Srejber: Monetary policy and inflation expectations
I should like to begin by thanking you for the invitation to come here. Today I intend to talk about what guides the Executive Board when we make decisions on monetary policy and what role predictability in the central bank's actions plays for maintaining price stability. Then I shall take a brief look back at the economic situation immediately following the tragic events of 11 September and summarise the economic situation today. Finally, I should like to explain my views on monetary policy and why I entered a reservation against the decision to leave the repo rate unchanged at the last Executive Board monetary policy meeting.
Monetary policy strategy
I shall begin with a reminder of what guides us when we make decisions on monetary policy; i.e. how monetary policy strategy has been formulated.
The Riksbank has been assigned by law the task of maintaining price stability. We also have the task of promoting a safe and efficient payment system. The banks play a central role in both of these tasks. The monetary policy objective has been defined by the Riksbank as a target for CPI inflation of 2 per cent with a tolerated deviation interval of +/- 1 percentage point. The steering interest rate is the instrument at the Riksbank's disposal for maintaining price stability. It is also possible to use other methods, e.g. foreign exchange interventions, as we did in the summer, as a means of achieving our objective. However, the repo rate is the monetary policy instrument normally used.
Our monetary policy strategy can be described, somewhat simplified, with the rule of thumb that when inflation one to two years ahead is estimated, taking into account the risk scenario, to deviate from the inflation target, the steering interest rate is adjusted. The Riksbank's assessment of the future inflation rate is communicated, for instance, in the Inflation Reports presented four times a year.
The Riksbank has chosen to set its inflation target in a medium-term perspective. This is because monetary policy influences inflation with some time lag. An ambition to try to ensure that inflation also developed in line with the target in the short term would probably require significant changes in the steering interest rates and thereby contribute to unnecessarily large fluctuations in demand. The fact that the inflation target is expressed in terms of a medium-term perspective also plays down the importance of temporary price changes. Consumer prices are sometimes affected by various disturbances that have temporary effects on inflation. Political decisions, for instance, on tariffs and indirect taxes, can also have a temporary effect on the rate of price increase.
It may also be useful here to describe briefly how monetary policy affects inflation. A change in the interest rate has effects on the rate of price increase via a number of different channels. For instance, the interest rate influences the general public's decisions regarding investment and consumption. A higher interest rate normally means that growth in demand is slightly subdued, which is assumed to entail that the amount of free resources in the economy increases slightly in the long term. The effects on demand often come after a few quarters and thus also subdue price increases at an even later stage.
A change in the interest rate also affects the exchange rate. An increase means that the return on assets in the home country's currency is slightly higher. This also means that expectations of growth and inflation are subdued somewhat relative to developments in the rest of the world, which according to the textbook can contribute to a slightly stronger exchange rate. If the krona appreciates, it becomes slightly cheaper to import to Sweden, which can be assumed to lead to import prices on input and consumer goods increasing at a slower rate. A stronger krona also subdues global demand for Swedish export goods, which helps subdue growth in demand. On the other hand, if there is little confidence in the country's economic policy, an increase in the interest rate risks resulting in a weaker exchange rate rather than the reverse, because confidence in growth is low.
Monetary policy can also affect price developments via households' and companies' expectations of future inflation. In this context it is important to point out that there are a number of different measures of inflation expectations. The Inflation Reports contain different measures of how households and companies expect the rate of price increase to develop. Pricing on the financial markets can also provide some information here. A further type of inflation expectation is reflected in the various forecasters' assessments of developments. The Riksbank bases its estimations on a total assessment of these different measures of inflation expectations.
Stable and low inflation expectations affect both companies' pricing behaviour and wage formation, as well as having a beneficial effect, as far as monetary policy is concerned, on how consumers react to price changes.
Companies' pricing behaviour is based partly on the development of marginal costs, the competitive situation, demand and inflation expectations. If a company has confidence in the Riksbank's monetary policy, its prices will not be affected as much by temporary changes in, for instance, supply, demand or the exchange rate. In this situation, the company can assume that inflation will return to the long-term trend once the effects of the disturbance subside. A credible inflation target can thus contribute to maintaining the monetary policy target at lower cost in terms of fluctuations in demand and interest rates.
One means of inspiring confidence is to show a high degree of openness and clarity with regard to how monetary policy is conducted. The fact that the Riksbank describes the considerations regarding economic prospects that form the basis for the interest rate decisions makes it easier for external analysts to understand and predict monetary policy. Moreover, the fact that the
considerations are reported openly may help increase confidence in the assessments and conclusions.
Consistency in monetary policy actions also increases confidence in the price stability target, which in turn should have a stabilising effect on inflation expectations.
The role of predictability in monetary policy
It is important the Riksbank is clear and explains the bases on which the interest rate decisions are made. As the effect of monetary policy on price developments has a time lag, the interest rate is changed on the basis of forecasts of future price developments. However, it is not always easy to interpret correctly what is happening in the economy. In recent years it has been difficult to find a clear relationship between the demand situation and inflation in Sweden. GDP growth was high from 1997 to the middle of 2000. Since then the growth in demand has been subdued. During 1997 and 1998 inflation fell, but the rate of price increase has tended to accelerate thereafter. At the same time, inflation expectations have remained at a relatively stable level.
There are many explanations for this development and the Riksbank's view has been described, for instance, in various Inflation Reports. Let me here expound some personal reflections on this issue.
If the Riksbank's monetary policy inspires confidence, inflation expectations will be established around the inflation target. This means that resource utilisation in the economy could vary slightly more, without having an immediate impact on inflation. Stable, low inflation expectations based on the assumption that the central bank shall act to control the inflation rate can mean that inflation will not become too high, even if resource utilisation increases above its potential. Conversely, if the confidence is lower, high inflation expectations can mean that prices do not fall, even if resource utilisation declines. There are also a number of studies indicating the relationship between inflation and growth is not stable, but is due to the degree of resource utilisation and to what the general public thinks of the central bank's actions. A central bank that has a high level of credibility as safeguarder of price stability can allow demand to be higher in relation to potential growth than a central bank with the opposite reputation.
The fact that both wage setting and price setting behaviour are affected by inflation expectations means that it takes some time before a reduction in resource utilisation has an impact in terms of falling prices and lower wage hikes. In the same way as it takes time to change inflation expectations and pricing behaviour downwards, it is difficult to change them upwards. As I mentioned above, this is partly because the general public expects certain behaviour from central banks, and partly because raising prices involves costs for the companies. These are both purely administrative costs and also reputation costs. Reputation costs in the sense that it is more damaging to good will to be among the first to raise prices, than to be one of many during a wave of price increases. This means that once some companies have started raising prices, many others soon follow. For this reason, if the central bank misunderstands the situation and allows a positive inflationary pressure to prevail as long as the inflation rate remains low, the general public's image of the central bank's behaviour will change. This in turn will affect pricing behaviour and once the increases have started, more will follow in the shadow of the first. This is why inflation often comes at a late stage during an economic upturn and it can mean that inflation remains too high during a period when economic activity is declining, despite falling demand and production. It is something of a paradox that a credible central bank can have greater difficulty seeing the signs of price pressure in time.
In my opinion, there is a risk that Swedish inflation has been held down through deregulation and increased competition in such a way that we have been lulled into thinking that the inflationary tendencies in the economy are lower than they actually are. When companies' scope to reduce their prices as a result of, for instance, deregulation and increased competition, begins to be exhausted and demand pressure is finally expressed in rising prices, there is a risk that the price increases will be both greater and introduced more rapidly than expected by us analysts.
Monetary policy must be based on forecasts of inflation and resource utilisation. If monetary policy is based on an overly optimistic assessment of the interplay between growth and inflation, it can lead to monetary policy becoming overly expansive. If inflation expectations have become more slow moving, it could mean that there is time for considerable imbalances to build up before policy has an impact in the form of higher inflation. Examples of this are Japan in the 1980s, the USA in 1927-29 and again in the 1990s.
The example from Japan indicates that even if inflation does not rise during the forecast period with which monetary works, it is important to take a broader outlook both throughout the entire economy and further forward in time, and to envisage the risks that could exist with regard to financial stability. In Japan the cost of borrowing was very low during a long period of time to borrow money to invest in shares and property. As the productivity of the economy increased, confidence in the future grew and this made both investors and credit institutions more willing to take risks. The result was that asset prices were pushed up. A raise in the interest rate in this type of situation, where over-optimism was rife and both inflation and interest rates were low, could have meant that investors, property buyers and credit institutions would have acted rather more cautiously than if the interest rate had remained unchanged at a low rate and thus at least subdued the price bubble somewhat. The costs to society of waiting too long to change the direction for monetary policy can be high if imbalances are allowed to grow and the fall in production can be drastic when the imbalances are corrected.
It is therefore important that monetary policy should act in a preventive manner and that great vigilance is exercised with regard to the temperature of the entire economy, both the real and the financial parts.
Let us return, after this outlook, to our own, calmer circumstances. Low, stable inflation expectations cannot be used in all situations as an argument in favour of retaining monetary policy unchanged. It could be possible that the expectations are based on the general public assuming that the Riksbank will act if and when required. The fact that inflation expectations remain stable could then reflect expectations of future monetary policy. If these expectations are not fulfilled, then inflation expectations, and later on the rate of price increase, will rise. The example from Japan also shows that it is not certain that inflationary pressure will be visible in CPI inflation; sometimes price pressure is first seen in other prices than consumer prices. If the Riksbank waits to adjust the repo rate until inflation expectations and inflation have already increased, there is a risk that the delay will be too great. In this type of situation larger interest rate adjustments may be required than if the Riksbank had acted at an earlier stage.
This brings me to the current economic situation. However, first I would like to take a brief look back at the events that occurred in connection with the acts of terrorism in September.
The Riksbank's actions following the terrorist attacks on 11 September 2001
After the terrorist attacks in the USA, the entire world was shocked to realise that something so dreadful could happen. The Executive Board made an assessment that the uncertainty caused by the acts of terrorism risked further subduing demand internationally, mainly through the effects on companies' and households' confidence in the future, which could result in a more profound and prolonged economic downturn. This was why the Riksbank lowered its interest rate, together with the central banks in the USA, the rest of Europe and several other countries.
Prior to the terrorist attacks in the USA on 11 September, new statistics indicated that the global economic downturn could reach its lowest level at the end of 2001. However the picture was not clear-cut. As a result of the terrorist attacks, uncertainty increased even further regarding the time scale for a recovery. If the consumers' confidence with regard to the future were to decline, together with rising unemployment and a weak development in the stock markets, there was a great risk that activity in the US economy would be subdued. Companies often react to uncertainty with a "wait and see" policy, which means that they postpone investments. In addition, there was uncertainty as to whether the financial system could cope with the problems arising as a result of the devastation in New York. On top of this, certain sectors in the global economy would be hit directly, for instance, insurance companies, airlines, tourism, etc. All of this risked worsening the economic downturn both in the USA and in other countries.
The role of a central bank in this type of situation is to contribute to creating confidence and promoting financial stability. It is then important to act quickly and resolutely. This may be considered to contradict the claim that it is important for the Riksbank to act predictably. However, for those who have observed the central banks' actions on other occasions when problems in the financial system have arisen and confidence in the future has declined, such as in 1987 and 1998, it should be no surprise that central banks act to support the economy with a lower interest rate and greater supply of liquidity. This is to counteract a downward spiral in the global economy as the result of a decrease in the willingness to take risks. Interest rate reductions around the world can be seen to be an insurance against a really poor outcome.
It is important to remember that Sweden already had a very expansive economic policy on 11 September, with a low interest rate, weak krona and expansive fiscal policy. Given this, it is natural to withdraw the 'insurance' interest rate reduction when confidence and expectations have stabilised and the risks of a severe negative spiral have declined.
The current economic situation
So, can we say that confidence has returned or stabilised? To answer that question, I would like to describe how we see the current economic situation. The fact that willingness to take risks has increased can be seen in Standard & Poor's credit spreads, which have returned to the same levels as prior to the terrorist attacks. In the USA, swap spreads had fallen since mid-September as no financial crisis occurred, however they have increased slightly since the start of the year, probably because of the Enron scandal. In Europe and Sweden, on the other hand, swap spreads have continued to fall to lower levels than prior to 11 September.
Much of the statistics received since the beginning of December support the picture of international developments painted in the December Inflation Report. The signals that the subduing effect on demand is declining and that the level of activity is stabilising in the US economy have become clearer. Restraining factors, such as destocking and an adjustment in investments, appear to be gradually abating in strength. On the other hand, the uncertainty over developments in the labour market, lasting imbalances in savings and volatile stock market developments comprise a continued risk factor in the forecast.
There are also signs that the industrial sectors in many other OECD countries are entering a phase of recovery.
One particular risk factor in the developments is, in my opinion, that equity prices have been pushed up in the USA, given the economic situation. At the same time, the degree of indebtedness in the private sector remains relatively low. The low real interest rate entails risks for future developments in asset prices, particularly for the stock market. These risks need not be impending in the current situation, but the problems could become tangible in a couple of years' time. This means that there is a risk that economic policy, which is very expansive in order to avoid a strong downward spiral, would postpone the adjustment of imbalances in the economy. Some parallels can be drawn with the events that took place after the crisis in 1987, when a financial bubble recurred. A further example of an asset bubble way back in time is what happened in the USA in 1927. At that time, the US central bank pursued an expansive monetary policy in order to stabilise the gold standard. This contributed to an over-valuation of assets and later to the crash in 1929. If the financial imbalances in the USA are not corrected slowly in the near future, there is a risk that the economic upturn will be more short-lived and the downturn more powerful when the correction is finally made.
On the other hand, the continued developments in the USA will depend on whether domestic demand picks up and is maintained. If the growth in productivity has really reached a bottom level at around 1.5 per cent in an economic downturn, this could be a sign that the long-term sustainable growth rate in the USA has moved upwards and that a recovery will not merely be temporary.
Looking at Sweden, it was assessed in the December Inflation Report that the GDP had shown weak growth during 2001 and that it would then rise during 2002 and 2003. The main features of the picture we pained in the Inflation Report still stand. However, according to the National Accounts for the first three quarters of last year, the development in private consumption was weaker than had been assumed. There are other statistics that indicate that the GDP may have been stronger during 2001 than shown in the National Accounts, for instance, credit growth, money supply, tax payments and developments in employment. Regardless of what is correct, we appear to have had a relatively moderate slowdown during 2001. Industrial activity has been very weak, but this has been balanced by the expansive economic policy contributing to other sectors in the Swedish economy developing well. The question is how much of the slowdown in the industrial activity will have time to spread to other parts of the economy before there is an upturn in industrial activity. For the future we see a very good development in disposable income, some recovery in household wealth and continuing low oil prices, which should provide a stronger growth in consumption both this year and next year.
The situation on the labour market gives mixed signals. Open unemployment rose somewhat between December 2001 and January this year. Employment levelled off during the second half of last year, but the increase was still 80,000 persons, between 2000 and 2001. Wage rises were slightly higher than expected last year, despite weaker economic activity. However, developments on the labour market are divided in two parts. Employment in the manufacturing industry has fallen since the beginning of last year, while it is only now possible to detect a slowdown in the services sectors. Similarly, the number of redundancy notices has been highest in the manufacturing industry. In the public sector, healthcare and welfare, there have been no notices of redundancy and the number of new vacancies registered has increased. The fact that the weak growth in the manufacturing industry has not spread to the private and public services sectors is probably due to the expansive economic policy.
Forward-looking economic indicators, such as confidence indicators for households and companies, imply that the downturn in activity is abating and that developments are beginning to stabilise.
With regard to prices, inflation increased slightly more than expected in 2001. This can be partly explained by unexpected price increases on, for instance, fruit and vegetables; the price of these product groups will probably continue to increase slightly more than expected in the short term. Imported inflation has increased more than expected in both the producer and consumer areas, which could be due to the exchange rate's effect on prices being slightly higher than assumed. However, price increases have also been broader and higher than expected for other goods and services.
My view of the current situation
So how do I assess the situation in which we find ourselves today? Firstly, I feel that confidence in the future and willingness to take risks appear to have recovered from the shock of 11 September. But what will happen to inflation? There are three reasons why inflation should fall this year. The first is that growth in demand and utilisation of resources were both lower than expected last year, which subdues price pressure for the near future. The second is that inflation expectations are disposed towards a decline in inflation, according to recent surveys. And the third is that the krona has appreciated slightly, which means a lower risk of imported inflation. On top of this, the effects of last year's price rises on certain goods and services, including food and energy, will gradually abate.
However, there is also a risk that the price rises will not abate on their own. I can see, unfortunately, a couple of reasons why inflation may remain higher than we have hoped and why the price increases might continue to spread.
Firstly, only half of the increase in the inflation rate could be explained by temporary effects. Many prices have risen more than expected. This is probably due to resource utilisation in the economy being relatively high as demand was sufficiently high to enable implementation of price increases. The fact that competition is weak in certain sectors probably reinforces this, which facilitates passing on the costs to the consumers. A further sign is that the increase in employment has been good. There has been an adjustment downwards in working hours rather than in the number of jobs. The division on the labour market, with increases in employment in sectors less exposed to competition has meant that wages have risen more than expected. Thanks to expansive fiscal and monetary policy, unemployment is expected to rise only marginally. As economic activity appears to be increasing during 2002, it is difficult to imagine that the pressure on the labour market will decline significantly.
Secondly, my personal opinion is that it is possible that inflation expectations, reflected as an acceptance of price rises, may have increased. I believe that the temporary price rises that have occurred have clouded our view and it has been possible to undertake more price rises under cover of these. There have been reports in the media, for instance, on importers who can no longer withstand the cost increases caused by the weak exchange rate. Mad cow disease and foot and mouth disease have made us get used to the idea of higher meat prices. Poor harvests, worm-eaten apples, etc. have meant that the resistance to higher fruit and vegetable prices is weak. The fact that certain trade union representatives are also saying in the media that they may have to break wage agreements if inflation remains high, indicates that they do not fully rely on price stability being maintained. The implicit inflation expectations, which can be seen in the pricing in the interest rate market, also show that inflation expectations two years ahead are rather high.
It is, of course, worrying that inflation is higher than expected. The lowering of the repo rate by 0.5 percentage points in September was partly to contribute to reducing uncertainty and raising confidence in the future in an unusually uncertain situation. However, the future prospects are no longer as uncertain and the economic slowdown appears to be in its final stages. There is therefore no longer a need for a large insurance against a severe decline.
It is important to react quickly with interest rate policy when there are downside risks, but it is equally important to react quickly to upside risks. As can be seen in the minutes of the Executive Board meeting on 7 February 2002, I entered a reservation against the decision to leave the interest rate unchanged. I would rather have seen an increase in the interest rate of 0.25 percentage points, which is not any severe tightening of policy, but should rather be seen as not stepping so hard on the gas.
