House price changes and land leverage

Citation

Hoesli, M. (2010), "House price changes and land leverage", Journal of Property Investment & Finance, Vol. 28 No. 4. https://doi.org/10.1108/jpif.2010.11228daa.001

House price changes and land leverage

Article Type: Editorial From: Journal of Property Investment & Finance, Volume 28, Issue 4

After years of soaring prices, housing markets have experienced severe price declines in several countries. The US housing market has dropped by 25 to 30 percent from the peak of the market in the second quarter of 2007 to the end of 2009, with the Miami market, for instance, experiencing a decline by almost 50 percent. In the UK, house prices went down by almost 20 percent between 2007Q2 and 2009Q1. During the last three quarters of 2009, however, the UK market exhibited some signs of recovery, with an increase of approximately 8 percent. Other housing markets have also suffered in recent quarters, this being especially true of the Irish and Spanish markets. The French housing market has also started to decline, although price changes have so far remained quite contained in comparison with the changes seen in the above-mentioned countries.

Obviously, the housing sector has significant impacts on the overall economy. One example is the role of housing equity, which is directly related to the change in the value of a house, on household consumption. Another example is the impact that mortgage defaults can have on the stability of the financial sector. The recent US experience is a good example of the tight linkages between the housing and financial sectors, and of the problems that occur because of those linkages. Given the importance of the housing sector, much research has focused on the determinants of house prices. Factors such as gross domestic product, interest rates, and changes in the population have typically been used to explain house price changes.

Assessment of the determinants of house prices could potentially benefit by recognizing that a house is a combination of a lot and a structure. Indeed, fundamentals affecting the prices of the two components are not the same. In equilibrium, structure value will equal depreciated replacement cost; if value exceeds cost, developers will construct more houses profitably and if the reverse is true construction will cease. The construction of new houses will depress prices on existing houses toward replacement cost, and the lack of construction, in the face of population and income growth, will eventually raise price to replacement cost. Thus, structure values move by and large in line with a construction cost index.

The determinants of residential land values are more complex. The economic might of a metropolitan area or region is the primary determinant of the demand for land. This can be measured in terms of employment and income per employee or population and gross domestic product per capita. Real interest rates are also relevant because they relate expected future land rents to current values. Real mortgage interest rates are an appropriate measure because they reflect the cost of financing owner-occupied housing. The demand for urban land is somewhat easier to measure than the supply. Supply is constrained by topography and land use regulations, both of which can have substantial impacts on land values and house prices.

If the relative value of land and structures were constant over time and locations, disentangling the two components would be of little or no importance. However, land leverage, i.e. the ratio of land to property value, will be greater in more highly populated areas than in rural regions, and also greater at the peak than at the trough of a housing cycle.

Recent research has focused on this feature of properties. Bosticet al. (2007) and Davis and Heathcote (2007) were among the first to emphasize the relationship between land leverage and house price appreciation, a focus that was apparently motivated by a desire to explain the disparities in house price appreciation rates in different parts of the US during the late 1990s and early 2000s. Bostic et al.’s empirical study demonstrates that land leverage is positively associated with house price inflation. Davis and Heathcote (2007) show that from 1975 to 2006 the real price of residential land in the US nearly quadrupled, while the price of structures grew by only one-third. By 2006, land accounted for 46 percent of aggregate residential property value, compared with 35 percent in 1975.

Similar trends exist for other countries. Bourassa et al. (2009), for instance, use data for three New Zealand cities and find that the prices of houses with greater land leverage are more volatile over the course of the property cycle. Also, a study of France shows that land leverage for single-family houses is almost 50 percent in the Paris region, and as low as 18 percent in more remote parts of the country (Commissariat général au développement durable, 2009). A study that we are currently conducting also suggests wide disparities across locations and through the real estate cycle in Switzerland.

Of particular interest in explaining house price changes across both locations and time is modeling the two components of houses (i.e. land and structure) separately. Davis and Heathcote’s (2007) strategy is to regress real house, land, and structure prices on a set of fundamentals that includes real per capita income, the nominal three-month Treasury Bill rate, and the inflation rate. A second set of regressions adds three additional variables: population, the percentage of the population aged 35 to 54 (the primary home-buying category), and the spread between 30-year fixed mortgage rates and the three-month Treasury Bill rate. The land price regressions in particular perform better than house price regressions, and the estimates for the house price regressions appear to be a weighted average of the estimates for the land and structure price regressions.

Research on land leverage is often hampered by the lack of appropriate land price indexes and/or accurate measures of the ratio of land to total house value. However, in countries where large databases are available, for instance because hedonic mass appraisal tools are in place, this does not constitute a problem. My view is that research on land leverage will be high on the agenda of many researchers, in particular in the UK where hedonic models are widely used. Again, such research will help to understand better why house price changes vary so substantially across locations and time.

Martin HoesliUniversities of Geneva (Switzerland) and Aberdeen (UK) and BEM Management School (France)