More Spanish Mortgage Data

Indexing INE mortgage issuance data to the beginning of the series produces the chart on the left. Some things stand out.

1). From early 2005 through at least late 2007, one can note a pronounced deviation of totals loaned from number of mortgages contracted. This reflects price increases, the greater willingness of bankers to finance beyond 80% of the value and the vibrancy of the higher end market;

2). The marked volatility of the same two series from mid-2005 through most of 2008 is (aside from the expected effects of Easter, Christmas and August vacations) was a result of the sheer volume of pre-sold new projects coming on line. Work is finished, approvals are granted and suddenly a 200 unit promotion represents 200 new mortgages. The data will have a tendency to clump*;

3). The noticeable improvement in both numbers and value of mortgages contracted in recent months may be the result of pent up legitimate demand usatisfied during the credit crisis. The lag between approval and the actual depositing of the money in a creditor's account is likely in the range of three months. But it probably also represents more certainty on the part of financial institutions concerning their ability to assess risk and a greater willingness of potential buyers to proceed. Clearly, these do not expect more negative surprises on the macro-economic, or personal, front;

4). Despite the fact that both numbers and total value of mortgages have stabilized in recent months, their average value continues to fall. This is possibly representative of price adjustments, the new found common sense of bankers and the utter collapse of the luxury trade.

-------------------------

*A perfect example of this effect is the latest offering from the multiple loudspeakers of Edward Hugh's re-education camp. His proposal that the fact that the assessed values of new dwellings have tanked in various (but by no means all) municipalities in Madrid is representative of anything at all reveals his utter lack of comprehension of the data presented.

The assessment of new projects is done on a mass basis. This is to say that when a new project of, say, 100 units, is finished the assessor moves in and does them all. Hence, the series is skewed by the sudden appearance of that number of near identical (in price per square metre terms) homes - the totality of these statistically, in the real world, being only one data point.** This does not happen in the case of second hand residences, which thus constitute a much larger, more reliable and independent sample. The difference is evident in the chart above. Note the near tenfold difference in the results for Colmenar Viejo - and the, interesting in its own right, coincidence of the two categories in many of the lower income suburbs to the south of Madrid city.

As an aside, the offending entry itself was a repeat, without any further analysis, or probably even bothering to look at the source data, of an article in idealista.com - a website exclusively dedicated to marketing residential properties. The reader, and Mr. Hugh, would best consider this a mild, and legal, form of bait-and-switch.

The Ministerio de Vivienda source spreadsheet can be found here.

**Magnifying this effect was the continued diminution of apartment size over the course of the period, and the corresponding increase in numbers built and later assessed per buildable surface, as developers tried to keep price entry points constant without prejudicing their margins.

---------------------------------