Improving lender performance by 28% using Finexos and transactional data
Finexos was engaged by a consumer lender who was experiencing a high percentage of defaults in their UK-based consumer loan portfolio. We were tasked with conducting an examination of a book of more than 70,000 loans, totalling £23m, which was in run-off (i.e. the loan capital was not being re-lent, as the debt matured).
Analysis of the data quickly brought to light several pivotal insights:
neither the loan decision process nor the scoring mechanism was automated
there was a notable absence of open banking or transactional data integration for enhanced affordability evaluation
these factors culminated in the fact that most of the loans could be categorised as ‘bad loans’* and should have been rejected, at the point of origination
this capital could have instead been been lent to more capable borrowers
Inputting the available data (including three months of transactional data) into the Finexos risk engine revealed that, had the UK lender utilised Finexos technology and analysis in their decision-making, it would have achieved the following outcomes:
a net improvement in lender performance exceeding £6.5m or 28% of the total value of the loan book
a swing of the book from a £4.5m loss to over £2m in profit
84% of good loan were detected
39% of bad loans were detected
higher rates of application fraud could have been detected at the point of origination
In addition to enhancing the loan book performance, several other benefits could have been achieved:
high costs of manual underwriting and loan management driven would have been significantly reduced through Finexos automation of processes
decision times would have been made dramatically faster for borrowers
improved outcomes would have helped the lender meet the Consumer Duty active monitoring of consumer behaviour
using Finexos throughout the product lifecycle would have allowed for continuous vulnerability monitoring
*A bad loan is a non-performing loan (NPL) that is subject to late repayment or is unlikely to be repaid by the borrower in full.