WILMORE PAPER COMPANY

Advertising Budget Model

Time Series Analysis by Burcu Arikan, Spring 2008

Introduction

Time-series analysis allows researchers to predict future values of time series variables. In time-series analysis, researchers look for systematic patterns in the data in order to make inferences about future data outcomes. The belief that the past behavior of a variable may continue into the future constitutes the rational basis behind the use of time series forecasting. In the business world, companies must plan for the future on a permanent basis if they are to capitalize on market opportunities and timely address potential problems. Time series can make a valuable contribution in this quest.

In this study, 30 quarters of advertising expenditures and sales data for Anheuser-Busch (see below) were analysed with the purpose of predicting future Anheuser-Busch sales.

To attain our objective, a four-step building process is followed: (I) Specification, (II) Estimation, (III) Verification, and (IV) Validation. Each one of these stages is detailed in the corresponding linked page of this website.

TABLE 1: ADVERTISING EXPENDITURES AND SALES