The crypto-exchange playbook, again

If you watched crypto exchanges grow up, the last two years of prediction markets will feel like a rerun. A product lives for years in a grey zone — clever, faintly disreputable, run mostly offshore. Then a U.S. regulator draws a line and says what the thing legally is. The moment it becomes legitimate, the rest of the world stops ignoring it and starts licensing it — country by country, one regulator at a time. And in each market that opens, the first venue to show up with a license tends to capture most of the volume and then prove almost impossible to dislodge.

That is not a story about prediction markets. It is the story of crypto exchanges, and before that of regulated derivatives, and of stock exchanges before that. Prediction markets are just the newest asset class to run it. This piece is about why the pattern repeats — and why, if it holds, the thing worth owning is not the matching engine. It's the license.

The playbook, in five moves

Strip the specifics away and the pattern is mechanical. It runs in the same order every time.

One: a new asset class appears in a grey zone. Something genuinely new starts trading before anyone has decided what it is. Bitcoin traded for years before a regulator would call it anything. Event contracts traded on Intrade and offshore books long before they were legal at scale in the U.S. The product works; its legal status is a question mark; serious capital stays on the sidelines because you cannot build on a question mark.

Two: the US legitimizes it first. A U.S. regulator is usually the one that ends the ambiguity, because the U.S. has the deepest capital markets and the most-watched rulebook. Crypto spent the 2010s being slowly defined by the SEC and CFTC — what's a security, what's a commodity, who may list it. For event contracts the moment was cleaner and earlier: in 2020 the CFTC recognized them as a regulated asset class,¹ turning a grey-zone wager into a licensed financial product. (I traced that decision in detail in The CFTC and the birth of an asset class.) Legitimacy doesn't make the product better. It makes it bankable.

Three: once legitimized, every country regulates it — one at a time. A U.S. green light is a starting gun, not a global rulebook. Every other jurisdiction still has to decide for itself: its own statute, its own licensing regime, its own supervisor. This is the slow part, and it never happens all at once. It took years for crypto exchanges to assemble the patchwork — a BitLicense in New York, a payments license in Singapore, registration in Japan, MiCA across the EU. Each market opens on its own clock.

Four: exchanges are winner-take-most. An exchange is not a normal business — it's a liquidity network, and liquidity is reflexive. Traders go where the order book is deepest because depth means tight spreads and reliable fills; and the book is deepest where the traders already are. The advantage compounds on itself.² Volume begets tighter spreads begets more volume. In a market with that structure, leads don't erode — they widen, and second place is a long way back.

Five: the first licensed venue in each market wins, and stays won. Put moves three and four together and the conclusion is forced. In each jurisdiction, the venue that arrives first with a license starts the liquidity flywheel before anyone else is legally allowed to spin one. By the time a competitor clears the same regulator, the incumbent's book is already the deepest in the country — and the network effect that made it deep now defends it. First-and-licensed is a very hard position to take from.

We have run this script before

The cleanest precedent is the one most people lived through: crypto exchanges. Bitcoin began as the ultimate grey-zone asset — borderless, permissionless, regulated by no one. Then the U.S. started drawing lines, and a telling thing happened. The venue that won the American market was not the most permissionless or the cheapest. It was Coinbase — the one that chose, early and deliberately, to operate inside the U.S. regulatory perimeter, collecting money-transmitter licenses state by state and courting oversight that its offshore rivals treated as a threat. Being regulated-first was slower and more expensive. It also made Coinbase the exchange American institutions and ordinary users trusted, and it became the U.S. leader and the first crypto company in the S&P 500.³ Compliance wasn't a tax on the business. It was the moat.

Now zoom out from the U.S., because the global picture is where the playbook really shows its hand. Crypto did not consolidate into one planetary exchange. It fractured along regulatory borders. There was a global, lightly-regulated venue — Binance — that chased volume everywhere at once. But inside individual regulated markets, local licensed champions emerged and dominated their home turf. The textbook case is Korea: a tightly-regulated market where domestic rules pushed trading onshore, and a single licensed exchange, Upbit, came to handle the overwhelming majority of the country's volume — a share offshore venues simply could not contest from outside the perimeter.⁴ One asset class, but the map of winners was drawn by who held which license where.

That is the shape to keep in mind: a regulated-first leader in the anchor market, a global lightly-regulated alternative, and a patchwork of per-country licensed champions underneath. It is not unique to crypto. Regulated derivatives consolidated onto a handful of licensed exchanges; national stock markets each have their dominant venue. Whenever a tradable asset class becomes regulated, the market organizes itself around licenses, and the network effects of liquidity make each licensed incumbent hard to move. Prediction markets are simply the next entry in a long list.

The same map, redrawn for event contracts

Hold the crypto template up against prediction markets today and the correspondence is almost uncomfortable.

The regulator that legitimized the asset class is, once again, an American one — the CFTC, which has supervised event contracts as a recognized class since 2020. The regulated-first U.S. leader is Kalshi, the first federally-licensed event exchange, which chose the slow path of operating inside CFTC oversight rather than offshore — and is now valued around $22B.⁵ The global, on-chain, lighter-touch alternative is Polymarket — the Binance-shaped role in this analogy: bigger reach, looser perimeter, and now wired into traditional finance through the Intercontinental Exchange, owner of the NYSE, which has committed up to $2B — not to gamble, but to become the global distributor of Polymarket's event data.⁶ And the per-country licensed champions — the Upbits of prediction markets — mostly do not exist yet. That is the part of the map that is still blank.

Crypto exchanges

The script, last time

• Regulator that legitimized itSEC / CFTC (US)
• Regulated-first US leaderCoinbase
• Global / lighter-touch venueBinance
• Per-country licensed championUpbit — Korea

Prediction markets

The script, again

• Regulator that legitimized itCFTC (US) — 2020
• Regulated-first US leaderKalshi — ~$22B
• Global / on-chain venuePolymarket — ICE-backed for data
• Per-country licensed champion— still unclaimed —

That blank row is the entire thesis. The U.S. moves of the playbook have already happened — the asset class is legitimate, the regulated-first leader is built, the global alternative is funded by the people who own the New York Stock Exchange. What has not happened is stages three and four playing out across the rest of the world: every other market deciding to license event contracts, and a first licensed venue claiming each one. If the pattern holds, those seats get filled the same way they did in crypto — by whoever shows up first with the compliance to operate inside the local rules.

Why the moat is the license, not the tech

It's worth being blunt about what this implies, because it cuts against startup instinct. The matching engine, the order book, the market-maker, the resolution pipeline — all of it is well-understood and, frankly, copyable. Nothing in the mechanism of a prediction market is a durable secret; I've spent most of this series explaining exactly how it works, in public. If the technology were the moat, the category would be a commodity race to the lowest fee.

But the binding constraint isn't technical. In most countries, a venue that pays out real money on real-world events is a licensed financial product, and the right to operate one is granted one jurisdiction at a time — slowly, with capital requirements, surveillance obligations, and someone accountable for the integrity of the book. That right does not transfer across borders, it can't be forked, and it can't be shipped in a sprint. It is the one input a competitor cannot simply clone. Which is why, in every prior run of this playbook, the durable winners were the ones who treated regulation as the product surface — and why I'd bet the same is true here.

What's hard about this bet

I want to be honest about the ways this thesis can be wrong, because "it rhymes with crypto" is an argument, not a proof.

Regulation is per-country and slow. The same fragmentation that creates the opportunity is the thing that makes it grueling. Every market is its own multi-year campaign — a fresh statute to read, a fresh supervisor to satisfy, fresh capital to lock up — and many jurisdictions will treat event contracts as gambling, or simply refuse to act for years. The prize is real; the path to it is measured in regulatory cycles, not product sprints.

Licenses don't transfer. This cuts both ways. It's what makes a won market defensible — and it's what makes the next market start from zero. There is no continental shortcut, no single approval that unlocks a region. Winning ten markets means running the gauntlet ten times. A strategy built on this has to be comfortable compounding slowly.

The incumbents will fight for it. These seats are not unguarded. The established derivatives exchanges — CME and Cboe — are rolling out their own event-contract products,⁵ and they already hold licenses, balance sheets, and regulator relationships in dozens of countries. Crypto-native venues will push in from the other side, armed with global liquidity and on-chain rails. A newcomer is racing both a well-capitalized traditional bloc and a fast crypto one for the same per-country licenses.

And the honest caveat that runs through this whole series: most of what trades on these venues today is sports — roughly 80% of Kalshi's volume since contracts launched in mid-2024.⁵ The license thesis does not depend on pretending otherwise. It rests on the claim that the mechanism — a crowd-priced, real-time probability, and the hedging and data layer built on it — is where the durable value sits, and that the sports volume is the on-ramp that funds getting there. ICE paid up to two billion dollars for the data feed, not for a sportsbook. But if you'd told me the category was pure forecasting infrastructure today, I'd tell you to look at the order book.

Why this is the bet I'm making

I'll keep this short, because the point of the piece is the pattern, not the pitch. This is the exact playbook we're running at Seeker: treat the license as the product, win it one market at a time, and build the surveillance and settlement in from the first trade rather than bolting them on after. We're starting with Vietnam — a launchpad, not the destination — for the same reason Upbit started in Korea: a single licensed venue inside a regulated perimeter is a position the offshore world can't contest from outside it. The MVP and demo are live; the license is the goal, not a thing we already hold. The wager is simply that prediction markets run the crypto-exchange playbook again, and that the venue people trust in each market is the one that showed up early enough to win the license rather than route around it.⁶

You don't have to care about any one company to find the pattern interesting. A new asset class became legitimate in the U.S.; the rest of the world will spend the next several years deciding how to license it; and in a winner-take-most market, the first venue through each door tends to keep the room. We have seen this movie. The only open question is who plays the lead in each market — and most of those parts haven't been cast.