• Fintech Brainfood
  • Posts
  • Fintech 🧠 Food - Nov 8th - Whatsapp does payments, DOJ vs Visa, Square crushes earnings and why A2A payments matter

Fintech 🧠 Food - Nov 8th - Whatsapp does payments, DOJ vs Visa, Square crushes earnings and why A2A payments matter

Hey everyone 👋, thanks so much for coming back for more brain food, covering four fintech's that caught my eye this week, an in-depth look behind some of the biggest stories and best content of the week. It's the "I know kung-fu" version of what happened in fintech. Apparently. :)

If you haven't joined yet, you can subscribe right here.

This week Credit Kudos launched an excellent report on the power of alternative credit data; you should check it out :)

I've seen a lot this week about how regulators make changing a core banking system hard. Twaddle.  Being regulated is hard; changing a core banking system is hard, but that's not the regulators doing.

For the uninitiated, "core banking" can refer to several things, but generally, people mean the system of record (SOR), which is the golden source of truth for a financial institution. The SOR is the truth and where the money sits, typically containing account numbers balances, and this is the data that gets reported to the regulator.

Core banking can also mean payment systems and any number of surrounding services (like regulatory reporting, product management (checking, deposits, lending), payments integration, and more. Some people even use "core" to refer to any internal critical service. How you define "core" is essential because that impacts how you might change it or solve its problems. For that reason, I like to think of the core as the system of record.

Core banking software gets a bit of a bad rap. Banks spend billions every year upgrading their 1970s core, written in languages that pre-date Fortran and carry significant operational risk from lumbering all of this legacy around them. But the issue isn't the language they're written in or that they're running on a mainframe. There are two problems 

  1. Banks built with the best tech they had at the time 

  2. Banks are subject to Conway's law

Until the early 1990s, it was quite common to have file sizes limit how much information can be transferred (and it's why payment file standards like ISO 8583 have such weird encoding). Networks didn't allow for more data. Engineers did amazing things with those constraints and built a global financial system that was astonishing for its time. The passage of time has not been kind to these systems, however. Over the years, they have expanded to millions of lines of code as each new regulation, product, or complexity came along. The fact that these systems continue to run and support banks with 10s of millions of customers is a marvel.  Early in my career, one senior IT architect described changing the "core" as performing open-heart surgery inside a Nuclear reactor in a populated area.

Conway's law says that companies' IT real estate always resembles their org design. For banks, this is no different. Multiply this by the number of countries they operate in, decades of M&A, and duplication of systems for various products, and it gets messy, fast. To this day, the payments system vendor is different from the loan system, which is different from the deposit system.

Now banks find themselves with mainframes that no employee in the bank wrote the code for, in 30 to 60 different markets, with concentric circles of technologies added around the edges over the years, often with point-to-point integrations. These banks have to keep the lights on, reduce costs, and keep up with digital. It's no surprise then that the easy way out is to "outsource your soul" (as Jamie Dimon put it) and let someone else deal with the burden (as Deutsche Bank just did).

When a bank *does* bite the bullet and go for a core transformation, it rarely ends well. Banks would historically migrate from the old system to the new with a "big bang" migration. Banks tell the regulator the date at which their "system of record" will change and then spend a massive amount of energy and money trying to switch everything across in a weekend "outage."  It's a bit like trying to re-locate the entire population of the United States to Alaska and have the postal service redirect everything correctly, first-time x 1000. Big bang migrations are an enormously complex task. These are the types of migrations that make bankers and regulators nervous.

The cautionary tale that keeps bank CEO's up at night is TSB, the UK bank that tried a core migration and failed so severely that customers couldn't access their accounts, and some customers saw the wrong amount in their accounts. TSB caused a national uproar. The CEO was pulled into Parliament to explain what was going on. Then lost their job, along with much of the executive team.

Why would any bank do this? There are almost no examples of "successful core transformations" except one.  Commonwealth Bank Australia migrated from 10 core banking systems to one in the late 2000s and nearly halved their IT operating costs. The exec behind that, Michael Harte, is a good friend, mentor and former boss who's frustration at that process was that there were no excellent core banking solutions to put in the place of the legacy. Just updated versions of old ones. That's not to slur Temenos or Oracle, who've made incredible strides in their latest versions of their software. But they've all built ever faster versions of the 1970s mental model of what "core banking" is.

Now here's the mind-bending bit. To quote 11:FS CTO Ewan Silver, "Banks are state machines." 

Therefore banks need an event-driven architecture. Many banks have pockets of microservices or event-driven architecture but no coherent long term strategy for how this cuts across the organization or reshapes the very idea of how you model bank products and asset in / asset out.

The reason banks used to do "big bang" migrations is because they worried about the cost of running two core systems side by side (the vendors aren't cheap and hook banks into long contracts). There is no "genuinely SaaS" core banking solution (even the newer ones on the market's commercial model looks very much like the older vendors). Most modern core banking solutions are banking software that happens to be in the cloud. That's not the AWS of banking. (as an aside, most "AWS of banking" solutions are APIs over an old bank.)

If banks are state machines, everything they do can be described by an event-driven architecture.  The trick then is finding areas of in-year return and committing to drive that architecture throughout the bank without stopping. It can be done.  

Many of the Neobanks in the UK have a microservices, event-driven core, proving the newer architectures can be a regulated store of value. You have to know how to make it work. Regulators are looking for confidence that a bank can meet regulations. A shortcut and lazy way to do that is to use software someone else did, but like TSB, that may still be an incredibly risky strategy.

The core transformation is dead. Remaking banks as event-driven state machines is the goal. Explaining how this meets regulation and benefits customers is the hard yards banks and fintechs now have to do.

4 Fintechs 🤑

1. Sugar - Clearbanc for Game Studios (UK) 🎮

  • Sugar provides a range of non-equity-based finance solutions for game creators.  Like Clearbanc, Sugar uses data to give access to the capital early for tax rebates, funding for marketing and platform scale, as well as early access to funds from app stores or platforms. Sugar is the type of niche solution that the banks do not offer.  Games are a massive and growing industry with predictable revenues from various well-known platforms (e.g., Steam, Apple, Google, Sony, etc.).

  • This type of "ARR as an asset class" startup has done well, with companies like Pipe, Clearbanc, and Uncapped all starting to show real promise. Gaming has it's own unique data challenges that a generic e-commerce site doesn't, and Sugar gets that. If you're bullish on the metaverse, solutions like Sugar are super interesting over a ten-year time horizon.  If Epic is willing to go to war with Apple overpayment fees to feed game creators, imagine if they added something like Sugar to their portfolio?  Monzo investor Passion Capital led their £1.5m seed round.Yes 👏

2. Trilo.io - If Fast, Plaid, and Amex rewards had a baby (UK) 👶

  • Trilo is the latest in a line of companies in Europe to allow merchants to integrate account to account (A2A) payments. The user gets to the checkout, clicks the Trilo button, and gets directed to their bank app (like Plaid would). The user then authorizes the payment, and they're presto all set.  

  • Card payments in Europe are about to get quite a bit more painful and lower conversion. Merchants are about to get whacked with a European regulation called "strong customer authentication" (SCA). SCA will add friction at checkouts. But, A2A / noncard payments could be a solution. A2A is a competitive space with companies like Vyne and Banked getting a lot of momentum for the "European bank account driven checkout button."  

  • What I like about Trilo is they've thought about user acquisition and the merchant data play. They're looking at creating a loyalty/reward & data-driven experience. The Affirms and Klarna have done this well; they drive repeat spend to merchants. They're selling revenue, which is the easiest sale there is. Trilo is early, but exciting :).

3. Strivebanking - The kid's account that grows with them (UK) 🧒

  • Strive sets up a savings account for kids 0 and up, then introduces a piggy bank age 5+, then chores and pocket money service from 8+.  One of the issues with the "pocket money accounts" has always been they're relatively niche and don't drive a ton of revenue. With a million users paying $5.99 a month, there's a healthy top-line revenue (~$71m), but the cost of running these programs is high unless you're living off the interchange revenues (which in Europe isn't feasible)

  • Strive has an alternative; it's building a few other routes to monetize that child-parent relationship.  The savings account (if it includes investments) can generate revenue passively, and there's room for other financial products in here. I'd love to think about the journey from a pocket money account to a student account.  That transition is challenging and potentially includes a lot of student debt. Has anyone solved that?

4. Vivid - Another European Neobank 🤷‍♂️

  • Vivid looks a lot like European Neobanks. Metal card, manage subscriptions, zero fees sending money abroad. It's also offering free ATM withdrawals and sorting finances into pockets.  They just raised 17.6m Euro led by Ribbit Capital (who are not morons), but I don't get it. We've seen this movie.  European interchange isn't a good revenue line, the market is crowded, and the path to profitability is unclear. I'd love to know what I'm missing because this is a great deal for consumers.  Sorry to be that guy, but when they're doing at least 20 euros of cashback every month, how will this thing ever make money?

  • I usually hate when bankers bang on about profitability for such an early-stage company, and there's a lot to like here—Vivid runs on Solaris bank, the most complete "Bank as a Platform" player out there. Unlike Marqeta or Synapse, Solaris has a banking license. In theory, Solaris could switch on lending for someone like Vivid, and then it becomes a game-changer.

Things to know 👀

  • "By acquiring Plaid, Visa would eliminate a nascent competitive threat that would likely result in substantial savings and more innovative online debit services for merchants and consumers." Visa's response was basically, "Plaid isn't a payments company."

  • The DoJ also pointed to Mastercard's inability to gain more than 1/4 of market share as a reason why Visa is a monopoly.  This image has been circulating everywhere as part of the reason why Visa was allegedly so concerned.

  • 🤔 My analysis: The issue isn't that Plaid threatened Visa as a data business. It's that Plaid would threaten Visa IF it became meaningful money moving business. By acquiring Plaid, Visa can effectively prevent that from happening. It might seem far fetched that Plaid would become a money mover in the USA, where there's now another ACH or payments startup every week but look at Europe. We now have (even this week) companies that are exactly that—account to account (A2A) payment providers.

  • 🤔 My analysis: Plaid had a real shot at being the new money rails.  With the startups now arriving, one or two well-placed acquisitions could have created a meaningfully competitive rail.  Imagine if Fast and Plaid and Dwolla had a baby.  

  • 🤔 My analysis: This is a pretty lovely outcome for Mastercard, committed to becoming "multi-rail" and has already made acquisitions of companies like Vocalink who run the real-time payment infrastructure in the UK, Singapore, and Australia. Mastercard's incoming CEO is also German and may bring a new perspective to the job. Visa has made real in-roads into B2B payments in the past few years, pushing into territory that SWIFT would have historically held. Visa could get on the front foot and try to own the A2A space, but the bank lobby will be the issue.  

  • 🤔 My analysis: The bank lobby will be in Visa's corner; banks don't want to change their systems and dislike Plaid much more than Visa would have. There's no regulatory reason why account to account payments should take off in the USA, but there was also no regulatory reason why Plaid had to become a success. It just did. However, the big merchants (think Amazon, Walmart) would love a competitor to card payments. Don't look only at Plaid or Visa; look at everyone circling it.

  • WhatsApp pay can go live using the homegrown, multibank Unified Payments Interface, the National Payments Corporation of India said in a statement Thursday. The US firm can gradually expand its UPI base starting with as many as 20 million users

  • India's payments market is crowded with domestic pioneer Paytm, Alphabet Inc.'s Google Pay, Walmart Inc.'s PhonePe, Amazon.com Inc.'s Amazon Pay, and dozens of other startups. Yet, given its enormous user base of more than 400 million, WhatsApp has the potential to compete with the leaders and reshape digital payments in India — a market slated to grow to $1 trillion by 2023.

  • 🤔 My analysis: This is massive. But by no means a slam dunk. Whatsapp has several shop features (product catalogs, shops, and customer service tools), but they're far from the only ones doing this. The nature of India's infrastructure (UPI, Aadhaar, etc.) means Facebook doesn't "own" the payment in the way they might in other markets.  The WhatsApp payment "rail" can't create any ecosystem lock-in.

  • 🤔 My analysis: Facebook has been trying to get payments done in India forever. It feels almost like India got it done for them (not long after the Reliance Jio investment).  Two reasons I think this is more India succeeding than Facebook. 1) The India Stack means India can prevent a monopoly on the payment rail from emerging. 2) Facebook isn't Ali or Tencent.  If you zoom out and look at the Ali and Tencent attempts to own India's payments, they've been locked out as geopolitical tensions escalate. India appears intentionally and knowingly picked a way forward that works for it by building world-class, open-source infrastructure rails.

  • 🤔 My analysis: Now, let's see if Facebook can get its payments shit together anywhere else. At all. Ever. It's astonishing how little Facebook payments have achieved in the near-decade they've been at it.

  • ICYMI, Bitcoin steamed up over $15,000 (and at the time of writing is still hovering around there).  Some commentators suggested the political uncertainty fed into a core narrative for Bitcoin. Others point to the ongoing money printing by central banks globally as the pandemic continues to hit the global economy.

  • Square also CRUSHED it's Q3 earnings, with Square Cash app being the bright spot, more than 220% YoY growth, with more than $1.6bn in Bitcoin revenues.  Square cash, unlike Venmo, before it continues to drive gross margin too as a profitable business.

  • 🤔 My Analysis: There's possibly no more significant evidence of Bitcoin being mainstream than the massive revenue Square Cash App did in Q3.  Purely speculating, but I wouldn't be surprised if this were significantly more than Coinbase did in the same period from USD/BTC (I have no evidence for this; it's just the sheer scale)

  • 🤔 My Analysis: Square Cash app is a phenomenon.  Banks worried about big tech stealing their customers, but the real threat is the Square Cash app. It's 70m + users, engagement, and cross-sell is the beginning of something extraordinary.

Good reads 📚

  • Marc gives an overview of Central Bank Digital Currency - "It's electronic, it exists as a digital object, and a central bank backs it. He then dives into the politics of CBDC.

  • "As it expands abroad, DC/EP may be used as a mechanism to disintermediate existing global financial infrastructure like SWIFT. The SWIFT system last year facilitated the transfer of $77 trillion."

  • 🤔 My analysis: Marc is spot on (and I had noted the same in BF 23rd August). The dollar being the world's global reserve currency essentially means that the west can control and manage who gets access to this system. While the system isn't particularly effective at preventing criminality, it's quite useful for ceasing political opponents' assets in other countries (e.g., China, Russia, Iran).

  • 🤔 My analysis: China / Africa trade is significantly increasing, China is still a net exporter, and much of this trade is happening in Wechat, Ali, or even JD.com.  A Chinese CBDC that works globally for international trade would reduce China's reliance on the dollar. In that world, China's internal risk is that the big tech companies like Ali have too much control (see: Ali's IPO halted after Jack Ma has been quite outspoken about the global financial system). Control the currency, and the rails, control trade.

  •  🤔 My analysis: Watching all of this, Europe also has an opportunity to assert the Euro's role in the global economy. That said, it's taken Europe more than a decade to get the integration of its payment systems working (SEPA and TARGET 2), and it doesn't always work. China's advantage is execution.  DC/EP is scaling, and it works.

Tweets of the Week 

Recommend following the thread from Oscar here if you’re a stablecoin fan ☝

That's all, folks. 👋