On A Generalization of the Box-Jenkins Airline Model

• Activity Number: 204
• Type: Topic Contributed
• Date/Time: Tuesday, August 13, 2002 : 10:30 AM to 12:20 PM
• Sponsor: Business & Economics Statistics Section*
• Author(s): Donald Martin*+ and David Findley
• Affiliation(s): U.S. Census Bureau/Howard University and U.S. Census Bureau
• Keywords: Time series models ; seasonal adjustment ; signal extraction

Abstract

The Box-Jenkins "airline" model is the most widely used ARIMA model for seasonal time series. However, in this model, the nonseasonal moving average factor serves the modeling of nonseasonal components, but the "seasonal" moving average factor plays a dual role in that it models nonseasonal as well as seasonal components. This is intuitively unappealing. Also, in the context of seaonal adjustment, it creates a "confounding" effect that makes it difficult to interpret the model's coefficients, particulary when the seasonal moving average coefficient is rather small. In this paper, we examine a slight generalization of the airline model that has a more restricted seasonal moving average factor with only seasonal zeros and has a second-order nonseasonal moving average factor. We focus particularly on the properties of model-based seasonal adjustment filters obtained from this model.