Tom Purves – Product Management, Fintech & AI

So I was initially a little intimidated to be on the same conference agenda as folks like Eric Schmidt, Pat Gelsinger, John Rettig (President of Bill.com), Ed McLaughlin (President of MasterCard). But, there I was, invited by the good folks at Deutsche Bank, to talk to their investor clients about the future of fintech.  Over a couple of hours, we held a wide ranging conversation, but one of the major themes kept coming back to the prospects for the major public card networks. My message was consistent with what I’ll always tell my advisory clients: the biggest threats to the established payment networks are also their biggest opportunities.

Regulators (and the DOJ)

Regulators can and will keep shaking up the economics, rules, and market access of the card networks. The trendline is piecemeal and stochastic, but the major networks do expect global interchange to trend ever-downwards in the long run. The forever-pending (and rather flawed) Credit Card Competition Act (aka Durbin 2.0) could significantly disadvantage Visa and MasterCard vs Amex, Discover. But who knows what final form that legislation could take, if it ever did get back on the legislative agenda. But wherever you have disruption, you also have new opportunity. When Durbin1.0 regulated and set caps on debit interchange, arguably Visa came out ahead. Network-level processing fees weren’t overly impacted while lower blended interchange helped grow acceptance. At the same time, that regulation carved out interchange exemptions for smaller banks. That boost for smaller banks helped the networks balance the market concentration of their clients while also helping to spur an innovation explosion of fintechs and neobanks growing new debit programs, sponsored by those smaller FIs.

[edit. US DOJ today released a significant anti-trust complaint against Visa related to perceived monopolization of US Debit. I will say this, there is at least one majorly flawed assumption underlying both the DOJ complaint and the Durbin acts. The assumption that either the low cost PIN debit networks were capable of innovating enough to usefully serve ever-evolving needs of the market -or- that new tech-company-led networks potentially capable of out-innovating the card networks (say Apple, Amazon, Afirm/Klarna etc.) would actually charge lower swipe fees. One only has to look at the eye-watering take-rates of the apple-app store, of FBA stores on Amazon, or the high discount rates of BNPL forms of tender etc.)

But regulators don’t just regulate payment networks, they also increasingly regulate banks, processors, fintechs and technology providers. And where regulators create disruption and compliance challenges, those industries suddenly need value-add compliance solutions. Often non-optionally and on a deadline.  In many parts of the world, banking and payments regulators have been a boon for the sudden creation of demand for solutions in strong customer authentication,  identify verification, data protection, realtime consumer notifications and disclosures, data security and residency etc. FIs and large enterprises struggle just to keep up with regulatory, security mandates at the same time as tech debt let alone have room on the roadmap for innovation whereas Visa, Mastercard and their partner network have the economies of scope and scale to offer more and more VAS to clients with few other options.

New Payment Rails

The networks are scared of UPI in India and Pix in Brazil. In both cases, central banks created a new payment rail that was cheap, fast, modern and essentially mandated into existence as a public infrastructure. In many other markets we also see closed loop wallets and super apps capturing big chunks of local commerce. Realtime account to account payments are meanwhile well established in Europe. In the US we have closed loop flows in PayPal, Venmo, cashapp. And then we have new forms of tender like BNPL and pay-by-bank. So the possibility of course is that new rails capture much of the future global growth in global payments, or that they start to meaningfully cannibalize traditional C2B retail commerce as well.

For now, the card networks are a little less afraid of US new RTP rails of real-time ACH and FedNow. These bank-designed RTP systems in the US are accidentally/on-purpose closer to bare-metal payment rails than sufficiently-featured payment services. These systems lack essential features such as user-friendly payment aliases, standardized user experiences, network rules for buyer/seller protection, and transaction economics that incentivize buyers. As a result, there’s an opportunity for networks (or other wallets/players) to build significant value-creating overlays on-top of these new payment rails in the US. Visa and Mastercard are leaning into their respective open banking acquisitions (Finicity and Tink respectively) to monetize over-top of otherwise cheap/free new realtime rails. Visa’s strategy with Visa+ also holds interesting promise for becoming a universal bridge between global p2p payments apps and schemes.

New Payment Flows

With new rails, also comes vast opportunities in ‘new payment flows’. The numbers here are potentially staggering. Christopher Newkirk (Visa EVP) put it really well at the event “In the markets that we operate in, there’s $200 trillion of flows we are going after…$145 trillion of that is B2B, and $55 trillion is the other sets of kind of disbursement, payout, P2P, push payments payouts. In the $145 trillion of B2B, go spend time with like the
function inside your company that is doing buyer supplier payments, and you will just, you might be surprised at how much manual work is still going on there.”. The same comments were echoed by John Rettig of Bill who talked about a massive opportunity ahead to properly ‘digitize’ the vast flows of B2B invoicing and bill payments in check and ACH. And I see this opportunity too as potentially the next great secular shift in global payments that could sustain continued growth of the big networks, just as the last great secular shift from cash to card has sustained them the last few decades. But only so far as the networks can actually (see above) get in on that action, and add meaningful value to justify meaningful payments yield.  Meanwhile the threat can also go the other way as comoditized realtime rails build scale in ‘new flows’ but ultimately spill over and cannibalize volumes and yields in retail commerce.

The Cracking Open of Banking

Open Banking makes pay-by-bank easier. Open Banking makes ACH rails better / less bad. Players like Plaid offer great service out of pre-verifying accounts and available balances before pushing or pulling funds, doing a good part of the job done by a credit card network’s ‘authorization’ transaction. Any open banking layers that helps to make alternate rails more useful, or helps to commoditize pay-by-bank services is a threat to the major networks.

Meanwhile, the coming combo with faster/cheaper new payment rails means that open banking isn’t just for ‘read only’ usecases any more. We could expect to see an explosion of innovation in cash management and financial product innovation for consumers and small to med-size business segments. Especially when combined with the next two trends I’ll mention (the increasing embedding of payments in vertical software and AI applications and vice versa).

But the card networks have a play in Open Banking too. MasterCard has Finicity, Visa has Tink (a consolation prize for not scoring plan A which was Plaid, but still). See Visa’s recent announcement “Visa to upgrade pay-by-bank service in UK next year” an interesting play combining their acquired open banking infra with UK’s new payment rails (UK Faster Payment System) to offer pay-bank services, even if it’s technically disruptive to some traditional Visa Debit use cases.

Software eating the world

I am bullish on embedded fintech, payments and finance. Shopify has become a juggernaut in ecommerce thanks to their better software and developer-friendly ecosystem. And because of having better software, they win the payments business of all the merchants on their platform. Same with Toast, who could be on track to becoming one of the single largest merchants in the US. The same with Squire for barbershops, Zenoti for spas, Jobber for home services etc. Not to mention the gig economy giants AirBnB, Uber, Lyft etc. In the last decade, Stripe has created something like 65B of market capitalization largely just on the strength of making payments easier to integrate with software.

Historically, the networks have passively benefited from the rise of vertical software platforms accelerating the shift to digital (and mostly card). But as these platforms grow, they can start to concentrate the seller market and demand larger discounts on network fees and network VAS. That is a threat to the networks’ margins. On the other hand, these same embedded payments players are also driving massive growth in new flows (not just retail purchasing but b2b, earned wage access, disbursements, issuing, cross-border etc.) and have more complex needs (bridging multiple payment rails, geographies, more complex fraud and compliance challenges). If the networks can significantly expand vertically into the software domain, significant growth opportunities remain.

AI eating Software

AI really is going to change everything. The first areas to be transformed by AI are customer service and software engineering. By lowering the cost of customer servicing, underwriting, loan servicing the number of economically-bankable customers globally could expand significantly. Smaller ticket lending and BNPL becomes more economical (see Klarna’s recent positive results spurred by AI efficiency gains). But as AI-driven efficiencies quickly become tablestakes for bankers and merchants, the mid and longer tails of these banks and merchants struggle to keep up and increasingly turn to their vendors or the networks for a larger share of SaaS services. This is the bull case. Accelerating software efficiencies drives more secular shifts to digital payments as well as creating boundless more opportunity for AI-enhanced SaaS value-add services.

But the (credit) card networks have one really big problem when it comes to AI. Data is the fuel that powers the AI economy. Visa and Mastercard will brag about the enormous breadth of their payment data. But the problem is, their data is a trillion transactions wide, but it’s only an inch deep. The networks are operating on a data-starved messaging standard, ISO8583, which severely limits their ability to capture context like SKU data or loyalty info. Except in a few narrow cases, card networks have very little contextual information about each of the payments they transmit. ISO8583 doesn’t have room for sku level data, loyalty data or even basic contextual information about the buyer and the seller. And AI, as well as many other potential value-add services need context in order to really do anything interesting. The networks have structural challenges both technically and incentivationally (can we make that a word?) to carry more payment data.

By contrast, competing new RTP payment rails, built on ISO20022, have built-in support for rich payment data. Initially, that rich data is already a huge boon for commercial and b2b payments.

The over-concentration of both giants and minnows

At the small end, small banks (particularly in the North America) keep disappearing. Fewer banks means fewer copies of key value-add-services like debit processing that the likes of Visa can sell to coop and community FIs. Small banks also have small volumes, but also punch way above their weight in pricing and margin contribution. I believe this mix-shift to fewer-but-larger clients is a key contributor to growing incentive payouts and margin pressure. On the other end of the scale,  consolidation of the larger issuers, acquirers, processors and merchants gives that side of the network more power too. The biggest clients keep getting bigger as do the global tech giants, and so do their bargaining power. I am also watching what happen as vertical SaaS platforms continue to hoover up smaller merchants behind their platforms. Consolidation there too brings better bargaining power, incentives and volume discounts and these factors will continue to squeeze on margins of the big card networks. There’s a thesis out there for example that Toast could quietly emerge as one of the largest single merchants in America, perhaps giving a million smaller merchants behind them much more economic bargaining power on their cost of payments. 

Looking for the opportunity side here, the overall super trends of the ever-increasing digitization and interconnectedness of global commerce means that network effects of the biggest networks just become more and more valuable. Sometimes you don’t really need brilliant execution, the most modern/best standards just so long as you have a universal standard that’s good enough. The powerful forces of scale & scope economies, compounded by network effects, rewards returns on scale to the largest participants and that includes the networks. Lastly, [and to be fair, the DOJ has certain bones to pick with this angle] consolidation of powerful platforms across the industry creates an opportunity for Visa, MC, Amex to strike grand strategic partnership deals with each of these players across multiple markets.  

---

Hey if you’ve read this far and want to dive deeper on any of the above topics, feel free to setup a call. I offer paid rates for anyone in the investor community, or for founders or product leaders first 30min of expert feedback or advisory is always free.