Stable Tape

The independent daily record of US dollar stablecoin supply.

All sections

Every day at 16:00 London, we record how many US dollar stablecoins exist, and publish the Close from 16:30. Once published, a day’s record is never changed.

About Stable Tape

Stable Tape is the independent daily record of US dollar stablecoin supply. Read at 16:00 London. Every source named. Never changed.

About Stable Tape

The independent daily record of US dollar stablecoin supply. Read at 16:00 London. Every source named. Never changed.

Every day at 16:00 London, Stable Tape records what the main sources say about US dollar stablecoins, sets them side by side, and keeps that record unchanged.

What it is

Ask the Federal Reserve how much money there is and it answers: which money? M1 and M2 are both right; they measure different things. Stablecoin supply works the same way. Each source decides what to count, so 2 respected sources can give different totals for the same day.

Stable Tape shows each source’s number under its own name, every day, and never changes a saved day. The mistake is quoting a number without saying whose it is.

What the check has found

  • The sources moved much closer together in October, and almost all of that change came on 13 to 15 October, when the market barely moved.
  • Sources can change without notice. A source can change which stablecoins it counts, or its own past figures, at any time. On 14 and 15 October 8 stablecoins left 1 source’s list and its past figures changed back to 2023. Stable Tape keeps what each source said at 16:00 on the day, unchanged, so the earlier figures stay on the record.
  • A pricing fault, found and reported. We found a pricing fault at a major source that could move its total by hundreds of millions of dollars overnight, and reported it.

Why did the sources change in October? One possibility is that they are tightening what counts as a stablecoin as US rules arrive. We don’t know yet, so we’re asking them, and we’ll publish what they say.

See the evidence on Supply →

How we work

  • 1 time, every day. The Close is read at 16:00 London and published with its evidence.
  • Never rewritten. A mistake is corrected in a new, dated record. The original stays.
  • We remember what the sources said. When a source rewrites its own history, we keep what we saved, with its fingerprint.
  • Never averaged. Each source keeps its own number and its own name.
  • Independent. No one can pay to change a result, and every payer is named. No one who pays us has any say over what we measure or publish, and any payment tied to a stablecoin is said on that stablecoin’s page.

How the record works: dots, sources and Signal →

What comes next

Under the GENIUS Act, regulated US issuers must publish their own supply every month. The GENIUS Act takes effect on 18 January 2027. We will check each published report against the chain, and every stablecoin sits on 1 of 4 steps.

  1. ObservedIt is in the sources.
  2. TrackedWe’ve confirmed who issues it.
  3. ReportingThe issuer publishes a monthly report. We keep every version.
  4. Proof of SupplyWe set our chain reading beside the issuer’s report. The ring seals when they agree.
  • Reading the chain ourselves. Supply read from each issuer’s own contracts, beside the sources and the issuers’ own reports.
  • Pages for issuers. Claim your stablecoin’s page, add your official links, and get a live badge for your site.
  • Proof of Supply. Starting with the GENIUS Act in 2027. Then the EU’s MiCA rules, and other countries as their rules arrive.
  • Signal. Alerts when supply makes a move that matters, now in testing.

How Proof of Supply works → What the GENIUS Act changes →

Stablecoins are growing up. Help us keep the record.

If you work on stablecoins, research them, regulate them or want to support independent data, we’d like to hear from you.

Support and partner →

About Stable Tape

How the record works.

We remember what the sources said

Every day at 16:00 London, Stable Tape saves each source’s answer exactly as it arrived, with a fingerprint (a hash) that would show if it were ever altered. The Close carries its fingerprint publicly. The full record is not public, so a reader cannot recompute it.

Sources sometimes rewrite their own history. When they do, we keep what we saved. The original stays, and the source’s new version is recorded beside it with the date we saw it. Over time this becomes something no source keeps for itself: what each one said on each day, and when and how it later changed its mind.

Comparing old and new versions automatically is being built now. Until it is public, changes such as the October list change are written up by hand on Supply.

Why our charts use dots

Most charts join each reading to the next with a line. A line makes 2 claims: that each reading is right, and that the movement between them is real. For stablecoin supply, nobody can promise either, because measuring it is much harder than it looks.

What makes a reading uncertain, and when a line is earned

What makes a reading uncertain

  • Price. A stablecoin can be counted at $1 or at its market price. When a source loses a stablecoin’s price, it may count every token at $1 again. In October, that kind of fault could move a total by hundreds of millions of dollars overnight. Stablecoins that have lost their peg can still be counted.
  • Supply. Sources differ on which tokens are really in circulation. Some count tokens the issuer still holds itself, or stablecoins that have stopped working.
  • Copies across chains. The same stablecoin can live on many blockchains, and bridged or wrapped copies can be counted twice.
  • What counts as a stablecoin. Each source decides whether to include tokens that pay a return, synthetic dollars or stablecoins tied to other currencies.
  • The list changes. Sources add and remove stablecoins, often without notice. On 14 and 15 October, a change to 1 source’s list looked like a fall of billions.
  • History is revised. Sources rewrite their own past figures. On those same days, that source changed its figures back to 2023.
  • Timing. Sources update at different moments, and a total can swing from hour to hour. We read each one within minutes of 16:00 London.
  • Capture. A source can be slow, return an error or change its format at the moment we read it. When a reading fails, we leave a gap rather than fill it.

So each dot is exactly what a source said at the Close, kept as it was. We don’t join the dots with lines. Where we can show that a change is like for like, we say so in words.

When a line is earned

Our own chain readings will be different. Each one counts the tokens on an issuer’s own contracts at a named block, so anyone can check it. And every change from one day to the next can be matched to the tokens created and destroyed in between. When every step is proven like that, the readings can be joined with a line.

The line only ever means what was on the chain. What the law counts as a stablecoin’s supply is a separate question, and different regulators may answer it differently. So an issuer’s own report stays a dot beside the line, labelled with the rule it follows. If our method changes, the line stops, a dated marker shows why, and a new line begins. The old line stays as it was.

Signal, now in testing

Signal is an alert for when stablecoin supply makes a move that matters. Most days supply drifts a little and there’s nothing to say. Signal stays quiet until supply breaks clearly out of its recent range, then tells you, with the numbers behind it.

How we know it works

Its rules are fixed in advance, so nobody decides after the fact what counts.

Why we think it works

An independent test went through the first 2 months of the record, from June to mid August 2026, without looking at Signal’s answers first.

  • It knows when to stay quiet. Across 50 daily readings it spoke 3 times. Of the 16 moves that looked big, it correctly let 10 pass.
  • The check protects it. On 13 August a faulty reading showed a jump of $85bn. We withdrew that reading, so Signal stayed quiet. Without the check, it would have sent a false alarm.
  • The test found a weakness too. One of the 3 alerts came from a pricing fault at a source. Supply had not really moved. That’s the kind of fault our daily check now catches, and fixing it comes before Signal goes public.

Sign up for the weekly check to get early access while we test it.