Hyperliquid funding rate: what is normal and what is a spike
On Hyperliquid's largest markets, normal funding is the base rate of 0.00125% an hour, or a little under it. From November 2025 to October 2026, BTC paid exactly the base rate in 47.9% of hours, was negative in 19.2% and was above the base in only 0.78%. Funding of 0.003% an hour, a level many people call a spike, came in 11 hours all year. On HYPE it came in 286.
Free key14 days of history at 60 requests a minute. Without a key, the API reaches 24 hours back.
By requesting a key you agree to the Terms and API terms; see the Privacy Policy. We may email you when your key keeps hitting its limits; you can opt out at any time.
The short version#
- Base rate: 0.00125% an hour, which is 0.01% per 8 hours or 0.03% a day. Positive means longs pay shorts. How it is set: Hyperliquid funding rates.
- BTC, ETH, SOL: anything above the base is already rarer than 1 hour in 100.
- HYPE, XRP, DOGE: above-base hours are two to eight times as common as on BTC, and 0.003% came in 84 to 286 hours of the year.
- Negative funding is normal too: 16% to 36% of hours on the majors over the year.
- Normal moves. BTC funding was negative in 47% of hours in April 2026 and in 4% in July.
A year, by market#
Every settled hourly rate, 1 November 2025 to 31 October 2026. Rates in percent per hour.
| Market | At the base | Above the base | Negative | Hours at or above 0.003% | 99% of hours at or below | Highest | Lowest |
|---|---|---|---|---|---|---|---|
| BTC | 47.9% | 0.78% | 19.2% | 11 | 0.00125% | 0.00898% | -0.00694% |
| ETH | 52.8% | 0.56% | 16.2% | 9 | 0.00125% | 0.00834% | -0.00767% |
| SOL | 36.1% | 0.63% | 36.1% | 9 | 0.00125% | 0.00975% | -0.01532% |
| HYPE | 74.8% | 6.29% | 9.2% | 286 | 0.00601% | 0.02174% | -0.01008% |
| XRP | 41.9% | 1.74% | 30.8% | 84 | 0.00274% | 0.01582% | -0.01070% |
| DOGE | 50.7% | 2.00% | 21.5% | 90 | 0.00304% | 0.03795% | -0.00762% |
The rest of the hours sit between zero and the base. Hours at the base are the ones where the premium was small enough for the formula's clamp to cancel it.
The last 90 days#
3 August to 31 October 2026, 2,160 hours per market. Funding ran closer to the base and was negative less often than over the full year.
| Market | At the base | Negative | Hours at or above 0.003% | 99% of hours at or below |
|---|---|---|---|---|
| BTC | 63.0% | 8.0% | 8 | 0.00225% |
| ETH | 66.2% | 7.4% | 7 | 0.00191% |
| SOL | 56.2% | 11.5% | 9 | 0.00221% |
| HYPE | 71.4% | 9.4% | 95 | 0.00838% |
| XRP | 72.4% | 8.0% | 63 | 0.00620% |
| DOGE | 74.2% | 3.6% | 79 | 0.00790% |
Normal changes month to month#
Share of hours with negative BTC funding:
| Nov 25 | Dec | Jan 26 | Feb | Mar | Apr | May | Jun | Jul | Aug | Sep | Oct |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 6% | 17% | 15% | 32% | 40% | 47% | 31% | 16% | 4% | 9% | 8% | 6% |
A level that is rare in one month can be common three months later. Pick thresholds from the data you test on, not from a rule of thumb.
Over the whole year the average BTC rate was 0.00066% an hour. Added up, a long held the entire year would have paid about 5.8% of its size in funding.
What to call a spike#
From the year above:
- BTC, ETH, SOL: above 0.00125% is already the top 1% of hours. 0.002% came in 40, 25 and 33 hours. 0.003% came in about 10.
- HYPE: the top 1% starts at 0.006% an hour. 0.003% came in 286 hours.
- XRP and DOGE: the top 1% starts near 0.003%.
- Deep negative: at or below -0.003% an hour, SOL had 381 hours, ETH 57 and BTC 26.
What happened when we traded the spikes#
Same rules for each: $1,000 a trade, 2% stop, 3% take profit, close after 24 hours, fees, funding and slippage charged. Tested 1 November 2025 to 14 July 2026, holdout 14 July to 1 November 2026.
| What we typed | Tested | Holdout |
|---|---|---|
| short $1k of ETH when funding is above 0.003% an hour | +$26.39, 2 trades | +$38.77, 3 trades |
| short $1k of HYPE when funding is above 0.005% an hour | -$78.39, 42 trades, beat 43% of random | +$50.59, 23 trades, beat 87% of random |
| buy $1k of SOL when funding is below -0.003% an hour | -$310.06, 74 trades, beat 21% of random | +$5.09, 2 trades |
The ETH bot is the trap of a level that almost never happens: 5 trades in a year. The SOL bot shows that deep negative funding was not a buy signal in that period. The HYPE bot is the only one worth another look, and it lost money over the 256 tested days before it made money on the holdout.
Check the levels yourself#
- Free files: the funding dataset has every settled rate since May 2023, no key needed. New months appear 90 days after the data date.
- API:
/v1/fundingand SQL over the same tables. A free key reads 14 days, Explorer 90 days, Builder a year. See the Data API. - MCP: ask your agent;
get_funding_statsandget_funding_ratesare onmcp.ironflow.sh/mcp.
curl "https://api.ironflow.sh/v1/funding?market=HYPE-PERP&from=-24h"Questions#
What is a normal funding rate on Hyperliquid?#
0.00125% an hour, the base rate, or slightly below. BTC paid exactly that in 47.9% of hours from November 2025 to October 2026.
What is a high funding rate on Hyperliquid?#
On BTC, ETH and SOL, anything above 0.00125% an hour is in the top 1% of hours. On HYPE the top 1% starts at about 0.006% an hour.
How often is Hyperliquid funding negative?#
On BTC, 19.2% of hours over the year, from 4% in July 2026 to 47% in April 2026.
Related guides#
- Hyperliquid funding rates: formula, history and API
- How to backtest a Hyperliquid strategy in plain English
- Hyperliquid historical data
Think a funding level is a signal? Test it before it costs you money. Type short $1k of ETH when funding is above 0.002% an hour on ironflow.sh.
Past results do not predict future ones. Not financial advice.
Free key14 days of history at 60 requests a minute. Without a key, the API reaches 24 hours back.
By requesting a key you agree to the Terms and API terms; see the Privacy Policy. We may email you when your key keeps hitting its limits; you can opt out at any time.