Overview of Rolling Forecast:
Dynamic business environment and economic challenges has made businesses to re-think their planning strategies. In such scenarios, rolling forecast methodology helps to deal with such uncertainties effectively. The annual budget becomes redundant once the actuals of first month or quarter become available, which makes a re-forecast a necessity. With re-forecast you will have most recent (real time) updated numbers leading to more informed decisions.
Rolling forecast is forward looking method based on the most current data. The granularity of these forecasts is not too high, thus taking shorter time to prepare them. Rolling forecasts can be done both on monthly and quarterly basis. Both have their own pros and cons. While a monthly forecast make require more efforts, a quarterly forecast doesn't give out timely information. Whether a forecast should be done monthly or quarterly should depend on the nature of the business. Businesses thriving in highly dynamic business environment might require a monthly rolling forecast. While a business in fairly static environment and one which follows a cyclical trend can opt for quarterly forecasts.
Traditional forecasting vs Rolling forecast
Traditional forecasting process focuses on current year and takes place monthly or quarterly. The value derived from prior plan, actual or forecast data but is not the most recent one and process also takes 2-3 days. Hence this process is not the accurate one and do not provide quality data to have business decisions.
Rolling forecast is a continuous process focusses on forward movement of data as it looks for next 12-18 months data and hence no annual planning and forecasting is required.
Why Rolling Forecasts?
- It is forward looking process and hence eliminates need for annual planning/forecasting.
- This process is ongoing and revised regularly based on most recent data hence you have better picture of what's going on in business to check impacts of previous strategy and make necessary corrections.
- It is driver based planning and forecasting method so it allows you to have value derived from the other values.
- It also reduces resource effort and cost and hence increases productivity and profitability.
We can say that rolling forecast is more proactive approach, focuses more on factors and analysis than data gathering, involving more teams/departments and hence providing most recent and accurate data for analysis and decision-making.
Rolling forecast in Hyperion Planning
On the user side, implementing a rolling forecast in Hyperion Planning it would affect the data source (re-forecast of future periods). The application administrator has to provide for functionality that addresses frequent changes to dimension members and create calculation logic to provide for the rolling functionality.
Below changes will be required to implement the rolling forecast in a Hyperion planning application:
i. Changes in substitution variable
In this first step we are setting the substitution variables for respective scenarios.
A substitution variable becomes very important in a rolling forecast implementation. A substitution variable is a placeholder for members which changes regularly. Most common usage of these variables is for setting periods such as current month or current year. Once the value of these substitution variables change, the update happens in all locations where the particular variable is being used. This saves a lot of effort and makes application maintenance easier.
You can have a MAXL script to set the value for each scenario and every month a new MaxL would be run which changes the value of the substitution variable accordingly.
ii. Changes in planning application/ web forms
As you need to bring in the new scenario in planning application, the layout of the data form also need to update accordingly so that the user can view the new scenarios. Changes in web form includes putting year dimension in Page and put all the twelve months in columns to make it simpler for user to move from one year to another.
For a year-over-year forecast, there will be multiple columns in web form to show the periods which can extend up to multiple years. For example, In case of 12 month rolling forecast, if we forecast in month of Jan then forecast period will be till December of same year while in case of forecasting for other months, it will extend to next year.
iii. Changes in Business Rules/member formula/calculation scripts
Changes in calculations is also required so that it can accommodate the calculation to roll year over year. Usage of substitution variables is consider as best practice and is very helpful for rolling forecast implementation.
Conclusion:
Rolling forecast is an innovate approach used by the business nowadays so as to remove the burden of an annual plan. Although the approach is similar to the tradition method but the logic is completely different. It is ongoing, driver based, forward looking process to allow focus on data which they can control.
Thus looking at the pros concept of Rolling Forecast model we successfully implemented it into our existing system.