Automated underwriting

Automated vs manual underwriting

Automated underwriting is much discussed in the specialist finance market with the industry sharing differing views on the topic.

The procedure is described as a computer-generated loan underwriting decision which uses completed loan application information; this allows the automated system to retrieve the relevant data before arriving at a logic-based loan decision.

Stuart Law, chief executive of Assetz Capital, said it was not in favour of automated underwriting.

“Assetz Capital's loans are typically in excess of £100,000, with the vast majority being between £200,000-3m,” said Stuart.

“This means that automated underwriting is highly risky, and this is one of the main reasons why we employ a nationwide team of regional relationship directors to directly liaise with active and potential borrowers.

“While we use a variety of technologies to speed up some parts of the loan process – of which an important part is to gather insight and data on a loan – the final underwriting always sits with an experienced professional who can make highly informed decisions.”

Nonetheless, Tim Mycock, development director at Reditum Capital, believed that automated underwriting would be great for efficiency.

“In 99% of cases, speed is of the essence and these engines would generate instant answers to save a great deal of time,” explained Tim.

“Of course, the success will depend on the quality of the data inputted to the system and what criteria are set, but it does have the potential to save significant time.”

As with most things in life, there are pros and cons regarding the system. 

But for something to be worth the trouble, the pros should certainly outweigh the cons, and this may be the case with automated underwriting.

“Well it will potentially save a lot of time, put everything in a standardised format and remove potential human error,” added Tim.

“The main positive is it will create a quicker, streamlined process for clients.

“But quite often even when all the criteria are met for a deal, an additional factor can be found which prevents the deal from proceeding. 

“It would in effect miss the human ‘sniff’ test. 

“A lot will depend on the type of information captured – but I can’t see it being suitable for the more complicated transactions – there will be a lot of room for error in that regard. 

“Often, we will assess a deal which might be declined on the facts presented, but when we apply our experience and expertise, we can see a way to structure the deal differently that will work.”

“Deals like this might be missed by a machine.” 

Michael Strange, managing director at Funding 365, said there was an argument that automated underwriting would be the most efficient approach to loan origination in the regulated bridging space for the most vanilla cases.

“In the regulated space, the loans typically are of relatively low LTVs and the ability to be defrauded is somewhat reduced by the fact that the owner is likely to live in one of the security properties.

“In the unregulated space, I cannot foresee any scenario where it is sensible to move to fully automated underwriting.

“Unregulated bridging typically involves a relatively complicated fact pattern – perhaps having to understand the extent of a property’s refurbishment requirements or to unravel the rationale behind a borrower’s poor credit record.

“An underwriter needs to understand the full complexities of a particular scenario to appreciate the risk profile of a loan.

“I believe that any lenders that attempt to walk along the path of automated underwriting in the unregulated bridging space will find that they are actually walking through a minefield.”

James Bloom, managing director of development finance at Masthaven, added: “In the short-term lending world, we believe that individual manual underwriting is a key ingredient – both from a risk factor and from a customer experience point of view.

“At Masthaven, each case is assigned to an individual underwriter. 

“We believe this helps us deliver successful outcomes and allows us to help people in a timely fashion.

“There are certain parts of the process where automation can help, but the fundamental parts of the process need manual and human intervention, and we believe there is no substitute for that.”

Leave a comment