Ingves: Introduction on monetary policy
I would like to begin in roughly the same way as I did when I was last here – by saying a few words about the objective and basic purpose of monetary policy. It is in the light of this that we make our interest rate decisions. And our actions should be assessed against this background. Our most recent economic assessment and the material for assessing monetary policy are also what we are here to discuss today.
We must always bear in mind that the basis for monetary policy is the price stability objective which the Swedish people, through the Riksdag, have delegated to the Riksbank. And this is actually not merely an objective, it is also a means. By keeping inflation low and stable, monetary policy shall help to create the best possible conditions for good and sustainable economic development.
To do this in the best possible manner we must constantly learn from what is happening and try to become better at what we do. Part of this process involves, of course, examining, assessing and discussing monetary policy in a constructive manner. Here the Riksdag Committee on Finance plays an important role, both in examining and in developing the examination process.
The Committee has recently taken several good initiatives. The most recent example was ordering the more detailed material for assessing monetary policy which the Riksbank published about a week ago. I shall return to this assessment later. But first I would like to say a few words about current monetary policy.
Our most recent assessment
Our most recent interest rate decision was not an easy one. There is considerable uncertainty, for instance, on which direction international economic activity will take. And, as you are aware, we also had slightly differing opinions within the Executive Board. However, we must nevertheless make a decision and the majority of us reached the assessment that the repo rate should be raised to 4.25 per cent. We are forecasting that the repo rate will remain around this level over the coming year. This type of interest rate development will contribute to inflation falling back and being around the 2 per cent target a couple of years ahead, at the same time as production and employment develop in a balanced manner.
Different forces are pulling in different directions
In our decision we weighed together many different factors. A lot has happened; both events that could push up inflation and events that could hold it back.
On the one side were conditions implying a higher interest rate. Inflation rose rapidly in 2007 and we are expecting it to remain high over the coming year. Expected inflation has also risen and cost pressures are high. Economic activity in Sweden remains good and the labour market is strong. Although GDP growth will slacken and employment will increase more slowly this year, resource utilisation will nevertheless be higher than normal.
On the other side there was an increased risk of weaker growth in the world economy. There is considerable uncertainty over which direction developments in the US economy will take and how this will affect the rest of the world. And the financial markets have continued to be marked by unease and turmoil.
When all of these factors were weighed together, the majority of us reached largely the same view regarding the repo rate as expressed in our December forecast, that is, that the interest rate needed to be raised in order for us to attain the inflation target. However, there is still considerable uncertainty regarding economic indicators and inflation prospects. It therefore appears particularly important to point out that the interest rate path is a forecast. It is not a promise! And the interest rate decisions that are made in the future will of course depend on how new information affects the prospects for inflation and economic activity in Sweden.
Communication in connection with the interest rate decision
It can hardly have escaped the notice of anyone here that the Riksbank received a good deal of criticism in connection with this interest rate increase. What has been criticised is, perhaps, not so much the increase as such, more that the interest rate decision took many by surprise.
There have been comments that the Riksbank has been too silent and not prepared market agents and others for a raise. We must, of course, consider this criticism. A natural part of monetary policy is to consider whether there are things we could do better. This is the way we always try to work.
There is one aspect of this of which it is important to be aware. This is that we quite simply cannot send clear signals as to how the interest rate will be set two to three weeks ahead of a monetary policy meeting. This would require us to be fairly certain about the majority opinion and that it will not change. The decision-making situation is often unclear to the very last. And, I would say that this was definitely the case this time.
Now, I do believe that there is widespread understanding that we cannot in advance state how the interest rate will be set. It may be the case that as a result of the recent turmoil there has been more of a need than usual that the Riksbank comments on the course of events, without anticipating the interest rate decision.
Assessing monetary policy
Now let us look back and discuss the material for assessing monetary policy, which came about on the initiative of the Committee on Finance. It is for me quite special this time. This is the first occasion on which the material for assessing monetary policy actually covers decisions that I have been party to.
I do not intend to say a great deal about any conclusions in the material in this brief introduction. The point is that it is the Committee on Finance that assesses us. However, I would like to raise several issues of principle about assessment in general and about assessing monetary policy in particular. These issues have in various ways left their mark on the formulation of the material.
For several reasons, it is important that we are assessed. As I mentioned earlier, one reason is that we become more efficient and develop our work methods. Monetary policy acts with a lag and must therefore be based on forecasts. If we are to be as proficient as possible in forecasting future developments then we must continuously take heed and incorporate new knowledge, both new theories and everything else that is happening in the economy “right now”. One might say that it is similar to navigating with a nautical chart that is continuously being redrawn a bit. But it is not enough to have the latest shipping forecast and an up-to-date chart in order to be a skilful skipper. In order to learn just how much the rudder needs to be adjusted in rough seas, one has to learn from experience. Hopefully, one becomes more skilful at maintaining one’s course each time one has to parry. And, for monetary policy to be as efficient as possible, we must, in the same way, learn from past events.
Another reason concerns our role as an independent authority. It presupposes that the general public has ample opportunity to insight into our activities and that we are regularly examined. So assessment is important to us – both from an efficiency perspective and a democratic perspective. However, for the assessment to meet these purposes, there are a couple of factors to be taken into account.
What can monetary policy achieve?
To begin with, the demands – the actual base for the assessment – should be founded on what monetary policy can actually achieve. Monetary policy is able to secure an inflation rate that is in line with the inflation target over a number of years. It can also contribute to stabilising cyclical fluctuations in production and employment. However, monetary policy cannot achieve a permanently higher level of employment or growth in the economy. Both previous experience and economic theory have shown this to be the case. And this was one of the reasons why it was decided when drawing up the Sveriges Riksbank Act that it would be inappropriate for the Riksbank to have an explicit target regarding, for instance, GDP growth or employment, in addition to the price stability objective.
It would then appear natural to assess monetary policy by looking at the way in which the inflation outcome relates to the Riksbank’s inflation target. But there are some factors here too that should be borne in mind.
In retrospect
One such factor is that efficient inflation targeting by necessity must be based on forecasts of economic development. When a central bank changes its policy rate it takes time before, for instance, inflation, production and employment react. Assessing monetary policy in retrospect is, therefore, not as simple as it may seem. It is not enough to merely make a direct comparison of outcome and forecast, or inflation outcome and target. The inflation and production that are observed will have been affected by various events, which in many cases could not be predicted when the forecast was made. It is important, therefore, to consider what information was available at the time the decisions were made when one discusses whether the analysis and interest rate decisions were reasonable. A large forecasting error can imply that the forecast was poor. But, it can also be due to a shock, which could not have been predicted, having a major effect on the economy.
What then, are reasonable demands to make of our forecasts? One reasonable demand is that they do not on average have poorer accuracy than forecasts made by other economic analysts. Should our forecasts be ”systematically” inferior, it would evidently have been possible for us to make better assessments on the whole. This could also mean that our monetary policy decisions probably were not as good as they could have been.
A long period of examination is necessary to be able to say anything more definite about the accuracy of the forecasts. The economy is constantly exposed to random shocks that lead to forecasts being incorrect. However, over a long period of time these errors offset one another and on average good forecasts should therefore not over-estimate or under-estimate the actual events. For shorter periods it may be interesting to study individual forecasting errors in greater detail to try to analyse their causes. But even if such an analysis is valuable it cannot be used to draw general conclusions about the forecasting performance.
While accuracy is very important, even more is required for a forecast to be good in a monetary policy context. A good forecast for inflation, for instance, should correspond with the forecasts for other variables. And the driving forces behind the forecasts should be explained in a comprehensible manner. Why does the forecast look the way it does? What are the mechanisms acting as a motor for this development? These qualities are also required for a forecast to function as part of the decision-making base for practical monetary policy.
Consideration to the real economy
Another factor to take into account in an assessment of monetary policy is that there may sometimes be reason to deliberately deviate from the inflation target. Normally we focus our policy on attaining the inflation target within two years. This is partly due to it taking time before monetary policy has its full effect. But it also has to do with the Riksbank thereby being able to contribute to a more stable development in production and employment. However, it does assume that the inflation target is not threatened.
To be able to assess monetary policy during a certain year, it is important to remember that such considerations also lie behind the decisions. In order to gain a picture of the initial situation for a certain year, it is necessary to study the course of inflation in earlier years. The fact that inflation has been allowed to deviate from the target may be due to poor monetary policy decisions. But it may also have been part of a deliberate strategy by the Riksbank to be able to balance production and employment. If we do take such considerations, we shall however have clearly explained this in connection with our decisions.
So how did it turn out?
After this general discussion, allow me to conclude with a very brief description of the considerations that monetary policy faced during the assessment period. The development we envisaged in 2005 and 2006 was that economic activity would continue to strengthen. Cost pressures and resource utilisation were expected to rise gradually, but a strong development of productivity was nonetheless assumed to hold back inflationary pressures. The most important consideration was the rate at which the repo rate should be raised. It was a question of the low inflation rate initially rising towards the target, but in a way that did not involve a risk of an overly rapid increase in the rate of inflation later on.
At the beginning of 2006 we began a phase with increases, which continued in 2007. The policy was expected to contribute to inflation being below the target during a relatively long period, to then be around 2 per cent, measured in terms of the CPI, during 2007. Underlying inflation, the CPIX, was expected to be around 1.5 per cent. The inflation outcome was close to these forecasts - at 2.2 and 1.2 per cent respectively. As I just observed, one cannot rely too heavily on individual forecasts. But I cannot deny that it, psychologically at least, feels far more satisfying than if we had been further away.
But, naturally, the actual assessment of how well we have managed is the responsibility of the Committee on Finance. And the Committee’s initiative to the independent material for assessment that we have now published will hopefully make it easier to examine our work. I hope it will contribute to a good discussion on monetary policy.
Thank you!
