Carry, explained in ten minutes

Most people meet crypto as a bet on direction: buy, hope, sell. Carry is not that. Carry is the payment a market makes to whoever is willing to take the side nobody else wants.

The imbalance is always the same shape. One crowd wants leverage — it wants to be long, bigger than its cash allows, right now. Somebody has to be short against it. Another crowd wants protection against a crash and will pay for it; somebody has to sell that protection. In both cases the eager side pays the patient side, and the payment is set by how eager the crowd is, not by where the price goes next.

A carry strategy takes the patient side, on terms set in advance. Drag the demand below and watch what the impatient side has to pay to get what it wants.

A perpetual future never expires, so nothing drags its price back to the spot market. Exchanges solve that with funding: every eight hours, whichever side is crowded pays the other. Three payments a day, every day, for as long as the position is open.

The rates look tiny and that is the trap in both directions. 0.04 % per eight hours is 0.12 % a day, and a rate that repeats a thousand and ninety-five times a year is not tiny at all. But the costs repeat too — trading fees on both legs, borrowing, rebalancing the hedge when it drifts, slippage — and the strategy bears those whether or not funding shows up.

So the rate alone never decides anything. The Carry Engine states the test as a gate: expected funding income must beat all expected costs plus a risk buffer, strictly, before a position is allowed to open. Drag the rate below the gate and watch the verdict change.

A dated futures contract settles on a known day. Until then it can trade above the spot price, and that difference has a name: basis. It is the market quoting you a price for time.

Cash-and-carry is what you do about it. Buy the asset on the spot market, sell the more expensive future against it, and hold both. At expiry the future settles into the underlying, so the two prices meet. The gap you sold is the gap you collect — and the rate was decided the moment both legs went on, not by anything that happens afterwards.

What happens afterwards still matters, which is why the lesson has a slider rather than a number. Before expiry the spread can widen: a position entered at a good rate can show an unrealised loss for weeks and still settle exactly where it was always going to. The Carry Engine says it plainly — convergence is a tendency, not a guarantee.

Hold the asset and you are exposed to its price. Sell the same amount through a perpetual and you are exposed to its price in the opposite direction. Put both on at once, in the same size, and the two exposures cancel: that is delta-neutral. Delta is sensitivity to price; neutral means there is none left.

What survives the cancellation is the thing you actually wanted. The funding payments keep arriving three times a day, the basis keeps converging, and neither of them asked where the price went. Drag the market below through a 40 % crash and a 40 % rally — the two legs swap places, their sum stays on zero, and the only line that grows is the carry.

Two honest footnotes. The legs cancel exactly only while they are the same size and the two prices move together; they drift, and re-sizing the hedge costs money, which is the rebalancing line in the costs. And a hedge removes direction, not risk: execution, liquidity, collateral, venue and volatility all survive it. That is lesson 6.

An option is insurance. The buyer pays a premium for the right to a price; the seller takes the premium and accepts a defined obligation. Sold on its own, that obligation has no ceiling — which is why a strategy that sells premium sensibly never sells it on its own.

The structure below is the sensible version. The strategy sells protection at one level and buys it back at a higher one, so the worst case is a number that can be calculated before the trade exists: the distance between the two levels, less the premium kept. The market can do anything above the far strike and the loss stops there. That is what «defined risk» means, and it is why the wing is bought rather than saved on.

This is the third source of carry, alongside funding and basis, and it runs on a clock: a premium is sold for a term and is earned as that term passes. CarryLane’s second strategy, the «Carry Option», is built on that clock — quarterly option cycles written on top of the same hedged book, two in its 180-day lock and four in its 365-day one. Drag the settlement below and watch the premium turn into payoff.

Everything above describes income. None of it describes safety, and a course that stopped one lesson earlier would be a sales page. Carry has a specific failure mode: it pays a little, reliably, for a long time, and then asks for a lot back at once. The whole craft is in the controls that sit between those two sentences.

Here is the list, with how the strategy handles each item. Two things it does not contain: numbers about how often any of this has happened, and a promise that it will not. Targets are objectives, not guarantees, and carry can print negative days.