Global Advances in Business Studies (GABS), 2026.
Vol.
5 No.
1, 40-52
GLOBAL ADVANCES IN BUSINESS STUDIES (GABS)
http://journal-gabs.
org/gabs Analysis of ESG scores on Malaysian banking performance: a panel data approach Naufal Shadiq1* Faculty of Economics and Business Universitas Negeri Jakarta.
Indonesia Article info Article history:
Received: 5 February 2026 Accepted: 19 April 2026 Published: 29 April 2026 Keywords:
ESG.
banking performance.
Malaysia.
nonlinear effects.
firm heterogeneity Abstract This study study examines the relationship between environmental, social, and governance (ESG) scores and bank performance in Malaysia over 2019Ae2024.
We focus on profitability, measured by return on assets adjusted for risk-weighted assets (ROA RWA), and firm value, proxied by TobinAos Q.
Using a fixed effects framework as the primary specification, complemented by nonlinear and interaction models, we find no statistically significant linear or lagged effect of ESG on profitability or firm value.
However, we document a nonlinear .
nverted U-shape.
relationship between ESG and firm value, indicating diminishing returns at higher ESG levels.
Furthermore, the interaction between ESG and firm size is negative and significant, suggesting that ESG has stronger valuation effects for smaller banks.
These findings imply that ESG is not a universally value-enhancing factor but operates through nonlinear and conditional mechanisms.
The study contributes to the ESGAebanking literature by reconciling mixed prior evidence and highlighting the importance of model specification and firm heterogeneity.
JEL classifications: G21.
M14 Citation:
Shadiq.
Analysis of ESG scores on Malaysian banking performance: a panel data approach.
Global Advances in Business Studies, 5.
, 40-52, https://doi.
org/10.
55584/Gabs.
*Corresponding author:
shadiqnaufal027@gmail.
E-ISSN: 2828-8394
org/10.
55584/Gabs.
Introduction The increasing emphasis on environmental, social, and governance (ESG) practices has reshaped expectations toward financial institutions, particularly banks.
As highly regulated entities that rely heavily on public trust, banks are no longer evaluated solely based on financial performance, but also on their sustainability, governance quality, and social responsibility (Friede et al.
, 2.
This shift reflects the growing recognition that long-term financial stability is closely linked to responsible and transparent business practices.
Consequently.
ESG has emerged as a strategic framework that not only enhances firm value and stakeholder trust but also supports risk management and institutional resilience.
In particular, the governance and risk management components embedded within ESG play a critical role in strengthening financial stability in the banking sector (Menicucci & Paolucci, 2.
From a theoretical perspective, the resource-based view (RBV) and dynamic capabilities frameworks conceptualize ESG as a strategic intangible asset that enhances reputational capital, strengthens stakeholder trust, and reduces information asymmetry (Azmi et al.
, 2021.
Wernerfelt, 1.
Dynamic capabilities theory further suggests that ESG reflects a firmAos adaptive capacity to respond to regulatory, environmental, and societal pressures, potentially transforming sustainability initiatives into long-term value creation (Aydomu et al.
, 2022.
Teece et al.
, 1.
However, despite strong theoretical foundations, empirical evidence on the ESGAe financial performance nexus remains inconclusive, particularly in emerging markets.
While some studies report positive associations between ESG and financial performance, others find weak, insignificant, or even negative relationships (Buallay.
McWilliams & Siegel, 2.
These mixed findings suggest that the ESGAe performance relationship may not be strictly linear and may depend on firm-specific characteristics as well as the level of ESG engagement.
In addition, measurement and identification challenges remain important concerns in this literature.
ESG performance and financial outcomes may be jointly determined, as financially stronger banks possess greater capacity to invest in sustainability initiatives (Azmi et al.
, 2021.
Wintoki et al.
, 2.
Furthermore, the choice of performance measures is critical.
Conventional return on assets (ROA) does not fully capture risk exposure in banking, whereas return on assets based on risk-weighted assets (ROA RWA) provides a more appropriate risk-adjusted measure of profitability.
Meanwhile.
TobinAos Q reflects forward-looking market valuation (Aydomu et al.
Nizam et al.
, 2.
Against this backdrop.
Malaysia provides a relevant empirical setting.
As part of ASEANAos evolving financial landscape.
Malaysian banks face increasing regulatory pressure for sustainability disclosure, while the depth and effectiveness of ESG implementation remain heterogeneous (Mohammad & Wasiuzzaman, 2.
This institutional context allows for an examination of whether ESG functions as a genuine value-creating mechanism or primarily reflects compliance and existing firm Accordingly, this study investigates the relationship between ESG performance and bank performance in Malaysia over the period 2019Ae2024.
Unlike prior studies that primarily focus on linear relationships, this study extends the analysis by incorporating nonlinear and conditional effects to capture potential diminishing returns and firm-level heterogeneity.
By focusing on risk-adjusted profitability (ROA RWA) and firm value (TobinAos Q), this study contributes to the literature by providing evidence that ESG effects are not uniform, but depend on both the level of ESG engagement and firm characteristics.
Shadiq Global Advances in Business Studies 2026, 5.
, 40-52
E-ISSN: 2828-8394
org/10.
55584/Gabs.
Conceptual framework ESG as a strategic resource Environmental, social, and governance (ESG) performance reflects a firmAos commitment to sustainable and responsible business practices beyond traditional financial metrics.
In recent years.
ESG has gained increasing importance as a framework for evaluating corporate performance, particularly in relation to long-term value creation and risk management (Friede et al.
, 2.
In the banking sector.
ESG integration is especially relevant due to high regulatory scrutiny, systemic risk exposure, and strong reliance on stakeholder trust.
From the resource-based view (RBV), firms are conceptualized as bundles of heterogeneous resources that generate competitive advantage when they are valuable, rare, and difficult to imitate (Barney, 1991.
Wernerfelt, 1.
ESG practicesAisuch as governance transparency, risk management systems, and stakeholder engagementAican be interpreted as intangible strategic resources embedded within organizational processes (Azmi et al.
, 2.
These resources enhance reputational capital, reduce information asymmetry, and may improve access to financing.
However, the RBV perspective alone may not fully explain firm behavior in dynamic financial environments.
The dynamic capabilities framework extends this view by emphasizing a firmAos ability to adapt to evolving regulatory, environmental, and societal pressures (Teece et , 1.
Accordingly.
ESG can be interpreted not only as a strategic resource but also as an adaptive capability that supports risk management, strengthens institutional resilience, and aligns long-term strategic objectives (Menicucci & Paolucci, 2023.
Nizam et al.
, 2.
Thus.
ESG can be conceptualized as a strategic capability that influences firm outcomes through both internal efficiency and external market perception.
ESG and banking financial performance The relationship between ESG and financial performance in banking operates through two key channels: risk-adjusted profitability and market valuation.
First, profitability in banking must be evaluated relative to risk exposure.
Traditional return on assets (ROA) does not fully capture differences in asset risk.
Therefore, return on assets based on risk-weighted assets (ROA RWA) provides a more appropriate measure of managerial efficiency within regulatory capital frameworks (Menicucci & Paolucci, 2023.
Nizam et al.
, 2.
ESG practices may enhance internal controls, improve risk monitoring, and strengthen credit quality, which in turn may influence risk-adjusted profitability.
Second, firm value is better reflected through forward-looking market measures such as TobinAos Q (Aydomu et al.
, 2.
ESG performance may serve as an indicator of long-term stability and governance quality, thereby reducing information asymmetry and influencing investor expectations.
As a result.
ESG may affect market valuation differently from accounting-based profitability.
Nonlinear and conditional effects of ESG performance Despite strong theoretical arguments, empirical evidence on the ESGAeperformance relationship remains mixed (Buallay, 2019.
McWilliams & Siegel, 2.
This inconsistency suggests that the ESGAeperformance relationship may not be strictly linear and may vary depending on firm characteristics and the level of ESG engagement.
From an economic perspective.
ESG investments may exhibit diminishing marginal Initial ESG adoption improves governance quality, transparency, and stakeholder trust, thereby enhancing firm value.
However, as ESG engagement increases, additional investments may generate lower incremental benefits while incurring higher compliance costs and managerial complexity (Azmi et al.
, 2.
This trade-off is consistent with prior findings that suggest a nonlinear relationship between sustainability practices and financial performance.
Shadiq Global Advances in Business Studies 2026, 5.
, 40-52
E-ISSN: 2828-8394
org/10.
55584/Gabs.
where the benefits of ESG may decline beyond a certain threshold (Nollet et al.
, 2.
While ESG contributes to improvements in organizational processes and long-term orientation (Eccles et al.
, 2.
, its marginal impact on firm value may weaken at higher levels of engagement.
This implies a nonlinear .
nverted U-shape.
relationship, where ESG contributes positively up to an optimal level, beyond which its marginal effect declines.
In addition.
ESG performance may function as a signaling mechanism in financial markets characterized- by information asymmetry.
According to signaling theory, firms use observable indicators to convey private information about their underlying quality to external stakeholders (Connelly et al.
, 2011.
Spence, 1.
In this context.
ESG disclosures may signal governance quality, risk management capability, and long-term strategic orientation.
This signaling effect is likely to be stronger for smaller banks, where information asymmetry is higher and credibility needs to be established.
In contrast, for larger banks.
ESG practices may be perceived as standard expectations, thereby reducing their incremental signaling value.
Taken together, these arguments suggest that ESG effects are both nonlinear and conditional, rather than uniformly Literature review and research gap The relationship between environmental, social, and governance (ESG) performance and corporate financial performance remains inconclusive, with studies reporting positive, negative, and insignificant results across sectors, including banking (Aydomu et al.
, 2022.
Nollet et al.
While bankAos ESG practices may enhance stakeholder trust and access to finance (Nizam et al.
, 2.
, their implementation costs can initially outweigh benefits, particularly in the banking sector (Menicucci & Paolucci, 2.
To explain these mixed findings, prior research suggests a nonlinear relationship.
While some studies argue for a U-shape (Barnett & Salomon, 2.
, in the heavily regulated banking sector where compliance costs are steep.
ESG investments are more likely to exhibit diminishing returns, forming an inverted U-shape where excessive ESG engagement eventually penalizes market valuation.
From a theoretical perspective, signaling theory explains how ESG disclosures reduce information asymmetry by signaling firm quality and risk management capability to stakeholders (Connelly et al.
, 2011.
Spence, 1.
This mechanism is particularly relevant in banking, where transparency and trust are critical (Menicucci & Paolucci, 2.
However.
ESG effects are not uniform, as aggregated scores may obscure differences across environmental, social, and governance dimensions, which can produce heterogeneous financial impacts (Aydomu et al.
, 2.
In emerging markets such as Malaysia.
ESG adoption remains uneven despite increasing regulatory pressure, and its impact on firm value has been shown to depend on firm-specific conditions (Mohammad & Wasiuzzaman, 2.
However, prior Malaysian banking studies have predominantly relied on linear models without adequately controlling for unobserved firm heterogeneity via fixed effects, leaving a significant methodological gap.
Accordingly, this study addresses this gap by examining the nonlinear and conditional effects of ESG on riskadjusted profitability (ROA RWA) and firm value (TobinAos Q) in the Malaysian banking sector.
Research methods Research design and objective The primary objective of this study is to examine the relationship between environmental, social, and governance (ESG) performance and the financial performance of Malaysian commercial banks.
ESG performance is primarily measured using ESG combined score (ESGC), which captures overall sustainability performance across environmental, social, and governance dimensions.
Specifically, the study investigates whether ESG performance Shadiq Global Advances in Business Studies 2026, 5.
, 40-52
E-ISSN: 2828-8394
org/10.
55584/Gabs.
risk-adjusted profitability measured by return on assets based on risk-weighted assets (ROA RWA) and .
market valuation measured by TobinAos Q.
In addition to examining the direct relationship, this study further explores whether the ESGAeperformance relationship is nonlinear and varies across firm characteristics.
This approach is motivated by prior literature suggesting that ESG effects may not be uniform and may depend on the level of ESG engagement as well as firm-specific conditions.
This research adopts a quantitative approach using secondary panel data.
The population consists of financial institutions listed on Bursa Malaysia.
However, the accessible population is restricted to commercial banks with consistent and publicly available ESG and financial data during the 2019Ae2024 period.
A purposive sampling technique is applied based on the following criteria: .
listed on Bursa Malaysia, .
classified under the banking sector, .
availability of ESG scores from Refinitiv, and .
availability of audited annual financial statements.
Based on these criteria, eight Malaysian commercial banks meet the selection requirements.
The final dataset forms an unbalanced panel covering the period 2019Ae2024, with the number of observations varying across models due to ESG data availability.
ESG data, including the combined score and individual pillar scores .
nvironmental, social, and governanc.
, are obtained from the Refinitiv ESG database.
Financial data are collected from banksAo audited annual reports and financial statements, while macroeconomic variablesAisuch as GDP growth and inflation rateAiare sourced from Bank Negara Malaysia and the World Bank database.
To ensure robustness and mitigate the influence of extreme values, all continuous variables are winsorized at the 1st and 99th percentiles.
The operational definitions and measurement formulas of all variables are presented in Table 1.
Tabel 1.
Research variables Variable Dependent variable ROA RWA TobinAos Q Definition Measurement Measures the bankAos ability to generate profit from riskweighted assets.
Captures market valuation relative to book value of ROA RWA = Net income Risk Oe weighted assets TobinA s Q = MV equity BV of debts BV of assets Independent variable ESG combined score Overall sustainability performance based on environmental, social, and governance dimensions.
ESG score .
Ae.
Control variable Firm size Firm leverage Loan-to-deposit ratio (LDR) Capital adequacy ratio (CAR) GDP growth Inflation rate Reflects bank scale and operational capacity.
Indicates reliance on external debt financing.
Measures liquidity and lending Reflects capital strength relative to risk exposure.
Captures macroeconomic expansion effects.
Represents changes in general price levels affecting banking Firm size = ln(Total asset.
Totaldebt Total assets Total loans LDR = Total deposits Tier 1 capital Tier 2 capital CAR = Risk-weighted assets GDPt Oe GDPtOe1 GDP growth = GDPtOe1 Firm leverage = Inflation rate = IHKt Oe IHKtOe1 IHKtOe1 Research hypotheses Drawing upon the resource-based view (RBV), dynamic capabilities framework, and Shadiq Global Advances in Business Studies 2026, 5.
, 40-52
E-ISSN: 2828-8394
org/10.
55584/Gabs.
signaling theory.
ESG performance is conceptualized as a strategic capability that may influence banking financial performance through improved risk management, reputational capital, and information signaling mechanisms.
However, prior literature suggests that the ESGAeperformance relationship may not be strictly linear and may vary across firm Accordingly, this study formulates the following hypotheses:
H1: There is a significant association between ESG performance and banking financial performance .
isk-adjusted profitability and market valuatio.
H2: The relationship between ESG performance and market valuation is nonlinear, exhibiting an inverted U-shaped pattern.
H3: The impact of ESG performance on market valuation is conditional upon firm size, with smaller banks experiencing stronger positive effects.
Empirical models To examine the relationship between ESG performance and banking financial performance, this study employs panel data regression models.
The primary specification is estimated using fixed effects (FE) model with robust standard errors, which controls for unobserved timeinvariant heterogeneity across banks.
The baseline models for risk-adjusted profitability (ROA RWA) and market valuation (TobinAos Q) are specified as follows:
ROA_RWAycnyc = 0 1 ESGCit Ock Controlsit i t Ait TQ it = 0 1 ESGCit Ock Controlsit i t Ait where i represents unobserved bank-specific effects, t captures year fixed effects to account for macroeconomic shocks, and Ait is the error term.
To capture potential nonlinear effects of ESG, a quadratic term (ESGC .
is introduced into the baseline specification.
This allows for testing whether ESG investments exhibit diminishing or nonlinear marginal returns:
ROA_RWAycnyc = 0 1 ESGCit 2 ESGCit2 Ock Controlsit i t Ait TQ it = 0 1 ESGCit 2 ESGCit2 Ock Controlsit i t Ait Furthermore, to examine whether the ESGAeperformance relationship varies across firm characteristics, an interaction term between the ESG score and firm size is included:
ROA_RWAycnyc = 0 1 ESGCit 2 SIZEit 3 (ESGCit y SIZEit ) Ock Controlsit i t Ait TQ it = 0 1 ESGCit 2 SIZEit 3 (ESGCit y SIZEit ) Ock Controlsit i t Ait To address potential endogeneity concerns, a two-stage least squares .
SLS) approach is applied as a robustness check.
The first-stage regression predicts the ESG score using instrumental variables:
ESGCit = 0 1 LagESGCit 2 PeerESGCit Ock Controlsit i t it where LagESGC represents the one-year lagged ESG score, and PeerESGCrepresents the C it ) leave-one-out industry mean ESG score.
In the second stage, the predicted ESG values (ESGC are substituted back into the main performance equations:
C it Ock Controlsit i t Ait ROA_RWAycnyc = 0 1 ESGC C it Ock Controlsit i t Ait TQ it = 0 1 ESGC Shadiq Global Advances in Business Studies 2026, 5.
, 40-52
E-ISSN: 2828-8394
org/10.
55584/Gabs.
For the 2SLS robustness check, instrument strength and validity were confirmed using the KleibergenAePaap Wald F-statistic for instrument relevance and the SarganAeHansen J-test for overidentification exogeneity.
This study is based on a relatively small panel of eight Malaysian commercial banks observed over a six-year period .
9Ae2.
Consequently, the empirical models may be subject to limited statistical power and potential overfitting, particularly in specifications involving multiple control variables and instrumental variables.
As a result, the findings should be interpreted with caution and viewed as indicative rather than definitive.
Findings Descriptive statistics Table 2 presents the descriptive statistics of the main variables.
The average ESG combined score (ESGC) is 67.
86, indicating a moderate level of sustainability performance among Malaysian banks.
Risk-adjusted profitability (ROA RWA) has a mean of 0.
0149, while TobinAos Q averages 0.
868, suggesting relatively stable market valuation across the sample.
The dispersion of ESGC is moderate, while TobinAos Q shows relatively low variation, reflecting a relatively homogeneous banking environment.
Table 2.
Descriptive statistics Variable Obs Mean
ESGC
Std.
Dev.
Min Max
ROA RWA 46
SIZE
FLEV
CAR
LDR
INF
Pre-estimation diagnostic Table 3 reports the Pearson correlation matrix.
ESGC is positively correlated with ROA RWA .
= 0.
389, p < 0.
but shows no significant linear correlation with TobinAos Q.
Notably.
ESGC is strongly correlated with firm size .
= 0.
860, p < 0.
, suggesting that larger banks tend to exhibit higher ESG performance.
To ensure this high bivariate correlation does not compromise the regression models, a formal multicollinearity diagnostic was required.
The initial set of control variables included GDP growth (GDPG) as a proxy for macroeconomic conditions.
However, during the preliminary diagnostic phase.
GDPG was found to exhibit a high bivariate correlation with inflation (INF).
To ensure model parsimony and prevent potential multicollinearity issues that could bias the standard errors.
GDPG was excluded from the final empirical specifications, leaving INF as the sole macroeconomic control variable.
As shown in Table 4, despite the strong bivariate correlation between the ESG combined score and firm size, the variance inflation factor (VIF) analysis revealed a maximum VIF of 49 for Firm Size and 5.
08 for the ESG score, with a Mean VIF of 2.
Because all individual VIF values remain safely below the strict conventional threshold of 10, multicollinearity does Shadiq Global Advances in Business Studies 2026, 5.
, 40-52
E-ISSN: 2828-8394
org/10.
55584/Gabs.
not pose a severe threat to the stability of the baseline models.
However, as standard econometric best practice, to maintain structural stability and minimize non-essential multicollinearity in the interaction model, the continuous variables ESGC and SIZE were meancentered prior to generating the interaction term.
Table 3.
Pearson correlation matrix
Variable ESGC
ESGC
ROA RWA TQ
SIZE
FLEV
CAR LDR INF
ROA RWA 0.
389*** 1
SIZE
860*** 0.
FLEV
320** 1
CAR
450*** 1
LDR
INF
Note : ** p < 0.
05, *** p < 0.
Table 4.
Variance inflation factor (VIF) diagnostic Variable VIF
1/VIF
Firm size (SIZE) ESG combined score (ESGC) Firm leverage (FLEV) Capital adequacy ratio (CAR) Loan-to-deposit ratio (LDR) Mean VIF Table 5.
Hausman test
ROA RWA
Tobin's Q Hausman Degrees of freedom p-value Preferred estimator Fixed effects Fixed effects Note : ** p < 0.
A Hausman specification test was employed to adjudicate between random effects (RE) and Fixed Effects (FE) estimators, determining whether unobserved firm-specific heterogeneity Shadiq Global Advances in Business Studies 2026, 5.
, 40-52
E-ISSN: 2828-8394
org/10.
55584/Gabs.
is systematically correlated with the independent variables (Table .
For risk-adjusted profitability (ROA RWA), the test strongly rejected the null hypothesis (N^2 = 14.
36, p = 0.
, confirming that unobserved heterogeneity is indeed correlated with the regressors.
This statistically mandates the use of a Fixed Effects model.
For market valuation (TobinAos Q), the Hausman test did not reject the null hypothesis (N^2 = 4.
39, p = 0.
, suggesting a Random Effects model could be adequate.
However, to maintain methodological consistency across the study, the Fixed Effects estimator was retained as the primary specification for all dependent variables.
Beyond consistency, this decision is theoretically justified by the purposive sampling nature of the dataset.
Because the panel consists of a specific, non-random cohort of 8 Malaysian commercial banks, the bank-specific effects are more appropriately treated as fixed parameters to be estimated rather than random draws from a broader, unknown population.
Baseline and fixed effects estimation Table 6 presents the baseline and main regression results.
The pooled OLS estimates indicate that ESGC has a positive and statistically significant effect on ROA RWA ( = 000302, p < 0.
, suggesting that higher ESG performance is associated with improved riskadjusted profitability.
This finding is consistent with prior studies that argue ESG enhances internal governance, risk management, and operational efficiency in banking institutions (Azmi et al.
, 2021.
Menicucci & Paolucci, 2.
Table 6.
Baseline and fixed effects results Variables ROA RWA (OLS) ROA RWA (FE) ESGC 0.
000302*** .
TQ (OLS)
TQ (FE)
SIZE
032305** .
027533*** .
FLEV
042950** .
509674*** .
938999*** .
CAR
185220* .
LDR
INF
150320** .
Constant 520000 .
Note : *p < 0.
10, ** p < 0.
05, *** p < 0.
However.
ESGC does not significantly affect TobinAos Q in the baseline model, indicating that ESG performance may not be immediately reflected in market valuation, which is typically forward-looking and influenced by broader investor expectations (Aydomu et al.
, 2022.
Nizam et al.
, 2.
When unobserved bank-specific heterogeneity is controlled using fixed effects, the effect of ESGC becomes statistically insignificant for both ROA RWA and TobinAos Q.
This shift suggests that the positive association observed in the OLS model is likely driven by timeinvariant firm characteristics rather than ESG performance itself, consistent with concerns of endogeneity and omitted variable bias in ESG studies (Wintoki et al.
, 2.
Table 7 reports the nonlinear regression results.
For TobinAos Q.
ESGC exhibits a positive linear term and a negative quadratic term, indicating an inverted U-shaped relationship between ESG performance and firm value.
This finding suggests that ESG enhances market valuation up to an optimal level, beyond which additional ESG engagement yields diminishing returns.
Shadiq Global Advances in Business Studies 2026, 5.
, 40-52
E-ISSN: 2828-8394
org/10.
55584/Gabs.
Such a pattern is consistent with prior literature documenting nonlinear ESGAeperformance relationships driven by trade-offs between benefits and implementation costs (Barnett & Salomon, 2012.
Nollet et al.
, 2.
Nonlinear effects In contrast, the nonlinear specification for ROA RWA does not yield statistically significant coefficients, indicating that nonlinear effects are more pronounced in market-based performance than in accounting-based profitability.
This supports the argument that market valuation is more sensitive to strategic and reputational signals associated with ESG (Aydomu et al.
, 2.
Table 7.
Nonlinear effects Variables ROA RWA
ESGC
001949* .
ESGCA
000017** .
SIZE
FLEV
874352*** .
CAR
LDR
INF
Note : *p < 0.
10, ** p < 0.
05, *** p < 0.
Interaction effects Table 8.
Interaction effects Variables ROA RWA TobinAos Q ESGC (Centere.
000735* .
SIZE (Centere.
ESGC y SIZE 000240 .
000550** .
LDR
CAR
277414* .
325927** .
FLEV
868475*** .
INF
Observations R-squared (Withi.
Note : *p < 0.
10, ** p < 0.
05, *** p < 0.
Table 8 presents the results of the interaction model used to examine conditional effects.
The interaction term between ESGC and firm size is negative and statistically significant for Shadiq Global Advances in Business Studies 2026, 5.
, 40-52
E-ISSN: 2828-8394
org/10.
55584/Gabs.
TobinAos Q.
The marginal effect of ESG on firm value can be expressed as OCTQ/OCESG = CA CCSIZE, indicating that the impact of ESG decreases as firm size increases.
This result suggests that ESG provides relatively stronger valuation benefits for smaller banks, which is consistent with signaling theory.
In environments characterized by higher information asymmetry.
ESG disclosures serve as a stronger signal of credibility and governance quality, particularly for smaller institutions (Connelly et al.
, 2011.
Spence, 1.
Two-stage least squares estimation Table 9 reports the instrumental variable results.
Although the Hansen J-test suggests that the instruments are not rejected on overidentification grounds, the KleibergenAePaap F-statistic indicates a weak instrument problem.
Consequently.
ESGC is not statistically significant in explaining either ROA RWA or TobinAos Q.
Table 9.
Two-stage least squares .
SLS) results Variables ROA RWA
TobinAos Q
ESGC
SIZE
FLEV
119523*** .
CAR
LDR
Observations Kleibergen-Paap rk Wald F-statistic Hansen J-Statistic .
-valu.
First-Stage Diagnostics Note : *** p < 0.
These results suggest that the evidence for a causal relationship between ESG and financial performance is weak.
This is consistent with the broader ESG literature, which highlights the difficulty of establishing causality due to reverse causality and omitted variable bias (McWilliams & Siegel, 2000.
Wintoki et al.
, 2.
Discussion and conclusion Building on the empirical findings, this study provides a nuanced understanding of how ESG performance relates to banking outcomes in the Malaysian context.
The results indicate that the relationship between ESG and financial performance is not straightforward, but depends on model specification, firm characteristics, and the nature of the performance metric.
Under the baseline pooled OLS specification.
ESG performance demonstrates a positive and significant association with risk-adjusted profitability (ROA RWA).
However, this relationship becomes statistically insignificant when applying the Fixed Effects (FE) estimator.
This suggests that the positive association observed in OLS is largely driven by time-invariant, unobserved bank characteristics rather than ESG performance itself.
In addition.
ESG scores in the banking sector tend to be relatively persistent over time, such that the within-transformation in the FE model absorbs much of the variation needed to identify a significant linear effect (Wintoki et al.
, 2.
As a result, the evidence for a direct linear relationship between ESG and accounting-based performance remains inconclusive.
Shadiq Global Advances in Business Studies 2026, 5.
, 40-52
E-ISSN: 2828-8394
org/10.
55584/Gabs.
The sample period also includes the COVID-19 pandemic .
0Ae2.
, which had a substantial impact on both banking performance and ESG reporting practices.
This may introduce additional noise into the estimation and partially contribute to the weak and inconsistent linear relationships observed across models.
A similar absence of a robust linear effect is observed for market valuation (TobinAos Q).
However, the nonlinear specification reveals a significant inverted U-shaped relationship, indicating that ESG enhances firm value up to an optimal level, estimated around an ESG score of approximately 57, although this threshold should be interpreted cautiously given the small sample size.
Beyond this point, the marginal benefits of ESG diminish, suggesting that excessive ESG engagement may introduce compliance costs and managerial complexity that outweigh incremental gains.
This finding is consistent with prior studies documenting diminishing returns to ESG investments (Barnett & Salomon, 2012.
Nollet et al.
, 2.
and aligns with the resource-based view, where ESG operates as a strategic capability that yields decreasing marginal benefits at higher levels of investment (Barney, 1991.
Wernerfelt, 1.
Furthermore, the interaction model provides evidence of conditional effects.
The marginal effect of ESG on firm value is conditional on firm size and can be expressed as OCTQ/OCESG = CA CCSIZE.
The negative interaction coefficient indicates that the marginal impact of ESG decreases as firm size increases, implying that ESG provides relatively stronger valuation benefits for smaller banks.
This is consistent with signaling theory, where ESG disclosures serve as a stronger credibility signal for firms facing higher information asymmetry (Connelly et al.
, 2011.
Spence, 1.
While this study focuses exclusively on ESG combined score (ESGC), prior literature suggests that governance is often the most influential pillar in banking due to its direct link with risk management and regulatory compliance (Menicucci & Paolucci, 2.
The absence of strong linear ESGC effects may therefore reflect aggregation bias, where the combined score masks heterogeneous effects across individual ESG dimensions (Aydomu et al.
, 2.
This highlights a limitation of aggregated ESG measures and suggests that future research may benefit from disaggregating ESG components.
Finally, the instrumental variable .
SLS) results provide only weak support for a causal The KleibergenAePaap F-statistic indicates a weak instrument problem, likely driven by the small sample size, and the second-stage coefficients are statistically insignificant for both performance measures.
Therefore.
ESG should not be interpreted as a robust causal driver of financial performance, but rather as a strategic factor whose effects depend on firm characteristics and model specification (McWilliams & Siegel, 2.
In conclusion.
ESG performance does not exhibit a uniformly positive linear impact on either risk-adjusted profitability or market valuation in the Malaysian banking sector.
Instead, its effects are better characterized as nonlinear and conditional.
These findings suggest that ESG should be implemented strategically, with attention to optimal levels of engagement and firmspecific conditions.
For regulators and investors, the results highlight that ESG signals are not uniformly priced, particularly within MalaysiaAos evolving regulatory environment.
Future research should extend the sample period and further explore the role of individual ESG pillars, while accounting for structural shocks such as the COVID-19 pandemic.
References