How Perpetual Futures Funding Rates Work
NDA-Trade Research · · 8 min read
A funding rate is a periodic payment exchanged between traders holding long and short positions in a perpetual futures contract. It is the mechanism that keeps a perpetual’s price anchored to the underlying asset’s spot price: when the perpetual trades above spot, longs pay shorts; when it trades below, shorts pay longs. The payment flows between traders — the exchange sets the rate and settles the transfer, but does not (in the standard design) collect it.
That paragraph is the whole idea. The rest of this post unpacks why the mechanism exists, how the rate is calculated and settled, how to annualize it for comparison across venues, and why the same contract can carry very different funding rates on different exchanges — which is what makes funding rate arbitrage possible.
Why Perpetuals Need Funding at All
A traditional futures contract has an expiry date, and at settlement it resolves against the underlying — so its price is forced to converge with spot. The expiry date is the anchor.
A perpetual futures contract (or “perp”) never expires. You can hold a leveraged long or short position indefinitely, which is why perps became the dominant instrument in crypto derivatives. But removing the expiry removes the anchor: nothing structurally forces the perp’s price to track spot.
Funding is the substitute anchor. Instead of convergence-at-expiry, the exchange imposes a recurring cash flow that penalizes the crowded side of the market:
- If the perp trades at a premium to its index price (a reference price derived from spot markets, usually blended across several exchanges), the funding rate goes positive. Longs pay shorts, holding longs becomes costly, and the pressure pushes the perp back toward the index.
- If the perp trades at a discount to the index, the rate goes negative and the flow reverses.
Funding turns price deviation into a carrying cost, and the market arbitrages that cost away — making it one of the few structural yield streams in crypto.
Who Pays Whom
The sign convention is worth committing to memory, because every funding dashboard assumes you know it:
| Funding rate | Perp vs. index | Who pays | Who receives |
|---|---|---|---|
| Positive | Perp above index (premium) | Longs | Shorts |
| Negative | Perp below index (discount) | Shorts | Longs |
The payment is typically calculated on position notional, not margin. If you hold a hypothetical $100,000 long and the interval’s funding rate is +0.01%, you pay $10 at settlement regardless of collateral — leverage amplifies funding cost relative to capital just as it amplifies price P&L. Funding is also zero-sum between traders: long open interest equals short open interest, so every dollar paid by one side is received by the other.
The General Shape of the Funding Rate Calculation
Every exchange publishes its own formula, and the details genuinely differ — do not assume the mechanics on one venue carry over to another. That said, most designs share a common skeleton:
funding rate ≈ premium component + clamp(interest component − premium component)
The moving parts:
- Premium (basis) component. A measure of how far the perp’s traded price sits from the index price, usually averaged over the interval (often using order-book impact prices rather than the last trade, to resist manipulation). This is the term that actually does the anchoring work.
- Interest rate component. A small fixed or slowly-varying term representing the interest differential between the quote and base currency — the cost-of-carry logic inherited from traditional futures pricing. On many venues it is a small constant.
- Clamps and caps. The interest term is typically clamped so it only matters when the premium is small, and the final rate is capped within a band (both per-interval caps and limits on how fast the rate can change). Caps keep funding from becoming ruinous during liquidation cascades or index dislocations.
Some venues — particularly newer perp DEXs — simplify or modify this structure: pure premium-based formulas, different averaging windows or cap regimes, or funding accrued continuously rather than at discrete timestamps. The shape is universal; the exactness is venue-specific. Always read the contract specification of the exchange you actually trade on.
Settlement Intervals: How Often Funding Is Paid
Historically, the standard on centralized exchanges has been an 8-hour funding interval — three settlements per day at fixed UTC times. That convention is still common, but it is no longer universal:
- Several CEXs settle some or all contracts on shorter intervals (4-hour and 1-hour schedules exist), and exchanges sometimes shorten the interval on volatile contracts specifically.
- Many perp DEXs, such as Hyperliquid and other on-chain venues, adopted 1-hour funding as their norm, and some protocols accrue funding continuously and settle it whenever a position is touched.
Two practical warnings. First, intervals vary by venue and by contract, and exchanges change them — verify any specific interval (including those mentioned here) against the venue’s current documentation. Second, on most CEX designs you pay or receive funding only if you hold the position at the settlement timestamp; closing a minute before means no payment for that interval. Continuous-accrual designs remove that timing game.
The interval also matters economically: 0.01% per 8 hours and 0.01% per hour are very different yields, which is why serious comparison requires annualization.
Annualizing a Funding Rate
To compare funding across venues with different intervals — or against any other yield — convert the per-interval rate to an annual percentage rate:
APR (%) = per-interval rate × settlements per year × 100
Settlements per year is just (24 / interval hours) × 365. An 8-hour interval settles 1,095 times per year; a 1-hour interval settles 8,760 times.
Worked hypothetical. Suppose a perp on an 8-hour venue shows a funding rate of +0.01% (0.0001 in decimal). Then:
APR = 0.0001 × 1,095 × 100 = 10.95%
A short position held constantly at that rate would collect roughly 10.95% annualized on notional (ignoring compounding, fees, and the fact that rates move constantly — real realized yield is never this clean).
| Per-interval rate | APR at 8h interval (1,095×/yr) | APR at 1h interval (8,760×/yr) |
|---|---|---|
| 0.001% | 1.10% | 8.76% |
| 0.005% | 5.48% | 43.80% |
| 0.01% | 10.95% | 87.60% |
| 0.05% | 54.75% | 438.00% |
| −0.01% | −10.95% | −87.60% |
All numbers above are illustrative, not observed market rates. The table makes the interval point vividly: the same headline per-interval number is roughly eight times more significant on an hourly venue than on an 8-hour venue. Comparing raw per-interval rates across exchanges without normalizing is a category error.
Predicted vs. Realized Rates
Most exchanges display two numbers, and conflating them causes real mistakes:
- The predicted (or estimated) funding rate is a live projection of what the next settlement will charge, computed from the premium observed so far in the current interval. It updates continuously and can move materially before the settlement timestamp.
- The realized funding rate is the final rate actually applied at settlement — the one that hits your account.
Predicted rates are what you screen on; realized rates are what you earn. A spiky predicted rate can settle at something far more modest once the interval’s average is taken, so historical realized funding is the honest dataset for evaluating any carry strategy.
Why Funding Rates Diverge Across Exchanges
If funding merely tracked one global premium, every venue would show the same rate. In practice, rates for the same asset routinely differ across exchanges, sometimes by sign. The reasons are structural:
- Isolated order flow. Each perp is its own market. A directional crowd on one venue moves that venue’s premium — and therefore its funding — with no obligation on other venues to follow.
- Different user bases and leverage demand. Retail-heavy, market-maker-dense, and DeFi-native venues attract different flow; demand for leveraged longs on one exchange can coexist with balanced positioning elsewhere.
- Listing and market-structure differences. A token may be newly listed and hotly demanded on one venue, mature elsewhere. Formula details, caps, intervals, and index composition also differ, so even identical order flow would not produce identical rates.
This divergence is not noise to be ignored — it is the raw material of funding rate arbitrage: go long where funding is deeply negative (you get paid to be long), short the same asset where funding is high positive (you get paid to be short), and collect both legs while remaining market-neutral on price.
Reading Funding Data in Practice
When you scan funding across venues, the single most useful derived number is the net spread: the difference between the funding rate on the best venue to be long and the best venue to be short, for the same asset, expressed as an annualized rate. A net spread of, say, 15% APR (hypothetically) means a delta-neutral pair of positions across those two venues would accrue roughly that carry before fees, slippage, and rate drift.
Doing this manually across Binance, Bybit, OKX, KuCoin, Hyperliquid, Lighter, Extended, Aster, Nado and other venues means normalizing different intervals, sign conventions, and symbols; NDA-Trade computes and ranks these live net spreads at /funding-rates. For definitions of terms used here — index price, basis, open interest, carry — see the glossary.
The core skill is always the same: normalize to APR, distinguish predicted from realized, and remember that a spread is only as durable as the positioning imbalance that created it.
FAQ
What does a negative funding rate mean?
A negative funding rate means the perpetual is trading below its index price, and shorts pay longs at settlement. Holding a long position earns the funding payment; holding a short costs it. Persistent negative funding usually signals crowded short positioning or heavy hedging demand on that venue.
How often is funding paid?
It depends on the exchange and sometimes on the specific contract. Eight-hour intervals have been the traditional CEX standard, while many perp DEXs settle hourly or accrue funding continuously. Intervals change over time, so check the venue’s current contract specifications rather than relying on remembered defaults.
Can funding rates be predicted?
Partially. Exchanges publish a predicted rate for the upcoming settlement based on the premium observed so far in the interval, and it is usually a reasonable short-horizon estimate. Beyond the next interval, funding depends on future positioning and price action, so forecasts degrade quickly; historical realized rates are more useful for judging whether a level tends to persist.
Why are funding rates different on each exchange?
Because each exchange is an isolated market with its own order flow, user base, leverage demand, formula, caps, and settlement interval. A positioning imbalance on one venue moves its premium and funding without moving others. These persistent differences are what cross-exchange funding arbitrage strategies harvest.
This article is for educational purposes only and does not constitute financial, investment, or trading advice. All numbers shown are hypothetical illustrations, not observed market data.