Why Sizing Matters as Much as Selection

The mathematical reality of compounding is brutal on oversized losing positions. A 50% loss on a position requires a 100% gain to recover. A 10% position that loses 50% costs 5 percentage points of total portfolio wealth — meaningful but survivable. A 30% position with the same 50% loss costs 15 percentage points — potentially catastrophic, triggering behavioral panic selling and preventing the portfolio from recovering through subsequent gains on correct positions.

The principle that concentrating in high-conviction ideas maximizes returns is correct within limits — but those limits are defined by the loss scenario. An investor who is right 60% of the time on individual stocks but sizes each position at 20-30% of the portfolio will likely underperform one who is right 50% of the time but sizes at 5-7%, simply because one large wrong position overwhelms a string of smaller correct ones. Position sizing is portfolio survival engineering.

→ Check AAPL's ATR and volatility profile for position sizing

Volatility-Based Sizing: The Foundation

The most robust general-purpose position sizing method calibrates size to the stock's actual volatility using Average True Range (ATR) as the stop distance. The formula:

Stop Distance = 2 × ATR (14-day)
Position Size = (Portfolio Value × Risk Per Trade %) ÷ Stop Distance

Example:
Portfolio = $100,000, Risk per trade = 1% → $1,000
Stock ATR = $4.00, Stop distance = $8.00 (2×ATR)
Position Size = $1,000 ÷ $8.00 = 125 shares

This approach automatically produces smaller positions in more volatile securities and larger positions in stable ones — keeping the dollar loss on each trade roughly constant regardless of the individual stock's price behavior. A highly volatile semiconductor stock gets a 2% position; a stable consumer staples stock gets a 6% position. Both lose approximately 1% of the portfolio if the stop is hit. Without this normalization, a volatile stock's random daily swings dominate portfolio outcomes while the stable stock barely registers.

The Kelly Criterion: Optimal Long-Run Growth

For investors who can estimate their win rate and average win/loss ratio, the Kelly Criterion provides the position size that maximizes long-run compound wealth growth. Kelly Fraction = (p × b - q) / b, where p = probability of profit, q = 1 - p, and b = average win divided by average loss. If you win 55% of trades with an average win of 1.5× the average loss: Kelly = (0.55 × 1.5 - 0.45) / 1.5 = 25% of capital per trade.

Full Kelly (25% in this example) is rarely practical — it produces severe drawdowns during losing streaks that are mathematically acceptable on an infinite time horizon but psychologically and practically destructive in real investing. Half Kelly (12.5% here) reduces median drawdowns dramatically while maintaining approximately 75% of the maximum compound growth rate. Quarter Kelly (6.25%) is conservative and appropriate when there is uncertainty about the underlying win rate estimate. Most systematic investors use some fraction of Kelly, not full Kelly.

→ Compare TSLA vs JNJ: volatility contrast for sizing intuition

Correlation Adjustment: Avoiding Double Risk

Standard position sizing methods assume positions are independent. They are not. If two positions have a correlation of 0.85, adding the second provides almost no risk diversification despite appearing to be a separate bet. The effective number of independent positions in a portfolio is calculated as N² / Σ(correlations), which can be dramatically lower than the position count if the portfolio is clustered around a common factor.

The practical implication: reduce individual position sizes when adding to an already-concentrated correlation cluster. If your portfolio already has 15% in high-growth semiconductor stocks, adding a 5% position in a correlated AI hardware name effectively increases the cluster exposure to 20% — not creates a new independent risk. Sizing the new position at 3% and capping the semiconductor/AI cluster at 20% of total equity maintains the intended risk budget.

→ Open Portfolio Attribution to identify your correlation clusters

Practical Sizing Workflow

• 1. Set your risk budget: determine what percentage of portfolio value you are willing to lose on a single position (0.5-2% is typical for diversified portfolios).
• 2. Define the stop distance: use 2× ATR below entry (volatility-based) or below a structural support level — whichever places the stop beyond normal noise.
• 3. Calculate shares: Position Size = (Portfolio × Risk %) / Stop Distance.
• 4. Apply correlation check: if the position increases an existing cluster above your concentration cap, reduce size proportionally.
• 5. Apply maximum position cap: no single name should exceed 10-15% of portfolio regardless of signal strength, to prevent single-stock catastrophe.
• 6. Reassess when volatility changes: if a position's ATR doubles post-entry, the implied stop distance has changed — consider whether the position size still reflects your actual risk tolerance.

Common Pitfalls

• Equal-dollar sizing without volatility adjustment: allocating the same dollar amount to a 15% annualized volatility stock and a 45% annualized volatility stock creates wildly unequal risk contributions.
• Sizing from conviction alone: high conviction without a stop and a risk budget is the recipe for catastrophic loss — especially in single-stock positions where company-specific risk (fraud, competitive disruption, earnings collapse) can cause 80%+ losses regardless of macro environment.
• Ignoring correlation clusters: five semiconductor positions each at 4% of portfolio look diversified by count but represent 20% cluster exposure to the same set of macro and sector risks.
• Moving stops after entry without resizing: widening a stop to avoid a loss converts a 1% risk position into a 3% risk position — either accept the new risk explicitly or re-enter at the original size with the new wider stop.

Apply Position Sizing In AlgoVestIQ