Heikensten: Inflation and monetary policy
Thank you for the invitation to talk here about Sweden's economy and monetary policy. First I shall say something about the principles for a monetary policy that target's inflation. I shall then consider the Riksbank's current assessment of inflation's future course and the implications of this for the coming monetary stance. After that and partly in view of the major share price increases in recent years, I shall consider the significance of asset prices for monetary policy and then round off with some conclusions for monetary policy.
1. Monetary policy with an inflation target
The construction of Swedish monetary policy, based on the objective of price stability, is entrusted to the Riksbank. More precisely, the aim is to keep the annual change in the CPI at 2 per cent. Deviations of ±1 percentage point from this target can be accepted, however.
By avoiding excessively rapid changes in the general price level, monetary policy can help the economy to function smoothly. A stable price level makes it easier for economic agents—consumers, savers and investors—to reach decisions about consumption and investment. A price increase for a certain product then signals a shortage of that particular item rather than general inflation. Price stability accordingly creates favourable conditions for a reasonable allocation of resources and good economic growth. It also reduces the risk of an arbitrary redistribution of income and wealth.
The target variable for monetary policy is thus the overall change in consumer prices. It takes time, however, for a monetary policy measure to affect the economy and this variable. This monetary policy time lag is commonly considered to be one to two years.
Monetary policy decisions by the Riksbank therefore have to be based, not on the most recent figure for inflation but on assessments of how inflation is likely to develop in the coming twelve to twenty-four months.
The Riksbank's assessment of inflation in the Swedish economy is a continuous process and the assessments are published in the quarterly inflation reports. A broad selection of indicators provides information about inflationary pressure in the economy and the implications of this for the monetary stance.
Elements of risk in inflation's future path are also assessed. Even when the assessment of future inflation is unchanged, increased uncertainty can warrant greater caution in the implementation of monetary policy. The Riksbank may then defer interest rate adjustments until more is known about the situation. This tends to reduce movements in market rates of interest, besides benefiting economic development in general.
2. The Riksbank's assessment of inflation
This brings me to the Riksbank's current assessment of future inflation. Will the price stability target be met in twelve to twenty-four months time without this calling for a repo rate adjustment? The assessment I shall be presenting today is based in the main on the inflation report the Riksbank published three weeks ago. Our assessment in that report concerned the period from the second quarter of 1998 to the end of the first quarter of 2000.
Broadly speaking, inflation is conditioned by the following factors: developments in the rest of the world and their impact on the Swedish economy, for instance via the exchange rate; the relationship between domestic supply and demand; other cost shocks and transitory effects; and expectations of future inflation. I shall now present our view on each of these in turn.
International activity and inflation
Developments in the rest of the world affect price movements in Sweden via import prices as well as through their impact on the demand situation in general. The degree to which external price movements show up in prices expressed in Swedish kronor is dependent both on how the exchange rate develops and on the extent to which importers pass on external price changes in their Swedish prices. Moreover, world market price movements can alter the competitive position of Swedish exports and thereby affect the domestic rate of wage increases and other production costs.
A highly topical issue in the international discussion is, of course, the prospects in Asia. The effects of the Asian crisis and its global economic repercussions are difficult to gauge but our working assumption is that, after very weak growth this year, the Asian economies will recover by degrees. In that case, international development in the coming years should be relatively good, for instance because activity in Western Europe is becoming increasingly strong. Even so, external inflationary pressure is judged to be moderate. Downward effects on primary product prices, for example from the Asian crisis, as well as on other product prices, are contributing to this.
Exchange rate
The Swedish krona's exchange rate is an important consideration in the Riksbank's assessment of price stability. In the past six months the krona has tended to weaken, though the opposite has been the case since the middle of February. A depreciation that becomes permanent affects inflation prospects and thereby the construction of monetary policy.
In this context I should like to elaborate on the Riksbank's appraisal of the Swedish krona. In the recent debate it has been suggested that the Riksbank's exchange rate policy includes a target rate for the krona. That is a misunderstanding. The Riksbank does not target the krona's exchange rate. But as one of the factors behind inflation, the exchange rate is important for monetary policy in a flexible exchange rate regime. A considerable period with a weak exchange rate might lead to a forecast rate of inflation that exceeds the target, in which case the Riksbank has to respond in order to meet this target. Similarly, the Riksbank would have to take appropriate measures if the krona becomes permanently strong.
But a stable exchange rate is, of course, fundamentally desirable. It makes economic decisions easier for everyone and, all else equal, should thereby promote higher growth. Moreover, Sweden is a part of a wider European context. As a member of the European Union, we have undertaken to treat exchange rate policy as a matter of common interest. There is also the possibility of Sweden joining the euro area; in its decision last year on EMU, the Riksdag (Sweden's Parliament) made it clear that Sweden might then participate in the exchange rate mechanism (ERM). In this perspective, reasonable exchange rate stability would be an advantage.
Today there are many indications that the single monetary policy in EMU will be conducted with an inflation target that is very close to Sweden's, that is, 2 per cent.
Under such circumstances there would be a good possibility of the exchange rate between the Swedish krona and the euro being stable.
Demand relative to supply
A level of demand that exceeds long-term production capacity and supply will generate inflationary pressure. Monetary policy therefore has to counter tendencies towards such an imbalance. The Riksbank uses a variety of indicators to form an overall picture of how future inflation is likely to be affected by demand in relation to supply.
In the inflation report the Riksbank considered that economic growth in Sweden will be somewhat lower than expected earlier. But the profile is unchanged, with a rising trend that is generated to an increasing extent by domestic demand. GDP growth is expected to be just over 2.5 per cent this year and about 3 per cent in 1999. If our prediction of a relatively favourable development of international activity is correct, growth in 2000 should also be good.
The accelerating growth of domestic demand stems from increased private consumption as well as rising industrial investment. A continued increase in private consumption in both 1998 and 1999 is indicated by a number of factors. A major addition to household wealth has resulted from rising share prices in recent years and, more recently, from rising property prices. I shall be discussing this further on. Moreover, in 1998 and 1999 fiscal policy will become gradually less restrictive. A marked increase in household disposable income is therefore likely in 1998, for the first time since 1994. Households are also positive about their own economic situation, though these expectations have become somewhat more subdued. The dual situation in the Swedish economy—a strong export sector but weak domestic demand—should therefore go on receding.
Since the publication of the inflation report, some new statistics that throw light on cyclical activity have been presented, including the national accounts for the 4th quarter of 1997. The overall picture supports the Riksbank's assessment in the inflation report: activity is rising but the tendency is not dramatic.
Other cost shocks and transitory effects
Inflationary impulses can also be generated by increases in specific costs that are unconnected with either increased international inflation or high domestic demand. Rising primary product prices are one example. Similar inflationary impulses can come from fiscal policy in the form of changes to indirect taxes and subsidies. While these kinds of inflationary impulse are normally transitory, they can influence inflation expectations and thereby alter the long-term inflation process. In order to gauge trend inflation (inflation excluding transitory impulses) and thereby arrive at a deeper understanding of the inflation process, the Riksbank uses various indicators of underlying inflation, for example.
Transitory effects have played a prominent part in CPI changes in recent years. The drop in interest rates has had a considerable downward effect on CPI inflation. During 1997 this was countered by the tobacco tax, for example. In the inflation report it was estimated that in the coming twenty-four months, changes in taxes and interest rates will largely cancel out. In other words, transitory effects would have no sizeable impact on the CPI. The downward tendency in primary product prices this winter is of interest in the present context. To some extent, at least, it is presumably a temporary phenomenon.
Inflation expectations
High demand prompts producers to raise prices and employees to bargain for wage increases. But inflationary price and wage increases can also stem from high inflation expectations as such, in that economic agents strive to maintain or improve their real income. In this way, inflation expectations are liable to be self-fulfilling. They are therefore a central feature of the price formation process. In Sweden, the clearest example is the wage-price spiral in the 1970s and '80s.
The overall picture of inflation expectations shows a low level that is relatively well in line with the inflation target. According to most indicators—implied forward interest rates as well as data from surveys and forecasters—inflation expectations have recently tended to move downwards. A low-inflation regime is being established.
Inflation forecast
Rising economic activity is foreseen in the coming twenty-four months. In the latest inflation report we therefore considered that during 1998 and 1999 the rate of inflation will move up from current levels, though the increase was expected to be slight, partly as a consequence of good productivity growth. Annual CPI inflation was judged to be 1.6 per cent this year and 2.1 per cent in 1999.
In the light of the February inflation figures and the development of oil prices, it is conceivable that CPI inflation in 1998 will be somewhat lower than envisaged in the inflation report. To date we have seen a limited pass-through from the oil price fall to consumer prices. At the same time, the level at which oil prices will stabilise is uncertain. Moreover, underlying domestic inflation has been unaffected by the lower rate of CPI inflation.
Although the lower CPI change for 1998 is expected to be mainly a result of transitory effects, the fact remains that in recent years the rate of price increases in Sweden as well as abroad has been below the anticipated level. There are many explanations for this. For Sweden, one of the most important is the transition to a low-inflation regime, with all that this implies for price setting, lower inflation expectations and so on. Structural reforms and increased competition have certainly contributed, too.
3. Monetary policy and asset prices
As I mentioned earlier, rising asset prices are expected to contribute to increased domestic demand. This price trend has been pronounced. Low and stable consumer price increases in most western industrialised countries have been accompanied by rapidly rising asset prices. In Sweden in the 1990s, consumer prices have risen 30 per cent, while share prices have moved up about 170 per cent. The increase in share prices calls for a combination of very good economic development and persistently low interest rates.
This development poses questions for monetary policy. Do the rising asset prices imply that inflationary pressure is adopting a different guise from rising consumer prices? Or do rising asset prices entail other risks? And how should monetary policy respond to these changes?
The wealth effects from rising asset prices and their impact in turn on demand and inflation have been cited as a reason for paying more consideration to asset prices in the construction of monetary policy. Another reason is the financial instability that may be caused by exaggerated price increases. Here, then, is a closer look at these relationships.
Effects on demand
Consumer prices can be affected by asset price increases via consumption as well as investment. Appreciating assets imply increased household wealth and this may lead to higher consumption. Rising share prices can also lead to increased investment in that it becomes cheaper for firms to raise funds.
Experience shows that, at least in the long term, household behaviour is more responsive than business investment to rising asset prices. It seems reasonable that in the longer run, households regard rising share prices as a real increment to wealth that adds to their lifetime income and accordingly provides more scope for consumption.
In the short run, however, the relationship seems to be weaker for several reasons.
One is that of the share-owning households, a large proportion have relatively high incomes and considerable wealth. Their marginal consumption propensity is relatively low; they invest for the long term and are probably less concerned about short-run price fluctuations. Another reason is the large proportion of share holdings that represent saving for retirement. The long-term element in saving accordingly exists in other groups besides the high-income categories.
Effects on financial stability
Another argument that has been heard for giving asset prices a more prominent role in the construction of monetary policy is the potential effect of these prices on the financial system's stability. The bank crisis in Sweden in the early 1990s and the crisis in Asia are two clear examples of how the bank system can be affected when a financial bubble bursts, though in those cases the exposure was primarily to the real estate market.
Such crises are liable to exact massive social costs, which is a reason for central banks to move at an early stage in order to prevent such exaggerated price movements.
There are other reasons, however, which suggest that a central bank should refrain from deploying monetary policy at an early stage to counter asset price increases.
Above all, of course, there is the difficulty in telling which price tendencies are, in fact, exaggerated to the point of constituting a price bubble. There are usually some fundamental factors which can be seen as plausible causes of the price rise.
Determining just how much of an upward price movement is warranted in such a situation is extremely difficult.
Moreover, the argument for an early response to share price bubbles is weaker than in the case of property prices. As I mentioned, in most countries banks are more exposed to risks in the real estate market than they are to stock market risks. A stock exchange crash should then have less effect on financial system stability and thereby on economic activity in the medium term.
Asset prices and monetary policy
The issue of how monetary policy should handle and respond to changes in asset prices is inevitably complex.
The Riksbank's current target for price stability is the CPI, which does not include prices of financial and real assets like shares and property. The Riksbank certainly monitors the development of asset prices closely both to arrive at a general assessment of inflationary pressure and to detect any risks of financial instability. But Swedish monetary policy does not respond directly to rising asset prices to the same extent as it does to price increases for goods and services, for instance.
Difficulties arise, however, should the Riksbank respond to asset prices more directly. One major problem is that movements in these prices are not necessarily an indication of future inflation, so that it is difficult for central banks to identify the situations in which a policy response is called for. But one can always contribute to a broader discussion of asset prices and highlight the risks associated with a valuation of shares and real estate which presupposes that future economic development is highly favourable.
4. Conclusion
In the assessment of monetary policy's construction the Riksbank looks ahead to estimate the path of inflation in the coming twenty-four months. Our objective is price stability.
In the latest inflation report we still considered that in the coming twenty-four months the rate of inflation would tend to rise but somewhat more slowly than we had expected earlier. The uncertainty in the assessment led to the conclusion that the Riksbank should be alert to future tendencies. The prospects pointed to a cautious tightening of the monetary stance in the coming twelve months.
New statistics have been published since the inflation report was presented but they do not entail any substantial changes in our outlook. The new information on inflation expectations and wage agreements confirms our earlier assessment. The exchange rate tendency—with some appreciation—is also more in line with our prediction in the inflation report. The same applies to the new information about cyclical activity.
In some respects, however, the picture is weaker than expected. International inflation is low and CPI inflation in Sweden in February was lower than anticipated earlier. The latter was mainly a result of lower prices for clothing and footwear together with some fall in the long bond rate. In addition, the fall in primary product prices may have some effect on the CPI figures for coming months. However, this is no doubt largely a temporary phenomenon and to date it has had little effect on underlying domestic inflation.
In conclusion, then, inflationary pressure in the short run may be marginally lower than we considered in the inflation report. But the statistics do mostly support the assessment of the somewhat longer term.
The conclusion in the latest inflation report was that the repo rate should not be altered at present but that a cautious tightening would probably be considered in the year ahead. The information that has been obtained since the publication of the inflation report does not alter that conclusion.
