2008 - 613,512A short note, directed to those who might be swallowing this story whole, about how their truly stupid methodology guarantees that Spain's housing stock will never meaningfully diminish - no matter how much of it is actually sold - would be in order.
2009 - 688,044
2010 - 687,523
2011 - 676,038
Fomento's methodological guidelines with respect to new home sales defines them as 'dwellings of new construction'. Unsatisfied by this unhelpfully obvious (and of no legal standing) definition, the writer proceeds to dig and finds out that the same ministry's house price index methodology defines them as 'between 0 and 1 years of age', an oddly limiting restriction that is later amplified to a more acceptable level in the link to the actual spreadsheet as prices for homes 'up to two years old'. Finally, the unsold housing stock is calculated by subtracting new sales from construction completion approvals as (properly) recorded by the College of Technical Architects.
The blue line shows new home sales as a percentage of total home sales, according to Fomento. Bearing in mind that the national statistical agency defines new as 'obra nueva' - a legal description used by the property registrars that provide the data and which has no relationship whatsoever to the actual age of the structure - the red line shows the same ratio according to the INE.
The correlation between completions lagged two years and Fomento's relegation of sales by age is evident. To date the effect has been to assign 62,000 INE new sales* to the Fomento used bin, even as total sales approximately coincide. But that 56,000 of those are attributable to the first three quarters of 2011 should give the reader a taste of things to come as housing starts four years earlier, currently at less than 10 percent of their peak, become the factor that controls Fomento's calculation of new home sales.
*A future entry will deal with legalities dictating the property registry description of a transaction that, given the massive intervention of banks in the market, also radically skew sales stats towards the second hand column.