1 Introduction

The sustainability of economic and social activities is a widespread concern and is of major importance nowadays. The United Nations proposed the 2030 Agenda for Sustainable Development, and announced 17 Sustainable Development Goals (UN, 2015). Although the plan presented to achieve these goals, the climate emergency and the uncertainty surrounding its effects pose new challenges to all economic actors, businesses, households, and governments. In particular, climate change will lead to financial losses that threaten financial stability and financial sector actors (BIS, 2020; Rudebusch, 2021). Recently, the ECB reviewed its monetary policy strategy considering the threat to environmental sustainability as an important challenge for the conduct of monetary policy (ECB, 2021a). Furthermore, the EU developed the European Action Plan on sustainable finance, which aims to encourage financial institutions to channel more capital into sustainable economic activities, imposing the measurement of their weight in the overall economic activities (Brühl, 2023).

Climate change induces physical risk due to the increasing frequency and severity of weather-related events, due to the long-term effects of change in climate patterns, and the transition risk associated with the uncertain financial impacts required to produce fewer carbon emissions (BIS, 2020; ECB, 2021b; Rudebusch, 2021). Both risks are interconnected (most pronounced for the energy sector) and lead to business disruptions, lower productivity, and income. In addition, the financial consequences include losses in financial and credit markets, falling equity and bond prices, and reduced property values (Rudebusch, 2021). Countries that address environmental and social issues are on the right path to ensure healthy economic growth (Peiró-Signes et al., 2022; Wang et al., 2020).

According to Yip and Bocken (2018), sustainable banking is defined as “the delivery of financial products and services, which are developed to meet the needs of people and safeguard the environment while generating profit” (see Aracil et al., 2021 for a literature review on sustainable banking). The intermediation activity of the banking sector is considered relatively clean. However, banks can improve the management of scarce resources if they use the information about the impact of financial activities on ecosystems (Scholtens, 2017). Bank products (e.g. credit) can considerably affect the environment (Jeucken, 2004). Also, the external environment can interfere with the activities of bank customers (Jeucken, 2004), creating credit risk due to the deterioration of borrowers’ ability to repay debt and the loss of collateral value, as well as an operational risk if the bank is physically affected and also other types of risk.

There is a growing trend in the literature that recognizes the importance of Environmental, Social, and Governance (ESG) scores in the valuation of assets and financial contracts and in determining the credit risk of countries and companies (Capelle-Blancard et al., 2019; Hoepner et al., 2016; Avci & Sungu-Esen, 2022; Hübel, 2022; Razak et al., 2020; Hock et al., 2020; Bannier et al., Rock 2021).

Moreover, the literature shows that country-level sustainability affects cross-border banking flows. For example, Avci and Sungu-Esen (2022) show that an increase in a given country’s sustainability scores increases banking fund flows from developed countries to that country, controlling for other determinants of international fund flows.

Our paper provides complementary evidence on the impact of country-level sustainability efforts on banking activity. First, we focus on the domestic credit market. We aim to show whether domestic sustainability efforts foster (or hamper) domestic lending by banks to the domestic non-financial sector and the impact on loan portfolio quality. Moreover, we focus on the cross-sectional differences between countries in recent times and ascertain the effect of the overall financial development and regulatory credit-favorableness regulatory framework on banks’ activity. Specifically, this paper aims to provide an empirical answer to the question: are country-level sustainability efforts related to banking activity? We focus on two dimensions of banking activity. The first is the amount of credit granted to the domestic sector. The second is the relative importance of nonperforming loans. We complement the work of Birindelli et al. (2022), who rely on the moderating role of countries’ environmental performance on banks’ credit risk. We also consider the sustainability pillars in addition to the overall score.

We find a positive association between country-level sustainability performance and the amount of domestic credit granted and a negative association between sustainability performance and the importance of nonperforming loans (NPL) on commercial banks’ balance sheets. Some of the pillars of sustainability also seem to have a relevant positive impact on credit activity.

Our work contributes to the literature in three different ways. First, to the best of our knowledge, it is the first paper to present a worldwide cross-country study that aims to document the importance of country-level sustainability for banks’ domestic lending and banks’ credit risk. Second, our approach focusing on market conditions complements the one that looks at individual banks’ sustainability concerns and practices, concerning risk assessment and loan pricing. Third, it explores the impact of specific pillars of sustainability, thus providing a better picture of the influence of specific aspects of sustainability on the banking sector’s credit and credit risk.

Next, the paper is organized as follows: Sect. 2 presents the related literature, and Sect. 3 describes the data, the variables, and the methodology. Section 4 presents and discusses the results. Finally, Sect. 5 provides the conclusion.

2 Related literature

Banks’ sustainable practices and internal dimensions of sustainability can affect their default risk and systemic risk and performance (e.g. Olmo et al., 2022; Moufty et al., 2021; Forcadell et al., 2020; Buallay et al., 2021; Yin et al., 2021). Lending decisions influence credit risk, and Weber et al. (2010) support incorporating sustainability criteria into credit rating and loan grant decisions. Carbon-neutral lending is negatively correlated with banks’ default risk (Umar et al., 2021), and Chinese private banks’ green lending practices reduce risk (Yin et al., 2021). Borrowers with lower carbon emissions have a better capacity to repay their debt: they are likely to have lower earnings and cashflow volatility due to their sustainable business model, thus benefiting banks with lower credit risk, lower loan loss provisions, and economic capital requirements (Umar et al., 2021).

Banks’ sensitivity to environmental protection contributes to higher efficiency in the use of resources and lending to more sustainable and competitive firms. It also improves banks’ reputation, attracting depositors, keeping their confidence, and increasing the intermediation margin (Gangi et al., 2019). Di Tommaso and Thornton (2020) find evidence of a minor negative relationship between ESG scores and bank risk, conditional on executive board characteristics, and between ESG scores and bank value.

A different approach states that the environmental and sustainability framework affects debt’s characteristics, costs, and risks, affecting the countries’ financial and banking sectors. Climate factors can damage corporate assets or reduce their profitability (Rudebusch, 2021) contributing to the deterioration in the borrower’s ability to service its debt. The ESG performance of a country is relevant to sovereign debt (Capelle-Blancard et al., 2019). It signals a country’s long-term commitment and may act as a buffer against shocks. Countries with higher ESG performance levels face lower risk and lower financing costs. ESG scores are also associated with flatter credit curves, signaling that these markets expect lower future increases in sovereign credit risk (Hübel, 2022) and the country’s legal system impact green bond yield spread (Frecautan and Ivashkovskaya, 2024). Higher country sustainability is associated with higher cross-border bank flows by Avci and Sungu-Esen (2022), who argue that sustainability is perceived as a positive signal of a better government and a less risky investment environment. Countries with better ESG performance potentially have more projects to finance, positively impacting bank lending. Banks factor climate risk into their decisions, reducing credit supply in polluted regions (Aslan et al., 2022). Therefore, country sustainability signals good borrowing quality to international lenders (Avci & Sungu-Esen, 2022).

Hoepner et al. (2016) show that higher sustainability in borrower countries is associated with a lower cost of bank loans. This mechanism protects the borrower from the operational and reputational hazards occurring from systemic social and environmental challenges and ultimately reduces its default risk. Therefore, country sustainability is a determinant of bank loans (Hoepner et al., 2016). Javadi and Masum (2021) argue that banks consider climate risk in their lending decisions. They find that banks consider climate risk when lending to firms. Thus, loans granted to borrowers more exposed to climate risk carry a significantly higher cost. From the above discussion, it is clear that country sustainability and climate risk influence the countries’ and firms’ cost of capital and that banks consider them when making lending decisions.

The climate conditions of the regions also impact households’ debt. This is particularly true for the case of mortgage loans (the majority of loans to households), as their collateral consists of immovable assets and is especially vulnerable to physical risks (Duan & Li, 2024; Calabrese et al., 2024). Increased disaster risk reduces the supply of mortgage credit, especially when securitization is not possible (Ouazad & Khan, 2022). Lenders reduce credit in the regions vulnerable to climate change, by lowering the rate of approval of mortgage applications, originating a lower amount of loans (Duan & Li, 2024). Mortgage interest rates are higher in the regions with more expected hot days (Baranyai & Banai, 2022).

Furthermore, ESG country scores are a possible determinant of bank risk. Environmental aspects related to conditions external to companies and households can threaten the borrower’s repayment capacity (Jeucken, 2004). For example, climate change and extreme events can cause high costs to economic activities. The value of collateral pledged by clients to the bank can be adversely affected (Jeucken, 2004). Country sustainability matters in the relationship between corporate sustainability and credit risk (Razak et al., 2020). Sustainability performance helps build up internal resources, reduces cash flow volatility, and improves credit risk profile (Razak et al., 2020). In the case of household loans, extreme weather events such as cyclones and heavy rainfall have a significant impact on mortgage defaults (Calabrese et al., 2024).

According to Capelle-Blanchard et al. (2019), the ESG performance of a country signals its long-term vision and is a buffer against shocks. Avci and Sungu-Esen (2022) view country sustainability as a signal of good borrowing quality. Other aspects of sustainability, such as social, labor, human rights, and cultural practices, reduce the vulnerability of economic activities and the population. Eliwa et al. (2021) show that country sustainability characteristics like the legal framework and the cultural system have a moderating effect on the relationship between firms’ ESG practices (performance and disclosure) and their cost of debt. The literature also reports that more sustainable and socially responsible companies have lower credit risk (Hock et al., 2020; Bannier et al., 2021). Banks that incorporate environmental sustainability into their lending policy should be able to select less risky and more profitable borrowers and alleviate the lender-borrower information gap (Gangi et al., 2019).

The literature already addresses the relationship between country sustainability and their debt (e.g. Capelle-Blancard et al., 2019) and between sustainability and international banking flows (e.g. Avci & Sungu-Esen, 2022). Sovereign debt markets affect corporate credit and equity markets (Hubel, 2022). The state of sustainability in a country moderates the banks’ credit risk (Birindelli et al., 2022).

However, to the best of our knowledge, the relationship between country-level sustainability and banks’ domestic lending activity (both the importance of loans and the loan portfolio quality) has never been directly tested in the literature. Banks’ financial intermediation is crucial for the economy since they mediate between agents with surpluses and those with deficits, providing the latter with the necessary funds for investment/consumption. Banks play a role in reducing asymmetric information issues. Climate risk and other sustainability issues affect the extent to which companies are affected by information asymmetry and banks may be unable to properly assess the magnitude of the additional costs and decrease credit supply (Aslan et al., 2022). Therefore, they can incorporate ESG criteria in their assessment of potential borrowers and projects. As a result, better countries’ ESG performance contributes to a higher number of projects to be financed, which is enhanced by international lenders (Avci & Sungu-Esen, 2022). This is in line with the growing trend that recognizes the importance of ESG in the valuation of every asset class and financial contract (Hoepner et al., 2016). Nevertheless, the impact of country-level sustainability on domestic bank lending has never been explored.

Therefore, we posit that banks factor sustainability risks into their lending decisions and that a positive relationship exists between country-level sustainability and loans granted by the domestic banking sector to the domestic private sector.

On the other hand, as discussed above, sustainability issues are a possible determinant of the quality of bank loans, measured by nonperforming loans (NPL). Moreover, following the Global Financial crisis and the European sovereign debt crisis, the literature on the macroeconomic determinants (e.g. GDP growth, inflation, unemployment, among other variables) of NPL expanded (e.g. Castro, 2013; Dimitrios et al., 2016; Louzis et al., 2012; Beck et al., 2015; Radivojević et al., 2019). In addition, natural disasters also impact NPL positively (in the short and long term). Still, the impact is not significant in OECD countries, equipped with better infrastructure and supporting financial systems (Chen et al., 2022). Furthermore, Birindelli et al. (2022) report that a country’s environmental performance index moderates the banks’ NPL: banks located in countries with higher environmental performance index tend to reduce their credit risk more substantially than banks located in countries with a lower environmental performance. Therefore, we expect a negative relationship between country sustainability and NPL. If we confirm this hypothesis, we can conclude that sustainability contributes to banking and financial sector stability.

3 Research design

We use three primary data sources. First, for country-level sustainability scores, we use SolAbility Sustainable Intelligence data (https://solability.com). The Global Sustainable Competitiveness Index (GSCI) is computed based on data provided by the World Bank, various United Nations agencies, the International Monetary Fund, and other non-governmental organizations. According to SolAbility, this index incorporates information that ensures business success and shapes future potential/decline (SolAbility, 2020). Another advantage of SolAbility indexes is their comparability across countries. Five pillars compose the GSCI: (i) the Natural Capital pillar characterizes the physical environment of a country; (ii) the Social Capital pillar recognizes that nations need a minimum level of social cohesion and solidarity; (iii) the Intellectual Capital pillar measures the capability to generate wealth and jobs through innovation and value-added industries; (iv) the Governance pillar characterizes the physical and non-physical attributes of the country’s framework as the basis for business; and (v) the Resource Intensity pillar measures the efficiency of using available resources.

The GSCI scale and the sub-indexes scale range from 0 to 100, the highest values signal improved performance in the respective dimension (see SolAbility (2020) for more details).

Second, for data on Loans (domestic loans of deposit money banks as a percentage of GDP) and NPL (nonperforming loans as a percentage of gross loans), we rely on the World Bank’s WDI - World Development Indicators database (Weill, 2011; Bougatef, 2016; Acheampong & Elshandy, 2021).

Considering the data availability of the two previous databases, we obtained our initial samples: one with 162 countries for Loans and the other with 130 countries for NPL.

The countries’ macroeconomic data, GDP growth and inflation (e.g. Radivojević et al., 2019; Capelle-Blancard et al., 2019), are also drawn from the WDI database (except for the data on Inflation for the Union of Comoros and the Republic of Uzbekistan, which come from the International Monetary Fund Country Information website). Data on two variables that control for institutional quality are also drawn from the WDI database (Bougatef, 2016). Specifically, (i) depth of credit information index (CredInfo) - rules affecting the scope, accessibility, and the quality of available credit information; and (ii) strength of legal rights index (LegRight) - measures the degree to which collateral and bankruptcy laws protect the rights of borrowers and lenders. Adding the macroeconomic data marginally reduced our samples to 156 countries (Loans sample) and 129 countries (NPL sample).

Finally, we rely on the Global Financial Development (GFD) database, also from the World Bank, to source the data on the controls for the characteristics of each country’s banking sector. Specifically, we control for banking sector concentration (assets of the five largest banks as a share of total commercial banking assets), accessibility (number of Automated Teller Machines (ATM) per 100,000 adults), and financial stability (the z-score compares the buffer of a country’s commercial banking system - capitalization and returns - with the volatility of those returns). As the available data on these additional controls significantly reduced our samples to 112 countries (Loans sample) and 92 countries (NPL sample), we opted to include them in the second stage of additional tests.

Because SolAbility raises severe concerns about directly comparing sustainability scores before 2018 due to index methodology changes (see SolAbility (2018), and to leave the impact of the COVID-19 health crisis aside, we look at a cross-section defined by the average values for the two years 2018 and 2019. With this procedure, we focus on the effect of interest (cross-sectional variation), also eliminating the impact of idiosyncratic short-term fluctuations (e.g. Weill, 2011), and interpret this recent cross-section as representative of a period of relative worldwide stability in the banking sector. Table 1 summarizes our variables and data sources.

Table 1 Description of variables
Full size table

For estimation, we run the following cross-sectional OLS regression (base model):

$$\matrix{{{{\rm{Y}}_{\rm{i}}}\;{\rm{ = }}\;\alpha \;{\rm{ + }}\;{{\rm{\beta }}_{\rm{1}}}{\rm{Scor}}{{\rm{e}}_{\rm{i}}}\;{\rm{ + }}\;{{\rm{\beta }}_{\rm{2}}}{{\rm{X}}_{\rm{i}}}\;{\rm{ + }}\;{ \varepsilon _{\rm{i}}}} \hfill \cr {\rm{ }} \hfill \cr}$$
(1)

where Yi refers either to Loans or nonperforming loans (NPL) of country i and Scorei is our variable of interest, the score of a given country i on either the Global Sustainable Competitiveness Index or on each one of its components. Xi is a vector of control variables (see Table 1 for details), and εi is the error term. Following the literature, we expect β1 to be positive in the Loans regression and negative in the NPL regression, thus corroborating the hypothesis that country-level sustainability efforts foster commercial banking lending activity. All t-statistics (t-stat) are estimated with heteroscedasticity robust standard errors.

4 Results and discussion

Table 2 presents the main results. Overall, we find a positive association between country-level sustainability performance and the amount of credit granted and a negative association between sustainability performance and the importance of NPL on commercial banks’ balance sheets. For example, specification 1 in Table 2 highlights that an increase by one point in the global sustainability index is associated with an increase of approximately 3.6% points in the relative amount of domestic loans (as a percentage of GDP). In other words, our results suggest that stronger sustainability frameworks foster lending activity.

Table 2 Global sustainability and lending activity
Full size table

Likewise, specification 3 of Table 2 shows that an increase of one point in the GSCI is associated with a decrease in the proportion of nonperforming loans of 0.61% points. Thus, our results suggest that higher sustainability performance helps hamper bad loans.

Moreover, our results are robust to the effect of several controls. Regarding the macroeconomic controls, inflation shows a negative (and statistically significant) influence over bank loans. This result is compatible with the findings of Weill (2011) and reflects the reduction in productivity levels that inflation causes. The relationship between our dependent variables and GDP growth is not statistically significant, in line with the results of Capelle-Blanchard et al. (2019) and Hübel (2022)Footnote 1.

The depth of credit information index presents a positive (and statistically significant) effect on loans and a negative (and statistically significant) effect on NPL (similar to Bougatef, 2016), suggesting that the availability of data that reduces information asymmetry between lenders and borrowers facilitates domestic credit and improves its quality.

The dummy variable indicating G20 membership shows a positive effect on the importance of loans and a negative effect on the quality of loans, suggesting that the banking system of the G20 countries is more developed, contributing to a strong and stable loan market. This result remains statistically significant in all specifications. On the other hand, GDP growth and the strength of legal rights control do not present statistical significance in the estimated models.

As the adjusted R2 results show, we have more explanatory power in the Loans model than in the NPL model, and with magnitude in line with other published results (e.g. Eliwa et al., 2021; Birindelli et al., 2022; Sol Murta & Gama, 2022). This reinforces the objective of establishing an empirical relationship between country-level sustainability efforts and the domestic banking market activity, and does not fully explain the cross-sectional variation in Loans or NPL.

Table 3 looks at the impact of specific pillars of sustainability on Loans (columns under Loans) or NPL (columns under Nonperforming loans). To conserve space, we present the results for the model using the largest sample. However, unpublished results, available upon request, show that when we add the bank-specific controls, thus reducing sample size, the statistically significant relationships between sustainability pillars and Loans identified in Table 3 remain qualitatively similar.

Table 3 Sustainability pillars and lending activity
Full size table

Concerning Loans, the social capital, the intellectual capital, and the governance pillars have a relevant positive influence on credit activity, similar to that of the overall global sustainability index. The social capital pillar has the highest coefficient amongst the coefficients of those three statistically significant pillars. The influence of social cohesion in a country and the ability to generate jobs and wealth are similar. These results are consistent with Avci and Sungu-Esen (2022), who find that sustainability’s environmental, social, and governance pillars have a statistically significant link with their dependent variable. Our study finds no statistically significant coefficients for the natural capital and resource intensity pillars.

Concerning NPL, only the governance efficiency and the social capital pillars show a statistically significant negative relationship with loan quality.

Tables 4 and 5, and 6 present the results of six robustness tests focusing on the overall country-level sustainability scoreFootnote 2.

Table 4 Robustness tests I
Full size table
Table 5 Robustness tests II
Full size table
Table 6 Robustness tests III
Full size table

In Table 4, first, to shed further light on comparable samples regarding country coverage, we look at the subsample of the 123 countries for which the SolAbility data, both our dependent variables and the macroeconomic controls, are available. Second, because of the significance of the indicator variable for G20 membership, we look at the subsample of non-G20 countries (data availability renders the estimation for the G20 subsample unfeasible).

Results presented in Table 4 corroborate our basic conclusion for a positive effect of sustainability scores on the amount of loans and a negative effect on the (low) quality of the loan portfolio, both in the same sample size specification as in and in the non-G20 countries specification, however slightly reduced, compared to the results in Table 2. Concerning the control variables, Credinfo loses statistical significance in the Loans model, and G20 in the NPL model.

In Table 5, first, to check the robustness of our results to the particular definition of sustainability captured by the Global Sustainable Competitiveness Index from SolAbility, we replace the GSCI score with the average Global Index Score from the Sustainable Development Report for the two years from 2018 to 2019 (see Sachs et al., 2018, 2019). Second, to gain further insight into the causal effect of sustainability performance on credit market activity, we re-estimate our model using the GSCI for 2017 (Lagged Sustainability).

Again, either the different sustainability indicator or the country’s lagged sustainability corroborates our baseline conclusions, and with increased explanatory power in the case of the Loans model. Moreover, the effect of the control variables remains qualitatively similar, when compared to Table 2.

In Table 6, we first add the aforementioned banking sector-specific controls to the regression, which significantly reduces our sample size. Second, to gain further insight into potential endogeneity bias due to reverse causality or omitted variable bias, we follow the intuition of Jiraporn et al. (2014), Boubaker et al., 2020, Wang et al. (2020), and Avci and Sungu-Esen (2022), and use as an instrument for sustainability the 2017 hand-calculated average sustainability score of land border countries (for islands we use the maritime border countries), and estimate the complete model using two-stages least squares. The sustainability efforts of neighbouring countries in the previous year are likely to be related to a country’s sustainability efforts, as sustainability requires internationally coordinated efforts (Wang et al., 2020), but are less likely to drive a country’s domestic banking sector performance in the upcoming years. As Jiraporn et al. (2014) show, CSR scores of neighbouring firms relate positively to the CSR scores of a given firm but not to a financial dimension, credit ratings. Moreover, the highly significant Kleibergen-Paap robust F-statistic shows that our instrument does not suffer from the weak instrument problem.

The results presented in Table 6, corroborate our baseline conclusion of a positive effect of sustainability efforts on the amount of domestic credit granted by banks, and a negative effect on the (low) quality of bank loans are highly robust. Moreover, concerning the banking sector-specific controls, the z-score, an indicator of banks’ financial health, shows a statistically significant relationship with lending activity. Financial health is positively associated with the amount of credit granted and negatively associated with the proportion of bad loans. Our results also show that the number of ATMs per 100,000 adults has a positive relationship with credit granted and that bank concentration helps hamper bad loans.

For consistency with previous larger samples (156 observations for loans and 129 observations for nonperforming loans), in unpublished results to conserve space but available upon request, we show that the positive (negative) statistically significant relationship between country-level sustainability and Loans (NPL) remains statistically significant with z-stat 3.44 (-1.92) and Kleibergen-Paap robust F statistic for weak identification of 77.06 (99.61), respectively.

In summary, our results highlight the importance of sustainability in driving the domestic credit market: it facilitates lending and reduces the importance of NPL, thus contributing to the quantity and the good quality of loans.

5 Conclusion

This paper addresses the importance of countries’ sustainability performance for the banking sector activity by looking at bank loans’ quantity and quality. Using a sample of sustainability scores for a large cross-section of countries, we present empirical evidence that the countries’ sustainability efforts help to foster the domestic credit market in two ways: by increasing the amount of domestic credit granted and by improving the quality of loans.

Therefore, the sustainability of countries fosters funding of economic activities and contributes to reducing banks’ credit risk. Our results suggest important policy implications. Firstly, governments, central banks, and non-governmental organizations should continue to reinforce their concerns about sustainability practices because, in what concerns banking organizations, a framework of higher sustainability fosters the domestic credit market, thus contributing to banking stability and enhanced economic development. Secondly, our results suggest that those policy measures that contribute to social cohesion and solidarity (e.g. reinforcement of women in management positions), innovation and value-added industries (e.g. increase in education expenditure), and the country’s governance performance (e.g. quality of public services), have a relevant impact on the smooth functioning of credit markets. Thirdly, it is also important that these organizations continue to compute global sustainability indicators directly comparable across countries and regions, thus contributing to the study of sustainability issues over a longer period, as these will become increasingly important in the coming years.

Future studies could look at different measures of country sustainability and further analyze the importance of the components of sustainability measures for the lending activity of financial intermediaries, thus accurately identifying which environmental, social, and governance aspects are the most relevant. In addition, the impact of sustainability on firms’ financial policies, namely the debt policy, would allow for a better understanding of the demand side of credit markets.