The Effect of Regulatory Liquidity Measure on Bank Capital Structure

Abstract

This study investigates the effects of liquidity regulation, specifically the liquidity coverage ratio (LCR), on the capital structure of South African banks, with a focus on debt maturity composition. Using panel data covering the period 2015–2024, the analysis applies the Generalized Method of Moments (GMM) estimator to address potential endogeneity concerns. The findings reveal a significant positive relationship between LCR and banks’ total and long-term debt ratios, indicating a shift towards more stable funding structures. In contrast, the LCR is negatively associated with short-term debt. These results suggest that stricter liquidity requirements encourage banks to rely less on short-term funding and more on long-term debt instruments. Although the analysis is limited to a small sample of leading South African banks, the findings provide important insights into the structural implications of liquidity regulation. The study highlights the need for regulators to consider how liquidity requirements shape banks’ financing decisions within broader macroprudential frameworks. By promoting stable funding structures, liquidity regulations enhance banking sector resilience, protect depositors, and support sustainable credit provision. This study contributes novel evidence from an emerging market and addresses a gap in the post-crisis financial regulation literature by linking liquidity regulation to debt maturity profiles.

1. Introduction

The COVID-19 pandemic and the global financial crisis of 2007–2009 revealed major shortcomings in banks’ funding structures and liquidity management procedures, demonstrating how a significant reliance on unpredictable short-term funding and excessive maturity transformation can greatly increase systemic risk (Brunnermeier and Oehmke 2013; Gorton et al. 2010). Recognising that market-based mechanisms alone are insufficient to guarantee a reliable and strong supply of liquidity during periods of financial stress, international regulators have prioritized liquidity risk as a crucial component of the financial stability agenda (Allen and Gale 2018). Therefore, the post-crisis banking regulation has increasingly focused on the relationship between liquidity ratios and the balance sheet structure of banks, especially how liquidity ratios affect the allocation of capital (Hellwig and Admati 2014). While capital adequacy and liquidity regulation are considered to be complementary tools of prudential regulation, the joint impact of these two regulations on banks’ capital structure decisions is still not well understood, particularly in emerging market economies, where the institutional, structural, and macroeconomic factors can affect the transmission channels of regulations (Brůha and Kočenda 2018; Wu and Xia 2016).

As a reaction to the global financial crisis, the Basel Committee on Banking Supervision (BCBS) established two reinforcing liquidity benchmarks as part of the Basel III framework: the liquidity coverage ratio (LCR) and the net stable funding ratio (NSFR). The LCR is designed to ensure sufficient liquidity in the near term, while the NSFR is intended to address underlying funding vulnerabilities by encouraging more robust and durable long-term financing practices (BCBS 2010, 2014; Husodo et al. 2024). Ultimately, these regulations aim to reduce the impact of abrupt deposit outflows and unforeseen liquidity disruptions arising from off-balance-sheet activities.

Banks play a crucial role in the financial system by channelling funds from those with excess capital to those in need, acting as essential financial intermediaries (Obadire et al. 2023). Consequently, oversight of banks’ capital and funding decisions is paramount for mitigating systemic risk and fostering financial system stability (Obadire et al. 2023). As the foremost global entity responsible for overseeing banking practices, the Basel Committee on Banking Supervision (BCBS) has formulated the Basel Accords, currently in their third version, to mitigate weaknesses in banks’ capital frameworks, including excessive reliance on debt and short-term obligations. Basel III introduced more rigorous capital definitions and improved quality benchmarks, thereby directly affecting how banks construct their capital composition (Obadire et al. 2023). Although considerable academic attention has been given to the connection between regulatory capital and the generation of liquidity (for example, Umar et al. 2018; Mazioud Chaabouni et al. 2018; Gupta et al. 2023), comparatively little empirical investigation has focused on the inverse relationship: the effects of liquidity rules on banks’ capital structures. This research aims to fill this void by examining the influence of the liquidity coverage ratio on the capital structure of South African banks from 2015 to 2024. Notably, empirical studies that concurrently assess the regulatory metric and the liquidity coverage ratio remain scarce, particularly in the context of developing economies and across differing debt maturity structures (Mdaghri and Oubdi 2022).

This study addresses a critical gap in the banking and financial regulation literature by shifting the analytical focus from the conventional perspective, in which capital is treated as a determinant of liquidity. Here, the focus is on a relatively underexplored reverse relationship, in which liquidity regulation shapes banks’ capital structure decisions. Existing studies largely emphasise how capital adequacy influences liquidity creation, risk-taking, and financial stability (Berger and Bouwman 2009; Distinguin et al. 2013), thereby overlooking the role of binding liquidity constraints introduced under the Basel III framework. Although emerging research has begun to examine liquidity regulation, it predominantly focuses on lending behaviour, asset allocation, and bank performance rather than on how instruments such as the liquidity coverage ratio (LCR) affect leverage, funding composition, and maturity transformation (Banerjee and Mio 2018). Consequently, the mechanisms through which liquidity requirements influence capital structure adjustments such as shifts toward more stable but costlier funding sources remain insufficiently theorised and empirically tested. By explicitly modelling liquidity regulation as a determinant of capital structure in this context, this study extends the existing literature and provides a more comprehensive understanding of bank behaviour in the post-crisis regulatory environment.

The study seeks to address the following research question:

• What is the relationship between the liquidity coverage ratio (LCR) and the capital structure of banks?

This paper examines the impact of liquidity regulation on the capital structure decisions of banking institutions and makes three primary contributions to the existing literature. First, the study advances theoretical discourse by integrating financial regulation with capital structure theory. Specifically, it investigates how key liquidity standards, namely the liquidity coverage ratio (LCR), influence banks’ capital structure decisions. In doing so, it addresses a notable gap in the literature, which has predominantly focused on the effect of capital on liquidity, rather than examining the reverse relationship.

Second, the study provides novel empirical evidence from South Africa, an emerging economy that adheres to international regulatory frameworks while operating under distinct structural and macroeconomic conditions. By analysing how banks adjust their leverage, funding strategies, and asset–liability management in response to liquidity requirements, the findings offer valuable insights for policymakers and regulatory authorities in comparable emerging market contexts.

Third, the study employs a comprehensive panel dataset spanning the period 2015 to 2024 to evaluate the effects of LCR on bank capital structures. This longitudinal approach enables the identification of potential unintended consequences of liquidity regulation, including a shift towards higher-cost funding and adjustments in risk management practices. Consequently, the study contributes to a deeper understanding of the broader implications of liquidity regulation for financial stability, credit provision, and the mitigation of systemic risk.

Ultimately, this research improves our comprehension of the interplay between liquidity regulation and capital structure choices. Furthermore, it identifies potential unforeseen outcomes, such as a transition towards costlier funding sources or changes in asset-liability management strategies, with wider ramifications for the formulation of regulatory policies. The findings aim to inform policymakers and regulatory bodies in refining liquidity standards to effectively minimise systemic risks without hindering banks’ fundamental intermediation activities.

The remainder of this article is organised as follows: Section 2 offers a review of the relevant literature concerning liquidity regulation and capital structure. Section 3 details the data sources and methodological framework used in the analysis. Section 4 presents and discusses the empirical results obtained. Finally, Section 5 concludes the paper and provides relevant policy recommendations.

2. Literature Review

2.1. Theoretical Framework

This research is based on the three pertinent theories of capital structure that describe how bank financing decisions work. The three underlying theories are the pecking order theory, the agency cost theory, and the trade-off theory. The examination of these theories clarified and offered insight into how financial services companies finance themselves. The trade-off hypothesis states that the optimal capital structure is a debt-to-equity ratio that maximises the company’s worth. As Lerner and Flach (2022) state, interest costs are tax-deductible. Dividends from shareholder equity, on the other hand, are not. In particular, interest on debts owed to third parties is the source of this, and Modigliani and Miller (1963) observed that a business’s income tax obligations decrease with its level of leverage. Consequently, companies seek the optimal debt-to-income ratio while balancing tax benefits and the costs associated with financial challenges. Myers and Majluf (1984) shows in his analysis of the trade-off theory that its value increases as a firm’s debt levels increase. However, financial difficulties become increasingly costly as debt accumulates, until the firm’s value is maximised (Lerner and Flach 2022).

The pecking order theory, which asserts that management favours using internal funds over raising external ones, was introduced by Myers and Majluf (1984). According to the pecking order, a company would rather use its funds than borrow money (Barclay and Smith 2020). In essence, companies issue external debt, issue stock money from outside sources, and use their funds first. The pecking order hypothesis, according to Myers and Majluf (1984), does not establish an optimal or clear level of debt. Additionally, it is evident from the pecking order hypothesis that when a company’s internal resources are inadequate to meet its investment needs, it will turn to outside finance more frequently (Shyam-Sunder and Myers 1999). The pecking and trade-off models both explain the financial conduct of specific companies, and neither can be proven incorrect (Fama and French 2002). Yet, there is no universal theory of capital structure, and there is no justification for assuming that all models of capital structure are restricted as stated by Myers (2003).

Furthermore, firm managers might not always operate in the best interests of firm owners, according to Jensen and Meckling’s (1976) agency cost theory. Thus, companies use more debt capital to make sure that managers’ actions are in line with shareholders’ interests. This is due to the fact that managers’ motivations, as well as their investment and operational choices, are directly impacted by the debt financing decision (Barclay et al. 2003).

The theoretical foundation of this study is based on the bidirectional causality between solvency and liquidity, as discussed by Diamond and Rajan (2001, 2005). They argue that banks face vulnerabilities not only from leverage but also from aggregate liquidity shortages. Such issues occur when banks struggle to meet short-term obligations, despite being fundamentally solvent. Consequently, bank leverage and solvency can be influenced by exacerbating liquidity shortages, creating a feedback loop that drives further adjustments to the gearing ratio. As a result, banks tend to maintain excess liquidity for several regulatory and theoretical reasons. Regulatory requirements mandate that banks maintain specified levels of the liquidity coverage ratio (LCR) and the net stable funding ratio (NSFR). In some cases, authorities also want banks to demonstrate they can handle unfavourable situations, which leads them to maintain more liquidity than is required. In practice, banks employ their internal liquidity stress models to assess their risks differently than regulators do. As a result, the outputs of their stress-testing models will be used to determine the appropriate liquidity levels for each bank. The cautious incentives, market frictions, asymmetric information, and safety trade-offs are among the theoretical explanations for this excess liquidity. Therefore, there is a strong need to examine the relationship between banks’ gearing ratios and regulatory liquidity requirements.

The traditional capital structure theories remain relevant to banks despite stringent regulatory frameworks, because regulatory compliance does not eliminate the underlying trade-offs and behavioural responses that shape banks’ financial decisions. While liquidity requirements and safeguards such as those introduced under Basel III and national deposit protection schemes are designed to reduce the likelihood of bank runs and enhance depositor confidence, they simultaneously impose binding constraints on balance sheet management. In practice, banks actively manage liquidity not only to meet regulatory thresholds, such as those introduced under Basel III and national deposit protection schemes, designed to reduce the likelihood of bank runs and enhance depositor confidence, like the Liquidity Coverage Ratio (LCR) and Net Stable Funding Ratio (NSFR), but also to optimise funding costs, profitability, and risk exposure. These regulatory pressures can induce adjustments in leverage, encourage substitution between short-term and long-term funding, and incentivise shifts toward more stable yet potentially costlier sources of finance. Consequently, liquidity management is not merely a compliance exercise but a strategic determinant of capital structure choices.

2.2. The Effects of Regulatory Liquidity Measures on Bank Capital Structure and Hypothesis Development

The financial literature describes bank capital structure as the result of the interaction among specific firm characteristics, market factors, and regulatory limitations, rather than as isolated influences. Various studies highlight key factors, including profitability, asset risk, size, and regulatory capital requirements. However, the effects of these factors often vary based on broader institutional and macroeconomic conditions (Berger et al. 1995; Gropp and Heider 2010; Sibindi and Makina 2018). It is important to understand that banks do not simply react to changes in leverage; instead, they actively optimise their capital buffers in response to market discipline and regulatory demands.

From a theoretical perspective, the role of capital extends beyond merely providing funding; it also includes aligning incentives and reducing risks. Berger et al. (1995) demonstrate that higher levels of capital can help mitigate moral hazard by aligning the interests of shareholders, managers, and regulators, thereby encouraging responsible risk-taking. The effectiveness of regulatory frameworks, such as the Basel Accords, largely depends on their design. If the requirements are not properly calculated, they can lead to unintended consequences, such as regulatory arbitrage or poor capital allocation. This dual role of capital as both a stabilising buffer and a potential constraint creates a foundation for conflicting empirical predictions.

Empirical evidence on traditional determinants further highlights this conflict. In line with the pecking order theory, Gropp and Heider (2010) find a negative relationship between profitability and leverage, suggesting that banks prefer to use internal funds when available. At the same time, they find that banks with higher risk maintain elevated capital ratios, suggesting that market forces and regulatory oversight drive the accumulation of precautionary capital. Moreover, the negative relationship between bank size and capital buffers implies that larger banks, due to implicit assurances and better funding access, can operate with reduced capital reserves. Overall, these findings indicate that capital structure decisions are influenced by trade-offs among risk, profitability, and market access, reinforcing the notion that bank-specific characteristics consistently affect leverage decisions.

Although these factors largely mirror those in non-financial companies, their impact is heightened in the banking industry due to regulatory oversight and concerns about systemic risk. Sibindi and Makina (2018) affirm that traditional firm-level factors still matter for banks, but their research also emphasizes that capital structures are sensitive to systemic shocks, such as the global financial crisis from 2007 to 2009, during which banks reduced leverage to regain stability. This procyclical modification underscores the importance of regulatory liquidity frameworks, particularly the liquidity coverage ratio (LCR) and net stable funding ratio (NSFR), which may require banks to realign their capital and liquidity positions simultaneously. Thus, capital structure cannot be examined separately from liquidity factors.

An expanding body of research focuses on the relationship between capital and liquidity, highlighting two contrasting theoretical frameworks: the risk-absorption hypothesis and the financial fragility crowding-out hypothesis. The risk-absorption hypothesis asserts that higher capital levels enhance banks’ capacity to generate liquidity by improving their resilience to financial shocks. In contrast, the financial fragility crowding-out hypothesis suggests that excessive capital can hinder liquidity creation by reducing banks’ incentive to convert illiquid assets into liquid ones.

Empirical evidence provides mixed support for these conflicting perspectives on the relationship between capital and liquidity creation. On one hand, research by Umar et al. (2018) and Mdaghri and Oubdi (2022) indicates a negative relationship, aligning with the crowding-out effect. On the other hand, Mazioud Chaabouni et al. (2018) show that banks with strong capital are better positioned to provide liquidity during stressful times, reinforcing the risk-absorption perspective.

Recent studies have attempted to resolve these contradictory findings by highlighting non-linear relationships. For instance, Gupta et al. (2023) identify a U-shaped relationship in developed economies, while an inverted U-shaped relationship is observed in developing markets. This suggests that the impact of capital on liquidity formation varies with capitalization levels and institutional context.

The results suggest a dynamic, non-linear relationship between capital and liquidity. Lower capital levels can enhance liquidity creation by stabilizing banks (the risk-absorption effect). However, beyond a certain threshold, higher capital may hinder liquidity by reducing banks’ incentives for maturity transformation (the crowding-out effect). This underscores the non-linear nature of the relationship between capital and liquidity. This paper tests the following research hypothesis:

H₁: 

There is no significant relationship between LCR and bank capital structure.

2.3. The South African Banking Environment

The banking landscape in South Africa stands out for its robust regulatory systems, high institutional concentration, and integration with the global financial network. The country, which adheres to Basel standards, has adopted the LCR and NSFR, with oversight of their implementation by the South African Reserve Bank (SARB) (2021). As a result, major South African banks have adjusted their funding strategies and liquidity management to align with these international benchmarks. Nevertheless, several core challenges within the local economy could impact the effectiveness of these liquidity regulations. These challenges include economic uncertainty, exchange rate fluctuations, limited capital markets, and a scarcity of long-term financing options. These factors make South Africa unique compared to developed countries and may affect how local banks respond to liquidity rules.

Obadire et al. (2023) suggest that local factors, such as market maturity and institutional strength, often affect the effectiveness of regulations in developing countries. Since there is little research on this, the real-world consequences of LCR compliance for capital structure decisions in South Africa remain largely speculative. The importance of exploring this reverse connection, how liquidity constraints influence capital structure, is highlighted by Mdaghri and Oubdi (2022), especially in settings where financial markets are not fully developed, and banks face significant operational hurdles.

2.4. Identifying a Research Gap

While there have been many studies on how capital adequacy relates to bank liquidity, little attention has been given to how liquidity regulations influence banks’ capital structure decisions. This is particularly evident in developing markets, where local institutional factors and market imperfections can skew regulatory choices. Specifically, there is a lack of empirical analysis concerning the impact of the LCR on how South African banks structure their capital. This study intends to fill that void by providing an in-depth look at how Basel III’s liquidity regulations shape capital structure decisions within South Africa’s banking sector. Through this examination, it aims to enhance our understanding of regulatory changes following the financial crisis and their broader implications on banks’ financial behaviour.

3. Data and Materials

This paper focuses on 16 regulated local banks operating in South Africa. Yet, the sample examined in this article is drawn from the group of these 16 banks, comprising 11 South African-registered banks spanning 2015 to 2024. Due to hurdles in obtaining financial information for the mentioned period, the five smaller banks were not included. These banks are believed to provide a reliable reflection of South Africa’s licensed banking sector during the 2015 to 2024 timeframe. By doing so, this study contributes to a deeper understanding of the regulatory changes that followed the financial crisis and their broader impact on banks’ financial behaviour. The list of South African banks approved under Bank Act 94 of 1990, as of 31 December 2020, was sourced from the South African Reserve Bank’s website. Moreover, financial and economic data were regularly gathered from the South African Reserve Bank (SARB) on a monthly and yearly basis. Over a decade, this study collected information from 11 banks, resulting in 110 entries for the banking dataset. As mentioned earlier, the procedures of the licensed banks under review varied, though the focus was solely on banks authorised within South Africa.

Like other researchers, this study uses three key indicators of capital structure as dependent variables: the overall debt ratio, the long-term debt ratio, and the short-term debt ratio (Mabandla and Marozva 2025a; Mohammad 2022). According to Rajan and Zingales (1995), the ratios of short-term debt, long-term debt, and total debt to total assets are more effective measures of financial leverage compared to the ratio of liabilities to total assets. They provide a clearer understanding of the firm’s potential to face financial difficulties soon and offer a more accurate reflection of its past financing strategies. The liquidity coverage ratio (LCR) was formulated to ensure that participants in the banking sector hold adequate liquid assets during a hypothetical 30-day period of significant stress on the liabilities side of the balance sheet (Roberts et al. 2018).

Table 1. Summary of variables and proxies.

Variables	Proxies and Definitions	Proxies by	The Expected Sign of the Coefficient
Capital structure proxies (Dependent variable)
Total debt ratio at book value (TDRB)	TDRB represents the proportion of the firm’s total debt compared to its total assets, using their book values.	Mohammad (2022), Lazarus et al. (2024)
The long-term debt ratio (LTDR)	LTDR represents the proportion of long-term debts in comparison to the overall assets.	Mabandla and Marozva (2024) and Lazarus et al. (2024)
The short-term debt ratio (STDR)	STDR represents how much of a firm’s assets are tied up in short-term debts.	Mabandla and Marozva (2025b) and Lazarus et al. (2024)
Independent variables
Liquidity coverage ratio (LCR)	$LCR={\displaystyle \frac{High quality liquid assets}{Cash outflows-Cash inflows}}$	Sidhu et al. (2022) and Vu (2024)	Negative
Control variable’s
Economic growth measured by Gross Domestic Product (GDP)	GDP: The growth rate of Real Gross Domestic Product.	Zeitun et al. (2017) and Khan et al. (2023)	Positive or negative
Inflation rates	Annual Consumer Price Index (CPI).	Khan et al. (2023) and Kebede (2024)	Positive or negative
Interest rates	Effective interest rate.	Karpavičius and Yu (2017)	Negative
Size	Size—the natural logarithm of total assets.	Bandyopadhyay and Barua (2016) and Obadire et al. (2023)	Positive or negative
COVID-19	Dummy variable, 1 for the COVID-19 period and 0 for the non-COVID-19 period.	Mabandla and Marozva (2025a)	Negative

Source: Researchers’ own compilation.

below provides the details of the dependent and independent variables and their data sources.

Model Specification

The research employed the Generalised Method of Moments (GMM). The typical dynamic GMM approach is structured as follows:

$$Y_{it}={\alpha y}_{i t-1}+{\beta X}_{it}+{\mu }_i+{\epsilon }_{it}$$

(1)

where

$Y_{it}$ denotes the book value of capital structure indicators for bank i at time t; x represents the vector of explanatory variables for bank i at time t, indicating the bank-specific variable; ${\alpha }_0$ indicates an invariant term; $\beta$ is the elasticity of the independent factors, namely, the gradient of these variables; ${\mu }_i$ represents fixed effects within financial institutions; ${\epsilon }_{it}$ is a stochastic disturbance term; the subscript i denotes the cross-sectional dimension; and t signifies the temporal sequence scale.

This study employs the two-step system Generalised Method of Moments (GMM) estimator developed by Arellano and Bover (1995) and Blundell and Bond (1998), which extends the earlier difference GMM framework of Arellano and Bond (1991). The choice of this estimator is motivated by the dynamic nature of bank capital structure and the need to address potential endogeneity arising from the interaction between leverage and liquidity regulation. The estimation relies on internal instruments constructed from lagged values of the endogenous variables, specifically lagged levels used as instruments for equations in first differences and lagged differences used as instruments for equations in levels. This instrumenting strategy exploits the panel structure of the data and allows the model to control for simultaneity and dynamic persistence. In addition, the application of forward orthogonal deviations (Helmert transformation) removes unobserved bank-specific effects while preserving the orthogonality conditions between the transformed error term and the lagged regressors, thereby ensuring the validity of the moment conditions and improving estimation efficiency.

The treatment of endogeneity is central to the empirical strategy, particularly given the likelihood of reverse causality between leverage and liquidity measures such as the liquidity coverage ratio (LCR). To address this, the study classifies variables into endogenous, predetermined, and strictly exogenous categories, with LCR and selected macroeconomic variables treated as endogenous and instrumented using their appropriate lags. Year dummies are specified as strictly exogenous and are not instrumented, consistent with standard practice. The validity of the internal instruments and the imposed exclusion restrictions is assessed using the Hansen and Difference-in-Hansen tests, following the approach outlined by David Roodman (2009), while also ensuring that instrument proliferation is controlled to maintain the reliability of the estimates. Through this approach, the two-step system GMM estimator effectively mitigates biases arising from omitted variables, measurement error, and simultaneity, thereby producing consistent and robust parameter estimates.

Thus, the two-step system GMM was employed to examine the impact of liquidity regulation on capital structure. This article exclusively utilised South African data as it constituted the primary focus of our investigation. The research examined the critical factors affecting leverage within the South African banking sector by regressing leverage (TDR, STDR, and LTDR) against the variables outlined in question 2.

$$\Delta {\mathrm{C}\mathrm{S}}_{\mathrm{i}\mathrm{t}}=\left(1-\mathsf{\alpha }\right)\Delta {\mathrm{C}\mathrm{S}}_{\mathrm{i}\mathrm{t}-1}+{\mathsf{\beta }}_1{\Delta \mathrm{L}\mathrm{I}\mathrm{Q}}_{\mathrm{i}\mathrm{t}}+{\mathsf{\beta }}_{\mathrm{j}}{\sum }_{{\displaystyle \frac{j}{t}}=1}^n{\Delta X}_{ij}+\Delta {\mathsf{\epsilon }}_{\mathrm{i}\mathrm{t}}$$

(2)

where

${\mathrm{C}\mathrm{S}}_{\mathrm{i}\mathrm{t}}$ represents the capital structure for bank i at time t as measured by LTDR, TDR and STDR.

${\mathrm{L}\mathrm{I}\mathrm{Q}}_{\mathrm{i}\mathrm{t}}$ is the liquidity measure for bank i at time t as measured by LCR.

$X_{ij}$ is a set of macro and microeconomic control variables, including GDPR, inflation rates (IF), and interest rates (INTR), that are analysed in the conclusion.

4. Results and Discussions

4.1. Descriptive Statistics

Table 2. Descriptive statistics.

Variables	Mean	Median	Maximum	Minimum	Std. Dev.	Skewness	Kurtosis	Jarque–Bera	Observ
TDR	0.88	0.91	0.94	0.63	0.07	8.12	73.11	23,520.27	110
LTDR	0.18	0.21	0.31	0.03	3.52	8.11	72.87	23,367.05	110
STDR	0.52	0.54	0.81	0.02	0.13	(0.33)	4.44	11.43	110
LCR	1.66	1.62	2.33	1.11	0.26	0.64	3.19	7.54	110
RGDP Billion	4488.39	4474.21	4599.26	4310.33	82.18	(0.61)	2.92	6.83	110
INTR	3.43	3.28	5.86	0.51	1.52	(0.17)	2.40	2.18	110
SIZE	8.00	8.12	9.24	9.24	6.63	0.98	0.02	14.53	110
CPI	4.20	4.10	5.60	3.10	0.81	0.44	1.97	8.36	110

provides an overview of the summary statistics for the variables used in this study, offering significant insights into their distribution characteristics, as well as measures of central tendency and variability, across a panel of 11 banks spanning the period from 2014 to 2024.

During the evaluation period, the metric for capital structure, represented by the total debt ratio (TDR), which signifies the average proportion of a bank’s assets funded through deposits and alternative borrowings, exhibited an average value of 0.88 with a standard deviation of 0.07. The TDR varied from a minimum of 0.63 to a maximum of 0.94, indicating extensive variability with an overall range of 94. In comparison, the long-term debt ratio (LTDR) displayed an average of 0.18 with a standard deviation of 3.52, spanning from a minimum of 0.03 to a maximum of 0.31. Meanwhile, the short-term debt ratio (STDR) recorded an average of 0.52 and a standard deviation of 0.13, ranging from 0.02 to 0.81. These observations suggest that some banks maintained less than 2% of their total obligations in short-term debt, underscoring potential differences in debt maturity structures across institutions.

However, the liquidity coverage ratio (LCR) exhibited an average of 1.66, accompanied by a standard deviation of 0.26, which signifies relatively low variation within the examined sample. The LCR spanned from a minimum of 1.11 to a maximum of 2.33. This range indicates that all financial institutions maintained liquidity reserves that were significantly above the Basel III minimum of 1.00. These findings suggest robust short-term liquidity standings were maintained by the banks throughout the period under review.

The Real Gross Domestic Product (RGDP) had an average of R4488.39 billion and a standard deviation of 82.18. This suggests a relatively low level of variability throughout the period under examination. The RGDP fluctuated between a low of R4320.33 billion and a high of R4599.26 billion, indicating moderate economic expansion during the timeframe analysed.

On the other hand, the interest rate (INTR) exhibited an average value of 3.43, with a standard deviation of 1.52, suggesting moderate variability throughout the duration of the study. The INTR ranged from 0.51 to 5.86, highlighting the diversity in monetary policy conditions during the observed period.

The typical bank size, as indicated by a mean of 8.00 and a standard deviation of 0.98, suggests moderate variability in bank size within the examined sample. The bank sizes ranged from 6.63 to 9.24, demonstrating a diverse range of operational scales among the analysed financial institutions.

Lastly, the Consumer Price Index (CPI) had an average of 4.20 and a standard deviation of 0.81, suggesting moderate inflation rate variability throughout the analysed period. The CPI fluctuated between 3.10 and 5.60, highlighting changes in the overall price level in the economy.

4.2. Correlation Analysis

As illustrated in

Table 3. Correlation analysis.

Variables	TDR	LTDR	STDR	LCR	RGDP	INTR	SIZE	CPI
TDR	1.00
LTDR	1.00	1.00
STDR	−0.21 **	−0.21 **	1.00
LCR	(0.03)	(0.02)	−0.23 **	1.00
RGDP	0.01	0.01	0.09	0.03	1.00
INTR	−0.21 **	−0.20 **	0.01	(0.01)	0.47 ***	1.00
SIZE	(0.16)	(0.14)	0.06	0.46 ***	0.03	(0.03)	1.00
CPI	(0.17)	−0.16 *	0.10	(0.03)	0.02	0.49 ***	(0.07)	1.00

Level of significance. * p < 0.05, ** p < 0.01, and *** p < 0.001.

, the correlation analysis reveals the associations between the independent and dependent variables in the banking sector.

The research indicates that there is a negative and significant correlation between both the TDR and the LTDR with the STDR. This finding implies that South African banks, when embodying higher amounts of total or long-term indebtedness, are inclined to lessen their dependence on short-term funding options, reflecting a preference for more stable, long-term financial structures. In addition, there is a negative and significant connection between the LCR and STDR, indicating that institutions with greater short-term liquidity reserves generally tend to minimise their reliance on short-term debt instruments.

The analysis of the correlation between Real Gross Domestic Product (RGDP) and TDR, LTDR, and STDR reveals a positive yet statistically non-significant connection. On the other hand, there is a negative and statistically significant relationship between interest rates (INTR) and both TDR and LTDR.

Regarding bank size, the findings suggest a negative but statistically non-significant association with TDR and LTDR, which implies that larger banks in South Africa do not necessarily depend more on long-term debt. Nonetheless, a positive relationship was observed between the size of banks and STDR.

Ultimately, the research reveals a negative yet statistically insignificant relationship between the Consumer Price Index (CPI) and TDR, and a negative, statistically significant relationship between CPI and LTDR. Conversely, there exists a positive, yet statistically insignificant, association between CPI and STDR.

The correlation coefficients among the predictor variables were all below 0.7, suggesting no significant multicollinearity (Siddik et al. 2017). Furthermore, the Variance Inflation Factor (VIF) values ranged from 1.30 to 3.10, well below the widely accepted threshold of 10, further corroborating the absence of multicollinearity. In addition, no substantial associations were observed among the parameters within the equations. Consequently, variables with minimal explanatory capability were omitted from the final model to improve parsimony and overall model effectiveness.

4.3. Empirical Analysis

In the context of the findings outlined in

Table 4. Effect of LCR on capital structure.

Models	2-Step System GMM	2-Step System GMM	2-Step System GMM
Variables	TDR	LTDR	STDR
LagTDR	0.444 ***
	(0.0119)
LagLTDR		0.659 ***
		(0.0717)
LagSTDR			0.0174 *
			(0.00743)
LCR	0.134 ***	0.174 **	−0.834 ***
	(0.0149)	(0.0447)	(0.0712)
LRGDP	0.139 ***	0.0669	0.615 ***
	(0.026)	(0.254)	(0.122)
CPI	−0.0342 ***	−0.0444 ***	−0.0202 *
	(0.00361)	(0.00549)	(0.00834)
INTR	−0.0493 ***	0.00281 *	−0.00811 *
	(0.00184)	(0.00121)	(0.00281)
SIZE	−0.377 ***	−0.961 ***	0.153
	(0.0349)	(0.0586)	(0.0839)
COVID_19	0.0952 ***	0.0162 ***	−0.000300
	(0.00133)	(0.00148)	(0.000402)
_cons	−1.162	0.784	0.3813
	(2.865)	(3.142)	(1.1120)
N	110	110	99
Groups	11	11	11
Instrument	9	9	9
Arellano–Bond AR(1)	−2.11	−1.26	−1.66
Arellano–Bond AR(2)	1.33	1.44	1.54
Sargan test	1.96	8.70	2.39
Hansen test	2.85	5.46	5.48

Robust standard errors in parentheses. * p < 0.05, ** p < 0.01, and *** p < 0.001.

, the research reveals a negative, statistically significant association between indicators of capital structure and their preceding values, particularly for TDR and LTDR. This implies a proclivity among financial institutions to adapt their leverage positions over time, aligning with the principles of the dynamic trade-off theory, which asserts that firms incrementally progress towards achieving optimal capital structures (Flannery and Rangan 2006). Conversely, the relationship observed between the STDR and its lagged value was positive yet lacked statistical significance, suggesting less robust adjustment dynamics in short-term borrowing decisions.

The analysis demonstrates a positive and significant correlation between the LCR and the TDR and LTDR, while the LCR is negatively and significantly related to the TDR. These findings concur with the trade-off theory, suggesting banks weigh the advantages of indebtedness against the risks associated with financial distress (Kraus and Litzenberger 1973; Myers and Majluf 1984). Furthermore, the results are in alignment with the principles of the Basel III framework, which recommends maintaining robust liquid assets while minimising dependence on short-term funding to bolster stability (Basel Committee on Banking Supervision 2011a). Consequently, banks with substantial liquidity reserves are inclined to prefer long-term debt structures to satisfy regulatory and risk management stipulations (Gropp and Heider 2010; Flannery and Rangan 2006). These results are consistent with the findings of Rao et al. (2017), who identified a positive linkage between debt and liquidity, as determined by current ratios.

In terms of control variables, the results of the study reveal a positive and significant association between real GDP and the indicators of capital structure, TDR, LTDR, and STDR among banking institutions. This implies that during periods of economic growth, banks enhance their use of leverage to address increasing credit demand and facilitate investment activities. However, there was a negative and significant correlation between CPI and the TDR and the LTDR. This indicates that higher inflation tends to discourage banks from taking on long-term debt, likely due to increased uncertainty and higher borrowing costs.

In addition, the interest rate (INTR) demonstrates a positive and significant correlation with the TDR. This implies that financial institutions are likely to augment their overall leverage in response to escalating interest rates, possibly as a strategy to secure borrowing before any additional increases in rates. While a positive yet statistically insignificant connection is identified with the LTDR, a contrasting negative and significant association is observed between INTR and the STDR. On the other hand, the bank size was negatively and significantly related to both TDR and LTDR, indicating that larger banks tend to rely less on long-term and total debt. However, the relationship with STDR is positive but not significant. Lastly, the COVID-19 pandemic had a notable and positive impact on both TDR and LTDR, indicating a greater reliance on both overall and long-term debt during this time. In contrast, the pandemic’s effect on the STDR was negative but not significant.

5. Conclusions and Policy Implications

This research examined the impact of liquidity regulations, specifically the NSFR and the LCR, on the capital structure of South African banks, using data from 2015 to 2024. The study employed the Generalised Method of Moments (GMM) estimator to address potential endogeneity issues frequently encountered in financial structure research. The findings of the study demonstrated a positive and significant connection between the LCR and with both the TDR and LTDR, while its relationship with the STDR was significantly negative.

The findings presented in this research have significant implications for financial regulators and policymakers. Liquidity requirements, such as the LCR, not only enhance resilience during times of financial distress but also influence banks’ decisions regarding their capital structures by encouraging long-term, stable funding practices. Therefore, regulators, particularly the South African Reserve Bank (SARB), should recognise the dual role of liquidity regulations and view them as both prudential and structural mechanisms. This evidence supports the case for maintaining stringent liquidity standards as an essential part of a broader macroprudential policy framework. Furthermore, regulatory bodies should consider developing tailored supervisory guidance that connects liquidity regulation with capital structure outcomes, thereby promoting long-term financial stability.

Future research has the potential to expand on the current analysis in several ways. First, including both book value and market value metrics of leverage could provide a more comprehensive understanding of capital structure dynamics. Second, comparative studies across different financial sectors, such as insurance, asset management, or fintech, may reveal whether the effects of liquidity regulation vary by institution type. Third, cross-national studies that examine the interaction between different regulatory frameworks and firm-specific characteristics would enhance the generalisability of these findings. Finally, further academic inquiry could investigate the relationship between liquidity regulation and other macroprudential tools, such as capital buffers or countercyclical policies, to determine their combined effect on financial stability and capital structuring decisions.