Can I ask a stupid question?

Generali

Ok so here goes.

The biggest difference is between the types of averaging used.

A geometric mean is a bit complicated. Basically we should all know what an arithmetic mean is: it's what most people mean when they say 'average', it's a load of numbers added together divided by the number of numbers.

A geometric mean is of 2 numbers is the square root of the two numbers multiplied together (square root of 4 is 2 as 2x2=4). Of 3 numbers you would multiply the 3 numbers together and take the cubed root (cube root of 8 is 2 as 2x2x2 = 8, cube root of 27 is 3 as 3x3x3 = 27). Of n numbers you multiply them all together and take the nth root.

Each month over time, small differences in the type of averaging will add up to a big difference. The point is though that the indices aren't really meant to be measuring the average price of a house, they're meant to be measuring the rate of changing of the price of the average house sale so that the average is so different doesn't really matter as long as the rates of change are similar. If the rates of change are similar then it shows the indices are all equally right or equally wrong. If, as now, some are up and some are down you have to look at what is being measured and the different things being measured will be what is causing the difference.

AIUI, a geometric mean is generally seen as being better for measuring things that will tend to change exponentially like house prices. have no idea why that should be. Anyone interested should apply to do a stats degree.

The difference is not caused by Haliwide measuring realised prices while Rightmove measure asking prices. Not many get 40% off the asking price!!!

Graham_Devon

Really2 wrote: »

Table at the bottom has all them on, just it has the LR as quarterly.

No, I mean it doesn't explain the disparity between LR and rightmove prices.

LR is non mix adjusted, wheres Haliwide (thanks Joe, I'm using it all the time now) are.

However, LR is still 40ish% out, which excludes the cash sale stuff too.

Although cleavers unconvinced by my argument (and I am too, to an extent), so far, no ones been able to dispute it....quite proud of that Though I'm leaning far more towards properties being listed, but never selling.

Generali

Graham_Devon wrote: »

No, I mean it doesn't explain the disparity between LR and rightmove prices.

LR is non mix adjusted, wheres Haliwide (thanks Joe, I'm using it all the time now) are.

However, LR is still 40ish% out, which excludes the cash sale stuff too.

Although cleavers unconvinced by my argument (and I am too, to an extent), so far, no ones been able to dispute it....quite proud of that Though I'm leaning far more towards properties being listed, but never selling.

LR doesn't need to be mix adjusted as it measures like for like sales.

I suspect LR use the same averaging as Acadametrics.

Graham_Devon

Generali wrote: »

The difference is not caused by Haliwide measuring realised prices while Rightmove measure asking prices. Not many get 40% off the asking price!!!

I'm not sure anyone actually get;s what I'm trying to say here.

So I shall put it into pictures. Need propertybee to see the evidence, but....

House picked at random in Ashford, Kent.

Initial asking price £379,000.

Final asking price before going STC... £310,000

ACTUAL sold price? Don't know, but say 8% off that = Final selling price of £285,000.

Difference between selling price (recorded in index's) and INITIAL asking price, (recorded on RM index) = 25%.

Property listing number 2.
http://www.rightmove.co.uk/property-for-sale/Ashford.html?minBedrooms=4&includeSSTC=true&_includeSSTC=on&index=70

---

Initial asking price: £350,000

Final asking price "Offers above £220,000".

So say final selling price = £225,000

Difference between selling price (recorded in index's) and INITIAL asking price, (recorded on RM index) = 36%.

House no.3 on listings
Clicky

Can't really be bothered listing any more. Realise not all houses are like this, and the weight of the other houses brings the percentages down, however, having more of an affect, is those listed, which never sell, due to the price being high and the seller not having to sell, so pulls it.

If the above, really have no affect on the difference between RM andactual sold price index's, then I dunno what to say. Evidence is above though. No ones got 25 or 36% off the asking price. The asking price at the time was lower than the INITIAL asking price. But it's the initial asking price which RM looks at, not the asking price before a sale.

HAMISH_MCTAVISH

Cleaver wrote: »

HAMISHHHH!!! ARE YOU THERE??!?!? ARE YOU SNOWED IN?!!??! PLLEASSEE!!! HELP ME!!!!

The difference between average sale prices and rightmove average asking prices is purely down to the different way the indices measure average price.

Indeed, the difference between the various sold price indices can be as much as 50K, just because of methodology.

A comparitive look at index methodology can be found here, along with explanations for why the different indices have different average prices.

http://www.acadametrics.co.uk/Index%20Monitor.pdf

Generali

A difference of very little, something like 0.2% per month would make a difference of 40% over a decade or so.

JonnyBravo

sjaypink wrote: »

You mean Haliwide take out all the country mansions and plush city penthouses? Or skew the figures towards 3 bed semis etc?

ETA: But hang on, LR is way off RM figures but would still include these wouldn't it?

Yes they skew the figures towards the "typical 3 bed semi".


LR is yet another method. It excludes many types of sale and also only allows repeat sales into its data.

Ultimately they are all measuring different things but everyone bundles them together and assumes they must be the same.
They are similar but not the same.
As they are similar they move in the same way but need not equal each other.

ash28

Haliwide exclude properties of a £1m or more Rightmove and Acadametrics don't - in our county alone (Berkshire) of 100 pages of houses for sale 19 pages are upwards of £1m and for Surrey there are 100 pages of property and you get to £2m - those properties have to make a difference and as Generali says Haliwide both have a "notional" or "standard" house -

I look at them all - always interesting - but I wouldn't buy a house based on any one of them.

HAMISH_MCTAVISH

JonnyBravo wrote: »

As they are similar they move in the same way but need not equal each other.

This is the key point.

Over the course of a house price cycle all the indices will show very similar movement, or at least they have until now as in the past the whole market tended to move more or less in a similar way.

But the starting price and end price are often different as they are measuring different subgroups of the data, or the same data in a different way.

As was said earlier, the "average house" price, is a different thing to the "average house price".

In practical terms the differences haven't usually added up to much over the short term, and so it hasn't mattered a lot until now.

But with continuing signs of a split market, I suspect the differences in the indices will be getting more attention in the future.

Generali

JonnyBravo wrote: »

Ultimately they are all measuring different things but everyone bundles them together and assumes they must be the same.
They are similar but not the same.
As they are similar they move in the same way but need not equal each other.

I think that's penultimately.

What they are actually trying to measure is the rate of change of house prices on average, not the average house price I believe.

Similarly, but more clearly, the FTSE100 Index looks to measure the rate of change of the average share price contained in the index not the average share price of the shares in the index.