# Hyperliquid funding rate: how it works, what it costs, and 90 days of data

Funding is the hourly payment between longs and shorts on every Hyperliquid perp. Here is the formula, including the clamp that pins it to a baseline, real rates across markets and over 90 days, what it costs a $10,000 position, and how traders use it.

- URL: https://www.ethyai.app/blog/hyperliquid-funding-rate
- Author: Ethy Team
- Published: November 15, 2026; updated: November 15, 2026
- Publisher: Ethy AI, https://www.ethyai.app/

**The short answer:** Hyperliquid pays funding every hour between longs and shorts to keep each perp close to its spot oracle price. The rate is one eighth of an 8-hour formula built from the premium and a 0.01% interest rate, capped at 0.5% an hour. Positive means longs pay shorts. Near spot, it rests at 0.00125% an hour, about 11% a year.

## What is the funding rate on Hyperliquid, and who pays whom?

[A perpetual contract, or perp](https://www.ethyai.app/blog/perpetual-futures), is a contract on a price with no expiry date. A future that expires is pulled to the spot price on its last day. A perp has no last day, so it needs another pull, and that pull is funding: a payment between traders who are long and traders who are short, sized to push the perp’s price back toward the spot price.

On Hyperliquid, funding is paid every hour, on every perp. It is peer to peer: Hyperliquid’s [funding docs](https://hyperliquid.gitbook.io/hyperliquid-docs/trading/funding) say “no fees are collected on the payments,” so the exchange takes no cut. The sign tells you who pays:

- **Positive rate.** The perp traded above its spot reference, so longs pay shorts.
- **Negative rate.** The perp traded below it, so shorts pay longs.
- **The resting rate.** 0.00125% an hour, paid by longs. A market pays it whenever its perp stays close to spot, which is most of the time (the next section shows why).

> **0.00125%** The hourly funding rate on a Hyperliquid perp that trades close to spot: 0.01% per 8 hours, paid by longs to shorts, about 11% a year. (Hyperliquid Docs: Funding, read November 15, 2026)

Each payment is your position size times the oracle price times the rate. The oracle is Hyperliquid’s spot reference price, and the docs stress that funding uses it, “not the mark price,” the price behind your profit and loss. Two things follow. Funding is charged on the whole position, not on your margin, so leverage makes it bigger next to the money you put up. And it arrives every hour whether the trade is winning or losing, added to or taken from your balance.

## How is funding calculated on Hyperliquid?

Hyperliquid designed its funding to work like the big centralized exchanges’: it works out an 8-hour rate, then pays one eighth of it every hour. The rate has two parts.

1. **The premium (P).** How far the perp trades from the oracle. Every 5 seconds, Hyperliquid measures the average price you would get on a market buy and a market sell of a set size, the impact notional, which its [contract specifications](https://hyperliquid.gitbook.io/hyperliquid-docs/trading/contract-specifications) put at $20,000 for BTC and ETH and $6,000 for every other market. If the impact bid sits above the oracle, or the impact ask below it, that gap divided by the oracle is the premium. The samples are averaged over the hour.
2. **The interest rate (I).** Fixed at 0.01% per 8 hours, the usual setting on centralized exchanges, and paid to shorts. The docs say it represents “the difference in cost to borrow USD versus spot crypto.”

The formula is F = P + clamp(I − P, −0.05%, +0.05%). Each hour you pay or receive F ÷ 8, and the hourly rate is capped at 0.5% on every market. The oracle is a weighted median of spot prices on Binance, OKX, Bybit, Kraken, KuCoin, Gate, MEXC and Hyperliquid, which validators publish about every 3 seconds, per the [oracle docs](https://hyperliquid.gitbook.io/hyperliquid-docs/hypercore/oracle). For HYPE, whose main spot market is Hyperliquid’s own, the oracle leaves out outside exchanges until they have enough liquidity.

### Why the rate so often sits at exactly 0.00125%

The clamp is the part most guides skip, and it explains most of what you see. Whenever the hour’s average premium sits between −0.04% and +0.06%, the clamp gives back exactly the gap between I and P, and the formula collapses to F = I. Funding is then the 0.01% baseline, 0.00125% an hour, whatever the premium did inside that band. The rate goes negative only when the premium falls below −0.05%, and rises above the baseline only when the premium tops +0.06%.

| Average premium over the hour | 8-hour rate (F) | Paid each hour (F ÷ 8) | Who pays |
| --- | --- | --- | --- |
| −0.20% | −0.15% | −0.01875% | Shorts pay longs |
| −0.05% | 0% | 0% | Nobody |
| −0.04% to +0.06% | 0.01% | 0.00125% | Longs pay shorts (the baseline) |
| +0.10% | 0.05% | 0.00625% | Longs pay shorts |
| +0.50% | 0.45% | 0.05625% | Longs pay shorts |
| +4.05% or more | 4% (capped) | 0.5% | Longs pay shorts (the cap) |

We checked the formula against real numbers. Hyperliquid’s API returns each hour’s rate together with its average premium, and for BTC, ETH and HYPE from August 17 to November 14, 2026 the formula gives back 6,480 of the 6,480 hourly rates exactly. The baseline is the normal state: BTC paid exactly 0.00125% in 68% of those hours, ETH in 68% and HYPE in 73%. A rate of 0.00125% tells you the perp stayed close to spot that hour. It is not a signal.

## How do you read Hyperliquid funding: hourly, annualized and predicted?

On [Hyperliquid’s trade screen](https://www.ethyai.app/blog/how-to-use-hyperliquid#trade-screen), the market header shows “Funding / Countdown”: the current rate and the time left until the next payment. Its tooltip calls it “the hourly rate at which longs pay shorts (if negative, shorts pay longs).” The app rounds it, so the 0.00125% baseline shows as 0.0013%. Next to the price chart, a Funding tab plots the history by hour, 8 hours or day, with a running total.

To compare rates across intervals, multiply them out:

| Per hour | Per 8 hours | Per day | Per year (simple) |
| --- | --- | --- | --- |
| 0.00125% (the baseline) | 0.01% | 0.03% | 10.95% |
| 0.005% | 0.04% | 0.12% | 43.8% |
| 0.01% | 0.08% | 0.24% | 87.6% |
| 0.05% | 0.4% | 1.2% | 438% |
| 0.5% (the cap) | 4% | 12% | 4,380% |

Two cautions on yearly numbers. The docs give the baseline as “11.6% APR,” which compounds the payments; multiplying the hourly rate by 24 and by 365 gives 10.95%, the simple figure this guide uses. And a yearly figure taken from a single hour is a thermometer reading, not a forecast, since the rate can change every hour.

### Comparing Hyperliquid with Binance and Bybit

Binance and Bybit usually quote funding per 8 hours and Hyperliquid per hour, so a raw side-by-side is off by a factor of eight. Hyperliquid’s [Funding Comparison page](https://app.hyperliquid.xyz/fundingComparison) lines the three up in 8-hour terms by default, with columns for the gaps (“Binance-HL Arb” and “Bybit-HL Arb”), and can switch to hourly, daily or yearly figures.

### The predicted rate

For the next payment, the API has a request, `{"type": "predictedFundings"}`, that returns Hyperliquid’s predicted rate next to Binance’s and Bybit’s, each with its interval in hours. It covers the main exchange only: the [info endpoint docs](https://hyperliquid.gitbook.io/hyperliquid-docs/for-developers/api/info-endpoint/perpetuals) say predicted rates are “only supported for the first perp dex,” so trade.xyz and other HIP-3 markets are left out. This is what it returned on November 15, 2026, at 10:36 UTC, turned into yearly rates so the venues compare:

| Market | Hyperliquid | Binance | Bybit |
| --- | --- | --- | --- |
| BTC | 10.95% | 0.85% | 3.81% |
| ETH | 3.09% | 0.08% | −2.07% |
| SOL | 7.39% | −4.08% | −21.37% |
| XRP | −0.38% | −3.77% | −24.97% |
| DOGE | −12.86% | −5.55% | −13.56% |
| HYPE | 10.95% | Not in the response | Not in the response |

Same coins, very different readings: SOL’s predicted rate was positive on Hyperliquid and negative on both centralized exchanges. Each venue’s rate follows its own premium. For the past, `{"type": "fundingHistory", "coin": "BTC", "startTime": …}` returns each hour’s rate and premium, 500 rows per call, so you page forward from the last timestamp. Your own payments come from `userFunding`. Our [Hyperliquid API guide](https://www.ethyai.app/blog/hyperliquid-api-guide-building-automated-trading-strategies) shows the code.

## Which Hyperliquid markets have the highest and lowest funding?

Funding changes every hour, so any ranking is a snapshot. This one comes from Hyperliquid’s public API on Saturday, November 14, 2026, at 8:52 PM UTC, on a down day: BTC was 5.2% lower than 24 hours before and HYPE 13.3% lower. Of 350 live markets, we kept the 132 that traded at least $1 million in the previous 24 hours, since thin markets can show extreme rates on very little money, and ranked them by hourly rate. Annualized means the hourly rate times 24 times 365.

| Market | What it tracks | Hourly rate | Annualized | 24h price change | 24h volume |
| --- | --- | --- | --- | --- | --- |
| xyz:PURRDAT | Hyperliquid Strategies Inc. stock (Nasdaq: PURR) | −0.058% | −508% | −9.8% | $1.6M |
| BERA | Crypto token | −0.0521% | −456% | −18.6% | $5.1M |
| SOPH | Crypto token | −0.0326% | −285% | +22.6% | $5.0M |
| XMR | Crypto token | −0.0289% | −253% | −11.2% | $42.7M |
| xyz:SNXX | Tradr 2X Long SNDK Daily ETF | −0.0259% | −227% | −9.5% | $4.0M |
| xyz:IREN | IREN Limited stock | −0.0259% | −227% | −7.4% | $1.4M |
| GRAM | Crypto token | −0.0216% | −189% | −6.4% | $9.8M |
| IMX | Crypto token | −0.0212% | −185% | −11.4% | $1.6M |
| INIT | Crypto token | −0.0205% | −180% | −15.5% | $2.1M |
| xyz:KORU | Direxion Daily MSCI South Korea Bull 3X ETF | −0.0191% | −167% | −9.4% | $6.8M |

Read the table for what it says about the day, not as a list of trades. 130 of the 132 markets were down, and 97 paid negative funding: their perps traded below their spot references, so shorts paid longs almost everywhere. Selling was heavier in the perps than in spot. Only one market paid more than the baseline: xyz:TLT, the trade.xyz perp on the iShares 20+ Year Treasury Bond ETF, at 0.00867% an hour (76% a year). Another 17 sat at exactly the 0.00125% baseline, HYPE among them, and 6 trade.xyz markets at their own baseline of 0.000625%, explained below.

A rate of −0.058% an hour, −508% a year, looks like free money for a long. It lasts as long as the discount lasts, which can be one hour, and the price fell 9.8% that day.

## What has Hyperliquid funding looked like over the last 90 days?

We pulled every hourly rate for BTC, ETH and HYPE from August 17 to November 14, 2026, 2,160 hours each, from Hyperliquid’s public API. The chart shows each day’s average, annualized.

*Chart: Daily average of the hourly funding rate, annualized (× 24 × 365), on Hyperliquid’s BTC, ETH and HYPE perps, August 17 to November 14, 2026. The baseline is 10.95% a year. (Hyperliquid public API (fundingHistory), fetched November 15, 2026)*

| Market | 90-day average, per year | Hours positive | Hours at the baseline | Highest day | Lowest day |
| --- | --- | --- | --- | --- | --- |
| BTC | 9.1% | 93% | 68% | Sep 21: 18.2% | Aug 18: −2.1% |
| ETH | 8.5% | 92% | 68% | Sep 21: 17.0% | Oct 30: −12.9% |
| HYPE | 11.1% | 91% | 73% | Oct 30: 50.2% | Aug 29: −1.7% |
| xyz:NVDA | 5.7% | 92% | 68% | Nov 12: 34.8% | Nov 8: −26.8% |

Three things stand out. Funding on the big crypto perps was positive almost all the time: BTC in 93% of hours, which means longs paid shorts. Its average, 9.1% a year, still sits below the 10.95% baseline, because when BTC funding left the baseline it went below it (in 29% of hours, negative or not) far more often than above it (2.5%). And the spikes are brief: HYPE’s highest day, October 30, averaged 50.2% a year, and funding was back at the baseline by November 4. Holding a long through a spike like that costs real money; holding it through a normal month costs about the baseline or a little less.

xyz:NVDA, [trade.xyz’s Nvidia perp](https://www.ethyai.app/blog/hyperliquid-nvda), runs on half the rate, so its baseline is 0.000625% an hour. It is in the table for the section on trade.xyz below.

## How much does funding cost a position?

Multiply the position’s value by each hour’s rate and add up the hours. With real rates: a BTC long worth about $10,000 throughout, held for the 168 hourly payments from 00:00 UTC on November 8, 2026, to 23:00 UTC on November 14, paid rates that added up to 0.1645%, about $16.45. A short of the same size received it. Had the rate sat at the baseline all week, the bill would have been $21.00: 0.00125% times 168 hours.

| Market | Sum of the 168 hourly rates | A $10,000 long paid | A $10,000 short received |
| --- | --- | --- | --- |
| BTC | 0.1645% | $16.45 | $16.45 |
| ETH | 0.0658% | $6.58 | $6.58 |
| HYPE | 0.1445% | $14.45 | $14.45 |
| xyz:NVDA | 0.1759% | $17.59 | $17.59 |

Two things change the picture.

- **Leverage.** Funding is charged on the $10,000, not on your margin. At 10x leverage, with $1,000 of margin, the BTC long’s $16.45 was 1.6% of its margin in one week. Each payment is added to or taken from your balance, so over weeks it changes how much margin stands behind the position.
- **Extreme hours.** On Sunday, November 8, at 21:00 UTC, xyz:NVDA’s rate was −0.0484% for a single hour, so shorts paid $4.84 per $10,000 in that hour, 77 times its baseline hour.

For a trade you close the same day, funding is usually a rounding error. For a position you hold for weeks, it is a running cost or income that can outweigh the trading fees. If you [test a strategy](https://www.ethyai.app/blog/how-to-backtest-a-trading-strategy) that holds positions for days, count funding as a cost; many backtests leave it out.

## How do traders use funding, and what can go wrong?

### Holding the side that gets paid

When funding is strongly positive, a short collects it every hour; when it is negative, a long does. That sounds like income, but funding is small next to price. At the baseline, a $10,000 position collects $0.125 an hour, and a 1% move against it costs $100, as much as 800 hours of baseline funding. And the rate is set again every hour, so it can change sign from one payment to the next.

### Funding rate arbitrage

The classic version, often called cash and carry, buys a coin on a spot market and shorts the same amount of its perp. Price moves roughly cancel out, and the pair collects funding while the rate stays positive. The cross-exchange version holds the perp long where funding is lower and short where it is higher: the gap that the “Arb” columns of Hyperliquid’s comparison page show. What goes wrong:

- **Fees.** The trade takes four fills: open and close, on both legs. At Hyperliquid’s [base fee rates](https://hyperliquid.gitbook.io/hyperliquid-docs/trading/fees) as of November 2026, a taker pays 0.045% per perp fill and 0.070% per spot fill, so opening and closing a spot and perp pair costs 0.23% of the position, about 7.7 days of baseline funding.
- **The rate turns.** Funding is paid hour by hour, at whatever the premium does. On November 14, 2026, 97 of 132 liquid markets were negative: a short that had been collecting was paying.
- **One leg gets liquidated.** The perp short needs its own margin. A sharp rally can liquidate it while the spot leg is fine, which leaves you long and unhedged.
- **Two venues, two sets of risks.** The cross-exchange version needs margin on both, deals with different payment times (hourly here, every 8 hours on Binance and Bybit by default) and depends on transfers that take time when one side needs topping up.
- **HIP-3 markets differ.** On trade.xyz, funding is scaled down and moves most when the stock market is closed, as the next section shows.

### Funding as a sign of a crowded trade

High positive funding means longs are paying to keep their positions; deep negative funding means shorts are. Traders read a lasting extreme as a crowded trade, which can unwind sharply when it turns. Read it next to open interest, the value of positions still open: open interest and funding rising together say new money is piling onto one side. It is context, not a timing signal, and a crowded trade can stay crowded for a long time. Our [Hyperliquid whale tracker guide](https://www.ethyai.app/blog/hyperliquid-whale-tracker) shows how to check which side the largest accounts are on.

> **Funding is not a yield** It is a transfer between traders, set again every hour. A position that collects it still carries price risk, liquidation risk and trading fees, and the rate can turn against it at the next payment.

> **Ask Ethy:** [What is funding doing on ETH right now, and who is paying whom?](https://agents.ethyai.app/chat?landing=blog&post=hyperliquid-funding-rate&prompt=What%20is%20funding%20doing%20on%20ETH%20right%20now%2C%20and%20who%20is%20paying%20whom%3F)

## Does funding work the same on trade.xyz and other HIP-3 markets?

The plumbing is the same. HIP-3 markets, the perp exchanges that outside teams deploy on Hyperliquid, run on its trading engine, which pays their funding every hour under the same cap. What differs are the settings. According to Hyperliquid’s [HIP-3 deployer docs](https://hyperliquid.gitbook.io/hyperliquid-docs/for-developers/api/hip-3-deployer-actions), the team that deploys a market can set, per market: a multiplier between 0 and 10 that scales the rate, the interest rate (per 8 hours, from −1% to 1%), and the clamp (per 8 hours, up to 1%, with a default of 0.03%). HIP-3 markets also measure the premium a little differently, from the midpoint of the impact bid and ask.

Here is what [trade.xyz, the largest HIP-3 exchange](https://www.ethyai.app/blog/trade-xyz), had set on November 15, 2026, according to Hyperliquid’s API (the `perpDexs` request):

- **Half the rate on almost everything.** A multiplier of 0.5 on 120 of its 122 live markets, and its [specification index](https://docs.trade.xyz/perpetuals/specifications-and-schedules/specification-index) lists 0.5 for each market. trade.xyz announced the change in its [December 19, 2025 changelog](https://docs.trade.xyz/perpetuals/changelog/funding-rate-formula-updates): it lowers the baseline to “~5.5% annualized,” which is 0.000625% an hour, and gives “less aggressive funding during weekend price discovery.” Our guides to [oil perps on Hyperliquid](https://www.ethyai.app/blog/hyperliquid-oil) and [the S&P 500 perp](https://www.ethyai.app/blog/hyperliquid-sp500) track funding on those markets since they listed.
- **Almost none on the pre-IPO markets.** xyz:ANT and xyz:OURA, priced before those companies list, use a multiplier of 0.005, a hundredth of the usual rate.
- **No interest on the currencies.** The interest part is set to 0 on EUR, GBP and JPY, so their resting rate is zero: funding only appears when the premium moves past the clamp.
- **The default clamp.** trade.xyz sets no clamp of its own, so the 0.03% default applies. Its changelog writes the formula with Hyperliquid’s 0.05%, but the 0.03% default fits 2,160 of the 2,160 hourly xyz:NVDA rates in our data and 0.05% fits 1,464. In practice, the xyz baseline holds while the premium stays between −0.02% and +0.04%.

### Funding on stock perps moves most on weekends

trade.xyz prices most US stocks from outside markets from Sunday 8:05 PM to Friday 8:00 PM New York time. The rest of the week, the oracle follows the perp’s own order book, as our guide to [trading stocks on Hyperliquid](https://www.ethyai.app/blog/how-to-trade-stocks-on-hyperliquid) explains. That is when xyz:NVDA’s funding swung hardest. Over our 90 days, that weekend window held 28% of the hours but 18 of the 20 largest hourly rates, and the average rate inside it, in either direction, was 1.7 times its size during the rest of the week. For BTC, ETH and HYPE, the same ratio was between 0.98 and 1.04: their funding was no bigger on weekends.

*Chart: Hourly funding rate on xyz:NVDA, trade.xyz’s Nvidia perp, October 18 to November 14, 2026. The tallest swings fall on weekends, when the oracle follows the perp’s own order book. (Hyperliquid public API (fundingHistory), fetched November 15, 2026)*

For a stock perp you hold over a weekend, that is the funding risk to plan for: a quiet week at 0.000625% an hour, then a Sunday afternoon at −0.0484%.

## How can Ethy help you check funding before you trade?

Ethy is an AI trading agent you ask before you trade. For funding, that means you can ask in plain English what a market’s funding, open interest and positioning look like right now, who is paying whom and whether the crowd leans long or short. Then you can read the chart on four timeframes (daily, 4-hour, 1-hour and 15-minute) and find out why the market moved, with a web search that cites its sources.

If you decide to trade, Ethy places perp orders on Hyperliquid’s main exchange and on trade.xyz, with [stop loss and take profit](https://www.ethyai.app/blog/hyperliquid-stop-loss) as native Hyperliquid orders that stay on the exchange even when Ethy is closed. One USDC balance backs both, through Hyperliquid’s unified account mode. Hyperliquid charges or pays funding on those positions every hour, like on any other.

**Check funding before you hold a perp** Ask Ethy what funding and open interest say about a market, read the chart on four timeframes, and decide with the whole picture in front of you. [Start free](https://agents.ethyai.app/?landing=blog&post=hyperliquid-funding-rate)

## Summary

- Hyperliquid pays funding every hour, peer to peer: a positive rate means longs pay shorts, a negative rate means shorts pay longs.
- The rate is one eighth of an 8-hour formula, F = P + clamp(I − P, −0.05%, +0.05%) with I = 0.01%, capped at 0.5% an hour.
- The clamp pins funding at 0.00125% an hour, about 11% a year, while the premium stays between −0.04% and +0.06%: BTC sat there in 68% of hours from August 17 to November 14, 2026.
- Payments are size times oracle price times rate, on the whole position: a $10,000 BTC long paid about $16.45 for the week to November 14, 2026.
- In the sell-off of November 14, 2026, 97 of 132 liquid Hyperliquid markets had negative funding.
- trade.xyz halves the rate (about 5.5% a year at rest), nearly zeroes it on pre-IPO markets, and its funding swings most on weekends.
- Funding trades carry fees, sign flips and liquidation risk. Read funding as context on a crowded trade, not as income.

## Questions people ask

### How often is funding paid on Hyperliquid?

Every hour, on every perp, including HIP-3 markets such as trade.xyz. The formula produces an 8-hour rate, and each hour you pay or receive one eighth of it. Binance and Bybit usually pay every 8 hours, so divide their rate by 8, or multiply Hyperliquid’s by 8, before comparing.

### Why is Hyperliquid funding so often exactly 0.00125%?

Because of the clamp in the formula. Whenever the hour’s average premium sits between −0.04% and +0.06%, the clamp cancels it, and the rate equals the 0.01% per 8 hours interest part: 0.00125% an hour. From August 17 to November 14, 2026, BTC paid exactly that in 68% of hours.

### What does a negative funding rate mean?

The perp traded below its spot oracle, so shorts pay longs. On Hyperliquid that takes an average premium below −0.05% over the hour. It is common in sell-offs: on November 14, 2026, 97 of 132 liquid markets were negative. A long collects it, which does not make the long a good trade.

### Is the funding rate in the Hyperliquid app hourly or per 8 hours?

Hourly on the trade screen: the “Funding / Countdown” figure is the hourly rate, rounded, so the 0.00125% baseline shows as 0.0013%. The Funding Comparison page shows 8-hour rates by default, to line up with Binance and Bybit, and can switch to hourly, daily or yearly figures.

### Where can I see Hyperliquid funding rate history?

In the app, the Funding tab next to the price chart plots it by hour, 8 hours or day. Through the API, send a fundingHistory request with the coin and a start time in milliseconds to https://api.hyperliquid.xyz/info. It returns each hour’s rate and premium, 500 rows per call, so page forward from the last timestamp.

### Do stock perps on trade.xyz pay funding?

Yes, every hour, through the same engine. trade.xyz scales Hyperliquid’s formula by 0.5, so the resting rate is 0.000625% an hour, about 5.5% a year. Its two pre-IPO markets use a hundredth of the normal rate, and its currency markets have no interest part. Rates swing most on weekends, when prices come from trade.xyz’s own order book.

### Does leverage change how much funding I pay?

Not in dollars: funding is charged on the position’s full value at the oracle price. Leverage changes how large that cost is next to your margin. The $16.45 a $10,000 BTC long paid in the week to November 14, 2026 was 1.6% of a $1,000 margin at 10x, and 0.3% of a $5,000 margin at 2x.

### What is funding rate arbitrage on Hyperliquid?

Holding two offsetting positions so price moves cancel and funding is collected: long spot and short the perp, or long the perp on one exchange and short it on another. It takes four fills of fees, the rate changes every hour, and the short leg can be liquidated in a rally. At Hyperliquid’s base taker fees, one round trip on a spot and perp pair costs about 7.7 days of baseline funding.

## Sources

- [Hyperliquid Docs: Funding](https://hyperliquid.gitbook.io/hyperliquid-docs/trading/funding)
- [Hyperliquid Docs: Contract specifications (funding impact notional)](https://hyperliquid.gitbook.io/hyperliquid-docs/trading/contract-specifications)
- [Hyperliquid Docs: Oracle](https://hyperliquid.gitbook.io/hyperliquid-docs/hypercore/oracle)
- [Hyperliquid Docs: Info endpoint for perpetuals (fundingHistory, predictedFundings, perpDexs)](https://hyperliquid.gitbook.io/hyperliquid-docs/for-developers/api/info-endpoint/perpetuals)
- [Hyperliquid Docs: HIP-3 deployer actions (funding multiplier, interest rate, clamp)](https://hyperliquid.gitbook.io/hyperliquid-docs/for-developers/api/hip-3-deployer-actions)
- [Hyperliquid Docs: Fees](https://hyperliquid.gitbook.io/hyperliquid-docs/trading/fees)
- [Hyperliquid: Funding Comparison](https://app.hyperliquid.xyz/fundingComparison)
- [trade.xyz: Funding rate formula updates (December 19, 2025)](https://docs.trade.xyz/perpetuals/changelog/funding-rate-formula-updates)
- [trade.xyz: Specification index](https://docs.trade.xyz/perpetuals/specifications-and-schedules/specification-index)
