News Tone Dispersion and Investor Disagreement
56 Pages Posted: 21 Dec 2012 Last revised: 3 Jun 2017
Date Written: June 3, 2017
We exploit linguistic analysis of firm-specific news to measure aggregate disagreement, based on the notion that investors disagree more when news tone is highly dispersed across firms. Consistent with theories of disagreement, we find that news tone dispersion i) is negatively related to aggregate discount-rate shocks, ii) is positively related to volatility and turnover, and iii) predicts returns negatively on the aggregate market, on high-beta stocks, and on short-sales constrained stocks. Using news tone dispersion to measure disagreement is advantageous as it can be measured over any interval, even days, and reflects new information in a timely fashion.
Keywords: disagreement, news analytics, predictability, stock returns, volatility
JEL Classification: G12, G14, G17
Suggested Citation: Suggested Citation