Stops & targets

It is 10:05 ET (22:05 ICT, winter). NQ has balanced between 24,960 and 25,030 all morning; a 5-minute bar closes at 25,042 with its low at 25,021. Two traders buy 25,040 on the breakout. Trader A stops at 25,020, under the breakout bar’s low, and targets 25,100. Trader B stops at 25,030 because “$200 is what I can afford”, and sets no target. At 10:20 price dips to 25,027, then rallies to 25,100 by 11:10. A makes $1,200 before costs on one NQ. B loses $200 on the correct read. Same idea, same entry: the stop’s location decided both outcomes. Synthetic example.

Key terms

• Structure stop (dừng lỗ theo cấu trúc): a stop just beyond the price whose loss proves the trade idea wrong Definition. The opposite is a dollar-comfort stop, chosen for what you can stand to lose.
• Target (mục tiêu): the price where you planned, before entry, to take the profit Definition. It is a planned exit, not a forecast.
• R:R (tỷ lệ lợi nhuận trên rủi ro): reward ÷ risk, in points or dollars, computed before entry Definition. Trader A: 60 ÷ 20 = 3:1.
• Break-even win rate (tỷ lệ thắng hòa vốn): the share of trades that must win for the method to pay for itself: (risk + cost) ÷ (risk + reward) Definition. RSK1.01 owns the formula; this lesson applies it to stops and targets.

Term you need: σ (sigma). One standard deviation of the day’s move, read from implied volatility: price × VXN ÷ √252 for NQ. At a 25,000 settle and VXN 24, σ is about 378 points. OPF4.03 owns the ruler; M4.08 puts it on your morning sheet.

Chart first: what costs do to the bar you must clear

The curve is the mechanism of this lesson: the stop sets the risk, the target sets the reward, and the cost is a fixed dollar amount that weighs more the smaller the risk is.

How it works

Stops: placed where the idea is wrong

The stop is decided first, even though it triggers last. Ask: what price would prove this trade idea wrong? For Trader A’s breakout, the idea is “price accepts above the balance”. A trade back under the breakout bar’s low says that acceptance failed, so the stop goes just under it. That is a structure stop: below support on longs, above resistance on shorts.

Leave a small buffer beyond the level, a few ticks, rather than sitting exactly on it House rule. Stops tend to cluster just beyond obvious round prices, and in currency markets those clusters measurably sped up moves once they were hit (Osler 2003, 2005) MeasuredOther market. Whether NQ stop clusters behave the same way is not yet measured.

A stop chosen for dollar comfort is a number with no connection to the idea. Trader B’s 10-point stop was about 0.03σ on a 378-point day, well inside ordinary noise, so it was likely to be hit whether or not the read was right. When the structure stop costs more than your risk budget, change the size (fewer contracts, or MNQ instead of NQ), not the stop. M6.09 and RSK1.05 compute the contracts from the stop distance.

Time stops, stop orders and the one direction a stop moves

Time stop (dừng lỗ theo thời gian): if the idea needed acceptance within a set time and nothing happened, exit House rule. Write the time before entry (“not working by 10:40 ET, I’m out”); a dead trade ties up risk and attention.

Stop orders Definition: a stop-market order becomes a market order when price trades at the trigger, so it fills at the next available price, which can be worse than the trigger in a fast market or a gap. A stop-limit caps the price but may not fill at all, leaving you in the trade. For intraday index futures we default to stop-market House rule. MKT2.06 covers brackets and order hygiene.

Never widen a stop (nới dừng lỗ) House rule. Moving it further away replaces the plan’s defined cost with hope. A stop you only keep in your head is not a stop: enter it as a resting order with the entry, as one bracket. Moving the stop toward breakeven and trailing it are management decisions, covered in M6.11.

Targets: planned exits, checked against the ruler

A trade with no target has no R:R, so you cannot know what win rate it needs. Take targets from structure: the opposite edge of the balance, the next reference area on your sheet (M4.07), or a projection of the balance height. Treat each as a planned exit, not as a price the market owes you House rule. “I’ll trail and see” is not a target; trailing belongs to management after the first target (M6.11).

Myth / hypothesis: “After a breakout, price travels the height of the balance (the measured move).” Not yet measured in our lab on NQ Lore. Pre-market levels as a group were touched and respected no more often than random prices at the same distance (our lab, 721 sessions) Measured: no edgeNasdaq index. Use structure to place a target you can plan around, not because the level pulls price.

Then check the target’s distance against the σ ruler. Trader A’s 60-point target is 0.16σ: distance is no obstacle on a normal day, and the real question is whether it comes before the stop. A target 450 points away is about 1.2σ. On the Nasdaq-100 index, the day’s high or low got beyond 1σ from the prior close on about 38% of days, counting both directions MeasuredNasdaq index, so one direction alone is rarer still. A target beyond 1σ needs an unusual day; write it down as such.

Break-even with costs, and one paragraph on expectancy

Compute R:R before entry, then the break-even win rate with your round-trip cost (commission, fees and slippage, entry and exit together). With a $400 risk and $15 a round trip on one NQ:

The cost is fixed in dollars, so a tighter stop makes it a bigger share of the risk. Trader B’s $200 stop at 1:1 needs 53.8%, not 51.9% Definition. M10.02 converts costs into ticks per contract.

Expectancy (kỳ vọng) = win rate × average win − loss rate × average loss − cost per trade Definition. It is the number that decides whether a method makes money over many trades; RSK1.01 owns it, with the streak and drawdown lessons that follow.

Our minimum: 2:1 planned House rule. It keeps the bar you must clear, and the cost’s share of the reward, low. It is a filter, not an edge. If a setup has no directional edge, a wider target is simply hit less often: in one trader’s pre-registered NQ study, every one of 18 stop and target shapes lost money after $15 costs PlausibleOne trader's study (one practitioner’s study, intervals not clustered by session, Plausible at most; M5.14 has the details).

Win rate alone means nothing. 60% winners at 0.5:1 loses money before costs (it needs 66.7%); 30% winners at 3:1 makes money (it needs 25%) Definition. A win rate quoted without R:R and costs is half an equation.

What the data says

Worked example: judging a method by its break-even

Synthetic numbers. A rejection-short method on 1 NQ, planned at 2.5:1. After 60 logged trades: 24 wins (40%), average win $820 (41 points), 36 losses, average loss $340 (17 points), $15 a round trip.

1. Realized R:R: 820 ÷ 340 = 2.4:1, a little under the plan, as early exits and slipped stops usually make it.
2. Break-even with costs: (340 + 15) ÷ (340 + 820) = 30.6%.
3. Expectancy: 0.40 × 820 − 0.60 × 340 − 15 = 328 − 204 − 15 = +$109 a trade.
4. How sure? A 95% interval for 24 wins in 60 runs from about 29% to 53% (EV3.01). The break-even, 30.6%, sits just inside it. The method looks good, and it is not yet proven.

The first 20 trades went 5 wins and 15 losses (25%), below break-even. A 95% interval for 5 of 20 runs from about 11% to 47%: it contains both a losing and a winning method. Quitting at trade 20 would have “proven” a working method broken on noise. Expectancy needs samples, and samples need survival: size small enough to reach them (M6.09, M7.07, M7.08).

The mirror and the failure

Short mirror. Same balance, 24,960–25,030. A 5-minute bar closes at 24,948 with its high at 24,969. Short 24,950, stop 24,970 (above the breakdown bar’s high), target 24,890 (the balance height projected down). Risk 20 points, reward 60, 3:1, break-even 25.9% with $15 costs.

The failure. Same long as Trader A, another day. Price dips to 25,019, one point through A’s stop, then rallies to 25,100. A is out for −$415 with costs, on a right read and a correct process. No stop placement avoids this: a structure stop is right about where the idea fails, not about every probe. That is why the method is judged on its break-even over many trades, and why a stop you widen after this day is a different method from the one you measured.

Prop-firm note. In an evaluation or funded account your stop distance is what the firm’s loss limits see, so size from it (RSK1.05) and enter the stop and target as one bracket with the entry (MKT2.06; PLAT2.08 for trading a Topstep account). Simulated accounts fill orders by the simulator’s rules, which can be kinder than live fills (MKT2.03); firms forbid strategies that exploit simulated fills (PROP1.04). Read your firm’s help-center articles before you trade.

Common mistakes

Common mistake: Setting the stop where the pain feels affordable and skipping the target. The affordable stop sits inside normal noise and loses on right reads; the missing target means winners are closed by feeling, early when scared, late when greedy.

Common mistake: Reading a planned 3:1 as an edge. R:R sets the bar; only a measured win rate above it, after costs and on enough trades, is an edge. With no directional edge, the wider target is hit less often and the expectancy stays below zero after costs.

Common mistake: Tightening the stop “to save money”. The dollar risk falls, but the fixed cost becomes a bigger share of it (53.8% to break even at 1:1 on a 10-point NQ stop), and a stop inside the retest zone gets hit more often. Cheaper per loss can mean more losses.

Drill: price one short before you take it

Synthetic. NQ balanced between 25,180 and 25,250; price broke above, failed, and you plan a short at 25,248. The failed high is 25,262; your stop goes one point above it, at 25,263. Target: the balance low, 25,180. One NQ, $20 a point, $15 a round trip. Prior settle 25,200, VXN 22.

1. Risk and reward in points and dollars, and the R:R.
2. Break-even win rate before and after costs.
3. σ for the day, and the target’s distance in σ.
4. A friend says “use 25,255 instead, it’s cheaper”. Compute the new break-even and say what it hides.

What we don’t know yet

• Whether NQ stop clusters beyond round numbers speed up moves the way they did in currency markets: not yet measured.
• Whether measured-move or balance-edge targets are reached more often than placebo targets at the same distance: not yet measured.
• The σ ruler is measured on the NDX index; on NQ futures, and from the entry price rather than the prior close, it is extrapolated.

Recap

• The stop goes just beyond the price that proves the idea wrong, as a resting order entered with the trade; resize when it is too expensive, never move it away.
• The target is a planned exit from structure, checked against the σ ruler; a target beyond 1σ needs an unusual day.
• Break-even win rate = (risk + cost) ÷ (risk + reward): about 51.9% at 1:1, 34.6% at 2:1 and 25.9% at 3:1 with $15 on a $400 NQ risk; tighter stops raise it.
• The 2:1 minimum is a house rule, a filter rather than an edge; win rate without R:R and costs is meaningless.
• Judge a method against its break-even with an interval; 15 or 20 trades cannot separate a working method from a broken one.

Quiz

Further reading

• Osler, C. L. (2003), “Currency orders and exchange rate dynamics: an explanation for the predictive success of technical analysis”, Journal of Finance 58(5). Where stop and take-profit orders cluster, and what happens when they are hit.
• Osler, C. L. (2005), “Stop-loss orders and price cascades in currency markets”, Journal of International Money and Finance 24(2).
• Order Flow, Measured: home. A public, pre-registered practitioner study on NQ, 2018–2026. Read the headline cost line and the page “Does a Wider Target Help?”.
• Our lab: the implied-move ruler on 6,482 Nasdaq-100 index days (2001–2026): closes beyond 1σ on 23%, highs or lows beyond it on 38%.
• RSK1.01 for expectancy and break-even; M10.02 for costs in ticks; M5.14 for why R:R cannot create an edge.

Next: M6.09 turns the stop distance into a position size, so the risk this lesson defines fits your account.

Educational content only. Not financial advice. Futures trading involves substantial risk of loss.