The situation of Spain’s housing sector is defined by a building sector that is at all-time lows, housing prices that are bottoming out, nearing 0% growth, and a market where transactions are relatively stable at a minimum figure of 300,000 transactions per year. Over the last few years the demand for housing in the market has accumulated due to the potential households that are not being created and the high degree of mobility amongst the population. Homeownership is highly affordable, despite two important limitations that restrict entry to the market: credit (limited by means of high cash requirements equivalent to 45% of the house’s price) and the lack of stable employment, where salaries are frozen. Consequently, prices have not recovered and the supply of newly-constructed housing is not growing, despite the fact that it is needed.
The short-term prospects for the evolution of this sector depend directly on the economy’s recovery as a whole. If the negative forecasts for the third recession do not prove true, housing prices will again recover slowly and the building sector will most likely follow their lead with a difference of about six months. Creating incentives for building in places where demand has accumulated would work as a mechanism to accelerate the process, which will, in any case, be slow over the next few years.
- Different indicators for housing prices
One of the most controversial questions that has arisen over the past few months is the disparity that exists in the explanations for housing price indices, depending on the statistic source used and also on the opinions of the companies that generate their own price databases. Although it is not a question of much dispute in Spain, it is a well-known point of discussion in publications related to housing prices. One question that has raised a lot of arguments is the methodology used to construct these pricing indices. Most countries where such indices have been developed use three types: weighted indices, those based on hedonic pricing, and those involving repeated sales adjusted by hedonic methods. Having several indices provides one with more tools for understanding the evolution of housing prices. For example, the United States uses the following housing price indices: the NAR (weighted), the OFHEO (repeated sales of mortgages on homes in the secondary market, adjusted for quality) and the Case-Shiller index (repeated property transactions, adjusted for quality).
The underlying reason why there are multiple indices is because it is difficult to measure prices accurately for housing, as it is an asset that is heterogeneous by nature. Housing prices vary radically, depending upon their location, age, and the different ways their characteristics are assessed, depending upon on the culture and/or conditions of the demand that exists in the different markets. When this is what happens to an asset, then the evolution of the relevant price may be radically different from the average if the data used as the basis for the statistics is all concentrated in just a few markets. Therefore, the first requisite that is necessary for an index and its generality to be acceptable is that the observations that are used to calculate such prices are actually representative of each market included in the index. If they are not, prices and price evolution are going to be biased towards the relevant market or markets observed.
Weighted indices do not take quality into account, in other words, they do not take into account the differences in home quality when determining price. They only evaluate the price observed/obtained. This is the case of the index issued by the Ministry of Public Works, although the weighting used therein, based on the assessments provided to it by the AEV members, is very precise from the perspective of location, which makes it a high-quality index.
The index issued by the “INE” (Spanish National Institute for Statistics) uses a hedonic base, therefore, housing quality is taken into account when constructing the index. This index also includes (and weighs on the basis of) the location, thus making it an index that reflects such characteristic very well. The methodology that it applies is a common methodology used by other indices in the European countries, thus making it possible to compare the evolution of housing prices calculated in this way with those of the EU.
Being that the INE index takes quality into account, it can detect changes that arise in relation to surface area, rooms, type of home, etc. This is the main difference it holds with the index issued by the Ministry of Public Works, in addition to time lags and the use of data declared in public deeds. The unavoidable changes that arise in housing quality throughout the building cycle can cause hedonic indices to be more volatile than weighted ones, which means that their results can vary at certain points in the cycle.
This difference is less noticeable when there is a large volume of transactions in the market, but it is much more noticeable in small markets. This occurs because when there are many elements observed, their characteristics tend to converge with the average and price valuation coincides with the strictly mathematical average. But when there are few elements observed, said elements tend to become concentrations of units with similar characteristics (or similar locations), which makes the averages calculated to reflect quality, different from those using weighting. Said differences arise and they cause the calculated indices to be different. This is what happens in Spain’s two price indices, as illustrated by Figure 8.
As one can see, the different volatility in these two indices is significant and the index issued by the Ministry of Public Works is the most stable of the two. Said volatility could indicate a situation where there are changes in the quality of the homes during the observation period (although it can also come from the contracting parties’ tax behavior), so that said changes absorb their part and accentuate the contraction in the price component that does not take into account the “cost of housing quality”. In other words and according to the latest data, a deeper plunge in this index would seem to imply that it is measuring higher quality units that are unidentified by the market and that, if this component is eliminated, it produces an adjusted price that falls faster than if it were not adjusted for the component, and vice versa. Another reason for the variance in evolution is found in the lower number of observed elements in all the typologies, which could cause some specific elements to “weigh” more than others within a hedonic index. For example, some studies find that in periods of recession transactions are concentrated in those involving smaller and cheaper homes, while in periods of expansion, all types of homes are sold. This is a cause for disparity that would be reflected in a hedonic index but not in a weighted one.
The INE index shows that prices have stopped falling in 2014, whereas the weighted index issued by the Ministry of Public Works continues to reflect a negative trend. If one assumes that both indices have the same observed elements, the data of the INE index could indicate that most homes included in the index in 2014 are of lower quality (smaller, located in secondary areas, with less services…) and the data of the Ministry’s index would only calculate an average per location for all the homes in the sample.