ETHUSDT · Perpetual futures
Ethereum open interest and funding
Open positions, the cost of carrying a long and account positioning for the ETHUSDT perpetual across three venues, next to a month of history.
9,90 Md $US sits in open ETH positions on Bybit, Binance et OKX·67 % of accounts sit long·a long costs +3,0 % a year on average, from +1,9 % on OKX to +4,1 % on Bybit
Our own series, written once a minute since the collector started.
Where the money sits
The same contract behaves differently on each venue: its own book, its own crowd, its own payout interval. A gap in the annualized column between two rows is a gap somebody is already trading.
| Exchange | Open interest | In coins | Funding | Annualized | Interval | Long / short |
|---|---|---|---|---|---|---|
| Bybit | 2,16 Md $US | 804 962 ETH | +0,0037 % | +4,1 % | 8h | 2,03 |
| Binance | 6,14 Md $US | 2 279 988 ETH | +0,0028 % | +3,1 % | 8h | 2,72 |
| OKX | 1,59 Md $US | 592 533 ETH | +0,0017 % | +1,9 % | 8h | 1,32 |
Why ether funding runs hotter than bitcoin
Ether funding sits above bitcoin funding more often than not, and the reason is on the spot side. Staked ether already pays a yield, so a desk that hedges a long perpetual against spot gives that yield up to do it. The perpetual has to pay more before the trade is worth putting on, and the difference shows up as a persistently higher rate.
The second source is the calendar. Ether reacts to protocol upgrades and to flows into funds that hold it, and both arrive on announced dates. Open interest builds into those dates and unwinds within hours of them, which is why the ether series shows steps where bitcoin shows drift.
Where ether and bitcoin diverge
When ether funding runs well above bitcoin funding and both are rising, leverage is concentrated here and the crowded side is long. That configuration resolves quickly in either direction, and it resolves harder than the same setup on bitcoin because the book underneath is thinner.
The quieter case is the more useful one: ether open interest growing while funding stays flat. Nobody is paying up for the exposure, which usually means the growth is hedged rather than directional.
Is this the same contract on all three venues?
It is the same underlying and the same USDT margin, but three separate contracts with three separate books. Contract sizes, payout intervals and leverage limits differ, which is why the rows above never match exactly.
What happens at the payout time?
Every payout period each venue settles funding between longs and shorts at the rate shown. Nothing goes to the exchange. Holding through the payout is what the annualized column prices: at twelve percent a year, a long that sits for a month gives up about one percent of its notional.
Why is the long/short ratio above one while price is falling?
The ratio counts accounts, not money. Retail crowds tend to lean long into weakness, so a ratio above one during a drawdown is ordinary and says more about positioning than about the next move.
Other contracts