Projection Inputs – Slope Software

Projection Period & Tags

The leftmost header section is where you can set the start date and duration of the projection to be run in addition to informational tags used for organizational purposes. Click on the Edit button in upper right corner of this section to modify the settings.

Projection Run Settings

The projection run settings section is where you set the scenarios to run and other options about how to process the projection. Click on the Edit button in this section to modify any of these settings.

SCENARIO TABLE

You must select one Scenario Table to use for the projection. That scenario file will contain one or more scenarios to be run during execution. You also have an option to run a subset of the scenarios contained in the Scenario Table. On the edit screen for the scenario table, you can enter a comma-separated list of scenario numbers or ranges (e.g. 1-3,5,8). SLOPE will automatically run each scenario on a separate virtual machine (VM) in order to optimize the overall run time of the model.

OUTPUT ALL RESULTS

The output all results option acts as an override for the Output Variable property on each individual variable. When selected only calculated variables will be written to the results database including time periods outside the projection period and will be available for reporting. This option can also be helpful for diagnosing problems with calculations as it will let you see all intermediate value results. Running projections with this option enabled will slow down the overall run time, so it is suggested that you only turn this on during testing and debugging processes using a smaller set of data than full runs.

OUTPUT FORCE RECALC RESULTS

The output force recalc results option provides a way to see the intermediate results of Force Recalc loops within the model calculations. When selected, every variable calculated as part of a force recalc loop (i.e. which has the Force Recalc variable property set) will be saved and available for reporting in the special force recalc results tables. This option can also be helpful for diagnosing problems with calculations as it will let you see all intermediate value results. Running projections with this option enabled will slow down the overall run time, so it is suggested that you only turn this on during testing and debugging processes using a smaller set of data than full runs.

DISABLE MODEL POINT OUTPUT

The disable model point option turns off the writing of individual model point results to the results database from this run. Model point level output is the largest volume of data produced during a run, therefore turning this option on can result in significant run-time improvements. However, this option should only be used if results are only needed at the product level or higher.

FORCE STATIC PROCESSING

The force static processing option overrides the normal behavior of the calculation engine Processing Order and forces all products to be run in static processing mode. This option should be used in situations where a product has optional dependencies on Portfolio or Company level results, but those dependencies are not being used in the current run. For example, this option allows liability only runs to bypass the normal portfolio return dependency and use a portfolio assumed rate from a table or static value. SLOPE will not automatically detect that products can be run in static processing mode, even when there are IF conditions or other logical operations that would bypass the need for dynamic processing in the variables. The Force Static option is the only way to force a product to be run in static mode, even when Portfolio or Company references exist on a given product.

Turning this option on can result in faster runs and reduced memory usage, thus lowering the compute environment requirements for the run. However, if the product variable and assumptions are not properly set up to skip the portfolio or company dependencies, then this option could result in unintended results in the calculations.

NUMBER OF CORES

The number of cores option allows you to select the number of virtual CPUs used for each scenario run. Smaller runs can use a smaller number of CPUs (reducing run cost) while large runs can utilize the larger available capacity. The edit settings screen will display all compute environment options available for use. If the available options do not fit your needs, other options can be made available by submitting a request to support@slopesoftware.com.

Assumption Sets

The assumption sets section is where you select Assumption Sets to be used in the current projection. You may select as many assumption sets as desired for each projection. The assumption sets will automatically set the table inputs on the projection (see more below). If multiple assumption sets are selected that both contain values for the same Table Structure or Decrement Table, the assumption set that appears last (bottom) on the list will prevail (Last In Wins). The table input section of the projection input screen will also display the name of the Assumption Set that it received its value from so there is no confusion on what assumptions are being used.

Variable Scoping Controls

Variable tags from model development can be selected in order to filter the variables that are executed during the model run. Variables with no tags specified can also be included or excluded. The Variable Scoping Controls can be used to reduce the output for performance reasons or to help focus on the results of only a handful of variables. Any variables that are excluded by a tag or the absence of a tag (when Variables without tags is set to excluded) will only be calculated if they are referenced by another calculated variable.

If Included Variable Tags are set, then everything else with a tag that is not listed in the Included Variable Tags section will be excluded by default. However, if a variable has both an included tag and an excluded tag, it will be excluded. This allows you to exclude a subset of the included tags. For example: If a variable has both "Valuation" and "US GAAP" tags in the screenshot, then it would be excluded, but all other variables with "Valuation" would be run.

Note: If a variable is included via the tags, it must still be set to always calculate (or be referenced by another variable that is set to always calculate) in order to be calculated.

Add Portfolio(s)

Each projection must contain at least one portfolio but may contain more. In the Included Portfolios section of the projection details page, click on the Add/Remove Portfolios button to add one or more Portfolios to the run. On each Portfolio, you can set each of the following:

Select the Portfolio you want to run.
Select a sales allocation if you are using dynamic sales (this can be omitted if not)
Select one or more Investment Strategy to apply (this can be omitted if no asset purchase/sale is desired)
Add Products to the portfolio for each in-force product to be run (assets & liabilities).
Set Model Point Files and Output Groupings for each Product
Set Asset Scaling Variable, if applicable. This variable is used to scale assets within this portfolio at the start of the projection. The selected variable should return a decimal ratio that will be used to scale variables on all assets that scale on purchase/sale at the end of time zero. If left blank, no asset scaling will occur.

Investment and New Business

The investment strategy and new business profile are defined by the Sales Allocation and Investment Strategies selected. See the corresponding articles for more details on each of these.

Set Assumptions

For each Table Structure and Decrement Table in the model, you must select a corresponding Data Table to use for the current run. This is the table that will be read during execution by all the table read formulas defined in the model. These can also be set by using assumption sets to assign multiple tables at once (see above). Any table that has been set by an assumption set will no longer be editable.