Open this calculator on its own page
🔗 Correlation-Adjusted Risk Calculator
Five positions at 1% each can only ever lose 5%. Correlation does not raise that ceiling; it decides how often you hit it. This works out how often.
• Same sector stocks (e.g., 5 tech): 0.70-0.90
• Different sectors, same direction: 0.40-0.60
• Longs + Shorts (hedged): 0.00-0.30
• Different asset classes: -0.20-0.20
💡 What This Means
Effective Risk = Nominal Risk × √(N × Avg Correlation)
If your positions move together, they'll HIT STOPS together. Correlation amplifies risk.
Why Correlation Destroys "Diversification"
You have 5 positions. You risk 1% on each. Total risk: 5%, right?
WRONG.
If those 5 positions are correlated (same sector, same trend direction), they move TOGETHER. When one dumps, they ALL dump.
🚨 The Formula
Effective Risk = Nominal Risk × √(N × Average Correlation)
Where:
- N = Number of positions
- Nominal Risk = Sum of individual position risks
- Average Correlation = Average correlation between positions
Example: 5 Tech Stocks
Your positions: AAPL, MSFT, GOOGL, NVDA, AMD
Risk per position: 1%
Nominal risk: 5% (1% × 5)
Average correlation: 0.80 (tech stocks move together)
Effective Risk = 5% × √(5 × 0.80) = 5% × √4 = 5% × 2.0 = 10%
You THINK you're risking 5%.
You're ACTUALLY risking 10%.
One sector rotation, one Fed speech, one macro event. And ALL your stops get hit in the same hour.
💎 The Brutal Truth
Diversification is a myth if your positions are correlated.
5 tech stocks = 1 bet on tech sector
3 crypto positions = 1 bet on crypto
4 energy longs = 1 bet on oil prices
Diversification is a claim about correlation, and it is measurable: run the numbers rather than counting tickers.
Worked example: four rules, one book
This is the book Lesson 72 prices, so every figure below is reproducible from that page and from the calculator above.
Four positions. Two per cent risked on each. A pairwise correlation of 0.9472, which is what Lesson 71 measured between two rules running on the same instrument.
What the three numbers are
| Nominal heat | 4 × 2% = 8% | The most the book can lose. Not a standard deviation. |
| Daily standard deviation | 7.84% | 98% of the maximum, because the four barely differ. |
| All four against you | 40.53% | One day in 2.5. Independent, it would be one day in 16. |
What correlation actually does
It does not make the 8% bigger. Nothing can: four stops at two per cent each lose eight per cent when they all hit, and that is the end of the arithmetic. What correlation changes is how often eight per cent is the outcome.
At a correlation of zero, all four going against you on the same day is one day in 16. At 0.9472 it is one day in 2.5. The worst case did not move. It just stopped being rare.
Why that matters more than a bigger number would
- A maximum you meet twice a year is a tail. A maximum you meet twice a week is your ordinary variance, and you have to be able to sit through it.
- Four correlated rules are carrying about 1.06 independent bets, which is what Lesson 74 measures. You are not running four positions. You are running one, four times.
- Any figure that reports a loss larger than every stop being hit at once is describing something that cannot happen, and should be treated as a mis-specified model rather than a warning.
Educational only. Trading involves substantial risk of loss. Not financial advice. Past performance does not guarantee future results.
How to Avoid Correlation Disasters
Solution 1: Diversify Across Uncorrelated Assets
Bad: 5 tech stocks (correlation ~0.80)
Better: 2 tech, 1 energy, 1 financial, 1 healthcare (correlation ~0.40)
Best: Stocks + crypto + forex + commodities (correlation ~0.10)
Solution 2: Reduce Position Size When Correlated
If you MUST trade correlated positions (e.g., day trading tech during earnings season), reduce risk per trade.
Normal risk: 1% per trade, 5 positions = 5% heat
High correlation (0.80): 0.5% per trade, 5 positions = effective 5% heat
Solution 3: Hedge with Inverse Positions
If you're long 3 tech stocks, short 1 tech ETF (QQQ) or buy tech puts as insurance.
Reduces correlation from 0.85 to 0.40-0.50
Cost: Premium/spread. Benefit: Sleep at night.
Solution 4: Calculate Correlation Weekly
Use this calculator EVERY Sunday before the trading week:
- List all open positions
- Estimate average correlation (use the guide)
- Calculate effective risk
- If > 8%, close or hedge positions
Correlation by Asset Class
| Portfolio Type | Avg Correlation | Verdict |
|---|---|---|
| 5 tech stocks (same sector) | 0.70-0.90 | 🚨 Dangerous |
| 3 stocks different sectors | 0.40-0.60 | ⚠️ Moderate |
| Stocks + bonds + commodities | 0.10-0.30 | ✅ Good |
| Long stocks + short stocks | -0.20-0.20 | ✅ Hedged |
✅ Action Plan
- Open your broker platform right now
- List all your current positions
- Estimate average correlation (use table above)
- Run this calculator
- If effective risk > 8%, close the most correlated positions
Do this every week. A book whose positions all move together is one position, however many tickers it holds.