The European Commission's targeted consultation on the review of the Markets in Crypto-Assets Regulation closes on 30 September 2026, and one sentence in it will shape this category more than any of the questions that follow it. Section 4.4 of the consultation document introduces the subject like this: prediction markets are markets "where parties take a bet on a future development in almost any area, often far removed from financial assets or services". Counted across the published document, that is the only place the word "bet" appears anywhere in it. "Betting", "gambling" and "wager" do not appear at all, and neither does "binary option" or "event contract".
That sentence is not an attack and it is not inaccurate. It reads like what it is, a short description written by an institution that has not been handed a better one. It matters anyway. Four public authorities now use four different names for the same instrument, and this industry has spent far more effort arguing about which name is correct than writing the description that all four of them actually run their tests on. The name is the part we do not control. The description is the part we do.
Brussels borrowed our name and a gambling word in the same paragraph
The section is headed Prediction markets and perpetual futures and sits in PART 4 of the Commission's document, the part titled "Policy areas beyond the current scope of MiCA". The heading uses the name the industry chose for itself. The sentence underneath uses a word that belongs to a different rulebook. Both are true at the same time, and neither is a slip of the pen.
The same paragraph also fixes what is being asked about, and the boundary it draws is technological rather than commercial. The Commission writes that these markets "are increasingly conducted on or are structured through DLT, smart contracts and crypto-assets", and Question 70 asks only about prediction markets "where they are DLT enabled and facilitated through smart contracts". A venue running an ordinary matched order book with no token anywhere near it is not clearly inside the question at all, which is worth noticing before anyone decides the file does not concern them. Question 71, which asks what substantive requirements should apply to providers of prediction market services, arrives with no answer options of any kind, a point we set out in an earlier piece on the empty requirements box.
Washington put the name in a heading and the test in the sentence below it
The American frame looks like the opposite of the European one. Structurally it does the same thing. The Commodity Exchange Act at 7 U.S.C. section 7a-2(c)(5)(C) carries the heading "Special rule for review and approval of event contracts and swaps contracts", and the clause under it is headed "Event contracts". The operative words never use the term. What they describe is "agreements, contracts, transactions, or swaps in excluded commodities that are based upon the occurrence, extent of an occurrence, or contingency", followed by a list of subjects: activity unlawful under federal or state law, terrorism, assassination, war, gaming, and other similar activity the Commission may add by rule.
The rule built on that statute repeats the pattern exactly. Commission Regulation 40.11 is headed "Review of event contracts based upon certain excluded commodities", and paragraph (a)(1) then forbids a registered entity to list "an agreement, contract, transaction, or swap based upon an excluded commodity" that "involves, relates to, or references terrorism, assassination, war, gaming, or an activity that is unlawful under any State or Federal law". The name sits in the heading. The test is a mechanism plus a list of things the contract may point at.
That is also why the event contracts proposal published on 12 June 2026 by the Commodity Futures Trading Commission turns out to be an argument about two ordinary words, "involves" and "gaming", rather than about what the product ought to be called. We wrote about the measuring stick that definition hands an exchange when the proposal first appeared.
ESMA wrote down in one line that the name does not count
The plainest statement of the whole problem comes from the European Securities and Markets Authority, in a public statement issued on 3 July 2026 under reference ESMA35-243228190-8148, announced the same day in a short press release. The statement says it outright: "the commercial name provided by firms (e.g. 'event contracts') is irrelevant for the categorisation under MiFID II of products distributed, marketed or offered to clients".
Having put the name aside, ESMA supplies a description instead. Event contracts, it writes, are "agreements whose financial outcome is binary (a fixed payout or no payout at all) and depends on a yes-or-no answer to a question about a future event". The classification then runs on the definition of a binary option carried over from the 2018 product intervention measures: settled in cash, paying only at close-out or expiry, and paying either a predetermined fixed amount or zero according to whether the underlying meets a predetermined condition. Every element of that test is a fact about how the contract is built, and none of them is a fact about what it is called.
The footnote carries the consequence. Footnote 1 of the same statement records that an event contract "may also classify as a bet under national gambling legislation" and, where it takes token form and is not a financial instrument, as a crypto-asset regulated under MiCA. One instrument, three possible rulebooks, and the sorting is done on construction. We mapped that fork out in the piece on which rulebook applies when an event contract is a token.
Britain had a shelf for this twenty years before the product existed
The Gambling Commission published its answer early, in a blog post of 4 February 2026 by Brad Enright, its Director of Strategy. The description it uses is the market's own: a prediction market is "a platform which enables participants to trade event-based contracts on markets including financial, sports, and political events". The classification that follows is not. Their "core aspects are akin to what in the UK would be described as a 'Betting Exchange'", the post says, and it "would appear current products would fall within the definition of a 'Betting Intermediary' under UK legislation".
Reaching that conclusion required no new drafting at all. Section 9 of the Gambling Act 2005 defines betting as making or accepting a bet on the outcome of "a race, competition or other event or process", on "the likelihood of anything occurring or not occurring", or on "whether anything is or is not true". Section 13 defines a betting intermediary as "a person who provides a service designed to facilitate the making or acceptance of bets between others", which is a serviceable description of a matched book. Words written in 2005 reach a product launched two decades later because they describe a mechanism rather than a market segment. How Britain then divides the same book between two regulators according to what each contract points at is a separate story, and we told it in August 2026.
The four names disagree and the four tests barely do
Line them up as they stood on 13 September 2026. Brussels says prediction markets and describes a bet, with distributed ledger technology as the boundary of the question. Washington says event contract in the heading and then tests what the settlement is based upon and what activity the contract involves. Paris says the name is irrelevant and tests whether the payout is binary. London says betting intermediary and tests whether the service facilitates bets between other people.
Four vocabularies, and underneath them very nearly one question, asked three ways. What does this contract pay, and when. What does it point at. Who is on the other side of it. An operator who can answer those three in writing, with evidence, has answered every one of the four authorities at once, whatever word each of them writes at the top of the page.
Where the four genuinely diverge is in the consequence, and the consequences are not small. In the United States the answer routes a contract into listing review and a public interest test. Under MiFID II a binary payout means the product cannot be marketed, distributed or sold to retail clients at all, and ESMA reminds firms in the same statement that distributing only to non-retail clients still requires authorisation. In Great Britain the answer is an operating licence with the conditions attached to it. Under MiCA the token form may pull in a third set of obligations. Same three questions, four different bills at the end.
The description is the only part of this an operator can actually write
If the classification runs on construction rather than vocabulary, then the useful thing to file before 30 September is a construction document, not a naming argument. Four things belong in it, and a venue that cannot produce them has found something out about itself.
- What the contract pays and when. Binary, scalar or combination. What happens to a combination when one leg cannot resolve. Whether payout occurs only at expiry or close-out. This is the exact ground on which ESMA does its sorting, so leaving it vague does not avoid the test, it just means somebody else performs it on your order book.
- What the settlement source is and who may change it. Named source, published before the market opens, with the rule for what happens when the source is silent or contradicted. Our Directive 01 says resolution rules are published before a market opens and are not revised while money is at risk, which is a sentence written to be checked rather than believed.
- Who stands on the other side. The Commission's sentence says parties take a bet, and that is the one part of it that is testable. Either the book is matched between users, or the venue or an affiliate takes principal risk, or both happen in different products. Whichever it is, it belongs on the surface where the trade happens, which is what Directive 02 asks for.
- How a person reaches the venue and what protects them once inside. Jurisdictions reached, the access controls actually in force, and whether deposit limits, cool-off and self-exclusion are built into the account or granted on request. That distinction is Directive 03, and it is visible from outside in a way that a policy document is not.
The Commission has written down what it will treat as a useful answer, which saves guessing: responses "are expected to be most useful where they present a clear and detailed narrative, demonstrated by data (where possible), concrete examples, legal references and qualitative evidence", with suggestions for solutions welcome. Data, examples and legal references are exactly the materials a description of this kind is made from, and exactly the materials a naming argument does not need.
The name you object to is the one nobody described first
There is a version of the next two years in which this category spends its energy objecting. Objecting that a prediction market is not a bet, that an event contract is not a binary option, that a betting exchange is a different animal. Some of those objections are even correct. None of them is a reply to what any of the four authorities actually asked, because none of the four asked what the product is called.
Our manifest puts the alternative in one line: "The market that survives regulation is the one that was already behaving as if regulation had arrived." Directive 05 makes the same point about jurisdictions, that standards are set against the one that regulates next rather than the one that has not yet. A description written by operators, filed while the file is open, with numbers in it, is what that looks like in practice on a specific date. If that description is accurate, then being called an event contract in Washington and a betting intermediary in London stops being a threat, because the same paragraph answers both.
The Commission's consultation page records that the window was extended to 30 September 2026 at 23:59 CEST, and that only responses submitted through the online questionnaire are counted. After it shuts, the next fixed point is set by Article 140(1) of the Regulation, quoted in the consultation's own footnote: by 30 June 2027 the Commission, having consulted the European Banking Authority and ESMA, presents its report to the European Parliament and the Council, accompanied where appropriate by a legislative proposal. Between those two dates the description of this category will be written by whoever bothered to write one. Firms that want to put their name to a floor while the file is open can do so by signing the commitment.