Employer Credit Risk Model at AU Small Finance Bank

Vedansh joined the credit risk and policy team at AU Small Finance Bank as an intern. The team's function was drafting credit policies that determined whether a customer qualified for a credit card. The existing model treated all applicants primarily on income, without accounting for the stability of their employment.

His manager identified a gap: government employees and military personnel had more stable income than employees at startups or small private companies, but the existing model did not distinguish between them. A government employee with a modest income could be a better credit risk than a higher-earning startup employee, but the model was rejecting them at the income screening stage.

Vedansh was asked to build a categorization system that incorporated employer financial health as a credit signal.

Building the Categorization Framework

He researched existing employer categorization frameworks available internally and from external sources that published employer creditworthiness ratings. He built a four-tier system:

• Super A: central government employees and military personnel, highest income stability
• Category A: PSU workers and state government-associated firms
• Category B: large-cap private sector companies with strong employee cost-to-revenue ratios
• Category D: employees at smaller or higher-risk private employers

For each tier, he assessed job security, income predictability, and debt servicing capacity relative to income.

Testing and Implementation

He built a demo dataset of customers across all four categories and ran them through the proposed policy changes. The team reviewed the outputs and refined the thresholds before implementation.

The policy was piloted in Jaipur and Jodhpur in Rajasthan. Super A and Category A customers saw an 80% reduction in credit card approval time. Credit limits were increased by 20% for Super A customers and 10% for Category A customers relative to the baseline. Terms were also relaxed for stable-employer categories.

The pilot showed an increase in customer conversion and satisfaction, and a reduction in credit risk as more stable-income customers moved through the approval process.