With the property market seemingly spiralling out of control, the Mortgage Market Review (MMR) has had a huge amount of publicity in recent months.
The new regulations have caused lenders to become a lot more forensic in their underwriting and they’re being a lot more detailed in the questions they ask and the checks they make in order to properly ascertain affordability and determine the applicant’s financial profile.
But there is an argument that the lenders are barking up the wrong tree.
OK, it’s fair to say that the more detailed the data lenders can get their hands on, whether internally or via credit reference agencies like Equifax or Experian, has a value – ‘structured’ data like this can help them make more prudent lending decisions.
However, lenders are kidding themselves if they believe that the data supplied by an applicant, or via a credit referencing agency, is an adequate form of underwriting in the digital age. The reality is that it tells only part of the story.
Today, the most revealing insights into whether applicants are really able to afford a mortgage, or have risks attached to them that are not obvious through standard underwriting measures, are found not in structured data such as credit history, proof of income or a set of accounts. Instead, they are found in ‘unstructured’ data.
So what exactly is ‘unstructured’ data? Unstructured data is data that comes in all kinds of irregular and hard-to-read formats. It’s the toughest to analyse.
For the purposes of this article – lender underwriting and the MMR – unstructured data would be, for example, an applicant’s online footprint, specifically that generated through Facebook updates, Tweets, blog posts and other digital activity.
It’s this kind of data that can offer key insights into an individual’s real, as opposed to paper-based risk profile.
For example, people who are perceived to be low risk via structured underwriting (salary, outgoings, credit history, say) may appear a far greater risk once they are looked at through the lens of unstructured data. Unstructured data, if you like, offers a direct line straight to the real person rather than the person that officially exists on paper.
The problem, to date, has been that unstructured data has been difficult to analyse manually due to its sheer scope. This is a cause for concern as it’s been estimated that around 80% of business-relevant data is unstructured.
The long and the short of it is that lenders need to start analyzing unstructured data and using it as a key part of their underwriting processes. Thankfully, this is where Big Data can help.
Big Data platforms are able to scan vast unstructured data ecosystems in rapid timeframes and can effortlessly uncover important insights, connections and activities relating to a specific individual (the applicant) – including fraud or other forms of financial crime.
They can also continue to undertake this analysis in real-time, which represents a ‘living Know Your Customer’, and be embedded alongside more traditional underwriting techniques that focus on structured datasets.
In short, the secrets hidden within unstructured data are now available to mortgage lenders, and can offer a depth of underwriting previously unseen.
With the arrival of cost-effective and cloud-based unstructured data analysis, the future of lender underwriting in the years ahead is set for irreversible change.
