The Role of Firm Attributes in Shaping Value Relevance: Evidence from Saudi Arabia

Abstract

This study examines the moderating effect of firm attributes on the value relevance of accounting information in Saudi Arabia. Using a sample of 630 firm-year observations from 126 Saudi listed firms over 2018–2022, the research evaluates whether audit quality, size, leverage, growth potential, board diversity, and profitability complement the valuation role of earnings per share (EPS) and book value per share (BVPS) and if so then which direction of the attribute gave greater value relevance. Results reveal that all the firm attributes tested have a significant moderating effect on value relevance. Lower leverage, higher growth potential, greater board diversity, and profitability all lead to higher predicted market value for given EPS and BVPS. Big 4 audit quality and larger firm size are found to moderate the value relevance of accounting information rather than to influence share price directly. Both attributes strengthen the value relevance of earnings per share (EPS)—the EPS coefficient is significantly higher for firms audited by a Big 4 firm and for larger firms—while weakening the value relevance of book value per share (BVPS), with the BVPS coefficient being significantly lower in both cases. The combined effect is that earnings carry greater pricing weight, and book values carry lesser pricing weight, when audit quality is high and when firms are larger. Results also reveal that cohorts with Big 4 auditor, larger size, lower leverage, higher growth potential, more diverse boards, and profitability all have greater value relevance (higher R²) than cohorts with the alternative for each attribute. Hence, tests provide evidence that these attributes strengthen the association between selective accounting figures (EPS and BVPS) and share prices. The findings contribute to agency, information asymmetry, and value-relevance theory by showing that firm attributes condition the EPS and BVPS pricing weights rather than affecting price directly. The results have implications for regulators and firms seeking to improve financial reporting credibility and usefulness amid concentrated ownership. This study contributes timely empirical evidence on the multifaceted drivers of value relevance in an under-researched Middle Eastern emerging market.

1. Introduction

Financial reporting reduces information asymmetry between insiders and outside investors, a function grounded in agency theory (Jensen & Meckling, 1976), information-asymmetry theory (Akerlof, 1970; Healy & Palepu, 2001), and signalling theory (Ross, 1977; Spence, 1973). The degree to which accounting numbers are reflected in equity prices—termed value relevance—is therefore a central indicator of the informational quality of financial statements (Barth et al., 2023). However, the strength of this association is not invariant: it is conditioned by the credibility of the reporting process, the fundamentals of the reporting firm, and the broader information environment in which investors interpret disclosures (Leuz & Wysocki, 2016).

These contingencies are particularly salient in emerging markets, where concentrated ownership, weaker enforcement, and limited analyst coverage amplify reliance on firm-level signals (Hearn, 2011; Chehade & Procházka, 2024). Saudi Arabia represents a theoretically meaningful setting in this respect. Following mandatory IFRS adoption in 2017 and the governance reforms embedded in Vision 2030, the Kingdom combines a rapidly modernising regulatory architecture with persistent features—family-controlled ownership, an evolving audit market, and limited board diversity—that intensify the role of firm attributes in shaping investor inferences (Alomair et al., 2022; Al-Shattarat, 2021). Despite this, evidence on which firm-level factors strengthen or weaken the price–accounting relationship in Saudi Arabia remains fragmented.

Prior international research has typically examined determinants of value relevance in isolation—most often auditor type, firm size, or leverage—and predominantly within developed markets (Collins et al., 1997; Barth et al., 2023; Vander Bauwhede & Van Cauwenberge, 2022). Saudi-focused studies have confirmed the joint relevance of earnings and book values post-IFRS (Al-Shattarat, 2021; Alomair et al., 2022) but have not integrated governance, fundamental, and informational contingencies within a unified framework. This study addresses that gap.

Building on the theoretical anchors above, the firm attributes examined are organised into two conceptual categories. Governance attributes—audit quality and board gender diversity—operate through monitoring and credibility-enhancing channels (Adams & Ferreira, 2009; DeFond & Zhang, 2014). Financial-fundamental attributes—firm size, leverage, profitability, and growth—condition the persistence and reliability of reported numbers (Collins et al., 1997; Hayn, 1995; Frank, 2002).

The study pursues two objectives. First, it establishes the baseline value relevance of earnings per share (EPS) and book value per share (BVPS) in Saudi Arabia’s post-IFRS environment. Second, it investigates whether and how firm attributes moderate the relationship between accounting information and share prices.

The study makes three contributions. Theoretically, it advances contingent value-relevance research by integrating governance, fundamental, and information-environment perspectives within a single framework anchored in agency, signalling, and information-asymmetry theories, thereby moving beyond the single-attribute focus of prior work. Empirically, it offers post-IFRS evidence from a strategically important yet under-researched Middle Eastern market and helps to reconcile inconsistencies in prior findings (e.g., Frank, 2002; Acaranupong, 2021; Dimitropoulos & Koronios, 2021). Contextually, the focus on Saudi Arabia is theoretically motivated rather than merely descriptive; the coexistence of recent IFRS adoption, Vision 2030 governance reforms, concentrated ownership, and a developing audit market provides a setting in which the credibility and signalling channels postulated by theory are likely to be most observable.

The remainder of the paper is structured as follows. Section 2 develops a unified theoretical framework and the hypotheses. Section 3 outlines the methodology. Section 4 presents the results and discussion. Section 5 concludes.

2. Literature Review and Hypothesis Development

This section proceeds in three parts. Section 2.1 sets out the conceptual foundations of value relevance. Section 2.2 develops a unified theoretical framework drawing on agency, information-asymmetry, and signalling theories, which is applied consistently across all attributes examined in the study. Section 2.3 synthesises the empirical evidence and develops the six hypotheses, each explicitly mapped to the theoretical channels identified in Section 2.2.

2.1. Conceptual Foundations of Value Relevance

Value relevance is defined as the statistical association between accounting numbers and equity market values, conditional on those numbers reflecting information that investors use in pricing securities (Barth et al., 2008, 2023; Francis & Schipper, 1999). The construct rests on the premise that if accounting information is decision-useful, it should be reflected in the market’s valuation of the firm. Following Ball and Brown (1968), value relevance has become a central empirical lens through which the informational quality of financial reporting is assessed, and a key benchmark used by standard-setters and regulators to evaluate the consequences of accounting regimes (Barth et al., 2023).

Operationally, value relevance has been examined under two complementary specifications: the returns model, which links unexpected earnings to abnormal returns (Ball & Brown, 1968), and the price model developed by Ohlson (1995), in which share price is expressed as a linear function of book value of equity and earnings. The price-model specification is preferred when the research question concerns the level of association between accounting numbers and equity values rather than information arrival timing and is therefore adopted in the present study (see Section 3).

A substantial empirical literature confirms that earnings and book values are jointly value-relevant across both developed and emerging markets (Collins et al., 1997; Khader & Shanak, 2023; Alomair et al., 2022). However, this association is far from uniform. Its strength varies systematically across firms, time periods, and institutional environments (Barth et al., 2023; Imhanzenobe, 2022). The accounting information set is therefore best understood as conditionally informative, with the degree of condition depending upon characteristics of the reporting firm, the credibility of its reporting process, and the broader information environment in which investors interpret its disclosures. Identifying the firm-level moderators of this conditional informativeness is the core task of contingent value-relevance research and is the focus of this study.

2.2. Unified Theoretical Framework

Three complementary theories explain why and how firm attributes alter the price–accounting relationship. Although each emphasises a distinct mechanism, they share a common premise: investors rationally adjust the weight they place on accounting numbers in proportion to those numbers’ credibility and informativeness. Together they form the unified analytical framework adopted in this study.

Agency theory (Jensen & Meckling, 1976; Fama & Jensen, 1983) holds that the separation of ownership and control creates incentives for managerial opportunism—including earnings management, value-destroying investment, and selective disclosure—that financial reporting can either mitigate or obscure. Mechanisms that strengthen monitoring of managerial behaviour, such as high-quality external auditing and effective board oversight, reduce agency costs and improve the credibility of reported numbers. Investors should therefore place greater weight on accounting information produced under stronger monitoring regimes.

Information-asymmetry theory (Akerlof, 1970; Healy & Palepu, 2001) emphasises that disparities in information between insiders and outsiders generate adverse-selection costs that depress equity values and increase the cost of capital. Disclosure mitigates these costs, but only to the extent that outsiders perceive the disclosed information as reliable and informative about firm fundamentals. Firm characteristics that lower the cost of acquiring or verifying accounting information—such as larger size (which attracts analyst and media coverage) or sustained profitability (which reduces uncertainty about earnings persistence)—should therefore amplify the value relevance of reported numbers.

Signalling theory (Spence, 1973; Ross, 1977) posits that, in environments of asymmetric information, informed parties convey their type to less-informed parties through costly, credible signals. Firm attributes that constitute or substantiate such signals—for example, the voluntary appointment of a Big 4 auditor, the choice to maintain board gender diversity, or the pursuit of growth strategies that require external financing—function as commitment devices that reinforce the informativeness of the accompanying accounting disclosures.

These three lenses are complementary rather than competing. Agency theory speaks to the monitoring channel, information-asymmetry theory to the asymmetry-reduction channel, and signalling theory to the credibility channel. Each firm attribute examined in this study operates through one or more of these channels, as set out in

Table 1. Mapping of firm attributes to theoretical channels.

Attribute	Category	Primary Theoretical Channel	Secondary Channel
Audit quality	Governance	Monitoring (Agency)	Credibility (Signalling)
Board gender diversity	Governance	Monitoring (Agency)	Credibility (Signalling)
Firm size	Fundamentals	Asymmetry reduction	Monitoring (Agency)
Firm leverage	Fundamentals	Asymmetry reduction (negative)	Monitoring (Agency)
Profitability	Fundamentals	Asymmetry reduction	Credibility (Signalling)
Firm growth	Fundamentals	Credibility (Signalling)	Asymmetry reduction

below and applied consistently throughout Section 2.3.

Within this framework, the six attributes are organised into two conceptual categories:

Governance attributes—audit quality and board gender diversity—which operate primarily through the monitoring (agency) and credibility (signalling) channels.

Financial-fundamental attributes—firm size, leverage, profitability, and growth—which operate primarily through the asymmetry-reduction (information-asymmetry) channel by conditioning the persistence and reliability of reported numbers, with secondary signalling effects.

2.3. Empirical Evidence and Hypotheses

2.3.1. Audit Quality

External auditing is the cornerstone monitoring mechanism in modern corporate governance and is central to mitigating both agency conflicts and information asymmetry (DeFond & Zhang, 2014; Knechel et al., 2013). Big 4 auditors are widely associated with superior technical expertise, greater independence, and stronger reputational incentives that raise the cost of issuing an unqualified opinion on misleading statements (Azizkhani et al., 2023). The voluntary engagement of a Big 4 auditor therefore functions as a credible signal of reporting quality (Ross, 1977), while the audit process itself constitutes the monitoring mechanism predicted by agency theory.

Empirically, the value-relevance literature consistently reports stronger price–accounting associations for firms audited by Big 4 firms across both developed and emerging markets (Alali & Foote, 2012; H. L. Lee & Lee, 2013; Al-Hiyari, 2024; Chehade & Procházka, 2024). Evidence from the Middle East and North Africa region similarly indicates that audit quality enhances the credibility of accounting numbers in settings characterised by concentrated ownership and developing audit markets (Alomair et al., 2022). On this basis:

H1. 

Big 4 auditing positively moderates the relationship between accounting information and share prices.

2.3.2. Firm Size

The literature offers two competing arguments regarding the moderating role of firm size. The “competing-information” view holds that larger firms generate richer non-accounting information through analyst coverage, media scrutiny, and voluntary disclosure, which reduces investors’ marginal reliance on financial statements (Vander Bauwhede & Van Cauwenberge, 2022). The “earnings-persistence” view, grounded in agency and information-asymmetry theory, argues instead that larger firms have more diversified operations, more persistent earnings, more developed internal controls, and more stable accounting estimates, all of which enhance the informativeness of their reported numbers (Collins et al., 1997; Dang et al., 2018).

The empirical evidence is mixed (Collins et al., 1997; Alali & Foote, 2012; Brimble & Hodgson, 2007), but a meaningful pattern emerges: in emerging markets, where alternative information channels are sparse and analyst coverage is limited, the asymmetry-reduction channel dominates, and accounting numbers carry greater weight for larger firms (Hearn, 2011; Alomair et al., 2022). Given that the Saudi market exhibits these features, we reformulate the previously non-directional hypothesis as directional:

H2. 

Firm size positively moderates the relationship between accounting information and share prices.

2.3.3. Firm Leverage

Leverage is theoretically associated with both heightened agency conflicts and increased information asymmetry. From an agency perspective, highly leveraged firms face greater incentives to engage in earnings management to avoid debt-covenant violations and to preserve access to external financing (Acaranupong, 2021; Dimitropoulos & Koronios, 2021). From an information-asymmetry perspective, leverage is associated with greater perceived bankruptcy risk, more volatile residual cash flows to equity, and a less stable mapping between accounting numbers and equity values (Acaranupong, 2021). Although creditor monitoring may, in principle, improve disclosure quality, this counter-effect is typically weaker in markets where bond financing is limited, and creditor enforcement institutions are still developing—features characteristic of the Saudi setting. The empirical balance therefore favours a negative moderating effect (Dimitropoulos & Koronios, 2021; Chehade & Procházka, 2024). Accordingly:

H3. 

Firm leverage negatively moderates the relationship between accounting information and share prices.

2.3.4. Firm Growth

Signalling theory (Ross, 1977; Spence, 1973) provides a coherent rationale for growth-related variation in value relevance. Firms with strong growth opportunities are typically dependent on external capital, and have powerful incentives to provide credible, informative accounting signals to attract financing on favourable terms (Putri & Panggabean, 2020). Investors, in turn, rely heavily on book values and current earnings as anchors for projecting future profitability when forward-looking opportunities are abundant.

The competing argument advanced by Frank (2002)—that earnings of high-growth firms are more transitory and therefore less relevant—has received only limited support in subsequent emerging-market studies, which find that growth opportunities raise rather than diminish the weight investors place on reported numbers (Putri & Panggabean, 2020; Rennekamp, 2012). On the weight of evidence, and consistent with signalling theory, we reformulate the previously non-directional hypothesis as directional:

H4. 

Firm growth positively moderates the relationship between accounting information and share prices.

2.3.5. Board Gender Diversity

Board gender diversity strengthens the monitoring function predicted by agency theory. Empirical evidence indicates that female directors attend board meetings more regularly, are more likely to sit on monitoring committees, and exert stronger oversight over management (Adams & Ferreira, 2009). Diverse boards have been linked to higher earnings quality (Ammer & Ahmad-Zaluki, 2017), reduced earnings management (Gul et al., 2011), reduced tax aggressiveness (Mungai et al., 2020), and ultimately higher value relevance of reported numbers (Cimini, 2022; Wang et al., 2022; Al-Hiyari, 2024). The presence of female directors may also operate as a credible signal of governance commitment, particularly in settings such as Saudi Arabia where female board representation has historically been uncommon and is now rising under Vision 2030 reforms.

The earlier draft framed this hypothesis under institutional theory; we have removed that framing because the empirical mechanism documented in the literature is monitoring-based, not isomorphism-based. Accordingly:

H5. 

Board gender diversity positively moderates the relationship between accounting information and share prices.

2.3.6. Profitability

The information content of earnings depends on their persistence (Hayn, 1995; Collins et al., 1997). Profitable firms produce earnings that are more persistent and more predictive of future cash flows, while losses are typically perceived as transitory because shareholders hold a liquidation option that bounds downside exposure (Hayn, 1995). From an information-asymmetry perspective, this reduces the informativeness of reported earnings for loss firms; from a signalling perspective, sustained profitability functions as a credible signal of reporting reliability and operational stability.

The empirical literature consistently reports higher earnings response coefficients and higher value relevance for profitable firms (Hayn, 1995; Roychowdhury et al., 2019; Al-Hiyari, 2024). On this basis:

H6. 

Profitability positively moderates the relationship between accounting information and share prices.

3. Research Design

This research utilises a quantitative methodology to examine the moderating role of firm attributes in shaping the value relevance of accounting information in Saudi Arabia. In particular, panel regressions evaluate the interactions between earnings, book values, and firm-specific variables like size, leverage, and share prices within the Saudi context.

3.1. Sample Selection and Data Collection

This research examines a 5-year period from 2018 to 2022. This period starts after the implementation of IFRS in Saudi Arabia in 2017, in order to mitigate the influence of the transition to IFRS on the results. Prior studies, such as Alomair et al. (2022), which examined periods before IFRS (2015–2016) and after IFRS (2017–2018) in the Saudi market, indicate that accounting information is jointly value relevant post IFRS. By focusing on the period from 2018 to 2022, this study aims to provide a better understanding of the influence of firm-specific attributes on value relevance in the post-IFRS period.

The sample includes 126 non-financial Saudi firms continuously listed during 2018–2022, providing a balanced panel of 630 firm-year observations. Financial firms were excluded due to their distinct governance norms. Firms with non-calendar fiscal years were also omitted to ensure consistent comparisons (Al-Abbas, 2009). The balanced panel controls for time-invariant firm attributes, enhancing internal validity (Yildirim, 2021).

Data on firm attributes and financial variables was gathered from the Saudi Stock Exchange and Bloomberg databases. This recent dataset allows assessing associations between newly disclosed firm attributes, accounting information, and market value in the contemporary Saudi context.

3.2. Empirical Analysis

This research employs a two stage panel data regression analysis which combines two approaches found in prior literature to assess the contingent role of firm attributes in shaping accounting information relevance (Badu & Assabil, 2022). The analysis expresses a firm’s market value as a function of book value of equity and earnings per share, consistent with valuation theory (Ohlson, 1995). The panel structure allows analysing behavioural differences across firms and time periods using a general linear model approach (Barth et al., 2023). The basic model is:

$$P_{it}=a_0+b_1{BVPS}_{it}+b_2{EPS}_{it}+e_{it}$$

(1)

where $P_{it}$ is the share price of firm i; ${BVPS}_{it}$ is the reported book value of equity per share for firm i in year t; ${EPS}_{it}$ is reported earnings per share for firm i in year t; and $e_{it}$ is the residual term.

The first stage of the analysis is to examine the moderating influence of firm attributes, by incorporating interaction terms between accounting information and firm attributes (Dang et al., 2018). The basic model is:

$$P_{it}=a_0+b_1{BVPS}_{it}+b_2{EPS}_{it}+b_3{FATTR}_{it} {+{b_{4}BVPS}_{it}\ast {FATTR}_{it}+b_5{EPS}_{it}\ast {FATTR}_{it}+e}_{it}$$

(2)

where ${FATTR}_{it}$ represents the relevant moderating firm attribute variable for firm i in year t.

The interaction coefficients $b_4$ and $b_5$ are of central interest, indicating whether specific firm features meaningfully alter the relationship between accounting figures and market value. If the interactions are statistically significant, it indicates that differentiating the relationship between accounting information and market value based upon the firm attribute adds value to the analysis.

The second stage of the analysis is performed whenever the first stage reveals that utilising knowledge of a firm’s attributes can significantly impact market value prediction. The second stage tests which direction (category) of the attribute gives greater value relevance for the accounting information. Which direction (category) of the attribute has greater value relevance is determined by looking at the R² values for how well BVPS and EPS predict share price, based upon the estimate of Equation (2), for the two subsets of the data divided by the categories of the attribute.

The analysis will be performed six times with the firm variables—audit quality, size, leverage, growth, board diversity, and profitability to analyse the distinct moderating impact and direction of impact on value relevance of each attribute. This approach permits examining how different firm attributes shape the value relevance of financial information based on the interaction effects and R²’s. Through focusing on the contingent role of firm-specific factors, the study can assess the hypotheses on how various corporate attributes impact the relevance of accounting data in the Saudi context.

All regressions were run in SPSS (version 26) using the General Estimating Equations approach to deal with the panel data and setting the covariance matrix type to Exchangeable to deal with possible autocorrelation. No fixed effects were included so as to maintain consistency with the seminal Collins et al. (1997) approach widely used across the existing literature.

3.2.1. Measurements of Variables

The dependent variable is the firm’s market value per share.

• Market value per share is measured through a firm’s share price (P), measured as share price four months after the financial year end, following Barth et al. (2023) and Ohlson (1995).
• The main independent variables are firm-specific accounting numbers:
• Book value of equity per share (BVPS)—the reported book value of equity divided by the number of outstanding shares, following Badu and Assabil (2022), Barth et al. (2023), B. Chen et al. (2020), and Ohlson (1995).
• Earnings per share (EPS)—the reported earnings per share as used in Barth et al. (2023), Jara et al. (2019), Kwon and Wang (2020), and Ohlson (1995).
• The moderating variables are firm’s attributes:
• Audit quality (BIG4) is a binary variable based on the size of the external auditors (Chehade & Procházka, 2024), where firms audited by big 4 auditors (i.e., PricewaterhouseCoopers, Deloitte, Ernst & Young, and KPMG) take 1, and 0 otherwise.
• Firm size (FSIZE) is a binary variable based on the median of a firm’s total assets (Collins et al., 1997), and for a robustness test total sales (Vander Bauwhede & Van Cauwenberge, 2022), where large firms take 1 and small firms take 0.
• Firm leverage (FLEV) is a binary variable based on the median of firm’s debt-to-assets ratio (Dimitropoulos & Koronios, 2021), and for a robustness test debt-to-equity ratio (Acaranupong, 2021; Brimble & Hodgson, 2007), where high leverage firms take 1 and low leverage firms take 0.
• Firm growth potential (FGRTH) is a binary variable based on the median of firm’s market-to-book ratio (Rennekamp, 2012), where high growth potential firms take 1 and low growth potential firms take 0.
• Board diversity (FBDIV) is a binary variable based on the presence of at least one female director (Cimini, 2022), where firms with gender-diverse boards take 1, and 0 otherwise.
• Profitability (ESIGN) is a binary variable based on the sign of earnings (Collins et al., 1997; Hayn, 1995), where profit-making firms take 1 and loss-making firms take 0.

While it is common practice in many studies to represent some of these variables by continuous variables this cannot be done in this study. Since the basis of comparing value relevance for different values of a firm attribute is to compare the R² of different subsets of the data split based upon levels/categories of the firm characteristic (stage 2 of the analysis) this can only be done if the attribute is measured on a categorical scale.

3.2.2. Regression Assumptions Assessment

The valuation models in this study assume linear relationships between share price and the explanatory variables such as earnings and book value, making it essential to assess standard regression assumptions. With 630 firm-year observations, the sample exceeds thresholds for statistical power (Cohen et al., 2013). Linearity and homoscedasticity were confirmed via scatterplots, and autocorrelation is mitigated by the 5-year balanced panel (Petersen, 2008). Outliers were addressed through winsorization (Barth et al., 2023), and normality is assumed based on the central limit theorem. Multicollinearity was assessed using VIFs (O’brien, 2007). Results indicated that for some firm attributes the results of Equation (2) using all variables had multicollinearity issues as a result of the correlation between the dummy and the cross-product term dummy × BVPS. To address this, two alternative models for Equation (2) were run for each firm attribute; one dropping the dummy and one dropping the cross-product term. In all cases this produced results with acceptable VIF values. Overall, regression assumptions were reasonably met or addressed, supporting the robustness of the analysis and the validity of inferences drawn.

3.3. Robustness Tests

To evaluate the reliability of the findings, this study conducts robustness checks by re-estimating the models using alternative measurements for two firm attributes: size and leverage. Size is re-measured using total sales rather than total assets (Vander Bauwhede & Van Cauwenberge, 2022), while leverage is re-measured using the debt-to-equity ratio rather than total debt to total assets (Acaranupong, 2021; Dimitropoulos & Koronios, 2021). Testing the sensitivity of results to different variable proxies enhances the rigour and validity of the inferences by providing a specification check on the original models. Comparing the outcomes across alternative size and leverage metrics allows assessing the consistency of the findings on how firm attributes moderate the relevance of accounting information within the Saudi context (Petersen, 2008).

4. Analysis and Discussion

• Descriptive statistics

As shown in

Table 2. (a) Descriptive statistics for continuous variables. (b) Descriptive statistics for binary variables.

(a)
Variable	Minimum	Maximum	Mean	Std. Deviation
P	1.03	133.40	31.27	23.20
BVPS	−0.64	46.00	14.68	6.65
EPS	−8.00	11.00	0.55	2.04
TOTAL ASSETS (Billion Riyal)	0.02	485.45	13.20	52.21
TOTAL SALES (Billion Riyal)	0.00	198.47	4.66	16.93
TOTAL DEBT-COMMON EQUITY RATIO	−2291.46	1702.28	82.66	174.65
TOTAL DEBT-TOTAL ASSETS RATIO	0.00	80.93	25.34	19.78
MARKET-TO-BOOK RATIO	−19.00	117.25	2.91	5.67
EARNINGS (Million Riyal)	−3.78	23.07	0.38	1.89
RETURN (%)	−96.28	459.22	18.72	63.28
(b)
Variable	0		1
BIG4	357 (56.7%)		273 (43.3%)
FBDIV	569 (90.3%)		61 (9.7%)
ESIGN	202 (32.1%)		428 (67.9%)

a,b, the sample is well-distributed across various attributes. One notable exception is the very low percentage of gender-diverse boards, with only 9.68% of the sample having at least one female director, highlighting the limited gender diversity on Saudi boards.

One concern when analysing the attributes individually is possible relationships between them. This is addressed in

Table 3. Pattern among firms’ attributes.

Firms’ Attributes		Firm’s Size		Firm’s Leverage		Firm’s Potential Growth		Board Gender Diversity		Firm’s Profitability
		Small	Large	Low	High	Low	High	Only Male	Diverse	Loss	Profit
Audit quality	Non-big4	184	173	193	164	183	174	324	33	130	227
	Big4	131	142	122	151	132	141	245	28	72	201
Firm’s size	Small			158	157	194	121	291	24	104	211
	Large			157	158	121	194	278	37	98	217
Firm’s Leverage	Low					159	156	276	39	88	227
	High					156	159	293	22	114	201
Firm’s Potential Growth	Low							298	17	116	199
	High							271	44	86	229
Board Gender Diversity	Only Male									193	376
	Diverse									9	52

and

Table 4. Correlation between variables.

	P	BVPS	EPS	BIG4	FSIZE1	FSIZE2	FLEV1	FLEV2	FGRTH	FBDIV	ESIGN
P	1.000
BVPS	0.489	1.000
EPS	0.541	0.433	1.000
BIG4	0.009	0.038	0.071	1.000
FSIZE1	0.082	0.174	0.271	−0.019	1.000
FSIZE2	0.029	−0.001	−0.015	−0.006	0.011	1.000
FLEV1	−0.142	−0.234	−0.204	−0.016	0.251	0.044	1.000
FLEV2	−0.078	−0.210	−0.193	−0.003	0.238	0.017	0.867	1.000
FGRTH	0.233	−0.124	0.033	0.017	−0.152	−0.082	−0.037	−0.002	1.000
FBDIV	0.188	0.090	0.181	0.017	0.066	0.122	−0.091	−0.048	−0.011	1.000
ESIGN	0.290	0.313	0.645	0.038	0.259	0.058	−0.088	−0.109	−0.041	0.121	1.000

. Table 3 shows the pattern among firm attributes while Table 4 shows the correlations between the variables used in the models. There are no correlations amongst the explanatory variables that raise serious concerns about the independence of the attributes.

• The overall value relevance

The baseline price regression results in

Table 5. Results for GLM analysis with full model and two reduced models for each firm attribute.

Attribute	Intercept	BVPS	EPS	Dummy	BVPS × Dummy	EPS × Dummy	R2 (%)	Max VIF	R2 (%) Dummy = 0	R2 (%) Dummy = 1
Baseline	12.677 **	1.094 **	4.607 **				37.3	1.231
Audit	9.602 **	1.374 **	3.815 **	6.699	−0.626 *	1.802 *	38.2	8.564	37.0	39.5
Audit	12.615 **	1.192 **	4.001 **		−0.225 *	1.373	37.9	2.372
Audit	13.562 **	1.090 **	4.186 **	−2.000		0.913	37.5	2.194
Size	14.477 **	1.108 **	1.545 *	−3.328	−0.106	4.738 **	41.0	9.440	14.9	52.3
Size	12.603 **	1.221 **	1.423 *		−0.305 **	4.952 **	41.0	4.070
Size	15.289 **	1.047 **	1.597 *	−4.764 **		4.617 **	41.0	4.010
Leverage	11.842 **	1.018 **	6.413 **	5.452	−0.233	−3.428 **	39.7	7.423	49.1	16.4
Leverage	14.134 **	0.889 **	6.595 **		0.108	−3.693 **	39.5	2.709
Leverage	13.028 **	0.935 **	6.586 **	2.306		−3.669 **	39.6	2.535
Growth	3.510	1.231 **	1.561 **	13.348 **	0.112	3.144 **	51.0	8.538	38.9	43.9
Growth	10.948 **	0.798 **	2.083 **		0.913 **	2.224 **	49.7	3.618
Growth	2.667	1.288 **	1.480 *	14.907 **		3.309 **	51.0	3.372
Board Diversity	11.947 **	1.119 **	3.938 **	9.482	−0.819	6.935 **	40.0	13.540	34.4	58.4
Board Diversity	12.530 **	1.084 **	3.976 **		−0.244	6.364 **	39.9	2.550
Board Diversity	12.621 **	1.071 **	4.004 **	−1.629		5.498 **	39.8	1.897
Loss/Profit	12.104 **	0.565 **	−2.059 **	−2.797	0.130	12.657 **	53.7	12.664	10.2	56.0
Loss/Profit	10.294 **	0.966 **	−2.191 **		−0.038	14.782 **	53.6	4.448
Loss/Profit	11.056 **	0.657 **	−2.045 **	−1.246		12.724 **	53.7	4.706

* and ** indicate significance at the 0.05 and 0.01 levels respectively.

provide evidence that both earnings per share (EPS) and book value per share (BVPS) are positively and significantly related to share prices for the 630 Saudi firms from 2018 to 2022. The model explains 37.3% of the price variation based on the adjusted R-squared, which indicates reasonably good explanatory power in the Saudi context. The EPS and BVPS coefficients are 4.607 and 1.094, respectively, and both significant at the 1% level. This confirms the value relevance of both earnings and book values in Saudi Arabia, aligning with extensive prior research by Alomair et al. (2022), Khader and Shanak (2023), and Alfaraih and Alanezi (2011).

These findings align with recent Saudi studies, including Al-Shattarat (2021) and Alomair et al. (2022), who reported adjusted R² values of around 40% pre-IFRS and over 50% post-IFRS, reinforcing the current study’s relevance.

Globally, these findings align with evidence from developed markets—Filip et al. (2021) in the US, and Almaqtari et al. (2022) in the Gulf—both reporting significant links between earnings, book values, and stock prices. Despite being an emerging economy, Saudi Arabia shows comparable value relevance, affirming the usefulness of accounting information for investors. These results provide updated post-IFRS evidence, indicating that IFRS adoption has preserved financial statement relevance in Saudi Arabia, consistent with H. L. Lee and Lee (2013) and Imhanzenobe (2022). Overall, the baseline results confirm the valuation role of accounting information in a context where empirical research remains limited.

• The moderating effect of firm attributes

Table 5 presents regression results testing the moderating effects of six different firm attributes on the value relevance of accounting information. Each attribute is examined through a separate model of Equation (2) with the variables interacting to assess contingent relationships. For all attributes the initial estimate of Equation (2) led to relatively high levels of multicollinearity (maximum VIF > 5) and in all cases the major cause of this was the correlation between the dummy variable and the cross-product variable with BVPS. To assess whether the insignificance of these variables in some models was a result of multicollinearity all models were rerun first excluding the dummy variable and then excluding the BVPS cross-product variable. This additional analysis does reveal that in some models initial insignificance of these variables was a result of multicollinearity. For the initial estimates of model 2 for each attribute the forecast results are also divided into two groups corresponding to the 0 and 1 values of the dummy variable with an R² value calculated for each group. The series of models allows comprehensive analysis of whether knowing different firm-specific attributes can improve the usefulness of accounting information in explaining market valuations (assessed by the significance of the dummy or cross-product variables). Examination of the R² values associated with the categories of the dummy variable can also give an assessment of which category of the attribute makes the accounting information most value relevant.

4.1. Audit Quality

The R² of 38.2% represents only a slight improvement over the baseline model. The interaction BVPS*BIG4 is negative and the interaction EPS*BIG4 is positive with both being significant at the 5% level. This supports that there is value in differentiating between Big 4 and non-Big 4 audited firms when predicting market value using a firm’s book value and earnings with the R² values for the two groups indicating that the accounting information has greater value relevance for firms using Big 4 auditors. However, it does mean that the direction of impact on market value of having a Big 4 auditor is dependent upon the values of EPS and BVPS. This is consistent with (Visedsun et al., 2025), which shows its impact is dependent upon the value of EPS. Overall, these results support hypothesis H1 that Big 4 auditors enhance the value relevance of financial information.

These findings support prior research that indicates higher value relevance with Big 4 auditors (Alali & Foote, 2012; H. L. Lee & Lee, 2013) or value in utilising type of auditor to predict market value or returns (Al-Hiyari, 2024; J. Chen et al., 2001; C. Lee & Park, 2013; and Siekkinen, 2017). Big 4 auditors provide greater assurance of reliability and accuracy due to their advanced technical capabilities, independence, and reputation (DeFond & Zhang, 2014), supporting agency theory’s link between higher audit quality and reduced information asymmetry.

Critically, these results are particularly relevant in the post-IFRS adoption period after the switch from Saudi GAAP in 2017. The continued positive moderating effect of Big 4 auditors in the post-reform period aligns with studies by Imhanzenobe (2022) and H. L. Lee and Lee (2013), which attribute Big 4 firms’ substantial expertise and experience with IFRS implementation globally to their greater capabilities under IFRS standards relative to non-Big 4 auditors. This implies that Big 4 audits help mitigate the uncertainties and valuation complexities arising from major accounting changes like IFRS adoption.

4.2. Firm Size

The R² of 41.0% represents a moderate improvement over the baseline model. The interaction EPS*FSIZE is positive and highly significant (1% level). FSIZE or the interaction BVPS*FSIZE are negative and highly significant (1% level) once the multicollinearity between them is removed. This supports that there is value in differentiating between smaller and larger firms when predicting market value using a firm’s book value and earnings with the R² values for the two groups indicating that the accounting information has substantially greater value relevance for larger firms. However, it does mean that the direction of impact on market value of being a larger firm is dependent upon the values of EPS and BVPS. Overall, these results support hypothesis H2 that firm size moderates the relationship between accounting information and value relevance.

The finding of the positive impact of company size on value relevance aligns with the findings of Alali and Foote (2012), although only for market value, and Brimble and Hodgson (2007), although only for the earlier years in their study. Contrasting with this are negative impact findings of Alali and Foote (2012), for returns, Ater et al. (2025), Brimble and Hodgson (2007), for the later years in their study, and Collins et al. (1997). The significant positive moderation of EPS impact aligns with studies arguing that larger firms have more persistent earnings that attract greater investor emphasis (Collins et al., 1997) and with results of prior research (J. Chen et al., 2001; Habib & Azim, 2007; and Habib & Weil, 2008). With respect to the significant negative moderation of BVPS there are differing results of prior research with J. Chen et al. (2001) having consistent results while the results of Habib and Azim (2007), with a positive effect, and Habib and Weil (2008), not significant, are in conflict. A key point of difference with much of the prior research that relates company size to value relevance is that many prior articles use company size as a control variable using a continuous variable form such as natural log of total assets. This is a substantial difference to the binary measure of company size used in this article, due to the desire to also consider the split sample approach, that could explain the difference in findings compared with some past research. The significantly higher R² for the larger firm group supports the claim that accounting information better predicts future performance for larger firms, as discussed in the literature review.

4.3. Financial Leverage

The R² of 39.7% represents only a slight improvement over the baseline model. The interaction EPS*FLEV is negative and highly significant (1% level). This supports that there is value in differentiating between firms with less versus more leverage when predicting market value using a firm’s book value and earnings with the R² values for the two groups indicating that the accounting information has substantially greater value relevance for firms with lower leverage. It predicts higher market value for firms with less leverage but equal EPS and BVPS. Overall, these results support hypothesis H3 that financial leverage negatively moderates the value relevance of accounting information.

The finding of greater value relevance of firms with lower leverage agrees with the work of Ater et al. (2025), and Chehade and Procházka (2024), which was also based on Saudi Arabian companies. The negative moderating effect of leverage on EPS aligns with prior studies such as Habib and Azim (2007), which argue that higher debt creates greater incentives for earnings management and income smoothing to avoid debt covenant violations. However, it conflicts with the results of Habib and Weil (2008), which finds a positive moderating effect, and C. Lee and Park (2013), which finds no significant moderation. This perspective is rooted in agency theory, which links leverage to information asymmetry and managerial opportunism (Dimitropoulos & Koronios, 2021). Therefore, the diminished relevance of earnings supports the view that leverage-related incentives lower the reliability of earnings information.

However, the insignificant impact of leverage on BVPS contradicts this agency theory explanation, as prior studies show mixed empirical evidence overall.

4.4. Firm Growth Potential

The R² of 51.0% represents a substantial improvement over the baseline model. The interaction EPS*FGRTH is positive and highly significant (1% level). FGRTH or the interaction BVPS*FGRTH are positive and highly significant (1% level) once the multicollinearity between them is removed. This supports that there is value in differentiating between firms with lower versus higher growth potential when predicting market value using a firm’s book value and earnings with the R² values for the two groups indicating that the accounting information has greater value relevance for firms with higher growth potential. It predicts higher market value for firms with higher growth potential but equal EPS and BVPS. Overall, these results support hypothesis H4 that growth opportunities moderate the value relevance of accounting information.

The significant positive moderation of BVPS relevance aligns with studies by Frank (2002) and Rennekamp (2012), which argue that high growth creates greater unpredictability in earnings due to firms’ early life cycle stage. High growth indicates transitory earnings volatility that diminishes emphasis on current period earnings.

However, the insignificant EPS result contradicts this perspective on growth and earnings volatility. As highlighted in the literature review, prior empirical evidence is mixed, with some studies like Rennekamp (2012) finding a positive effect of growth on earnings relevance. The asymmetric effects imply that growth contingently impacts particular accounting information.

4.5. Board Gender Diversity

The R² of 40.0% represents only a slight improvement over the baseline model. The interaction EPS*FBDIV is positive and highly significant at the 1% level. This supports that there is value in differentiating between firms with no gender diversity versus gender diversity on the board when predicting market value using a firm’s book value and earnings with the R² values for the two groups indicating that the accounting information has substantially greater value relevance for firms with gender diversity on the board. It predicts higher market value for firms with more diverse boards but equal EPS and BVPS. Overall, these results support hypothesis H5 that board gender diversity enhances the value relevance of financial information.

The significant positive moderation of EPS relevance aligns with prior studies arguing that diverse boards enhance oversight and governance, improving earnings quality and relevance (Gul et al., 2011; Cimini, 2022). Varied perspectives result in more diligent monitoring and conservative reporting (Adams & Ferreira, 2009). The higher EPS relevance also supports institutional theory, linking diversity to reputational integrity and stakeholder trust (Lucas-Pérez et al., 2015).

However, the insignificant impact of gender diversity on BVPS suggests that diversity may impact accounting information asymmetrically. As highlighted by Cimini (2022), evidence is limited to specific contexts and measures, indicating nuanced, contingent effects of gender diversity. While there were other studies that have found board diversity has a positive impact on market value (Al-Hiyari, 2024; Cimini, 2022; Qureshi et al., 2019) the authors could find no other study that had tested the direction of impact of gender diversity on value relevance.

4.6. Profitability

The R² of 53.7% represents a substantial improvement over the baseline model. The interaction EPS*ESIGN is positive and highly significant (1% level). This supports that there is value in differentiating between firms with negative and positive earnings when predicting market value using a firm’s book value and earnings with the R² values for the two groups indicating that the accounting information has substantially greater value relevance for firms with positive earnings. Overall, these results support hypothesis H6 that profitability enhances the value relevance of accounting information.

The significant positive moderation of EPS relevance aligns with studies arguing that profitability indicates greater earnings persistence and future cash flow predictability (Ater et al., 2025; Hayn, 1995; Collins et al., 1997). Investors emphasise current earnings more for profitable firms when predicting future performance, supporting theoretical predictions of higher earning multiples for profitable firms.

However, the insignificant BVPS result contradicts the view that book value becomes more relevant for loss firms as a proxy for liquidation value, as highlighted by Roychowdhury et al. (2019). The asymmetric effects imply that profitability contingently impacts particular accounting figures. Prior studies have found that profitable firms, compared to loss making firms, have predicted higher market value for the same levels of earnings and book value (Al-Hiyari, 2024; Collins et al., 1999) while Collins et al. (1997), Hayn (1995) and Mashoka (2025) also found greater value relevance for profitable firms, as in this study.

4.7. Robustness Analysis

Robustness tests using alternative measures for firm size and leverage confirmed the associated results. Using total sales (FSIZE2) instead of total assets (FSIZE1) and the debt-to-equity ratio (FLEV2) instead of the debt-to-assets ratio (FLEV1) yielded similar findings. These consistent results across different proxies reinforce the robustness of the conclusions regarding the moderating effects of size and leverage on value relevance in Saudi Arabia.

5. Conclusions

This study examined whether and how six firm attributes—audit quality, size, leverage, growth, board diversity, and profitability—moderate the value relevance of accounting information for Saudi-listed firms. All attributes influenced the use of EPS for prediction, while only some affected the use of BVPS. Positive EPS interactions with audit quality, diversity, and profitability, and a negative interaction with leverage, highlight how governance, performance, and investor perceptions enhance the association between earnings and market value. Notably, combining all attributes increased R² to 66.4%, a substantial improvement over the baseline model.

Beyond these summary results, the findings carry three marginal contributions to the unified theoretical framework set out in Section 2.2. With respect to agency theory, the results show that monitoring mechanisms—Big 4 audit quality and board gender diversity—do not raise share prices directly but instead re-weight the components of accounting information that investors price, strengthening the EPS pricing weight while leaving the BVPS weight unchanged or weakened, with the dummy variables themselves remaining statistically insignificant. This refines the conventional agency-theoretic prediction that governance “raises value”, showing that in the Saudi market the mechanism operates through enhanced credibility of the earnings signal rather than through a level shift in price.

With respect to information-asymmetry theory, the differential moderating effects across firm fundamentals confirm that the same accounting numbers carry different information content under different signal-to-noise conditions: larger firms exhibit a substantially stronger EPS pricing weight (R² rising from 14.9% to 52.3% across size groups), highly leveraged firms exhibit a weaker EPS pricing weight (R² falling from 49.1% to 16.4%), and loss firms exhibit a striking EPS sign reversal accompanied by a sharply elevated BVPS weight, consistent with the abandonment-option interpretation under high information asymmetry (Hayn, 1995). These patterns refine information-asymmetry-based predictions by showing that firm attributes condition not only the magnitude but also the sign of accounting coefficients. Finally, with respect to the Ohlson (1995) framework, the systematic variation in EPS and BVPS coefficients across firm attributes provides empirical support for treating these coefficients as conditional rather than fixed parameters, in line with recent calls in the literature (Barth et al., 2023).

The findings offer timely empirical evidence for Saudi Arabia, showing that firm-level factors can enhance earnings relevance under IFRS. High-quality audits and diverse boards improve oversight, while profitability and positive sentiment signal reliability—thereby reducing information asymmetry in a market characterised by concentrated ownership and limited disclosure. These results support agency theory and add to contingent value relevance literature in emerging markets.

From a policy perspective, the findings translate into specific recommendations aligned with Saudi Vision 2030’s Financial Sector Development Program and the Capital Market Authority (CMA) reform agenda. First, regulators should accelerate the ongoing tightening of audit-oversight rules under SOCPA and consider incentives for engagement of high-quality auditors among smaller listed firms, given the evidence that audit quality strengthens the pricing weight of earnings. Second, the CMA’s board-diversity initiatives should be reinforced through binding minimum-representation thresholds rather than voluntary targets, since gender diversity is shown here to enhance the credibility of the earnings signal. Third, disclosure requirements should be expanded for loss-making and highly leveraged firms—whose earnings are priced least reliably—through mandatory forward-looking and segmental information, complementing Vision 2030’s broader transparency objectives.

For listed companies, the results suggest concrete actions to enhance the valuation usefulness of financial information: engaging Big 4 or comparable-quality auditors, broadening board composition, and providing richer narrative and segmental disclosures—particularly for smaller, highly leveraged, or loss-making firms whose accounting signals are otherwise priced with greater noise. For investors, the evidence cautions against applying uniform EPS/BVPS multiples across the Saudi market and instead supports conditioning valuation models on firm attributes: placing greater weight on EPS for large, well-governed, and profitable firms, and shifting weight toward BVPS for loss-making and highly leveraged firms in line with the abandonment-option interpretation. This attribute-conditioned approach offers a practical route to judging the credibility of accounting information and improving valuation accuracy in an emerging-market setting.

Limitations include the use of binary proxies for complex constructs. Future research should apply more refined measures, such as audit fees, IFRS expertise, or multi-year growth trends, and explore longer post-IFRS periods. Cross-country comparisons could also help identify institutional factors that influence the strength and direction of moderating effects. Despite these limitations, the study advances understanding of how firm attributes shape the usefulness of accounting information for valuation in emerging markets.