Consumer Confidence and Asset Returns

Investor sentiment is a key topic in financial research, with many studies exploring the potential of consumer sentiment to predict financial asset returns. However, a consensus on this relationship has yet to be reached. This paper extends previous research by utilizing the Long Short-Term Memory (LSTM) model, a powerful predictive tool, to examine the relationship between consumer confidence and stock, bond, and futures markets returns. Unlike prior studies that that focused primarily on linear relationships, particularly between investor sentiment and stock returns, this research considers the non-linear dynamics across multiple asset classes. The analysis uses monthly data from January 1978 to December 2021, including CCI from the University of Michigan official website, stock, bond and future market returns from main financial databases. The findings indicate that the CCI demonstrates significant predictive power for bond market returns (BMR), a noticeable ability to predict stock market returns (SMR) but limited forecasting capability for future market returns (FMR).