Internet Search Investor Sentiment And Non Equity Markets

In this work quantitative internet search data from Google Trends is used as a proxy to examine investor sentiment in the US gold, commodity and housing market. The results of multiple regression models for daily and weekly data between January 2014 and December 2018 exhibit several implications. First, investor sentiment plays an inferior role for non-equity markets, compared to stock markets. Second, while sentiment measured with stock market-based proxy VIX is a strong predictor for contemporaneous non-equity market returns, internet search sentiment data provides significant results about return reversals for lagged market returns. Third, commodity and housing markets are more susceptible to search-based investor sentiment compared to the gold market. Finally, when relating internet search data to market returns, the results strongly depend on the used time frame. Although there are limitations to internet search data, the findings are broadly consistent with investor sentiment theory.