Stops that are decisions, not hopes
A stop belongs where your reason for being in the trade stops being true. Put it where your discomfort begins and the market will find it for you.
There are two completely different things people call a stop.
The first is a price at which your thesis is wrong. You bought because the level held; the stop sits below the level, because if it breaks, the reason you are in the trade has gone.
The second is a price at which you will have lost as much as you can stand. It is placed by working inwards from the pain, not outwards from the chart.
Only the first one is a stop. The second is a risk budget wearing a stop's clothing, and conflating them is how traders end up with stops that are simultaneously too tight to survive and too loose to protect them.
The order of operations
This is the fix, and it is the whole chapter in two lines:
- The chart decides where the stop goes.
- The stop distance then decides how big the position is.
Never the reverse. If the structurally correct stop is ₹18 away and ₹18 × your intended quantity is more than you can risk, the answer is a smaller quantity — not a closer stop. Moving the stop inwards to fit the position you wanted does not reduce your risk. It converts a trade with a sensible invalidation level into one that will be stopped out by noise, repeatedly, while being wrong about nothing.
Tight stops are not safer
It feels as though a closer stop must mean less risk. Per trade, it does. Over a hundred trades, usually not.
Halving your stop distance lets you double your size for the same rupee risk — so far, neutral. What is not neutral is the hit rate. A stop at half the distance is hit far more than twice as often, because price spends most of its time making small moves and only occasionally makes large ones. You have kept the risk per trade and roughly doubled the number of times you pay it, while keeping the same number of chances at the large move.
That is the real trade-off, and it is why there is no universally correct stop distance — only one that is consistent with how far the instrument ordinarily travels. A stop inside the instrument's normal daily range is not a stop, it is a toll.
A stop is an instruction, not a guarantee
Three ways a stop fails to do what you assumed, all of them routine.
It gaps through. Your stop is at ₹102 and the instrument opens at ₹94. There was no trade at ₹102 for you to exit at. Your loss is 8 rupees worse than the number you budgeted, and no order type prevents this. This is why position size must leave room for a loss somewhat larger than the stop implies, and why a single position should never be large enough that the gap alone would be serious.
It triggers and does not fill. A stop-loss with a limit price becomes a limit order once triggered. If price is moving fast, that limit may never trade, and you are left holding a position you believed you had exited — the worst of both worlds, because you have stopped watching it.
The order type you wanted does not exist. On NSE, stop-loss orders with a market condition (SL-M) were discontinued for option contracts with effect from 27 September 2021, under circular NSE/FAOP/49677 dated 21 September 2021; such orders are rejected by the exchange. The stated purpose was to limit the damage from freak trades. Stop-loss with a limit condition (SL-L) remains available. BSE later discontinued SL-M across its segments from 9 October 2023.
That last point has a direct practical consequence, and it is the reason this platform's order ticket offers Limit and Stop-Loss only, and sends no market orders: in options, the order type that guarantees a fill is not available to you. Every protective order you place in options is a limit order with a trigger, and a limit order can fail to fill.
| What you want | What exists in NSE options | What it does |
|---|---|---|
| Exit at any price, now | nothing — SL-M discontinued 27 Sep 2021 | — |
| Exit when price reaches X | SL-L: trigger at X, limit at Y | becomes a limit order at Y when X trades |
| Enter or exit at a price or better | Limit | fills at your price or not at all |
The workaround everybody uses is to set the limit price some distance beyond the trigger — trigger at ₹102, limit at ₹99 — which buys you a much better chance of filling at the cost of accepting a worse price. That is a deliberate trade and worth making consciously. Setting the limit equal to the trigger is the common mistake: it is the version most likely to trigger and never fill.
Where the stop level comes from
Three defensible methods, and one that is not.
Structure. Below the swing low, beyond the range, through the level that defined your setup. The best method when your thesis is about a level, because the stop and the thesis are then the same statement.
Volatility. A multiple of a recent average range. Honest, mechanical, and the right default when your entry is not about a specific level. It adapts: the same multiple gives a wider stop in a violent week, which is correct, because the noise is wider too.
Time. "If this has not worked within forty minutes, I am wrong." Underused and often excellent, especially for mean-reversion setups where a thesis that is going to work tends to work quickly. A time stop also has the rare property of being immune to the noise problem entirely.
A round number of rupees. Not a method. "I'll risk two thousand rupees" is a risk budget, and it belongs in the quantity calculation, not on the chart.
Two traders, same entry at ₹120, same ₹6,000 risk budget.
Trader A decides the position first: "750 units feels right." That forces the stop to ₹112 — eight rupees, well inside where this option moves in an hour. Over twenty such trades they are stopped out fourteen times by noise and catch the move six times.
Trader B reads the chart first: the level that matters is under ₹102, so the stop is ₹18 away, so the position is 333 units, rounded to 300. Over the same twenty trades they are stopped out seven times and catch the move thirteen times.
Trader B holds less than half the size and makes considerably more money, because they are paying their risk budget half as often for the same set of opportunities. Trader A is not taking more risk per trade — they explicitly are not — they are taking the same risk twice as often, which is the same thing said in a way that is harder to notice.
Moving a stop
One rule, and it is worth being absolute about: a stop moves in the direction of the trade, never against it.
Tightening a stop as a position works is a legitimate strategy — it is what trailing stops do, and the module on options covers how profit protection modes formalise it. Widening a stop because price is approaching it is not a strategy. It is the decision you already made, unmade at the worst possible moment, with the least information you will ever have about the trade.
If you find yourself reasoning that this particular stop was badly placed, you may well be right. Honour it anyway, and place the next one better. The cost of honouring a bad stop is one bounded loss. The cost of learning that stops are negotiable is every loss after it.
What to take from this chapter
- A stop is where your thesis is wrong. A risk budget is how much you can lose. They are different things and only one of them belongs on the chart.
- The chart decides where the stop goes; the stop distance then decides the quantity. Never the reverse.
- Tight stops do not reduce risk so much as raise the frequency of paying it — and frequency is what moves your break-even hit rate.
- A stop is an instruction, not a guarantee: it can gap through, or trigger and never fill.
- NSE discontinued SL-M for options from 27 September 2021 (NSE/FAOP/49677), so every protective order in options is a triggered limit order. Setting the limit equal to the trigger is the version most likely to fail to fill.
- Derive the level from structure, volatility or time. A round number of rupees is a budget, not a level.
- Stops move with the trade, never against it. One bad stop honoured costs one loss; a stop treated as negotiable costs all of them.