Lenders need to realise that they may currently be at the relatively thin end of what will become a very wide wedge.
Once you accept that this is where we are headed, you begin to appreciate that access to evermore data streams and ability to analyse them rapidly is going to be key.
Data systems boost needed
Banks are going to have to quickly access much wider data streams and have the capability to extract, digest and analyse at pace.
It is pretty obvious that most lenders’ data processing capabilities are not currently up to this standard. Many have information effectively spoon fed to them from data sources, including credit rating agencies.
They have become used to data bureaus responding to their requests in the form of ‘compiled characteristics’ that only summarise raw client data. This could simply comprise the number of CCJs a prospect has, number of defaults, or extent of arrears over a 24-month period.
This and other forms of rather standardised data provision needs to be replaced by sophisticated company-specific analysis.
It will no longer be a case of making do with the credit scores that the bureau provides. They may be predictive, but as a minimum you are going to have to understand why they are predictive for your specific customer base. To be able to say ‘the data is predictive when these questions are asked and these elements are there’.
From such a position a lender can finesse a scorecard over time so that it becomes fine-tuned to required customer demographics.
In this way, they can extract the intelligence that they need for their business, rather than take the pieces of data plus the predetermined questions that a bureau has arbitrarily decided they need.
So, in our example above, rather than a bureau merely telling the lender that an applicant has missed X payments over the past two years, the lender should be able to take 24-months of raw data from the bureau and create their own characteristics to decision against.
We have to move to a situation where all lenders have the ability to interrogate across all their existing and newly-developing data streams.
Traditionally asking CRAs to undertake such extra analysis has come at a significant cost, and, if you wanted to undertake higher level analysis in-house, you would need the sort of dedicated IT resource that only big banks have.
Thankfully, advances in data technology are making sophisticated in-house analytics achievable for every lender, even those that lack IT and data science knowhow.
Lenders need to realise that their ability to thrive in the future is going to increasingly depend on them becoming experts in data handling and analysis.
