Time series analyses of stock market and consumer confidence indices: A Granger causality test

Abstract:

Following the methodologies proposed by Otoo (1999), Jansen and Nahuis (2003), and Bremmer (2008), I used the vector autoregression (VAR) approach and conducted Granger-causality testing to investigate the possible causality between stock returns and two confidence indices (Consumer Confidence Index and Business Confidence Index) in several American, European, and Asian stock markets: Brazil, China, Germany, Greece, Japan, Korea, Mexico, Spain, Thailand, Turkey, the UK, and the US. I concluded that even though the relationship between stock returns and consumer and business confidence has its own idiosyncratic properties in each country, stock returns still generally Granger-cause consumer and business confidence but not vice versa, with Japan and Korea being the two main exceptions. Stock returns Granger-cause confidence implies that rising (declining) stock markets may make consumers and businesses feel better (worse) about the future, and therefore induce them to spend and/or invest more (less). If stock returns can affect the economic output via the market confidence channel, stock price indices might also be used as another leading economic indicator, especially in emerging economies where leading economic indicators are relatively scarce, not timely, and not reliable (Mauro 2003). On the other hand, the unidirectional effects running from confidence indices to stock returns in Japan and Korea might indicate strong herd-like investment behaviors and/or less efficient regulatory institutions and market integrity together with relatively weak market participation.