A funded trader once told us their account blew its trailing-drawdown limit on a single stop-loss. The stop was set at a clean 4995.00 on E-mini S&P 500 (ES); a 8:30 a.m. data print gapped price straight to 4988 and the fill came in eight points worse than planned. Nothing was broken. The trader had simply chosen an order type that traded price control for a guarantee they did not understand. Order types are not trivia. On a leveraged, rule-bound futures account, the order type you pick is the difference between a controlled loss and a violation.
This is a reference you can come back to. We build from the two primitives every other order is made of, walk up through the stop variants and their exact trade-offs, then unpack what a bracket and an OCO actually do under the hood, how to choose by your prop firm's rules, and what a server-side copier has to do to mirror all of it faithfully.
Market and limit orders, the baseline
A market order guarantees execution but not price, and a limit order guarantees price but not execution. Every other futures order type is a rule layered on top of these two. A market order fills immediately at the best available price (or prices, if you sweep more than one level); in a fast market that fill can be meaningfully worse than the last print you saw, which is slippage. A limit order sets the worst acceptable price: a buy limit fills at your limit or lower, a sell limit fills at your limit or higher. It can fill better, never worse, but if the market never trades at your price it simply does not fill.
One detail surprises people who assume "market order" means a raw market order. On CME, a plain market order is usually handled as a market with protection: it enters as a market order but is capped by a protection price band around the best price, so a thin book cannot fill you at an absurd level. The remainder that cannot fill inside the band rests as a limit at the band edge. So even the simplest order on the menu has a built-in price guard.
Fill certainty versus price certainty. Market gives you the first and sacrifices the second. Limit gives you the second and sacrifices the first. Stops, brackets, and trailing stops are just rules deciding which one you get and when.
Stop-market vs stop-limit
A stop order is dormant until the market trades at or through its trigger price, and the difference between stop-market and stop-limit is what it turns into at that moment. Unlike a limit, a stop is not on the book when you submit it; it is held and released only when price reaches the trigger. A buy stop sits above the market, a sell stop sits below it, which is the opposite of a limit and why stops are used for exits and breakouts rather than for getting a better entry.
When triggered, a stop-market becomes a market order: virtually guaranteed to fill, but at whatever price liquidity allows, which can be far from your trigger in a spike. A stop-limit becomes a limit order at a second price you specify: it caps your slippage but can miss the fill entirely if price runs through the limit before anyone trades there. That is the whole contrast, and the comparison in the diagram below lays out the three common stop behaviors side by side on fill guarantee, price control, and gap exposure.
Here is the insider correction to a common myth: on CME a stop-market is not a naked market order either. It triggers into a stop-with-protection, which becomes a market order capped by a product-specific protection point. For ES that protection point is 3 index points, which is 12 ticks, at the value published at our time of research. So a triggered ES stop can fill no worse than the trigger plus or minus 3 points. If price has already printed past that band in a single move, the order rests as a limit at the protection price and can sit there unfilled. Your "guaranteed-fill" stop is guaranteed only inside the band.
Stop-market and stop-limit are the same trade-off seen from two sides. One accepts slippage to guarantee you are out. The other caps slippage but can leave you in during the exact move you were trying to escape.
When a stop-limit leaves you unprotected (gap risk)
A stop-limit can fail to protect you precisely when you need it most, because a gap through the limit price means there is no counterparty at your price and the order does not fill. Gap risk is structural in futures, not a freak event. CME Globex runs Sunday 5:00 p.m. CT through Friday 4:00 p.m. CT, with a daily maintenance break from 4:00 to 5:00 p.m. CT (Monday through Thursday) when nothing executes. Price can reopen at 5:00 p.m. at a level well away from the 4:00 close, and a scheduled release can gap price intrabar in milliseconds.
Walk the worked numbers on ES, where 1 tick is 0.25 points worth $12.50, 1 point is 4 ticks worth $50 per contract. You are long 1 ES from 5000.00 with a sell stop at 4995.00, planning a 5-point, $250 risk. A news release prints ES from 4996 straight to 4988, an 8-point gap with nothing trading in between.
| Stop type at 4995.00 | What happens on the gap | Approx. result on 1 ES |
|---|---|---|
| Stop-market | Triggers, fills at next available ~4988 | -12 pts = -$600 ($350 worse than planned) |
| Stop-limit (limit 4994.75) | Triggers, no buyer at 4994.75 or better, does not fill | Still long into the drop, bleeding until you act |
| CME stop-with-protection | Capped by 3-pt band, fills no worse than ~4992; if price already below band, rests as limit at 4992 | About -8 pts = -$400, or unfilled at 4992 |
Same 8-point gap, three very different outcomes, decided entirely by the order type. The stop-limit, the one that felt safest because it controls price, is the only one that left the position completely open. This is why overnight protection is the worst place to use a naked stop-limit: a reopening gap through the limit skips it, and you wake up long into a hole.
Trailing stops and how the trail is calculated
A trailing stop is a stop that follows price by a fixed distance in your favor and never moves backward. You set a trail distance, in ticks, points, or a percentage; as price moves your way the stop ratchets along tick by tick at that distance behind the best price seen since activation. When price reverses, the stop stays put. It moves only in the favorable direction. When the trailing level is finally hit, it fires as a stop, typically a stop-market, so it inherits the same slippage behavior as any stop on trigger.
Concretely, long 1 ES from 5000.00 with a 5-point (20-tick) trailing stop: the stop starts at 4995.00. Price runs to 5010.00 and the stop ratchets up to 5005.00, locking in 5 points. Price pulls back to 5006.00 and the stop holds at 5005.00. If price then drops through 5005.00, you exit around there for a +5-point, +$250 gain, minus any slippage on the trigger.
Traders treat a trailing stop as guaranteed locked profit. It is not. It locks the trigger level. The actual exit still slips on a fast reversal, and a trailing stop held only on your local platform stops trailing the second your machine or connection drops. Confirm whether yours is broker-side or local.
If you trade the smaller contracts to manage that slippage and risk per tick, the sizing math is worth knowing cold; we cover it in ES vs NQ vs MES vs MNQ. And because trailing stops interact directly with trailing-drawdown evaluation rules, it pays to understand how trailing, static, and end-of-day drawdown differ before you wire one to a funded account.
What a bracket order really is
A bracket order is an entry order with a stop-loss and a take-profit attached, all defined in one ticket before you commit capital. When the entry fills, the two exits activate together and are linked as an OCO, so the moment one exit fills the other is cancelled. The value is atomicity: there is no window where you hold the position without a stop, and no manual step to clean up the unused exit. You define full risk and reward up front, then let the structure run.
The diagram below shows the shape: an entry in the middle, a stop-loss bracketing it below, a take-profit bracketing it above, and the arrows showing the OCO link where one filling cancels the other. Run the numbers on ES. You buy 1 ES at 5000.00 and bracket it with a take-profit (sell limit) at 5010.00 and a stop-loss (sell stop) at 4995.00. That is +10 points ($500 target) against -5 points ($250 risk), a clean 2:1 reward-to-risk, fully defined before the trade is even live.
- Case A, target hits: price trades 5010.00, the sell limit fills for +10 points = +$500, and the OCO immediately cancels the 4995.00 stop so you are not left accidentally short.
- Case B, stop hits in an orderly market: price trades down through 4995.00, the sell stop-market triggers and fills near 4995.00 for about -5 points = -$250, and the OCO cancels the 5010.00 target.
- Case C, gap on news: as in the table above, the stop-market still fills but slips; the order type inside the bracket decides the damage.
OCO: the mechanism behind brackets
An OCO (one-cancels-other) links two opposing orders so that when either fills, the other is automatically cancelled. In a bracket it is the take-profit and the stop-loss that are linked. The point is to make sure you never hold both a live target and a live stop after the position is already closed, which would otherwise leave a resting order that could open a new, unwanted position.
Here is the part most traders never learn: CME Globex does not natively support OCO. The exchange has no concept of "cancel that other order when this one fills." The OCO logic lives entirely in your broker or trading platform, which watches for the fill and sends the cancel itself. That has a real consequence. If the platform, your connection, or a relay hiccups in the gap between the fill and the cancel, both legs can momentarily be live, briefly doubling your exposure. This is the actual reason brackets occasionally misbehave, and it is not a bug in the exchange, it is a feature implemented above it.
Brackets and OCO are not exchange objects. They are software promises kept by your platform. That distinction is invisible until the day a disconnect leaves both legs working at once.
Choosing order types for funded-account rules
For a funded account, choose the order type that protects you from the rule that ends your account, not the one that saves you a few ticks. Prop firms punish unbounded losses and drawdown breaches far harder than they care about slippage, so the order-type calculus on a funded account is different from a personal account. Below is the decision rule we hand to traders.
- Default to a bracket on every entry. A stop must exist the instant the entry fills. Never carry a position with the stop "in your head."
- Use a stop-market (or CME stop-with-protection) for the risk-defining stop. Getting out beats getting a perfect price. A bad fill is a bad day; an unprotected position into a fast move is a blown account.
- Reserve stop-limit for liquid, orderly conditions only. If you use one, never leave it as your sole overnight protection, and never across a news release.
- Set the take-profit as a limit. You want price control on the way out for profit, where non-execution just means you stay in a winning trade.
- Check whether your trailing stop is broker-side. A local trailing stop is not protection if your platform can drop.
- Verify the firm's order-type rules. Some firms restrict holding through news or overnight, and trailing-drawdown logic can interact with how and when your stop fires.
Where is server-side not the answer? If your entire risk model is one discretionary chart you watch live and never leave, a heavyweight bracketed, copied setup is overkill; a single manual stop-market does the job. Order-type discipline matters most when you are scaling decisions across accounts or away from the screen.
How a trade copier replicates these orders
A trade copier must re-implement bracket and OCO logic on every follower account, not simply mirror one exchange object, because those structures are platform-managed and not native to the exchange. Since CME Globex does not hold the OCO link itself, a copier cannot lean on the exchange to keep a follower's stop and target in sync. To mirror a leader faithfully, the copier has to replicate the entry, the stop-loss, and the take-profit on each follower, and maintain the cancel-the-other logic per follower, so that when one exit fills on a given follower the matching leg is cancelled on that same follower.
Get that wrong and a follower ends up either double-protected, which is harmless, or carrying an orphaned working order after the position is already closed, which is not. This is exactly the kind of correctness that a server-side engine running close to the brokers handles better than a home PC managing dozens of accounts over consumer internet. Thor replicates entry, stop, and target per follower at roughly 17ms server-side, keeping each follower's bracket internally consistent rather than hoping the exchange does it. For the deeper mechanics of how replication and latency work, see what is copy trading, and for how the leading tools compare, the best futures trade copiers. If you route through a specific stack, the NinjaTrader, Tradovate, and general futures copier pages cover the per-platform specifics.
To verify any of the exchange-side figures here, CME publishes the current protection points, non-reviewable ranges, and session calendar on cmegroup.com; those values change with market conditions and by product, so treat the ES numbers above as illustrative and re-check before quoting them as current.
Frequently asked questions
What is the difference between a stop and a stop-limit?
A stop (stop-market) becomes a market order when triggered and is virtually guaranteed to fill but at whatever price liquidity is available, while a stop-limit becomes a limit order when triggered and only fills at the limit price or better. The stop-market accepts slippage to guarantee you are out; the stop-limit caps slippage but can miss the fill entirely if price blows through the limit, leaving you in the position.
Do trailing stops work overnight?
It depends on where the trailing stop is held. A trailing stop managed at the broker or exchange keeps trailing overnight, but a trailing stop managed only on your local platform stops trailing the moment your machine or connection drops. Even a server-side trailing stop fires as a stop on trigger, so a reopening gap after the CME daily maintenance break can still slip the exit well past the trailing level.
Is a bracket order the same as OCO?
No, a bracket order contains an OCO but adds an entry order. A bracket packages three orders in one ticket: an entry, plus a stop-loss and take-profit that activate when the entry fills and are linked as an OCO so that filling one cancels the other. An OCO on its own is just the two linked exit orders without the entry.
Which order type is safest for a funded account?
For risk-defining stops on a funded account, a stop-market (or CME stop-with-protection) is usually safest because it prioritizes getting you out over getting a perfect price. Prop-firm rules punish unbounded losses far harder than a few ticks of slippage, so being filled badly beats being left unprotected during a fast move. Always pair it with a bracket so a stop exists the instant the entry fills.
Can a trade copier copy bracket and OCO orders?
Yes, but only if the copier re-implements the bracket and OCO logic per follower account. Because brackets and OCO are platform-managed and not native to CME Globex, the copier must replicate the entry, stop-loss and take-profit and maintain the cancel-the-other logic on each follower, so that when one exit fills on a follower the other is cancelled on that same follower. Thor mirrors entry, stop and target per follower at roughly 17ms.
Why didn't my stop fill at my price?
A stop does not guarantee your exact price, it only guarantees a trigger. A stop-market fills at the next available price after the trigger, which slips in a fast market; a stop-limit only fills at your limit or better and can be skipped entirely on a gap. On CME, a triggered stop becomes a stop-with-protection capped by a product-specific protection band, so if price prints past that band in one move it can rest unfilled at the protection price.