FORECASTING VIX INDEX AS A MEASURE OF MARKET VOLATILITY BY THE USE OF GOOGLE QUERIES

Abstract

Modelling human behavior is rather challenging as imitating it with proxy variables is not straightforward. In recent years, search engines collect and provide us with a plethora of data, which might be a rather effective way of analyzing or forecasting human behavior. Although several authors tried to answer various questions on the usage of Google Trends data in financial markets, to the best of our knowledge, there are no previous studies carried out to forecast VIX Index using the Google searches on oil and related terms. In this paper we use Google searches on oil and related terms as a proxy variable for human expectations to model the CBOE Volatility Index. To that end, tradetional ARDL modelling was applied. The results indicate that there is statistically signify-cant relationship between Google queries on oil and market volatility. We explain this from the perspective of decision making since certain search activities on Google reveal the urge to show certain behavior and, on the other hand, the same behavior affects the market volatility.