These terms appear in the charts and statistics on this page:

Page Layout

The page has these main sections from top to bottom:

• 1. Header

  Shows dataset name, model type, average correlation value, and daily change. Help (?) and Video icons are here.

• 2. Correlation Matrix

  Interactive grid of colored cells. Click any cell to add its time series to the charts below.

• 3. Time Series Charts

  Shows correlation evolution over time. The date picker chart shows the average. Additional lines appear when you click matrix cells.

• 4. Summary Table

  Statistics for the most recent date: Lower Correlation, Upper Correlation, 1st Factor, 2nd Factor.

• 5. Related Models Panel (Right Side)

  Lists other correlation models available for this dataset (e.g., DCC, DECO, DCC-NL). Click a link to view the same data analyzed with a different model.

• 6. Estimation Info Panel (Right Side)

  Shows the date range used for model estimation and model parameters. Click to expand and view detailed estimation information.

Using the Correlation Matrix

The matrix is the main interactive element on this page:

• Click Off-Diagonal Cells

  Click any cell (not on the diagonal) to add that asset pair's correlation time series to the chart. Click again to remove it.

• Click Diagonal Cells

  Click a diagonal cell (where an asset meets itself) to add ALL correlation pairs involving that asset to the chart.

• Read the Colors

  Darker red means higher correlation. Lighter colors mean lower correlation. Hover over any cell to see the asset pair names displayed above the matrix.

• Clear All Series

  Click 'CLEAR ALL' above the matrix to remove all selected pairs from the chart and start fresh.

• Large Datasets (20+ Assets)

  For datasets with more than 20 assets, the matrix displays the 20 most correlated assets by default. Use the dropdown to switch to the 20 least correlated assets, or select a custom set of assets. You can also save custom datasets for later use.

• Finding Datasets by Ticker

  To find correlation data for a specific asset, select 'Correlation' and your preferred model, then search by ticker (e.g., SPX:IND). Note that search works by ticker only, not by dataset name. A ticker may appear in multiple datasets if the asset belongs to more than one group.

Summary Table Columns

The table below the charts shows statistics for the most recent date:

• Lower Correlation

  The 5th percentile of all pairwise correlations. This is the lower bound. Only 5% of pairs have correlations below this value.

• Upper Correlation

  The 95th percentile of all pairwise correlations. This is the upper bound. Only 5% of pairs have correlations above this value.

• 1st Factor

  How much of total variance is explained by the first principal component (usually the 'market factor'). Higher means assets move more as a group.

• 2nd Factor

  How much is explained by the second principal component. Added to 1st Factor, this shows cumulative explained variance.

Using the Time Series Charts

The charts show how correlations changed over time:

• View Dropdown

  Switch between 'Median Correlation' (with 5th/95th percentile bands) and 'Principal Factors' (1st and 2nd factor percentages) using the dropdown.

• Date Range

  Drag the handles at the bottom to zoom into specific time periods. The chart above updates to show only that range. Note: the maximum selectable range is 10 years.

• Preset Buttons

  Use 6M, 1Y, 2Y, 5Y, 10Y, or All buttons to quickly select common time ranges.

The charts and statistics help you understand how assets in this group relate to each other and how those relationships change over time.

Reading the Time Series

Key patterns to look for in the charts:

• Correlation Spikes

  Sharp upward moves in the correlation lines indicate stress periods (2008, 2020) when all assets moved together.

• Median vs Quantile Gap

  When the 5th and 95th percentile lines are far apart, correlations vary widely across pairs. When they're close, all pairs behave similarly.

• Factor Dominance

  If the 1st Factor line is consistently high (above 70-80%), the market factor dominates and diversification within this group is limited.

• Individual Pair Lines

  When you click matrix cells, each pair appears as a separate colored line. Pairs with lines consistently below the group offer better diversification potential. Pairs consistently above the group move together more than typical.

• Correlation Trends

  A gradual upward trend over months or years suggests assets are becoming more interconnected, often due to globalization or sector convergence. A downward trend suggests increasing independence between assets.

• Comparing Multiple Pairs

  Select several pairs to compare side-by-side. During stress periods (2008, 2020), most correlations spike upward. Pairs whose correlations increase less during crises maintain diversification benefits when you need them most.

The correlations shown on this page are updated daily using dynamic models that allow relationships to change over time based on recent market behavior.

Daily Updates

Each day, the model is re-estimated with the latest data. After volatile days, correlations typically increase. After calm days, they tend to decrease.

Common Questions

How do I select correlation pairs to compare?

Click any off-diagonal cell in the matrix to add that pair's time series to the chart. Click a diagonal cell to add all pairs involving that asset. Click 'CLEAR ALL' to remove all selections.

How do I download the data?

First select the pairs you want by clicking matrix cells. Then click the 'DOWNLOAD' button above the time series chart. You must be logged in with a V-Lab account.

How do I switch between Median and Factor views?

Use the dropdown selector above the time series charts. 'Median Correlation' shows the median line with 5th and 95th percentile bands. 'Principal Factors' shows factor exposure percentages.

Understanding the Charts

Common questions about interpreting what you see:

Why do correlations spike suddenly during certain periods?

During market stress (2008, 2020, etc.), fear and forced selling affect all assets simultaneously. Assets that normally move independently start moving together. This is visible as sharp upward spikes in the correlation lines.

What if the matrix shows dark red (high correlations) everywhere?

High correlations throughout mean the assets in this group move together most of the time. Diversification within this group is limited since assets tend to rise and fall together.

What does it mean when 1st Factor is above 80%?

When the first principal component explains 80%+ of variance, a single 'market factor' dominates all assets. They essentially move as a group. Diversification within this asset set is limited.

What are the Related Models links on the right?

V-Lab offers different correlation models for each dataset. DCC (Dynamic Conditional Correlation) is the standard model that tracks how correlations change based on recent market movements. DCC-NL (Nonlinear) is designed for datasets with many assets, where it provides more stable estimates. DECO (Dynamic Equicorrelation) assumes all assets in the group share the same average correlation level, useful for seeing overall market trends. Click the links to see the same dataset analyzed with a different model and compare the results.

Here's how to quickly analyze correlations using this page:

Finding Historical Patterns

Scroll through the time series to find crisis periods. Look for correlation spikes around 2008 (financial crisis), 2011 (European debt), 2020 (COVID). Notice how correlations spiked and then gradually returned toward normal levels.

Applying to Portfolio Decisions

When evaluating a potential investment, compare its correlation with your existing holdings during both calm and stressed periods. An asset with low correlations during normal times but high correlations during crises (2008, 2020) provides less protection when you need it most. Focus on pairs whose correlations remain relatively low even during stress periods.