Digital Money Taxonomy: Why the ECB Says Crypto Terminology Is Fundamentally Broken

The Hidden Power of Digital Money Terminology

Language does more than describe reality — it shapes it. When Ludwig Wittgenstein wrote that “to imagine a language means to imagine a form of life,” he captured a truth that ECB Working Paper 3022 applies with devastating precision to the world of digital money and finance. Published by Ulrich Bindseil, Charles-Enguerrand Coste, and George Pantelopoulos, this paper argues that the terminology used across crypto, DeFi, and CBDC discussions is not merely imprecise — it is systematically biased in ways that distort public perception, misguide policy debates, and influence investment decisions.

The stakes are enormous. When retail investors hear “decentralised finance,” they imagine a system free from the control of powerful intermediaries. When policymakers discuss “smart contracts,” they may assume these instruments carry legal weight. When regulators define “crypto-assets” so broadly that a tokenised government bond falls into the same category as Bitcoin, they create regulatory confusion that benefits incumbents at the expense of clarity. The paper demonstrates that five factors drive this terminological chaos: the speed of innovation, technical complexity, vested financial interests, path dependencies once terms are enshrined in law, and the inherently technical nature of the underlying systems.

Six Principles for Sound Financial Terminology

Drawing on ISO 704:2022, the international standard for terminology work, the ECB paper establishes six criteria that good financial terms should meet — and systematically shows how digital money terminology fails nearly all of them.

Monosemy requires that one term corresponds to one concept. The term “tokenisation” spectacularly violates this principle, carrying at least four distinct meanings across different contexts and institutions. Transparency demands that a term’s meaning should be inferable without additional context — yet “smart contract,” “mining,” “gas fee,” and “burning” are opaque to anyone outside the crypto ecosystem.

Consistency requires terms to integrate into existing concept systems. The paper highlights how “retail CBDC” and “wholesale CBDC” use the word “digital” to mean entirely different things — electronic for retail, DLT-based for wholesale. Neutrality prohibits judgemental connotations, yet “fiat money” (implying central bank money is imposed by decree rather than trusted), “TradFi” (dismissing centuries of financial evolution as merely “traditional”), and “DeFi” (claiming decentralisation that does not exist) all carry heavy ideological baggage.

The paper also applies four quality criteria for definitions: non-circularity (definitions must not be tautological), accuracy (neither too narrow nor too broad), singleness (describing only one concept), and conciseness. The MiCA regulation’s definition of crypto-asset is flagged as tautological — essentially defining a crypto-asset as an asset that uses cryptography, which is circular and unhelpfully broad. These failures matter because once flawed terminology enters legislation, it becomes extremely difficult to correct. To explore how financial institutions are navigating these definitional challenges, see our interactive library on financial regulation.

Crypto-Asset: When a Bond on DLT Is Still a Bond

Perhaps the most consequential terminological confusion concerns the term “crypto-asset” itself. The ECB paper draws a fundamental distinction between native crypto-assets — digital objects created within a DLT environment that have no existence outside it, like Bitcoin — and tokenised traditional assets — real-world financial instruments like bonds, equities, or real estate whose ownership records are maintained on a DLT platform.

The paper argues forcefully that calling a tokenised government bond a “crypto-asset” is as misleading as calling a bearer bond a “paper-asset.” The economic nature of the asset does not change based on the technology used to record ownership. A bond on a DLT platform is still a bond — it carries the same rights, obligations, and risk profile. Lumping it together with unbacked speculative tokens like Bitcoin under the umbrella of “crypto-assets” creates false equivalences that confuse investors and complicate regulation.

The authors propose qualified naming instead: “bond represented on a DLT platform” preserves the asset’s economic identity while noting the technological innovation. For truly native, unbacked digital objects, they suggest alternatives like “virtual assets” or “virtual ledger entries” — terms that accurately describe what these objects are (entries on a digital ledger) without the misleading connotations of “crypto” (which implies security and sophistication) or “asset” (which implies inherent value). The ECB Working Papers series provides the full academic context for these proposals.

Smart Contracts Are Neither Smart Nor Contracts

The term “smart contract” ranks among the most misleading in all of digital finance. Coined by Nick Szabo in 1994, the concept was later implemented through Ethereum’s programmable blockchain. But as the ECB paper demonstrates, these instruments are neither “smart” in any meaningful sense nor “contracts” in any legal sense.

Smart contracts cannot anticipate all future states — they execute predetermined code when conditions are met, relying on oracles (third-party data feeds) for real-world information. This dependency on external data sources means they are fundamentally limited in their ability to handle the complexity and ambiguity inherent in real contractual relationships. As legal scholar Elaine Mik concluded, “Smart contracts are contracts in name only.” Werbach and Cornell reinforced this: “smart contracts are not intended to be legally enforceable.”

The “smart” label is equally problematic. These are simple if-then programs — they cannot exercise judgment, interpret ambiguous situations, or adapt to unforeseen circumstances. The trust supposedly eliminated by removing human intermediaries is merely transferred to code developers who face no formal auditing requirements or accountability frameworks. Even Vitalik Buterin, Ethereum’s creator, has acknowledged the gap between implementation and intent in smart contract execution.

The paper argues that the term is functionally redundant with programmability — the ability to embed conditional logic in financial transactions. And programmability is not even new to DLT: traditional payment systems have long embedded program code through direct debits, APIs, and standing orders. What DLT adds is the specific execution environment, not the concept of automated conditional transactions.

The Stablecoin Misnomer: Coins Without Stability

The etymology of “stablecoin” reveals its marketing origins. The term first appeared on BitcoinTalk forums between 2011 and 2014, gaining traction with the launch of early projects like BitUSD, NuBits, and Tether in 2014. From the beginning, the name was aspirational rather than descriptive — a branding exercise designed to contrast with the notorious volatility of Bitcoin and other cryptocurrencies.

The ECB paper identifies two fundamental problems with the term. First, “coin” wrongly suggests a bearer instrument — a physical object that carries its value independently, like a gold coin or banknote. In reality, stablecoins are ledger entries: digital records in a database that have no existence outside their DLT platform. Calling them “coins” evokes a physicality and independence they do not possess.

Second, the term implies stability that is anything but guaranteed. Algorithmic stablecoins — which attempted to maintain their peg through code-based mechanisms rather than asset backing — have been decisively proven unviable, most dramatically with the collapse of TerraUSD. Even the Financial Stability Board hedges with unusual language, saying stablecoins “aim to maintain” a stable value rather than asserting they actually do.

The paper’s proposed alternative is revealing in its simplicity: “e-money on a distributed ledger” or “e-money represented on a DLT platform.” Fully backed stablecoins are functionally equivalent to electronic money — a well-understood concept with established regulatory frameworks. The EU’s MiCA regulation already takes this approach, deliberately avoiding the term “stablecoin” in favour of “e-money token” and “asset-referenced token.” This is not mere academic pedantry: the terminology determines which regulatory framework applies, which in turn determines what consumer protections exist.

Tokenisation: A Revolutionary Concept or Redundant Term?

Tokenisation is widely celebrated as one of the most promising applications of distributed ledger technology. But the ECB paper argues the term carries so many different meanings that it has become “largely redundant and misleading.” The authors identify at least four distinct historical interpretations of what tokenisation means, used inconsistently by different institutions, in different documents, and sometimes within the same document.

At its core, tokenisation represents the process of recording ownership on a ledger. But this is precisely what traditional financial infrastructure already does. The ECB’s own TARGET2-Securities (T2S) system settles securities by updating ownership records in a centralised ledger — functionally identical to what DLT-based tokenisation claims to innovate. The difference is the underlying technology (distributed vs. centralised ledger), not the concept of recording ownership.

Historically, “token” connoted a bearer instrument — something that existed outside a ledger, carrying its value physically. A subway token or a casino chip are classic examples. But in DLT usage, “tokens” exist only as ledger entries. This contradiction means the term simultaneously evokes bearer-instrument independence (suggesting tokens are self-contained) while describing ledger-dependent records (tokens that cannot exist outside their DLT platform). The once-prominent debate about “token-based vs. account-based CBDC” was rightly abandoned by around 2023 when this confusion became untenable. For further analysis of digital finance innovations, explore our interactive library.

The DeFi Decentralisation Illusion Exposed

Perhaps no term in digital finance carries more ideological weight than “DeFi” — decentralised finance. The ECB paper, drawing on research by Aramonte, Huang, and Schrimpf (2021), argues that DeFi contains a “decentralisation illusion” where the label obscures significant centralisation at every layer of the stack.

At the settlement layer, permissionless blockchains appear decentralised in theory but exhibit concentration in practice. Mining pools and stake pools aggregate computational or financial power, creating de facto gatekeepers. The governance of blockchain protocols is controlled by small groups of software developers who make consequential decisions — as demonstrated when a handful of Bitcoin developers resolved the 2013 chain fork, effectively deciding the future of a supposedly decentralised network.

At the interface layer, users interact with DeFi through centralised service providers — websites, mobile apps, and exchanges that are as centralised as any traditional financial institution. When El Salvador adopted Bitcoin as legal tender, users primarily transacted through the Chivo wallet — a centralised, government-controlled application that settled transactions outside the Bitcoin network entirely. The “decentralised” label served as marketing rather than technical description.

The paper draws on multiple competing taxonomies of the DeFi “stack” (Schär 2021, Auer et al. 2024, IOSCO 2022) to show that even academics cannot agree on how to describe the layers of DeFi — further evidence that the terminology is not fit for purpose. The term’s real function, the authors suggest, is marketing: it positions crypto-native financial services as inherently superior to “TradFi” (traditional finance), despite the reality that DeFi introduces new risks — smart contract vulnerabilities, oracle manipulation, governance attacks — without eliminating the centralisation it claims to abolish. Research from the Bank for International Settlements corroborates these concerns about DeFi’s structural limitations.

CBDC Terminology: Digital vs Electronic Money

The terminology around Central Bank Digital Currencies reveals a particularly instructive case of inconsistency. The ECB paper traces the term’s origin to around 2015, when Bank of England’s Andy Haldane and the Federal Reserve’s David Andolfatto (who coined “Fedcoin”) began discussing the concept. The first major institutional report using “CBDC” was the CPMI-MC publication of 2018.

The fundamental problem: “digital” means different things in “retail CBDC” and “wholesale CBDC.” For retail CBDC, “digital” essentially means electronic — a form of central bank money accessible to the general public through electronic means. For wholesale CBDC, “digital” specifically implies DLT-based settlement. This inconsistency within a single term family violates the consistency principle and creates confusion about what CBDC actually represents.

The paper’s proposed replacements are deliberately precise. For retail use, it suggests “retail central bank electronic money” (rCBEM) — replacing “digital” with “electronic” to avoid the DLT connotation, and “currency” with “money” since “central bank money” is the established term (whereas “central bank currency” is not standard usage). For wholesale use, the paper proposes simply “wholesale central bank money represented on a DLT platform” — no new qualifier needed since wholesale central bank money is inherently electronic.

Most provocatively, the paper demonstrates that retail CBDC is not actually a new concept. The Bank of Finland experimented with pre-funded central bank money payment smart cards in 1987, launched the system in 1992, and operated it for three years. In the 1990s, central bank employees in France, Germany, Switzerland, and the UK held electronic current accounts at their respective central banks. What is marketed as revolutionary innovation has historical precedents stretching back decades — obscured by terminology that implies novelty. The International Monetary Fund’s fintech research provides additional context on the global evolution of digital money concepts.

How Flawed Digital Money Terminology Shapes Policy

The ECB paper’s analysis is not academic abstraction — it has direct consequences for how billions of euros in financial regulation are crafted and applied. When the EU’s MiCA regulation defines “crypto-assets” broadly enough to encompass both Bitcoin and tokenised government bonds, it creates a regulatory framework that may either over-regulate traditional assets represented on new technology or under-regulate speculative instruments by treating them as equivalent to familiar financial products.

The paper identifies a crucial distinction between the object layer (what an asset is economically) and the settlement layer (how it is transferred). A government bond remains a government bond whether its ownership is recorded on paper, in a centralised database, or on a distributed ledger. Conflating these layers — as much crypto terminology does — leads to regulation that targets technology rather than economic substance.

The decline of cash across major economies provides additional context: euro area cash payments at point of sale fell from 79% to 59% between 2016 and 2022. US cash use dropped from 40% in 2012 to 19% in 2020. Sweden’s cash use fell from 33% to 10% over the same period. This rapid digitisation makes getting the terminology right increasingly urgent — the terms we use today will shape the regulatory frameworks governing how most people interact with money for decades to come. Our interactive library features additional analysis of regulatory frameworks in digital finance.

Fixing the Language of Digital Finance

The ECB paper’s overarching conclusion is both sobering and constructive: public sector institutions have a responsibility to use sound, neutral terminology, and the current lexicon of digital money fails this standard comprehensively. Five key recommendations emerge from the analysis.

First, separate the object from the platform. A financial instrument’s economic nature should be described independently of the technology used to record it. “Bond on a DLT platform” is clearer and more accurate than “tokenised bond” or “crypto-bond.” This principle applies across all asset classes and eliminates the false novelty that current terminology creates.

Second, retire misleading terms. “Smart contracts” should be replaced with “programmable conditions” or simply “code-based automation.” “DeFi” should be qualified as “crypto-native financial services” — removing the unearned claim of decentralisation. “Stablecoins” should become “e-money on a distributed ledger” for fully-backed instruments.

Third, harmonise CBDC terminology. The retail/wholesale distinction should use consistent qualifiers: “electronic” for the general concept of non-physical central bank money, and “DLT-represented” when the distributed ledger technology is specifically relevant. The proposed rCBEM term deserves serious consideration.

Fourth, resist path dependency. Once flawed terms enter legislation, they become extremely difficult to correct. The EU’s approach in MiCA — deliberately avoiding “stablecoin” in favour of “e-money token” — shows that better choices are possible even when industry terminology pushes in the opposite direction.

Fifth, recognise that terminology is not neutral. Terms like “fiat money” and “TradFi” carry implicit value judgments that position crypto-native systems as superior. Public institutions should actively counter this framing by using precise, descriptive language that accurately represents both the capabilities and limitations of new technologies. As Wittgenstein warned, “A picture held us captive” — and the pictures drawn by current digital money terminology systematically favour the interests that created them.