The Law Of The Many Zeros

Despite the obvious hidden cost of later being forced yet again to shave the palm of our right hand, Mr. Sánchez' comment to our last post that perhaps house sales in Spain were continuing to fall precipitously has us thinking - and not just about exactly how we are going to defend our assessment that the line was on the verge of flattening out, evidence notwithstanding. But first we must deal with The Law of the Many Zeros.

Judging by the frequency with which the number appears, it seems to have been widely accepted that there are one million (1,000,000) unsold new homes in Spain. And if one is to believe this blog entry - dated December 25, 2008 - and this - published September 14th last - this figure has now passed from the realm of the considered estimate to that of the universal constant. Such is the near-Einsteinian attractiveness of a number that is not a zero followed by a long string of those that are. The dark matter binding Spanish real estate to its fate of eternal entropy.

The power of Many Zeros is sufficient to utterly negate the fact that 187,387 new homes were sold in the above period, work was started on approximately none and huge numbers of works-in-progress were abandonned for lack of funding or market. This is because when a pundit tacks a huge line of zeros behind another number, the uncritical and desperate-for-truth everyman cuts the commentator a tremendous amount of slack when it comes time to render accounts. As an approximation, one might consider what is our guess as to this effect.

1). An estimate of 1,000,000 (six zeros) will be considered to have been correct if the 'truth' turns out to be anywhere between 800,000 and 1,200,000;

2). Another of, say, 900,000 will likely only get the benefit of the doubt if the value turns out to be 100,000 above or below;

3). However, if some fool were to go out and actually count the number of unsold new homes and come up with 943,627, the rabid masses will have his balls on a platter, hang a dead cat from his lintel and pelt his person with overripe tomatoes whenever he ventures to leave the shelter of his home for a loaf of bread if he misses by anything over 10,000;

4). Estimates, on the other hand, over one million are merely redundant unless one uses the geometric multiplier effect of amplifying the first number. Thus, guesses like Edward Hugh's 3,000,000 will probably be considered to have been accurate if the actual sum turns out be anything over one hundred and three.

Returning to the Random Spaniard's comment, we thought it might be half interesting (at least for Gillette's bottom line) to calculate to what degree current new home sales were managing to decompose Spain's stock of new housing. To this end - on the, we believe reasonable, basis that virtually no new inventory has been added in the last year - we've done a monthly calculation of the number of years it will take to deplete the offer at trailing one year sales volumes for the last twelve months... and we've done it for proposed housing inventories of 1.2 million, 1 million and 800,000 units.

The procedure is as follows:

1). Calculate the rolling 12-month sum of sales for each month;

2). Divide, in month 1, the inventory by the above figure;

3). In month 2 (and ensuing months), subtract prior period monthly sales from previous stock and divide the result by the yearly sales for the current period.

The end result of this is an ongoing approximation of how many years it will take to work, at current levels, the inventory of houses to the impossible and undesirable level of zero. And the surprise is that, for a starting figure of 1 million homes, the resulting number is incredibly stable - ranging between 3.53 and 3.62 years over the stretch. For 1.2 million the number trends slightly upward, rising by 0.18 years to 4.46 from November to October. For 800,000, the figure falls by 0.18 years to 2.67.

Not just a matter of doing our bit to reglue and polish up the warped and tarnished veneer of science once sported by the study of economics, the relative constancy of these results - they keep selling houses and the number of years supply doesn't budge - leads us to believe that Spanish housing is, in fact, a radioactive element. We've called it habitatium and calculated, with the help of 1728.com, the half life for the three versions so far discovered.

Habitatium 1,200,000 - 3.72 years

Habitatium 1,000,000 - 3.04 years

Habitatium 800,000 - 2.36 years

So, 'Woooo!', says the reader. 'At the end of October, 2011 there are still going to a half a million unsold new homes!'. Undeniable yes, but the real question to be asked is 'Where are they going to be?'.

The last time we made note of one these estimates (and admittedly not knowing if the expert was an adherent to the Many Zeros faith), it was stated somewhere that there were 450,000 finished and unsold homes along the Mediterranean coast - Spain in its entirety to many readers from northern Europe as well as Telegraph readers who deny such affiliation.

These places, by all accounts, are still unsellable. So, some guess as to how many years supply of new Spanish dwellings not located on the Med (of which there might only be 550,000) might be appropriate. We'll do that.

Arbitrarily assigning 70% of all sales to the 55% of properties that are non-Mediterranean turns a stock of 2.6 years. Raising it to a not entirely unrealistic (given the lack of foreign buyers) 80% lowers this to 2.16 years - product of the lessening of the inventory from 550,000 to 386,000 units over the period.

Half-life? Not applicable because the rate of decay is increasing.

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