Create a Carve Out

A carve out is a subset of a portfolio’s assets including a cash portion. As you build a new portfolio that combines a cash return and a carved out node’s return from a particular portfolio, you can select the cash node from a different source entity than the one used to select the carved out node. You can create a carve out using a Blended or Floating type of custom benchmark.

For example, you can carve out the Equity portion of a balanced portfolio that includes the components, Equity, Fixed Income, and Cash. The carve out can use the Equity component of the balanced portfolio plus a cash return that represents the Cash node of the balanced portfolio or cash from another source. You can also use a representative cash return that represents the Cash node of the balanced portfolio or cash from another source for internal analysis.

For GIPS, you can use the equity portion of a balanced portfolio but you must also use the equity cash portion of the balanced portfolio. For example, you can select the Asset Class Model as the target dictionary. This model includes a Total level and an Investment Type level (Cash, Equity, Fixed Income, Unknown). You can apply the Cash component a 7% weight based on the 3-Month Treasury Bills index, applying a weight of 93% to the Equity component.

For the Equity component in the target dictionary, you select the Eagle Test Portfolio 1 as the source entity, applying a weight of 93%. You select a source dictionary of Asset Class Model. A value of Total initially appears in the Dictionary Node field, but you want to use only the Equity component of that portfolio for the custom benchmark.

To create a carve out:

Follow the procedure to create a Blended Weighted benchmark or a Floating Weight benchmark.
Select a value from the Target Performance Model drop-down list.
A grid appears at the bottom of the workspace with column headings that are based on the model you selected.
Double-click a cell in the grid to define values for the blended index.
A dialog box appears with Entity, Source, Model, Model Node, and Weight fields.
Enter values for Entity, Source, Model, and Weight.
If you want to change the Model Node, double-click the Total section.
The Dictionary Node Selection dialog box appears showing the structure of the target performance model selected. If the performance model has a level below Total, the level 2 grouping is displayed.
Select the appropriate node in the tree structure and click Save.
The selected node name appears in the Total dialog box.
Enter a numeric value in the Weight text box and click OK.
The dialog box closes. After you assign values for each node, those values appear in the target dictionary grid.
Select other entities to define using the procedure outlined above, as needed.
When you have completed your entries in the grid, click Save.
The custom index attribute information is saved in the database in the RULESDBO.CUSTOM_INDEX_ATTRIBUTES table.
Additional pages are available for Benchmark, Peer Group Association, Disclosures, and Permissions, but they are not required for the custom benchmark setup.