How to calculate correlation
r_xy = Cov(X, Y) ÷ (σ_X × σ_Y)
Returns: rₜ = (Pₜ − Pₜ₋₁) ÷ Pₜ₋₁
Correlation is calculated on returns, not prices — two unrelated assets that both trend upward can show a misleadingly high price correlation.
Worked example
- Paste 40 daily closes for each asset; the calculator converts them to 39 daily returns.
- Returns of A and B have covariance 1.2 and standard deviations 1.5% and 1.0%.
- r = 1.2 ÷ (1.5 × 1.0) = 0.80 — strongly positive.
How to interpret correlation
- +0.7 to +1: move together — little diversification benefit.
- −0.3 to +0.3: largely independent — good diversifiers.
- Below −0.5: tend to offset each other — natural hedges.
Correlations rise in market crises exactly when diversification is needed most. Check how much diversification your portfolio actually achieves with the diversification ratio calculator.
Limitations and common mistakes
- Make sure each series covers the same dates; misaligned data understates correlation.
- Correlation measures linear co-movement only, and it is unstable over time.