Correlation-aware allocation goes beyond simple diversification by mathematically measuring how assets move together and sizing positions accordingly. Two highly correlated positions (e.g., ES and NQ) effectively concentrate risk. By adjusting allocation based on measured correlations, traders build portfolios with better risk-adjusted returns.
Pearson correlation coefficient (-1 to +1). +1 = perfectly correlated (same direction). -1 = perfectly negatively correlated. 0 = uncorrelated. Rolling correlation: measure over recent periods (30, 60, 90 days). Regime changes: correlations shift during different market environments. Cross-asset correlations: BTC/stocks, gold/real rates, USD/commodities.
Reduce combined position size when trading two correlated instruments. If BTC and ETH have 0.9 correlation, treat them as ~1.8 positions, not 2.0. Include negatively correlated assets as hedges. Monitor rolling correlations to detect regime shifts.
If two positions' correlation increases above 0.7, reduce combined exposure. Hedge with negatively correlated assets during uncertain periods. Reassess allocation when regime changes are detected (e.g., BTC decoupling from stocks).
Weekly review of rolling correlations. Portfolio-level application.
During crises, the thing that usually saves portfolios is cash and genuinely uncorrelated assets. True decorrelation is rare — during 2020 and 2022, almost everything correlated briefly. The only true hedge is position sizing.
Fixed-fractional position sizing risks a fixed percentage of your account on every trade (commonly 1-2%). This ensures that a string of losses reduces position sizes proportionally, protecting capital during drawdowns. It's the most widely recommended position sizing method for discretionary traders because it's simple, sustainable, and mathematically sound.
The Sharpe ratio measures risk-adjusted return by dividing excess return (above the risk-free rate) by the standard deviation of returns. The Sortino ratio improves on Sharpe by penalising only downside deviation, recognising that upside volatility is desirable. Both ratios help traders compare strategies on a level playing field and identify which approaches deliver the best returns for their level of risk.
Volatility-based position sizing adjusts trade size according to an asset's current volatility, ensuring that each position carries a consistent dollar risk regardless of how volatile the instrument is. By dividing fixed risk by the asset's average true range or standard deviation, traders normalise risk across different assets and market conditions, preventing outsized losses during volatile periods.
The risk-reward ratio (R:R or RRR) compares the potential loss (distance to stop loss) to the potential gain (distance to target) for each trade. A 1:2 R:R means you risk $1 to potentially make $2. By maintaining favourable risk-reward ratios, a trader can be profitable even with a win rate below 50%.