Like the winter weather, referenda results and Arsene Wenger’s future, forecasting property values is a risky business.
We can all be experts on past transactions but predicting the fair price for a property today, yet alone tomorrow, needs more than just the ‘I know my patch' gut feel that influences the majority of property transactions.
In the past rising markets have covered over a trail of over-optimistic estimations, minimising Professional Indemnity Insurance claims. But with uncertainties likely to continue to affect the market for several years, and with billions of pounds put at risk, much smarter and objective advice is demanded by a customer base that has increasing thirst for more information, thanks to mobile Apps and widgets on their smartphones.
Artificial Intelligence (AI) and Machine Learning (ML) have been around for over a decade in the banking sector and yet have made very little impact on the world of property. That is until now.
When these are paired with Big Data and powerful Cloud based processing capacity, the next generation of valuation toolkits can be delivered to users, which are far smarter and responsive than those built with MRA and standard matrix analysis, methods which currently form the basis of most AVMs (Automatic Valuation Model).
The ML is constantly reacting and learning from changes in and around the marketplace and can be trained to look at other marketplaces to broaden the range of issues to be considered within a single valuation. Thus the components affecting the valuation criteria are being constantly adjusted as millions of bits of new data are assimilated.
As soon as a property sale is recorded on the Land Registry, a new public transport route announced, a new school opened, or an increase in local air pollution levels registered, the AI system will model the impacts, produce a modified valuation figure and then model the outcomes in a number of different ways, advising on the most likely scenarios.
These scenarios can also be further processed within industry specific hybrid models that combine specific levels of sophistication, maximising the ability to use the information creatively, yet ensuring accuracy and reliability. This allows, for instance real estate developers to work within much clearer risk parameters.
Does a developer continue to build residential blocks to sell, revert to a wholly or partial rental scenario, or sell on and move onto the next project? All scenarios can be modelled, stress tested and risk assessed.
There is now no need to wait until the market conditions ‘have picked up post Brexit’ or the ‘overseas investor becomes more comfortable’ or even that ‘new government initiatives will re-ignited a stagnant marketplace’.
Decisions can be made now on a vastly more intelligent prognosis than ‘gut feel’ and a list of apocryphal or redacted sales comparison list.
Who knows, with the ability to assimilate vast amounts of hitherto undreamt of information, AI and ML make real estate valuation a science and not just an art. A science that does not depend on the hunger of the buyer, the desperation of the seller. or the charm and salesmanship of the broker.
Arsene’s next career?
|Philip Challinor is the chairman of Houseprice.AI and was part of the architects team at Denning Male Polisano who helped convert Highbury, the former home of Arsenal FC into 700 homes for local people.|
If you would like to know more information about Houseprice.AI , Horizon, or access to our API please feel free to contact us at firstname.lastname@example.org.