# Risk Management — The Operating System of a Trading Business
Risk management isn't a strategy. It's not a stop loss calculator. It's the operating system your trading business runs on. If the OS is solid, any reasonable strategy can work. If the OS is broken, the best strategy in the world will fail.
The Three Layers
Layer 1: Pre-Trade Risk
Decisions made before you enter a position:
- Account allocation: How much of your net worth is at risk in futures at all?
- Max daily loss: A hard stop. You lose X dollars in a day, you're done. Go outside. Come back tomorrow.
- Max consecutive losses: After Y losses in a row, you stop. Full review before resuming.
- Market condition filters: Don't trade during certain events (FOMC, NFP first 5 minutes, etc.)
Layer 2: Per-Trade Risk
The structure of each individual trade:
- Entry: Where you get in
- Stop loss: Where you get out if you're wrong
- Position size: How much you risk
- Take profit: Where you exit if you're right (or how you trail)
Layer 3: Post-Trade Risk
What happens after the trade closes:
- Journaling: Every trade recorded. Entry rationale, exit rationale, emotional state.
- Performance review: Weekly analysis. Are your trades matching your model's expectations?
- Regime detection: Has the market changed? Is your strategy still appropriate?
- Account adjustment: Recalculate position sizes based on new account equity.
Most traders only care about Layer 2. That's a mistake. Layer 1 prevents you from blowing up. Layer 3 tells you when to stop.
Stop Losses: The Good, The Bad, The Ugly
The Good
A stop loss defines your maximum acceptable loss before you enter. It's non-negotiable. Without it, a $500 risk becomes a $5,000 loss on a bad gap.
Where to place a stop:
- Below a structural level (swing low, prior resistance-turned-support)
- Below a volatility band (2 ATR from entry)
- At a maximum dollar loss (immutable, even if the level doesn't align perfectly)
The correct mindset: A stop loss is not "where I'll be proven wrong." It's "where I've decided the trade hypothesis is invalid." If your stop keeps getting hit and price reverses immediately, your hypothesis was still right — your entry or stop placement was wrong. Tighten the entry, not the stop.
The Bad
Too tight: You get stopped out on noise. A 3-point stop on ES during a news release will get hit 90% of the time before the move goes your way. Your stop needs room to breathe.
Too loose: Your risk per trade balloons. A 30-point stop on 2 ES contracts is $3,000 risk. That's too much for most accounts.
The compromise: 0.5-1.0 ATR for the stop distance. For ES with ATR(14) around 40 points, a 20-40 point stop is appropriate. That's $1,000-$2,000 per contract. If that's too much risk, trade fewer contracts — not a tighter stop.
The Ugly
Gapping stops: Your stop is at 5,400. Price gaps down to 5,380. Your stop fills at 5,380 (or worse). You lost 20 more points than planned. Gaps happen. They happen more than you think.
Stop hunting: Institutions know where retail stops cluster — below obvious swing lows. They push price through those levels, trigger the stops (adding liquidity for their larger orders), and reverse immediately. Your stop got hit. Your trade was right. Welcome to the game.
Mitigation: Don't place stops at obvious levels. Place them 1-2 ticks beyond. It won't save you from a gap, but it reduces stop hunting.
Maximum Drawdown Rules
Drawdown isn't optional. Every strategy hits a losing streak. The question is whether you survive it.
Drawdown limits by strategy type:
| Strategy Type | Expected Max Drawdown | Warning Level | Stop Level |
|—————|———————|—————|————|
| Trend following | 30-40% | 20% | 40% |
| Mean reversion | 10-15% | 10% | 20% |
| Momentum | 15-25% | 15% | 30% |
| Statistical arb | 5-10% | 8% | 15% |
Our XGBoost model targets: max drawdown under 15%. Warning at 10%. Full review at 12%. Strategy pause at 15%.
What to do in drawdown:
| Stage | Action |
|——-|——–|
| 0-5% | Normal operation |
| 5-10% | Reduce position sizes by 25%, increase signal threshold |
| 10-12% | Half position sizes, review model performance vs backtest |
| 12-15% | Full stop. This is the "did something change" zone. |
| 15%+ | Zero positions. Back to simulation until the issue is identified. |
Maximum Daily Loss
This is your most important risk rule. It's a simple hard stop: you lose X dollars in a day, you stop trading.
Why it works:
- After a loss, your judgment is impaired. You're chasing. The next trade will be worse.
- After a big loss, you'll break your own rules to "get it back." That's when the account gets blown.
- One bad day is one bad day. One bad day followed by revenge trading is a blown account.
Setting your daily loss limit:
- Prop firm evaluation: 50% of the max allowed drawdown (e.g., if max is $3,000, daily max is $1,500)
- Personal account: 3x your average trade risk. If you risk $500 per trade, your daily max is $1,500.
Correlation and Concentration Risk
If all your trades are in ES, you have no diversification. If all your strategies buy on the same signal, you have no diversification. If all your capital is in one broker, you have concentration risk.
Diversification that matters:
- Instrument: ES, NQ, RTY, CL, GC — different assets, different drivers
- Timeframe: Short-term and longer-term strategies can balance each other
- Direction: Long and short at the same time isn't diversification (it's a hedge), but a strategy that trades both directions is less exposed to directional risk
- Strategy: Momentum + mean reversion + carry — the holy grail of strategy diversification
Reality check: Most retail traders don't have the capital for genuine diversification with futures. If your account is $50k, you can't effectively trade ES, NQ, and GC while managing risk across all three. Your best hedge is your position size.
The Risk Management Checklist
Before every trade:
- [ ] Is this within my daily loss limit? (Current P&L + this trade's risk)
- [ ] Is this within my max position limit? (Contracts in this + contracts in other open positions)
- [ ] Is the market in a known high-risk event? (FOMC, NFP, OPEX week)
- [ ] Does this trade match my model's signal? (Not discretionary, not a "feeling")
- [ ] Am I revenge trading? (Did I just lose? If yes, stop.)
- [ ] Have I checked my model's recent performance? (Is it in a drawdown or winning streak?)
Prop Firm Risk
Trading against a prop firm evaluation adds a new layer:
The Goal: Pass the evaluation phase (usually a profit target with a max drawdown limit) and unlock the funded account.
The Trap: The evaluation creates perverse incentives. You need to make a certain return without exceeding a certain drawdown. This encourages small, high-probability trades — but also encourages over-trading on easy days and freezing on drawdown days.
Our approach: Same risk rules apply. The prop firm evaluation is a *constraint* on our position sizing (cannot exceed their drawdown limit), but it doesn't change our risk fundamentals. If the evaluation constraints force us outside our normal risk profile, we don't take the evaluation.
The Bottom Line
Risk management is boring. It doesn't make you money. It prevents you from losing money you already have. Every professional trader treats it as the most important part of their business. Every amateur treats it as an afterthought.
Be boring. Live to trade another day.