Persson: Monetary policy yesterday, today and tomorrow
Thank you for the invitation to come to Aktiespararna Lidingö this evening to talk about monetary policy, which is for me a job and – I assume – for you an interest, as you save in shares. It is of course important for those investing in companies to obtain a good (risk-adjusted) return on their investment. Monetary policy that contributes to low and stable inflation leads to less uncertainty, clearer game rules and better conditions for growth. I intend to begin with a look back over monetary policy during the past decade and a description of the sequence of events when we changed over to inflation targeting after the economic crisis in the early 1990s. After that I shall briefly discuss the current economic situation, mention the changed conditions I envisage and what they may entail for monetary policy in the future.
From economic crisis to 10 years of inflation targeting
Many of you will remember and will have been affected in some way by the economic crisis at the beginning of the 1990s. That was when the foundations of the current monetary policy were laid; when the fixed exchange rate was abandoned as a monetary policy objective and replaced by an explicit inflation target. This process can probably be most easily understood in the light of the economic policy conducted during the decades prior to the crisis so let me therefore begin with a brief look back.
The economic policy conducted during the 1970s and 1980s meant that nominal wages and prices rose more rapidly in Sweden than in other countries. At the same time, our aim was to maintain a fixed exchange rate, which assumed that inflation in Sweden had developed at roughly the same rate as in the countries that were our most important trading partners. Price and wage trends led time and time again to cost crises that finally forced Sweden to make devaluations of the krona and undermined the credibility of the fixed exchange rate regime. The exchange rate adjustments were also one of the reasons behind a failure to implement necessary rationalisations and streamlining. Although the devaluations strengthened our competitiveness temporarily, they provided the economy with new inflation impulses and by the time the price and wage spiral had turned around a few times, a new cost crisis had arisen. The fact that developments did not favour real growth in Sweden or real wages becomes clear in comparisons with both other time periods and other countries, such as the Nordic countries, Germany and Austria (Figure 1).
When the fixed exchange rate target was abandoned in November 1992, after an unsuccessful attempt to defend the krona – you probably all remember the increase in the marginal rate to 500 per cent – it followed on from yet another cost crisis. At the beginning of 1993, the outlook for the Swedish economy was bleak - we were in the midst of an international economic slowdown, the banking crisis was a fact, interest rates were high and unemployment was soaring. During the period with a fixed exchange rate fiscal policy had borne the entire responsibility for stabilisation policy, while monetary policy had been tied to defending the krona rate. It was evident that fiscal policy had not been sufficiently tightened in connection with the overheating to which credit deregulation contributed. It was high time to break with history – but how should this be done?
A new stabilisation policy regime
A comprehensive change in regime was implemented in order to take Sweden out of the crisis and to create the conditions for a more stable and more favourable economic development. Central government finances were consolidated and fiscal policy was subjected to regulations regarding an expenditure ceiling and surplus target to attain a long-term balance in public finances. Monetary policy was aimed at safeguarding price stability. The interest rate, which had been a tool for balancing currency flows during the period with an exchange rate target, would now be used to affect demand in the economy with the aim of stabilising price trends.
The new target for monetary policy, to aim for a low and stable inflation rate, was announced in January 1993 and applied in terms of the consumer price index (CPI) with effect from 1995. The fact that there was some delay was partly due to the economy having just been exposed to strong inflation impulses, mainly due to the weakening of the krona after the fixed exchange rate was abandoned.
The target for monetary policy, which was further manifested by being written into the Sveriges Riksbank Act in 1999, is to “maintain price stability” and the Riksbank’s interpretation of this is that inflation should be retained around 2 per cent a year, measured as the annual change in the consumer price index (CPI). There is also a tolerance interval of plus/minus 1 percentage point around this target. Maintaining inflation at exactly 2 per cent all of the time would hardly be possible and constantly fine-tuning the rate for this purpose would probably risk reinforcing fluctuations in the economy rather than having a stabilising effect.
When the Riksbank adjusts the key rate, its repo rate, this affects total demand and ultimately also inflation through various “channels” in the economy – the cost of borrowing, incentives to save and exports and imports, via exchange rate changes. It takes time for an interest rate change to have full effect via these channels; one usually estimates that it takes around 1-2 years. Monetary policy is therefore conducted on the basis of this time horizon. By influencing demand in the economy, we try to steer inflation 1-2 years ahead towards the target of 2 per cent.
However, we conduct what is known as flexible inflation targeting. The Riksbank can choose to disregard effects that contribute to inflation deviating temporarily from the target. This occurs on a regular basis with regard to, for instance, indirect taxes, household mortgage interest expenditure, or supply shocks that are assessed as temporary. In practice, monetary policy is now primarily based on a measure of inflation, UND1X, which is adjusted for mortgage interest expenditure and the direct effects of changes in indirect taxes and subsidies.
Moreover, if a shock should contribute to inflation deviating substantially from the target two years ahead, an assessment is made of how quickly it should be returned to target, in order to avoid unnecessary fluctuations in the real economy, that is, growth and employment. The Riksbank published a clarification of monetary policy in 1999, which repeated the original target wording but also made clear how these temporary effects would be managed, as well as substantial deviations from the target.
Why an inflation target?
The main purpose of introducing a price stability target was to break the pattern of inflation shocks, devaluation and poor real growth. If we look further back in time, periods of economic prosperity appear to have been connected with a low, stable inflation rate (Figure 2). Price stability does not only provide better and more secure conditions for making economic policy decisions. A long-term, credible commitment to meet the inflation target also reduces uncertainty when assessing future income, which ultimately determines the (fundamental) value of various types of asset, such as shares. The larger the risk that one will not receive one’s invested funds plus the desired return on the investment, the greater the compensation one requires for investing/lending one’s money. Compare, for instance, with the situation in Sweden during parts of the 1980s (after deregulation) when the system of tax relief on interest, combined with inflation shocks, meant that in principle one could earn money by borrowing while it cost money to save! The conditions for making wise economic decisions will thus become better when the price stability target has been met.
The same reasoning can be applied to wage-setting. During the 1970s and 1980s wage-earners were repeatedly forced to make large demands for nominal wage increases as inflation was “eating up” their purchasing power. Real wage increases were almost non-existent, while nominal wage increases pushed up prices, resulting in recurring cost crises. A low, stable inflation rate provides entirely different conditions for more stable and better developments in real wages.
Monetary policy and the stock market
As I have the privilege of speaking to a group of shareholders, I would like to take the opportunity to make a small digression into the connection between monetary policy and the stock market. I have already mentioned that a low, stable inflation rate creates good conditions for favourable economic growth and long-term developments in the stock market. However, the reverse can also apply: stock market developments also have some significance for the shaping of monetary policy. The price of various assets, such as shares, is important in itself for the dynamics of a country’s growth and inflation trends. If, for instance, share prices rise, this stimulates household consumption and corporate investment, and vice versa. This in turn affects price trends. Thus, the stock market, like the housing market, can affect the Riksbank’s monetary policy decisions.
Since the 1960s and 1970s, Swedes have been saving more and more in shares, both directly and via mutual funds and insurance, and the market has developed considerably during this period. The return on shares has been very good, seen over the past 30 years, but it has also been something of a roller coaster ride. Here I am thinking of periods of overvaluation, which have been followed by sharp price falls, for instance in connection with the financial crisis and the later IT bubble. Despite these periods of falling share prices, the percentage of shares in household wealth has increased, and stock market developments have thus become increasingly important for households’ consumption decisions and for inflation.
Another reason to be interested in the stock market is its significance for efficiency and thereby for growth. An efficient securities market fulfils many important functions, for instance, as information bearer, channel for risk management and as mediator of capital. A stock market that functions in the optimal manner, with rational investors, helps to allocate investment to the projects with the best expected yield. Of course, this requires rational behaviour and active, competent and well-informed investors. The development of the stock market has progressed rapidly and, in addition to the purely technical factors, regulations, better information and education, for instance through associations like the Swedish Shareholders Association, all contribute to creating a more efficient stock market. However, the market does not always function perfectly and it is not difficult to find episodes of herd behaviour and overreaction when pricing has not reflected fundamentals. One example I am thinking of is the IT bubble at the end of the 1990s.
What has happened since the changeover to inflation targeting?
The Riksbank became the 4th central bank in the world to conduct monetary policy with an explicit, low inflation target. First of all was New Zealand, three years ahead of Sweden, followed by Canada and the United Kingdom(1). If one looks back at developments since the changeover to inflation targeting, it must be said that monetary policy has worked well in Sweden. Following an initial period with relatively high interest rates and a tighter policy, the inflation rate has moved downwards and stabilised at a low level (Figure 3).
One important difference compared with the 1970s and 1980s is that credibility has been established for the monetary policy target, which can be said in turn to be a necessary condition for the policy to succeed. The fact that confidence in the inflation targeting policy strengthened fairly rapidly from the mid-1990s onwards is very clear, if one examines inflation expectations during this period (Figure 4). From relatively high levels indicated in the surveys made immediately after the changeover to an inflation target, the expected inflation rate gradually stabilised at around the target of 2 per cent. Nor does it seem to be the case that price stability was attained at the cost of stronger fluctuations in the real economy than before, which was a fear expressed by some prior to the changeover. The fluctuations in the real economy appear to have declined in comparison with the decades prior to the crisis at the beginning of the 1990s, and we have had much more positive developments with regard to real growth and real wages.
Has inflation been too low?
During the period from 1995 to September 2005, CPI inflation has averaged at around 1.1 per cent, which is within our tolerance interval but clearly below the target of 2 per cent. However, the average UND1X inflation rate has been higher, around 1.9 per cent between 1995 and 2004, which is mainly due to interest rates and thereby also mortgage interest expenditure gradually falling during the period. During the past two years the rate of price increase measured as UND1X has been lower, on average around 0.9 per cent. There are several reasons for this, for instance, stronger productivity growth than expected and surprisingly low import prices. The reasons for the low inflation rate are to a large extent factors that are beyond the Riksbank’s control. Monetary policy decisions are based on a forecast of future developments and it is in the nature of things that unforeseen events can and do occur.
However, with the benefit of hindsight, it is in my opinion clear that there would have been scope for the Riksbank to conduct more expansionary monetary policy than it did. However, we were by no means alone in making incorrect forecasts of inflation. In principle all of the other forecasters, albeit to different extents, made roughly the same forecasting errors. We all underestimated effects of the global slowdown in inflation and the low import prices and we incorrectly assessed productivity growth, which was much higher than expected.
When assessing the situation and answering the question of whether excessively tight monetary policy has held back demand, it is important to remember that the forecasting errors are primarily due to supply-side factors and that monetary policy in recent years has nevertheless been expansionary, which is reflected not least in the increase in household borrowing and in rising asset prices, primarily with regard to housing, but also in equity. Long-term interest rates are also very low, companies in general have very good liquidity and good profitability.
It can also be noted that in connection with the Riksbank cutting its repo rate in June as a result of weaker economic growth than expected at the beginning of the year, some voices were raised warning that the low interest rates could lead to an excessive credit boom and overly high house prices. The consideration that has to be made between growth in demand over the coming two years and the risks linked to high indebtedness, for instance, are still relevant and current, and I will return to this shortly.
Monetary policy, productivity and jobs
One consequence of the high level of productivity is that employment has not improved very much, despite several years of good growth. It is not likely that the link between economic activity and employment has disappeared but - and this has been clear during the current economic upturn – the relationship may very well vary from one economic cycle to the next. We have produced larger quantities but using fewer employees. One possible reason for the high productivity growth is earlier investment in IT, in particular. Another is that the stiffer competitive pressure, which among other things originates from the internationalisation of the economy, has forced companies to make their production more efficient. This is essentially positive, as productivity gains contribute to increased income, investment and growth, which should in the long term lead to new job opportunities. There are many indications that an increasing number of companies are now approaching the limit where they can no longer meet the increase in demand without recruiting new staff. We are thus beginning to see signs of a turnaround, but the question of the labour market and the significance of monetary policy for employment remains important.
To what extent can the problems in the labour market be said to be due to the Riksbank having conducted an excessively tight monetary policy? If we had succeeded in predicting the high productivity we would probably have conducted a slightly more expansionary policy, but my answer is nevertheless that one is overestimating the power of monetary policy if one believes that it could have solved the problems in the labour market. A large part of unemployment is due to structural reasons and therefore does not vary according to the economic cycle. I do not believe that it would have led to so much higher consumption and investment if the repo rate had, for instance, been on average half a percentage point lower during 2002 and 2003. My assessment is thus that it would probably have led only to marginally lower unemployment.
If one wishes to achieve a lasting improvement in employment, measures are required that will stimulate the functioning of the labour market in different ways and lead to investment and a good corporate climate. One reason behind the problems in the labour market is that we have not really been able to keep up with the structural change process that follows in the wake of globalisation or, expressed in another way, that we have not been quick enough in adapting to the increased pressure for change when, for instance, the manufacturing industry is challenged by low-cost countries in eastern Europe and Asia.
To summarise: our look back in time leads to the conclusion that the change in monetary policy regime was a success, despite the fact that the period since the mid-1990s was a fairly turbulent time for conducting monetary policy. There have been, for instance, periods of considerable financial unease and extraordinary events of various types. A few of the many examples are: the Asian crisis, the Russian crisis, the IT bubble and the acts of terrorism on 11 September 2001. In addition, food prices and various energy prices, in particular oil prices, have periodically varied substantially for varying reasons, such as war, natural disasters and dry summers.
I would now like to give you a brief summary of the current economic situation and then conclude with a few thoughts on future challenges for monetary policy.
The current economic situation
As you have probably noted, we Executive Board members chose to hold the repo rate unchanged at our most recent monetary policy meeting in late October. This decision was also in line with market expectations.
The outlook for the Swedish economy is bright. Our forecast is that economic activity will strengthen in the near future, a development which is supported by the fact that economic policy remains expansionary. There are some causes for concern; primarily the high oil price and global imbalances, which contribute to the uncertainty over international economic developments. Our assessment in the October Inflation Report was that the risks for inflation were balanced, that is, the risks of weaker price development were balanced by the risks of higher inflation. As before, there is also reason to take into account the fact that household borrowing and house prices are continuing to rise at a rapid rate.
I agree with the analysis in the Inflation Report and with the decision to hold the repo rate unchanged. At the same time, in my opinion there is reason for some concern with regard to the sustainability of the current economic trend. There are already signs of a slowdown in the United States, which are probably linked to the severe storms the country has suffered, as well as to the fact that inflation appears to be accelerating and to the Federal Reserve’s interest rate increases.
It has long been clear that a slowdown in domestic demand is necessary in the United States in order to correct global imbalances. However, there is a risk that the adjustment of the imbalances will occur too rapidly and will be combined with a rapid fall in property prices. This type of development could in turn lead to a rapid braking of consumption. If this occurs without a simultaneous rise in consumption in the euro area, there could be a problem for the Swedish economy, as such a large part of our exports are to the euro area. The signs of an upturn in Europe are so far few and growth in the United Kingdom has slackened considerably. Such a situation would require more measures than a single monetary policy in Europe.
Weaker growth in the United States could also lead to a decline in interest from China in investment in US government securities, which could further push up long-term market rates. This would occur during a downturn, while the effects of oil price rises continued to push up inflation. In my opinion, there is thus an evident risk that international economic activity would decline, possibly as soon as next year or the following year. The dilemma is that in this case it would occur in a situation where interest rates were already low in many countries, which limits the scope for monetary policy to counteract developments.
If we look at Sweden alone, the economic outlook is favourable and the conditions for a positive economic development are in place. Companies are showing good profitability, investment is beginning to pick up and there are signs of improvement in the labour market. It is the international picture and its effects on Sweden that I consider to be worrying. A possible international slowdown in the coming two years could lead to growth failing to reach the level indicated in the main scenario of the Riksbank’s Inflation Report.
Another question that affects the Riksbank’s monetary policy considerations is developments in the property market. Exaggerated increases in asset prices can lead to subsidence in the future. With a larger percentage of debt in relation to income and with shorter periods for fixing interest rates on loans, it is probable that monetary policy would have a stronger and faster impact on household consumption. Moreover, the Riksbank’s tasks include not only safeguarding price stability, but also financial stability. This is a further reason for closely following developments in indebtedness in the economy and price trends in the property market. In the event of economic shocks it might be necessary to weigh price stability, real economic stability and financial stability against one another; it might not be possible to achieve all of the objectives at the same time.
Future challenges for monetary policy
Some factors that have monetary policy significance do not change according to the Swedish business cycle. This includes the effects of globalisation. As I mentioned earlier, globalisation and increased competition from low-cost countries like China and India have contributed to lower inflation by way of import prices, increased competition and greater pressure for structural change. This makes demands not only of monetary policy, but also structural and fiscal policy. The moving of manufacturing and services to other countries need not mean higher unemployment and lower growth if innovative forces are equally strong in the country from which the jobs are disappearing. In addition, flexibility in the labour market needs to be good. Measures to facilitate for the labour force to move between sectors are crucial, in order to avoid long-term unemployment, for example. Skills development and retraining opportunities must be offered to those losing their jobs in the sectors that are closing down.
The adjustment process can be painful for those involved, but structural change is also a positive force. Lower costs and productivity gains enable investment and employment opportunities to move to other, more refined products and services. Old sectors or companies that are pushed out of the market create scope for new employment in areas where the country concerned has a comparative advantage or within domestic services production that does not compete to the same extent as the manufacturing industry. It should also be borne in mind that many rapidly-growing low-cost countries like China will probably increase their imports as much as their exports. What can we offer these countries in the future? The areas where we have comparative advantages contain future export potential and employment opportunities.
Monetary policy challenges
The fact that prices of imported goods have tended to be lower than we had earlier expected over a period of time raises the question of what measure of inflation we should concentrate on. One question that has been raised is whether the Riksbank should disregard imported inflation and instead focus on domestically-generated inflation. This argument is based on the assumption that as the Riksbank is unable to affect imported inflation, it should aim monetary policy towards the areas that can be influenced. Let me give a simple example of why I do not consider this viewpoint to be valid.
Of course, the price in a foreign currency that a foreign producer receives for its goods is entirely beyond the Riksbank’s control. However, when the product is imported to Sweden, Swedish importers, for instance, wholesalers, have to exchange foreign currency in order to pay for the goods. In other words, monetary policy plays some role in the price in Swedish krona that the importer has to pay for the goods through its effects on the exchange rate. When the importer then wishes to sell the goods in the Swedish market, it must have a network of retailers who will sell the product. This means that Swedish costs, such as wages and rent paid for premises, also play a part in the price the retailer charges for the goods. Naturally, the demand for the product is also an important factor. In other words, the traditional supply and demand relationship in the Swedish market is important to the end-price that Swedish consumers have to pay for the imported goods. There are studies indicating that domestic cost components may comprise as much as 75 per cent of the end-price(2). Arguing that the Riksbank cannot affect imported inflation is therefore simplifying rather too much. On the other hand, it may be the case that during the prevailing changeover process it may be quite reasonable to allow inflation to be below (or in a different situation, above) the inflation target for a period of time. This lies within the boundaries of the flexible inflation targeting the Riksbank conducts. A policy of flexible inflation targeting means that we always consider how large a deviation from the inflation target might be acceptable, and whether we can allow a little more time before bringing inflation back in line with the target.
Another challenge is the low interest rates. At present, the Riksbank’s key rate is at an historically low level, although we are in the upturn phase of the economic cycle, at least this appears to be the case in our view of recent growth and our forecast for future growth. At the same time, one cannot rule out the possibility that inflation might remain low for a long period of time. If economic activity should decline substantially, the problem will be that there is little scope for further stimulation using monetary policy. There is then a risk that the Riksbank would lack the means to manage its stabilisation policy task. On the other hand, if we continue to have very low interest rates at the same time as the growth rate is high, there is a risk of imbalances arising in the housing market.
Conducting flexible monetary policy means being prepared to deal with this type of economic shock. The bank is therefore working on further clarifying the principles published in 1999 that explain how the Riksbank’s inflation target will be applied and the considerations that sometimes need to be made between inflation, stability in the real economy and stability in the financial system. This does not involve any fundamental changes in monetary policy strategy but may make it easier to understand how monetary policy is applied.
Thank you!
Notes:
1) Chile and Israel also introduced an inflation target at an early stage, 1990 and 1991 respectively, but inflation and the targets were much higher there than in the other "pioneering” countries.
2) Burstein, Ariel, Martin Eichenbaum, and Sergio Rebelo (2005), "Large Devaluations and the Real Exchange Rate", Journal of Political Economy, Vol. 113, No. 4, August 2005, pp. 742-784.
