House prices have more than tripled since 1992. After having fallen during the last part of 2002 and the beginning of 2003, house prices rose by more than 20 per cent from May 2003 to November 2004. We analyse factors underlying the pronounced rise in house prices using an empirical model. We find that interest rates, housing construction, unemployment and household income are the most important explanatory factors for house prices. The analysis indicates that house prices react quickly and strongly to changes in interest rates. Thus, a considerable portion of house price inflation since May 2003 can be explained by the fall in interest rates in the last two years. Conversely, the fall in interest rates will only make a modest contribution to house price inflation in 2005. An interest rate increase in line with the interest rate path in Inflation Report 3/04 can in isolation lead to a 3-3 1/2 per cent fall in house prices per year in 2006 and 2007. However, this interest rate path reflects an expected decline in unemployment and an expected increase in the growth of wage income. The model implies that house prices will increase by 2-4 per cent per year in the period 2005-2007 if interest rates, unemployment, income and housing construction develop in line with the analyses in Inflation Report 3/04. We find no evidence that house prices are overvalued in relation to a fundamental value determined by interest rates, income, unemployment and housing construction.
Developments in house prices may be important for activity in the Norwegian economy. First, house prices affect activity in the construction sector. New housing construction projects will be profitable if house prices increase in relation to building costs. This stimulates housing investment. Second, house prices affect household demand. Higher house prices mean an increase in wealth for homeowners and some owners will want to extract some of this gain to increase consumption. This effect is amplified by the fact that homeowners increasingly have the possibility of raising mortgage-secured loans when house prices rise--at interest rates that are often far lower than for other types of loans.
Developments in house prices also affect household borrowing for house purchases. An increase in house prices will fuel debt accumulation for a long period (see Jacobsen and Naug 2004), reflecting the fact that only a small portion of the housing stock changes hands each year. Even if house prices gradually level off, there will be a long period when selling prices are higher than the last time the dwelling changed hands.
Mortgage-secured loans account for more than 80 per cent of banks' lending to households. If house prices decline, collateral values can fall below the value of the housing loan for some households. Banks' loan losses will increase if these households are unable to service their debt. As a result, banks may become more reticent about providing loans to households and house prices may fall further. A fall in house prices will also reduce household wealth and the possibility of raising a mortgage-secured loan. This will curb private consumption and the level of activity in the Norwegian economy. Consumption may also become less interest rate sensitive than when households can borrow large amounts through mortgage-secured loans.
House prices have more than tripled since 1992. After having fallen during the last part of 2002 and the beginning of 2003, house prices rose by more than 20 per cent from May 2003 to November 2004. Developments in the housing market have contributed to a 10-11 per cent increase in household debt per year since 2000. The debt burden for low- and middle-income households is now close to 50 per cent higher than the last peak in 1987. The high accumulation of debt has made households more vulnerable to negative economic disturbances.
The sharp rise in house prices in the last year and a half may prompt the question of whether there is a bubble in the housing market, i.e. whether house prices are far higher than a fundamental value determined by interest rates, income and other fundamental explanatory factors for house prices. A house price bubble can arise if (i) many individuals want to purchase a dwelling today (putting an upward pressure on prices) because they expect house prices to rise in the period ahead and (ii) these expectations are not based on fundamentals. If there is a price bubble in the housing market, prices may fall sharply if price expectations change. Prices may show a particularly sharp decline if price expectations change as a result of a change in fundamentals. In this case, banks may experience that the value of the collateral falls below the value of the loan and that households increasingly have difficulty repaying (very high) debt. This can, as described above, lead to an economic downturn (see IMF (2003) and Borio and Lowe (2002)).
House price inflation since May 2003 may, however, reflect changes in fundamentals. In particular, it is likely that the fall in interest rates since the end of 2002 has contributed to the rise in prices. The current low interest rate level is unlikely to continue, however. If interest rates have a strong impact on house prices, we would therefore expect house price inflation to be relatively subdued when interest rates gradually normalise. As long as interest rates increase gradually, there is nevertheless reason to believe that price adjustments will be fairly slow. Nor will house prices necessarily fall when interest rates gradually increase, since the interest rate increases may reflect rapid growth in wages and employment.
It follows that indicators and models that measure whether house prices are overvalued in relation to fundamentals, or whether the fundamentals have been responsible for the high house prices, may be useful when monitoring financial stability. Understanding how and to what extent house prices depend on various fundamentals is also important for projecting house price developments.
The ratio of house prices to income and the ratio of house prices to house rents is commonly used to measure whether house prices are overvalued in relation to long-term fundamental values (see, for example, The Economist (2003) and a box in Financial Stability 1/03). Such measures may indicate that house prices in Norway are high in relation to fundamentals (see Charts 1 and 2). These measures are incomplete, however, since they do not measure whether house prices are high (in relation to income or house rents) due to a bubble or due to developments in fundamentals. An alternative approach is to estimate an econometric model of house prices using fundamental variables as explanatory factors. Then, under certain conditions, one can use the deviation between actual and fitted house prices as a measure of whether or not house prices are overvalued in relation to fundamental explanatory factors. IMF (2004), Foley (2004) and McCarthy and Peach (2004) have used such an approach.
In this article, we try to answer the following questions:
* What are the most important fundamental explanatory factors for house prices?
* How quickly and strongly do house prices react to changes in these factors?
* Is there a price bubble in the housing market?
* What has driven developments in house prices in recent years?
* What will happen to house prices if interest rates and the Norwegian economy develop in line with the analyses in Inflation Report 3/04?
We estimate a model of house prices on quarterly data for the last 14 years. The analysis indicates that interest rates, housing construction, unemployment and household income are the most important explanatory factors for house prices. We find that house prices react quickly and strongly to changes in interest rates. Thus, a considerable portion of house price inflation since May 2003 may be explained by the fall in interest rates in the last two years. The model implies that house prices will increase by 2-4 per cent per year in the period 20052007 if interest rates, unemployment, income and housing construction develop in line with the analyses in Inflation Report 3/04. We find no evidence that house prices are overvalued compared with a fundamental value determined by interest rates, income, unemployment and housing construction.
In the next section, we discuss factors that may affect house prices. We then investigate the relevance of these factors by estimating a model of house prices (section 3). In sections 3 and 4, we use the model to discuss the questions raised above. The model was presented in Financial Stability 1/04.
What can affect house prices?
House prices are determined by housing supply and housing demand. Housing supply, measured by the housing stock, is fairly stable in the short term, since building new dwellings takes time and housing construction per year is low in relation to the total housing stock. In the short term, therefore, house prices will generally fluctuate with changes in demand. The housing stock will adapt to demand over time, however. A long-term model of house prices should therefore contain explanatory factors for developments in the housing stock, such as construction and building site costs and prices for new dwellings. Here, we restrict the analysis to explain house price movements for a given housing stock.
Housing demand consists of two components: household demand for owner-occupied dwellings and demand for dwellings as a pure investment instrument. It is reasonable to assume that the first component is clearly larger than the second. We will therefore place greatest emphasis on the demand for owner-occupied dwellings.
Households may consume housing services either by owning or renting a dwelling. In this analysis, we consider the demand for housing services from owner-occupied dwellings (including flats in housing cooperatives). We also assume that...