Position sizing & risk management for leveraged crypto

Most traders spend their mental energy choosing entries. They research catalysts, read order-flow, dial in indicators. Then they open a trade by feel — "about 20% of the account" — and wonder why a solid thesis still bleeds out. The entry is not the hard part. Position sizing is.

Sizing decides whether a string of normal losing trades is recoverable or account-ending. A system with a 45% win rate and positive expectancy can go 10 losses in a row without being unusual — the question is only whether you survive those 10 trades. At 2% risk per trade, a 10-loss streak costs you about 18% of equity. At 10% risk per trade, the same streak takes 65%. One version keeps you in the game. The other may not.

Fixed-fractional risk: the baseline

The cleanest sizing rule for retail traders is fixed-fractional risk: risk a fixed percentage of current equity on every trade. The percentage is usually between 0.5% and 2%. Aggressive but defensible approaches go to 3%; beyond that, the math starts to work against you in drawdown scenarios.

Why a percentage of current equity rather than a fixed dollar amount? Because after losses your equity shrinks, so the raw dollar risked also shrinks — the system auto-scales down when you are running cold and scales back up when you recover. It is a natural form of drawdown protection without any manual intervention.

Sizing off the stop

The formula is:

qty = (equity x risk_fraction) / |entry_price - stop_price|

Work through a concrete example. You have $50,000 in equity. You are willing to risk 1% per trade ($500). You want to buy BTC at $68,400 with a stop at $66,900 — a $1,500 stop distance.

qty = 500 / 1500 = 0.333 BTC

At $68,400 per coin, that position is worth about $22,800 — roughly 45% of your account in notional exposure. Whether you use 1x, 3x, or 5x leverage to carry that notional exposure is a separate question. Your maximum loss if stopped out is still $500, or 1% of equity, regardless of which leverage multiple you choose.

Leverage does not change the risk

This is the point most traders get backwards.

At 1x, holding $22,800 of BTC requires $22,800 of initial margin. At 10x, the same $22,800 notional requires $2,280 of margin. Your dollar loss if the stop-loss is hit is still $500 in both cases. Where leverage matters for risk is in a different, more dangerous way: it shrinks the distance between your entry and your liquidation price. At 10x, a 10% adverse move liquidates you. If your stop is only 2.2% away, that margin is fine. If you move the stop or remove it entirely, 10x leverage becomes genuinely dangerous.

The rule is simple: always know your liquidation price before you open a leveraged position, and make sure your stop sits well above (for longs) or below (for shorts) that level.

The Kelly criterion and why to use less of it

The Kelly criterion gives the theoretically optimal fraction of equity to bet on each trade to maximize long-run geometric growth:

f = (win_rate x avg_win - loss_rate x avg_loss) / avg_win

For a system where you win 50% of trades and your average win is 2x your average loss, Kelly suggests betting about 25% of equity. That number sounds aggressive because it is. Full Kelly maximizes expected log-wealth, but the path there involves drawdowns that are psychologically and practically brutal. A single string of losses at full Kelly can cut your equity in half.

The standard practice is to bet a fraction of Kelly — typically half-Kelly or quarter-Kelly. Half-Kelly on the example above is 12.5%, which is still well above the 1-2% range most systematic traders use. Fractional Kelly is not timid; it is the approach that actually gets you to the long-run outcome Kelly promises, because you stay in the game long enough.

Portfolio heat and correlation

Running multiple positions simultaneously creates a compound risk question. If you have five positions open at 1% risk each, your total "portfolio heat" is 5%. That is manageable. If all five positions are highly correlated — say, five long positions on large-cap altcoins during a BTC-led selloff — they will all hit their stops in the same hour, and you realize the full 5% loss at once.

The fix is to treat correlated positions as part of the same exposure. A rough rule: if two assets have a 30-day correlation above 0.7, allocate their combined risk as a single unit. So two correlated longs at 0.5% each count as one 1% risk unit, not two independent bets. This keeps your true worst-case scenario legible.

Open interest can be a useful read on crowding. When open interest is elevated on a directional move, a reversal tends to be violent because many participants are adding risk in one direction — exactly when your correlated long book is most exposed.

Funding eats into your real risk math

In perpetual futures, the effective entry cost is not just the fill price. The funding rate is a recurring drag (or tailwind) on every hour of holding time. If you are long in a market where longs pay 0.03% every 8 hours, you are paying roughly 32.8% annualized carry just to hold the position.

This means your sizing formula needs a holding-time adjustment. If you plan to hold for a week and funding is running hot, the real expected loss is the stop-loss dollar amount plus the accrued funding over seven days. Funding can meaningfully shift a positive-expectancy setup into a marginal or negative one if you are on the wrong side of it for long enough.

Use isolated margin when you want hard limits on how much of your account a single position can consume. Use cross margin only when you understand that the exchange can draw from your full balance to keep a position alive — that changes the liquidation math entirely.

Putting it together

A practical pre-trade checklist:

1. Define the stop level before the entry. The stop is not adjustable after you are in.
2. Apply the formula: qty = (equity x risk%) / |entry - stop|.
3. Verify the liquidation price for the leverage multiple you plan to use. Confirm the stop is at least 20-30% above (long) or below (short) liquidation.
4. Count your portfolio heat. Add this trade's risk% to all open positions. Stay under a comfortable total — 5% is a reasonable cap for most traders.
5. Check funding direction and rate. If you are paying high carry, reduce hold time or size down.

None of these steps take more than two minutes. Together, they decide whether a correct thesis actually produces a correct trade outcome.