TL;DR. A winning edge does not pay if you blow up before it plays out. Risk management is the small set of rules that keep your account alive long enough for the math to work. The rules are simple, the discipline is hard, and the cost of skipping them is your entire account.
What you'll be able to do
- Calculate position size for any trade given your account, your risk tolerance, and your stop distance.
- Understand why drawdowns are mathematically non-linear and why the third rule on this page exists.
- Set up the kill-switch and daily-loss-limit infrastructure that prevents an ordinary bad day from becoming a career-ending one.
Why this matters
A strategy with a real edge — say, a 0.4 expectancy and a 55% win rate — will eventually make money. Eventually might be ten weeks away. In those ten weeks the strategy might lose ten trades in a row, because losing streaks are normal even for winning strategies. If your position size is small enough to survive a 10-loss streak, you collect the long-run edge. If it isn't, you don't, and the strategy's lifetime expectancy means nothing to you.
Every blown account in retail systematic trading is a variation of this story. The strategy might have been real; the sizing wasn't.
The core idea
Three numbers govern survival:
- Position size — how much of your account is at stake on a single trade.
- Stop distance — how far the market has to move against you before you exit.
- Account loss limit — the total drawdown at which you stop trading entirely and re-evaluate.
The first two together determine your per-trade risk — the dollar amount you lose if a single trade hits its stop. The third determines your portfolio survival floor — the line below which you accept the strategy might be broken and pull capital out.
Get those three right and your account survives almost any losing streak that doesn't involve a bug. Get any one wrong and you're vulnerable to events that other systematic traders would shrug off.
The math of ruin
Drawdowns are non-linear. The percentage gain required to recover from a drawdown is always larger than the drawdown itself, and the relationship gets ugly fast:
| Drawdown | Gain required to recover |
|---|---|
| 5% | 5.3% |
| 10% | 11.1% |
| 20% | 25.0% |
| 30% | 42.9% |
| 50% | 100.0% |
| 70% | 233.3% |
| 90% | 900.0% |
A 50% loss requires a 100% gain to break even. Once you're past 30%, recovery becomes mathematically improbable in any reasonable time frame. The point of risk management is to keep the worst-case scenario above 20%, where recovery is still in normal territory.
The 1% rule (and what it really means)
The most fundamental sizing rule in systematic trading is fixed fractional risk. Instead of trading a constant number of contracts, you size every trade so that hitting the stop costs the same percentage of your current account.
The formula:
Position size = (Account equity × Risk %) / (Entry price − Stop price)
Example. $100,000 account, risking 1% per trade ($1,000). Trading MES, stop placed 10 points away from entry. MES is $5 per point. Per-contract risk = 10 × $5 = $50. Position size = $1,000 ÷ $50 = 20 contracts.
If the stop moves to 20 points away: position size = $1,000 ÷ ($20 × $5) = 10 contracts. Same dollar risk, half the size.
Two things this gets you that fixed-contract sizing doesn't:
- Compounding when winning. As your account grows, your 1% grows with it, so position size scales up automatically.
- Self-protection when losing. As your account shrinks during a drawdown, your 1% shrinks with it, so position size scales down. A 10-loss streak at 1% risk leaves you with ~90% of your account. A 10-loss streak at flat 10-contract sizing on a $100k account ends in disaster.
The "1%" number is a starting point, not a law. Conservative discretionary traders run 0.25–0.5%. Aggressive systematic traders sometimes go to 2%. Above 2% per trade, a 5-loss streak (which is statistically routine) starts to do real damage; above 5%, an ordinary losing month can ruin you.
Hard stops vs. dynamic stops
A hard stop is a fixed price. You enter at 4290, stop at 4280, stop is 4280. Forever. Simple, easy to backtest, easy to mentally commit to.
A dynamic stop adapts to the market. The two most common types:
- Volatility-based. Stop distance =
N × ATR(period). Wide when the market is volatile, tight when it's calm. Wire ATR through a multiplier into Set Stop Loss in the Visual Builder. - Time-based. If the trade hasn't reached its target in N bars, close it regardless. Prevents capital from sitting in stagnant trades indefinitely.
There are also trailing stops (the stop follows price up but not down, locking in profit) and breakeven stops (the stop moves to entry price once the trade reaches some profit threshold, eliminating risk on the remainder). Both are useful patterns; both are easy to wire in the builder.
The best stop is the one that's tight enough to limit losses but wide enough that ordinary market noise doesn't take you out before your edge plays. For volatile instruments, that's usually some multiple of ATR rather than a fixed point distance.
Per-trade vs. per-day vs. per-account limits
Position sizing protects you trade-by-trade. The two layers above it protect you when the combination of trades goes wrong.
- Daily loss limit. The dollar amount you're willing to lose in a single day. When breached, the strategy flattens all positions and refuses to trade again until the next session. Common range: 2–4% of account. For prop-firm accounts this is non-negotiable and set by the firm.
- Account drawdown limit (the "uncle point"). The total drawdown from peak equity at which you stop the strategy entirely and re-evaluate it. Typically 1.5× to 2× the strategy's worst historical drawdown. Hitting this doesn't mean the strategy is dead — it means the assumption "live behavior will resemble backtest behavior" has been falsified enough that you owe the strategy a real audit before risking more capital.
Both limits are configurable in the exported executable as automatic kill-switches. The point is that they fire whether or not you're watching.
Portfolio exposure and correlation
A single-trade view of risk is incomplete. If you're long three different tech stocks simultaneously, you're not diversified — you're 3× leveraged on the tech sector, and a sector-wide selloff hits all three at once.
Two related rules:
- Cap total open exposure. Across all strategies, the sum of position sizes shouldn't exceed some multiple of your account. 2× is conservative; 4× is normal; above 6× requires real conviction that your strategies are uncorrelated.
- Watch portfolio-level correlation. The Portfolio Builder's correlation matrix shows how strategies' daily returns move together. Two strategies with +0.85 correlation aren't a portfolio — they're one trade you've made twice. The strategy correlation page covers crisis-correlation (the tendency for correlations to spike during stress, exactly when diversification is most needed).
A worked example
Three accounts, same starting equity, same strategy, three different sizing rules. Strategy: 55% win rate, average win = 1.5R, average loss = 1R, ~150 trades over the test period.
Sizing — same edge, three outcomes
What this shows. The strategy's edge is identical in all three columns. The 1% sizing produces a smooth equity curve with manageable drawdowns. The 3% sizing produces a more impressive peak but a drawdown deep enough that most traders would have quit the strategy mid-test. The fixed-contract sizing fails in a third specific way: when the account drops 20%, it's still risking the same absolute dollar amount, making recovery much harder.
Same edge, three different outcomes. Sizing is the variable.
Common misunderstandings
- "Stops slip too often, so I trade without them." Trading without stops means accepting unbounded losses in exchange for slightly fewer slippage events. The math of that trade is awful. Use stops; use realistic slippage in your backtest so you've already priced the cost.
- "My system has a 75% win rate, so I can size up." Even a 75% win-rate strategy will have a 6-loss streak roughly every 4,000 trades — and you'll experience several 4-loss streaks per year. If your sizing can't survive an ordinary streak, you'll quit a winning strategy at the worst moment.
- "I'll use bigger size when I'm confident." Discretionary sizing inside a systematic framework is the most common way systematic traders end up with discretionary results. Pick a sizing rule, encode it in the strategy graph, and let it size every trade automatically.
Prop-firm specific notes
If you're trading a prop-firm evaluation, the daily loss limit and trailing max drawdown rules are set by the firm and are non-negotiable. The most common failure mode in evaluations is not a bad strategy — it's an aggressive strategy that briefly violates the daily limit and triggers an automatic failure even though the cumulative P&L was positive.
The prop firm traders use case walks through the specific numbers for FTMO, TopStep, Apex, and the rest, and shows how to encode the rules as kill-switches in the strategy graph.
When risk management matters most
- During winning streaks. This is the moment when discipline is hardest. The strategy is "obviously working," confidence is high, the temptation to size up is strong. Resist. Sizing rules exist precisely so you don't have to make sizing decisions in moments of high emotion.
- At account peaks. The next drawdown starts here. Your kill-switch threshold should be set as a percentage of peak equity, not starting equity — otherwise you're carrying more dollar risk after every winning month.
- When deploying multiple strategies. The strategies that look uncorrelated on paper often correlate during stress. Size each one as if the others might bleed simultaneously.
You can survive almost any losing streak that's smaller than your account. You can't survive a single losing streak that's larger than your account. Position sizing is the lever that controls which category every losing streak falls into. Get it right and your edge gets to play out. Get it wrong and your edge doesn't matter.
Pro pattern: write down your kill numbers before you trade
Before you deploy a strategy live, write down — in a file, on paper, anywhere you can read it later — the three numbers: per-trade risk %, daily loss limit, account drawdown limit. Then encode them as kill-switches in the exported executable so they fire automatically. The point is to commit to the rules in a moment of clarity, before the strategy is in drawdown and you're tempted to renegotiate.
How this fits the workflow
You design strategies in the Visual Builder, validate them in the Backtest Engine, and combine them in the Portfolio Builder. Risk management is a layer that runs through all three — the position sizing page covers the strategy-level mechanics, the portfolio builder handles the multi-strategy view, and the kill-switches in the exported executable enforce the limits in production. The Level 3 position sizing & risk lesson covers the same material at curriculum depth.