Position sizing for crypto traders: the 1% rule revisited

Ask ten profitable traders for the single most important variable in their P&L and most of them will say the same thing: size. Get the entry wrong and you lose a planned amount. Get the size wrong and you can lose the account in a single bad trade. The 1% rule — never risk more than 1% of equity on any single position — is the default answer, but it was written for equities. Crypto is different.

What the 1% rule actually says

The rule is about risk per trade, not position size. You're not saying "my position is 1% of my account." You're saying "if my stop is hit, I lose 1% of my account." The position size is a function of where your stop is.

risk_per_trade = account_equity * risk_pct
stop_distance  = entry_price - stop_price          # for a long
position_size  = risk_per_trade / stop_distance    # in units of asset

# Example: $10,000 account, 1% risk, BTC long at $40,000 with stop $39,000
# risk_per_trade = $100
# stop_distance  = $1,000
# position_size  = 0.1 BTC ($4,000 notional, 0.4x leverage)

Notice what this protects against: a single losing trade costs at most 1% of equity. Notice what it doesn't protect against: a correlated string of losses. Five 1% losses in a row is a 5% drawdown. Twenty in a row is 18%. A run of bad trades is mathematically guaranteed at some point, and the only defence is the next layer of risk management — daily loss caps, max concurrent positions, and reduced size after drawdowns.

ATR-based sizing

Fixed-percentage stops (e.g. "stop 2% below entry") are easy to read but lousy in practice — they're indifferent to volatility. ATR-based stops scale with how volatile the market currently is. The stop distance becomes K × ATR(14), where K is typically 1.5 to 3. In a quiet market your stop is closer, your size larger; in a volatile market your stop is wider, your size smaller. The risk in dollars stays constant.

# Same formula, but stop_distance = K * ATR
K = 2.0
stop_distance = K * ATR(14)
position_size = risk_per_trade / stop_distance

# Quiet BTC: ATR = $400 -> stop_distance = $800 -> position = 0.125 BTC
# Volatile BTC: ATR = $1,500 -> stop_distance = $3,000 -> position = 0.033 BTC

This is the default in every professional crypto desk's sizing engine. The idea is older than crypto — Richard Dennis's Turtles used it in the 80s — but it survives because it's a clean answer to the volatility-clustering problem. Markets that just had a big move are likely to have more big moves; size down.

Why crypto needs tighter sizing than equities

Three reasons. First, crypto trades 24/7 — no overnight gap risk per se, but also no daily circuit breakers to slow a panic. Second, leverage is widely available and cheap, so retail size grows in good times and gets liquidated in bad ones. Third, single-asset moves of 20% in a day still happen on majors, and 50-90% in a day still happens on smaller caps. A 1% risk-per-trade rule that uses a 2% stop puts you at 0.5x notional leverage — but that 2% stop is one normal candle on a typical alt.

Practical adjustment: cap notional exposure at 1.0x account equity for majors and 0.25x for non-majors, even when your stop-based sizing would allow more. Liquidation cascades exist; you don't want to be in a position that becomes uncloseable.

Kelly, and why not to use full Kelly

The Kelly criterion gives you the size that maximises geometric growth given a known edge. For a binary bet with win probability p, payout odds b, the fraction is f = (bp - (1-p)) / b. For a strategy with 55% win rate and a 1.5:1 reward-risk ratio, full Kelly is about 25%. That's catastrophically large. Why? Because you don't actually know your win rate or payout — you have a noisy estimate, and full-Kelly sizing on an over-estimated edge produces ruin with high probability.

The standard adjustment is fractional Kelly — typically 0.25× to 0.5× full Kelly. For the strategy above, that's 6-12% per trade, still higher than the 1% rule. For a strategy with measurement-error baked in, 1-2% is honest. Most professional shops live in the 0.5-2% range; if a sizing rule tells you to bet more, your edge estimate is probably wrong.

The daily loss cap

The single highest-impact rule beyond per-trade sizing is the daily loss cap. If the account is down 3% on the day, no more trades today. This converts a tail-risk single-day blowup into a bounded event and forces you to come back the next day with a clear head. Set it in code, not in willpower.

risk:
  size:
    method: atr
    risk_pct: 0.75            # 0.75% of equity per trade
    atr_period: 14
    atr_mult: 2.0
  exposure:
    max_notional_majors: 1.0  # 1x account for BTC/ETH/SOL
    max_notional_alts:    0.25
  daily_cap:
    loss_pct: 3.0
    action: halt_until_next_day

Compounding the percentage

A subtle point: the 1% rule with risk computed off current equity (not initial equity) compounds your size down on drawdowns and up on winners. That's actually what you want — you're shrinking your bet exactly when your edge estimate is being challenged. The opposite (anti-Martingale) is what blows up traders.

Next steps

The platform's strategy designer has ATR-based sizing and daily-cap primitives built in — you compose them in the same YAML rule set as your entry / exit logic, and they apply identically in backtest and live. Start with the getting-started guide or jump straight to sign-up.