Stablecoin supply fell $10B while volume hit a record. Read the split.

The most-quoted stablecoin number this month is the wrong one. Total supply fell about $10 billion since its May peak, a 3% drop and the largest in dollar terms since the Terra collapse in 2022. Most coverage read that as crypto liquidity draining out. Then June transaction volume hit a record $1.79 trillion, up 63% from May and 125% from a year earlier. Both numbers are real. They point in opposite directions because they measure two different businesses.

Supply is the float: the dollars parked in an issuer reserve, earning Treasury yield the issuer keeps. Volume is the flow: the dollars actually moving through the rails, settling payments and trades. For most of stablecoin history the two rose together, so people used market cap as the single scoreboard. In June they split. The float shrank while the flow set a record. And each side has a different leader. USDT holds about 59% of supply. USDC moved 67% of June volume. The biggest balance sheet and the busiest rail are now two different companies.

The float shrank, and it costs real money

Total stablecoin supply ended June near $312 billion, down $7.7 billion on the month and about $10 billion from the May record. Tether USDT slipped from roughly $190 billion in May to about $184 billion. Circle USDC fell from a March peak just under $80 billion to around $73 billion. On a percentage basis the drop is small, 3% against the 26% wipeout of the 2022 bear market, and analysts have called it a pullback rather than a crisis.

Small in percent does not mean small in dollars. The float is where issuers make their money. An issuer mints tokens on demand, holds the backing dollars in short-dated Treasury bills and overnight repo, pays holders nothing, and keeps the yield. At a Fed funds rate near 4 to 5%, that float is a printing press. Tether earned more than $10 billion in profit in 2025 on about $187 billion in reserves. Circle, on a smaller base near $76 billion, earns roughly $3 billion a year the same way.

Do the arithmetic on a redemption. Every billion dollars that leaves the float takes about $40 to $50 million a year of reserve income with it. A $10 billion contraction is not a liquidity footnote. It is a direct cut to the money machine that funds both companies. This is why who gets to issue under the GENIUS Act matters so much: the license is a license to earn Treasury yield on other people money.

The flow set a record, and USDC owns it

June moved a record $1.79 trillion in adjusted volume, narrowly past the $1.78 trillion set in February, up 63% from May and 125% year over year. The word adjusted matters. The figure comes from Visa Onchain Analytics, built with Allium, Artemis, and Castle Island, and it strips out trading bots, exchange rebalancing, and repeated smart-contract loops to approximate real payments and commerce. Raw on-chain volume would be several times larger. This is an estimate of organic activity, not an audited figure like reserves, and it belongs in a different column from the float numbers above.

On that adjusted basis, USDC did 67% of June volume, about $1.21 trillion, mostly on Solana and Base. USDT did 32%, about $576 billion. Across the first half of 2026 the gap is just as wide: USDC around 70% of adjusted volume, USDT around 25%. The token with a quarter of the market cap is doing two-thirds of the actual moving.

The reasons are structural. USDC is US-regulated, integrated with banks, and native to the fast, cheap chains where payment volume lives. The regulatory split helped too: when MiCA pushed USDT off EU exchanges, Circle picked up the compliant rail by default. USDC is built to be spent. USDT is built to be held.

Two kings, two economies

This is the split worth internalizing. USDT is the king of the float: the biggest balance sheet, roughly 59% of supply, and the largest reserve-income stream in the business. Its moat is inertia and dollar demand in emerging markets, where USDT is the cash people actually hold. That is also why USDT is expanding back onto Bitcoin rails: defend the places the float lives.

USDC is the king of the flow: two-thirds of adjusted volume, the settlement layer institutions reach for. Its moat is regulation and banking access, not size. One token wins the balance-sheet game, the other wins the usage game, and in June those games stopped moving together.

The honest caveat: the two crowns rest on different evidence. USDT float is attested reserves. USDC flow is a filtered estimate of organic activity, and estimates can be wrong at the edges. Treat the float as a fact and the flow as a well-built approximation, and the divergence still holds. It is too large to be a rounding error.

Circle bet on the rail, not the float

Circle spent the month acting like a company that has read this split. On July 10 it won final OCC approval to open First National Digital Currency Bank, a national trust bank. It starts with federally regulated custody and is designed to eventually manage the USDC reserve directly, under OCC oversight. The stock jumped on the news.

Circle has a reason to look past the float. It shares up to half of its reserve income with Coinbase and other partners, so its float economics are structurally worse than Tether, which keeps nearly all of its own. So Circle leans into the part it leads: payment-network fees and settlement volume that do not depend on where the Fed sets rates. The trust charter is a flow-side bet. If rates fall, reserve income falls for everyone, and the durable business is being the rail that money runs on rather than the vault it sits in.

What to watch now

The flow is now big enough that incumbents are defending it. Wall Street has started pushing back on the trillion-dollar stablecoin story, and banks are building their own always-on settlement networks in response. You do not build a defense against a rail that no one uses.

Two things decide how this split resolves. The first is rates. As long as Treasury yield sits near 4 to 5%, the float stays lucrative and USDT reserve-income lead is worth defending. Cut rates and the flow business becomes the whole story, which is the outcome Circle is building toward. The second is whether USDT can turn its float lead into rail usage before that happens, or whether USDC rail lead lets it survive a lower-yield world that would gut the pure reserve-income model.

For a reader, the takeaway is a habit. Stop judging a stablecoin issuer by market cap alone. Ask which economy it is winning: the float, where the money is made today, or the flow, where the money moves. In June those two answers stopped being the same company, and that is the more useful number than the $10 billion everyone quoted.

Sources

• Stablecoin market cap has shrunk by $10 billion since May, but analyst sees no reason to panic (CoinDesk)
• Stablecoin market loses $10B as crypto liquidity quietly contracts (crypto.news)
• Visa reports record $1.79T stablecoin transaction volume in June (CryptoBriefing)
• Visa Onchain Analytics reports record $1.79 trillion in adjusted stablecoin volume for June 2026 (Solana Compass)
• Stablecoins are moving more money while crypto cash pile gets smaller (CryptoSlate)
• How do stablecoin issuers make money? (Transak)
• Circle receives final OCC approval to establish national trust bank (Circle)
• Circle gets an OCC bank charter as stablecoin competition heats up (CNBC)
• Wall Street mounts pushback on trillion-dollar stablecoin boom (Bloomberg)

This is not financial advice.