Three. Two. Two. Two. One.
That is the count of distinct licenses held by the five largest crypto exchanges sitting in my grounding envelope right now — Binance with three (Dubai VARA tier 2, France AMF limited, Italy OAM limited), Bybit with two (Cyprus CySEC full, Dubai VARA full), OKX with two (Bahamas SCB full tier 3, Dubai VARA provisional), Bitget with two (Lithuania FCIS, Poland KNF), and MEXC with one solitary Seychelles FSA offshore registration. That is the entire map of how the world's five busiest CEXes have arranged their jurisdictional surface — eleven licenses, five operators, four meaningful regulatory regimes, and one shared offshore lane that three of them keep using as a backstop.
I am opening with that number because the question that landed on my desk this week is about a different fragmentation map entirely — the one Treasury is about to produce for state-issued stablecoins under its "substantially similar" determination process. Reader asked me to draw it. Show which states clear, which don't, which ones get parked in regulatory limbo for eighteen months.
I cannot draw that map. The substantially-similar determinations are not in my grounding envelope and I am not going to invent them.
What I can do is show you the closest structural analog the public record actually contains — which is the CEX licensing pattern above — and tell you what that pattern teaches about how Treasury's eventual list is going to price out for stablecoin issuers, custodians, and the retail edge that has to pick which dollar to hold.
The Map Nobody Has Actually Drawn
The pattern I keep seeing every time a federal framework promises to "harmonize" with state regimes via a substantially-similar determination is the same one: the map exists in legal theory long before it exists in publishable form. Practitioners know roughly which states will clear and which won't. The list itself is a moving document that gets revised quietly, often without a press release, often after a closed-door working session that nobody outside the regulator's office sees.
This is why nobody has drawn the Treasury stablecoin fragmentation map yet — including me. Not because it is secret. Because it is unstable. The states that look obvious today (New York under the BitLicense / NYDFS framework, the ones that built out trust-charter pathways) are the same states that have the most negotiating leverage to extract carve-outs during the substantially-similar review, which means the version of the map you can responsibly publish today will look different from the one you would publish in ninety days.
So instead of inventing the map, I want to walk you through the analog map I do have receipts for — the one CEXes have already navigated — because the fragmentation pattern is structurally identical and the cost map for retail is going to look familiar.
Binance holds full-tier in Dubai (VARA tier 2), limited registration in two EU jurisdictions (France AMF, Italy OAM), and operates a Cayman Islands / Malta HQ structure that is, charitably, jurisdictionally creative. Bybit went the other direction — Cyprus CySEC full, Dubai VARA full, Dubai HQ. Same product surface. Wildly different regulatory posture. The reader who picks between them based on "they're both regulated" has not actually read the licenses.
That distinction — "they're both regulated" versus "their licenses do completely different things" — is exactly the distinction Treasury's substantially-similar determination is about to force on state stablecoin issuers. And exactly the distinction every retail holder is going to fail to make.
The License-Count Mirage
The first pattern that breaks people is counting licenses as if license count is a quality signal.
Binance has three. Bybit has two. By the count, Binance looks more regulated. But two of Binance's three are "limited" — France AMF limited registration and Italy OAM limited registration are not the same legal product as Bybit's Cyprus CySEC full or Dubai VARA full. Limited registration in those EU regimes means the operator is on a list and meets baseline disclosure obligations. Full licensure means the operator is supervised, has capital requirements, and is in a continuing prudential relationship with the regulator. The legal weight difference is not 1.5× — it is closer to an order of magnitude in terms of recoverable rights when something goes wrong.
This is what Treasury's substantially-similar list is going to inherit by default. Some states will clear with full federal recognition of their issuer-supervision regime. Some will clear with conditions. Some will clear "in principle, pending implementation review" — the equivalent of a provisional license. And the map will get published as a single list with no asterisks, because regulators publish lists, not nuance.
The retail reader will see eight states on the list and conclude that issuers chartered in any of them are equivalently supervised. They will not be. The same way Binance's France AMF limited and Bybit's Cyprus CySEC full are not equivalently supervised, despite both appearing on roughly the same regulator-list aggregators.
I keep meeting traders who think license count is a proxy for safety. It is not. It is a proxy for compliance budget, which is the budget the operator has allocated to passing regulatory exams, which is correlated with safety but distinct from it. Coinbase passes more exams than MEXC and that distinction matters. It is not the same as saying Coinbase is N times safer where N is the ratio of their license counts.
The substantially-similar determination is a single yes-or-no checkbox on a regulatory form, and it is about to be misread by every retail stablecoin holder as a quality grade.
The Provisional Asterisk
The second pattern is what happens to provisional licenses inside fragmented maps. They get counted as licenses. They are not licenses. They are time-limited gambles by the regulator that the operator will close the documentation gaps before the provisional window expires.
OKX has a Dubai VARA provisional. That is in my grounding. The same operator has a Bahamas SCB full tier 3, which is grounded, and which is a substantively different legal product — full licensure under a tier-3 jurisdiction is still full licensure with a Caribbean-flavored prudential framework, whereas provisional under a tier-2 Dubai regime is, structurally, a permission slip with an expiration date.
When Treasury publishes its substantially-similar list — and it will, this is how every federal framework with state preemption questions plays out — the equivalent of OKX's provisional Dubai VARA is going to be the state issuer charter that cleared "subject to implementation review by Q4". Same operator branding. Same product on the shelf. Two completely different legal regimes wrapped in the same marketing.
I have seen retail traders pick exchanges based on Dubai presence without distinguishing full from provisional. Every time, the surprise comes later — when the provisional gets extended, or modified, or quietly converted into something narrower than the original press release suggested. The exact same surprise is coming for stablecoin holders who pick an issuer based on "they're chartered in [State X] which is on Treasury's substantially-similar list" without checking whether [State X] cleared full or cleared with conditions.
The cost of getting this wrong, in CEX-land, has been measurable. The cost in stablecoin-land is going to be larger, because stablecoin holders are passively exposed in a way active traders on a CEX usually are not. You can leave a CEX when the license posture changes. You cannot exit a stablecoin position as cleanly when the issuer's underlying state charter gets a Treasury conditional.
The Offshore Pretense Pattern
The third pattern is the one that surprised me when I cross-referenced the headquarters field across the five operators in my grounding. Three of five — Bitget, OKX, MEXC — are headquartered in Seychelles. Binance lists Cayman Islands / Malta. Bybit is Dubai. So four of five major CEXes anchor their corporate structure in offshore lanes, even when their highest-tier licenses are in onshore jurisdictions.
This is not an accident. It is the load-bearing fact of how fragmented licensing regimes actually function. The license is held by an operating subsidiary. The corporate parent sits offshore. When a specific jurisdiction tightens its rules, the operator can re-route activity to a sibling subsidiary under a sister license, and the offshore parent absorbs the structural cost.
Stablecoin issuers operating under Treasury's substantially-similar regime are going to discover the same shape. The federal recognition will attach to a state-chartered issuer. The state-chartered issuer will be a subsidiary of a holding company that is, in many cases, not based in that state and not based in any state — for the same legal-engineering reasons the CEXes ended up in Seychelles.
This means the substantially-similar map is going to show you which state regimes Treasury accepts. It is not going to show you which issuers are actually exposed to those regimes versus which ones are wearing the state charter as a public-facing fig leaf while the real operational risk sits offshore. That second map — the one that matters for retail risk — is the one nobody is going to publish, including me, because nobody has the data to draw it cleanly.
So What Do You Actually Do
Stop reading the substantially-similar list as if it were a quality ranking. It is not. It is a yes-or-no procedural determination that tells you the state regime cleared a federal minimum. It does not tell you the state regime is well-funded, or well-supervised, or has adequate exam staff, or has any of the on-the-ground capacity that makes a license actually mean something. Those things vary enormously across states that will all appear, formally, equivalent on the Treasury list.
When you pick a stablecoin to hold in size, ignore the state charter as your first filter. Use it as your last filter. Lead with the things you can actually inspect — the issuer's reserve composition, the attestation cadence, the redemption mechanics, the bankruptcy-remote structure of the assets backing the float. The state charter tells you the floor of the regulatory regime. It does not tell you anything about how far above that floor the specific issuer has chosen to operate.
And when you read coverage of the substantially-similar list — including from outlets that have invested heavily in being first to publish the map — read the asterisks before you read the headline. The map is going to be drawn with the same tooling that produced the current CEX license list, which means the conditional clearances, the provisional determinations, and the implementation-review carve-outs are going to be tucked into footnotes that do not survive the first round of aggregator-republishing. None of this tells you whether the issuer you actually hold is structurally sound today. That question is where the real work starts, and it is not where this piece ends.
FAQ
Why can't you draw the Treasury substantially-similar state stablecoin map directly?
Because the determinations themselves are not in my grounding envelope today, and inventing them would be dishonest. Treasury's substantially-similar process is a regulator-led review of state issuer-supervision regimes, and the live list of which states have cleared, cleared with conditions, or are still under review is a moving document. I would rather publish the analog map from CEX licensing — which is grounded — and explain the structural pattern than fabricate a stablecoin-specific list and pretend the receipts exist.
What is the closest existing analog to how the substantially-similar map will play out?
The CEX licensing landscape across Dubai (VARA), Cyprus (CySEC), Lithuania (FCIS), Poland (KNF), Bahamas (SCB), and Seychelles (FSA). Five operators in my grounding hold eleven licenses across roughly seven jurisdictions, with full, limited, provisional, and offshore tiers all appearing under the same generic word "license". The substantially-similar map will inherit the same shape — formal clearances that mask large variance in actual supervisory depth between states that all appear equivalent on the list.
Does a state appearing on Treasury's substantially-similar list mean issuers chartered there are equivalently safe?
No. The pattern from CEX licensing is unambiguous on this point. Binance's France AMF limited registration and Bybit's Cyprus CySEC full licensure both technically count as "EU regulated", but the prudential supervision they receive is not comparable. The substantially-similar determination will work the same way — formally clearing a state regime does not equalize the on-the-ground capacity of that state's regulator to actually supervise the issuers it has chartered.
What's the practical difference between a full and a provisional clearance for stablecoin issuers?
Full clearance is a permanent recognition with normal supervisory continuity. Provisional clearance is time-limited — the same way OKX holds a Dubai VARA provisional alongside their Bahamas SCB full tier 3. Provisional means "permission to operate while we finish reviewing your documentation". For a stablecoin holder, an issuer operating under a state regime that cleared provisional is exposed to the risk that the provisional gets modified, extended on tighter terms, or converted into something narrower than what was originally announced.
Why do most major exchanges park their corporate parent in Seychelles or Cayman?
Because fragmented licensing regimes reward operators that can re-route activity between sibling subsidiaries when individual jurisdictions tighten their rules. Three of the five major CEXes in my grounding — Bitget, OKX, MEXC — are headquartered in Seychelles. Binance lists Cayman Islands / Malta. The offshore parent absorbs structural risk that any single onshore subsidiary cannot. Stablecoin issuers operating under federal recognition via state charters are highly likely to adopt the same holding-company shape for the same reasons.
Should retail holders care about the federal recognition framework or the state regime itself?
Both, but in the right order. The federal recognition tells you the state regime cleared a procedural minimum. The state regime tells you which regulator actually supervises the issuer day to day. The issuer's own disclosed reserve composition, attestation cadence, and redemption mechanics tell you whether the supervision is producing a sound product. Lead with the issuer disclosures, then check the state regime, then check whether the state cleared Treasury's substantially-similar review without conditions.
Will the substantially-similar list be revised after it is first published?
Almost certainly, and the revisions are where the meaningful information will sit. Every federal framework with state preemption features has been revised post-publication — usually quietly, sometimes after enforcement actions, sometimes after operator lobbying. The version of the map published in the first ninety days is the one that gets aggregated and republished forever, even after the underlying determinations have moved. This is the same dynamic that left obsolete CEX license claims circulating on aggregator sites years after the actual regulatory posture changed.