Crystal Intelligence found $4.6 billion less Tether USD (USDT) at 19 exchanges and lenders on Oct. 1 than in early July, even as the token’s circulating supply barely changed.
Crystal’s market-cap data put circulating USDT at about $184 billion on Sept. 30, down $647 million, or 0.4%, from June 30. Its holder data show a narrower shift: the 19 firms whose USDT balances declined added about $1.9 billion in other stablecoins. Those changes do not establish that customers exchanged the departing USDT for rival tokens.
USDT held in wallets Crystal could not link to named holders rose by about $1.5 billion. Crystal said some exchange customers may have moved tokens into their own wallets, though the data do not account for every change at the selected firms.
Bybit had the largest USDT decline in the group, at $1.66 billion, while adding $315 million in Ripple USD (RLUSD). USD Coin (USDC) balances grew $223 million at Binance and $273 million at OKX. Their USDT holdings fell $987 million and $899 million, respectively.
Bybit announced an RLUSD rewards program on July 20, giving customers an incentive to keep the token on its platform. Bybit said Ripple service fees funded the rewards. The timing coincides with Bybit’s higher RLUSD balance, but the balance data alone do not establish what caused the increase.
Separately, USDC deposits on Hyperliquid rose $1.5 billion to about $7.2 billion by quarter-end. Hyperliquid identifies Coinbase as the treasury deployer for USDC on its platform and says its protocol receives most of the associated reserve yield under AQAv2. Reserve-yield accrual began Aug. 26.
USDT still represented roughly two-thirds of the approximately $85 billion in stablecoins Crystal assigned to named exchanges at quarter-end. A separate Oct. 17 stablecoin market snapshot put USDT’s market value at $183.43 billion; that market-wide measure differs from balances held at particular venues.
Bybit’s third RLUSD rewards phase ended Oct. 18, and a fourth began Oct. 19. Its continuation means later balances will not yet show whether holdings persist after the rewards program ends.
