Although the mortgage market paused in the initial lockdowns, it remained incredibly strong throughout the remainder of the pandemic and adapted quickly to a temporary state of operating with staff working from home and assessing properties without visiting them.
But how should mortgage lenders balance the opportunities of digitisation we were prepared to accept during the pandemic against risk in the longer term?
The challenges of lockdown and social distancing triggered a shift away from physical valuation toward mainly desktop valuations – where a person assesses the property without going to see it. However, a number of lenders also increased their use of the automated valuation model (AVM).
To truly replace physical or desktop valuations involving a person, you need more than the current generation of AVMs which have been increasingly used for lending on particular types of mortgages and lower loan to values for 20 years.
The digitisation spur
Before the pandemic, data scientists had already been investigating how to advance automated property valuation and mortgage risk decisioning beyond the AVM. This work accelerated when it became harder to physically inspect a property due to the lockdown and restrictions around person-to-person contact.
Although not ready in time for the first lockdowns, new tools are now available that use machine learning and analytics to automate decisions in seconds for more mortgage applications than has ever been possible.
The most significant benefit of this is faster time to ‘yes’ for borrowers, literally in real time in an increasing number of cases.
And for lenders, there is a significant operation cost saving by using digital solutions. There is also the benefit of a reduced drop out in the mortgage application process. With larger high street lenders tending to spend around £20-40m a year on valuing properties, there is a massive opportunity to lessen expenditure.
The most updated solutions still rely on the AVM for valuations, but they also go beyond this by assessing the confidence in the customer’s estimated valuation as well as identifying and resolving property risks. This is enabled by using the enormous amount of property data that has been acquired by over the years and used for training in machine-learning based models.
The Covid effect
The unprecedented challenges of the Covid-19 pandemic have helped us all be more innovative and determined to help people move home regardless of any restrictions. This has only happened by helping lenders through the last 20 months with automated valuations.
The next generation of digital valuation and risk decisioning is now perfectly placed to support lenders, brokers and borrowers in a post-pandemic era.