Now for a quick detour into the economics of credit cards. The global electronic payment processing industry has two large players (Visa and Mastercard) and a few smaller players (American Express, Discover and JCB). Large players partner with local banks to issue cards, while smaller players used to issue cards independently, although this is also changing now. Merchants are charged for each transaction involving the use of debit or credit cards (merchant discount rate or MDR), and the proceeds are shared by the initiating bank, the terminal bank and the intermediary.
The economics of a typical card issuer in India – SBI Cards, for example – looks like this. Depending on the channel, the company spends 3,000 to 4,500 rupees to acquire a cardholder. On average, a cardholder spends Rs 135,000 per year. Of this, she repays Rs 112,500 at the end of the billing cycle, and the balance (Rs 22,500) is converted into loans.
Of this 22,500 rupees, 11,300 rupees is deferred by paying only, say, 5% now, and the balance later; Rs 8,200 is converted into a term loan (with EMI) and Rs 3,000 is paid in the coming months. Credit card companies charge an interest rate of 30% and above to those carrying the balance and 20% for EMIs.
The card industry’s revenue stream is twofold: net interest earned and fees and other. On average, each card earns net interest of Rs 3,700 and fees of nearly Rs 4,400. On the other hand, the costs are at three levels: expense-based costs, marketing and collection costs and receivables irrecoverable. The first and second above are Rs 3,300. Historically, credit costs were only 5% of assets (or Rs 1,100), bringing the after-tax gain per card to Rs 2,700.
What a fabulous company. An investment of 4,000 rupees generates 2,700 rupees after tax, breaking even in 1.5 years.
This is what made the card business so special: India’s digital penetration would increase, and in that context, card penetration would increase. And who doesn’t want a fast-growing company that breaks even in a year and a half, right?
But things started to change a few years ago. India’s digital revolution had already been underway for years under the umbrella of India Stack (a set of open APIs and digital public goods), and it got a huge boost with the introduction of the interface card (UPI) in 2016. Compared to credit cards (1.5 to 3% MDR) or debit cards (0.5 to 1% MDR), UPI does not charge merchants. Customers tie their bank accounts to UPI and there are no credit reporting requirements. Much of India has no credit history. Therefore, UPI dominates, at least in small transactions.
Here are some numbers to consider. Over the past two years, digital retail payments have grown from $420 billion per month to over $620 billion today. Of that extra $200 billion, more than $70 billion came in the form of higher UPI payments. How much did credit cards contribute? Less than $6 billion.
For January 2022, credit cards’ share of retail digital payments was less than 2%, while UPI’s was 18%. It was below 9% in 2020.
But then, we’re just beginning our digital journey. India Stack’s end goal is to deliver digital services by providing solutions to four key challenges: establishing customer identity and eliminating tedious paperwork; eliminate the dominance of physical silver; protect user data and penetrate currently unbanked areas with low-cost credit.
The Jan Dhan, Aadhaar and Mobile (JAM) trinity, together with UPI, solved part of the problem. Over the next decade, as additional layers are implemented, the “lending ecosystem” will be completely redefined.
The next two layers in the stack are the Account Aggregators (AA) and the Open Credit Enabling Network (OCEN). The Reserve Bank of India (RBI) already regulates NBFC AAs which will operate as aggregators of user data (which is currently hosted individually by lenders) and use the data to “democratize credit” through the OCEN, which has was launched in 2020.
Suppose an SME with a limited credit history wishes to borrow an amount too small for banks (or NBFCs), the OCEN network would be used by Lending Service Providers (LSPs) in a way that reduces operational costs and with greater efficiency. Once fully operational, the innovators plan to bring over a billion people into India’s formal credit network.
This presents a huge opportunity for existing players, but it also comes with huge risks. Once data is democratized and shared at will, the “fluke” – customer data collected over years of operation – disappears.
Listed shares of card companies have been under pressure lately due to a spike in bad debt due to Covid-19. The cost of credit for SBI cards, for example, has more than doubled, to 11% in 2021, from the standard 5%). As India’s ecosystem begins to normalize now, so would credit costs. In the short term, this can provide respite. But a bigger challenge looms.
For us, the credit card product is a combination of payments and loans. For low-priced items, UPI already beats the cards hands down. The question is: can it be reproduced for larger quantities? In advances, a blended interest rate of 21% and a business that generates more than 25% ROE is extremely high. As credit begins to democratize, credit card companies will likely lose their fluke (customer data collected over years of operations) and new fintech players will eat away at their highly lucrative market.
The entire payment ecosystem in India, over the next decade, is likely to see a sea change once the entire Indian stack is fully implemented. New business models will emerge and existing players will be forced to change or exit. The business model of credit cards, which can essentially be best served by breaking it down into payments and credits, is likely to be widely debated. While the short-term pain will certainly ease, as credit costs come down, it would be very interesting to see what happens to this business in the longer term.