Wednesday 22 July 2009

Fun with numbers

An excellent bit of number crunching by Cliff D'Arcy at Yahoo:

I compared the quarterly unemployment data from the Office for National Statistics with house-price data from Nationwide BS, over the past 26 years. The result: one goes up, the other goes down

According to my nifty spreadsheet, the correlation coefficient in my example was -0.81. In other words, when unemployment rose, house prices dropped 81% of the time , and vice versa. This is quite a strongly negative relationship.

Alas, what this calculation doesn't tell us is whether this is a causal relationship. In other words, we can't claim that falling house prices cause higher unemployment, nor the other way around. After all, both could be the result of another factor, such as rising or falling economic growth, and there may well be no direct link between the two.

Then again, as I said earlier, unemployment is a lagging indicator: it takes time for joblessness to hit consumer confidence and, ultimately, house prices. Introducing an 18-month time lag between the two sets of data increased the correlation coefficient to -0.84, producing an even stronger association.

In summary

If unemployment does rise from today's 2.38 million to as much as 3.2 million, as forecasters predict, then I suspect that this spells bad news for house prices.

Indeed, I don't expect to see any sustained increase in house prices until unemployment peaks and starts to fall. This could happen in the second half of 2010, but I'd be mightily surprised if we saw any lasting recovery in property prices before then.

In short, while we may see a few upwards blips in the monthly house-price data, don't expect a true dawn for house prices in 2009...


H/t K8te at HPC.

5 comments:

CityUnslicker said...

There is a difficulty in that the severity of the crash last year will make numbers look better at the end of this year, and indeed is doing so now.

As there was a period of virtually no sales and no mortgage approvals, any YOY comparison post October this year is going to be skewed.

Also if activity dropped below what i would call minimum (i.e. people who have to move, moving), which it likley did, then activity and prices may pick up relative to the overall trend.

None od this will deny the reality of much lower volumes, mortages and ulitmately prices; but it will make the stats awfully confusing for the rest of this year and q1 next.

Mark Wadsworth said...

JH, widget added.

CU, yes, there is all sorts of mucking about with statistics (in which I am world champion). But housing markets are supremely localised - as long as prices in my area are dropping by more than the amount we pay in rent, I shall continue to sit tight.

Lola said...

"If unemployment does rise from today's 2.38 million to as much as 3.2 million, as forecasters predict, then I suspect that this spells bad news for house prices.
"

Or from my point of view, good news for house prices.

AntiCitizenOne said...

Good news for house affordability.

Mark Wadsworth said...

L, AC1, indeed, it's just a shame that affordability comes at the price of a million people losing their jobs ... now if it were a million quangista losing their jobs, that would be a different matter.