Gender Dynamics and Banks’ Performance: Does Cybersecurity Disclosure Matter? Evidence from Jordan

Abstract

Purpose: Rapid bank digitisation has heightened cybersecurity risks and increased stakeholder expectations for transparent cyber risk governance and disclosure. However, research on whether women’s board involvement enhances financial success varies and depends on the context, particularly within different institutional settings. Therefore, this study investigates the impact of Women on Boards (WIB) on Earnings per Share (EPS) of Jordanian banks during 2010 to 2022 and further examines the moderating effect of Cyber Security Disclosure (CSD) on the relationship between WIB and EPS. Design: Combining manual content analysis of each Jordanian bank’s annual report with regression analysis to assess the correlation between EPS, WIB, and CSD. The study also controls for audit quality estimates, financial leverage, bank age, and size. Findings: Our results reveal a negative correlation between EPS and the increasing number of women on boards; thus, simply having more women on boards does not necessarily lead to higher EPS. Additionally, there is a positive interaction between WIB and CSD on EPS, indicating that strong cybersecurity practices can mitigate the negative effects of gender diversity on the board. The ongoing negative association between board diversity and EPS underscores the complexity of gender relations in corporate governance issues. Originality: This research is the first to examine both gender diversity and cybersecurity practices within the same context, as they jointly influence corporate governance and financial performance in new ways. It emphasises the importance of viewing cybersecurity disclosures as a strategic component that can positively impact the financial outcomes of board diversity.

1. Introduction

Growing awareness of gender diversity in corporate leadership, especially within the banking sector, has become a notable trend recently (Birindelli et al., 2019; Galletta et al., 2022; Moreno-Gómez et al., 2018). Company boards tend to be more effective and better governed when they include both men and women (Francoeur et al., 2008). At present, there are considerably fewer women on boards compared to their male counterparts, reflecting systemic barriers and biases that hinder women’s participation in board-level decision-making (Adams & Ferreira, 2009; Terjesen et al., 2016; Afroze et al., 2025). Although efforts have been made to increase women’s access to board positions, women remain substantially underrepresented in global corporate governance (Pletzer et al., 2015; Brieger et al., 2019), especially within the banking sector in Jordan (Mohammad et al., 2018). Most research examining the link between women’s presence on boards and company performance, often measured by earnings per share (EPS), in Jordanian banks (Alshirah et al., 2022; Saidat et al., 2019), has not explicitly explored the moderating role of Corporate Social Responsibility (CSD); additional research is needed to understand how CSD influences this relationship.

Several recent studies have shown a positive relationship between the number of women on boards of directors worldwide and improved corporate financial performance (Francoeur et al., 2008; Wellalage & Locke, 2013). Furthermore, research indicates that gender diversity on boards can positively influence the impact of political ties on a bank’s financial performance, as well as enhance the performance and stability of banks (for example, Moreno-Gómez et al., 2018).

However, empirical findings remain mixed, and the effectiveness of board gender diversity seems to depend on the surrounding governance and disclosure environment. In this context, cybersecurity disclosure has become a vital governance tool that reduces information asymmetry, boosts stakeholder confidence, and improves board oversight (Mansour et al., 2026; Bruno et al., 2025). Recent evidence further shows that board characteristics—especially female representation and critical mass—play a key role in shaping the quality, credibility, and strategic focus of cybersecurity disclosure (Elnahass et al., 2025; Afroze et al., 2025). As a result, CSD may act as a moderating factor that affects how women’s participation on bank boards contributes to better financial performance. Emerging governance research also indicates that board gender diversity may impact not only firm results but also disclosure-related governance practices, especially in areas marked by increased risk and information asymmetry, such as cybersecurity, thereby providing a theoretical foundation for exploring CSD as an interactive moderating mechanism (Elnahass et al., 2025; Afroze et al., 2025).

In conclusion, previous research shows that banks experience decreasing marginal returns in financial performance, with gender diversity playing a significant role in this decline (Birindelli et al., 2019).

A significant gap exists in the current understanding of how the presence of female directors on bank boards influences their EPS in Jordan from 2010 to 2022. While existing Jordanian studies primarily focus on direct relationships between board characteristics and bank performance (Saidat et al., 2019; Alshirah et al., 2022), recent international research emphasises the relevance of cybersecurity disclosure as a governance tool at the board level (Mansour et al., 2026; Bruno et al., 2025). However, no prior study has combined women on bank boards and cybersecurity disclosure within a unified empirical framework in the Jordanian banking context.

This research helps bridge the gap by examining whether Women on Bank Boards (WIBs) influence earnings per share (EPS) and whether Cybersecurity Disclosure (CSD) moderates that relationship. The period from 2010 to 2022 is suitable for Jordanian banks because it covers the growing digitisation of banking operations and culminates in the years when cybersecurity disclosure in banks’ annual reports became more prominent, alongside a documented increase in cyberattacks and related regulatory measures (Elsayed et al., 2024).

To accomplish this objective, a highly structured quantitative data-collection approach was employed, including the review of financial and annual reports from the Amman Stock Exchange.

Although Jordan has made progress towards achieving gender equality and empowering women (Alshirah et al., 2022), the effort to ensure women occupy leadership roles within their institutions, especially in the banking sector, still has a long way to go (Mohammad et al., 2018). This study adopts an institution- and governance-based view on gender diversity, avoiding any attribution of performance outcomes to inherent gender traits and instead emphasising the structural and informational environments in which boards operate.

Therefore, to effectively address this issue, collaboration between decision-makers and policymakers in both the private and public sectors, as well as civil society, is essential to challenge traditional views of gender roles, eliminate barriers often created by structural changes to promote women’s inclusion, and foster inclusive and diverse workplace cultures. Furthermore, the representation of gender diversity by women in corporate leadership positions is a significant concern for many Jordanian citizens (Saidat et al., 2019).

Within the Jordanian institutional and regulatory context, this study makes a significant contribution to shaping future policy development, governance practices within the banking sector, and the formulation of strategies relating to banking. The study will help deepen understanding of how the variables of CSD and gender diversity influence business performance, thereby enabling the creation of more effective governance frameworks that can accommodate a broader range of industry sectors.

The structure of this study is outlined as follows: first, preparation for developing research hypotheses and how these hypotheses are developed, along with presenting the theoretical foundations; second, preparation for conducting an empirical study; third, carrying out the empirical investigation; fourth, evaluating the results of the empirical study; and fifth, discussing the implications of the findings and providing recommendations for future research.

2. Theoretical Background, Literature Review, and Hypotheses

2.1. Theoretical Background

This study investigates how gender impacts the performance of the banking industry using agency theory and resource dependency theory, as well as their connections to contemporary corporate governance issues organisations encounter today. The study will utilise these two theoretical frameworks as guides to better understand the significance of the composition of the board of directors in enhancing organisational performance, particularly within banks. This research will emphasise how these two related theories contribute to increasing the productivity and profitability of the banking sector.

Board independence is vital in bridging the gap between management and shareholder interests, and therefore agency theory highlights an important role for boards of directors (Fama & Jensen, 1983; Jensen & Meckling, 1976). One way boards achieve independence is through diversity among members (gender or otherwise), which can introduce different perspectives on ethical issues (Campbell & Mínguez-Vera, 2008; Baker et al., 2020) that influence decision-making processes that are traditionally male-dominated. The inclusion of diverse members is expected to improve the overall quality of governance and, in turn, to positively affect banks’ financial performance. Furthermore, through the application of agency theory, it is assumed that a manager will voluntarily provide information about the company to shareholders to reduce the information gap between management and shareholders, which should ultimately enhance transparency and governance quality (Watson et al., 2002; Healy & Palepu, 2001).

Additionally, Resource Dependence Theory states that boards’ effectiveness relies on their ability to generate and utilise external resources and networks (Hillman et al., 2000; Hillman & Dalziel, 2003; Pfeffer & Salancik, 1978). From this viewpoint, gender-diverse boards are believed to have better access to external resources and greater capacity to mobilise them than non-gender-diverse boards. The argument that diverse perspectives are essential for developing an effective strategy to counter cybersecurity threats and for providing stakeholders with valuable information about their company’s vulnerabilities is especially relevant in cybersecurity. Having access to multiple contact lists is one way to ensure a diverse perspective and to formulate strategies that incorporate diverse viewpoints. Unlike signalling theory, a gender-diverse board is likely to provide additional information about its governance practices, thereby enhancing its reputation, building trust with stakeholders, and increasing its access to critical external resources (Elshandidy et al., 2018; Mazumder & Hossain, 2023).

This research proposes a central hypothesis that banks with gender-diverse boards will demonstrate improved financial performance, with CSD practices acting as a moderating factor between gender diversity and financial results. Gender-diverse banks are generally better equipped than their less diverse counterparts to manage and disclose cyber risks, leading to the development of an informed and trustworthy stakeholder community and enhancing their position in the financial market, which ultimately boosts their financial outcomes. Evidence indicating that organisations with good governance, evidenced by transparency and comprehensive disclosures, are preferred by stakeholders supports this hypothesis. This evidence also reinforces the idea that board diversity fosters better organisational performance and subsequently enhances financial results.

2.2. Gender Diversity and Bank Performance

Examining the impact of women’s inclusion in banking (WIB) shows how female members of corporate boards can influence firm performance. Furthermore, the ways in which countries have implemented either legal or voluntary initiatives to promote WIB globally form part of the evidence indicating a widespread commitment by many organisations to increase the number of women in governance roles and to ensure greater representation of women as advisors (Adams & Ferreira, 2009; Brahma et al., 2021; Conyon & He, 2017; Gharbi & Othmani, 2023).

Female presence on corporate boards positively correlates with several financial performance metrics, such as Return on Equity (ROE), Return on Assets (ROA), and operating profit ratios (Conyon & He, 2017). Recent studies have provided greater clarity on this issue. For instance, Bhatia and Gulati (2021) and Birindelli et al. (2024) confirmed that women on boards positively influence bank performance, with the latter study further suggesting that women on boards help reduce the negative effects of greenwashing on the financial performance of banks in Europe. Conversely, Adusei et al. (2017), Shehata et al. (2017), and Chijoke-Mgbame et al. (2020) have identified a negative impact of gender diversity on the EPS of banks.

Proença et al. (2020) suggest a threshold (>14% gender diversity) that determines when gender diversity positively impacts the link between political connections and bank performance. The advantage of gender diversity on bank efficiency and overall performance has been thoroughly documented in subsequent studies, including those by Boadi et al. (2022). Elgadi and Ghardallou (2022) reported mixed or negative results in some cases, which are the only exceptions to this trend.

Recent studies by Hazaea et al. (2023) and Adeabah et al. (2019) demonstrate a “threshold effect” whereby gender diversity positively influences bank efficiency. It has been suggested that, beyond this point, the effect may either stabilise or diminish. The findings of Mazzotta and Ferraro (2020) and Gharbi and Othmani (2023) indicate a positive relationship between gender diversity and financial performance and support gender quotas as a mechanism to increase the fair representation of women on boards. Additionally, Andoh et al. (2023), Sbai and Ed-Dafali (2023), and Paolone et al. (2024) identified a positive impact of gender diversity on bank performance, attributable to improved risk management and ESG (environmental, social, and governance) performance. Conversely, however, Adusei et al. (2017), Shehata et al. (2017), and Chijoke-Mgbame et al. (2020) have found contradictory results regarding the positive influence of gender diversity on bank performance and the extent to which the impact of gender diversity is moderated by the environment or organisation in which it occurs.

Given the breadth of evidence and the theoretical grounding in agency and stakeholder theories, this analysis posits the following hypothesis:

H1-a. 

There is a significant relationship between EPS and WIB.

2.3. CSD and Bank Performance

Technology has transformed the way we bank digitally; however, that transformation has not only increased innovation but also introduced many new cyber risks to the banking industry. These new risks can jeopardise the operational reliability and confidence of banking institutions. Disruption in operations through either lost access or misappropriated access to digital banking services can lead to a decline in consumer confidence and affect how well banks operate. Data breaches create legal issues by exposing client data. Cyber and operational risks are interconnected due to failures of governance and oversight, as illustrated by research conducted by Jin et al. (2023), Benaroch et al. (2012), and Walton et al. (2021).

Cyber incidents can influence various aspects of banking by affecting how banks manage their finances and overall operations. Studies by Xu et al. (2019) and Banker and Feng (2019) show changes in how banks handle their earnings through real earnings management when they have experienced cyberattacks, along with an increase in turnover for chief information officers, respectively. Therefore, cyber threats impact all parts of the banking system, from depositors to investors.

Bakker and Streff (2016) state that the impact of cybersecurity on a bank’s profitability, along with the effectiveness of risk reporting, are key topics of discussion regarding cybersecurity. Cyber espionage and cyberattacks from the past highlight the importance of addressing vulnerabilities within the financial services industry as part of national security. Jo (2017) and Tweneboah-Kodua et al. (2018) provide information on how security-related investments are influenced by both the dynamic nature of the marketplace and the economy.

Cyber threats require both institutional response strategies and the industry’s wider move towards increasing cash reserves as a precaution (Garg, 2020). In Jordan, banks encounter cyber vulnerabilities caused not only by existing cybersecurity measures but also by employee cyber behaviours. A comprehensive approach to managing the risks related to cybersecurity governance is essential to reduce these risks. Current research (Hamour, 2023) continues to highlight the importance of a persistent commitment to managing organisational cybersecurity risks, despite ongoing challenges.

A superior level of cybersecurity is a crucial strategic tool for banks to remain competitive. According to the Resource-Based View (RBV), cybersecurity represents a company’s resources that can offer a competitive edge for the bank by safeguarding customers’ access to banking services from cyberattacks. Banks will be able to reduce digital risks and protect their digital assets from cyber threats, as well as identify and capitalise on digital technologies for greater efficiency, increased profitability, and reduced risk using advanced cybersecurity technology (Kraaijenbrink et al., 2010; Furr & Eisenhardt, 2021).

The agency theory framework can be used to study how corporate governance affects financial performance through an examination of the principal-agent relationship between interested parties and the potential differences in their objectives, as well as their asymmetry of information, which may lead to conflicts of interest. Recent studies by Cortez and Dekker (2022) and Jiang et al. (2022) highlight the role of cybersecurity risk disclosure in reducing information asymmetry and aligning management and shareholder interests.

As demonstrated by both Eaton et al. (2019) and D’Arcy and Basoglu (2022), stakeholders are increasingly demanding greater transparency and accountability in managing cyber risks. Employing governance mechanisms, such as independent directors and IT governance, is essential within the agency theory framework, as highlighted in both studies. Research further indicates that companies offering higher levels of disclosure generally perform better financially, as positive perceptions by investors—linked to transparency—boost investor confidence (Albitar et al., 2020).

Empirical evidence from the banking sector backs this perspective: Elsayed et al. (2024) found that in the MENA region, including several GCC markets, banks with higher cybersecurity disclosures—indicating better cyber-risk governance and oversight—tend to be more profitable (ROA). This highlights cybersecurity as an important governance mechanism that adds value.

Although detailed discussions have examined the impact of cybersecurity on various aspects of banking operations, no recognised study has yet linked CSD to the bank’s performance. The absence of such research has led to the hypothesis that a significant relationship exists between the bank’s performance and its CSDs. This study will help explain how cybersecurity practices and disclosures influence the bank’s financial performance and contribute to the debate on corporate governance and operational integrity in the digital age.

Considering this, the study posits the following hypothesis:

H1-b. 

There is a significant relationship between bank performance and CSD.

2.4. The Moderating Role of CSD

Research has established that including female board members enhances ethical standards and stakeholder-inclusive decisions in businesses, especially regarding cybersecurity risk, which emphasizes managing and disclosing such risks. The studies by Chen et al. (2016) and Sila et al. (2016) clearly show that female board members bring unique perspectives and values that positively shape board governance. Additionally, Liao et al. (2015) and Radu and Smaili (2021) offer further evidence that boards with a higher proportion of female directors are more inclined to adopt a transparent approach to cybersecurity, environmental, and social governance.

As cyber threats increase and stakeholders demand more detailed reporting and accountability for their investments, the importance of cyber governance is also rising. Pendley (2018) highlighted a shift towards a culture of transparency in cyber governance, which will require sufficient representation of women on boards to address emerging cyber issues, as women tend to have heightened risk awareness and adopt a more moralistic leadership style.

The theory of agency, the theory of resources, and other theorists believe that having women on the board increases stakeholders’ confidence because of increased engagement with the board and increased accountability to stakeholders regarding cybersecurity (Krus, 2012; Mazumder & Hossain, 2023). In Canada and the United States, for example, regulatory entities have mandated that board members be knowledgeable and take appropriate actions to mitigate the risk presented by cybersecurity (Gao et al., 2020; Radu & Smaili, 2021).

Research from Mazumder and Hossain (2023) shows that having more female directors on corporate boards and banks enhances transparency around cybersecurity. This suggests that increasing female representation positively influences board governance concerning cybersecurity risks, thereby improving cybersecurity oversight and fostering a more open culture on cyber risk management through female director involvement. In this context, gender diversity in the boardroom significantly boosts cybersecurity oversight and is therefore vital to the success of any online financial institution today.

Given the evidence and theoretical grounding, the following hypothesis is posited:

H2. 

There is a moderate impact of the CSD on the relationship between banks’ performance and WIB.

3. Research Design

3.1. Study Sample

This study examines banks in Jordan to determine the proportion of women serving on boards, how this influences bank performance, and the moderating role of CSD. Jordan’s banking sector has been selected due to its strategic position amidst evolving governance standards and regional financial strategies (Ismail & Ahmed, 2023). Additionally, the sector is a significant area for research on gender diversity in boardrooms, especially given the increasing importance of cybersecurity. The reasons for choosing Jordanian banks are twofold: first, the diverse banking institutions provide a unique opportunity to explore governance issues from a holistic perspective owing to their varied structures (Bhagat & Bolton, 2019); and second, the swiftly changing cybersecurity landscape heightens the importance of studying the link between board composition and financial performance.

The Jordanian banking sector comprises 15 banks with valid Central Bank of Jordan licences during the period of this study. All banks in our sample are from this list, comprising 12 traditional banks and three Islamic banks. Including all these banks in the sample enhances its internal validity for generalising findings about the Jordanian banking industry and helps prevent sample selection bias. This panel covers the years from 2010 to 2022 and includes 195 observations (the number of banks times the number of years), providing a balanced longitudinal dataset that enables analysis of variations within each bank over time.

This study has been conducted from 2010 to 2022, a period selected because it follows the financial crisis and reflects the growing importance of cybersecurity due to technological advancements (Mohammed, 2015), which directly impact corporate governance and financial reporting. Consequently, this is an ideal point to explore the relationship between a bank’s board membership, cybersecurity measures, and the performance of operations.

Because the data were collected manually to ensure accuracy and authenticity, the dataset was assembled carefully, providing a solid foundation for the study. Annual reports containing metric analysis were manually compiled from multiple sources. Extensive analysis was conducted to identify all disclosures relevant to the study and its findings on cybersecurity metrics. The financial performance metric was compiled from detailed financial statements. Regarding the collection of metrics, the study emphasises empirical rigour, which will be discussed in more detail in the Variable Measurements section of this paper.

3.2. Study Models and Variable Measurements

The models presented in this section are used to examine the relationships among earnings per share, female representation on boards, and cybersecurity disclosure. Specifically, the first equation tests H1 by analyzing the association between women on boards and bank performance. The second equation examines the relationship between cybersecurity disclosure and bank performance, thereby paving the way for the third equation. The third equation tests H2 by assessing whether cybersecurity disclosure moderates the relationship between bank performance and women on boards.

$${\mathrm{EPS}}_{\mathrm{it}}={\mathsf{\alpha }}_{\mathrm{it}}+{\mathsf{\beta }}_1{\mathrm{WIB}}_{\mathrm{it}}+{\mathsf{\beta }}_2{\mathrm{Audit}\mathrm{quality}}_{\mathrm{it}}+{\mathsf{\beta }}_3{\mathrm{Leverage}}_{\mathrm{it}}+{\mathsf{\beta }}_4{\mathrm{Age}}_{\mathrm{it}}+{\mathsf{\beta }}_5{\mathrm{Size}}_{\mathrm{it}}+\mathsf{\epsilon }$$

(1)

$${\mathrm{EPS}}_{\mathrm{it}}={\mathsf{\alpha }}_{\mathrm{it}}+{\mathsf{\beta }}_1{\mathrm{CSD}}_{\mathrm{it}}+{\mathsf{\beta }}_2{\mathrm{Audit}\mathrm{quality}}_{\mathrm{it}}+{\mathsf{\beta }}_3{\mathrm{Leverage}}_{\mathrm{it}}+{\mathsf{\beta }}_4{\mathrm{Age}}_{\mathrm{it}}+{\mathsf{\beta }}_5{\mathrm{Size}}_{\mathrm{it}}+\mathsf{\epsilon }$$

(2)

$${\mathrm{EPS}}_{\mathrm{it}}={\mathsf{\alpha }}_{\mathrm{it}}+{\mathsf{\beta }}_1{\mathrm{WIB}}_{\mathrm{it}}+{\mathsf{\beta }}_2{\mathrm{CSD}}_{\mathrm{it}}+{\mathsf{\beta }}_3{\mathrm{CSD}}_{\mathrm{it}}\times {\mathrm{WIB}}_{\mathrm{it}}+{\mathsf{\beta }}_4{\mathrm{Audit}\mathrm{quality}}_{\mathrm{it}}+{\mathsf{\beta }}_5{\mathrm{Leverage}}_{\mathrm{it}}+{\mathsf{\beta }}_6{\mathrm{Age}}_{\mathrm{it}}+{\mathsf{\beta }}_7{\mathrm{Size}}_{\mathrm{it}}+{\mathsf{\beta }}_8{\mathrm{Islam}}_{\mathrm{it}}+\mathsf{\epsilon }$$

(3)

where:

• Dependent variable:

EPS (Earnings per Share) is one of the most important measures used to assess bank performance. EPS is calculated by dividing a bank’s net income by the total number of shares outstanding (issued), enabling a precise evaluation of its overall financial condition and profitability.

• Independent variables:

A bank’s Board of Directors (also known as WIB) can be evaluated by the proportion of women on the board. WIB is defined as the percentage of women board members relative to the total number of board members and reflects how banks currently conduct governance and diversity practices. (Consler et al., 2011; Lückerath-Rovers, 2013; Nielsen & Huse, 2010).

CSD is analysed using content analysis of bank annual reports, a painstaking process that counts the available keyword variations. The following steps illustrate the procedures:

-

keyword-based content analysis and score construction

The operation of CSD involves systematic, manifested content analysis through keyword frequency counts found in annual reports published by each bank. Established content analysis protocols, particularly those concerning coding (keyword dictionaries) and counting rules, were formulated beforehand to enhance the objectivity and reproducibility of the collected data (Neuendorf, 2002). The CSD dictionary was developed from prior cybersecurity disclosure and risk-reporting literature such as Gao et al. (2020), Li et al. (2018), Mazumder and Sobhan (2021), Radu and Smaili (2021), and Mazumder and Hossain (2023), and contains 50 keyword variants capturing cybersecurity risks, incidents, and safeguards (e.g., “cyber”, “cyber-attack”, “data breach”, “malware”, “phishing”, “encryption”, and “firewall”; full list in Appendix A).

-

The content analysis procedure involves manual, rule-based counting.

The annual report for each bank year was manually downloaded as a PDF and analysed using a standardised keyword search procedure (as detailed in Appendix A). Each keyword was then searched individually using a standard, built-in search function at the top of the PDF, and the number of occurrences was recorded. The rule-based manual approach reduces interpretive discretion in code decisions because they are based on predefined terms rather than a subjective classification system. The approach also avoids issues such as algorithmic problems and preprocessing that can affect automated text methods (e.g., extraction errors; “algorithm errors”) (McKenny et al., 2018). Moreover, measuring the frequency of these specified terms enables us to create a more reliable database, as it is less affected by variations in structure and grammatical errors in the sentences conveying information from these reports (Mazumder & Sobhan, 2021; Unerman, 2000; Mazumder & Hossain, 2023).

-

Deriving the CSD score from keyword frequencies.

Let $Coun{t}_{i,t,k}$ denote the number of occurrences of the keyword $k$ in bank $i$. The annual report in year $t$, where $k=1,\dots ,K$ and $K$ is the number of keywords in Appendix A. The baseline CSD score is computed as the total number of keyword hits across the dictionary:

$$CS{D}_{i,t}= {\displaystyle \sum }_{k=1}^{k}Coun{t}_{i,t,k}$$

This aggregation yields a continuous disclosure-intensity measure that captures the extent to which cybersecurity-related terminology appears in the report. Because the dictionary is fixed and the counting rule is mechanical, the measure can be replicated by any researcher applying the same keyword list and counting protocol (Neuendorf, 2002).

-

Reliability considerations and justification

Reliability for content analysis refers to the degree to which a measurement method produces the same result after being used multiple times. When human judgment is used to measure something, the most common way to evaluate this reliability is through intercoder agreement (Neuendorf, 2002). However, in our setting, the CSD is generated from a manifest and a rule-based frequency count that relies on a predetermined dictionary rather than subjective interpretations of meaning. The principal threats to consistent results are procedural, e.g., the use of separate keyword lists or inconsistent counting rules, rather than arising from differences in opinion regarding the significance of the written material. This involves offering full transparency for the keyword dictionary provided in Appendix A and providing clear documentation of the basic counting protocol, so that others can easily replicate our methods. Additionally, reliability researchers advise that although reliability coefficients can assist or mislead analysts, their accuracy depends on how they are utilised; these coefficients can mislead when applied inappropriately or in situations where agreement is artificially created by the design of measurement instruments (Krippendorff, 2004). Therefore, for deterministic keyword counts, transparency of the coding protocol and replicability of the procedure are the most informative safeguards for measurement robustness (Neuendorf, 2002).

• Control variables:

The control variables for this study have been operationalised based on prior academic research, including Larmou and Vafeas (2010), Margaritis and Psillaki (2010), Rossi (2016), and Dao and Ta (2020). Audit as a proxy for audit quality is a binary variable coded with ‘1’ for banks examined by one of the ‘Big Four’ accounting firms and ‘0’ for all other banks. Financial leverage is calculated as total liabilities divided by total assets, indicating the extent of a bank’s financial leverage. A bank’s age is measured using the natural logarithm of the total number of years since establishment, allowing an evaluation of its institutional maturity. Similarly, the total asset value of a bank is used to determine its size through the natural logarithm, establishing a measure of operational scale. Islam is a binary variable coded 1 for Islamic banks and zero otherwise.

Gender is a sensitive attribute; therefore, we regard any WIB results strictly as governance associations specific to their context, and we avoid language that suggests inherent gender differences. We believe that observable effects are more likely to be due to institutional or organisational factors (e.g., role allocation, tokenism, governance practices) rather than to gender alone. We present our findings in a non-stigmatising way that aligns with established research ethics principles.

4. Empirical Findings

4.1. Descriptive Results

As shown in

Table 1. Descriptive analysis. (A) Continuous variables (B) Binary variables.

Panel A
Variables	Mean	SD	Min	Max	Q1	Median	Q3
EPS	0.2054	0.1388	−0.0300	0.6610	0.1100	0.1660	0.2950
WIB	0.0737	0.0776	0.0000	0.2700	0.0000	0.0760	0.0900
CSD	8.9158	11.8457	0.0000	56.0000	0.0000	3.5000	14.0000
Leverage	0.8668	0.0381	0.6716	0.9304	0.8448	0.8654	0.8941
Age	3.6668	0.4008	2.6391	4.5109	3.4657	3.6889	3.9318
Size	21.6207	0.8931	19.6746	24.0416	21.0240	21.5563	21.8642
Panel B
Variables	If banks are audited by the Big Four			If banks are audited by non-Big Four
Audit quality (freq)	184			11
Audit quality (%)	94.36			5.64
Variables	If banks are commercial			If banks are Islamic
Islam (freq)	156			39
Islam (%)	80			20

Note: Variables trimmed at 1st and 99th percentiles. Source: Authors’ creation.

, EPS (Earnings Per Share) reveals that the average across all firms is 0.2054, with a standard deviation of 0.1388, suggesting that these firms are experiencing an unusually high level of profitability. The EPS figures for firms within this group vary widely, likely due to a combination of their market position and operational efficiency.

The presence of WIB, on average, shows that firms’ governance structures include 8% WIB, suggesting that firms may have different policies regarding diversity and inclusion. The CSD variable (8%) has a mean score of 8.9158; however, CSD scores exhibit a high degree of variability, with a standard deviation of 11.8457, indicating that companies adopt substantially different practices in communicating about cybersecurity. Additionally, the range of CSD scores (0 to 56) reveals that while some companies openly communicate their cybersecurity practices, others are less transparent, resulting in significant variation between firms and suggesting either a diversity of approaches based on perception or the potential influence of regulation over the CSD score.

The leverage ratio indicates a preferred level of debt financing within the banking industry; the average leverage was recorded at 0.8668 with a low standard deviation of 0.0381. Consequently, maintaining a similar leverage level across the industry might suggest a consensus among banks regarding the ideal capital structure. Including the age and size of a firm (averages: age = 3.6668, size = 21.6207) offers deeper insights into the composition of banks. With age (SD = 0.4008) and size (SD = 0.8931) displaying moderate variance, it is evident that the industry consists of both established banks and new entrants of varying sizes, creating a dynamic market with diverse growth trajectories and strategic focuses. The dataset also shows that a large majority of firms (94.36%) opt for Big 4 auditors, highlighting the industry’s strong reliance on reputable auditors and possibly underscoring the importance of quality audits and robust governance standards. Conversely, a small proportion of firms (5.64%) engage auditors outside the Big 4, indicating different audit preferences or strategic considerations. Additionally, the sample comprises 80% commercial banks and 20% Islamic banks.

4.2. Correlation Analysis

This study examines EPS and its relationship with several other independent variables, showing how governance, financial, and firm characteristics interact in a complex way. As shown in

Table 2. Pairwise Correlation Matrix for EPS, WIB, CSD, and Control Variables.

Variables	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)
(1) EPS	1.000
(2) WIB	−0.124 *	1.000
(3) CSD	−0.040	0.282 ***	1.000
(4) CSD × WIB	−0.068	0.585 ***	0.760 ***	1.000
(5) Audit Quality	−0.166 **	−0.113	0.147 **	0.090	1.000
(6) Leverage	0.081	0.062	0.289 ***	0.228 ***	−0.124 *	1.000
(7) Age	0.231 ***	−0.038	0.290 ***	0.133 *	−0.114	−0.078	1.000
(8) Size	0.661 ***	0.017	0.298 ***	0.162 **	−0.011	0.214 ***	0.542 ***	1.000
(9) Islam	0.032	0.216 ***	0.001	0.038	−0.011	−0.235 ***	0.223 ***	0.157 **	1.000

Note: This table presents Pearson correlation coefficients between the study’s main variables. Significance levels are indicated as *** p < 0.01, ** p < 0.05, * p < 0.1. Source: Authors’ creation.

, there is a statistically significant negative correlation between WIB and EPS at −0.124 *, which, although not very strong, suggests a link between having more women on the board and lower earnings per share. This negative correlation may relate to differences in governance or decision-making processes used by companies with women on the board compared to those without women. Similarly, as with CSD and EPS, CSD has a non-significant negative correlation with EPS (−0.040), indicating that high transparency in cybersecurity practices may not directly affect EPS or financial returns. Furthermore, the correlation between CSD and WIB is positive and statistically significant (0.282 *), suggesting that banks with higher female board representation tend to have higher levels of cybersecurity disclosure. In contrast, the interaction term (CSD × WIB) shows a weak and non-significant negative correlation with EPS (−0.068).

By examining the interaction among audit quality, leverage, company age, company size, and Islamic banking focus, we can gather more evidence about these various relationships. The correlation is negative (−0.166 **) between audit quality and earnings per share (EPS), which could suggest that higher audit quality results in lower EPS due to conservative accounting practices or possibly because of audit adjustments stemming from conservative decisions.

Furthermore, the link between leverage and earnings per share, although weak (0.081), indicates that leverage does have an impact on EPS but only slightly. Conversely, both company age (0.231 ***) and company size (0.661 ***) are positively related to EPS, suggesting that larger and older companies tend to generate higher EPS, which is probably due to their stronger market presence and more efficient operating practices compared to younger or smaller competitors.

Regarding the focus on Islamic banking, there is a small but positive relationship with EPS (0.032), suggesting little difference in earnings between commercial banks and Islamic banks. Therefore, the results emphasise the complex influence of various variables on a company’s overall financial performance, highlighting the significance of audit quality, governance structures, and other company-specific attributes in determining earnings performance.

The relationships between the independent variables in Table 2 indicate that there are no major collinearity issues initially; however, it is important to note that a comprehensive collinearity test was conducted, and the results show that, in terms of the average VIF values across the various models in the study, these values exceed the commonly used threshold of 5. This suggests the possibility of multicollinearity issues (Hair et al., 2019). Therefore, as a first step, to minimise the influence of extreme values on the initial set of independent variables, they were set to their 1st and 99th percentile values. Alongside the interaction terms for the regression models, the independent variables were standardised to allow for direct comparison, thereby reducing the potential for collinearity (García et al., 2016). Subsequently, following these adjustments, a further assessment of the collinearity diagnostics indicated a notable improvement. The average VIF values for all the regression models are now below the threshold of 5, demonstrating the effectiveness of the measures implemented.

4.3. Regression Results

The Breusch-Pagan Lagrange Multiplier test, developed by Breusch and Pagan (1980) and utilised by Zaman (2000) and Halunga et al. (2017), was performed to determine the most appropriate estimation method for each hypothesis in this study. Due to the results of the Breusch-Pagan test—a statistically significant Prob > chi-bar2 at the alpha = 0.01 level—Random Effects should be adopted as the estimation framework for this research. Hausman specification tests were then carried out to compare the Random and Fixed Effect models (Amini et al., 2012). According to the Hausman test, Random Effects were again found not to differ significantly from Fixed Effects since Prob > chi2 was not greater than the alpha = 0.05 level for most comparisons. The research should, therefore, continue using Random Effects as the estimation method. This testing approach was designed to provide a comprehensive and scientifically robust basis for the empirical findings of this study.

According to the Random Effects regression models, there are significant findings regarding the relationship between banks’ EPS and various independent variables; this data supports the hypotheses of this research. Model 1 of

Table 3. Random effect results.

Variables	Model (1)	Model (2)	Model (3)	Model (4)
WIB	−0.1410 **		−0.2162 ***	−0.2129 ***
	(0.0552)		(0.0766)	(0.0774)
CSD		−0.1911 ***	−0.2742 ***	−0.2779 ***
		(0.0577)	(0.0837)	(0.0842)
CSD × WIB			0.1815 **	0.1811 **
			(0.0915)	(0.0917)
Audit quality	−0.9273 ***	−0.6048 **	−0.7486 ***	−0.7411 **
	(0.2665)	(0.2837)	(0.2854)	(0.2886)
Leverage	−0.1216 *	−0.0836	−0.0783	−0.0849
	(0.0679)	(0.0705)	(0.0701)	(0.0722)
Age	−0.3316 ***	−0.2357 *	−0.2385 *	−0.2380 *
	(0.1177)	(0.1284)	(0.1307)	(0.1392)
Size	0.8454 ***	0.8938 ***	0.9320 ***	0.9513 ***
	(0.1215)	(0.1297)	(0.1325)	(0.1413)
Islam				−0.1396
				(0.3120)
Constant	0.8750 ***	0.5675 **	0.6878 **	0.7922 **
	(0.2718)	(0.2891)	(0.2907)	(0.3843)
R-square	11.4%	13.61%	17.33%	17.44%
Wald chi2	68.50 ***	66.97 ***	74.30 ***	69.51 ***
N	195	190	190	190

Note: *** p < 0.01, ** p < 0.05, * p < 0.1. Standard errors are presented in parentheses. Source: Authors’ creation.

shows a strong negative correlation between WIB and EPS, with a coefficient of −0.1410 that is statistically significant at the alpha = 0.05 level. Based on the hypothesis developed for this research (H1-a), which posits a relationship between EPS and WIB but does not specify its direction, this result suggests that an increase in women on boards is associated with a decrease in EPS. Consequently, this provides empirical support for H1-a, indicating a strong link between women on boards and companies’ financial performance.

Research hypothesis H1-B is supported by Model (2). It tests whether there is a relationship between CSD and EPS. The model shows a coefficient of −0.1911 at a significance level of one percent; statistical analysis indicates that this negative coefficient confirms H1-b, and that the relationship between CSD and EPS is significant. The results suggest that a high level of CSD contributes to lower-than-expected EPS. Therefore, based on these findings, H1-b should be supported, as there is strong evidence of a meaningful relationship.

A key hypothesis of the current study is outlined in H2, which examines the role of CSD as a moderator in the relationship between WIB and EPS. Evidence from Model (3) supports this main hypothesis, showing a significantly negative coefficient (−0.2162) for WIB at the 1% level, and a large positive coefficient (0.1815) for the interaction term CSD × WIB at the 5% level. Thus, although WIB has a direct negative effect on EPS, the presence of CSD reduces this negative impact. However, the overall negative effect (−0.2742 + 0.1815 = −0.0927) remains negative when compared to the individual effect of WIB (−0.2162). These results support the hypothesis that CSD can diminish, to some extent, the adverse effect of WIB on EPS. This further suggests that having a diverse board regarding gender and providing transparency about cybersecurity practices can lessen the harmful or neutral impact of WIB on financial performance. Consequently, H2 is supported; CSD appears to be a significant moderating factor in the WIB-EPS relationship. These findings underscore the complex and interconnected nature of the elements influencing banks’ financial performance.

Including the control variable of Islamic Banking in Model (4) confirms what was previously found in the past three Models regarding WIB, CSD, their interactions, and EPS. The introduction of the variable “Islam,” which indicates the difference between an Islamic Bank and a Conventional Bank, does not significantly alter the previously established correlation between WIB and CSD, nor does it change the relationship between WIB, CSD, and EPS. The consistency of results throughout this study provides evidence of reliability and validity; therefore, the influence that Board Gender Diversity and Cybersecurity Transparency have on Financial Performance remains the same, regardless of the type of Banking Model used to evaluate them. Consequently, the findings of this study are applicable across different types of Banking Systems.

Examining the control variables, the results reveal a negative correlation between EPS and audit quality across all models (with coefficients ranging from −0.9273 in Model (1) to −0.7411 in Model (4)), significant at the 1% to 5% level, indicating a strong relationship. It appears that when audits are conducted with greater care, there tends to be lower EPS due to conservative adjustments and identification of financial discrepancies that result from high-quality audits.

Table 3 indicates that the relationship between Leverage and EPS, as outlined in Model (1), was marginally significant at the 10% level, with a coefficient of −0.1216. However, the findings suggest that this relationship was not significant in any of the other three models; this implies that the connection between these two variables may depend on the inclusion of additional variables or on model dynamics. Additionally, the bank’s age consistently exhibits a negative relationship with EPS across all models. The coefficient in Model (1) was −0.3316, whereas in Model (4) it was −0.2380, with significance levels varying from 1% to 10%. It seems that although older banks have a larger market share, they tend to produce lower levels of EPS; this could be due to several factors, including ongoing costs associated with older banks’ operations, a slower response to changes in the marketplace, or the natural progression through various stages of their business development.

For each of the models analysed above, there was an overall positive, statistically significant relationship between bank size and EPS; the coefficients ranged from 0.8454 to 0.9513, all statistically significant (p < 0.01). Therefore, as banks’ sizes increase, their EPS also tends to be higher; this may be because larger banks benefit from economies of scale, monopoly power, and/or access to diverse sources of revenue (which results in higher financial returns).

To ensure the stability of our results, we replaced our primary measure of firm performance with two alternative measures—return on assets (ROA) and return on equity (ROE)—and re-estimated all datasets across these four different models. Although this adds an extra layer of analysis for each model, the results remain qualitatively similar to those reported earlier and continue to show the same overall directionality; all results from the full robustness set are available from the author(s) on reasonable request.

5. Discussion

The current study explores the complex relationship between gender diversity on bank boards and bank performance, with cybersecurity disclosure acting as a moderating factor. The empirical findings reveal a significant interaction among variables, especially with EPS (earnings per share), which is influenced by transparent practices within banks’ cybersecurity functions.

Our study revealed a significant negative correlation between the presence of women on corporate boards and earnings per share, contradicting the popular belief that board gender diversity positively influences governance and financial performance. In our view, the previously held assumption that board gender diversity offers a positive benefit needs to be reconsidered. While many researchers have provided evidence that women on boards enhance financial and technical efficiencies, including: García-Meca et al. (2015); Bhatia and Gulati (2021); Elnahass et al. (2023); Al-Jaifi (2020); Mazzotta and Ferraro (2020); Chijoke-Mgbame et al. (2020); Scholtz and Kieviet (2018); and Boadi et al. (2022); the findings here suggest that the actual impact of gender diversity may vary considerably depending on the industry, particularly in the banking sector. These differences can be further influenced by country-specific institutional, cultural, and regulatory factors, such as those present in the Jordanian banking environment, where governance roles, board influence, and decision-making authority may differ in practice.

This emphasises the need for further research to better understand the overall impact of gender diversity in corporate governance, as current studies may not fully account for the influence of sector-specific factors on corporate governance.

From an agency theory perspective, increased gender diversity on bank boards may lead to stronger monitoring and greater risk aversion, which can constrain short-term performance measures such as earnings per share. This effect is likely intensified in the banking sector due to high regulatory pressure and limited managerial discretion. From a resource dependence theory standpoint, the performance benefits of board diversity are context-dependent; in highly regulated environments, additional expertise and external linkages may not immediately result in higher earnings. Consequently, the negative relationship between women on boards and EPS reflects a sector-specific and conditional outcome rather than ineffective governance.

Consistent with agency theory and resource dependence theory, these findings indicate that the effectiveness of board gender diversity depends on how governance roles are performed within particular institutional environments, particularly when regulatory oversight and risk considerations predominate in board decision-making.

Our research expands on the claims made by Ghosh and Ansari (2018), as well as the results of Proença et al. (2020), which suggest that CSD is a significant moderating variable that influences the relationship between WIB and financial performance. Consequently, despite an extensive body of literature confirming the strong inverse relationship (negative correlation) between WIB and a company’s EPS, the impact of gender diversity may go beyond what is demonstrated by the direct effect alone on EPS. Furthermore, the coefficient of determination for gender diversity also indicates the interaction between gender diversity and EPS. Therefore, there may be circumstances where gender diversity can positively affect financial performance.

Additionally, this research has enhanced our understanding of existing research on cybersecurity transparency beyond what Bakker and Streff (2016) accomplished in their study, where they identified a link between formalised risk reporting processes and improved financial performance. While this demonstrates that formalised, detailed disclosures can lead to positive outcomes, they may also emphasise potential cybersecurity risks, which could negatively influence investor perceptions.

Theoretical and Practical Implications

Theoretical implications. Besides providing insights on board diversity and financial performance, this study highlights the importance of a more comprehensive approach to understanding how diverse boards operate and generate results in relation to their firm’s disclosure practices. Hence, it challenges the idea that the only benefit of diversity is single-dimensional (i.e., financial performance) and instead stresses the need to consider the overall functioning of diverse boards and the multiple aspects of diversity that contribute to a company’s overall success.

Practical implications: Due to the lack of comprehensive cybersecurity declaration guidelines for financial institutions, this research helps clarify what different banking institutions should anticipate from Board Gender Diversity in terms of their performance when faced with perceived risk, as demonstrated through Study-Based Risk Reporting. Additionally, providing detailed cybersecurity information to stakeholders can be a valuable tool for banks to counteract negative perceptions of Board Gender Diversity and to leverage its contributions to improve organisational governance. Therefore, the synergy between Board Gender Diversity and disclosure practices is crucial, significantly influencing how effectively banks are governed. Based on this finding, regulatory authorities and policymakers should promote or mandate more detailed cybersecurity disclosures from banks to strengthen governance and, consequently, their financial performance.

6. Conclusions

This research examined 15 Jordanian banks from 2010 to 2022 to understand the complex relationship between the WIB and EPS, and how the CSD moderates that relationship. The findings provide empirical support for a statistically significant negative relationship between the proportion of women on bank boards (WIB) and earnings per share (EPS). However, the study also confirms that cybersecurity disclosure (CSD) plays a critical moderating role, as it significantly reduces the negative impact of board gender diversity on financial performance.

The empirical findings provide a broader understanding of the Jordanian banking sector, offering more detailed insights into the relationship between corporate governance, gender diversity, and cybersecurity transparency.

We find that higher WIB is associated with lower EPS in Jordanian banks during 2010–2022; this observation reflects governance dynamics in this context and does not suggest inherent performance differences based on gender. Additionally, the analysis shows that boards with women have a more complex relationship with banks’ EPS than those with fewer female directors. This indicates that increasing female representation without supporting mechanisms—such as inclusion in key committees or access to strategic information—may restrict their influence on performance indicators like EPS.

CSD plays a vital role in the transparency and risk management processes of banks; therefore, CSD maintains a close relationship with both areas of bank operations. An interesting point is that CSD is a factor that alleviates the negative impact of board diversity on EPS. When banking institutions are transparent about their cybersecurity disclosures, this diminishes the negative effect that a board’s gender diversity has on the bank’s earnings. The reason for this reduction in the adverse impact of a gender-diverse board on EPS is attributed to increased investor confidence and a lower level of risk linked to a banking institution’s transparent cybersecurity practices.

These results highlight the importance of integrating board diversity initiatives with broader governance improvements, particularly in heavily regulated sectors like banking. The strength of these connections, particularly with respect to banking-related characteristics, demonstrates that these findings are applicable across all types of banking. Consequently, this research not only emphasises the significant impact of board structure and transparency on financial performance but also recognises cybersecurity disclosure as a strategic governance tool in a rapidly changing digital environment.

To conclude, the findings of this research offer further insight into current debates around corporate governance and how increasing the number of women on corporate boards may impact the earnings per share of companies, subject to their specific circumstances. We have demonstrated that strategic cyber transparency can act as a link between governance processes and financial performance objectives. We suggest establishing comprehensive policies for board diversity and strategic cyber transparency.

Based on these results, we recommend that Jordanian banks take practical steps to align their diversity and cybersecurity strategies. First, banks should integrate cybersecurity oversight into the board’s responsibilities, especially within audit and risk committees. Second, female board members should be actively involved in strategic decision-making, particularly in areas related to technology and governance, to increase the effectiveness of their participation. Third, banks are encouraged to develop and adopt standardised cybersecurity disclosure frameworks to enhance transparency and build trust with stakeholders. Finally, regulators such as the Central Bank of Jordan should consider issuing clear guidelines on minimum cybersecurity disclosure standards, thereby ensuring consistency and improving the governance climate across the banking sector.

The limitation on the number of observations in this study arises from the limited number of banks located in Jordan. Consequently, this restricts the broader applicability of the results for generalising across banking sectors. Future researchers can enhance the generalisability of their findings by expanding their geographic scope or investigating alternative locations to obtain potential results for other regions. Additionally, further qualitative research will provide additional context regarding the different systems of corporate governance and transparency when establishing complex relationships developed in this study.