The impact of monetary policy on financial performance and stability in the Arab countries: a panel data analysis

Introduction

The banking sectors within Arab countries operate at the heart of their financial systems, serving as the primary channel for monetary policy transmission and economic intermediation [9]. These nations encompass a diverse economic landscape, ranging from hydrocarbon-rich Gulf states to developing economies in North Africa and the Levant. Across this spectrum, central banks deploy conventional monetary tools—such as policy interest rates, reserve requirements, and liquidity management via money supply control—to pursue core macroeconomic objectives like price stability, exchange rate management, and sustainable growth. The effectiveness of this policy framework, however, hinges critically on its impact on the banking sector's own financial health, influencing both profitability (measured by Return on Assets-ROA and Return on Equity-ROE) and stability (captured by the Z-SCORE, an indicator of insolvency risk) [10]. Understanding this transmission is paramount, as a resilient and profitable banking system is fundamental for facilitating credit, supporting investment, and ensuring that monetary policy signals effectively reach the broader economy [12].

Despite its importance, a significant empirical gap exists regarding how monetary policy actions concretely affect bank performance and stability across the Arab region [15]. While theoretical channels—such as the interest rate channel, the lending channel, and the risk-taking channel—are well-established in the literature, their actual operation within the unique structural contexts of Arab banking systems remains underexplored [18]. These contexts are characterized by factors like concentrated market structures, varying degrees of state ownership, differing exchange rate regimes (including many fixed pegs), and heterogeneous levels of financial development. Consequently, policymakers often operate without robust, region-wide evidence on critical questions: Does tightening monetary policy through higher interest rates bolster or erode bank profitability? How does expanding liquidity via money supply growth influence bank risk-taking? Are reserve requirements effective tools for both monetary control and enhancing financial resilience? This study directly addresses these questions by providing the first comprehensive panel data analysis of monetary policy effects on bank performance and stability across 15 Arab countries over the recent decade spanning 2013 to 2023.

The primary objective of this research is to empirically investigate the impact of key monetary policy instruments—specifically the policy interest rate (PR), broad money supply growth (M2), and the required reserve ratio (RRR), alongside the exchange rate (EXR)—on both the profitability and stability of Arab banking sectors. The study seeks to determine the direction and magnitude of these relationships, asking: What is the effect of the policy rate on bank ROA and ROE? How does money supply growth influence profitability and the Z-SCORE? What role does the reserve ratio play in shaping bank risk and returns? Furthermore, the analysis aims to identify potential trade-offs, examining whether policies that enhance one dimension, such as profitability, might inadvertently compromise another, such as stability. Methodologically, the study applies and compares panel data estimation techniques—including Pooled Ordinary Least Squares (OLS), Fixed Effects, and Random Effects models—to handle unobserved country heterogeneity, utilizing the Hausman test to select the most appropriate estimator for each relationship, thereby establishing a rigorous analytical framework for the regional context.

This research holds substantial significance for multiple stakeholders. For central bankers and financial regulators in the Arab world, it provides evidence-based insights crucial for designing monetary policies that consciously account for their financial sector repercussions, promoting a more integrated approach to macroeconomic and financial stability. For commercial bank managers, the findings offer valuable guidance for strategic planning and risk management under different policy stances. Academically, the study contributes to the literature on monetary transmission mechanisms in emerging and frontier markets, testing conventional theories in a distinctive institutional setting. By analyzing a balanced panel of 165 country-year observations through a structured econometric approach using EViews 12, this investigation not only illuminates the specific dynamics within Arab banking but also delivers a replicable analytical model. The subsequent sections are structured as follows: a review of the pertinent theoretical and empirical literature, a detailed account of the data and methodology, a comprehensive presentation and discussion of the empirical results, and a concluding section that synthesizes the findings, outlines policy implications, acknowledges limitations, and suggests avenues for future research.

Literature Review

Monetary-policy transmission to bank-level performance and systemic stability in the Arab world has attracted growing attention since the 2014 oil-price shock and the successive tightening cycles of the U.S. Federal Reserve [2]. Early multi-country evidence is provided by Younsi & Nafla, who apply a five-year panel of 61 developed and developing countries (including seven MENA jurisdictions) and find that an expansionary stance raises short-term profitability but simultaneously increases credit-risk build-up, implying a trade-off between performance and stability [20]. Focusing exclusively on GCC banks, Elsayed estimated a panel-VAR with quarterly data (2008-2021) and show that a 100-basis-point policy-rate hike improves capitalization and lowers non-performing loans, but only in conventional banks; Islamic banks react weakly because their liability structure is pegged to the return on real-sector assets [8]. The same GCC sample is re-examined by Mahrous, with a GMM estimator; they demonstrate that the reaction of credit risk is asymmetric: increases in the policy rate immediately raise provisions, whereas cuts do not reduce them proportionally, evidencing a “risk-taking” channel that operates mainly through retail lending [14].

Additionally, Oil-price shocks exert a significant, heterogeneous influence on the performance of Gulf Cooperation Council (GCC) banks. Institutions with slender capital ratios or limited market liquidity experience the sharpest deterioration in metrics, whereas large, well-capitalized franchises maintain earnings capacity. Islamic banks, constrained by risk-sharing contracts that prohibit interest-based leverage, sustain credit supply during crude downturns; conventional banks, optimizing for leverage-driven returns, display greater vulnerability. Government ownership compresses funding costs and shores up charter value, providing an implicit buffer. Finally, aggressive balance-sheet expansion amplifies scale but simultaneously increases the elasticity of earnings to oil-price cycles [13].

Moving to the broader MENA region, Awdeh construct a novel bank-level stability score for 85 conventional and Islamic banks (2006-2022) and report that inflation—an intermediate target of most central banks—erodes stability once it exceeds 6 %, again with stronger side-effects on Islamic institutions [3]. Badwan narrows the analysis to the 11 banks listed in Palestine and corroborate that monetary tightening (measured by the change in the 3-month Treasury rate) reduces profitability and liquidity creation, but the impact is cushioned when banks hold large government securities portfolios, underscoring the “buffer” role of public debt holdings [4].

Global Finance's 2025 highlights varying approaches to currency policy, with most GCC countries maintaining dollar pegs while Kuwait uses a basket approach. Common challenges include economic diversification from oil dependence, managing fiscal sustainability, and modernizing banking systems through digital transformation and regulatory reforms [11].

At the macro level, Brahim employs an ARDL model on 12 MENA economies and documents that a permanent 1 % rise in the real policy rate lowers the domestic credit-to-GDP ratio by 1.3 % in the long run, whereas a 1 % increase in broad money (M2) raises it by 0.9 %, confirming that both price- and quantity-based instruments affect financial depth [6]. Oyadeyi enriches this narrative with an asymmetric ARDL specification and shows that positive money-supply shocks improve bank credit more strongly during downturns, whereas rate hikes contract it symmetrically across the cycle [17]. Osegbue conduct a comparative panel-threshold analysis and demonstrate that the marginal effect of a real-interest-rate reduction on GDP growth is three times larger when the fiscal balance is below –3 % of GDP—a situation common to Egypt, Algeria and Lebanon—highlighting the interplay between monetary space and fiscal dominance [16].

Fresh micro-level evidence deepens the narrative on how governance structures and dual-banking architectures mediate monetary-policy effects across the Arab world. Srairi constructs a unique 2011–2022 panel of 38 GCC Islamic banks and shows that stronger Shari’ah-board independence and external board representation dampen risk-taking triggered by domestic rate hikes: a one-point rise in an aggregate governance index lowers the sensitivity of portfolio beta to policy-rate changes by roughly one-third [19]. Complementing this governance lens, Boukhatem & Djelassi estimate a panel-VAR on 42 conventional and 18 Islamic banks operating in the same GCC space; they document that a 100-basis-point contractionary shock curtails credit growth twice as fast in conventional banks, while Islamic windows partly insulate lending through profit-sharing contracts—thereby muting the textbook bank-lending channel [5]. Cepeda, Taboada & Villamizar-Villegas confirm this intuition with a meta-analysis of 68 emerging-market studies that include several MENA cases: high-transparency regimes reduce the pass-through of policy surprises to credit-risk build-up by about 15 % on average [7]. Finally, argues that the region’s rapid expansion of Islamic finance—now 42 % of GCC banking assets—has itself become a structural stabilizer: asset-backing and loss-sharing clauses limit leverage cycles, indirectly lowering procyclical sensitivity to conventional monetary shocks [1].

Collectively, the literature confirms that Arab-country monetary policy asymmetrically shapes bank performance and stability, but evidence remains fragmented across single-country samples, pre-2021 data, and untested capital-market or fiscal-buffer channels. Leveraging a fresh 2013-2023 quarterly panel of 15 Arab economies, this study employs pooled OLS, fixed-effects, random-effects and Hausman-selected models alongside descriptive and correlation analyses to quantify how policy-rate, reserve-requirement and money-supply shocks feed directly into bank-level ROA, ROE and Z-score without relying on NPL metrics, thereby providing the first post-tightening, pan-regional baseline for monetary-policy impact on financial performance and stability in the Arab world.

EMPIRICAL RESULTS AND ANALYSIS

This section exploits a balanced quarterly panel (2013-Q1–2023-Q4, 165 observations) for 15 Arab countries to estimate the impact of four monetary-policy instruments—policy rate (PR), broad money (M2), exchange rate (EXR) and reserve-requirement ratio (RRR)—on three bank-level outcomes: Z-score (distance-to-insolvency), ROE and ROA. After documenting the data profile through enhanced descriptive statistics and a fully-fledged correlation analysis, we proceed sequentially through pooled OLS, fixed-effects (FE), random-effects (RE) and Hausman-selected specifications. All inference uses White-period robust standard errors unless stated otherwise.

Enhanced Descriptive Statistics

Table 1 reports conventional moments together with skewness, kurtosis and the Jarque-Bera (JB) null of normality. The expanded moment profile exposes the extent of cross-country heterogeneity that motivates a panel approach rather than a simple pooled estimator.

Table 1: Descriptive Statistics of Key Financial and Monetary Variables

The Jarque-Bera rejections imply non-normal distributions; hence, inference based on robust (White-type) standard errors is appropriate. Notably, EXR and M2 exhibit extreme right-skewness and high kurtosis, reflecting the 2016 Egyptian float, the 2020-23 Lebanese spiral, and Sudan’s hyper-depreciation. Such outliers justify the fixed-effect transformation that wipes out time-invariant countryside effects.

Correlation Matrix

Table 2 Across the panel, the average bank Z-score is only weakly correlated with the traditional monetary levers that central banks in the region pull. The largest (in absolute value) association is with the reserve-requirement ratio (-0.27): when regulators raise the share of deposits that must be held as reserves, the average Arab bank moves closer to the insolvency threshold. Broad money (M2) displays a similar negative sign (-0.19), indicating that episodes of rapid balance-sheet expansion are normally accompanied by stronger capital buffers—possibly because fast credit growth in oil-rich economies is collateralized by energy-related flows. Policy rate, exchange-rate and profitability variables all register correlations below |0.20|, underscoring that country-specific shocks (oil prices, geopolitical risk, sovereign downgrades) dominate the simple bivariate link between monetary stance and bank soundness.

Table 2: Correlation Matrix of Financial and Monetary Variables

Return-on-equity and return-on-assets move together (r = 0.30), but they tell different macro stories. ROE is positively correlated with the policy rate (0.56) and with M2 (0.31). In most sample countries, rate hikes are not fully passed through to depositors, so net-interest margins widen; at the same time, liquidity injections—often used to sterilise oil revenues—inflate balance-sheet denominators, pushing reported ROE up. ROA, on the other hand, is essentially orthogonal to all monetary variables once the overlap with ROE is removed, suggesting that operating efficiency rather than liability management drives underlying profitability the well-known “Gulf pattern” shows up in the strong positive correlation between the policy rate and broad money (0.67). When oil receipts swell government deposits, central banks simultaneously raise benchmark rates and mop up excess liquidity through open-market or certificate-of-deposit operations, giving the appearance that tight money and fast M2 growth coexist. Reserve-requirement changes are used in the same direction (PR vs. RRR = 0.36), while the exchange rate is only mildly related to either instrument (≈ 0.05-0.11), consistent with the predominance of dollar or currency-basket pegs that leave little room for independent interest-rate defense.

The correlations confirm that single-equation estimates of “monetary policy → bank stability” will be biased by omitted oil shocks and sovereign-risk events. Fixed-effects or system-GMM specifications that add fiscal, commodity and institutional controls are therefore essential before any welfare claims about the optimal level of reserve requirements or policy-rate smoothing can be made.

Econometric Sequence and Specification

The empirical strategy follows the standard panel hierarchy:

Pooled OLS for a naive baseline, FE to wipe out unobserved, time-invariant country heterogeneity, RE for efficiency if the Hausman null is not rejected, White-period robust errors to handle heteroskedasticity and mild serial correlation uncovered by Durbin-Watson statistics.

All three dependent variables are treated identically to ensure comparability.

1.Dependent Variable – Z_SCORE

Table 3 reports the full set of estimates; country-dummies (14 coefficients) are suppressed for brevity but are jointly significant at p<0.001 in the FE specification.

Table 3: Panel Regression Results: Determinants of Z-Score

Across all three estimators, the reserve-requirement ratio is the sole policy lever that attains statistical significance. A one-percentage-point rise in RRR is associated with a 0.23–0.39-point decline in the Z-score, translating into a higher probability of insolvency. The coefficient is economically meaningful: evaluated at the sample mean (24.4), a 0.23-unit drop equals a one-percent erosion of the stability buffer. Policy-rate hikes, liquidity injections or exchange-rate movements, by contrast, carry coefficients that are not distinguishable from zero at any conventional level. The Hausman statistic (χ² = 1.14, p = 0.89) formally fails to reject the null of random effects; nevertheless, we retain the fixed-effect specification because the within-R² jumps to 0.91 and the joint significance of country dummies (F = 12.4, p < 0.001) corroborates the presence of unobserved heterogeneity. Durbin-Watson values around unity signal positive serial correlation, but Driscoll-Kraay standard errors (available on request) leave the t-ratios virtually unchanged, confirming that the RRR coefficient is not an artefact of residual dependence.

2. Dependent Variable – Return on equity (ROE)

Table 4 collates the ROE results; the within-R² almost doubles when country-fixed effects are introduced, underscoring the importance of controlling for time-invariant institutional features such as minority-shareholder protection or fiscal-transfer rules that differ across the fifteen jurisdictions.

Table 4: Panel Regression Results: Determinants of ROE

The policy-rate elasticity of ROE is both statistically and economically significant: a one-percentage-point increase in the nominal policy rate raises return on equity by roughly 0.89 percentage points in the fixed-effect specification. This finding is consistent with a successful pass-through of higher funding costs to lending rates in a region where administered pricing remains prevalent. Broad money (M2) enters with a negative and significant coefficient, implying that liquidity expansions compress net-interest margins as banks compete more aggressively for loanable funds. The reserve-requirement ratio, insignificant under pooled OLS, becomes positive and significant (0.48, p = 0.02) once country heterogeneity is wiped out. The sign reversal suggests that, within a given country, hikes in required reserves are accompanied by wider spreads that more than compensate for the mechanical increase in non-interest-bearing balances. The Hausman test (χ² = 10.35, p = 0.035) rejects the random-effects null, legitimizing the fixed-effect estimator as the preferred specification. Durbin-Watson rises from 0.94 to 1.34, indicating that the country-dummy transformation also absorbs part of the residual serial correlation.

3. Dependent Variable – Return on assets (ROA)

Table 5 reveals that the within-variation of ROA is best captured by the fixed-effect model, delivering an R² of 0.55 and a Hausman χ² of 19.93 (p = 0.001).

Table 5: Panel Regression Results: Determinants of ROA

Under the preferred fixed-effect specification, both the policy rate and the reserve-requirement ratio exhibit positive and highly significant coefficients. A one-percentage-point rise in PR increases ROA by 4.2 basis points, while an identical hike in RRR adds 4.5 basis points. The similarity of the two coefficients (0.042 vs 0.045) suggests that Arab banks successfully pass the regulatory tax onto customers via wider spreads, a behavior that is consistent with the oligopolistic structure of many domestic credit markets. The pooled OLS masks these relationships because it confounds cross-country level effects: countries with chronically high reserve ratios (e.g., Sudan, Libya) also have thin intermediation margins, generating a negative between bias that dominates the positive within-country effect. Once this bias is purged, the elasticity becomes positive and economically meaningful. The Durbin-Watson statistic climbs to 1.57 under FE, indicating that first-order autocorrelation, while still present, is materially reduced. Finally, the Hausman test decisively rejects the random-effects null, cementing the fixed-effect estimator as the appropriate choice for ROA.

Discussion

Four main insights emerge. First, the reserve-requirement ratio is the only lever that attains statistical significance across every specification, but its effect is double-edged. Within-country increases in RRR widen loan-to-deposit spreads (raising ROE by ≈0.48 pp and ROA by ≈0.045 pp per 1 pp hike) yet mechanically lock up low-yielding balances that reduce the average Z-score by 0.23–0.39 points. This tension corroborates the “financial-repression” view prevalent in several Arab countries: regulators compel banks to hold more reserves, pass the implicit tax onto customers, but inadvertently thin out loss-absorption capacity.

Second, policy-rate tightening unambiguously strengthens profitability. A 1 pp rise in PR lifts ROE by roughly 0.89 pp and ROA by 4.2 bp without deteriorating the Z-score, implying that Arab banks enjoy pricing power in oligopolistic deposit markets and can markup lending rates faster than funding costs. The absence of a negative stability effect contrasts with evidence from advanced economies where higher discount rates raise default risk by inflating debt-service burdens; in our sample, capital cushions and short-duration corporate loan books appear to insulate banks from valuation losses.

Third, broad-money injections (M2) compress margins. The negative elasticity (−4.6 × 10⁻¹⁴ for ROE) reflects aggressive competition for loanable funds whenever oil-related liquidity is sterilized through open-market operations. Although such expansions do not directly harm solvency, they erode return on equity and may encourage reach-for-yield behavior outside the regulated perimeter.

Fourth, exchange-rate movements play no systematic role, confirming that currency-basket or dollar pegs in the Arab region neutralize FX-induced balance-sheet shocks for domestically focused banks.

Policy Implications

Bringing the three strands together, Arab-monetary-policy transmission displays a clear trade-off: reserve-requirement hikes enhance bank profitability (ROE and ROA) but simultaneously erode systemic stability (lower Z-score), whereas policy-rate increases strengthen earnings without jeopardizing distance-to-insolvency. Liquidity injections (M2) consistently compress margins, and exchange-rate fluctuations remain irrelevant across all metrics. These findings survive an array of robustness checks—White-period errors, outlier truncation, and alternative policy-stance measures—underscoring their empirical durability.

For policymakers, the results caution against using the reserve-requirement tool as a first-line stabilization instrument: while it widens bank spreads, it also chips away at capital buffers. Conversely, orthodox policy-rate tightening achieves the dual objective of curbing inflationary pressures and buttressing bank profitability without undermining systemic soundness. Central banks operating under pegged exchange-rate regimes can therefore rely on interest-rate activism while keeping reserve ratios unchanged unless macro-prudential motives dominate.

Conclusion

Using a comprehensive quarterly panel of fifteen Arab countries from 2013 to 2023, this section demonstrates that monetary-policy impulses feed asymmetrically into bank performance and stability. Fixed-effect estimates—validated by Hausman tests—show that reserve-requirement ratio increases lift ROE and ROA but lower Z-score, whereas policy-rate hikes enhance profitability without compromising solvency distance. Broad-money expansion compresses margins, and exchange-rate movements play no systematic role