Traditional credit ratings are fairly accurate in predicting future loan performance, and that’s why lenders have tended to focus on clients with already a good credit rating, as screening them is less pricey. However, curiosity about other ways to recognize potential good borrowers that lack credit rating keeps growing, specifically in countries in which a non-trivial fraction of people remains unbanked.
Elevated computing power and accessibility to big data have led the way for brand new technologies that predict repayment behavior using information not the same as the typically collected. Data from bills, tax payments, employment records or perhaps cell phones or social systems for example Facebook is presently getting used for this finish.
One particular alternative tool produced by EFL, a fintech company founded this year, depends on psychometric testing.[1] With different personality and behavior test adopted a touchscreen device, the tool creates a 3-digit score that within a few minutes predicts financing applicant’s future repayment behavior. The tool includes a number of psychometric questions that extract attitudes that predict credit risk and future default behavior. The applying extracts info on how and what you solutions. For instance, the tool records time it requires a customer to reply to each question, if your client hesitated to reply to, or maybe she altered solutions.
Inside a recent paper, we partnered and among the biggest banks in Peru to review the potency of it for screening SME loan applicants. The aim of our partner bank ended up being to expand its SME portfolio, and also the EFL tool offered them an alternate choice to better screen the SME segment. Their conventional screening method (which trusted a fico score from Equifax Peru along with a site trip to the firm) was more appropriate for bigger firms and led to a higher rejection rate within the SME segment.
Throughout the pilot, all SME loan applicants required the EFL oral appliance received a psychometric credit rating. Applicants who achieved a score greater than the usual threshold set by our partner bank were offered financing, to be able to make use of a regression discontinuity (RD) design to judge the potency of the tool.
According to several RD methods round the EFL score threshold, we first study how effective the EFL tool is at growing use of credit. If insufficient credit information isn’t a barrier for SMEs searching for loans, SMEs by having an EFL score underneath the threshold may potentially obtain financing elsewhere, by which situation loan use might not differ round the EFL score threshold. Inside a second exercise, we evaluate if SMEs which were offered financing in line with the EFL tool exhibit worse repayment behavior than SMEs which were offered financing in line with the conventional screening method.
Since loan applicants who lack credit rating will benefit more from technologies like the EFL tool, we classify our loan applicants in 2 groups: thin files versus thick files. As the Equifax lots of thick file applicants derive from their credit rating, the lots of thin file applicants derive from other resources because of the lack of credit rating.
Within the six several weeks following a pilot, we discover the EFL tool elevated the prospect of getting a brand new SME loan from the lender by as much as 19 percentage points (figure 1). This increase being greater among applicants with thin credit agency files, with additional as much as 59 percentage points.
To determine repayment behavior we first consider the fraction of applicants who’d financing in arrears in excess of two months inside the 24 several weeks following a application for the loan. Thin file applicants screened through the EFL tool were as apt to be in arrears as thin file applicants screened through the conventional method. However, for thick file applicants, being screened through the EFL tool elevated the prospect of finding yourself in arrears.
We check out the Equifax credit rating four years following the application for the loan. Again, worse repayment behavior was discovered only among applicants with thick files (figures 2 and three). One good reason with this finding might be that even applicants with poor credit histories received loans when they had high EFL scores.
Overall, we discover that psychometric credit rating seems to become a viable screening way of loan applicants with no credit rating: they are more inclined to obtain loans, but don’t pay back worse, than underneath the traditional screening method. Our findings also highlight the strength of credit agency information: a job candidate with a low credit score history will probably have repayment problems later on, regardless of what the psychometric score states.
Using psychometric tools in SME Finance and other associated topics is going to be discussed within the approaching Overview Span of Financial Sector Issues around the globe Bank. We encourage you to definitely get more information at details about the program and the way to register.