The moment that ruins a two-sided trade is almost never the moment you planned for. You wanted a spread, or a basis, or a pairs position, something with two legs that mostly cancel out and leave you exposed to one specific thing. Then one leg fills and the other sits there. For a few seconds, or a few minutes, you are not in the trade you designed. You are just long or short the market, naked, in whatever direction the filled leg pointed. That gap is legging risk, and it is where a surprising amount of money quietly disappears.
I think about this more than I probably should, because the trades that look cleanest on paper are often the ones most exposed to it. A cash-and-carry basis trade or a calendar spread has almost no directional risk once both legs are on. The risk lives entirely in the seam between the two fills.
Why the seam is worse than it looks
The intuition people carry is that if a trade is roughly market-neutral when complete, being half-on for a moment is only half as risky. It is usually worse than that. When you leg into a spread, you tend to fill the easy side first, because the easy side is liquid and fills instantly. The hard side, the one with the wide book and the thin depth, is the one that lags. So you end up holding the naked position in exactly the direction that is hardest to hedge out if the market moves against you.
And the market tends to move against you at the worst time, because the same event that widens the spread you were trying to capture is often the thing making the illiquid leg gap. You reach for the second leg and the price you assumed is gone. Now you either chase it, which eats the edge you were trying to capture, or you sit half-hedged and hope. Neither is the trade you wanted.
There is a subtler version of this in crypto. If you are long spot on one venue and short a perp on another to capture funding, an outage or a withdrawal freeze on one side can leave you structurally legged for hours, not seconds. The position looks neutral in your spreadsheet and is anything but on the exchanges.
Sequencing: work the illiquid leg first
The single most useful habit here is counterintuitive the first time you hear it. Work the hard leg first. Establish the fill you are least sure of, then hedge the easy side against it once you know your actual price.
The logic is straightforward once you sit with it. The liquid leg is liquid precisely because you can get it whenever you want at close to the price you see. It is the reliable one. So you do not need to lock it in early. What you need to lock in is the uncertain leg, and once that is done, you know your exact basis and can complete the trade at a price you control rather than a price you assumed.
People do the opposite because the liquid leg is satisfying. It fills instantly, it feels like progress, and now you are committed to finishing. But committing to the easy half first is what forces you to chase the hard half. Flip the order and the pressure flips with it.
A few refinements I have found worth keeping:
- Price the illiquid leg as a limit order, then compute where the second leg needs to fill for the spread to still make sense. If the market has moved past that point by the time your first leg fills, you were going to lose the edge anyway, so it is better to know before you are committed.
- Size the two legs to actually offset. It sounds obvious, but futures multipliers, perp contract sizes, and lot conventions differ, and a spread that is off by ten percent in notional is a directional bet wearing a spread's clothes.
- If both legs are illiquid, you do not have a legging problem, you have a position you probably should not be legging into manually at all. Consider a smaller size or an exchange-native spread order if one exists.
Abort rules that fire before you can talk yourself out of them
The hardest part is not knowing when to abort. It is aborting. Once the first leg is on, every instinct pushes you to complete the trade, because unwinding feels like admitting the setup failed. So the rule has to be set before you enter, in numbers, not vibes.
The version I use is a maximum acceptable slippage on the second leg, decided up front. If the illiquid leg is on and the liquid leg cannot be filled within that band, the position gets unwound, not completed. Yes, that means eating a small loss on the leg you already have. That loss is the price of not carrying naked directional risk you never wanted. It is almost always cheaper than the alternative, which is completing a bad spread and then being stuck in it.
Two more that earn their place:
- A time-based abort. If the second leg is not on within some short window, treat it as a failed entry and flatten. Legging risk grows with time held naked, so a stale half-position is a bleeding one.
- A move-based abort. If the underlying moves more than a set amount while you are half-on, flatten regardless of your slippage band. A fast move is exactly when the second leg gets expensive and the naked exposure gets dangerous, and it is the moment your discipline is weakest.
The reason to write these down is that half-legged in a moving market is the worst possible state to be making a judgment call. You are stressed, the price is running, and every option looks bad. A rule you set while calm is the only thing that reliably beats the you who is panicking.
Keeping the two sides honest
Most of managing legging risk is really about visibility. You need to see both legs, across whatever venues they live on, as a single position rather than two unrelated line items. When they drift apart, whether from a partial fill, a rejected order, or an exchange going dark, you want to know in seconds, because seconds are the whole timescale of the problem. This is a chunk of why we built the execution and monitoring side of Blockcircle the way we did, so a two-sided position reads as one thing across the 150-plus venues it might span rather than something you reassemble by hand.
None of this makes legging risk go away. Any time you build a position out of two fills that do not happen at once, you are exposed in the gap. What you can do is shrink the gap, put the uncertain leg first, and decide in advance what you will do when the second leg refuses to cooperate. Do that consistently and the trades that used to quietly bleed in the seam start behaving like the neutral positions you thought you were putting on.