Heikensten: Introduction on monetary policy
The whole speech
Let me begin as usual by thanking you for the opportunity to come to the Riksdag Committee on Finance and give an account of monetary policy. The committee's public hearing of the governor is an important part of the transparency the Riksbank wishes to uphold. These meetings illustrate the clear link between the bank and the democratically-elected body in Sweden. Now these meetings are a part of our legislation.
As is usual on these occasions, I intend to begin by commenting on our current assessment of inflation and our view of monetary policy. Then I shall briefly discuss events in the Swedish economy last year and the results of the monetary policy we conducted.
The current assessment
World economy recovering
As Sweden is a small country, very dependent on events in the outside world, it is natural to begin a discussion of the current economic situation with a description of developments in the world economy. Global economic growth has largely followed the path anticipated by the Riksbank and most other analysts during the autumn and winter. Compared with the picture we envisaged in December, we are now assuming that international growth will be slightly stronger. This is mainly due to a slightly more positive assessment of developments in the United States and Asia, but also to more favourable developments in Latin America, and central and eastern Europe. On the other hand, developments in the euro area, particularly in the larger countries - Germany, France and Italy - have been more subdued than expected and a slightly weaker development is foreseen in the near future than was previously assessed.
Although international demand has developed largely as we expected, the same cannot be said for prices. The subdued international price trend in recent years has probably been linked to the earlier economic slowdown. However, there are also other explanatory factors. One possible cause of the weak price pressure is the globalisation of trade and production patterns that has followed in the wake of deregulation and the spread of information technology. Many new markets are being integrated into the global economy, including countries in central and eastern Europe and Asia.
We are now seeing weak international price trends for manufactured goods, while there is a sharp rise in commodity prices. This could also be at least partly due to changed international patterns in production and trade. The best example of this is China, with its huge size and rapid industrialisation. The country is now one of the world's largest commodity importers, particularly crude oil, copper and certain agricultural products. China's rapid industrial expansion, with an average annual growth rate of 8 per cent, has contributed to a sharp rise in, for instance, steel and cement prices. Meanwhile, China is literally churning out cheap manufactured goods, produced mainly by labour with a low level of education and very low wages.
To summarise, we assume in the Inflation report that international growth will remain good. Growth in the OECD area is expected to amount to around 3 per cent a year over the coming two years. International price pressure on manufactured goods is expected to rise in future as economic activity strengthens, although not as rapidly as historical patterns indicate. On the other hand, oil price forecasts have been revised upwards for the entire forecast period, mainly as a result of increased demand. Nevertheless, international price pressure is expected to be slightly lower during the entire forecast period than assumed in earlier assessments.
Swedish economic picture remains the same
Economic activity in Sweden during the autumn and winter has largely followed the path we anticipated. The downturn that we now appear to have left behind us has been relatively mild, with an average growth rate during the past three years of 1.5 per cent. Important reasons why economic activity did not weaken more than this are the expansionary fiscal policy pursued and low interest rates, which contributed to upholding household consumption. At the same time, the relatively weak krona has contributed to favourable developments in exports.
We can now see that the economic upswing is continuing to strengthen gradually. Industrial activity has made some progress; export demand is accelerating and firms are relatively optimistic about the future. Capacity utilisation is also continuing to increase. Households' expectations regarding their own finances and the Swedish economy are once again more optimistic, after a decline at the end of last year. Retail trade sales indicate a good level of consumption. A slightly stronger development in real disposable incomes, partly due to lower inflation, continuing low interest rates and rising asset prices, is expected to benefit consumption in the near future.
Given this background, we have found no reason to make more than a few marginal changes to the overall picture of growth we have been outlining since last autumn. We now believe that economic growth will amount to just over 2½ per cent on average over the coming two years, which is a tenth of a percentage point more than our previous estimates. However, the annual outcomes differ slightly more because of the fact that the figures for the current and coming years are affected by changes between the years in the number of working days. This has no major significance for our view of resource utilisation, as both production and demand are affected. The conclusion of our assessment of economic activity is that the economy will approach full resource utilisation at the end of the forecast period.
Weaker price and cost pressure
Although we have not changed our views of developments in economic activity, during the autumn and at the beginning of this year we have gradually changed our view of inflationary pressure in the Swedish economy. This is connected with the labour market showing a weaker development than was forecast in December. Another factor is that productivity has increased more rapidly than expected, resulting in lower domestic cost pressure. A further factor is that the recent decline in inflation appears to have had some effect on inflation expectations, which have fallen slightly. Some of these are factors we pointed to already in December. At our meeting in early February we were able to base our decision on statistics from a further couple of months which reinforced the picture. Inflation was expected to be below target level and we therefore cut the repo rate by 0.25 percentage points. Let me go into these factors more closely and comment on them in the light of events since February. This is because it is essentially these factors that lie behind the decision we made yesterday.
During the latter part of 2003 unemployment rose relatively sharply. A number of factors contributed to this. Demand for labour declined both in the business sector and in the public sector, and there were cutbacks in labour market policy measures. As yet, there are few signs of a turnaround in the labour market and business tendency surveys indicate that the upturn in the private sector will be relatively weak this year. However, some improvement may be on the way, although it is difficult to determine exactly when the trend will turn. Recently, the number of new job vacancies registered has increased somewhat and the number of redundancy notices has declined. At the same time, the strained local government finances indicate a weak upturn in employment in this sector in the near future.
The situation in the labour market has caused us to revise down our assessments of future wage cost increases. This was also included in the picture in February. The central wage agreements signed in the manufacturing and retail trade sectors have confirmed this development. Our new forecasts indicate wage cost increases of 3.5 per cent over the coming two years, which is slightly lower than in December.
Last year there were increasing signs of more rapid productivity growth. While employment, measured as the number of hours worked in the business sector, developed weaker than anticipated, growth was largely as expected. This is reflected in more rapid productivity growth. We drew attention to this picture already in December. Since then, these developments have become clearer and been reinforced.
The stronger productivity growth is probably connected with economic activity; at the beginning of a recovery the available production resources are used to a greater degree and it is only when the recovery is secure that investment and new recruitment accelerate in earnest. However, there are numerous indications that productivity has risen more than is usual for the economic situation, a pattern that has been even more evident in the United States. It is conceivable that this development is linked to the factors I mentioned earlier; globalisation, deregulation and increased competition. However, more lasting changes in productivity may also be linked to the large investments in IT and telecommunications at the end of the 1990s. It is possible that the full impact from the investments made and technological advances is first achieved in connection with cost-cutting and rationalisation. We have a special boxed article on this subject in the report, which I can recommend reading. It is not possible to determine with any certainty what has caused the improvement in productivity. However, given this more favourable picture, we have nevertheless anticipated a rather higher level of productivity growth in the near future.
The new inflation figures received also indicate a slightly more subdued inflationary pressure than we had estimated. The outcome shows not only that electricity prices have fallen back as expected, but also that prices have increased at a slower rate in general. It is thus not only import prices, but also domestic price pressure that has increased more slowly. This could be connected with the higher productivity I mentioned. In this case, it would mean that the economy could grow at a slightly faster rate without inflation accelerating. Particularly as inflation expectations have also declined slightly.
Inflation picture and monetary policy
When we look ahead somewhat, we can conclude that the temporary fluctuations in energy prices, which have had a strong effect on inflation since the end of 2002 and pushed it down significantly in recent months, are expected to have abated by mid-2005. As Swedish and international economic activity improve, price pressure is also expected to increase. However, in the main scenario inflation is expected to be below the Riksbank's target both one and two years ahead, mainly due to lower wage cost increases and improved productivity growth. During 2004 UND1X inflation is expected to amount to an average of 0.6 per cent, which is 0.5 percentage points lower than the December forecast. The corresponding figure for 2005 is 1.2 per cent, which entails a revision of 0.4 percentage points for the year. At the end of the two-year period that comprises the main focus of the forecasts, inflation according to UND1X is expected to be 1.6 per cent, on the assumption of an unchanged repo rate.
So what are the risks that might upset the picture I have just presented? We make it clear in the Inflation Report that we consider the risks of a poorer development in Swedish and international economic activity to have declined as the recovery has progressed. Experience also shows that it is easy to underestimate the force of an upturn once it has found a foothold. Developments in the labour market in both Sweden and the United States are one factor pointing in the other direction. In addition, developments in the euro area remain sluggish and in the United States there are problems with imbalances in foreign trade in particular. Our opinion is nevertheless that, compared with the picture we present in the main scenario, there is an upside risk for inflation from economic activity in Sweden and abroad, with links to price trends.
There are arguments that we could have either overestimated or underestimated cost pressure. Compared with the December assessment, we have revised up our basic view of productivity, although there is considerable uncertainty in this area, given the surprising developments over the past six months. On the other hand, the uncertainty regarding wages has diminished somewhat, as central wage agreements have now been concluded for a large part of the labour market. All in all, our assessment is that the risks in this field are balanced.
The conclusion of the discussion on risks in the Inflation Report, and which I have only briefly touched on here, is that we saw reason to make an upward adjustment in the main scenario inflation forecast of around a tenth of a percentage point. This does not alter the conclusion that inflation in terms of both CPI and UND1X will be below the target level during the coming two years. During the first year we believe that inflation will be significantly below two per cent. We have also assessed that inflation will be marginally below the target at the end of the two-year period. This is the background to the decision we made yesterday to cut the repo rate by 0.5 percentage points.
Inflation in line with target in 2003
That covers the discussion in the Inflation Report and our interest rate decision. Let me now move on to discuss monetary policy from a retrospective view. We have as usual in connection with the publication of the first Inflation Report of the year produced a background report to facilitate the assessment of monetary policy over the past three years. This report is included in a special box in the report.
Measured as an annual average, the inflation rate was very close to the 2 per cent target in 2003. The CPI increased by exactly 2 per cent, while UND1X inflation, which is adjusted for the direct effects of indirect taxes and interest rates, amounted to 2.3 percent. However, these averages hide relatively strong fluctuations over the course of the year.
At the beginning of 2003 inflation rose rapidly and during a couple of months it actually exceeded the upper tolerance limit of 3 per cent. However, it had fallen back by May and was once again in line with the 2 per cent target. The downturn continued throughout the remainder of the year, particularly for CPI inflation. The fact that CPI inflation fell more than UND1X inflation was due to the interest rate cuts made earlier in the year. In December, CPI inflation was 1.4 per cent, while UND1X was 1.7 per cent.
The upturn at the beginning of the year was mainly due to rising electricity prices caused by supply problems in the hydroelectric power companies' reservoirs and by unusually high demand resulting from a cold winter. Oil prices also rose and contributed to the higher inflation rate. The high energy prices had an impact on several sub-components of the CPI. For instance, housing costs and various prices linked to transportation activities rose unusually quickly during the spring. When adjusted for energy prices, the path of inflation in 2003 was much more stable and remained within the interval of 1.2-1.9 per cent.
Small deviations from target
Despite the fact that inflation fluctuated substantially over the year, we can conclude that the average CPI deviation from target level was relatively small in an historical perspective; around half a percentage point, compared with approximately one percentage point during the period 1995-2002. In a comparison with a number of other inflation-targeting countries, it is evident that Swedish inflation was closest to the target level last year, measured as an annual average, and that it was fairly average with regard to deviations from the target level.
As monetary policy's effect on inflation has a time lag, the outcome of inflation in 2003 is largely a result of the monetary policy conducted in 2001 and 2002. Let me therefore report some of our considerations in the light of the developments and forecasts made then. At the beginning of 2001 the repo rate was 4 per cent. During the spring there was a severe depreciation of the krona. Despite the fact that economic activity in Sweden and abroad was in a weakening phase, domestic resource utilisation was at a high level, while inflation had increased significantly during the spring and exceeded the upper tolerance limit. Inflation expectations had also begun to rise. Given this, we raised the repo rate in summer 2001. Economic developments and monetary policy considerations in Sweden differed from those in many other countries, where monetary policy in many cases had begun to be more expansionary. The decision to raise the interest rate in July 2001 was not self-evident. Three Executive Board members entered reservations against the decision, with reference to the weakening economic activity.
The terrorist actions in September 2001 caused us to cut the rate by half a percentage point. Our assessment was that confidence among both households and firms would deteriorate. There was also a risk to the stability of the financial system. The economic slowdown that had begun could thus become more profound and prolonged than was previously anticipated. The easing of economic policy after the terrorist acts contributed to some recovery in international and domestic economic activity. In spring 2002 we made the assessment that inflation one to two years ahead would exceed the target level. A gradual upturn in the underlying inflation rate, combined with unexpectedly high wage increases in the services sectors in particular, indicated that resource utilisation could be higher than we had assumed. The repo rate was raised in two stages.
Steering interest rates in several other small countries were also raised at this time. However, during the summer and autumn we saw more and more signs that the recovery had come to a halt. Industrial activity weakened and geopolitical unease increased, as did the concern that the heavy stock market fall that began in 2000 would affect households and firms more than expected. The Riksbank, like the ECB and the Federal Reserve, began to reappraise the prospects for economic activity and inflation. Towards the end of the year, we made a downward adjustment in the inflation forecasts and the repo rate was cut by as much as it had been raised during the spring.
With hindsight, it could perhaps be claimed that raising the interest rate in spring 2001 and 2002 was unnecessary. However, this would not lead to a fruitful discussion. It is always easy to know how one should have acted afterwards, when the result is a fact. The interesting part of an assessment is whether our decisions are understandable in the light of the picture available at the time the decision was made. When we analyse this question, we can conclude that we have acted consistently on the basis of the inflation forecasts we made, much as we usually do.
A further question is how good we are at making forecasts. The accuracy of the forecasts for price trends as a whole in 2003, which were made during the two preceding years, appears relatively good. However, the general assessment hides a number of large forecast errors for individual components. For instance, domestic inflation was underestimated, while international prices and imported inflation were overestimated. We thus have some work left to do to achieve better precision in the individual components that are important to the total assessment. Despite this, I can observe that our inflation forecasts for 2003 were slightly better than the average of other forecasters.
Some concluding reflections
Let me now, in conclusion, emphasise a couple of issues connected with the present decision-making situation.
As I have already mentioned, economic activity has essentially followed the paths we have been anticipating since early last autumn. On the other hand, new information has been received, which indicates lower price and cost pressure. This information concerns international prices, productivity in Sweden, the labour market situation and wage costs as well as inflation outcomes and expectations. It is not always easy to interpret all this information. All in all, it has led to a significant change in our forecasts since October. This has in turn made us cut the interest rate by a total of 0.75 percentage points. Similarly, we made a couple of cuts in a row last spring, when it became obvious that prospects for economic activity weakened more than expected and our inflation forecasts were therefore adjusted downwards. This is the way we work. New information affects our assessments and sometimes leads to a reappraisal of our inflation forecasts. If this happens, it also has an impact on the policy we conduct.
The difference between the present situation and the situation last spring is that the uncertainty then also encompassed developments in the real economy. At that time there appeared to be a substantial risk that economic activity would continue to weaken. To this extent, the present situation, despite the sometimes rather heated debate, is not more troubled, but actually less troubled. Growth has increased over the past six months to one year and most indications are that it will continue to do so. This makes the need for interest rate cuts appear less acute. At the same time, inflationary pressure is lower. It is important to take the inflation target equally seriously whether it requires cutting or raising the repo rate. There is usually no reason to have a higher interest rate than the situation requires, this is part of the credibility of the policy conducted. The result of our endeavour to act consistently and symmetrically has been fairly good; inflation has remained in line with the target level during the years of inflation-targeting policy.
Nevertheless, I would like to point out that we do not follow our rules of action mechanically. Our focus is on the inflation rate one to two years ahead, and in this perspective we are clearly below the target level and today's decision is therefore well-motivated. In addition, inflation will probably be well below target level for the greater part of the coming year, while it is difficult to envisage any strong reasons why it should accelerate rapidly beyond the forecast horizon. This is connected with the fact that there is presently a high level of unutilised resources in the economy and that resource utilisation is expected to increase slowly. Given this background, there should be plenty of time to take action and raise the rate in future; when and if there is reason to believe that inflationary pressure is growing.
We also try to take into account in our actions risks to financial stability. This is primarily done on the basis of information on how lending and so on affects the banks, and through them, the payment system. However, there is also a more difficult problem concerned with how credit expansion and asset prices affect the conditions for the stability of the payment system, as well as for developments in the real economy in a slightly longer-term perspective. These are questions that have been discussed more at the Riksbank than many other central banks, given the background of the crisis at the beginning of the 1990s.
However, in today's concrete decision-making situation we have not considered the build-up of credit and increased indebtedness in recent years to be sufficiently serious to motivate a departure from our usual means of conducting monetary policy, namely on the basis of expected inflation trends a couple of years ahead. At present, we see no actual problems for the banks and the stability of the financial system. Nor do we see any major problems for the household sector as a whole to meet its obligations, which does not prevent problems for individual households from arising when the interest rate sooner or later begins to rise. All in all, there is good reason to closely monitor loan developments and the build-up of debt in society.
Thank you.
