12May
Developing & Designing of Credit Underwriting Algorithm 0 Comments | By admin
Business problem:
The bank had no scorecard for underwriting process for its WC, BL, PL, Agri –TL products; leading to a non-standardized underwriting process. This further led to high TAT, low productivity, low FTR % etc.
The mandate was to develop a customized credit underwriting algorithm and parameterized scorecard for new customers and renewal customers of WC, BL, PL, Agri –TL, affordable housing products, by doing comprehensive testing of historical data, defining weighted average scoring model, defining critical Go / No Go parameters and creating final scorecard solution
Solution Approach:
Current state analytics on credit underwriting process and identifying dominant parameters which have a strong influence on the decision-making process
Back testing of historical data and relative performance analysis of all those parameters and study its impact and influence on underwriting process
Develop a multi-factor parametric equation by studying various dominant parameters of the process and defining them in to an implicit equation
Designing banking and turnover surrogate policies
Post finalization of parameters by relevant stakeholders, we start developing the algorithm on the platform as required by the client
Once the product is developed, we do a rigorous pilot testing based on various test cases before going live
Providing coaching, hand-holding, mentoring and on call training sessions with the users
Key Outcome and benefits:
Reduced sanction TAT (by over 97%) from 3 / 4 days to 40 Mins for sanction/ rejection for cases up to INR 2 cr.
A robust real time tracking mechanism which helps senior management monitor and track parameter performance, auto approved cases, auto reject cases, deviation approval tracking and monitoring rejects, load sharing, TAT, FTR etc.
