Open AccessInvestigating Earnings Management Within Maltese Limited Companies
View Full Article|
Download PDFAbstract:
This study aims (i) to assess the prevalence of Earnings Management among non-financial Maltese Listed Entities; (ii) to explore the underlying motivations and drivers that give rise to such practices; and (iii) to investigate the methods and techniques currently employed by the auditee or auditor to prevent or detect Earnings Management within Maltese Listed Entities. A sequential two-phase explanatory mixed-methods approach was employed: first, the accrual-based model was applied to assess the presence of Earnings Management, followed by 20 semi-structured interviews with Audit Partners and Chief Financial Officers. While Earnings Management sector-specific behaviours were observed, no statistically significant differences in the distribution of Earnings Management across sectors were found, suggesting overall consistency. Despite its presence, Earnings Management remains ambiguous, with diverse interpretations creating opportunities for exploitation. The principles-based nature of IFRS facilitates Earnings Management, allowing subjective judgment to serve managerial interests. Motivations for the practice include company-level capital pressures and contractual obligations, with auditors seen as key deterrents owing to their commitment to professional standards. While current preventative measures are effective, the study calls for stronger scrutiny of management and auditors. It also highlights opportunities for local regulatory bodies to enhance consistency and depth in their approach to addressing complex Earnings Management techniques. Lastly, External Auditors face challenges such as quality gaps between Big 4 and non-Big 4 firms, and client resistance during efforts to detect Earnings Management. The study has sought to understand the Earnings Management phenomenon within the Maltese context, given its negative implications on Financial Reporting.1. Introduction
The Conceptual Framework for Financial Reporting states that financial statements (FS) support users in making economic decisions related to investing, lending, or governance oversight (IFRS Foundation, 2018). International Financial Reporting Standards (IFRS) adopt a principles-based approach that relies on professional judgement to represent economic substance. However, this same flexibility creates opportunities for manipulation, allowing Earnings Management (EM) through subjective accounting choices (Callao & Jarne, 2010; Jeanjean & Stolowy, 2008; Toumeh & Yahya, 2019).
EM has been central to major corporate scandals, including Enron, WorldCom, and Arthur Andersen (Callao et al., 2021; Rani et al., 2013). Despite extensive academic debate, its ethical boundaries remain contested, with no universally accepted definition (Beneish, 2001; Dechow & Skinner, 2000; Franceschetti, 2018; Kamau & Murori, 2024; Ronen & Yaari, 2008) EM is a point of contention because it undermines the reliability of reported performance, weakens investor confidence, and threatens the credibility of financial reporting (Bin Khidmat et al., 2018; Burlacu et al., 2024; Goel, 2016). As Turner (2006) notes on page 384, a common feature of such scandals is “a lack of transparency in financial reporting, resulting in numbers being reported to investors that do not reflect economic reality.”
In Malta, research on EM is limited and largely outdated. Earlier studies report mixed findings: Cardona (2003) identifies high EM levels, whereas Mercieca (2004) finds stronger earnings quality. These inconsistencies are notable given that Maltese market characteristics, concentrated ownership, small market size, and comparatively weaker enforcement, are linked to higher EM risk (Leuz et al., 2003). This highlights the need for updated empirical evidence.
To address this gap, this study evaluates the presence of EM in Maltese Listed Entities (MLEs) and examines the motivations and deterrents influencing both auditors and auditees. The research objectives are to (i) assess the prevalence of EM among non-financial MLEs; (ii) explore the underlying motivations and drivers that give rise to such practices; and (iii) investigate the methods and techniques currently employed by auditees or auditors to prevent or detect EM within MLEs.
Overall, the study aims to identify sector-level EM patterns in Malta, explain their drivers and constraints, and propose reforms to strengthen financial reporting integrity within non-financial MLEs.
Although Malta represents a small capital market, its institutional characteristics make it a particularly informative setting for examining earnings management. As a small, IFRS-based EU jurisdiction with concentrated ownership structures and evolving regulatory enforcement, Malta provides a micro-regulatory laboratory in which the interaction between managerial discretion, audit oversight, and principles-based financial reporting can be observed in a relatively transparent manner (Grima et al., 2021; Pavia et al., 2021).
Studying earnings management in this context yields insights that extend beyond Malta, offering implications for other small and medium-sized capital markets operating under IFRS, where limited market depth, close stakeholder relationships, and developing enforcement mechanisms may heighten both the incentives for and constraints on earnings management.
Prior research demonstrates that institutional characteristics, such as ownership concentration, market size, and enforcement strength, significantly influence earnings management behaviour (Ball et al., 2000; Francis & Wang, 2008; Leuz et al., 2003). Small capital markets are frequently characterised by closer relationships among managers, auditors, and regulators, which may simultaneously constrain opportunistic behaviour and create conditions conducive to financial reporting discretion.
Within this stream of literature, Malta represents a micro-regulatory environment in which the effects of IFRS discretion and audit oversight can be examined with reduced noise from large-market complexity. As such, the Maltese context offers a valuable setting for extending earnings management research into small, IFRS-adopting EU markets that remain under-represented in the empirical literature.
2. Literature Review
Earnings remain the principal indicator of financial performance (Curtis et al., 2021; Lynch & Rothchild, 2000; Nissim, 2021). They influence investor decisions, executive compensation, debt agreements, and wider contractual arrangements (Francis et al., 2008). As a result, poor earnings quality can distort capital allocation and undermine economic efficiency (Baker et al., 2019; Tahat et al., 2022).
Although both GAAP and IFRS allow professional judgement, this flexibility creates scope for manipulation (Adhikari et al., 2021; Callao & Jarne, 2010; Hepworth, 1953). Consequently, EM has emerged as a means by which firms shape perceptions of financial health within the boundaries, or ambiguities, of accounting standards.
Despite its prominence, EM lacks a single accepted definition (Beneish, 2001; Franceschetti, 2018). Healy and Wahlen’s (1999) description remains influential, framing EM as the use of judgment in reporting or structuring transactions to mislead stakeholders or influence contractual outcomes.
Ronen and Yaari’s (2008) review of literature categorises EM into three types: beneficial EM, which improves transparency (Beneish, 2001; Suh, 1990); neutral EM, which is legally permissible but opportunistic (Fields et al., 2001; Mitschow, 2003); and harmful EM, intended to mislead stakeholders (Levitt, 1998; Schipper, 1989; Tzur & Yaari, 1999). Nonetheless, the boundary between EM and fraud is frequently obscured (Dechow & Skinner, 2000).
EM techniques may be income-increasing, inflating earnings, or income-decreasing, deferring earnings, and may occur through accrual-based discretion or through real EM methods, such as delaying expenditure (Almahrog et al., 2016; Guillamon-Saorin et al., 2010). Empirical studies, including those by Healy (1985) and DeFond & Jiambalvo (1994), demonstrate management’s consistent use of accruals to alter reported earnings within the latitude allowed by accounting standards.
Two dominant perspectives shape EM research. The opportunistic view sees managers manipulating numbers for personal benefit (Levitt, 1998; Schipper, 1989; Tzur & Yaari, 1999), whereas the information perspective argues that discretion can signal private expectations of future performance (Beneish, 2001; Demski et al., 1984; Sankar & Subramanyam, 2001; Suh, 1990).
Agency Theory is the dominant lens in EM research. The separation of ownership and control creates incentives for managers to act in self-interest, generating agency costs and distorted financial reporting (Almahrog et al., 2016; Armstrong et al., 2025; Ater & Hansen, 2020, Bathala & Rao, 1995; Eisenhardt, 1989; Jensen & Meckling, 1976; Ronen & Yaari, 2008; Shapiro, 2005; Tahat et al., 2022).
Signalling Theory adds that managers may manipulate disclosures to influence market perceptions, especially in the presence of information asymmetry (Abed et al., 2022; Grougiou et al., 2014; Miller, 2002). Lastly, Stakeholder Theory extends such perspectives by recognising a broader set of parties affected by reporting decisions (Freeman, 2010).
Auditee incentives to exploit accounting flexibility depend on firm-specific and industry-specific circumstances (Callao et al., 2021). Evidence from Malaysia shows EM practices vary across industries, with retail, real estate (RE), and services demonstrating higher manipulation (Wasiuzzaman, 2018).
Several research papers (Almahrog et al., 2016; Bansal 2024a; Mangala & Isha, 2017; Nelson et al., 2002) refer to Healy and Wahlen (1999), who identify three core motivations:
• Capital market incentives-Managers manipulate earnings to meet analyst forecasts, sustain share prices, or reduce earnings volatility (Abarbanell & Lehavy, 2003; Brown & Caylor, 2004).
• Contractual incentives—Earnings affect debt covenants and bonus schemes, encouraging manipulation to avoid covenant breaches or maximise remuneration (Almahrog et al., 2016; DeFond & Jiambalvo, 1994; Franz et al., 2014).
• Regulatory incentives - Entities adjust earnings to exploit regulatory benefits or reduce scrutiny (Chen et al., 2010; Collins et al., 1995; Friedlan 1994).
External auditing reduces agency costs by providing independent assurance on the truthfulness of FS (Aschauer & Quick, 2018; DeAngelo, 1981). Nonetheless, auditors also face competing incentives.
On one hand, economic pressures encourage maintaining positive client relationships and avoiding conflict (McCracken et al., 2008). Contrastingly, professional standards and litigation risk promote scepticism and high-quality audits (Azad et al., 2023; Carrington, 2010; Larcker & Richardson, 2004; Nelson et al., 2002; Ruhnke & Schmidt, 2019). This tension influences how aggressively auditors challenge potential EM.
A country’s legal environment significantly influences reporting behaviours (Ball et al., 2000; Francis & Wang, 2008; Leuz et al., 2003). In Malta, the Accountancy Board, the Malta Institute of Accountants, and the Malta Financial Services Authority oversee the implementation of accounting and auditing standards (Accountancy Board, 2024). The Quality Assurance Unit established by the Accountancy Board also conducts periodic inspections of audit firms.
Despite these mechanisms, Maltese regulation is still maturing, and further strengthening of oversight is needed (Fabri, 2016).
EM prevention strategies largely revolve around effective corporate governance (CG) (Bansal, 2024b; Boachie & Mensah, 2022).
Empirical studies associate independent boards with reduced EM (Alves, 2023; Kjærland et al., 2020;Klein, 2002). Audit committees (AC) with financial expertise also enhance oversight (Turegün & Kaya, 2016). However, board characteristics may correlate with unobservable factors that also influence EM (Xie et al., 2003).
Strong internal audit (IA) functions contribute to reduced EM by enhancing compliance and internal control monitoring (Abbott et al., 2016; Prawitt et al., 2009). From a technological perspective, information systems may similarly reduce opportunities for manipulation by improving accuracy and timeliness, though they can also expand avenues for sophisticated manipulation (Bartov et al., 2017; Sułkowski et al., 2019).
The effectiveness of EM detection depends largely on audit quality, defined as the probability of detecting and reporting a material misstatement (DeAngelo, 1981). High-quality audits are associated with lower EM (Infuehr, 2022), with Big-4 audit firms generally providing superior audit quality due to greater expertise and resources (DeFond & Zhang, 2014).
Although ISAs do not address EM directly, ISA 240 outlines auditor responsibilities relating to fraud, offering general guidance relevant to EM practices (Czakowska, 2020; Shbeilat, 2024). Nonetheless, managers may employ audit management tactics that shift manipulation into less scrutinised areas, reducing detection probability (Luippold et al., 2015)
3. Research Methodology
To achieve the study’s objectives, a sequential two-phase explanatory mixed-method approach was employed. The quantitative data were first collected and analysed using the accrual-based Yoon et al. (2006) model. The preliminary analysis of the quantitative data provided a basis for the design of the interview schedules, which explored the motivations, perceptions and detection practices surrounding EM. Such an approach provided greater depth to the research, in line with Hurmerinta-Peltomäki & Nummela (2006), who evidenced increased value in international business research when employing both quantitative and qualitative methods.
The study population consisted of non-financial MLEs that had equities or bonds listed on the Malta Stock Exchange (MSE) as at 31st December 2023. Financial MLEs, including banking and insurance entities, were excluded in line with DeFond & Jiambalvo (1994) and Becker et al. (1998), who highlight that such firms are controlled by supplementary regulations leading to different environments in which EM operates, making the computation of discretionary accruals (DA) problematic. In addition, any industry with fewer than ten firms per industry-year observation was omitted, as were MLEs lacking any of the required variables for the Yoon et al. (2006) model.
EM is commonly analysed by examining DA (DeAngelo, 1986; Dechow et al., 1995; Healy, 1985; Jones, 1991; Yoon et al., 2006). While the Modified Jones model is the most common aggregate accrual model used in research to estimate DA (Islam et al., 2011; Peasnell et al., 2000; Stubben, 2010), its application to local data yielded a very low goodness-of-fit, with the explanatory variables lacking statistical significance.
Consistent with the approaches of Islam et al. (2011), Alareeni and Aljuaidi (2014) and Yoon et al. (2006), all of whom encountered similar challenges with the Modified Jones model, this study adopts the cross-sectional variant of the Yoon et al. (2006) model, which possesses a significantly enhanced goodness-of-fit and statistical significance of explanatory variables. This is illustrated in Table 1.
Econometric models, as adopted in this study, hold that Total Accruals (TA) are comprised of DA and Non-discretionary accruals (NDA):
When no EM is present, TA is equal to NDA. Thus, DA are the difference between TA and NDA, where DA equates to zero, in the absence of EM.
In econometric accrual models, TA are regressed on predictors of NDA; the unexplained residual represents DA (Dechow et al., 1995), as expressed in the following form:
The Yoon et al. (2006) model includes the following variables:
where,
TA = Accounting Earnings – Cash Flow from Operations
REV = Net Sales Revenue
REC = Receivables
EXP = Sum of Cost of Goods Sold and Selling and General Administrative Expenses excluding Non-Cash Expenses
PAY = Trade Payables
DEP = Depreciation Expenses
PEN = Retirement Benefits Expenses
Δ = Change Operator
$\hat{\beta}_0,\hat{\beta}_1,\hat{\beta}_2,\hat{\beta}_3$ = Estimated Coefficients
Following the econometric approach, DA were derived as the portion of TA exceeding expected NDA. Using the estimated coefficients from Eq. (3), DA were computed as the residual from the fitted model:
Such residuals quantify the divergence between actual TA and expected NDA for each observation and serve as the proxy for EM. Positive values indicate income-increasing EM, while negative values indicate income-decreasing EM.
Modified Jones Model | Yoon et al. (2006) Model | ||||||||
Sector | Explanatory Variable | p-Value | Sig. | R2 (%) | Explanatory Variable | p-Value | Sig. | R2 (%) | |
CD | $\frac{1}{Assets_{t-1}}$ | 0.475 | No | 5.4 | $\frac{(\triangle R E V i-\triangle R E C i)}{R E V i}$ | <0.001 | Yes | 87.9 | |
$\frac{\Delta Rev-\Delta Rec}{Assets_{t-1}}$ | 0.425 | No | $\frac{(\triangle E X P i-\triangle P A Y i)}{R E V i}$ | 0.001 | Yes | ||||
$\frac{PPE}{Assets_{t-1}}$ | 0.302 | No | $\frac{(D E P i+P E N i)}{R E V i}$ | 0.033 | Yes | ||||
RE | $\frac{1}{Assets_{t-1}}$ | 0.545 | No | 1.9 | $\frac{(\triangle R E V i-\triangle R E C i)}{R E V i}$ | 0.035 | Yes | 17.5 | |
$\frac{\Delta Rev-\Delta Rec}{Assets_{t-1}}$ | 0.911 | No | $\frac{(\triangle E X P i-\triangle P A Y i)}{R E V i}$ | 0.629 | No | ||||
$\frac{PPE}{Assets_{t-1}}$ | 0.645 | No | $\frac{(D E P i+P E N i)}{R E V i}$ | 0.514 | No | ||||
Indust-rials | $\frac{1}{Assets_{t-1}}$ | 0.359 | No | 18.8 | $\frac{(\triangle R E V i-\triangle R E C i)}{R E V i}$ | 0.561 | No | 12 | |
$\frac{\Delta Rev-\Delta Rec}{Assets_{t-1}}$ | 0.715 | No | $\frac{(\triangle E X P i-\triangle P A Y i)}{R E V i}$ | 0.676 | No | ||||
$\frac{PPE}{Assets_{t-1}}$ | 0.977 | No | $\frac{(D E P i+P E N i)}{R E V i}$ | 0.438 | No | ||||
The group-level data for the explanatory variables was collected from the audited FS of non-financial MLEs for the financial years 2023 and 2022. In accordance with the cross-sectional method, the data from the MLEs were categorised by their respective industries. Since the MSE does not provide an Industry Classification Benchmark Code (ICB), MLEs were manually assigned to industries using the FTSE Russell (2024) ICB Code Description. When an MLE spanned multiple industries, it was included in each applicable industry. Table 2 details the industry observations.
The segmented data were analysed using IBM SPSS Statistics Version 29. Normality was tested using the Shapiro-Wilk test owing to the small sample size. The consumer discretionary (CD) and real estate (RE) industries violated normality, while the Industrials sector satisfied it. Both the CD and RE datasets exhibited right-skewness. A 1% winsorization procedure, consistent with Yoon et al. (2006) and Peasnell et al. (2000), was considered; however, it did not meaningfully improve the distributional properties or the robustness of the model, and was therefore not adopted. A Generalised Linear Model was considered but rejected because it removed negative values and failed to converge. Linear regression was therefore retained for its interpretability, preservation of data characteristics, and consistency with prior EM research.
Industry | Industry-Observations |
CD | 26 |
RE | 31 |
Industrials | 10 |
Total | 67 |
Twenty semi-structured interviews were conducted with 12 Chief Financial Officers (CFOs), and 8 Audit Partners (“APs”), 4 of whom were each representative of the big-4 firms, and the remaining 4 were non-big-4 APs. This enabled a dual-perspective approach in which CFOs act as key financial decision-makers, while APs provide a critical external view.
Given the small size and concentrated nature of the Maltese capital market, this study adopted an expert-based purposive sampling strategy focusing on key financial decision-makers and audit professionals within non-financial Maltese Listed Entities. The qualitative sample comprised Chief Financial Officers and Audit Partners who are directly involved in financial reporting judgments and earnings management oversight, ensuring high informational richness per interview.
Data collection continued until thematic saturation was achieved, at which point no substantively new codes, concepts, or relationships emerged in the later stages of the interviews. Saturation was observed before the completion of the full interview schedule, with subsequent interviews confirming and reinforcing previously identified themes rather than introducing novel insights. This aligns with established qualitative research guidance indicating that saturation in expert and relatively homogeneous populations is commonly achieved within 12–25 interviews.
In line with qualitative research conventions, the findings are not intended to be statistically generalisable. Instead, they offer analytical and contextual generalisability, whereby insights derived from Malta as a small, IFRS-adopting EU jurisdiction may be transferable to other small and medium-sized capital markets characterised by concentrated ownership structures, close stakeholder relationships, and evolving regulatory enforcement (Hennink et al., 2017; Guest et al., 2006; Guest et al., 2012; Saunders et al., 2018).
Semi-structured interviews offered the flexibility needed for in-depth probing, which is particularly valuable in exploratory research (Karatsareas, 2022; Adams, 2015).
Two separate interview schedules were compiled for the APs and CFOs. Parallel themes were utilised in both interview schedules to facilitate the identification of similarities and divergences in perspectives between the two respondent groups. Both schedules consisted of a synopsis of the research topic, followed by three sections, each addressing a research objective through open-ended questions and one closed-ended question on a 5-point Likert scale, where ‘1’ indicated strongly disagree and ‘5’ indicated strongly agree.
The qualitative data were analysed by reading transcripts several times, grouping the interview transcripts by question, and subsequently identifying pertinent themes in each question using the coding function in NVivo (Release 15.1.1). Thus, thematic analysis was utilised for qualitative data and topic-wide themes were identified. In addition, the Likert-scale question was imported into IBM SPSS Statistics Version 29 for analysis.
While qualitative findings are analytically transferable to comparable small-market contexts, they are not statistically generalisable beyond similar IFRS-based jurisdictions (Flyvbjerg, 2006; Patton, 2015; Yin, 2018).
Despite substantial efforts to conduct a thorough analysis of the research topic, the study still faces certain limitations.
From the quantitative perspective, while cross-sectional accruals-based models are widely used to estimate DA as a proxy for EM, they are prone to measurement error, as model residuals may capture noise or omitted firm-specific factors rather than true discretion (Dechow et al., 1995).
The inclusion of $\frac{(D E P i+P E N i)}{R E V i}$ As a proxy for non-current accruals, it may also understate discretionary elements (Yoon et al., 2006). The small size of the MSE limited the number of observations, increasing sensitivity to outliers (Dechow et al., 1995; Kothari et al., 2005). Moreover, the normality assumption was violated in two sectors, potentially affecting the precision of the estimates. At the same time, the explanatory power of the RE and Industrials models was reduced because other EM determinants were unaccounted for.
From a qualitative perspective, reliance on self-reported information about EM introduces the risk of social desirability bias (Atkinson & Willis, 2007; Spokes & Denham, 2019). This was mitigated by triangulating perspectives from both auditors and auditees; however, participation was limited to individuals who consented to interviews.
Moreover, the study focused specifically on the CD, RE and Industrials sectors of non-financial MLEs, with quantitative analysis restricted to 2022 and 2023 FS and qualitative data collected up to 7 March 2025. Consequently, findings may not be generalisable beyond these sectors, time frames or the Maltese context.
Therefore, despite methodological triangulation, the findings, while insightful, should be interpreted cautiously as they may not be generalisable or relevant across all contexts.
4. Findings
Linear regression was performed in SPSS to estimate the Beta-coefficients for each explanatory variable of the Yoon et al. (2006) model. The change agent represents the difference between 2023 and 2022 FS figures. Once the model was regressed, the following results ensued:
The CD Sector:
All three explanatory variables were statistically significant, as illustrated in Table 3.
$\frac{(\Delta R E V i-\Delta R E C i)}{R E V i}$ Had a negative coefficient (-3.581, p < 0.001), indicating accelerated revenue recognition.
$\frac{(\triangle R E V i-\triangle R E C i)}{R E V i}$ Was positive (1.330, p < 0.001), evidencing expense deferral.
$\frac{(D E P i+P E N i)}{R E V i}$ Was also positive (3.870, p = 0.033), suggesting non-cash expense deferral.
Unstandardized B | t | Sig. | |
$\frac{(\triangle R E V i-\triangle R E C i)}{R E V i}$ | -3.581 | -12.786 | <0.001 |
$\frac{(\triangle E X P i-\triangle P A Y i)}{R E V i}$ | 1.330 | 3.788 | <0.001 |
$\frac{(D E P i+P E N i)}{R E V i}$ | 3.870 | 2.259 | 0.033 |
Constant | -0.033 | -0.138 | 0.891 |
Dependent Variable = $\frac{TA i}{R E V i}$ | |||
The above results in the following model:
$\frac{T A i}{R E V i}=-0.033-3.581 \frac{(\triangle R E V i-\triangle R E C i)}{R E V i}+1.330 \frac{(\triangle E X P i-\triangle P A Y i)}{R E V i}+3.870 \frac{(D E P i+P E N i)}{R E V i}+\varepsilon i$
The RE Sector:
One variable was statistically significant, as illustrated in Table 4.
$\frac{(\triangle R E V i-\triangle R E C i)}{R E V i}$ was negative $(-0.941, p=0.035)$, suggesting accelerated sales similar to the CD sector.
$\frac{(\triangle E X P i-\triangle P A Y i)}{R E V i} \quad(0.178, p=0.629)$ and $\frac{(D E P i+P E N i)}{R E V i}(-3.351, p=0.514)$ were not significant, indicating limited evidence of expense deferral.
The above results in the following model:
$\frac{T A i}{R E V i}=0.322-0.941 \frac{(\triangle R E V i-\triangle R E C i)}{R E V i}+0.178 \frac{(\triangle E X P i-\triangle P A Y i)}{R E V i}-3.351 \frac{(D E P i+P E N i)}{R E V i}+\varepsilon i$
The Industrials Sector:
None of the explanatory variables were statistically significant, and thus there is no evidence of a meaningful relationship with the dependent variable, as illustrated in Table 5.
The above results in the following model:
$\frac{T A i}{R E V i}=0.096+0.186 \frac{(\Delta R E V i-\triangle R E C i)}{R E V i}-0.195 \frac{(\triangle E X P i-\triangle P A Y i)}{R E V i}-1.306 \frac{(D E P i+P E N i)}{R E V i}+\varepsilon i$
Unstandardized B | t | Sig. | |
$\frac{(\triangle R E V i-\triangle R E C i)}{R E V i}$ | -0.941 | -2.221 | 0.035 |
$\frac{(\triangle E X P i-\triangle P A Y i)}{R E V i}$ | 0.178 | 0.489 | 0.629 |
$\frac{(D E P i+P E N i)}{R E V i}$ | -3.351 | -0.661 | 0.514 |
Constant | 0.322 | 0.640 | 0.528 |
Dependent Variable = $\frac{TA i}{R E V i}$ | |||
Unstandardized B | t | Sig. | |
$\frac{(\triangle R E V i-\triangle R E C i)}{R E V i}$ | 0.186 | 0.615 | 0.561 |
$\frac{(\triangle E X P i-\triangle P A Y i)}{R E V i}$ | -0.195 | -0.440 | 0.676 |
$\frac{(D E P i+P E N i)}{R E V i}$ | -1.306 | -0.831 | 0.438 |
Constant | 0.096 | 0.487 | 0.644 |
Dependent Variable = $\frac{TA i}{R E V i}$ | |||
The sector-specific models were then utilised to estimate DA through Eq. (4). Since DA serves as the proxy for EM, a positive value indicates income-increasing EM, while a negative value suggests income-decreasing EM. Table 6, Table 7, Table 8 present the DA calculations for the three sectors.
Code (CDn = CD Entity no.n) | DA | Type of EM |
CD1 | 1.05 | Income Increasing |
CD2 | -0.52 | Income Decreasing |
CD3 | -0.12 | Income Decreasing |
CD4 | -0.64 | Income Decreasing |
CD5 | -0.03 | Income Decreasing |
CD6 | -0.39 | Income Decreasing |
CD7 | -1.32 | Income Decreasing |
CD8 | 0.30 | Income Increasing |
CD9 | 0.07 | Income-Increasing |
CD10 | -0.12 | Income Decreasing |
CD11 | 0.66 | Income Increasing |
CD12 | 0.81 | Income Increasing |
CD13 | 0.12 | Income Increasing |
CD14 | -1.03 | Income Decreasing |
CD15 | 0.46 | Income Increasing |
CD16 | 0.24 | Income Increasing |
CD17 | 0.08 | Income Increasing |
CD18 | 0.17 | Income Increasing |
CD19 | -0.35 | Income Decreasing |
CD20 | -0.17 | Income Decreasing |
CD21 | -0.16 | Income Decreasing |
CD22 | 0.49 | Income Increasing |
CD23 | -0.05 | Income Decreasing |
CD24 | 0.45 | Income Increasing |
CD25 | -0.28 | Income Decreasing |
CD26 | -0.71 | Income Decreasing |
Code (REn = RE Entity no.n) | DA | Type of EM |
RE1 | -0.32 | Income Decreasing |
RE2 | -2.83 | Income Decreasing |
RE3 | -0.12 | Income Decreasing |
RE4 | -0.16 | Income Decreasing |
RE5 | -1.69 | Income Decreasing |
RE6 | 0.73 | Income Increasing |
RE7 | 0.16 | Income Increasing |
RE8 | -0.25 | Income Decreasing |
RE9 | -0.11 | Income Decreasing |
RE10 | -0.26 | Income Decreasing |
RE11 | 6.44 | Income Increasing |
RE12 | 1.74 | Income Increasing |
RE13 | 0.05 | Income Increasing |
RE14 | -0.06 | Income Decreasing |
RE15 | -2.30 | Income Decreasing |
RE16 | -0.72 | Income Decreasing |
RE17 | -4.74 | Income Decreasing |
RE18 | 0.00 | Income Increasing |
RE19 | -0.19 | Income Decreasing |
RE20 | 0.16 | Income Increasing |
RE21 | 0.23 | Income Increasing |
RE22 | 0.66 | Income Increasing |
RE23 | 0.13 | Income Increasing |
RE24 | 0.31 | Income Increasing |
RE25 | 0.10 | Income Increasing |
RE26 | 0.76 | Income Increasing |
RE27 | 0.31 | Income Increasing |
RE28 | 0.75 | Income Increasing |
RE29 | 0.17 | Income Increasing |
RE30 | 0.66 | Income Increasing |
RE31 | 0.37 | Income Increasing |
Code (In = Industrials Entity no.n) | DA | Type of EM |
I1 | -0.09 | Income Decreasing |
I2 | 0.21 | Income Increasing |
I3 | -0.27 | Income Decreasing |
I4 | 0.00 | Neutral |
I5 | 0.61 | Income Increasing |
I6 | -0.13 | Income Decreasing |
I7 | -0.19 | Income Decreasing |
I8 | 0.01 | Income Increasing |
I9 | -0.08 | Income Decreasing |
I10 | -0.06 | Income Decreasing |
Descriptive Statistics were employed to assess the extent and distribution of EM across the sectors (Kaliyadan & Kulkarni, 2019). The descriptive statistics are presented in Table 9.
The CD sector reported a mean DA of –0.038 and a median of –0.04, indicating a slight income-decreasing tendency and a relatively symmetric, near-normal distribution. Variability was moderate (SD = 0.546), with only mild skewness and kurtosis.
The RE sector showed minimal average EM (mean DA = -0.001), yet the median DA of 0.1 and the highest SD (1.703) indicated substantial dispersion and a right-skewed distribution. The notably high kurtosis (7.987) suggests the presence of extreme values and significant EM outliers.
The Industrials sector displayed an almost neutral mean DA (0.001) and lower variability (SD = 0.246). However, strong positive skewness (1.827) and elevated kurtosis (3.917) indicated that a small number of firms accounted for unusually high EM, despite the sector's overall stability.
Despite the RE sector exhibiting the highest variability in DA, the Kruskal-Wallis test revealed no statistically significant differences in DA distributions across industries (p = 0.722). This suggests that while EM can vary significantly across industries, there are no consistent differences within them.
Statistic | Consumer Discretionary | Real Estate | Industrials |
Mean | -0.038 | -0.001 | 0.001 |
95% Confidence Interval–Lower Bound | -0.258 | -6.253 | -1.776 |
95% Confidence Interval–Upper Bound | 1.823 | 6.240 | 1.780 |
5% Trimmed Mean | -0.028 | -0.040 | -0.018 |
Median | -0.040 | 1.000 | -0.070 |
Variance | 0.298 | 2.900 | 0.062 |
Standard Deviation | 0.546 | 1.703 | 0.250 |
Minimum | -1.320 | -4.740 | -0.270 |
Maximum | 1.050 | 6.440 | 0.610 |
Range | 2.370 | 11.180 | 0.880 |
Interquartile Range | 0.700 | 0.620 | 0.210 |
Skewness | 0.296 | 0.923 | 1.827 |
Skewness Std. Error | 0.456 | 0.421 | 0.687 |
Kurtosis | 0.314 | 7.897 | 3.917 |
Kurtosis Std. Error | 0.887 | 0.821 | 1.334 |
No evidence of significant multicollinearity was found among the predictors, given that all condition indices and Variance Inflation Factor values remain well below the threshold of 10 (Mason, 1987).
The CD sector demonstrated a notably high explanatory power, with an R² of 87.9%, whereas the RE and Industrials sectors showed substantially lower R² values of 17.5% and 12% respectively, as illustrated in Table 10. Although the small sample sizes may partly account for these differences, the variation in explanatory power more likely reflects the influence of sector-specific dynamics and additional unobserved factors affecting DA (Bass et al., 1968). This outcome is consistent with prior literature, which notes that accrual-based EM models often yield low R² values (Dechow & Skinner, 2000).
Industry | Explanatory Variable | p-Value | R2 |
Dependent Variable = $\frac{TA i}{R E V i}$ | |||
CD | $\frac{(\triangle R E V i-\triangle R E C i)}{R E V i}$ | <0.001 | 87.9% |
$\frac{(\triangle E X P i-\triangle P A Y i)}{R E V i}$ | 0.001 | ||
$\frac{(D E P i+P E N i)}{R E V i}$ | 0.033 | ||
Intercept | 0.791 | ||
RE | $\frac{(\triangle R E V i-\triangle R E C i)}{R E V i}$ | 0.035 | 17.5% |
$\frac{(\triangle E X P i-\triangle P A Y i)}{R E V i}$ | 0.629 | ||
$\frac{(D E P i+P E N i)}{R E V i}$ | 0.514 | ||
Intercept | 0.528 | ||
Industrials | $\frac{(\triangle R E V i-\triangle R E C i)}{R E V i}$ | 0.561 | 12% |
$\frac{(\triangle E X P i-\triangle P A Y i)}{R E V i}$ | 0.676 | ||
$\frac{(D E P i+P E N i)}{R E V i}$ | 0.438 | ||
Intercept | 0.644 | ||
The first section of the interview schedule consisted of questions about respondents’ understanding of EM within MLEs.
In the first two questions, respondents were asked to provide their perceptions of the importance of earnings when assessing an entity's health. All participants (20/20) agreed that earnings are a highly significant Key Performance Indicator (KPI) and a core indicator of organisational performance. However, some APs (4/8) and CFOs (4/12) highlighted certain caveats. Firstly, the relative importance of earnings varies across stakeholder groups, and secondly, earnings importance may be less representative for different company types at different stages of the entity’s lifecycle. Despite this, most APs (7/8) and CFOs (8/12) noted that earnings should be analysed alongside other key metrics, such as liquidity, cash flow, and balance sheet metrics.
A follow-up question then explored whether the significance of earnings influences the interviewees’ respective processes. Most APs (6/8) indicated that the prominence of earnings influences audit strategy, particularly in materiality assessments and in increased scrutiny of high-risk areas. Conversely, other Aps (2/8) maintained that earnings’ importance should not affect the audit’s objective of presenting a true and fair view.
Among CFOs, most (7/12) stated that earnings’ importance does not alter the FR process, and that management should report on the true figures, whether they are satisfactory or not. Others (5/12) acknowledged that it may influence reporting behaviour or lead to the inclusion of more earnings-related disclosures.
When asked to define EM, all APs (8/8) and some CFOs (5/12) defined EM as an opportunistic practice in which leeway within accounting regulations is utilised in areas of judgement.
Other CFOs (5/12) defined EM as the manipulation of figures to achieve a desired outcome. The sentiment of such respondents evoked a stronger intent to mislead or achieve a specific financial outcome, rather than simply leveraging judgement-based flexibility. Lastly, a CFO (1/12) defined EM in terms of business adaptation and transformation. At the same time, another CFO (1/12) described EM as the practice of adjusting earnings to depict a true and fair view for the reporting period.
When respondents were queried on EM’s relevance to their respective industries, most APs (6/8) noted that the risk of EM heightens the need for sufficient audit evidence in judgment-intensive areas to ensure that accounting treatments reflect economic substance. Others (2/8) emphasised that EM primarily increases the level of professional scepticism required, which varies with the nature of the entity.
From the CFOs’ perspective, several respondents (5/12) stated that EM risk differs across industries, with sectors involving significant related-party transactions or high judgment, such as RE and pharmaceuticals, being particularly susceptible. Lastly, a minority of CFOs (3/12) also noted that, despite external pressures, engaging in EM conflicts with their role as organisational gatekeepers.
When asked to distinguish between EM and fraud, all APs (8/8) and most CFOs (10/12) acknowledged that the boundary between EM and fraud is unclear. Most APs (6/8) and CFOs (7/12) identified intent as the key differentiating factor; EM remains acceptable when used to present results favourably, but becomes fraudulent when used to deliberately mislead stakeholders.
Some APs (2/8) and CFOs (5/12) also noted that EM commonly arises in judgment-based areas. While such judgements may legitimately fall within EM when exercised in good faith, respondents noted that a consistent directional bias in these estimates signals a shift toward fraud. However, a minority of CFOs (2/12) rejected the notion of a grey area, arguing that any misleading accounting choice should be classified as fraud.
When participants were asked whether they perceived EM to be actively present in MLEs, most APs (6/8) and some CFOs (6/12) believed EM is present, noting that the local context is no exception to international trends. The remaining APs (2/4) and CFOs (6/12) claimed that their experiences led them to believe that there is no EM present within MLEs. The respondents (8/20) argued that the practice conflicts with ethical standards and noted that the high public scrutiny and stringent Listing Rules create an environment in which EM is unlikely to occur.
Participants were subsequently presented with the preliminary quantitative results of this study and asked to provide their perceptions of these findings. Participants’ perceptions largely aligned with their prior perceptions of EM in Malta, with two exceptions (2/20).
Some CFOs (4/12) attested that such findings are unexpected and concerning, given that MLEs are expected to exemplify robust FR. In addition, some APs (5/8) and CFOs (4/12) attributed these findings to external pressures and market dynamics, noting that the subjective nature of certain transactions creates grey areas that enable EM. Some APs (2/8) and CFOs (2/12) highlighted that the findings indicate the need for stronger oversight, and some (3/20) perceived the EM observed as income-smoothing rather than aggressive earnings inflation.
Despite differing views on its impact, most APs (6/8) and CFOs (8/12) agreed that EM undermines stakeholder trust in FS, with one CFO (1/12) observing that recurring EM patterns, such as property revaluations, can serve as red flags, reducing confidence in reported figures.
The second section of the interview schedule consisted of questions about the motivations and drivers of EM practices.
The next question requested participants to indicate whether they believe the IFRS framework grants flexibility to facilitate EM. All APs (8/8) and most CFOs (8/12) agreed that the IFRS framework provides flexibility that can facilitate EM. Some APs (2/8) and CFOs (2/12) attributed this to the conceptual nature of IFRS, which lacks prescriptive rules, while a few CFOs (2/12) argued that IFRS is sufficiently clear to limit opportunities for EM. Other CFOs (2/12) noted that the scope for flexibility depends on industry context or preparers’ intent.
Participants who viewed IFRS as flexible (16/20) highlighted practical examples, with IFRS 9 most frequently cited due to the significant judgement involved in expected credit loss models. Table 11 summarises the IFRS principles most frequently referenced by respondents to support their views on flexibility.
Principle/Aspect of IFRS Framework | APs (n/8) | CFOs (n/12) | Total (n/20) |
IFRS 9 | 3 | 2 | 5 |
Material Estimations & Provisions | 2 | 2 | 4 |
Capitalisation vs Expensing of Costs | 2 | 1 | 3 |
Depreciation Choice & Useful Life Estimates | 2 | 1 | 3 |
IFRS 16 | 0 | 2 | 2 |
IFRIC 12 | 0 | 2 | 2 |
When discussing the primary motivations behind EM practices within MLEs, respondents (20/20) highlighted a range of financial and strategic drivers. Most APs (6/8) and CFOs (8/12) identified market pressures as a key driver, noting that the choice of income-increasing or income-decreasing practices depends on the image the company aims to convey. A few respondents (3/20) also highlighted a heightened risk of EM during company valuations in a takeover process.
Contractual pressures were also highlighted, with some APs (4/8) and CFOs (3/12) noting that performance-based bonuses tied to earnings or debt covenants create conditions in which violations carry significant financial consequences. To avoid such outcomes, management is motivated to engage in EM to ensure these obligations are met.
At an aggregate level, most APs (5/7) and CFOs (7/12) emphasised corporate-level pressures as a more significant driver than personal incentives, citing the broader impact and potential for collusion. A minority of respondents indicated that personal incentives and corporate pressures are equally influential (4/20). At the same time, a few CFOs (2/12) suggested that personal incentives may play a dominant role due to management’s influence and short-term focus.
CFOs (12/12) were then asked to indicate their level of agreement with six statements to assess their motivations for EM practices. The descriptive statistics of the responses are presented in Table 12, in descending order of agreement.
CFO responses indicated that contractual obligations, such as debt covenants (x̄ = 3.33), were the most significant motivator for EM. Industry-specific requirements (x̄ = 2.75) and regulatory incentives (x̄ = 2.67) received mixed responses, slightly leaning towards disagreement. Market perceptions (x̄ = 2.58) were similarly met with indecision leaning towards disagreement, while compensation targets (x̄ = 1.92) and analyst expectations (x̄ = 1.67) were generally disagreed with.
CFOs | N | Mean | Std. Dev | Mode | Min | Max |
S4: Contractual obligations, such as debt covenants, create pressure to manage earnings in specific ways. | 12 | 3.33 | 1.30 | 4 | 1 | 5 |
S6: Entities sometimes adjust earnings to meet industry-specific regulations. | 12 | 2.75 | 1.14 | 4 | 1 | 4 |
S5: Government subsidies or regulatory incentives impact an entity’s reporting practices. | 12 | 2.67 | 1.30 | 2 | 1 | 5 |
S1: EM practices are sometimes necessary to avoid adverse market perceptions. | 12 | 2.58 | 1.56 | 1 | 1 | 5 |
S3: Compensation targets and bonuses for management significantly influence financial reporting choices. | 12 | 1.92 | 1.08 | 1 | 1 | 4 |
S2: Meeting financial analysts’ earnings expectations is a priority, even if it means adjusting reported earnings. | 12 | 1.67 | 0.99 | 1 | 1 | 4 |
1 = Strongly Disagree; 5 = Strongly Agree | ||||||
Auditors were asked to provide their perspectives on the determinants of EM in MLEs. Most Aps (6/8) observed a misalignment between management and shareholder objectives. In addition, some Aps (4/8) reported witnessing managers engaging in EM to present favourable results ahead of key disclosures or events. In contrast, other APs (4/8) indicated they had not observed such practices and lacked insight into these behaviours.
APs (8/8) were asked about the investigation of misstatements in MLEs, highlighting how materiality judgements can lead to the potential to overlook EM. Most APs (6/8) indicated that both quantitative and qualitative criteria determine whether a misstatement warrants investigation, whereas a minority (2/8) relied solely on clearly trivial thresholds. Despite practical audit constraints, several APs (4/8) emphasised the importance of maintaining high-quality audits, with one AP (1/8) noting that deadlines and budgets should not compromise audit integrity. Some non-big-4 APs (2/4) highlighted the role of technology in efficiently identifying unusual transactions, while some big-4 APs (2/4) stressed that rigorous planning helps manage workload and maintain audit quality.
In parallel to CFOs, APs (8/8) were asked to indicate their level of agreement with five statements about motivators to overlook EM. Table 13 presents the descriptive statistics obtained, in descending order of agreement. APs strongly agreed with “S4” (x̄ = 4.75) and “S2” (x̄ = 4.375), while strongly opposing “S5” (x̄ = 1.375), “S3” (x̄ = 1.25) and “S1” (x̄ = 1.125). This implies that APs are not influenced by client pressures or firm culture in compromising audit quality and remain motivated to uphold rigorous audit standards.
Aps | N | Mean | Std. Dev | Mode | Min | Max |
S4: I am more likely to carry out further testing if the risk of misstatement is material. | 8 | 4.75 | 0.5 | 5 | 4 | 5 |
S2: In a situation where I am pressured to overlook a minor misstatement, I am likely to raise the issue further. | 8 | 4.375 | 0.746 | 4 | 3 | 5 |
S5: If my firm discourages strict insistence on unadjusted misstatements, I am less likely to pursue them. | 8 | 1.375 | 0.519 | 1 | 1 | 2 |
S3: I am likely to ignore misstatements even if I believe non-correction could lead to legal risks. | 8 | 1.25 | 0.463 | 1 | 1 | 2 |
S1: I am likely to accept unadjusted misstatements to maintain a positive client relationship. | 8 | 1.125 | 0.353 | 1 | 1 | 2 |
1 = Strongly Disagree; 5 = Strongly Agree | ||||||
The final section of the interview schedule consisted of questions relating to the prevention and detection of EM methodologies.
When interviewees (20/20) were asked about the role of local accounting and auditing regulatory bodies in addressing EM, most APs (7/8) and some CFOs (6/12) expressed confidence in the regulators’ diligence. A CFO (1/12) noted that these bodies provide guidance, training, and conferences to support robust FR. However, a subset of CFOs (5/12) and an AP (1/8) criticised regulatory oversight for limited initiative and investigation, citing a lack of practical engagement in monitoring EM.
When APs were questioned about the local regulator's quality assurance visits, most (5/8) raised concerns about the focus on compliance rather than substantive review of areas involving significant management judgment. Nonetheless, some (3/8) acknowledged that such quality assurance visits served as an important additional safeguard against EM.
When participants (20/20) were questioned on measures or strategies to prevent the practice of EM from ensuing, most Aps (7/8) and CFOs (11/12) highlighted that a strong system of internal controls and good CG were part of a focal strategy, ensuring that policies and procedures are duly adhered to, and that an ethical culture is fostered within the entity. CFOs (11/12) noted that internal controls and strong CG serve as watchdogs for the entity, ensuring it strictly adheres to IFRSs through investigative efforts, with the AC and ancillary committees providing additional scrutiny.
The next question asked CFOs (12/12) about the use of specialised technology to prevent EM. Several CFOs (6/12) reported using automation in financial controls to detect anomalies, while the remaining respondents (6/12) relied on the robustness of existing processes or the engagement of big-4 auditors or external advisors.
To capture a dual perspective, APs were asked about their perceptions of the effectiveness of current EM preventive techniques in MLEs. All APs (8/8) considered current EM preventive measures effective, though some APs (3/8) highlighted that smaller MLEs may face gaps in their application.
To further strengthen EM mitigation efforts in MLEs, most APs (6/8) recommended deploying stronger ACs and engaging directors. Other APs cited appropriate recruitment and training to reinforce ethical standards (3/8) or a greater educational effort to manage unrealistic growth expectations of MLEs (2/8).
The next question posed to APs (8/8) concerned red flags participants perceived as indicative of potential EM. Most APs (6/8) claimed that a key giveaway is when areas of judgment are consistently overly prudent or aggressive, skewing earnings in a desired direction. In addition, some APs (6/8) mentioned the presence of outliers in analytical reviews, while others (7/8) mentioned industry-specific complexities.
Furthermore, APs (8/8) were then questioned about the audit methods and the technology used to detect EM. Respondents discussed the following methods:
Technology: All APs (8/8) agreed that technology is an essential tool for detecting EM by facilitating the review of trends, statistics, and ratios for potential abnormalities. While all Big-4 APs (4/4) confirmed that their firms have adopted Artificial Intelligence (AI) and Data Analytics technologies within their audit procedures, all non-Big-4 APs (4/4) stated that such software is yet to be fully deployed in the coming years. Despite this, all APs (8/8) agreed that AI and Data Analytics can enhance EM detection by improving the efficiency and effectiveness of audit work.
ISA 240: All Aps (8/8) agreed that while ISA 240 is not primarily geared towards EM, the standard’s guidelines help in shaping audit procedures for EM detection. Most APs (5/8) argued that the standard discusses professional judgement and scepticism, which serve as guiding principles when determining whether accounting practices comply with IFRS or constitute fraud. Other APs (4/8) commended the standard's risk-based approach for identifying potential EM.
Journal entry testing: Some APs (2/8) noted this test as an effective tool, since it allows scrutiny of certain keywords, the individual posting the transaction, and several other factors.
The next question asked respondents to relay the challenges APs face in identifying EM. All APs (8/8) agreed that a pertinent challenge is the uncertainty inherent in EM practices, which operate in a grey area. In addition, some APs (4/8) also noted that uncooperative clients hinder the proper identification of EM.
To satisfy the dual-perspective approach, CFOs (12/12) were asked about their views on audit effectiveness in detecting EM. Most CFOs (7/12) believed that audits are generally effective, describing them as rigorous and characterised by strong challenge from auditors. However, some CFOs (3/12) questioned auditors’ technical depth, suggesting that while clear instances of manipulation would be detected, more subtle practices might not. A few CFOs (3/12) also perceived big-4 audits as more robust, noting that firms intending to engage in EM may preferentially choose smaller audit firms.
Overall, most CFOs (9/12) emphasised that open communication and continuous consultation with auditors facilitated effective collaboration and reduced the likelihood of year-end disputes.
5. Discussion
Consistent with Lynch and Curtis et al. (2021), Nissim (2021), and Rothchild (2000), participants unanimously reaffirmed earnings as a central KPI. However, they emphasised that earnings’ relevance varies across stakeholders, organisational types and lifecycle stages, implying that EM's incentives are highly contextual. This echoes Callao et al. (2021), who note that both firm and macro-level conditions influence the pursuit of EM.
APs stated that earnings’ prominence affects audit planning—particularly materiality and audit effort—but differed on whether stakeholder sensitivity to earnings should influence audit strategy. CFOs also varied: while most claimed neutrality, some admitted adjusting disclosures under earnings pressure, indicating potential bias as highlighted by Fang et al. (2017).
These observations reinforce the metric’s centrality but also the risks of low-quality earnings noted by Baker et al. (2019) and Tahat et al. (2022). Excessive reliance on earnings may distort performance signals and increase agency costs, consistent with Almahrog et al. (2016), Bathala & Rao (1995), and Jensen & Meckling (1976). APs’ increased scrutiny in earnings-sensitive areas, together with recognition that managed earnings reduce reporting quality, supports the need for further EM research in the Maltese context.
Participants demonstrated varying interpretations of EM. Many described it in line with Healy & Wahlen (1999) as discretion within IFRS, while others viewed it as manipulation consistent with the “harmful” category defined by Ronen & Yaari (2008). A dominant theme was the blurred boundary between EM and fraud, supporting critiques by Dechow & Skinner (2000), and that intent is subjective and that current definitions inadequately distinguish the two.
This ambiguity mirrors findings by Beneish (2001), Dechow and Skinner (2000), Franceschetti (2018), Kamau and Murori (2024), and Ronen and Yaari (2008), who note the absence of a universal definition. Conceptual uncertainty weakens ethical boundaries, may permit misleading yet IFRS-compliant practices, and complicates regulatory intervention, as highlighted by Benkraiem et al. (2022). Participants’ concerns indicate the need for more explicit EM guidance to protect stakeholders.
Both quantitative and qualitative findings confirm the presence of income-increasing and income-decreasing EM across Maltese non-financial MLEs, supporting the findings of Cardona (2003). Regression analysis revealed sector patterns: CD firms showed significance across revenue acceleration, delayed payments, and non-cash adjustments; RE firms showed significance only in revenue-related coefficients; and Industrials firms showed no significance. Descriptive statistics indicated greater volatility in RE firms, while qualitative commentary similarly portrayed complex sectors, including RE and pharmaceuticals, as more susceptible to EM.
Despite these tendencies, the Kruskal-Wallis test indicated no statistically significant differences in DA across industries, contradicting the findings of Wasiuzzaman (2018). This discrepancy may reflect the Yoon et al. (2006) model’s limited explanatory power in specific sectors or the use of alternative EM techniques.
Participants widely acknowledged EM as a tolerated norm, implying that reported earnings may not fully reflect economic performance. Such distortions reduce informational value (Baker et al., 2019; Tahat et al., 2022) and may lead to suboptimal capital allocation, misaligned incentives and agency costs (Armstrong et al., 2025; Jensen & Meckling, 1976). These findings highlight the need for stronger regulatory oversight, enhanced scepticism and specialised auditor training. The tolerance of EM also raises ethical concerns, underscoring threats to faithful representation and MSE credibility.
Participants noted that IFRS’s judgement-based principles, though intended to promote transparency, provide opportunities for EM. This aligns with Adhikari et al. (2021), Callao & Jarne (2010), and Hepworth (1953). Preparers may comply with the letter rather than the spirit of IFRS, undermining credibility and investor trust in line with Bin Khidmat et al. (2018) and Burlacu et al. (2024). These concerns suggest structural issues in standard-setting, where heavy reliance on professional judgement can compromise the integrity of FR.
Corporate-level pressures emerged as the dominant drivers of EM, aligning with Healy and Wahlen (1999). Market expectations were particularly influential, with managers manipulating earnings to meet investor benchmarks and avoid volatility penalties, consistent with signalling theory (Abed et al., 2022; Grougiou et al., 2014; Miller, 2002). This dynamic reduces FR reliability and perpetuates market intolerance toward natural fluctuations.
Contractual pressures—debt covenants and, to a lesser extent, performance-linked bonuses—also motivated EM, supporting agency theory consistent with Almahrog et al. (2016), Armstrong et al. (2025), Ater and Hansen (2020), Bathala and Rao (1995), Eisenhardt (1989), Jensen and Meckling (1976), Ronen and Yaari (2008), Shapiro (2005), Tahat et al. (2022). While personal motivations were less pronounced, systemic organisational pressures indicate deeper governance shortcomings. Effective mitigation, therefore, requires systematic-level reforms that foster long-term value alignment, transparency and stronger board oversight.
Auditors acknowledged conflicting pressures but emphasised adherence to professional standards when misstatements surpass materiality or pose legal risk, consistent with Azad et al. (2023), Carrington (2010) and Ruhnke & Schmidt (2019). Specialised technology also supports audit quality, reducing the risk of overlooking EM, aligning with Infuehr (2022). However, the possibility of social desirability bias suggests the need for continued training and internal safeguards to maintain EM detection effectiveness.
Participants acknowledged Malta’s established regulatory framework but expressed mixed views regarding its effectiveness. While some cited proactive guidance and training, others noted enforcement weaknesses, consistent with Fabri (2016). Concerns also arose around the QAU’s emphasis on administrative compliance rather than substantive judgements where EM risk is highest. Strengthening sector expertise and prioritising substance over form would enhance EM deterrence, consistent with the findings of Ball et al. (2000), Francis and Wang (2008), and Leuz et al. (2003).
Participants highlighted internal controls, CG and AC oversight as central EM deterrents, consistent with Bansal (2024b) and Boachie & Mensah (2022). Internal audit functions also play a key role, consistent with Abbott et al. (2016) and Prawitt et al. (2009)
Although APs affirmed the general effectiveness of current preventive measures, they emphasised the need for stronger AC and director engagement, improved recruitment of skilled personnel, and more consistent application of governance, indicating that existing controls lack sufficient depth and consistency to curb EM effectively. Strengthening these mechanisms is critical to improving earnings quality and maintaining stakeholder confidence in FR within the Maltese context.
Consistent with Czakowska (2020) and Shbeilat (2024), findings indicate that, although ISA 240 is not explicitly designed for EM detection, auditors consistently rely on its principles to guide their evaluation of discretionary accounting choices.
CFO perceptions that non-Big 4 firms provide comparatively weaker assurance support the assertion by DeFond & Zhang (2014) of higher perceived audit quality among Big 4 firms, suggesting vulnerabilities that could be exploited by entities seeking to conceal EM.
Despite robust methodologies, auditors emphasised the inherent intangibility of EM, which complicates its detection and aligns with concerns raised by Luippold et al. (2015) regarding audit management tactics. While EM and fraud share some commonalities, the two practices are conceptually distinct, rendering the reliance on ISA 240 insufficient. These findings reinforce the need for a universal conceptual understanding of the practice and for the development of a dedicated ISA that specifically addresses EM.
6. Conclusion
This study concludes that the earnings metric remains central as a KPI, magnifying the importance of understanding the EM phenomenon in Malta. Findings strongly indicate the presence of EM within non-financial MLEs, as evidenced by variations in the timing of revenue recognition, payment patterns, and non-cash adjustments. Although EM techniques varied across sectors, the overall extent of EM was found to be broadly consistent. Despite its presence, EM remains surrounded by ambiguity, as evidenced by the diverse interpretations of its meaning, purpose, and acceptance, creating grey areas that can be exploited.
The study further concludes that the principles-based nature of IFRS standards creates opportunities for EM, as subjective judgment can be leveraged to serve managerial interest rather than faithful representation. From the auditees’ perspective, the agency problem and signalling theory were evident, with capital forces and contractual obligations as the primary motivators of EM, and corporate-level pressure trumping individualistic incentives. Auditors emphasised the misalignments between management and shareholder objectives to support this. Positively, auditors were found to be deterrents to EM, given their prioritisation of professional standards and integrity over economic pressures.
Moreover, despite the effectiveness of current preventative measures, namely the robust internal controls, effective CG structures and the vigilance of ACs, the findings highlight the need for strengthening the accountability, independence and depth of these bodies’ scrutiny, to effectively challenge management and auditors when earnings quality is not up to standard, given that EM is still slipping past current safeguards. Moreover, while general confidence was expressed in Maltese regulatory bodies, their operational impact was criticised for a lack of consistency and robustness needed to deter complex EM techniques effectively. This was particularly accentuated in the quality assurance visits, which were evidenced to overemphasise compliance rather than scrutinising professional judgement, an area most susceptible to EM.
Lastly, the EA engagement, while validated as a crucial tool for reducing agency costs, demonstrated limitations that hindered its full potential. These include quality gaps between Big-4 and non-Big-4 firms, the absence of a dedicated ISA targeting EM, the use of audit management tactics by auditees, and limited client cooperation.
This study makes several contributions to the accounting, auditing, and financial reporting literature.
First, it contributes new empirical evidence on EM within a small, IFRS-adopting capital market, addressing a notable gap in the literature. While prior EM research has primarily focused on large or emerging markets, evidence from small jurisdictions such as Malta remains limited and outdated. By examining non-financial Maltese listed entities, this study enhances understanding of how EM manifests in concentrated ownership environments with evolving regulatory enforcement, offering insights relevant to other small- and medium-sized capital markets operating under similar institutional conditions.
Second, the study contributes methodologically by employing a sequential explanatory mixed-methods design. The use of the accrual-based model provides a quantitative assessment of discretionary accruals, while in-depth semi-structured interviews with Chief Financial Officers and Audit Partners offer qualitative insights into the motivations, perceptions, and constraints surrounding EM. This triangulated approach advances prior EM research that relies predominantly on archival data by integrating practitioner perspectives, thereby strengthening the interpretability and practical relevance of the findings.
Third, the findings contribute to the IFRS and audit quality debate by highlighting how the principles-based nature of IFRS, while intended to promote faithful representation, creates opportunities for discretionary judgment that may facilitate EM, particularly in judgment-intensive areas. The study further sheds light on the role of external auditors as key deterrents to EM, while also identifying challenges related to audit quality differentials between Big-4 and non-Big-4 firms, client resistance, and the absence of auditing standards explicitly addressing EM. These insights extend the auditing literature by illustrating how professional judgement, materiality, and scepticism operate in practice within smaller audit markets.
Finally, the study offers practical and regulatory contributions. By identifying gaps in current preventive and detection mechanisms, particularly in relation to governance oversight and regulatory quality assurance practices, the research provides evidence-based insights for regulators, standard setters, audit practitioners, and corporate governance bodies. The findings support calls for enhanced scrutiny of judgment-based accounting areas, stronger audit committee engagement, and more substantive regulatory reviews beyond procedural compliance, thereby contributing to ongoing discussions on improving financial reporting quality and investor confidence.
Viewing Malta as a micro-regulatory laboratory highlights how earnings management can persist even in highly visible markets operating under IFRS. The findings suggest that small market size alone does not eliminate discretionary reporting practices; instead, it reshapes their form and detection. Regulators and standard setters in similar jurisdictions may therefore benefit from enhanced scrutiny of judgment-based accounting areas and more substantive quality assurance mechanisms that focus on professional judgment rather than procedural compliance alone.
Conceptualization, M.M. and N.T.; methodology, M.M. and N.T.; validation, P.J.B., N.T.; L.E., and S.G. formal analysis, P.J.B., N.T.; L.E., and S.G. investigation, P.J.B., N.T.; L.E., and S.G. resources, P.J.B., N.T.; L.E., and S.G. data curation, P.J.B., N.T.; L.E., and S.G. writing—original draft preparation, M.M. and N.T.; writing—review and editing, P.J.B. and S.G.; visualization, X.X.; supervision, N.T. and L.E. project administration, P.J.B. and S.G. All authors have read and agreed to the published version of the manuscript.
The data used to support the research findings are available from the corresponding author upon request.
The authors declare no conflict of interest.
Views: 139
Crossref: 0