Thursday, September 25, 2008

Tim Harford entertains and informs

Houses cost more in the summer. Here’s why.

UK.....house prices fell by 1.5 per cent in April, according to the Halifax index. Except: they didn’t. The Halifax’s own figures show that house prices rose in April, albeit by less than 0.2 per cent. The 1.5 per cent fall, widely reported, is the result of “seasonal adjustment”, an attempt to strip out predictable calendar patterns and report just the underlying trend. House prices usually surge in April, and this April the surge was disappointing enough to be reported as a fall.

Posted by jayk @ 12:15 AM (715 views)
Please complete the required fields.



9 thoughts on “Tim Harford entertains and informs

  • Trys to explain why house prices rise in summer quoting research.
    Might be a simpler answer. Maybe the type of house sold in the summer is more expensive.

    Reply
    Please complete the required fields.



  • Phew panic over… House prices DO only go up after all!!

    Reply
    Please complete the required fields.



  • Timmy T,

    Not what he says at all, but it’s not unusual for a HPC poster to misquote, misrepresent and misunderstand something which appears to be disagreeing with what we all want to be true.

    The simple fact is, the drops reported by HBOS and Nationwide are huge. Yet they are NOT ABSOLUTE. They apply their own seasonal adjustments which in real money terms – i.e. to you and me – mean absolutely nothing. OF COURSE house prices are dropping, nobody can deny that any longer, but not as fast as these two indices tell us. Nearly every other index is posted by a company or organisation whose current survival does not depend almost entirely on the mortgage market; nearly every other index is reporting a slower rate of decline. One regular accusation thrown around this place is “qui bono”, i.e. who benefits, who are the VIs, must ignore the VIs – etc etc ad infinitum ad nauseum. Well, is it not possible that HBOS and Nationwide are exaggerating the drops to try and influence a downward change in the Bank’s repo rate? Are they not VIs?!?!?! What is the model they use to “seasonally adjust”? How DO they turn a 0.2% increase in the value of an asset into a 1.5% decrease?

    Last month many HPC posters mocked HBOS when the latter complained that their figures were being misquoted: instead of their headline rate being quoted, most organisations chose to quote the [so-called] “real” rate (i.e. not spread across a three month sample as the bank chose to do). Yet I heard not a single HPC poster or media outlet say “Actually, let’s not quote the seasonally adjusted figures, but the ABSOLUTE figures in real money terms instead!”. Why? Because if they did that the the percentage drop would probably be halved, that’s why. And you and the Daily Mail don’t want to hear that. I for one want to see an absolute figure, in real money terms, not one massaged to show a drop because it’s freaking summertime!

    Perhaps you should sit down and think about that rather than immediately disregard Harford as another one of HPC.co.uk’s “house price crash deniers” (you know, the type of person hated as much as holocuast deniers around these parts).

    Reply
    Please complete the required fields.



  • Jayk,

    The only way to get ABSOLUTE numbers which aren’t distorted by any other factors would be to look at the change in price of the same house. These figures are published monthly so that would reduce the sample size to the number of houses that were sold twice in 2 consecutive months, by the same lender.

    I have sat down and had a think. You’re right I’m sorry. I’ll be on the naughty chair with my hands on my head.

    Reply
    Please complete the required fields.



  • You hit the nail on the head. You and many others it seems would rather see a figure which is magically turned from a small positive change into a big negative one using a totally ridiculous, arbitrary and secret formula because it allows them to say “look, prices ARE plummeting”.

    It does not need to be based on one house at all; already there is an adjustment based on property type and size, which is eminently sensible. What is not sensible is saying that, in addition to the adjustment for property type and size, and in a month in which prices actually rose according to their own data (albeit by only 0.2%), they’ll state that prices dropped significantly (at an annualised rate of about 18% I think) because in previous April’s they rose more than they did in this April! WTF?

    Also, your implied suggestion in your previous post that Harford was saying prices never go down was extremely silly. He was reporting a verifiable fact and revealing to his readers the very interesting method (sic) behind those statistics that so many people put such stock in. If you think this is a bad thing then, well, I don’t know what to say.

    I’ll shut up when HBOS and Nationwide publish both seasonally-adjusted and non-seasonally-adjusted figures.

    Reply
    Please complete the required fields.



  • Jayk loves the FT index because it delivers the correct figures, ie not falling as much as the others say.
    You can believe what you like with regard to the indices with all their fancy adjustments and manipulations, in the REAL world we all know house prices are plummeting and the effects on the housing market of last weeks shocks to the US and UK economies will be felt in the months and years to come.The time for studying figures by various VI,s is over.

    Reply
    Please complete the required fields.



  • Sorry Jayk but if you want absolute numbers like you said, then you do have to look at the change in price of the same house. All published data has obviously been adjusted – if you want to claim that some adjustments are legitimate but others aren’t then that’s your right. But I think that most people would agree that a 1.5% fall was far more realistic than a 0.2% rise in April, which points towards the fact that this adjustment stacks up. I think it far more likely that, on average, the houses bought in April were slightly larger (and therefore more expensive) than those bought in March. But that wouldn’t have made the headlines. And I never claimed that they did say that prices only go up – I was having a laugh. Get a grip.

    Reply
    Please complete the required fields.



  • Jayk

    HBOS and Nationwide do publish the underlying data in a SA and NSA format. It’s there on their websites, and you can use Excel to analyse the data if you so wish.

    For the record, the SA formula isn’t secret, it’s just too complicated for the layman to comprehend. And it’s boring. Pretty much like the rest of this message will be, but read on if you fancy………

    HBOS NSA for 1 Year to 31/08/08: -12.77%
    HBOS SA for 1 Year to 31/08/08: -12.74%

    No of months where NSA>SA = 6
    No of months where NSASA = 5 (average 0.43%)
    No of months where NSA

    Reply
    Please complete the required fields.



  • addendum

    seems that this site doesn’t like the left arrow (chevron) button and misses out text on the same line …….

    No of months where NSA is less than SA = 6 months (HBOS)

    No of months where NSA is less than SA = 7 months at an average of -0.33% (Nationwide)

    Reply
    Please complete the required fields.



Add a comment

  • Your email address is required so we can verify that the comment is genuine. It will not be posted anywhere on the site, will be stored confidentially by us and never given out to any third party.
  • Please note that any viewpoints published here as comments are user´s views and not the views of HousePriceCrash.co.uk.
  • Please adhere to the Guidelines

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes:

<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>