What is a bracket order?

The word "bracket" is literal. For a long position the target sits above the entry and the stop below it, so the trade is bracketed on both sides and cannot leave the space between them except through one of the two doors you built. The picture is simple enough that it invites a lazy reading, in which the three legs are three of the same thing at three different prices. They are not. They are three different order types with three different failure modes, and the differences decide everything downstream.

The entry can be a market order, a limit or a stop, depending on how you want to get in, and it is the only leg with any discretion in it. The target is a plain limit order resting above the market: it exists, it sits in the order book, it is visible to the exchange, and it waits. The stop is not an order in the book at all. It is a condition. Nothing exists in the market until the traded price touches your trigger, at which point an order is released and begins its life. That difference, an order already resting against an order still waiting to be born, is the source of almost every surprise a bracket order will ever produce, and it is worth understanding the underlying difference between a limit order and a market order before you rely on either.

There is a second asymmetry hiding in that, and it runs the wrong way for a trader. The target is a limit, so it can never fill worse than your price. It may not fill at all, but it cannot fill badly. The stop releases an order that can fill worse, and in the conditions where you most need it, usually will. So the bracket is lopsided: the good side is price-certain and fill-uncertain, and the bad side is fill-certain and price-uncertain. Hold that thought, because it is exactly why everything in the table below is a plan rather than a promise.

Notice the two entries in the R column marked "by definition". They are not a flourish. R is not a fourth thing you configure somewhere in the order ticket: it is created the moment you place the stop, and it is created whether or not you were thinking about it. The entry is 0R because it is the origin, and the stop is −1R because the stop is what R means. Only the target's R is a decision, and that is the one figure in the whole structure that expresses what you are actually willing to wait for.

What turns three orders into a bracket is not that they were typed at the same time. It is the linkage between the two exits, and that linkage does two distinct jobs. The exits are contingent, meaning neither of them exists until the entry fills, so an entry that is still pending or gets rejected never leaves loose orders lying around in the market. And they are mutually exclusive, meaning that the instant one of them executes, the same instruction retires the other. Target fills, stop cancelled. Stop triggers, target cancelled. The link runs in both directions and it is not something you have to remember to do.

The easiest way to see what that is worth is to take it away. Everything a bracket does mechanically, you could do by hand: buy, then type in a limit above, then type in a stop below. People do exactly this every day, and it is the honest fallback where the product is not offered. But run the two versions against the same session, order state by order state, and the difference is not a matter of convenience. It is two windows of exposure that the manual version cannot close, and one of them can put you into a position you never wanted.

So the link enforces two invariants across the life of the trade, and they are worth stating plainly because they are the entire mechanical case for the order type. First: from the moment the position exists to the moment it closes, there is no instant at which it is live and the stop is not. Second: no order of yours outlives the position it belonged to. Both are true by construction rather than by diligence. The manual version can satisfy both on a good day, when the fill is clean and you are paying attention and nothing else is happening. It cannot satisfy them by construction, and construction is the only kind of reliability that survives a bad day.

Strip the mechanics away and ask what the order type is really for, and the answer is not convenience. It is that the two exit decisions are the ones you are least able to make at the moment they arrive. Consider the same number at two moments. Sitting flat before the entry, ₹492 is a level on a chart. You can look at the base the price has been grinding in all morning, see the swing low at ₹493.40, and say without any strain at all that below ₹492 the reason you wanted to be long has stopped being true. Nothing is at stake. The judgement costs you nothing and is therefore worth something.

Twenty minutes later, with the position on and the price at ₹493.10, that number is no longer a level. It is ₹1,725 of your money and a question about whether you were wrong, arriving together, at speed, while the tape moves. The chart has not changed. The level has not changed. You have changed, and the specific way you have changed is that you now have a reason to want the level to be somewhere else. That is the moment the stop gets widened "just a little, to give it room", which is a sentence that has never once been said by a trader who was flat. A bracket order does not make you a calmer person. It simply ensures that by the time you would like to negotiate, there is nothing left to negotiate about, because the instruction is already in the market and the decision it encodes was made by a version of you who had no stake in the answer.

A bracket order does not make you disciplined. It makes your discipline load-bearing, by moving the two exit decisions to the only moment you were ever qualified to make them.

This is not a soft point, and the Indian context gives it a hard edge. SEBI's own study of the equity derivatives segment found that about 93% of individual traders in equity derivatives made net losses over FY22 to FY24, aggregate net losses exceeding ₹1.8 lakh crore (SEBI, September 2024). No single order type explains a number like that, and it would be dishonest to suggest one could fix it. But the figure is a useful corrective to the idea that the in-the-moment decision is a skill most people simply have. It is the decision that gets made under leverage, under time pressure, and with an open position generating a fresh opinion every second. The value of pre-commitment is not that it is clever. It is that it takes that decision off the table before the table exists.

And it is worth being equally clear about what pre-commitment does not buy, because the case for the order type is weakened, not strengthened, by overselling it. The bracket enforces levels. It does not choose them. If the entry has no edge behind it, a bracket is simply a well-organised way to lose money at a steady, disciplined rate. If the stop sits at a level that means nothing on the chart, the bracket will enforce that meaningless level with perfect fidelity. And if the size was picked before the stop, which is the subject of two sections from here, the bracket will cheerfully wrap a position far too large for the account and hold it there until one of the exits fires. The order type closes the gap between deciding and doing. Whether the deciding was any good is a separate question, and it is the harder one.

Return to the figure at the top of the page and look again at what happened when the stop went down at ₹492. A level was placed, obviously. But something else was created at the same instant, and it is the reason a bracket order pairs so naturally with the way professionals talk about trades. The distance from the entry to the stop, ₹8.00 a share, became a unit. One R. From that moment every other number in the trade can be read as a multiple of it, and most of them are more honest that way.

The reason to bother is that rupees do not compare and R does. A ₹16 gain against an ₹8 risk and a ₹160 gain against an ₹80 risk are the same trade, taken at different sizes on different instruments, and if you keep the record in rupees you will never see that. Keep it in R and a hundred trades across a dozen instruments and three timeframes collapse into one comparable series, which is the only form in which the question "does this method actually work" can even be asked. It also puts a floor under the reporting: a run of trades has an average outcome in R, and that number is your expectancy per unit of risk, which is a real statistic rather than a story about a good week.

Now look at what the bracket does with that. Because it demands the target up front, it will not let you submit the order without answering a question you might otherwise dodge indefinitely: how many R am I willing to sit through this for? That is a genuinely uncomfortable question and it has a real answer, and the answer determines the target rather than the other way around. It is why the target in the example lands on ₹516 rather than ₹515. ₹515 is a round number, which is to say a number your eye picked and your method never approved. ₹516 is what +2R looks like on this chart today, and it would look like something else tomorrow. The level is an output of the decision, not a substitute for it.

One extension of the structure follows naturally once you are thinking in units, and it is the reason most bracket implementations offer a trailing stop. A trail follows the price at a fixed distance as the trade moves your way and holds still when it pulls back, so it only ever ratchets in one direction. Buy at ₹100 with an initial stop at ₹96 and a trail of ₹4: at ₹105 the stop has climbed to ₹101, and if the price then falls back to ₹102 the stop stays at ₹101. Read in R, its job is precise. A fixed bracket caps the trade at +2R by design, and the trail is what removes the cap. Once the stop is above the entry the trade cannot lose; once it is a full R above the entry the worst available outcome is a winner. What it costs is precision, because a trail is a distance, and a distance tight enough to protect most of the peak is also tight enough to be hit by the instrument's ordinary movement. That movement is not a matter of opinion. It is measurable, and the next section measures it.

Read the sizing table backwards and the most expensive habit in retail trading falls out of it as a matter of arithmetic. Hold the risk budget fixed and the trade has exactly two free variables, the stop distance and the number of shares, tied together by a single constraint: distance times size equals the budget. Fix either one and you have fixed the other. There is no third option and no clever way around it. The only question left is which of the two you allow the chart to name, and the chart has an opinion about precisely one of them.

Done in the right order, the chart names the stop, because the chart is the only thing that knows where your idea stops being true. The budget and the stop then name the size, and the size arrives as an output that you find out about rather than decide. Done in the wrong order, and this is the common way, the size comes first. A round number, usually. A thousand shares, because the account allowed a thousand shares, or because the available margin did. The budget has not moved, so the arithmetic quietly takes its revenge on the only variable left: the stop distance is forced to ₹2.00, which puts the stop at ₹498.00, and nothing whatsoever on the chart happens at ₹498.00.

That is the whole failure, and it is worth naming it precisely because it is so rarely described as what it is. The trader who sized first did not take a small risk. They took the same ₹2,000 risk, with a far larger notional behind it, and then paid for the privilege by moving the stop to a place where ordinary market noise will find it several times an hour. The budget held. The plan did not. And the outcome is not a stop-out that tells you anything, because being stopped at ₹498.00 carries no information at all: the idea was never tested. If you want the size handed to you as an output rather than argued into existence, our position sizing calculator does that arithmetic in the correct direction.

So the causal chain runs one way and only one way, and it is worth committing to memory in this order. The chart names the stop. The stop and the risk budget name the size. The size is never an input to anything. Everything that goes wrong in this part of trading is some version of running that chain backwards, and the bracket order will not save you from it, because a bracket wraps whatever position you hand it with equal enthusiasm. Where the stop belongs on the chart is its own subject, and one worth more than a paragraph: the reasoning behind stop-loss placement is what makes the rest of the arithmetic mean anything.

Everything so far has treated the bracket as a two-door structure: the price goes up and out through the target, or down and out through the stop. That is how it is drawn, how it is taught, and how most people carry it in their heads. It is also incomplete, because the classic bracket order is an intraday instrument and was never designed to hold anything. Whatever is still open near the end of the session gets closed by the broker's own square-off process at the prevailing market price, and that process does not consult the plan.

So a bracket has three exits, not two. The target, the stop, and the clock. The third one deserves a name because it behaves nothing like the other two: it has no price attached to it at the moment you place the order, it is a market order you did not write, and it fires on a schedule rather than on a condition. It is also, in practice, extremely common. The days a trade runs cleanly to one of its two levels are the memorable ones. Plenty of sessions simply do not resolve, and those are the ones the clock settles.

Read the ledger under the chart and note the thing that makes the third exit genuinely dangerous rather than merely untidy: it does not care which side of your entry you are sitting on. A trade at −0.9R at 3:19 pm is squared off at roughly −0.9R, which is annoying but survivable. A trade at +1.9R at 3:19 pm, ten paise from the target it spent all day walking toward, is squared off at whatever the last few minutes of the session decide to do to it. The clock takes the worst of both worlds and hands it to you in exchange for nothing. There is no version of the intraday product in which this is optional.

The consequence for the method is sharper than it first looks, and it is the reason this section exists rather than being a footnote. An intraday bracket is a bet with a deadline attached, which means the target has to be reachable inside one session by an instrument that actually moves that far in a session. A +2R target on something that travels 1R in an average day is not a plan; it is a wish, and the clock will settle it for you at a price nobody chose, session after session, while you conclude that your stops are too tight or your entries are early. The instrument's daily range is a hard input to the target, not a detail. If you cannot get to your R multiple before the deadline, you do not have an intraday strategy. You have a swing strategy being squared off every afternoon.

This is the section that separates a description of a bracket order from an understanding of one, and it comes down to a single sentence. A stop-loss does not hold a price. It holds a condition: when the traded price crosses your trigger, an order is released into the market. Everything after that word "released" is the market's business, not yours. Your instruction has been carried out in full the moment the order leaves. What it fills at, and whether it fills at all, was never something you were in a position to specify.

The first fork in that road is what kind of order gets released. A stop-loss market order becomes a market order on trigger, so the exit is close to certain and the price is whatever is available; in a fast move, that can be a long way from your level. A stop-loss limit order becomes a limit order at a price you set, which caps how bad the fill can be but introduces a strictly worse failure: if the market goes straight through your limit and never trades back, the order does not fill at all, and you are still holding a losing position with no protection and a false sense that you had some. Capping the price can cost you the exit entirely. There is no third option that gives you both, and which of the two your broker and segment actually support has changed over time, so confirm it rather than assume it.

It is important to read that figure for what it says and not for more. It does not say stops are useless, and anyone who draws that conclusion has misread it. Column one is the ordinary case, it is the majority of days, and on those days the bracket does precisely what it promised at a cost of a few paise. What the figure says is narrower and more useful: the stop bounds the decision you made, not the outcome the market delivers. Minus one R is the loss you agreed to accept. It is not a loss the market ever agreed to hand you, and the market was not consulted when you wrote it down.

The practical consequences are not dramatic, but they are the ones that separate a trader who has thought about this from one who has not. Your stop is only ever as good as the book behind it, which makes liquidity a risk-management input rather than a convenience: the same ₹8 stop is a different instrument entirely on a counter that trades in a ₹1.40 spread. And your account has to be able to absorb an occasional −2R on a trade you planned at −1R, because the four columns above are not a warning about a rare disaster. They are just what a stop is, drawn honestly. If a single −2R would do structural damage, the problem is not the stop; it is that the size was too big for the instrument you chose to place it on.

The bracket belongs to a small family of orders that pair an entry or a holding with automated exits, and they are easy to confuse until you fix on two questions. How many exit legs does it have, and is it intraday or can it hold a delivery position? Every member of the family is described by those two answers, and the family is smaller than the terminology suggests.

The nearest relative is the cover order, which is a bracket with the target removed: an entry and a compulsory stop-loss, and nothing else. The compulsory part is the interesting bit, because it means a cover order cannot be placed without protection by construction, which is the same structural promise a bracket makes, minus the automation of the profitable exit. It is a narrower tool with a narrower claim, and the trade-off is real rather than obvious; the differences are worth working through properly in the guide to what a cover order is. At the other end sits the plain stop-loss, which protects a position but knows nothing about entries, targets, or linkage, and which is therefore the most available and the least structural of the three.

Which brings us to the question this page gets asked most, and to an answer that has a date on it. Everything above is the mechanism, and the mechanism is universal. The specific product Indian retail traders knew as a Bracket Order was something more particular: an intraday order that, because its compulsory stop defined the maximum loss in advance, came bundled with heavy leverage, sometimes several multiples of the capital committed. It is worth being honest about which half of that was the attraction. For most users it was not the bracketing.

It did not survive the regulator. Under SEBI's peak-margin framework, introduced by circular SEBI/HO/MRD2/DCAP/CIR/P/2020/127 dated 20 July 2020, brokers had to ensure a client's full margin was in place through the day, verified at random intraday snapshots, rather than only at the day's end. The requirement was phased in from December 2020 in steps and reached 100 percent by September 2021. Once full upfront margin was mandatory, the extra intraday leverage that made bracket and cover orders worth offering simply evaporated, and a representative Indian retail broker discontinued both rather than ship a hollow version. The tool changed. Note what did not: the discipline it encoded is still available to anyone willing to build it by hand, and the two windows in the figure earlier are the price of doing so.

Set the whole thing at altitude and a bracket order turns out to be a small, honest tool with one genuinely good idea in it. It belongs to the execution layer, the thin slice between deciding to act and managing what you did, and within that slice it is excellent. It removes the window in which you are exposed and unprotected. It retires the surviving leg so nothing outlives the position. And it settles the two exit decisions at the only moment you were calm enough to settle them. That is a real contribution and it is worth having. It is also the entire contribution, and the ledger below is deliberately symmetrical, because the right-hand column is not a disclaimer. It is the rest of the job.

Read the two columns together and the shape of the thing is clear enough. Everything on the left is enforcement, and everything on the right is judgement. The order type is very good at the first and structurally incapable of the second, and the reason it gets oversold is that enforcement is visible and judgement is not. A bracket order looks like a plan. It photographs like discipline. But the entry has to be worth taking, the stop has to sit where the idea is genuinely wrong rather than where the size needed it, the target has to express an R multiple the instrument can actually reach before the clock closes the session, and the size has to arrive as an output of the first two rather than as an ambition the leverage happened to permit.

That upstream work, the edge, the invalidation level, the R multiple and the sizing, is the part worth learning, and it is what the method we teach is built around. The bracket order is the easy half, and it will do its half perfectly on a plan that deserves it and equally perfectly on one that does not. A single instruction that carries its own exit is a genuinely elegant piece of engineering. What it carries is still up to you.

• SEBI peak-margin framework. Circular SEBI/HO/MRD2/DCAP/CIR/P/2020/127, dated 20 July 2020, on the framework to enable verification of upfront and peak intraday margin collection; phased from December 2020 and reaching 100 percent by September 2021, which removed the intraday leverage that the bracket and cover order products were built on. sebi.gov.in
• Indian retail derivatives outcomes. SEBI's study of the profit and loss of individual traders in the equity derivatives segment (September 2024) is the source of the FY22 to FY24 figure quoted in this guide, and the context for the section on why an in-the-moment exit decision is a poor thing to rely on. sebi.gov.in
• Exchange order types and trigger behaviour. The exchanges define the limit, market, stop-loss market and stop-loss limit order types, and the trigger behaviour that a bracket order's protective leg depends on. Which types are supported in which segment has changed over time, so treat the exchange as the source rather than any article. nseindia.com
• Price bands and trading halts. Daily price bands on individual securities and market-wide index circuit breakers are set by SEBI and the exchanges. They are the mechanism behind the third column of the failure figure, in which a released stop-loss order finds no counterparty because the security is locked at a band. sebi.gov.in
• Bracket-order product mechanics. The three-leg structure, the one-cancels-the-other exit linkage, the optional trailing stop and the intraday auto square-off reflect the standard specification of the product as offered by Indian retail brokers before its withdrawal. Partial-fill handling, linkage behaviour, square-off timing and current availability are broker-level facts that vary and change; verify them with your own broker rather than from any secondary description, including this one.