Applies to: @RISK 6.x and Up

 

Correlation Matrix in @RISK Fit Summary

• The correlation matrix shown in a fit summary is the Spearman (rank-order) correlation of the transformed historical data.

  • In @RISK 6.0, Pearson correlations were used; changed to Spearman in 6.1 for better results with differently shaped distributions.

• Transformed data are not visible as numbers, but you can graph them by doing a single fit with the same transformations.

How It Works:

1. Historical data are transformed (user-selected or Auto Detect).

2. @RISK computes Spearman correlations of the transformed data (displayed in the fit summary).

3. For projections, @RISK:

   • Generates "raw" projected numbers based on the transformed data.

   • Applies the correlation matrix to these raw numbers.

   • De-transforms the numbers to produce the final outputs visible in worksheet cells.

Key Point:
The correlation matrix affects the raw projections, not the de-transformed numbers in the worksheet.