The real estate market
|
|
|||
|
Real Estate Watch
|
|||
|
|
|||
|
An indicator of real estate crises
|
|||
|
|
|||
|
Paul van den Noord, an OECD economist has just published a working paper ("Are House Prices Nearing a Peak? A Probit Analysis for 17 OECD Countries") which proposes an econometric model for estimating the probability of property prices "peaking" in developed countries.
|
of the estimate and the limitations of the model as population is not taken into account, the positive evolution of the equilibrium price trend becomes a fundamental factor for explaining the existence of real estate investment funds and companies as long-term investment instruments.
|
||
|
|
|||
|
The model defines a "property peak" as one which is preceded by six quarters when the cumulative increase in the price of housing is more than 15% in real terms and is followed by six quarters when there are decreases in real housing prices. The model relates the probability of there being a real estate crisis with:
|
According to the model, with the latest figures drawn up by the Ministry of Housing where there is an increase of 12% in the price of the square metre of open-market houses appraised in the first quarter of 2006, the probability of real housing prices falling in the next six quarters is of just over 10%.
|
||
|
|
|||
|
a) past performance of housing prices (such as acceleration in the performance of the price),
b) the house price gap defined as the difference between the current price and the trend and
c) nominal long-term interest rates (which affect the cost of the debt).
The results for Spain are summarized in chart 1, which shows the evolution of the probabilities estimated by the model, the periods of "property peaks", the behaviour of real housing prices and the trend.
As can be observed in the chart, in Spain's case, the series shows three "peaks": one corresponding to the period 1973Q3-1974Q2, one to 1977Q2-1978Q1 and the last one observed in 1990Q4-1991Q3, all of them shaded in on the chart. In these periods, after observing cumulative increases of 17%-20% in real housing prices in the six preceding quarters, there was a turnaround in price behaviour and in the six successive periods cumulative downturns of around 10%-12% in real terms were observed.
Comparison of Crisis Probability vs Evolution of Housing Prices
|
A simple exercise the model enables us to do is to infer the possible probabilities of a real estate crisis in the second quarter of 2006, conditioned to future scenarios for housing prices in the said quarter (yet to be published).
Assuming that the future distribution of the price of housing is similar to what was observed in the past (characterized as normal distribution), the model makes it possible to estimate crisis probabilities for the coming quarters of 2006. Given the current interest rate levels and the recent performance of housing prices, the model indicates that there should be quarterly increases in housing prices of around 5% or annual increases of around 20% in 2006 so that the probability of an adjustment in prices occurring would be of 40%.
The results for the year are shown in the chart below. The left-hand axis shows the accumulated historical probability of each of the quarterly increases in housing prices in Spain, while the right-hand axis shows the probability of crisis estimated by the model for each of these increases.
Probability of Real Estate Crisis
Historical comparison and Probit model
|
||
|
|
|||
 - 0019-1.jpg) |
 - 0019-2.jpg) |
||
|
|
|||
|
This positive trend is a pattern that is repeated in 16 of the 17 OECD countries analyzed, with Germany the only exception in the period observed (1970Q1-2005Q4). To calculate long-term growth, a simple regression model was used, the Ordinary Squared Minimums, on the logarithm for the real housing price. Trend calculated in this way for Spain would be around 3.5% per annum. These appreciation trend is slightly higher than the average OECD increases, which are at an annual average of around 3%. Despite the simplicity
|
-8% -6% -4% -2% 0% 2% 4% 6% 8% 10% Increase 2006Q1-2006Q2
Source: BBVA
However, it is worth considering that although the model "gives" intuitive results when it comes to estimating the probabilities of crisis or real estate peaks (this grows with increases in interest rates and faster growth of housing prices), and fits in very well with the first two crises (19731977), it was not so accurate in the 1990 crisis, where it began to "show" significant crisis probabilities as from 1985, five years before the peak occurred.
Mikel Sanz Risks UCRAM - BBVA
|
||
Find beautiful russian women for marriage here! . Check Vienna hostels with guest reviws.