'UK house prices have a long way to fall yet
Anyway, back to Britain’s favourite obsession – house prices. The standardised average HBOS UK house price for November came in at £163,605, which takes house prices back below the January 2005 level, and marks an 18% drop from the August 2007 peak.

However, HBOS note that the house price to earnings ratio (which peaked at 5.8 times having come up from around 3 times in 1995) is now a more reasonable 4.56 times. So you could be forgiven for thinking that the house price correction, at least in terms of affordability, was approaching the half way mark. Unfortunately this appears to be more due to an apples and oranges comparison rather than more reliable metrics.



For example, first-time buyer (FTB) affordability based on FTB house prices (Nationwide data) and mean full-time worker take-home pay, is only just starting to turn better as house prices fall, but still has a very long way to go (see chart below).

The last decline, peak to trough, took seven years. Affordability this cycle was worst (highest) in the December 2007 quarter, so at best, we are just one year or 14% of the way through the affordability correction on this measure.

The HBOS house price to average earnings ratio was discontinued in the first quarter of 2006, when it recorded a then-record high 5.68X (see chart below). However, HBOS notes that the series peak was 5.84X in July 2007, even though HBOS’s own measure of house prices rose by more than 14% between 1Q 2006 and August 2007 (implying a peak like-for-like ratio of nearer 6.3X). In November this ratio had fallen to 4.56X, according to HBOS.

With the 25-year average around 3.9X and the low in December 1995 of 3.01X, the question begging an answer is: are these numbers directly comparable? The correction by this metric appears 40% complete after little more than a year, even though the last cycle took six and a half years from peak to trough.



However, given that we know that the HBOS standard average house price is now £163,605 on a seasonally adjusted basis, we can deduce from the earnings to price ratio that HBOS uses an average earnings figure of £35,878. If this looks high, that’s because it is.

Perhaps in order to make the multiple not sound too bad, HBOS uses the very highest measure of average earnings (namely males in full-time employment and including bonuses and inducements) as its average earnings number. Whether they did so before is not a matter of record.

The 2008 Annual Survey of Hours and Earnings (ASHE) released on 14th November shows that in April, mean weekly earnings for full-time adult employees in UK and Northern Ireland, were £578.20 a week. That of course equates to gross annual pay of £30,066.40. This exactitude is important because HBOS’ previous series was discontinued and there is a wide difference between average wages for all employees, full-timers and between men and women.

Nationwide data goes back further than the HBOS numbers and by using the closest National Statistics subset to the old dataset (full-time employees) we can create a house price to earnings ratio that is directly comparable with the past. On this basis, house prices (according to Nationwide) are actually now still 5.5 times earnings and were perhaps as high as 6.4X at the 2007 peak (see chart below).

So, on a proper apples to apples comparison, the house price correction is not actually 40% through, as HBOS’s data might have you believe, but rather only about a quarter of the way towards the 1995 floor ratio. Indeed, compared to the 1989-1995 correction, UK house price to earnings multiples need to fall a further 7% just to return to the previous cycle peak of 5.1X.

The silver lining in all this is that the HBOS house price data certainly appears to be ahead of the Nationwide data, perhaps because of HBOS’s higher London and South East exposure. So although that suggests that the Nationwide data will deteriorate quickly from here, Nationwide house prices are only just over 10% down from their peak. That means the final likely sell-off is less than if we applied the Nationwide ratios to the HBOS price falls. Even so, if the old floor of 2.95 times earnings from 1995 holds true again this cycle, that would imply house prices might fall to £98,000 over the next five years, down a further -40% (assuming 2% a year earnings growth through the recession).

Now, there’s no particular reason why the house price to earnings ratio would necessarily have to fall back to the lows seen in 1995. House prices in the UK are a (lagged) residual, all other things being equal, of the debt service burden as a percentage of after-tax income.

By late 1995, UK base rates had hit 6.75%, while RPI inflation was 3.9%. So real base rates were 2.8% and real mortgage rates were 225 basis points above that, say 5%.

Nominal mortgage rates were thus around 9% in 1995 and on a 25-year repayment, annual equivalent payments were around 10.1%, or 6.2% after inflation (a good proxy for earnings growth). 2.95 times salary at a real 6.2% mortgage rate equates to 18.3% of income. How does that compare with today?

Best-buy mortgage rates are currently around 5% (maybe falling to 4% after the rate cut) so annual repayments are equivalent to about 7%. But here’s the rub. Although officially RPI inflation is still 4.2%, it’s falling like a stone and could actually be negative (deflation) within a year. Meantime house prices are still 5.5 times earnings. For 2.95 times earnings to provide the floor, inflation must not go negative (4% mortgage rates equate to an equivalent 6.33% in annual repayment costs).

But there’s another problem. Banks are being forced to pass on base rate cuts to existing mortgage holders by political pressure. They are responding by ceasing new loan provision almost entirely (the CML warned that net mortgage lending will be negative next year). House prices are not determined by homeowners but by homebuyers. Many best buy (on price) mortgage products now require 40% deposits. If you don’t have a substantial deposit, the rate, should you qualify at all, is going to be substantially higher than this analysis assumes.

There is still every reason to think that the house price to earnings ratio floor this time could feasibly be 2.95 and quite possibly even lower'

I naturally am still using my back of fag packet figures for average salary of £20k.

James Ferguson of Moneyweek examining these issues in early december 2008.I think the use of mean averages is misleading in the extreme.As is the use of 'male' earnings-cos wimmin doesnt wanna to buy da house innit.

When this bubble finally settles,most people will be telling you never to invest in property as it only ever goes down etc etc and naturally that's when you should be buying.We are nowhere near yet IMO.