Position sizing is the decision that determines whether you survive long enough to profit. It's not your entry indicator, your chart pattern, or your market thesis — it's how many contracts you put on relative to your account. Get it wrong and even a 60% win rate strategy goes to zero. Here's the math, the frameworks, and the practical application for crypto derivatives.
Position sizing is the process of determining how many contracts or how much notional exposure to take on a single trade, given your account size, risk tolerance, and the trade's specific parameters.
It answers one question: How much?
Most losing traders spend 90% of their time on *what* to trade and *when* to enter. The professional edge lives in *how much*. As Vince's mathematics of money management demonstrates, the quantity decision is inseparable from your equity level — and getting it even slightly wrong can halve your long-run wealth.
The core insight from decades of position sizing research is this: there exists an optimal fraction of your equity to risk per trade. Risk too little and you underperform. Risk too much and you eventually go broke — even with a positive-expectancy system. The relationship isn't linear; it's a curve with a sharp peak and a steep drop-off on the overbetting side.
Every sound position sizing method starts with the same input: your maximum permissible loss per trade, defined by your stop-loss level. This is the foundation — as the Money Management Strategies framework puts it, the stop-loss price should be the level that "unequivocally confirms" the trade thesis is invalid.
Here's the formula:
Position Size (contracts) = Risk Amount / Risk Per Contract
Where:
Example: You have a $50,000 account. You're willing to risk 2% per trade. BTC is at $65,000, and your stop-loss is at $63,500 (a $1,500 distance). On a standard linear perpetual where 1 contract = 1 BTC:
That's it. You don't start with "I want to go 10× long" and work backward. You start with your stop distance and your risk budget and let the math tell you the size.
Vince's optimal f framework takes fixed fractional sizing to its mathematical conclusion. The idea: for any trading system with a known distribution of outcomes, there exists an optimal fraction `f` of equity to risk that maximizes the geometric growth rate of your account.
The key concepts:
The critical lesson from Vince's research: The TWR curve is asymmetric. Overbetting destroys wealth faster than underbetting sacrifices gains. In his canonical 50/50 coin flip example (win +$2, lose −$1):
| f (fraction risked) | TWR after 40 bets |
|---|---|
| 0.10 | 4.66× |
| 0.25 (optimal) | 10.55× |
| 0.40 | 4.66× |
| 0.50 | 1.00× (breakeven) |
| >0.50 | Ruin |
Risking 25% is optimal. Risking 40% gives the *same* result as risking 10%. Risking 50% — which feels "aggressive but not crazy" — yields zero growth. Anything above 50% guarantees ruin over time.
This is the most important chart in trading. Most retail crypto traders, using 20×–50× leverage on their full account, are operating deep in the right side of this curve — the ruin zone.
The risk of ruin formula gives you a sobering reality check. In the simplified model:
R ≈ (q/p)^k
Where:
With p = 0.55 (a solid edge), q = 0.45, and k = 10 (risking 10% of capital per trade):
R = (0.45/0.55)^10 ≈ 13.4% chance of ruin
That's a 1-in-7 chance of going broke, even with a 55% win rate. Now increase k to 20 (risking 5% per trade):
R = (0.45/0.55)^20 ≈ 1.8% chance of ruin
Same edge, half the risk per trade, and your ruin probability drops by 7×. This is why the professionals obsess over sizing, not indicators.
Crypto derivatives add three complications that amplify sizing errors:
Leverage is not position size. A 10× leveraged position on $5,000 margin gives you $50,000 notional exposure. But your *risk* depends on your stop distance, not your leverage. A 10× long with a 1% stop risks $500. A 3× long with a 5% stop risks $750. The 3× position is actually riskier in dollar terms despite lower leverage. Always size from stop distance, never from leverage.
Funding costs erode margin. A position held for multiple funding intervals accumulates carry costs. If funding is +0.05%/8h and you're long, you're losing ~0.15%/day of your notional. On a 10× leveraged position, that's 1.5%/day of your margin. Factor funding into your risk calculation for any position you plan to hold more than a few hours.
Liquidation ≠ stop-loss. Your stop-loss should be placed *above* your liquidation price with a meaningful buffer. If your liquidation price is $60,000, a stop at $60,500 gives you essentially zero margin for error — slippage during a cascade can blow past both levels. Professional sizing accounts for liquidation distance as a hard floor, with the stop-loss set well above it.
The full process, adapted from professional futures practice:
1. Rank opportunities by expected value, not excitement. Deploy capital to the highest-ranked first.
2. Set total exposure limit. Keep 50–80% of equity as margin buffer.
3. Allocate risk per trade. Standard: 1–3% of equity per trade. Never exceed 5%.
4. Define the stop-loss at the price that invalidates the thesis — not a round number.
5. Calculate contracts: Risk Amount / Risk Per Contract = Position Size.
Survival is the prerequisite. No edge compounds if you blow up first. The mathematics are unambiguous: overbetting destroys even positive-expectancy systems. In crypto, where 30% daily moves happen and cascading liquidations can gap through stop-losses, conservative sizing isn't cautious — it's the minimum viable strategy.
Sizing is your actual edge. Two traders with identical entry indicators but different sizing rules will have radically different equity curves. The one risking 2% per trade with optimal f awareness will compound steadily. The one risking 15% per trade will have spectacular wins followed by catastrophic drawdowns.
It removes emotion from the equation. When your size is determined by a formula — not by how confident you "feel" — you eliminate the biggest source of trading error. The formula doesn't care that you're "really sure this time." It sizes based on math.
Using leverage as a sizing tool. "I'll just use 20× this time" is not a sizing method. It's a recipe for ruin. Leverage determines your margin efficiency, not your risk. Size from stop distance.
Ignoring correlation. Five open BTC-correlated trades at 2% risk each is not 2% total risk — it's closer to 10% if they all move against you simultaneously. In crypto, correlation between assets during sell-offs approaches 1.0. Treat correlated positions as a single risk unit.
Adjusting size after losses to "make it back." Increasing size after a drawdown to recover faster is textbook gambler's ruin. The math is clear: your optimal fraction should *decrease* as equity drops, not increase. Let the formula do its job.
For most traders, 1–2% of total account equity per trade is the professional standard. Aggressive but survivable strategies might go to 3%. Anything above 5% per trade puts you on the wrong side of the optimal f curve — where overbetting degrades long-run returns regardless of win rate.
Leverage affects your margin requirement, not your risk. A 10× leveraged 1 BTC position and a 1× 1 BTC position have the same dollar exposure to BTC price moves. The difference is how much capital you post as collateral. Size based on the dollar risk (entry to stop-loss × position size), not the leverage multiplier.
Yes. When volatility expands, your stop-loss distance should widen (to avoid being stopped out by noise), which mechanically reduces your position size for the same risk budget. This is the correct, math-driven response: same dollar risk, wider stop, smaller position. Don't fight volatility with tighter stops — let the sizing formula adapt naturally.
---
*This article is part of The Codex — PARAGON's structured learning library.*