Forecasting the Index of Consumer Sentiment

Abstract

This paper attempts to build and to estimate statistically a forecasting model for the Index of Consumer Sentiment (ICS). Variables in the complex consumer environment are examined to determine which may act upon consumers collectivelyto influence ICS. A model is initially constructed and then tested against quarterly ICS data collected over a twenty-one year period. Coefficients of the variables in the fitted model are examined to indicate the strength of impact on the consumer of each variable. Results indicate that it is possible to forecast ICS with considerable accuracy by using appropriate economic and social data.