What is a funding rate?
A funding rate is a periodic payment between traders holding long and short positions on a perpetual futures contract. It is not a fee to the exchange - one side pays the other - and it exists to keep the perpetual's price close to the spot price. A positive rate means longs pay shorts; a negative rate means shorts pay longs.
What 1.8 million payments look like: across 522 Binance coins from 2024 to October 2026, 79% of payments were positive and 60% sat exactly at the 0.01% per 8 hours baseline. Rates of +0.10% or more were only 1.3% of payments, and, surprisingly, rates of -0.10% or lower were more common: 4.4%. Bitcoin averaged 6.9% a year in funding paid by longs.
Right now (per 8 hours): Binance +0.0084%, BingX +0.0100%, Bybit +0.0100%, OKX +0.0100% for BTC. See all coins on the funding rate table.
Why funding exists
A perpetual future has no expiry date, unlike a normal future that converges to spot on the settlement day. Something else has to hold its price near the spot price, and that something is funding. When the perpetual trades above spot, there are more buyers than sellers pushing it up; longs pay shorts, which makes holding a long more expensive and holding a short more attractive. When it trades below spot, the payment reverses.
How the rate is set
- Premium: the gap between the perpetual's price and an index of spot prices. A perpetual above the index gives a positive premium.
- Interest component: a built-in rate that reflects the cost of borrowing in one currency versus the other. Most major exchanges use 0.01% per 8 hours, roughly 11% a year, which is why a calm market sits at exactly that number.
- Clamping: a typical formula (Binance publishes one) is premium plus the interest-minus-premium difference limited to a band of plus or minus 0.05% - so when the perpetual trades close to the index the rate equals the interest component. Other exchanges use similar but not identical formulas.
- Caps: exchanges limit how high or low the rate can go; coins that hit the cap are on the funding limits page.
What a normal rate looks like: our data
We took every funding payment on 522 Binance USDT perpetuals from January 2024 to October 2026 (1,809,936 payments), converted each to a per-8-hours rate and grouped them:
| Rate per 8 hours | All coins: payments | Share | Bitcoin: share |
|---|---|---|---|
| <= -0.10% | 79,512 | 4.4% | 0.0% |
| -0.10% to -0.05% | 46,190 | 2.6% | 0.0% |
| -0.05% to 0 | 258,478 | 14.3% | 15.0% |
| 0 to 0.009% | 163,158 | 9.0% | 52.1% |
| baseline 0.01% | 1,094,728 | 60.5% | 26.4% |
| 0.011% to 0.05% | 96,389 | 5.3% | 5.8% |
| 0.05% to 0.10% | 47,895 | 2.6% | 0.8% |
| >= 0.10% | 23,586 | 1.3% | 0.0% |
Three things stand out. Most of the time funding is boring: 60% of payments were exactly at the 0.01% baseline, and 75% were between zero and +0.05%. Extremes are rare: only 3.9% of payments were +0.05% or more. Negative extremes are more common than positive ones (4.4% at -0.10% or lower against 1.3% at +0.10% or more), - one possible reason, not tested here, is that many small coins have thin books where shorts crowd in after drops. Bitcoin is calmer than the average coin: 85% positive, average +0.0063% per 8 hours, about 6.9% a year.
Descriptive counts of Binance data, tokenized stocks and commodities excluded, coins still trading on 21 October 2026 (delisted coins are not included). Payment intervals differ: about 74% of the payments came from contracts that pay every 4 hours, 22% from 8-hour ones and 4% from hourly ones; each is converted to the same 8-hour scale.
When it is paid, and what it costs you
- Timing: at fixed times - on 8-hour contracts commonly 00:00, 08:00 and 16:00 UTC. Only positions open at that exact moment pay or receive; close a minute before and you pay nothing for that period.
- Size: notional x rate. A 10,000 USDT position at 0.01% per 8 hours pays 1 USDT per payment, 3 USDT a day. The funding fee calculator does the arithmetic for any coin and holding time, with live rates.
- Leverage: funding is charged on the full position, not on your margin. At 10x, 0.01% per 8 hours is 0.1% of your margin per payment, about 110% a year on the margin if it never changed.
- Versus trading fees: for a one-day hold at the baseline rate, funding (0.03%) is smaller than a round-trip taker fee; for holds of a week or more at an elevated rate, funding dominates.
How to read a funding rate
- Near baseline (about 0.01%): no strong imbalance; the market is balanced or only mildly long-leaning.
- Elevated (+0.05% or more): a crowded long side willing to pay up. Holding a long gets expensive - in our data these were 4% of payments.
- Negative: shorts pay longs; often follows sharp drops, with shorts crowded in. Common on smaller coins.
- Across exchanges: the same coin can have different rates on different exchanges at the same time; that gap is what funding rate arbitrage trades. Compare in one table: funding heatmap.
Does funding predict price?
The popular belief: when funding is very high the crowd is long and will get squeezed, so short it - and when funding is very negative, buy. We tested it with the rules written before the data:
- Extreme funding alone, 522 coins: shorting +0.10% per 8 hours or more won 55% of 8,560 trades but averaged -0.33% after costs and funding; buying -0.10% or lower averaged +0.07%. No edge.
- Funding plus rising open interest: the extra condition changed nothing (short -0.34%, long +0.07%, against -0.37% and +0.10% for funding alone). No edge.
Funding is a good description of positioning and cost. As a standalone timing signal, it did not work in our tests. A few violent squeezes against the crowd outweighed many small winning shorts. Live status of both signals: signals today.
Common mistakes
- Comparing rates with different intervals. 0.01% every hour is 0.08% per 8 hours. Always normalize.
- Reading a high rate as a sell signal. It shows crowding; the crowd can stay right for a long time.
- Forgetting funding in a carry trade. Receiving funding on a short is income only while the rate stays positive and price does not squeeze you.
- Ignoring how funding stacks with leverage. The cost relative to margin grows with leverage, and so does liquidation risk: leverage guide.
FAQ
What is a funding rate in crypto?
A periodic payment between traders who are long and traders who are short on a perpetual futures contract. It is not a fee to the exchange. It keeps the perpetual's price close to spot: when the perpetual trades above spot, longs pay shorts; when below, shorts pay longs.
What does a positive funding rate mean?
Longs pay shorts. It means the perpetual trades above the spot index or that the baseline interest component dominates. In our data a positive rate is the normal state: 79% of 1.8 million payments across 522 coins were positive, and 60% sat at the 0.01% per 8 hours baseline.
What is a normal funding rate?
On most major exchanges the baseline is 0.01% per 8 hours (about 11% a year). In our data 60% of payments were at that baseline. Rates of 0.05% or more were only about 4% of payments, and 0.10% or more about 1.3%.
Is the funding rate a fee paid to the exchange?
No. Funding moves between traders: longs pay shorts or the reverse. The exchange only calculates and transfers it. Trading fees are a separate charge the exchange keeps.
How often is funding paid?
Every 8 hours on many contracts (commonly 00:00, 08:00 and 16:00 UTC), but not all. In our 2024-2026 data about 74% of payments came from contracts that pay every 4 hours and 4% from hourly ones, because exchanges move volatile coins to shorter intervals. Check the interval before comparing rates.
Does a high funding rate mean the price will drop?
Not reliably. In our pre-registered test on 522 coins, shorting funding of +0.10% per 8 hours or more won 55% of trades but averaged -0.33% after costs and funding. Adding rising open interest did not help. It shows a crowded side, not a timing signal.
Can funding be negative?
Yes. A negative rate means shorts pay longs and the perpetual trades below spot. In our data 21% of payments were negative, and strongly negative rates (-0.10% or lower per 8 hours) were more common than strongly positive ones: 4.4% against 1.3%.