The impact of internal audit system on performance: evidence from emerging markets

The study investigates how internal audit (IA) influences commercial bank performance, positioning IA as a core governance mechanism that strengthens internal control, risk management, and reporting quality. IA is defined by the IIA as an objective, independent function that adds value and improves operations. The introduction situates IA’s importance historically and practically, highlighting its role in ensuring reliable financial information, supporting strategy execution, and enhancing organizational sustainability. It notes unique contextual challenges in Arab countries (cultural, political, religious features, e.g., Islamic finance constraints) and a limited empirical understanding of IA’s role in these settings. The study aims to open the “black box” of IA by examining multiple IA factors together and assessing their direct and mediated effects on bank performance, thereby informing regulators and practitioners in emerging markets.

The paper integrates resource-based theory (RBV) to argue that IA capabilities (standards, independence, competence, organizational support) are strategic resources that enhance performance. It also draws on institutional theory (compliance with international IA standards), agency theory (mitigating information asymmetry via competent and independent auditors), and social categorization theory (gender diversity effects). Hypotheses developed: H1: Implementation of IA standards positively impacts bank performance, grounded in literature linking standard compliance to improved control, transparency, and risk identification. H2: IA independence/objectivity enhances performance, as independence is essential to audit quality and governance effectiveness. H3: Professional competence of internal auditors positively affects performance; prior findings are mixed, with some contexts reporting weak or no effect due to structural constraints. H4: Female representation in IA departments positively affects performance; literature is mixed, but several studies suggest positive impacts on reporting integrity and control quality. H5: IA size and frequency of meetings positively impact performance; evidence is mixed but suggests larger, more active IA units can improve effectiveness. Additional model paths: H6: Top management support for IA positively impacts performance. System-level hypotheses: H7: The internal audit system (IAS) positively affects internal auditors’ performance. H8: IAS directly and positively affects bank performance. H9: IA performance mediates the IAS–performance relationship. The framework (Fig. 1) reflects these relationships.

Design: Quantitative, cross-sectional survey using a closed-ended questionnaire (5-point Likert scale: 1=strongly disagree to 5=strongly agree). Measurement items adapted from prior literature with minor modifications; constructs include IA standards, independence/objectivity, professional competence, female representation in IA, IA size and meeting frequency, top management support, IA performance, and bank performance dimensions (profitability efficiency, social disclosure efficiency, operational efficiency). Demographics captured: gender, education, job role, experience, tenure at bank. Sample and data collection: Targeted 34 commercial banks across Saudi Arabia, Yemen, Sudan, Somalia, and Djibouti. Distributed 242 questionnaires to 170 branches via Google Forms and phone follow-up; received 172 responses. Sample deemed adequate for PLS-SEM. Respondent profile (n=172): 76.16% male, 23.84% female; education: 74.42% bachelor, 16.28% master, 9.30% diploma; roles: auditors 56.98%, accountants 13.95%, finance staff 25.58%, regulators 3.49%; experience mainly 3–6 years (67.44%). Analysis: Partial Least Squares Structural Equation Modeling (PLS-SEM) using SmartPLS 3.3.3. Assessed measurement model (indicator loadings mostly >0.60 after item trimming; AVE ≈0.50+ across constructs; CR 0.749–0.870; Cronbach’s alpha ≥0.60). Discriminant validity via Fornell–Larcker, cross-loadings, and HTMT (<0.90). Multicollinearity diagnostics: VIFs <3.3 and inter-construct correlations <0.80. Structural model assessed with bootstrapping for path significance, R2, f2, Q2; overall GoF computed as √(mean AVE × R2). Also tested a mediation model where IA performance mediates the IAS–bank performance link.

- Measurement validity and reliability met conventional thresholds: AVE generally >0.50; CR 0.749–0.870; CA ≥0.60; HTMT <0.90; VIF <3.3. - Direct effects on bank performance (FP) from Table 5: • H1 (IA standards -> FP): β=0.176, t=2.755, p=0.006 (accepted). • H2 (Independence/objectivity -> FP): β=0.164, t=2.325, p=0.020 (accepted). • H3 (Professional competence -> FP): β=0.036, t=0.489, p=0.625 (not accepted). • H4 (Female representation in IA -> FP): β=0.298, t=5.166, p<0.001 (accepted). • H5 (IA size and meeting frequency -> FP): β=0.139, t=2.044, p=0.041 (accepted). • H6 (Top management support -> FP): β=0.205, t=2.424, p=0.015 (accepted). - Model explanatory power: R2 (FP)=0.477 (moderate). Predictive relevance: Q2=0.142. Global GoF ≈0.507 (high per study’s benchmark). - Mediation (Table 6): • IAS -> IA performance (IAP): β=0.385, t=4.642, p<0.001 (significant). • IAP -> FP: β=-0.024, t=0.389, p=0.697 (not significant). • IAS -> FP (direct): β=0.671, t=11.980, p<0.001 (significant). • Indirect effect (IAS -> IAP -> FP): β=-0.009, t=0.369, p=0.712 (not significant). Thus, no mediating effect of IAP. - Overall: All proposed IA dimensions significantly and positively affect bank performance except professional competence; IAS directly improves both IA performance and bank performance, but IA performance does not mediate the IAS–performance link.

The findings confirm that multiple facets of the internal audit system—adherence to international IA standards, auditor independence/objectivity, gender diversity within IA, adequate IA size with frequent meetings, and strong top management support—collectively and directly enhance bank performance in emerging Arab markets. This supports RBV by showing IA-related capabilities as valuable organizational resources. The strong positive effect of female representation suggests diversity improves oversight, reporting integrity, and internal controls, addressing mixed prior evidence and adding novel insights in culturally conservative contexts. Conversely, professional competence did not significantly impact performance, which the authors attribute to contextual constraints (e.g., family ownership, political interference, limited investment in IA) that may inhibit competent auditors from translating skills into performance gains. The non-significant mediation via IA performance implies that IAS exerts predominant direct effects on performance rather than operating through perceived auditor performance. For regulators and bank leaders, the results underscore the importance of institutionalizing IA independence and standards, investing in IA capacity and diversity, and ensuring top-level support to realize performance benefits.

The study advances understanding of how internal audit systems affect commercial bank performance in five Arab countries. It provides evidence that compliance with IA standards, auditor independence/objectivity, gender diversity, appropriate IA size and meeting frequency, and strong top management support significantly improve performance, while professional competence alone does not show a significant effect in this context. IAS also directly enhances both internal auditors’ performance and bank performance, though IA performance does not mediate the IAS–performance relationship. The paper contributes by integrating multiple IA dimensions within a unified framework and offering cross-country evidence from an under-researched region. Future research should expand to more countries and sectors, incorporate additional contextual factors (political systems, cultural differences), and conduct comparative studies across regions (e.g., Europe, Africa) to examine how institutional environments shape IA effectiveness and performance impacts.

- Geographic and sectoral scope limited to 34 commercial banks in five Arab countries (Saudi Arabia, Yemen, Sudan, Somalia, Djibouti), which may constrain generalizability. - Cross-sectional, survey-based design using closed-ended questionnaires may introduce common method bias and limits causal inference. - Focused on commercial banks only; results may differ for other financial and non-financial institutions. - Measured a defined set of IA factors; omitted contextual variables (e.g., political systems, cultural nuances) that may affect IA effectiveness and performance.