Land Registery Average House Prices Calculation?

westv

Would someone be able to explain, in simple terms, how the Land Registery calculates the figures it shows for the average house prices it lists for areas shown in its monthly report? Is the calculation any different to other average house prices shown for the same locations on other property sites (which also state they use Land Reg data).
Many thanks.

Generali

http://pro.landmarkanalytics.co.uk/Land-Registry-House-Price-Index-Methodology-1995.pdf

Basically the Land Registry tracks the resale of existing houses and compares the price to the previous time it was sold. That's why it doesn't include new houses in the index.

The Nationwide & HBOS HPIs both track sold prices where they are the lender. They also apply something called 'hedonic regression' which tries to mitigate for the fact that houses get better over time. They try to remove the increase in value due to houses getting nicer (indoor bathrooms, central heating, double glazing, insulation etc) and just measure the change in actual underlying values.

westv

Generali wrote: »

http://pro.landmarkanalytics.co.uk/Land-Registry-House-Price-Index-Methodology-1995.pdf

Basically the Land Registry tracks the resale of existing houses and compares the price to the previous time it was sold. That's why it doesn't include new houses in the index.

The Nationwide & HBOS HPIs both track sold prices where they are the lender. They also apply something called 'hedonic regression' which tries to mitigate for the fact that houses get better over time. They try to remove the increase in value due to houses getting nicer (indoor bathrooms, central heating, double glazing, insulation etc) and just measure the change in actual underlying values.

Ok thanks.
Would I be correct in saying that the monthly HPI index figure for any particular area, say, Cornwall is then applied to the "average price" figure (say, £100,000) they already have for that area and the figure is then revised up or down? The average isn't calculated directly from house sales that month?

HAMISH_MCTAVISH

Which land registry index? (There are two, and they use different methodology)

What exactly is it you're trying to figure out?

westv

HAMISH_MCTAVISH wrote: »

Which land registry index? (There are two, and they use different methodology)

What exactly is it you're trying to figure out?

I thought the question was clear enough.

Anyway, just to clarify. The LR issues a free report each month which lists a figure for the "average price" for all English counties and some other areas. I was interested in knowing (in layman's terms) how these average figures are calculated and how (if at all) that calculation method differed from other sources that give "average price" reports for the same areas using LR information.

HAMISH_MCTAVISH

westv wrote: »

I thought the question was clear enough.

Anyway, just to clarify. The LR issues a free report each month which lists a figure for the "average price" for all English counties and some other areas.

OK.

In layman's terms, the LR monthly series does not report either the median or mean average of sales prices for that month.

What is does is calculate the monthly change of a notional "average house", via repeat sales regression, of the properties that sell within it's dataset and that have a previous sold price within the data range.

Essentially it's a geometric average based back to April 2000.

I was interested in knowing (in layman's terms) how these average figures are calculated and how (if at all) that calculation method differed from other sources that give "average price" reports for the same areas using LR information.

This has a pretty good explanation for them all.

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

For Land Registry read Section 1.2.2.2 - Standardised average prices

westv

Excellent info all round.
Helps to explain why LR has the average price for Kingston upon Hull as £69,592 - 23% below peak - whereas LSL/Acadata show the average as £105,862 or 1.90% below peak. A huge difference.