What Funding Rate Arbitrage Actually Is
Funding rate arbitrage is the trade of holding offsetting positions so you collect the payment that perpetual futures traders hand each other every eight hours, without caring which way the price moves. It is a market-neutral strategy, which means the goal is to be flat on price direction and positive on funding cash flow.
Funding arbitrage works because a perpetual futures contract has no expiry, so the exchange uses a funding rate to hold its price near spot. When funding is positive, longs pay shorts.
You take the short side against a long spot position, and the payment lands in your account every eight hours while your net exposure to the price is roughly zero. The edge is the funding, and the discipline is keeping the legs matched.
Because the payment flows from the crowded side to the other side, the strategy pays you to stand where the leverage crowd is pushing against. That is the whole idea, and everything else is execution.
I treat funding arbitrage as a yield strategy, not a directional one, and I screen it the way I would screen a bond: what does it pay, over what horizon, and what breaks it. The deep treatment of what a funding rate is and how it is calculated lives in the explainer on crypto funding rates, which is the definitional companion to this strategy page.
Why the Edge Exists: The Funding Mechanism
A perpetual future never settles, so nothing mechanically forces its price back to the real spot price of the coin. The funding rate is the exchange's solution: a periodic cash transfer between the long and short sides that pulls the contract back toward spot whenever it drifts.
Coinbase's derivatives guide describes the funding rate as two pieces added together: an interest component tied to the cost of borrowing the underlying, and a premium component that measures how far the perp sits from the spot index. When the perp trades above spot, the premium is positive, longs pay shorts, and the payment drags the perp back down toward spot.
That payment is the arbitrage edge. If you are short the perp and long spot when funding is positive, you receive the funding on the short leg while the spot leg holds your value flat, and the net position is close to delta-neutral.
The size of the edge scales with how stretched the crowd is, and I treat the crowding as both the opportunity and the expiration date. At the baseline the capture is modest, and at extremes it is large, which is also when the trade is most likely to be disrupted by a violent unwinding.
The Spot-Future Basis Trade, Step by Step
The classic execution is the spot-future basis trade, sometimes called cash and carry, and it has a fixed mechanical structure. You buy the coin on spot, you short the same amount on the perpetual, and you hold both legs until the funding you capture meets your target or the setup breaks.
Notional matching is the detail that makes the position neutral. If the spot leg and the perp leg are not the same dollar size, the residual is a directional bet layered on top of the funding capture, which defeats the point of the trade.
The basis itself is a second edge. When the perp trades above spot, the spread between them is the basis, and a basis that narrows back to zero pays the basis trader on top of the funding.
The two captures are related but distinct, and conflating them is a common source of confused returns.
I run the basis trade with both legs placed at the same moment so I am never holding naked exposure while I wait for the second fill. The mechanics of posting the short leg on margin are covered in the guide to crypto margin trading, because the perp leg is a leveraged position that inherits every margin rule.
The Funding Capture Math
The capture is the funding rate times the notional, paid each interval, and the arithmetic is simple once you trust the inputs. The baseline funding rate near 0.01 percent per eight-hour interval, sourced from CoinGlass's aggregate data, annualizes to roughly 11 percent on the paying side over a full year of holds.
Worked at a notional of $100,000, that baseline rate pays the receiving side about $10 every eight hours, or roughly $30 a day across the three daily settlements. Held for a full year with the rate static, the gross capture approaches the annualized figure before any cost is subtracted.
In crowded markets the rate runs far higher. MetaMask's guidance flags funding above 0.3 percent per eight-hour interval as an extreme that signals over-leverage, and Zipmex tracked Bitcoin funding near 70 percent annualized in January 2026, which would have paid the short side that annualized rate on the notional for as long as the crowding held.
Those extreme rates are also when the trade breaks, because a rate that rich usually means the book is one squeeze away from a forced unwinding. I treat anything near the MetaMask extreme band as a signal to take profit and leave, not a reason to size up.
The Entry Signals I Check Before I Open
Before I put the trade on, I confirm three things on the venue's data: the funding direction, the basis width, and the order-book depth. If any one of the three is weak, the capture is not worth the execution risk.
| Signal | What it tells me | My threshold to enter |
|---|---|---|
| Funding direction | Which side pays whom | Positive and stable for the side I will short |
| Basis width | How far the perp sits from spot | Wide enough to absorb a fee round-trip |
| Order-book depth | Whether my size fills without slippage | Several times my intended size on both sides |
| Next funding timestamp | When the next payment settles | Enough time to enter before the next tick |
The basis tells me whether the entry pays me twice, once on the funding and once if the spread narrows. The depth tells me whether my fill will slip enough to erase the day's capture before it accrues.
I check the next funding timestamp last, because entering minutes before a settlement locks in a capture I can then evaluate against the execution cost without waiting a full interval to see it.
The Risk Rules I Run on Every Position
The risk rules matter more than the entry, because funding arbitrage accounts that blow up almost always fail on risk, not on reading the rate wrong. I run the same three rules on every position, and they have kept me out of more trouble than any entry signal ever has.
I cap any single arbitrage position at a modest share of total account equity, so a basis shock on one trade cannot take the account with it. Diversification across pairs and venues is the point of the cap, not timidity.
I set a basis reversal stop, because if the spread swings against me by more than a planned percentage of spot, the capture is no longer earning its keep and the directional risk is growing. The funding payment is small per interval and the basis move can be large, so the stop protects the larger number.
I exit on a funding flip that lasts two consecutive periods. A single interval flip is noise, but two in a row usually means the crowd has repositioned and I am now on the paying side, which is the exact opposite of the trade I intended to hold.
Position Sizing and Leverage for the Trade
Sizing decides whether a run of adverse basis moves ends the account, and it is the lever most arbitrage traders underuse. I risk a small, fixed percentage of the account on any single leg and derive the notional from the funding I want to capture rather than from greed.
The useful identity is that desired funding per interval equals notional times the funding rate per interval. Rearranged, the notional equals the funding I want divided by the rate, which gives me a sizing target grounded in the capture rather than in a leverage slider.
Leverage on the perp leg is usually modest in this strategy, between 2x and 5x, because the point is to hold a matched position, not to amplify a directional view. The best leverage for crypto futures guide frames the same choice for directional traders, and the arbitrage framing lands at the conservative end of that band by design.
I keep the actual margin on the perp leg comfortably above the maintenance floor, because a basis adverse move that pushes the short leg toward liquidation is how a neutral trade becomes a forced loss.
What Can Go Wrong: The Honest Risk List
The strategy looks clean on a spreadsheet and gets messy in execution, and the failures cluster into four modes. Adverse basis drift, where the perp moves further from spot instead of narrowing, turns a capture trade into a directional loss on the short leg.
A funding flip turns the receiver into the payer, sometimes within a single interval if sentiment shifts hard. The two-period exit rule exists for exactly this failure mode, and ignoring it is how a winning capture becomes a funding bill.
Fee drag and slippage are the quiet killers. A round-trip on a large notional can cost more than a day of baseline funding, so entering with limit orders and on deep books is the difference between a profitable capture and a slow bleed.
The tail risk is a venue or stablecoin event on the legs, and the historical liquidation record is the cautionary data. CoinDesk Research estimates more than $19 billion wiped in the April 2025 cascade tied to a stablecoin depeg, and FTI Consulting tracks roughly $19 billion liquidated on October 10, 2025.
The full mechanics of how those forced sales cascade are in the guide to how liquidation works, and they are why I never run the trade on a thin or questionable venue.
Funding Arbitrage vs Basis Arbitrage: The Two Flavors
Two related but distinct trades get called funding arbitrage, and conflating them muddles the returns. Funding arbitrage proper captures the periodic payment between longs and shorts, and it pays as long as the funding direction holds, regardless of where the basis sits.
Basis arbitrage, or cash and carry, captures the spread between the perp and spot narrowing back to zero. It pays when the basis compresses, and the funding payment along the way is a bonus or a drag depending on direction.
In practice the two trades overlap, because a perp that trades rich usually also carries positive funding, so the basis trader often gets paid funding while waiting for the spread to close. The distinction matters for how you exit: the funding trader leaves when the rate flips, and the basis trader leaves when the spread narrows.
I think of them as two exit rules layered on the same structure, and I decide which rule governs the trade before I enter so I am not rationalizing the exit after the fact.
A Worked BTC Example With 2026 Numbers
The cleanest way to see the trade is with real numbers, so I will walk through one. Say Bitcoin spot sits at $100,000, the perpetual trades at a slight premium, and the eight-hour funding rate is positive at the 0.01 percent baseline.
The structure is long one BTC spot and short one BTC perp, for a notional of about $100,000 on each leg.
The funding capture on the short leg is 0.01 percent of roughly $100,000, which is about $10 every eight hours, or roughly $30 a day. Held flat for seven days across twenty-one settlements, the gross funding capture approaches $210 before any cost is subtracted.
Now the costs. A round-trip on each leg at typical taker fees runs a small fraction of the notional, and if the basis drifts adversely instead of narrowing, the short leg loses more than the funding accrues.
The honest net return over a clean hold lands in the mid-single digits annualized on the notional, which is why the strategy attracts capital that accepts a bond-like return for a bond-like risk profile.
The example shows both why the trade works and why it is not a get-rich vehicle. The capture is real and sourced, the costs are real and drag on every fill, and the edge survives only with disciplined entry and exit.
To pressure-test execution before sizing up, the futures trading groups on Whop include rooms where leveraged and neutral strategies are workshopped live.
Common Mistakes That End Arbitrage Accounts
The mistakes that end funding arbitrage accounts are mechanical and repeatable, which is the frustrating part of writing them down, and I have watched each of these four blow up an otherwise sound capture trade. Over-sizing a single position, mismatching the notional on the two legs, ignoring the fee round-trip, and holding through a funding flip are the four that quietly turn a capture trade into a loss.
Over-sizing is the most common because the strategy feels safe, and the safety is real until a basis shock on an oversized position takes the account. The cap on single-position exposure exists for this failure mode specifically.
Mismatched notionals are the subtlest mistake, because the position still receives funding while quietly carrying a directional residual. The two legs have to match in dollar size, not in unit count, and verifying that before entry is a one-line check that saves the trade.
Holding through a funding flip is the avoidable one. The trade was entered to receive funding, and once you are paying it for two periods running, you are running the opposite of the strategy you planned.
The full futures trading strategy framework treats exit discipline as the part that survives losing streaks, and funding arbitrage is no exception to that rule.