> ## Content Index
> Fetch the complete content index at: https://thepaymentsoperator.com/llms.txt
> Use this file to discover other available public pages before exploring further.

# The Fed Mapped the Future of Payments. The Part That Is Broken Was Not on the Map.
- URL: https://thepaymentsoperator.com/the-fed-mapped-the-future-of-payments-the-part-that-is-broken-was-not-on-the-map/
- Published: 2025-08-25T03:29:00.000Z
- Updated: 2026-11-18T11:22:35.000Z
- Description: A Fed governor told a blockchain conference there was nothing to fear in the technology. He was right. Correspondent banking — the thing that is actually broken for everyone outside the dollar — never came up.
- Author: Armando

*A Federal Reserve governor spent a speech telling a room of blockchain people there was nothing to be afraid of. He was right. The problem is what he never had to mention.*

On 20 August 2025, Christopher Waller of the Federal Reserve Board spoke at the Wyoming Blockchain Symposium in Teton Village. The speech was titled "Technological Advancements in Payments." It was reported, almost everywhere, as a pro-crypto speech by a Fed governor — the "nothing scary" line travelled furthest, and it travelled without its sentence.

I read the speech twice. It is a better document than its coverage, and the most interesting thing in it is not an argument. It is an omission.

## What he actually said

Waller's frame is deliberately unglamorous. Every payment, he argues, consists of three things: an exchange of value, a transaction technology, and a record of the transaction and of ownership. Crypto transactions have the same three parts. Therefore new technology should not be feared merely for being new. Distributed ledgers are, in his words, "simply new technology to transfer objects and record transactions," and there is "nothing scary about this."

That is a reduction, and reductions are how a central banker says something is permitted. The subtext of a Fed governor telling a blockchain conference that their technology is ordinary is not an endorsement of the technology. It is a refusal to treat it as a category requiring its own fear.

On stablecoins, he is more specific. They began as a tool for crypto trading, he says, but they have uses beyond it — among them giving people in high-inflation countries access to US dollars. He believes they could maintain and extend the dollar's international role and improve retail and cross-border payments. The GENIUS Act, which he describes as the first major crypto-asset legislation to become law, is "an important step."

And then a line that is easy to read past: stablecoins, he notes, are a private-sector innovation that still connects to legacy payment services for funding and redemption.

Hold on to that sentence. It is the whole essay.

On artificial intelligence he is briskly deflationary. AI is an enabling technology, not payments infrastructure; payment firms have used machine learning for fraud detection and forecasting since the early 1990s. Large language models improve fraud systems, compliance and file reconciliation. Agentic AI — systems that plan and execute multistep processes — he calls the apparent next wave.

That is the line in the speech most likely to age fastest, and it is worth marking today rather than later. A governor of the Federal Reserve has named agentic AI as the next wave while most of this industry is still arguing about whether it is a wave at all. If he is right, the question that follows is the one the speech does not reach: when software plans and executes a multistep process that ends in a payment, who is the party that authorised it, and who pays when the process was executed correctly and the outcome was still wrong?

## The sentence almost nobody quoted

The most consequential passage in the speech is about who should build.

Private-sector innovation, Waller says, should be the default model. Platforms built by the Federal Reserve itself should be used only rarely — where there is a specific market need that the private sector has not met. The Fed's job, in his telling, is to provide core clearing and settlement, to convene the industry around standards, to conduct technical research, and to engage with innovators.

That is not a crypto position. It is a doctrine about the ownership of payment infrastructure, and it is the single most important thing in the speech for anyone outside the United States, because the rest of the world has been running the opposite experiment and publishing the results.

## The other experiment

In Brazil, the central bank did not convene the industry around a standard. It built the rails, mandated participation by the largest institutions, made the basic service free for individuals, and shipped. In India, a public utility company owned by the banking system did substantially the same thing. In the European Union, the regulator eventually stopped waiting and legislated instant transfers into existence.

In the United States, the Federal Reserve waited, and the waiting produced the payments system Americans actually have: a card duopoly funded by interchange, an ACH that moves in batches, cheques that still clear in meaningful volume, and a real-time service that arrived in 2023, long after a dozen smaller countries had one.

And then there is México, which is the case I know best and the one that complicates the story.

Banco de México built SPEI and has operated it since 2004 — a real-time interbank transfer system, run by the central bank, years before most of the countries now celebrated for instant payments had anything. By Waller's taxonomy, México did the thing the Fed says should be rare, and did it early. Banxico then went further and built the layers above the rail: CoDi, a QR-based payment request standard, and later DiMo, which lets a transfer be addressed to a telephone number instead of an eighteen-digit account identifier.

If building the rail were sufficient, México would be the global case study. It is not, and every Mexican in payments knows why.

SPEI works. SPEI has always worked. What SPEI did not have for most of its life was a reason for an ordinary person to use it, because the banks that connected to it had no commercial interest in making transfers easy, free and prominent, and nobody made them. Brazil's central bank mandated participation and set the terms of presentation. Banxico built the rail and then asked the banking system to please carry people onto it.

That is the lesson I would put to Waller, and it cuts against both of us. Building the rail is necessary and not sufficient. Convening is sufficient for nothing. The variable that actually determined outcomes in the last decade was neither ownership nor convening — it was whether the regulator was willing to compel distribution. Brazil was. México, for a long time, was not. The United States has not been asked.

I want to be careful not to overclaim the general case. The United States did not get a worse payments system because the Fed was passive. It got a different one, and the private rails it built are, in several respects, extraordinary — the card networks solved cross-border acceptance, dispute rights and risk-based authorisation at a scale no public scheme has matched. A Brazilian can pay a Brazilian instantly. A Brazilian cannot pay a Mexican instantly. The card networks solved the second problem and charge for it. Nobody has solved it for free.

But the doctrine has a cost, and the cost is distributional. When the state builds the rail, the price of basic movement of money tends toward zero and competition moves upward, to the products that sit on top. When the private sector builds it, the price of basic movement of money is a margin somebody earns, and that margin is paid disproportionately by the people and the merchants with the least negotiating power.

Waller is not unaware of this. He notes that the card networks, though privately built, rely on Fed-operated settlement for their interbank obligations — which is to say the private rail runs on a public foundation and always has. The doctrine of rare intervention is therefore less a principle than a judgment about where the line falls. Reasonable people in reasonable central banks have drawn it in different places and got different countries.

## The omission

Here is the thing I cannot stop noticing.

It is a speech about technological advancement in payments, delivered by a governor of the institution that sits at the centre of the dollar system, and correspondent banking never comes up.

Stablecoins are mentioned as having potential to improve cross-border payments. That is the extent of it. There is no diagnosis of why cross-border payments need improving — no account of the correspondent chain, of the hops, of the relationships that have been withdrawn from entire countries over the last fifteen years because the compliance cost of serving a small market exceeded the revenue from serving it.

I do not read this as carelessness. I read it as the accurate reflection of a vantage point. From inside the dollar, the correspondent network is plumbing that works. You do not think about the pipes in your own house. You think about them in somebody else's house, when you are standing in it and the water is brown.

From México, cross-border is not a feature of payments. It is the payments question. The United States–México corridor is among the largest remittance flows on earth, and the cost of moving money along it has been a subject of policy attention for twenty years without ever becoming cheap enough to stop being a subject of policy attention. Behind that cost sits the correspondent chain, the compliance burden on it, and the steady withdrawal of dollar clearing relationships from institutions judged too small or too risky to be worth serving.

And this is exactly where that quiet sentence about stablecoins comes back. They are a private-sector innovation that still connects to legacy payment services for funding and redemption. Which means the instrument offered as the solution to the correspondent problem still has to enter and exit through the correspondent problem. You can move a tokenised dollar from Mexico City to Lagos in seconds. Getting the dollars in at one end and the pesos out at the other runs through the same banks, the same relationships, the same withdrawal decisions.

The middle got faster. The ends did not move.

## From a Mexico City desk

México is a useful place from which to read a Fed speech, because we are neither inside the dollar nor independent of it. We have our own currency, our own central bank, and a monetary policy that is genuinely ours. We also have an economy whose largest trading partner, largest remittance source and effective unit of account for anything imported is the United States. The peso is sovereign. The dollar is the weather.

So a doctrine that says the private sector should build payment infrastructure by default reads differently here than it does in Teton Village. For domestic peso payments, México ignored the doctrine thirty years ago and was right to. Banxico built SPEI. The question of whether the central bank should own the rail was settled here before the Fed started asking it.

For anything touching dollars, the doctrine applies to us whether we accept it or not, because the dollar infrastructure we depend on is not ours to instrument. Mexican institutions do not get to build a dollar rail. We rent access to one, from American banks, under American rules, subject to American risk appetite. That is the asymmetry a Fed governor has no particular reason to think about and that we have no ability to stop thinking about.

The practical consequence for anyone running payments in this region: do not wait for the rail. Waller has told you, in the politest possible language, that the institution at the centre of the dollar intends to build rarely and convene often. Convening is not a settlement guarantee. Plan for a world in which the dollar infrastructure you depend on is owned by companies, priced by companies, and withdrawn by companies.

And the second consequence, which is the one I would put in front of a Mexican board: SPEI's history says that the hardest part of a payment system is not the technology and not the ownership. It is distribution. Whoever is designing the next thing — a stablecoin corridor, an agentic checkout, a tokenised settlement layer — should spend less time on the rail and more time on the question of who will be compelled, incentivised or shamed into carrying customers onto it. We have a working national instant payments system that spent fifteen years underused. The rail is never the bottleneck.

## My judgment

The speech is right about the technology and silent about the geography, and the silence is the more important half.

Waller's reduction — value, transaction technology, record — is correct and useful, and I wish more payments executives in this region would adopt it, because it cuts through an enormous amount of vendor theatre. If a distributed ledger is just another way to record ownership, then the question about any blockchain proposal is the same question you ask of any system: what does it record, who can change the record, and what happens when it is wrong. That is a procurement question, not a philosophical one.

But "nothing to be afraid of" is a statement about technology risk, and the thing people in emerging markets are afraid of is not technology risk. It is dependency risk. A tokenised dollar does not reduce a Mexican institution's dependence on the United States. It increases it, elegantly, by replacing a chain of correspondent relationships with a direct claim on a US issuer under US rules. That may well be an improvement — fewer hops, cheaper, faster, more transparent. It is not independence, and the two keep getting conflated by people selling stablecoin rails into this region.

My second judgment is the SPEI lesson generalised, and I hold it firmly: the ownership debate is largely a distraction. Public rails that nobody is compelled to distribute lose to private rails with a sales force. Private rails without a public settlement floor do not exist. What separates Brazil from México is not that one central bank built and the other did not — both built. It is that one of them was willing to tell the banks how the product would be presented to customers, and the other was not.

The third is uncomfortable and I will state it anyway. For cross-border, the dollar is the shared sovereign, and that is why a tokenised dollar is plausible as a global rail where every public cross-border project has failed. Cross-border rails require a shared sovereign and there is only one on offer. Those of us who depend on the dollar without issuing it should therefore read every speech like this one twice — once for what it says, and once for the map it is drawn on.

## Three questions for the people who own this

**1.** If your cross-border strategy depends on stablecoins, where exactly do the dollars enter and the local currency exit, and who owns those two relationships? If the answer is a correspondent bank, what did you actually change?

**2.** Name the three pieces of dollar infrastructure your institution depends on that could be withdrawn by a decision taken outside your country. Who reviews that list, and when was it last reviewed?

**3.** For whatever new rail you are building or buying: who will be compelled or paid to put customers on it? If the answer is "it will be obviously better," you have not answered the question, and SPEI is sitting there as the proof.

Reply to this email with a line or two. The sharpest answers will shape the next issue.

## Sources

**Christopher J. Waller, Federal Reserve Board, 20 August 2025** — "Technological Advancements in Payments," delivered at the Wyoming Blockchain Symposium 2025, Teton Village, Wyoming. Source for everything attributed to him here: the three-part framing of a payment as an exchange of value, a transaction technology and a record; that distributed ledgers are "simply new technology to transfer objects and record transactions" and that there is "nothing scary about this"; that private-sector innovation should be the default model and Federal Reserve-built platforms used only rarely, where there is a specific market need; that the Fed's role is core clearing and settlement, convening the industry around standards, technical research, and engagement with innovators; that card networks, though privately built, rely on Fed-operated settlement for interbank obligations; that stablecoins began as a crypto-trading tool but have other uses including dollar access in high-inflation countries, could maintain and extend the dollar's international role, and still connect to legacy payment services for funding and redemption; that the GENIUS Act is "an important step"; and that AI is an enabling technology rather than payments infrastructure, with agentic AI the apparent next wave. Waller stated the views were his own and not necessarily those of his Board colleagues.

**On what the speech does not say.** I have checked: correspondent banking is not discussed. The only cross-border point in the text is that stablecoins have the potential to improve cross-border payments. The argument I build on that absence is mine, and it is an argument about vantage point, not an accusation of error.

**On the comparisons.** The characterisations of Brazil, India, the European Union, the United States and México are structural, not statistical. I have deliberately cited no adoption figures, transaction volumes, remittance totals or cost comparisons, because I did not open a primary source for any of them while writing this and I will not publish numbers I have not read. What I assert is verifiable in each system's own public design documents: that Brazil's instant rail is operated by its central bank with mandated participation and free basic use for individuals; that India's is run by a bank-owned public utility; that the European Union legislated instant transfers rather than waiting; that the United States' real-time service arrived in 2023; and that Banco de México has operated SPEI since 2004 and subsequently introduced CoDi and DiMo as layers above it. My characterisation of SPEI's adoption history is a judgment formed from working in Mexican payments, not a citation. Anyone who wants the numbers should go to the central banks for them, as I will before I write about the comparison at length.

**A note on what is mine.** The reading that the speech's doctrine of rare intervention is its most consequential passage; that the real variable in the last decade was willingness to compel distribution rather than willingness to build; that stablecoins increase rather than reduce dollar dependency for emerging markets; and that cross-border rails require a shared sovereign and the dollar is the only one on offer — those are arguments, not findings, and I expect serious people to disagree with at least two of them. Nothing here reflects the position, portfolio, clients or vendors of my employer.