Who’s managing earnings? A personality study of earnings management - Using algorithms to analyze how CEO personality traits engage in earnings management

Abstract

Extensive research explains the underlying reasons to why firms engage in value-destroying and risky earnings management. There is subjectivity involved in these decisions, and to explain why managers engage in earning management differently, recent research has started to go beyond the traditional motives to explain why CEOs manage earnings differently. Postulating that earnings management is a costly and risky activity, we use a sample of firms suspected of managing earnings to analyze CEO personality traits association to earnings management decisions. The Big Five Personality Traits are scored from earnings conference call (ECCs) transcripts using a newly developed open-language machine learning algorithm specifically trained and validated on CEOs. We find long-term focused conscientious CEOs, who often want to avoid accountability for suboptimal outcomes, to manage earnings the most to reach benchmarks and increase both real- and accruals management. Similarly, insecure and risk-averse neurotics want to avoid confrontation causing them to increase real-earnings management, whilst accruals management is regarded as too risky. Risk-seeking extraverted CEOs who are often great rhetorics, and non-conforming open CEOs who are often great visionaries both decrease real-earnings management since it is value-destroying. Thus, we find interesting associations between personality traits and earnings management decisions. We also see potential for further research using personality within accounting and other research domains.