Man v Machine

Automation, Big Data and the threat to the Estate Agency profession.

 

The creep of technology is picking up at an almost exponential rate, posing a threat to those jobs that can performed based on a machine algorithm.

The dynamic of the role of ‘Appraiser’ or ‘Valuations Consultant’ is a fairly straight forward one, to quantify the value of a property to stack the purchase for a lender to approve a loan, otherwise known as a mortgage. So, could this role conceivably be taken over be a machine? That is the question that we will attempt to answer. 

These days everyone is aware of the value and importance of ‘Big Data’, the aggregating and assimilating of multi-set streams of data to allow you to model, predict and even automate future events down to an extremely high point of accuracy.  The fact is we have lived with automation for most of our lives, it has always been there silently and obediently set to task pleasing corporate shareholders by facilitating the reduction of costs, reducing pay rolls, eliminating down time attributed to the human frailties of sickness, vacations, and going off-line, otherwise known as ‘returning home’. 

But the creep-effect of technology has now moved beyond perfunctory tasks and has done so for more than 3 decades, with the automation of car assembly lines seeing the first visible displacement of the human workforce in favour of the efficiency, low cost and 24/7 ability of machines to be ‘online’, performing repetitive tasks and keeping production flowing. Now we can see evidence of this human takeover displacing workers in professional consultancy roles, where high levels of human calculation, intuition and even sentiment can be measured, coded and replicated by machine. 

These days third party property aggregators and valuation sites such as Zoopla, Prime Location and even spareroom.com utilise algorithms and machine learning to capture data from both sellers and buyers, along with open market property data that is compared and contrasted, listings are analysed, price indicators such as interior fabric, fixtures and fittings are quantified and a value is calculated and levied on the appraisal to enable price discovery. 

Now even banks and other mortgage originators have confidence enough in transactions that remove the need for a human providing a property valuation. To understand this we will need to defer back to the 2008 global financial implosion and the blame game that followed. It needs little coverage the fact that the housing bubble stemmed from the volume of bad loans approved for lending by valuations consultants that were pressurised to hit targets and to manipulate values to achieve parity with loans. So, when the market crashed and loan books were effectively near worthless the finger pointer was firmly directed at the Appraisers and Valuations experts. Little wonder that banks are keen to automate and control as much of the pricing element now as possible. Which could be claimed is a fair commercial motive. 

Banks are now more confident than ever in trusting an estimate from an algorithm, especially given that the valuations are instantaneous and cost much less then employing an industry professional.  

But what about considerations of inventory, checking the fabric of the the property and the many other on-site visual cues that valuations experts look for? Rising to this challenge Data Engineers are busy addressing this, writing code that can process real estate photos to identify, value and discriminate between properties based on the interiors and writing or discounting a value in.

That said it is necessary to consider that automation tools employed for appraising properties and producing valuations have their limitations, such as properties situated in remote locations, or new build multi unit properties located in areas where such a structure sets a precedent, which means that an algorithm is unable to compare and connect with other properties and associated metrics to arrive at a valuation. 

In conclusion it seems apparent that machines employed by lenders are unlike to replace valuations experts any time soon, they clearly have their limitations, with exceptions perhaps made for properties that have a high process rate within a defined locality.  It seems that Valuations and Appraisals experts can rest easy, for the moment.

 

 
Mathilde Bock