This study examines the implications of top executives' narcissism on auditor appointment and remuneration, focusing on the association between narcissistic leadership traits and audit outcomes in the context of an emerging economy, Bangladesh.
Using a hand-collected sample of 436 firm-year observations from publicly listed firms in Bangladesh over the period 2018–2021, this study employs panel regression to examine the impact of executive narcissism on auditor selection and audit fees. Endogeneity is addressed using a combination of entropy balancing, a two-stage least squares instrumental variable approach, and the Heckman two-stage selection model.
The findings indicate that narcissistic chief executive officers and chairs tend to appoint geographically proximate auditors while paying higher audit fees, potentially to exert greater influence over the audit process. This behavior is more pronounced in family firms and linked to faster audit report issuance, with no evidence of improved audit quality. However, Big 4 auditors weaken the observed effect. The results remain robust across alternative measures of executive narcissism and auditor–client distance, offering novel insights into how executive traits shape corporate governance and audit outcomes.
The study underscores the need for regulatory interventions to safeguard auditor independence and uphold audit quality, particularly in settings where narcissistic executives may prioritize personal influence over organizational integrity. It further highlights the importance of auditors to carefully assess risks associated with executive personality traits when determining audit fees and providing audit services.
This research is among the first to examine the interplay between executive narcissism and auditor–client geographic proximity within an emerging economy. By extending the literature on leadership traits, corporate governance and audit outcomes, the study provides novel insights of relevance to academics, policymakers and practitioners.
1. Introduction
Corporate governance and financial reporting quality are pivotal to the integrity and efficiency of capital markets. Central to these mechanisms is the auditing process, which serves as a critical safeguard against financial misrepresentation and fraud. However, the effectiveness of auditing is not solely determined by regulatory frameworks or technical expertise; it is also influenced by the behavioral and psychological traits of corporate executives. Among these traits, narcissism—a personality characteristic marked by grandiosity, entitlement, dominance, and a desire for admiration—has emerged as a significant factor shaping corporate decision-making and financial outcomes (Chatterjee and Hambrick, 2007; Olsen et al., 2014; Kim and Anderson, 2024). Narcissistic executives, such as CEO and Chairman, often exhibit overconfidence, risk-seeking behavior, and a propensity for manipulating information to enhance their personal image (Ham et al., 2018). These tendencies can have profound implications for auditing practices, particularly in the areas of auditor-client geographic proximity and audit fees (Johnson et al., 2013).
The concept of auditor-client geographic proximity has gained attention in recent years due to its potential impact on audit quality and efficiency. Geographic proximity between auditors and clients can facilitate better communication, deeper local knowledge, and more effective monitoring, all of which contribute to higher audit quality (Choi et al., 2012; Dong et al., 2018). However, proximity can also create opportunities for undue influence, particularly when powerful executives seek to manipulate audit outcomes. Narcissistic executives, with their desire for control and admiration, may exploit their geographic closeness to auditors to sway audit decisions in their favor, thereby compromising audit integrity (Johnson et al., 2013). In addition to geographic proximity, audit fees represent another critical dimension of the auditor-client relationship. Audit fees reflect the complexity, risk, and effort associated with an audit engagement, and they are often influenced by factors such as client size, industry, and governance quality (Hay et al., 2023). However, the presence of narcissistic executives can introduce additional complexities. For instance, narcissistic CEO or Chairman may demand more extensive audits to create an illusion of transparency and accountability, thereby driving up audit fees. Conversely, they may pressure auditors to reduce fees by leveraging their influence or exploiting relational proximity, potentially compromising audit quality (Johnson et al., 2013). These dynamics are particularly pronounced in emerging economies like Bangladesh, where concentrated ownership structures and weak regulatory enforcement create an environment conducive to executive overreach (Rajpurohit and Rijwani, 2022).
The interplay between narcissistic executives, auditor-client geographic proximity, and audit fees is critical in the context of emerging economies. These economies are often characterized by institutional weaknesses, regulatory gaps, and a prevalence of family-owned businesses, where executives wield significant influence over corporate decisions (Siddiqui, 2024). In these contexts, geographic proximity between auditors and clients can carry both advantages and risks: while closer distance may improve communication and audit efficiency, it can also facilitate undue influence or relational pressure from powerful executives. When combined with audit fee negotiations, these proximity dynamics can place auditors in the difficult position of balancing professional independence with the demands of influential clients. The stakes are high, as the outcomes of such interactions have direct implications for financial transparency, investor confidence, and the overall stability of capital markets.
This study is motivated by a need to fill critical gaps in the intersection of auditing, corporate governance, and executive behavior. While a substantial body of literature links executive narcissism to outcomes such as earnings management, aggressive financial reporting, and corporate risk-taking (Olsen and Stekelberg, 2016; Ham et al., 2018), its influence on the audit process itself remains largely underexamined. Auditors serve as gatekeepers of financial credibility, yet little is known about how personality-driven behaviors of senior leaders shape auditor selection, fee structures, and other engagement terms. Furthermore, although auditor–client geographic proximity and audit fees have been studied extensively in developed markets (Choi et al., 2012; Dong et al., 2018), there is a paucity of research in emerging economies, where concentrated ownership structures, relational contracting, and less stringent oversight may amplify the effects of executive traits. Only Rahaman et al. (2025) investigated the association between auditor-client geographic proximity and audit characteristics such as audit fees, quality and audit report lag using the context of an emerging economy like Bangladesh. This gap is significant because institutional context can alter not only the strength of these relationships but also the underlying mechanisms driving them. By investigating these dynamics in Bangladesh, this study extends existing literature on executive behavior and auditing, offering fresh insights into how individual-level psychological traits interact with structural features of the governance environment to shape audit-related outcomes.
Using a dataset comprising 436 firm-year observations from 2018 to 2021 for the listed companies in Bangladesh, we document that top executive narcissism is positively associated with both auditor–client geographic proximity and audit fees. Further, the association is more pronounced in family firms. However, Big 4 auditors mitigate the Chair/CEO narcissism effect on engagement proximity and audit fees. These results remain robust after accounting for endogeneity concerns and employing alternative measures of executive narcissism. Additionally, further analysis indicates that while chair narcissism does not significantly affect audit quality, CEO narcissism is negatively and significantly associated with audit report lag, suggesting a potential influence on the timeliness of financial reporting.
This study makes several significant contributions to the literature on executive narcissism and audit traits. First, the study provides the first-hand empirical evidence linking executive narcissism to auditor–client geographic proximity. Prior research on auditor location decisions has primarily focused on economic, institutional, and relational determinants such as client size, audit market competition, and industry specialization (Choi et al., 2012; Dong et al., 2018; Rahaman et al., 2025). However, these studies have largely treated auditor selection as a rational, cost-based decision, overlooking the psychological characteristics of executives who often influence such choices. By integrating the perspective of upper echelons theory (Hambrick and Mason, 1984), this paper introduces an individual-level lens to an economic inquiry. It shows that narcissistic executives may deliberately prefer geographically proximate auditors to facilitate greater interaction, assert dominance, and influence audit outcomes. This study insight challenges the implicit assumption that auditor choice is primarily driven by rational cost–benefit considerations. Instead, auditor selection may also reflect executive preferences for prestige, control, or influence, particularly when the executives exhibit narcissistic tendencies. This perspective moves beyond the structural determinants explored in previous work and introduces executive personality as a novel determinant of auditor proximity and thus addresses an unexamined gap in both the audit geography and executive behavior literature.
Second, this study enriches the literature on executive traits and audit pricing by investigating the relationship between executive narcissism and audit fees within an emerging economy characterized by relationship-based governance and weaker institutional enforcement. Existing studies, such as Xiang and Song (2021) have documented higher audit fees for narcissistic CFOs in China, but the current study extends this inquiry in two important ways: (1) by examining both CEOs and board chairmen (two distinct yet powerful leadership roles) and (2) by using the context of Bangladesh, where informal networks, personal influence, and relational contracting play a much stronger role in audit negotiations than in developed markets. This context-specific evidence demonstrates that the behavioral and institutional dynamics through which narcissism affects audit pricing differ meaningfully across governance environments and adds depth to the cross-country understanding of executive psychology and auditing. Emerging markets may have more concentrated ownership, family control, fewer independent boards, and laxer auditor monitoring. These institutional characteristics provide managers more discretion, giving narcissistic executives more power to choose auditors and determine audit fees. This contributes to the research on developing markets by highlighting that executive behavioral variation must be taken into account to fully understand governance outcomes, particularly in environments with limited institutional controls.
Third, the study advances literature by jointly examining narcissism in two distinct leadership roles (CEOs and board chairs) rather than focusing exclusively on one. A CEO typically influences operational and strategic decisions, while a chairman plays a key role in board oversight and external relationship management. This dual-role perspective allows us to assess whether narcissism operates differently depending on the nature of the leadership position and offers a more comprehensive understanding of how top-level personalities influence auditor relationships.
Fourth, our findings suggest that the influence of executive narcissism on auditor selection is more pronounced in family firms and when firms engage Big-4 auditors. Concentrated ownership and relational governance in family firms may increase CEOs' discretion, making it easier for narcissistic qualities to influence auditor selection. On the other hand, Big-4-affiliated auditors reduce personality-driven effects by enforcing strict external monitoring and reputational discipline. These results extend upper echelons theories by highlighting how governance structures condition behavioral influences.
Fifth, by extending the analysis to audit quality and audit report lag, the study provides a richer understanding of the broader consequences of executive narcissism for the audit process. While we find no significant effect on conventional measures of audit quality, the negative association between CEO narcissism and audit report lag suggests that narcissistic leaders may expedite the release of audited financial statements, potentially as a means of showcasing efficiency or responsiveness to stakeholders. This opens a new line of inquiry into the timeliness of financial reporting as a potential channel for executive image management.
Finally, the study extends the upper echelons theory into the audit setting by demonstrating that executives' psychological traits influence not only strategic and operational decisions but also the nature of external assurance relationships. By empirically showing that executive narcissism affects auditor proximity, fee structures, and reporting timeliness, the study positions auditing outcomes as behavioral reflections of executive personality. Thus, this study enriches the explanatory power of the upper echelon theory in external governance contexts.
The findings of this study carry several important practical and policy implications for regulators, boards, and audit practitioners, particularly in emerging-market settings characterized by relationship-based governance and concentrated executive power. Regulators may strengthen oversight of auditor appointment and remuneration by requiring firms to formally document and publicly disclose the rationale behind auditor selection, geographic proximity, and audit fee negotiations. Periodic regulatory reviews could be instituted for cases in which auditor–client proximity or fee increases appear unusual. Given the disciplining role of Big-4 auditors in emerging markets, policies that strengthen auditor independence and audit committee authority are especially important where Big-4 expansion is limited. Boards and audit committees can mitigate the influence of narcissistic executives by reinforcing committee independence, centralizing auditor interactions within the audit committee, and incorporating behavioral considerations into executive selection. Audit firms, in turn, may incorporate executive personality risk indicators into client acceptance and engagement planning, adjust audit team structures, and strengthen documentation to safeguard independence, particularly in relationship-based governance environments.
The remainder of this study is structured as follows: Section 2 discusses the nature of narcissistic executives and auditing practices reforms in Bangladesh. Section 3 outlines the theoretical framework, whereas Section 4 provides a comprehensive review of prior literature to support the development of hypotheses. Section 5 details the research methodology, including the sample structure, research design, and model development. Section 6 presents the findings, encompassing descriptive statistics, bivariate analysis, and multivariate analysis with detailed discussion. Finally, Section 7 concludes the study, highlighting its implications, limitations, and avenues for future research.
2. Narcissistic executives and auditing practices reforms in Bangladesh
Bangladesh provides a timely and relevant setting for examining the interplay between narcissistic executives and auditing practices due to its combination of regulatory reforms and persistent institutional weaknesses. Over the past decade, the country has introduced significant corporate governance and auditing reforms, including the Financial Reporting Act 2015, the establishment of the Financial Reporting Council (FRC), minimum audit fee guideline 2016 by the Institute of Chartered Accountants of Bangladesh (ICAB) and the adoption of Bangladesh Standards on Auditing (BSA), largely modeled on International Standards on Auditing (ISA). These initiatives were designed to enhance transparency, strengthen auditor independence, and align corporate practices with global standards. However, despite these formal advancements, the enforcement of audit reforms remains weak due to capacity constraints, political influence, and systemic institutional voids (Zafarullah and Rahman, 2008). Karim et al. (2025) report that, despite the existence of minimum audit fee guidelines in Bangladesh, compliance remains extremely low, with only about 7% of companies adhering to the prescribed standards. The World Bank's Worldwide Governance Indicators (2023) consistently rank Bangladesh low on regulatory quality and rule of law, highlighting the difficulties of implementing reforms effectively. Such conditions create an environment in which auditors struggle to maintain independence, particularly when interacting with powerful executives who are entrenched in both corporate and political networks.
The governance landscape in Bangladesh is characterized by concentrated ownership and a prevalence of family-controlled businesses, which magnifies the influence of executive personality traits on corporate outcomes (Khan et al., 2015). Narcissistic leaders may exploit weak institutional checks to shape decision-making and organizational practices in their favor (Chatterjee and Hambrick, 2007; Ham et al., 2018). Research indicates that narcissistic executives can simultaneously produce value-enhancing outcomes, such as strategic boldness, innovation, and visionary leadership, and value-diminishing outcomes, including opportunistic reporting, earnings manipulation, and ethical violations (Brunell et al., 2008; Kashmiri et al., 2017; Olsen et al., 2014). In emerging economies like Bangladesh, where enforcement mechanisms are fragile, the potential for negative manifestations of narcissism is heightened, increasing risks for financial transparency and governance integrity.
These behavioral dynamics intersect critically with the auditing process. Auditors in Bangladesh face structural limitations, including resource constraints, limited technological adoption, and weak enforcement of independence requirements (Nurunnabi, 2018; Rahaman et al., 2025). Simultaneously, client firms that are often led by dominant, narcissistic executives can exert informal influence over auditors through personal relationships or fee negotiations. Geographic proximity between auditor and client, common due to limited auditor distribution outside major cities like Dhaka and Chattogram, may facilitate communication but can also reinforce familiarity that undermines auditor independence (Rahaman et al., 2025). Narcissistic executives may exploit such proximity to steer audit outcomes in their favor, either by pressuring auditors to reduce fees or by selectively employing high-fee audits to signal transparency and prestige. In both cases, audit decisions reflect the executives' personal goals as much as organizational needs, rendering the audit function vulnerable to symbolic compliance rather than substantive oversight.
The Bangladeshi context also provides a practical field for studying the interaction between executive personality and auditing in a partially reformed regulatory environment. While formal rules exist, enforcement gaps allow dominant executives to exert influence over audit practices, providing a clear setting to examine whether institutional reforms can mitigate the behavioral risks associated with narcissistic leadership. Events such as the 2024 formation of an Anti-Corruption Reform Commission led by Transparency International Bangladesh highlight ongoing efforts to address governance challenges, underscoring that regulatory compliance alone is insufficient without cultural and behavioral alignment (Corruption Perceptions Index, 2024).
By situating this study in Bangladesh, we contribute to the literature on executive behavior and auditing by extending evidence beyond developed economies, where enforcement is generally stronger and governance mechanisms more robust. Politically connected, family-controlled firms in corruption-prone settings make the role of auditing a critical mechanism through which executives can exert influence. Understanding how narcissistic executives operate in such environments illuminates broader implications for financial transparency, investor confidence, and market stability, not only in Bangladesh but across similarly structured emerging economies (Ham et al., 2018; Nurunnabi, 2018).
3. Theoretical literature review
This study investigates the influence of executive narcissism on auditor–client geographic proximity and audit fees through the lens of upper echelons theory. Originally developed by Hambrick and Mason (1984), this theoretical framework posits that organizational outcomes are reflections of the values, experiences, and personality traits of top executives, as strategic decisions are shaped by executives' cognitive interpretations and psychological characteristics. Under this perspective, executives act as key decision filters, translating their personal traits into organizational choices. In settings where managerial discretion is relatively high—such as emerging economies characterized by weaker board oversight and less stringent governance mechanisms—executive personality traits are likely to exert a particularly strong influence. Consequently, narcissistic tendencies among top executives may play a critical role in shaping auditor-related decisions, including auditor selection, engagement management, and the negotiation of audit scope and fees.
Narcissism, characterized by grandiosity, entitlement, dominance, and a desire for admiration, is particularly consequential under the upper echelon theory. Narcissistic executives often overestimate their capabilities, seek to control relationships with key stakeholders, and engage in behaviors aimed at enhancing their status and visibility (Chatterjee and Hambrick, 2007). Prior research links narcissistic leadership to aggressive reporting, impression management, and risk-taking behaviors (Ham et al., 2018; Olsen and Stekelberg, 2016), suggesting that such leaders shape not only the internal governance environment but also the firm's interactions with external monitors. These tendencies imply that narcissistic executives may influence audit relationships in ways that reflect their psychological motives, not merely organizational needs.
However, audit-related decisions also have established economic and logistical determinants. Auditor–client geographic proximity is often linked to efficiency benefits such as reduced travel time, easier access to client premises, and improved local knowledge (Choi et al., 2012). Audit fees commonly reflect factors such as engagement risk, auditor expertise, and market competition (Hogan and Wilkins, 2008; Simunic, 1980). A complete framework must therefore recognize these conventional explanations while considering how executive narcissism interacts with or amplifies them. Upper echelons theory suggests that narcissistic leaders may assign greater weight to relational and symbolic aspects of auditor choice, such as the ability to monitor, influence, or showcase alignment with reputable auditors, than other executives would.
Impression management theory further clarifies these behavioral motives. Narcissistic executives seek positive social evaluation and often construct environments that reinforce their desired image (Goffman, 1959; Schlenker, 1980). Geographic proximity enables more frequent interaction with auditors, allowing narcissistic leaders to shape narratives, signal engagement, and project an image of transparency or competence. At the same time, behavioral agency theory suggests that narcissistic executives may favor arrangements that reduce personal accountability risks, including closer or more relational auditor ties that they believe can soften scrutiny or provide leverage in negotiations (Wiseman and Gomez-Mejia, 1998). Together, these perspectives help explain why proximity decisions may reflect psychological motives even when efficiency considerations are held constant.
These effects are likely to be more pronounced in emerging economies, where governance institutions often provide fewer constraints on executive behavior. Weak board oversight, concentrated ownership, and relationship-based contracting increase the capacity of dominant executives to influence auditor appointment and interaction processes directly. In such settings, narcissistic traits can shape decisions regarding the selection of proximate auditors who are more accessible, more susceptible to interpersonal influence, or better positioned to reinforce executives' reputational objectives within local networks.
Similar dynamics apply to audit fees. The literature shows that narcissistic executives contribute to riskier reporting environments through overconfidence, earnings manipulation, and governance override (Ham et al., 2017; Judd et al., 2017). Auditors, sensitive to these behaviors, respond with heightened skepticism and expanded audit effort, which increases audit fees. However, beyond risk considerations, narcissistic executives may strategically select high-reputation or resource-intensive auditors to enhance their own prestige, aligning with their self-enhancement motives. Impression management theory suggests that such choices allow narcissistic leaders to associate themselves with prestigious external validators, even at higher cost. Thus, audit pricing becomes both a reflection of auditors' risk responses and executives' symbolic preferences.
4. Empirical literature review and hypothesis development
4.1 Executive narcissism and auditor-client proximity
The literature on auditor–client proximity reveals a central debate. On one hand, proximity can improve audit quality by reducing information asymmetry and giving auditors better access to client information (Choi et al., 2012; Francis et al., 2022; Harymawan et al., 2022). On the other hand, close geographic ties may weaken independence by fostering familiarity, social connections, or undue influence (Beck et al., 2019; Lennox and Li, 2012; Rahaman et al., 2025). This contradiction appears clearly in prior findings. Some studies show that proximity helps auditors detect risks and reduces earnings management (Zhang, 2020), while others, especially in emerging markets like Bangladesh, find that proximity can lower audit quality by weakening auditor independence (Rahaman et al., 2022, 2025). These contrasting results emphasize the importance of contextual factors such as regulatory strength and cultural norms in shaping how proximity affects audits. However, most existing studies still concentrate on structural factors like market competition and rarely consider how executives' behavioral traits may also influence auditor selection (DeAngelo, 1981; Simunic, 1980).
Upper echelons theory suggests that executives' psychological traits shape strategic decisions, including their interactions with auditors (Chatterjee and Hambrick, 2007; Hambrick and Mason, 1984). Narcissistic executives are likely to exert greater influence on governance mechanisms, including auditor selection (Cohen et al., 2010; Ouyang et al., 2015). Behavioral agency theory adds that executives respond not only to organizational risks but also to personal risks related to reputation and status (Wiseman and Gomez-Mejia, 1998). In this regard, narcissistic leaders may prefer proximate auditors because proximity offers more opportunities to influence audit decisions, reduce perceived threats, or engage in opinion shopping (Kooti, 2024; Lennox and Li, 2012).
Impression management theory also helps explain these dynamics, as narcissistic leaders often try to shape how others perceive them (Bolino et al., 2008). In this context, they may prefer nearby auditors because choosing a local audit firm can signal legitimacy and strong ties within the business community. This is an important factor in collectivist environments like Bangladesh, where social networks strongly influence business practices (Kim and Anderson, 2024; Nurunnabi, 2018; Qi et al., 2017). At the same time, some argue that proximity may also support transparency, since closer auditors can conduct more thorough and timely audits that reduce agency problems (Fan and Wong, 2005). Recent research further illustrates these dynamics. Some studies find that narcissistic traits within audit committees are associated with lower disclosure quality. For example, audit committee chairs with stronger narcissistic tendencies often reduce transparency. Other work shows that antagonistic narcissism can weaken auditor skepticism, which increases risks to auditor independence (Kaszak et al., 2025).
A significant gap remains in understanding how executive narcissism shapes auditor–client proximity, especially in emerging economies where weak governance allows personality-driven influences to play a larger role (Rahaman et al., 2022). There exist some studies on executive traits but these studies often treat leaders as similar in their decision-making and thus, offer limited insight into how behavioral biases may intensify familiarity threats or influence auditor independence. This study addresses this gap by examining how narcissistic executives affect proximity choices in Bangladesh's family-controlled and relationship-based business environment. Given the behavioral tendencies of narcissistic leaders and the relational nature of emerging-market auditing environments, it is reasonable to expect such executives to prefer geographically proximate auditors. Therefore, the following hypothesis has been developed:
There is a positive association between executive narcissism and auditor-client geographic proximity.
4.2 Executive narcissism and audit fees
Audit fees are primarily determined by auditors' assessments of client risk, the competitiveness of the audit market, and the bargaining dynamics between the auditor and the client. In their risk assessment, auditors evaluate management competence, the “tone at the top,” and the susceptibility of financial reporting to misstatements (Duellman et al., 2015; Hammersley et al., 2011; Patelli and Pedrini, 2015; Schmidt, 2014). A higher perceived risk necessitates more extensive audit procedures and evidence gathering, which in turn increases audit costs (Simunic, 1980).
A central question in recent research is whether executive traits contribute to higher risk. Narcissistic leaders often increase inherent risk through overconfident decisions such as aggressive earnings management or tax avoidance (He et al., 2020; Mitra et al., 2019; Olsen and Stekelberg, 2016; Zhu and Chen, 2014). They may also raise control risk by overriding internal procedures or weakening oversight systems (Campbell et al., 2011; Ham et al., 2017; Resick et al., 2009). Most empirical studies find that these behaviors lead to higher audit fees, reflecting auditors' need for more effort and caution (Judd et al., 2017; Xiang and Song, 2021). However, some research shows that narcissistic executives may occasionally encourage faster corrective actions (Campbell and Green, 2008; Shen et al., 2024), and strong governance mechanisms such as independent audit committees can limit certain risk-enhancing behaviors, including tax aggressiveness. Recent work continues to highlight the risks associated with narcissism, linking it to insolvency concerns in banks (Khanchel et al., 2025) and opportunistic insider trading (Jiang et al., 2025).
In addition to risk-related factors, impression management can also influence audit fees. Narcissistic executives often prefer well-known or high-profile auditors because these affiliations enhance their legitimacy and signal competence to outside stakeholders (Xiang and Song, 2021). This reflects their tendency to seek external validation. As a result, higher audit fees may arise not only from auditors' responses to risk but also from executives' preference for more prestigious audit services. Behavioral agency theory further suggests that narcissistic leaders may accept higher organizational costs if doing so helps protect their personal reputation, especially in settings where monitoring is weak (Ham et al., 2025).
Existing research has seldom integrated these behavioral mechanisms with economic models of audit pricing, and almost no empirical evidence exists from emerging economies where institutional constraints are weaker and executive discretion is greater. The present study addresses this gap by examining whether narcissistic executives are associated with higher audit fees in Bangladesh, a setting where auditors must navigate both client risk and the interpersonal dynamics of executive leadership. Building on the theoretical arguments described above, we hypothesize:
There is a positive association between executive narcissism and external audit fees.
5. Research design
5.1 Data and sample
Our dataset covers all companies listed on the Dhaka Stock Exchange (DSE) from 2018 to 2021. The year 2018 marks a transformative period with significant reforms in auditing practices, particularly the restructuring of audit reports and the introduction of KAM disclosures (Rahaman and Karim, 2023), while 2021 represents the most recent year of data collection. Data were meticulously sourced from the annual reports of these listed entities to ensure robustness and reliability. Panel A of Table 1 presents the sample selection procedure, detailing the step-by-step process from the initial population to the final sample. The initial dataset included 915 firm-year observations. However, we excluded 57 observations due to the non-availability of the annual reports, 43 observations due to unavailability of narcissism data, and an additional 379 observations due to missing control variables. Consequently, our final analytical sample comprises 436 firm-years, capturing over 60% of the total market capitalization.
Sample design and sample distribution
| Panel A: sample selection | Firm-year observations (N) |
|---|---|
| Initial sample from dhaka stock exchange listed firms (2018–2021) | 915 |
| Less- Observations excluded due to unavailability of annual reports | (57) |
| Less- Observations excluded due to unavailability of narcissism data | (43) |
| Less- Observations excluded due to missing control variables | (379) |
| Final sample | 436 |
| Panel A: sample selection | Firm-year observations (N) |
|---|---|
| Initial sample from dhaka stock exchange listed firms (2018–2021) | 915 |
| Less- Observations excluded due to unavailability of annual reports | (57) |
| Less- Observations excluded due to unavailability of narcissism data | (43) |
| Less- Observations excluded due to missing control variables | (379) |
| Final sample | 436 |
| Panel B: Sample distribution by industry | N | % |
|---|---|---|
| Bank | 56 | 12.84 |
| Cement | 16 | 3.67 |
| Ceramic | 9 | 2.06 |
| Engineering | 63 | 14.45 |
| Financial Institutions | 38 | 8.72 |
| Food and Allied | 13 | 2.98 |
| Fuel and Power | 41 | 9.40 |
| Insurance | 60 | 13.76 |
| IT sector | 14 | 3.21 |
| Pharmaceuticals and Chemicals | 41 | 9.40 |
| Textile | 71 | 16.28 |
| Others | 14 | 3.21 |
| Total | 436 | 100 |
| Panel B: Sample distribution by industry | N | % |
|---|---|---|
| Bank | 56 | 12.84 |
| Cement | 16 | 3.67 |
| Ceramic | 9 | 2.06 |
| Engineering | 63 | 14.45 |
| Financial Institutions | 38 | 8.72 |
| Food and Allied | 13 | 2.98 |
| Fuel and Power | 41 | 9.40 |
| Insurance | 60 | 13.76 |
| IT sector | 14 | 3.21 |
| Pharmaceuticals and Chemicals | 41 | 9.40 |
| Textile | 71 | 16.28 |
| Others | 14 | 3.21 |
| Total | 436 | 100 |
| Panel C: Sample distribution by year | N | % |
|---|---|---|
| T1 | 169 | 36.47 |
| T2 | 146 | 33.49 |
| T3 | 131 | 30.05 |
| Total | 436 | 100 |
| Panel C: Sample distribution by year | N | % |
|---|---|---|
| T1 | 169 | 36.47 |
| T2 | 146 | 33.49 |
| T3 | 131 | 30.05 |
| Total | 436 | 100 |
Note(s): Table 1 presents the sample design and sample distribution. Panel A articulates the sample selection. Panel B shows industry-wise sample distributions. Panel C splits the sample by year. We use the data from the first three years post the recent and significant audit reform to avoid the effect of pre-reform trends. Some firms adopted the new reporting model for the first time in their December year-end financial statements in 2018, while other firms adopted it in their June year-end financial statements in 2019. Therefore, T1 denotes the first year of the new extended auditor reporting (2018–2019). Similarly, T2 and T3 denote the 2nd (2019–2020) and 3rd year post new reporting model (2021–2021)
We acknowledge that nearly 50% of observations were missing due to data unavailability, which may raise concerns about sample selection. To address this, we incorporate industry and year fixed effects. Panel B of Table 1 shows the sample distribution across various industry sectors, with textiles representing the largest proportion at 16.28%, followed by engineering (14.45%), insurance (13.76%), and banking (12.84%). Collectively, the remaining sectors account for approximately 40% of the sample. Panel C of Table 1 presents the annual distribution, with 169 observations for 2018–2019, 146 for 2019–2020, and 131 for 2020–2021.
5.2 Dependent variable
This study examines the impact of executive narcissism on two key audit outcomes: the choice between local and distant auditors and the determination of audit fees. Accordingly, two dependent variables are employed to test the corresponding hypotheses: auditor–client geographic proximity and audit fees. Auditor–client proximity is operationalized using both geographic distance (in kilometers) and travel time (in minutes) between the head offices of auditor and client, consistent with prior studies (Beck et al., 2019; Choi et al., 2012; Dong et al., 2018; Francis et al., 2022). Audit fees are measured as the natural logarithm of audit fees, following standard practice in the auditing literature.
5.3 Independent variable
The key independent variable in this study is executive narcissism, which reflects the extent of self-admiration and self-importance exhibited by top executives. Since executives are generally unwilling to disclose psychological traits through surveys, this study relies on observable proxies—signature size and photograph prominence—to capture narcissism (Chatterjee and Hambrick, 2007; Ham et al., 2017; Rijsenbilt and Commandeur, 2013; Agnihotri and Bhattacharya, 2019; Azevedo et al., 2024; Li et al., 2025; Rajabalizadeh and Schadewitz, 2025). These proxies are widely used and validated in the literature across different cultural contexts, supporting their applicability and validity in the Bangladeshi setting. Signature size is calculated as the rectangular area of the executive's signature (length × width in cm) divided by the number of letters in the executive's name. This standardized measure reflects exaggerated self-expression in written form (Ham et al., 2017, 2018). Photograph prominence is assessed on a scale from 0 to 4, following prior literature (Chatterjee and Hambrick, 2007; Rijsenbilt and Commandeur, 2013; Zhu and Chen, 2014). A score of 0 indicates no photograph, 1 represents a group photo with other executives, 2 corresponds to a small image, 3 denotes half-page coverage, and 4 reflects full-page coverage in the annual report. Because top executives exercise strict control over annual report content, narcissistic executives are more likely to ensure prominent placement of their photographs to reinforce their leadership importance (Chatterjee and Hambrick, 2007, 2011).
Consistent with prior research (Cragun et al., 2019; Ham et al., 2017; Junge et al., 2024; Rijsenbilt and Commandeur, 2013), both measures are standardized and aggregated to form a single narcissism index for both the Chair and CEO. This composite approach captures both written and visual dimensions of narcissistic expression, providing a more comprehensive measure of executive narcissism than relying on a single indicator.
5.4 Control variables
In addition to the key independent variable, the model incorporates several control variables to account for governance, audit, and firm-level factors that may influence auditor choice and audit fees. Chair and CEO gender are included as prior research suggests gender diversity in leadership can affect governance quality and auditing practices (Adams and Ferreira, 2009; Harjoto et al., 2015). Board-level diversity and monitoring strength are further controlled for through the female director ratio (FDR) and the independent director ratio (IDR), consistent with evidence that gender representation and board independence enhance oversight and audit outcomes (Lai et al., 2017). Audit committee size is included to capture monitoring capacity, as larger audit committees may strengthen internal governance and audit fees (Karim et al., 2020; Rahaman and Karim, 2023). Audit-specific characteristics are also controlled for: audit report lag, measured as the number of days between fiscal year-end and the audit report date, reflects timeliness of audit completion; audit tenure accounts for the length of the auditor–client relationship, which may affect auditor independence and reporting quality; and a Big 4 indicator variable captures the effect of auditor reputation and quality (Dong et al., 2018; Rahaman and Karim, 2023). Firm characteristics include firm size, which influences audit complexity and fees; firm age, as older firms may have more stable reporting environments; profitability (ROA), controlling for performance-related effects on audit risk; and leverage, reflecting financial risk and the demand for monitoring (Dong et al., 2018; Judd et al., 2017; Waresul Karim and Hasan, 2012). Finally, a fiscal year-end indicator is added to capture seasonality effects, while industry and year fixed effects control unobserved heterogeneity across sectors and time (Rahaman and Karim, 2023). Collectively, these controls align with prior auditing and governance literature and strengthen the validity of the estimated relationship between executive narcissism and audit outcomes.
5.5 Regression model
We use the following regression specifications to test our hypotheses, following prior studies (Dong et al., 2018; Judd et al., 2017; Lai et al., 2017).
Equation (1) examines the effect of Chair or CEO narcissism on auditor–client geographic proximity, while Equation (2) evaluates its implications for audit fees. To provide a comprehensive understanding, the definitions and measurements of all variables used in these models are detailed in Appendix. The study employs ordinary least squares (OLS) regression to test the hypotheses across all models. In addition, entropy balancing and two-stage least squares (2SLS) regression are utilized to address potential endogeneity concerns and enhance the robustness of the results.
6. Empirical results and discussion
6.1 Descriptive statistics
Table 2 provides descriptive statistics for the study variables based on 436 observations, highlighting their central tendencies and variability. Auditor-client geographic distance measures have means of 1.469 kilometers and 2.763 min, respectively, with wide ranges, while audit fees average 6.071 with a standard deviation of 0.91. These statistics indicate variations across the firms in terms of appointments of local or distant auditors and payment of audit fees. Chair and CEO narcissism scores are relatively low, averaging 0.118 and 0.108, respectively, with subcomponents like signature size and picture prominence showing similar trends. Gender diversity is limited, as reflected by mean values of 0.154 for Chair gender and 0.050 for CEO gender. Governance metrics, including female and independent director ratios, average 0.161 and 0.232 respectively, while audit characteristics show an average lag of 122.64 days (log = 4.757) and tenure of 2.284 years, with Big 4 auditors involved in 20.4% of cases. Firm-level variables such as size, age, and leverage average 16.073, 3.197, and 0.57, respectively, while return on assets (ROA) remains modest at 3%. The year-end variable indicates a balanced distribution of fiscal year observations. These statistics highlight the extensive variability across the firm, auditor, and top executives' characteristics used in the sample, which is essential for a comprehensive analysis of the implications of top executives' narcissism on the appointment and remuneration of auditors.
Descriptive statistics
| Variables | Observations | Mean | Standard deviation | Minimum | Maximum |
|---|---|---|---|---|---|
| Dependent variable | |||||
| DisKM (actual) | 436 | 6.76 | 7.72 | 0.03 | 50.00 |
| DisMN (actual) | 436 | 22.28 | 20.10 | 1.00 | 120.00 |
| DisKM (log) | 436 | 1.47 | 1.02 | −3.51 | 3.91 |
| DisMN (log) | 436 | 2.76 | 0.87 | 0.00 | 4.79 |
| Aud_Fee | 436 | 6.07 | 0.91 | 2.30 | 8.94 |
| Independent variable | |||||
| Chair_Narc | 436 | 0.12 | 0.10 | 0.00 | 0.63 |
| CEO_Narc | 436 | 0.11 | 0.11 | 0.00 | 0.60 |
| Chair_Narc_Sig | 436 | 0.02 | 0.02 | 0.00 | 0.08 |
| CEO_Narc_Sig | 436 | 0.02 | 0.02 | 0.00 | 0.06 |
| Chair_Narc_Pic | 436 | 1.78 | 1.41 | 0.00 | 4.00 |
| CEO_Narc_Pic | 436 | 1.66 | 1.38 | 0.00 | 4.00 |
| Control variables | |||||
| Chair_Gen | 436 | 0.15 | 0.36 | 0.00 | 1.00 |
| CEO_Gen | 436 | 0.05 | 0.22 | 0.00 | 1.00 |
| FDR | 436 | 0.16 | 0.14 | 0.00 | 0.62 |
| IDR | 436 | 0.23 | 0.10 | 0.00 | 0.83 |
| AC_Size (log) | 436 | 1.39 | 0.27 | 0.69 | 3.76 |
| Aud_Lag (log) | 436 | 4.76 | 0.34 | 3.29 | 5.70 |
| Aud_Ten | 436 | 2.28 | 1.09 | 1.00 | 5.00 |
| Big4 | 436 | 0.20 | 0.40 | 0.00 | 1.00 |
| Firm_Size (log) | 436 | 16.07 | 1.91 | 12.50 | 20.08 |
| Firm_Age (log) | 436 | 3.20 | 0.41 | 1.95 | 4.13 |
| ROA | 436 | 0.03 | 0.04 | −0.14 | 0.24 |
| Lev | 436 | 0.57 | 0.30 | 0.00 | 1.83 |
| Yearend | 436 | 0.628 | 0.483 | 0.00 | 1.00 |
| Variables | Observations | Mean | Standard deviation | Minimum | Maximum |
|---|---|---|---|---|---|
| Dependent variable | |||||
| DisKM (actual) | 436 | 6.76 | 7.72 | 0.03 | 50.00 |
| DisMN (actual) | 436 | 22.28 | 20.10 | 1.00 | 120.00 |
| DisKM (log) | 436 | 1.47 | 1.02 | −3.51 | 3.91 |
| DisMN (log) | 436 | 2.76 | 0.87 | 0.00 | 4.79 |
| Aud_Fee | 436 | 6.07 | 0.91 | 2.30 | 8.94 |
| Independent variable | |||||
| Chair_Narc | 436 | 0.12 | 0.10 | 0.00 | 0.63 |
| CEO_Narc | 436 | 0.11 | 0.11 | 0.00 | 0.60 |
| Chair_Narc_Sig | 436 | 0.02 | 0.02 | 0.00 | 0.08 |
| CEO_Narc_Sig | 436 | 0.02 | 0.02 | 0.00 | 0.06 |
| Chair_Narc_Pic | 436 | 1.78 | 1.41 | 0.00 | 4.00 |
| CEO_Narc_Pic | 436 | 1.66 | 1.38 | 0.00 | 4.00 |
| Control variables | |||||
| Chair_Gen | 436 | 0.15 | 0.36 | 0.00 | 1.00 |
| CEO_Gen | 436 | 0.05 | 0.22 | 0.00 | 1.00 |
| FDR | 436 | 0.16 | 0.14 | 0.00 | 0.62 |
| IDR | 436 | 0.23 | 0.10 | 0.00 | 0.83 |
| AC_Size (log) | 436 | 1.39 | 0.27 | 0.69 | 3.76 |
| Aud_Lag (log) | 436 | 4.76 | 0.34 | 3.29 | 5.70 |
| Aud_Ten | 436 | 2.28 | 1.09 | 1.00 | 5.00 |
| Big4 | 436 | 0.20 | 0.40 | 0.00 | 1.00 |
| Firm_Size (log) | 436 | 16.07 | 1.91 | 12.50 | 20.08 |
| Firm_Age (log) | 436 | 3.20 | 0.41 | 1.95 | 4.13 |
| ROA | 436 | 0.03 | 0.04 | −0.14 | 0.24 |
| Lev | 436 | 0.57 | 0.30 | 0.00 | 1.83 |
| Yearend | 436 | 0.628 | 0.483 | 0.00 | 1.00 |
6.2 Correlation analysis
Table 3 demonstrates the correlation matrix, focusing on the relationship between executive narcissism measures, auditor-client distance, and audit fees. Executive narcissism metrics, including combined and individual measures based on signature and image size, are negatively and significantly (mostly at 1% level) correlated with both measures of auditor-client distance, indicating that higher narcissistic executives appoint local auditors. Similarly, these narcissism measures are positively and significantly (at 1% level across all measures) linked to audit fees, suggesting that firms led by more narcissistic executives may pay higher audit fees compared to their counterparts. This pattern could reflect auditors' strategic responses to executive characteristics, possibly related to perceived influence or negotiation dynamics (Olsen and Stekelberg, 2016).
Pairwise correlation
| Variables | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 |
|---|---|---|---|---|---|---|---|---|---|---|---|
| (1) DisKM | 1 | ||||||||||
| (2) DisMN | 0.92*** | 1 | |||||||||
| (3) Aud_Fee | −0.14*** | −0.13*** | 1 | ||||||||
| (4) Chair_Narc | −0.19*** | −0.23*** | 0.25*** | 1 | |||||||
| (5) CEO_Narc | −0.12** | −0.20*** | 0.28*** | 0.60*** | 1 | ||||||
| (6) Chair_Narc_Sig | −0.19*** | −0.21*** | 0.20*** | 0.77*** | 0.35*** | 1 | |||||
| (7) CEO_Narc_Sig | −0.14*** | −0.18*** | 0.31*** | 0.42*** | 0.78*** | 0.49*** | 1 | ||||
| (8) Chair_Narc_Pic | −0.12** | −0.14*** | 0.34*** | 0.62*** | 0.56*** | 0.26*** | 0.32*** | 1 | |||
| (9) CEO_Narc_Pic | −0.06 | −0.11** | 0.26*** | 0.50*** | 0.62*** | 0.22*** | 0.29*** | 0.87*** | 1 | ||
| (10) Chair_Gen | 0.09* | 0.11** | −0.10** | −0.17*** | −0.05 | −0.19*** | −0.10** | −0.07 | 0.01 | 1 | |
| (11) CEO_Gen | 0.05 | 0.03 | −0.09* | 0.03 | 0 | 0.08* | −0.03 | −0.08* | −0.08* | −0.04 | 1 |
| (12) FDR | 0 | −0.05 | −0.08 | 0.06 | 0.11** | 0.09* | 0.09* | 0.02 | 0.03 | 0.22*** | 0.22*** |
| (13) IDR | 0.27*** | 0.27*** | −0.16*** | −0.18*** | −0.18*** | −0.16*** | −0.20*** | −0.14*** | −0.10** | 0.18*** | 0.25*** |
| (14) AC_Size | −0.04 | −0.10** | 0.10** | 0.06 | 0.12** | 0.03 | 0.14*** | 0.10** | 0.09* | −0.15*** | −0.07 |
| (15) Big4 | −0.15*** | −0.14*** | 0.38*** | 0.10** | 0.05 | 0.01 | 0.07 | 0.19*** | 0.09* | −0.11** | 0.01 |
| (16) Aud_Ten | 0 | 0.06 | 0.17*** | 0.04 | 0.08* | 0.09* | 0.11** | 0.07 | 0.11** | 0.02 | 0 |
| (17) Aud_Lag | −0.03 | −0.03 | −0.18*** | −0.11** | −0.18*** | 0.01 | −0.12** | −0.19*** | −0.19*** | 0.05 | −0.06 |
| (18) Firm_Size | −0.13*** | −0.11** | 0.60*** | 0.21*** | 0.18*** | 0.16*** | 0.22*** | 0.33*** | 0.25*** | −0.17*** | −0.12** |
| (19) Firm_Age | −0.11** | −0.13*** | 0.03 | 0.02 | 0.02 | 0.09* | 0.13*** | −0.06 | −0.10* | −0.05 | −0.02 |
| (20) Lev | −0.11** | −0.10** | 0.24*** | 0.05 | 0 | 0.05 | 0.06 | 0.15*** | 0.09* | −0.17*** | −0.08* |
| (21) ROA | −0.04 | −0.03 | 0.08 | −0.03 | −0.02 | −0.04 | 0.05 | −0.07 | −0.11** | −0.01 | 0 |
| (22) YearEnd | 0.25*** | 0.27*** | −0.28*** | −0.21*** | −0.24*** | −0.19*** | −0.28*** | −0.23*** | −0.19*** | 0.22*** | 0.09* |
| Variables | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 |
|---|---|---|---|---|---|---|---|---|---|---|---|
| (1) DisKM | 1 | ||||||||||
| (2) DisMN | 0.92*** | 1 | |||||||||
| (3) Aud_Fee | −0.14*** | −0.13*** | 1 | ||||||||
| (4) Chair_Narc | −0.19*** | −0.23*** | 0.25*** | 1 | |||||||
| (5) CEO_Narc | −0.12** | −0.20*** | 0.28*** | 0.60*** | 1 | ||||||
| (6) Chair_Narc_Sig | −0.19*** | −0.21*** | 0.20*** | 0.77*** | 0.35*** | 1 | |||||
| (7) CEO_Narc_Sig | −0.14*** | −0.18*** | 0.31*** | 0.42*** | 0.78*** | 0.49*** | 1 | ||||
| (8) Chair_Narc_Pic | −0.12** | −0.14*** | 0.34*** | 0.62*** | 0.56*** | 0.26*** | 0.32*** | 1 | |||
| (9) CEO_Narc_Pic | −0.06 | −0.11** | 0.26*** | 0.50*** | 0.62*** | 0.22*** | 0.29*** | 0.87*** | 1 | ||
| (10) Chair_Gen | 0.09* | 0.11** | −0.10** | −0.17*** | −0.05 | −0.19*** | −0.10** | −0.07 | 0.01 | 1 | |
| (11) CEO_Gen | 0.05 | 0.03 | −0.09* | 0.03 | 0 | 0.08* | −0.03 | −0.08* | −0.08* | −0.04 | 1 |
| (12) FDR | 0 | −0.05 | −0.08 | 0.06 | 0.11** | 0.09* | 0.09* | 0.02 | 0.03 | 0.22*** | 0.22*** |
| (13) IDR | 0.27*** | 0.27*** | −0.16*** | −0.18*** | −0.18*** | −0.16*** | −0.20*** | −0.14*** | −0.10** | 0.18*** | 0.25*** |
| (14) AC_Size | −0.04 | −0.10** | 0.10** | 0.06 | 0.12** | 0.03 | 0.14*** | 0.10** | 0.09* | −0.15*** | −0.07 |
| (15) Big4 | −0.15*** | −0.14*** | 0.38*** | 0.10** | 0.05 | 0.01 | 0.07 | 0.19*** | 0.09* | −0.11** | 0.01 |
| (16) Aud_Ten | 0 | 0.06 | 0.17*** | 0.04 | 0.08* | 0.09* | 0.11** | 0.07 | 0.11** | 0.02 | 0 |
| (17) Aud_Lag | −0.03 | −0.03 | −0.18*** | −0.11** | −0.18*** | 0.01 | −0.12** | −0.19*** | −0.19*** | 0.05 | −0.06 |
| (18) Firm_Size | −0.13*** | −0.11** | 0.60*** | 0.21*** | 0.18*** | 0.16*** | 0.22*** | 0.33*** | 0.25*** | −0.17*** | −0.12** |
| (19) Firm_Age | −0.11** | −0.13*** | 0.03 | 0.02 | 0.02 | 0.09* | 0.13*** | −0.06 | −0.10* | −0.05 | −0.02 |
| (20) Lev | −0.11** | −0.10** | 0.24*** | 0.05 | 0 | 0.05 | 0.06 | 0.15*** | 0.09* | −0.17*** | −0.08* |
| (21) ROA | −0.04 | −0.03 | 0.08 | −0.03 | −0.02 | −0.04 | 0.05 | −0.07 | −0.11** | −0.01 | 0 |
| (22) YearEnd | 0.25*** | 0.27*** | −0.28*** | −0.21*** | −0.24*** | −0.19*** | −0.28*** | −0.23*** | −0.19*** | 0.22*** | 0.09* |
| Variables | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 |
|---|---|---|---|---|---|---|---|---|---|---|---|
| (1) DisKM | |||||||||||
| (2) DisMN | |||||||||||
| (3) Aud_Fee | |||||||||||
| (4) Chair_Narc | |||||||||||
| (5) CEO_Narc | |||||||||||
| (6) Chair_Narc_Sig | |||||||||||
| (7) CEO_Narc_Sig | |||||||||||
| (8) Chair_Narc_Pic | |||||||||||
| (9) CEO_Narc_Pic | |||||||||||
| (10) Chair_Gen | |||||||||||
| (11) CEO_Gen | |||||||||||
| (12) FDR | 1 | ||||||||||
| (13) IDR | −0.03 | 1 | |||||||||
| (14) AC_Size | 0.01 | −0.16*** | 1 | ||||||||
| (15) Big4 | −0.02 | −0.12** | 0.13*** | 1 | |||||||
| (16) Aud_Ten | −0.01 | −0.02 | 0.04 | 0 | 1 | ||||||
| (17) Aud_Lag | −0.03 | 0.04 | −0.04 | −0.26*** | 0.02 | 1 | |||||
| (18) Firm_Size | −0.22*** | −0.16*** | 0.04 | 0.33*** | 0.18*** | 0.02 | 1 | ||||
| (19) Firm_Age | 0.11** | −0.14*** | 0.01 | 0.06 | 0.05 | 0.18*** | −0.10** | 1 | |||
| (20) Lev | −0.17*** | −0.07 | −0.03 | 0.22*** | 0.03 | 0.05 | −0.05 | 0.15*** | 1 | ||
| (21) ROA | 0.06 | −0.04 | −0.05 | 0.08* | −0.01 | 0.03 | −0.03 | 0.53*** | 0.18*** | 1 | |
| (22) YearEnd | −0.02 | 0.38*** | −0.19*** | −0.16*** | −0.07 | −0.01 | −0.14*** | −0.11** | −0.10** | −0.30*** | 1 |
| Variables | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 |
|---|---|---|---|---|---|---|---|---|---|---|---|
| (1) DisKM | |||||||||||
| (2) DisMN | |||||||||||
| (3) Aud_Fee | |||||||||||
| (4) Chair_Narc | |||||||||||
| (5) CEO_Narc | |||||||||||
| (6) Chair_Narc_Sig | |||||||||||
| (7) CEO_Narc_Sig | |||||||||||
| (8) Chair_Narc_Pic | |||||||||||
| (9) CEO_Narc_Pic | |||||||||||
| (10) Chair_Gen | |||||||||||
| (11) CEO_Gen | |||||||||||
| (12) FDR | 1 | ||||||||||
| (13) IDR | −0.03 | 1 | |||||||||
| (14) AC_Size | 0.01 | −0.16*** | 1 | ||||||||
| (15) Big4 | −0.02 | −0.12** | 0.13*** | 1 | |||||||
| (16) Aud_Ten | −0.01 | −0.02 | 0.04 | 0 | 1 | ||||||
| (17) Aud_Lag | −0.03 | 0.04 | −0.04 | −0.26*** | 0.02 | 1 | |||||
| (18) Firm_Size | −0.22*** | −0.16*** | 0.04 | 0.33*** | 0.18*** | 0.02 | 1 | ||||
| (19) Firm_Age | 0.11** | −0.14*** | 0.01 | 0.06 | 0.05 | 0.18*** | −0.10** | 1 | |||
| (20) Lev | −0.17*** | −0.07 | −0.03 | 0.22*** | 0.03 | 0.05 | −0.05 | 0.15*** | 1 | ||
| (21) ROA | 0.06 | −0.04 | −0.05 | 0.08* | −0.01 | 0.03 | −0.03 | 0.53*** | 0.18*** | 1 | |
| (22) YearEnd | −0.02 | 0.38*** | −0.19*** | −0.16*** | −0.07 | −0.01 | −0.14*** | −0.11** | −0.10** | −0.30*** | 1 |
Importantly, no multicollinearity is observed among the control variables, as none of the pairwise correlations exceed commonly accepted thresholds (Toubiana and Maruenda, 2021), ensuring the robustness of the estimated relationships. While high correlations are present between the alternative measures of auditor-client geographic distance (coefficient = 0.92) and executive narcissism measures (maximum coefficient = 0.77), the maximum correlation among all other variables is 0.60, which remains below conventional thresholds, indicating that multicollinearity is not a concern for the core analysis.
6.3 Baseline results
Table 4 presents the baseline results, exploring the impact of executive narcissism on auditor-client distance and audit fees. The coefficients for both Chair and CEO narcissism measures are consistently negative and statistically significant (at 1% level mostly) with both measures of auditor-client distance (kilometers and minutes), indicating that higher levels of narcissism in executives correspond to closer auditor-client proximity. Similarly, these narcissism measures positively and significantly (at 1% level) associated with audit fees [1]. This suggests that auditors may demand higher fees when dealing with more narcissistic executives, possibly reflecting perceived risk or increased negotiation difficulty. The findings of this study are highly congruent with prior literature (Choi et al., 2012; Olsen et al., 2014; Judd et al., 2017; Dong et al., 2018). Thus, we can accept both of our hypotheses.
| Variables | DV=Auditor-client distance | DV=Audit fee | ||||
|---|---|---|---|---|---|---|
| DisKM | DisMN | DisKM | DisMN | Aud_Fee | Aud_Fee | |
| Independent variable | ||||||
| Chair_Narc | −1.343*** | −1.408*** | 0.745** | |||
| (−2.741) | (−3.403) | (2.157) | ||||
| CEO_Narc | −0.784* | −1.193*** | 1.003*** | |||
| (−1.569) | (−2.832) | (2.881) | ||||
| Control variables | ||||||
| Chair_Gen | −0.033 | −0.001 | 0.036 | 0.069 | 0.033 | −0.002 |
| (−0.233) | (−0.008) | (0.253) | (0.588) | (0.333) | (−0.022) | |
| CEO_Gen | −0.159 | 0.027 | −0.184 | 0.005 | −0.317* | −0.316* |
| (−0.636) | (0.128) | (−0.731) | (0.025) | (−1.804) | (−1.806) | |
| FDR | 0.415 | 0.052 | 0.344 | 0.019 | 0.350 | 0.371 |
| (1.070) | (0.158) | (0.877) | (0.059) | (1.280) | (1.358) | |
| IDR | 1.541*** | 1.357*** | 1.556*** | 1.349*** | −0.187 | −0.212 |
| (2.823) | (2.943) | (2.805) | (2.888) | (−0.486) | (−0.549) | |
| AC_Size | 0.155 | 0.004 | 0.145 | 0.005 | 0.090 | 0.065 |
| (0.848) | (0.024) | (0.784) | (0.034) | (0.698) | (0.504) | |
| BIG4 | −0.268** | −0.243** | −0.300** | −0.275** | 0.394*** | 0.394*** |
| (−2.063) | (−2.210) | (−2.264) | (−2.462) | (4.300) | (4.271) | |
| Aud_Ten | 0.006 | −0.000 | 0.003 | −0.002 | 0.036 | 0.041 |
| (0.093) | (−0.005) | (0.047) | (−0.031) | (0.749) | (0.880) | |
| Aud_Lag | −0.219 | −0.178 | −0.249 | −0.228* | −0.214* | −0.171 |
| (−1.411) | (−1.357) | (−1.566) | (−1.702) | (−1.955) | (−1.544) | |
| Firm_Size | 0.045 | 0.059 | 0.031 | 0.055 | 0.235*** | 0.229*** |
| (0.896) | (1.397) | (0.629) | (1.305) | (6.697) | (6.588) | |
| Firm_Age | −0.204* | −0.189* | −0.221* | −0.212** | −0.233*** | −0.218** |
| (−1.702) | (−1.864) | (−1.810) | (−2.063) | (−2.754) | (−2.569) | |
| Lev | −0.232 | −0.144 | −0.265 | −0.188 | −0.067 | −0.105 |
| (−1.073) | (−0.790) | (−1.207) | (−1.019) | (−0.443) | (−0.689) | |
| ROA | 0.448 | 0.350 | 0.722 | 0.609 | 1.229 | 1.043 |
| (0.332) | (0.307) | (0.505) | (0.506) | (1.291) | (1.049) | |
| YearEnd | 0.544* | 0.284 | 0.591* | 0.324 | −0.488** | −0.522** |
| (1.673) | (1.036) | (1.808) | (1.178) | (−2.129) | (−2.294) | |
| Constant | 1.398 | 2.607** | 1.777 | 2.961*** | 4.504*** | 4.438*** |
| (1.065) | (2.351) | (1.340) | (2.651) | (4.864) | (4.809) | |
| Year_FE | YES | YES | YES | YES | YES | YES |
| Industry_FE | YES | YES | YES | YES | YES | YES |
| Observations | 436 | 436 | 436 | 436 | 436 | 436 |
| Adjusted_R2 | 0.173 | 0.182 | 0.160 | 0.183 | 0.486 | 0.494 |
| Variables | DV=Auditor-client distance | DV=Audit fee | ||||
|---|---|---|---|---|---|---|
| DisKM | DisMN | DisKM | DisMN | Aud_Fee | Aud_Fee | |
| Independent variable | ||||||
| Chair_Narc | −1.343*** | −1.408*** | 0.745** | |||
| (−2.741) | (−3.403) | (2.157) | ||||
| CEO_Narc | −0.784* | −1.193*** | 1.003*** | |||
| (−1.569) | (−2.832) | (2.881) | ||||
| Control variables | ||||||
| Chair_Gen | −0.033 | −0.001 | 0.036 | 0.069 | 0.033 | −0.002 |
| (−0.233) | (−0.008) | (0.253) | (0.588) | (0.333) | (−0.022) | |
| CEO_Gen | −0.159 | 0.027 | −0.184 | 0.005 | −0.317* | −0.316* |
| (−0.636) | (0.128) | (−0.731) | (0.025) | (−1.804) | (−1.806) | |
| FDR | 0.415 | 0.052 | 0.344 | 0.019 | 0.350 | 0.371 |
| (1.070) | (0.158) | (0.877) | (0.059) | (1.280) | (1.358) | |
| IDR | 1.541*** | 1.357*** | 1.556*** | 1.349*** | −0.187 | −0.212 |
| (2.823) | (2.943) | (2.805) | (2.888) | (−0.486) | (−0.549) | |
| AC_Size | 0.155 | 0.004 | 0.145 | 0.005 | 0.090 | 0.065 |
| (0.848) | (0.024) | (0.784) | (0.034) | (0.698) | (0.504) | |
| BIG4 | −0.268** | −0.243** | −0.300** | −0.275** | 0.394*** | 0.394*** |
| (−2.063) | (−2.210) | (−2.264) | (−2.462) | (4.300) | (4.271) | |
| Aud_Ten | 0.006 | −0.000 | 0.003 | −0.002 | 0.036 | 0.041 |
| (0.093) | (−0.005) | (0.047) | (−0.031) | (0.749) | (0.880) | |
| Aud_Lag | −0.219 | −0.178 | −0.249 | −0.228* | −0.214* | −0.171 |
| (−1.411) | (−1.357) | (−1.566) | (−1.702) | (−1.955) | (−1.544) | |
| Firm_Size | 0.045 | 0.059 | 0.031 | 0.055 | 0.235*** | 0.229*** |
| (0.896) | (1.397) | (0.629) | (1.305) | (6.697) | (6.588) | |
| Firm_Age | −0.204* | −0.189* | −0.221* | −0.212** | −0.233*** | −0.218** |
| (−1.702) | (−1.864) | (−1.810) | (−2.063) | (−2.754) | (−2.569) | |
| Lev | −0.232 | −0.144 | −0.265 | −0.188 | −0.067 | −0.105 |
| (−1.073) | (−0.790) | (−1.207) | (−1.019) | (−0.443) | (−0.689) | |
| ROA | 0.448 | 0.350 | 0.722 | 0.609 | 1.229 | 1.043 |
| (0.332) | (0.307) | (0.505) | (0.506) | (1.291) | (1.049) | |
| YearEnd | 0.544* | 0.284 | 0.591* | 0.324 | −0.488** | −0.522** |
| (1.673) | (1.036) | (1.808) | (1.178) | (−2.129) | (−2.294) | |
| Constant | 1.398 | 2.607** | 1.777 | 2.961*** | 4.504*** | 4.438*** |
| (1.065) | (2.351) | (1.340) | (2.651) | (4.864) | (4.809) | |
| Year_FE | YES | YES | YES | YES | YES | YES |
| Industry_FE | YES | YES | YES | YES | YES | YES |
| Observations | 436 | 436 | 436 | 436 | 436 | 436 |
| Adjusted_R2 | 0.173 | 0.182 | 0.160 | 0.183 | 0.486 | 0.494 |
Note(s): Table 4 reports the baseline regression results estimated using OLS. Coefficients are presented with t-statistics in parentheses. Variable definitions are provided in Appendix. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively. DV stands for dependent variable
These findings align with the lens of upper echelons theory, which posits that organizational outcomes are strongly shaped by the characteristics and behaviors of top executives (Cragun et al., 2019; Hambrick and Mason, 1984; Liu et al., 2022). Narcissistic executives, driven by their desire for dominance, validation, and control, may prefer geographically proximate and familiar auditors as a means of exerting influence, reducing perceived risks, or maintaining stricter oversight. This behavioral tendency is particularly pronounced in Bangladesh, where top executives often hold highly centralized power, thereby explaining the observed negative association between narcissism and auditor–client distance. Simultaneously, the positive relationship between executive narcissism and audit fees suggests that auditors perceive narcissistic executives as a source of increased audit risk (both inherent and control risks) or complexity (Olsen and Stekelberg, 2016; Xiang and Song, 2021). Narcissistic leaders' propensity for overconfidence (Navis and Ozbek, 2016), weak internal control system (Salehi et al., 2021), earnings management/fraud (Rijsenbilt and Commandeur, 2013; Buchholz et al., 2019), aggressive tax strategies (García-Meca et al., 2021), higher agency cost (Kim and Anderson, 2024), and potential lawsuit (O'Reilly et al., 2018) may lead auditors to charge premium fees as compensation for heightened audit challenges or reputational risk. Thus, we can conclude that narcissistic executives appoint local auditors and pay higher audit fees to address the complexities and risks associated with their leadership style.
The study has incorporated several control variables that help explain the relationships between executive narcissism and auditor-client distance and audit fees. Chair and CEO gender have no significant impact, likely reflecting weak board oversight and limited gender-based decision-making influence on audit practices in Bangladesh (Khan et al., 2015). Firm size is positively associated with both auditor-client distance and audit fees, indicating that larger firms tend to engage auditors from farther away and incur higher fees (Judd et al., 2017). Conversely, firm age negatively impacts both variables, suggesting older firms prefer recruiting local auditors by maintaining cost-effective auditing relationships. Independent directors are positively associated with both auditor-client distance and audit fees, implying that more independent boards may seek greater oversight and higher audit costs (Karim et al., 2020). The use of Big-4 auditors also shows a significant negative association with distance and a significant positive association with audit fees, indicating that these firms are more likely to be located closer to their clients while charging premium fees, reflecting their reputation, expertise, and the perceived value they bring to the audit process (Judd et al., 2017; Waresul Karim and Hasan, 2012). Audit tenure and audit lag, however, do not significantly affect the outcomes, and the fiscal year-end variable shows a negative association with audit fees, potentially reflecting seasonal trends. Overall, these control variables emphasize the complex interplay of factors influencing auditor-client proximity and audit fees.
6.4 Robustness test
Table 5 presents the results of robustness tests to confirm the consistency of the relationships between executive narcissism, auditor-client distance, and audit fees using alternative measures of narcissism. We solely use the signature size and picture size for both the CEO and chairman of the company to measure narcissism following prior literature (Chatterjee and Hambrick, 2007, 2011; Olsen et al., 2014; Ham et al., 2017, 2018). The findings reveal that both measures of narcissism for both CEO and chair exhibit significant negative associations with auditor-client distance (both in kilometers and minutes) and positive associations with audit fees, either at 1% or 5% levels. These robust findings strengthen the study's conclusions, indicating that narcissistic traits, irrespective of measurement method, influence auditor selection and fee structures.
Robustness test
| Variables | DV=Auditor-client_Distance | DV=Audit_Fee | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| DisKM | DisMN | DisKM | DisMN | DisKM | DisMN | DisKM | DisMN | Audit_Fee | Audit_Fee | Audit_Fee | Aud_Fee | |
| Chair_Narc_Sig | −7.43** | −7.86*** | 5.21** | |||||||||
| (−2.14) | (−2.67) | (2.17) | ||||||||||
| CEO_Narc_Sig | −7.04*** | −6.32*** | 7.95*** | |||||||||
| (−2.73) | (−2.89) | (3.35) | ||||||||||
| Chair_Narc_Pic | −0.079** | −0.079** | 0.06** | |||||||||
| (−2.057) | (−2.415) | (2.40) | ||||||||||
| CEO_Narc_Pic | −0.08** | −0.1*** | 0.05** | |||||||||
| (−2.09) | (−2.97) | (1.98) | ||||||||||
| Constant | 2.75** | 3.76*** | 1.53 | 2.85** | 1.50 | 2.71** | 1.37 | 2.71** | 4.52*** | 4.34*** | 4.45*** | 4.39*** |
| (2.14) | (3.45) | (1.15) | (2.52) | (1.14) | (2.43) | (1.03) | (2.38) | (5.08) | (4.91) | (4.81) | (4.72) | |
| Baseline controls | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES |
| Year_FE | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES |
| Industry_FE | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES |
| Observations | 436 | 436 | 436 | 436 | 436 | 436 | 436 | 436 | 436 | 436 | 436 | 436 |
| Adjusted R2 | 0.142 | 0.154 | 0.152 | 0.167 | 0.159 | 0.157 | 0.155 | 0.159 | 0.474 | 0.482 | 0.465 | 0.471 |
| Variables | DV=Auditor-client_Distance | DV=Audit_Fee | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| DisKM | DisMN | DisKM | DisMN | DisKM | DisMN | DisKM | DisMN | Audit_Fee | Audit_Fee | Audit_Fee | Aud_Fee | |
| Chair_Narc_Sig | −7.43** | −7.86*** | 5.21** | |||||||||
| (−2.14) | (−2.67) | (2.17) | ||||||||||
| CEO_Narc_Sig | −7.04*** | −6.32*** | 7.95*** | |||||||||
| (−2.73) | (−2.89) | (3.35) | ||||||||||
| Chair_Narc_Pic | −0.079** | −0.079** | 0.06** | |||||||||
| (−2.057) | (−2.415) | (2.40) | ||||||||||
| CEO_Narc_Pic | −0.08** | −0.1*** | 0.05** | |||||||||
| (−2.09) | (−2.97) | (1.98) | ||||||||||
| Constant | 2.75** | 3.76*** | 1.53 | 2.85** | 1.50 | 2.71** | 1.37 | 2.71** | 4.52*** | 4.34*** | 4.45*** | 4.39*** |
| (2.14) | (3.45) | (1.15) | (2.52) | (1.14) | (2.43) | (1.03) | (2.38) | (5.08) | (4.91) | (4.81) | (4.72) | |
| Baseline controls | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES |
| Year_FE | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES |
| Industry_FE | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES |
| Observations | 436 | 436 | 436 | 436 | 436 | 436 | 436 | 436 | 436 | 436 | 436 | 436 |
| Adjusted R2 | 0.142 | 0.154 | 0.152 | 0.167 | 0.159 | 0.157 | 0.155 | 0.159 | 0.474 | 0.482 | 0.465 | 0.471 |
Note(s): Table 5 reports the robustness test results estimated through OLS using alternative measures of CEO and Chair narcissism. Coefficients are presented with t-statistics in parentheses. Variable definitions are provided in Appendix. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively. DV stands for dependent variable
6.5 Heckman's (1979) two stage model
We implement a Heckman (1979) two-stage model to address potential bias arising from sample attrition due to missing variables or bias arising from non-random observability of CEO narcissism. In the first stage, we estimate a probit model to predict the likelihood that CHAIR/CEO narcissism is observable as a function of firm characteristics, industry, and year, and obtain the inverse Mills ratio (IMR). In the second stage, we examine auditor distance and audit fees conditional on firm characteristics, incorporating the inverse Mills ratio to correct for selection bias. As reported in Table 6, the inverse Mills ratio is statistically insignificant, indicating that sample inclusion is not systematically correlated with the error term of the auditor selection or audit fee equation. This result suggests that selection bias due to data unavailability is unlikely to drive our findings.
Second stage estimation of Heckman two-stage selection model
| Variables | DV=Auditor-client distance | DV=Audit fee | ||||
|---|---|---|---|---|---|---|
| DisKM | DisMN | DisKM | DisMN | Aud_Fee | Aud_Fee | |
| Chair_Narc | −2.458*** | −2.497*** | 1.953*** | |||
| (−0.951) | (−0.804) | (0.636) | ||||
| CEO_Narc | −2.011** | −2.433*** | 2.706*** | |||
| (−0.948) | (−0.799) | (0.617) | ||||
| IMR | 0.736 | 0.381 | 0.771 | 0.408 | −0.573 | −0.537 |
| (1.126) | (0.951) | (1.120) | (0.945) | (0.762) | (−0.747) | |
| Constant | −0.035 | 1.987** | −0.003 | 2.046*** | 5.358*** | 5.017** |
| (−2.988) | (2.524) | (−2.968) | (2.504) | (2.004) | (1.963) | |
| Year_FE | YES | YES | YES | YES | YES | YES |
| Industry_FE | YES | YES | YES | YES | YES | YES |
| Controls | YES | YES | YES | YES | YES | YES |
| Observations | 436 | 436 | 436 | 436 | 436 | 436 |
| Adjusted_R2 | 0.169 | 0.181 | 0.169 | 0.184 | 0.499 | 0.512 |
| Variables | DV=Auditor-client distance | DV=Audit fee | ||||
|---|---|---|---|---|---|---|
| DisKM | DisMN | DisKM | DisMN | Aud_Fee | Aud_Fee | |
| Chair_Narc | −2.458*** | −2.497*** | 1.953*** | |||
| (−0.951) | (−0.804) | (0.636) | ||||
| CEO_Narc | −2.011** | −2.433*** | 2.706*** | |||
| (−0.948) | (−0.799) | (0.617) | ||||
| IMR | 0.736 | 0.381 | 0.771 | 0.408 | −0.573 | −0.537 |
| (1.126) | (0.951) | (1.120) | (0.945) | (0.762) | (−0.747) | |
| Constant | −0.035 | 1.987** | −0.003 | 2.046*** | 5.358*** | 5.017** |
| (−2.988) | (2.524) | (−2.968) | (2.504) | (2.004) | (1.963) | |
| Year_FE | YES | YES | YES | YES | YES | YES |
| Industry_FE | YES | YES | YES | YES | YES | YES |
| Controls | YES | YES | YES | YES | YES | YES |
| Observations | 436 | 436 | 436 | 436 | 436 | 436 |
| Adjusted_R2 | 0.169 | 0.181 | 0.169 | 0.184 | 0.499 | 0.512 |
Note(s): Table 8 reports the regression results of second stage estimation of Heckman two-stage selection model. Coefficients are presented with t-statistics in parentheses. Variable definitions are provided in Appendix. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively. DV stands for dependent variable
6.6 Addressing endogeneity
Endogeneity is a major concern for our study, as the appointment of auditors and the determination of audit fees may not be entirely exogenous. These decisions could be influenced by unobserved firm-specific factors, such as management's strategic preferences, the complexity of financial reporting, or inherent risk profiles, which simultaneously affect the observed auditor-client distance and audit fees. Furthermore, there is a possibility of reverse causality, where firms with certain characteristics—such as those engaging in aggressive financial reporting practices or possessing weaker internal controls—may both attract higher audit fees and prefer auditors located nearby to facilitate oversight or communication. These intertwined relationships complicate the identification of causal effects and necessitate robust econometric techniques to mitigate biases and ensure the validity of the findings. We apply entropy balancing (EB) and two-stage least squares (2SLS) with an instrumental variable approach to address the endogeneity concern of this study.
For EB, we consider the top quartile of Chir_Narc and CEO_Narc as narcissistic executives and the remaining as non-narcissistic executives. We then match narcissistic groups with non-narcissistic groups so that the first and second moments of all covariates are the same between narcissistic and non-narcissistic subsamples. We do not report the reweight statistics for brevity. Table 7 presents the results using entropy-balanced samples. The findings indicate a significant negative relationship between executive narcissism and auditor-client distance. Additionally, the results show a positive and significant relationship between executive narcissism and audit fees. This suggests that our baseline results still hold under the entropy-balanced sample.
Endogeneity test using entropy balancing (EB)
| DV=Auditor-client_Distance | DV=Audit_Fees | |||||
|---|---|---|---|---|---|---|
| Variables | DisKM | DisMN | DisKM | DisMN | Aud_Fee | Aud_Fee |
| Chair_Narc | −0.26*** | −0.26*** | 0.19** | |||
| (−2.67) | (−3.20) | (2.25) | ||||
| CEO_Narc | −0.21** | −0.21** | 0.25** | |||
| (−2.10) | (−2.36) | (2.48) | ||||
| Constant | 5.19*** | 5.33*** | 7.04*** | 6.92*** | 2.24 | 4.32*** |
| (3.02) | (4.07) | (3.85) | (5.28) | (1.51) | (3.88) | |
| Controls | YES | YES | YES | YES | YES | YES |
| Year_FE | YES | YES | YES | YES | YES | YES |
| Industry_FE | YES | YES | YES | YES | YES | YES |
| Observations | 436 | 436 | 436 | 436 | 436 | 436 |
| Adjusted R2 | 0.222 | 0.217 | 0.174 | 0.171 | 0.500 | 0.440 |
| DV=Auditor-client_Distance | DV=Audit_Fees | |||||
|---|---|---|---|---|---|---|
| Variables | DisKM | DisMN | DisKM | DisMN | Aud_Fee | Aud_Fee |
| Chair_Narc | −0.26*** | −0.26*** | 0.19** | |||
| (−2.67) | (−3.20) | (2.25) | ||||
| CEO_Narc | −0.21** | −0.21** | 0.25** | |||
| (−2.10) | (−2.36) | (2.48) | ||||
| Constant | 5.19*** | 5.33*** | 7.04*** | 6.92*** | 2.24 | 4.32*** |
| (3.02) | (4.07) | (3.85) | (5.28) | (1.51) | (3.88) | |
| Controls | YES | YES | YES | YES | YES | YES |
| Year_FE | YES | YES | YES | YES | YES | YES |
| Industry_FE | YES | YES | YES | YES | YES | YES |
| Observations | 436 | 436 | 436 | 436 | 436 | 436 |
| Adjusted R2 | 0.222 | 0.217 | 0.174 | 0.171 | 0.500 | 0.440 |
Note(s): Table 7 reports the regression results estimated using OLS using entropy balanced sample. Coefficients are presented with t-statistics in parentheses. Variable definitions are provided in Appendix. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively. DV stands for dependent variable
Furthermore, to address endogeneity arising from omitted variable bias and simultaneity, this study employs a two-stage least squares (2SLS) approach with an instrumental variable. Specifically, industry-average narcissism metrics for both CEOs and Chairs are used as instruments. While prior studies have successfully applied industry-average measures as instruments to mitigate endogeneity concerns (Chan et al., 2012; Jiraporn and Chintrakarn, 2013), it is acknowledged that this approach has certain limitations, which are discussed in the summary and conclusion section.
Table 8 reports the results of the 2SLS regression using an instrumental variable approach to address endogeneity concerns in this study. The first-stage regression results show strong positive associations of the instruments with the narcissism measures of the respective executives, confirming the validity and relevance of the instruments. In the second-stage regression, the results demonstrate a significant (at 1% level mostly) negative relationship between executive narcissism and auditor-client distance, measured in both kilometers and minutes to travel, and a significant (at 5% level) positive association between executive narcissism and audit fees. These findings provide robust validation of our baseline results, demonstrating their consistency and reliability even after accounting for potential endogeneity concerns.
2SLS Regression using instrumental variable
| Variables | First-stage_regression | Second-stage_regression | ||||||
|---|---|---|---|---|---|---|---|---|
| DV=Auditor-client distance | DV=Audit fees | |||||||
| Chair_Narc | CEO_Narc | DisKM | DisMN | DisKM | DisMN | Aud_Fee | Aud_Fee | |
| Ind_Avg_Chair_Narc | 0.88*** | |||||||
| (2.73) | ||||||||
| Ind_Avg_CEO_Narc | 0.73*** | |||||||
| (2.85) | ||||||||
| Chair_Narc | −2.58*** | −2.56*** | 5.522** | |||||
| (−2.77) | (−3.27) | (2.287) | ||||||
| CEO_Narc | −1.72* | −2.10*** | 7.947*** | |||||
| (−1.83) | (−2.66) | (3.301) | ||||||
| Constant | −0.169** | −0.085 | 1.437 | 2.576** | 1.950 | 3.098*** | 4.334*** | 4.262*** |
| (−2.007) | (−1.058) | (1.028) | (2.185) | (1.433) | (2.707) | (4.834) | (4.647) | |
| Controls | YES | YES | YES | YES | YES | YES | YES | YES |
| Year_FE | YES | YES | YES | YES | YES | YES | YES | YES |
| Industry_FE | YES | YES | YES | YES | YES | YES | YES | YES |
| Observations | 436 | 436 | 436 | 436 | 436 | 436 | 436 | 436 |
| Adjusted_R2 | 0.191 | 0.250 | 0.175 | 0.183 | 0.168 | 0.181 | 0.475 | 0.482 |
| Variables | First-stage_regression | Second-stage_regression | ||||||
|---|---|---|---|---|---|---|---|---|
| DV=Auditor-client distance | DV=Audit fees | |||||||
| Chair_Narc | CEO_Narc | DisKM | DisMN | DisKM | DisMN | Aud_Fee | Aud_Fee | |
| Ind_Avg_Chair_Narc | 0.88*** | |||||||
| (2.73) | ||||||||
| Ind_Avg_CEO_Narc | 0.73*** | |||||||
| (2.85) | ||||||||
| Chair_Narc | −2.58*** | −2.56*** | 5.522** | |||||
| (−2.77) | (−3.27) | (2.287) | ||||||
| CEO_Narc | −1.72* | −2.10*** | 7.947*** | |||||
| (−1.83) | (−2.66) | (3.301) | ||||||
| Constant | −0.169** | −0.085 | 1.437 | 2.576** | 1.950 | 3.098*** | 4.334*** | 4.262*** |
| (−2.007) | (−1.058) | (1.028) | (2.185) | (1.433) | (2.707) | (4.834) | (4.647) | |
| Controls | YES | YES | YES | YES | YES | YES | YES | YES |
| Year_FE | YES | YES | YES | YES | YES | YES | YES | YES |
| Industry_FE | YES | YES | YES | YES | YES | YES | YES | YES |
| Observations | 436 | 436 | 436 | 436 | 436 | 436 | 436 | 436 |
| Adjusted_R2 | 0.191 | 0.250 | 0.175 | 0.183 | 0.168 | 0.181 | 0.475 | 0.482 |
Note(s): Table 8 reports the two-stage least square (2SLS) regression results using instrumental variable. Coefficients are presented with t-statistics in parentheses. Variable definitions are provided in Appendix. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively. DV stands for dependent variable
6.7 Additional analyses
We conduct additional analyses to offer more insights into the observed association between Chair/CEO narcissism and audit distance and audit fees. Our main results show that Chair/CEO narcissism is negatively associated with audit distance, while they are positively associated with audit fees. We argue that the association between Chair/CEO narcissism and audit distance and audit fees will be more pronounced in family firms. In family firms, CEOs/Chairs may face less external pressure due to concentrated ownership and weak boards (Khan et al., 2015). To examine our anticipation, we conduct a subsample analysis. Table 9 Panel A presents the results. We observe that the coefficients on Chair/CEO narcissism are more robust for the family firms' sample, suggesting that the association between Chair/CEO narcissism and audit distance and audit fees is relatively stronger in family firms.
Cross-sectional analyses
| Variables | Non-family firms | Family firms | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| DV=Auditor-client distance | DV=Audit fee | DV=Auditor-client distance | DV=Audit fee | |||||||||
| DisKM | DisMN | DisKM | DisMN | Aud_Fee | Aud_Fee | DisKM | DisMN | DisKM | DisMN | Aud_Fee | Aud_Fee | |
| Panel A: Family firms, executive narcissism, audit distance, and audit fees | ||||||||||||
| Chair_Narc | −1.013 | −0.958 | 0.031 | −1.540** | −1.581*** | 0.764* | ||||||
| (−1.203) | (−1.399) | (0.050) | (−2.482) | (−2.886) | (1.969) | |||||||
| CEO_Narc | −0.922 | −1.696** | 0.577 | −1.317* | −1.185** | 0.913** | ||||||
| (−1.154) | (−2.651) | (0.976) | (−1.949) | (−1.975) | (2.222) | |||||||
| Constant | 1.458 | 2.126 | 1.736 | 2.479 | 3.747** | 3.653** | 1.517 | 3.170 | 2.195 | 3.803 | 1.215 | 1.314 |
| (0.587) | (1.052) | (0.697) | (1.243) | (2.035) | (1.985) | (0.542) | (1.282) | (0.791) | (1.542) | (0.694) | (0.778) | |
| Year_FE | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES |
| Industry_FE | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES |
| Observations | 214 | 214 | 214 | 214 | 214 | 214 | 222 | 222 | 222 | 222 | 222 | 222 |
| Adjusted_R2 | 0.147 | 0.148 | 0.151 | 0.176 | 0.486 | 0.491 | 0.133 | 0.136 | 0.123 | 0.121 | 0.524 | 0.545 |
| Variables | Non-family firms | Family firms | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| DV=Auditor-client distance | DV=Audit fee | DV=Auditor-client distance | DV=Audit fee | |||||||||
| DisKM | DisMN | DisKM | DisMN | Aud_Fee | Aud_Fee | DisKM | DisMN | DisKM | DisMN | Aud_Fee | Aud_Fee | |
| Panel A: Family firms, executive narcissism, audit distance, and audit fees | ||||||||||||
| Chair_Narc | −1.013 | −0.958 | 0.031 | −1.540** | −1.581*** | 0.764* | ||||||
| (−1.203) | (−1.399) | (0.050) | (−2.482) | (−2.886) | (1.969) | |||||||
| CEO_Narc | −0.922 | −1.696** | 0.577 | −1.317* | −1.185** | 0.913** | ||||||
| (−1.154) | (−2.651) | (0.976) | (−1.949) | (−1.975) | (2.222) | |||||||
| Constant | 1.458 | 2.126 | 1.736 | 2.479 | 3.747** | 3.653** | 1.517 | 3.170 | 2.195 | 3.803 | 1.215 | 1.314 |
| (0.587) | (1.052) | (0.697) | (1.243) | (2.035) | (1.985) | (0.542) | (1.282) | (0.791) | (1.542) | (0.694) | (0.778) | |
| Year_FE | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES |
| Industry_FE | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES |
| Observations | 214 | 214 | 214 | 214 | 214 | 214 | 222 | 222 | 222 | 222 | 222 | 222 |
| Adjusted_R2 | 0.147 | 0.148 | 0.151 | 0.176 | 0.486 | 0.491 | 0.133 | 0.136 | 0.123 | 0.121 | 0.524 | 0.545 |
| Variables | Non-big 4 auditors | Big 4 auditors | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| DV=Auditor-client distance | DV=Audit fee | DV=Auditor-client distance | DV=Audit fee | |||||||||
| DisKM | DisMN | DisKM | DisMN | Aud_Fee | Aud_Fee | DisKM | DisMN | DisKM | DisMN | Aud_Fee | Aud_Fee | |
| Panel B: Big 4 auditors, executive narcissism, audit distance, and audit fees | ||||||||||||
| Chair_Narc | −1.429** | −1.339*** | 0.915** | 0.102 | 0.181 | −0.723 | ||||||
| (−2.547) | (−2.752) | (2.330) | (0.103) | (0.266) | (−0.769) | |||||||
| CEO_Narc | −0.930* | −1.336*** | 1.155*** | 0.212 | 0.305 | −0.459 | ||||||
| (−1.713) | (−2.859) | (3.084) | (0.145) | (0.305) | (−0.326) | |||||||
| Constant | 0.104 | 1.411 | 0.061 | 1.409 | 5.007*** | 4.986*** | 2.958 | 4.245** | 2.718 | 3.380 | 3.678 | 3.844 |
| (0.067) | (1.043) | (0.039) | (1.050) | (4.585) | (4.636) | (1.042) | (2.162) | (0.905) | (1.645) | (1.356) | (1.329) | |
| Year_FE | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES |
| Industry_FE | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES |
| Observations | 348 | 348 | 348 | 348 | 348 | 348 | 88 | 88 | 88 | 88 | 88 | 88 |
| Adjusted_R2 | 0.224 | 0.211 | 0.216 | 0.216 | 0.418 | 0.430 | 0.261 | 0.390 | 0.257 | 0.393 | 0.335 | 0.314 |
| Variables | Non-big 4 auditors | Big 4 auditors | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| DV=Auditor-client distance | DV=Audit fee | DV=Auditor-client distance | DV=Audit fee | |||||||||
| DisKM | DisMN | DisKM | DisMN | Aud_Fee | Aud_Fee | DisKM | DisMN | DisKM | DisMN | Aud_Fee | Aud_Fee | |
| Panel B: Big 4 auditors, executive narcissism, audit distance, and audit fees | ||||||||||||
| Chair_Narc | −1.429** | −1.339*** | 0.915** | 0.102 | 0.181 | −0.723 | ||||||
| (−2.547) | (−2.752) | (2.330) | (0.103) | (0.266) | (−0.769) | |||||||
| CEO_Narc | −0.930* | −1.336*** | 1.155*** | 0.212 | 0.305 | −0.459 | ||||||
| (−1.713) | (−2.859) | (3.084) | (0.145) | (0.305) | (−0.326) | |||||||
| Constant | 0.104 | 1.411 | 0.061 | 1.409 | 5.007*** | 4.986*** | 2.958 | 4.245** | 2.718 | 3.380 | 3.678 | 3.844 |
| (0.067) | (1.043) | (0.039) | (1.050) | (4.585) | (4.636) | (1.042) | (2.162) | (0.905) | (1.645) | (1.356) | (1.329) | |
| Year_FE | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES |
| Industry_FE | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES |
| Observations | 348 | 348 | 348 | 348 | 348 | 348 | 88 | 88 | 88 | 88 | 88 | 88 |
| Adjusted_R2 | 0.224 | 0.211 | 0.216 | 0.216 | 0.418 | 0.430 | 0.261 | 0.390 | 0.257 | 0.393 | 0.335 | 0.314 |
Note(s): Table 9 reports the results for cross-sectional analyses. Panel A presents how family firms moderate the association between executives' narcissism and audit distance and audit fees. Panel B shows how Big 4 auditors moderate the association between executives' narcissism and audit distance and audit fees. Coefficients are presented with t-statistics in parentheses. Variable definitions are provided in Appendix. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively. DV stands for dependent variable
Nevertheless, we predict that firms' willingness to engage a Big 4 auditor weakens the demonstrated association between Chair/CEO narcissism and auditor distance and audit fees. Big 4 affiliated audit offices may not always be located in the vicinity of the client firms. Therefore, companies opting Big 4 auditors may need to go far distance, avoiding local non-Big 4 auditors. Moreover, executives with narcissistic traits, who intend to influence audit procedures, may not offer higher audit fees to Big 4 auditors because they are less likely to compromise quality for money and have a reputation concern. To test our supposition, we conduct a sub-sample analysis. Table 9 Panel B reports the results. As we see, the coefficients are significant for non-Big 4 auditors but insignificant for the Big 4 sample, indicating that Big 4 auditors reduce the effect of Chair/CEO narcissism on auditor distance and audit fees.
Further, to validate our conclusions on the key findings, we extend our analysis by examining the broader implications of executive narcissism on critical audit outcomes, specifically audit quality and audit report lag (ARL). Audit quality is measured using discretionary accrual models, consistent with prior literature, where these models have been widely applied and validated as reliable proxies in auditing and accounting research (Bartov et al., 2000; Choi et al., 2012; Krishnan, 2003; DeFond and Zhang, 2014). Audit report lag is measured in terms of the number of days taken for issuing audit reports from the accounting year-end date (Krishnan and Yang, 2009; Schwartz and Soo, 1996). The definitions of audit quality measures and ARL are provided in Appendix.
Table 10 reports the results on the impact of executive narcissism on audit quality measures and audit report lag (ARL). The findings reveal that executive narcissism is generally negatively associated with audit quality proxies and ARL, though these relationships are not statistically significant, except for CEO narcissism, which shows a strong negative association with ARL at the 1% significance level. This suggests that narcissistic executives may prioritize speed in the audit reporting process, expediting report issuance at the potential expense of audit quality (Amin et al., 2025). Such behavior reflects a symbolic rather than substantive compliance, where the focus is on obtaining audit reports quickly to project efficiency rather than ensuring the robustness of audit quality.
Executive narcissism, audit quality and audit report delay
| Variables | DV=Audit quality | DV=Audit_Report_Lag | ||||||
|---|---|---|---|---|---|---|---|---|
| Jones91 | Stubben10 | SP | Jones91 | Stubben10 | SP | Aud_Lag | Aud_Lag | |
| Chair_Narc | −0.033 | −0.012 | −0.201 | −0.191 | ||||
| (−1.15) | (−0.64) | (−0.151) | (−1.192) | |||||
| CEO_Narc | −0.011 | −0.013 | 0.852 | −0.565*** | ||||
| (−0.35) | (−0.64) | (0.60) | (−3.56) | |||||
| Constant | 0.116 | 0.090 | −30.701 | 0.111 | 0.070 | −26.595 | 5.255*** | 5.170*** |
| (1.318) | (1.640) | (−0.010) | (1.236) | (1.247) | (−0.017) | (15.027) | (14.877) | |
| Controls | YES | YES | YES | YES | YES | YES | YES | YES |
| Year_FE | YES | YES | YES | YES | YES | YES | YES | YES |
| Industry_FE | YES | YES | YES | YES | YES | YES | YES | YES |
| Observations | 223 | 217 | 344 | 220 | 214 | 340 | 436 | 436 |
| Adjusted_R2 | 0.13 | 0.23 | 0.271 | 0.12 | 0.23 | 0.303 | 0.221 | 0.246 |
| Variables | DV=Audit quality | DV=Audit_Report_Lag | ||||||
|---|---|---|---|---|---|---|---|---|
| Jones91 | Stubben10 | SP | Jones91 | Stubben10 | SP | Aud_Lag | Aud_Lag | |
| Chair_Narc | −0.033 | −0.012 | −0.201 | −0.191 | ||||
| (−1.15) | (−0.64) | (−0.151) | (−1.192) | |||||
| CEO_Narc | −0.011 | −0.013 | 0.852 | −0.565*** | ||||
| (−0.35) | (−0.64) | (0.60) | (−3.56) | |||||
| Constant | 0.116 | 0.090 | −30.701 | 0.111 | 0.070 | −26.595 | 5.255*** | 5.170*** |
| (1.318) | (1.640) | (−0.010) | (1.236) | (1.247) | (−0.017) | (15.027) | (14.877) | |
| Controls | YES | YES | YES | YES | YES | YES | YES | YES |
| Year_FE | YES | YES | YES | YES | YES | YES | YES | YES |
| Industry_FE | YES | YES | YES | YES | YES | YES | YES | YES |
| Observations | 223 | 217 | 344 | 220 | 214 | 340 | 436 | 436 |
| Adjusted_R2 | 0.13 | 0.23 | 0.271 | 0.12 | 0.23 | 0.303 | 0.221 | 0.246 |
Note(s): Table 10 reports the impact of executive narcissism on audit quality and audit report delay. We use three proxies for audit quality: Jones91 (Jones, 1991), Stubben10 (Stubben, 2010), and SP (small profit), where Jones91 and Stubben10 are continuous variables, and SP is an indicator variable equal to 1 if profit is less than 1% of total assets and 0 otherwise, following prior literature (Dechow et al., 2010). The mean values of Jones91, Stubben10, and SP are 0.042, 0.027, and 0.49, respectively. Audit report lag (ARL) is measured as the number of days from the accounting year-end to audit report issue and has a mean of 4.76 days. Coefficients are presented with t-statistics in parentheses. Variable definitions are provided in Appendix. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively. DV stands for dependent variable
This finding aligns with prior literature, which highlights the detrimental influence of executive narcissism on corporate governance and reporting practices. For instance, narcissistic executives are often linked to heightened risks of earnings management and financial misstatements, which undermine audit quality (Chatterjee and Hambrick, 2007; Rijsenbilt and Commandeur, 2013). Moreover, their overconfidence and desire for rapid recognition may lead to expedited audit processes, resulting in shorter audit report delays but potentially compromising the thoroughness and accuracy of audit procedures (O'Reilly et al., 2018; Salehi et al., 2021). These findings suggest a plausible mechanism wherein narcissistic executives may strategically opt for geographically proximate auditors to exercise heightened influence over the auditing process. This proximity potentially facilitates greater oversight and enables the exertion of pressure to expedite the delivery of audit reports. However, this preference for speed and control appears to come at a premium, as evidenced by the payment of higher audit fees. Such a dynamic, while expedient, raises concerns about a potential trade-off, wherein the emphasis on proximity and rapid reporting may undermine the rigor and overall quality of the audit process.
7. Summary and conclusion
This study examines whether narcissistic executives appoint geographically proximate auditors and pay higher audit fees, drawing on the lens of upper echelon theory, which posits that organizational outcomes are shaped by the traits and behaviors of top executives. The findings reveal that firms led by narcissistic CEOs and Chairs exhibit a strong preference for engaging auditors located in closer geographical proximity, while simultaneously incurring higher audit fees. These results are consistent and robust across alternative measures of executive narcissism and auditor-client distance, reinforcing the validity of the observed associations. Importantly, the results also withstand rigorous tests for potential endogeneity, including entropy balancing and 2SLS with an instrumental variable approach. Further analysis highlights two potential channels: narcissistic executives appear to demand expedited audit reports, thereby reducing audit report lag, but often at the expense of audit quality. By appointing familiar, geographically proximate auditors and compensating them with higher fees, narcissistic executives seem to prioritize symbolic compliance and control over substantive improvements in audit quality.
The findings of this study make notable contributions to the existing literature in several significant ways. First, this research represents a pioneering effort to investigate the influence of top executives' narcissism on both the selection and remuneration of auditors. While prior studies have examined the relationship between executive narcissism and audit fees (Judd et al., 2017; Xiang and Song, 2021), little attention has been given to the types of auditors preferred by narcissistic CEOs and Chairs. This study fills that gap by documenting a clear tendency of narcissistic executives to appoint geographically proximate auditors, a dimension of auditor choice that has been largely overlooked.
Second, the study provides new insights into the negative implications of executive narcissism for audit quality and audit report lag. Unlike previous research that primarily linked narcissism to firm-level outcomes such as performance, investment, or corporate social responsibility (Cragun et al., 2019; Ham et al., 2018; Marquez-Illescas et al., 2019; Olsen et al., 2014), this study uniquely connects narcissistic leadership traits to compromised audit quality and expedited audit reporting processes. The findings highlight a trade-off, where narcissistic executives prioritize speed and control over auditors at the expense of substantive audit quality, thereby undermining the overall integrity of financial reporting.
Third, by applying upper echelon theory, this study extends its explanatory power into the auditing domain. It demonstrates how the psychological characteristics of top executives influence not only firm-level outcomes but also critical audit-related decisions. This theoretical extension underscores the relevance of executive traits in shaping governance processes, auditor independence, and financial reporting quality.
Fourth, the results provide actionable insights for regulators, corporate boards, and auditors. The findings call for stronger regulatory safeguards in emerging economies like Bangladesh, where narcissistic executives may prioritize symbolic compliance (e.g. faster audit report issuance) over substantive audit quality. Regulators should enhance the enforcement of minimum audit fee guidelines, strengthen auditor independence requirements, and promote transparency in auditor selection and remuneration to mitigate undue executive influence. Similarly, boards and nomination/audit committees may factor in psychological traits such as narcissism during recruitment and succession planning, given their implications for governance and oversight. Auditors can incorporate executive narcissism as a potential risk factor when designing audit strategies, negotiating fees, and assessing audit independence.
Finally, by demonstrating how executive narcissism can compromise audit quality, the study speaks to broader societal concerns about the reliability of financial reporting. Stronger governance mechanisms and improved audit practices are vital to ensuring investor protection, market confidence, and overall trust in financial systems.
While the study offers original contributions, it is not without its limitations. First, the research is conducted in the context of an emerging economy characterized by political connections, weaker governance structures, family business, and less stringent regulatory oversight (Khan et al., 2015; Muttakin et al., 2018). These contextual factors may limit the generalizability of the findings to developed economies, where governance frameworks and market dynamics significantly differ. The nuances of institutional environments in advanced economies, such as stronger legal protections and auditor independence requirements, may alter the observed relationships. Second, the study focuses exclusively on publicly listed companies. It is plausible that the dynamics of top-level executive narcissism could manifest differently in private firms, where ownership structures, stakeholder expectations, and governance mechanisms vary. Exploring these differences could provide a richer understanding of narcissism's impact across diverse organizational contexts. Third, the study employs observable proxies for narcissism, such as image prominence and signature size, to capture executive narcissistic tendencies. While these measures are widely recognized and validated in the literature (Chatterjee and Hambrick, 2007; Cragun et al., 2019; Ham et al., 2018), they may not fully reflect the multidimensional nature of narcissism. As such, these proxies could oversimplify or inadequately capture the psychological complexity of narcissistic traits in executives. Finally, we acknowledge the limitations of using industry-average narcissism measures as an instrument in the 2SLS analysis, as identifying an ideal instrument is inherently challenging. Similarly, we recognize the limitations of discretionary accrual-based measures of audit quality, which, despite their wide application in prior literature, remain indirect and subject to potential misspecification.
Future research can address the limitations of the current study in several ways. Comparative studies across developed and emerging economies could provide valuable insights into how institutional differences moderate the effects of executive narcissism on audit outcomes. Examining the private sector, where corporate governance and reporting requirements differ substantially from public firms, may also yield unique findings. Additionally, adopting richer methodologies—such as psychometric assessments, longitudinal behavioral analyses, or triangulated mixed-methods approaches—could capture the psychological complexity of executive narcissism more comprehensively. Finally, methodological refinements, including the use of alternative econometric strategies and complementary measures of audit quality, could enhance robustness and strengthen construct validity in future investigations.
Appendix
Definition and measurement of the variables
| Variables | Description |
|---|---|
| Dependent_Variable | |
| DisKM | The natural log of the distance (in kilometers) between the auditor's and client's offices |
| DisMN | The natural log of the distance (in minutes) between the auditor's and client's offices |
| Aud_Fee | The natural logarithm of the amount of audit fees paid by the firm |
| Independent_Variables | |
| Chair_Narc | A composite score derived from multiplying the chair's signature size and picture size scores (see below) |
| CEO_Narc | A composite score derived from multiplying the_CEO's signature size and picture size scores (see below) |
| Chair_Narc_Sig | Size of Chair's signature calculated as the rectangle's area (length × width in inch) divided by the number of letters in the Chair's name |
| CEO_Narc_Sig | Size of CEO's signature calculated as the rectangle's area (length × width in inch) divided by the number of letters in the CEO's name |
| Chair_Narc_Pic | A 0–4 score for the Chair's picture size in the annual report, ranging from no picture to a group photo with other executives or CEO, half-page, or full-page coverage |
| CEO_Narc_Pic | A 0–4 score for the CEO's picture size in annual reports, ranging from no picture to a group photo with other executive or chair, half-page, or full-page coverage |
| Control variables | |
| Chair_Gen | An indicator variable of 1 if the chair of the firm is female; and 0 otherwise |
| CEO_Gen | An indicator variable of 1 if the CEO of the firm is female; and 0 otherwise |
| FDR | Female director's ratio measured as the proportion of female members on the board |
| IDR | Independent director's ratio measured as the proportion of independent directors on the board |
| AC_Size | The size of the audit committees in numbers |
| Aud_Lag | The number of days between accounting year end date and date of signing the audit report |
| Aud_Ten | The number of years current audit firm is auditing the client |
| Big4 | An indicator variable of 1 if the audit firm is one of the big-4 auditors; and 0 otherwise |
| Firm_Size | Natural logarithm of total assets indicating the size of the firm |
| Firm_Age | Natural logarithm of number of years in business from the listing year |
| ROA | Return on Asset measured as the net profit divided by total assets |
| Lev | The ratio of debt to total assets |
| YearEnd | An indicator variable coded 1 if the reporting year ended on December; and 0 otherwise. |
| Mediating variables | |
| Jones91 | Discretionary accruals model following Jones (1991) as the proxy for audit quality |
| Stubben10 | Discretionary accruals model following Stubben (2010) as the proxy for audit quality |
| SP | Small profit is an indicator variable equal to 1 if profit is less than 1% of total assets; and 0 otherwise |
| ARL | Audit report lag is measured as the number of days from the accounting year-end to the issuance of the audit report |
| Variables | Description |
|---|---|
| Dependent_Variable | |
| DisKM | The natural log of the distance (in kilometers) between the auditor's and client's offices |
| DisMN | The natural log of the distance (in minutes) between the auditor's and client's offices |
| Aud_Fee | The natural logarithm of the amount of audit fees paid by the firm |
| Independent_Variables | |
| Chair_Narc | A composite score derived from multiplying the chair's signature size and picture size scores (see below) |
| CEO_Narc | A composite score derived from multiplying the_CEO's signature size and picture size scores (see below) |
| Chair_Narc_Sig | Size of Chair's signature calculated as the rectangle's area (length × width in inch) divided by the number of letters in the Chair's name |
| CEO_Narc_Sig | Size of CEO's signature calculated as the rectangle's area (length × width in inch) divided by the number of letters in the CEO's name |
| Chair_Narc_Pic | A 0–4 score for the Chair's picture size in the annual report, ranging from no picture to a group photo with other executives or CEO, half-page, or full-page coverage |
| CEO_Narc_Pic | A 0–4 score for the CEO's picture size in annual reports, ranging from no picture to a group photo with other executive or chair, half-page, or full-page coverage |
| Control variables | |
| Chair_Gen | An indicator variable of 1 if the chair of the firm is female; and 0 otherwise |
| CEO_Gen | An indicator variable of 1 if the CEO of the firm is female; and 0 otherwise |
| FDR | Female director's ratio measured as the proportion of female members on the board |
| IDR | Independent director's ratio measured as the proportion of independent directors on the board |
| AC_Size | The size of the audit committees in numbers |
| Aud_Lag | The number of days between accounting year end date and date of signing the audit report |
| Aud_Ten | The number of years current audit firm is auditing the client |
| Big4 | An indicator variable of 1 if the audit firm is one of the big-4 auditors; and 0 otherwise |
| Firm_Size | Natural logarithm of total assets indicating the size of the firm |
| Firm_Age | Natural logarithm of number of years in business from the listing year |
| ROA | Return on Asset measured as the net profit divided by total assets |
| Lev | The ratio of debt to total assets |
| YearEnd | An indicator variable coded 1 if the reporting year ended on December; and 0 otherwise. |
| Mediating variables | |
| Jones91 | Discretionary accruals model following |
| Stubben10 | Discretionary accruals model following |
| SP | Small profit is an indicator variable equal to 1 if profit is less than 1% of total assets; and 0 otherwise |
| ARL | Audit report lag is measured as the number of days from the accounting year-end to the issuance of the audit report |
Note
As the banking sector is subject to stricter regulatory requirements compared to non-financial firms, we re-estimated the models excluding banks (12.84% of the sample). The results remain qualitatively and statistically robust, indicating that the main findings are not driven by the inclusion of the banking sector. We don't report the results for brevity.

