A new approach to identify bad loans is needed: Credgenics co-founder

The pile-up of bad loans is a pain point and the pressure is mounting, said Rishabh Goel

Credgenics answers to a unique profile — it is the country’s only technology-enabled platform for resolution of non-performing assets. It recently raised $25 million in a Series-A funding round from WestBridge Capital and Tanglin Venture Partners, which saw the participation of Accel Partners at a valuation of over $100 million. Rishabh Goel, co-founder, spoke to Raghu Mohan. Edited Excerpts. 

How has your firm been imagined?

I was with Blackrock on the distressed assets portfolio and was familiar with the pricing of transactions in the US and UK markets. While this was not a very big problem in the developed markets, it was a black-hole when it came to India and south Asia — no data-points, visibility, or clarity on what is to be done. Now when it comes to the pricing of risk, you can only do it if you have knowledge on the collections part. What are the chances of default? What’s the average amount that can be collected after it goes into default? Hence, we thought the immediate thing that can be given to the market is to bring in a technology-driven platform. To automate collections, try to bring data-learning that can offer insights as to how to prioritise accounts. Or, how to be sympathetic and do away with this practice of harassing borrowers as it doesn’t eventually lead to actual collections. And also be able to automate a lot of procedures and collect more. Now, when it comes to underwriting, credit penetration is around a tenth of  what it is in the US. This has to increase. So, our aim, firstly as a company, is that every payment which has to be collected is done through the Credgenics platform, which will also impact the speed of lending and improve credit penetration.

How do you see the buy-now-pay-later (BNPL) business coming along, especially the spectre of delinquencies? Are you working with BNPL firms?

BNPL is the newest segment gaining substantial ground in the fintech ecosystem. Apart from its utility in e-commerce purchases, it is increasingly getting popular in the offline world as well because it eliminates the need to carry a credit card when your mobile phone can help with the same. In BNPL, the amount is directly settled to the merchant’s account against a purchase, same as credit card, hence the delinquency pattern is not very different compared to a card portfolio. Yes, we are working with several BNPL players — with BharatPe, Capital Float, and ePayLater, among others.

Many asset reconstruction companies (ARCs) are getting into the retail part of the game. How do you see your role in it?

The pile-up of bad loans is a pain point and the pressure is mounting. For bad loans in the high days-past-due buckets, which are typically bought by ARCs, you have litigation issues involved along with the usual collection efforts. While our digital collection efforts have been proving the mettle with each of our banking, and non-banking financial companies, the legal modules on the platform, along with our e-mediation and online dispute resolution offerings, would also be very relevant for the ARCs. We are glad to have already partnered with large ARCs such as Reliance ARC and the Asset Reconstruction Company of India.

With all the technology on hand, one would have assumed that lenders would be mapping their customer’s behaviour better. Where exactly do you think is the shortcoming on their part?

The adoption of technology in banking and finance has been recent, so yes, we can expect better results in the future. Features such as voicebots, however, ride on the database where the common questions or doubts raised by the borrowers, or bank customers are recorded. The process is gradual and we can expect revolutions in this space with time. Fintechs are, however, doing better at mapping the borrower’s behaviour, spending and repayment histories, their loan demands, among other such metrics. For instance, we also tap the borrower’s behaviour and specifically their nervousness factor by monitoring how many times they have clicked the digital-notice URL.

How do you go about managing this data and what are its sources? Can you pull in, say, data from within a consortium and share it?

We are into the retail side, not the corporate loan distress part. Again, we are not merging any data from banks. We are bound by strict regulations. And separate instances of bank data can’t be merged, but you can use the intelligence generated from one portfolio to another. I’ll tell you how. We generate very good data points at the pin-code level. And if there is a huge difference in the collection rates of a lender versus the industry standard, it essentially means that the problem is with that lender, and it’s not pin-code level problem. We pass on all the insights at the industry level to lenders. Now if you (as a lender) have a two-wheeler portfolio in Odisha with issues in it, and we also have a client with a two-wheeler portfolio in the state, we can at least help you replicate the practices of the better-managed portfolio. You also have to bear in mind that updates in the credit bureaus take three to four months