Crypto Perpetual Funding Carry

Thesis (edge)

A perpetual future is a crypto contract that never expires. To keep its price tethered to the underlying spot price, exchanges use a funding mechanism: every eight hours, if the perpetual is trading above spot, everyone holding a long position pays a fee to everyone holding a short. If it trades below spot, the shorts pay the longs.

In crypto, the perpetual has usually traded above spot, because the market is full of retail traders who want leveraged long exposure and are willing to pay for it. That persistent demand to be long creates a persistent payment to whoever is willing to be short.

The trade is straightforward. Buy the coin in the spot market. Short the same amount in the perpetual. Your net exposure to the price of the coin is roughly zero, so you do not care whether it goes up or down. What you collect is the funding payment, over and over, every eight hours.

At times this has produced eye-catching annualized yields, sometimes 20 or 30 percent or far more during a mania. It looks like free money. It is not, and the reasons why are the most important part of this page.

Where it works (regimes)

Funding is high when retail enthusiasm is high. Bull markets, breakouts, meme-driven rallies and periods of heavy leverage all push the perpetual above spot and drive funding up. The strategy earns most exactly when the market feels most speculative.

Funding compresses toward zero in quiet markets, and it goes negative in sharp selloffs when everyone wants to be short. When funding is negative you are paying to hold the position rather than collecting, and the correct response is to be out.

The uncomfortable truth is that the yield is highest when the risk is also highest. Peak funding coincides with peak leverage in the system, which is precisely when an exchange is most likely to have problems, when liquidation cascades are most violent, and when the counterparty holding your collateral is most stressed.

Signals

• Funding rate: the headline signal. Exchanges publish it, usually as a small percentage charged every eight hours. Convert it to an annualized figure to compare across venues, but be careful: a rate that is high right now is not a rate you will earn for a year. Use a rolling average of what was actually paid, not a snapshot annualized to a scary-looking number.
• Basis: the gap between the perpetual price and the spot price. It is the underlying cause of the funding, and a persistent premium is the condition that keeps funding positive.
• Open interest and positioning skew: heavy long open interest with thin shorts means the funding is likely to stay positive, and it also means a liquidation cascade could be severe.
• Cost threshold: the entry rule should compare funding not to zero but to the full round-trip cost of putting on and taking off the trade, including spot fees, perpetual fees, spread and withdrawal costs. A funding rate that barely exceeds costs is not a trade.

Portfolio construction

The position is a pair: long spot, short perpetual, sized so the two offset. The delta is close to zero, so the profit and loss is dominated by the funding stream and by the small drift in the basis.

The collateral question is where most of the design effort belongs. Your short perpetual position requires margin on the exchange. If the price of the coin rallies hard, your short loses money on paper and the exchange demands more margin, even though your spot position is gaining an offsetting amount. If the spot is held somewhere else, or in cold storage, you may not be able to move collateral fast enough, and you get liquidated on a position that was economically flat the entire time.

That scenario, being liquidated on a hedged trade, is the single most common way this strategy fails in practice. Hold generous excess margin. Assume you will need it on the worst possible night.

Spread across multiple venues, but understand what that buys you. It reduces the damage from any one exchange failing. It does not remove it.

Risk model

Exchange risk is the risk. Not price risk, not volatility, not the funding rate. The exchange.

To collect funding you must have assets on a crypto exchange. Crypto exchanges have failed, frozen withdrawals, been hacked, and in the case of FTX, simply taken customer assets and lost them. People running exactly this delta-neutral, apparently risk-free funding trade lost everything, not because the trade went wrong but because the venue holding it ceased to exist. Any model of this strategy that does not put a meaningful probability on total loss of the exchange balance is not a model, it is a marketing document.

Beyond that:

• Liquidation risk on the short leg, discussed above, which is a collateral management problem rather than a market problem.
• Funding reversal, where the rate flips negative and the trade starts costing money. This is manageable if you have a clean exit rule and painful if you do not.
• Auto-deleveraging, where an exchange forcibly closes your profitable position because the counterparty on the other side blew up and the insurance fund ran dry. You can be right, hedged and still closed out.
• Stablecoin risk, where the collateral you posted is a token that is supposed to be worth one dollar and briefly is not.
• Oracle and index risk, where the exchange's own price feed misbehaves and triggers liquidations at prices that never traded.

Size this strategy as if each exchange balance could go to zero, because it can.

Costs & implementation

Trading fees on both legs, spot spread, perpetual spread, and the cost of moving assets between venues. During volatile periods spreads widen and slippage on unwinding a large position can consume weeks of accumulated funding in a single afternoon.

Capacity is real but limited. Shorting the perpetual in size pushes the perpetual price down toward spot, which reduces the very funding rate you were trying to capture. The trade competes with itself.

Operationally this is one of the more demanding strategies to run. It requires continuous monitoring, automated collateral management, and the ability to act at three in the morning on a Sunday, because crypto never closes and the liquidation engine does not observe weekends.

Failure modes

• Treating the annualized funding number as a yield you will actually receive. It is a snapshot, and it decays.
• Chasing the highest funding, which is usually offered by the smallest and least solvent exchange.
• Running thin collateral and getting liquidated on a fully hedged position during a rally.
• Ignoring the cost of entry and exit, which for short-lived funding spikes can exceed the funding collected.
• Assuming delta-neutral means risk-neutral. It removes price risk and leaves every other risk fully intact.
• Backtesting on funding data without modelling fees, slippage, liquidation events or exchange failure, which produces a beautiful equity curve that bears no relationship to what a real book would have experienced.

Our Notes & Suggestions

Be very clear about what is actually being earned here. This is not an arbitrage. It is payment for providing leverage to enthusiastic buyers, and for accepting custody and counterparty risk on venues with weak regulation and, in several documented cases, no meaningful controls at all. The yield is compensation for exactly that. It is not free.

If you run it, run it on the largest and most reputable venues, keep balances small relative to total capital, keep excess margin at all times, and have a written plan for the day an exchange stops processing withdrawals. That day has happened repeatedly, and the participants who survived it were the ones who had assumed in advance that it would.

For education and research this is a superb strategy to study, because it teaches the single most important lesson in systematic trading. A position can be perfectly hedged against the risk you were thinking about and still be exposed to the one that actually kills it.