Predicting Performance Using Consumer Big Data
Log in to collectAcademic paper
Predicting Performance Using Consumer Big Data [Click to Open PDF]
Strategy in a nutshell
The strategy trades 330 US stocks using consumer data—web visits, in-store activity, and brand interest—to form long-short quintile portfolios. The top quintile is held long, the bottom quintile short, and positions are equally weighted across three proxies.
Economic rationale
Alternative consumer data provides an edge over traditional information sources. Web activity predicts earnings surprises, while brand and in-store data forecast revenue growth, allowing the strategy to capture predictable stock returns efficiently.
Backtest performance
Annualised return9.29%
Volatility14.29%
Sharpe ratio0.62