Multiple Time Series Analysis of Competitive Marketing Behavior

Abstract Competitive marketing behavior is analyzed and modeled using the framework of multiple time series analysis (MTSA). MTSA is proven to detect causality of marketing mix variables and sales (market share), and the lag structure of carryover effects of the sales and advertising relationship. An empirical analysis is conducted by applying this framework to comprehensive time series data of a particular industry where three firms are competing in an oligopolistic market. In comparison with the traditional MTSA approaches used by market researchers that use bivariate causality testing, the MTSA approach proposed in this study based on the vector autoregressive moving average (VARMA) method has advantageous characteristics for marketing analysis as it permits causality testing of all variables simultaneously. Yet its potential capability to analyze marketing problems has not yet been explored by marketing researchers. This study shows that VARMA modeling is capable of capturing the dynamic competitive structure of the market. Major findings of the empirical analysis indicate that media advertising, promotion, and price are the primary arenas of competition in this industry and that the VARMA models outperform univariate time series models in goodness-of-fit measures as well as in forecasting performance.