cyberquantic logo header
EN-language img
FR-language img
breadcrumbs icon
Banking & Finances

Credit Lending & Scoring

Credit scoring using KYC data

Credit scoring using KYC data

Goal:
Other
Problem addressed
Assigning a risk score to every loan applicant in real time, using just KYC data,
which would ensure both new-to-credit and mature customers can be assessed
for their creditworthiness, and offered loans on appropriate terms.
Scope of use case
Building a risk scorecard for loan applicants using KYC data for better risk
management and high population coverage.
Description
Financial institutions find it much easier to assess customers
with an existing credit history, or those living in urban areas.
There are also several credit bureaus that assist them in this
endeavour. However, these frameworks do not work as well
for new-to-credit customers, especially in rural areas.
If only industry wide models or simple heuristics were used
to score such customers, many deserving loan applicants
would end up not getting a loan or not getting it on
reasonable terms. Instead, if a good risk scorecard is built
using KYC data, which is collected from every loan applicant
as a routine and regulated process, it is possible to ensure
every applicant receives an objective score.
To tackle this problem, non-linear models such as Random
Forest and XGBoost are being used which can accommodate
many parameters, including categorical ones, and are
reasonably resistant to noise in the data.
Interested in the same or similar project?
Submit a request and get a free project evaluation.