Every argument in this industry eventually reaches for the same number, so once a quarter I like to take it apart and see what’s actually inside. Total U.S. dollar stablecoin supply stands near $309 billion this week, according to DefiLlama.
Before anything else, definitions, because sloppy method is endemic in this corner of the data. Supply means circulating tokens at par, excluding wrapped and bridged duplicates that would double-count the same dollars. It is not market capitalization, a term that adds nothing to an asset pinned at a dollar. Volume means reported 24-hour exchange turnover, from CoinGecko, and it measures trading, not payments. The two numbers answer different questions and belong in different sentences.
The tape
Total supply near $309 billion, up roughly $3.4 billion over 30 days. Reported daily volume in the mid-$90 billions. USDT around $183.4 billion, USDC near $74.6 billion. The middle tier: USDS about $6.6 billion, DAI $4.8 billion, USDe $4.3 billion, USD1 $4.2 billion, USDG $3.2 billion, PYUSD near $2.7 billion after its summer drift. The tokenized-fund adjacents, BUIDL and USYC, sit around $2.8 billion and $2.7 billion. Euro-denominated tokens, all issuers combined, stay below $2 billion.
Do the dominance arithmetic and the top two hold roughly 83%, the top five pass 90%, and everything below tenth place sums to less than USDC alone. Whatever the licensing era redistributes, it starts from that table.
Three reading rules keep the tape honest, and I hold myself to them. Rankings below the top five reshuffle on flows that would not move the leaders a basis point, so mid-table position changes are noise until sustained across quarters. Bridge inventories and exchange cold wallets wobble daily prints, so 30 days is the shortest window I’ll editorialize on. And issuer-published numbers run on their own calendars, so the tape sometimes disagrees with a press release by a billion; where it matters, I’ll cite both and say which clock I’m reading.
The duopoly that will not move
Here’s what strikes me. USDT at 59% and USDC at 24% have now weathered a year in which both were supposedly under siege: consortium tokens, bank ventures, yield-bearing challengers, three continents of new law. The combined share barely moved.
The mechanism is boring and decisive. Liquidity begets liquidity. The pair everyone quotes is the pair everyone must hold, exchanges and market makers and corridors standardize on the deepest books, and depth compounds. Every challenger of the past five years has discovered that distribution deals move single-digit billions while the network effects move nothing. I keep waiting for the exception. It keeps not arriving.
This month’s net growth decomposed the way it has all year: concentrated in the leader, with USDT minting through weekends in its emerging-market corridors, USDC’s float oscillating with institutional flows and increasingly tied to network usage rather than exchange balances, and the middle tier netting roughly flat as PYUSD’s decline offset USDG’s and USD1’s accumulation. The category’s net growth is currently a Tether statistic with a Circle garnish.
The middle tier, name by name
The stalls have different causes, which matters more than the similar sizes. USDS and DAI serve a constituency no licensing regime can, collateralized governance-run dollars, and their combined $11 billion is best read as that constituency’s structural size, stable and unexpansive. USDe is the yield experiment, its float tracking the profitability of the basis trade behind it, and the Ethena Pay neobank launch is an attempt to bolt consumer distribution onto a markets-driven balance.
USD1’s fast arrival into the $4 billion range shows what political adjacency and exchange listings buy quickly, and what they haven’t bought yet, which is institutional treasury adoption. USDG is my sleeper pick: Paxos’ consortium dollar is small in float but accumulating exactly the licensed, bank-adjacent distribution the new regimes reward, Singapore issuance, European clearing-bank listings.
And PYUSD remains the market’s standing lesson that a hundred million users and a token in their app are different things. None of the six is failing. All six keep discovering that the duopoly’s moat is liquidity, and liquidity has no shortcut.
The boundary problem
A word on category hygiene, because the trackers now list instruments that stretch it. BUIDL and USYC are tokenized money-market funds: they pay yield, restrict transfer and are stablecoins only in the loosest chart-making sense. Their growth is real, the institutional cash-management market moving on-chain, and it was faster in percentage terms this month than any payment token’s.
But folding them into stablecoin supply muddies the category’s defining trade. Payment stablecoins under every 2026 regime hold full reserves and pay holders nothing, while tokenized funds pay yield and are securities. The convergence of the two lists tells you where yield went after the interest bans. It tells you nothing about payment-token growth, and the boundary between the lists is where the next regulatory fight sits.
The run rate against the forecasts
Now the back-of-the-envelope part, which is why we’re all here. Thirty-day growth of roughly 1.1% annualizes near 14%, which puts the market around $350 billion a year from now and $400 billion sometime in 2028.
The forecasts quoted at conferences run to $2 trillion by 2030, a figure popularized in this year’s consortium launch materials. The gap is arithmetic. At 1% a month the market doubles in six years; the forecasts need it to double in under three.
For them to land, growth has to accelerate severalfold on catalysts not yet visible in supply data: the GENIUS era unlocking corporate treasuries, the consortium tokens shipping with real distribution, Samsung-class consumer shelves converting, machine payments arriving on the purpose chains. I don’t have any inside information, and any of those may deliver. The supply curve’s testimony is that none has yet.
Where the dollars sit
Composition tells a different story from the marketing. The largest share of supply still works as trading-market cash: exchange balances, market-maker inventory, derivatives collateral. The payments share, corridor working capital, corporate treasury, card floats, is the growing edge, visible in clearing-bank listings and settlement pilots, and still the minority of the reservoir.
Chains mirror the split. Supply concentrates on Ethereum, the institutional settlement layer, and Tron, the informal economy’s workhorse, whose cheap transfers carry an outsized share of emerging-market flow. Solana’s consumer share climbs, and the purpose-built payment chains, Tempo live since March, Arc arriving mid-September, are only beginning to bid for residence. A year from now this report will carry a chain column worth arguing about.
One more ratio worth keeping: at mid-$90 billions of daily reported volume, the entire supply turns over roughly every three days on exchanges alone, before counting payment traffic that never touches an order book. The quarter’s most consequential flows are the ones this dashboard measures worst: network settlement pilots, clearing-bank conversions, corridor traffic, Circle’s payments network compounding toward $15 billion annualized.
Volume, qualified
The volume figure needs its asterisks stated once. Reported exchange turnover includes venues whose numbers the industry has long treated skeptically, double-counts assets that hop venues intraday and misses entirely the over-the-counter and on-chain payment flows where the category’s actual work happens. I quote it because it’s the only continuously published flow series, and because its trend is informative even where its level is soft.
Better flow measures exist in fragments: on-chain transfer volumes adjusted for internal shuffling, the card networks’ settlement disclosures when they come, corridor operators’ occasional numbers. I’ll assemble them as they mature. Until then, the honest statement is that the industry measures its reservoir well and its rivers badly, and every strong claim about stablecoin payments adoption should be read with that gap in mind.
The euro line, again
The series will keep printing one number until it changes: euro stablecoins, all issuers combined, below $2 billion, a rounding error after four years of MiCA clarity and one of the strangest facts in the category. Euro rates spent a decade making float economics unattractive, European treasurers have less reason to flee their own currency than emerging-market savers do, and dollar network effects operate inside Europe too. A Rotterdam importer paying a Shenzhen supplier prices in dollars regardless.
The line matters now because two of the year’s biggest announcements bet against it. The 21-bank consortium names a euro token as its second product, and bank-issued euro instruments like EURI already sit on European clearing shelves waiting for volume. Either the corporate demand the banks perceive is real and this line finally moves, or Europe’s tokenized-money future runs through the digital euro and dollar tokens, with private euro stablecoins as a footnote. I’ll report the verdict either way.
The chart the category needs and lacks
If a single new data series would improve public understanding, it’s supply split by function: trading balances against payment working capital against treasury holdings. Nobody publishes it because nobody can, cleanly; the same USDC is exchange collateral on Monday and corridor float on Wednesday. Proxies exist, exchange-held share, addresses touching known payment processors, chain mix, and each is partial.
The gap matters because every strategic argument in the industry, the trillion-dollar forecasts included, is secretly a claim about that split. The bulls are claiming the payment share compounds until it dominates. The skeptics are claiming the reservoir stays trading-shaped with payments at the margin. The supply total I audit here cannot arbitrate between them, and readers should notice how often commentary pretends it can.
House rules for this series
Two disciplines, stated once so readers can hold me to them. Price-of-one is not information: a pegged asset oscillating a few basis points around par is microstructure, and 30-day price changes will never appear here as growth. And projections do not become baselines: the $2 trillion figures are scenarios with assumptions, quoted only beside the run-rate arithmetic that tests them.
What the series will do is keep definitions sharp as the category blurs: payment tokens against tokenized funds, supply against flow, circulation against activity. The most useful thing a data column can do in a maturing market is refuse to let the numbers become adjectives.
The calendar
The next report’s diary is already written. Mid-September: Arc’s mainnet, and first residence data for an issuer-owned chain. Mid-October: Singapore’s consultation closes, the first read on how recognition regimes define comparable. Through autumn: Treasury’s comment file on the GENIUS issuance rule, where the definitions deciding USDT’s American surface get argued line by line. Year-end: Hong Kong’s licensed HKD launches, and whatever first disclosures emerge from the consortium ventures.
Most of those events will barely dent the supply curve, and that asymmetry is itself a finding. The reservoir changes slowly, which is what reserves are for. The plumbing around it is changing as fast as it ever has, and plumbing is where the next tape will differ from this one.
What would move the tape
Four dials into year-end. The January 2027 GENIUS effective date, the first hard filter the market has faced, with the share math changing if the largest issuer’s U.S. surface narrows. The consortium tokens, whose launches would test whether owned distribution can dent the liquidity moat. Samsung’s wallet rollout, the largest consumer shelf ever pointed at the category. And the tokenized-fund boundary, where regulators will decide what these tables even count.
So: $309 billion, five-sixths of it in two names, adding a percent a month while the industry finances plans that assume ten times the size inside four years. That tension between the curve and the conviction is the state of stablecoin supply in September 2026, and the next edition lands after Arc’s launch and Singapore’s Oct. 16 consultation close. Small new entries, probably. But this category’s history keeps being written in numbers everyone later wishes had been kept more carefully, so I’ll keep them.