A stop-loss is the simplest risk tool in trading and the one traders break most often. Not because they don't know how to place one, but because they place it in the wrong spot, on the wrong thing, and then move it the moment it matters. Order types are the easy part. Here's the part that actually decides whether your stop protects you.
What a stop-loss is, briefly
A stop-loss is an instruction to exit a trade once it reaches a price that proves you wrong. Its job is not to predict; it's to cap. It turns an open-ended loss into a known, survivable one, decided in advance while you're calm rather than in the moment while you're not.
The order mechanics are worth knowing but simple. A stop-market order exits at the best available price once your trigger hits, so it fills but the price isn't guaranteed. A stop-limit order only fills at your limit or better, so the price is controlled but in a fast move it may not fill at all. On Indian platforms these are your SL-M and SL-L orders. Which matters more than it sounds, and we'll come back to it.
The real question: where do you put it?
This is where most stops go wrong, and where options traders go wrong worst. The instinct is to set the stop on the option's premium: "I'll exit if this ₹120 option hits ₹90." The problem is that premium is noisy. Implied volatility swings, bid-ask spreads widen, and the premium can lurch 20 or 30% on a move in the underlying that's completely normal. A premium-based stop gets hit by ordinary noise, not by being wrong.
The fix: set your stop on the underlying, not the option. Decide the Nifty or Bank Nifty level (or the chart structure) at which your trade idea is actually invalidated, and exit when that level breaks, whatever the premium happens to be. Your thesis was about the underlying's direction; your stop should be too. The premium is just the vehicle. This single change turns a stop that whipsaws you out of good trades into one that only fires when you're genuinely wrong.
The fast-market trap
Now back to order types, because for options it's a real risk. On a sharp move or a gap, a stop-limit (SL-L) order can fail to fill: price jumps straight past your limit and your "stop" does nothing, leaving you holding a losing option that's still falling. In fast, illiquid options especially, a stop that doesn't fill is the same as having no stop at all.
For most options buyers, a stop-market (SL-M) order is the safer default: you give up control of the exact exit price in exchange for actually getting out. A bad fill is almost always better than an unbounded loss on a position you thought was protected.
The rule that makes a stop work: don't move it
You can place a perfect stop on the right level with the right order type and still lose, if you do the one thing that undoes all of it: move it wider when it's about to be hit.
This is the most common and most expensive stop-loss mistake. Price approaches your level, the loss becomes real, and you tell yourself "just give it a little more room." You drag the stop down. An hour later the position is far worse. That isn't risk management anymore; it's hope trading, and the stop you so carefully set became theatre.
A stop only works if it's untouchable once placed. The decision was made by your clear-headed self before the trade; the version of you watching the loss doesn't get to overrule it. Tightening a stop to lock in profit is fine. Widening one to avoid a loss is the exact behaviour stops exist to prevent. If you can't trust yourself not to move it, use a hard SL-M order so the platform enforces what you won't.
The stop defines the "risk" in your risk-reward ratio and the distance that sets your position size. It's the foundation the rest of your risk management is built on. Place it on the level, use an order type that fills, and never widen it.
SubTrades reads your tradebook and finds where your stops fail in practice: the exits that drifted wider, the losers held past any reasonable level, the trades where a stop would have saved the day. Auto detection of psychological patterns. Import your Zerodha, Dhan, Upstox, or Angel One trades and see it on day one. Free during the founding beta.