Evaluating Stock Market Performance Using Aggregated Employee Reviews

Abstract

Investors are constantly driven by a desire to outperform the market. While investments may perform well in the short-term, there is great difficulty in achieving long-term success. We show that using aggregated employee reviews left on Glassdoor can result in significant performances spanning multiple years. A portfolio managed as the top quintile of firms by their Glassdoor rating has a significant Four-Factor alpha and excess returns above the S\&P500. Strong positive changes in employee's perception on "Senior Leadership" and "Culture \& Values" can also be used to create significant portfolio performance, with its significance additionally tested through a Fama-MacBeth regression. Lastly, using a text classification model to categorize employee reviews, we show that the dimensions "Agility," "Customer Orientation," and "Engagement" are indicative of significant positive performance.