Underwriting is the process by which lenders decide if a person is eligible for a guarantor loan, credit card or mortgage. Eligibility is measured in a number of ways, essentially in order to decipher if a person is a low-risk customer and are trustworthy enough to make repayments each month.
A loan provider will make their approval or decline decision based on the characteristic profile of that applicant, and will do this by investigating whether customers with similar profiles have repaid or defaulted. Behavioural underwriting and using past data effectively can become a very good competitive advantage for one loan company over another, and we review some of the methods below.
Age: Do particularly age groups have higher or lower default rates than others? Is there a correlation between age and repayment likelihood?
Gender: Do some genders repay better than others? Is there a pattern which suggests one gender pays in a more timely fashion than another? There are some brands that have tried to target females more, believing that they are better customers such as Cash Lady.
Location: Are some areas of the UK better at paying on time than others? A recent study by This is Money showed that areas of Northern Ireland, Wales and West Midlands have higher level of debts than other parts of the UK. This may cause lenders to be more restrictive when lending to applicants from these regions.
Residential status: Is lending to those that live at home or with friends riskier than granting a loan to a tenant or homeowner? Several underwriters believe that homeowners are better candidates to lend to since they would have already gone through the rigorous credit and affordability checking to get a mortgage. Plus, they have security in their home that they could also use to raise finance for repayments such as second mortgages, equity release and remortgaging – but this doesn’t mean that a homeowner could still have debt and be behind on repayment.
Employment: Does working part-time rather than full-time mean one will less able to repay a loan? What about if someone is on benefits, or works a significantly low-income job?
Credit History: Most lenders will have a minimum credit score required to be eligible for a loan. This data is accessed by their partnership with a credit reference agency like Equifax, CallCredit or Experian and the lender will pay something like £1 or £3 each time to review their records. Some lenders have a minimum score to be eligible for the next stage of underwriting. For guarantor loans, it is actually not the main borrower but more their guarantor who needs a good credit history and if they have one, it will maximise their chances of approval.
Most lenders will usually work these aspects of underwriting into the initial criteria that you see on a website or their basic application. For instance, companies will often say from the outset that applicants need to be over 18 or in employment as a way to quickly narrow down their ideal candidate.
Some sophisticated lenders use more specific metrics as a way of understanding their customer’s behaviour and thus their eligibility for a loan. They may look at:
Time spent on the site: Is it possible that users that spend more time on the site researching and understanding the product will be more reliable than someone who clicks and applies instantly?
How much they ask to borrow: Some lenders track the loan calculators on their websites and then deem customers that try to borrow the maximum amount as their first port of call ‘riskier’ than those who pick a specific amount.
Whether they read the terms and conditions: Wonga.com used to say that they would calculate the time a customer spent on certain pages of the site. This would involve eye-tracking technology to measure the length of time spent on reading important information such as the terms and conditions and the loan agreement. Logically, a customer that understands all the terms of their loan will be more responsible than one who does not. Obviously, time-spent reading and ‘understanding’ may not be correlative values.
Repeat visitors: Is someone who visits the website for the first time going to be more reliable than someone who visits it for the 5th time? Is it better that they have reviewed the website and come back or does coming back several times suggest that they have disorganized behavioural patterns?
Can you Underwrite Purely on Behaviour?
It is possible to underwrite purely on behaviour but it could be breaching regulation which requires UK lenders to carry out credit and affordability checks.
Historically, there have been lenders that have funded payday loans purely based on behavioural profiles. For instance, they might receive an application where the borrower is a 35-year-old female who owns a home and is full-time employed and so the lender decides that there is a 95% chance she will repay her loan on time (according to their statistics) and thus fund her loan in 15 minutes – but what about the 5% that default without any real affordability checks carried out? This is something that was used by some of the most well-known payday loan companies in the UK and lead to fines running in the millions following the introduction of regulation in January 2015.
A Combination Approach
In practice, the best underwriting approach is a combination of behavioural analysis and the use of credit checking to measure proper affordability. It only takes a small fee for a lender to go through a credit reference agency like Equifax or Experian to ensure that their customers are properly risk-assessed.
As per FCA regulation, all loan companies in the UK are required to carry out credit and affordability checks. The use of behavioural metrics can be extremely useful in creating a more complete picture of prospective customers, and with the right technology and analysis, it can lead to very successful underwriting by securing low default rates for companies.