· Mis à jour
Apprenez la formule de taille de position, le placement du stop, l'espérance et les maths du drawdown qui protègent un compte de trading.
Auteur: Charts Quest
Créez un compte ChartsQuest gratuit pour lire l'article complet — et obtenez les deux premiers niveaux de la quête.
Inscription gratuiteIf you only ever learn one thing about trading, learn this: the traders who survive long enough to get good are not the ones with the cleverest entries. They are the ones who decided, before they ever clicked a button, exactly how much they were willing to lose on any single idea — and then never broke that promise to themselves.
That sounds almost boring. It is supposed to be boring. Risk management is the unglamorous plumbing of trading, and like plumbing, you only notice it when it fails. This guide walks through the actual arithmetic — the position sizing formula, stop placement, expectancy, drawdown recovery, and risk of ruin — in plain English, with worked numbers at every step. None of it requires anything beyond basic multiplication and division. All of it matters more than any chart pattern you will ever learn.
One framing note before we start: nothing here is a promise that trading will work out for you. Most beginners lose money, and risk management does not turn a bad strategy into a good one. What it does is far humbler and far more important — it makes sure that while you are learning, no single mistake, and no single bad week, can take you out of the game.
Beginners almost always study trading in the wrong order. They start with entries: which candlestick pattern to buy, which indicator crossover to wait for, which level to watch. Entries are the fun part. They feel like the skill.
But think about what actually determines whether an account survives. Imagine two traders using the exact same entry signals on the exact same charts:
Now suppose the strategy hits a perfectly normal rough patch: seven losing trades in a row. This is not a disaster scenario — for a strategy that wins half the time, a streak of seven losses will show up eventually if you trade long enough, the same way seven heads in a row eventually shows up if you flip a coin all afternoon.
Trader A finishes the streak down roughly 6.8% (each loss compounds slightly, so seven 1% losses do not quite add to 7%). She is annoyed, but functional. Her account is at about 93.2% of where it started, and she needs roughly a 7.3% gain to get back to even — entirely plausible.
Trader B finishes the same streak down roughly 52%. Each 10% loss compounded: after one loss he has 90% of his account, after two he has 81%, after three 72.9%, and after seven about 47.8%. He now needs to more than double his remaining money just to get back to where he started. Same strategy. Same signals. Same market. One trader is fine; the other is effectively finished.
That is the whole argument in one example. Your entry method decides how often you are right. Your risk management decides whether being wrong — which will happen constantly, forever, no matter how good you get — is survivable. Survival is the prerequisite for everything else, which is why it comes first.
There is a second, quieter reason risk comes first. In trading, almost nothing is under your control. You do not control whether price goes up or down after you enter. You do not control news, other participants, or volatility. The one number you control completely, on every single trade, is how much you lose if you are wrong. It would be strange to obsess over the things you cannot control and improvise the one thing you can.
The foundation of everything that follows is a single rule: decide a fixed percentage of your account that you are willing to lose on any one trade, and never exceed it. This is your risk budget, sometimes called your "R" or your risk per trade.
For most people learning, the sensible range is 0.5% to 1% of the account per trade. On a $5,000 practice account, that means $25 to $50 per idea. Not per share, not per day — per trade, total, if your stop is hit.
One percent feels insultingly tiny when you first hear it. Here is why experienced traders treat it as a ceiling rather than a floor:
Losing streaks are normal, not rare. A strategy that wins 50% of the time has roughly a 1-in-128 chance of seven straight losses on any given run of seven trades. Trade a few hundred times and a streak like that becomes close to inevitable. At 1% risk, seven straight losses costs you about 6.8% — uncomfortable but recoverable. At 5% risk, the same streak costs about 30%. At 10%, about 52%. The streak is identical; only the sizing decides whether it is a bruise or an amputation.
Small risk keeps your judgment intact. There is a practical psychological point hiding in the math. When a single trade can cost 10% of your account, you will feel it in your body. You will move stops, exit winners too early, hesitate on valid setups, and "revenge trade" after losses. At 0.5–1%, individual trades stop feeling like life-or-death events, which is exactly what lets you follow your plan. The budget protects your account and your decision-making at the same time.
You are paying tuition while you learn. A beginner's early trades are practice, and practice has an error rate. A 1% risk budget means your first hundred trades — the ones where you will make most of your mistakes — cannot cost you more than a fraction of your account even if they go badly. You are buying yourself a long runway.
Treat the risk budget the way you would treat a household budget category that genuinely cannot flex — rent, not entertainment. The budget is set once, in advance, when you are calm. It is never adjusted mid-trade, never doubled because a setup "looks especially good," and never increased to win back a loss faster. Every rule in the rest of this guide assumes the budget is fixed. The moment it becomes negotiable, all the math below stops protecting you.
Here is the single most useful equation in retail trading:
Position size = risk budget ÷ stop distance
In words: take the dollar amount you are willing to lose on this trade (your risk budget), and divide it by the amount the price would have to move against you before you admit you are wrong (the distance from your entry to your stop-loss). The result is how many units — shares, coins, contracts — you can hold.
Notice what this formula does and does not let you choose. You choose your risk budget (fixed, in advance). The chart chooses your stop distance (more on that in the next section). The position size is then calculated, not chosen. This is the inversion that separates disciplined traders from gamblers: a gambler picks a position size that feels exciting and hopes the loss is tolerable; a disciplined trader fixes the tolerable loss and derives the size.
Position size = $100 ÷ $2.00 = 50 shares.
Check it: if the stop is hit, you lose 50 shares × $2.00 = $100. Exactly 1% of the account. The position is worth 50 × $50 = $2,500 — a quarter of the account in market exposure, but only 1% of the account at risk, because you have a defined exit.
Same account, same $100 budget, same $50 entry — but this time the structure lets you place a stop at $49.20, only $0.80 away.
Position size = $100 ÷ $0.80 = 125 shares.
If stopped out: 125 × $0.80 = $100. Same loss as before, even though the position is two and a half times larger. This is the formula's quiet elegance: tighter stops allow larger positions at identical risk, and wider stops force smaller positions. Risk stays constant; size floats.
Position size = $100 ÷ $160 = 0.625 units.
Total position value: 0.625 × $2,000 = $1,250 — only 12.5% of the account. The volatility of the asset, expressed through the wide stop, automatically shrank your exposure. You did not need a separate rule saying "be careful with volatile assets." The formula handled it.
Position size = $20 ÷ $12 = 1.67 shares. If your platform only allows whole shares, one share risks $12 (fine, under budget) and two shares risk $24 (over budget — not allowed). And if a sensible stop produced a size smaller than the minimum tradable unit, the correct answer is to skip the trade entirely. "The math does not allow this trade at my account size" is a complete, legitimate, professional reason to pass. The formula is not just a sizing tool; it is a filter.
The formula needs a stop distance as input, and the quality of that input matters enormously. There are two ways people choose stops, and only one of them makes sense.
An arbitrary stop is set by a number that has meaning to you but none to the market: "I'll risk 2%," "I'll use a $1 stop because it is round," "I'll put it where my loss equals $50." The problem is that price does not know or care where your pain threshold is. An arbitrary stop frequently lands in the middle of the price territory where the asset normally fluctuates, so you get stopped out by routine noise on trades where your actual idea was never proven wrong. You end up taking real losses on ideas that went on to work — the most demoralizing outcome in trading.
A structural stop sits at the price where your reason for taking the trade is objectively invalid. If you bought because price bounced off support at $48, then the trade idea is "buyers defend $48." The structural stop goes slightly below $48 — say $47.60, leaving a small buffer for noise and false pokes — because if price closes below that, buyers did not defend the level and your thesis is simply wrong. Likewise, if you entered on a breakout above resistance, the structural stop sits back below the broken level, because a return below it means the breakout failed.
A structural stop answers the question "where would I no longer want to be in this trade?" rather than "how much am I willing to lose?" The risk budget already answered the second question; the stop's only job is to answer the first.
Put together, every trade follows the same sequence:
Beginners do this backwards: pick a size that feels good, then look for somewhere to hide the stop. The backwards version guarantees that your stop is in a meaningless place, your risk is uncontrolled, or both.
Here is the most common way a beginner's account actually dies. Not in one dramatic blowup — in a specific, repeatable behavioral failure: moving the stop after the trade is open.
The script is always the same. You enter long at $50 with a stop at $48. Price drifts down to $48.30. The trade is nearly dead — and suddenly your brain produces a flood of reasons why $48 was too tight: "the bigger support is really $46," "I don't want to get wicked out," "it just needs more room." You move the stop to $46. Sometimes price bounces and you feel vindicated. That is the worst possible outcome, because the lesson you learn is that breaking the rule works.
Run the numbers on what actually happened. Your original position was sized for a $2 stop: $100 ÷ $2 = 50 shares, risking $100. The moment you moved the stop to $46, your stop distance became $4 — but your size was still 50 shares. Your risk on the trade silently doubled to $200, or 2% of the account. You did not decide to risk 2% when you were calm and planning. You decided it while losing, under stress, in the exact mental state where humans make their worst financial decisions. And if price grinds to $46.30, the same brain will offer to move the stop to $44 — now risking $300 on a trade budgeted for $100.
Notice the asymmetry: people widen stops on losers but rarely let winners run with equal patience. So the habit selectively enlarges your losses while leaving your wins the same size. A strategy that was breakeven on paper becomes steadily losing in practice — entirely because of this one behavior. In trade journals of struggling traders, "moved my stop" is the single most common entry next to the biggest losses.
The fix is procedural, not motivational: the stop is placed at the structural invalidation point at entry, and it may only ever move in the direction that reduces risk (for a long trade, upward — never downward). If the original stop was genuinely wrong, the correct response is to exit, re-plan, and re-enter as a fresh trade with a fresh, properly sized position. Stops are one-way doors. The moment they become adjustable under pressure, the position sizing formula — and everything it protects — is fiction.
Once losses are capped and sized, you can think clearly about a question that confuses almost every beginner: how often do you need to be right?
The answer comes from expectancy — the average amount you make or lose per trade, over many trades:
Expectancy = (win rate × average win) − (loss rate × average loss)
Suppose you risk $100 per trade (1% of $10,000), your average winner makes $200 (a 2-to-1 reward-to-risk ratio), and you win 40% of the time.
You are wrong six times out of ten and still come out ahead over a large sample, because the average win is twice the average loss. Now flip it: a 60% win rate where winners average $50 and losers average $100 gives 0.60 × $50 − 0.40 × $100 = $30 − $40 = −$10 per trade. Right most of the time, losing money anyway.
Three lessons fall out of the arithmetic:
Win rate alone tells you nothing. A high win rate with big losses loses; a modest win rate with controlled losses and decent winners can be fine. The two numbers only mean anything together.
Capping losses is half of expectancy. The "average loss" term is the part you control directly with stops and sizing. Every time you widen a stop, you inflate that term and drag expectancy toward negative — which connects this section to the previous one.
Expectancy only emerges over many trades. Twenty trades tell you almost nothing; randomness dominates small samples. This is one more argument for small risk per trade: you need to survive long enough for a large sample to play out before you even know whether your approach has positive expectancy at all. And be honest with yourself about the hard part — most untested ideas have negative expectancy, and no amount of sizing discipline turns a negative-expectancy strategy profitable. Sizing keeps you alive while you find out.
Here is the cruelest piece of arithmetic in trading, and every beginner should be able to recite it: losses and gains are not symmetric. A percentage loss requires a larger percentage gain to recover, and the gap explodes as the loss deepens.
Start with $10,000 and lose 10%: you have $9,000. To get back to $10,000 you need to gain $1,000 — but $1,000 from a $9,000 base is an 11.1% gain. Lose 10%, need 11.1%. Mildly unfair. Now watch it accelerate:
| Drawdown | Account becomes | Gain needed to recover |
|---|---|---|
| −5% | $9,500 | +5.3% |
| −10% | $9,000 | +11.1% |
| −20% | $8,000 | +25% |
| −33% | $6,700 | +49.3% |
| −50% | $5,000 | +100% |
| −75% | $2,500 | +300% |
| −90% | $1,000 | +900% |
The relationship is not linear — it is a cliff. At −10% the recovery is a normal task. At −50% you must double your money just to be back where you started, which is a feat most professionals never achieve quickly, attempted here by a trader who just demonstrated they can lose half an account. At −90%, even a spectacular +100% run leaves you down 80%.
This table is the real reason risk per trade is kept tiny. The goal is to make deep drawdowns arithmetically unlikely. At 1% risk per trade, reaching a 20% drawdown takes on the order of 22 consecutive losses — astronomically rare for any strategy that wins even a third of the time. At 10% risk per trade, a 20%+ drawdown takes just three bad trades, and a 50% drawdown takes seven. The table also explains why your reaction to a drawdown should never be to size up to "win it back faster": from −20%, doubling risk to recover faster also doubles the speed at which −20% becomes −40%, where the recovery requirement jumps from +25% to +66.7%. The hole gets harder to climb out of precisely as you dig faster.
The discipline is to do the opposite of instinct: when in a drawdown, keep risk the same or reduce it, and let small consistent process climb you out.
"Risk of ruin" is the probability that an account eventually hits a level it cannot recover from — zero, or a drawdown so deep the trader quits or can no longer trade sensibly. You do not need the formal formula; you need its three intuitions.
First: with oversized risk, ruin is close to certain, not just possible. Think of betting on coin flips with an edge in your favor — say you win 55% of flips. If you bet half your money on each flip, you will still go broke with near certainty given enough flips, because a short unlucky run wipes you out before the edge can express itself. The edge is real, but ruin arrives first. Risk of ruin is the formal version of "the casino can be beatable and still bankrupt you if you bet too big."
Second: ruin probability falls off a cliff as bet size shrinks. The relationship between per-trade risk and ruin probability is dramatically nonlinear. Going from 10% risk to 5% does not halve your ruin odds; it can cut them by orders of magnitude. Going to 1% can push ruin from "likely within a year" to "would require a losing streak so long it implies your strategy is broken anyway" — at which point the stop on your career (stop trading, reassess) triggers long before the account is gone.
Third: ruin is absorbing. Every other state in trading is recoverable. Drawdowns end, bad strategies get fixed, bad habits get unlearned. Ruin is the one state with no next move. That is why every rule in this guide is shaped around making ruin nearly impossible rather than making profits fast: an unlikely outcome that is irreversible deserves more respect than a likely outcome that is temporary.
A practical companion rule many traders add: a daily or weekly loss limit (for example, stop trading for the day after losing 2–3% of the account, roughly two to three full losses). This caps the damage from tilt — the bad decisions that cluster after losses — and acts as a circuit breaker between you and the ruin math.
All of the arithmetic above assumes you actually do what you planned. The journal is how you find out whether that is true.
A useful journal is not a diary of feelings (though noting your emotional state helps). It is a record that lets you audit yourself with numbers. For every trade, log:
After thirty or fifty trades, patterns emerge that are invisible day to day. You discover your actual average loss is $140 on a $100 budget — meaning you widen stops more than you admit. You discover your win rate on one setup is double another's. You discover that your three biggest losses all share the same note: "moved stop." You compute your real expectancy from real data instead of guessing. The journal converts trading from a stream of emotional events into a dataset you can improve, and it is the cheapest edge available to any beginner. Review it on a schedule — weekly is plenty — and look for one behavior to fix at a time, not ten.
Here is the entire guide compressed into a checklist you can run in under a minute, every time, no exceptions:
A trade that fails any step is not a smaller trade — it is no trade. The routine feels mechanical because it is supposed to be. The creativity in trading belongs in finding setups; the execution should be as boring as filing taxes.
It feels slow, but run the expectancy math: at 1% risk with a modest positive expectancy of 0.3R per trade (entirely realistic for a working strategy), forty trades a month compounds meaningfully over a year — while keeping any single mistake survivable. The traders who try to shortcut this with 5–10% risk are not growing faster on average; they are sampling from a distribution where a normal losing streak ends the account. Slow and alive beats fast and ruined, and the table in the drawdown section is the proof.
No — that is an arbitrary stop. A fixed percentage ignores both the structure of the chart and the volatility of the asset. On a quiet stock, 2% might be far beyond any sensible invalidation point; on a volatile asset, 2% might sit inside ordinary hourly noise and get hit constantly on ideas that were never wrong. Let the chart set the stop at the price that invalidates your idea, then let the formula set your size. The stop answers "where am I wrong?"; the budget answers "how much can being wrong cost?" — keep those two jobs separate.
Always round down, never up. If the formula says 1.67 shares, trade 1 — your risk comes in under budget, which is fine. Rounding up to 2 puts you over budget, which defeats the entire system. If rounding down takes you below the minimum tradable size, the trade is too large for your account at that stop distance, and the correct move is to skip it. Many platforms now support fractional shares, which removes the problem for stocks.
Use a real order. A "mental stop" works exactly until the first time it matters: price hits your level, your brain produces a reason to wait one more candle, and the controlled $100 loss becomes an uncontrolled $300 one. The entire value of a stop is that it executes when you are at your worst — distracted, asleep, or rationalizing. One honest caveat: in fast or gapping markets, a stop order can fill worse than its trigger price (slippage), so the cap is firm in normal conditions but not absolutely guaranteed. That is a reason to size conservatively, not a reason to skip the stop.
More than feels intuitive. Twenty trades are nearly meaningless — a coin-flip strategy can easily go 13–7 by luck. By fifty trades you have a rough sketch; by one hundred or more, your journal numbers start to mean something, provided you traded the same approach consistently throughout. This is another argument for tiny risk per trade: gathering a hundred trades of honest data at 1% risk costs you a survivable amount even if the strategy turns out to be mediocre. Treat your first hundred trades as paid market research, not as your fortune-building phase.
Yes — and paper trading is exactly when to build the habits, because the habits are the hard part. Run the full routine: fixed budget, structural stop, calculated size, journal entry, no mid-trade stop widening. The one thing paper trading cannot teach is how different the same decisions feel with real money attached, which is why the standard advice is to move to real money only in very small size at first, keeping the identical process. If your discipline collapses when stakes become real, better to learn that at 0.5% of a small account.
Position sizing and stop discipline are the foundation, but they work best alongside an understanding of where sensible stops live on a chart — which means reading support, resistance, and structure. That is exactly what the interactive lessons at ChartsQuest are built to teach, one small concept at a time with practice charts. A good next read is our guide to false breakouts and liquidity traps, which covers why the "obvious" stop placements get targeted — and how structural thinking helps you place yours better.
This article is educational content only, not financial advice; trading involves substantial risk of loss, and you should never trade with money you cannot afford to lose.
ChartsQuest est fourni à des fins éducatives uniquement. Rien ici ne constitue un conseil financier, juridique ou de trading.
Certains contenus de ChartsQuest peuvent être créés, édités ou accélérés à l'aide d'outils d'IA.
Tilt, trades de revanche, et stops déplacés tuent plus de comptes que la mauvaise analyse. Un guide pratique sur les checklists, le journal et la pensée orientée processus.
Lire →
MembresLes quatre figures les plus courantes : anatomie, règles de confirmation, objectifs mesurés et — surtout — comment chacune échoue.
Lire →
MembresUne approche structurelle des tendances : comment les définir objectivement, tradez les pullbacks plutôt que chasser, et reconnaître la cassure qui les termine.
Lire →