Neil Williams (pictured), managing director of LendingMetrics, argues that mortgage lenders and networks now have no excuse for not switching to automated underwriting

The reality of underwriting is that at some mortgage lenders it is stuck in the Stone Age.

You might think that this is overstating it a bit but, really, think about it. How many lenders still rely on paper bank statements and manual processes when making lending decisions?

In the age of Big Data, when terabytes of data are being accumulated on a daily basis, there remain lenders employing, collectively, thousands of underwriters who have to manually pour over paper copies of utility bills and bank statements.

Why is this still happening? The short answer to this is that expert IT is not in plentiful supply at most lenders. To move over from a legacy system of manual proofing to software utilising Open Banking is not something that you can do in-house overnight.

Your IT resource will have to spend months planning and implementing the change.

Even then – as last year’s TSB debacle goes to prove – things can still go horribly wrong.

Who can blame some CEOs for hesitating when the potential downside of reshaping their back office can be so huge.

Well, such delay cannot continue for much longer – given the seismic improvement in underwriting quality and massive cost savings that new technology is capable of delivering.

Until recently, lenders have only had ‘credit history’ and ‘historic’ proofs of status – a print out of a recent bank statement, a pay slip, and proofs of address on other documents. All these ‘indications’ are time-consumingly produced by the applicant and worryingly vulnerable to fraud.

For a price, lenders can access information periodically provided by the major banks about individuals’ income, but even this is not definitive.

Thankfully, Open Banking means an end to all this.

No more over-reliance on credit history and paper proofs. And, significantly, no more need to estimate a person’s ability to afford the mortgage that they are applying for.

As we all know by now, Open Banking requires banks to share their line-by-line detailed customer statement data with other parties. This treasure trove – which has always been jealously guarded by the major banks – is now available to all lenders (with the agreement of the finance applicant) via Account Information Service Providers (AISP).

So, with an applicant’s permission, months of transactions can be captured. Read-only data comes directly from the bank that holds the current account to the AISP.

And it is supplied in real time. Someone applies for a loan, agrees to limited timeframe and read-only access to their accounts, and thousands of lines of transactions can be analysed. In milliseconds, this can pull out salary details and all financial commitments, to be then tested by an algorithm to determine whether the loan should be granted or not.

With Open Banking, lenders – for the first time – can make an affordability assessment that is extremely accurate.

In the past, the problem has often been that the borrower has had a rose-tinted or incomplete view of their finances. Time-stretched individuals often spend far too little time completing the ‘outgoings’ section of their application form.

With real-time data to examine, this is not an issue. The data and subsequent analysis provide insights that often the borrower will be ignorant of.

The technology can pick up on things that would otherwise be overlooked: patterns that if continued would lead to budgeting problems, such as spending that indicates a gambling habit.

And all this is done digitally and, as mentioned, in milliseconds. No more paper applications, no more laborious supplying of paper ‘proofs’, no manual underwriting of loans, no more two-week wait for a decision.

The technology can place the lender’s own rules around the raw intelligence. There is either a ‘decline’ or ‘referral’ outcome. And with referrals, instead of the underwriter having to go through a long list of standard ‘tasks’, only those tasks relevant to the applicant are flagged.

Mistakes are eliminated.

Most lenders will admit that, while they might have a team of say 10 underwriters, six may be following the lead of three high experienced long-standing employees who know the scorecard inside out. This creates a big potential for oversight or mistakes. Auto-decisioning means an end to this. Indications of fraud are better detected and quickly picked up by the technology.

Loan management is made much easier. Open Banking means a borrower’s bank transactions are usually obtainable within a 90-day timeframe, so, if a first payment is missed, the account can be ongoingly interrogated so a request payment call can be timed when there is sufficient funds in an account. Collection rates can be improved by only phoning late payers with sufficient money in their account.

Signs of stress can be straighforwardly picked up. A monthly check can trigger conversations with those borrowers struggling with managing their finances. Payments can then be reorganised rather than missed.

As we can see, the quality of every lending decision is light years away from what it has been. The decisioning is embedded into the online application process.

The automation also brings big cost savings. Underwriting that used to require a team of people, no longer does. And the loan book can increase year-on-year without the need for any more underwriting heads.

Alongside the potential salary cost savings, the automated decisioning removes the need to pay for credit searches, AML, KYC or income checks. Given that each credit search alone can cost one pound fifty, the total savings made are considerable – even for a small lender.

Ah, I hear you say, the downsides (of potential TSB-type miss-steps) remain. And then there is the question of implementation timescale and IT expertise availability.

Well, neither of these need apply any longer. There are platforms – such as LendingMetrics’ OpenBankVision (OBV) – that are available off-the-shelf. They can be ‘white labelled’ and made to work for you within hours. We can have OBV working on the day a contract is signed.

Such platforms have the benefit of being tried and tested. They are already in use by ‘early-adopter’ lenders. No extended implementation timeframe, no IT expertise vulnerability. They will work from day one.

The platform could even be free (OBV is). I am often asked for ‘proof of concept’ or a trial, because lenders are concerned about the costs of auto decisioning. I have to repeat to them – because they find it hard to believe – that our platform can be used free. (We derive our profit from selling parallel services.)

The lenders that are using this sort of knowhow have been able, in an instant, to rationalise their underwriting resource, improve decisioning and take the brakes off any growth that might have been choked-off by manual processes.

So, for those who have yet to make the leap, all I can say is, as time moves on, you will become increasingly disadvantaged. Can you really afford not to embrace change?

Originally published on:
Mortgage Introducer, February 12, 2019