Predicting financial markets with Google Trends and not so random keywords

摘要

We check the claims that data from Google Trends contain enough data topredict future financial index returns. We first discuss the many subtle (andless subtle) biases that may affect the backtest of a trading strategy,particularly when based on such data. Expectedly, the choice of keywords iscrucial: by using an industry-grade backtesting system, we verify that randomfinance-related keywords do not to contain more exploitable predictiveinformation than random keywords related to illnesses, classic cars and arcadegames. We however show that other keywords applied on suitable assets yieldrobustly profitable strategies, thereby confirming the intuition of Preis etal. (2013)