The ECB's Pontes launch is a digital euro for banks, not wallets

The first thing I would get straight about the European Central Bank's Pontes launch is this: nobody in Europe has a digital euro in their pocket. What went live on September 21, 2026 is a wholesale system built for banks. The retail central bank digital currency, the kind that could end up in ordinary wallets, is a separate project, and the rules being written for it now are the ones that will govern it later.

What the ECB launched with Pontes

Christine Lagarde, the ECB's president, gave the date: Monday, September 21, 2026. Pontes, in plain terms, is a digital euro made available to banks so they can transact among themselves using tokenized assets and distributed ledger technology. It sits on the institutional side of the payment system, where banks and other financial firms settle with each other.

Maria Zeee of Zeee Media draws the line at that boundary: "The retail piece is coming later."

Pontes means bridge, and the system is meant to connect institutions that today settle through a mix of central bank reserves, messaging networks and correspondent relationships. Lagarde's pitch is speed and less friction, a change she said the banks have to make.

Households get no wallet and no app for buying bread out of this. It is a set of rails for moving value between institutions, and the ECB is developing the consumer version separately. If you saw a headline saying the digital euro had arrived, read it again with the word wholesale in mind.

Central bank digital currency, in plain words

A central bank digital currency is a digital form of a country's money issued directly by its central bank. The balance in your bank account is a promise from a commercial bank; a CBDC is a claim on the central bank itself, which is why supporters call it the digital equivalent of cash. Retail CBDCs are built for households and businesses. Wholesale CBDCs, the kind Pontes belongs to, are built for institutions settling large payments among themselves.

A tokenized asset is a representation of a bond or a deposit recorded on a shared ledger, so ownership and payment can move together. Distributed ledger technology is a record of transactions kept in sync across many computers instead of one central database, so no single party holds the only copy.

Put the two together and you get settlement in one step, sometimes called atomic settlement: the asset and the payment change hands at the same moment instead of moving through a chain of intermediaries over a day or two. That is the efficiency argument Lagarde makes, and it is why central banks have tested the idea for years.

For a reader who will never touch Pontes, the design choices still matter. Holding limits, who can see a transaction and what a payment can be programmed to do are settled while the rails are being laid. Rules written into code at that stage tend to survive the move to consumer products.

Why banks get the rail first and people get the wallet later

The ECB's public position, as Zeee describes it, is that a retail digital euro would be complementary. People would keep ordinary accounts at private banks, cash would remain available, and a CBDC account would sit alongside. The bank says it does not want to cause a run on deposits, so it plans a cap on how many digital euros one person can hold.

The cap is the detail I keep coming back to. The ECB would set the per-person limit inside an overall ceiling agreed through the EU's political process, which makes a political decision look technical. It will be fought over.

Timing matters too. Zeee says the retail pilot is targeted for 2027, with possible issuance around 2029. Pilots have run in Australia and elsewhere since the COVID years, she notes, and the projects she follows generally aim for issuance before 2030.

You can read the two-track design two ways. Zeee sees a staging post: running a CBDC system beside private banking is expensive, and the only reason to pay for both is a plan to move people across later, with Know Your Customer rules as the lever. That argument rests on motive, and it assumes a rule written for one purpose cannot stay written for that purpose. The holding cap points the other way, since a limit on balances is a strange thing to build if your goal is to pull every deposit into a central bank account.

The limits and the numbers behind them

Zeee puts the working range for a per-person holding limit at €500 to €3,000. She describes it as the range usually used in official analysis and in past proposals, not a number the ECB has fixed. No final figure has been announced.

Two of those rows need care. The holding range comes from analysis and earlier proposals, not from a final rule, and the basic income figures are academic models, not a program in force. There is no EU-wide universal basic income, as Zeee says herself.

Universal basic income, usually shortened to UBI, is a payment made to everyone regardless of means, and the numbers enter the story because of how Zeee connects them to the cap. A 2026 European Citizens' Initiative called for it to be at least the national poverty line, defined as 60% of median income in each country. Academic EU models often land between €570 and €860 a month.

Zeee uses those figures to sketch what a capped CBDC could mean if a basic income were ever paid through one. That scenario is worth keeping separate from announced policy: it is built from separate pieces, and the ECB has not proposed replacing wages or savings with a capped digital balance. Check the central bank's published documents before treating any of these figures as settled.

Zeee's warning about programmable money

Her forecast goes further than any published design. She describes a sequence: a basic income paid only in CBDC, private money squeezed or collapsing, and a capped, programmable account left as the only option. In that scenario the monthly ceiling would sit around €3,000, the balance would carry an expiry, and the code would block certain purchases. "It is going to have expiration dates," she says, adding that the money would have to be used by the end of the month.

The restrictions she lists are specific. The programming could stop people from buying meat or dairy, and travel would be limited to one short-haul flight per person every three years, capped at 1,500 km. She links those restrictions to statements she attributes to the World Economic Forum about meat, dairy and ammunition, and to what Catherine Austin Fitts calls the third lock, the point at which the system can cut a person off in real time instead of through a bank branch or a court order.

I think it helps to separate the machinery from the forecast. A CBDC balance is a record on a ledger, a record can carry conditions, and a condition can be applied instantly at scale. Zeee says she is not a technical expert, and her case rests on incentives rather than on code. The forecast is hers, assembled from pilots, past proposals and the logic of programmable money.

Her case is strongest on the direction of travel. A holding cap, a political ceiling and a parallel system all keep the option of restriction open, whatever the intentions of the people writing the rules. It is weaker on the timetable. Payment habits, bank balance sheets and election cycles move slowly, and a visible cap on savings would be a political problem in any democracy. The scenario also needs several things to go wrong at once, including a collapse of private money that no central bank has said it wants.

Cash, stablecoins and the American version

The ECB says cash stays. Zeee is not convinced. She points to the COVID years, when officials described banknotes as dirty and a virus risk, and treats that as a template for future excuses to make cash awkward to use. She agrees with Catherine Austin Fitts that more people are moving toward cash, buying direct from producers and supporting local businesses, though she doubts the trend is strong enough to push back in cities. A farmer she spoke to in Central Texas accepts payment in gold and silver.

Then there is the United States. Michael Selig, chairman of the Commodity Futures Trading Commission, said US financial markets need to prepare for mass tokenization and 24/7 on-chain finance. Zeee reads that as the American version of the same shift, with a different corporate shape.

Her argument there runs through public-private partnerships. During the COVID years, she notes, the government could not switch off a person's social media account itself, but it could ask a platform to do it. She says the Genius Act and the Clarity Act, the two US laws covering stablecoins, let the government reach programmable digital money through private issuers in the same way.

For readers who have not used one, a stablecoin is a token issued by a company and usually backed by reserves, designed to hold a steady value against a currency such as the dollar. It is a claim on the issuer, not on a central bank, and that difference matters here. Zeee's point is that the practical effect for a user can land close to a CBDC: the issuer decides who can hold it, where it can be spent and whether an address gets frozen, under rules the government writes.

That reading is contested, and the two do not become the same thing because they share features. Stablecoins sit inside the banking and securities rules of the country that regulates them, and their issuers can fail. Her broader question still lands: if a private token can be programmed and blocked, the absence of a government-issued digital dollar changes less than it seems.

She also dismisses the US executive order against a CBDC, saying such orders are not worth much once a presidency ends.

Ukraine, Diia and the test-bed argument

Ukraine was the most concrete part of the conversation. Zeee describes the Diia app, which holds digital identity and payments in one place and, in her account, was supported by USAID. She calls the country a test bed for a single app that carries money, identity and a channel for reporting on neighbours.

Her example goes further. She says United24, the official fundraising platform, released a promotional video about Ukraine becoming the first fully AI-run country after reconstruction, with an AI judiciary and an AI government, and that the video was deleted after she reported on it. She names Uber among the backers of the rebuilding effort.

I cannot verify a deleted video from the outside, and one promotional clip is thin evidence for a plan. What makes the argument worth taking seriously is that wartime settings do produce digital identity and payment experiments that outlast the emergency. Once a population is used to holding identity and money in one app, the app becomes the obvious place for anything a government wants to attach to either.

What to watch, and what I would check first

Wholesale is the word to hold onto. Every headline about the digital euro between now and the retail launch should be read with that word inserted, and if it does not fit, the headline is probably ahead of the facts.

The number to remember is the holding limit. When the ECB settles on a per-person cap, that figure will tell you more about the design than any launch event. A generous limit points to a payment tool that sits beside bank deposits. A tight one points to a system built to keep money where the banks can lend it.

Before relying on any claim in this area, read the primary document from the central bank or the regulator instead of the summary, because pilot, issuance and launch mean different things and get mixed up in coverage. Watch what the legislation says about programmability, privacy and who can freeze an address, in both the EU and the US, since those clauses outlast the press release. And keep some capacity to pay outside the digital system, whether that is cash, a local supplier who takes it or knowing which of your bills can only be paid online. You do not need a theory about a global plot to notice when a payment option disappears.

If you do end up with a stablecoin or a CBDC wallet, check the site address before you connect anything and start with an amount you can afford to lose. The pilot Zeee expects in 2027 is where the per-person cap will stop being a range and start being a number.