Predicting Future Buyer Behavior with Consumers' Confidence and Sentiment Indexes

Knowledge of consumer attitudes and expectations is an indispensable marketing tool and may be useful for predicting changes in consumer spending and saving patterns. However, most empirical research on the value of consumer confidence (sentiment) indexes for forecasting future buyer behavior has focused on the statistical correlation between these indexes and several measures of consumer spending and business activity. As Huth et al. (1994) correctly pointed out, a more appropriate approach is to focus on the direction of Granger-causality between these variables. Prior to testing for Granger-causality, we investigate the stationarity properties of alternative measures of these variables and we employ a flexible lag structure. Our empirical results support the use of these indexes to forecast business activity but not consumer purchases or overall economic activity.