The economics of card led BNPL
The previous blog post decoded the economics of UPI led BNPL. If you haven’t already read that, please do so now, as it provides the necessary context for this post.
However, UPI is not the only form factor to offer BNPL. If anything, NPCI has taken a view that UPI cannot be used to offer BNPL without an underlying bank account (an overdraft account). On the other hand, the regulator has restricted the issuance of credit cards to only the banks (there are exceptions to this rule. RBI has permitted two non-bank entities, to issue cards: BOB Cards and SBI Cards). Thus, fintech companies such as Slice, Moneytap, and Uni have issued what they call “credit line cards” to their customers that are nothing but prepaid cards with clever fund management mechanisms. This works almost like a credit card from the user’s point of view. They don’t use the term “credit card” because the regulator has expressly barred such usage to avoid any confusion for the customers.
The fundamental reasons why companies are increasingly preferring this route:
a) Flexibility on billing cycle definition (weekly, biweekly, fortnightly, monthly)
b) Total control on the approval rate
c) Faster GTM (getting a prepaid card program approval takes just a day or two)
The economics of this is not very different from that of the UPI-led BNPL model that we had explored earlier. There are few additional factors to keep in mind here:
- Printing & shipping a card involves cost
- Every time someone uses the card at a merchant store, there is revenue to be made from interchange
- KYC (read the new regulations here). Specifically, full KYC as per banks would mean a bigger commitment on your part.
Let’s understand these and then take a look at the total cost.
Card Printing & shipping cost
This actually depends on the type of card. A typical card with an NFC chip costs about 120 rupees. If you want a fancy metal card like what Onecard has, then it could easily go to a few thousands of rupees. Let’s leave that out of our consideration. A conservative estimate of Rs.110/- per card is more realistic. Let’s add another 20 rupees for packaging and another 30 rupees for shipping (blended rate across pincodes).
This creates a fixed cost of Rs.160/- for every customer that you onboard.
Given that the cards expire and taking into consideration of customer churn, let’s suppose that an average customer stays for 2 years. This would work out to ~80 rupees per customer per year.
Revenue from Interchange
The blended interchange revenue that can be availed in the market is about ~0.7%. Let’s make a generous assumption of 0.8% (as we can assume that you will receive the GST paid as input credits and can offset it against other GST payments). In the previous post, we assumed an average spend of 3000 rupees per cycle.
Utilised value = 3000 rupees
Interchange share from bank = 0.8%
Revenue from interchange per month = 24 rupees
Total Revenue and Cost
Basis the above and the numbers from my earlier blog post (15-day billing cycle, 5 days grace period, spend of 3000 rupees spend per cycle) below is the new calculation. The numbers:
Revenue items
Balance conversion to EMI revenue / customer / year = 180 rupees
Late payment revenue per customer p.a = 65 x 2 = 130 rupees
Transactional EMI revenue = 240 rupees
Interchange share revenue p.a. = 300 rupees
Cost items
Card cost / customer / year = 80 rupees
Cost of capital / customer / year = 15 x 26 cycles = 390 rupees
Revenue earned per customer p.a. = 180 + 130 + 240 + 300 — (390 + 80) = 380 rupees
At 12% cost of capital (and you choose not to pass on the benefit of lower cost of capital to the customer), the cost of capital reduces to 260 rupees. Accordingly, the revenue per customer increases:
Statement EMI revenue = 240 rupees
Late payment revenue = 160 rupees
Transactional EMI revenue = 360 rupees
Interchange share revenue p.a = 300 rupees (no change)
Revenue earned per customer p.a. @ 12% CoC = 240 + 160 + 360 + 300 — (260 + 80) = 720 rupees
With card as a form factor, we see that the profitability at a customer level increases quite a bit without adding any significant risk item. However, please keep in mind that our calculation is based on a fully engaged user. In reality, there is a significant non-usage of cards that are shipped to customers. Industry average activation rates are 70%. Any significant reduction in utilization % or increased customer churn can result in an increased fixed cost burden thereby rendering the margins low.
Things to be careful about
a) Designing your own card also requires you to make a commitment to the printer by way of minimum order quantity. If you are just getting started, a safer number to go with could be 5000 cards. Even though you will be paying a higher cost, it is a much better option than committing to some large number only to find out that user response is muted.
b) Scheme (Visa, Mastercard, Rupay) transaction costs are different with Rupay being the cheapest. These costs are fixed in nature and impact the revenue that you make from the interchange. Sometimes, you may end up paying the scheme than the other way around.
c) Cards tend to get damaged / lost / natural expiry. Replacements would have to be offered free of cost.
d) Card systems have elaborate dispute resolution mechanisms and you will incur charges.
e) Engineering cost involved in making the tech investments and achieving PCI DSS compliance (as required).
There is no single solution for credit. So we continue to see growth in Credit Cards, BNPL, and Prepaid cards (as lending form factor).
PS: Our company Hyperface is a card issuance platform that will help you to focus on the things that matter: better activation rate, better card designs, better customer UX, score cards, underwriting, collections and customer support.
To know more about our offering, head to hyperface.co or write to us here: business@hyperface.co.