When Tax Avoidance Meets Sustainability: ESG Disclosure and Firms’ Cost of Debt

Abstract

Understanding the factors that shape firms’ borrowing costs has become increasingly important amid growing concerns about corporate transparency, sustainability, and tax practices. Specifically, the study investigates the relationship between tax avoidance (TA) and the cost of debt (CoD) among non-financial firms listed on an emerging Stock Exchange, while examining the moderating effect of environmental, social, and governance disclosure (ESG disclosure) based on the Panel-Corrected Standard Error (PCSE) approach. The findings indicate a significant positive association between TA and CoD, suggesting that lenders perceive aggressive tax practices as a source of additional risk, which consequently increases borrowing costs. The results further show that ESG disclosure plays a moderating role in this relationship, as the positive effect of TA on borrowing costs becomes weaker among firms with higher levels of ESG disclosure. This implies that stronger ESG disclosure improves transparency and alleviates creditors’ concerns about firms’ tax-related behavior. The findings were further validated through robustness checks using alternative CoD measures, fixed-effects regression, and dynamic panel GMM estimations to address endogeneity concerns. The study contributes to the literature by providing evidence from an emerging market and highlighting the role of ESG disclosure in mitigating the negative financial effects of TA.

1. Introduction

The growing public scrutiny of corporate taxation and a series of high-profile tax-related scandals have intensified concerns regarding the implications of tax avoidance (TA) for firms’ financing decisions and risk assessments by creditors [1]. In particular, the cost of debt (CoD) has emerged as an important mechanism through which debt markets evaluate and discipline corporate tax behavior, especially in countries where tax revenues constitute a major source of government funding [2,3].

Simultaneously, the proliferation of environmental, social, and governance (ESG) information has provided lenders with a new information platform [4], in which ESG information is increasingly employed as an input for credit decisions [5,6]. As a result, TA, ESG disclosure, and the CoD have become increasingly interconnected within the broader discussions of sustainable finance, corporate transparency, and stakeholder accountability [6,7,8].

TA generally refers to the use of legal tax-planning strategies that reduce firms’ tax liabilities through the exploitation of loopholes and ambiguities in tax regulations [3,9,10]. From a shareholder value perspective, TA may enhance firm value by increasing after-tax cash flows, improving internal financing capacity, and reducing financing constraints, which implies implicitly lower CoD [11]. Under this view, tax savings strengthen firms’ ability to meet financial obligations and may therefore be viewed favorably by creditors. However, the benefits of TA are not universally accepted. While tax savings may improve liquidity, tax avoidance often involves complex and opaque transactions that can increase uncertainty regarding firms’ true financial positions. From an agency theory perspective, such complexity may exacerbate information asymmetry between managers, shareholders, and creditors, making it more difficult to distinguish efficient tax planning from opportunistic managerial behavior [12], which can increase the likelihood of audits and legal consequences [13,14,15]. Consequently, TA may be a function of successful tax planning and solvency measures or a reflection of agency and enforcement risks [16] that increases borrowing costs for firms while creditors remain unsure of the actual risk profile of tax-avoiding firms [17,18,19,20].

Existing research relevant to the present study can be grouped into three interrelated streams. The first stream examines the relationship between TA and debt financing costs. However, the evidence remains inconclusive. One group of studies supports the tax-saving perspective, suggesting that tax avoidance improves after-tax cash flows, strengthens internal liquidity, and enhances firms’ ability to meet debt obligations, thereby reducing borrowing costs [21,22,23]. Conversely, another group of studies supports the risk-exposure perspective, arguing that tax avoidance increases information asymmetry, regulatory uncertainty, audit risk, and concerns regarding managerial opportunism, leading creditors to demand higher risk premiums [13,15,20]. More recent evidence generally supports the latter argument, documenting a positive association between TA and CoD [3]. That is, adopting a well-developed non-financial disclosure can mitigate the tax sanction by narrowing the information gap between managers and creditors and improving the ease of understanding the tax risk–return profile [3,24]. Nevertheless, the persistence of conflicting findings suggests that creditor responses may depend on additional informational mechanisms that influence risk assessment.

The second stream investigates the role of ESG disclosure in debt markets. Prior studies consistently show that greater ESG disclosure improves transparency, reduces information asymmetry, and enhances stakeholder confidence, resulting in lower bond spreads and more favorable borrowing conditions [4,6,25,26]. These findings indicate that creditors increasingly incorporate non-financial information into lending decisions [3]. However, studies examining the relationship between ESG disclosure and tax avoidance provide mixed evidence. Some researchers argue that ESG disclosure constrains aggressive tax practices by strengthening accountability and stakeholder monitoring [27,28], whereas others suggest that ESG initiatives may be strategically employed to enhance legitimacy while firms continue to engage in opportunistic tax behaviors [29,30,31]. These mixed findings suggest that ESG disclosure may play a more complex role than simply promoting or discouraging TA.

The third and most closely related stream to the present study examines ESG disclosure as a mechanism that influences stakeholders’ interpretation of corporate actions. Emerging evidence suggests that ESG disclosure can mitigate the negative consequences of TA and other governance-related concerns by reducing information asymmetry and enhancing transparency [30,32,33,34]. However, most prior studies focus on firm value, market performance, corporate reputation, or alternative reporting mechanisms such as integrated reporting [30,32,34]. Comparatively little evidence exists regarding whether ESG disclosure influences creditors’ evaluation of TA risk and, consequently, the CoD.

Accordingly, while prior research has separately examined TA, ESG disclosure, and CoD, limited attention has been devoted to understanding their interaction. Moreover, the existing literature rarely distinguishes between the direct effects of TA and ESG disclosure and their combined influence on debt financing decisions. Although extensive research separately investigates TA, ESG disclosure, and debt financing costs [35], relatively limited attention has been paid to whether ESG disclosure moderates the relationship between TA and the CoD.

Prior research has taken an independent approach to TA and ESG disclosure, or emphasized alternative forms of disclosure, such as integrated reporting rather than ESG disclosure [3,36,37]. Also, most studies applied in developed markets with high investor protection and well-developed debt markets, such as the US, Western Europe, and China, to some extent [18,24,25,38]. In contrast, very few studies have been applied in emerging markets, such as India, Malaysia, Indonesia, the Middle East, South Africa, and the North African regions [7,32,34,39,40]. More specifically, evidence remains scarce regarding whether ESG disclosure moderates the association between TA and CoD, particularly in emerging economies characterized by elevated information asymmetry, evolving sustainability reporting practices, and bank-centered financial systems, like Egypt. Addressing this gap constitutes the primary objective of the present study.

Unlike prior studies that examine the TA–firm value relationship from the perspective of shareholders [32], this study adopts a creditor-oriented perspective by investigating how TA influences borrowing costs. Shareholders often view TA favorably because it reduces tax expenses and allows firms to retain more resources, which can improve cash flows and increase firm value [32,41,42]. In contrast, creditors may perceive aggressive TA as a signal of higher risk, as it can reduce transparency, increase managerial discretion, and expose firms to potential regulatory scrutiny [22,43,44]. As a result, lenders may require higher interest rates to compensate for the additional risk associated with such practices.

The Egyptian context provides a unique setting for examining these relationships. Egypt has recently experienced substantial reforms in both taxation and sustainability reporting. The introduction and development of the S&P/EGX ESG Index have increased the visibility of ESG practices among listed firms, while regulatory initiatives have encouraged greater sustainability disclosure [32,34]. Simultaneously, repeated amendments to tax regulations and the transition toward a self-assessment tax regime have heightened tax-related uncertainty and regulatory scrutiny [32]. These developments have increased the importance of transparency and risk assessment for both firms and creditors. In addition, Egypt’s financial system remains predominantly bank-centered, making creditors a critical stakeholder group in corporate financing decisions [30,32].

Taken together, the literature suggests that the effect of TA on debt financing costs cannot be fully understood without considering the informational environment in which creditors evaluate firms. Since ESG disclosure has emerged as an important transparency mechanism, examining its moderating role provides a natural extension of the existing literature and directly addresses the inconsistencies documented in prior studies. The distinction between ESG disclosure level and ESG disclosure quality is also important. While disclosure quality reflects the accuracy, credibility, and completeness of reported information, the present study focuses on ESG disclosure level, which captures the extent of ESG information made available to stakeholders. This focus is particularly relevant in emerging markets, where the availability and breadth of ESG information often represent the first step toward reducing information asymmetry and improving stakeholder decision-making [4,5].

To explain these relationships, this study adopts a multi-theoretical framework grounded in agency theory and signaling theory. Agency theory suggests that tax avoidance may increase agency conflicts and information asymmetry, leading creditors to demand higher risk premiums [3,14,45]. Signaling theory further argues that both tax practices and ESG disclosures communicate important information about managerial quality, transparency, and firm risk to external stakeholders [46,47]. Accordingly, ESG disclosure may influence how creditors interpret tax avoidance by providing additional signals regarding governance quality and long-term sustainability [4,5,47]. Taken together, these perspectives suggest that ESG disclosure can moderate the impact of tax avoidance on the CoD by reducing information asymmetry, improving perceived transparency, and shaping creditors’ risk assessments.

This study advances the existing body of knowledge by investigating the impact of TA on debt costs within an emerging market setting that features high levels of fiscal dependence on tax revenues and informationally sensitive debt markets [3,24]. Moreover, this study examines whether ESG disclosure serves as a moderating mechanism in the association between TA and the CoD, specifically examining whether ESG disclosure has a disciplining impact on tax behavior from the perspective of creditors.

By focusing on an emerging economy characterized by strong reliance on tax revenues, ongoing tax reforms, increasing ESG institutionalization, and a bank-centered financial system, the study contributes to the literature in several ways. First, it extends the understanding of how creditors evaluate TA in emerging markets. Second, it examines ESG disclosure as a moderating mechanism linking tax behavior and debt financing costs. Third, it integrates agency theory and signaling theory to provide a comprehensive framework for understanding the interaction between tax avoidance, non-financial disclosure, and creditor risk assessment [5,40], as well as offering policy implications that are relevant to ESG reporting and tax planning within emerging markets [32,34].

The rest of the study is organized as follows. Section 2 discusses the Egyptian contextual ramifications. Section 3 develops the underlying theoretical arguments and critically reviews prior research on TA, ESG disclosure, and the CoD, leading to the formulation of the study’s hypotheses. Section 4 describes the research design, sample selection, and variable measurement. Section 5 presents the research findings, while Section 6 presents the robustness tests and further analysis. The last section concludes the paper by summarizing the main results, highlighting their implications, and providing directions for future research.

2. Egyptian Institutional Context and Research Relevance

Recent tax reforms and the government’s growing dependence on tax revenues have significantly influenced how corporate tax practices are viewed in Egypt. Although TA may provide firms with immediate financial benefits through tax savings, stronger regulatory oversight has increased concerns about the potential consequences of aggressive tax planning (Please refer to https://www.pwc.com/m1/en/publications/shedding-light-on-egypts-shadow-economy.html (accessed on 7 May 2026)). The Egyptian tax environment has undergone substantial changes in recent years. The move toward a self-assessment tax system, along with repeated amendments to the Unified Tax Law, has increased both the complexity of tax compliance and the uncertainty surrounding firms’ tax positions [32]. Given the government’s reliance on tax revenues and the growing focus on tax enforcement, aggressive tax strategies may attract greater regulatory scrutiny and expose firms to compliance and reputational risks. As a result, creditors may associate such practices with higher uncertainty, greater legal exposure, and possible reputational costs, leading them to adopt a more cautious approach when evaluating lending decisions [32,48,49].

Egypt’s bank-centered financial system places creditors at the core of corporate financing decisions. Since firms rely predominantly on bank debt rather than capital market financing, lenders become key stakeholders in assessing corporate risk. Banks typically engage in ongoing monitoring of borrowers and therefore rely heavily on transparent and reliable information when evaluating creditworthiness [32]. When tax strategies become complex or insufficiently transparent, information asymmetry between firms and creditors may increase, creating uncertainty regarding firms’ actual risk exposure. This uncertainty can influence lenders’ risk assessments and ultimately affect the CoD charged to borrowing firms.

Egypt has made significant progress in promoting corporate sustainability and transparency over the past decade. The introduction of the S&P/EGX ESG Index, together with regulatory initiatives encouraging sustainability reporting, has increased the availability and visibility of ESG information in the capital market [32]. As ESG reporting becomes more widespread, investors and creditors increasingly view such disclosures as an important source of information about a firm’s long-term performance, transparency, and commitment to sustainable business practices. At the same time, Egypt has undertaken several corporate governance reforms aimed at strengthening accountability, improving disclosure quality, and enhancing investor confidence. These reforms have encouraged firms to adopt more robust governance structures and reporting practices. Consequently, ESG disclosure has emerged as an important tool through which companies can demonstrate transparency, effective oversight, and responsible management, thereby helping to alleviate concerns related to agency conflicts and managerial opportunism.

The combination of ongoing tax reforms, increasing regulatory scrutiny, expanding ESG reporting practices, and a bank-centered financial system creates a unique institutional setting in Egypt. These factors jointly influence how creditors evaluate corporate risk. On one hand, stronger tax enforcement may increase concerns about the regulatory, financial, and reputational risks associated with aggressive tax planning. The complexity of the tax system, ongoing economic reforms, and evolving stakeholder expectations contribute to information asymmetry between firms and creditors [50]. On the other hand, enhanced ESG disclosure provides creditors with additional information regarding transparency and risk-management practices. Consequently, the impact of TA on debt financing costs may not be uniform across firms but may depend on the level of ESG information available to creditors. This pattern suggests that ESG disclosure has the potential to alter creditors’ interpretation of TA activities and, therefore, moderate their effect on borrowing costs.

Accordingly, the relationship between TA, ESG disclosure, and the CoD is particularly important in emerging markets such as Egypt. The Egyptian context offers a distinctive setting due to the economy’s strong dependence on tax revenues, the dominant role of banks in corporate financing, and the growing emphasis on sustainability reporting. These institutional characteristics make Egypt an appropriate environment for examining how creditors interpret firms’ tax strategies and ESG disclosures when assessing lending risk [32,50].

3. Theoretical Background, Literature Review, and Hypotheses Development

This section synthesizes existing research and theoretical frameworks to derive the study’s hypotheses on the association between TA, ESG disclosure, and the CoD in the context of emerging markets [51]. Based on agency and signaling theories, existing research proved that TA may have implications for creditors’ perceptions of risk, and ESG disclosure may have implications for the interpretation of TA strategies by external financiers [2]. Prior empirical research presents mixed results without a clear consensus on whether TA ultimately adds value or increases risk, and whether ESG disclosure reduces the CoD. The subsequent subsections will discuss the dominant streams of research and derive testable hypotheses on the association between TA and the CoD, and the moderating effect of ESG.

3.1. The Impact of TA on the CoD

TA and the CoD have been examined from two competing perspectives in the literature [44]. The first perspective, commonly referred to as the risk exposure effect, suggests that TA increases creditors’ perceptions of risk and consequently raises borrowing costs [3,13,15,20]. The second perspective, known as the tax saving effect, argues that TA may enhance after-tax cash flows and improve debt repayment capacity, particularly in firms characterized by strong governance, high transparency, and effective monitoring mechanisms [1,2,21,22]. Consistent with agency theory, the risk exposure perspective argues that TA can intensify agency conflicts and increase information asymmetry between managers and creditors because tax-related activities are often implemented through complex and opaque organizational structures [45]. As a result, creditors may face difficulties distinguishing between efficient tax planning and managerial opportunism, leading them to perceive aggressive tax practices as a source of uncertainty and risk [3,14].

Empirical evidence from South Africa, China, Germany, and other emerging markets largely supports this view. Prior studies report that firms exhibiting higher levels of TA—measured through indicators such as lower cash effective tax rates, larger book-tax differences, or greater tax payment volatility—tend to face higher loan spreads, increased bond yields, and more restrictive debt covenants [3,15,24]. Creditors often associate aggressive TA with greater tax audit exposure, potential penalties, volatile future cash flows, and lower reporting quality, prompting them to demand higher risk premiums and impose stricter contractual protections [20]. In addition, studies examining capital market reactions show that disclosures related to tax shelters and tax investigations are associated with negative abnormal stock returns and higher required returns on both equity and debt securities, suggesting that investors frequently view aggressive TA as a source of economic and reputational costs rather than a value-enhancing strategy [52,53]. Prior Egyptian evidence suggests that transparency and reporting quality influence debt financing costs and creditor risk assessments [54,55].

While agency theory emphasizes the risk implications of TA, signaling theory provides a more nuanced perspective [56]. According to signaling theory, TA may convey either favorable or unfavorable signals to creditors. On the one hand, efficient tax planning can signal managerial competence, effective resource utilization, and enhanced cash flow generation, thereby supporting the tax saving effect and potentially lowering financing costs [1,21,22]. Such benefits are likely to be more evident in highly profitable, financially stable, and well-governed firms, where creditors may interpret tax savings as evidence of sound managerial decision-making rather than opportunistic behavior.

On the other hand, signaling theory also suggests that in environments characterized by substantial information asymmetry and limited transparency, creditors may find it difficult to distinguish value-enhancing tax planning from aggressive TA. In many emerging markets, where disclosure quality and investor protection mechanisms remain relatively weak, low tax payments may instead signal opacity, earnings management, excessive risk-taking, or potential future regulatory disputes [3,14,57]. Consequently, creditors may respond conservatively by requiring higher borrowing costs to compensate for the additional uncertainty [13,15,20,24].

Taken together, these arguments suggest that the relationship between TA and the cost of debt is likely to depend on firm-specific characteristics and institutional conditions. Creditors may evaluate TA more favorably in firms characterized by strong profitability, effective governance, and high transparency. Nevertheless, evidence from emerging markets generally indicates that the risk exposure effect tends to outweigh the tax saving effect when information asymmetry is elevated [3,24].

Collectively, agency and signaling approaches, along with the existing body of empirical research, suggest that TA tends to be interpreted as an activity that amplifies risk rather than one that creates shareholder value within an emerging market setting like Egypt [13,20]. Therefore, tax-evading firms are expected to face a higher CoD funding for both bank loans and bond issuances when compared to tax-compliant firms, given that investors will seek compensation for tax-related and reputational risks connected with opaque tax avoidance practices [3,15].

Accordingly, although creditors may perceive TA differently depending on profitability, governance quality, and transparency, the Egyptian institutional environment suggests that, on average, higher levels of TA are more likely to be interpreted as indicators of risk, uncertainty, and increased monitoring costs. Therefore, the study expects a positive association between TA and the cost of debt. Thus, the first hypothesis is stated as follows:

H1. 

Higher levels of TA are associated with higher debt financing costs.

3.2. The Moderating Impact of ESG Disclosure

Prior research suggests that ESG disclosure can influence how creditors evaluate firms’ TA activities by providing additional information regarding governance quality, transparency, and stakeholder commitment [4,6,7,25,26]. Consequently, ESG disclosure may affect the extent to which lenders perceive TA as a source of risk and incorporate such perceptions into debt pricing decisions.

From an agency theory perspective, more extensive ESG disclosure serves as an additional monitoring mechanism that reduces information asymmetry and constrains managerial opportunism. Hence, it enables creditors to better judge whether tax benefits are being utilized to enhance value and improve long-term solvency [36,58,59]. More extensive ESG reporting enables creditors to gain a clearer understanding of a firm’s governance structure and risk management practices. As a result, creditors may be better able to distinguish between tax strategies that enhance firm efficiency and those that increase agency conflicts.

Signaling theory provides a complementary explanation. ESG disclosure can convey information about management quality and commitment to sustainable business practices [37,47,60]. Although disclosure breadth does not directly measure disclosure credibility, in emerging markets such as Egypt, ESG disclosure remains relatively voluntary and is not yet uniformly adopted across firms. Consequently, firms that provide more extensive ESG information incur additional reporting effort and expose themselves to greater stakeholder scrutiny, making such disclosures informative to creditors. Therefore, firms that provide more extensive ESG disclosure may be perceived as more transparent and less likely to engage in opportunistic behavior.

Nevertheless, alternative perspective argue that ESG initiatives may not always reflect genuine commitment to sustainability [61]. In certain cases, firms may use ESG activities and disclosures to improve their public image while continuing to engage in aggressive tax practices or other questionable actions. From this perspective, ESG disclosure may be viewed as a symbolic exercise rather than a true indicator of transparency and accountability, which could reduce its effectiveness in alleviating creditors’ concerns about firm risk [31]. Although this concern cannot be completely dismissed, agency and signaling theories suggest that creditors place greater weight on the additional information provided by ESG disclosure when assessing firm risk. Consequently, the monitoring and informational benefits of ESG disclosure are expected to outweigh potential concerns regarding symbolic reporting. Empirical evidence generally supports this argument. Firms with stronger ESG disclosure and performance tend to enjoy lower borrowing costs, improved creditworthiness, and more favorable debt contract terms [62,63]. This advantage arises because creditors prefer these firms due to their lower exposure to environmental, social, reputational, and governance risks [25,37,38,64].

Moreover, evidence indicates that enhanced non-financial reporting can help creditors differentiate between tax planning and aggressive TA, thereby reducing the risk premium associated with tax-related activities. In the South African context, Medhioub and Boujelbene [3] demonstrates that assurance of integrated reports helps creditors differentiate between prudent, low-risk tax planning and aggressive, risk-increasing TA strategies, and more extensive non-financial information reduces the positive association between TA and corporate CoD. Notwithstanding the differences in the context of application, evidence from both a developing country (Egypt) and a developed economy indicates consistent findings. The studies of Elamer et al. [30] and Alomair and Metwally [32] showed that ESG ratings and ESG disclosure moderate the adverse relationship between TA and firm value. These findings suggest that ESG-related information influences how stakeholders interpret the risks associated with TA, as ESG dimensions can change the nature of TA from a risk-enhancing activity to one whose implications depend on the firm’s sustainability context.

Further, other studies applied in different contexts, such as China and Indonesia, have proved that strong ESG performance alleviates financing constraints and lowers corporate loan and bond costs. On the other hand, ESG failures and greenwashing risks may result in higher debt costs due to creditors’ apprehensions about unknown risks stemming from environmental degradation and social misconduct [5,25,65]. To sum up, studies conducted in both developed and emerging markets also suggest that ESG practices can mitigate the adverse consequences associated with TA by improving transparency and stakeholder confidence [30,32].

Therefore, ESG disclosure is expected to weaken creditors’ negative assessment of tax avoidance by reducing information asymmetry and signaling stronger governance and risk-management practices. As a result, the increase in borrowing costs associated with TA should be less pronounced among firms with higher levels of ESG disclosure. In contrast, firms characterized by both aggressive tax behavior and weak ESG disclosure are more likely to be viewed as opaque and risky, leading creditors to demand higher borrowing costs [30,32]. Building on the above arguments, ESG disclosure is expected to reduce the extent to which TA increases borrowing costs. The study’s theoretical framework, which incorporates the research variables and proposed linkages in a logical structure, is shown in Figure 1. Accordingly, the following hypothesis is proposed:

H2. 

ESG disclosure negatively moderates the positive relationship between TA and the CoD.

4. Research Methodology

4.1. Data Set

This research looks into the effects of TA on the CoD of the non-financial Egyptian firms and how ESG disclosure (ESG) acts as a moderator. A sample of publicly listed firms from the Egyptian Stock Exchange (ESE) from 2018 to 2023 was examined. The initial sample comprised all non-financial companies registered on the Egyptian Stock Exchange; financial organizations were not included because of their different reporting and regulatory frameworks. Firms with incomplete data across the study variables were excluded at the firm level to ensure data completeness over the entire observation period [32,33,34,39,66,67].

All continuous variables were winsorized at the first and 99th percentiles to lessen the impact of extreme values. To lessen the impact of outliers without removing legitimate observations, this method was applied to all major continuous variables, such as tax avoidance measures, ESG disclosure ratings, cost of debt, and Control Variables. And the S&P/EGX ESG index score, which is built using a standardized divisor approach used by S&P Dow Jones Indices to ensure comparability across organizations, provides the basis for the ESG disclosure variable.

Therefore, the dataset was constructed as a balanced panel consisting of 110 companies producing 660 firm-year observations over a 6-year period (2018–2023), as shown in

Table 1. Sample structure.

Industry	Companies	Observations	%
Trade and Distributors	5	30	4.55
Paper and Packaging	4	24	3.64
Travel and Leisure	7	42	6.36
Basic Resources	10	60	9.09
Textile and Durables	7	42	6.36
IT, Media and Communication Services	4	24	3.64
Real Estate	16	96	14.55
Healthcare and Pharmaceuticals	4	24	3.64
Food, Beverages, and Tobacco	20	120	18.18
Building Materials	13	78	11.82
Industrial Goods, Services, and Automobiles	10	60	9.09
Contracting and Construction Engineering	10	60	9.09
Total	110	660	100

. To collect the data that used in this research, we used company websites, annual reports, Mubasher Egypt, and Bloomberg Asharq.

4.2. Variable Measurements

This study’s empirical model, which explains the connection between tax avoidance, ESG disclosure, and debt cost, is based on well-known financial and accounting theories. In particular, agency theory and information asymmetry theory contend that tax avoidance raises perceived risk and information opacity, which prompts creditors to seek more returns in the form of increased borrowing rates. Because ESG disclosure is essential in lowering information asymmetry and improving stakeholders’ capacity to evaluate corporate transparency, governance quality, and long-term risk exposure, it is added as a moderating variable. Therefore, by enhancing creditors’ perception of risk, ESG disclosure is anticipated to reduce the positive correlation between tax avoidance and loan costs. Therefore, the interaction-based panel data model is used to capture both the direct and combined effects of tax avoidance and ESG disclosure on financing results, particularly in an emerging market setting, when institutional disparities, disclosure quality, and information environments change greatly between businesses and over time.

4.2.1. The Cost of Debt (CoD) (Dependent Variable)

In our research, we calculate the CoD by dividing a company’s interest expenses by the average amount of short-term and long-term debt incurred throughout the year, in order to capture borrowing costs faced by the firm over the fiscal year [3,68,69,70]. Consequently, the following is how we measured CoD:

CoDit = Interest expensesit ÷ Average (short-term debtit + long-term debtit)

4.2.2. Tax Avoidance (TA) (Independent Variable)

According to many studies in the literature, TA encompasses both illicit and legitimate activities that a firm uses to reduce its tax obligations [57,71]. In light of previous research, TA is determined by the Effective Tax Rate (ETR), where ETR is commonly used as a trustworthy stand-in to identify instances of TA in academic research [72,73]. Our definition of tax avoidance in this study, which is in line with previous studies [10,74], includes actions that lower a company’s taxes in relation to its pre-tax accounting revenue. Thus, scaled tax expenses by pre-tax income produce the ETR. In this study, we multiplied (ETR × −1) to develop an increasing indicator of tax avoidance, as a higher ETR indicates less aggressive taxation [31].

4.2.3. ESG Disclosure (ESG) (Moderating Variable)

ESG disclosure was measured based on the extent of firms’ sustainability disclosure in accordance with the S&P/EGX ESG framework rather than index membership status [34,66,74]. The S&P/EGX ESG index relies substantially on publicly available corporate information and evaluates firms based on environmental, social, and governance dimensions disclosed to stakeholders. Therefore, although the index may reflect aspects of both ESG practices and outcomes, it is employed in this study as an external and standardized indicator of the extent of ESG-related information communicated by firms [75].

4.2.4. Control Variables

We added a collection of control variables that may significantly influence CoD, in keeping with earlier research. These factors include the firm size (Size), the current ratio (CR), leverage (Lev), profitability (ROA), and firm loss (L). We anticipate that Size and CoD will be negatively correlated. According to Eliwa et al. [76], Erragragui [77], and Hasan et al. [78], firms of a larger size are supposed to have greater resources for external financing at a lower cost. We expect that Lev and CoD will be positively correlated, where those companies with a lower level of Lev are anticipated to have better solvency and a cheaper interest rate [57,77]. Additionally, a negative correlation between ROA and the CoD is anticipated. Businesses with a high return on assets are better off financially and frequently obtain loans with lower interest rates [79,80,81]. Firms that have a higher CR are able to settle their current debts [3]. Therefore, we anticipated that CR and CoD would have a negative association. Businesses that declare a loss this year are viewed by bankers as riskier borrowers [57,59,82]. As a result, we anticipated that L and CoD would positively correlate.

Table 2. Variables’ definitions.

Variable	Measurement
Dependent variable
Cost of Debt (CoD)	CoDit = Interest expensesit ÷ Average (short-term debtit + long-term debtit)
Independent variable
Tax avoidance (TA)	TAit = −1 × (Income tax ÷ Pre-tax income).
Moderating variable
ESG disclosure (ESG)	A composite score of the three factors (environmental, social, and governance), where the S&P/EG X ESG Index provides the ESG ratings that are used to calculate the ESG disclosure.
Control variables
Firm Size (Size)	=log (total assets)
Leverage (Lev)	$={\displaystyle \frac{total debt}{total assets}}$
profitability (ROA)	$={\displaystyle \frac{net profit}{total assets}}$
Firm loss (L)	A dummy variable that, if a company reported negative earnings during the fiscal year, is equal to 1, and if not, it is equal to 0.
Current Ratio (CR)	$={\displaystyle \frac{\mathrm{current}\mathrm{assets}}{\mathrm{current}\mathrm{liabilities}}}$

below lists the control variables and their explanations.

4.3. The Study Models

To analyze the established hypotheses in our research, we used a Panel-Corrected Standard Error (PCSE) model as the baseline empirical specification. The panel data characteristics were assessed using suitable diagnostic tests, such as tests for autocorrelation, heteroskedasticity, and cross-sectional dependency, prior to model estimation. These tests’ findings show that these econometric problems exist, which supports the PCSE estimator’s use [83,84]. As a result, PCSE is used because it offers accurate and efficient estimates when there is serial correlation in panel data sets, heteroskedasticity, and contemporaneous correlation across businesses. In addition, Driscoll–Kraay Standard Errors, Fixed Effects and System GMM estimations are used as robustness checks to ensure the consistency of the results across alternative econometric specifications.

In terms of the sample’s industrial makeup, Table 1 shows a fairly unequal distribution across sectors, with stronger representation in the Real Estate sector (14.55%) and the Food, Beverages, and Tobacco sector (18.18%), while other industries show relatively smaller proportions. The empirical models use industry fixed effects to account for unobserved sectoral variation in order to allay this worry. This method guarantees that the predicted connections are not skewed by variations in the cost of debt, tax avoidance, and ESG disclosure policies between industries. As a result, the possible impact of industry concentration on the primary conclusions is successfully reduced, improving the empirical results’ validity and dependability.

In this study, to test our hypotheses, we create multiple regressions. To test H1 and H2, respectively, we produced Models 1 and 2.

Model (1):

CoD_it  = β₀   + β₁TA_it + β₂Size_it + β₃Lev_it + β₄ROA_it + β₅L_it + β₆CR_it + β_t + β_ind + ε_it

(1)

Model (2):

CoD_it = β₀ + β₁TA_it + β₂TA × ESG_it + β₃Size_it + β₄Lev_it + β₅ROA_it + β₆L_it + β₇CR_it + β_t + β_ind + ε_it

(2)

where CoD denotes the cost of debt; TA represents tax avoidance; ESG denotes ESG disclosure; TA × ESG is the interaction term capturing the moderating effect of ESG disclosure; Size is firm size; Lev is leverage; ROA is return on assets; L is firm loss; CR is current ratio; β_t represents year fixed effects; β_ind represents industry fixed effects; and ε_it is the error term.

5. Main Results

5.1. Descriptive Statistics

The descriptive statistics for the main variables—TA, ESG, CoD, and company characteristics—for the final sample are shown in

Table 3. Summary statistics.

Variables	N	Min	Max	Mean	S.D
CoD	660	0.02	0.12	0.07	0.02
TA	660	−3.02	−0.03	−0.15	0.39
ESG	660	0.41	1.00	0.60	0.17
Size	660	10.84	17.56	13.93	1.83
ROA	660	−0.17	0.28	0.053	0.08
Lev	660	0.05	0.98	0.52	0.29
CR	660	0.83	2.763	1.837	0.86
L	660	0.00	1.00	0.20	0.38

. At the 1% and 99% percentiles, we winsorized all continuous variables. CoD spans from 0.02 to 0.12, with an average of 0.07. Additionally, TA shows an average value of −0.15; the negative sign reflects the measurement approach (TA = −1 × Income Tax/Pre-tax Income), and higher absolute values indicate greater tax avoidance intensity. With an average of 60%, the ESG ranges between 0.41 and 1 in the sample. The mean value of Lev is 52%. The mean of Size is 13.93. The mean of ROA is 5.3%. The mean of CR is 1.83. The sample’s average company loss is 20%, which equals 132 observations (660 × 20%).

5.2. Correlation Matrix

The correlations between the primary variables in our study are displayed in

Table 4. Correlation.

Variables	CoD	TA	ESG	Size	ROA	Lev	CR	L	VIF
CoD	1								----
TA	0.106 ***	1							1.04
ESG	−0.084 **	−0.159 ***	1						1.05
Size	−0.069 *	−0.089 **	0.093 **	1					1.25
ROA	−0.097 **	0.010	0.042	0.055	1				1.94
Lev	−0.045	0.002	0.031	0.352 ***	−0.464 ***	1			1.58
CR	−0.104 ***	0.046	−0.057	0.076 *	0.061	0.055	1		1.02
L	0.017	0.023	−0.055	−0.038	−0.316 ***	0.338 ***	−0.004	1	1.64

Notes: *** p < 0.01, ** p < 0.05, * p < 0.1.

. It is noted that CoD is positively correlated with tax avoidance. Furthermore, it is negatively correlated with ESG disclosure, ROA, CR and Size. Further, we conducted an analysis using the Variation Inflation Factor (VIF) test. The results presented in Table 4 indicate that the VIF score ranged from 1.02 to 1.94; being less than 10, the score indicates the absence of multicollinearity.

5.3. Baseline Regression Results

5.3.1. Test of H1: TA and CoD

Table 5. Multivariate results.

Variables	Model-1 Direct Effect		Model-2 Moderating Effect
	Coef.	Z	Coef.	Z
TA	0.0062 ***	4.14	0.0078 ***	3.39
ESG	----	----	−0.0067 **	−2.20
TA × ESG	----	----	−0.0123 **	−2.36
Size	−0.0018 **	−2.24	−0.0016 **	−2.10
ROA	−0.0075 **	−2.16	−0.0115 **	−2.25
Lev	−0.0012	−0.15	−0.0017	−0.35
CR	−0.0013 **	−2.12	−0.0014 **	−1.96
L	0.0014	0.46	0.0016	0.52
Constant	0.0992 ***	6.32	0.1012 ***	6.62
Industry and Year	Included
Adj. R2	0.424		0.443
Chi2	33.417 ***		45.219 ***

Notes: *** p < 0.01, ** p < 0.05.

, Model (1) reports the estimation results of Equation (1), which examines the direct relationship between tax avoidance (TA) and the cost of debt (CoD). The results reveal a positive and statistically significant association between TA and CoD. Specifically, the coefficient on TA is 0.0062 and is significant at the 1% level (t = 4.14), indicating that firms engaging in higher levels of tax avoidance face higher borrowing costs. This finding supports H₁ and is consistent with prior studies reporting that creditors perceive aggressive tax practices as a source of additional risk [3,13,15,20]. From a creditor perspective, TA may increase concerns regarding information opacity, earnings sustainability, managerial opportunism, and potential regulatory scrutiny. Unlike shareholders, who may view tax savings as a source of additional cash flows, creditors focus primarily on downside risk and the firm’s ability to meet its debt obligations. Consequently, lenders may interpret aggressive tax practices as a signal of heightened uncertainty and demand a higher risk premium when extending credit [85,86]. This argument is particularly relevant in Egypt, where ongoing tax reforms, evolving governance practices, and relatively high information asymmetry may amplify creditors’ risk perceptions. Accordingly, lenders may respond by increasing borrowing costs and imposing stricter lending conditions. These findings are consistent with agency theory and support the risk-exposure effect proposed in this study.

To assess economic significance, the estimated coefficient of TA (β = 0.0062) was multiplied by its standard deviation (SD = 0.39). The results indicate that a one-standard-deviation increase in TA is associated with an increase of approximately 0.0024 in the cost of debt. This finding suggests that the impact of TA is not only statistically significant but also economically meaningful.

The observed relationship can be explained through the risk-exposure channel. Higher levels of TA may reduce financial transparency and increase information asymmetry between firms and creditors. As monitoring costs and uncertainty rise, lenders require additional compensation for bearing the perceived risk. Consequently, borrowing costs increase as creditors incorporate this risk premium into debt pricing decisions. This interpretation aligns with agency theory and information asymmetry arguments, which predict that creditors demand higher returns when firm risk becomes more difficult to assess.

In the context of Egypt, lower levels of financial transparency and more lax tax enforcement methods exacerbate these impacts by increasing information asymmetry between creditors and businesses. Because of this, banks and other lending organizations are more inclined to view tax avoidance as an indication of high risk, which raises borrowing prices to offset the perceived risk exposure.

Regarding the control variables, the results show that Size, ROA, and CR are negatively associated with the cost of debt. Larger firms generally benefit from lower borrowing costs due to their greater stability and lower perceived default risk. Similarly, more profitable firms and firms with stronger liquidity positions are viewed as more creditworthy, enabling them to obtain debt financing at a lower cost [3,6,13,22].

5.3.2. Test of H2: ESG Disclosure Act as Moderator

To test H2, we introduced an interaction term between TA and ESG disclosure in Equation (2). The results reported in Table 5, Model (2) show that ESG disclosure is negatively and significantly associated with the CoD. This finding suggests that firms with more extensive ESG disclosure generally enjoy lower borrowing costs and improved access to external financing [6,7,26].

More importantly, the interaction term (TA × ESG) is negative and statistically significant at the 5% level (β = −0.0123), providing support for H2. The negative coefficient indicates that ESG disclosure weakens the positive relationship between tax avoidance and the cost of debt. In other words, the increase in borrowing costs associated with tax avoidance is less pronounced among firms with higher levels of ESG disclosure. Therefore, ESG disclosure plays a moderating role in the TA–CoD relationship. As a result, our H2 is confirmed beyond statistical significance, and we assessed the economic significance of this moderating effect by multiplying the interaction coefficient (−0.0123) by one standard deviation of ESG disclosure (SD = 0.17). The results indicate that a one-standard-deviation increase in ESG disclosure reduces the marginal effect of tax avoidance on the cost of debt by approximately 0.0021. This finding suggests that ESG disclosure not only has statistical significance but also meaningfully mitigates the adverse financing consequences associated with tax avoidance.

The result is consistent with agency theory. ESG disclosure provides creditors with additional information regarding a firm’s governance practices, risk management systems, and long-term strategic orientation, thereby reducing information asymmetry and monitoring concerns [87,88]. As creditors obtain a more comprehensive understanding of the firm’s operations, they are better able to distinguish between tax strategies that enhance efficiency and those that increase financial and regulatory risks. The findings are also consistent with signaling theory. In the Egyptian market, where creditors often rely on supplementary information beyond traditional financial statements [89], more extensive ESG disclosure may serve as a signal of transparency, accountability, and responsible management practices [87,90]. Such information becomes particularly valuable when evaluating firms engaged in tax avoidance activities because it helps creditors assess the broader governance environment in which these activities occur. Consequently, creditors may be less likely to interpret tax avoidance as a signal of elevated risk when it is accompanied by strong ESG disclosure.

The moderating role of ESG disclosure appears particularly relevant in Egypt, given ongoing tax reforms, increasing regulatory scrutiny, and growing stakeholder expectations regarding corporate transparency. While creditors may associate tax avoidance with regulatory, reputational, and cash-flow risks, firms that provide more extensive ESG disclosure are likely to be perceived as more transparent, better governed, and more accountable to stakeholders. These characteristics can reduce creditors’ perceptions of risk and weaken the positive effect of tax avoidance on borrowing costs.

6. Robustness Tests

6.1. CoD Alternative Measurement

We employed a different proxy to measure the CoD in order to address the potential constraints of the existing proxy and increase the internal validity of our study. In accordance with previous studies [91,92], we examined the ratio of financial expenses to total liabilities. As provided in

Table 6. Alternative measure of CoD.

Variables	Model-1A Direct Effect		Model-2A Moderating Effect
	Coef.	Z	Coef.	Z
TA	0.0097 ***	3.18	0.0108 ***	3.22
ESG	----	----	−0.0123 **	−2.17
TA × ESG	----	----	−0.0194 **	−2.22
Size	−0.0029 **	−2.10	−0.0027 **	−1.92
ROA	−0.0149 **	−2.38	−0.0226 **	−2.05
Lev	0.0087	0.11	0.0025	0.35
CR	−0.0020 *	−1.88	−0.0023 *	−1.82
L	0.0022	0.43	0.0026	0.52
Constant	0.0331	1.33	0.0293	1.24
Industry and Year	Included
Adj. R2	0.256		0.274
Chi2	32.721 ***		42.738 ***

Notes: *** p < 0.01, ** p < 0.05, * p < 0.1.

, the coefficient signs and significance levels remain largely consistent when an alternative measure of CoD is employed. This consistency suggests that the documented relationship is not sensitive to the specific operationalization of debt financing costs. Furthermore, the similarity in coefficient magnitudes indicates that the economic interpretation of the findings remains stable across alternative model specifications.

6.2. TA Alternative Measurement

To find out how sensitive or resilient our main conclusions are to using an alternative TA measure, we replicated our main study using the book-tax difference (BTD), which is computed by dividing the difference between the book value of income and taxable income by total assets, as an alternative indicator of corporate TA [30,93]. As demonstrated in

Table 7. Alternative measure of TA.

Variables	Model-1B Direct Effect		Model-2B Moderating Effect
	Coef.	Z	Coef.	Z
BTD	0.0026 ***	2.74	0.0022 **	2.16
ESG	----	----	−0.0115 **	−2.20
BTD × ESG	----	----	−0.0149 **	−2.35
Controls	Included
Industry and Year	Included
Adj. R2	0.419		0.414
Chi2	28.950 ***		27.830 ***

Notes: *** p < 0.01, ** p < 0.05.

, re-estimating the model using book-tax differences as an alternative proxy for TA leads to results that are consistent with the baseline findings. This consistency reduces concerns that the reported relationship is driven by the limitations of a single measure and suggests that the findings remain stable under alternative specifications of TA.

6.3. ESG-Sensitive and Non-ESG-Sensitive Industries

The sample was split into two subgroups according to industry characteristics in order to capture possible changes across institutional and operational contexts [94,95]. Industries were divided into ESG-sensitive (high environmental and operational risk industries) and non-ESG-sensitive (service-oriented industries) categories based on previous research. To investigate whether the moderating influence of ESG disclosure on the connection between TA and CoD varies across sector settings, separate regressions were calculated for each category.

Regarding the moderating effect of ESG disclosure across industry categories, the findings in

Table 8. Industry heterogeneity and ESG moderation effect.

Variables	ESG-Sensitive Industries Moderating Effect		Non-ESG-Sensitive Industries Moderating Effect
	Coef.	Z	Coef.	Z
TA	0.0074 **	2.19	0.0054 ***	3.08
ESG	−0.0022 **	−2.24	−0.0008 **	−1.98
TA × ESG	−0.0104 **	−2.12	−0.0078 **	−2.08
Controls	Included
Industry and Year	Included
Adj. R2	0.403		0.468
Chi2	22.220 ***		93.061 ***

Notes: *** p < 0.01, ** p < 0.05.

reveal that in both groups, the interaction term between TA and ESG disclosure is statistically significant and negative, suggesting that ESG disclosure mitigates the positive correlation between TA and CoD. The stronger moderating effect observed in ESG-sensitive industries suggests that creditors place greater emphasis on ESG-related information when evaluating firms operating in sectors exposed to higher environmental, social, and reputational risks. In these industries, ESG disclosure provides valuable insights into firms’ governance quality, risk-management practices, and commitment to sustainable operations. As a result, creditors may rely more heavily on such information when assessing the risks associated with TA. This finding indicates that the role of ESG disclosure in shaping creditor perceptions varies across industries and becomes particularly important in sectors facing greater sustainability-related challenges.

6.4. Lagged ESG Disclosure Analysis

The research uses a delayed formulation of ESG disclosure (ESG t − 1) to address possible reverse causation and reduce simultaneity problems between ESG disclosure and cost of debt. By ensuring that ESG data is predefined in relation to financial results, this method strengthens causal interpretation and improves the temporal ordering of variables [96].

The findings in

Table 9. Lagged ESG disclosure results.

Variables	Lagged ESG Moderating Effect
	Coef.	Z
TA	0.0016 **	2.15
ESG	−0.0067 **	−2.07
TA × ESG	−0.0541 ***	−2.79
Controls	Included
Industry and Year	Included
Adj. R2	0.442
Chi2	43.090 ***

Notes: *** p < 0.01, ** p < 0.05.

reveal that even after taking temporal ordering into account, the interaction term between TA and lagged ESG disclosure is still statistically significant and negative, indicating the moderating effect of ESG disclosure. Although the use of lagged ESG disclosure cannot fully eliminate endogeneity concerns, the persistence of the moderating effect after introducing temporal separation between the explanatory and dependent variables provides additional support for the stability of the documented relationship and reduces concerns regarding simultaneity bias.

6.5. Lagged TA Analysis

Tax avoidance and cost of debt may be jointly determined, raising concerns about reverse causality. To address this issue, we mitigate endogeneity concerns by incorporating lagged tax avoidance (TA_[t − 1]) in the empirical specification to reduce simultaneity bias. The findings in

Table 10. Lagged TA results.

Variables	Lagged TA Direct Effect		Lagged TA Moderating Effect
	Coef.	Z	Coef.	Z
TA	0.0043 **	2.08	0.0030 *	1.88
ESG	-----	-----	−0.0069 **	−2.17
TA × ESG	-----	-----	−0.0736 ***	−2.27
Controls	Included
Industry and Year	Included
Adj. R2	0.406		0.436
Chi2	34.02 ***		38.93 ***

Notes: *** p < 0.01, ** p < 0.05, * p < 0.1.

reveal that the results remain consistent across all specifications, indicating that the main findings are robust to endogeneity and reverse causality concerns.

6.6. Alternative Estimation Method

To ensure the robustness and stability of the baseline findings, an alternative estimation approach based on firm fixed effects (FE) is also employed. We used the fixed effects (FE) regression to manage the company’s unobserved characteristics that could be connected to the outcome variable. Reguera-Alvarado et al. [97] claim that the firm FE approach eliminates possible endogeneity issues and improves the efficiency of estimates. The outcomes shown in

Table 11. FE regression.

Variables	Model-1C Direct Effect		Model-2C Moderating Effect
	Coef.	Z	Coef.	Z
TA	0.0037 ***	3.22	0.0088 **	2.36
ESG	----	----	−0.0085 **	−2.05
TA × ESG	----	----	−0.0138 **	−2.18
Controls	Included
Industry and Year	Included
Adj. R2	0.158		0.177
F-statistics	11.950 ***		9.753 ***

Notes: *** p < 0.01, ** p < 0.05.

indicate that our findings are unaltered, suggesting that our analyses are reliable and effective. The results’ agreement across PCSE and FE specifications attests to the key conclusions’ statistical and economic robustness, as they are unaffected by the estimate approach selected.

6.7. Endogeneity Test

In

Table 12. Dynamic panel GMM.

Variables	Model-1D Direct Effect		Model-2D Moderating Effect
	Coef.	Z	Coef.	Z
CoD lag (1)	0.4935 ***	5.08	0.4463 ***	4.06
TA	0.0206 **	2.18	0.0311 **	1.98
ESG	-------	-------	−0.0132 **	−2.03
TA × ESG	-------	-------	−0.1109 **	−2.33
Controls	Included
Industry and Year	Included
Chi2	3698.01 ***		3653.82 ***
Hansen test (p-value)	0.231		0.346
AR1 (p-value)	0.000		0.001
AR2 (p-value)	0.143		0.157
Number of instruments	12		14

Notes: *** p < 0.01, ** p < 0.05.

, in order to account for endogeneity across variables, we employed the dynamic panel GMM. Reverse causality, simultaneity bias, and missing potential variables are all addressed by the GMM method [32]. The results shown in Table 12 suggest that our findings are still reliable even after accounting for endogeneity problems. Standard diagnostic tests are used to evaluate the GMM specification’s validity. For both models, the Hansen J test of overidentifying constraints produces p-values over 0.05, confirming the suitability of the chosen instruments and showing that the null hypothesis of instrument validity cannot be rejected. Furthermore, the Arellano–Bond test for first-order serial correlation [AR(1)] yields statistically significant p-values (p < 0.05), which is expected in first-differenced equations and confirms the validity of the transformation used in the system GMM estimation. In contrast, the Arellano–Bond test for second-order serial correlation [AR(2)] yields insignificant p-values (p > 0.05), indicating that the differenced error terms do not exhibit second-order serial correlation and supporting the validity of the model specification. After accounting for endogeneity issues, the empirical results show that the primary conclusions are still valid. While the coefficients of tax avoidance and ESG disclosure (as well as their interaction term) continue to be consistent with the baseline results, the lagged dependent variable is statistically significant, indicating the dynamic character of the model.

6.8. Using Driscoll–Kraay Standard Errors

To ensure the robustness of our findings, we re-estimated all baseline models using Driscoll–Kraay standard errors, which are robust to heteroskedasticity, autocorrelation, and cross-sectional dependence in panel data settings with large N and small T. The results remain qualitatively consistent with the baseline estimations (see

Table 13. Results of Driscoll–Kraay estimator.

Variables	Model-1B Direct Effect		Model-2B Moderating Effect
	Coef.	Z	Coef.	Z
TA	0.0017 ***	2.02	0.0028 *	1.98
ESG	----	----	−0.0085 **	−2.12
TA × ESG	----	----	−0.0135 ***	−5.13
Controls	Included
Industry and Year	Included
Adj. R2	0.116		0.125
F	81.82 ***		84.16 ***

Notes: *** p < 0.01, ** p < 0.05, * p < 0.1.

). This confirms that the main empirical findings are not sensitive to the choice of standard error correction method. Accordingly, the robustness checks further strengthen the validity of the results and support the reliability of the estimated relationships reported in the study.

Overall, these results offer strong evidence that the estimated associations are not influenced by endogeneity bias, supporting the internal validity and dependability of the empirical findings.

Table 14. A comparative summary of the robustness tests.

Robustness Test	Objective	Alternative Specification	Main Result
Alternative CoD Measure (Table 6)	Verify sensitivity to CoD measurement.	Financial Expenses/Total Liabilities	Coefficient signs and significance remain unchanged.
Alternative TA Measure (Table 7)	Verify sensitivity to TA measurement.	Book-Tax Difference (BTD)	Findings remain consistent with baseline results.
Industry Heterogeneity (Table 8)	Examine industry differences.	ESG-sensitive vs. non-ESG-sensitive industries	ESG moderation remains significant in both groups and is stronger in ESG-sensitive industries.
Lagged ESG Disclosure (Table 9)	Address reverse causality concerns.	ESG t − 1	Interaction term remains significantly negative.
Fixed Effects Estimation (Table 10)	Control for unobserved firm heterogeneity.	FE regression	Findings remain unchanged.
Dynamic Panel GMM (Table 11)	Address endogeneity concerns.	System GMM	Main coefficients remain significant; diagnostic tests are satisfied.

shows a comparative summary of the robustness tests utilized.

7. Conclusions

7.1. Summary

This study examines the impact of TA on the CoD and investigates the moderating role of ESG disclosure in an emerging economy. Using panel data from firms listed on the Egyptian Stock Exchange during the period 2018–2023 and employing the Panel-Corrected Standard Error (PCSE) estimation technique, the findings provide strong evidence that higher levels of TA are associated with a higher CoD. This suggests that, within Egypt’s bank-centered financial system and evolving regulatory environment, creditors perceive aggressive tax practices as indicators of greater risk and uncertainty regarding future cash flows. The study further demonstrates that ESG disclosure significantly weakens the positive relationship between TA and CoD. In the Egyptian setting, where creditors place considerable weight on transparency and risk assessment, ESG disclosure functions as a credibility-enhancing mechanism that mitigates concerns associated with TA.

From an agency perspective, ESG disclosure serves as an additional monitoring and transparency mechanism that reduces information asymmetry, enhances accountability, and alleviates creditors’ concerns regarding managerial opportunism. By providing stakeholders with broader information about governance quality, risk management practices, and long-term strategic objectives, ESG disclosure helps mitigate the agency risks associated with tax avoidance. From a signaling theory perspective, extensive ESG disclosure acts as a credible signal of a firm’s commitment to transparency, ethical conduct, and sustainable value creation. Such disclosures convey positive information about management quality and the firm’s long-term orientation, enabling creditors to distinguish between firms engaging in tax planning for efficiency purposes and those pursuing excessively aggressive tax strategies. As a result, ESG disclosure improves creditors’ perceptions of firm risk and reliability, reducing the risk premium embedded in borrowing costs.

Overall, the findings suggest that ESG disclosure functions as a credibility-enhancing mechanism that not only mitigates agency concerns but also sends positive signals to debt providers regarding the firm’s governance and sustainability practices. These findings highlight the growing importance of sustainability-related disclosures in shaping financing decisions and reducing financing frictions within emerging markets such as Egypt.

7.2. Limitations and Future Research

Several limitations should be acknowledged, which also open avenues for future research. First, while this study examines the moderating role of ESG disclosure in the relationship between TA and the CoD, it does not directly test the specific mechanisms through which ESG disclosure influences creditors’ assessments. The theoretical arguments suggest that ESG disclosure may operate through several channels, including reducing information asymmetry, mitigating agency conflicts, and improving creditors’ perceptions of firm risk. Future research could extend the present study by explicitly examining these transmission pathways using mediation analysis and direct proxies for information asymmetry, agency costs, and credit risk perceptions. In addition, future studies may investigate whether the moderating effect of ESG disclosure varies across firm characteristics, such as profitability, governance quality, ownership structure, and financial constraints, to provide a more nuanced understanding of the conditions under which ESG disclosure influences the nexus of TA and CoD.

Second, the study’s focus on non-financial firms listed on the Egyptian Stock Exchange may limit the extent to which the findings can be generalized to other sectors, particularly financial institutions, which operate under different regulatory and reporting environments. In addition, the exclusive focus on an emerging market context does not allow for comparison with more developed or differently structured markets, such as those in the GCC region. Future research could address these gaps by expanding the analysis to include other sectors and by conducting cross-country comparisons, which would provide richer and more robust evidence of the effect of institutional differences.

Third, the sample period (2018–2023) may not fully capture long-term structural changes in ESG reporting frameworks, tax regulations, and market behavior. Extending the time horizon or conducting longitudinal studies could provide deeper insights into how these relationships evolve over time. Future research may also examine the costs and unintended consequences of ESG implementation, including greenwashing risks, compliance burdens, financing constraints, and potential trade-offs between tax planning and corporate liquidity.

Finally, although the study employs dynamic GMM estimation to address potential endogeneity concerns, issues related to omitted variables and measurement error may still persist. Future research could strengthen the robustness of findings by integrating traditional econometric techniques with alternative approaches, such as quasi-experimental designs. In addition, the use of advanced methods such as machine learning may enhance the accuracy of predictions. This study adopts a linear specification of the relationship between tax avoidance and cost of debt and does not examine potential nonlinear or threshold effects; therefore, exploring possible nonlinear dynamics is suggested as a direction for future research.