By Benjamin Roure L'Hermitte
A stop loss is an order that closes your position automatically at a price you chose in advance, to cap the loss on a trade. It is your seatbelt, but placed wrong it becomes a loss magnet: too tight out of fear, too wide out of hope, or exactly on a round number where many other orders already sit. This guide lays out a stop loss strategy: where to place it, why, what the order really guarantees, and how to stop getting knocked out right before price runs your way.
- The stop goes where your thesis is invalidated, not at an arbitrary distance or where it 'makes a nice ratio'.
- A too-tight stop gets you out on market noise; a too-wide stop wrecks your ratio.
- The stop price is not guaranteed: once hit, the order becomes a market order, filled at the available price.
- A stop placed right on a round number joins a zone where many orders already sit, which one study measured on the currency market: shift it.
- Never move a stop to give the trade 'one more chance'. Moving it against yourself is the worst habit.
A good stop loss answers one question: at what price is my trade idea wrong? As long as you answer that, your stop makes sense. The moment you place it for other reasons (capping the loss at a round number, matching your position size, avoiding being wrong too often), you sabotage it.
The stop is also the piece linking your risk management to your psychology. It's what makes your loss known in advance, and therefore bearable. Without a stop, every losing trade becomes a negotiation with yourself, held at the worst possible moment to win it. Let's see how to place it cleanly.
What a stop loss is, and what it really does
Technically, a stop loss is a stop order. According to the SEC's investor bulletin on stop, stop-limit and trailing stop orders, it is an order to buy or sell once the price reaches a specified level, the stop price, and at that moment it becomes a market order. A sell stop sits below the market for a long position, a buy stop above it for a short position. The same bulletin warns that the stop price is not the guaranteed execution price: it is a trigger, and the price you get can deviate significantly from your level depending on the liquidity available at that moment. In France, the AMF's guide to placing a stock market order describes the same order as a 'trigger threshold' order and says the same thing: once the threshold is reached, you do not control the execution price.
The stop loss has two functions, and it pays to keep them apart. The first is technical: it marks the price at which your analysis is refuted. If you're long because price is holding a support, your stop goes below that support: if it breaks, your reason to be long is gone. The second function is mental: it turns a potentially infinite loss into a known loss, decided coldly, before emotion enters.
A trader without a stop isn't a brave trader, it's a trader who delegated their worst decision to their in-the-moment emotional self. 'It'll come back' is the phrase the stop exists to keep you from saying.
Where to place the stop: the logic
The rule is simple to state: the stop goes where, if price reaches it, your scenario is dead. Concretely it depends on your method, but the healthy references are always structure levels: below the low that validated your entry, beyond the zone you expect a reaction from, on the other side of the breakout you were following. Never a round number 'because that's 20 pips'.
Once that level is identified, you add a small margin to absorb the noise (spread, wicks, normal volatility). Price breathes, it doesn't move in a straight line. A stop glued to the exact level gets swept by a single wick before the move resumes. The margin is calibrated on the instrument's volatility, not on your wish to risk little.
Four ways to set the distance, and what each is worth
Traders set their stop distance in four ways, and no study read for this guide establishes that one beats the others: each has a logic, and your journal will tell you which one holds on your trades.
| Method | Where the stop goes | Its logic | Its limit |
|---|---|---|---|
| Under the structure | Below the low or beyond the zone that validated the entry, plus a margin | The level means something: if it gives way, the scenario is wrong | The margin still has to be set, and an obvious level gathers other orders |
| ATR multiple | At 1.5 or 2 times the ATR from the entry, for example | The distance follows current volatility | A multiple chosen by habit, not by an established rule |
| Fixed percentage | At 1% or 2% below the entry, for example | Simple, constant from one trade to the next | Ignores the instrument's structure and volatility |
| Time limit | Exit if the trade has not moved after a set delay | A trade that stalls ties up capital | Does not cap the loss in price: it complements a price stop, it does not replace it |
The first two combine well: the structure gives the level, the ATR gives the margin. The percentage stop is the easiest to stick to and applies the same distance to a calm market and a turbulent one. The time stop is an exit rule, not a protection: it sits on top of a price stop.
The two symmetrical mistakes
The too-tight stop
Driven by fear of losing, you glue your stop just below your entry to 'risk less'. Result: you get knocked out on normal market noise, again and again, even though your analysis was right. FINRA's page on stop orders in volatile markets describes exactly that mechanism: rapid price movement during a short period of time can trigger a stop order, and the stock might later rebound and resume trading at its prior level. You pile up small frustrating losses and watch price reach your target without you. A too-tight stop doesn't reduce your risk, it increases your loss frequency.
The too-wide stop
Conversely, to 'let the trade breathe', you place a huge stop. Your win rate rises (price has room), but each loss is massive and your ratio collapses. And since size has to shrink to keep the same money risk, the trade pays less when it wins. The wide stop is the favorite disguise of the trader who never wants to be wrong.
| Stop type | Effect on win rate | Effect on ratio | Trap |
|---|---|---|---|
| Too tight | Drops sharply | Improves in appearance | Knocked out on noise |
| Well calibrated | Stable | Coherent | None |
| Too wide | Rises | Collapses | Massive losses |
The obvious-stops trap
Round numbers and visible levels attract stops, and that can be measured. In her study for the Federal Reserve Bank of New York (Staff Report No. 125), Carol Osler examined almost 9,700 stop-loss and take-profit orders placed at a large foreign exchange dealing bank between September 1999 and April 2000, in dollar-yen, dollar-pound and euro-dollar: the requested execution rates cluster strongly at round numbers, and stop-loss orders tend to intensify trends once those levels are crossed. Traders who follow orderflow watch that wave arrive. What the study does not measure is who triggers what: the 'stop hunt', the idea that some player goes after those stops, is a reading, not an established fact. It is laid out, with what supports it and what it lacks, in the guide to liquidity in trading.
The counter isn't to remove the stop, it's to shift it. Place it a bit beyond the obvious level, where its break truly has directional meaning. You risk a little more in distance, but your stop no longer depends on a single round trip through a level that many orders share.
Never move a stop against yourself
Here's the rule that separates disciplined traders from the rest. Once your stop is placed, you never pull it back to give the trade 'one more chance'. Moving a stop against yourself turns a planned small loss into an unpredictable disaster: you abandon your plan at the worst moment, gripped by hope. The only allowed direction for a stop is in your favor, to secure a gain once the trade has run your way.
The trader who moves their stop isn't trying to win, they're trying to be right. The market charges dearly for that difference.
Mental stop or stop in the market?
Some traders use a mental stop: they decide the level but don't place the order. No study read for this guide measures the results of those who stick to it. What has been tested is the effect of an order placed in advance: in the laboratory experiment by Fischbacher, Hoffmann and Schudy on stop-loss and take-gain orders, participants who could program automatic sales showed a significantly smaller disposition effect than the others, because they realized more of their losses. It is a laboratory result, not a measurement on real accounts, and it points one way: the order placed in advance helps you accept the loss you decided on cold. Until your data shows you honor your levels, place a real stop in the market.
Stop order or stop-limit order, the choice changes what is guaranteed the day price jumps past your level. What each order promises, and what it does not, is detailed in the guide to order types.
The stop is only one of the two bounds of a trade. The other, the profit target, and how to set the two together are covered in the guide to take profit and stop loss.
The trailing stop: securing without smothering
Once a trade runs your way, the exit question arises, and the trailing stop is one of the most useful tools for it. The SEC bulletin cited above defines it as a stop order whose stop price is not a fixed price: it trails the market by a defined percentage or amount as long as price moves in your favor, and stays put when price turns. You gradually lock in part of the gain without ever moving the stop against yourself, which is exactly the opposite of the reflex to cut too early.
The trailing stop's trap is gluing it too close to price: the SEC warns, in the same bulletin, that short-term market fluctuations can activate it. As with the initial stop, you must leave margin for noise. The trailing stop should accompany the trend at a distance, not follow every candle. It entrusts your exit to a mechanical rule rather than your fear of the moment.
A trailing stop can also come out of a calculation. Welles Wilder's Parabolic SAR plots a level under price that follows the rise and never moves back down; the lesson on the Parabolic SAR works that calculation by hand and shows what it costs in a market with no trend.
Adapting the stop to volatility
A fixed-distance stop (always 20 pips, for example) has a built-in flaw: market volatility changes constantly. In a calm phase, 20 pips leave comfortable margin; in a turbulent phase, the same 20 pips get swept by the slightest jolt. The right stop is calibrated on the volatility of the moment: wider when the market moves hard, tighter when it's calm, always based on structure and not an arbitrary number.
The classic tool for that is the ATR (average true range). According to TradingView's help page on the Average True Range, the indicator was introduced by J. Welles Wilder in 1978, it measures volatility only and not price direction, and a 14-period lookback is the most common. Placing your stop at a multiple of the ATR, 1.5 or 2 times for example, is a traders' habit, not an established rule. On an asset at 100 with an ATR of 1.5, a two-ATR stop sits at 97: with 100 of risk, size is 33 units (100 ÷ 3, rounded down), for a real risk of 99. And since your size is computed from that distance, your risk in money stays constant no matter what.
The stop is also a planning tool
We see the stop as protection, but it's also a revealer of your trade's quality. If placing your stop where your thesis is invalidated gives a huge distance, it's the sign your entry is far from its invalidation point, so your risk is poorly placed. A good setup is precisely characterized by an entry close to its invalidation, which allows a tight stop and a favorable ratio. The stop thus becomes a quality filter.
Concretely, if a trade's logical stop seems too wide for a decent ratio, the right answer isn't to widen your target or shrink your stop artificially, it's to skip the trade or wait for a better entry. The stop, by forcing you to locate the invalidation precisely, obliges you to take only the trades where the risk/reward ratio is genuinely good. It's one of the most underrated roles of the stop: it helps you say no to mediocre trades before even taking them.
A worked example: stop, size and ratio
Take a trader with 10,000 who accepts risking 1% per trade, or 100. They spot a support at 100, place their stop at 98 with a small safety margin, a risk distance of 2. Their position size is calculated directly from that distance: 100 of risk divided by a distance of 2 gives 50 units. Their target, based on market structure, sits at 106, a distance of 6: the trade's ratio is therefore 3 to 1, and its potential gain 300 for a risk of 100.
This example shows why the stop is never an isolated decision: it mechanically determines your position size and therefore your risk/reward ratio. If that same trader had placed their stop at 90 'to be wider', the distance would be 10, their size 10 units to keep the same 100 of risk, and their potential gain 60 instead of 300, with the target still at 106. The stop is never just a technical detail, it's the pivot around which the whole geometry of the trade is built.
Going from that distance to a number of lots or shares, and how the point value changes it from one market to the next, is detailed in the guide to calculating position size.
Stops and gaps: the risk a stop can't eliminate
A classic stop loss never guarantees your exit price: it triggers a market order, filled at the available price. In a liquid, continuous market, that price is close to your level. During a gap, a major economic announcement or a market open after a significant event, price can jump straight past your stop without ever trading at the planned level. Traders who deliberately trade the gap and go accept that risk knowingly. Your order then fills at the first available price, sometimes far worse than your stop: that's slippage. In the example above, a stop at 98 filled at 96 costs 200 instead of 100: the planned −1 R becomes −2 R.
Two orders move that risk without removing it. The stop-limit order sets a floor on the execution price, but, the SEC writes, it may not be executed at all if price moves away from the limit. The guaranteed stop, offered by CFD brokers, promises execution at the chosen level even across a gap: broker IG's glossary entry on the guaranteed stop states that the broker then accepts the risk of slippage, for a premium IG only charges if the stop is triggered. That is a broker's documentation of its own product, to check with yours. The rest is handled upstream: reduce your size before major economic releases, avoid opening a position right before a high-impact announcement, and accept that, on leveraged markets, an ordinary stop is not an absolute guarantee.
Moving to breakeven: doing it right, not too soon
Moving your stop to your entry point (the classic 'breakeven') once a trade is in profit is a traders' practice, not a research result, and it has a logic: it removes the risk of turning a winner into a loser. Its trap is doing it too soon, as soon as price has moved a few points your way: you end up stopped out at zero by the market's normal breathing, right before the real move starts.
Good practice is to give the trade enough room to breathe before securing breakeven, generally after it's covered a significant portion of the distance to your target, not at the first positive points. Moving to breakeven should protect real progress in the trade, not prematurely neutralize an idea that needs time to develop.
How to calculate the break-even price, costs included, and what this move gains and loses have their own guide: break-even in trading.
What research has measured about stops, and on what
Two published studies test stop rules on long series, and their population has to be read before drawing anything from them. In 'When do stop-loss rules stop losses?', by Kaminski and Lo (Journal of Financial Markets, 2014), the rule studied reduces a whole portfolio's exposure after a threshold of cumulative losses. The theoretical framework says that if prices follow a random walk, such a rule always lowers expected return, and that it can only add value if returns show some persistence. On daily futures price data, applied to a buy-and-hold strategy on index futures, certain stop policies increase expected return and substantially reduce volatility at the longer sampling frequencies.
'Taming Momentum Crashes: A Simple Stop-Loss Strategy', by Han, Zhou and Zhu (working paper, SSRN) applies a stop to a momentum strategy on US stocks from January 1926 to December 2013: each position in the portfolio is closed as soon as its loss reaches 15% from the start of the month. The worst monthly loss of the equal-weighted version goes from −49.79% to −17.43%, that of the value-weighted version from −64.97% to −22.10%, and the Sharpe ratios more than double.
What these two studies do not say: nothing about a stop placed under a structure low on an intraday trade in currencies or futures, nothing about the distance to choose, nothing about the methods in the table above. They are portfolio rules, tested on daily or monthly data of US stocks and indices. They establish that a stop changes the shape of the results, by cutting the worst losses, not that one way of placing it beats the others.
Tracking and measuring it with Tradoshi
The stop doesn't just cap the loss, it carries precious information: your initial risk, and therefore your R. Tradoshi reads the stop from your orders and automatically derives the percentage of capital risked and the R-multiple of each trade, without you entering anything.
- Real risk computed: your stop sets your risk in percentage, read directly from your orders.
- R-multiple per trade: you see whether your winners truly pay several times your risk.
- Missing-stop flagging: trades without a stop stand out, they're your most dangerous ones.
- Consistency over time: your average risk tells you whether you're holding your management or improvising.
Frequently asked questions
Where do I place my stop loss?
At the point that invalidates your trade thesis, not at an arbitrary distance. If you're long because a support is holding, your stop goes below that support, plus a small margin for noise. The idea is simple: if price reaches your stop, your reason to be in the trade is gone.
What stop distance is correct?
There's no universal distance: it depends on market structure and the instrument's volatility. The right stop is the one that lets the trade breathe while keeping a decent ratio. Calibrate the margin on volatility, never on your wish to risk a small amount.
Why do I always get stopped out right before price reverses?
Two causes to check in your journal: a stop that is too tight, which gets hit on the market's normal noise, or a stop placed right on a level that many orders share, such as a round number. Glue your stop less to price, add a margin calibrated on volatility, shift it beyond the obvious level, then count in your journal whether noise exits go down.
Should I move my stop loss during a trade?
Only in your favor, to secure a gain once the trade has run your way (trailing stop, moving to breakeven). Never against yourself to give the trade one more chance: pulling a stop back turns a planned loss into a disaster and is the worst habit a trader can build.
Mental stop or real stop in the market?
Place a real stop in the market until your data shows you honor your levels. The mental stop relies on your ability to act at the critical moment, and nothing guarantees it; what has been tested in the laboratory is that an order placed in advance helps you realize a loss you decided on cold.
What is a stop hunt?
It's the name given, in one reading of the market, to price running through a zone where stops cluster, below a support or a round low, followed by a reversal. What research measures is the clustering of stop orders at round numbers and the intensification of the move once those levels are crossed, on the orders of one foreign exchange dealing bank; nothing in it measures who triggers what. The counter stays the same: don't place your stop right on the level everyone is watching.
Can you trade without a stop loss?
Technically yes, but it's playing Russian roulette. Without a stop, a single position can wipe out weeks of gains, and 'it'll come back' becomes your only plan. The stop makes your loss known and decided coldly; giving it up means handing your worst decision to your in-the-moment emotional self.
How does the stop determine my position size?
The distance between your entry and your stop, combined with your accepted money risk per trade, mechanically sets your size. If you risk 100 and your stop is 2 away, your size is 50 units. A wider stop shrinks your size and your potential gain proportionally: the stop is never an isolated detail, it's the pivot of the trade's whole geometry.
Does a stop loss protect against gaps and economic announcements?
Not completely. A gap can make price jump past your stop without executing at the planned level: the order becomes a market order and fills at the available price, which can turn a planned −1 R into a real −2 R. A stop-limit order caps the price but may not fill; a guaranteed stop, at brokers that offer one, fixes the price for a premium. For the rest, reduce your size before major economic releases.
Do studies show that a stop loss improves results?
On portfolio rules, partly: on index futures, Kaminski and Lo find that certain rules increase expected return and reduce volatility, and Han, Zhou and Zhu that a 15% stop cuts the worst months of a momentum strategy on US stocks. None of these studies covers a stop placed on an intraday trade, nor compares placement methods.