News You Can Use: "Using Machine Learning to Measure CEO Depression"

With earnings season kicking off we have a natural experiment before us. And we will be paying particular attention to the bankers.

From the Journal of Accounting Research, January 8, 2025:

ABSTRACT
We introduce a novel measure of CEO depression by applying machine learning models that analyze vocal acoustic features from CEOs' conference call recordings. Our research was preregistered via the Journal of Accounting Research's registration-based editorial process. In this study, we validate this measure and examine associated factors. We find that greater firm risk is positively associated with CEO depression, whereas higher job demands are negatively associated with CEO depression. Female and older CEOs show a lower likelihood of depression. Using this novel measure, we then explore the relationship between CEO depression and career outcomes. Although we do not find any evidence that CEO depression is associated with CEO turnover, we find some evidence that turnover-performance sensitivity is higher among depressed CEOs. We also find limited evidence of higher compensation and higher pay-performance sensitivity for depressed CEOs. This study provides new insights into the relationship between CEO mental health and career outcomes.....

....MUCH MORE (the paper is open access)

 

Yeah, yeah, CEO career outcomes, blah, blah, blah. 

What we want is insight into what they are seeing but not saying. I mean the CEO is not going to be talking "Deep, ineffable sadness," or quoting Nietzsche:

He who fights with monsters might take care lest he thereby become a monster.
And if you gaze for long into an abyss, the abyss gazes also into you.

But the CEO might not have the iron discipline presumed by Voltaire's comment:

 

"Ils ne se servent de la pensée que pour autoriser leurs injustices,
et emploient les paroles que pour déguiser leurs pensées" 

—François-Marie Arouet--'Voltaire', Dialogue xiv. Le Chapon et la Poularde (1766).

"Men use thought only to justify their wrong doings, and employ speech only to conceal their thoughts"

Yes, AI and machine learning can help us detect, translate and interpret things like the CEO of RH (formerly Restoration Hardware) saying during his earnings call last week: ‘Oh, sh—’ as the stock dropped 40%.

—NBC New York: https://www.nbcnewyork.com/news/business/money-report/oh-sh-rh-ceo-reacts-live-to-stock-tanking-on-tariffs-poor-earnings/6211001/

So we look forward to the conference calls we will hear over the next month or so, and the embedded information we can glean from them.