Profitability Drivers in European Banks: Analyzing Internal and External Factors in the Post-2009 Financial Landscape

Abstract

The paper examines the key determinants of European banks’ profitability by analyzing the return on assets (ROA), return on equity (ROE), net interest margin (NIM), and the risk-adjusted measures of profitability, RAROAA and RAROAE, across 34 European countries during the period from 2013 to 2018—a time characterized by economic recovery and significant regulatory reforms, including the implementation of Basel III standards. Using the Generalized Method of Moments (GMM) approach and data of 3076 European banks, the research addresses the complex interplay between internal (bank-specific) factors and external factors, including macroeconomic and industry-specific factors. The results show that profitability is positively associated with a higher capital adequacy, liquidity risk, and income diversification, but not for risk-adjusted profitability ratios. Credit risk, management efficiency, and excessive size have a negative effect on all studied profitability measures. Macroeconomic conditions, in particular, GDP growth and inflation, also have a significant impact on profitability. The findings offer valuable insights for policymakers, regulators, and financial institutions aiming to enhance profitability while maintaining the stability of the European banking sector.

1. Introduction

In the wake of the global financial crisis in 2009, the European banking sector underwent significant changes as financial institutions and regulators reassessed the fundamentals of bank stability and profitability. The crisis highlighted vulnerabilities within the financial system and sparked interest in understanding the determinants of bank profitability across Europe, particularly given the significant differences in recovery trajectories and profitability metrics between European banks and their US counterparts. Profitability is a critical determinant of bank resilience; it enables institutions to withstand economic downturns and absorb shocks, thereby promoting stability in the broader financial ecosystem (Borroni and Rossi 2019). Low profitability can heighten systemic vulnerabilities and increase the susceptibility of financial systems to crises by constraining banks’ capacity to extend credit, foster economic growth, and address market disruptions (Saif-Alyousfi 2022). In this context, examining key profitability measures provides valuable insights for policymakers, regulators, and financial practitioners.

This paper examines the determinants of profitability of 3076 banks from 34 European countries over the period of 2013 to 2018, a period characterized by economic recovery and regulatory changes, such as the implementation of Basel III standards. The paper seeks to explore how selected bank-specific, industry-specific, and macroeconomic factors affected various measures of profitability, by using traditional measures such as the Return on Average Assets (ROAA), Return on Average Equity (ROAE), and Net Interest Margin (NIM), as well as measures which are adjusted for risk, including the Risk-Adjusted Return on Assets (RAROAA) and Risk-Adjusted Return on Average Equity (RAROAE). To study the determinants of the profitability of European banks, we use the Generalized Method of Moments (GMM). The bank-level data were collected from BankFocus database, while the macroeconomic variables were from the World Bank and OECD databases.

Our results suggest that bank size, credit risk, and the cost-to-income ratio have a negative effect on bank profitability, while capital adequacy, liquidity risk, and income diversification have a positive effect. However, when the profitability measures are adjusted for risk, capital adequacy and liquidity risk show a negative effect, which may suggest that better capitalized and more liquid banks take less risk, leading to lower returns on a risk-adjusted basis. Macroeconomic conditions, in particular, GDP growth, have a positive effect on profitability, reflecting the economic recovery during the period and the increased demand for credit.

The paper adds to the literature in five ways. First, it focuses on the post-crisis period of stable GDP growth and regulatory changes brought about with Basel III and CRD IV, a period that has been analyzed only by a few studies (see, for example, Korytowski (2018) and Davis et al. (2022)). This period gives us an opportunity to explore the effects on bank profitability in a setting that reflects both the impact of the post-crisis economic recovery and the adjustments made by banks to meet stricter capital and liquidity requirements. These requirements were introduced to strengthen the resilience of the banking sector and reduce the likelihood of future financial crises by ensuring that banks maintain sufficient capital buffers. However, this regulation has also had an impact on profitability by increasing operating and compliance costs, especially for smaller banks (Gržeta et al. 2023). Second, in addition to the traditional profitability measures commonly used in the literature, the paper also applies to risk-adjusted measures, RAROAA and RAROAE, which are designed to provide a clearer view of the relationship between financial performance and the risks taken to achieve that performance. Third, the analysis is based on the GMM which enables addressing endogeneity issues in estimating the relationship between profitability and its determinants, thereby increasing the reliability of the results. Fourth, the study uses data from a large number of countries with distinct economic and regulatory landscapes, thereby increasing its regional relevance: it analyzes not only banks in the European Union (EU) Member States, which are commonly studied in the literature, but also includes data for the Western Balkan countries, which have been studied by only a few studies (for example, Radovanov et al. 2023). Lastly, by providing insights into the dynamics of bank profitability under different economic conditions, our study aims to support the development of regulatory frameworks that balance profitability and stability in the European banking sector.

The paper is structured as follows: After the Introduction, an overview of the empirical research on European bank profitability is summarized in Section 2. Section 3 presents the data and methodology, including a description of the econometric model and the specification of the variables included in the model. Section 4 presents the results and discussion. Finally, the conclusion summarizes the findings and provides recommendations for further research.

2. Literature Review

Since the 2009 global financial crisis, bank profitability has remained a significant challenge globally (Bank for International Settlements 2018). While banks in the North Atlantic region have largely recovered, European banks have lagged in profitability compared to their American counterparts over the past decade (Elekdag et al. 2020). To address these variations, scholars have extensively analyzed the determinants of bank profitability across different regions and periods, employing diverse methodologies. This section reviews the literature by dividing the determinants into two categories: internal or bank-specific factors, and external factors, which include industry and macroeconomic factors, with a particular focus on studies for European countries.

2.1. Internal or Bank-Specific Factors

One of the most common internal determinants of bank profitability is bank size; yet, the findings are divided. Several studies have demonstrated a positive correlation between bank size and profitability. For instance, research on 44 Indonesian banks found that an increase in bank size resulted in an increase in ROA over the period of 2016–2018 (Anggari and Dana 2020). Similarly, a study on 2446 banks across 47 Asian countries over the period of 1995–2017 confirmed that larger banks benefit from economies of scale and scope, positively influencing various profitability measures like the ROA, NIM, and profit before tax (Saif-Alyousfi 2022). Similar findings were reported also by Koroleva et al. (2021) for state-owned Chinese commercial banks. Shehzad et al. (2013) showed that, in the OECD countries, in the period of 1988 to 2010, bigger banks were more profitable than small banks. As regards European banks, positive effects of banks’ size on their profitability were reported by Menicucci and Paolucci (2016) and Radovanov et al. (2023) for banks in the West Balkan countries. Blaga et al. (2024) showed that bank size has a positive impact on ROA for banks at the lowest profitability level, with the effect becoming statistically insignificant at higher profitability quantiles. On the other hand, several studies report that bank size negatively affected the profitability of the European banks (for example, Goddard et al. (2004), Pasiouras and Kosmidou (2007), and Koutsomanoli-Filippaki et al. (2012)), indicating that larger banks may suffer from inefficiencies that reduce their profitability. Among recent studies for European banks, Davis et al. (2022) reported that bank size negatively affected the ROAA and ROAE over the period of 1990–2018, whereas Korytowski (2018) found only a negative effect on the ROAA over the period of 2011–2015.

An important determinant of bank profitability is credit risk exposure. Several studies report that higher credit risk exposure, measured either by loan-loss provisions to total loans or non-performing loans to gross loans, is associated with the lower profitability of European banks (Athanasoglou et al. 2006; Blaga et al. 2024; Borroni and Rossi 2019; Davis et al. 2022; Horobet et al. 2021; Menicucci and Paolucci 2016; Mirović et al. 2024; Petria et al. 2015; Radovanov et al. 2023). Nevertheless, Korytowski (2018) found no significant effect of credit risk on European banks in terms of the ROAA and ROAE over the post-crisis period. Djalilov and Piesse (2016) found that, over the 2000–2013 period, for banks in early-transition CEE countries, credit risk positively affected profitability, whereas, in late-transition countries of the former USSR, credit risk negatively impacted profitability.

Another critical determinant of bank profitability is liquidity risk exposure, often measured by the loan-to-deposit ratio. The literature offers mixed views on its impact on profitability. Studies on banks in Jordan (2005–2011) and Vietnam (2013–2018) found a positive correlation between liquidity and profitability (Al Nimer et al. 2015; Doan and Bui 2021). Similar findings are reported also by some European studies—positive and significant effects of bank profitability were reported by Petria et al. (2015), Mergaerts and Vander Vennet (2016), Korytowski (2018), Borroni and Rossi (2019), Davis et al. (2022), and Blaga et al. (2024). In contrast, Budhathoki et al. (2020) noted that higher liquidity could lead to an increase in non-performing loans (NPLs), which negatively affects profitability. Radovanov et al. (2023) also demonstrated a negative impact of the loan-to-deposit ratio and the non-performing loans-to-total-loans ratio on both the ROA and ROE. This negative impact was corroborated by a study on euro-area banks (2006–2017), where a decrease in NPLs significantly improved the ROA (Elekdag et al. 2020). Lastly, a strand of literature finds no significant relationship between liquidity and bank profitability, such as in the case of Bangladeshi banks between 2006 and 2011 (Akter and Mahmud 2014), as well as in the case of 16 Chinese commercial banks between 2008 and 2020 (Jigeer and Koroleva 2023).

Other important bank-specific factors affecting bank profitability include capital adequacy, credit quality, and management efficiency. For example, well-capitalized banks tend to exhibit higher profitability, as shown in several studies of European banks (Ercegovac et al. 2020; Goddard et al. 2004; Petria et al. 2015; Radovanov et al. 2023). Davis et al. (2022) found that the capital ratio had a positive impact of the ROAA over the 1990–2018 period; yet, it has a negative effect on ROAE. Negative relation between the capital and bank profitability was reported also by Hoffmann (2011) for the U.S. and Topak and Talu (2017) for Turkish banks. According to Davis et al. (2022), the negative effect can be explained by the evolving Basel Accord capital requirements. Similarly, capital ratios and asset management are significant determinants of profitability in GCC banks, with liquidity moderating the relationship between capital adequacy and performance (Al-Matari 2023). European Central Bank (2024) shows that a higher share of loans in bank assets supports profitability by increasing net interest margins, especially when interest rates are favorable. Banks with a larger loan proportion benefit from interest income, a key profit driver. However, this advantage depends on the economic cycle. During downturns, high loan exposure can lead to increased non-performing loans, reducing profitability due to higher loan loss provisions. As regards management efficiency, proxied by th ecost-to-income ratio, most studies report a negative effect of the cost-to-income ratio on the bank profitability of the European banks (Borroni and Rossi 2019; Davis et al. 2022; Elekdag et al. 2020; Korytowski 2018; Mirović et al. 2024; Petria et al. 2015).

Several works of research dealt with the income diversification of banks (Ammar and Boughrara 2019; Berger et al. 2010; Elekdag et al. 2020; Gambacorta and van Rixtel 2013; Goddard et al. 2013; Le and Ngo 2020; Nguyen et al. 2012; Petria et al. 2015). These studies collectively suggest that, while income diversification generally has a positive impact on bank profitability, the extent and nature of diversification matter significantly. Moderate diversification tends to yield better results, while excessive diversification can lead to risks that may negate profitability benefits.

2.2. Industry and Macroeconomic Factors

Apart from internal factors, bank profitability is affected also by industry and macroeconomic (i.e., external) factors.

One of the commonly studied industry-specific factors that affect bank profitability is market concentration, often proxied by the Herfindahl–Hirschman Index (HHI). Several studies for European banks reported a positive association between market concentration and bank profitability (Athanasoglou et al. 2008; Goddard et al. 2013; Petria et al. 2015). In contrast, Korytowski (2018) reported negative effects.

Among macroeconomic factors, GDP growth is a key variable influencing bank profitability. A positive relationship between GDP and profitability was observed in studies of banks in Asian countries (1995–2017) and city commercial banks in China (2008–2020), as economic growth typically boosts loan demand and profitability (Jigeer and Koroleva 2023; Saif-Alyousfi 2022). However, some studies show a negative correlation, particularly in highly competitive markets. For instance, Liu and Wilson (2010) found that GDP growth negatively affected Japanese banks’ profitability by lowering entry barriers and increasing competition. Similarly, Tan (2013), in a study on Chinese banks between 2003 and 2009, and Koroleva et al. (2021) from 2007 to 2019 showed that higher GDP growth leads to lower bank profitability. As for the European banks, the positive effect of GDP growth on bank profitability was reported by Borroni and Rossi (2019), Davis et al. (2022), Mirović et al. (2024), and Petria et al. (2015). Athanasoglou et al. (2008) found that fluctuations in real GDP per capita showed no significant effect on banks’ profits in the SEE region. On the other hand, Islam (2023) reported that GDP growth negatively affected the profitability of the UK banks over the 2015–2019 period.

Other macroeconomic factors, such as inflation and real interest rates, have shown varied effects on profitability. For example, while studies in Jordan and Vietnam (Al Nimer et al. 2015; Doan and Bui 2021) have indicated significant relationships between these variables and profitability, research on the UK banks found no significant impact (Islam 2023). This suggests that internal factors, such as operational efficiency and risk management, play a more critical role than external conditions in determining profitability in some markets (Islam 2023). No significant association between inflation and bank profitability was found also by Davis et al. (2022) and Petria et al. (2015). Korytowski (2018) showed that inflation had a negative effect on the ROAA and ROAE of European banks after the 2008 financial crisis. Mirović et al. (2024) similarly reported a negative association in the eurozone during the 2015–2020 period, while Horobet et al. (2021) identified a similar trend in CEE countries over the 2009–2018 period. Several other studies for European banks reported a positive association between inflation and bank profitability (Athanasoglou et al. 2008; Căpraru and Ihnatov 2014). Only a few studies analyze the relationship between bank profitability and long-term interest rates. For Europe, Borroni and Rossi (2019) reported that the long-term interest rate is positively associated with NIM; yet, there is no significant relationship with the ROAA and ROAE.

The role of regulation in shaping bank profitability has also been explored extensively. Stricter capital requirements, introduced under frameworks like Basel II and Basel III, have had mixed effects. A global study of 615 commercial banks (2000–2004) found that enhanced market discipline improved profitability, while stricter capital regulations reduced profit efficiency (Pasiouras et al. 2009). In Europe, the impact of these regulations has been more nuanced. A study of 433 banks (2006–2015) found that, while large- and medium-sized banks improved efficiency and profitability, smaller banks struggled due to increased regulatory burdens, which could lead to future mergers or failures (Gržeta et al. 2023).

The development of financial technology (fintech) is also an external factor that has a multifaceted impact on the European banking sector, especially in recent years, fostering innovation, competition, and transformation in business models, regulations, and customer experiences. Several research papers addressed the issue, for example, Yoon et al. (2023), using the World Bank Global Findex Database for 91 countries in 2014, 2017, and 2021. Their results imply that banks in less developed countries benefit most from investing in fintech innovation. Yudaruddin (2022) analyzed the impact of fintech startups on Islamic and conventional banking performance in Indonesia. His main finding was that fintech startups have a detrimental effect on bank performance. Dasilas and Karanović (2023) analyzed the impact of fintech on bank performance in the UK from 2010 to 2019 and reported positive effects.

The literature on bank profitability highlights a complex interplay between internal and external factors. Internal determinants, such as bank size, liquidity risk, capital adequacy, loans share, credit risk, cost efficiency, and income diversification, are significant, although their impact can vary depending on regions, periods, and specific market conditions. External factors like GDP growth, inflation, concentration, fintech, and regulation also play a role, particularly in shaping the broader competitive and operational environment. The diverse findings underscore the importance of tailoring regulatory and managerial strategies to bank size and market context to ensure sustainable profitability.

3. Data and Model Specification

3.1. Data

The empirical analysis of the factors influencing the profitability of European banks is based on a dataset comprising financial statements and macroeconomic information. The bank-specific variables were sourced from the BankFocus database, while the macroeconomic variables (GDP growth, inflation rate, and long-term interest rate) were obtained from the World Bank and OECD databases. The analysis focused on banks located in 34 European countries, encompassing the 27 EU Member States, Norway, the United Kingdom, and the Western Balkan countries for which data were available (i.e., Albania, Bosnia and Herzegovina, Montenegro, North Macedonia, and Serbia) for the period between 2013 and 2018.

The 2013–2018 period is significant for a deeper analysis for two primary reasons. First, 2013 marks the onset of the post-crisis phase of consistent economic growth, with the GDP starting to increase steadily for the first time since 2012. In the EU-28, the real GDP growth was 0.3% in 2013, climbed to 1.8% in 2014, rose further to 2.6% by 2017, and then gradually declined to 1.8% in 2018 (Eurostat 2024). Second, this period is also characterized by the introduction of Basel III and CRD IV regulations, which strengthened capital requirements and introduced new standards for leverage and liquidity, aimed at bolstering financial stability (Bank for International Settlements 2024).

3.2. The Model

The analysis employs an unbalanced sample of banks, comprising all banks with complete data for a minimum of four consecutive years. The use of an unbalanced sample avoids the potential issue of survivor bias that may arise in balanced panels. Moreover, banks with outlier variable values (i.e., the top and bottom 1% of variable values) were excluded from the analyses to minimize the impact of extreme outliers. Additionally, the sample was restricted to include only those banks with recent and comprehensive financial data, as well as to exclude branches and specialized financial institutions, in addition to national central banks. The final sample comprised a total of 3076 banks.

To investigate the determinants of European banks’ profitability, we employ the Generalized Method of Moments (GMM), originally introduced by Arellano and Bond (1991). This method has since become a widely used approach in the field, particularly in European studies, as evidenced by works such as Athanasoglou et al. (2008), Horobet et al. (2021), Mirović et al. (2024), and Radovanov et al. (2023). The GMM model is frequently employed in literature due to its efficacy in addressing endogeneity by using lagged values of endogenous variables. It is also characterized by robustness to heteroskedasticity and autocorrelation, as well as suitability for large cross-sectional dimensions with shorter time series, according to Arellano and Bond (1991), Baltagi (2008), and Blundell and Bond (1998). Two versions of the GMM estimator are available: the difference GMM and the system GMM. The difference GMM, as developed by Arellano and Bond (1991), employs lagged levels of endogenous variables as instruments for first-differenced equations, thereby removing time-invariant fixed effects. The system GMM, introduced by Blundell and Bond (1998), enhances the earlier method by integrating first-difference equations with the original-level equations. This reduces biases due to weak instruments and is particularly suited to models with a large cross-section and short time series. In addition, by using both levels and first differences, the system GMM provides lower standard errors and more reliable parameter estimates. The system GMM treats endogeneity more comprehensively and captures dynamic relationships more effectively than the difference GMM. In this way, we avoid the problem of weak instruments which can bias coefficients, inflate standard errors, and undermine the credibility of the model.

We applied the following two-step system GMM equation:

$${\mathsf{\Pi }}_{it}=\alpha {\mathsf{\Pi }}_{it-1}+\sum _{j=1}^J{\beta }_j{\mathsf{{\rm X}}}_{it}^j+\sum _{k=1}^K{\beta }_k{\mathsf{{\rm X}}}_{it}^k+\sum _{m=1}^M{\beta }_m{\mathsf{{\rm X}}}_{it}^m+c_t+{\epsilon }_{it}$$

(1)

where ${\mathsf{\Pi }}_{it}$ is the dependent variable, i.e., the profitability or risk-adjusted profit measure of bank i at year t. The ${\mathsf{\Pi }}_{it-1}$ is the one-year lagged dependent variable. We apply the one-year lagged dependent variable to capture the dynamic nature of bank profitability. The profitability measures, namely, are persistent and often depend on their own past values. The explanatory variables are grouped into bank-specific (${\mathrm{X}}_{it}^j$), banking industry-specific (${\mathrm{X}}_{it}^k$), and macroeconomic (${\mathrm{X}}_{it}^m$) ones. $c_t$ is the constant term and the ${\epsilon }_{it}$ is the error or disturbance term. The variables used are explained in more detail in Section 3.3.

To assess the validity of the instruments and the model, we use the Hansen test for overidentifying restrictions and the AR(1) and AR(2) tests to check for the absence of residual serial correlation.

3.3. Variable Specification

Table 1. Description of the variables.

Variable	Definition	Data Source	Expected Effect on Profitability
Dependent variables: Measures of bank profitability
ROAA	Net income/average total assets	BankFocus
ROAE	Net income/average total equity	BankFocus
NIM	Net interest margin: difference between interest income and interest expenses relative to the amount of interest-earning assets.	BankFocus
RAROAA	ROAA/σROAA	Authors’ calculation based on BankFocus
RAROAE	ROAE/σROAE	Authors’ calculation based on BankFocus
Independent variables
Bank-specific variables
Bank size	Natural logarithm of total assets	Authors’ calculation based on BankFocus	+/−
Capital	Equity/total assets		+/−
Loans share	Net loans/total assets		+/−
Credit risk	Loan loss provisions/loans		-
	Non-performing loans/gross loans		-
Liquidity risk	Liquid assets/total assets		+/−
	Loans/deposits and short-term funding		+/−
Management efficiency	Cost-to-income ratio: total operating expenses/total operating income	BankFocus	-
Income diversification	1 − [(Net interest income − other operating income)/(operating income)]	Authors’ calculation based on BankFocus	+/−
Banking-industry-specific variable
Market concentration	Herfindahl–Hirschman index	Authors’ calculation based on BankFocus	+/−
Macroeconomic variables
Economic growth	Real GDP growth, annual	World Bank	+
Inflation	Inflation rate measured by consumer price index, annual	World Bank	+
Long-term interest rate	Yield on 10-year government bonds	OECD, national central banks	+

presents the variables utilized in the regression modelling, their definitions, data sources, and, based on the literature review, their anticipated effect on bank profitability. As illustrated in Section 2, the literature classifies factors influencing bank profitability into two categories: internal or bank-specific factors and external factors, which encompass banking industry-specific and macroeconomic factors. In continuation of this section, we explain variables used in the regression modelling.

3.3.1. Dependent Variables: Measures of Bank Profitability

We use three different variables to proxy the bank (return on average assets (ROAA), return on average equity (ROAE), and net interest margin (NIM)) and two measures for the risk-adjusted return (risk-adjusted return on average assets (RAROAA) and risk-adjusted return on average equity (RAROAE)). The RAROAA and RAROEA are calculated as the ROAA and ROEA, respectively, divided by the standard deviation (volatility) of the ROAA and ROAE, respectively, over the observation period. While the first three measures are commonly used in the literature (see Section 2), the RAROAA and RAROEA are an extension of the underlying measures, designed to provide a clearer view of the relationship between financial performance and the risks taken to achieve that performance (see Borroni and Rossi (2019) for further details).

3.3.2. Bank-Specific Factors

Bank size, proxied by the natural logarithm of total assets, is one of the most extensively researched internal factors of bank profitability. The effect of bank size is ambiguous because, on the one hand, it may create economies of scale (Shehzad et al. 2013), while, on the other hand, the benefits from the economies of scale might decrease with increasing bank size, as large banks may be subject to rigidities (Athanasoglou et al. 2008; Goddard et al. 2004).

The second factor is financial structure, which shows how banks’ assets are financed and their ability to absorb losses. To assess the effect of the financial structure of banks, we use the capital adequacy ratio (or leverage ratio), defined as the ratio of equity to total assets, a measure commonly used across literature. A higher capital adequacy ratio indicates a stronger capital buffer of the bank and lower risk, with a potentially positive effect on bank performance. On the other hand, a higher capital ratio implies that the bank relies less on borrowed funds, limiting its leverage and increasing its financing costs, which could have a negative impact on its profitability (Andersen and Juelsrud 2024).

The third factor that we consider is the loan share, defined as the ratio of net loans to total assets. This variable reflects the bank’s business model, with the effect on profitability depending on the economic cycle and conditions. In periods of economic growth, the effect is mainly positive, whereas, in periods of economic crisis, the effect is unclear, although several studies report a negative effect (Borroni and Rossi 2019).

To proxy the credit risk, we use the ratio of loan-loss provisions to total loans. This variable represents the quality of the bank’s assets and the efficacy of the bank’s risk management practices. A higher ratio indicates a higher level of perceived credit risk within the bank’s loan portfolio and is associated with lower levels of bank profitability (Horobet et al. 2021; Menicucci and Paolucci 2016). To ensure the robustness of our regression model, we also consider the ratio of non-performing loans to gross loans, a measure commonly applied in literature. For this purpose, we have re-tested the above specified regression model by applying an alternative measure of credit risk. The results are reported separately.

The ratio of short-term liquid assets to total assets is employed as a measure of liquidity risk. A higher ratio is associated with a lower probability of a bank failing in a situation of distress. An increase in liquidity should enhance the profitability of banks. However, as evidenced by the findings of Bordeleau and Graham (2010), there is a point at which further liquid assets have a detrimental impact on profitability, due to the opportunity costs involved. To ensure the reliability of our results, we have also employed an alternative measure of liquidity risk, defined as the ratio of loans to deposits and short-term funding. Moreover, in this case, we have re-tested the above specified regression model by using an alternative measure of liquidity risk and reported the results separately.

As a proxy for management efficiency, we use the cost-to-income ratio, which should be negatively associated with bank profitability. Namely, improved management, which results in reduced operating expenses, should increase efficiency (Akbaş et al. 2012; Petria et al. 2015).

Furthermore, the model incorporates a measure of income diversification, which represents an approach that banks can utilize to enhance risk management. The impact of income diversification on banks’ profitability remains inconclusive. On the one hand, it has the potential to enhance profitability by generating new revenue streams. However, on the other hand, the volatility associated with these new sources of revenue may outweigh the benefits of diversification (Borroni and Rossi 2019; Stiroh 2004).

3.3.3. Banking-Industry-Specific Factors

To account for industry-specific factors, the market concentration is proxied by the Herfindahl–Hirschman Index. The Herfindahl–Hirschman Index is calculated by taking the sum of the squares of the market share of banks. The impact of the market concentration on bank profitability is inconclusive. On the one hand, a bank operating in a highly concentrated market may possess market power and achieve economies of scale. Conversely, they may be less inclined to enhance efficiency or more susceptible to engaging in riskier activities (Beck et al. 2006).

3.3.4. Macroeconomic Factors

As a first macroeconomic factor, we use the real GDP growth. As demonstrated in Section 2, empirical studies indicate that bank profitability follows cyclical trends, with an increase in bank activity and revenues leading to a positive effect on profitability during periods of growth and an opposite effect during periods of economic downturn. Another macroeconomic variable that we include is (current) inflation. When inflation is higher than anticipated, banks pass on this increase to their customers in the form of higher interest rates on loans, which consequently increases bank profitability. The third factor that we apply is long-term interest rates, which are measured as the yield on 10-year government bonds. This incorporates future inflation expectations and should be positively associated with profitability (Borroni and Rossi 2019).

Table 2. Descriptive statistics, 2013–2018.

Variable	Mean	Median	Standard Deviation	Min	Max
ROAA	0.41	0.31	0.47	−3.93	3.99
ROAE	3.97	3.27	4.43	−36.42	29.81
NIM	2.06	1.97	0.76	0.05	7.50
RAROAA	5.29	3.51	6.83	−6.76	132.65
RAROAE	4.19	3.27	4.36	−5.76	97.98
Bank size	13.32	13.17	1.64	9.74	19.05
Capital adequacy ratio	10.45	9.64	4.31	1.19	84.89
Loans share	60.07	61.69	15.65	1.26	95.27
Loans-loss provisions to total assets	0.32	0.15	1.27	−29.56	73.71
Non-performing loans to gross loans	5.66	2.88	7.79	0.00	100.00
Liquidity ratio	17.24	12.55	14.45	0.28	97.44
Loans-to-deposits ratio	74.67	72.49	107.15	1.36	9317.88
Cost-to-income ratio	70.31	69.97	15.41	16.63	179.08
Income diversification	0.46	0.48	0.43	−13.95	1.98
HHI	1316.50	1143.89	761.48	501.14	6146.61
GDP growth	1.73	1.79	1.33	−6.55	24.48
Inflation	1.07	1.03	0.78	−2.10	7.69
Long-term interest rate	1.32	1.16	1.16	0.09	10.05

Source: Authors’ calculations.

shows the descriptive statistics of the variables included in the econometric analysis over the 2013–2018 period. The ROAA and ROAE indicate moderate profitability, with mean values of 0.41% and 3.97%, respectively, although the ROAE shows high variability. The NIM averaging 2.06% reflects the difference between interest income and expenses relative to assets, with values of up to 7.5%, indicating varied profitability across banks. Risk-adjusted returns, the RAROAA and RAROAE, display higher means at 5.29% and 4.19%, although both exhibit significant volatility, suggesting occasional high returns adjusted for risk. The liquidity ratio has a broad spread, with an average of 17.24%, indicating differing levels of asset liquidity and resilience across institutions. The CIR, averaging 70.3%, underscores operational efficiency, while the non-performing-loans-to-gross-loans ratio is 5.66%.

We further examined the Pearson correlations between the explanatory variables that are included in the regression models. The correlations between explanatory variables are mostly weak, implying that there are no multicollinearity issues (see

Table A1. Pearson correlation coefficients.

	ROAA	ROAE	NIM	RAROAA	RAROAE	Bank Size	CAR	Loans Share	LLP to Total Assets	Liquidity Ratio	CIR	Income Diversification	HHI	Inflation	GDP Growth	LT Interest Rate	NPL to Gross Loans	Loans to Deposits Ratio
ROAA	1.00	0.87 *	0.19 *	0.07 *	0.20 *	−0.01	−0.01 *	0.05 *	−0.17 *	0.14 *	−0.42 *	0.09 *	0.20 *	0.06 *	0.17 *	−0.08	−0.06	0.04 *
ROAE		1.00	0.12 *	−0.01	0.20 *	0.08 *	−0.05 *	0.05 *	−0.14 *	0.13 *	−0.44 *	−0.03 *	0.13 *	0.12 *	0.12 *	0.06 *	−0.17 *	0.03 *
NIM			1.00	0.02 *	0.02 *	−0.23 *	0.10 *	0.06 *	0.12 *	−0.00	−0.08 *	−0.16 *	0.10 *	0.12 *	0.11 *	0.37 *	−0.17 *	0.00
RAROAA				1.00	0.68 *	0.10 *	−0.00	0.08 *	−0.14 *	−0.17 *	−0.02	0.04 *	0.04 *	0.02	0.00 *	−0.18 *	−0.20 *	0.00
RAROAE					1.00	0.03 *	−0.22 *	0.11 *	−0.14 *	−0.13 *	−0.13 *	−0.05 *	0.09 *	0.10 *	0.04 *	−0.18 *	−0.26 *	0.01
Bank size						1.00	−0.22 *	0.09 *	0.04 *	−0.15 *	−0.15 *	0.10 *	−0.01	0.13 *	−0.07 *	−0.06 *	−0.07 *	0.02 *
CAR							1.00	−0.05 *	0.04 *	0.13 *	−0.05 *	0.22 *	−0.01	0.13 *	0.07 *	−0.22 *	−0.05 *	0.12
Loans share								1.00	−0.10 *	−0.54 *	−0.07 *	−0.16 *	0.08	−0.14 *	0.16 *	−0.19 *	−0.19 *	−0.04
LLP to total assets									1.00	−0.06 *	−0.26 *	−0.04 *	−0.11 *	0.05 *	−0.11 *	−0.07 *	−0.22 *	−0.16 *
Liquidity ratio										1.00	0.09 *	0.13 *	0.07	−0.12	−0.08 *	0.04 *	−0.25 *	−0.21 *
CIR											1.00	0.09 *	−0.16 *	0.16 *	−0.12 *	−0.25 *	0.11 *	−0.18 *
Income diversification												1.00	0.08 *	0.18	−0.09 *	−0.19 *	−0.24 *	−0.20 *
HHI													1.00	0.04 *	0.04	0.03 *	0.01	−0.10
Inflation														1.00	−0.04	−0.03	−0.21 *	0.07 *
GDP growth															1.00	−0.19 *	−0.08 *	−0.02 *
LT interest rate																1.00	0.45 *	−0.00
NPL to gross loans																	1.00	−0.06 *
Loans-to-deposit ratio																		1.00

Note: Robust standard errors in parentheses. Statistical significance: * 10%.

in Appendix A). Nevertheless, the results provide some interesting insights. There is a strong positive correlation (0.87 *) between the ROAA and ROAE, expectedly indicating that higher returns on average assets often accompany higher returns on equity, suggesting consistent profitability across both measures. The NIM and long-term interest rate also have a significant positive correlation (0.37 *), implying that, as long-term interest rates increase, net interest margins tend to widen, likely benefiting institutions that rely on interest-based income. Another interesting correlation is between the long-term interest rate and the ratio of non-performing loans to gross loans (0.45 *), suggesting that, as long-term rates rise, the proportion of non-performing loans may increase, possibly due to higher borrowing costs affecting loan repayment. The GDP growth and long-term interest rate exhibit a negative correlation (−0.19 *), hinting that economic growth may slow as long-term rates rise, potentially due to tighter monetary conditions. Lastly, the share of loans in total assets and liquidity ratio have a notable negative correlation (−0.54 *), suggesting that institutions with a higher proportion of loans relative to assets may maintain lower liquidity, likely as more funds are committed to lending rather than liquid reserves.

4. Results and Discussion

In this section, we report the results of the two-step system GMM estimations. First, we present the results of the GMM estimations for the profitability measures (ROAA, ROAE, and NIM) in

Table 3. Results of the system GMM estimations for profitability measures as dependent variables.

	(1)	(2)	(3)
	ROAA	ROAE	NIM
Dependentt−1	0.134 ***	0.146 ***	0.731 ***
	(0.029)	(0.029)	(0.026)
Bank size	−0.182 ***	−1.456 ***	−0.093 ***
	(0.031)	(0.325)	(0.027)
Capital	0.031 ***	0.192 **	0.014 ***
	(0.009)	(0.077)	(0.004)
Loans share	0.001	−0.007	0.009 ***
	(0.001)	(0.016)	(0.001)
Credit risk (loan loss provisions/loans)	−0.247 ***	−2.312 ***	−0.015
	(0.036)	(0.334)	(0.010)
Liquidity risk (liquid assets/total assets)	0.002 **	0.019 **	−0.000
	(0.001)	(0.008)	(0.001)
Cost-to-income ratio	−0.018 ***	−0.174 ***	−0.005 ***
	(0.001)	(0.016)	(0.001)
Income diversification	0.195 ***	0.493	−0.213 ***
	(0.049)	(0.465)	(0.064)
HHI	0.000	−0.000	−0.000 ***
	(0.000)	(0.000)	(0.000)
Inflation rate	−0.033 ***	−0.231 ***	−0.023 ***
	(0.005)	(0.051)	(0.006)
GDP growth	0.007	0.056	0.011 ***
	(0.005)	(0.047)	(0.003)
Long-term interest rate	0.045 ***	0.373 ***	0.076 ***
	(0.009)	(0.081)	(0.008)
Constant	3.583 ***	33.465 ***	1.517 ***
	(0.475)	(4.847)	(0.366)
Observations	13,615	13,615	13,615
Number of banks	3045	3045	3045
AR1	0.000	0.000	0.000
AR2	0.078	0.26	0.301
Hansen test (p-value)	0.158	0.152	0.089

Source: Authors’ calculations. Note: Robust standard errors in parentheses. Statistical significance: *** 1% and ** 5%.

, followed by the estimations for the risk-adjusted performance measures in

Table 4. Results of the system GMM estimations for risk-adjusted profitability measures as dependent variables.

	(1)	(2)
	RAROAA	RAROAE
Dependentt−1	0.645 ***	0.381 ***
	(0.041)	(0.039)
Bank size	−0.435 ***	−0.814 ***
	(0.112)	(0.119)
Capital	−0.031 *	−0.128 ***
	(0.017)	(0.044)
Loans share	−0.023 ***	−0.035 ***
	(0.007)	(0.007)
Credit risk (loan loss provisions/loans)	−0.733 ***	−0.845 ***
	(0.157)	(0.151)
Liquidity risk (liquid assets/total assets)	−0.032 ***	−0.018 ***
	(0.005)	(0.004)
Cost-to-income ratio	−0.058 ***	−0.060 ***
	(0.005)	(0.005)
Income diversification	−0.062	0.013
	(0.139)	(0.176)
HHI	0.000	0.000 ***
	(0.000)	(0.000)
Inflation rate	−0.042 **	−0.035
	(0.021)	(0.024)
GDP growth	0.013	0.035 *
	(0.020)	(0.020)
Long-term interest rate	−0.062	−0.017
	(0.040)	(0.038)
Constant	14.007 ***	21.097 ***
	(1.904)	(2.113)
Observations	13,600	13,600
Number of banks	3042	3042
AR1	0.000	0.000
AR2	0.301	0.13
Hansen test (p-value)	0.122	0.098

Source: Authors’ calculations. Note: Robust standard errors in parentheses. Statistical significance: *** 1%, ** 5%, and * 10%.

.

As expected, the results for all three profitability measures exhibit a comparable pattern in terms of the coefficients and statistical significance. The lagged values of all three profitability measures are positive and statistically significant, indicating persistence in bank profitability. This finding is consistent with the results of several other studies on European banks (see, for example, Borroni and Rossi (2019), Goddard et al. (2013), and Horobet et al. (2021)). Djalilov and Piesse (2016) observed that profitability tends to persist over time, with the factors influencing bank profitability differing across transition countries. The coefficient is higher for the NIM, which may indicate that the net interest margin is a more significant source of profitability. This finding is also reported by Horobet et al. (2021) for CEE countries.

Regarding bank-specific measures, econometric estimations indicate a negative and statistically significant coefficient for bank size in terms of asset size, particularly for the ROAE. Similar findings have been reported by Athanasoglou et al. (2008), Berger (1995), Davis et al. (2022), Goddard et al. (2004), Korytowski (2018), and Pasiouras and Kosmidou (2007). The negative relation may indicate that large banks record diminishing returns to scale, potentially due to the greater complexity and higher fixed costs (Berger 1995; Goddard et al. 2004). This implies that, when a bank exceeds an optimal size, it becomes less efficient in utilizing resources, resulting in lower returns. Excessive size can also be an obstacle in adapting to a changing market and regulatory conditions, which are characteristic of the period under consideration. The results contrast with those of Menicucci and Paolucci (2016) and Radovanov et al. (2023), who identified positive effects of bank size on profitability.

Furthermore, estimations show that banks with a higher capital adequacy ratio tend to generate higher profitability, with the coefficient being stronger for the ROAE. Several recent studies reported a positive relationship for the ROAA and NIM, but not for the ROAE (Borroni and Rossi 2019; Davis et al. 2022). Djalilov and Piesse (2016) found that better-capitalized banks tend to be more profitable in early-transition countries, while Căpraru and Ihnatov (2014) demonstrated that growth in capital adequacy positively impacts bank profitability. This can be attributed to the fact that well-capitalized banks are more stable and better equipped to absorb economic shocks. As reported by the European Central Bank (2019), high bank capitalization improves banks’ resilience, preventing cuts in credit and the kicking-in of adverse second-round effects.

Regarding the loans share, the relationship is not statistically significant for the ROAA and ROAE; yet, it is positive for the NIM. The latter is expected, as a higher share of loans adds to the net interest margin through higher interest income. This is further supported by the findings of the European Central Bank (2024), particularly under favorable interest rate conditions.

Credit risk, measured by the ratio of loan loss provisions to total loans, is negatively associated with the ROAA and ROAE; yet, the association is not statistically significant with the NIM. This finding aligns with the literature on European banks, where credit risk is often cited as a primary factor affecting banking profitability (Borroni and Rossi 2019; Davis et al. 2022; Petria et al. 2015; Zou and Li 2014). On the other hand, liquidity risk, measured as liquid assets to total assets, has a small but significant positive effect on the ROAA and ROAE, and no effect on the NIM. This suggests that banks with higher liquidity buffers may be more resilient, and it can support banks to better respond to short-term funding needs. Similar effects, but stronger, were reported by Davis et al. (2022). In contrast, Kalanidis (2017) found that all liquidity measures had a negative impact on the ROAA, and the ROAE of the EU banks.

Furthermore, the cost-to-income ratio proxies the ability of banks to efficiently employ their resources to generate income. Our estimations show that the cost-to-income ratio is negatively associated with all three measures of bank profitability, which was an expected result and reported also by some other studies for European banks (Borroni and Rossi 2019; Davis et al. 2022; Elekdag et al. 2020; Goddard et al. 2013; Korytowski 2018; Mirović et al. 2024; Petria et al. 2015).

Income diversification is, according to our estimations, positively associated with the ROAA; yet, it has a negative relationship with the NIM. This aligns partially with the findings of Ammar and Boughrara (2019), Elekdag et al. (2020), Gambacorta and van Rixtel (2013), Goddard et al. (2013), Le and Ngo (2020), Nguyen et al. (2012), and Petria et al. (2015), who demonstrated that income diversification generally has a positive impact on bank profitability, although the specific type of diversification plays a role. This may suggest that banks that diversify their income sources (e.g., non-interest income) may achieve higher overall profitability but at the expense of a lower net interest margin. This could reflect a shift from traditional interest-based income to fee-based income, which may not have the same margin characteristics (Meslier et al. 2014).

In terms of external factors, our estimations show that market concentration, a banking-industry factor, negatively impacts the NIM; yet, the coefficient is very small. The negative relationship between market concentration and bank profitability was reported also by Athanasoglou et al. (2006), Căpraru and Ihnatov (2014), and Demirgüç-Kunt and Huizinga (2010). Economic conditions also play a significant role in influencing banks’ profitability. According to our estimations, GDP growth is positively associated with the NIM, aligning with the findings of Borroni and Rossi (2019), Davis et al. (2022), Mirović et al. (2024), and Petria et al. (2015), which identified a positive relationship between GDP growth and bank profitability. Our results highlight that economic growth stimulates loan demand, thereby directly enhancing both the quantity and quality of bank operations. Inflation, on the other hand, has a negative effect on profitability measures, a finding reported also by Korytowski (2018). Lastly, the long-term interest rate is positively associated with profitability measures. This is in line with the findings of Borroni and Rossi (2019), one of few studies that also includes interest rates in its estimations.

To check the robustness of our econometric specifications, we employed two alternative measures of risk—credit risk, measured as the ratio of non-performing loans to gross loans, and liquidity risk, measured as the ratio of loans to deposits and short-term funding. As above, also the alternative measure of credit risk had a significant negative effect on all three measures of profitability (the regression coefficients were −0.015 for the ROAA, −0.120 for the ROAE, and −0.007 for the NIM). The sign and statistical significance of other explanatory variables included in the model did not change compared to the estimates in Table 3.1 Similarly, the alternative measure of liquidity risk, i.e., the ratio of loans to deposits and short-term funding, also had the same sign as in the baseline regression; yet, the effect was significant only for the NIM.

Table 4 shows the results of the econometric estimation for the risk-adjusted performance measures RAROAA and RAROAE. The regression estimates are generally stable, with estimates for most explanatory variables like those for the profitability measures presented in Table 3. Specifically, bank size is negatively correlated with risk-adjusted profitability, suggesting that smaller banks achieve higher levels of risk-adjusted profitability. In contrast to the estimates in Table 3, the capital adequacy ratio and the share of loans in total assets are negatively associated with risk-adjusted profitability. This may indicate that banks with a higher capital adequacy ratio take less risk, which leads to a lower return on a risk-adjusted basis. Moreover, banks with a higher share of loans in total assets are more exposed to credit risk and may lack diversification into other activities. Similar findings have also been reported by Borroni and Rossi (2019). Credit risk, liquidity risk, and management efficiency have a negative effect on risk-adjusted profitability. These estimates align with the findings presented in Table 3, as well as with Borroni and Rossi (2019), with the exception for liquidity risk. The negative association between liquidity risk and risk-adjusted profitability measures may indicate that banks with a higher liquidity risk are less able to efficiently manage their short-term funding needs, leading to higher liquidity management costs, reduced financial stability and, ultimately, lower risk-adjusted profitability. GDP growth is only positively associated with the RAROAE, while inflation is negatively associated with the RAROAA. Long-term interest rates do not appear to have a statistically significant effect on risk-adjusted profitability measures.

The robustness check using two alternative measures of risk (credit risk, measured as the ratio of non-performing loans to gross loans, and liquidity risk, measured as the ratio of loans to deposits and short-term funding) confirms our baseline econometric estimates. Both the ratio of NPLs to gross loans and the ratio of loans to deposits have a negative effect on risk-adjusted profitability (the regression estimates for credit risk are −0.132 for the RAROAA and −0.114 for the RAROAE, both statistically significant; the regression estimates for liquidity risk are −0.01 for the RAROAA and −0.022 for the RAROAE, but only the second measure is statistically significant). The sign and statistical significance of the other explanatory variables included in the model did not change compared to the estimates in Table 4.2

In both Table 3 and Table 4, we report the Hansen and AR tests p-values. The Hansen test p-value > 0.05, indicating that the null hypothesis of instrumenst validity cannot be rejected, suggesting that the instruments used in the model are valid. Similarly, the AR(2) test p-values confirm that the residuals show no problematic serial correlation.

5. Conclusions

The findings of this study provide valuable insights into the determinants of profitability for European banks, with a particular focus on the period from 2013 to 2018, a period marked by significant economic recovery and regulatory transformation. This period, following the 2009 financial crisis, represents a critical phase as European banks adapted to the new Basel III regulatory requirements while navigating a gradually improving economic landscape. By examining bank-specific indicators such as capital adequacy, credit risk, bank size, and liquidity, along with industry and macroeconomic factors like the market concentration, GDP growth, and inflation, this study identifies the key drivers influencing profitability during this post-crisis recovery phase.

Capital adequacy stands out as a critical contributor to profitability, highlighting the importance of capital buffers for resilience under the new regulatory frameworks. However, when profitability is adjusted for risk, capital adequacy has a negative effect. While larger banks benefit from economies of scale, excessive size can reduce efficiency, a trend that has been exacerbated by the complexity of evolving regulations. In addition, credit risk, as measured by loan loss provisions and non-performing loans, has a negative impact on profitability, confirming that risk management remains essential for sustainable performance as banks face new regulatory pressures. Liquidity risk, on the other hand, only has a negative impact on risk-adjusted profitability, suggesting that maintaining liquidity buffers reduces returns on a risk-adjusted basis. The cost-to-income ratio also shows a negative effect, pointing out that operational inefficiencies have a direct impact on a bank’s ability to generate high returns. Macroeconomic conditions, in particular, GDP growth, have a positive effect on profitability, reflecting the economic recovery during the period and the increased demand for credit. The mixed effect of inflation and the positive correlation of long-term interest rates with profitability further underline how these banks responded to changing economic cycles during this recovery period. The selected period thus illustrates the dual challenge faced by European banks: increasing profitability while adapting to stricter capital and liquidity standards aimed at ensuring the stability of the sector.

Understanding the factors that influence profitability is not only of academic interest but also of practical relevance for regulatory policy and financial stability and also for bank management. Identifying which bank-specific, industry-specific, and macroeconomic factors have the most significant impact on profitability is crucial for designing effective policies that can help banks manage their risks while remaining profitable. For bank management, the findings can provide lessons for designing their resource allocation and risk management strategy to decrease their exposure to adverse shocks, with a positive impact on banks’ competitiveness.

This study provides valuable insights into the profitability dynamics of European banks during the post-2009 financial recovery. However, several areas need further exploration, presenting opportunities for future research. First, future studies should extend the analysis to include more recent periods, particularly examining the economic disruptions caused by the COVID-19 pandemic and the ongoing effects of geopolitical tensions. Second, a deeper investigation into the evolving regulatory framework is needed to understand its impact on bank profitability. Third, future research could explore the influence of monetary policy on profitability more comprehensively. Lastly, further studies could enrich the existing body of knowledge by examining the effects of fintech innovations on bank profitability.