The global license map

Two ideas from earlier in this series only matter when you put them on a map. The first: prediction markets are winner-take-most — liquidity attracts liquidity, so the deepest venue keeps getting deeper and the runner-up bleeds out.¹ The second: the right to run one of these markets is granted one country at a time — the mechanism is universal and copyable, but the license is neither.² Stack those two facts together and you don't get a product question. You get a geography question. Whoever is first to be legal and liquid inside a given market tends to keep that market — and "a given market" means a country, with its own regulator, its own politics, its own idea of what an event contract even is.

So the real contest in this category is not a better order book. It is a race across a map, fought regulator by regulator, and most of the map is still blank. This piece is an attempt to read it as it stands at the end of April 2026 — honestly, including the large parts that are genuinely unsettled. Treat every line that follows as a snapshot of a moving picture, not a finished atlas.

The framework: onshore beats offshore

Before the regions, the logic that governs all of them. Start from a fact that is true almost everywhere: the demand already exists. People want to take positions on elections, sports, rates, weather, and whatever else has a clean answer — and if no licensed venue will let them, they go to an offshore one that will. That is the status quo in most of the world right now: real money, real volume, sitting on platforms a national regulator cannot see, supervise, or tax.

Given that, a regulator has essentially two end-states to choose between. Offshore and invisible: the activity happens anyway, on platforms beyond reach, with no surveillance, no consumer protection, and no record. Onshore and surveilled: the activity happens on a licensed venue with position limits, audited settlement, identity checks, and a regulator who can pick up the phone. Stated that way, the long-run preference of most serious regulators is not mysterious. Banning the demand does not remove it; it exports it. Licensing it — carefully, with conditions — is how you get jurisdiction back.

This is not a prediction that every country welcomes event contracts. Plenty will not, and some will ban them outright and mean it. It is a claim about direction: where these markets have moved from grey to legal, it has usually been because a regulator decided that supervising the thing beat pretending it wasn't happening. The map below is mostly the story of who has reached that decision, who is fighting about it, and who hasn't started.

Region by region

Four broad zones, from the most settled to the most open. The honest caveat up front: regulatory status here ranges from "decided in court" to "genuinely nobody knows yet," and I have tried to label which is which. Where a country's posture is unclear, I say so rather than inventing a license that doesn't exist.

United States — the most advanced

The US is the furthest along, and it is the reason this category exists in its current form. The Commodity Futures Trading Commission — the federal derivatives regulator — treats event contracts as derivatives, the same legal family as a futures contract on oil or wheat. That framing dates to the CFTC's 2020 recognition of event contracts as a distinct, regulated asset class,³ and it is the hinge the whole market turns on: once these were derivatives rather than wagers, a federally-regulated exchange could list them. Kalshi is exactly that — a CFTC-regulated exchange, recently valued around $22B.⁴

"Advanced" does not mean "uncontested." The sharpest fight was over election markets: whether a CFTC-regulated venue could list contracts on US election outcomes was litigated, hard, before it resolved in favor of the exchange being allowed to offer them.⁵ That episode is the template for what every market goes through — a real legal argument about whether this specific product is allowed, settled in a specific venue, with a specific answer. The US has more of those answers on the books than anyone. It is the closest thing to a finished tile on the map, and even it is still being colored in.

Europe — fragmented

Europe is where the single-market intuition breaks. The EU does have a unified crypto framework — MiCA, the Markets in Crypto-Assets regulation, which came into force across 2024–2025 and finally gave crypto-assets a common European rulebook.⁶ The trap is assuming MiCA covers prediction markets. Largely, it does not. An event contract is not obviously a crypto-asset, and in much of Europe it lands instead under national gambling regimes — which are not harmonized at the EU level and differ sharply country to country.

The cleanest example is the United Kingdom, which regulates betting through the Gambling Commission.⁷ Whether a given event contract is a financial instrument (one regulator, one rulebook) or a bet (a different regulator, a different rulebook) is precisely the kind of line that varies by jurisdiction — and the answer determines who can offer it, to whom, and under what conditions. Across the rest of Europe the picture is genuinely mixed: some national regulators may treat these instruments through a financial lens, others through a gambling lens, and for many the status is simply not yet settled. I won't pretend to a country-by-country license map for Europe, because an honest one does not exist yet. The reliable statement is the structural one: Europe is fragmented, the relevant authority is usually national rather than pan-European, and the financial-versus-gambling question is live in a lot of places at once.

Asia — highly varied

Asia is the widest spread on the map — the same continent contains both the hardest bans and some of the largest latent demand. Several jurisdictions prohibit most forms of betting and online wagering outright, and an event-contract venue would run straight into those prohibitions. Others are conservative but not closed: financial centers like Singapore and markets like Japan run tightly-regulated regimes where novel instruments are approached cautiously, through existing financial and gambling law, rather than waved in.

I am deliberately not assigning specific licenses to specific Asian countries here, because the regimes are varied, fast-moving, and in several cases unsettled enough that any precise claim would be a guess dressed up as a fact. What is safe to say is the shape: posture varies enormously — outright bans in some places, cautious frameworks in others — while the underlying demand across the region is large and, today, mostly served offshore. That gap, between heavy latent demand and little licensed supply, is the whole reason the region matters to anyone building in this space.

Emerging markets — the open field

This is where the map is most blank, and therefore most interesting. Across much of the emerging world there is no settled regime for event contracts at all — not a ban, not a license, just an absence. Demand is present (often large and young); supply is offshore; and over the coming years many of these markets will have to decide how to regulate the category, because the activity is already happening and a "no rule" state is unstable. That decision window is the opportunity. In a market that has not yet drawn its lines, a licensing-first operator — one that shows up early, builds surveillance and settlement in from the first trade, and works with the regulator to define the rules rather than around them — can win the position before the rules harden.

Reading the map

Here is the whole thing as one picture: four regions, the primary regulator or framework in each, and an honest posture label. The colors are deliberate — accent for where a licensed path clearly exists, amber for contested or gambling-framed, red for grey or banned. Note how little of the map is solid blue.

One honest reading of that figure: exactly one row is solid blue. The category's home market is settled; everywhere else ranges from "fragmented and contested" to "blank." That is not a weakness of the thesis — it is the thesis. A finished map has no land left to claim.

The thesis: a land grab

Now combine the two forces this piece opened with. Winner-take-most says that within any single market, liquidity concentrates onto one venue.¹ Per-country licensing says that "any single market" is a country with its own gate, and you have to be let through that gate to compete at all.² Multiply them and you get the structure of the whole opportunity: a land grab, where the first venue to be both licensed and liquid in a given market is very hard to dislodge — because by the time a second venue clears the same regulator, the liquidity has already pooled with the first, and depth is the one advantage you cannot simply copy.

The sequence is the same in every market, which is why it's worth drawing as a funnel. First the asset class gets legitimized — a regulator decides event contracts are a thing it will permit, on terms. Then an operator wins the license for that specific market. Then liquidity concentrates onto the first credible licensed venue, because traders go where the depth is and depth begets depth. And then that lead becomes durable: the network effect that made the venue deep also makes it nearly impossible to displace. Legitimize → license → concentrate → hold. Run that loop once per country, and the map fills in one tile at a time.

First licensed venue Everyone else

This is, not coincidentally, the playbook crypto exchanges ran a decade ago: the asset class got legitimized somewhere, then every market opened it one regulator at a time, and the first licensed venue in each tended to take most of the liquidity and keep it. The deeper mechanics of that pattern — why it concentrates, what the natural experiments show — are their own piece, and so is the network-effect engine that makes winner-take-most the rule rather than the exception. This piece is just what happens when you lay those two pieces over a globe.

What's hard — and uncertain

I have been hedging the whole way through, on purpose, because the honest version of this thesis comes with large caveats. Here are the real ones, stated plainly.

Regulation is slow, and fragmented. A license is not a product launch you control. It moves at the pace of a regulator's calendar, can take years, and has to be won separately in every market — there is no single approval that unlocks the world. The map fills in slowly, if it fills in at all, and a business built on it has to survive that pace.

It is politically fraught. Two flashpoints recur. The first is gambling: many regulators look at event contracts and see betting, full stop — and given that sports has been roughly 80% of the volume on the leading US venue since these markets launched, that view is not a misreading of what people actually trade today.⁸ The argument that the mechanism is a forecasting tool, not a casino, is real and it's the one I believe — but it runs straight into the lived fact that most of the volume looks like a sportsbook. Any honest map has to hold both. The second flashpoint is election markets, which are politically sensitive everywhere and were contested even in the US before they resolved.⁵ A regulator nervous about either of these can slow a market to a crawl or shut the door.

Some markets will simply ban it — and mean it. The onshore-beats-offshore logic is a tendency, not a law. Plenty of jurisdictions will decide the cleanest answer is prohibition, and they are entitled to; the map will have permanent red on it. A thesis about direction is not a guarantee about any one country.

So take this for exactly what it is. The map is being drawn in real time, most of it is still blank, and the parts that aren't are often contested. I am confident about the shape — winner-take-most plus per-country licensing makes this a land grab, and the early, licensed, liquid venue in each market has a structural edge. I am not confident about the schedule, the boundaries, or which specific tiles ever turn blue. That's the honest position: a thesis about direction, not a finished atlas — and the surest thing on the page is that the version a year from now will look different.