CEO Personality and Firm Policies

Anastasia Zakolyukina is Assistant Professor of Accounting and Neubauer Family Faculty Fellow at University of Chicago Booth School of Business. This post is based on a recent paper authored by Professor Zakolyukina; Ian D. Gow, Assistant Professor of Business Administration at Harvard Business School; Steven N. Kaplan; Neubauer Family Distinguished Service Professor of Entrepreneurship and Finance at University of Chicago Booth School of Business; and David F. Larcker, Professor of Accounting at Stanford Graduate School of Business.

In the paper, CEO Personality and Firm Policies, which was recently made publicly available on SSRN, we use two samples of high quality personality data for chief executive officers (CEOs) and the way they speak during question-and-answer (Q&A) portion of earnings conference calls to develop a measure of CEO personality in terms of the Big Five traits: agreeableness, conscientiousness, extraversion, neuroticism, and openness to experience. These personality measures have strong out-of-sample predictive performance and are stable over time. Our measures of the Big Five personality traits are associated with financing choices, investment choices, and firm operating performance.

Recent research in economics and finance has started to explore the relations between individual traits of senior executives, the investment and financing choices made by these executives, and firm performance (Bertrand and Schoar, 2003). These studies have examined traits such as overconfidence (e.g., Malmendier and Tate, 2005, Ben-David et al., 2013), optimism (e.g., Heaton, 2002, Graham et al., 2013), risk-taking (e.g., Graham et al., 2013), and past experiences and credentials (e.g., Adams et al., 2005, Benmelech and Frydman, 2014, Falato et al., 2014).

Perhaps the other most obvious characteristics are personality traits, which are defined as “patterns of thoughts, feelings, and behaviors that reflect the tendency to respond in certain ways in certain circumstances” (Roberts, 2009, p.140). But, with the exception of Kaplan et al. (2012) and O’Reilly et al. (2014b), very little research examines how (or whether) differences in personality traits are associated with executive decision-making and firm performance. This gap in the literature is puzzling given the extensive literature on personality psychology (e.g., Goldberg, 1993) and the fact that personality is a predictor of work performance for different occupations, including (usually lower level) executives (e.g., Barrick, 2005, Judge et al., 2002).

A fundamental challenge for research into the effects of executive personality is the cost and difficulty of measuring personality. Measuring personality traits typically requires the administration of costly instruments or detailed interviews, which are difficult to implement for large samples of top level executives. The purpose of this paper is to suggest an alternative approach. In this paper, we (i) develop and validate measures of CEO personality that can be obtained for a large sample of observations and (ii) examine the associations of these personality measures with investment choices, financing choices, and firm performance. We develop predictive models for CEO personality by statistically linking linguistic features of the CEOs during the Q&A portion of conference calls to their known personality traits from Kaplan et al. (2012) and Kaplan and Sorensen (2016). We validate our measures of CEO personality traits by showing that they exhibit substantial out-of-sample predictive ability using an independent sample obtained from O’Reilly et al. (2014b).

Our approach relies on research in psychology and linguistics (e.g., Pennebaker and King, 1999) that explores whether differences in the way language is used by individual executives reflects differences in personalities. This research finds that language is not only predictive of personality (e.g., Mairesse et al., 2007, Schwartz et al., 2013, Park et al., 2015), but that language-based personality scores are also stable over time (Park et al., 2015). Our findings are consistent with these studies. Using the O’Reilly et al. (2014b) sample of 28 CEOs, we find out-of-sample correlations between predicted and actual personality scores that range from 0.23 for agreeableness to 0.49 for neuroticism. These out-of-sample correlations are similar to the out-of-sample correlations that range between 0.21 and 0.54 reported in prior studies linking personality to language in different contexts, such as student essays and random conversation snippets (Mairesse et al., 2007) and Facebook messages (e.g., Schwartz et al., 2013, Park et al., 2015).

Having validated these models, we apply them to 70,329 conference calls from StreetEvents, a product of ThomsonReuters, to estimate scores for Big Five personality traits for 4,591 individual CEOs with no personality data and examine the associations between these estimated personality traits and investment choices, financing choices, and performance for the CEOs’ firms. Openness is positively associated with R&D intensity and negatively with net leverage. Conscientiousness is negatively associated with growth. Extraversion is negatively associated with both contemporaneous and future return on assets and cash flows. Although our results are suggestive of possible sources of “manager effects,” it is important to emphasize and caution that, like Bertrand and Schoar (2003), our paper is meant to be descriptive.

The full paper is available for download here.

Both comments and trackbacks are currently closed.