Correlation Matrix
Calculate pairwise correlation between assets
First row is headers. Each subsequent row is a data point. Use commas to separate values.
What is this tool?
A correlation matrix shows the pairwise correlation coefficients between multiple assets or data series. Values range from -1 (perfect negative correlation) through 0 (no correlation) to +1 (perfect positive correlation). Diversification works best when portfolio assets have low or negative correlations. When all assets move together (high positive correlation), diversification benefits vanish. This tool computes Pearson correlation from historical price or return data and visualizes it as a heat map.
How to use
- 1
Enter asset data
Input price or return data for each asset (one per row, aligned by date).
- 2
Select method
Choose Pearson, Spearman, or Kendall correlation method.
- 3
View matrix
See the correlation matrix with color-coded heat map for easy interpretation.
Frequently Asked Questions
Why is correlation important for portfolio management?
Correlation determines diversification effectiveness. A portfolio of stocks with 0.9 correlation behaves almost like a single stock during market events. True diversification requires assets with correlations near 0 or negative. Gold, bonds, and cash often have low or negative correlation with equities, making them effective hedges.
What are the different correlation methods?
Pearson measures linear relationships and is most common. Spearman ranks values first, detecting monotonic but non-linear relationships. Kendall is more robust for small samples and ordinal data. For financial data, Pearson is standard, but Spearman is useful when relationships are non-linear (e.g., volatility clustering).
Does correlation imply causation?
No — this is the most common statistical mistake. Two variables can be highly correlated without any causal relationship. Stock prices and ice cream sales are correlated (both rise in summer), but one does not cause the other. In finance, use correlation for portfolio construction, not as proof of causation.