Nyberg: Developments in the property market
I would like to begin by thanking you for the opportunity to come to Malmö and speak about the Riksbank's view of developments in the property market. Of course, these developments are not only of interest to those of us present today. They are important to all people, as movements in property prices have significance both for the real economy and financial developments in Sweden. When property prices rise, we feel more financially secure and increase our consumption, which leads to a rise in employment and growth. We might even borrow a bit more from the bank if we can. Normally, there is no harm in this. But there is a limit, of course, to the burden of debt households can bear without giving rise to worrisome risks. Should the debt of commercial property companies also swell, it could entail risks for the stability of the entire banking system, a point that was illustrated so well during the crisis at the beginning of the 1990s.
On account of this, the Riksbank monitors developments in the property market carefully. We usually divide it into two segments, which give rise to different questions and reflections.
The first segment is the housing market. Since 1997, housing prices have risen by just over 50 per cent, measured in current prices. The price boom has continued in spite of the generally weak economic activity of the last two years. This is a new experience which differs from previous business cycles, when prices have tended to recede during periods of economic decline. Similar developments have occurred in other countries, for example in the US, England and Holland.
The situation could appear worrying, particularly given that the rise in housing prices has been accompanied by an increase in household indebtedness.
The upsurge in prices has raised questions about the healthiness of this trend. Are the price rises in the Swedish housing market the result of a bubble? Are the higher levels of household indebtedness sustainable? The Riksbank's view is that the current price developments in the housing market do not reflect any speculative expectations of future price rises which would warrant the term "bubble". Nor is household indebtedness a bigger problem than usual for this stage of the business cycle. I will explain this in a moment.
The other segment is the market for commercial property. Here, it is primarily prices for office premises which have displayed a different trend to that seen in the housing market. Commercial property prices have adjusted to the economic slowdown. In Sweden, both prices and rents have fallen since 2000 at the same time as vacancy rates have risen. I myself do not believe that these price declines are a concern from an economic perspective, as was very much the case with the property crisis at the beginning of the 1990s. Although some similarities do exist between the two, the differences nevertheless outweigh them, and I will be coming back to these later. I will also be looking ahead and briefly discussing the driving forces behind price developments in the property market.
Developments in the housing market
In principle, there are two different methods to explain rapid price changes. The first approach is based on the fundamental driving forces that govern prices, and explains price changes through changes in these variables. Such fundamental factors include interest rate levels, household income and the production of new homes.
The second approach assumes instead that prices are not governed by these fundamental driving forces, but rather are based on expectations of future price rises. Such a situation is usually referred to as a "bubble". A recent example of this is the surge in prices seen in IT shares during the latter part of the 1990s. The rate of increase observed then proved gradually to be unrelated to fundamental driving forces but rather to unrealistic expectations of a trend that never materialised. When this eventually became evident, prices plummeted.
So the price rises seen during a bubble do not mirror fundamental factors but simply reflect expectations of future price increases. Investors become involved in a process of speculation where estimates of future cash flow are no longer of importance and price rises are instead assumed to be the trend. If you buy today, you will always be able to sell at a better price tomorrow.
Where the housing market is concerned, however, there is reason to question whether a bubble is possible at all. This is because most people do not purchase a home primarily as an investment but rather as a place to live. Consequently, a house buyer neither thinks nor behaves as an investor speculating on a quick and persistent rise in prices. There can hardly be many people who buy a home solely because they expect to gain financially through a rise in house prices! So assuming that speculation on future price increases doesn't exist, a bubble should quite simply not be able to occur. Some of you might now be thinking that this is just a convenient academic approach to the issue. After all, we have witnessed rapid price increases and even faster declines in the housing market, for example for tenant-owned apartments in metropolitan areas during the crisis at the beginning of the 1990s. In principle, I would be prepared to agree, if it proved impossible to explain the price rises with fundamental factors.
In order to see if there are fundamental factors that could explain the recent price increases, it makes sense to look at the driving forces that could have been important for house prices, i.e. interest rates, primarily the post-tax real interest rate, household disposable income, the supply of housing and the demand that results from people moving to metropolitan areas in particular.
As regards interest rates, both real and nominal rates have fallen continuously since the mid 1990s. For households this has entailed significantly lower mortgage costs. At the end of 1992 the banks' average nominal lending rate to households was just over 15 per cent and inflation stood at just over 2 per cent. Now, the average lending rate is around 6 per cent, while inflation has remained at about the same level. So households can incur more than double the amount of debt today and still have lower debt-servicing costs.
The favourable economic climate from the mid 1990s also contributed to a drop in unemployment, on the one hand, and an increase in wages on the other. All in all, this led to a rise in household disposable income. Moreover, disposable income has continued to increase during the period of relatively weak economic activity in recent years. This has been mainly due to unemployment remaining low, in spite of the economic slowdown.
So low interest rates and high growth in household disposable income are two fundamental factors that could explain the rise in prices in the housing market.
Furthermore, there is no excess supply in the housing market similar to that seen at the beginning of the 1990s when housing construction was encouraged through various kinds of grants and tax subsidies. Instead, there is very little construction of housing today, and discussions are concerned rather with how to step this up. In addition to the low level of housing construction, demand for housing has risen mainly in metropolitan areas due to the greater numbers moving to these regions during the latter part of the 1990s. It is not surprising that a relatively constant, low supply, as well as high demand, should have contributed to the substantial price rises.
Thus, low interest rates, an increase in disposable income and a low supply of new housing, together with high demand, are the fundamental factors that I believe have driven the rate of price increases from the mid 1990s. They also suffice to show that current price rises should not be described as an expectations-driven "bubble". The same analysis can be performed for several other countries experiencing similar developments in their housing market, in particular the US.
So did we have a housing market bubble during the crisis years at the beginning of the 1990s? I don't really believe that we did, at least not in the sense that homes were being bought for the purpose of speculation with a view to being sold once prices had risen further. Yes, prices rose sharply, especially in metropolitan regions, but these were due to reasons of a fundamental nature. Households' willingness to pay increased in line with their income during the economic upswing, the deregulation of the credit market led to the removal of some financing obstacles, and the supply of housing was, in the short term, highly inflexible. This is enough to fuel a steep rise in prices. When demand subsequently waned as a result of a deterioration in household finances, at the same time as real interest rates rose and the supply of housing - both new and old - increased, prices fell again. Even substantial price movements can have natural explanations and do not have to be due to unrealistic expectations that are first raised and then dashed.
Households have continued to borrow from banks and mortgage institutions throughout the economic decline, which is not the normal pattern of behaviour, but this is of course due to low interest rates. Their debt burden has grown and raised questions about whether we are heading back towards an imbalance that sooner or later will have to be corrected. However, I myself do not believe that household debt per se constitutes any immediate source of concern. Households' debt ratio, i.e. debt as a percentage of income, is just over 110 per cent. However, as a result of the historically low interest rates, it appears that households' debt-servicing costs will continue to be manageable. The interest ratio, i.e. households' debt-servicing costs in relation to disposable income, is today barely 5 per cent.
The situation for Swedish households is therefore considerably more favourable than at the end of the 1980s and beginning of the 1990s. At that time, the credit expansion which was caused partly by the deregulation of the credit market resulted in a debt ratio of around 130 per cent, while the interest ratio increased rapidly over a couple of years from around 6 per cent to almost 10 per cent.
However, there is no doubt that the recent rise in indebtedness has made households increasingly sensitive to changes in interest rates. Households tend to focus on their cash flow and are unperturbed as long as their debt-servicing costs can be accommodated in their monthly budget. But an interest rate rise of just a few percentage points would have a much bigger impact on cash flow today than when levels were around 15 per cent. An interest rate rise of four to six per cent would cause monthly costs to increase by 50 per cent. Higher mortgage rates could therefore prompt households to increase their saving and reduce their consumption more than expected or intended. Hopefully we will be able to avoid sharp, quick interest rate rises in our conduct of monetary policy, although mortgage rates are also influenced by events outside Sweden. To some extent though, the effect of any increases in rates could be alleviated by the fact that an increasing number of households have recently been switching over to fixed-rate loans.
Developments in the commercial property market
Prices in the commercial property market have been driven by somewhat different forces than in the housing market. The trend here has also been more closely correlated to the level of economic activity, with prices falling over the last three years. However, the price decline in Malmö [9 per cent] has been considerably gentler than in Stockholm [24 per cent] or Gothenburg [18 per cent]. These declines have been mainly caused by lower rents, which essentially have been a result of lower demand for office premises due to the subdued economic activity. It also makes sense that Malmö has been affected less than the other major cities given that the slowdown has not been as severe here.
The rapid rise in prices and their subsequently rapid decline points to a trend not all that dissimilar from that experienced during the property crisis at the beginning of the 1990s. Nevertheless, there are many differences between the two situations. There is no doubt that we experienced a bubble at the beginning of the 1990s, when prices largely lost their relationship to fundamentals and were instead underpinned by expectations of future price rises. For a period of time the direct yield on property was 8 percentage points lower than the yield on a government bond. So investors must have expected prices to increase a good deal more than 8 per cent in order to compensate them for the risk of investing in property rather than bonds. In this case, you can really say that expectations of price rises were high! This is not the situation today.
Speculation in construction was also less prevalent during the upswing at the end of the 1990s than during the rise in activity a decade earlier. Production of property has been relatively low, and projects have had pre-contracted tenants to a greater extent. In addition, the property crisis at the beginning of the 1990s was affected significantly by the structural changes that occurred during the period. The deregulation of the credit market together with the changes in economic policy required substantial adjustments in the market.
So the present situation is not a cause for concern in the same way as the situation at the beginning of the 1990s.
Some thoughts on the future
The fact that the situation is not a source of concern does not mean that I am predicting a significant recovery in property prices. An increase in economic activity will of course stimulate demand, which sooner or later will apply upward pressure on prices. But at the same time I can see several factors which could restrain a rise in property prices, especially with regard to office premises.
Firstly, there is a risk that demand for office premises will be weaker than the general level of demand in the economy. From the mid 1990s, there was a gradual improvement in economic activity in Sweden, which resulted in higher employment in general, but particularly within office-intensive businesses such as IT and telecommunications. So the increase in demand for office premises was partly attributable to economic conditions in general, but also to especially favourable conditions within office-intensive businesses in particular. The subsequent slowdown in activity also hit some of these businesses hardest, which affected demand for office premises. Thus, a recovery in demand for office premises would not only be associated with a revival in general economic activity but also in activity within these office-intensive businesses.
Irrespective of business area, it also appears that companies today, unlike at the end of the 1990s when expectations of growth and expansion were high, are seeking to reduce the area of office space per employee. This is one way for them to adapt their costs to new, more strained economic conditions.
All in all, these aspects suggest that demand for office space may not recover at the same rate as the general level of activity.
Secondly, unlike single-family dwellings for instance, commercial properties are a form of investment. Consequently, prices of commercial properties are affected more than single-family dwellings by the return on other investments.
In times of uncertainty, property has often been seen as an attractive alternative to equities in particular, and investors tend to increase the share of property in their portfolios, thus leading to upward pressure on prices. According to Swedish Property Index, the return on property in Sweden over the past three years has exceeded that on equities. Moreover, in 2000 and 2001, the return on property surpassed that on all alternative forms of investment. Last year, however, the return on property was exceeded by that on bonds and Treasury bills.
Thus, in a more optimistic economic climate, there is a risk that investment in property will appear less attractive than in equities, for example. In such a situation, as investors reduce the share of property in their portfolios, there is a risk that prices will not rise as much as warranted by the higher demand alone. Consequently, this is a factor that could restrain property prices during an upswing.
Summary and conclusions
The markets for housing and commercial property differ in terms of both their developments in recent years and the challenges facing them in the near future.
In the housing market, prices have continued to rise despite the recent economic slowdown. The market for commercial property has instead undergone a period of adjustment towards lower prices and rents since 2000.
By all appearances, the price rises in the housing market have been due to changes in the fundamental factors of significance for house prices. To speak of a housing market bubble is therefore not very constructive. Persistently low interest rates, strong growth in disposable income and low growth in the supply of residential property justifies current price levels, in spite of the weak economy. In the event of an increase in activity and higher income, which we all hope and believe will happen, there is no reason to expect a broad decline in housing prices. Of course, local differences in the production of new homes, for instance, could have an impact on prices.
In the commercial property market, where rents and prices have already adjusted to a different demand climate, the price increases that naturally follow a rise in economic activity could take time, particularly as there appears to be considerable excess supply of office space in various locations. Moreover, property as a form of investment is usually considered less attractive when the economy turns around in earnest.
