Can Google Search Volume Data Help Predict Future Stock Measures?

Sammanfattning: We hypothesise that search volume data from large search engines, such as Google, can be used to predict higher stock returns, higher liquidity and higher volatility. To test the validity of our hypothesis, we empirically test the relation be-tween the public Search Volume Indices provided by Google Inc. and weekly stock returns, stock liquidity and stock volatility in a sample of OMX Nordic stocks from 2004 to 2011. We do this by performing a Fama-MacBeth regression analysis and conducting two event studies. We find that abnormal search volumes can help predict abnormal returns and abnormal liquidity in the two subsequent weeks. Moreover, we find that these relations are stronger for smaller firms than larger firms. Furthermore, we find that the predicted abnormal return is highest when high growth in search volume is observed. The results are robust to the inclusion of control variables and to the methodology. Our findings suggest that it is possible to construct a market neutral trading strategy and gain up to 11.5 percent return per year, before transaction costs and taxes, by investing each week in firms with highest search volume growth each.