Funding rate history: how much did it really cost to hold a BTC or ETH perpetual for a year?

01The one-year bill, side by side

Every few hours the exchange works out the funding payment and moves it between the two sides of the contract — that moment is called a settlement. We took every settlement these exchanges published for their BTC and ETH perpetuals over the year to 20 October 2026 and simply added the rates together. The sum is the share of your position size that a long paid out (a plus) or received (a minus) by sitting in the trade the whole time. OKX publishes only about the last three months, so it appears in the 90-day columns only.

A useful number to hold these against: traders often quote 0.01% every 8 hours as the "normal" funding rate. Three payments a day for 365 days would come to about 11% of your position size in a year. Real BTC and ETH funding came in far below that on all four exchanges — which means the quiet fear that funding slowly eats your balance was, for this particular year, much bigger than the actual bill. If you are not sure yet who pays whom and why, the basics are on the funding page; this page is only about what it measured out to.

02How we counted it

The data comes straight from each exchange's own public funding history for the USDT-margined BTC and ETH perpetuals (USDC on Hyperliquid), from 20 October 2025 to 20 October 2026. Binance, Bybit and OKX settle these contracts every 8 hours; Hyperliquid settles every hour, with smaller rates each time. Because we are adding up the actual published settlements rather than annualising one snapshot, the different intervals already end up comparable — a year is a year, whether it came in 1,095 payments or 8,760.

We summed the rates without compounding and assumed a constant position size, so the result is "percent of position size per year", not a return on your deposit. Funding is a payment between traders holding longs and shorts, not a fee the exchange collects. If you want to turn an hourly rate into an 8-hour one in your head, that conversion gets its own short page.

03Which months were expensive, and which paid you

The year was lumpy. August 2026 was the heaviest month for longs on Binance and Bybit: +0.62% on Binance BTC, +0.44% on Bybit BTC, +0.48% and +0.43% on the two ETH contracts. On Hyperliquid the peaks landed elsewhere — July 2026 for BTC at +0.80% and September 2026 for ETH at +0.84%. On Binance and Bybit, that single heavy month was close to a fifth of the whole year's bill; on Hyperliquid the cost was spread a bit more evenly across the months.

Several months went the other way. April 2026 was negative for BTC longs on Binance (-0.18%), Bybit (-0.08%) and Hyperliquid (-0.07%), and February 2026 paid ETH longs on Binance (-0.31%). When lots of people pile into longs, the perpetual drifts above the normal market price of bitcoin (the exchange calls that the index price), and longs start paying shorts to pull it back. When sellers are the bigger crowd, it goes the other way and shorts pay. Across the year, longs paid in about 70–84% of settlements, so the direction is tilted but not fixed.

It is also coin-specific. On Binance, the SOL perpetual summed to -0.47% over the same year: a 1,000 USDT long would have received about 5 USDT in total instead of paying. Do not assume what you measured on bitcoin carries over to the altcoin you actually want to trade.

04What this means for a trade you actually plan to hold

Say you open a 1,000 USDT long on bitcoin at 5x leverage, so 200 USDT of your money is tied up as margin. At the pace of the last 30 days on Binance (+0.29% a month), one week of holding is about 0.7 USDT, two weeks about 1.5 USDT, four weeks about 2.9 USDT. Against your 200 USDT of margin — the money that is actually yours — four weeks is around 1.5%. At Hyperliquid's recent pace (+0.71% a month) the same four weeks is about 7 USDT, roughly 3.5% of that margin.

That is the honest size of the thing: real, worth knowing, and far smaller than a 2% move in price. The payment leaves or enters your futures balance each settlement, so after a week, look at your open trade in the app and check the margin and liquidation price again — they will not be exactly where they were on day one. How long a position is worth keeping open at all is a separate question, covered on how long you can hold a futures position.

Put your own size and holding time into the calculator above, then check today's rate for your contract. And if a short starts looking attractive only because longs are paying, that trade-off deserves a look first.

05What these numbers do not cover

Several things are deliberately left out. We ignored trading fees, which you pay separately on entry and exit and which often outweigh a week of funding — those are collected on the costs section. We ignored compounding, price moves and partial periods, and we assumed your position size never changed. Results also differ between USDT and USDC contracts on the same coin, and exchanges can change funding intervals or rate caps, especially in violent markets.

One more honest note: receiving funding on a short is not income. On Hyperliquid's ETH perpetual, a short would have collected about 6.4% over the year — and in that same year any decent move in ETH's price would have swamped that number in either direction. Funding decides whether holding is cheap or annoying. Price decides whether you make money.

If you want to check this for your own contract, every exchange publishes the past settlements for each perpetual in its futures data pages. The quickest way to get a feel for it is to open a Binance Futures account — check during sign-up that futures are available where you live, since each of these exchanges serves a different list of countries and some readers are served by a local entity with different products — then find the funding history for BTCUSDT and scroll back a few months before you place anything. Our notes on which exchanges work in which country are worth a minute first.