, but probably not more so than during Q3. Both the CBI Distributive Trades Survey and the CIPS Services Index pointed to a moderation of activity. Consumer confidence measures were positive but had been steady for about eight months, and in particular had not been driven upwards along with the rapid rise in house prices. And some indicators of housing market activity, such as particulars delivered, had ticked down slightly over the past two months. The Committee judged, however, that the recent strength in house prices would tend to support continuing buoyant consumption growth. House prices were generally expected to continue rising over the next year or so at an annual rate which was higher than mortgage interest rates. In the view of some members, the resulting negative own-real rate of interest should, other things being equal, make house purchase attractive. This could alternatively be thought of as expected real capital gains more than offsetting the real cost of borrowing (measured by mortgage interest rates less expected consumer price inflation). 6 Other parts of the economy were, however, facing quite different own-real interest rates. In particular, the nominal cost of borrowing for companies was still considerably higher than manufacturing output price inflation, which might hold back investment spending.

mpc: 3 7 More generally, since August the sterling yield curve had steepened, with rates in the two-to-five year range up by at least 50 basis points and the gap between fixed and variable mortgage rates having closed. This had increased the cost of borrowing for both companies and fixed rate mortgage borrowers. This would help to restrain demand in the medium term, and on one view was not fully reflected in the forecast. Others felt that the effects were likely to be small. (The Committee's discussion of whether the shift in market interest rate expectations had more general implications for policy setting is summarised below under `Tactical considerations'.) 8 The Committee agreed that final domestic demand could not continue to grow at the current rate without jeopardizing achievement of the inflation target in the medium term. Even without further monetary tightening some of the factors mentioned above were likely to restrain domestic demand growth. The Committee's latest projection incorporated a slowdown to around 3% by the end of the two-year forecast horizon, but that would probably still place pressures on supply capacity unless the contribution to GDP growth from stockbuilding and net trade was negative. 9 Stocks had been reduced in the first part of the year, consistent with reports of excess inventory holdings at the end of 1998. More recent reports suggested that there was little reason to expect a continuing negative contribution to GDP growth from this source in the short run. 10 The principal influences on net trade pointed in different directions. On the one hand, sterling continued to be strong, and was higher than in August. On the other hand, the world economy was recovering more quickly than expected. Exports of goods (excluding oil and erratics) to countries outside the EU had grown at an annual rate of 11% between Q2 and Q3; and total goods exports grew by 6% in the three months to August compared with the previous three months. Surveys suggested that manufacturers' expectations of export volume growth were higher than for around three years. 11 While it was possible that the strength of the recent exports data partly reflected temporary factors, it was also possible that exporters had to some extent adjusted to a level of sterling that they now expected to persist. If so, the equilibrium real exchange rate might have risen, but it was too early to be at all confident about that possibility. 12 To the extent that export strength were to persist, a negative contribution to GDP growth from net trade would have to come from a relatively faster increase in imports, so that the UK would in effect be relying on spare capacity abroad to contain incipient inflationary pressures from strong domestic

mpc: 4 demand growth. But while there was probably still spare capacity abroad, some members felt that it would be eroded as the world economy recovered, tending to put upward pressure on import prices. 13 On balance, the Committee had assumed that net trade would make a negative contribution to GDP throughout the forecast period, reflecting the further rise in the sterling exchange rate and a likely increase in imports.

The labour market

14 There was some evidence that the labour market had tightened further. Over the quarter, claimant count unemployment had fallen to 4.2%, and the Labour Force Survey (LFS) measure to 5.9%. Employment had continued to rise, as had total hours worked. Some forward-looking indicators suggested continued tightening. Employment intentions, as measured by surveys, were up in all sectors and were above average historical levels. Skill shortages persisted according to the Bank's regional Agents and surveys, but with differences across regions and different types of skill. The picture from surveys was mixed. The BCC measure of recruitment difficulties had fallen slightly in both manufacturing and services. The CBI survey showed skill shortages rising slightly but with the level still below its historical average. 15 Real earnings growth calculated on the basis of the Average Earnings Index (AEI) had been rising for some time, with the recent rate faster than would have been expected on the basis of the changes in unemployment and employment. As noted above, one possible explanation for this was that a greater-than-generally-expected decline in inflation had boosted ex post real earnings. Another possible explanation was that the `natural' rate of unemployment, the rate consistent with stable inflation, was higher than the Committee had assumed in its August projections. Some members thought it unlikely that the `natural' rate was below the current actual rate or that it was continuing to fall at the same rate as actual unemployment, so that there was good reason to think that labour market pressures had been increasing. Other members noted that the Reward Index showed earnings growth still trending lower and the New Earnings Survey also gave a more benign picture of earnings growth than the AEI. Although there were different interpretations of the latest data, the Committee agreed to take as its central projection a higher assumption for real earnings growth than in August. 16 The recent nominal pay data presented a second puzzle. On the one hand, settlements had remained around 3.5%. On the other hand, the AEI had risen at an annual rate of 4.9% in the three

mpc: 5 months to August. Allowing for month-to-month volatility, the AEI series had been edging upwards in recent months. If the difference was a good guide to wage drift, it was possible that the recent rise in nominal earnings would be consolidated in forthcoming settlements. A survey by Industrial Relations Services pointed in the opposite direction, however. In its latest survey, 43% of respondents had expected to make lower settlements than last year, 46% the same, and 11% higher. Last year, only 23% had expected lower settlements than the previous year, and 26% had expected higher. It was suggested that the apparent fall in expected settlements this year might reflect the low recent RPI outturns (1.1% in August and September) and expectations that RPIX would remain below the 2½% target for a while. Anecdotal evidence on this, based partly on Committee members' own regional visits, varied. There were suggestions both that RPI was the starting point in negotiations, and that the benchmark was the higher of RPI or RPIX. Some employers had suggested that the recent outturn of around 1% was not a feasible starting point; before allowing for productivity improvements, it was easier to argue either for zero if necessary on the basis of a particular company's circumstances, or alternatively for 2½% reflecting the inflation target and so medium-term inflation expectations. The Committee noted that there would not be much hard data on settlements until the new year, although the Bank's regional Agents would monitor, through their contacts, the emerging evidence. 17 Whatever the position, it was possible that settlements - both backward-looking data and forward-looking surveys - were now less reliable indicators of labour market conditions given the decline in collective bargaining, and the increase in personalised remuneration, over recent years. If so, relatively more weight needed to be placed on the path of broader measures of earnings. 18 The Committee as a whole agreed that its assessment of labour market conditions was highly uncertain. Structural changes and the new monetary regime meant that behaviour may have changed, but it was impossible to know by how much. Looking forward, the quantity data, particularly the LFS series, would be important indicators of whether there were any further tightening, and possibly of its speed. But what mattered was the extent of pressure on labour supply, for which there was no direct measure, and the extent to which this was translated into nominal earnings growth and settlements.

Prices and costs

19 The main issue concerning prices was the prospective effect on the aggregate price level, and thus on measured inflation in the short-to-medium term, of various expected developments. In particular, the Committee had made allowance in its Inflation Report projections for falls in price-cost margins

mpc: 6 reflecting an increase in competition; utility prices changes; and the Government's intention, announced in its pre-Budget Report published on 9 November, of removing the automatic over- indexation of duties on tobacco and fuel. Taken together, on the agreed assumptions these reduced the central projection for RPIX inflation by around half a percentage point at the two-year horizon (and more before that). 20 Some members were content with the assumed effects for the first year but preferred a smaller effect in the second year. This was for two reasons. First, they were concerned that the Committee had focussed on some one-off price falls, whereas there might be some one-off upward pressures as well; for example, it was conceivable that, as real incomes and wealth rose, consumers were switching from cheaper, standard goods to higher-margin, customised or designer products. Second, there were doubts about reflecting expected changes in relative prices in the aggregate nominal price level at longer horizons given that, by targetting inflation, the Committee influenced nominal, not real, demand growth over the medium term. While the particular downward pressures might plausibly be unanticipated in the near-term, they would more likely be anticipated as time passed. In that case, the effects of relative price changes might not show up as a decline in the overall rate of inflation over the medium run. 21 While agreeing that this analysis applied in the medium to long run, those members who supported the central assumption in the forecast argued that there was sufficiently clear evidence that the downward influences on prices would increase during at least the next two years. They agreed that it was difficult to calibrate the short-run impact of longer-term structural changes such as intensified competition. The Committee agreed that the size and persistence of the specific effects were inevitably uncertain, making it important to monitor the specific effects reflected in the published projections.

Money, credit and asset prices

22 M4 growth had continued to fall, the twelve-month rate reaching 2.8% in Q3, the slowest growth rate since 1963 (when the series began). But this reflected continued falls in the money holdings of non-bank financial institutions (OFCs). OFC money and credit were generally regarded as being most relevant to analysis of monetary conditions during periods of financial market instability. 23 Excluding OFCs, annual broad money growth had been 6.1% in Q3. Within this, household M4 growth was 6.4%. M0 and household Divisia growth were each around 7%. Growth in net credit to

mpc: 7 households was 8.6% in Q3, the highest rate since 1991. These data were broadly corroborative of the real-side evidence of continuing robust consumption growth. 24 The FT All Share Index had risen by about 2½% over the past month, and was at about the same level as in August. House prices continued to rise rapidly, with the twelve month rate on both the Halifax and Nationwide indices above 10%. As already noted, since the August Inflation Report the sterling yield curve had steepened out to medium maturities. 25 The sterling exchange rate index had risen by over half a percentage point since the Committee's October meeting and was about three percentage points above the path assumed in the August fan chart projections. The Committee agreed to incorporate in its published November projections an assumption that sterling would be half way between its current rate and a path related to market interest rate differentials, with the sensitivity of the projections to both shown in the Inflation Report. Some members inclined towards the former, and some preferred the latter.

The world economic outlook

26 The Committee noted the assumptions it had made in its latest projections. The outlook in each of the US, the euro area and Japan was somewhat better, and prospects among the emerging market economies had generally improved. The overall picture was therefore slightly stronger than assumed in the Committee's August projection, although there remained clear downside risks from the possibility of an equity price correction in the USA and persistent current account imbalances. 27 Against this background, Committee members agreed that the outlook for world prices looked less benign than at the time of the August Inflation Report. While there was still spare capacity in parts of the world economy, this was being utilised as a generalised recovery progressed. Imported price inflation had been weak (or negative) over the past few years, but had recently begun to rise.

The November inflation and output growth projections

28 The Committee agreed the projections to be published in the Inflation Report on 10 November. 29 On the assumption of constant official interest rates of 5.5%, the central projection in the published fan chart for activity was for slightly more rapid growth in the short term than in August, levelling off to slightly above trend. The level of activity was higher than projected in August for almost all of the

mpc: 8 forecast period on account of the recent stronger-than-expected outturns. The balance of risks was on the downside, mainly reflecting risks to the world economic outlook. 30 The Committee's best collective judgment of the prospects for inflation was for RPIX to remain below the 2½% target for nearly all of the forecast period, dropping just below 2% for a while, before increasing quite sharply to end up at around the target at the two-year horizon, when it was still rising. This acceleration in prices reflected the accumulating pressures on supply capacity from the period of above-trend demand growth and the unwinding of the various temporary downward influences on the price level and so on measured inflation. The trough was lower than in August. The projection at the end of the forecast period was almost the same as in August, but with a higher sterling exchange rate and with the official interest rate at 5.5% rather than 5.0%. For the same reasons as for the output projection, the balance of risks to inflation was on the downside. 31 As already described, there was considerable uncertainty in the Committee about the inflation outlook, and also a range of preferred assumptions for the path of the nominal exchange rate, earnings growth, and price-cost margins; these were presented in Table 6.B on page 58 of the Inflation Report published on 10 November. Different members preferred different combinations of these assumptions, either raising or lowering the inflation projection at the two-year horizon by ¼ to ½ percentage points, so that there was a range of approaching one percentage point between the lowest and highest preferred central projections of individual Committee members.

Tactical considerations

32 The Committee discussed three other issues potentially relevant to its policy decision: the significance of the steepening in the sterling yield curve; the relevance of ECB and FOMC policy decisions; and the implications of the shape of the inflation projection. 33 On one view, the rise over recent months in short and medium-maturity market interest rates, as evidenced by for example short sterling futures contracts and the gilt yield curve, meant that monetary conditions were materially tighter than reflected in the fan chart projections. In consequence, the market was doing some of the Committee's work for it, so that other things being equal the Committee could operate with a lower official repo rate than would otherwise be the case. Against this, it was argued that the Committee would take material risks with its credibility if it failed to tighten in line with market expectations, where it regarded those expectations as well based. If demand were

mpc: 9 restrained by higher market rates but the Committee then failed to raise rates, the market's view of the Committee's reaction function would change, storing up trouble for the future. 34 The Committee agreed that it should not place great weight on expectations of what the ECB and FOMC might do at their meetings on 4 November (the same day as the MPC's meeting) and 16 November respectively. It was noted that a clear majority of market participants expected the MPC to raise its rate by 25 basis points, in which case a move of that size would probably not have much effect on market prices. 35 The shape of the inflation projection had changed materially over the past year. A year ago, it had been humped; in May and August it had had a gentle saucer shape; now inflation was projected to fall further below the target, before rising more steeply towards the end of the forecast period. This reflected the view that accumulating inflationary pressures would surface after a period in which measured inflation was held back by a series of specific price level changes. In particular, after falling back in the short run earnings growth was likely to pick up further out in the forecast period. On the one hand, the degree of uncertainty about the outlook was considerably greater at the two year horizon, so that the Committee could be less sure about the projected steep rise than it could be about the nearer- term prospect of inflation remaining below target. On the other hand, monetary policy changes would have a greater effect on inflation at around two years than on nearer-term prospects. The Committee agreed that, other things being equal, the shape of the projection meant that, compared with a profile in which inflation rose steadily, there was more time to tighten policy in order further to restrain demand growth if that still proved necessary.

The immediate policy decision

36 The Committee agreed that the analysis of the outlook for inflation had changed over the past few months. The news on demand and activity had been stronger than expected, whereas RPIX inflation had been weaker than expected. The short-term relationship between output and prices had for some time been uncertain, and if anything that uncertainty had increased. For a long time strong domestically-generated inflationary pressures had been offset by strongly benign external influences: the price and net trade effects of the rising pound and of the shocks to world activity. Those external influences were now smaller. Sterling had recently risen slightly, but by less than during 1997 and 1998, and the world economy was recovering, with consequent rises in some input costs, for example oil. The key issues were now on the one hand the size and persistence of various downward influences

mpc: 10 on prices, and on the other hand the more-rapid-than-expected pick up in output growth and the extent to which domestic demand growth would slow without further monetary policy tightening. Members' analysis of the immediate policy choice varied according to how they viewed those factors. 37 A variety of arguments were identified in favour of an immediate increase in interest rates. Activity had recovered faster than expected, and confidence indicators had remained strong despite the September tightening of monetary policy and sterling's recent strength. A further decline in the velocity of circulation was needed for narrow money, household Divisia and household M4 growth to be consistent with the inflation target. Final domestic demand growth was projected to slow but still to be above trend at the forecast horizon, supported amongst other things by a strong housing market. A significant negative contribution to GDP growth from net trade was effectively being relied upon to avoid aggregate demand exceeding the economy's supply capacity. The labour market seemed again to be tightening, and there was consequently a risk that unit labour cost growth would not in fact moderate. AEI earnings growth had risen over recent months, and the Bank's regional Agents had reported greater concerns about skill shortages increasing wage growth. Although there was particular uncertainty about the relationship between output growth and inflation, members were influenced by the rise in wage pressures and the pick up in survey-based measures of pricing intentions. Input price pressures were gathering. The inflation projection was still rising at the forecast horizon so that, assuming no news in the meantime, the February projection for 2002 Q1 would be higher than the target. While RPIX inflation was likely to be contained for a while by the various downward pressures on prices, a further tightening of policy was needed to restrain the medium-term inflationary implications of continued strong domestic demand. The risks from tightening and not tightening were not symmetric. An immediate 25 basis point increase would not carry many risks to output as growth was above trend and the level of activity was already probably not much, if at all, below the trend level. But there would be risks in not increasing interest rates now. The news on activity had been significantly stronger than expected since the Committee tightened pre-emptively in September, so that not tightening again risked increasing inflation expectations, entailing a larger tightening later than would otherwise have been necessary. 38 As well as weighing the considerations set out in paragraph 37, those members of the Committee who favoured a higher central projection than that assumed - whether because of different preferred assumptions on the path of the exchange rate or price-cost margins, or on both - regarded the risks of overshooting the target as correspondingly greater. The choice for them now was between 25 and 50 basis points. While a case could be made for 50 basis points in order to counter the strength of

mpc: 11 underlying inflationary pressures, the following considerations led them to prefer a smaller rise. As the Committee had discussed, there were signs, for example in surveys, that there might be some slowing in demand growth. There was time to wait to gather evidence on this as inflation was projected to be below the 2½% target for nearly all of the forecast period. Separately, given market expectations, an increase of more than 25 basis points risked putting undue upward pressure on the exchange rate, which would tend to restrain net trade and import prices whereas policy action was needed to address domestically-generated inflationary pressures. 39 One member explored the arguments for an interest rate increase of less than 25 basis points. The profile for inflation implied that the pressures on the economy's supply capacity needed to be restrained in the medium term but that there was time to do this, particularly given the uncertainties about the outlook. An increase of less than 25 basis points would assist that process and would, in particular, provide a clear signal of the Committee's resolve in addressing those medium-term concerns, while recognising the downward pressures on inflation in the near term, when it was projected to drop materially below the 2½% target. It would be easier for the Committee to introduce smaller-than-25 basis point changes now having already tightened policy in September. However, as this would be the first time rates changed by less than 25 basis points, there was a significant risk that the reasons would be misunderstood. The better course was, therefore, to raise rates by 25 basis points. 40 Various arguments were advanced for maintaining the official interest rate at 5.25%. Inflation was below target and was set to remain so over the next two years. The outlook depended to a significant degree on two material factors: the exchange rate, and the balance of aggregate demand and the economy's supply capacity. Recent export outturns added weight to the view that sterling would be stronger than the path assumed as the best collective judgment; and survey measures showed capacity utilisation falling slightly, perhaps because investment during this upswing had been strong. There were significant downward pressures on prices, reflecting not only one-off regulatory measures but also an ongoing intensification of competition throughout the economy. This would help to restrain earnings growth below the path assumed as the best collective judgment. Monetary conditions, as reflected in the yield curve, had tightened significantly in recent months, and were probably bearing down on demand by rather more than had been assumed in the published fan charts. Taking these considerations together, the steep rise in the central projection of inflation 18 months hence was highly uncertain and there were clear downside risks to it. With a symmetric inflation target, a further tightening was not yet needed.

mpc: 12 41 The Governor invited members of the Committee to vote on the proposition that the Bank repo rate be increased by 25 basis points to 5.50%. Eight members of the Committee (the Governor, Mervyn King, David Clementi, Willem Buiter, Charles Goodhart, Ian Plenderleith, John Vickers and Sushil Wadhwani) voted for the proposition. DeAnne Julius voted against, preferring to maintain interest rates at 5.25%. 42 The following members of the Committee were present: Eddie George, GovernorMervyn King, Deputy Governor responsible for monetary policyDavid Clementi, Deputy Governor responsible for financial stabilityWillem BuiterCharles GoodhartDeAnne JuliusIan PlenderleithJohn VickersSushil Wadhwani 43 Gus O'Donnell was also present as the Treasury representative.

mpc:

ANNEX: SUMMARY OF DATA PRESENTED BY BANK STAFF

A1 This Annex summarises the analysis presented by Bank staff to the Monetary Policy Committee on 29 October, in advance of its meeting on 3-4 November 1999. At the start of the Committee meeting itself, members were made aware of information that had subsequently become available, and that information is included in this Annex.

I The international environment

A2 US GDP had risen by 1.2% in Q3, according to the provisional estimate. Consumption and investment had remained strong, while inventories had made a positive contribution to GDP growth, following a negative contribution in Q2. Net exports had continued to fall in Q3, albeit by less than in Q2. Definitional and statistical changes to the National Accounts had raised the average annual GDP growth rate by 0.4 percentage points between 1992 and 1998. The National Association of Purchasing Managers' (NAPM) Index had fallen slightly in September, but had remained at a level consistent with an expansion of manufacturing output. A3 The quarterly growth in the Employment Cost Index measure of labour costs had declined slightly in Q3. Annual producer price inflation had increased to 3.1% in September, and annual consumer price inflation to 2.6%. A4 GDP growth in the euro area had been revised up to 0.5% in Q2. Both consumption and investment growth had been revised up. Industrial production had risen by 2.5% in the year to August, reflecting improving business confidence, particularly in France. M3 had risen by 6.1% in the year to September. Private sector credit growth had been higher. Annual inflation had remained unchanged in September, at 1.2%. Excluding energy, food, alcohol and tobacco, inflation had continued to fall. A5 Despite a decline in September, industrial production in Japan had risen by 3.8% in Q3. However, the September Tankan survey indicated that companies were continuing to cut fixed investment. Volumes of imports and exports had continued to strengthen, particularly to other Asian countries. Consumption indicators had been mixed and retail sales growth had remained negative. Base money had risen by 6.1% in the year to September, but broad measures of money growth had been lower. Survey evidence had suggested that banks were becoming more willing to lend, but the stock of outstanding loans (adjusting for write-offs and securitisations) had continued to fall. The rate of

mpc: 2 unemployment had fallen to 4.6% in September, although this had partly reflected a fall in the participation rate. The consumer price index had fallen by 0.2% in the year to September. A6 Trade volumes had risen in the United States, Japan and the euro area. Non-oil commodity prices had remained stable. Oil prices had fallen very slightly, despite the most recently available evidence suggesting a higher degree of compliance among OPEC members with the agreement to limit supply. However, the continued strength in oil prices had contributed to an increase in petrol and producer prices in all countries. A7 There had been little movement in the main dollar, yen and euro exchange rates. Equity markets in the main economies had recovered slightly from falls earlier in the month. There had been a further increase in euro-area interest rates implied by Euribor futures contracts, with markets expecting around 50 basis points of tightening by next summer. A8 The trend in industrial production growth in most emerging market economies had been positive. It had remained negative in Latin America.

II Monetary and financial conditions

A9 Provisional figures had suggested that narrow money growth had rebounded in October. The one- month growth rate of notes and coin had risen to 0.8%, compared with an upwardly revised 0.1% in September. The twelve-month growth rate (after adjusting for the introduction of the new 50 pence and £2 coins) had risen to 7.6% in October from 7.0% in September. A10 M4 had fallen by £2.9 billion (0.4%) in September, with the annual growth rate falling back to 2.8%, the lowest since the series began in 1963. M4 lending (excluding the effects of securitisations) had risen by £4.2 billion (0.4%) in September. A11 Households' M4 deposits had risen by 1.3% in Q3. In real terms, households' M4 deposits were estimated to have increased by 4.8% in the twelve months to Q3. And households' real Divisia growth was estimated to have picked up to 5.3%, consistent with strong consumption growth. Households' M4 lending had grown strongly in Q3. Annual growth had been 8.6%, the highest since 1991. Within total lending to individuals, the annual growth rate of secured lending had been 7.5%, reflecting strength in the housing market. Loan approvals had also been strong. Total unsecured lending had

mpc: 3 increased by 3.3% in Q3, in line with its increases over the year. Credit card lending had continued to grow in excess of 20% per annum. A12 Private non-financial corporations' (PNFCs) deposits had remained broadly flat in Q3, though deposits had risen by £1.4 billion (1.2%) in September. The twelve-month growth rate had fallen from 7.4% in Q2 to 4.9% in Q3. PNFCs' M4 lending had picked up in Q3. Lending had been particularly strong in September, increasing by £1.8 billion (0.9%). But the annual growth rate had fallen from 4.7% in Q2 to 4.2% in Q3, compared with an average annual rate of 5.8% in 1998. A broader measure of PNFCs' borrowing (that also included all bond and equity issues as well as non-sterling lending by banks) had been weaker in Q3 than in Q2: the average monthly flow had increased by £3.3 billion compared with £5.9 billion in Q2. A13 Other financial corporations' (OFCs') M4 deposits had fallen sharply in Q3, and the annual growth rate had continued the downward trend since 1998 Q1. The average monthly flow had been -£2.7 billion in Q3, compared with -£1.3 billion in Q2. The monthly flow in September had been particularly weak (-£7.0 billion). OFCs' M4 lending had increased slightly in Q3, but the flow in September had been -£2.1 billion. A14 Short-term interest rate expectations implied by short sterling futures contracts had fallen since the previous MPC meeting. The falls had been between 20 and 30 basis points in the September 2002-March 2003 contracts (less at shorter maturities). Nominal forward rates at broadly comparable maturities had also decreased (by 40-50 basis points after 3 to 3.5 years). A15 Short sterling futures and nominal forward rates derived from the gilts market had pointed to different profiles for the expected level of interest rates. In particular, short sterling futures had pointed to a peak of 7.3% in interbank rates in December 2001, around 100 basis points greater than the peak of 6.3% for gilt yields in the forward curve in the second half of 2001. The three-month LIBOR/general collateral interest rate spread had picked up sharply at the beginning of October as the date for year-end LIBOR delivery had passed. But spreads further out along the curve had not looked unusual given recent experience. A16 There had also been falls in nominal yields at the long end of the forward curve during October. This had been mirrored in falls in long corporate bond yields. Since the beginning of the year, the yield

mpc: 4 on 25 year corporate bonds had also fallen relative to yields on shorter maturity bonds, reflecting movements in similar maturity gilt yields. And it was noticeable that there had been a sharp pick up in issuance of very long corporate bonds (with a maturity in excess of 15 years) over this period. So any technical factors that may have helped to explain the fall in relative gilt yields at the long end appeared to have reduced the funding costs for some firms. A17 The pass-through of September's base rate rise to retail savings rates had remained incomplete in October. Interest rates on fixed-rate mortgages had increased, following continued increases in swap rates at all maturities. But the narrowing in the spread of five-year fixed mortgage rates over five-year swap rates observed this year had continued in October. A18 Implied inflation expectations had fallen by around 15-30 basis points out to fifteen years. The fall had been smaller at longer maturities. The October biannual Consensus Economics survey had pointed to a fall in inflation expectations over the medium to long term (four years and further out) between April and October. By contrast, over the same period, inflation expectations implied using index-linked gilts had risen out to eight years, though they had fallen further out. A19 The sterling effective exchange rate index (ERI) had appreciated, by 0.7% over the month. Within this, sterling had appreciated by 1.4% against the euro, but had depreciated by 0.7% against the US dollar and by 3.0% against the yen. Real exchange rate expectations calculated using the biannual Consensus Economics survey of long-run inflation and nominal exchange rate expectations had suggested that the markets increasingly expected the strength of sterling to be sustained. The real exchange rate expected at end-2003 had increased by 2.8% between the October 1998 and 1999 surveys. The FTSE All-Share index had risen by 2.4% on the month. As in recent months the index for smaller capitalisation stocks had performed more poorly, falling by 0.8% on the month.

III Demand and output

A20 The preliminary ONS estimate of GDP growth in 1999 Q3 had been 0.9%, the strongest for two years. The annual growth rate had risen to 1.8%. Service sector output had grown by 1.0% in Q3 and had been 2.6% up on a year earlier - close to its long run average. Within services, the distribution, hotels and catering sector had grown by 0.9%. Manufacturing output had risen by 1.0% in Q3, following revisions to data in July and August.

mpc: 5 A21 Construction new orders had fallen by 5.0% in the three months to September but the CIPS index of construction activity had indicated strong growth since January 1999. A22 Retail sales volumes had risen by 0.1% in September. Growth in Q3 had been 1.2%. The GfK consumer confidence index had been flat in October, but remained above its historical average. The new quarterly Consumers' Association Consumer Trends Survey consumer confidence index had been +34 in October, compared with +33 in July. According to this survey, households expected house prices to rise by 4% over the next year. A23 Both the Halifax and Nationwide house price indices had shown rises in annual house price inflation in October, to 10.8% and 11.6% respectively. The Halifax index had grown particularly sharply in October, rising by 2.8%, following zero growth in September. Particulars delivered had fallen to 122,000 in September - the second consecutive monthly fall - but remained 7% higher than a year earlier. The Royal Institute of Chartered Surveyors (RICS) survey had shown a similar picture of housing transactions. Housing starts had risen by 2.9% in Q3. A24 Vehicle sales had been very volatile, given the change to the registration letter system. Private car registrations in Q3 had fallen by 16.6% on a year earlier. In the first 9 months of 1999, they had been 0.8% lower than the same period a year earlier. A25 The change in inventories had contributed negatively to GDP growth in Q2. The CIPS survey of manufacturing had shown that stocks of finished goods had risen in October for the first time since July 1998. The CBI Quarterly Industrial Trends Survey had shown that the balance of finished goods inventories had risen to -10 in October from -20 in July. A26 The BCC and CBI surveys had shown a pick-up of manufacturing investment intentions in Q3 compared to very low levels a year ago, though the CBI measure had fallen slightly to -11 in Q3 from -8 in Q2. The BCC survey had shown an increase in investment intentions (plant and machinery), to 11 in Q3 from 3 in Q2. But as investment intentions are forward-looking it seemed likely that actual investment would not strengthen until 2000. A27 The public sector net cash requirement had been £2.0 billion in September - with cumulative cash borrowing for 1999/2000 similar to the previous year.

mpc: 6 A28 The deficit on trade in goods had narrowed to £1.9 billion in August from £2.2 billion in July. The EU deficit had widened slightly to £0.4 billion, but the non-EU goods deficit had narrowed to £1.5 billion. The non-EU deficit had narrowed further to £1.2 billion in September. Excluding oil and erratics, export volumes had risen by 5.8% in the three months to August and imports had risen by 3.8%. Although export growth over the past half-year had been partly driven by erratic items, underlying export growth (excluding oil and erratics) had also been positive. Based on monthly data, the 1999 Q3 net trade contribution to GDP seemed likely to be positive. A29 Looking ahead to Q4, survey evidence had suggested continuing strong growth. The CIPS purchasing managers' survey of manufacturing showed a slight improvement in the headline index, to 54.2 in October, with the output index above its no-change point of 50 for the seventh consecutive month. The CBI Quarterly Industrial Trends survey had shown a sharp improvement of manufacturing sector confidence in Q3. The BCC survey for Q3 had shown a marked improvement in domestic and overseas orders, in both the manufacturing and services sectors. Domestic and overseas orders balances had appeared to be moving in line over recent quarters - though growth in domestic orders remained stronger. The CIPS Services Survey had shown an increase in activity in October (though the rate of growth had been the slowest since March). The CBI Distributive Trades Survey had shown a rise in retail sales in October (though with growth slowing, and sales considered to be slightly below average for the time of year). The improved outlook had also been broadly consistent with other surveys such as those by the Institute of Directors and the Engineers' Employers Federation. The latest surveys when taken together suggested quarterly growth in GDP continuing in Q4 at broadly the same rate as in Q3.

IV The labour market

A30 LFS employment had increased by 99,000 (0.4%) in the three months to August compared with the previous three months - up sharply on the increase in the three months to May. In full-time equivalent terms, growth had been the strongest since the summer of 1998. Total hours worked had risen by 0.4% during the three months to August over the previous three months. Average hours worked had risen by 0.1%, although they had remained weaker than in the same period the previous year.

mpc: 7 A31 According to the CIPS survey, manufacturing employment had stopped falling, employment growth in services had eased, while employment growth in the construction sector had been broadly stable. Looking ahead, surveys of employment intentions had all indicated further employment expansion. A32 Surveys of skill shortages and recruitment difficulties had provided a more mixed picture of the degree of labour market tightness. The CBI Industrial Trends Survey had reported that skill shortages had increased in the September quarter, but remained below their historical average. The BCC Quarterly Economic Survey had reported that recruitment difficulties had lessened over this period, but remained high by historical standards. The stock of unfilled vacancies in job centres had remained high in September. However, the National Press Recruitment Advertising Index had continued its downward trend. A33 Unemployment levels had fallen on both the LFS and claimant count measures. LFS unemployment had fallen by 83,000 to 5.9% in June-August compared with the previous three months. The claimant count had fallen by 5,400 in September, with the rate unchanged at 4.2%, the lowest since the early 1980s. But the participation rate had still not exceeded its peak in the early 1990s. A34 Inactivity had risen by 23,000 in June-August compared with the previous three months, with the inactivity rate unchanged at 21.2%. However, it remained about 30,000 below its level at the same time the previous year. A35 August's average earnings data had included returns from a new sample, implementing a recommendation of the Turnbull-King report following the suspension of the index in late 1998. Headline average earnings growth in August had risen by 0.3 percentage points to 4.9%. Stronger growth in private sector earnings had more than offset a fall in public sector earnings growth. Earnings growth had in recent months been affected by changes in the timing of bonus payments, though the latest month's figure had appeared to be free from such distortions. The implementation of the National Minimum Wage (NMW) may also have had an impact on recent AEI outturns. The New Earnings Survey had shown that gross weekly pay had grown by 3.6% in the year to April 1999, slightly lower than the AEI over the same period (4.0%).

mpc: 8 A36 Other evidence on earnings growth had been mixed. The Reward Index had risen by 0.1 percentage points in September to 3.4%, but it remained well below the AEI. The FRES survey in October had shown stronger earnings growth for temporary staff, but little change in salary growth for permanent staff. The Bank's regional Agents reported that overall pay pressures had remained subdued. There had been little new information on settlements in September, because relatively few settlements were made at this time of year. The Bank's AEI-weighted twelve-month mean settlement had remained at 3.5% in September. After allowing for inflation, settlements appeared to have stopped growing, but remained at a high level. A37 Wage drift, on the basis of these data, had increased in recent months, particularly in the service sector. Service sector wage drift was now around 1 percentage point above its average level since 1994. Drift in the production sector had, however, been more subdued. A38 The Bank's regional Agents had conducted a survey of their contacts to help assess the nature of the skill shortages that firms were facing. A sample of firms thought likely to be experiencing skill shortages had been selected. Of the 123 firms included in the survey, 84% had reported that shortages had worsened since last year, primarily in specialist areas. A similar proportion of firms expected skill shortages to persist both in specialist areas and more generally. Comments had included reference to poor general educational standards, as well as the legacy of the past run-down in vocational training, the erosion of work culture in some areas, and constraints on the mobility of labour. A39 Around 80% of firms sampled had expected to have to pay higher wages to obtain the skills they required, including more than two-thirds of the group for whom skills recruitment had become no worse in the past year. More than a quarter had said there would be an increased focus on training by management. Just under 10% had mentioned an impact on output, either via a constraint on output levels, or on the development of their business (both from their own skill shortages and shortages of IT services), or via poorer levels of service to customers. A number had also mentioned increased pressure on existing staff, leading to higher staff turnover.

V Prices

A40 The Bank's oil-inclusive commodity price index had risen by 2.1% in September and by 13.5% on a year earlier, mainly as a result of continued strengthening in crude oil prices. Excluding oil,

mpc: 9 commodity prices had been flat over the month and had been 3% higher compared with the previous year. The contrast between different types of commodities had continued: metal prices had risen further, whereas domestic food prices had fallen again. A41 Seasonally adjusted manufacturing input prices had risen by 0.7% in September to give 5.8% annual inflation. This had largely reflected higher oil prices. The latest CIPS survey input price index had risen to 56.2 in September, the highest for four years. The CBI Quarterly Industrial Trends output price expectations balance had strengthened further in Q3. Although cost pressures from materials had been mounting since early this year, output price inflation had remained subdued. A42 The latest CIPS services survey had shown a sharp rise in input prices in October, with the index rising to 59.2, its highest since May 1997. The survey had also indicated that average prices charged had risen for the first time in four months. A43 Export and import prices had risen by 1% and 0.6 % respectively in the three months to August. Stripping out the oil component, export prices had fallen and import prices had remained unchanged. A44 Comparing price inflation with weighted costs had suggested that retailers' and manufacturers' margins had fallen over the past year. The Euler Trade Indemnity Survey had indicated that competitive pressures had somewhat reduced respondents' profitability since early 1999. A45 RPIX inflation had remained at 2.1% in September. Whereas harvest-related seasonal food prices had risen on the month, non-seasonal food prices had fallen. The annual rate of change of the retail sales deflator had remained at -0.5% in September. This had mainly reflected falls in food and non-food store prices earlier in the year.

VI Reports by the Bank's Agents

A46 The Bank's regional Agents had reported a continued recovery in economic activity. Growth had continued to be driven mainly by domestic demand, although many firms had also reported a pick-up in export orders. However, although manufacturing output continued to rise, the recovery had been far from universal. Construction activity had remained strong. The service sector continued to record strong growth, although it had not increased further. IT activity had remained strong, as Y2K work had

mpc: 10 been replaced by firms investing in more developmental projects, such as Internet sites. Agricultural production had remained weak in all regions. A47 Retail spending had improved further, although growth had been moderate. The pattern of consumer spending had remained similar to previous months, with department store sales weaker than in discount and out-of-town stores. A48 There had been evidence of further tightening in the labour market. Most regions had reportedincreased skill shortages, with particular concern reported in the south and south-east of England. Many contacts had noted that unskilled labour had also become more difficult to recruit, especially for retail positions. Pay pressures had remained subdued, with most settlements lower than a year ago. However, pressures had appeared slightly stronger than in previous months, especially in southern regions. A49 Input prices had picked up further, particularly for oil-related products and some metals, although overall prices remained broadly flat. Many manufacturers had continued to find it difficult to raise prices, and retail prices remained flat or falling. Furthermore, a number of firms thought that consumers had become more sensitive to price differences within each sector, and had begun to bargain over price to an increasing extent. A50 Growth in house prices had remained more pronounced in the southern regions, although there had been evidence that growth here had peaked. The buy-to-let market had slowed. In the northern regions of the United Kingdom, growth in housing activity had continued to increase steadily.

VII Market intelligence

A51 The effective exchange rate index for sterling had been little changed since the October MPC meeting, but sterling had appreciated against the euro from the middle of the month. This strengthening had coincided with the announcement of merger and acquisition activity. It was possible that anticipation of potential sterling purchases had encouraged others to bring forward their purchases of sterling. Such influences on sterling's exchange rate were expected to be temporary.

mpc: 11 A52 Market participants had continued to comment that movements in US equities were an influence on the dollar's exchange rate against sterling and other currencies. There had been some evidence of a greater correlation between the Dow Jones Industrial Average and dollar exchange rates, including that between the dollar and sterling. However, the wider implications for sterling's effective exchange rate had been less clear. A53 The difference between interest rates implied by futures and economists' interest rate forecasts had widened over the three months to October. This was more likely to be related to increased risk- aversion, the unwinding of long-standing trading positions and mortgage-related swap market activity, rather than a genuine difference of view between traders and economists. In terms of the immediate rate decision, market commentators on balance attached a high probability to a 25 basis point increase in the Bank's repo rate but did not expect changes in rates in December or January.

mpc: