Wednesday, 27 February 2013

Wildly misleading statistic of the day: "41% of homes sold at a loss since 2007"

This got an "oh woe is us" write up in some of the papers, but let's refer back to the original press release from Castle Trust:

Over 130,000 families have sold their homes at a loss since 2007, according to exclusive analysis of housing transactions by housing investment and shared equity provider, Castle Trust.

The initial research, which tracks the proportion of properties selling at a profit or loss, includes an analysis of properties in England and Wales which were bought and sold between January 2007 and January 2013. Of these properties, 40.7% (131,442) were sold at a loss, with the average shortfall being £24,430 (on average 11.0% of the house price).

Over the same period, 55.6% (179,689) of homes sold for a profit generating an average return of £45,199 per transaction (on average 20.4% of the house price) and the remaining 3.7% (12,051) sold for the purchase price.

Let's gloss over the fact that lower house prices do not represent "a loss" for the honest hard working population of this country: the result of lower selling prices is that the purchaser saves more in mortgage repayments than the vendor loses in (negative) return on the cash he invests from the sale. So from our point of view, that's a significant gain and it's only a loss from the banks' point of view.

Let's focus on those two headline numbers: "130,000 families" and "41 per cent".

Readily available statistics, for example HMRC's Property transactions completed in the UK with value £40,000 or above show that there were 5,433,160 sales in the six calendar years 2007 to 2012, plus an unknown number of sales for £40,000 or less.

So either 41% is correct and about 2,200,000 were sold at a loss; or 130,000 is correct and 2.4% were sold at a loss. Or, more likely, both figures are completely wrong.

RETHINK: unless of course they mean homes which were bought after January 2007 and then re-sold before January 2013, in which case the 41% figure is probably about right, seeing as on the whole across England & Wales, house prices have been flat for the last seven or eight years; we'd expect half to have re-sold for a higher and half to have re-sold for a lower price.


The Cowboy Online said...

I do wonder how those losses have been calculated as well. Would re-mortgages, other loans secured against the property, or simply comparing the sale against peak property value in the period prior to sale be included in the calculation?

I wonder if John Prescott's PathFinder folly was part of the calculation?

Mark Wadsworth said...

TCO, apart from the fact that the figures are out by a factor of ten in either direction, they say it means actual purchases minus actual sales.

mombers said...

I think the discrepancy comes about because not all of the 5,433,160 sales were sales of homes bought in the Jan 2007 to Jan 2013. E.g. a house sold in 2012 but bought in 2006 would not be counted

Mark Wadsworth said...

M, yes, see the post script headed "rethink".