Monetary policy and the exchange rate
Since November 1992, when the defence of the krona's fixed exchange rate was abandoned, much of the discussion concerning monetary policy has involved the exchange rate. At this traditional meeting with the Riksbank's primary dealers, I should like to consider four issues connected with Sweden's currency.
One of these issues is the function of the krona in a monetary policy regime that targets inflation with a flexible exchange rate. Another issue is the actual path of the exchange rate in recent years and the conclusions it warrants. The euro's introduction at the beginning of January has revived the question of Sweden's participation in EMU; this raises the other two issues, namely the process towards ERM membership, if this is envisaged, and aspects of setting the future ERM rate that would then ensue.
1. Monetary policy significance of the exchange rate
The floating exchange rate regime that was introduced for the Swedish krona in November 1992 was regarded by many people as a fairly temporary arrangement. They considered that for Sweden, as a small country dependent on foreign trade, a fixed exchange rate was most advantageous, so the Riksbank ought to try to reintroduce a fixed rate regime as soon as possible.
Since its introduction in 1993, however, the inflation target has proved to be a good compass for achieving and maintaining price stability. Confidence in the monetary policy regime has gradually grown and now seems to be firm. Inflation and inflation expectations are low. Since the cyclical upturn in the summer of 1993, annual GDP growth has averaged almost 3 per cent, government finances have been rapidly consolidated and Sweden has begun to repay external debt.
Important forecasting variable
During lengthy periods of financial unrest and problems with confidence in Sweden's economy the Riksbank judged that the krona was undervalued. At times, moreover, the exchange rate fluctuations have been relatively marked, even after it seemed that economic policy's credibility had been established.
In the period since the fixed exchange rate was abandoned the Riksbank has not targeted the exchange rate. For monetary policy, the primary consideration since 1992 has been the picture of inflation prospects in relation to the targeted rate of 2 per cent. But the exchange rate is still one important factor in the Riksbank's assessment of inflation. An exchange rate movement that is judged to have consequences for inflation prospects will accordingly influence the formation of monetary policy, ultimately the level at which the Riksbank sets the repo rate.
The exchange rate normally affects inflation through import prices as well as via foreign trade and resource utilisation. All else equal, a weak exchange rate will lead in theory to higher import prices and thereby to higher consumer prices. In practice, however, the weak krona has not affected either import or consumer prices as much as the Riksbank had anticipated. There may be a number of explanations for this.
- One may be that the Riksbank has overestimated resource utilisation and hence demand pressure in the economy. A weaker demand situation probably makes it less possible for firms to let the weak exchange rate pass through to consumer prices.
- Another factor may be downward price pressure from the increased international competition. Although the price effects of this are presumably of a one-off nature in principle, they may affect inflation over a long period in that it takes time for them to work through the economy.
- A further explanation may be that the move to a flexible exchange rate has altered pricing behaviour. When the krona was devalued in the Seventies and Eighties, the new exchange rate was perceived as permanent and prices were set in relation to its weaker level. Since 1992, many of the krona's depreciations have probably been perceived as transient and therefore resulted in more limited price adjustments.
Not a mechanical relationship
The relationship between the exchange rate and inflation prospects has evidently changed over time. No straightforward mechanical link can be said to exist between the exchange rate, inflation prospects and the repo rate. The monetary policy response to an exchange rate shift also depends on the underlying cause, the perceived duration of the shift and the economic situation in other respects.
- If the krona weakens on account of portfolio shifts that stem in turn from fears of rising inflation, for example, it is natural for this to elicit a tighter monetary stance. That was in fact a major argument for the Riksbank's policy from 1993 to 1995.
- If, on the other hand, the exchange rate weakens at a time when overall inflationary pressure is low, for example in connection with declining international demand, this does not necessarily imply a threat from inflation. The weaker exchange rate is then rather a part of a real economic adjustment. Such mechanisms were presumably involved last autumn.
Monetary policy therefore has to be based on an assessment of inflation prospects that is as comprehensive as circumstances permit. The exchange rate is one important factor in this assessment but normally it should not play an isolated role. That is why the Riksbank has chosen not to use a monetary index—a weighted combination of interest rates and the exchange rate—as an intermediate target.
The Riksbank's view of the exchange rate has altered a good deal since 1992. We have gradually placed more emphasis on the krona's forecast path and been less concerned about its occasional fluctuations. As I see it, a number of things have contributed to this. It has proved difficult to show that attempts at short-run exchange rate stabilisation, using either the repo rate or interventions, have been particularly successful. Neither does experience from other countries with a flexible exchange rate provide all that much support for such a policy. At the same time, increased credibility has made the gradual modification of the Riksbank's view easier.
The Riksbank has accordingly chosen to be comparatively inactive in the currency market. At the same time, an exchange rate that is highly volatile and displays pronounced fluctuations clearly has drawbacks; the uncertainty it generates is liable to spread to other financial markets, besides complicating corporate and personal decisions about investment, production and consumption. In recent years the Riksbank has therefore occasionally chosen to intervene in the foreign exchange market in order to underscore its appraisal of the situation. It need hardly be said that if such interventions had been judged to conflict with the inflation target, they would never have been considered.
As long as the Riksbank targets inflation directly, I see no reason to change the current principles for our behaviour in the currency market. But I also want to underscore that the approach to the exchange rate in an inflation targeting strategy is by no means self-evident. Some central banks that target inflation—the Bank of Canada is one example—seem to react to short-run exchange rate fluctuations more frequently, sometimes guided by a monetary index, partly with the aim of stabilising markets and expectations. Their behaviour in this respect is more reminiscent of how the Riksbank acted in the early stage of the flexible exchange rate regime. The operations of other central banks, in the UK and Australia, for instance, are more akin to the Riksbank's current practice. Comparisons of exchange rate fluctuations in different countries with a flexible exchange rate provide little evidence of any clear effects from differences in the monetary policy strategies. The Canadian dollar, for example, does not seem to have been more stable even though the Bank of Canada has focused more on the monetary conditions and made greater use of interventions.
2. Forces behind the flexibles exchange rate
At the time of the departure from the krona's fixed exchange rate in November 1992, the real trade-weighted exchange rate was overvalued. Inflation in Sweden had been higher than elsewhere for a considerable period and this had successively eroded competitiveness.
When the flexible regime was introduced, the krona depreciated markedly; in the first six months the nominal TCW index weakened about 20 per cent. In that inflation in Sweden had already been brought more or less into line with the rest of the world and there were signs of expectations that it would remain relatively low, such a marked nominal depreciation exceeded what was needed to restore a reasonable competitive situation. Having previously been overvalued, the real exchange rate accordingly swung abruptly to an undervalued level and remained there for a long time.
Other currencies that had been obliged to float in connection with the crisis in 1992—the Finnish markka and the Italian lira—also weakened more than was motivated by long-term considerations.
Gradual stabilisation
A major factor behind the krona's initial depreciation was no doubt uncertainty about the Swedish economy and economic policy. The grave problems with government finances, together with uncertainty about inflation's future path, contributed to this.
In the early years with the flexible exchange rate the path of the krona was less susceptible to current inflation or demand. A simple regression analysis shows that the confidence of market agents in Swedish economic policy and price stability—measured as the long-term interest rate differential with Germany—can explain almost a third of the nominal exchange rate's variability since 1992. In the period from 1994 to 1996 the swings in credibility, measured in this way, actually explain almost two-thirds of the exchange rate movements.
In this uncertain environment, with a lack of credibility, the impact of international turbulence on Swedish interest rates and the exchange rate was accentuated. The nervous mood in international financial markets in connection with the Federal Reserve's monetary tightening in 1994 and the crisis in Mexico at the end of that year added to the krona's depreciation and widened interest rate differentials. A similar pattern was observed in other countries with a history of high inflation and problems with government finances—all the high-interest countries were treated in much the same way.
The situation in Sweden's financial markets then improved as the measures for consolidating government finances began to bite, prospects for Sweden's economy brightened and international developments became more stable. During 1995 the Swedish krona became markedly stronger and long-term interest rates declined. This trend continued during 1996, when the Riksbank rapidly lowered the repo rate. At the end of 1996 the level of the TCW exchange rate index was 114, which was inside the Riksbank's estimated interval for long-term equilibrium.
For about two years, from the summer of 1996 to the summer of 1998, the krona stayed inside a relatively narrow interval in relation to the German mark, about 3 per cent around a mean level of 4,40. In effective TCW terms, however, the krona weakened. There are good reasons for supposing that this was due to other factors than problems with credibility. The link between exchange rate movements and the long-term interest rate differential was no longer discernible, which indicates that the krona's path now had less to do with aspects of credibility. The main explanation for the weaker exchange rate in this period lay rather in cyclical differences. Stronger economic growth in the UK and the United States caused the dollar and sterling to appreciate against the krona.
Turbulent autumn
In the autumn of 1998 the krona weakened again, mainly, in the Riksbank's opinion, on account of international turbulence. Growing financial problems during 1998, culminating after Russia's suspension of debt payments and the near failure of an American hedge fund, increased the risk aversion of international investors and prompted a flight to more secure assets. This hit the krona even though the economic fundamentals in Sweden were mostly good. One reason why the general increase in risk premia weakened the krona was probably the persistently high level of external debt. Other countries with large foreign debts were similarly affected. In Sweden's case, however, some part was also played by domestic factors; the krona tended to weaken, for example, at the time of the general election in September.
When the financial market turbulence then subsided around the turn of last year, the krona strengthened once more. This suggested that low inflation and the rapid consolidation of government finances were again playing a relatively larger part in exchange rate movements. The stronger krona also had to do with political signals early this year of Sweden's adherence to the monetary union. Market players believed that the euro conversion rate for the krona would be stronger than the current market rate and that reinforced the krona's appreciation.
During the discussion about the krona's depreciation during 1998 there were two arguments that call for some comments.
One argument was that the krona weakened relative to the German mark on account of structural problems in Sweden. The Swedish economy is unquestionably beset by problems of a structural nature and it is not inconceivable that a link exists between the exchange rate and some of these problems, for example the structure of taxation. If it is true that the lack of industrial renewal which seems to be discernible in the statistics from the Seventies and Eighties still applies, this is another factor that can affect the exchange rate. The discussion would benefit, however, from more specific arguments and a thorough analysis. For one thing, it is by no means clear that the structural problems in Sweden are so much greater than those in Germany, for example, that this could account for the krona's depreciation against the German mark. For another, it seems remarkable that long-term problems of this kind would affect the exchange rate so suddenly.
Another argument ascribes the weakening of the krona to discrepancies between Sweden's inflation target and the monetary policy strategy of the ECB. This, too, seems unlikely. The monetary policy differences seem to be slight. Since the beginning of the Nineties, moreover, inflation in Sweden has been below the EU average; in recent years this situation has been accentuated, with Sweden as one of the countries with the lowest inflation in Europe.
In conclusion, it can be noted that for much of the period with a flexible exchange rate the krona has been weaker than was considered desirable. In the early years this was mainly due to a lack of confidence in economic policy. In the period since the krona's recovery in 1995 to '96, however, the exchange rate has been comparatively stable, except during the widespread turbulence in connection with last autumn's international financial crisis. But even with today's relatively strong government finances and low inflation, a flexible exchange rate does seem to entail exchange rate fluctuations that are greater than many of us believed when the krona fell in the autumn of 1992. This is an experience we share with other countries that target inflation in a flexible exchange rate regime, for example the UK, Canada and Australia.
3. Sweden's participation in ERM2
The introduction of the euro has revived the question of Sweden's adherence to the economic and monetary union. Political statements and a more positive opinion at times have led some observers to talk of the possibility of Sweden joining the euro area as early as from 1 January 2002, after a Social Democrat Congress next spring followed by a referendum later in the year.
This in turn has led to talk of Sweden joining the new European exchange rate mechanism ERM2. The matter is being increasingly discussed in market letters and the media and that is why I want to consider some of its aspects with you this evening.
The Riksdag (Sweden's parliament) has already stated that membership of ERM2 can be regarded as a direct preparatory step before entry into the monetary union. The Government may therefore raise the question of membership of the fixed exchange rate mechanism when future participation in the monetary union is judged to be imminent.
Government decides exchange rate system, the Riksbank its application
As regards ERM2, what are the Government and the Riksbank, respectively, responsible for? The Riksbank Act, amended as of 1999, states that the exchange rate system is decided by the Government and applied by the Riksbank. This is the same arrangement as in virtually the whole of Europe. The decision as to whether and when the krona is to be linked to ERM2 is thus a matter for the Government. The Riksbank in turn determines how the exchange rate system chosen by the Government is to be applied, which involves setting the central rate and band width in a fixed exchange rate system and deciding any interventions.
In practice, if Sweden is to participate in ERM2, the law requires the involvement of both the Government and the Riksbank. They must also agree on the conditions. Moreover, ERM2 membership would not be decided unilaterally by Sweden; it is something that has to be negotiated with the other EU countries.
In the event of Sweden wishing to join ERM2, what would be the procedure? The negotiations would be conducted in the Economic and Financial Committee, which has succeeded the Monetary Committee. Each country is represented there at the level of under-secretary in the Finance Ministry and central bank deputy governor.
Experience shows that in the process of arriving at a reasonable central rate, the factors which are considered include the country's macroeconomic situation, various indicators of its competitive position, the exchange rate's historical path and its average levels in recent years. When other currencies have been linked to the exchange rate mechanism, this has usually been done either at or fairly close to the prevailing market rate.
ERM membership could stabilise exchange rate expectations
A factor of importance for participation in the euro area is the krona's stability. Membership of ERM2 could also facilitate Sweden's entrance to the monetary union.
It may therefore be appropriate to say something about the timing of an adherence to ERM2. This, as I mentioned, is a matter for the Government to decide. In the Spring Budget the Government has declared that participation in ERM2 is not being considered at present.
The choice of an entry date presumably depends in part on the duration of ERM membership that other countries expect from Sweden before qualifying for participation in the euro area. Another central consideration is whether ERM membership as such can contribute to exchange rate stability and smooth the path to the euro area. In my opinion, ERM2 membership could help to stabilise exchange rate expectations if and when it is clear that Sweden does intend to join the euro area. That in turn would simplify the krona's conversion into euro. At the same time, I believe this would presuppose a reasonably balanced economic situation in Sweden, with strong confidence in economic policy, as well as broad political support for ERM2 membership.
While there may be some cause for concern about conflicts between an exchange rate target in ERM2 and the direct inflation target, this risk should not be exaggerated. The broad bands for the exchange rate mechanism, together with experience in other countries in the run-up to the monetary union, show that no serious conflicts need arise.
If market players are confident that the central rate in ERM2 will be chosen as the conversion rate for participation in the euro area—as they have every reason to be in that this was the case for each of the eleven countries that currently make up the monetary union—then the krona's market rate is unlikely to deviate all that much from its central rate. This was clearly the case for the currencies that have already been converted into euro. Once it was clear, for example, that Finland would be joining the euro area, the markka remained stable and very close to its central rate.
In practice this would mean that during ERM membership, Swedish interest rates with relatively long maturities would converge on the levels that apply in the euro area. As the date for joining the union approaches, the very short market rates would be increasingly affected. This would successively reduce the Riksbank's possibilities of influencing domestic inflation prospects with the repo rate.
4. The SEK/EUR exchange rate
Another item in the discussion in financial markets is the rate at which the krona would be converted if Sweden were to join the euro area. There may therefore be reason to also say something about this.
Arriving at an appropriate conversion rate is not a simple matter. It involves trying to pin down many complex economic relationships and assessing the future long-term paths of various central variables.
Estimates of the long-term equilibrium exchange rate
For several years now we have been engaged at the Riksbank in analysing the krona's fundamental value and the likely path of the exchange rate in the somewhat longer term. This work has been needed quite apart from the question of future ERM membership, for instance as a contribution to assessments of the exchange rate's importance for inflation prospects. The study results provide information about the direction and approximate strength of the krona's future trend.
An assessment of the appropriate value for the krona in the event of ERM membership can start by considering what the exchange rate would be if the Swedish economy were in macroeconomic equilibrium. This is a situation were resources are fully utilised without generating higher inflation and the level of external debt is such that the economy can cope with it in the longer run. Some years ago we published estimates of the exchange rate in this equilibrium; we also indicated an interval for the real TCW exchange rate index that has subsequently been cited by many journalists and market players. Estimates of that type are not, however, a direct guide to the krona exchange rate that would be appropriate for ERM2 membership. Other factors that have to be considered include the economic situation in Sweden and the rest of the world. Moreover, the estimates are expressed in terms of the TCW index, which covers all the rates for Sweden's foreign trade, not just the euro area. The assessment must also allow for the fact that, at least in the coming years, the Swedish economy will not be in long-term equilibrium, which implies both internal and external balance. It should also be borne in mind that however sophisticated, calculations of this type do not yield precise results.
In recent months the TCW index has fluctuated between 124 and 125. This is some way from the estimated long-term equilibrium range. The main conclusion from our earlier studies—that in the longer run it is reasonable to expect the krona to strengthen—still holds. But it does seem reasonable to count on this being a gradual process that emerges as the economy approaches a situation with both internal and external balance. On various occasions in recent years it has turned out that we were over-optimistic about the adjustment's future rate. Perhaps we have also made less than full allowance for the imbalances that remain even though the economy is developing well in the right direction. The Swedish economy still has unutilised resources and external debt is both high and above the level in many other countries.
The major part of Sweden's external debt accumulated in the Seventies and Eighties in connection with large and continuous current-account deficits. Such deficits need not be a problem if, for example, they stem from high investment. In those decades, however, they were mainly generated by unduly high consumption. As a result, a proportion of our gross domestic product has to be reserved today for servicing foreign debt. The economy's capital stock and production capacity are not larger than they would have been without the debt burden. Neither can the current-account deficits and low saving be attributed to demographic factors.
Since the krona's depreciation in 1992, the current account has shown sizeable surpluses. They reflect the combination of good export growth and a markedly subdued increase in domestic demand; the latter came in turn from a series of years with high interest rates and the extensive measures to consolidate government finances. The financial surpluses are reasonable in view of the large interest burden from high external debt and the circumstance that some way into the next century the economically active proportion of Sweden's population will be shrinking. The adjustment of external debt will no doubt require continued current-account surpluses in the years ahead, which presupposes that the krona is weaker than its long-term equilibrium level. Moreover, the domestic demand trend needs to be balanced to prevent it from generating unduly high inflation. As central government debt is sizeable, it is also important that fiscal restraint continues, with the focus on debt repayment.
However, a continuation of current-account surpluses of the magnitude we have seen in recent years is not to be expected. As domestic demand strengthens, the economy approaches full capacity utilisation and the monetary stance accordingly becomes less expansionary, the krona should appreciate, accompanied by diminishing surpluses on the current account.
You would, of course, like to know the rate at which the krona would be linked to the euro in the event of Sweden opting to join the monetary union. Unfortunately it is still too early to answer that question. The rate will depend, for example, on Sweden's competitive position relative to the countries that have already adopted the euro. We would also have reason to aim for a central rate that puts the Swedish economy on the right track and promotes a continued adjustment to internal and external balance. The optimum central rate might differ somewhat—though probably not all that much—from the market rate in the event of Sweden joining ERM2 if, for instance, Sweden's monetary policy and cyclical position were to deviate from the euro countries' prior to Swedish actually joining the monetary union. Such a difference would also be motivated if the market rate were to be affected by short-term shocks in financial markets.
In the light of what we know today, however, the direction of the krona's path seems to be clear: ERM2 membership, if it occurs, will be arranged at an exchange rate that is stronger than the recent level.
