Managerial resilience and financial reporting quality : neural network model

Abstract

Using a Word2Vec neural network model on a large sample of firm-year text data from the question and answer section of top managers’ earnings call transcripts, this study measured top managers’ resilient tone (hereafter, managerial resilience). The study investigated the relationship between managerial resilience and financial reporting quality. In today’s business environment, adversities (e.g., financial crises, the COVID crisis, remote work environments, increasing global competition, and the looming threat of war) are inevitable. Relying on the psychological capital and upper-echelon theories, this study argues that resilient managers are more effective at discharging their job responsibilities of internal control over financial reporting. The study found that managerial resilience is positively associated with financial reporting quality. The main results were robust, controlling for managerial characteristics, alternative measures, measurement errors, selection bias, change analysis, reverse causality, and omitted variable bias. The study further found that the relationship between managerial resilience and financial reporting quality becomes stronger in firms with high complexity and weak corporate governance. These findings suggest that managerial resilience is more important in certain situations.