One for sasguru to cast his stastical eye over

[Foxy Moron]

No, it implies the rate of decline is slowing. It still predicts more decline toward the end of the year. Still seems a long way until HPI remains stable (0)might indicate the bottom.

[d000hg]

So if I'm reading right, it just happens to be claiming right now is the bottom of the drop?
Also, why does the 3rd graph continue into the future but the top one does not?

[Foxy Moron]

Better switch off your mum's computer now. She'll be expecting you to be doing your homework when she comes back from work.

how else will I get that top job in Actuary and get to work with Gurus!

[Doggy Styles]

This predictor, does it work like ice-dancing judges where you knock out the top and bottom scores?

[sasguru]

ewww

Better switch off your mum's computer now. She'll be expecting you to be doing your homework when she comes back from work.

[chris79]

Wish I was a cash buyer, would probably be more motivated to investigate this.

[Foxy Moron]

* licked my finger and stuck it up my arse

ewww

[sasguru]

I've just run a complicated analysis* and the results are just in. House prices will continue to plummet..















* licked my finger and stuck it in the air

[Foxy Moron]

I've done some time series modelling, often the best predictor is, what happened today is what will happen tomorrow or in other words the autoregressive moving average. Which (basically) appears to be what they are doing.

HTH

[DimPrawn]

http://www.houseprices.uk.net/articl...ice_predictor/

The House Price Predictor

At HousePrices.uk.net we have developed a new housing market predictor that provides a robust and timely view of the UK housing market. It determines the current level of house prices and the rate of house price inflation by combining measures of both house prices and estimates of house price inflation (HPI) derived from the BoE mortgage approvals for house purchase figures along with projections from historical data.

Kalman Estimation

The state estimation X = (price, hpi) is done using a Kalman filter, and it avoids the problems caused by trying to measure small changes in noisy price data over short times. We think that the Kalman filter approach combining both price and HPI/approvals data is the best way to get an accurate and timely handle on what's happening in the housing market.

A useful feature is its ability to make good a priori estimates of the Halifax prices. Good short term predictions follow naturally by projecting the much more stable trends in the mortgage approval figures. Also included is the house price trajectory implied by the currrent buy & sell prices from financial spreadbetting data (Cantor Spreadfair, house price spreadbetting) and residential property derivatives trading (Tradition, Future HPI).


What does the learned collective think?