Skip to main content
Top
Published in: Review of Quantitative Finance and Accounting 1/2024

Open Access 06-11-2023 | Original Research

CFO overconfidence and conditional accounting conservatism

Authors: Lu Qiao, Emmanuel Adegbite, Tam Huy Nguyen

Published in: Review of Quantitative Finance and Accounting | Issue 1/2024

Activate our intelligent search to find suitable subject content or patents.

search-config
loading …

Abstract

This study investigates the association between Chief Financial Officers (CFOs) overconfidence and conditional accounting conservatism. Relying on upper echelons and overconfidence theories and based on a large sample of US-listed firms’ data from 1992 to 2019 (21,626 firm-year observations), we find a statistically and economically significant negative relationship between CFO overconfidence and conditional accounting conservatism, suggesting that overconfident CFOs tend to diminish conditional accounting conservatism. These findings persist in a series of robustness tests. In the mechanism analysis, we predict that overconfident CFOs aim to convey private information by reducing conditional accounting conservatism. We prove this conjecture by observing that overconfident CFOs who adopt lower levels of conditional accounting conservatism increase earnings informativeness (i.e., the amount of information about future cash flows or earnings contained in current stock returns) and reduce their precautionary incentives to save cash. We further rule out another mechanism (i.e. compensation concerns) that may motivate overconfident CFOs to reduce conditional accounting conservatism. Moreover, we show that overconfident CFOs with higher powers are more able to minimize conditional accounting conservatism. Our study highlights the significance and motivation of overconfident CFOs in determining asymmetric recognition of good and bad news.
Notes

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

1 Introduction

Many studies suggest that managers hide or delay recognizing negative news to seek private benefits based on the traditional agency theory perspective, which assumes that managers can rationally estimate investment value at any time (e.g., Bleck and Liu 2007; Kim et al. 2011; Wang et al. 2018).1 However, this assumption only partially matches real-world managers’ decision-making. For example, Graham (2022) recently provides survey evidence that irrational managers have optimistic views of their firms’ value and performance, which might lead managers to perceive their firms’ stock to be undervalued. Consistent with the findings of the survey, the literature suggests that overconfident CEOs, due to the overestimation of the profitability of existing projects, are less likely to timely disclose bad news (Ahmed and Duellman 2013; Kim et al. 2016; Pierk 2021).2 These studies concentrate on CEOs, whereas other executives’ overconfidence has received less attention. This single focus might result in false attribution because it neglects the role of other managers who are accountable for particular areas, especially the firm’s accounting and finance (Malmendier et al. 2022).
Inspired by the above concern, our study focuses on the role of overconfident Chief Financial Officers (CFOs) in conditional accounting conservatism for two reasons. Firstly, CFO overconfidence is prevalent and plays a critical role in the firm’s financing and accounting decisions for which the CFOs are primarily responsible (Bertrand and Schoar 2003). For instance, CFO overconfidence outweighs CEO overconfidence in financing (Malmendier et al. 2022), cost stickiness (Chen et al. 2022), and earnings management (Qiao et al. 2023).
Secondly, given that information asymmetry exists between managers and outsiders, financial reporting is an essential tool for managers to communicate firm performance with investors (Jensen and Meckling 1976; Healy and Palepu 2001). Conditional accounting conservatism significantly affects earning quality, which continually attracts attention from investors, boards, and policymakers (Mora and Walker 2015).3 However, much less is known about the association between CFO overconfidence and conditional accounting conservatism, and the motivations that drive overconfident CFOs to influence conservative reporting.
Based on upper echelons and overconfidence theories, we predict a negative link between CFO overconfidence and conditional accounting conservatism since overconfident CFOs seek to communicate private information by diminishing conditional accounting conservatism (i.e., Mensah et al. 2004; Heflin et al. 2015; Qiao et al. 2023). However, from an agency perspective, the opposite outcome is possible. When the board of directors and investors are aware of the flaws of managerial overconfidence (e.g., delayed response to negative news), they might require more conservative information to accelerate the timely recognition of bad news (i.e., Francis and Martin 2010; Louis et al. 2012; Hsu et al. 2017). Accordingly, it is an open empirical question.
Empirically, we test the relationship between CFO overconfidence and conditional accounting conservatism using US-listed firms’ data from 1992 to 2019. To measure conditional accounting conservatism, we use two methods proposed by Khan and Watts (2009) and Banker et al. (2016) regarding the specific estimation of the timeliness of bad news. Following Malmendier and Tate (2005), Campbell et al. (2011), and Chen et al. (2022), we use managers’ option-exercising behaviors to measure CFO overconfidence. Our study finds a statistically and economically significant negative association between CFO overconfidence and conditional accounting conservatism. These findings are robust under difference-in-differences estimation based on propensity-score matching (PSM-DID), using an alternative overconfidence measure, and considering strong governance. In our robustness test, we also suggest that CFO overconfidence has an independent effect on conditional accounting conservatism.
In the analysis of mechanisms, we find that lower conditional accounting conservatism enhances the positive relationship between CFO overconfidence and earnings informativeness (i.e., the amount of information contained in the current stock return concerning future earnings or cash flows) and mitigates the positive relationship between an overconfident CFO and precautionary motives to hold cash. These findings show that overconfident CFOs reduce conditional accounting conservatism to communicate private information, which can give investors more valuable information to predict future cash flow. The released information asymmetry provides managers with better access to external funding and alleviates their precautionary motives (Mensah et al. 2004; Lee 2010; Heflin et al. 2015). It is possible that overconfident CFOs reduce conditional accounting conservatism due to seeking private benefits (Peng and Röell 2008; Chava and Purnanandam 2010; Jiang et al. 2010). We rule out this alternative mechanism and reinforce the validity of the private information explanation. Furthermore, we find that CFO power strengthens the negative association between CFO overconfidence and conditional accounting conservatism, consistent with the notion that managers’ power influences the extent to which their cognitive biases affect decision-making (Hambrick 2007).
Our paper makes several contributions to the literature. Since most studies on CFOs (e.g., Francis et al. 2015; Muttakin et al. 2019) focus on the link between CFOs’ demographic characteristics (e.g., gender and tenure) and their conservative accounting choices, our study supplements these studies by considering the CFOs’ psychological traits (i.e., overconfidence). Moreover, our findings provide fruitful evidence for overconfidence studies. Given that Ahmed and Duellman (2013) only focus on the association between CEO overconfidence and conditional accounting conservatism, we extend their study to reveal the effect of CFO overconfidence. Besides, different from Ahmed and Duellman (2013), we consider the influence of unconditional accounting conservatism in the regression on conditional accounting conservatism as conditional accounting conservatism can be preempted by unconditional accounting conservatism (Beaver and Ryan 2005), which increases the robustness of inference.4 Furthermore, our findings suggest that overconfident CFOs would convey private information by reducing conditional accounting conservatism, complementing Ahmed and Duellman (2013) by investigating the motivation for overconfident managers to affect timely bad news disclosure.5
In addition, our findings provide additional empirical evidence validating the integration of the upper echelons and overconfidence theories to investigate the influence of managerial overconfidence (Chen et al. 2022; Malmendier et al. 2022; Qiao et al. 2023). Furthermore, we discover that overconfident CFOs with higher power are more likely to lower conditional accounting conservatism, supporting the updated upper echelons theory (Hambrick 2007). Moreover, we find no evidence that overconfident CFOs adopt less conservative accounting information for their private benefit (Kim et al. 2011). Our findings imply that the incentive of overconfident CFOs is to act in the best interests of shareholders by reducing conditional accounting conservatism, which is inconsistent with the explanation of agency theory.
Our findings also have significant implications for policy and practice. The requirement for accounting conservatism was removed from the joint conceptual framework of the International Accounting Standards Board (IASB) and the Financial Accounting Standards Board (FASB) in 2010 due to their belief that neutrality was a more appropriate financial quality objective (FASB 2010; IASB 2010). However, this decision has created considerable and ongoing debate. Our findings suggest that managerial cognitive biases influence conservative reporting, thus providing additional evidence for standard-setters to make or adjust decisions. In addition, our research can help investors better understand the overconfident behavior of managers and bridge the inconsistent understanding of managerial rationality in academia and practice (Ben-David et al. 2013; Graham 2022).
The remainder of this study is organized as follows. Section 2 presents the theoretical framework, literature review, and hypothesis development. Section 3 shows the research design. Section 4 displays the baseline analysis and some robustness tests. Section 5 shows the analysis of mechanisms. Section 6 discusses the moderating effect of CFO power. Section 7 concludes the paper.

2 Theoretical framework, literature review, and hypothesis development

2.1 The effect of CFO overconfidence on conditional accounting conservatism

Cognitive biases, suggested by overconfidence theory, describe how people overestimate their ability to make accurate judgments, foresee the future, and influence events (Weinstein 1980; Alicke 1985). Managers’ cognitive biases influence their behaviors according to upper echelons theory (Hambrick and Mason 1984). To capture the impact of managers’ cognitive biases on accounting conservatism, early empirical studies focus on demographic characteristics. For instance, they find that managers’ gender (Ho et al. 2015), individual beliefs (Ma et al. 2020), and tenure (Muttakin et al. 2019) can explain firms’ conservative accounting reporting.
Some research, however, contends that managers’ demographic traits have limited explanatory power for cognitive differences (Ge et al. 2011) and that overconfidence is a prevalent characteristic among executives (Goel and Thakor 2008). These findings inspire researchers to explore managerial behavior from the perspective of overconfidence. Since CEOs set the tone at the top, early overconfidence studies concentrate on this executive type. They show that firms with overconfident CEOs tend to overinvest (Malmendier and Tate 2005), make overly optimistic forecasts (Hribar and Yang 2016), engage in more earnings management (Hsieh et al. 2014) and tax avoidance (Chyz et al. 2019), and adopt less accounting conservatism (Ahmed and Duellman 2013).
As the CFO’s role has expanded beyond typical accounting and financial choices and into strategy and operation decisions, recent studies shed light on the impact of CFO overconfidence (Ben-David et al. 2013; Chen et al. 2022; Malmendier et al. 2022). For instance, Ben-David et al. (2013) find that firms with overconfident CFOs tend to make aggressive decisions, such as overinvesting and reducing dividend payouts, because overconfident CFOs are more likely to overestimate their firms’ performance. Consistent with this notion, Chen et al. (2022) suggest that overconfident CFOs are more likely to extend resources while sales expand and keep extra resources when sales decrease since they overestimate future sales growth and underestimate risks. Furthermore, Malmendier et al. (2022) demonstrate that overconfident CFOs tend to use more debt than equity finance when they need external funds. They explain that overconfident CFOs deem their firms undervalued by the market, and thus equity funding is more costly. Therefore, based on upper echelons theory, we conjecture that CFO overconfidence plays a non-trivial role in conditional accounting concervatism.
Overconfident CFOs overestimate the future profits of current projects and believe that investors underestimate the firms’ performance (Ben-David et al. 2013; Chen et al. 2022; Malmendier et al. 2022). As key executives responsible for reporting accounting information, overconfident CFOs may find a way to communicate private information and improve investor valuations (Qiao et al. 2023). There is evidence that conditional accounting conservatism makes it difficult for investors to utilize accounting information to forecast future cash flows. To be specific, conditional accounting conservatism requires more stringent verification standards for recognizing gains than losses, which diminishes earnings informativeness (Barth et al. 2014) and persistence (Heflin et al. 2015), and increases information asymmetry (Mensah et al. 2004; Heflin et al. 2015). Accordingly, combining the evidence on upper echelons and overconfidence theories, we predict that overconfident CFOs might accelerate gain recognition and delay loss recognition to convey private information, adding more information for investors to evaluate firm performance (Barth et al. 2014; Heflin et al. 2015). The foregoing discussion leads to our first hypothesis:
H1
There is a negative association between CFO overconfidence and conditional accounting conservatism.
However, we might find a positive relationship between CFO overconfidence and conditional accounting conservatism from the agency perspective. Prior studies have established that conditional accounting conservatism can be used to monitor managers’ behaviors and alleviate agency problems. Specifically, the board of directors and investors use conditional accounting conservatism to urge managers to timely abandon negative net present value (NPV) projects (Francis and Martin 2010; Louis et al. 2012; Hsu et al. 2017) and improve the veracity of reported earnings (Watts 2003). In addition, since lenders’ upside is limited to contractual interest payments, lenders are more concerned with debtors’ losses than profits when settling debt contracts (Watts 2003). Therefore, lenders might demand more conditional accounting conservatism because it gives them more timely loss information, improves the detection of default risk, and increases the efficiency of debt contracts (e.g., Ahmed et al. 2002; Watts 2003; Wittenberg-Moerman 2008; Zhang 2008). When the board of directors and investors are aware of the shortcomings of managerial overconfidence (e.g., deferring response to bad news), they might demand more conservative information to accelerate the timely recognition of bad news. In view of this counterargument, it is an empirical question whether conditional accounting conservatism is negatively associated with CFO overconfidence.

2.2 CFO power, CFO overconfidence, and conditional accounting conservatism

Based on the updated upper echelons theory, CFO power may moderate overconfident CFOs’ conditional accounting conservatism decisions. The updated upper echelons theory contends that managers’ powers enhance the association between managerial cognitive biases and their decisions (Hambrick 2007). Consistent with this notion, some empirical evidence suggests that a powerful CFO is more likely to influence firm decisions. For example, Chen et al. (2022) suggest that the positive association between CFO overconfidence and cost stickiness is enhanced when firms have powerful CFOs. Besides, Florackis and Sainani (2021) find that powerful CFOs resist excessive pressure from CEOs who have the incentive to manipulate earnings. As we predict that overconfident CFOs tend to reduce conditional accounting conservatism, we further conjecture that overconfident CFOs with more power have a more pronounced influence on conditional accounting conservatism.
Although directors and investors might demand more conditional accounting conservatism to monitor overconfident CFOs, prior studies find that CFOs can outweigh directors in accounting-related decisions (Beck and Mauldin 2014). Thus, we predict that overconfident CFOs with higher power might be able to resist the demand for conditional accounting conservatism made by directors and investors. The foregoing discussion leads to our second hypothesis:
H2
Overconfident CFOs with greater power are more likely to reduce conditional accounting conservatism.

3 Research design

3.1 Sample selection and data sources

We use US-listed firms’ data collected from multiple databases. The accounting data is mainly collected from the CRSP/Compustat merged (CCM) and CRSP databases. The CFOs’ and CEOs’ compensation information comes from the ExecuComp database. For the robustness test, we use institutional investors and corporate governance data collected from Thomson Reuters Institutional (13f) Holdings and BoardEx databases, respectively. The analysts’ information is collected from the I/B/E/S database. Our study’s sample period is from 1992 to 2019. The sample starts in 1992 as managers’ compensation information has been available in the ExecuComp database since then. We exclude financial firms (SIC:6000–6999) and utility firms (SIC:4900–4999). To reduce the potential impact of outliers, we winsorize all continuous variables at the 1st and 99th percentiles. Our main sample has 21,626 firm-year observations. The firm-year observations change when testing the hypothesis related to CFO power, which depends on the available data for each proxy of CFO power. The detailed data sample construction is shown in Table 1.
Table 1
Sample selection
Steps
Observations
Total number of observations for US-listed firms with CCM and ExecuComp data
41,153
Exclude: Financial firms (sic: 6000–6999)
(7,154)
Exclude: Utility firms (sic: 4900–4999)
(2,151)
Exclude: Missing value of the variable used to test the H1 and H2
(10,222)
Total number of observations to test the H1
21,626
Total number of observations to test the H2
The different sample sizes depend on the number of observations of the independent variable

3.2 Variable measurement

3.2.1 The measurement of conditional accounting conservatism

Based on the definition of conditional accounting conservatism (i.e., recognition of bad news is more timely than recognition of good news), Basu (1997) proposes using the contemporaneous sensitivity of earnings to negative and positive returns to measure it.6 The measures of Basu (1997) include the industry-year measurement which assumes that all firms in the same industry are homogeneous and the individual firm measurement which assumes that the firm’s operating traits are stable. However, Khan and Watts (2009) argue that both measures of Basu (1997) have limitations because firms’ conditional accounting conservatism is affected by time- and firm-specific factors.7 Thus, Khan and Watts (2009) modify the method of Basu (1997) and develop a firm-specific estimation of the timeliness of bad news (C-Score) and good news (G-Score). The C-score refers to conditional accounting conservatism. The G-Score and C-Score are estimated as follows.
$$ NI_{i,t} = \beta_{1} + \beta_{2} D_{i,t} + \beta_{3} RET_{i,t} + \beta_{4} D_{i,t} \times RET_{i,t} + \varepsilon_{i,t} , $$
(1)
$$ G\_score\it{1}_{i,t} = \beta_{3} = \mu_{1} + \mu_{2} MV_{i,t} + \mu_{3} MTB_{i,t} + \mu_{4} LEV_{i,t} + \varepsilon_{i,t} , $$
(2)
$$ C\_score\it{1}_{i,t} = \beta_{4} = \lambda_{1} + \lambda_{2} MV_{i,t} + \lambda_{3} MTB_{i,t} + \lambda_{4} LEV_{i,t} + \varepsilon_{i,t} , $$
(3)
where, NIi,t refers to net income before extraordinary items deflated by the market value of equity at the beginning of fiscal year t; RETi,t refers to the annual buy and hold return starting from the fourth month after the previous fiscal year-end; Di,t refers to an indicator variable that equals one if RETi,t is negative, and zero otherwise; MVi,t refers to the log of the market value of equity; MTBi,t refers to the market value of equity divided by the book value of equity; LEVi,t refers to total debt divided by total assets.
Next, replacing β3 and β4 from Eqs. (2) and (3) into regression Eq. (1) yields Eq. (4).
$$ \begin{aligned} NI_{i,t} & = \beta_{1} + \beta_{2} D_{i,t} + RET_{i,t} \times \left( {\mu_{1} + \mu_{2} MV_{i,t} + \mu_{3} MTB_{i,t} + \mu_{4} LEV_{i,t} } \right) \\ & \quad + D_{i,t} \times RET_{i,t} \times \left( {\lambda_{1} + \lambda_{2} MV_{i,t} + \lambda_{3} MTB_{i,t} + \lambda_{4} LEV_{i,t} } \right) \\ & \quad + \delta_{1} MV_{i,t} + \delta_{2} MTB_{i,t} + \delta_{3} LEV_{i,t} + \delta_{4} D_{i,t} \times MV_{i,t} \\ & \quad + \delta_{5} D_{i,t} \times MTB_{i,t} + \delta_{6} D_{i,t} \times LEV_{i,t} + \varepsilon_{i,t} , \\ \end{aligned} $$
(4)
The firm-specific conditional accounting conservatism (C_score1i,t) is calculated by applying the estimates from Eq. (4) to Eq. (3). The higher value of C_score1i,t indicates more conditional accounting conservatism.
In addition, Banker et al. (2016) argue that cost stickiness biases the measurement of Khan and Watts (2009).8 Thus, Banker et al. (2016) modify the measure of Khan and Watts (2009) by considering the potential confounding effect of sticky costs as follows.
$$ NI_{i,t} = \beta_{1} + \beta_{2} D_{i,t} + \beta_{3} RET_{i,t} + \beta_{4} D_{i,t} \times RET_{i,t} + \beta_{5} S_{i,t} /MKT_{i,t - 1} + \beta_{6} DS_{i,t} + \beta_{7} S_{i,t} /MTK_{i,t - 1} \times DS_{i,t} + \varepsilon_{i,t} , $$
(5)
$$ G\_score\it{2}_{i,t} = \beta_{3} = \mu_{1} + \mu_{2} MV_{i,t - 1} + \mu_{3} BM_{i,t - 1} + \mu_{4} LEV_{i,t - 1} + \varepsilon_{i,t} , $$
(6)
$$ C\_score\it{2}_{i,t} = \beta_{4} = \lambda_{1} + \lambda_{2} MV_{i,t - 1} + \lambda_{3} BM_{i,t - 1} + \lambda_{4} LEV_{i,t - 1} + \varepsilon_{i,t} , $$
(7)
where, MVi,t−1 refers to the log of the market value of equity in the fiscal year t−1; BMi,t−1 refers to the book value of equity divided by the market value of equity in the fiscal year t−1; LEVi,t−1 refers to total debt divided by total assets in the fiscal year t−1; Si,t /MKTi,t−1 is the changes in sales divided by market value of equity in the fiscal year t−1; DSi,t equals one if Si,t is negative, and zero otherwise.
Next, replacing β3 and β4 from Eqs. (6) and (7) into regression Eq. (5) yields Eq. (8).
$$ \begin{aligned} NI_{i,t} & = \beta_{1} + \beta_{2} D_{i,t} + RET_{i,t} \times \left( {\mu_{1} + \mu_{2} MV_{i,t - 1} + \mu_{3} BM_{i,t - 1} + \mu_{4} LEV_{i,t - 1} } \right) \\ & \quad + D_{i,t} \times RET_{i,t} \times \left( {\lambda_{1} + \lambda_{2} MV_{i,t - 1} + \lambda_{3} BM_{i,t - 1} + \lambda_{4} LEV_{i,t - 1} } \right) \\ & \quad + \beta_{5} S_{i,t} /MKT_{i,t - 1} + \beta_{6} DS_{i,t} + \beta_{7} S_{i,t} /MKT_{i,t - 1} \times DS_{i,t} + \delta_{1} MV_{i,t - 1} + \delta_{2} BM_{i,t - 1} + \delta_{3} LEV_{i,t - 1} \\ & \quad + \delta_{4} D_{i,t} \times MV_{i,t - 1} + \delta_{5} D_{i,t} \times BM_{i,t - 1} + \delta_{6} D_{i,t} \times LEV_{i,t - 1} \\ & \quad + \delta_{7} S_{i,t} /MKT_{i,t - 1} \times MV_{i,t - 1} + \delta_{8} S_{i,t} /MKT_{i,t - 1} \times BM_{i,t - 1} + \delta_{9} S_{i,t} /MKT_{i,t - 1} \times LEV_{i,t - 1} \\ & \quad + \delta_{10} DS_{i,t} \times MV_{i,t - 1} + \delta_{11} DS_{i,t} \times BM_{i,t - 1} + \delta_{12} DS_{i,t} \times LEV_{i,t - 1} \\ & \quad + \delta_{13} S_{i,t} /MKT_{i,t - 1} \times DS_{i,t} \times MV_{i,t - 1} + \delta_{14} S_{i,t} /MKT_{i,t - 1} \times DS_{i,t} \times BM_{i,t - 1} \\ & \quad + \delta_{15} S_{i,t} /MKT_{i,t - 1} \times DS_{i,t} \times LEV_{i,t - 1} + \varepsilon_{i,t} . \\ \end{aligned} $$
(8)
The firm-specific conditional accounting conservatism (C_score2i,t) is measured by applying the estimates from Eq. (8) to Eq. (7). The higher value of C_score2i,t indicates more conditional accounting conservatism.

3.2.2 The measurement of overconfidence

The mainstream method of measuring managers’ overconfidence in the current literature is the option-based method proposed by Malmendier and Tate (2005). This method is based on the logic that managers’ long-term holding of options demonstrates a consistent unwillingness to limit their personal exposure to firm-specific risks, which shows their overconfidence. Our study, following the method of Malmendier and Tate (2005), regards CFOs or CEOs as overconfident managers when they are reluctant to exercise options that exceed 67% in the money. Malmendier and Tate (2005) use detailed information on each option package. As ExecuComp does not have detailed data on managers’ options holdings and exercise prices for each option grant before 2006, this study follows the method of Campbell et al. (2011) using the average moneyness of managers’ option portfolios as proxies of overconfidence. The average moneyness is calculated as follows.
$$ The\;realizable\;value\;per\;option_{i,t} = \frac{The\;total\;realizable\;value\;of\;the\;exercisable\;options_{i,t}\;}{{The\;number\;of\;exercisable\;options_{i,t}}} $$
(9)
$$ \begin{aligned} The{\text{ estimate of the average exercise }}price \, of \, options_{i,t} = \\ & The \, stock \, price \, at \, the \, fiscal \, year \, end_{i,t} - & & & & & & \, The \, realizable \, value \, per \, option_{i,t} \\ \end{aligned} $$
(10)
$$ \begin{gathered} The\;average\;percent\;moneyness\;of\;the\;options_{i,t} = \hfill \\ \frac{The\;\;realizable\;value\;per\;option_{i,t}}{{The\;estimate\;of\;the\;average\;exercise\;price\;of\;the\;options_{i,t}}}\;\;\;\;\;\; \hfill \\ \end{gathered} $$
(11)
Therefore, Holder67CFOi,t (Holder67CEOi,t) equals one from the first time that CFOs (CEOs) hold vested options that are at least 67% in the money to the end of their tenure, and zero otherwise (Chen et al. 2022).9

3.3 Baseline regression model design

We examine the relationship between CFO overconfidence on conditional accounting conservatism using the following model.
$$ \begin{gathered} C\_score_{i,t} = \beta_{0} + \beta_{1} Holder{\it{67}}CFO_{i,t} + \beta_{2} Holder{\it{67}}CEO_{i,t} + \beta_{3} CFO\_male_{i,t} + \beta_{4} CFO\_ownership_{i,t} + \beta_{5} MTB_{i,t} \hfill \\ + \beta_{6} SaleGrowth_{i,t} + \beta_{7}Volsale_{i,t} + \beta_{8} CashFlow_{i,t} + \beta_{9} FirmSize_{i,t} + \beta_{10} Leverage_{i,t} + \beta_{11} RDAD_{i,t} \hfill \\ + \beta_{12} Big\_four_{i,t} + \beta_{13} High\_LIT_{i,t} + \beta_{14} UnAC_{i,t} + Firm\;fixed\;effects + Year\;fixed\;effects + \varepsilon_{i,t} , \hfill \\ \end{gathered} $$
(12)
where, C_scorei,t is one of two conditional accounting conservatism proxies (C_score1i,t and C_score2i,t); Holder67CFOi,t, the proxy of overconfident CFOs, is the variable of interest; β1 captures the relationship between CFO overconfidence and conditional accounting conservatism. If the β1 is significantly positive, it indicates a positive association between CFO overconfidence and conditional accounting conservatism (hypothesis H1).
In the regression, we also include control variables based on previous conditional accounting conservatism studies. Specifically, Ahmed and Duellman (2013) find a negative relationship between CEO overconfidence and conditional accounting conservatism decisions. To control for the effect of CEO overconfidence, we include CEO overconfidence (Holder67CEOi,t). In addition, we control for some other CFOs’ characteristics. We control for CFO gender (CFO_malei,t) since male CFOs tend to use fewer conservative accounting principles than their female counterparts (Francis et al. 2015). In addition to CFO gender, we control for the influence of CFO ownership (CFO_ownershipi,t) as top managers’ ownership significantly affects conditional accounting conservatism (LaFond and Watts 2008). Given that firms are not static across time, we include firm performance in the regression. Prior studies suggest that firms’ growth opportunities, sale growth rates, operating uncertainty, and profitability affect conditional accounting conservatism (e.g., Smith and Watts 1992; Ahmed et al. 2002; Ahmed and Duellman 2007, 2013). As such, we use the market-to-book ratio (MTBi,t), sales growth (SaleGrowthi,t), sales volatility (VolSalei,t), and cash flows from operations (CashFlowi,t) to capture them, respectively. We also control for firm characteristics, including firm size (FirmSizei,t), leverage (Leveragei,t), R&D and advertising expenses (RDADi,t), following Ahmed et al. (2002), Givoly et al. (2007), and Ahmed and Duellman (2013). Considering the monitoring effect, we include auditor type and litigation risk because a high-quality auditor (Big_fouri,t) and high litigation risk (High_LITi,t) affect firms’ conditional accounting conservatism decisions (Basu et al. 2001; Watts 2003; Cano-Rodríguez 2010). In addition, Beaver and Ryan (2005) suggest that unconditional accounting conservatism should be controlled for in the study of conditional accounting conservatism as conditional accounting conservatism can be preempted by unconditional accounting conservatism. In response to their call, we control for unconditional accounting conservatism (UnACi,t). Finally, we use firm-fixed effect models and include year dummies to control for the potential omitted time-invariant effects of firms and years. Detailed variable measurements are provided in the Appendix.

4 Empirical results

4.1 Descriptive statistics

Table 2 displays descriptive statistics for all variables in the baseline regression. Panel A of Table 2 shows the descriptive statistics of the full sample. Following Ahmed and Duellman (2013), we use the conditional accounting conservatism measurement (C_score1i,t) proposed by Khan and Watts (2009) as our first conditional accounting conservatism measurement. The mean value of C_score1i,t is 0.062, which is similar to the result of Ahmed and Duellman (2013). Following Khalilov and Osma (2020), our second conditional accounting conservatism measurement (C_score2i,t) uses the approach of Banker et al. (2016). The mean value of C_score2i,t is 0.182, which is consistent with the findings of Khalilov and Osma (2020). Our overconfidence measure is consistent with Chen et al. (2022). Although our sample period differs from that used by Chen et al. (2022), we have comparable results: the mean values of overconfident CFOs and overconfident CEOs in their sample are 0.510 and 0.593, respectively, and the mean values of Holder67CFOi,t and Holder67CEOi,t are 0.548 and 0.650 in our sample. Over half of CFOs and CEOs are overconfident, which is in line with previous findings that overconfidence is a common trait among top managers (Goel and Thakor 2008; Malmendier et al. 2022), indicating that CFO overconfidence cannot be ignored.
Table 2
Descriptive statistics
 
N
Mean
SD
10P
25P
Median
75P
Panel A—Full sample
C_score1i,t
21,626
0.062
0.151
− 0.130
− 0.026
0.068
0.154
C_score2i,t
21,626
0.182
0.187
− 0.035
0.072
0.178
0.293
Holder67CFOi,t
21,626
0.548
0.498
0
0
1
1
Holder67CEOi,t
21,626
0.650
0.477
0
0
1
1
CFO_malei,t
21,626
0.921
0.269
1
1
1
1
CFO_ownershipi,t (%)
21,626
0.087
0.234
0.000
0.000
0.000
0.059
MTBi,t
21,626
3.447
3.709
1.051
1.559
2.394
3.840
SaleGrowthi,t
21,626
0.106
0.239
− 0.114
− 0.004
0.075
0.177
VolSalei,t
21,626
0.144
0.124
0.038
0.063
0.106
0.182
CashFlowi,t
21,626
0.100
0.085
0.014
0.059
0.101
0.147
FirmSizei,t
21,626
7.318
1.582
5.378
6.172
7.200
8.348
Leveragei,t
21,626
0.213
0.169
0.000
0.056
0.205
0.328
RDADi,t
21,626
0.067
0.126
0.000
0.000
0.022
0.080
Big_fouri,t
21,626
0.864
0.343
0
1
1
1
High_LITi,t
21,626
0.100
0.300
0
0
0
0
UnACi,t
21,626
0.014
0.051
− 0.034
− 0.010
0.009
0.031
Panel B—Subsample
Variable
Holder67CFOi,t = 0 (N = 9,775)
 
Holder67CFOi,t = 1 (N = 11,851)
 
T-statistics for tests of difference in means
  
 
Mean
 
Mean
 
(Non-overconfident CFOs-Overconfident CFOs)
  
C_score1i,t
0.079
 
0.047
 
0.032***
  
C_score2i,t
0.210
 
0.159
 
0.051***
  
Holder67CEOi,t
0.398
 
0.858
 
− 0.459***
  
CFO_malei,t
0.922
 
0.920
 
0.002
  
CFO_ownershipi,t
0.062
 
0.107
 
− 0.045***
  
MTBi,t
2.829
 
3.957
 
− 1.129***
  
SaleGrowthi,t
0.062
 
0.143
 
− 0.081***
  
VolSalei,t
0.143
 
0.145
 
− 0.002
  
CashFlowi,t
0.087
 
0.111
 
− 0.025***
  
FirmSizei,t
7.351
 
7.290
 
0.061**
  
Leveragei,t
0.225
 
0.203
 
0.021***
  
RDADi,t
0.069
 
0.066
 
0.004
  
Big_fouri,t
0.860
 
0.867
 
− 0.007
  
High_LITi,t
0.109
 
0.092
 
0.017***
  
UnACi,t
0.018
 
0.011
 
0.007***
  
Notes: Panel A of Table 2 shows the descriptive statistics of the full sample. Panel B of Table 2 shows the descriptive statistics of the subsample sample. T-tests are conducted to test for differences in the means between the overconfident CFOs sample and the non-overconfident CFOs sample. Detailed variable information is shown in Appendix. *, **, and *** represent significance levels of 0.10, 0.05, and 0.01, respectively. Variables of interest are marked in bold
Panel B of Table 2 shows the descriptive statistics of the subsample. T-tests are conducted to test for differences in independent and control variables between the overconfident CFO sample and the non-overconfident CFO sample. Univariate comparisons show that the mean values of C_score1i,t and C_score2i,t in the overconfident CFOs sample are significantly lower than the corresponding values for their non-overconfident CFOs sample.

4.2 Pairwise correlations

Table 3 shows the Pearson correlation coefficients. The correlations between the independent and control variables are less than 0.5, and the untabulated values of variance-inflating factors (VIFs) are lower than the cutoff of 10 (Gujarati et al. 2012), indicating that multicollinearity is not a problem when analyzing the regression results. Besides, Holder67CFOi,t and Holder67CEOi,t have a significant association (a correlation coefficient of 0.479), indicating that both should be included in one regression to avoid the omitted variables problem. Holder67CFOi,t has significant negative correlations with C_score1i,t and C_score2i,t without controlling other variables, which is in line with our hypothesis H1 that overconfident CFOs tend to reduce conditional accounting conservatism compared with non-overconfident CFOs.
Table 3
Correlations
 
Variable
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(1)
C_score1i,t
1
       
(2)
C_score2i,t
0.748***
1
      
(3)
Holder67CFOi,t
− 0.107***
− 0.136***
1
     
(4)
Holder67CEOi,t
− 0.078***
− 0.117***
0.479***
1
    
(5)
CFO_malei,t
0.021***
0.033***
− 0.004
− 0.030***
1
   
(6)
CFO_ownershipi,t
0.184***
0.131***
0.095***
0.053***
0.031***
1
  
(7)
MTBi,t
− 0.325***
− 0.356***
0.151***
0.136***
− 0.014**
− 0.034***
1
 
(8)
SaleGrowthi,t
− 0.074***
− 0.144***
0.168***
0.159***
0.008
− 0.018***
0.125***
1
(9)
VolSalei,t
0.147***
0.122***
0.009
0.021***
0.003
0.035***
0.007
0.065***
(10)
CashFlowi,t
− 0.273***
− 0.222***
0.145***
0.109***
− 0.041***
− 0.013*
0.201***
0.056***
(11)
FirmSizei,t
− 0.587***
− 0.469***
− 0.019***
− 0.041***
− 0.022***
− 0.186***
0.051***
− 0.061***
(12)
Leveragei,t
− 0.011
− 0.080***
− 0.063***
− 0.064***
0.050***
− 0.040***
0.107***
− 0.035***
(13)
RDADi,t
0.025***
− 0.013*
− 0.015**
0.025***
− 0.006
− 0.029***
0.163***
0.049***
(14)
Big_fouri,t
− 0.096***
− 0.082***
0.009
0.001
− 0.032***
− 0.066***
0.022***
− 0.049***
(15)
High_LITi,t
− 0.015**
0.047***
− 0.027***
− 0.015**
− 0.004
− 0.048***
− 0.054***
− 0.061***
(16)
UnACi,t
0.074***
0.057***
− 0.067***
− 0.049***
− 0.017**
0.008
0.089***
− 0.005
 
Variable
(9)
(10)
(11)
(12)
(13)
(14)
(15)
(16)
(9)
VolSalei,t
1
       
(10)
CashFlowi,t
− 0.088***
1
      
(11)
FirmSizei,t
− 0.175***
0.123***
1
     
(12)
Leveragei,t
− 0.041***
− 0.153***
0.355***
1
    
(13)
RDADi,t
− 0.037***
− 0.318***
− 0.214***
− 0.196***
1
   
(14)
Big_fouri,t
− 0.036***
0.040***
0.240***
0.070***
− 0.014**
1
  
(15)
High_LITi,t
0.086***
− 0.091***
0.184***
0.099***
0.036***
0.052***
1
 
(16)
UnACi,t
0.048***
0.003
− 0.064***
− 0.055***
0.186***
0.053***
0.042***
1
Table 3 shows the Pearson correlation coefficients. Detailed variable information is shown in Appendix. *, **, and *** represent significance levels of 0.10, 0.05, and 0.01, respectively. Variables of interest are marked in bold

4.3 Baseline resultsregression analyses of Hypothesis 1

In this section, we use fixed-effect models to test the relationship between CFO overconfidence and conditional accounting conservatism.10 Table 4 shows the regression results of the estimation of Eq. (12). The dependent variables are C_score1i,t measured using the method of Khan and Watts (2009) (columns (1) and (3)) and C_score2i,t measured using the approach of Banker et al. (2016) (columns (2) and (4)). The variable of interest is CFO overconfidence (Holder67CFOi,t) calculated following Campbell et al. (2011) and Chen et al. (2022).
Table 4
The impact of CFO overconfidence on conditional accounting conservatism
 
(1)
(2)
(3)
(4)
Variable
C_score1i,t
C_score2i,t
C_score1i,t
C_score2i,t
Holder67CFOi,t
− 0.016***
− 0.025***
− 0.012***
− 0.018***
 
(0.002)
(0.002)
(0.002)
(0.003)
Holder67CEOi,t
  
− 0.011***
− 0.019***
   
(0.002)
(0.003)
CFO_malei,t
0.006
0.008
0.006
0.008
 
(0.004)
(0.006)
(0.004)
(0.006)
CFO_ownershipi,t
0.016***
0.031***
0.016***
0.032***
 
(0.005)
(0.008)
(0.005)
(0.008)
MTBi,t
− 0.010***
− 0.008***
− 0.009***
− 0.008***
 
(0.001)
(0.000)
(0.001)
(0.000)
SaleGrowthi,t
− 0.014***
− 0.024***
− 0.013***
− 0.023***
 
(0.003)
(0.004)
(0.003)
(0.004)
VolSalei,t
− 0.014*
− 0.048***
− 0.013*
− 0.045***
 
(0.008)
(0.012)
(0.008)
(0.012)
CashFlowi,t
− 0.099***
− 0.119***
− 0.097***
− 0.115***
 
(0.012)
(0.017)
(0.012)
(0.017)
FirmSizei,t
− 0.070***
− 0.060***
− 0.069***
− 0.058***
 
(0.002)
(0.003)
(0.002)
(0.003)
Leveragei,t
0.202***
0.090***
0.200***
0.086***
 
(0.009)
(0.012)
(0.009)
(0.012)
RDADi,t
− 0.048***
− 0.088***
− 0.047***
− 0.086***
 
(0.018)
(0.030)
(0.018)
(0.030)
Big_fouri,t
− 0.014***
− 0.018***
− 0.014***
− 0.018***
 
(0.003)
(0.005)
(0.003)
(0.005)
High_LITi,t
0.020***
0.014***
0.020***
0.014***
 
(0.003)
(0.004)
(0.003)
(0.004)
UnACi,t
0.074***
0.110***
0.071***
0.104***
 
(0.016)
(0.022)
(0.016)
(0.022)
Constant
0.424***
0.517***
0.422***
0.514***
 
(0.015)
(0.028)
(0.015)
(0.028)
Observations
21,626
21,626
21,626
21,626
Firm fixed effects
Yes
Yes
Yes
Yes
Year fixed effects
Yes
Yes
Yes
Yes
Adj. R2
0.797
0.751
0.798
0.752
The results of the impact of CFO overconfidence on conditional accounting conservatism are shown in Table 4. The standard errors clustering at the firm level are displayed in parentheses. Detailed variable information is shown in Appendix. *, **, and *** represent significance levels of 0.10, 0.05, and 0.01, respectively. Variables of interest are marked in bold
The coefficient of CFO overconfidence (Holder67CFOi,t) is significant and negative in column (1), suggesting a negative relationship between CFO overconfidence and conditional accounting conservatism. Given that the mean value of conditional accounting conservatism (C_score1i,t) is 0.062, there appears to be a sizeable difference (25.806%).11 As shown in column (2), when we change the dependent variable from C_score1i,t to C_score2i,t, Holder67CFOi,t remains a significantly negative coefficient, indicating that the finding is robust under an alternative conditional accounting conservatism measure.
Given that CFO overconfidence (Holder67CFOi,t) and CEO overconfidence (Holder67CEOi,t) have a high correlation, and in response to the call made by Black and Gallemore (2013) and Malmendier et al. (2022) that CFO overconfidence and CEO overconfidence should be jointly considered in corporate decision-making, we control for the effect of Holder67CEOi,t in columns (3) and (4). Holder67CEOi,t has a significantly negative coefficient in columns (3) and (4), which is consistent with the finding of Ahmed and Duellman (2013). The coefficient on Holder67CFOi,t remains negative and statistically significant in columns (3) and (4). Besides, the coefficients on Holder67CFOi,t suggest that firms with overconfident CFOs cause a decrease in C_score1i,t by 19.355% (in comparison to the mean C_score1i,t) in column (3) and a decrease in C_score2i,t by 9.890% (in comparison to the mean C_score2i,t) in column (4), showing that these findings are also economically significant.12
In terms of the control variables, the market-to-book ratio (MTBi,t) has a negative and significant coefficient in columns (1) to (4), which is consistent with the findings of Roychowdhury and Watts (2007) that firms with high growth opportunities adopt low conditional accounting conservatism. CashFlowi,t has a negative and significant coefficient in each column, showing that high-profit firms reduce the application of conditional accounting conservatism (Ahmed and Duellman 2013). The coefficient on firm size (FirmSizei,t) is negative and significant in each column, suggesting that larger firms are less likely to recognize losses timely (Givoly et al. 2007). Leveragei,t has a positive and significant coefficient in columns (1) to (4), indicating that high-leverage firms increase their accounting conservatism. This finding is in line with Ahmed et al. (2002) in that high-leverage firms have more debt-holder and equity-holder conflicts, which increase conservatism levels. The coefficient of high litigation risk (High_LITi,t) is significantly positive in columns (1) to (4), indicating that firms under high litigation risk adopt more conditional accounting conservatism (Watts 2003).
Overall, we find a statistically and economically significant association between CFO overconfidence and conditional accounting conservatism, supporting Hypothesis 1.

4.4 Robustness tests

So far, our baseline regression results show that there is a negative association between CFO overconfidence and conditional accounting conservatism. In this section, we will conduct a series of tests to mitigate the endogeneity issues in our main findings.

4.4.1 Reverse causality concern

Someone might be concerned that overconfident CFOs might self-select into firms with less conservative accounting rather than have the incentive to reduce conditional accounting conservatism. To address this concern, we lag all the independent variables and control variables in Eq. (12) by one year relative to the conditional accounting conservatism measures. The results are shown in Table 5. Holder67CFOi,t has a significantly negative coefficient in columns (1) and (2), proving that overconfident CFOs are able to reduce future conditional accounting conservatism.
Table 5
The impact of CFO overconfidence on future conditional accounting conservatism
 
(1)
(2)
Variable
C_score1i,t+1
C_score2i,t+1
Holder67CFOi,t
− 0.010***
− 0.021***
 
(0.002)
(0.003)
Constant
0.359***
0.565***
 
(0.020)
(0.028)
Observations
17,074
17,074
Controls in Eq. (12)
Yes
Yes
Firm fixed effects
Yes
Yes
Year fixed effects
Yes
Yes
Adj. R2
0.779
0.785
The results of the impact of CFO overconfidence on future conditional accounting conservatism are shown in Table 5. The standard errors clustering at the firm level are displayed in parentheses. Detailed variable information is shown in Appendix. *, **, and *** represent significance levels of 0.10, 0.05, and 0.01, respectively. Variables of interest are marked in bold
To further mitigate the reverse causality concern, this study uses the PSM-DID method. Following Ahmed and Duellman (2013), we use CFO turnover as an exogenous shock and require that the outgoing CFO was in office for no less than four years, and that the incoming CFO remains in office for no less than three years to satisfy that CFOs have sufficient time to influence firms’ decisions. To eliminate the potential influence of CEOs, we remove the simultaneous CEO and CFO turnovers. Our study matches treatment and control groups based on the control variable included in Eq. (12) and uses the one-to-one nearest-neighbor method for which the caliper does not exceed 5% to find the matched sample. The Treati equals one if both outgoing CFO is non-overconfident and the incoming CFO is overconfident, and zero if both the outgoing CFO and incoming CFO are non-overconfident. The Posti,t equals one when observations occur in the first year after CFO turnover, and zero when observations occur in the last year before CFO turnover. The DID estimation model is as follows:
$$ C\_score_{i,t} = \beta_{0} + \beta_{1} Treat_{i} \times Post_{i,t} + Controls + Firm\;fixed\;effects + Year\;fixed\;effects + \varepsilon_{i,t} , $$
(13)
where, the interaction term, Treati × Posti,t, is our variable of interest. We include firm fixed effects and year fixed effects, so we exclude Treati and Posti,t to avoid multiple collinearities. Control variables are consistent with Eq. (12). Detailed variable information is shown in the Appendix.
Panel A of Table 6 shows the results of one-to-one nearest-neighbor PSM. Panel B of Table 6 describes the result of the DID estimation. In the DID test, we use a fixed-effect model to compare the impacts of firms’ changing their CFOs from non-overconfident to overconfident and firms’ changing their CFOs from non-overconfident to non-overconfident on conditional accounting conservatism decisions. As shown in Panel A of Table 6, the P values of control variables in the matched sample are larger than 0.1, indicating that these control variables are not significantly different between the treatment and control groups in the matched samples. The coefficient on Treati × Posti,t is significantly negative in Panel B of Table 6, suggesting that the negative relationship between CFO overconfidence and conditional accounting conservatism remains under PSM-DID estimation.
Table 6
PSM-DID
Panel A—PSM
Variable
Unmatched
Mean
Bias
T-test
Matched
Treated
Control
%bias
%reduct |bias|
t
p >|t|
(1)
(2)
(3)
(4)
(5)
(6)
(7)
Holder67CEOi,t
U
0.610
0.326
59.100
 
4.820
0.000
 
M
0.610
0.598
2.500
95.700
0.160
0.874
CFO_malei,t
U
0.890
0.909
− 6.100
 
− 0.510
0.612
 
M
0.890
0.890
0.000
100.000
0.000
1.000
CFO_ownershipi,t
U
0.056
0.032
22.600
 
2.110
0.036
 
M
0.056
0.042
13.300
41.100
0.930
0.354
MTBi,t
U
2.717
2.449
16.800
 
1.250
0.213
 
M
2.717
2.789
− 4.500
73.300
− 0.220
0.824
SaleGrowthi,t
U
0.107
0.035
31.300
 
2.210
0.028
 
M
0.107
0.108
− 0.400
98.800
− 0.020
0.984
VolSalei,t
U
0.134
0.136
− 2.000
 
− 0.180
0.860
 
M
0.134
0.149
− 11.300
− 477.200
− 0.740
0.461
CashFlowi,t
U
0.099
0.083
22.300
 
1.580
0.114
 
M
0.099
0.102
− 3.100
85.900
− 0.250
0.803
FirmSizei,t
U
7.679
7.795
− 7.700
 
− 0.620
0.538
 
M
7.679
7.689
− 0.700
91.100
− 0.050
0.962
Leveragei,t
U
0.208
0.231
− 15.500
 
− 1.260
0.209
 
M
0.208
0.202
3.600
76.800
0.240
0.814
RDADi,t
U
0.060
0.089
− 11.900
 
− 0.780
0.437
 
M
0.060
0.049
4.200
64.800
0.760
0.449
Big_fouri,t
U
0.927
0.919
3.100
 
0.240
0.807
 
M
0.927
0.963
− 13.600
− 342.800
− 1.030
0.307
High_LITi,t
U
0.037
0.085
− 20.200
 
− 1.470
0.141
 
M
0.037
0.049
− 5.100
74.600
− 0.380
0.701
UnACi,t
U
0.011
0.016
− 12.500
 
− 0.960
0.338
 
M
0.011
0.005
12.300
2.000
0.950
0.342
Panel B—DID
 
    
(1)
 
(2)
 
Variable
   
C_score1i,t
 
C_score2i,t
 
Treati × Posti,t
   
− 0.022*
 
− 0.025**
 
    
(0.012)
 
(0.010)
 
Constant
   
0.919***
 
0.826***
 
    
(0.174)
 
(0.139)
 
Observations
   
120
 
120
 
Controls in Eq. (13)
   
Yes
 
Yes
 
Firm fixed effects
   
Yes
 
Yes
 
Year fixed effects
   
Yes
 
Yes
 
Adj. R2
   
0.922
 
0.949
 
Panel A of Table 6 reports the results of PSM. Panel B of Table 6 shows the results of DID. The standard errors clustering at the firm level are displayed in parentheses. Detailed variable information is shown in Appendix. *, **, and *** represent significance levels of 0.10, 0.05, and 0.01, respectively. Variables of interest are marked in bold

4.4.2 Measurement error

Our first overconfidence measure uses 67% as the threshold value. However, some studies adopt other threshold values to measure overconfidence. Thus, we repeat our analysis using an alternative overconfidence proxy. Following Campbell et al. (2011) and Hirshleifer et al. (2012), we change the threshold value of option moneyness such that Holder100CFOi,t (Holder100CEOi,t) equals one from the first time that CFOs (CEOs) hold vested options that are at least 100% in the money to the end of their tenure, and zero otherwise. In addition, to rule out the concern that CFOs and CEOs exercise their options late because they are risk-tolerant rather than overconfident, our study, following Huang et al. (2016), controls for CFO and CEO risk tolerance (CFO_vegei,t and CEO_vegai,t). As shown in Table 7, the Holder100CFOi,t has a significant negative coefficient, suggesting that the negative relationship still holds when using the new proxy of overconfidence and controlling for CFOs’ and CEOs’ risk tolerance.13
Table 7
The impact of CFO overconfidence on conditional accounting conservatism-alternative measurement of overconfidence
 
(1)
(2)
(3)
(4)
Variable
C_score1i,t
C_score2i,t
C_score1i,t
C_score2i,t
Holder100CFOi,t
− 0.010***
− 0.019***
− 0.011***
− 0.021***
 
(0.002)
(0.003)
(0.002)
(0.003)
Holder100CEOi,t
− 0.011***
− 0.017***
− 0.010***
− 0.018***
 
(0.002)
(0.003)
(0.002)
(0.003)
CFO_vegai,t
  
0.001
− 0.003**
   
(0.001)
(0.001)
CEO_vegai,t
  
− 0.001
− 0.003**
   
(0.001)
(0.001)
Constant
0.416***
0.510***
0.490***
0.595***
 
(0.015)
(0.028)
(0.019)
(0.033)
Observations
21,388
21,388
18,601
18,601
Other controls in Eq. (12)
Yes
Yes
Yes
Yes
Firm fixed effects
Yes
Yes
Yes
Yes
Year fixed effects
Yes
Yes
Yes
Yes
Adj. R2
0.797
0.752
0.799
0.760
The results of using an alternative overconfidence measure are shown in Table 7. The standard errors clustering at the firm level are displayed in parentheses. Detailed variable information is shown in Appendix. *, **, and *** represent significance levels of 0.10, 0.05, and 0.01, respectively. Variables of interest are marked in bold

4.4.3 The effect of CEO characteristics

Some studies suggest that CFO overconfidence might not independently affect corporate decisions because CFOs employ their decision-making authority under the CEOs’ explicit or implicit policy direction. For example, Hsieh et al. (2018) and Chyz et al. (2019) find the interactive impact of CEO overconfidence and CFO overconfidence on tax-avoidance decisions, but overconfident CFOs do not have an independent effect on tax avoidance. However, a few studies argue that CFO overconfidence outweighs CEO overconfidence in affecting financing method choices (Malmendier et al. 2022), earnings management (Qiao et al. 2023), and stock price crashes (Qiao et al. 2022).
To examine if there is a joint effect of CFO overconfidence and CEO overconfidence or an independent effect of CFO overconfidence on conditional accounting conservatism, we add the interaction term, Holder67CFOi,t × Holder67CEOi,t, into Eq. (12). As shown in Table 8, the coefficient on Holder67CFOi,t × Holder67CEOi,t is not significant, suggesting that we do not find a statistically significant joint effect of CFO and CEO overconfidence on conditional accounting conservatism. The coefficients on Holder67CFOi,t and Holder67CFOi,t + Holder67CFOi,t × Holder67CEOi,t are significantly negative, indicating that overconfident CFOs have an independent effect on conditional accounting conservatism no matter whether CEOs are overconfident or not. These findings are held after controlling for CEO gender (CEO_malei,t) and CEO ownership (CEO_ownershipi,t), which are shown in columns (3) and (4).
Table 8
The impact of CEO overconfidence on the relationship between CFO overconfidence and conditional accounting conservatism
 
(1)
(2)
(3)
(4)
Variable
C_score1i,t
C_score2i,t
C_score1i,t
C_score2i,t
Holder67CFOi,t
− 0.015***
− 0.024***
− 0.015***
− 0.024***
 
(0.003)
(0.005)
(0.003)
(0.005)
Holder67CEOi,t
− 0.011***
− 0.021***
− 0.012***
− 0.021***
 
(0.003)
(0.004)
(0.003)
(0.004)
Holder67CFOi,t × Holder67CEOi,t
0.002
0.005
0.002
0.005
 
(0.004)
(0.005)
(0.004)
(0.005)
CFO_vegai,t
0.001
− 0.003**
0.001
− 0.003**
 
(0.001)
(0.001)
(0.001)
(0.001)
CEO_vegai,t
− 0.001
− 0.003**
− 0.001
− 0.003**
 
(0.001)
(0.001)
(0.001)
(0.001)
CEO_malei,t
  
0.013
0.004
   
(0.008)
(0.011)
CEO_ownershipi,t
  
0.000
0.000
   
(0.000)
(0.001)
Constant
0.496***
0.599***
0.482***
0.594***
 
(0.019)
(0.033)
(0.021)
(0.034)
Holder67CFOi,t + Holder67CFOi,t × Holder67CEOi,t
− 0.013***
− 0.019***
− 0.012***
− 0.019***
 
(0.002)
(0.004)
(0.002)
(0.004)
Observations
18,824
18,824
18,824
18,824
Other controls in Eq. (12)
Yes
Yes
Yes
Yes
Firm fixed effects
Yes
Yes
Yes
Yes
Year fixed effects
Yes
Yes
Yes
Yes
Adj. R2
0.800
0.761
0.800
0.761
The results of the impact of CEO overconfidence on the relationship between CFO overconfidence and conditional accounting conservatism are shown in Table 8. The standard errors clustering at the firm level are displayed in parentheses. Detailed variable information is shown in Appendix. *, **, and *** represent significance levels of 0.10, 0.05, and 0.01, respectively. Variables of interest are marked in bold

4.4.4 The effect of governance

Someone may argue that the strength of governance might mitigate our documented findings on the negative relationship between CFO overconfidence and conditional accounting conservatism as the board of directors and investors might restrict overconfident CFOs’ decisions (Qiao et al. 2022). We examine whether the association between CFO overconfidence and conditional accounting conservatism varies with the strength of governance.
We follow Ahmed and Duellman (2013) to measure strong governance. Firms have strong governance (StrongCGi,t) if they meet three of the four criteria listed below: First, the CEO is not simultaneously the Chairman. Second, the percentage of outside directors is greater than the sample’s median figure. Third, the institutional ownership percentage is higher than the sample’s median figure as institutional investors significantly affect conditional accounting conservatism (Lin 2016). Fourth, the percentage of inside directors is lower than the median value of the sample. As shown in columns (1) and (2) of Table 9, we additionally control the StrongCGi,t in Eq. (12). Holder67CFOi,t remains a significantly negative coefficient. Besides, we add the interaction term, Holder67CFOi,t × StrongCGi,t, into Eq. (12). As shown in columns (3) and (4), the coefficient on the interaction term is not significant. However, the coefficients on Holder67CFOi,t and Holder67CFOi,t + Holder67CFOi,t × StrongCGi,t are significantly negative, suggesting that strong governance cannot moderate the association between CFO overconfidence and conditional accounting conservatism. These findings are consistent with the findings of Schrand and Zechman (2012) and Ahmed and Duellman (2013) that governance mechanisms do not mitigate the effect of overconfident managers.
Table 9
The impact of strong governance on the relationship between CFO overconfidence and conditional accounting conservatism
 
(1)
(2)
(3)
(4)
Variable
C_score1i,t
C_score2i,t
C_score1i,t
C_score2i,t
Holder67CFOi,t
− 0.016***
− 0.025***
− 0.016***
− 0.025***
 
(0.003)
(0.004)
(0.003)
(0.004)
Variable
C_score1i,t
C_score2i,t
C_score1i,t
C_score2i,t
StrongCGi,t
0.002
0.007
0.002
0.008
 
(0.003)
(0.004)
(0.004)
(0.006)
Holder67CFOi,t × StrongCGi,t
  
0.000
− 0.003
   
(0.004)
(0.007)
CFO_vegai,t
0.001
− 0.002
0.001
− 0.002
 
(0.001)
(0.002)
(0.001)
(0.002)
CEO_vegai,t
− 0.000
− 0.001
− 0.000
− 0.001
 
(0.001)
(0.001)
(0.001)
(0.001)
CEO_malei,t
0.009
0.012
0.009
0.012
 
(0.011)
(0.017)
(0.011)
(0.017)
CEO_ownershipi,t
0.000
0.000
0.000
0.000
 
(0.001)
(0.001)
(0.001)
(0.001)
Constant
0.577***
0.509***
0.577***
0.508***
 
(0.028)
(0.053)
(0.028)
(0.053)
Holder67CFOi,t + Holder67CFOi,t × StrongCGi,t
  
− 0.016***
− 0.028***
   
(0.005)
(0.007)
Observations
11,361
11,361
11,361
11,361
Other controls in Eq. (12)
Yes
Yes
Yes
Yes
Firm fixed effects
Yes
Yes
Yes
Yes
Year fixed effects
Yes
Yes
Yes
Yes
Adj. R2
0.807
0.769
0.807
0.769
The results of the impact of strong governance on the relationship between CFO overconfidence and conditional accounting conservatism are shown in Table 9. The standard errors clustering at the firm level are displayed in parentheses. Detailed variable information is shown in Appendix. *, **, and *** represent significance levels of 0.10, 0.05, and 0.01, respectively. Variables of interest are marked in bold

5 Analysis of mechanisms

5.1 CFO overconfidence, conditional accounting conservatism, and conveying information

We hypothesize that overconfident CFOs tend to communicate private information by reducing conditional accounting conservatism because they perceive the market underestimates their firms’ performance. Then, our baseline results confirm a robust negative relationship between overconfident CFOs and conditional accounting conservatism. In this section, we further test whether overconfident CFOs can convey private information by reducing conditional accounting conservatism.
First, following prior studies (e.g., Lundholm and Myers 2002), we determine that overconfident CFOs’ convey their private information to investors by reducing conditional accounting conservatism if the current stock prices are highly related to future earnings. We use the future earnings response coefficient (FERC) to measure the informativeness of stock prices about future earnings (Shu 2021). To test the impact of accounting conservatism on FERC, we extend the basic FERC measurement as follows:
$$ \begin{aligned} R_{i,t} & = \beta_{0} + \beta_{1} Earn_{i,t - 1} + \beta_{2} Earn_{i,t} + \beta_{3} Earn_{i,t + 3} + \beta_{4} R_{i,t + 3} + \beta_{5} Holder\it{67}CFO_{i,t} \\ & \quad + \beta_{6} Holder\it{67}CFO_{i,t} \times Earn_{i,t - 1} + \beta_{7} Holder\it{67}CFO_{i,t} \times Earn_{i,t} + \beta_{8} Holder\it{67}CFO_{i,t} \times Earn_{i,t + 3} \\ & \quad + \beta_{9} Holder\it{67}CFO_{i,t} \times R_{i,t + 3} + \beta_{10} RevC\_score_{i,t} + \beta_{11} RevC\_score_{i,t} \times Earn_{i,t - 1} \\ & \quad + \beta_{12} RevC\_score_{i,t} \times Earn_{i,t} + \beta_{13} RevC\_score_{i,t} \times Earn_{i,t + 3} + \beta_{14} RevC\_score_{i,t} \times R_{i,t + 3} \\ & \quad + \beta_{15} Holder\it{67}CFO_{i,t} \times RevC\_score_{i,t} + \beta_{16} Holder\it{67}CFO_{i,t} \times RevC\_score_{i,t} \times Earn_{i,t - 1} \\ & \quad + \beta_{17} Holder\it{67}CFO_{i,t} \times RevC\_score_{i,t} \times Earn_{i,t} + \beta_{18} Holder\it{67}CFO_{i,t} \times RevC\_score_{i,t} \times Earn_{i,t + 3} \\ & \quad + \beta_{19} Holder\it{67}CFO_{i,t} \times RevC\_score_{i,t} \times R_{i,t + 3} + \beta_{20} Holder\it{67}CEO_{i,t} + \beta_{21} CFO\_male_{i,t} \\ & \quad + \beta_{22} CFO\_ownership_{i,t} + \beta_{23} CEO\_male_{i,t} + \beta_{24} CEO\_ownership_{i,t} + \beta_{25} MTB_{i,t} + \beta_{26} SaleGrowth_{i,t} \\ & \quad + \beta_{27} Volsale_{i,t} + \beta_{28} FirmSize_{i,t} + \beta_{29} FirmAge_{i,t} + \beta_{30} Leverage_{i,t} + \beta_{31} Invest_{i,t} + \beta_{32} PPE_{i,t} \\ & \quad + \beta_{33} Loss_{i,t} + \beta_{34} Analyst\_number_{i,t} + \beta_{35} CFO\_vega_{i,t} + \beta_{36} CEO\_vega_{i,t} + \beta_{37} StrongCG_{i,t} \\ & \quad + Firm\;fixed\;effects + Year\;fixed\;effects + \varepsilon_{i,t} , \\ \end{aligned} $$
(14)
where, the dependent variable is the stock return (Ri,t). We include last year’s earnings per share (Earni,t-1), earnings per share (Earni,t), the sum of earnings per share for fiscal years t + 1 to t + 3 (Earni,t+3), the sum of stock returns for fiscal years t + 1 to t + 3 (Ri,t+3) and CFO overconfidence (Hold67CFOi,t) in the regressions. To easily interpret our results, we multiply C_scorei,t (including C_score1i,t and C_score2i,t) by a negative one to generate RevC_scorei,t (including RevC_score1i,t and RevC_score2i,t). The greater the value of RevC_scorei,t the lower the conditional accounting conservatism. We also include interaction terms among them. The variable of interest is the triple interaction term, Hold67CFOi,t × RevC_scorei,t × Earni,t+3. In addition, following prior studies (e.g., Shu 2021) control CEO overconfidence (Hold67CEOi,t), CFO gender (CFO_malei,t), CFO ownership (CFO_ownershipi,t), CEO gender (CEO_malei,t), CEO ownership (CEO_ownershipi,t), Market to book ratio (MTBi,t), Sales growth (SaleGrowthi,t), Sales volatility (VolSalei,t), Firm Size (FirmSizei,t), Firm age (FirmAgei,t), Leverage (Leveragei,t), Investment (Investi,t), Property, Plant & Equipment (PPEi,t), Negative earnings per share (Lossi,t), Number of the analyst (Analyst_numberi,t), CFO risk tolerance (CFO_vegai,t), CEO risk tolerance (CEO_vegai,t), and Strong governance (StrongCGi,t). Detailed variable measurements are provided in the Appendix.
As shown in Table 10, the coefficients on the Holder67CFOi,t × RevC_score1i,t × Earni,t+3 and Holder67CFOi,t × RevC_score2i,t × Earni,t+3 are significantly positive in columns (1) and (2), respectively. These findings show that overconfident CFOs who reduce conditional accounting conservatism increase earnings informativeness, which is consistent with our prediction that overconfident CFOs tend to convey information by reducing conditional accounting conservatism.
Table 10
The impact of overconfident CFOs’ conditional accounting conservatism decision on earnings informativeness
 
(1)
(2)
Variable
Ri,t
Ri,t
Earni,t−1
− 0.170
− 0.074
 
(0.223)
(0.335)
Earni,t
− 0.319
0.535**
 
(0.222)
(0.233)
Earni,t+3
0.551***
0.674***
 
(0.094)
(0.115)
Ri,t+3
− 0.154***
− 0.150***
 
(0.020)
(0.028)
Holder67CFOi,t
− 0.060**
− 0.025
 
(0.030)
(0.043)
Holder67CFOi,t × Earni,t−1
− 0.023
− 0.502
 
(0.390)
(0.488)
Holder67CFOi,t × Earni,t
0.650**
0.598
 
(0.299)
(0.456)
Holder67CFOi,t × Earni,t+3
0.169
0.238*
 
(0.118)
(0.142)
Holder67CFOi,t × Ri,t+3
0.031
0.005
 
(0.023)
(0.031)
RevC_score1i,t
0.727***
 
 
(0.144)
 
RevC_score1i,t × Earni,t−1
1.625*
 
 
(0.886)
 
RevC_score1i,t × Earni,t
− 1.226
 
 
(0.754)
 
RevC_score1i,t × Earni,t+3
0.116
 
 
(0.406)
 
RevC_score1i,t × Ri,t+3
− 0.028
 
 
(0.100)
 
Holder67CFOi,t × RevC_score1i,t
− 0.360***
 
 
(0.135)
 
Holder67CFOi,t × RevC_score1i,t × Earni,t−1
− 0.499
 
 
(1.880)
 
Holder67CFOi,t × RevC_score1i,t × Earni,t
− 1.096
 
 
(1.645)
 
Holder67CFOi,t × RevC_score1i,t × Earni,t+3
1.106*
 
 
(0.623)
 
Holder67CFOi,t × RevC_score1i,t × Fi,t+3
0.107
 
 
(0.139)
 
RevC_score2i,t
 
− 1.114***
  
(0.123)
RevC_score2i,t × Earni,t−1
 
0.448
  
(0.929)
RevC_score2i,t × Earni,t
 
1.529*
  
(0.780)
RevC_score2i,t × Earni,t+3
 
0.348
  
(0.379)
RevC_score2i,t × Ri,t+3
 
0.037
  
(0.128)
Holder67CFOi,t × RevC_score2i,t
 
− 0.248*
  
(0.135)
Holder67CFOi,t × RevC_score2i,t × Earni,t−1
 
− 1.039
  
(1.313)
Holder67CFOi,t × RevC_score2i,t × Earni,t
 
− 0.498
  
(1.397)
Holder67CFOi,t × RevC_score2i,t × Earni,t+3
 
0.854*
  
(0.496)
Holder67CFOi,t × RevC_score2i,t × Ri,t+3
 
0.013
  
(0.146)
Constant
1.551***
1.181***
 
(0.272)
(0.274)
Observations
5,264
5,264
Controls in Eq. (14)
Yes
Yes
Firm fixed effects
Yes
Yes
Year fixed effects
Yes
Yes
Adj. R2
0.340
0.378
The results of the impact of overconfident CFOs’ conditional accounting conservatism decision on earnings informativeness are shown in Table 10. The standard errors clustering at the firm level are displayed in parentheses. Detailed variable information is shown in Appendix. *, **, and *** represent significance levels of 0.10, 0.05, and 0.01, respectively. Variables of interest are marked in bold
In addition, we predict that the ultimate purpose of overconfident CFOs conveying information via reducing conditional accounting conservatism is to improve financial flexibility as they are the principal executive in charge of financing (Malmendier et al. 2022). This notion will be approved if we can find that overconfident CFOs reduce precautionary motives to save cash by reducing conditional accounting conservatism. Specifically, as suggested by previous studies, when capital markets are frictionless, firms have less incentive to hold cash if they are able to obtain external funds (Miller and Orr 1966). However, firms’ precautionary motives increase when they face uncertainty about future access to external funds, thereby holding more cash (Opler et al. 1999; Denis and Sibilkov 2010). Overconfident managers overestimate the cost of obtaining external funds (Malmendier et al. 2022) and hold more cash for future investment (Chen et al. 2020). When firms adopt low-conditional accounting conservatism, investors will get more useful information for predicting future cash flows. This is a situation where firms can reduce the cost of external financing, and overconfident CFOs alleviate incentives to hold more cash (Lee 2010). Accordingly, we expect overconfident CFOs to reduce precautionary motives to save cash by adopting low- conditional accounting conservatism. We test this prediction using this model:
$$ \begin{gathered} Cash_{i,t} = \beta_{0} + \beta_{1} Holder\it{67}CFO_{i,t} + \beta_{2} RevC\_score_{i,t} + \beta_{3} Holder\it{67}CFO_{i,t} \times RevC\_score_{i,t} \hfill \\ + \beta_{4} Holder\it{67}CEO_{i,t} + \beta_{5} CFO\_male_{i,t} + \beta_{6} CFO\_ownership_{i,t} + \beta_{7} CEO\_male_{i,t} \hfill \\ + \beta_{8} CEO\_ownership_{i,t} + \beta_{9} MTB_{i,t} + \beta_{10} VolCashFlow_{i,t} + \beta_{11} Vwretd_{i,t} \hfill \\ + \beta_{12} SdVwretd_{i,t} + \beta_{13} NWC_{i,t} + \beta_{14} CAPEX_{i,t} + \beta_{15} ACQ_{i,t} + \beta_{16} R\& D_{i,t} \hfill \\ + \beta_{17} FirmSize_{i,t} + \beta_{18} Leverage_{i,t} + \beta_{19} Dividend_{i,t} + \beta_{20} CFO\_vega_{i,t} \hfill \\ + \beta_{21} CEO\_vega_{i,t} + \beta_{22} StrongCG_{i,t} + Firm\;fixed\;effects + Year\;fixed\;effects + \varepsilon_{i,t} , \hfill \\ \end{gathered} $$
(15)
where, the dependent variable is Cash (Cashi,t). The variable of interest is the interaction term, Hold67CFOi,t × RevC_scorei,t. We follow prior studies (e.g., Lee 2010) to control CEO overconfidence (Hold67CEOi,t), CFO gender (CFO_malei,t), CFO ownership (CFO_ownershipi,t), CEO gender (CEO_malei,t), CEO ownership (CEO_ownershipi,t), Market to book ratio (MTBi,t), Volatility of cash flow (VolCashFlowi,t), Value-weighted 12-month market-adjusted returns (Vwretdi,t), Returns volatility (SdVwretdi,t), Net working capital (NWCi,t), Capital expenditure (CAPEXi,t), Acquisition expenditure (ACQi,t), Research and development expenditure (R&Di,t), Firm Size (FirmSizei,t), Leverage (Leveragei,t), Dividend payer (Dividendi,t), CFO risk tolerance (CFO_vegai,t), CEO risk tolerance (CEO_vegai,t) and Strong governance (StrongCGi,t). Detailed variable measurements are provided in the Appendix.
In columns (1) and (2) of Table 11, we find that Holder67CEOi,t and Holder67CFOi,t have a significantly positive coefficient, respectively. These findings suggest that overconfident managers have more incentives to hold cash, which is consistent with the finding of Chen et al. (2020). The coefficient of Holder67CFOi,t × RevC_score1i,t (Holder67CFOi,t × RevC_score2i,t) is significantly negative in column (3) (column (4)), suggesting that overconfident CFOs reduce the incentive to hold cash by reducing conditional accounting conservatism. Overall, this evidence indicates that overconfident CFOs reduce precautionary motives to save cash, as less conditional accounting conservatism conveys more primary information and increases firms’ financial flexibility.
Table 11
The impact of overconfident CFOs’ conditional accounting conservatism decision on cash holding
 
(1)
(2)
(3)
(4)
Variable
Cashi,t
Cashi,t
Cashi,t
Cashi,t
Holder67CFOi,t
 
0.011***
0.007**
0.005
  
(0.003)
(0.003)
(0.003)
RevC_score1i,t
  
0.106***
 
   
(0.015)
 
Holder67CFOi,t × RevC_score1i,t
  
− 0.036***
 
   
(0.013)
 
RevC_score2i,t
   
0.080***
    
(0.011)
Holder67CFOi,t × RevC_score2i,t
   
− 0.022**
    
(0.010)
Holder67CEOi,t
0.005*
0.001
0.000
0.000
 
(0.003)
(0.003)
(0.003)
(0.003)
Constant
0.466***
0.463***
0.516***
0.505***
 
(0.058)
(0.058)
(0.059)
(0.060)
Observations
11,568
11,568
11,568
11,568
Controls in Eq. (15)
Yes
Yes
Yes
Yes
Firm fixed effects
Yes
Yes
Yes
Yes
Year fixed effects
Yes
Yes
Yes
Yes
Adj. R2
0.817
0.817
0.819
0.819
The results of the impact of overconfident CFOs’ conditional accounting conservatism decision on cash holding are shown in Table 11. The standard errors clustering at the firm level are displayed in parentheses. Detailed variable information is shown in Appendix. *, **, and *** represent significance levels of 0.10, 0.05, and 0.01, respectively. Variables of interest are marked in bold

5.2 Alternative explanation

Someone might contend that overconfident CFOs recognize good news more quickly than bad news due to their compensation concerns. From an agency perspective, managers’ stock and option holdings have been considered useful tools for aligning managers’ and shareholders’ interests (Bergstresser and Philippon 2006). Nonetheless, substantial evidence suggests that equity incentives may have the unintended consequence of motivating managers to manipulate the short-term stock price rather than enhancing the firms’ actual profitability (e.g., Peng and Röell 2008; Chava and Purnanandam 2010; Jiang et al. 2010). Specifically, Chava and Purnanandam (2010) and Jiang et al. (2010) suggest a positive association between CFO equity incentives and earnings management, and CFO equity incentives outweigh CEO equity incentives in affecting earnings management. Kim et al. (2011) extend this line of research by showing that CFO equity incentives dominate CEO equity incentives in hoarding bad news.
However, a few studies suggest that managerial overconfidence does not lead to agency problems because overconfident managers believe that their decisions are in the best interest of shareholders. For example, Kim et al. (2016) suggest that overconfident CEOs overestimate the future cash flows of negative NPV projects and continue to execute them. Qiao et al. (2023) find that overconfident CFOs overestimate the stability of their firms’ performance and demonstrate their belief to investors by smoothing earnings. To test whether overconfident CFOs reduce conditional accounting conservatism because they have greater equity incentives than their non-overconfident peers, we examine the moderating effect of CFO equity incentives on the relationship between CFO overconfidence and conditional accounting conservatism.
$$ \begin{aligned} C\_score_{i,t} & = \beta_{0} + \beta_{1}Holder\it{67}CFO_{i,t} + \beta_{2}CFO\_equityincentive_{i,t} + \beta_{3}Holder\it{67}CFO_{i,t} \times CFO\_equityincentive_{i,t} \\ & \quad + \beta_{4}Holder\it{67}CEO_{i,t} + \beta_{5}CFO\_male_{i,t} + \beta_{6}CFO\_ownership_{i,t} \\ & \quad + \beta_{7}CEO\_male_{i,t} + \beta_{8}CEO\_ownership_{i,t} + \beta_{9}MTB_{i,t} + \beta_{10}SaleGrowth_{i,t} \\ & \quad + \beta_{11}Volsale_{i,t} + \beta_{12}CashFlow_{i,t} + \beta_{13}FirmSize_{i,t} + \beta_{14}Leverage_{i,t} \\ & \quad + \beta_{15}RDAD_{i,t} + \beta_{16}Big\_four_{i,t} + \beta_{17}High\_LIT_{i,t} + \beta_{18}UnAC_{i,t} + \beta_{19}CFO\_vega_{i,t} \\ & \quad + \beta_{20}CEO\_vega_{i,t} + \beta_{21}StrongCG_{i,t} + Firm\;fixed\;effects + Year\;fixed\;effects + \varepsilon_{i,t}, \\ \end{aligned} $$
(16)
where, CFO_equityincentivei,t is the CFO equity incentive proxy; Holder67CFOi,t × CFO_equityincentivei,t is the variable of interest; β3 captures the moderating effect of CFO equity incentive on the relationship between CFO overconfidence and conditional accounting conservatism. If the β3 is significantly negative, overconfident CFOs with high equity incentives are more likely to reduce conditional accounting conservatism. Detailed variable measurements are provided in the Appendix.
As shown in Table 12, the coefficient on Holder67CFOi,t × CFO_equityincentivei,t is not significant in columns (1) and (2). There is no evidence to suggest that the CFO equity incentive stimulates the negative relationship between CFO overconfidence and conditional accounting conservatism. Overall, we rule out this alternative explanation and further consolidate our main mechanism.
Table 12
The impact of CFO equity incentive on the relationship between CFO overconfidence and conditional accounting conservatism
 
(1)
(2)
Variable
C_score1i,t
C_score2i,t
Holder67_CFOi,t
− 0.017***
− 0.025***
 
(0.004)
(0.005)
CFO_equityincentivei,t
− 0.016***
− 0.011***
 
(0.003)
(0.004)
Holder67CFOi,t × CFO_equityincentivei,t
0.006
0.003
 
(0.004)
(0.005)
Constant
0.567***
0.492***
 
(0.029)
(0.052)
Observations
11,710
11,710
Controls in Eq. (16)
Yes
Yes
Firm fixed effects
Yes
Yes
Year fixed effects
Yes
Yes
Adj. R2
0.801
0.765
The results of the impact of CFO equity incentive on the relationship between CFO overconfidence and conditional accounting conservatism are shown in Table 12. The standard errors clustering at the firm level are displayed in parentheses. Detailed variable information is shown in Appendix. *, **, and *** represent significance levels of 0.10, 0.05, and 0.01, respectively. Variables of interest are marked in bold

6 CFO power, CFO overconfidence, and conditional accounting conservatism–regression analyses of Hypothesis 2

This section will examine the moderating effect of CFO power on the relationship between CFO overconfidence and conditional accounting conservatism. According to prior studies, our primary measure of CFO power is based on their compensation (e.g., Finkelstein 1992; Bebchuk et al. 2011; Florackis and Sainani 2021). CFO_ranki,t, is an indicator variable that equals one if the CFO is among the three highest-paid managers, and zero otherwise (Florackis and Sainani 2021). This rank represents the value the board assigns to the CFO and the CFO’s authority and responsibility within the firm (Bebchuk et al. 2011; Feng et al. 2011; Baker et al. 2019).14 In addition to using CFO compensation, we also use a measure that reflects the CFO’s position in the management hierarchy to capture CFO power. In particular, earlier research has noted that managers who serve on boards are more likely to influence board and management team decisions (e.g., Beck and Mauldin 2014; Baker et al. 2019). Thus, the CFO board membership is widely used to measure CFO power (e.g., Caglio et al. 2018; Baker et al. 2019; Florackis and Sainani 2021). Besides, since the CEO sets the tone at the top, the close relationship between the CFO and the CEO increases the possibility and power of the CFO to participate in decision-making (e.g., Hsieh et al. 2018). The close relationship between the CEO and CFO is reflected in the long overlap in their tenure (Zenger and Lawrence 1989; Ancona and Caldwell 1992; Zhang 2019; Bowen et al. 2022). To reflect the CFO’s position in the management team, we create a dummy variable called CFO_PowerAlteri,t. It equals one if the CFO meets at least one of the following two conditions: being a member of the board of directors; working together with the CEO for a long time; otherwise, it equals zero.
$$ \begin{aligned} C\_score_{i,t} & = \beta_{0} + \beta_{1} Holder\it{67}CFO_{i,t} + \beta_{2} CFO\_power_{i,t} + \beta_{3} Holder\it{67}CFO_{i,t} \times CFO\_power_{i,t} \\ & \quad + \beta_{4} Holder\it{67}CEO_{i,t} + \beta_{5} CFO\_male_{i,t} + \beta_{6} CFO\_ownership_{i,t} \\ & \quad + \beta_{7} CEO\_male_{i,t} + \beta_{8} CEO\_ownership_{i,t} + \beta_{9} MTB_{i,t} + \beta_{10} SaleGrowth_{i,t} \\ & \quad + \beta_{11} Volsale_{i,t} + \beta_{12} CashFlow_{i,t} + \beta_{13} FirmSize_{i,t} + \beta_{14} Leverage_{i,t} \\ & \quad + \beta_{15} RDAD_{i,t} + \beta_{16} Big\_four_{i,t} + \beta_{17} High\_LIT_{i,t} + \beta_{18} UnAC_{i,t} + \beta_{19} CFO\_vega_{i,t} \\ & \quad + \beta_{20}CEO\_vega_{i,t} + \beta_{21} StrongCG_{i,t} + Firm\;fixed\;effects + Year\;fixed\;effects + \varepsilon_{i,t} , \\ \end{aligned} $$
(17)
where, CFO_poweri,t is one of two CFO power proxies (CFO_ranki,t and CFO_PowerAlteri,t); Holder67CFOi,t × CFO_poweri,t is the variable of interest; β3 captures the moderating effect of CFO power on the relationship between CFO overconfidence and conditional accounting conservatism. If the β3 is significantly negative, overconfident CFOs with more power are more likely to reduce conditional accounting conservatism. Detailed variable measurements are provided in the Appendix.
In columns (1) and (2) of Table 13, the coefficient on Holder67CFOi,t × CFO_ranki,t is significantly negative, suggesting that the CFO power enhances the negative relationship between CFO overconfidence and conditional accounting conservatism, which is in line with our second hypothesis (H2). The results are broadly similar when we use CFO_PowerAlteri,t to measure CFO power.
Table 13
The impact of CFO power on the relationship between CFO overconfidence and conditional accounting conservatism
 
(1)
(2)
(3)
(4)
Variable
C_score1i,t
C_score2i,t
C_score1i,t
C_score2i,t
Holder67CFOi,t
− 0.013***
− 0.018***
− 0.021***
− 0.020***
 
(0.004)
(0.005)
(0.004)
(0.004)
CFO_ranki,t
0.006**
0.014***
  
 
(0.003)
(0.004)
  
Holder67CFOi,t × CFO_ranki,t
− 0.007*
− 0.015***
  
 
(0.004)
(0.005)
  
CFO_PowerAlteri,t
  
0.007**
0.013***
   
(0.003)
(0.003)
Holder67CFOi,t × CFO_PowerAlteri,t
  
0.005
− 0.010**
   
(0.004)
(0.005)
Constant
0.566***
0.484***
0.614***
0.483***
 
(0.030)
(0.053)
(0.032)
(0.051)
Observations
10,679
10,679
11,890
11,890
Controls in Eq. (17)
Yes
Yes
Yes
Yes
Firm fixed effects
Yes
Yes
Yes
Yes
Year fixed effects
Yes
Yes
Yes
Yes
Adj. R2
0.804
0.764
0.743
0.764
The results of the impact of CFO power on the relationship between CFO overconfidence and conditional accounting conservatism are shown in Table 13. The standard errors clustering at the firm level are displayed in parentheses. Detailed variable information is shown in Appendix. *, **, and *** represent significance levels of 0.10, 0.05, and 0.01, respectively. Variables of interest are marked in bold

7 Conclusion

Our study provides statistically and economically significant evidence that overconfident CFOs tend to use less conditional accounting conservatism than non-overconfident CFOs, especially when they have great power. To address the potential endogeneity problems, we use the PSM-DID test to rule out the concern that overconfident CFOs self-select into firms with a low level of conditional accounting conservatism. Besides, we use an alternative overconfidence measurement and consider the effects of risk tolerance, CEO characteristics, and governance. These robustness tests show the consistency and reliability of our main findings. In the mechanism tests, we prove that overconfident CFOs who reduce conditional accounting conservatism aim to increase earnings informativeness and decrease precautionary motives rather than pursue private benefit.
Our findings contribute significantly to literature and theory. These findings extend the growing literature on the influence of CFO overconfidence on corporate decisions and provide a new understanding of the determinants of conditional accounting conservatism. Our study extends Ahmed and Duellman (2013), by showing that CFO overconfidence can determine asymmetric recognition of good and bad news. Theoretically, our findings provide more empirical support for the combination of upper echelons and overconfidence theories in explaining CFO behavior. Besides, we find that overconfident CFOs with high power are more likely to reduce conditional accounting conservatism, supporting the updated upper echelons theory. We rule out the agency theory-based view that overconfident CFOs reduce conservative accounting information for their own benefit. We suggest that overconfident CFOs believe they act in the best interests of shareholders by reducing conditional accounting conservatism to increase earnings informativeness and financial flexibility.
Our findings also have great implications for practice and policy. For policymakers, our findings provide more evidence from a CFO overconfidence perspective for the current discussion over FASB and IASB’s decisions, which removed conservatism from the joint conceptual framework. Furthermore, our findings can assist investors in better comprehending CFO overconfidence and bridge the gap between academic and practical understandings of managerial rationality.
Given that the effect of CFO overconfidence has received little attention within the current literature and CFOs have a growing role in firms’ strategic decision-making (e.g., Chen et al. 2022; Malmendier et al. 2022; Qiao et al. 2023), we suggest that future studies pay more attention to the effect of CFO overconfidence as well as the joint effect of CEO overconfidence and CFO overconfidence on corporate decision-making.

Acknowledgements

This paper was presented at the 2021 Academy of Sustainable Finance, Accounting, Accountability & Governance (ASFAAG) Annual Conference, 2021 British Accounting and Finance Association Annual Conference, and 2020 Nottingham University Business School (NUBS) TriCampus Conference. The comments of participants are gratefully acknowledged, in particular the critical feedback and positive encouragement of Professor Musa Mangena, Professor Sabur Mollah, Professor Niamh Brennan, Professor Henry Agyei-Boapeah, and Dr Hafez Abdo. Any errors that remain are our own.

Declarations

Conflict of interest

The authors have no relevant financial or non-financial interests to disclose.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://​creativecommons.​org/​licenses/​by/​4.​0/​.

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendix

Appendix: Variable information

Variable name
Variable measurement
Data source
Dependent variables
Conditional Accounting Conservatism (C_score1i,t)
The method of Khan and Watts (2009). Detailed information is provided in Sect. 3.2.1
CCM; CRSP
Conditional Accounting Conservatism (C_score2i,t)
The method of Banker et al. (2016). Detailed information is provided in Sect. 3.2.1
CCM; CRSP
Independent variable
CFO Overconfidence (Holder67CFOi,t)
The method of Campbell et al. (2011) and Chen et al. (2022). Detailed information is provided in Sect. 3.2.2
ExecuComp
Control variables
CEO Overconfidence (Holder67CEOi,t)
The method of Campbell et al. (2011) and Chen et al. (2022). Detailed information is provided in Sect. 3.2.2
ExecuComp
CFO gender (CFO_malei,t)
An indicator variable that equals one if CFO is male, and zero otherwise (Francis et al. 2015)
ExecuComp
CFO ownership (CFO_ownershipi,t)
The percentage of share outstanding excluded options held by the CFO (LaFond and Watts 2008)
ExecuComp
Market to book ratio (MTBi,t)
Equity market value divided by equity book value (Ahmed and Duellman 2013)
CCM
Sales growth (SaleGrowthi,t)
The percentage of annual growth in total sales (Ahmed and Duellman 2013)
CCM
Sales volatility (VolSalei,t)
The standard deviation of sales measured from fiscal year t-5 to fiscal year t-1 deflated by the total asset (Wang et al. 2018)
CCM
Cash Flow (CashFlowi,t)
Cash flow divided by total assets (Hsieh et al. 2014)
CCM
Firm Size (FirmSizei,t)
The natural logarithm of total assets (Jiang et al. 2010)
CCM
Leverage (Leveragei,t)
Total debt divided by total assets (Ahmed and Duellman 2013)
CCM
Research and development expense and advertising expense (RDADi,t)
Total research and development expense plus advertising expense deflated by total sales (Ahmed and Duellman 2013)
CCM
Audit firm (Big_fouri,t)
An indicator variable that equals one if the audit firm belongs to one of the big four auditor firms, Deloitte, EY, KPMG, and PwC, and zero otherwise
CCM
High litigation risk (High_LITi,t)
An indicator variable that equals one if firms’ litigation risk is in the top decile of the sample, and zero otherwise (Gao et al. 2020). The coefficients from the model (2) in Table 7 of Kim and Skinner (2012) are used to calculate the litigation risk
CCM; CRSP
Unconditional accounting conservatism (UnACi,t)
Income before extraordinary items minus cash flows from operations plus depreciation expense divided by total assets, and averaged over the preceding three years, multiplied by negative one (Ahmed et al. 2002)
CCM
Additional variables used in robustness tests
Treati
An indicator variable that equals one if the outgoing CFO is non-overconfident and the incoming CFO is overconfident, and zero if both the outgoing CFO and incoming CFO are non-overconfident
ExecuComp
Posti,t
An indicator variable that equals one when observations occur in the first year after CFO turnover, and zero when observations occur in the last year before CFO turnover
ExecuComp
Alternative CFO Overconfidence measurement (Holder100CFOi,t)
An indicator variable that equals one from the first time that CFOs hold vested options that are at least 100% in the money to the end of their tenure, and zero otherwise (Hirshleifer et al. 2012)
ExecuComp
Alternative CEO Overconfidence measurement (Holder100CEOi,t)
An indicator variable that equals one from the first time that CEOs hold vested options that are at least 100% in the money to the end of their tenure, and zero otherwise (Hirshleifer et al. 2012)
ExecuComp
CFO risk tolerance (CFO_vegai,t)
Natural logarithm of CFOs’ vega plus one (Huang et al. 2016). Vega is measured following the method of Core and Guay (2002)
CRSP
CEO risk tolerance (CEO_vegai,t)
Natural logarithm of CEOs’ vega plus one (Huang et al. 2016). Vega is measured following the method of Core and Guay (2002)
CRSP
CEO gender (CEO_malei,t)
An indicator variable that equals one if CEO is male, and zero otherwise (Francis et al. 2015)
ExecuComp
CEO ownership (CEO_ownershipi,t)
The percentage of share outstanding excluded options held by the CEO (LaFond and Watts 2008)
ExecuComp
Strong governance (StrongCGi,t)
Firms have strong governance if they meet three of the four criteria listed below: First, the CEO is not simultaneously the Chairman. Second, the percentage of outside directors is greater than the sample’s median figure. Third, the institutional ownership percentage is more than the sample’s median figure. Fourth, the percentage of inside directors is lower than the median value of the sample (Ahmed and Duellman 2013)
Thomson Reuters Institutional (13f) Holdings; BoardEx
Additional variables used in the analysis of mechanisms
Low conditional accounting conservatism (RevC_score1i,t)
The C_score1i,t multiply by − 1. The higher value of RevC_score1i,t indicates low conditional accounting conservatism
CCM; CRSP
Low conditional accounting conservatism (RevC_score2i,t)
The C_score2i,t multiply by − 1. The higher value of RevC_score2i,t indicates low conditional accounting conservatism
CCM; CRSP
Stock return (Ri,t)
The annual buy-and-hold stock return (Shu 2021)
CRSP
Last year’s earnings per share (Earni,t-1)
Earnings per share for fiscal year t-1 scaled by the stock price at the beginning of year t (Shu 2021)
CCM
Earnings per share (Earni,t)
Earnings per share for fiscal year t scaled by the stock price at the beginning of year t (Shu 2021)
CCM
The sum of earnings per share for fiscal year t + 1 to t + 3 (Earni,t+3)
The sum of earnings per share scaled by the stock price at the beginning of year t from fiscal year t + 1 to t + 3 (Shu 2021)
CCM
The sum of stock return for fiscal year t + 1 to t + 3 (Ri,t+3)
The sum of annual buy-and-hold stock return from fiscal year t + 1 to t + 3 (Shu 2021)
CRSP
Firm age (FirmAgei,t)
The sum of years from the firm that appears in the Compustat database (Shu 2021)
Compustat
Investment (Investi,t)
R&D expenses plus capital expenditures minus sales of fixed assets scaled by total assets (Shu 2021)
CCM
Property, Plant & Equipment (PPEi,t)
The net property, plant, and equipment scaled by total assets (Shu 2021)
CCM
Negative earnings per share (Lossi,t)
An indicator variable that equals one if the firm has negative earnings per share, zero otherwise (Shu 2021)
CCM
Number of analysts (Analyst_numberi,t)
The natural logarithm of one plus the number of analysts (Shu 2021)
I/B/E/S
Cash (Cashi,t)
The sum of cash and cash equivalents scaled by total assets (Lee 2010)
CCM
Volatility of cash flow (VolCashFlowi,t)
The standard deviation of cash flow measured from fiscal year t-5 to fiscal year t-1 deflated by the total asset (Wang et al. 2018)
CCM
Value-weighted 12-month market-adjusted returns (Vwretdi,t)
Value-weighted 12-month market-adjusted returns (Lee 2010)
CRSP
Returns volatility (SdVwretdi,t)
12-month market-adjusted returns volatility (Lee 2010)
CRSP
Net working capital (NWCi,t)
Working capital less cash scaled by total assets (Lee 2010)
CCM
Capital expenditure (CAPEXi,t)
Capital expenditure scaled by total assets (Lee 2010)
CCM
Acquisition expenditure (ACQi,t)
Acquisition expenditure scaled by total assets (Lee 2010)
CCM
Research and development expenditure (R&Di,t)
Research and development expenditure scaled by total assets (Lee 2010)
CCM
Dividend payer (Dividendi,t)
An indicator variable that equals one the firm is a dividend payer, and zero otherwise (Lee 2010)
CCM
CFO equity incentive (CFO_equityincentivei,t)
ONEPCT/(ONEPCT + Salary + Bonus). ONEPCT refers to the dollar change in the value of the CFOs’ stock and option holdings as a result of a 1% increase in the firm stock price (Bergstresser and Philippon 2006)
ExecuComp
Additional variables used in the analysis of CFO power
CFO power (CFO_ranki,t)
An indicator variable that equals one if the CFO is among the three highest-paid managers, and zero otherwise (Florackis and Sainani 2021)
ExecuComp
Alternative CFO power measurement (CFO_PowerAlteri,t)
An indicator variable that equals one if the CFO meets at least one of the following two conditions: being a member of the board of directors; working together with the CEO for a long time; otherwise, it equals zero (Florackis and Sainani 2021; Bowen et al. 2022)
ExecuComp
Footnotes
1
Agency theory suggests that information asymmetries between shareholders and managers cause managers to make decisions contrary to the shareholders’ best interests (Jensen and Meckling 1976).
 
2
Conditional accounting conservatism refers to accountants’ demand for a greater degree of verification to recognize good news than bad news; this is reflected in the accounting recognition of bad news being more timely than good news, where bad news (good news) represents negative (positive) expected cash flows (e.g., Basu 1997; Beaver and Ryan 2005).
 
3
The relationship between conditional accounting conservatism and financial reporting quality is controversial. The supporters of the ‘contracting efficiency view’ suggest that the primary purpose of financial reporting is to improve contract efficiency and monitoring (Holthausen and Watts 2001). From a ‘contracting efficiency view’, conditional accounting conservatism provides lenders with more relevant information through timely loss recognition, improving the detection of default risk (e.g., Ahmed et al. 2002; Watts 2003). However, adherents of the ‘value relevance view’ argue that monitoring is a secondary consideration for financial reporting and that financial reporting should be mainly concerned with the utility of forecasting future cash flows (Barth et al. 2001; Schipper 2005). From a ‘value relevance view’, conditional accounting conservatism makes it difficult for investors to use earnings information to forecast future cash flow (Mensah et al. 2004; Heflin et al. 2015).
 
4
Unconditional accounting conservatism refers to the constant under recognition of net assets (Beaver and Ryan 2005). Unconditional accounting conservatism includes expensing R&D and advertising expenditures, using accelerated depreciation methods, accumulating reserves above expected future costs, etc. (Beaver and Ryan 2005).
 
5
Similar to Ahmed and Duellman (2013), we also investigate the relationship between CFO overconfidence and unconditional accounting conservatism. Although the origins of conditional and unconditional accounting conservatism are distinct, their underlying causes are comparable (Ma et al. 2020). Specifically, we predict that overconfident CFOs overestimate their firms’ future performance, and thus they are more likely to recognize spending on expected future earnings as an investment and choose lower depreciation rates, reducing conditional accounting conservatism. Based on our untabulated result, we also observe a significant negative association between CFO overconfidence and the adoption of unconditional accounting conservatism.
 
6
Measuring news with firms’ stock returns (Basu 1997).
 
7
Since the method of Basu (1997) has a few limitations that are pointed out by many studies, such as Dietrich et al. (2007) and Khan and Watts (2009), we do not use this method in our main analysis. However, untabulated results show that there is a significant negative relationship between CFO overconfidence and conditional accounting conservatism when we use the method of Basu (1997).
 
8
Cost stickiness refers to earnings responding more to a decrease than an increase in sales (Banker et al. 2016).
 
9
Results remain robust if we require that CFOs (CEOs) hold vested options that are at least 67% in the money at least twice.
 
10
We use two-way fixed-effect models, firm-fixed effect and year-fixed effect, to control for the influence of time-invariant firm features and factors that are common to all firms for a particular fiscal year, respectively. Besides, the untabulated Fischer (F), Breusch-Pagan Lagrange Multiplier (LM), and Hausman tests indicate that fixed-effect models are the best choice for our study compared with random effect and pooled OLS models. The following regressions used the fixed-effect models for the same reasons, which we will not repeat for brevity.
 
11
25.806% equals the absolute value of the coefficient on Holder67CFOi,t (0.016) in column (1) of Table 4 divided by the mean value of C_score1i,t (0.062) in Table 2, then multiplied by 100%.
 
12
19.355% equals the absolute value of the coefficient on Holder67CFOi,t (0.012) in column (3) of Table 4 divided by the mean value of C_score1i,t (0.062) in Table 2, then multiplied by 100%. 9.890% equals the absolute value of the coefficient on Holder67CFOi,t (0.018) in column (4) of Table 4 divided by the mean value of C_score2i,t (0.182) in Table 2, then multiplied by 100%.
 
13
Although we have used two methods to capture firm-specific conditional accounting conservatism, we also use the method proposed by Basu (1997) to further increase robustness. Untabulated results show that overconfident CFOs tend to recognize good news more promptly than bad news.
 
14
We use two other measures of CFO relative pay share within top management teams to increase the reliability of using CFO pay to capture CFO power. First, CFO_topquartilei,t, is an indicator variable that equals one if the total CFO compensation is divided by the total compensation of the top three executives, excluding the CEO and CFO, in the top quartile of the sample, and zero otherwise (Baker et al. 2019). Second, CFO_payslicei,t equals CFO total compensation to the top five executives’ total compensation. If fewer than five executives’ compensations are reported, we assume that the remaining top five unreported executives get the same level of salary as the lowest-paid executive among those disclosed (Bebchuk et al. 2011; Feng et al. 2011). Using the above two CFO power measures does not qualitatively alter our documented findings. The results are not presented for brevity.
 
Literature
go back to reference Barth, ME, Landsman, WR, Raval, V, Wang, S (2014) Conservatism and the information content of earnings. Working paper, Stanford University and the University of North Carolina. Barth, ME, Landsman, WR, Raval, V, Wang, S (2014) Conservatism and the information content of earnings. Working paper, Stanford University and the University of North Carolina.
go back to reference Basu, S, Hwang, L-S, Jan, C-L (2001) Differences in conservatism between big eight and non-big eight auditors. Working paper, Baruch College, City University of New York Basu, S, Hwang, L-S, Jan, C-L (2001) Differences in conservatism between big eight and non-big eight auditors. Working paper, Baruch College, City University of New York
go back to reference Black, DE, Gallemore, J (2013) Bank executive overconfidence and delayed expected loss recognition. Black, DE, Gallemore, J (2013) Bank executive overconfidence and delayed expected loss recognition.
go back to reference Bowen, RM, Lyon, SC, Malhotra, S, Zhu, P (2022) CEO-CFO compatibility and audit engagement risk. Working paper, Chapman University Bowen, RM, Lyon, SC, Malhotra, S, Zhu, P (2022) CEO-CFO compatibility and audit engagement risk. Working paper, Chapman University
go back to reference Fasb (2010) Statement of financial accounting concepts no. 8: conceptual framework for financial reporting. Norwalk, CT: Author. Fasb (2010) Statement of financial accounting concepts no. 8: conceptual framework for financial reporting. Norwalk, CT: Author.
go back to reference Gujarati, DN, Porter, DC, Gunasekar, S (2012) Basic econometrics. Tata McGraw-Hill Education Gujarati, DN, Porter, DC, Gunasekar, S (2012) Basic econometrics. Tata McGraw-Hill Education
go back to reference IASB (2010). The conceptual framework for financial reporting. IFRS Foundation, London IASB (2010). The conceptual framework for financial reporting. IFRS Foundation, London
go back to reference Lee, J (2010) The role of accounting conservatism in firms’ financial decisions. Working paper, Singapore Management University Lee, J (2010) The role of accounting conservatism in firms’ financial decisions. Working paper, Singapore Management University
go back to reference Malmendier U, Pezone V, Zheng H (2022) Managerial duties and managerial biases. Manag Sci 69:3174–3201 CrossRef Malmendier U, Pezone V, Zheng H (2022) Managerial duties and managerial biases. Manag Sci 69:3174–3201 CrossRef
go back to reference Qiao L, Adegbite E, Nguyen TH (2023) Chief financial officer overconfidence and earnings management. In: Accounting Forum Forthcoming Qiao L, Adegbite E, Nguyen TH (2023) Chief financial officer overconfidence and earnings management. In: Accounting Forum Forthcoming
Metadata
Title
CFO overconfidence and conditional accounting conservatism
Authors
Lu Qiao
Emmanuel Adegbite
Tam Huy Nguyen
Publication date
06-11-2023
Publisher
Springer US
Published in
Review of Quantitative Finance and Accounting / Issue 1/2024
Print ISSN: 0924-865X
Electronic ISSN: 1573-7179
DOI
https://doi.org/10.1007/s11156-023-01188-7

Other articles of this Issue 1/2024

Review of Quantitative Finance and Accounting 1/2024 Go to the issue