Wyckoff Position Sizing — Phase-Based Risk Rules
The Math That Decides Survival

Position Sizing & Risk Management

Edge gets you to the table. Sizing decides if you stay long enough for it to play out. A profitable system at 10% risk per trade is mathematically guaranteed to blow up. This lesson is the math that separates traders who survive from traders who burn out — regardless of edge quality.

"Risk management is the art of staying in the game. Position sizing is the science of doing it." — Van Tharp, paraphrased

The Asymmetric Recovery Math

Losses and gains are not symmetric. The deeper the drawdown, the exponentially harder the recovery. This single chart explains why "small losses" matter more than "big winners".

5%50%90%
Drawdown
Required Gain
Severity

Recovery Required Per Drawdown

Loss Gain to Recover
-10%+11%
-20%+25%
-25%+33%
-33%+50%
-50%+100%
-66%+200%
-75%+300%
-90%+900%

Below ~25% drawdown, recovery is reasonable. Above 50%, the math fights you. Your sizing should make 25%+ drawdowns mathematically improbable.

Risk-of-Ruin Calculator

INTERACTIVE

Given your win rate, R-multiple, and risk per trade — what's the probability of ruin (account hitting -50%)? Most retail traders never run this number. They should.

Expectancy per trade:
E = ( × ) − ( × 1)
E = R per trade
Risk-of-Ruin Estimate

probability of -50% drawdown over 200 trades

Recommendation:

Survival Probability

Kelly Criterion Calculator

INTERACTIVE

Mathematically optimal sizing given your edge. Full Kelly is too aggressive for real trading (variance kills you) — practical traders use Fractional Kelly (¼ or ½). The calculator shows both.

f* = (b·pq) / b

f* = optimal fraction · b = R-multiple · p = win prob · q = 1−p

Live calculation:
f* = ( × ) /
f* = %
Full Kelly
DANGEROUS

Mathematically optimal but causes 50%+ drawdowns. Variance kills psychology before edge plays out.

Half Kelly (½)
BOLD

75% of Full Kelly's growth with much less variance. Acceptable for proven systems with high data confidence.

Quarter Kelly (¼) — Recommended
PRO

~50% of Full Kelly's growth with 1/4 of the variance. What most professional traders actually use.

Expectancy Calculator (Long-Term Edge)

INTERACTIVE

Plug in your actual trade data. Expectancy = expected profit per trade. If E ≤ 0, no amount of "trying harder" will make money. Fix your edge first; size second.

E = (Win% × Avg Win) − (Loss% × Avg Loss)
Win Rate
Profit Factor
$ / trade

100-trade projection:

$

Volatility-Adjusted Sizing (ATR-Based)

Fixed % of equity is good. Fixed % adjusted for instrument volatility is better. A 1% risk on EUR/USD ≠ 1% risk on a meme coin. ATR normalizes that.

The Formula
Position size = (Account × Risk%) / (ATR × Multiple)

Where:

  • ATR = Average True Range (typically 14-period) on your trading TF
  • Multiple = how many ATRs your stop-loss sits at (typically 1-3)
Why It Matters

• Crypto bar moves 5% intraday → tiny position

• Forex bar moves 0.3% → larger position

• Same dollar risk, scaled to instrument behavior

Result: P&L variance flattens across instruments. You stop blowing up on volatile names while underutilizing capital on quiet ones.

Correlation Risk — The Hidden Multiplier

5 long positions in tech stocks during a crash = 1 position with 5× the risk. Most retail discovers correlation the hard way.

Hidden Correlations (almost 1.0)

  • S&P 500 sectors during crashes
  • All major crypto vs BTC (during volatility)
  • DXY-driven FX pairs (EUR, GBP, AUD vs USD)
  • Risk-on assets during Fed announcements

Truly Diversified

  • Equities + Gold + Bonds (different drivers)
  • Long one currency pair, short another (uncorrelated)
  • Different timeframes (intraday vs swing)
  • Different setups (Spring + UTAD on different assets)

Practical rule: Treat positions in correlated instruments as one combined position for risk-budget purposes. Long EUR/USD + short USD/JPY = single dollar-bearish trade with 2× the heat.

The 4-Level Risk Hierarchy

Risk control isn't one rule — it's a stack. Each level is a circuit breaker for the next. Triggered = stop, no exceptions.

L1
Account Level — Max Drawdown 25%
If account hits -25% from peak → stop trading completely. Take 2 weeks off. Re-evaluate the entire system. This is the existential limit.
L2
Monthly Level — -15R Halve Size
Down -15R in a month → cut position size by 50% until recovered. Mathematical risk management overrides ego.
L3
Daily Level — -3R Stop For The Day
3 full-R losses today → no more trades today. No "one more setup". Cooling off prevents revenge trades.
L4
Trade Level — Max 1% Risk + Heat Limit 3%
No single trade risks more than 1% of account. Total open-position risk (heat) capped at 3% across all positions. Correlation counts as same position.

The hierarchy is a system, not a guideline. Each level should be enforced mechanically (broker-side limits where possible). The trade you're "sure about" that breaks the limit is the trade that breaks your account.

Sizing Mistakes That Kill Accounts

Most blowups aren't from one big bet. They're from these patterns compounding.

Sizing on win-streak ego
After 5 wins, doubling size for "the obvious next winner". The streak is statistical, not skill — variance reverts. Sized-up loss erases prior wins + capital.
Adding to losers ("averaging down")
Once SL is hit, the thesis is wrong. Adding more = compounding the wrongness. This is sunk cost fallacy in spreadsheet form.
Ignoring correlation across positions
3 long crypto positions = 1 position with 3× heat. Single market event wipes them simultaneously. Correlation is invisible until it isn't.
Risk % based on margin not equity
"Risk 1% of margin" sounds safe but margin can be 10× leveraged. Always size from account equity, never from buying power.
Skipping ATR — same % across all instruments
1% on EUR/USD = small move. 1% on a low-cap altcoin = bar moves 5%, gets hit instantly. Vol-adjusted sizing is non-negotiable.

Test Your Understanding

4 questions — instant feedback, no scoring stored.