Rosenberg: The property market, the economic situation and monetary policy
Thank you for your invitation to come to Karlstad and participate in this seminar on the property and rental market. To begin with, I intend to speak about the property market and why it is important for us at the Riksbank to monitor how it is developing. Then I will discuss the economic situation internationally and in Sweden and give my view on inflation prospects and monetary policy.
The property market
Developments in the property market are important both in view of the Riksbank's function to promote financial stability and of the monetary policy objective to maintain price stability. On account of that we regularly monitor the situation in both the commercial property market and the housing market. That the property market is significant for the Riksbank’s stability analysis is due to the clear linkages that exist with the financial sector. The most important is that property companies are major borrowers from the banks. Moreover, a large proportion of bank lending is secured on commercial and residential property. So if problems were to arise in the property market, they could spill over to the financial sector. One example of when this happened was during the banking crisis at the start of the 1990s. During the years 1992-93, fully 40 per cent of the banks’ loan losses stemmed from the commercial property market.
Let me begin by discussing how I view the developments in the commercial property market and then move on to the analysis of the housing market. Because the two markets have shown completely different developments in recent years.
The commercial property market
The equity market decline in 2000 was accompanied by a weakening in demand for office premises, and rents and prices for commercial property began to fall as the number of vacancies rose. In recent years, turnover of properties has risen again. Presently, the drop in prices seems to have decreased and in some regions prices have even risen. Even though it is difficult to determine whether there has been a definite turnaround yet in the price developments for commercial property, such a turnaround can be expected to come when demand for labour recovers more clearly. Vacancy levels are still high, though, which means that rents have not yet risen more generally. In the long run, a rise in property prices should be accompanied by an increase in rental income, since rental income is the underlying return on property investment.
The situation today differs considerably from the conditions seen at the time of the property crisis in the early 1990s, when property prices had risen rapidly over a long period without any corresponding upswing in rental income. At that time the price increases were driven instead by expectations that prices would continue to rise strongly in the future. When it eventually became evident that these expectations were not going to materialise, the prices fell quickly. Such speculative elements do not appear to exist today. At the same time, property companies’ earnings and debt-servicing ability are good. As things stand today, the property sector will hardly cause any large loan losses for the banks and so it does not constitute a threat either to the stability of the financial system.
Housing prices and household debt
Over the past one to two years we in the Riksbank’s Executive Board have devoted quite a bit of attention to the house market, where a rapid rise in prices has gone hand in hand with an increase in household debt. But it is important to distinguish between the conclusions we have come to regarding the significance for financial stability and the discussions we have had in connection with our monetary policy decisions. I will come back to this.
That house prices have risen rapidly in Sweden and many other countries can hardly have escaped anyone’s attention. In Sweden, house prices in the past ten years have more than doubled, and the upswing has been especially strong in metropolitan areas. Increasing house prices have been accompanied by an accumulation of household debt. Due to the fall in interest rates, households’ interest costs have dropped, despite the rise in debt. In historical terms, households’ interest costs are currently very low. That the low interest rates have increased households’ ability to take on bigger loans is one cause of the upturn in house prices. During the summer and autumn, the rate of increase in house prices appears to have been dampened somewhat in certain areas, but the rate of increase is still high. Monthly data show that the annual rate of growth for the whole country during the period July to September was 10 per cent.
The developments in Sweden are by no means exceptional. Internationally, there are examples of countries which have seen sharper rises in house prices, such as Australia and the UK. But there are also examples of countries where the rate of increase has been lower, such as Germany. Recently, however, the international situation has become more mixed in that house price inflation in some countries, e.g. Australia and the UK, has slackened considerably whereas in others, including the US, it has continued to pick up. In Australia and the UK, the rate of house price increases dropped when the central banks there raised key interest rates, although this occurred with a slight lag in Australia. In the US, the rate of increase has continued upwards in spite of rises in nominal short-term market interest rates. Longer-term rates, however, have not risen to the same extent since the Federal Funds rate began to be raised. Because it is more common for US households to fix the interest rate on their home loans for very long periods, they are mainly affected by changes in long-term interest rates. Moreover, the pick-up in inflation in the US has meant that real rates of interest have not increased in line with the rises in the Federal Funds rate.
Estimates at the Riksbank suggest that the upswing in Swedish house prices to date is largely attributable to rising incomes, low interest rates and low construction activity. Since house prices began to increase in the mid-1990s, nominal disposable incomes have risen by around 40 per cent and interest rates have dropped sharply. In 1996 the average nominal mortgage rate for households was just over 9 per cent. In June this year it was a little less than 4 per cent. Meanwhile, construction of new homes has been very low, even though low interest rates and rising prices have meant that construction has picked up relatively well in recent years.
My assessment is that so far there are no indications of any significant overvaluation of house prices on average in the country at current interest rate levels. Furthermore, Sweden essentially has no buy-to-let market, i.e. a market in which private investors buy homes solely with a view to letting them. That limits speculation in housing. The apparent absence of any marked overvaluation in the country overall does not mean of course that house prices in some parts of the country or in some submarkets have not been driven up more than is justified in the long term.
The fact that the price increases can be explained relatively well by rising incomes, lower nominal interest rates and low construction activity does not rule out the possibility that the rate of increase could slow down, or that prices could even fall in the future. The developments in these explanatory factors normally follow the business cycle. However, the adjustment should be less drastic than in a situation when prices have soared in a way that cannot be attributed to developments in underlying factors.
At present, it is especially important to remember that interest rates are low in historical terms and that they therefore very probably can be expected to rise in the period ahead. To get an idea of what constitutes a reasonable level for interest rates in the long term, one can look at countries that have had a floating exchange rate and low inflation for a longer time than Sweden. In the US and Germany, for instance, key interest rates have normally fluctuated between two and six per cent. When you have a regime with a successful inflation targeting policy I think that you normally can expect interest rates in this range. That also makes it clear that today’s repo rate of 1.5 per cent is uniquely low. So it is important that both households and mortgage institutions take account in their estimates of the possibility that interest rates will be higher in the future than at present. That also means that the rate of increase in house prices cannot be expected to remain as high in the long term as it is today.
The consequences of this assessment of house price developments differ somewhat depending on which of the Riksbank’s two tasks we are talking about. As regards payment system stability, i.e. financial stability, I do not believe that house prices and household debt pose any threat at present. The bulk of lending to households comprises loans secured on housing. A home is part of a household’s wealth and if the value of a home or apartment were to fall, it naturally would mean that the homeowner’s ability to pay would decrease. But the household’s income is not tied to the value of its home, and as long as the household is able to meet its borrowing costs, it does not represent a threat to the banks and the payment system.
Both higher incomes and lower interest payments have resulted in a good ability to pay in the household sector at present. Household data estimates that the Riksbank previously has presented in its Financial Stability Report show that most indebted households have good margins to cope with a loss of income or rises in interest rates in the future(1). Naturally, that does not mean that individual households could not run into problems when, for example, interest rates increase. For the banks, though, it means that credit risk in the household sector as a whole is low. Past experience also shows that the household sector seldom causes the banks to incur large loan losses. Not even during the deep crisis at the start of the 1990s, when house prices fell steeply, were losses on loans to the household sector a significant share of banks’ total loan losses.
From the point of view of monetary policy, the consequences of rising house prices and high debt levels are partly different. House prices influence, via wealth effects, households’ decisions regarding saving and consumption and thereby also demand in the economy overall. For indebted households, the scope for consumption is also affected if the share of their income used for interest payments changes. This is something we have to take account of when forecasting economic growth and inflation.
If interest rates rise, interest expenditure as a share of disposable incomes will increase, which tends to dampen consumption demand. That is exactly what you want to achieve if monetary policy is tightened. What happens to house prices is more uncertain. In the UK and Australia, for instance, house price inflation moderated, as I mentioned, when the central banks there began to raise key interest rates, but there has not been a general fall in house prices. In a scenario where the central bank begins to raise key interest rates to slow down a strong economy and market interest rates follow these upwards, it is plausible that the adjustment in house prices may be relatively soft, with a rate of increase that is dampened or levels off. But even a moderation of the rate of increase in house prices could help to reinforce the tightening impact on household consumption. Historically, households’ consumption ratio, i.e. the share of households’ disposable incomes that is used for consumption, has been fairly well-correlated with changes in the value of household wealth (see Figure 1)(2). Since the equity market decline in 2000 this relationship has been somewhat less clear, but it nonetheless suggests that house price inflation may only need to slow for the consumption ratio to drop.
The risks of a more dramatic downward adjustment in house prices would increase if economic activity were to weaken relatively sharply and household incomes were to fall as a result. Naturally, this could make itself particularly hard felt if the developments in house prices and debt beforehand had taken off far too quickly and households had taken on more debt than they can manage in the long run. In that case, we could witness a more drastic adjustment that would drive the economy into a deeper and more protracted slump that would be difficult for economy policy to manage. This is a risk scenario that the Riksbank’s Executive Board has devoted some attention to in recent years. It is also a risk that individual households have reason to take into account when they are about to take on new loans. So these mechanisms are something that we have to take into account when we make our monetary policy decisions.
The current economic situation
International economic activity still firm
International economic growth was high last year. In the latest Inflation Report in October, we noted that the slowdown in international economic activity earlier this year appeared to have been temporary. The oil price had turned out considerably higher than we previously expected, however, prompting us to sharply revise up our oil price forecast. This was expected to contribute to an increase in international price pressures and to a slight moderation of growth. These effects were forecast to be relatively limited, however. The picture of relatively firm economic growth abroad in the coming years was thus broadly unchanged. Consequently, Swedish export markets also were estimated to grow at a firm pace in the next few years.
In the US, the hurricanes and oil price increases in recent months have contributed to dampening the economy somewhat. The US Federal Reserve estimates that the rate of increase in industrial production has been held back a couple of percentage points in August and September overall. However, preliminary National Accounts data for the third quarter point to continued firm growth. Households and companies have continued to boost their consumption and investment at a fairly stable rate. The purchasing managers index for both the manufacturing sector and the services sectors in October indicate relatively firm developments at the beginning of the fourth quarter as well.
Elsewhere in the world, too, there have been a number of positive economic signals. In October, the purchasing managers index rose in several European countries and in Japan. The German IFO index also showed a positive trend in October. In the UK, however, the recent economic slowdown has continued, as shown, among other things, by preliminary GDP data for the third quarter.
Inflation abroad has turned out somewhat higher recently than we anticipated in the latest Inflation Report. In the US, inflation rose sharply in September, to almost five per cent. Excluding energy prices, though, it is still at roughly two per cent. The US Federal Reserve again raised the federal funds rate at its latest monetary policy meeting, bringing it up to 4 per cent, i.e. 3 percentage points higher than when the monetary tightening cycle was begun in June 2004.
In the euro area and some other countries, too, inflation has picked up more than anticipated recently. Here as well the upswing mainly seems to be associated with the direct impact of high energy prices. However, there are no real signs as yet that the high energy prices are going to have second-round effects on prices more generally in the economies abroad.
To sum up, recent data suggest that international economic activity is developing in line with the assessment in the latest Inflation Report. That implies continued firm growth this year and the coming two years. Meanwhile, it seems that inflation in the OECD area will be somewhat higher than we previously expected owing to the higher energy prices. Naturally, there are risks associated with the international economic performance and thus that inflation could deviate from our main scenario. However, in my opinion the risk outlook, like the main scenario, has not changed to any significant extent either since we published our latest Inflation Report in October.
New interest rate assumption as basis for forecast
Before I discuss the outlook for the Swedish economy, I would like to make some comments on the new repo rate assumption that was introduced in the Riksbank’s forecast in October. The forecast in the Inflation Report’s main scenario is now based on the assumption that the repo rate will develop in line with financial market expectations, as reflected in implied forward interest rates. In addition, the forecast horizon has been lengthened by one year so that the forecasts extend three years ahead.
Such an interest rate assumption has several advantages. One is that it normally gives a more realistic picture of future monetary policy. That makes it easier to evaluate our forecasts. It also becomes easier to compare our forecasts with those of other forecasters. The longer forecast horizon makes it clearer how inflation is being influenced by various temporary shocks. The reasons for the changes are to provide a more comprehensive set of data for monetary policy and to improve the pedagogical aspect of the monetary policy message. It is important to point out that the changes in methods do not entail any fundamental change in the Riksbank’s monetary policy strategy. Our interest rate decisions will continue to be guided by the ambition that inflation normally should be returned to target within two years.
The new repo rate assumption is a technical forecasting assumption, which should not be interpreted as the Riksbank’s stance regarding which interest rate path it considers to be most desirable. The interest rate path that underpinned the forecast in the October Inflation Report was judged to imply a reasonable development of inflation in the coming years. Our conclusion was that if this development were to continue in line with our assessment at the time, the repo rate would need to be raised in the period ahead. At the same time, we pointed out that there also were different conceivable paths for the future repo rate that could give roughly the same outcome for inflation.
Growth in the Swedish economy continuing to strengthen
For the Swedish economy, too, it was concluded in the Inflation Report that the economic upswing appeared to have gained impetus again after the slowdown at the start of the year. Compared with the previous assessment, GDP growth and resource utilisation were forecast to be somewhat higher in the years ahead, largely owing to more expansionary economic policy. All in all, the assessment was that growth would be relatively firm in the Swedish economy in the coming years.
The recent years’ high productivity growth has partly laid the foundation for this development. Companies have been able to boost output with their present workforces, which has improved their cost situation and held back price increases in the economy. The low inflation rate has enabled monetary policy to be expansionary for a comparatively long period. In addition, the high productivity and relatively good profitability is anticipated to create scope for comparatively favourable wage growth in the period ahead. This means that there are good prospects for firm growth in domestic demand over the coming years. Expansionary fiscal policy will also help to boost domestic demand, while more robust growth in Swedish export markets will contribute to stronger exports.
New information broadly supports the picture of a continued improvement in economic activity in Sweden. The National Institute of Economic Research’s (NIER) latest business tendency survey provides a relatively bright picture of economic conditions in most of the surveyed industries. The confidence indicator has risen for both the manufacturing and construction industries as well as for trade and services industries. New orders in manufacturing have picked up, mainly in the export market but there has also been a clear improvement in the domestic market. Monthly data for foreign trade in goods also indicate an increasingly bright picture for exports. Like the developments elsewhere in the world, the Swedish purchasing managers index for October also showed a positive development. New data for retail sales point to continued relatively strong rises. At the same time, the NIER’s consumer survey for October provides somewhat mixed signals. Households have become slightly less optimistic about the Swedish economy and unemployment, while their expectations regarding their own finances have improved.
As regards the labour market, we noted in the October Inflation Report that there were signs of an improvement in the situation. The strengthening of domestic demand and the announced measures in the Budget Bill were judged to lay the foundation for a rise in employment in the period ahead. However, the change in methods in the labour force surveys means that developments in the labour market are particularly hard to interpret at present. Employment growth appears perhaps to have been somewhat stronger than we expected in October. Developments in the number of new job vacancies and redundancy notices also point to an upturn in the labour market. In addition, the NIER’s latest quarterly business tendency survey contained a number of positive signals regarding companies’ hiring plans.
Taken together, the new information that we have received about the Swedish economy also indicates a development well in line with our forecast in the Inflation Report in October.
Inflation and monetary policy
Inflation in the Swedish economy has been low since the end of 2003. According to the most recent data, for September, annual inflation in terms of the CPI and UND1X stood at 0.6 per cent and 1.0 per cent, respectively(3). As we observed in the Inflation Report, the low inflation can be explained in large measure by favourable supply conditions – strong competition and fast productivity growth – and relatively low resource utilisation. At the same time, imported inflation has been restrained because the share of imports coming from low-cost countries has risen. Imported inflation has also been held back by the previous strengthening of the krona.
Developments in the krona exchange rate are one of all the factors that we have to consider when we forecast inflation. It is also one of the few areas where the performance since October has deviated considerably from what we expected at the time. Allow me therefore to comment briefly on the krona exchange rate. For a succession of years the Riksbank has been predicting that the krona will strengthen. This forecast has been very accurate up to the beginning of this year even though there have been some large short-term deviations on the way (see Figure 2). During the current year, however, the krona has weakened relatively sharply. In the markets, the focus at present seems to be very much on interest rate differentials in relation to abroad. Personally, I find the weakening of the krona somewhat surprising given that underlying factors such as the real economy, the relative inflation rate and so on point to the opposite. For that reason, we also expected in October that there would be an appreciation in the period ahead. I have difficulty seeing any reason for us now to change that assessment in our main scenario. But a weaker krona is a risk that we have to take into account. Were the krona exchange rate to remain weak, it would of course have implications for the inflation forecast.
In October, as before, we forecast that inflation would rise gradually in the period ahead. Resource utilisation was expected to pick up both in Sweden and abroad. We predicted that productivity growth would slacken somewhat and that cost pressures would gradually become higher. Substitution in favour of cheaper imported goods was expected to continue, but rising global resource utilisation was estimated to mean that global price pressures nevertheless would increase slightly on the whole. Against that background, our assessment was that UND1X inflation would rise from today’s low level and be close to the target a couple of years ahead. In my opinion, the information that has been received since October suggests that our assessment of inflation prospects at that time is holding up well so far. We in the Executive Board will convene for our next monetary policy meeting at the start of December. At that time, we will also have to factor in the information that has been released up to then.
Notes
1) See Financial Stability Report 2005:1.
2) Inflation Report 2005:3, page 43.
3) CPI data for October will be released today, 11 November.
