PRESENT: 
Lars Heikensten, Chairman
Eva Srejber
Villy Bergström
Lars Nyberg
Kristina Persson
Irma Rosenberg

Kerstin Alm
Claes Berg
Jörgen Eklund
Mithra Ericson
Mats Galvenius
Kurt Gustavsson
Kerstin Hallsten
Per Håkansson
Jyry Hokkanen
Pernilla Meyersson (§ 1)
Arvid Wallgren
Anders Vredin 

§ 1. The current inflation assessment

It was noted that Kurt Gustafsson and Arvid Wallgren would prepare draft minutes of §1 and 2 on the agenda for the meeting.

The discussion was based on the new data and analyses presented by the Monetary Policy Department. These analyses were based in turn on the assumption that the repo rate would remain unchanged at 2.0 per cent up to and including 2006 Q3.

1. Recent data regarding economic developments

The Monetary Policy Department reported on new data received since the final date for inclusion in the Inflation Report, 4 October.

Since that date the oil price had risen a further USD 4-5 per barrel. In the futures market, too, prices of crude oil for delivery up to 2 years ahead had increased by roughly the same. Some of the factors that had affected the recent price rise were a strike by Nigerian oil workers, threats of a conflict in the Norwegian oil fields and new assessments of how long it would take to restore production capacity in the Mexican Gulf after the hurricanes in the autumn. Since the beginning of the year the oil price had risen by around 60 per cent.

US jobs data, and to some extent relatively weak European statistics, had been the main factor influencing pricing in the financial markets since 4 October. In the United States, 96,000 new jobs were created in September compared with the previous month. Market participants had been expecting a rise of around 150,000. In Germany and France, industrial production fell short of market expectations.

In the United Kingdom the CPI outcome for September was lower than the consensus expectation. In addition, the European Central Bank (ECB) had decided to keep the refi rate unchanged. In explaining its decision, the ECB mentioned that there was uncertainty about economic developments owing to the oil price situation.

The yield on 10-year US government bonds fell by around 10 basis points following the release of jobs data for September. European yields, too, had dropped by roughly the same. Monetary policy expectations in the United States and the euro area had shifted somewhat towards later or smaller interest rate hikes. Swedish long-term interest rates had also fallen slightly since 4 October. The krona’s TCW-weighted exchange rate had strengthened somewhat.

Some domestic statistics had also been published since the start of October. Both new orders in manufacturing and industrial production had declined somewhat in the period June-August compared with the previous three-month period. Statistics Sweden’s activity index for August was down marginally on the month before. The National Labour Market Board's data for September gave a somewhat mixed picture, including fewer redundancy notices but also fewer job vacancies than the previous month. The extent of labour market policy measures had increased sharply. House prices rose by 3 per cent between the second and third quarter. This was a slight softening compared with the change between the first and second quarter. Consumer prices in September increased by 0.7 per cent on the month before and by 0.6 per cent year-on-year. In addition, underlying inflation (UND1X) rose by 0.8 per cent on the previous month and by 1.0 per cent compared with September last year.

2. The Executive Board discussion of economic prospects and the inflation outlook

Deputy Governor Irma Rosenberg presented the Inflation Report 2004:3 (Annex A to the minutes). The Inflation Report reflected the main features of the presentations and discussions at the Executive Board meetings on 29 September and 7 October.

Over the summer, the Riksbank’s assessments of international economic growth and Swedish export market growth had gradually become somewhat more optimistic. Although the high oil price was subduing the upturn somewhat in, for instance, the United States and the euro area, this was counterbalanced by the fact that growth in other countries was expected to be stronger than forecast in May.

The oil price had risen much higher than expected in the previous Inflation Report. As before, it was expected to fall back somewhat, but it was estimated to be higher over the coming years than forecast in May.

The forecast for GDP growth in Sweden had been revised up gradually at the monetary policy meetings since May. New data indicated unexpectedly strong exports so far this year, owing to high international economic growth. Consequently, export growth was expected to be stronger this year and next year. The proposals in the Government’s Budget Bill pointed to a more expansionary fiscal stance in the period ahead. This was expected to give rise to higher domestic demand in 2005 and 2006. The Riksbank now forecasted GDP growth to be 3.6 per cent this year and 3.2 per cent in the following two years.

Both employment and unemployment had developed largely in line with the assessment in May, despite the more robust GDP growth. Firms had boosted production without increasing staff numbers. Among other things, this reflected the fact that productivity growth had again become stronger than expected. The forecast for productivity had therefore been revised up somewhat for this year and the coming two years.

Inflation was currently low. Next year, inflation was expected to increase as a result of rising international and domestic cost pressures and higher capacity utilisation. Compared with the assessment in May, the forecast for inflation had been revised up slightly, owing to the upward revisions for the oil price and GDP growth. However, these inflationary impulses were counteracted somewhat by the relatively large amount of unutilised resources initially and by the fact that productivity was expected to be higher.

This time, the inflation risks were judged to be balanced around the main scenario. There was a risk that the oil price would be higher and that the effects of the high oil price on inflation would be greater than assumed in the main scenario. There were also risks of lower inflation and weaker economic activity. These risks stemmed partly from the large deficits in the US federal budget and current account and from substantial current account surpluses in Asia. At the same time, there was still uncertainty over productivity growth in the Swedish economy.

Taking into account the risk outlook, CPI inflation was expected to be 1.6 per cent one year ahead and 2.5 per cent two years ahead. The corresponding forecasts for UND1X inflation were 1.4 per cent and 2.0 per cent, respectively.

One member observed that indicators of economic developments abroad since the Inflation Report had, if anything, tended to be somewhat negative. Nevertheless, the assessment in the Inflation Report was unchanged. Rather, it was a question of slight differences in the risk assessment.

Another member noted that the assessment of developments in underlying inflation had been revised only marginally despite the fact that the assessments of economic growth and oil prices had changed markedly since the previous monetary policy meeting.

The first member pointed to several different explanations for this. So far this year productivity had been more robust than expected, and the forecast for coming years had also been revised up somewhat. In addition, resource utilisation was low. Even though oil prices so far this year had been markedly higher than anticipated and economic growth stronger, both employment and inflation had developed in line with the May forecast.

According to another member, not much had happened since the Inflation Report to warrant any change in the inflation assessment.

Another member noted that the only changes since 4 October were signs of marginally weaker international developments. In addition, it would be a while before inflation pressures accumulated in the Swedish economy. Even though growth had been high for some time, total unemployment was still around 7.5 per cent. The member saw no reason not to adopt the Inflation Report.

One member summarised the discussion and concluded that some recent information was pointing to slightly weaker real developments. Other members had noted some uncertainty about productivity. Nevertheless, all Executive Board members supported the assessment in the Report.

The Executive Board’s overall assessment was that the new information received since the final date for inclusion in the Inflation Report, 4 October, did not give reason for any change in the inflation outlook.

The Executive Board decided

• to adopt the Inflation Report as presented and
  
• to publish the Inflation Report on 14 October 2004 at 8 a.m.

This paragraph was confirmed immediately.

§ 2. Monetary policy discussion

1. Account of the monetary policy group’s view of the monetary policy situation (1)

Deputy Governor Irma Rosenberg presented the monetary policy group’s view of the monetary policy situation. According to the monetary policy group, inflation was currently being affected by a number of factors that were counteracting each other. High growth and increasing resource utilisation would contribute to a rise in inflation in the period ahead. But productivity growth was robust, holding down domestic inflation, while imported inflation was low despite the oil price rise. Compared with May, the inflation forecast had been revised up. Given that there was a relatively good deal of idle resources and that cost conditions were favourable, however, inflation was still forecast to be in line with the target a couple of years ahead. Against that background, the group’s view was that the repo rate should be left unchanged. Were economic activity to continue to strengthen, however, there was reason to expect gradually rising inflationary pressures. This suggested that the repo rate would eventually need to be raised.

As regards the impact of the increasing oil prices, it was judged that they would lead to only a limited rise in inflation and to fairly small effects on the real economy. Were oil prices to continue to increase, however, the situation could give rise to more delicate monetary policy judgements in the period ahead.

2. The Executive Board discussion

The Executive Board members agreed broadly with the monetary policy group’s view of the monetary policy situation.

One member argued that there was no need to adjust the repo rate at this meeting. Economic growth was indeed robust, but employment had not risen to the same degree and productivity growth had been somewhat stronger. Thus, it was likely that the economy was somewhat further from its capacity ceiling than the Riksbank had previously thought. However, the economy was approaching the point at which the vigorous growth should become evident in employment data and in capacity utilisation, thereby putting upward pressure on prices. Furthermore, oil prices, which hadn’t fallen as much as previously expected by the Riksbank, could feed through to inflation.

The member pointed to a number of uncertainty factors, including the strains that were building up because a number of Asian economies were to a large extent financing the US current account deficit by buying US Treasuries. This was not sustainable in the long run but instead would lead to a dollar adjustment, and hopefully also to an increase in US saving. However, there were currently no indications that this adjustment would have to be so abrupt that it would derail the cyclical upswing. The member also noted that there now were signs that the increase in house prices in Sweden was on the point of levelling off, which was positive since it reduced the risk of a price fall under disorderly conditions.

Another member pointed out that it was important not to change the interest rate level too late. This could result in a less smooth adjustment of interest rates later on. But the high growth and the high oil price was likely to result in increased price pressures. Thus, on the one hand, the repo rate could be raised on this occasion. On the other hand, cost pressures were low and there was therefore scope to wait and see.

The economic upswing would have an impact on inflation eventually, according to the next member as well. But three factors suggested that the inflationary pressures could be weaker than what had been usual in previous expansionary phases. Firstly, productivity growth had been strong, and the trend productivity growth rate appeared to have risen somewhat. This was likely an effect of the ongoing IT revolution that had spread to more and more sectors in the economy, and had become increasingly deeper in each sector. Secondly, globalisation was leading to an increased international division of labour and tougher competition. The dynamics of economic developments were increasing, and the costs falling. Finally, the Swedish labour market was becoming increasingly flexible, as reflected partly in larger variation in the number of hours worked per employee and in a greater frequency of project-based employment. Against this background, the member did not believe that there was any reason to hurry in raising interest rates. There were also clear risks in the shape of the large, growing deficits in the US current account and federal budget, which sooner or later would have to be corrected. The deficits were likely to be given greater political attention after the US presidential election in November. Were no measures to be taken now, the deficits could lead to dramatic effects later on.

According to the same member, however, it would be more desirable for Sweden to have less expansionary fiscal policy in the current economic climate. There was potential in the private service sector to boost employment. For this reason, the first step should be to remove obstacles in order to bring about such a development instead of stimulating employment in the public sector.

The next member agreed with the conclusions of the monetary policy group, but pointed out two things. Firstly, if the repo rate remained unchanged, monetary policy would be looser now than in last spring. The reason was that the Riksbank had revised up its forecast for productivity growth over the forecast horizon, which implied slightly higher potential growth than previously estimated. This could mean that it would take a bit longer before the economy’s resources became strained. That was only one side of the issue, however. Higher potential growth meant in turn that the neutral long-term repo rate was somewhat higher than before. Consequently, in order to maintain the same degree of expansion in monetary policy as last spring, the repo rate would have to be raised somewhat. In other words, keeping the repo rate unchanged was from this perspective synonymous with looser monetary policy, according to the member. Moreover, fiscal policy in the period ahead would be more expansionary than assumed in the last Inflation Report.

Secondly, continued the same member, the Riksbank had kept a low nominal and real key interest rate in Sweden so as to bolster demand during the economic slowdown and thus prevent inflation from being unnecessarily low. Demand in the economy had been dampened by firms’ problems after the IT boom and the equity price bubble, which resulted in a need to reduce costs and consolidate balance sheets. Low interest rates had facilitated firms' debt consolidation and made it cheaper for households to borrow, as reflected in the rate of increase in their borrowing and their debt levels. Cheaper loans boosted demand for housing and consumer durables, and sustaining consumption was also the intention of the Riksbank’s interest rate policy in the situation the Swedish economy had been in. A sluggish supply of housing caused residential property prices to rise when demand increased, which in turn meant that household consumption could rise further since households could increase their mortgages when their property rose in value. This transmission of interest rate policy was certainly not unique to Sweden. Exactly how it worked and what effects it had depended on institutional differences and in some countries it had resulted in clearly overvalued house prices. In Sweden, however, it was less evident.

According to the member, the corporate sector was now in a better situation thanks to firms’ stronger balance sheets. Some data indicated that listed companies as a group were more or less debt-free. Investment had been picking up for some time. This meant that the time was approaching to begin a normalisation of the key interest rate since there no longer was the same need as before to sustain household demand. It was possible that households would respond in the same way to interest rate hikes as to cuts, but there was a risk that they would react asymmetrically. In other words, movements in housing prices and households’ adjustment of their debt burden could occur faster when interest rates were raised than when they were lowered. In such a case it could give rise to undesired fluctuations in the real economy. The risk of asymmetry could increase when interest rates had been very low for a long time and housing prices and household debt had risen above a certain equilibrium level. Consequently, the danger of keeping interest rates very low for a long time suggested that there was reason not to wait too long before beginning the return towards more normal interest rate levels, concluded the member.

Another member said that the first line of argument was primarily long term in nature. The aspects that were relevant to, for example, productivity growth should already be taken into account in the current inflation assessment. The member also pointed out that the empirical knowledge of asymmetric reaction patterns in Sweden was limited.

As regards households’ reactions, one member commented that in the United States it was high-income earners - those households with the 20 per cent highest incomes – that had reduced their saving sharply. Four-fifths of US households had more or less maintained their saving ratio. It would be interesting to know what the corresponding figures were for Sweden. The member thought that the day’s discussion was slightly pessimistic given that total world growth according to the IMF was at its strongest for 30 years. The only substantial disappointment was that US growth may be somewhat weaker than expected last spring. Were resource utilisation not to pick up under such circumstances, it would be unique. Inflationary pressures would rise sooner or later; anything else was unlikely.

There was no information about what households were using their borrowing for, observed the next member. However, lending had risen concurrently with housing prices. At the same time, the housing market had undergone a structural change in recent years insofar as rental apartments had been converted into tenant-owned apartments, and this had also contributed to the rise in households' borrowing. Against this background, it was reasonable to believe that the increase in borrowing had been used to a large extent to finance housing purchases.

As regards cyclical inflationary pressures, the same member said that the investment upswing had only just begun and that the economy therefore was not that far into the investment cycle. Neither were employers having any difficulty in meeting their labour requirements. It was reasonable to believe that the demand for labour would rise in the period ahead, but it would not lead to inflationary impulses in the near term since resource utilisation in the labour market was too low at present.

One member summed up the discussion. The Swedish and international economy were in the midst of a cyclical upswing. Were the upswing to continue, it would be strange if it didn't eventually result in increased employment and mounting inflationary pressures. This in turn led to the conclusion that the key interest rate would need to be raised, which was also what many were expecting. When this would happen was more difficult to determine. There were signs that competition was generally tougher in the Swedish economy today than before. Productivity growth was also still robust and had contributed so far to curbing inflationary pressures. In addition, imported inflation had been low, also including the oil price.

So, to sum up, there were currently few arguments in favour of adjusting the repo rate; inflation was in line with the target. Exactly when a hike may be appropriate depended as usual on the Executive Board’s assessment of the inflation outlook.

§ 3. Monetary policy decision (2)

The Chairman summarised the monetary policy discussion under § 2 and noted that the members of the Executive Board were in agreement that UND1X inflation on this occasion provided the best picture of underlying inflationary pressures. The Chairman found that there was only one proposal: to keep the repo rate unchanged at 2.0 per cent.

The Executive Board decided

• that the repo rate would be left unchanged at 2.00 per cent and that this decision would apply from Wednesday, 20 October 2004, 
  
• that the lending rate would remain unchanged at 2.75 per cent and that the deposit rate would remain unchanged at 1.25 per cent with effect from Wednesday, 20 October 2004, and
  
• that the decision would be announced at 8.00 a.m. on Thursday, 14 October 2004, with the motivation and wording contained in Press Release no. 57, 2004.
  This paragraph was confirmed immediately.

Minutes by:
Kerstin Alm

Checked by:
Lars Heikensten
Eva Srejber
Villy Bergström
Lars Nyberg
Kristina Persson
Irma Rosenberg

Notes:
1) The monetary policy group is made up of Riksbank staff and is headed by Deputy Governor Irma Rosenberg. The main features of the group’s discussion are presented at the Board meeting. The opinions expressed in the minutes are not necessarily shared by all members of the group.

2) Board members who are present and do not enter a reservation have participated in and agreed to the Board's decision.