Firm-Level Cybersecurity Risk and Idiosyncratic Volatility

The authors propose a measure of firm-level cybersecurity risk developed by employing pattern-based sequence-classification method from computational linguistics to determine the proportion of time devoted to issues related to cybersecurity risk during earnings conference calls. Using their measure, they investigate the effect of cybersecurity risk on firm-level return volatility; they examine both idiosyncratic volatility and implied volatility and find that firm-level cybersecurity risk is positively correlated to idiosyncratic volatility on the days on which earnings calls are held. This suggests that the discussion of issues related to cybersecurity risk during earnings calls is related to an increase in the component of the volatility that responds only to firm-specific news. That positive relationship is robust to alternative measurements of the language in earnings call discussions and to industry classifications.