Philosophy on Payments | Part 5
A look forward into the grey malaise of what might be and the hopeful possibility of what ought to be.
Moving money is moving information.
To trust a money moving (i.e. payment) system, one must trust the information.
Trusting information requires agreed upon standards.
Standardizing is the key to commerce.
Standards require an authority to set and enforce them.
Authorities must balance self-organization and top-down management.
Those are the building blocks of my philosophy on payments, which I’ll be releasing soon. They’re an amalgamation of ideas borrowed from much smarter people, my study of payments history, and many, many walks with my patient wife.
I don’t claim to know the future, and what I lay out below is informed by research and observation. If I’ve said something wrong or that you don’t agree with, let’s have a conversation. The world only improves with dialogue and consensus, not monologues and isolation. Thank you!
The World as It Might Become
Today’s payment infrastructure, in the U.S. at least, is dated. With a few notable exceptions — namely P2P transfers and RTP + FedNOW — structural change has been glacial over the last 50 years. We still rely on ACH (est. 1973) for most of our payroll1 and electronic card networks (est. 19582) for most of our daily transactions.
These innovations — they really were innovations at the time — have become mundane with age. Rather than continuing to innovate, however, the mechanism of change has been gummed up by entrenched incumbents with incentives to maintain status quo. It’s why there exists an endless number of wrappers, orchestrators, and fintechs, but virtually no new networks.3 It’s these networks upon which our primary institutions for the managing and moving of money are built.4
A quick note: there has been a tremendous amount of innovation in the messaging and authorization methods surrounding electronic payments over the last 50 years. What has not changed much are the settlement methods which complete the transaction. That is my primary point of contention in this post.
Compounding the glacial innovation in settlement, ways in which they generate revenue is also based on these legacy networks, creating a strong incentive to be slow or outright hostile to change. This has been the case with adoption of faster payment methods.
There’s a whole generation of financial institutions and fintechs which are built on interchange revenue. Many of them are so reliant on it, that a world where people can make cheap, secure, instant payments without a card network is seen as an existential threat. So, the incumbents push back on changes to the established order as a matter of self-preservation.
None of this is to say that some of the services these organizations provide isn’t valuable. For example, the restaurant industry has benefitted from better software and easier payment acceptance.5 The issue is that in order to meet the “line go up” demands of the market, incumbent businesses far more often rely on returning the ‘cash cows’ instead of investing in the ‘question marks’.
Another serious issue for adoption of new methods is who’s in control of a network.
In the case of RTP, it is a consortium of the country’s largest banks. This disincentivizes smaller financial institutions from participating because they feel pressured. TCH and its owner banks also discouraged and stymied development of FedNOW.6 A similar story can be found with the wide-spread adoption of Zelle from Early Warning System (another, more concentrated big bank consortium).
As a result, change is likely to be glacial. Factors like the replacement cost — real and perceived — of legacy infrastructure and incumbent interests mean that, barring sweeping top-down regulations and reforms or customer agitation, we’re unlikely to see a system like PIX or UPI take off in America.
In the international realm, America faces all these problems plus one more: the US Dollar’s role as the world’s reserve currency. To ensure stability and maintain that status, changes to how USD moves within and outside the US must be measured and well-telegraphed to the rest of the world.7
The future of payments — over the next decade or so — is likely to look similar to today. The themes of increased focus on customer experience and finding new use cases for existing technology are going to be prevalent, while the uptake in new settlement methods will likely be muted outside specific use cases.
On the back-end:
Some Wire and ACH volumes will shift to RTP & FedNOW. I wouldn’t be shocked if the high fees charged stay the same, particularly for wire transfers, even as the funds flow across cheaper rails. There’s a chance that the Fed draws on the ACH launch playbook and just mandates participation in FedNOW for government payouts.
Broad ISO 20022 adoption will decrease error rates and improve interoperability, but only if the systems generating and ingesting the data are set up to use it. A lot of systems — banking, treasury, and ERP — will require upgrades to make the most of the messaging standard.
Digital currencies transfers will continue seeing small-scale adoption, but given the speculative nature of most coins, it is unlikely to really take off as a payment rail trusted by major institutions. The notable exception here are stablecoins which rely on a peg to an existing currency. USDC and USDT could really take off if financial institutions and fintechs can carefully thread the regulatory needle.8
AI for back-office tasks is going to be the next wave of fintechs.9 Certain tasks, such as data ingestion and matching, are really well suited to this, but I can see several situations where businesses try to build all their back-office payments tasks with AI that causes major issues. Just like other tools in accounting, AI tools will be sticky — really sticky. Once implemented, these stay in place long after a better option has become available.
On the front-end:
Virtual card issuing has been picking up steam over the last 5 years alongside the rise of digital wallet adoption like Apple Pay and Google Pay. These factors will lead to some changes, for example, in how we travel for work and receive healthcare benefits. Having the ability to immediately give a debit card to people might side-step the need for a traditional bank account.10
Push-to-card has begun to see some great use cases in the insurance industry. It saves the need to mail checks, gives faster access to funds, and provides a superior customer experience. Alongside the virtual card issuing, this could make access to the financial system much easier for underserved communities and (hopefully) eliminate the need for predatory institutions.
Biometric and other multi-factor authorizations are going to pick up steam in the coming years. It’s already taking place with the rise of TFA for suspicious transactions and Face ID for unlocking mobile wallets & accessing bank apps. The Orb, widely panned for its dystopian appearance, name, and purpose, is an example of how biometrics could be used to reduce the need for carrying traditional cards or currency notes.
Because of the benefits of float for payers and financial institutions, I expect most of the real-time payouts to instead be tied to some sort of short-term lending product. A prescient example is the push from labor for earned wage access and daily payroll instead of relying on payday lending. The fees, and there will be fees, for this type of solution might come from the payer, payee, or both — depends on the situation.
The world as it ought to be
When we think about how our current networks and systems came into being, it was during a period of global11 collaboration and growth. Technology was advancing at a rapid pace alongside global trade, and so problems were being solved with the latest innovations of the day…until it slowed down.
There were fits and spurts around the rise of the internet — particularly in online bill-pay and card processing. Most of these innovations were small tweaks to existing methods — WEB ACH transaction types and improved Card-Not-Present acquiring and authorization — versus systemic change (i.e. FedACH and other novel ways of settling transactions).
We’re in a similar era of technological and societal change as the mid-century. The world is more connected than ever and we’re seeing that legacy payment methods and institutions have trouble keeping up without an increasing number of middlemen, orchestration software providers, and aggregators. The issue here is the more layers present in the payments cake, the more prone it is to toppling over.
To simplify the movement of funds, it’s imperative that new settlement systems rely on modern standards (e.g. ISO 20022), be “always on”, and limit (not eliminate) ownership share of incumbents or those who may want to oppose further change down the line. Access to this system should be easy and not tied to any one institution, which will allow people to “vote with their wallets”.
On the back-end
Stablecoins and CBDCs are introduced and widely adopted. Stablecoins and CBDCs, Central Bank Digital Currencies, offer a method of exchanging value that is inherently detached from traditional systems. They have the benefits of security and can be integrated into the digital technologies all around us. For those skeptical about crypto, both methods still rely on a central bank: the currency to which the stablecoin is pegged or the central bank’s direct involvement.
A common, globally accepted messaging standard will remove frictions from international commerce and be a net benefit for everyone. ISO 20022 is a step in the right direction and will have an impact on global trade as information can be moved seamlessly across currencies and payment rails. The increased visibility into how funds are flowing — the path as well as the costs associated with them — will result in lower costs and more robust networks. The largest impact of such a unified system will be in how emerging markets are able to participate in the global economy.
Less reliance on traditional financial institutions via open banking. Payments tell part of the life story for people and businesses. Switching financial institutions ought not mean that some of the story is lost. This creates the incentive for financial institutions to innovate in the tools and services they offer and compete on service rather than convenience.
More collaboration and participation of payers, payees, financial institutions, and central banks. The original organization of Visa and SWIFT were amazing examples of stakeholders, no matter their relative position, ceding control to a central authority and then empowering authority to do what’s best for the whole rather than any single stakeholder. Decentralized financial networks are a step in the right direction, but I think they lack the balance that comes from institutional buy-in.
Reduced reliance on float. Payments companies today, that is to say those in a high interest rate environment, benefit immensely from the float that occurs because of the limitations in existing payment methods. As more transactions become real-time transfers, the delay in settlement and thus the float, will be removed. Financial institutions and fintechs will adapt to the new environment through focusing on value-add services (the thing they should be doing).
On the front-end
Connected finance and uniform standards will lead to smoother experiences for payers and payees, alike. Users will be able to access all their financial information in the environment of their choosing. This will lead to better insights and incredible innovations in financial management and access.
Heightened focus on analog security and authorization. The never-ending arms race between fraudsters and security professionals is not going away even in the world as it ought to be. Acquiring electronic data is (relatively) easy given how frequently and widely its used in the world, so a system of analog + biometric authorization methods may prove to be the most secure. Mastercard’s biometrics payment card that requires a thumbprint to become active is an excellent example of where things ought to be heading.
Multiple financial institution and fintech relationships. As information is more easily transferred and shared among financial service providers, people and business will be able to shop around for the best provider of each service. This á la carte approach will let niches and specializations form, producing better products and services for all.
Wrapping this up
Payments, at an individual level, tell a crucial part of the story of people’s lives — the mundane coffee purchases, major business decisions, or simply getting a paycheck12. In aggregate, the +204,500,000,000 messages sent in 2021 formed the frame of the U.S. economy.13
“Any sufficiently advanced technology is indistinguishable from magic” — Arthur C. Clarke, Profiles of the future: an inquiry into the limits of the possible
Today’s innovations like ApplePay may as well be magic to you and I. Few have a background in the 50+ years of innovations in electronic communication, network processing, and encryption which have occurred since the first credit card swipe in the late 1960s. As new tools like AI and distributed ledger technology get added to the mix, the well from which one must drink to gain comprehension only grows deeper.14
The magic tarnishes and becomes mundane with time and exposure, until you’re left with less of an innovation and more a piece of background infrastructure. It’s important, yes, but only insofar as it powers the technology resting upon it. As we “stand on the shoulders of giants”15 from the past, we must keep in mind that our shoulders are the ones upon which other will one day rest.
The technologies of the last 50 years have a place in our future, but in my view not as the primary means by which we conduct payments. The technology of the past should serve as the steppingstones — models for what works and what doesn’t work.
Just as the standardized shipping container didn’t get rid of pallets, crates, and boxes, the introduction of new payment systems won’t get rid of existing ones. The old will be integrated into the new and work in concert. That’s what building better systems, not just fancy wrappers will get you.
To get there, we have to have collaboration, not isolation.
Altruism, not selfishness.
Trust, not fear.
Get caught up on the previous posts:
Part 1 — Introduction
Part 2 — Clay Tablets to Checks
Part 3 — Checks to ACH
Part 4 — ACH to FedNOW
Fun fact: the reason banks can advertise “get payday two days earlier” is because most banks require an ACH file to be sent at least two days in advance of the settlement date. This dates back to a Fed rule from the 1970s as it was the longest amount of time it would take to correspond with a bank that didn’t use electronic messaging. They eliminated this rule in 1993, but the convention has stuck for the last 30+ years regardless.
To be a stickler, 1973 was the first time that the cards were processed electronically — without manual keying. I wrote about National Data Corporation who helped pioneer the keying technology and were early to the scene when processing the electronic POS data.
Stearns, David L. (2011). Electronic Value Exchange: Origins of the Visa Electronic Payment System. London: Springer. p. 89
There are a host of reasons starting a network is so difficult. Much of it comes down to solving what I previously called the “Payments Trilemma” of trust, acceptance, and speed.
On further examination, I don’t think speed was the right choice, but this is my public learning exercise so ¯\_(ツ)_/¯
The wildest thing to me is that banks still rely on COBOL, a +60-year-old, unstructured computer language. COBOL is still being used by 40% of banks in the US. That’s terrifying to me and represents a very large, very quiet systemic risk. All it takes is one error, one line of accidentally deleted code, for a bank’s core system to break down.
Yves Smith explores the potential systemic issues and why we don’t switch off of it in this 2016 piece.
I regularly chat with the staff about server and restaurant operator opinions around systems like Toast or Square. The staff loves it because of the simplicity and how easily they can enter, change, and close out order. Operators enjoy the ERP & payroll management services, though the common complaint of payment processing costs (fee + spread) does come up.
Alex Johnson has a much more in-depth and well-researched take on this. In particular his The Past, Present, & Future of Big Bank Consortiums and Big Banks’ Next Big Bet were really informative for me.
There’s a much deeper conversation to be had about the recent weaponization of SWIFT and the US dollar with respect to sanctions. In a way, the US Federal Reserve and Treasury have become another branch of the US military. In the same way we can wage physical war, we now actively participate in economic wars with countries like China, Russia, and Iran.
It’s a bit of a worry for me in my current position at a company whose raison d’être is facilitating international finance. Long term, I think treating the USD like this will be a net negative for the USA and a bit of a wash for the rest of the world.
Governments have an incentive to control the flow of their currency in order to regulate their domestic economy. USDT and USDC, two of the most popular USD stablecoins, are already being used as a store of value in places with high inflation or volatile currencies like Argentina (it’s in Spanish, you use google translate or learn it). Some people and businesses are even using these stablecoins to conduct domestic trade just like people have been doing for thousands of years with other strong foreign currencies.
The local governments have a few paths they can go: Some may might elect to be dovish and allow commerce to be conducted in this asset so long as taxes are properly assessed and paid. Others, fearing a loss of economic control, may take the hawkish route and ban the use of any stablecoins under threat of legal/criminal action.
In economic terms, the purchases of USDC and USDT amount to foreign outflows and serve to prop up the value of US Dollar. From a US regulatory standpoint: inflows are fine and large outflows will mean intervention. The use of USD based stablecoins means that the Dollar maintains its status as the de facto world currency and with everything being on public chains, the US can monitor flows for misuse or sanctioned individuals.
The BaaS and “we make money from debit-card interchange” era of Fintechs is likely heading towards a regulatory winter. Recent legislative proposals, legal rulings, and blowups reinforce this idea.
It has the added benefit of keeping people in the card ecosystem, further entrenching the existing infrastructure.
“Global” here is a bit of a stretch depending on your view of history. It was really NATO vs. Warsaw Pact. There were competing standards for almost everything. In the case of payment standards, the US really led the way because it was the dominant economic power of the era.
One of the largest uses of the value, however, was for direct deposit payroll (+18%). See the NACHA ACH statistics for more: https://www.nacha.org/content/ach-network-volume-and-value-statistics
I wrote the full length because it is an incomprehensibly large number. And that’s just electronic payments, it doesn’t consider the cash transactions which also occur.
Source: https://www.federalreserve.gov/paymentsystems/fr-payments-study.htm
Think of it like the story of Tantalus meets Mimir’s well. Odin must give up more and more body parts to gain understanding until, much like crossing an event horizon, he is wholly consumed and unable to escape.
This specific quote is from a 1675 letter Issac Newton wrote to Robert Hooke:
“What [René Descartes] did was a good step. You [, Robert,] have added much several ways, & especially in taking the colours of thin plates into philosophical consideration. If I have seen further it is by standing on the shoulders of Giants.”
However, the sentiment has been repeated hundreds, if not thousands, of times throughout history. Wikipedia has some examples.