Ingves: Introduction on monetary policy
Let me first of all thank you for the invitation to come here. This is my first hearing with the Committee on Finance and I hope that we will have a rewarding discussion on monetary policy. I shall begin with a brief report on our current assessment of the economic situation and future inflation prospects. Above all, I wish to clarify the reasoning as to why we are choosing to raise the repo rate now. Given that the recent monetary policy debate has revolved to a large extent around the objective of monetary policy, I would also like to take this opportunity to report our fundamental view of monetary policy strategy. My main message can be summarised thus: - There is one objective for monetary policy and that is to maintain price stability. - The task given to the Riksbank by the Riksdag and our method of working provide scope for taking other factors into account, such as growth and employment, as long as this does not jeopardise price stability. The inflation targeting policy has worked well in its present form and I therefore see no reason to change it. - The inflation forecast we publish today shows that UND1X inflation will be approximately in line with the target a couple of years from now. This assessment is based on, for instance, the assumption of firm growth in Sweden and abroad and gradual increases in the repo rate from its present low level. I will conclude my speech with a look back. The Riksbank has just completed its annual accounts for 2005 and it is therefore an appropriate time to discuss inflation developments over the past year, as is usual at the first hearing of the year. The most important conclusions are: - Inflation was unexpectedly low in 2005. This was mainly due to various changes on the supply side of the economy holding down inflationary pressure more than we had expected. Essentially, these changes were high productivity growth and weak growth in import prices. - These types of structural changes are difficult to capture at an early stage of the forecasting work. One indication of this is that in principle all forecasters overestimated inflation in 2005. The current assessment Before I discuss our assessment of the current economic situation, I would like to take up the issue of the link between our interest rate decisions and inflation. Why have we begun to conduct less expansionary monetary policy despite the fact that inflation is, and has been, low? To understand the most recent interest rate decisions, it is important to remember that we do not adjust the repo rate to affect the present inflation rate, but to influence inflation a couple of years ahead. It takes time for monetary policy to achieve its full impact on the economy. Monetary policy must therefore be based on expectations of the future, i.e. we are dependent on forecasts. It is therefore necessary that we occasionally ”lean against the wind”, that is, that we raise the repo rate despite the current inflation rate being below target level or that we lower the repo rate despite inflation being above the target level. If we were only to look at the present inflation rate, there is a risk that we would be forced to make more drastic adjustments further ahead and this type of roller-coaster interest policy would benefit no one. How do we assess future inflation prospects? The economic outlook is currently bright, both in Sweden and abroad. International growth has been both high and stable in recent years (diagram 1). Strong growth in the United States and parts of Asia has been a driving force for global economic activity. Increased participation in global trade by, among others, several Asian countries has contributed to this upswing and to stiffer international competition. So far, the increase in global demand has not had a visible impact on consumer prices. This is despite the oil price having almost doubled over the past two years and substantial price rises on other commodities. Most indications are that international economic activity will continue to develop well over the coming years. We are assuming that economic activity in the euro area will improve more and more and that growth in the United States and Asia will slow down somewhat. The Swedish economy has also shown good growth in recent years. The upswing in world trade has contributed to strong growth in Swedish exports. Over the past year, domestic demand has also developed favourably, partly due to stimulation from expansionary monetary policy. The labour market situation has at long last begun to improve. Our assessment is that growth in the Swedish economy will remain high over the coming years. Household consumption is expected to accelerate and employment to increase. We also envisage inflation rising as capacity utilisation increases (diagram 2). Over the past six months we have witnessed the beginning of a rise in the rate of price increase, albeit from a low level. This does not appear to be due to large price rises within a few product groups; it appears to be a broader upturn. The Inflation Report we are publishing today indicates that inflation will be close to the target level a couple of years from now. This assumes gradual increases in the repo rate. I would like to emphasise that there are factors which complicate this picture. Inflation was unexpectedly low in January. Dividing price trends into different categories shows that it was mainly some imported product groups that showed unexpectedly weak development. It is difficult to determine from the outcome of only one month whether or not this is a sign of a more lasting development. Naturally, it is something we will pay attention to in our future assessments of inflation. We can also observe that over the past few years inflation has been kept down by high productivity growth and low import prices. This is partly related to developments in economic activity. However, there are probably also lasting factors behind the low inflationary pressure. We can see similar patterns in some other countries. Amongst other things, it is probably a question of the effects of increased global competition. The durability of this type of structural change is difficult to determine. I intend to return to this in connection with my look back at inflation developments during 2005. A further complication is that house prices and household borrowing have increased rapidly (diagram 3). There are some risks with this type of development that can be difficult to quantify. One cannot entirely rule out the possibility that credit expansion and house prices are being driven up by over-optimistic expectations. This could lead to a situation in future with sudden downward adjustments in house prices and a large increase in household saving. Our assessment is that the probability of this type of scenario is small at present, but that if it were to occur, the negative effects on the economy would be substantial. Given our inflation forecast and after taking into account other factors, the Executive Board’s assessment is that it is reasonable to continue gradually raising the interest rate. At the same time, it is important to emphasise that we are easing up on the accelerator, not putting on the brakes. Monetary policy is still expansionary, which is clear from a historical comparison of repo rate levels (diagram 4). Even in an international comparison, it is clear that Sweden’s key rate is still very low. The link between interest rate decisions, house prices and household debts There have recently been claims that the Riksbank’s decision to raise the repo rate was largely because we have been concerned over house prices and increased household indebtedness. This is not the case. I would like to argue that the decisions to raise the interest rate now and in January are primarily justified on the traditional basis of inflation prospects two years ahead. Inflation has been assessed as being close to the target level a couple of years ahead, given gradual increases in the repo rate. We have therefore chosen to raise the repo rate. Our monetary policy considerations have of course taken into account considerations related to risks linked with house prices and the build-up of household debt. House prices affect demand through households' decisions regarding saving and consumption. Our inflation forecast therefore always includes an assessment of credit developments and the housing market. As I mentioned earlier, it may be difficult to quantify the risks linked to a rapid increase in house prices and indebtedness over a long period of time using our usual forecasting methods. As the expansionary monetary policy is to a large extent a driving force in developments in house prices and indebtedness, we have considered it reasonable in our monetary policy considerations to view these risks as an argument in favour of beginning to increase the repo rate now, rather than postponing the increases for a few months. However, this does not mean that we have set up any target for house prices. I intend to discuss the objective of monetary policy in more detail later. Let me first say a few words on future interest rate developments. Future interest rate developments The inflation forecast contains an assumption of how the interest rate will be adjusted in future. This is a relatively new method of making forecasts for the Riksbank. Like many other central banks, we have previously worked on the assumption of the interest rate remaining at the same level throughout the forecast period. The new method of forecasting also alters the communication and motivation of interest rate decisions somewhat. An important insight is that it is not possible to describe the strategy for monetary policy in terms of a simple policy rule - for instance, that the interest rate will be raised now if the inflation forecast two years ahead shows inflation above target level. This type of interpretation assumes that the forecast is based on an assumption that the interest rate will not change during the forecast period. With the new assumption, the inflation forecast is instead based on a whole series of future interest rate adjustments. The Riksbank’s forecasts currently assume that the series of interest rate changes will follow market expectations (diagram 5). I have recently raised the question of possibly going one stage further and instead reporting the Riksbank’s own forecast of interest rate developments. For my part, I consider it natural to make forecasts of future interest rates in the same way that we make forecasts for a number of other variables. Nor do I see any drama in this type of transition, as we already express opinions about one interest rate path. However, this is something we will discuss among the Executive Board members. Regardless of which alternative we choose, it is important to point out that the interest rate development we publish should not be regarded as written in stone. The future development of the economy is uncertain and interest rate developments are therefore also uncertain. The Riksbank is therefore unable to tie monetary policy to a particular sequence of future interest rate raises in advance. Our current assessment is that the forecasts described in the Inflation Report indicate low inflationary pressure and that the repo rate could perhaps be raised at a slightly slower rate in future than the interest rate path on which the forecasts are based. Market expectations of future interest rate increases have also been adjusted downwards in recent days, which thus appears reasonable – given the information we have at present. Prior to the next interest rate decision in April, we will of course need to make a new assessment and take into account the information available then. The future stance of monetary policy will thus depend as usual on new information received and on the Riksbank’s assessment of the inflation outlook. The principles for the Riksbank’s monetary policy The economic debate has recently focussed on monetary policy objectives. Some people have suggested that we should supplement our inflation target with other targets. Others have interpreted our most recent interest rate decision as meaning that we already have targets for other quantities and that we are trying to achieve too many targets. It is unfortunate that there appears to be a certain amount of confusion in this matter, so let me repeat my earlier comment: there is one objective for monetary policy and that is to maintain price stability. This is the objective set by law for the Riksbank’s monetary policy. In addition, we have been given the task of promoting a safe and efficient payment system. In connection with the work on legislation, it was discussed whether the Riksbank should also have objectives regarding GDP growth or employment, for instance. The Government and the Riksdag (parliament) decided that the Riksbank as an authority under parliament was obliged to support general economic policy goals such as sustainable growth and high employment. However, it was pointed out that this support should not come into conflict with the price stability objective. It is thus important to remember that the legislators considered it natural for the Riksbank to take into consideration the real economy, but that this consideration was not described as an objective for the Riksbank’s monetary policy. In my opinion, this was a wise decision and I see no reason to change the framework and introduce additional objectives to price stability. The Riksbank’s current method of working closely follows the views of the legislators. Monetary policy takes into consideration other aspects than inflation trends, and it should do so. How do we do this? As you know, the practical monetary policy work is governed by the objective that the increase in the consumer price index should be 2 per cent a year. This is how the Riksbank has chosen to specify the statutory objective of price stability. There is a tolerance interval of ± 1 percentage point around the target. The purpose of this interval is to explain that deviations from the target are unavoidable, but also that overly large deviations cannot be easily accepted if the inflation target is to remain credible. To make this clearer, monetary policy is guided by the principle that when there are deviations from the target, our aim is usually to bring inflation back to target level within two years. The tolerance interval together with this principle creates flexibility for monetary policy. This is necessary, as it is not possible for the Riksbank to control inflation with any precision in the short term. It is quite simply not within our power to maintain inflation at exactly 2 per cent all of the time. The flexibility also creates an opportunity for us to take into account real economic considerations as long as price stability is not jeopardised. Temporary deviations may thus be justified by taking into consideration growth and employment, for instance. The statutory objective of price stability must not be disregarded so it is important that the flexibility is not used in a way that would undermine confidence in the inflation target. One can say that the two-year horizon sets a limit as to how much consideration we can normally give to other aspects than price stability. The clarification of monetary policy published in 1999 also made it clear that under some circumstances it might be necessary to have an even greater degree of flexibility and allow inflation to take longer to reach the target than the usual two years. For instance, the economy could suffer a major supply shock that led to prices of commodities, energy and foods rising or falling substantially. In these situations it might be reasonable to accept a longer deviation from the target than two years if the strain on the real economy would otherwise be unacceptably high. Let me summarise: The Riksbank has one objective for monetary policy and that is to maintain price stability. The Riksbank's method of working encompasses possibilities within this framework to take into account different aspects of interest rate decisions, such as fluctuations in the real economy. However, such considerations do not mean that we have specific targets for other quantities than inflation. A look back - inflation 2005 Finally, I intend to take a look back at developments in inflation over the past year. This is to follow the usual custom at the first hearing of the year and give an account of the outcome of the monetary policy conducted. As you know, inflation was low last year (diagram 6). For the second year in a row, the inflation rate was below the Riksbank's tolerance interval for inflation. UND1X inflation was 0.8 per cent and CPI was 0.5 per cent on average for the year. The forecasts of inflation made by the Riksbank in 2003 and 2004 anticipated that inflation would be much higher than that in 2005. Why was inflation surprisingly low? To begin with, one can observe that it was not due to a weak economic upturn. On the contrary, the economic upturn accelerated in 2004 and growth was even stronger than expected. Last year the economy also continued to grow at a good rate. To understand the low inflation it is necessary to find other explanations than weak demand in the economy. The reasons should mainly be sought in the positive changes on the supply side of the economy, which I have already mentioned, namely high productivity growth and low import prices. High productivity growth has held down cost pressures in the economy more than expected. Production has been able to increase without any particularly high increase in companies’ labour costs per unit produced (diagram 7). During 2003 and 2004 unit labour costs in the business sector actually fell. At the beginning of an economic upturn it is normal to have a period when production increases at a more rapid rate through companies using existing labour force more intensively rather than through new recruitment. However, even taking into account the business cycle, productivity growth has been high in recent years. This indicates that productivity growth has increased in a more lasting manner. The Riksbank has also gradually increased its forecast for the productivity growth trend in recent years. Despite this, productivity has continued to be higher than forecast up to 2005. One difficulty in the forecasting work is that productivity growth usually fluctuates considerably even taking into account changes in economic activity. For instance, productivity growth fell heavily in 2001. Another supply factor that has contributed to low inflation is, as already mentioned, low increases in imported goods prices. Since early 2003, consumer prices for imported goods and services, excluding oil products, have fallen (diagram 8). The fact that the krona appreciated continuously from 2002 and up to the end of 2004 naturally played a role in the low import prices. However, the Riksbank’s forecasts for import prices included an assessment that the exchange rate would strengthen. Nevertheless, import prices excluding oil products have been surprisingly low. What can then explain the high productivity growth and the weak developments in import prices? This is, of course, difficult to know. One hypothesis often used to explain high productivity growth is that the extensive investment in information technology made earlier has begun to show results. Another explanation that I touched on earlier, and which has a bearing on both the rapid productivity growth and the depressed import prices, is the increase in competition. Stiffer competition from China and other low-cost countries has, for instance, contributed to increased pressure for rationalisation in many sectors. This, combined with outsourcing of work to low-wage countries, may have created opportunities to improve productivity at a faster rate than before. In addition, an increasing percentage of Swedish imports are coming from countries with low price levels. This shift in imports from countries with relatively high prices to countries where prices are much lower has probably contributed to keeping down prices of imports to Sweden more than the Riksbank had previously estimated. Competition within Sweden has also sharpened, which has increased the need for greater efficiency in many sectors. For example, our assessment is that increased establishment of low-price supermarket chains contributed to a fall in food prices in 2005. Should the interest rate have been cut further? It is clear that structural changes in the economy have in many ways held down inflation and created scope to cut the repo rate to historically low levels (diagram 9). From time to time there has been criticism which says that we should have cut the repo rate even more. The fact that such a discussion arises when inflation is well below the target level is not in itself surprising. However, it is important to have realistic expectations as to what monetary policy can achieve. Changes of a structural nature are difficult to capture at an early stage of the forecasting work as it takes some time before it becomes clear that the changes are not merely transient factors. This is one reason that needs to be recognised why inflation may deviate from the target level. One indication of the difficulties in making accurate forecasts is that inflation in 2005 was over-estimated by in principle all forecasters (diagram 10). A comparison indicates, if anything, that the Riksbank was among the first forecasters to take into account the effects that have subdued inflationary pressures. Some analysts have periodically advocated a somewhat faster rate for interest rate cuts than that chosen by the Riksbank. However, the difference has on average been small. Given the effects of monetary policy normally assumed, the difference in inflation and employment would have been marginal. For monetary policy to bring inflation much closer to the target level, a radically different policy would have been necessary. For instance, the Riksbank would probably have needed to cut the repo rate quickly and substantially a couple of years ago rather than making the gradual cuts that were implemented. It would have been difficult to justify this type of monetary policy given the view of economic prospects at that time. It is also clear that the monetary policy that was conducted has had a strong expansionary effect on the economy. For instance, GDP growth has been high. The expansionary monetary policy has also been evident in the rapid credit expansion in the household sector and the price increases in the housing market. It is doubtful whether it would have been wise to reinforce this development with even more expansionary monetary policy. Unfortunately, the strong growth has not resulted in a corresponding recovery in employment. On the contrary, the labour market has developed more weakly than might have been expected from a historical perspective, given the high GDP growth. To some extent this reflects the unexpectedly high productivity growth. However, a reasonable explanation is that there are also problems in the labour market that are not cyclical. This is something that should be resolved by other means than monetary policy. Let me conclude by saying that I welcome discussions on monetary policy. Monetary policy is to a great extent based on assessments and it is therefore natural that there are differing opinions as to what should be done with the repo rate – both outside of the Riksbank and within the Executive Board. However, it is also important that the discussion is based on a realistic picture of what monetary policy can achieve. Nor should the discussion be simplified by targeting one issue without taking into account all of the risks and other aspects that need to be weighed against one another in monetary policy decisions. We at the Riksbank will do all we can to contribute to this type of discussion climate.