Overview

@RISK allows you to use correlation coefficients from historical data to simulate distributions that reflect past relationships among variables.

Example:

• Inventory of homes ↑ → housing starts ↓

• Mortgage rates ↑ → mortgages sold ↓

Historical relationships are captured in an Excel correlation matrix, which can then be imported into @RISK to control correlated sampling.

Step-by-Step: Identify Correlations in Excel

1. Arrange your data in adjacent columns or rows.

2. Open the Data Analysis tool:

   • Excel 2010/2007: Data → Data Analysis

   • Excel 2003 or earlier: Tools → Data Analysis

   • If not available, enable Analysis ToolPak.

3. Select Correlation and click OK.

4. Specify Input Range and indicate whether data are in rows or columns.

5. Choose Output options: Output Range, New Worksheet, or New Workbook.

6. Click OK → Excel generates a correlation matrix with coefficients.

Step-by-Step: Import into @RISK

1. Create @RISK distributions from each data column/row.

2. Ensure distributions are ordered the same as in the Excel matrix.

3. In the @RISK Model Window, select the distributions to be correlated.

4. Right-click → Correlate Distributions.

5. Verify order matches the Excel correlation matrix.

6. Select all correlation coefficients in Excel (exclude labels), then Edit → Paste in @RISK.

7. Rename the correlation matrix if desired and click Apply.

8. @RISK updates:

   • Correlation icons appear beside correlated Inputs.

   • RiskCorrmat function is now linked to each correlated distribution.

   • A new worksheet with the @RISK correlation matrix is added in Excel.

Key Notes:

• The order of variables in @RISK must match the order in the Excel correlation matrix.

• This method ensures your simulations reflect real-world historical relationships between variables.