House Prices UP 3.5% YOY. Land Registry

Bestthingsinlifearefree

vishpatel wrote:

Anyone want to take a guess on what comes next? And after that....

There was always going to be a little resistance to HPI going negative. Vendors refuse to budge on their asking price, and so volumes go down.

Volumes ARE down... tick tock

Hi Vish,

If anything you maybe surprised when the next Land Reg figures come out they may well surge higher because there is a "lag / time delay" in the data of a couple of months.

These current figures would not have taken in the the recent upward turn ie the cut in interest rates, increased mortgage approvals and surveys from ODPM,Halifax, Nationwide all showing an upward movement in price growth in the last one to two months.

So alas those hoping for price falls / price crash in 2006 are likely to be severely disappointed (just as they were in 2003,2004,2005) as the momentum in the market is now upwards / positive.

Doozergirl

vishpatel wrote:

Anyone want to take a guess on what comes next? And after that....

There was always going to be a little resistance to HPI going negative. Vendors refuse to budge on their asking price, and so volumes go down.

Volumes ARE down... tick tock

Volumes are down YOY; since Q2, London has seen a 30% increase in completions, the South East over 41%!

devils_advocate

I believe that prices have fallen, and that the Land Reg figures are correct. A contradiction, you may think. But in order to interpret the data, you have to look at ALL the data.

Take a hypothetical postcode. In Q1, 2 houses are sold at 300k each, and 5 flats are sold at 100k each. Average price per Land Registry is 157k. In Q2, 2 identical houses are sold for 250k each, and 2 identical flats are sold for 75k each. Average price per land registry is 162k.

The conclusions from this example that many of you are drawing is that house prices have increased quarter on quarter. On the face of it, thats correct. But looking at the underlying detail, prices have fallen significantly.

So the fact that volumes are down is very significant, but you can't draw valid conclusions from the land registry data unless you know the mix of properties which have sold.

meanmachine_2

Whichever way both sides try to spin it, it ain't a crash and it ain't looking like a crash.

A shame since that'll mean a whole generation's wages are going into banks and agents' pockets - which eventually means armageddon on the high street no doubt.

But for now, again all it does is confirm my belief that Brits are barmy and obsessed with bricks rather than other matters such as language, learning and culture. Never has "bestthingsinlifearefree" been more of a misnomer. Clearly all he/she/it cares about is screwing as much money out of poor old FTBers (who are increasingly poor and old) as possible.

A real shame. And it puts me off buying even more now.

I think I'll sit this one out for 5-10 years and continue to rent at well below market value.

nelly_2

Funny you say that as I'm learning Italian.

zar

devils_advocate wrote:

I believe that prices have fallen, and that the Land Reg figures are correct. A contradiction, you may think. But in order to interpret the data, you have to look at ALL the data.

Take a hypothetical postcode. In Q1, 2 houses are sold at 300k each, and 5 flats are sold at 100k each. Average price per Land Registry is 157k. In Q2, 2 identical houses are sold for 250k each, and 2 identical flats are sold for 75k each. Average price per land registry is 162k.

The conclusions from this example that many of you are drawing is that house prices have increased quarter on quarter. On the face of it, thats correct. But looking at the underlying detail, prices have fallen significantly.

So the fact that volumes are down is very significant, but you can't draw valid conclusions from the land registry data unless you know the mix of properties which have sold.

Exactly - thanks for this example! One of pet peeves is the mean being used as the "average" in economic data such as this - salaries is another one that really annoys me. e.g. the "average" graduate starting salary is £22k or something ridiculous (usually based on just the blue chip graduate jobs as well!) and then all these 1000s of graduates are surprised when they don't get a job earning that much! :rolleyes:

In a lot of data the mean[1], the mode[2] and the median[3] are around the same, but I think (and I've never learned economics so I'm by no means an expert) that this can't be the case with house prices/salaries because you get people who spend £5m on a house or earn £1m a year but you never get people who earn -£1m to compensate. The mean is especially sensitive to this kind of situation.

[1] all the house prices added up and divided by the number of houses
[2] the price which the most houses sold for - you would need to group house prices into bands first e.g. £62001-£67000, £67001-£72000,...
[3] the price where half the houses cost less and half the houses cost more

I've never read any research into this but personally I reckon that if someone made an affordabilty index for houses that compared the mode house price with the mode salary rather than the means (or perhaps as well as) then that would give a better indication of whether the general market would rise stagnate crash or whatever! It would certainly give a better indication of whether FTBs are being priced out of the market and replaced with investors. Unfortunately for me in DA's example the second quarter is bi-modal (there are 2 houses at each price) but you can imagine that this would be unlikely to be the case in the real world and the mode on a similar dataset would fall from 100k to 75k, showing the real fall in prices.


Example taken from http://www.shodor.org/interactivate/discussions/sd1.html

Student: Which one is better: mean, median or mode?

Mentor: It depends on your goals. I can give you some examples to show you why. Consider a company that has nine employees with salaries of 35,000 a year, and their supervisor makes 150,000 a year. If you want to describe the typical salary in the company, which statistics will you use?

Student: I will use mode (35,000), because it tells what salary most people get.

Mentor: What if you are a recruiting officer for the company that wants to make a good impression on a prospective employee?

Student: The mean is (35,000*9 + 150,000)/10 = 46,500 I would probably say: "The average salary in our company is 46,500" using mean.

meanmachine_2

It is astonishing how statistics are fiddled.

Everything we read and hear and see has a motive behind its creation.

A well known bank recently worked out that FTBers are "not priced out of the market" - firstly by using incorrect "average price" figures, using an interest only mortgage and quoting incorrect average salary data.

So they underestimated houseprices, underestimated the actual cost of a mortgage and OVERestimated the average salary.

Put them altogether and you have a stat that is WAY off kilter.

Doozergirl

I don't want to pick apart the minutiae of your arguments but you both use very small data sets. Of course data will skew badly (sorry, not you meanmachine, it took me a little while to compose this!) too much white wine. I live in a relatively small village and know that the average house price data for a semi is not something I should base my marketing price on because it's useless. BUT, the larger the data set, the greater the accuracy.

I'm sure if you study the Land Registry's historical data, it is proven to be reliable, unless you're going to argue that the last crash was in fact a boom?

What's the alternative? Do we wave a finger in the air? If I wave my finger in the air here, there's a lot more Sold signs up than there were at the beginning of the year but the asking prices are going nowhere. What does that tell me? Not a lot, but the Land Registry figures back up my feeling. There ain't a great deal happening.

boysmum3

I'm New To All This But Can't Resist Chance To Put My Opinion In
From Personal Experience House Prices Are Stagnant - The Figure In The Estate Agents Window Is A Guide For Everyone - A House Is Only Worth What Someone Will Pay - Or What The Surveyor Values It At.
My House Was Undervalued By 4.5k By Surveyor - Or Was It??? When Ea Gave Us Valuation We Fell Over And Always Suspected It To Be To High!
But We Sold In 14weeks (longest Time I've Ever Waited To Sell A House) And We're Still Seeing Same Houses Week After Week In Local Paper

zar

Doozergirl wrote:

I don't want to pick apart the minutiae of your arguments but you both use very small data sets. Of course data will skew badly (sorry, not you meanmachine, it took me a little while to compose this!) too much white wine. I live in a relatively small village and know that the average house price data for a semi is not something I should base my marketing price on because it's useless. BUT, the larger the data set, the greater the accuracy.

Hi,
I am probably misunderstanding either you or the land register data here, in which case apologies, but "the larger the data set, the greater the accuracy" is not relevant in this case. Assuming that all house prices recorded by the land registry are the exact complete dataset with no imput errors then you are not sampling and so there are no error bars around the statistics, they are what they are no matter how big or small the dataset is.

The question I was trying to address (I'd had some white wine too! :beer: ) was given that we have the exact data, which summary statistics should we use to describe it to answer the questions that most people are asking? I see it as a fascinating problem because the data is so heterogeneous (i.e. very variable) - if you are in one of two flats in your village do you use the price of the semi's in the village that sold a year ago or the price of flats in the county to determine the worth of your flat? - probably a combination of the two.

I'm doing a phd in statistics and although it has nothing to do with economics in a weird way its a similar problem. I'm working with counts of ducks in winter, and how do you construct an index that tells you whether the national population is going up or down - conservationists want to know if there's a crash in one species, and fisherman want to know if there's a boom in another! Something that makes it a difficult problem is that different things happen in different parts of the country - just like with house prices.

No hand waving is required as an alternative to an index based on the mean, just more information. I'm sure that there are people that are employed to analyse this data properly, and perhaps thats what different organisations base their own predictions on what the market's going to do next. Give me a nice long report with plenty of histograms of the data broken down by region and house type (I know you can access tables of numbers like this on the web but I hate numbers!), and I'll be happy - but that wouldn't be what the "average" person who reads the paper wants! :eek:

*Don't blame statisticans for misuse of statistics blame everyone else :A *