What a funding rate is
A funding rate is a recurring payment between the two sides of a perpetual futures contract. When it is positive, everyone holding a long position pays everyone holding a short; when it is negative, shorts pay longs. The venue is not a party to it and takes none of it — it only moves the money.
Why perpetuals need one at all
An ordinary futures contract has an expiry date. On that date it settles against the spot price, and because everyone knows this in advance, the contract's price is pulled towards spot as the date approaches. Convergence is a consequence of settlement.
A perpetual never expires, so that mechanism does not exist. Left alone, the contract could trade at any distance from the underlying asset indefinitely. Funding is the replacement: when the contract trades above spot, longs pay shorts, which makes holding a long more expensive and holding a short more attractive, until the gap closes. It is a tether made of money rather than of time.
Reading the sign
| Rate | Direction | Usually means |
|---|---|---|
| Positive | Longs pay shorts | The perpetual is trading above spot. More capital wants leveraged long exposure than short. |
| Negative | Shorts pay longs | The perpetual is trading below spot — usually a crowded short, sometimes a spot bid the derivatives book has not followed. |
| Near zero | Almost nothing changes hands | The two sides are balanced. This is the resting state, not the exception. |
The sign is a fact about positioning, not a forecast. It says what traders have already done, and crowded positioning resolves in both directions.
How a venue arrives at the number
Each venue computes its own rate and publishes its own method; the shapes are similar and the details are not interchangeable. Broadly, a venue measures how far its contract is trading from an index of spot prices — the premium — adds a small fixed interest component, and clamps the result so a single dislocated minute cannot produce an enormous payment. What the clamps are, and how the premium is sampled, differs by venue and is stated in each venue's own documentation. The endpoints this site reads, and what each one returns, are listed on Data sources.
Two things are worth carrying away from that. First, the rate is a measurement of a price gap, so it moves with positioning rather than being set by anyone. Second, because each venue measures its own book against its own index, three venues can honestly report three different rates for the same coin at the same instant — which is the gap the carry calculator quantifies.
The comparison trap
This is the error that makes most published funding numbers unusable, and it is not subtle once seen. A venue quotes the rate it will charge per settlement, and settlements are not the same length everywhere. Hyperliquid settles hourly. Binance and Bybit settle on an eight-hour cycle. So an eight-hour rate of 0.01% and an hourly rate of 0.01% differ by a factor of eight in what they actually cost, and the two numbers look identical.
The fix is to annualise: multiply the quoted rate by the number of settlements in a year. That is 8,760 for an hourly contract and 1,095 for an eight-hour one. Simple, not compounded — compounding a rate that is re-measured every settlement would turn a measurement into a forecast. Here is the same contract, today, before and after:
| Venue | Quoted rate | Interval | Settlements per year | Annualised |
|---|---|---|---|---|
| Binance | 0.0001 | 8h | 1,095 | 10.95% |
| Hyperliquid | 0.0000125 | 1h | 8,760 | 10.95% |
| Bybit | 0.0001 | 8h | 1,095 | 10.95% |
BTC · rendered by the server from the snapshot stamped at the top of this page. The same normalisation for every covered contract →
What it costs to hold a position
Funding is charged on the notional value of the position, not on the margin posted against it. A position of $10,000 pays the same funding whether it was opened with $10,000 of collateral or with a tenth of that at ten times leverage — which is why leverage multiplies the cost relative to the capital at risk without changing the cost itself.
| Input | Value |
|---|---|
| Position notional | $10,000 |
| BTC funding, annualised, right now | 10.95% |
| Holding period | 30 days |
| A long pays | $90.00 |
$10,000 × 10.95% × (30 ÷ 365). The rate is today's and will not be tomorrow's, so this is what the position would cost if it stayed here — an arithmetic result, not a projection. Run it on your own numbers →
What a large funding rate does not tell you
A high positive rate is regularly reported as a signal that a reversal is due. It is not one, and the reason is mechanical: the rate is a measurement of a price gap that already exists. It tells you leveraged longs are crowded and are paying for the privilege. Crowded positioning does unwind, and it also persists for weeks in a strong trend while shorts pay the whole way. The rate is a cost, and treating a cost as a direction is the most common way this number is misused.
What it does tell you precisely is the price of carry: how much a position bleeds or earns per day for holding, before any move in the underlying. That is a real, computable number, and it is the one this site publishes.
Where funding is most expensive right now
The 12 covered contracts with the largest annualised rates in either direction. A large negative figure is not a bargain — it is a crowded short paying to stay short, and the same warning about direction applies to both columns.
| Coin | Annualised | Who pays | Open interest Hyperliquid |
|---|---|---|---|
| MON | -153.58% | Shorts pay longs | $38.34M |
| CASHCAT | 103.21% | Longs pay shorts | $35.19M |
| FARTCOIN | 27.61% | Longs pay shorts | $43.86M |
| VVV | 14.56% | Longs pay shorts | $30.19M |
| TAO | 14.51% | Longs pay shorts | $57.33M |
| ETHFI | 13.39% | Longs pay shorts | $17.29M |
| BTC | 10.95% | Longs pay shorts | $3.04B |
| HYPE | 10.95% | Longs pay shorts | $2.14B |
| SOL | 10.95% | Longs pay shorts | $644.49M |
| ZEC | 10.95% | Longs pay shorts | $386.10M |
| XRP | 10.95% | Longs pay shorts | $240.62M |
| PUMP | 10.95% | Longs pay shorts | $205.22M |