🛑 Stop losses: where to put them and what they cannot do

🎯 In 30 seconds: A stop loss is the price where you admit a trade is wrong and get out. You decide it before you buy, and it sets your position size. Good stops sit where the reason for the trade is clearly broken, often just past a recent low or a few average daily ranges away. In India a stop order is only valid for the day, it is a trigger and not a guaranteed price, it can jump straight past your level at the 9:15 open, and it cannot sell at all while a stock is frozen at its lower circuit.

What a stop loss is, and why you need one

A shopkeeper who stocks mangoes knows that by the third day, unsold fruit has to go at whatever price it fetches. Waiting another day does not bring the old price back. It just means more mangoes in the bin. A stop loss is the trader's version of that rule, decided while the fruit is still fresh.

A stop loss is a planned exit for a losing trade. It can be an actual order sitting with your broker or the exchange, or a price you watch and act on yourself.

On NSE, a stop-loss order waits in a separate book and is released into the market only when the last traded price reaches the trigger price you set. For a sell stop, that means the price has fallen to or below your trigger.

Stops do three jobs:

• They cap the loss. Without an exit, a 5% loss can become a 50% loss while you wait for it to come back. The math in Risk comes first shows why that is so costly.
• They set the size. You cannot use the position sizing formula without one: shares = (rupee risk - charges) / (entry - stop). See Position sizing.
• They take the decision away from the moment. Deciding where you are wrong while calm is easier than deciding while watching a red number grow in your broker's app.

Four ways to place a stop

The rule that matters most: put the stop where the trade idea is wrong, not where the loss feels uncomfortable. Here are four common methods, all using one hypothetical stock, Kapoor Paints Ltd, bought at ₹400.

1. Structure stops. Place the stop just beyond a price level that should hold if your idea is right, such as the most recent swing low or a support level. Say the stock bounced from ₹381 last week. If it falls below that, the bounce has failed. A stop at ₹379 gives a little room for noise.

2. Volatility (ATR) stops. The Average True Range (ATR) is the stock's typical daily range, including gaps. If ATR is ₹10.50, a stop 2 x ATR below entry sits at ₹400 - ₹21 = ₹379. The idea is to stay outside normal daily wiggle. See Volatility tools.

3. Percent stops. Exit if the stock falls a fixed percentage, such as 8%: ₹400 x 0.92 = ₹368. Simple and consistent, but it ignores both the chart and the stock's normal volatility. 8% is tight for a jumpy small cap and loose for a calm large cap.

4. Time stops. Exit if the trade has not worked within a set time, for example "if the stock is not up at least 1R within 10 trading days, sell." This frees money stuck in trades that are going nowhere and is often combined with a price stop.

In this example the structure stop and the 2 x ATR stop agree at ₹379. When two methods land in the same place, that is a good sign. When they are far apart, prefer the structure level and reduce the share count to keep the rupee risk the same.

SL and SL-M: the two stop orders in India

Indian brokers offer two kinds of stop order. They fail in different ways.

For a sell SL order you set the limit at or a little below the trigger. So a typical SL order on Kapoor Paints might be: trigger ₹379, limit ₹376. Once the last traded price reaches ₹379 or lower, NSE releases your order into the normal order book as a sell limit at ₹376, time-stamped at the moment it triggered. So it joins the queue behind sell orders already waiting at that price. It sells anywhere from ₹379 down to ₹376 if buyers are there. If the price drops straight from ₹380 to ₹372, it does not sell.

Why did the exchanges drop SL-M orders in places? Freak trades. A wrong order can knock a thin market down for a moment, trigger a cascade of stop-market orders and fill them at absurd prices. When BSE removed SL-M orders across segments in 2023, the move followed a freak trade in a Sensex option. A stop-limit with a sensible gap between trigger and limit protects you from selling at a freak price, at the cost of possibly not selling at all.

For a stop whose job is to get you out, many traders set the limit a little further below the trigger, say 1% to 2% for a liquid stock, so the order still fills in an ordinary fast move. Order types in general are covered in Order types.

Stops in India last one day

This catches many beginners. NSE's trading system supports DAY and IOC validity, and its own page notes that good-till-cancelled and good-till-date orders are not available, as per SEBI directives. A DAY order that does not execute is cancelled automatically at the end of the trading day.

So if you hold shares for delivery and place an SL order on Monday, it is gone by Monday evening. Tuesday's open is unprotected unless you enter a new one.

You have two ways to handle it:

• Re-enter the stop every morning before or right after 9:15. Put it in your daily routine.
• Use your broker's "good till triggered" (GTT) feature, if it has one. GTT is a broker service, not an exchange order type: the broker stores your trigger and sends an order to the exchange when the price reaches it. Read your broker's terms. Ask how it behaves when a stock opens below the trigger, whether it places a limit or market order, and what happens if their system has an outage.

Intraday (MIS) positions have their own problem: your broker will square them off near the close at the market price if you have not exited, stop or no stop. See Intraday trading.

The 9:15 gap: when the stop price is not the exit price

Indian markets are shut for about 17 and a half hours between the 3:30 p.m. close and the next morning. Company results, regulatory orders and global moves often land in that window. From 9:00 to 9:15 a.m. IST, NSE and BSE run a pre-open call auction that collects orders and sets a single opening price. If the news was bad, that opening price can be far below yesterday's close, and your stop cannot fill at a price where nobody traded.

Using the trade above: 88 shares, stop ₹379, planned loss about ₹1,992. The stock opens at ₹360.

You cannot remove gap risk, but you can manage it:

• Keep risk per trade small, so a 1.8R surprise is a bad day, not a disaster.
• Know the results date before you buy. Under SEBI's listing regulations, companies have to give the exchange advance notice of the board meeting that will consider results, and those notices appear in the corporate announcements on the NSE and BSE websites. Many swing traders close or shrink positions before results. See Results season.
• Watch the scheduled market events too: RBI policy days, the Union Budget, election results. See Trading news and events.
• Be careful with small, thinly traded stocks, where gaps are larger.

Circuit limits: when a stop cannot fill at all

Gaps are bad. Circuits can be worse. Most Indian stocks that do not trade in F&O have a daily price band set by the exchange: 2%, 5%, 10% or 20% either way from the previous close. Orders outside the band are not accepted. Stocks in the F&O segment have no fixed band. Instead, NSE applies an operating range of 10% to keep out orders at non-genuine prices.

When bad news hits a banded stock, it can fall straight to its lower circuit, the bottom of the band, and stick there. Everyone wants to sell at that price and almost nobody wants to buy. Your triggered stop joins a long queue of sell orders, matched by time priority as buyers trickle in. Often the day ends with most of that queue unfilled. The next day the band resets from the new lower close, and the same thing can happen again.

Say Kapoor Paints sits in the 5% band and is frozen at the lower circuit for three days. Each band is 5% below the previous close: about ₹372.40, ₹353.80 and ₹336.10. If your sell finally fills on day 3 near ₹336, the loss on 88 shares is about ₹5,760 with charges, close to 2.9R.

Some practical points:

• Check a stock's band before you buy. NSE publishes daily price bands, and your broker's quote screen usually shows the upper and lower circuit prices. Reading a stock quote shows where.
• Stocks in narrow 2% or 5% bands, and those under the exchanges' surveillance frameworks, are the ones most likely to lock. Many are small and illiquid. See The rules every Indian trader runs into.
• The same trap works in reverse for short sellers in an upper circuit, but retail investors cannot carry a short in shares overnight anyway. See Short selling.
• Size positions in banded stocks smaller than the formula says, because the true worst case is several bands, not one stop.

Mental stops

A mental stop is a price you plan to act on without placing an order. Some traders use them to avoid having their order caught by a quick dip.

The honest problem: mental stops only work if you are watching, and only if you act. Most beginners do not. Research on the disposition effect shows investors hold losing positions too long, and a mental stop leaves that tendency in charge. See Trading psychology.

If you use one, write it down, set a price alert in your broker's app at that level, and treat the alert as an order you must execute. If you have a job or college during market hours, a mental stop is not a stop.

Moving stops: breakeven and trailing

Moving to breakeven means raising the stop to your entry once the trade is up, say by 1R. In our trade, at ₹421 you would move the stop from ₹379 to ₹400. A reversal then costs you roughly the charges, about ₹150, not the full R.

The trade-off is real. A breakeven stop often sits right where normal pullbacks reach, so you get stopped out of trades that would have worked. Test it in your own records before making it a rule.

A trailing stop moves up as the price rises and never moves down. It can be a fixed rupee amount, a percentage, or a multiple of ATR. In India you usually trail by hand, raising your SL or GTT trigger as the price climbs, because exchange orders cannot trail themselves.

A separate example: a 10% trailing stop on a ₹400 stock starts at ₹360, so R is ₹40 a share and the formula gives 46 shares (₹1,850 / ₹40). The stock peaks at ₹470, so the stop ends at ₹423. Selling near ₹423 gives 46 x ₹23 = ₹1,058, less about ₹110 of charges, about ₹950, or roughly +0.5R.

Notice what the trailing stop cost: the stock was up ₹70 at the peak, and the trade kept ₹23 of it. Trailing stops protect gains but always give some back. Tighter trails give back less but get shaken out more often. There is no setting that avoids both.

A short stop-loss checklist

1. Find the level where the idea is wrong (structure, ATR, or both).
2. Add a small buffer beyond it.
3. Check the stock's price band and whether it is in F&O.
4. Size the position from that stop, after charges.
5. Check for results, a board meeting or other scheduled news before the planned exit.
6. Enter the SL order (trigger and limit) right after the entry fills, and re-enter it each day or use a GTT.
7. Only ever move it in the direction of the trade.

Try it

Open a daily chart of any liquid NSE stock. Mark the most recent swing low and find the stock's ATR (most charting tools show it). Write down three stops: structure, 2 x ATR and 8%. Compute the share count for each with ₹2,000 of risk and ₹150 for charges. Then look up the stock's price band on the NSE quote page and its next board meeting date in the corporate announcements. Finally, open your broker's order window (without placing anything) and find the SL order type: note which fields it asks for, and whether SL-M is offered for that stock.

Sources

• NSE: Trading system (order books, DAY and IOC validity, GTC and GTD not available)
• NSE circular (21 September 2021): SL-M orders for option contracts discontinued from 27 September 2021
• Business Standard (22 September 2023): BSE to discontinue stop loss market orders from October 9 to check freak trades
• NSE: Price bands
• NSE: Pre-open session
• NSE: SEBI turnover fees, STT and other levies
• SEBI (Listing Obligations and Disclosure Requirements) Regulations, 2015, as notified (regulation 29, prior intimation of board meetings)
• Odean, Journal of Finance (1998): Are investors reluctant to realize their losses?