Challenges with Forecasting Ratios in Hierarchies

• Activity Number: 247
• Type: Contributed
• Date/Time: Monday, July 30, 2012 : 2:00 PM to 3:50 PM
• Sponsor: Business and Economic Statistics Section
• Author(s): Bruce Elsheimer*+ and Michael J Leonard and Michele Trovero and Peter Dillman
• Companies: SAS Institute and SAS Institute and SAS Institute and SAS Institute
• Keywords: Time Series Forecasting ; Hierarchical Forecasting ; Forecasting Ratios

Abstract

Many business and regulatory scenarios require forecasting ratios of time series. This can be a challenging task, especially when the number of series is very large and/or the series are organized in a hierarchical fashion. Automatic time series model selection techniques can be used to forecast ratios formed by numerator and denominator series that aggregate to form a hierarchical time series. The goal is to achieve consistent reconciliation of the forecast ratios throughout the hierarchy. This presentation will compare methods for automatic forecasting of ratios of time series, possibly organized in a hierarchical fashion. SAS® High-Performance Forecasting, SAS® Forecast Server software and SAS/ETS® software are used to implement these methods.