Forecasting consumption using the consumer confidence index in Russia

Abstract

The paper considers short-term forecasting of household final consumption expenditures using the consumer confidence index in Russia. The article presents a comparative analysis of consumption forecasting approaches using a leading indicator in various countries. The author makes consumption nowcasts based on quarterly data for the period from 2000 to 2021. Unlike most studies based on seasonally adjusted time series, the current study uses seasonally differentiated time series. To determine the predictive power of the index, the author builds several models which include in turn the consumer confidence index, index lags, and consumption lags. The quality comparison of different specifications with the root mean squared forecast error demonstrates that the inclusion of the consumer confidence index in the model increases the accuracy of both out-of-sample and in-sample forecasts. The conducted statistical tests confirm that the inclusion of a leading indicator in the model improves the quality of forecasts.