House prices are still falling

[BrilloPad]

http://www.telegraph.co.uk/finance/5...l-falling.html

House prices are still falling
It may be true that the path to true love does not always run smooth. That is also true of house price declines. Last week the Nationwide reported that house prices rose by 1.3pc in July. Great, the slump in house prices is over. Actually, probably not.

Three of the last seven months of data from Nationwide have been down and four up. The latest data from the Halifax show four down months and only two up. So what is going on?

First, these data are seasonally adjusted by both the Halifax and the Nationwide. That is to account for the fact that there are regular monthly patterns every year. This is a problem now as patterns in the past when house prices were rising aren't very helpful in a period of falling prices.

Second, estimates of monthly house price changes aren't very accurate when the number of house sales are small, so you get lots of variation in prices each month. Over the period June 1989 to July 1995, for example, when average house prices fell from £70,095 to £60,965 there were 23 months where prices increased and three where they were flat and 48 when they fell.

How much further will house prices fall? The best guide is the ratio between average earnings and average house prices. This is a measure of affordability. Between 1983 and 2001, before house prices started to climb, the ratio averaged 3.62. By July 2007 the ratio had reached 5.84; it has subsequently fallen back to 4.33.

To get back to the long-run average of 3.62 from 5.84 implies a drop of 38pc. So far we are down 26pc, so it looks like there is more to go. The possibility is that house price falls will be even greater than that if the ratio falls below its long-run trend before recovering, as it did in the early 1990s.

House prices probably have a good way to drop yet. Lots of people may tell you otherwise − estate agents, mortgage brokers and bankers − because they have something to gain. My advice is just to look at the data.

============================================

A very good article : but one small point. Earnings growth is not totally flat.

[centurian]

Second, estimates of monthly house price changes aren't very accurate when the number of house sales are small, so you get lots of variation in prices each month.

The thing about all house price surveys is that they generally measure sales that have occurred, so they only count transactions where someone was prepared to pay the asking price.

It doesn't count all the transactions where the seller wouldn't lower their price to get a sale (i.e. how much buyers were prepared to pay).

You only need a few mugs who are so desparate to get onto the market that they will pay over the odds to get onto the ladder. Meanwhile almost all of those houses that are dropping in value don't get sold, so are not reported. This will skew the figures.

If sellers were forced to report their best of five offers in the past 3 months and these were counted, we'd quickly see the true market value.

[PM-Junkie]

The thing about all house price surveys is that they generally measure sales that have occurred, so they only count transactions where someone was prepared to pay the asking price.

It doesn't count all the transactions where the seller wouldn't lower their price to get a sale (i.e. how much buyers were prepared to pay).

You only need a few mugs who are so desparate to get onto the market that they will pay over the odds to get onto the ladder. Meanwhile almost all of those houses that are dropping in value don't get sold, so are not reported. This will skew the figures.

If sellers were forced to report their best of five offers in the past 3 months and these were counted, we'd quickly see the true market value.

Not true. Just one example - the land registry survey numbers are based on sales, and what was paid.

[BrilloPad]

Not true. Just one example - the land registry survey numbers are based on sales, and what was paid.

True. But repossessions are not included. Which is the point centaurian was trying to make?

[Menelaus]

Not true. Just one example - the land registry survey numbers are based on sales, and what was paid.

WHS.

The estate agent numbers are ridiculously over-inflated, hence using HMLR data in automated valuation models rather than estate agency data.

[BrilloPad]

WHS.

The estate agent numbers are ridiculously over-inflated, hence using HMLR data in automated valuation models rather than estate agency data.

Would you trust an index generated by estate agents?

This guys are nearly as bad as recruitment agents.

[Menelaus]

Would you trust an index generated by estate agents?

This guys are nearly as bad as recruitment agents.

Yep.

I managed to avoid saying "lying bastards" before but it was a close-run-thing.

[BrilloPad]

Yep.

I managed to avoid saying "lying bastards" before but it was a close-run-thing.

Dont say that. its unfair. to lying bastards.

[Menelaus]

Dont say that. its unfair. to lying bastards.

True dat.

[centurian]

Not true. Just one example - the land registry survey numbers are based on sales, and what was paid.

Exactly. What was paid on completed transactions. It doesn't capture what people were actually prepared to pay.

A subtle, but very important difference.

If you want to sell your house, you want to know what you can actually get for it - what is the market price.

The current figures are showing the price you might get - IF you are very, very lucky enough to find a mug of a buyer.