Insight, Impact, Innovation
HomeOur WorkCase StudiesHow do I safely book more small business loans?

How do I safely book more small business loans?

SME credit scoring analysis frees up working capital

The Business Challenge

PAE is one of India's largest auto components suppliers, sourcing from multiple manufacturers and selling these components to over 10,000 retailers across the country. Most retailers buy products from PAE on a 30 to 90 day extended credit period, which is critical to enhancing sales. However, credit extension was contributing to serious credit losses, resulting in direct erosion of profits and indirectly blocking working capital, which could otherwise have been deployed to enhance revenues.

The challenge was to create a process allowing PAE to extend credit to retailers with low default risk, while systematically reducing or refusing credit to the potential defaulters.

The Solution

Fractal Analytics recommended a scorecard-based approach for extending credit, which would score retailers' behavior on PAE business transactions on an ongoing basis according to their risk level as input for assessing default likelihood.

Integrating the data

Our analysts determined that the solution could be derived by identifying and integrating the best data to build the model. We combined billing data with repayment information to generate a monthly delinquency status of all retailers. Next, we integrated information that could be predictive of the retailers' credit behavior. Then, our database team collected information from multiple accounting systems to generate behavioral information on each retailer.

Exploration & analysis

Fractal's analysts studied the prepared data to identify patterns and factors that could map current data to future default events. Multiple aspects of the retailer's relationship with PAE were explored, including purchase frequency, nature of products purchased, purchase amounts, past payment behavior, average payment period, and size of business.

Through the application of advanced analytical techniques, Fractal constructed a robust, predictive model that powerfully segregated the potential defaulters from non-defaulters.

"The data visualization phase was the most intriguing and enlightening part of the engagement," said Pritam Doshi, Vice President PAE. "We learned a lot about the behavior of our customers that we didn't know over the past several decades! Furthermore, we are accruing strong benefits from Fractal's scoring strategies –– primarily through lower defaults, enhanced revenues, and better working capital management."

The Results

Initial results proved that Fractal Analytics' behavioral risk scorecard could very effectively segregate the defaulters from the non-defaulters.

A few months after implementation, subsequent model results also delivered very powerful results, capturing more than 50% of defaulters in the top score of retailers. PAE has successfully deployed the model to create enhanced credit extension decisions and today is able to allocate working capital across retailers more effectively and more efficiently.