In today’s marketplace the growing prominence of the automated valuation model (AVM) cannot be ignored. In its latest survey, the Council of Mortgage Lenders (CML) reported that in 2007 it expected AVMs to be used in 3 per cent of house purchases and 28 per cent of remortgages, with an anticipation that this will grow to 25 per cent of all house purchases and 55 per cent of remortgages by 2011.
Get the daily news delivered to your inbox
As we are well aware, used correctly, AVMs can prove extremely helpful in significantly reducing costs and drastically speeding up the lending process. However, while their potential is ever increasing, in their current form, the blanket use of AVMs across all residential properties is not realistic.
Current limitations
Despite their growing popularity and widespread use, there are a number of issues that are still hampering the growth and use of AVMs – namely the reliability of existing data. AVMs work on the basis of current datasets for property transactions over a period of years, typically only as far back as the year 2000, when online Land Registry data was first made available.
find the latest industry jobs
While the overall value and price supplied should reflect the condition of the individual property, there is little consistency in how some of the data is collected and introduced into the models. Inaccuracies can often occur, with properties classed in the wrong category or other information not input correctly and as a result, the reliability of the AVM is impaired.
Another concern among repeat sales regression models is that for residential property, the Land Registry only collects a limited dataset – the address and postcode of the property, the type of property, the tenure, whether the property is new or old and the recorded transaction price. With such a restricted data set available, the calculations made using this model are also limited – an AVM is only as good as the data that it is supplied with.
In the US and Canada, a much wider range of more detailed statistical information is gathered, including the property size, location and characteristics, aiding a much more accurate valuation. In the UK, a similar approach has been taken by the hedonic model. However, even these models fail to take a conclusive account of all the elements of a property which can impact on price and still suffer from limitations.
register for the next forum
Instrumental
Despite their limitations, given the models are used correctly, AVMs can prove instrumental, particularly for brokers in the initial stages of determining whether a client has a valid proposition or not.
- First steps
- Reducing the risk of error
register for 'adviser finder' here
An additional step to achieve further clarity is to link the property’s postcode with a mapping tool or aerial photograph, which allows the easy identification of similar property types nearby – the more similar the property, the more reliable the information. However, it is worth bearing in mind that if the property is, for example, fronting a major road, this should be reflected in the comparison with those in less busy locations.
- Broadening the search
- The role of the surveyor
Despite any future advancements, I am confident that surveyors will continue to play an important role in the lending and valuation process. From the valuer’s perspective, the AVM is not so much a lethal weapon, but more a tool to improve on quality and reduce risk to both parties.
