How do I use an assumption model?

This topic contains the following:

  • A brief overview of assumption models

  • Important considerations that might affect how you build and use assumption models

  • Steps for how to create, modify, and use an assumption model

  • A scenario that describes use of a simple assumption model

Assumptions are baseline data that apply to multiple models, especially financial models. The following are examples of assumptions:

Cost per unit

Tax rates

Revenue per unit (price list)

Material costs

Exchange rates

Growth factors

In Planning Business Modeler, you can store logical groups of assumptions in assumption models. These assumption models can be referenced by other models and they are typically used for forecasting activities.

Using an assumption model as a single source for assumption data provides the following key benefits:

  • Promotes data integrity. Each assumption has only one version. Business rules that reference global assumptions can be standardized. This eliminates the need for users to create ad hoc calculations in Microsoft Office Excel.

  • Facilitates maintenance. Assumption data, which is often variable, can be updated in only one location.

  • Streamlines modeling activities. Other models can be easily linked to existing assumption models.

Important considerations

We recommend that you review the following information before you work with assumption models:

  • You must belong to the Data Administrator or Modeler role to perform the tasks that are described in this topic. When possible, configure assumption models from the model site in which it was created. If you are working across model sites, make sure that you have the correct scope. For more information about the scope of administrative roles, see About permissions for administrative roles.

    Modifying or deleting an assumption model that is linked across models sites may have specific scope requirements. For more information, see the "Modifying an assumption model" section in this topic.

  • The level of detail of a member set (leaf level) in the assumption model must be higher than or equal to the member set in the main model. For example, you can link a main model that has a quarterly time granularity to an assumption model that has a yearly time granularity, but not the reverse.

    An exception is Exchange Rate assumption models, which contain currency exchange rates. Because exchange rate logic is very specific, you must have a separate exchange rate model to match each level of time period in your application that uses exchange rate assumptions. Otherwise, the resulting data might not be valid. For example, if you apply a monthly exchange rate to data that is tracked daily, the results will not account for variances that are caused by fluctuating exchange rates.

  • To write back to assumption data from PerformancePoint Add-in for Excel, the form must be based directly on the assumption model. Users cannot write back on assumption data from a form that is based on a model that is linked to an assumption model.

Creating an assumption model

To track your progress, follow the steps that are listed in the following table.

Step Task Related topic

1.

Determine the data that you want to include in the assumption model and how to organize it.

Not applicable

2.

Add the member sets to the model site.

Create a member set by using the wizard

3.

Create the assumption model. You can create two types of assumption models:

  • Global. A shared model that can contain any assumption data.

  • Exchange Rate. A shared model that supports predefined currency rules and is used to track exchange and currency information.

An assumption model can be edited only at the model site where it was created.

Create a model

4.

For exchange rate models, repeat for each level of granularity that you want to support.

The granularity of a member set that is used in an assumption model must be higher than or equal to the granularity of the member set in the main model.

Not applicable

Modifying an assumption model

You can change model properties in Planning Business Modeler just as you can in any other model. However, if you want to make structural changes such as adding a member set, the following restrictions apply:

  • Before you can save your changes to an assumption model, you must first remove the links to other models.

  • You cannot delete an assumption model that is linked to a model in a child model site.

If the assumption model is linked to models from other model sites in the application, the scope of the Data Administrator or Modeler role to which you belong determines whether you can see and remove all links. To see and remove links to models from other model sites, you must belong to one of the following roles:

  • The Data Administrator or Modeler role in the model sites that contain the linked models

  • The Data Administrator or Modeler role that has application-wide scope

To track your progress, follow the steps that are listed in the following table.

Step Task Related topic

1.

Remove links to the assumption model from other models.

The assumption models that are linked to a main model are shown in the Assumptions section on the Model Summary page. You will see only the linked models that are in the model sites to which you have Data Administrator or Modeler permissions.

Add or remove an assumption

2.

Check out the assumption model from the model site in which it was created, make your changes, and then check it in.

See "Check in and check out items in Planning Business Modeler" in the product help.

3.

Restore the links to the assumption model from other models.

Add or remove an assumption

4.

Deploy the model sites that contain the linked models.

Deploy a model site

Editing assumption data

You can edit assumption data by using one of two methods:

  • Perform a data load job from Planning Business Modeler for the assumption model and then redeploy the model site. For more information, see the topic About Data Load jobs.

  • Write back on the assumption model from PerformancePoint Add-in for Excel. You must write back on a form that is based on the assumption model; you cannot write back assumption data from a linked model. For more information, see the topics About permissions for business roles and "Assign forms to a user for data submission" in the product help.

Using an assumption model

You can link assumption models to models that are in the same model site or in its child model sites.

The following considerations may apply to your implementation:

  • You cannot link an assumption model to another assumption model.

  • A model can have only one exchange rate model linked to it at a time.

  • When the dimensions in the main model and assumption model differ by member set, assumption data will be available only where it can be calculated. This is demonstrated in the following examples:

    • If a dimension in an assumption model is not in the main model, the dimension is ignored in the main model.

    • If the main model and the linked assumption model contain different member sets from the same dimension, assumption data is added to leaf members of the main model where possible. If a member is not in the assumption model, it will remain blank. Parent members will be the result of the aggregated data from children that are correctly resolved.

To track your progress, follow the steps that are listed in the following table.

Step Task Related topic

1.

Review the overview to gain an understanding of assumption models.

About assumption models

2.

Identify assumptions that you want to access and the assumption models that contain them.

Not applicable

3.

Link appropriate assumption models to the main model.

When you link to an assumption model, a snapshot of the assumption data is created. This snapshot is what you can see from the main model in PerformancePoint Add-in for Excel.

Add or remove an assumption

4.

Optional. Reference the linked assumption model in a business rule.

The MODEL keyword is used to reference an assumption model in a rule. In addition to other factors, consider the following behavior when you determine which type of rule to use:

  • A SQL rule references the actual assumption data and obtains the most up-to-date data.

  • An MdxQuery or MdxScript rule references the snapshot of the assumption data.

The "Forecast sales based on an assumption" section in Assignment example: Creating a forecast

About rule implementations

5.

Optional. Reference the linked assumption model in a form from PerformancePoint Add-in for Excel.

For information about how to create forms, see PerformancePoint Add-in for Excel Help.

6.

Configure business roles by using the following considerations:

  • Users must be given Read permissions for at least one member in every dimension in the assumptions model, even if the dimension is not used in the main model. Otherwise, users will not be able to see any assumption data.

  • Writeback on assumption data will not succeed from a form that is based on the main model, regardless of role permissions or where the main model resides. In order for you to write back to assumption data from PerformancePoint Add-in for Excel, the form must be based directly on the assumption model.

  • Role permissions that are defined for member sets in an assumption model in a parent model site do not apply to a linked model in a child model site. When users connect to the linked model, the permissions that are defined at the child model site are enforced.

How do I create roles and configure permissions?

Scenario

Amy Rusko is a project manager for a bicycle manufacturing company. She wants to forecast the cost per unit for a line of mountain bikes monthly for the next year. She is concerned about the recent trend of increasing transportation costs.

In PerformancePoint Add-in for Excel, she creates a form that is based on the Production Units model, which is linked to a global assumption model called Component Costs. The assumption model contains the Shipping Rates assumption and other drivers of material costs. The Production Units model contains business rules that reference the assumption model to calculate a variety of operational metrics, such as cost per unit.

She adds a matrix that contains the product line and the gross profit margin for each SKU for the past six months. Then she adds a column that contains a business rule to calculate material costs based on a 1.45% monthly increase in shipping rates. When she views the projected data in PerformancePoint Add-in for Excel, she can evaluate the potential effect to overall production costs and identify at-risk products.

For another example of how an assumption model is used, see the section "Forecast sales based on an assumption" in the topic Assignment example: Creating a forecast.

See Also

Concepts

About assumption models