Does capital structure mediate the relationship between corporate governance compliance and firm performance? Empirical evidence from Pakistan

Does capital structure (CS) mediate the relationship between corporate governance (CG) compliance and firm performance (FP) in emerging countries? Previous studies have used formative measures (individual determinants) of CG to examine the relationships among CG and FP (Vo and Phan, 2013; Bhagat and Bolton, 2019; Ciftci et al., 2019), CG and CS (Haque et al., 2011; Morellec et al., 2012; Kieschnick and Moussawi, 2018) and CS and FP (Berger and Di Patti, 2006; Margaritis and Psillaki, 2010; Mathur et al., 2021). Fewer studies have investigated the interrelationships of CG, CS and FP. The literature on the mediating role of CS between CG and FP exclusively uses formative attributes of CG (Detthamrong et al., 2017; Ngatno et al., 2021), which may not comprehensively capture overall firm compliance with prescribed CG codes in respective countries. Additionally, this evidence is primarily confined to Thailand (Detthamrong et al., 2017), Mauritius and Saudi Arabia (Boshnak, 2023; Ronoowah and Seetanah, 2023) and financial companies in Bangladesh (Molla et al., 2023). Despite extensive research on CG and FP, there is a significant gap in understanding how CS mediates this relationship, particularly in emerging markets like Pakistan. This study aims to fill this gap by providing empirical evidence from Pakistan, where CG reforms are ongoing and crucial for market development. This study, therefore, aims to contribute to the literature by using two distinct summative measures of CG compliance of non-financial companies on the Pakistan Stock Exchange (PSX).

After large corporate scandals, the Sarbanes–Oxley Act of 2002 was introduced as a landmark reform for CG, prompting other countries to develop frameworks tailored to their respective markets. Public firms are obligated to adhere to CG guidelines established by regulators. The Securities and Exchange Commission of Pakistan (SECP) introduced a CG framework in 2002, and the regulations have undergone several revisions in the years since (Shakri et al., 2022). Notably, a shift from voluntary to mandatory compliance was implemented in 2012. These regulations are specified in the Code of Corporate Governance of Pakistan (CCGP). We use these guidelines as a benchmark to assess company compliance and calculate an overall compliance score. Using a summative measure of CG compliance offers several advantages over individual or formative measures. First, given the complex nature of CG with multiple facets, a comprehensive index captures the full spectrum of a firm’s governance mechanisms, facilitating comparisons of governance practices, quality and compliance trends across companies and over time. Second, because regulatory structures vary across countries, a standardised approach to measuring compliance allows for a deeper analysis of its influence towards financial decisions and FP. Third, a CG index provides practitioners with an approach to benchmark compliance, improving market transparency and investor confidence. Finally, it empowers regulators and firm managers to identify and prioritise areas of improvement in governance.

Our novel application of a CG compliance index for individual firms enables an examination of how CS mediates the impact of CG compliance on FP. Firms with strong CG compliance are more likely to adopt optimal CSs to enhance firm value, whereas those with weak CG compliance may make suboptimal CS decisions that erode FP (Javaid et al., 2023). Though debt offers tax advantages, increased borrowing amplifies capital costs and raises the risk of financial distress and bankruptcy (Ronoowah and Seetanah, 2023; Salehi et al., 2023). Consequently, a firm’s decisions on debt or equity financing may be influenced by the degree of CG compliance. Given the potential indirect impact of CG on FP (Detthamrong et al., 2017), exploring the channels through which CG compliance can affect FP is useful.

The research design is grounded on empirical foundations of three distinct relationships: CG and FP (Vo and Phan, 2013; Bhagat and Bolton, 2019; Ciftci et al., 2019), CG and CS (Haque et al., 2011; Morellec et al., 2012; Kieschnick and Moussawi, 2018) and CS and FP (Berger and Di Patti, 2006; Margaritis and Psillaki, 2010; Mathur et al., 2021). Although previous studies have explored the interconnections between these variables to infer the mediating roles of CG and CS on FP (Detthamrong et al., 2017; Ngatno et al., 2021), they have relied on individual CG measures. This research distinguishes itself by examining the influence of firms’ overall CG compliance.

A CG compliance index is constructed based on the CCGP. Extracting items from the CCGP, each firm is scored for compliance. This method, used in prior research (Ararat et al., 2017; Black et al., 2017), predicts firm-level governance and its relation to firm characteristics. It facilitates cross-country comparisons and complements Shakri et al.’s (2022) weighted index. Each item prescribed in the CCGP is assigned a binary score (1 for compliance and 0 otherwise). The index comprises five sub-indices: “board structure”; “responsibilities, powers and functions of the board”; “corporate and financial reporting framework”; “functions and responsibilities of audit committee”; and “eligibility and responsibility of external auditors.” Items within each sub-index are summarised in Table 1 

With a sample of 2,314 firm-year observations from 2008 to 2018, this study investigates the mediating role of CS in the relationship between CG compliance and FP within the context of Pakistan. Empirical findings indicate a significant positive association between CG compliance and FP. However, the results suggest that the influence of CG compliance on FP diminishes with higher debt levels. The robustness of these findings is confirmed through alternative model specifications, including the control of additional board-level variables and the exclusion of the Global Financial Crisis (GFC) period.

To address potential endogeneity concerns, several methods were used. A one-year lagged dependent variable was incorporated to account for model dynamics. Fixed effects (FE) estimation was used to mitigate omitted variable bias, with results remaining consistent. Then, system generalised method of moments (GMM) was applied to address simultaneity and unobserved heterogeneity. To further refine the analysis, propensity score matching (PSM) was conducted to differentiate between firms exhibiting high and low levels of CG compliance. The robustness of findings is confirmed across these various estimation techniques.

This study contributes to the literature in several aspects. First, building upon the growing literature on the channels of the relationship between CG and FP in emerging markets (Detthamrong et al., 2017; Ngatno et al., 2021; Ronoowah and Seetanah, 2023), this study adds another strand to the literature by empirically investigating the mediating impact of CS between CG compliance and FP. Our findings suggest that although CG compliance plays a significant role in improving financial performance (Shakri et al., 2022), CS decisions undermine this relationship.

Second, previous studies on the impact of CG (Haque et al., 2011; Morellec et al., 2012; Vo and Phan, 2013; Detthamrong et al., 2017; Kieschnick and Moussawi, 2018; Bhagat and Bolton, 2019; Ciftci et al., 2019; Ngatno et al., 2021) have mainly focused on individual aspects (formative measures) of CG, which may not fully capture firms’ overall compliance with prescribed CG codes in their respective countries. In contrast, this study uses two summative and comprehensive measures of CG compliance to better assess overall governance quality and its impact on FP via CS. This approach deepens our understanding of the complex interplay among CG, CS and FP.

Third, while the existing literature has examined the relationship between CG and FP in emerging markets, most evidence is concentrated in specific contexts, such as Thailand (Detthamrong et al., 2017), Mauritius and Saudi Arabia (Boshnak, 2023; Ronoowah and Seetanah, 2023) and financial companies in Bangladesh (Molla et al., 2023). This study, by focusing on Pakistani – an emerging South Asian economy that transitioned from voluntary to mandatory CG adoption in 2012 and continues to implement crucial CG reforms for market development – provides valuable insights into the impact of evolving CG practices in such a dynamic setting.

The practical implications of this study are significant for regulators, firms and investors. For regulators, while stringent CG compliance can improve market transparency and boost investor confidence, our findings from the Pakistan context indicate that higher compliance costs – particularly when coupled with elevated debt levels – can constrain FP. For firms, the results suggest that allocating resources to meet compliance requirements may lead to over-conservatism or under-investment, potentially limiting growth opportunities. Optimizing CS in alignment with CG practices is, therefore, essential to enhance financial performance. For investors, CG compliance can serve as a valuable criterion for investment decisions, helping them identify well-governed firms with balanced debt structures.

The remainder of this paper is structured as follows: Section 2 reviews the theoretical framework and empirical literature to build the hypotheses for testing the relationship between CG compliance, CS and FP. Section 3 discusses the research design. Section 4 presents the results followed by discussions and conclusions in Sections 5 and 6.

The following sub-sections provide a review of underlying theories of CG and CS and link the empirical literature related to emerging markets. It provides additional context to practices in Pakistan and outlines the hypotheses of this study.

Jensen and Meckling’s (1976) agency theory highlights the conflict of interest between firm shareholders and management, suggesting that managers may pursue personal objectives that conflict with firm value maximisation. This divergence underscores the need for governance mechanisms that align management’s interests with those of shareholders and facilitate oversight of management decisions. Effective CG frameworks provide such mechanisms, enhancing transparency and accountability, which are crucial for reducing agency problems.

CS decisions, meanwhile, are rooted in the Modigliani and Miller (1958) proposition that, under conditions of perfect markets, CS does not impact firm value. However, in reality – where information asymmetry, taxes and risk of default are present – subsequent research has shown that CS choices do influence firm value. Agency theory supports this relationship by proposing that effective CG can reduce agency costs, as it limits managerial tendencies toward over-investment and risky ventures that are inconsistent with shareholders’ goals.

For instance, well-governed firms often rely less on debt, as CG reduces the need for debt-related discipline mechanisms by curbing management’s excesses. When CG is robust, investor confidence improves, increasing access to equity financing, which lowers reliance on debt. Conversely, weak CG may drive firms to rely on debt financing, as investor confidence diminishes in such environments. This relationship is particularly relevant in markets like Pakistan’s, which is dominated by family-owned firms and state-owned enterprises where agency conflicts are more pronounced, and information asymmetry is high (Shakri et al., 2022). In such contexts, higher debt ratios can be a means to retain control, but at the expense of potentially higher financial risk.

Empirical evidence from both developed markets (Morellec et al., 2012; Kieschnick and Moussawi, 2018) and emerging ones, including Pakistan (Haque et al., 2011; Sheikh and Wang, 2012; Detthamrong et al., 2017; Javaid et al., 2023), supports the inverse relationship between CG and leverage. For firms in Pakistan, stronger CG compliance may help lower leverage by providing alternative governance mechanisms that mitigate agency problems, thereby reducing the need for debt as a disciplinary tool. Greater CG compliance can also improve market confidence in equity financing and reduce the probability of excessive risk from over-leveraging. CG compliance in these firms, therefore, can be expected to correlate with lower debt levels:

Agency theory (Jensen and Meckling, 1976) suggests that effective CG helps safeguard shareholder interests by curbing managerial discretion, ensuring that decisions align with shareholder goals. CG mechanisms, such as board oversight and independent directors, are designed to reduce agency costs, promote accountability and mitigate suboptimal decisions that could harm FP. Through structured monitoring, CG can, thus, contribute to higher profitability by aligning management’s actions with firm value maximisation.

Complementing this perspective, resource dependence theory (Pfeffer and Salancik, 1978) posits that diverse and well-composed boards offer firms access to valuable resources, including expertise, networks and external influence. These resources can strengthen the firm’s strategic capabilities, enabling it to capitalize on growth opportunities, manage risks and thereby enhance performance. In markets where firms have strong board diversity and governance structures, CG compliance may be especially beneficial, as it provides firms with a broader foundation for sustained profitability.

Institutional theory further suggests that CG compliance improves firm legitimacy and transparency, particularly in emerging or transitional markets (DiMaggio and Powell, 1983; Scott, 1987). As CG frameworks evolve in such markets, compliance becomes essential for attracting both domestic and foreign capital, thus enabling firms to access financing that fosters growth and profitability. In this regard, CG compliance not only satisfies regulatory standards but also aligns firms with investor expectations, enhancing trust and market efficiency.

Although numerous studies document the benefits of CG for FP (Talaulicar and Werder, 2008; Eberhart, 2012; Tariq and Abbas, 2013), less attention has been given to how varying degrees of CG compliance impact financial performance across different governance models. For example, Reddy et al. (2010) found a positive relationship between CG compliance and FP in New Zealand featuring the shareholder-oriented (Anglo-American), while Goncharov et al. (2006) reported that German firms with higher CG compliance were priced at a premium in Germany, which features the stakeholder-oriented (Continental European) model. However, CG compliance may also entail costs; some studies (Bauer et al., 2004; Aluchna and Kuszewski, 2020) suggest that high compliance costs could initially hinder profitability, especially where resources are directed toward compliance rather than immediate business growth. Bauwhede (2009) counters this is because of poorly governed companies using the available discretion over the timing of asset sales to cover up their inherently lower operating performance. Another reason may be that companies with low levels of governance compliance report less conservative profit margins for tax benefits. Overall, the impact of CG compliance on profitability, thus, depends on both the quality of CG practices and the context in which they are implemented.

In Pakistan, where family-owned firms dominate and investor confidence may be weaker because of single-tier decision-making, CG compliance is particularly important. The CCGP mandates independent directors, audit committees and separation of management and oversight roles to bolster transparency and monitoring. Adopting these measures enhances access to capital and supports investor trust, which are vital for firm growth and profitability in such a market:

The relationship between CS and FP has been a subject of much debate. Modigliani and Miller's (1958) theorem, which posits that in a perfect market, CS does not affect firm value, has been critiqued for its reliance on restrictive assumptions such as no information asymmetry and no transaction costs. In real-world markets, frictions, imperfections and agency problems do exist, which can influence the effect of CS on FP.

The pecking order theory (Myers and Majluf, 1984) suggests that firms prefer internal financing over debt and equity because external financing signals lack of internal resources or management confidence. Debt can signal management’s belief in the firm’s ability to repay and maintain performance, yet, at the same time, it introduces potential agency problems between shareholders and lenders. Agency theory (Jensen and Meckling, 1976) further explains that debt can serve as an external monitoring mechanism, aligning management’s interests with shareholders by imposing discipline through debt obligations. However, this can also lead to agency costs between debtholders and equity-holders, particularly when firms over-leverage, resulting in higher risk and potential conflict between stakeholders.

Trade-off theory (Kraus and Litzenberger, 1973) offers a more nuanced view, emphasizing the balancing act between the tax advantages of debt and the costs associated with borrowing, including the risk of financial distress. In emerging markets, where legal frameworks and investment protections may be weaker, the costs associated with debt, such as bankruptcy and agency issues, can be more pronounced, making high leverage more detrimental to profitability. Market timing theory (Myers, 1984; Baker and Wurgler, 2002) suggests that firms might take on debt based on short-term market conditions, but in markets with higher volatility and risks, this short-term approach can harm long-term profitability, especially if firms misjudge the optimal CS.

Empirically, the literature is mixed. In developed markets, many studies report a positive relationship between CS and FP (Berger and Di Patti, 2006; Margaritis and Psillaki, 2010; Gill and Mathur, 2011), suggesting that appropriate leverage levels can signal confidence and enhance firm value. However, in emerging and developing markets, the relationship tends to be less straightforward. Few studies (Ross, 1977; Majumdar and Chhibber, 1999; Abor, 2007; Boshnak, 2023) have found that higher leverage can negatively affect performance, particularly because of the high costs of debt, bankruptcy risks and a lack of investor protection.

The negative impact of high leverage on profitability can be exacerbated in environments where institutional protections and market efficiencies are limited. In such contexts, like Pakistan, firms may over-leverage in an attempt to mitigate agency costs, but the resulting financial distress, increased bankruptcy costs and the burden of excessive debt can lead to reduced profitability (Abor, 2007; Harris and Raviv, 2006). Furthermore, firms that are highly leveraged may become more conservative in their decision-making, avoiding risky but potentially profitable investments to preserve their financial stability, which ultimately harms profitability:

The existing literature has often examined the individual relationships between CG, CS and FP in isolation (Ngatno et al., 2021; Ronoowah and Seetanah, 2023). However, the joint effects of these factors, particularly how CG and CS interact to influence FP, have not been sufficiently explored. We propose that CS plays a mediating role in the relationship between CG and FP.

CG mechanisms, such as the appointment of independent directors and board oversight, are designed to reduce agency costs and align the interests of management and shareholders, which theoretically should improve FP. However, CG’s direct impact on performance is often contingent upon other external and internal factors (Detthamrong et al., 2017). For instance, appointing independent directors may not immediately lead to higher sales or profits. Instead, the board’s role in making optimal investment and financing decisions is critical. When governance mechanisms are strong, they may enable firms to make better CS choices, which, in turn, can enhance FP.

From a theoretical standpoint, both pecking order theory (Myers and Majluf, 1984) and market timing theory (Baker and Wurgler, 2002) argue that firms prefer debt financing over equity under conditions of information asymmetry, low borrowing costs and when debt serves as a signal of management’s confidence and the firm’s creditworthiness. These theories suggest that firms with strong governance are more likely to make informed CS decisions that align with long-term profitability (Shakri et al., 2022). Therefore, CG could influence the choice of debt financing, which, in turn, may enhance FP.

However, while agency theory (Jensen and Meckling, 1976) suggests that both CG and CS can support shareholder wealth maximisation, it also highlights that high leverage can lead to potential agency problems, where managers may become over-conservative or risk-averse, undermining performance. In this way, CG’s ability to influence CS decisions could be counterbalanced by the negative impacts of excessive debt, especially when it leads to higher compliance costs or bureaucratic inefficiencies. This suggests a dual effect: CG can either strengthen or weaken FP depending on how it interacts with CS.

Empirically, periods of financial crises, such as the 1997 and 2008 financial crises, demonstrate that despite the presence of strong CG frameworks in developed markets, excessive debt can diminish the potential positive effects of CG on FP (Bhagat and Bolton, 2019). In Pakistan’s context, political turmoil and volatile investor confidence have further complicated the relationship between governance and performance. While changes to the CG framework since 2002 have provided greater assurance to investors, they have also introduced higher compliance costs for firms (Shakri et al., 2024). These dynamics create a context in which CS could mediate the relationship between CG compliance and FP.

Thus, we argue that CS plays a mediating role by influencing the effectiveness of governance mechanisms on FP. In particular, when firms adopt optimal governance practices, they may make better CS decisions (e.g. choosing appropriate levels of debt financing), which, in turn, can enhance FP. However, when governance is not aligned with optimal CS decisions, the firm may face higher debt-related risks, leading to negative effects on performance:

To test the hypotheses, data on all non-financial companies from 2008 to 2018 were collected. Financial data was manually extracted from annual reports. The sample period commences in 2008 because of data availability and encompasses the 2012 revision of the CCGP where firm requirements changed from voluntary to mandatory compliance. This required continued scoring across two distinct compliance periods ( Appendix). A CG compliance index was constructed according to the CCGP’s prescriptions. Excluding firms with missing data, the final panel data set comprises 2,314 firm-year observations.

The primary variable of this study is CG compliance, measured as an unweighted CG compliance index (CGCI_UW). Following Ararat et al. (2017) and Black et al. (2017), CGCI_UW is constructed to predict firm-level governance and its associations. The unweighted index assigns equal weight to each CCGP compliance item (1 for compliance, 0 otherwise), and it is divided into five sub-indices: “board structure”; “responsibilities, powers and functions of the board”; “corporate and financial reporting framework”; “functions and responsibilities of audit committee” and ‘eligibility and responsibility of external auditors. Given the CCGP revision to guidelines in 2012, scores and observations for pre- and post-2012 periods are presented in the  Appendix. Additionally, a weighted index CGCI_W is adopted from Shakri et al. (2022) to assess robustness. There are some advantages to an unweighted index over a weighted approach, such as equal representation of all scoring categories, reduced concentration risk and computational simplicity in the appropriate determination of weights.

The dependent variable, FP is measured using return on assets (ROA). For robustness, other measures of FP are also used: return on equity and Tobin’s Q. Financial leverage, the mediating variable of CS, is measured with the debt-to-assets ratio (DTA).

To examine the mediating role of CS in the relationship between CG compliance and FP, this study adopts the mediation analysis framework of Detthamrong et al. (2017). The analysis involves testing for four conditions: a significant relationship between CG compliance (independent variable) and CS (mediator); a significant relationship between CS (independent variable) and FP (dependent variable); a significant relationship between CS (mediator) and FP (dependent variable); and a weak or no relationship between CG compliance and FP when CS is included as a mediator, to indicate partial or full mediation. Ordinary least squares regression is used to test these four hypotheses:

where Lev denotes financial leverage in a firm’s CS measured with the debt-to-assets (DTA) ratio. FP indicates firm profitability and performance, measured with ROA. CGCI denotes CG compliance using unweighted CG compliance index (CGCI_UW).

In equations (1)–(4), Controls denote a set of control variables that can impact the performance and profitability of the sample firms and are selected based on the literature (Nguyen et al., 2021; Shakri et al., 2022; Zaman et al., 2022). These financial variables include firm age (FIRM_AGE), firm size (FIRM_SIZE), fixed asset ratio (FAR), growth (SALES_GROW), market capitalisation (MARK_CAP) and property, plant and equipment (PPE). All estimations included industry and year effects. Table 1 presents the definitions of the study’s variables.

Table 2 presents the descriptive statistics. Mean ROA is 0.03, while CG compliance scores calculated using unweighted (CGCI_UW) index averaged 3.96. These findings align with previous research (Shakri et al., 2022; Wang et al., 2019).

The mean debt-to-asset ratio (DTA) of 0.58 indicates that 58% of assets are debt-financed, highlighting the significance of CS as the mediating variable. Averages of the financial control variables include firm age (3.46 years), the natural log of firm size (8.22), fixed asset ratio (0.54), sales growth (1.86), market capitalisation (2.82) and property, plant and equipment (3.47).

Table 3 presents the correlations among the study’s variables. Positive correlations between CGCI_UW with ROA offer preliminary support for a positive association between CG compliance and FP. Correlations between other dependent, independent, control and mediating variables remain below 0.50. To assess for multicollinearity, variance inflation factors were calculated, and all values were below the critical threshold of 10, indicating no multicollinearity concerns.

Table 4 presents the main results of the empirical models for equations (1)–(4) of Section 3.3, which test the relationships among CG compliance, CS and FP. Columns (1) and (2) test the relationship between CG compliance (CGCI_UW) and financial leverage (DTA), first with a multivariate regression, then with the inclusion of industry and year effects. The results show that CG compliance negatively impacts CS, supporting the first mediation condition and H1. Columns (3) and (4) test the relationship between CG compliance and FP (ROA), and a positive and significant association is found, satisfying the second mediation condition and supporting H2. Columns (5) and (6) demonstrate a negative relationship between CS and FP, consistent with H3 and the third condition for mediation.

The estimations that combine CG compliance, FP and CS as the mediator are shown in Columns (7) and (8) of Table 4. The inclusion of the mediator (DTA) has weakened the relationship between CG compliance and FP, indicating partial mediation by financial leverage and supporting the fourth mediation condition and H4. Among the control variables, firm size and property, plant and equipment positively influence financial leverage, while firm age, fixed asset ratio, sales growth and market capitalisation exhibit negative relationships.

Several robustness tests are performed to check the validity of the results, presented in Table 5. First, to address possible criticisms of adopting the methodology for an equally weighted index (Jiraporn and Chintrakarn, 2009; Benjamin and Mat Zain, 2015), a weighted index for CG compliance (CGCI_W) was used as an alternative proxy to the independent variable, following Shakri et al. (2022). The results, presented in Column (1) of Table 5, indicate a consistent positive relationship between CG compliance and FP, and the mediating variable is still negatively related to FP, supporting the earlier findings. Second, alternative proxies of the dependent variable are tested. The results for the impact of CG compliance on FP using Tobin’s Q in Column (3) of Table 5 are also positive and significant, as with the earlier tests using ROA.

Then, we re-tested the main analysis by including additional CG control variables, presented in Column (4). Given that the study’s sample period includes the GFC years of 2008 and 2009, it may impact FP and the leverage ratio. Column (5) excludes firm-year observations over the GFC period. When compared against the main analysis, both rounds of tests provide consistent findings of a positive relationship between CG compliance and FP, with CS as a mediating variable.

Empirical studies on CG and firm profitability are often affected by endogeneity concerns (Zaman et al., 2022). Therefore, a set of strategies was used to identify these issues. The techniques include a lead dependent variable (Lead DV), a FE model, system generalised method of moments (System GMM) and PSM, to control for potential endogeneity bias. The results are presented in Table 6 across Columns (1)–(6).

First, to address reverse causality, one-year lead values of the dependent variable were applied to the estimation of equation (4). The impact of CG compliance on FP is still significantly positive, while CS remains negative and significant. These results demonstrate that reverse causality does not drive the main findings. Furthermore, to control for omitted time-invariant and firm-specific factors, firm and year FE were incorporated into the model, with the resulting coefficients for CGCI_UW and DTA remaining significantly positive and negative, respectively, indicating robustness to alternative specifications.

The main results may be biased because of dynamic panel endogeneity that can drive the association between CG compliance (CGCI_UW) and firm profitability (ROA). Following Gull et al. (2023), we address this issue using the dynamic model system GMM estimation. Column (3) of Table 6 reports the system GMM, in which the CGCI_UW and ROA relationship are still positive at a 1% significance level. The Arellano–Bond test (AR1) for first differences and the Sargan test of over-identification required to validate the use of system GMM are reported. The Hansen overidentification test also shows that the instruments used in the regressions are exogenous. These tests validate the results and demonstrate that endogeneity does not drive our main results.

To address another endogeneity concern because of potential selection bias from firm-specific characteristics, the PSM approach (Rosenbaum and Rubin, 1983) was used. This approach refers to the possibility that improving or deteriorating financial performance is attributed to firm-level factors rather than CG compliance. To perform PSM, a dummy variable (CGCI_UW_DUMMY) was created based on the median CG compliance score. In this case, the dummy variable took the value of 1 if a firm’s CG compliance score is greater than the sample median or 0 otherwise. Treatment and control groups were then formed using this dummy variable. The treatment group included firms having a value of 1 in CGCI_UW_DUMMY, whereas firms with a zero value are included in the control group. Then, the predicted value of CG compliance was estimated with probit regressions for the dummy variable (CGCI_UW_DUMMY) on the same control variables used in the baseline model. This process yielded the propensity scores for all the firm-year observations. Next, two identical sub-samples were composed based on different criteria (the treatment and control groups) using the propensity scores. A matched sample was used to investigate the nexus of CG compliance and financial performance. Diagnostic tests were also conducted to ensure that PSM was applied correctly. Firstly, probit regressions were performed on CGCI_UW_DUMMY as a dependent variable on the original and matched sample. The results of the pre- and post-matched probit regressions using CGCI_UW_DUMMY as a dependent variable is reported in Columns (4) and (5) of Table 6. According to the results, in the pre-matched probit regressions, several firm-level variables were significantly associated with CGCI_UW_DUMMY, while none of the firm-level variables were significantly associated with CGCI_UW_DUMMY in the post-matched probit regressions. This implies that the PSM procedure was applied correctly. We also re-estimated our baseline model by using a matched sample. According to the results in Column (6) of Table 7, they are qualitatively similar to those reported in the main analysis (Table 4). This confirms that the negative (positive) association between levels of CG compliance (CGCI_UW) and financial performance (ROA) is not the result of firm-level characteristics.

Post-matched sample univariate analysis was also conducted to ensure that the treatment and control groups were comparable, and no significant differences were found between the two groups. Table 7 reports the results of the mean differences between the characteristics of the treatment and control groups. Based on the findings, both groups were identical, confirming the accuracy of the PSM procedure.

Following a mediation analysis framework, this study has established CS partially and negatively mediates the relationship between CG compliance and FP in Pakistan. In this section, we perform SEM path analysis based on Jollineau and Bowen (2023) and Sun (2023), to determine direct and indirect (mediated) path coefficients with the following set of equations:

Equation (5) includes DTA as a mediator for the relationship between CG compliance (CGCI) and FP (ROA), and equation (6) uses the link between DTA and CGCI to establish the mediation effect. The magnitude of a direct path from CGCI to ROA is β₁. The path coefficient from CGCI to DTA is X₁, and DTA to ROA is β₂. Thus, the path coefficient X₁ × β₂ captures the total magnitude of the indirect path from CGCI to ROA through DTA. This path analysis is illustrated in Panel A of Figure 1.

Panel B of Figure 1 reports the empirical results of the path analysis. The magnitude of the direct path, β₁, is 0.11, and it is statistically significant at the 1% level. For the mediated path from CGCI to DTA, X₁ is −0.15, and from DTA to ROA, β₂ is −0.22. Both mediated coefficients are also significant at the 1% level. The total indirect effect is −0.033. These results from the path analysis validate the baseline findings suggesting CS mediates the relationship between CG compliance and FP.

The key findings of this study are summarized as follows. Firstly, CG compliance is inversely related to financial leverage. This implies that Pakistani firms with greater compliance with CG requirements and practices are likely less reliant on debt for funding and as an external monitoring mechanism. These findings are consistent with the theoretical and empirical literature (Jensen and Meckling, 1976; Haque et al., 2011; Morellec et al., 2012). Emerging markets like Pakistan can benefit from firmer oversight and governance, which improve investor confidence in equity asset classes.

Secondly, CG compliance is positively related to financial performance. This provides additional support to the role and effectiveness of CG compliance in aligning the interests of firm management with those of shareholders (Ararat et al., 2017; Black et al., 2017; Tariq and Abbas, 2013). The additions to Pakistan’s CCGP since its introduction in 2002, significant revision in 2012 and subsequent updates to clarify the scope of firm compliance have contributed to overall improvements in the financial outcomes of the country’s corporate sector.

Thirdly, financial leverage is negatively associated with financial performance. Firms with more debt in the CS face decreases to profitability, especially when financial and trade-off costs are higher in emerging markets with a greater degree of informational asymmetry (Abor, 2007; Harris and Raviv, 2006; Zeitun, 2014). This indicates that CS can mediate the relationship between CG compliance and FP.

Finally, the results indicate that the relationship between CG compliance and FP decreases with the inclusion of CS, with evidence of partial mediation (Detthamrong et al., 2017). Firms in developing economies are likely to underestimate the costs of financial distress and bankruptcy, potentially leading to higher debt levels than necessary, undermining profitability. Over-leveraging can also lead to management adopting discretionary behaviours such as over-conservatism, short-termism and underinvestment. These behaviours would decrease FP over the longer term. Although lenders can contribute toward greater firm monitoring and help to mitigate agency problems, this is likely to be less effective for firm value, especially when a firm has strong CG mechanisms and compliance in place.

This study focused on firm compliance with regulations and guidelines on governance practices rather than individual mechanisms of CG. Higher compliance levels signal greater alignment with investors’ objectives and improve market confidence. This reduces firm reliance on debt sources for funding and the role of lenders as external firm monitors. Greater compliance costs coupled with higher levels of debt can also limit FP.

The findings of this study offer important implications for firms and their stakeholders including investors, regulators and lenders. Regulators are encouraged to provide greater prescriptions on best practices and guidelines to clarify recommended levels of compliance. Firms and lenders in emerging markets could consider the implications of firm compliance on the optimal mix of CS and the trade-offs associated with compliance costs and the costs of debt. Minority shareholders may also consider the informational signals associated with compliance levels when making investment decisions.

We identify avenues for future research arising from this study’s limitations. Firstly, though higher leverage can partially mediate the beneficial relationship between CG compliance and FP, more channels can be explored to further validate the positive effects of CG compliance. Secondly, this study focused on Pakistan, an emerging South Asian market that only recently adopted mandatory compliance to the prescribed CG framework and will benefit from future studies that can compare the effectiveness of adopting alternative CG models, especially across the region with similar corporate cultures.

Table A1