Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.

This study uses machine‐learning methods to predict next‐period change in profitability based on a model proposed by Penman and Zhang (2004, Working paper, Columbia University and University of California, Berkeley; “PZ”). We find that new machine‐learning methods predict out of sample substantially better than traditional regression methods and provide richer interpretations about the role and impact of different predictor variables through their nonlinear relationships and interaction effects. For example, our results contrast with previous research by showing that both components of the DuPont decomposition (change in profit margin and change in asset turnover) are informative of next‐period changes in profitability. Our results are robust across different performance metrics, alternative machine‐learning models, and software. Furthermore, an unconstrained machine‐learning model using a larger feature space could not significantly improve the performance of the PZ model. PZ variables alone accounted for most of the explanatory power of the unconstrained model, suggesting the PZ model is both well specified (in terms of feature selection) and robust in higher dimensional settings. With respect to the economic significance of this information, we find mixed results. The market appears to adjust its expectations more in line with the machine‐learning predictions relative to the PZ model but the portfolio returns are not significantly different.

We study whether the effectiveness of corporate governance mechanisms varies depending on the characteristics of the executives subject to these mechanisms, namely their “psychological type,” as proxied by their history of legal infractions. In particular, we examine insider trading, where we can compare the trading behavior of different types of executives in the same firm. We find that “recordholder” executives, that is, those with prior legal infractions, earn significantly higher profits from purchases and sales than nonrecordholder executives. Furthermore, the profitability of both purchases and sales is significantly increasing in the severity of the infraction. Governance mechanisms, such as blackout policies, lower profits of executives with only traffic infractions; however, profits for executives with serious infractions appear insensitive to blackout policies. Insiders with serious infractions are also more likely to trade during blackout periods and before large information events and are more likely to report their trades to the SEC after the filing deadline. Collectively, our evidence suggests that while governance mechanisms can discipline executives with minor offenses, they appear largely ineffective for those with more serious infractions.

In this paper, we propose that affective reactions are integral to accounting decision contexts like capital budgeting, and that researchers must jointly consider affect and cognition to better understand accounting decision makers' behavior. We argue that interpersonal relationships are characteristic of many capital-budgeting contexts, and that these relationships can lead to emotional affective reactions. For example, reactions such as frustration and anger may result if a manager is treated unfairly by another individual involved in a capital project. Drawing on relevant work in neurobiology and psychology, we then predict that these affective reactions can influence managers' capital-budgeting decisions. We report on four experimental scenarios that demonstrate the impact of affective reactions on capital-budgeting decisions. Consistent with our predictions, the results indicate that managers consider both financial data and affective reactions when evaluating the utility of an investment alternative. Our results suggest that researchers should consider both affect and cognition to more fully understand decision making in accounting contexts.

Discretionary bonus adjustments allow managers to restore the alignment of employee effort and compensation when bonus amounts are based on noisy objective performance measures. The implications of discretionary adjustments for employees' future efforts and fairness perceptions present important trade‐offs for managers to consider. Adjustments may be used to motivate different types of effort in future periods, but may also create perceptions of unfairness among employees who are not affected by negative events. This study examines the joint influence of the likelihood of future negative uncontrollable events and compensation interdependence (i.e., the extent to which one employee's compensation influences others' compensation) on managers' willingness to make adjustments for the effect of a negative uncontrollable event on a single employee. In our experiment, we manipulate the likelihood of future uncontrollable events and whether bonuses are determined individually or are drawn from a shared bonus pool. Results show that managers are less willing to adjust when the likelihood of future events is high to avoid setting a precedent, thereby motivating employees to adapt to changing conditions. We also find that managers are less willing to adjust, regardless of event likelihood, when compensation interdependence is high, to avoid demotivating unaffected employees. Finally, we find that participants' general attitudes toward compensation significantly influence their adjustment decisions beyond the effects of our independent variables. Our results highlight the unique nature of discretionary adjustments, help explain findings from previous research, and demonstrate important considerations managers must make when using the flexibility provided to them in pay‐for‐performance contracts.

Since the passage of the Sarbanes‐Oxley Act of 2002, many notable frauds have been tied to ineffective audit committee (AC) oversight. As a result, AC oversight is of continuing interest, and regulators continue to debate this issue, garnering a growing body of research focused on the role played by the AC. But little theoretical research exists to guide analytical and empirical researchers investigating AC oversight. The purpose of this study is to provide theoretical guidance by examining AC oversight in a strategic setting. We focus on the AC's role in overseeing internal controls (ICs) and the impact of whether the AC relies on management in designing the controls. We characterize how the nature of control risk changes and how IC strength is associated with the amount of managerial fraud, expected probability of fraud detection (which, on average, equates to audit effort), and audit quality (assessed as 1 − audit risk) across two settings defined by the degree of AC oversight. As one example that highlights the need for theoretical guidance, we consider the literature's presumption that IC strength is negatively associated with audit effort. We find that this association may be positive or negative as IC changes, where the association varies with the degree of direct AC oversight and the change in payoff parameters.

Two distinct lines of research have been dedicated to empirically testing how financial reporting quality (measured as the earnings response coefficient or ERC) is associated with management's choice of reporting bias and with audit quality. However, researchers have yet to consider how ERCs are affected by either the auditor's reaction to changes in the manager's reporting bias or the manager's reaction to changes in audit quality. Our study provides theoretical guidance on these interrelations and how changes in the manager's or the auditor's incentives affect both reporting bias and audit quality. Specifically, when the manager's cost (benefit) of reporting bias increases (decreases), we find that expected bias decreases, inducing the auditor to react by reducing audit quality. Because we also find that the association between expected audit quality and ERCs is always positive, changes in managerial incentives for biased reporting lead to a positive association between ERCs and expected reporting bias. When the cost of auditing decreases or the cost of auditor liability increases, we find that expected audit quality increases, inducing the manager to react by decreasing reporting bias. In this case, changes in the costs of audit quality lead to a negative association between ERCs and expected reporting bias. Finally, we demonstrate the impact of our theoretical findings by focusing on the empirical observations documented in the extant literature on managerial ownership and accounting expertise on the audit committee. In light of our framework, we provide new interpretations of these empirical observations and new predictions for future research.