The Digital Euro

A digital euro without a sovereign physical stack is a Position 4 product.

A Midnight in Sofia

Shortly after midnight on 1 January 2026, a forty-three-year-old man named Dimitar walked up to an ATM in central Sofia and withdrew one hundred euros. It was the first physical euro banknote ever issued in Bulgaria, and he smiled at the camera and said, “Great! It works!”

Bulgaria had joined the European Union in 2007 and committed to eventual adoption of the euro as part of the accession agreement. Moving through the convergence criteria took eighteen years of inflation management, fiscal consolidation, and holding the lev’s peg to the euro without deviation.

On 8 July 2025, the Council of the European Union adopted the final three legal acts fixing the conversion rate at 1.95583 leva to one euro and setting 1 January 2026 as the date the currency changed hands. Bulgaria became the twenty-first member of the monetary union.[1]

What Dimitar held in his hand that night was the oldest form of euros, and the only form the European Central Bank issues directly to members of the public.

Physical banknotes are a liability of the Eurosystem to the bearer, which means that when Dimitar took his hundred euros out of the ATM, his claim ran directly to the central bank rather than through the Bulgarian bank whose name was on the machine.

Most Europeans never experience this relationship, because most of the euros Europeans hold are not central bank money at all.

The Shape of European Money

A current account balance at ING or Deutsche Bank or UniCredit is not a euro in the way that a banknote is a euro. It is a claim on a commercial bank, denominated in euros, which the bank backs by holding reserves at the European Central Bank. Commercial banks issue the money that Europeans use for daily transactions.

The central bank issues the banknotes and the reserves, and both sit beneath the commercial money as its anchor.

This two-tier structure is common to every developed economy and has operated in Europe since the euro was introduced. Between five and eight per cent of euros in circulation exist as physical cash, and the remaining ninety-two to ninety-five per cent exists as commercial bank money. The smaller share does more work than its size suggests.

Cash is what makes commercial bank money credible, because a current account balance is a promise that the bank will deliver central bank money on demand, and the credibility of the promise rests on the fact that the delivery happens every day at every ATM in the monetary union.[2]

When cash retreats from daily use, the promise becomes abstract. The legal right to withdraw remains, but the operational reality in which withdrawal is a routine act of settlement erodes, and with it the lived experience of commercial bank money as exchangeable for something more fundamental than itself.

The Right and the Instrument

The European Parliament is, at the same time as it is legislating the digital euro, strengthening the legal status of physical cash. The Single Currency Package clarifies the legal tender obligations on merchants and public institutions, broadens the circumstances in which cash must be accepted, and places a formal duty on member states to ensure adequate access to cash infrastructure.

It is the most substantial codification of cash’s legal protections since the euro entered circulation in 2002.[3]

The codification is happening at the moment cash is disappearing from daily use. The European Central Bank’s Study on the Payment Attitudes of Consumers in the Euro Area records cash at 68 per cent of point-of-sale transactions in 2019, 59 per cent in 2022, and 52 per cent in 2024. The trajectory is continuous and the endpoint is not hard to extrapolate.

In the Netherlands cash accounts for 22 per cent of point-of-sale transactions today, and in Finland 27 per cent. These are where the euro area is heading.[4]

A right to pay in cash does not require that anyone pay in cash. It requires that those who wish to can, and that merchants and institutions do not refuse the legal tender of the monetary union. The legal protection therefore remains meaningful even as use declines, because protection of a minority practice is precisely what legal tender rules are for. The difficulty is that cash carries a bundle of properties that the legal right alone cannot preserve once the instrument itself has retreated from daily use. A right protected in statute does not automatically survive the withering of the infrastructure that makes the right liveable. The merchant who has not handled cash in a year loses the habit of handling it. The bank branch that closes does not reopen.

The ATM that is decommissioned is not replaced.

The digital euro is the proposal that the properties cash carries should be preserved in a second instrument, designed to carry them into the part of the economy where the first instrument can no longer reach. To evaluate that proposal, it helps to be clear about what the properties actually are.

The Properties Cash Carries

A banknote does five distinct things at once, and the combination is what makes cash what it is rather than any one property on its own.

It is public money held directly. When Dimitar takes his hundred euros out of the Sofia ATM, his claim runs to the European Central Bank, without any commercial institution remaining party to the holding. He does not need the cooperation of a bank, a payment processor, or a technology platform to continue holding it. This is the only form of money of which this is true for European citizens.

Cash transactions generate no institutional record. Two people exchanging banknotes produce no log, no report, no entry that sits in the custody of a third party. Either participant may record the transaction voluntarily, and authorities acting through legal process may investigate it, but the transaction itself is private in a way that no digital payment is private by default.

A banknote works when the surrounding infrastructure does not. When the electricity is out, when the internet is down, when a bank’s core systems fail, when a payment network is disrupted for reasons political or technical, cash continues to function. It depends on no third party to operate, which means it keeps working in exactly the conditions in which other forms of money stop.

Anyone can use it. Cash requires no device, no account, no credential, no digital literacy. It is available to the elderly, to children, to people who have lost access to their accounts, to visitors, to the digitally excluded, and to anyone whose relationship with formal financial infrastructure has broken down.

The holder chooses how much to keep. There is no legal ceiling on how much physical cash a person may possess. A citizen may hold a hundred euros or ten thousand, in whatever combination of denominations they prefer, and the instrument serves as a store of value and a means of exchange without distinction between the two functions.

These are the properties the legal tender regime protects by protecting the instrument. They are the properties the digital euro must preserve in digital form if it is to function as cash’s successor rather than a different kind of commercial bank money with a central bank label.

Whether the design before the European Parliament preserves them adequately, and where, on the specific points it does not, the design should be strengthened, is the question that follows.

The five properties are not all of the same kind. Properties two through five describe what the instrument does in operation: it generates no record, it works when infrastructure fails, anyone can use it, the holder chooses how much to keep. Property one describes what the instrument is.

A direct claim on the central bank, with no commercial intermediary in the chain of custody, is the architectural foundation that makes the other four hold structurally rather than procedurally.

Without property one, the operational properties become policy-dependent: an intermediary could be compelled to retain records, an intermediary could fail and bring the instrument down with it, an intermediary could decide who is included, an intermediary could bound the holding. With property one, the operational properties hold by design.

What follows assesses the proposed digital euro design property by property, beginning with the architectural foundation.

The ECB’s Five Goals

The ECB frames the digital euro around five purposes: countering the American card duopoly, countering dollar stablecoins, preserving direct central-bank issuance, protecting payment privacy, and ensuring resilience under infrastructure failure.

The first two and the last two are operational requirements that any successful digital euro must meet, and Papers 11 and 12 of this series document the underlying dependencies in detail.

Only the third purpose, preservation of direct central-bank issuance to citizens, explains why the central bank specifically must be the actor. No commercial entity can issue central-bank money, by definition. If direct issuance is preserved through the digital transition, it will be through a central-bank project. If it is not, it will be because the central bank did not act.

What follows assesses the design property by property, beginning with the architectural foundation.

Property One: Public Money Held Directly

The third sharpening is the hardest.

Circle Internet Financial, issuer of the USDC stablecoin, has on several occasions frozen euro-linked wallets on instruction from the US Office of Foreign Assets Control.

The action is technically voluntary, as Circle’s public statements emphasise, but the company has been consistent that it considers itself legally bound to comply with US sanctions, and the freezing capability is built into the stablecoin smart contracts as a matter of architecture.

This has produced a specific and recurring result: European citizens and institutions, holding digital assets denominated in euros, have had those assets rendered inaccessible by executive action of the American government, without the intervention of any European court, prosecutor, or regulator.[5]

The pattern extended in 2025 to categories of European institution the American political system had not previously chosen to sanction. Executive Order 14203, signed on 6 February 2025, authorised sanctions against International Criminal Court officials involved in investigations of US or allied personnel. The order was initially applied to Chief Prosecutor Karim Khan.

Over the course of the year, it was expanded three times, eventually covering eleven individuals, including four judges of the Court. Judge Beti Hohler of Slovenia, Judge Reine Alapini-Gansou of Benin, Judge Solomy Balungi Bossa of Uganda, and Judge Luz del Carmen Ibáñez Carranza of Peru were sanctioned in the June expansion.

Judge Sharon Prost of the United States, recently retired from the US Court of Appeals for the Federal Circuit and sitting as an ad hoc judge at the ICC, was added in August. Deputy Prosecutors Nazhat Shameem Khan and Mame Mandiaye Niang were added the same month.[6]

What happened to the sanctioned judges illustrates the mechanism with unusual clarity. Judge Prost, resident in Washington, found that her Amazon account stopped responding, her Visa and Mastercard credit cards stopped working at point of sale, and her Kindle library was deleted.

These consequences were not the result of any judicial finding, any criminal proceeding, or any lawful process in the jurisdiction where she lived. They were the automatic consequences of private American compliance infrastructure processing an American executive order. Judge Prost was never accused of a crime. She had been a federal appellate judge appointed by a Republican president.

She had agreed, in retirement, to sit as an ad hoc judge at the ICC. That was the entirety of the public conduct for which her commercial life in her own country was disrupted.[7]

For European judges on the same list, the consequences were the same in kind, with the additional feature that European banks, European card networks, and European payment processors implemented the American sanctions on European citizens living under European jurisdiction.

The sanctioned judges lost access to commercial banking services in the Netherlands, where the Court sits, without any Dutch legal process having authorised the loss.

Euroclear, BNP Paribas, and other major European financial institutions complied with the American list in the absence of any European judicial or regulatory instruction to do so,

because their correspondent banking arrangements, dollar clearing access, and technology infrastructure are sufficiently exposed to the American financial system that non-compliance would expose them to secondary sanctions.

The digital euro enters this environment as a specifically European instrument, issued by a European central bank, denominated in a European currency, settled on European infrastructure, and regulated under European law.

The question the design must answer is what this European character actually means when an American executive order arrives demanding that specific European holders be cut off from their euro balances.

One answer is the libertarian one: that money should be uncensorable in principle, that no authority should have the capacity to freeze anyone’s balance under any circumstances, and that the digital euro should be designed around cryptographic properties that make freezing technically impossible.

This is the position implicit in the original Bitcoin design, and it remains the position of parts of the cryptocurrency movement.

The libertarian position is not the right one for the digital euro, because European citizens are entitled to functioning law enforcement, functioning financial-crime capacity, and functioning judicial process, and a monetary instrument that renders European law enforcement incapable of acting within its own jurisdiction does not strengthen European sovereignty. It weakens it.

The defensible position is narrower. European money should be subject to European law. Holdings of digital euros should be accessible to freezing, seizure, or other constraint by European legal process, operating under European judicial oversight, in response to specific findings of fact by European authorities.

They should not be accessible to freezing, seizure, or other constraint by foreign executive action, operating through private compliance infrastructure, in response to political decisions taken outside any European legal framework.

This is the position already taken, in effect, by the seventy economists of the Piketty letter. The letter’s central question (whether Europeans will control their money in the digital age, or whether others will) does not ask for uncensorable money. It asks for European-censored money, by European legal process, accountable to European democratic institutions.

The ICC cases demonstrate in concrete terms what happens when the second of these conditions is absent. The digital euro is the vehicle through which the second condition can be established for a component of the European monetary system.

In design terms, this has specific implications. The digital euro must not rely on a compliance architecture that operates through private American intermediaries. It must not use cryptographic primitives, tokenisation standards, or settlement rails whose governance is ultimately subject to American legal authority.

It must not be built in a form that requires correspondent relationships with American financial institutions for ordinary operation. And it must be legally structured such that the mechanisms of freezing, seizure, and compliance reach through European courts, European regulators, and European law enforcement rather than through the secondary reach of American executive orders.

Some of this is already in the design.

The digital euro is issued by the European Central Bank, not by a private entity of any nationality, which removes the most basic vector of foreign jurisdictional reach. Its settlement rails are European by construction, and the offline architecture selected in October 2025 uses European hardware providers. These are necessary conditions.

They are not sufficient, because the surrounding payments ecosystem in which the digital euro will operate is saturated with American jurisdictional exposure, and the digital euro will only deliver on its sovereignty promise if its design actively resists the patterns of exposure rather than passively inheriting them.

The question is whether the regulation will state this commitment in specific terms. The current draft does not. Strengthening it on this point is the third of the three sharpenings.

Property Two: Privacy as Architecture

The current design of the digital euro treats payment privacy as a policy problem, to be managed through rules about what intermediaries may do with transaction data, what the ECB may see, and under which conditions information may be shared with tax authorities, financial intelligence units, or law enforcement.

These rules are real, and the proposed regulation contains strong language on data minimisation, on the prohibition of commercial profiling, and on the separation between the ECB and the payment service providers who handle customer relationships.

The difficulty is that policy rules can be changed, misapplied, or circumvented, and the infrastructure within which the rules operate determines how much difference a rule change would actually make.

If the underlying system generates a complete record of every transaction in a form that can be retrieved by the party that holds it, then privacy rests on the continued forbearance of that party and on the durability of the legal regime constraining them.

If the underlying system does not generate such a record in the first place, then privacy rests on the mathematics of the system itself, and policy changes cannot undo what the architecture did not do.

The distinction matters because the same digital euro will be used by Europeans in 2035, 2045, and 2055, under political conditions that cannot be predicted from Brussels in 2026. A design that is privacy-preserving only so long as current policy holds is a different kind of system from one that is privacy-preserving regardless of who is in government or what agency acquires access to the data.

The first is privacy as compliance. The second is privacy as architecture.

The technologies that would make architectural privacy possible are not speculative. Zero-knowledge proofs allow a transaction to be validated without the validator learning what the transaction contains. Blind signatures allow a token to be issued by one party and spent with another without the issuer being able to link the issuance to the spend.

Homomorphic encryption allows computations to be performed on encrypted data without the computing party seeing the underlying values. Each of these has been implemented at production scale in other contexts, and each has been evaluated in public research by the ECB and the Bank for International Settlements.

The offline form factor the ECB has selected is the most important design decision from the privacy perspective, because it is the part of the digital euro that has the strongest architectural case for cash-equivalent privacy. An offline transaction, by definition, does not transmit information to a central system at the moment it occurs.

Giesecke+Devrient, selected in October 2025 to deliver the secure-element hardware, has a European pedigree and the necessary engineering capacity.

The question is whether the cryptographic layer on top of the hardware is designed to preserve transaction privacy when the device eventually reconnects, or whether reconnection produces a delayed transaction record that is functionally equivalent to the online case.

The case for architectural privacy in the offline mode is strong. The offline digital euro is the direct replacement for cash at the point of sale, and it is the component most often discussed in terms of preserving the properties of cash.

A digital euro design that settled offline transactions anonymously, generated no retrievable transaction history at the wallet level, and allowed transfer between wallets without creating a persistent record would deliver architectural privacy for the subset of transactions that matters most to citizens in their daily lives.

The case for architectural privacy in the online mode is more contested.

Online transactions are where the regulatory concerns about money laundering, terrorist financing, and sanctions evasion concentrate, and the current proposal contemplates a model in which payment service providers hold transaction data, the ECB sees aggregate flows but not individual transactions, and authorities can access individual data through legal process.

This is a policy-based privacy model. Defensible, but weaker than the architectural model, and the weakness becomes material when political conditions change.

The digital euro regulation should, as a structural matter, commit to architectural privacy for the offline mode and push architectural techniques as far as they will go in the online mode, subject to the legitimate requirements of financial-crime enforcement. The current draft comes close to this on the offline side and falls short on the online side.

Closing the gap is the first of three sharpenings.

Properties Three and Four: Resilience and Universal Access

The Iberian Peninsula lost grid power for approximately ten hours on 28 April 2025. During those ten hours, card terminals stopped working, ATMs stopped working, and mobile payment apps stopped working. Citizens who had physical cash continued to transact; those who did not, could not.

The blackout was short enough to be inconvenient rather than catastrophic; a longer outage, a deliberate attack on payment infrastructure, or a co-ordinated failure across multiple systems would have produced harsher outcomes for the citizens without cash.

The offline form factor selected by the ECB in October 2025 (Giesecke and Devrient delivering the secure-element hardware) is the digital euro’s answer to property three, and an offline transaction by design does not depend on the surrounding network.

The cryptographic layer above the secure element determines whether the offline mode preserves working-when-infrastructure-fails or merely the appearance of doing so.[8]

Property four is harder. A banknote requires no device, no account, no credential, no digital literacy, and is available to the elderly, to children, to people who have lost access to their accounts, to visitors, to the digitally excluded, and to anyone whose relationship with formal financial infrastructure has broken down.

The design before the European Parliament addresses universal access through a physical card form factor in addition to phone-based wallets. The card form factor is necessary but not sufficient.

The wider design question is whether the digital euro will require an account relationship with a payment service provider for ordinary use, or whether it can be issued directly by member states or central banks to citizens who do not have such relationships.

Properties three and four are the two where the design choices already made come closest to preserving the cash properties they target. Strengthening either would produce useful improvement; the load-bearing design questions are properties one, two, and five, addressed in the sections immediately above and below this one.

Property Five: Holding Limits

The most politically contested element of the digital euro design is the question of how much of it any individual should be allowed to hold.

The ECB’s starting position, reiterated in multiple working papers and public statements by Piero Cipollone and other Executive Board members, is that holding limits are necessary to prevent the digital euro from functioning as a large-scale deposit alternative that would destabilise commercial banks.

The concern, modelled extensively by ECB and Bundesbank researchers, is that in a period of financial stress, depositors might shift significant balances out of commercial banks into digital euro wallets, draining bank funding at the moment banks are least able to absorb the shock.

A hard cap at a modest level (the working assumption has been around 3,000 euros per person) constrains the scale of any such shift.

The criticisms of this position are well-documented. A 3,000-euro cap makes the digital euro functionally useless for any transaction larger than a mid-sized consumer purchase, which pushes users back toward commercial payment rails for everything else and undermines the currency’s ability to establish itself as a meaningful medium of exchange.

It creates a political awkwardness, because the legal tender protections being codified in parallel are precisely protections against limits on the use of public money.

And it confuses the distinction between a transaction instrument and a store of value, treating a holding cap as a tool of bank-sector stability when the underlying concern is really about flow rates under specific stress conditions.

The German Bundesbank has been the most publicly active institution on this question, and the positions within it are more nuanced than the “Bundesbank opposes the digital euro” framing that has circulated in some commentary suggests. Bundesbank President Joachim Nagel has called the digital euro a priority for the Bundesbank.

Research by Burkhard Balz, an Executive Board member, has suggested that a limit as low as 500 euros would meet the financial-stability objective. Nagel’s own published research has argued for a range of 1,500 to 2,500 euros. Philipp Hachmeister, speaking for the German banking industry, has said that the industry could cope with a cap at 3,000 euros.

The European Central Bank’s own research, including papers by Ulrich Bindseil and others, has argued that even in extremely pessimistic scenarios for deposit substitution, a limit in the range of 3,000 euros would not produce systemic banking stress.[9]

The economists’ open letter of 11 January 2026, signed by Thomas Piketty, Paul De Grauwe, Daniela Gabor, José Leandro, and sixty-six others, pushed in the opposite direction.

The letter called for generous and gradually rising holding limits on the grounds that a meaningful digital euro requires meaningful balances, and that the digital euro will only function as a public good if citizens can hold it in amounts that matter to them.

The letter framed the question not as a technical dispute about financial stability but as a political question about whether Europeans will control their money in the digital age or whether others will control it for them.[10]

The debate as currently structured presents a false choice, because it treats holding limits as a single parameter applying uniformly to all digital euros held by a given citizen. A design that split the digital euro into two functionally different holdings could satisfy both the financial-stability concern and the cash-equivalence requirement without forcing either to give way.

The specific proposal is this.

Online digital euro holdings should continue to operate under a modest cap (3,000 euros is a reasonable placeholder) with an automatic “waterfall” arrangement that sweeps excess funds into a linked commercial bank account.

These are the balances most likely to behave as deposit substitutes in a stress scenario, and the cap plus waterfall directly addresses the ECB’s and the Bundesbank’s financial-stability concern.

They are also the balances most likely to be used for ordinary online transactions, where a 3,000-euro cap is not meaningfully constraining.

Offline digital euro holdings should operate under a different regime, closer to the regime cash itself operates under. No aggregate holding cap, but regulated flow rates that limit how much digital euro can be moved into offline wallets per month and per year.

Illustrative numbers in the range of 5,000 euros per month and 30,000 euros per year would allow meaningful offline balances to accumulate for citizens who want them, without creating the kind of sudden deposit flight the ECB is concerned about. The flow-rate constraint is the financial-stability instrument; the absence of an aggregate cap is the cash-equivalence instrument.

Both work because offline balances are, by design, outside the rapid online reallocation dynamic that produces deposit flight in the first place.

These numbers are illustrative. The structural point is that treating the holding-limit question as a single parameter forces a choice between two legitimate objectives that a two-parameter design can satisfy simultaneously. The Piketty letter is correct that a meaningful digital euro requires meaningful balances. The Bundesbank is correct that bank stability requires bounded deposit substitution.

A dual-track design meets both requirements.

The Anchor Customer

A digital euro designed as specified above, with architectural privacy in the offline mode, a dual-track holding-limit regime, and a legal framework that reaches through European courts rather than foreign compliance infrastructure, would be a substantial piece of sovereign financial infrastructure. It would also be expensive.

The ECB’s own estimates put the central cost of implementation at 1.3 billion euros, with bank integration costs adding a further 4 to 6 billion euros across the Eurosystem.[11]

These are not small numbers, and the political debate about the digital euro has often centred on whether the cost is justified given the specific use cases the system would serve. The answer changes if the digital euro is understood not only as a payment instrument but as an anchor customer for a broader European sovereign digital stack.

The European Central Bank is, by any standard, one of the largest potential buyers of sovereign digital infrastructure in Europe.

Its procurement decisions for the digital euro (which European hardware providers will produce the secure elements, which European software will run the validation layer, which European cloud providers will host the non-sensitive components, which European cryptographic primitives will be adopted as standards)

will establish reference customers for a cluster of European firms whose scale depends on having exactly this kind of anchor contract.

Paper 8 of this series made the argument that European sovereignty at the infrastructure layer requires someone to place the order,

and that the absence of anchoring government procurement is the single most consistent reason European technology companies fail to reach scale in direct competition with American firms whose early growth was underwritten by the CIA, NASA, and the Department of Defense.

Giesecke+Devrient, the October 2025 selection for secure-element hardware, is a German company that will now have an anchor contract.

The equivalent selections across the rest of the stack (and they have not all yet been made) are the places where the digital euro can operate as a sovereignty instrument in a second sense, procuring into existence the European supply chain that would otherwise not be built.

This frame changes the cost-benefit calculation for the digital euro project. The 1.3 billion euros of ECB implementation cost is the anchor commitment around which a European sovereign digital stack can be assembled.

The Bundesbank and other national central banks have made similar arguments in other contexts about the need for European anchor procurement in semiconductors, in cloud, in artificial intelligence. The digital euro is the payment-layer instance of the same logic.

What the Regulation Should Say

The European Parliament and Council are expected to reach political agreement on the Digital Euro Regulation during the second half of 2026, with implementation running on the timeline already announced: call for expression of interest to payment service providers closed 14 May 2026, pilot programme beginning in the second half of 2027, target issuance in 2029.

The text that emerges from trilogue will determine the character of the digital euro for at least the first decade of its operation, and probably longer.[12]

The text should do four things the current draft does not do explicitly enough.

It should state clearly that digital euro balances are reachable by European legal process and not reachable by foreign executive action. The specific language can be drafted by jurists with more facility in EU law than I have, but the substantive commitment is straightforward. The digital euro is European public money.

European authorities acting through European law can access it, freeze it, or seize it on findings of fact that would ordinarily support such actions. Foreign authorities cannot do so except through the recognised mechanisms of mutual legal assistance, subject to European judicial review, with protections for fundamental rights no weaker than those that apply to any other European asset.

It should commit to architectural privacy for the offline mode and make the commitment testable.

The regulation should specify that offline digital euro transactions generate no retrievable transaction record at the wallet or network level, that the cryptographic primitives selected preserve this property under reconnection, and that the ECB’s annual operational review will include an independent cryptographic audit of the privacy architecture.

A commitment in statute is the only kind of commitment that is robust to political change.

It should adopt the dual-track holding-limit structure. The online mode retains a modest aggregate cap with a waterfall arrangement; the offline mode removes the aggregate cap and substitutes regulated flow rates calibrated to prevent rapid deposit substitution.

The specific numerical values can be adjusted by ECB decision based on operational experience and financial-stability analysis, but the two-track structure should be fixed in primary legislation so that the political choice to preserve cash-equivalence in the offline mode cannot be reversed by later operational tightening.

It should commit the ECB and the national central banks to procuring digital euro infrastructure from European providers where European capacity exists and to investing in European capacity where it does not. The regulation need not specify vendors; it should specify the principle and the reporting mechanism by which Parliament can verify that the principle is being followed.

The digital euro is the largest new monetary-infrastructure project in Europe this decade, and the procurement choices it generates are among the most consequential industrial-policy choices the central bank will make in that time.

Completing the Project

The euro was introduced in 1999 as an electronic currency for financial institutions and in 2002 as a physical currency for citizens. It has outlasted a financial crisis, a sovereign-debt crisis, a pandemic, and several confident predictions of its imminent collapse.

It has expanded from eleven countries to twenty-one, and it is now shared by 357 million people across a quarter of the European continent. It remains, in 2026, the second most important reserve currency in the world.[13]

It has not, however, completed its project.

The retail rails over which euros move in daily commerce are mostly operated from outside Europe. The wholesale rails over which euros settle in tokenised form are increasingly assembling around a governance model domiciled in Delaware.

The stablecoin layer in which digital euros circulate outside central bank money is dominated by American issuers and, in its European variants, increasingly built on American infrastructure.

And the physical instrument through which European citizens have always held central bank money directly is retreating from daily use at a pace that makes its preservation, in its current form, a matter of years rather than decades.

The digital euro is the component of Europe’s response to this condition that sits inside the European Central Bank’s own institutional capacity.

The ECB cannot legislate the card duopoly out of existence or the stablecoin market into European shape, but it can issue a digital form of euros that is designed, from its architecture upward, to preserve the direct-issuance relationship between the central bank and European citizens. That is the project before the European Parliament now.

The project is the right one. The current design is close to right. Three specific sharpenings would make the difference between a digital euro that formally preserves the properties of cash and a digital euro that actually does so. A legal commitment that European money is subject to European law, and not to foreign executive action operating through private compliance infrastructure.

Architectural privacy in the offline mode. A dual-track holding-limit regime.

These are not radical proposals. They are extensions of positions already taken by the ECB’s own executive, by seventy of Europe’s most prominent economists in their open letter of January 2026, and by the design choices the project has already made. Making them explicit in the regulation is the political step that remains.

Paper 15 of this series examines the wholesale side of the same question: how euro settlement at institutional scale can be preserved under European governance as the wholesale payment layer moves to distributed ledger infrastructure governed, for the moment, from Delaware. The retail and wholesale arguments are parts of the same project.

The digital euro, in both its retail and wholesale forms, is the instrument through which Europe completes the monetary union it started in 1999: as a sovereign system, operated under European law, issued by European institutions, and accountable to European citizens.

Of the five properties cash carries, four are operational. They describe what the instrument does day to day. One is architectural. It describes what the instrument is.

The architectural property is property one, public money held directly with no commercial or technological intermediary in the chain of custody, and it is the property whose preservation makes the other four structural rather than procedural.

The Circle USDC freezings of euro-linked wallets and the ICC sanctions reaching judges through their Visa, Mastercard, and Amazon services demonstrate what intermediary-driven failure looks like for digital money.

A digital euro that is a direct claim on the European Central Bank, with no intermediary that can be compelled by foreign executive action, is the only digital form of money that preserves what cash preserves. That is the question the regulation must answer, and the answer is in the architecture or it is nowhere.

[1] Council of the European Union, three legal acts on Bulgaria’s adoption of the euro, 8 July 2025: Council Regulation (EU) 2025/1346 amending Regulation (EC) No 974/98; Council Regulation (EU) 2025/1347 amending Regulation (EC) No 2866/98 fixing the conversion rate at 1.95583 leva to one euro; Council Decision (EU) 2025/1345 authorising adoption.

[2] European Central Bank, Eurosystem balance sheet data and ECB Statistical Data Warehouse, 2025. Banknotes in circulation approximately 1.55 trillion euros against M3 broad money aggregate of approximately 17 trillion euros.

[3] European Commission, Single Currency Package, proposal COM(2023) 364 final of 28 June 2023, comprising the Regulation on the legal tender of euro banknotes and coins and the Regulation establishing the digital euro. Trilogue agreement on the legal-tender regulation reached December 2025.

[4] De Nederlandsche Bank, Cash Usage in the Netherlands, 2024 figures; Bank of Finland, Payment Statistics 2024.

[5] Circle Internet Financial, USDC Transparency and Reserve Reports, 2022-2026. Circle has publicly confirmed compliance with OFAC designations including the August 2022 Tornado Cash sanctions, freezing approximately 75,000 dollars of USDC held in 38 sanctioned wallet addresses within hours of the OFAC list update.

[6] Cumulative OFAC designations under Executive Order 14203 across three rounds (Karim Khan, June 2025 four judges, August 2025 expansion) totalling eleven individuals.

[7] Sharon Prost, public account of consequences of OFAC designation, reported in The Wall Street Journal and Reuters, August-September 2025.

[8] European Central Bank press release, ‘ECB selects digital euro service providers’, 2 October 2025. Framework agreement awarded to a consortium led by Giesecke+Devrient with Nexi and Capgemini for the offline digital euro solution.

[9] Ulrich Bindseil and ECB co-authors, ECB Working Paper Series on digital euro design and bank funding effects, 2022-2025.

[10] ‘The Digital Euro: Let the Public Interest Prevail’, open letter to the European Parliament from 70 economists, 11 January 2026, published by the Veblen Institute and the Sustainable Finance Lab.

[11] European Central Bank, Preparation Phase Closing Report, October 2025. Eurosystem development cost estimated at approximately 1.3 billion euros until first issuance, with annual operating costs of approximately 320 million euros from issuance onwards.

[12] European Central Bank, Governing Council decision of 30 October 2025 to move to the next phase of the digital euro project; ECB call for expression of interest to payment service providers, deadline 14 May 2026; pilot exercise targeted mid-2027; first issuance during 2029 conditional on adoption of the Digital Euro Regulation.

[13] European Central Bank and Eurostat, euro area population statistics, 2026. With Bulgaria’s accession on 1 January 2026, the euro area covers 21 member states and approximately 357 million people.