INFLUENCE OF GOOGLE TRENDS ON SHARES LISTED ON THE BRAZILIAN STOCKS EXCHANGE: EVIDENCE FROM PVAR MODELING

The proposed work analyzed how searches on the Google search engine influence the return, volatility and volume of the shares that make up the Ibovespa index, considering the period between 2015 and 2020. For this, the Auto-regressive Vector Panel (PVAR) modeling was applied. The results implies that the historical search volume for the company name and the ticker show a bidirectional relationship with the standard deviation of returns, volatility and traded volume, suggesting that the investor's demand for information is met, in part, by search engines, which that there is also increase in the traded volume, after a shock, in historical volume of search on the ticker. This evidence points to market efficiency, at least, for weekly situations, in which there is a possibility for the investor to research and understand each new market situation. However, the company's follow-up does not guarantee, at least in the long run, determining that the Efficient Market Hypothesis, in the semi-strong version, is indirectly obtained, by increasing the volume traded without altering the return in long run. Therefore, the use of Google Trends can, to some extent, improve the accuracy of forecasting models that seek to predict the return, volatility and volume of shares in short term.

Keywords:
Google Trends, Stock markets, Ibovespa index; PVAR