Open AccessArticle

Environmental, Social, and Governance Scores and Loan Composition Inside United States Banks

Silvia Bressan

Department of Economics and Management, Free University of Bozen, 39100 Bozen, Italy

Sustainability 2024, 16(18), 8075; https://doi.org/10.3390/su16188075

Submission received: 30 July 2024 / Revised: 11 September 2024 / Accepted: 13 September 2024 / Published: 15 September 2024

(This article belongs to the Special Issue Sustainability and Financial Performance Relationship)

Download Versions Notes

Abstract

We analyze the loan portfolios of United States banks from 2013 to 2023, showing that high environmental, social, and governance (ESG) banks have larger shares of consumer loans and commercial loans and smaller shares of construction loans and real estate loans. We also find that the governance pillar (G) is more tightly related to the bank loan composition compared to the environmental (E) and social (S) pillars. Furthermore, we show that construction loans and real estate loans decrease more considerably with bank ESG scores inside countries with high gas emissions, i.e., where ESG issues would arguably be more serious. Our interpretation is that sustainable banks are reluctant in lending money for real estate projects, exposing them to potentially high ESG risk. These findings contribute to developing a deeper insight about the relationship between ESG and bank lending, which, in the previous literature, has been treated more frequently in aggregate terms instead of separating loan types. Our outcomes suggest that sustainability is crucial for the availability of funds in the real estate sector, delivering important insights to bank and real estate managers, besides policy makers.

Keywords:

ESG scores; bank loans; loan composition; real estate loans; consumer loans; commercial loans

JEL Classification:

G21; M14

1. Introduction

The growing attention on the topic of corporate sustainability inside banks poses the question of whether Environmental, Social, and Governance (ESG) scores affect their lending. The United Nations-backed Principles for Responsible Investment recognized that ESG factors are important elements for assessing the creditworthiness of borrowers, as they can affect borrowers’ cash flows and the likelihood of defaulting on debt obligations (www.unpri.org/about-us/what-are-the-principles-for-responsible-investment (accessed on 24 August 2024)). The previous literature shows that also banks’ own ESG ratings have an impact on loans, having an influence on credit growth [1], credit quality [2], and long-term relationships with borrowers [3].

Nonetheless, we lack a more comprehensive understanding of how banks’ sustainability plays a role in lending. In particular, there is little knowledge about the association between the composition of banks’ loan portfolios and ESG performance [4]. While the previous literature uses aggregate loan amounts more frequently, we have no evidence of whether this relationship depends on the loan “type”, i.e., the specific business that the loaned money will finance. To contribute to filling this gap, we analyze United States banks from 2013 to 2023, separating their total loans into consumer loans, commercial loans, construction and development loans, real estate loans, foreign loans, and other unclassified loans. Our main result is that the shares of consumer loans and commercial loans in the loan portfolio increase in banks’ ESG ratings. In contrast, the shares of real estate loans and construction loans decrease. We also find that the governance pillar (G) is more tightly related to the bank loan composition compared to the environmental (E) and social (S) pillars. Moreover, we show that construction loans and real estate loans decrease more considerably with bank ESG scores inside countries with high gas emissions, i.e., where ESG issues would arguably be more serious. Our intuition is that high-ESG banks lend a lower amount of money to real estate projects compared to other investments because real estate businesses exhibit severe ESG-related risks.

These findings highlight the importance of addressing sustainability in the real estate sector, i.e., an industry pivotal for the transition to ESG. In fact, real estate assets have a serious impacts on climate change. A total of 20% of the total global greenhouse gas emissions originate from buildings. In addition, the real estate industries are impacted by climate change, consuming over 40% of global energy annually: 28% from operational emissions (i.e., energy needed to heat, cool, and power buildings) and 12% generated from materials and construction (www.weforum.org/agenda/2020/01/real-estate-finance-build-in-sustainability/ (accessed on 24 August 2024)). The need for more environmentally friendly practices in the real estate sector leads regulators across the globe to increasingly introduce new legislation around ESG and sustainability measures for real estate. For example, the World Economic Forum has developed a common set of environmental principles in partnership with the real estate industry. The goal of this effort is to ensure that the decision making and operations of real estate firms place a high priority on becoming environmentally sustainable (www.weforum.org/publications/environmental-sustainability-principles-for-the-real-estate-industry/ (accessed on 24 August 2024)). Practitioners too recognize the critical function of the real estate sector, not only for the environment, but also for the social element. For example, real estate has the power to positively impact local communities through communal facilities that help to create a positive image that can, for example, attract shoppers and bolster the local economy (kpmg.com/lu/en/blogs/home/posts/2023/03/esg-shaping-the-real-estate-landscape-to-come.html (accessed on 24 August 2024)).

Finally, our results are relevant in relation to the key role of the real estate sector for systemic stability. In fact, the European Systemic Risk Board publishes reports on potential systemic risks arising in the financial sector (www.esrb.europa.eu/pub/reports/html/index.en.html (accessed on 24 August 2024)). Based on regulatory concerns, previous research demonstrates that financial institutions are highly exposed to real estate risk [5,6,7], with negative consequences on systemic risk [8,9,10]. According to [11], the idiosyncratic risk exposure of banks also becomes larger when their share of real estate loans increases. Therefore, our insight is that improvements in socially responsible activities and ESG values in the real estate sector could increase the bank financing of real estates. This behavior would enhance the resiliency of firms and of the financial system too.

This paper is organized as follows. Section 2 reviews the literature related to this article. Section 3 presents the data and the research design. Section 4 outlines and discuss the results. Section 5 concludes.

2. Review of the Literature

The framework of our article is the relationship between ESG and banks’ lending. In this stream of research, a few papers have investigated the impact from corporate social responsibility and the loan quality. For example, ref. [12] studies the loan losses of United States banks in the years 1993–1994, finding evidence of a positive association between their financial performance and social performance. Analyzing worldwide banks from 2003 to 2009, ref. [13] shows that more responsible firms have lower amounts of non-performing loans. A similar result is documented based on observations of European banks [14] and United States banks [2]. Using Chinese banks, ref. [15] reports that banks with better ESG performances give more credit and have a lower probability of collateral requirements. In addition, the outcomes of [16] from United States firms display that good corporate social responsibility policies can improve credit ratings, potentially also improving the access of firms to bank funding. Nonetheless, the literature does not entirely agree on the fact that ESG has a positive impact on bank lending, as the evidence is heterogeneous. For example, ref. [17] finds that sustainable banks in Europe over the period 2002–2020 had a poor loan quality.

A few articles advance the hypothesis that the connection of ESG to bank lending may depend on market conditions, studying effects throughout different market phases. For example, ref. [1] shows that, during crises, the loan growth of sustainable banks increases more greatly than less sustainable institutions. The outcomes suggest that high-ESG banks are less affected by episodes of market distress in terms of credit risk, asset risk, and profitability, being also subject to cheaper funding costs. Similarly, ref. [18] finds that the lending of banks engaged in ESG activities is countercyclical, as high-ESG banks seem to stabilize credit during difficult times. Nonetheless, in contrast to this evidence, ref. [19] reports that ESG had a negative impact on banks’ credit during the recent COVID-19 pandemic.

A few articles report that the ESG profiles of the bank and customer contribute to shaping the lender-0borrower relationship. For example, ref. [3] finds that the bank’s own ESG standards affect the type of borrowers that it works with, and also has an effect on the borrower’s ESG performance over time. Corporate social responsibility seems to also be determinant for the cost of bank loans. For example, ref. [20] uses loans from United States banks to show that spreads are higher for customers with poor ESG quality and low levels of collateral, arguing that, in the absence of security, banks become more sensitive to corporate social responsibility concerns. Similarly, ref. [21] shows that low-ESG banks offer smaller loan spreads to sustainable borrowers in response to the pressure from stakeholders willing to incorporate ESG considerations. Nonetheless, the evidence is mixed, as [22] discovers that, for Chinese firms, the relationship between loan spreads and corporate social responsibility is non-linear, following an inverse U-shape, i.e., firms with extremely low or extremely high corporate social responsibility levels are subject to lower costs of bank credit. The lender–borrower relationship may also be affected by so-called “ESG washing”, which occurs when a bank with worse ESG performance intentionally lends to a firm with a better ESG performance to earn a better reputation. Ref. [23] demonstrates that this behavior results in low-ESG banks giving more loans at cheaper costs to high-ESG borrowers, also benefiting from a surge in the bank’s stock price when the loan deal is announced.

Finally, based on Chinese firms, ref. [24] proves that the institutional environment moderates the access to bank loans. In fact, in regions with less government intervention, firms with high corporate social responsibility quality have access to larger bank credit loans. In regions with high financial development instead, bank credit availability is not significantly related to corporate social responsibility. Furthermore, in China, [25] documents that the effect of corporate social responsibility on bank loans is stronger at firms with a higher degree of favoritism by senior officials, a particular kind of behavior in China’s relational society.

Evidently, the topic of ESG and bank loans is still debated in the literature, and many aspects are not covered extensively. The previous studies report evidence that bank ESG performance plays a role in lending in terms of the amount and quality of credit. However, none of the studies that we cite consider if this behavior may be related to the “type” of loans. Therefore, our article aims to fill this gap in knowledge, testing the impact from bank ESG scores on different loans, classified in terms of the specific type of business that the loaned money will finance.

3. Research Design

3.1. Data and Variables

We use data from S&P Capital IQ. Our data provider classifies companies according to industries and geographies. Therefore, we select banks in the United States, downloading their ESG scores from 2013 to 2023. The S&P Global ESG scores are based on the S&P Corporate Sustainability Assessment (CSA), on information provided directly to S&P and certified by S&P analysts, and on public domain information. The CSA is an annual evaluation of companies’ sustainability practices, focusing on sustainability criteria that are both industry-specific and financially material. For more information about the S&P CSA, see https://www.spglobal.com/esg/csa/policies-guidelines (accessed on 28 August 2024). The bank’s ESG score (

E S G

) goes from 0 to 100, and a higher score indicates that the bank has better ESG characteristics. We also test the separate environmental (E), social (S), and governance (G) pillars, which are all in the range 0–100.

From S&P Capital IQ, we download balance sheet figures with annual frequency. The focus of our analysis is the bank loan composition. Loan amounts (in USD) are classified according to the “type” of business. Specifically, we have separate items for consumer loans (

C O N S

), commercial loans (

C O M M

), construction and development loans (

C O N S T R

), real estate loans (

R E

), foreign loans (

F G

), and other unclassified loans (

O T

). For each loan type, we consider its total amount in percent of total bank loans. The definition of all our variables and their descriptive statistics are reported, respectively, in Table 1 and Table 2. We observe that the loan portfolios of our banks are more largely made by real estate loans (approximately 64% of total loans) and commercial loans (approximately 27% of total loans).

Our regressions will test the effect from ESG scores on the loan composition. In the set of controls, we include bank-specific aspects that may affect bank lending more broadly. Following [1], we control for profitability using the the return-on-assets (

R O A

), computed as net total income divided by total assets. Evidence shows that

R O A

reflects the quality of the bank assets [26,27,28]. A more profitable bank will be subject to fewer credit losses and will avoid increasing loss provisions considerably. This, in turn, could have a positive impact on bank credit. The ratio of total equity capital to total assets (

E Q U I T Y

) accounts for leverage, such as, for example, in [1,18]. A high level of capital (i.e., high

E Q U I T Y

) indicates that the bank is more solid and has good prospects of growth and loan availability. The variable

L O A N S

is the ratio of total loans to total assets and accounts for the firm business model [29]. When a bank has a high

L O A N S

, the business is mainly dedicated to loan making. A high number might mean that a bank’s liquidity is lower and more exposed to higher defaults [28].

3.2. Model

The goal of our model is to verify the effect from ESG scores on the bank loan portfolio. We refer to models in the literature that relate ESG ratings to measures for loan growth [1,18] or loan quality [2]. Nonetheless, in all these models, the dependent variables use aggregate measures of loans, while we now want to explain how different loan types depend on the bank ESG scores. Therefore, we estimate the following regressions, testing whether the ESG rating (

E S G

) of bank j at time t contributes to determining the share of the loan portfolio given to loan type l (i.e.,

C O N S

C O M M

C O N S T R

R E

F G

, and

O T

\begin{matrix} {L O A N T Y P E}_{l, j, t} = α_{l, j, t} + β_{l, j, t} E S G_{j, t} + {γ 1}_{l, j, t} E Q U I T Y_{j, t} + + δ_{l, j, t} R O A_{j, t} + + λ_{l, j, t} L O A N S_{j, t} + ζ_{l, t} + ϵ_{l, j, t} . \end{matrix}

(1)

The set of controls include profitability (

R O A

), leverage (

E Q U I T Y

), and the business model (

L O A N S

) that we outlined in the previous subsection. The term

ζ

captures year fixed effects, while

ϵ

is the regression error. Standard errors are clustered at the bank level. We checked that the results remain similar as

E S G

and the other regressors are one-period lagged.

4. Results

4.1. Regression Outcomes

Table 3 shows that the shares of consumer loans and commercial loans of our banks increase in their ESG scores. Oppositely, the shares of construction loans as well as real estate loans decrease. This impact is more pronounced for real estate loans, as the coefficient of

E S G

in the equation for

R E

(equal to −0.969) is seven times higher than the coefficient in the equation for

C O N S T R

(equal to −0.137).

Concerning the control variables, we notice that

E Q U I T Y

has the strongest effect. While real estate and construction loans are associated with high levels of capital, consumer and commercial loans follow the opposite pattern. This finding suggests that banks need to have huge equity cushions to finance real estate projects that are exposed to substantial real estate and ESG risks. The sign on

R O A

is significantly negative on

R E

, revealing that low-quality (i.e., less profitable) banks also have larger shares of real estate loans in their portfolios. Consistently, the positive sign from

L O A N S

R E

indicates that banks lending money for real estates also have a business model more focused on traditional lending, which could increase the exposure to default risk considerably.

We explore the assets of our banks more deeply, and, in Table 4, we separate the figure of consumer loans into home equity loans, credit card loans, and vehicle loans. The coefficient on

E S G

is always positive, but statistically significant only for credit card loans. In the same vein, in Table 5, we decompose real estate loans into the categories of one to four family loans, multifamily loans, and commercial real estate loans. We find that the negative sign on

E S G

is significant only in the regression for commercial real estate loans, while the impact on family mortgages is not significant.

We now focus on the loans for which the sign on

E S G

estimated by the previous regressions was significant (i.e., construction loans, real estate loans, commercial loans, and credit card loans) and test the separate impact from the environmental (E), social (S), and governance (G) pillars. In Table 6 and Table 7, the three variables are highly significant, with the governance and social elements having slight higher coefficients than the environmental element. These findings confirm that all aspects of bank sustainability (as measured by corporate ESG scores) are relevant to explaining bank loans, with the governance pillar having a primary role.

To provide further support to the argument that the banks’ ESG profiles explain loan portfolios, we use information about United States greenhouse gas emissions. We expect that, for states with greater greenhouse gas emissions, banks would perceive ESG-related risks to also be more acute. Thus, our conjecture is that high-ESG banks would be more reluctant to lend money for real estate and construction projects compared to states with lower gas emissions.

Using the information about greenhouse gas emissions in the United States from 2013 to 2021, in every year t, we separate low/high greenhouse gas emissions states. Greenhouse gases include carbon dioxide, fluorinated gases, methane, and nitrous oxide (cfpub.epa.gov/ghgdata/inventoryexplorer/#allsectors/allsectors/allgas/select/all (accessed on 27 August 2024)). Note that the sample that we used for all the previous regressions went until 2023. Here, given limitations in the data available, our sample goes until 2021. For low (high) greenhouse gas emissions states, the amount of emissions is below (above) the median greenhouse gas emissions across all of the United States at time t.

In Table 8 columns (1)–(2), we estimate regressions of construction loans and real estate loans on

E S G

interacted with a dummy variable that identifies low/high greenhouse gas emissions states. Concerning construction loans, we find that the negative sign on

E S G

is statistically significant only for high gas emissions states, while the sign is not relevant for low gas emissions states. Moreover, the impact from ESG scores on real estate loans is significantly more negative inside states with high greenhouse gas emissions. Finally, in columns (3)–(4), we notice that the effect from

E S G

on the shares of consumer loans and commercial loans is more positive inside states with high greenhouse gas emissions. Overall, these outcomes corroborate our argument that ESG-related issues lead sustainable banks to channel less credit to real estate projects, while giving more weight to consumer loans and commercial loans in their asset portfolios.

4.2. Discussion

We show that high-ESG banks give less credit for construction and commercial real estate investments. Overall, our findings cast attention on bank funding to the real estate sector. Our intuition is that sustainable banks are reluctant in financing real estate or constructions potentially harming their financial stability. In fact, the previous literature corroborates this argument, proving that real estate projects pose serious concerns about real estate risk [5,6,7] as well as ESG risk [3,4]. In turn, ESG plays a key role in solvency, as [30] demonstrates that ESG predicts financial distress. The authors report that good ESG profiles are inversely related to default probabilities and also reduce the likelihood of misclassifying distressed banks as healthy.

Moreover, existing research shows that high-ESG banks entail low idiosyncratic risk [31] and that real estate loans can instead increase the firm-specific risk exposure [11]. This evidence corroborates our argument that sustainable banks are not willing to take on additional risk by funding the real estate sector.

A second aspect to consider while discussing our findings concerns the role of reputation. A bank that lends to projects subject to high-ESG risks would be likely to also face higher costs for dealing with regulatory scrutiny, eventually harming its reputation and decreasing its opportunity to develop future business [3]. A concrete example may involve Bank of America, which announced that it would stop lending money to gun manufacturers that chose to continue the production of military-inspired firearms for civilian use after the high school mass shooting in Parkland (Florida) in February 2018 (money.cnn.com/2018/04/11/news/companies/bank-of-america-guns/index.html (accessed on 26 August 2024)). Therefore, our conjecture is that a bank with a good ESG standing would be reluctant regarding the funding of real estate projects that could affect its reputation in a negative way.

5. Conclusions

We analyze United States banks from 2013 to 2023 and analyze the relationship between their ESG scores and loan composition. We find that high-ESG banks have smaller shares of construction loans and real estate commercial loans compared to low-ESG banks. Furthermore, high-ESG banks also have larger shares of consumer loans and commercial loans. We observe that this pattern is more evident inside those states with high gas emissions, i.e., where ESG concerns would be more severe. Overall, our interpretation is that sustainable banks are reluctant in lending money for real estate investments that would increase exposure to ESG risk, besides real estate risk.

These findings raise the attention on the construction and real estate sector, i.e., an industry in which ESG has emerged as a high priority. In fact, from the construction of new properties to the upkeep and performance of existing ones, there are high levels of energy consumption and a strong reliance on fossil fuels. In addition, the social aspects of buildings in our society is not negligible. Our intuition is that the amount of bank credit available for real estate projects could be made larger by improving sustainability standards, making properties not only energy-efficient but also having a more positive social impact. Evidently, this insight is important for bankers and real estate managers.

Nonetheless, we acknowledge that our approach presents some limitations, and that few extensions could make our insights more robust. One possible avenue of research would be to examine ESG ratings of real estate (or construction) borrowers to test if ESG rating changes also lead to variations in bank credit. It will be interesting to test the hypothesis that upgrades in the ESG ratings of real estate enterprises determine an increase in bank credit. A second extension would involve the analysis of loan costs (i.e., loan spreads). Having information about the spreads applied on different loan types, one could verify whether there is a significant association between the bank ESG score and the cost of real estate loans.

We also recognize that our analysis used ESG ratings from one source and was limited to the United States. Therefore, follow-up analyses may use ESG ratings from other different providers and extend the research design to countries beyond the United States, taking into careful consideration the different accounting standards.

Finally, empirical research may also test if bank ESG scores are important for explaining measures of real estate loan growth, renegotiation, and securitization. Overall, collecting this evidence would help to have a broader overview of how bank sustainability is relevant for real estate financing.

Funding

This work was supported by the Open Access Publishing Fund of the Free University of Bozen.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

Data were downloaded from S&P Capital IQ.

Conflicts of Interest

The author declares that AI tools were not employed to edit this paper. The author declares no conflicts of interest. This work was supported by the Open Access Publishing Fund of the Free University of Bozen.

References

Danisman, G.O.; Tarazi, A. ESG activity and bank lending during financial crises. J. Financ. Stab. 2024, 70, 101206. [Google Scholar] [CrossRef]
Liu, S.; Jin, J.; Nainar, K. Does ESG performance reduce banks’ nonperforming loans? Financ. Res. Lett. 2023, 55, 103859. [Google Scholar] [CrossRef]
Houston, J.F.; Shan, H. Corporate ESG profiles and banking relationships. Rev. Financ. Stud. 2022, 35, 3373–3417. [Google Scholar] [CrossRef]
Bernardelli, M.; Korzeb, Z.; Niedziółka, P. Does fossil fuel financing affect banks’ ESG ratings? Energies 2022, 15, 1495. [Google Scholar] [CrossRef]
Allen, M.; Madura, J.; Wiant, K. Commercial bank exposure and sensitivity to the real estate market. J. Real Estate Res. 1995, 10, 129–140. [Google Scholar] [CrossRef]
Martins, A.M.; Serra, A.P.; Martins, F.V. Real estate market risk in bank stock returns: Evidence for 15 European countries. Int. J. Strateg. Prop. Manag. 2016, 20, 142–155. [Google Scholar] [CrossRef]
Mei, J.; Saunders, A. Bank risk and real estate: An asset pricing perspective. J. Real Estate Financ. Econ. 1995, 10, 199–224. [Google Scholar] [CrossRef]
Markose, S.; Giansante, S.; Shaghaghi, A.R. Too interconnected to fail financial network of US CDS market: Topological fragility and systemic risk. J. Econ. Behav. Organ. 2012, 83, 627–646. [Google Scholar] [CrossRef]
Iselin, M.; Liao, S.; Zhang, H. Common mutual fund ownership and systemic risk. Contemp. Account. Res. 2021, 38, 2157–2191. [Google Scholar] [CrossRef]
Pellegrini, C.B.; Cincinelli, P.; Meoli, M.; Urga, G. The role of shadow banking in systemic risk in the European financial system. J. Bank. Financ. 2022, 138, 106422. [Google Scholar] [CrossRef]
Bessler, W.; Kurmann, P.; Nohel, T. Time-varying systematic and idiosyncratic risk exposures of US bank holding companies. J. Int. Financ. Mark. Inst. Money 2015, 35, 45–68. [Google Scholar] [CrossRef]
Simpson, W.G.; Kohers, T. The link between corporate social and financial performance: Evidence from the banking industry. J. Bus. Ethics 2002, 35, 97–109. [Google Scholar] [CrossRef]
Wu, M.W.; Shen, C.H. Corporate social responsibility in the banking industry: Motives and financial performance. J. Bank. Financ. 2013, 37, 3529–3547. [Google Scholar] [CrossRef]
Di Tommaso, C.; Thornton, J. Do ESG scores effect bank risk taking and value? Evidence from European banks. Corp. Soc. Responsib. Environ. Manag. 2020, 27, 2286–2298. [Google Scholar] [CrossRef]
Qian, K.; Shi, B.; Song, Y.; Wu, H. ESG performance and loan contracting in an emerging market. Pac.-Basin Financ. J. 2023, 78, 101973. [Google Scholar] [CrossRef]
Baselga-Pascual, L.; Dimic, N.; Vähämaa, E. Corporate social responsibility and bank credit ratings. In Green Finance Instruments, FinTech, and Investment Strategies: Sustainable Portfolio Management in the Post-COVID Era; Springer: Cham, Switzerland, 2023; pp. 47–63. [Google Scholar]
Bruno, E.; Iacoviello, G.; Giannetti, C. Bank credit loss and ESG performance. Financ. Res. Lett. 2024, 59, 104719. [Google Scholar] [CrossRef]
Abdelsalam, O.; Azmi, W.; Disli, M.; Kowsalya, V. Bank lending cyclicality and ESG activities: Global evidence. Financ. Res. Lett. 2023, 58, 104541. [Google Scholar] [CrossRef]
Dursun-de Neef, Ö.; Forchieri, J.; Gehrig, T.; Schandlbauer, A. How did banks’ ESG conduct affect financial performance and lending during COVID-19? SSRN 2023. Available online: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4405578 (accessed on 24 August 2024). [CrossRef]
Goss, A.; Roberts, G.S. The impact of corporate social responsibility on the cost of bank loans. J. Bank. Financ. 2011, 35, 1794–1810. [Google Scholar] [CrossRef]
Shin, D.D. ESG shifts in lending and the cost of bank loans. SSRN 2020. Available online: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3740710 (accessed on 20 August 2024). [CrossRef]
Huang, J.; Duan, Z.; Zhu, G. Does corporate social responsibility affect the cost of bank loans? Evidence from China. Emerg. Mark. Financ. Trade 2017, 53, 1589–1602. [Google Scholar] [CrossRef]
Huang, K.J.; Bui, D.G.; Hsu, Y.T.; Lin, C.Y. The ESG washing in banks: Evidence from the syndicated loan market. J. Int. Money Financ. 2024, 142, 103043. [Google Scholar] [CrossRef]
Huang, G.; Ye, F.; Li, Y.; Chen, L.; Zhang, M. Corporate social responsibility and bank credit loans: Exploring the moderating effect of the institutional environment in China. Asia Pac. J. Manag. 2023, 40, 707–742. [Google Scholar] [CrossRef]
Cai, H.; Tan, C.; Wang, H.; Zhong, T. Does regional favoritism affect the relationship between CSR performance and the ability to obtain bank loans? Emerg. Mark. Financ. Trade 2019, 55, 1550–1565. [Google Scholar] [CrossRef]
Barros, C.P.; Ferreira, C.; Williams, J. Analysing the determinants of performance of best and worst European banks: A mixed logit approach. J. Bank. Financ. 2007, 31, 2189–2203. [Google Scholar] [CrossRef]
Chiorazzo, V.; Milani, C.; Salvini, F. Income diversification and bank performance: Evidence from Italian banks. J. Financ. Serv. Res. 2008, 33, 181–203. [Google Scholar] [CrossRef]
Trujillo-Ponce, A. What determines the profitability of banks? Evidence from Spain. Account. Financ. 2013, 53, 561–586. [Google Scholar] [CrossRef]
Köhler, M. Which banks are more risky? The impact of business models on bank stability. J. Financ. Stab. 2015, 16, 195–212. [Google Scholar] [CrossRef]
Citterio, A.; King, T. The role of Environmental, Social, and Governance (ESG) in predicting bank financial distress. Financ. Res. Lett. 2023, 51, 103411. [Google Scholar] [CrossRef]
Izcan, D.; Bektas, E. The relationship between ESG scores and firm-specific risk of Eurozone banks. Sustainability 2022, 14, 8619. [Google Scholar] [CrossRef]

Table 1. Variable definitions.

	Definition
$E S G$ (0–100)	ESG score
E (0–100)	E score
S (0–100)	S score
G (0–100)	G score
$C O N S$ (%)	Total consumer loans divided by total loans. Consumer loans include home equity loans, credit card loans, and vehicle loans
$C O M M$ (%)	Total commercial loans divided by total loans
$C O N S T R$ (%)	Total construction and development loans divided by total loans
$R E$ (%)	Total real estate loans divided by total loans
$F G$ (%)	Total foreign loans and leases divided by total loans
$O T$ (%)	Total unclassified loans divided by total loans
$H O M E E Q U I T Y$ (%)	Total consumer loans divided by total loans
$C R E D I T C A R D$ (%)	Total credit card loans divided by total loans
$V E H I C L E$ (%)	Total vehicle loans divided by total loans
$1 - 4 F A M I L Y$ (%)	Total 1–4 family first mortgage loans divided by total loans
$M U L T Y F A M I L Y$ (%)	Total 5+ family first mortgage loans divided by total loans
$C O M M R E$ (%)	Total commercial real estate loans divided by total loans
$R O A$ (%)	Return-on-assets computed dividing the total net income by total assets
$E Q U I T Y$ (%)	Total book value of equity capital divided by total assets
$L O A N S$ (%)	Total amount of net loans divided by total assets

Table 2. Summary statistics.

Mean	Median	Min	Max	sd	N
$E S G$	29.94	28	9	87	9.899	1185
E	24.02	22	0	96	13.54	1185
S	24.44	22	4	89	10.37	1185
G	35.76	33	11	84	10.76	1185
$C O N S$ (%)	9.336	5.91	0	39.89	9.53	1148
$C O M M$ (%)	26.82	22.62	0	95.19	16.07	1177
$C O N S T R$ (%)	7.444	6.1	0	76.85	6.308	941
$R E$ (%)	63.55	67.02	5.52	99.97	19.53	1179
$F G$ (%)	10.48	6.545	0.2	43.54	12.39	76
$O T$ (%)	2.226	0.76	0	16.57	3.321	221
$H O M E E Q U I T Y$ (%)	4.324	3.558	0	26.01	3.722	1081
$C R E D I T C A R D$ (%)	0.7898	0	0	23.77	2.738	1146
$V E H I C L E$ (%)	2.887	0.2061	0	35.54	5.363	1108
$1 - 4 F A M I L Y$ (%)	17.5	15.33	0.0320	97.63	13.52	1085
$M U L T Y F A M I L Y$ (%)	9.762	5.708	0	75.65	12.53	768
$C O M M R E$ (%)	32.55	33.07	0.6586	81.65	14.19	742
$R O A$ (%)	1.12	1.14	−6.86	7.1	0.623	1181
$E Q U I T Y$ (%)	10.78	10.54	2.61	20.48	2.428	1183
$L O A N S$ (%)	67.04	68.85	5.22	89.87	12.06	1183

Notes: See variable definitions in Table 1.

Table 3. Banks’ loan composition and ESG ratings.

	(1)	(2)	(3)	(4)	(5)	(6)
Variables	$CONS$	$COMM$	$CONSTR$	$RE$	$FG$	$OT$
$E S G$	0.318 ***	0.663 ***	−0.137 ***	−0.969 ***	0.294	−0.058
	(0.068)	(0.146)	(0.047)	(0.158)	(0.303)	(0.089)
$R O A$	0.559	2.809 *	1.063	−3.665 ***	−6.227	−0.590
	(0.527)	(1.578)	(0.877)	(1.264)	(4.611)	(1.411)
$E Q U I T Y$	−0.514 **	−1.194 ***	0.736 ***	1.658 ***	0.613	0.004
	(0.211)	(0.415)	(0.258)	(0.387)	(1.062)	(0.168)
$L O A N S$	−0.180 ***	−0.127	0.020	0.303 ***	−0.287	−0.002
	(0.049)	(0.085)	(0.044)	(0.075)	(0.164)	(0.046)
Constant	23.922 ***	31.989 ***	−2.089	45.733 ***	6.164	4.606
	(5.913)	(12.031)	(5.986)	(11.215)	(27.856)	(8.317)
Time fixed effects	Yes	Yes	Yes	Yes	Yes	Yes
N of observations	1147	1176	940	1178	76	221
R-squared	0.326	0.258	0.153	0.449	0.564	0.015

Notes: The table reports estimates for the regressions in (1). See variable definitions in Table 1. Standard errors in parentheses are clustered at the firm level. ***

p < 0.01

, **

p < 0.05

, *

p < 0.1

Table 4. Banks’ consumer loans and ESG ratings.

	(1)	(2)	(3)
Variables	$HOME EQUITY$	$CREDIT CARD$	$VEHICLE$
$E S G$	0.029	0.165 ***	0.068
	(0.032)	(0.047)	(0.064)
$R O A$	−0.179	0.128	0.682 **
	(0.246)	(0.141)	(0.278)
$E Q U I T Y$	−0.066	−0.105 **	−0.131
	(0.118)	(0.050)	(0.121)
$L O A N S$	−0.010	−0.060 **	−0.116 **
	(0.022)	(0.024)	(0.050)
Constant	10.987 ***	−0.323	10.614 **
	(2.815)	(2.050)	(5.068)
Time fixed effects	1080	1144	1106
N of observations	1080	1144	1106
R-squared	0.173	0.499	0.105

Notes: The table reports estimates for the regressions in (1). See variable definitions in Table 1. Standard errors in parentheses are clustered at the firm level. ***

p < 0.01

, **

p < 0.05

Table 5. Banks’ real estate loans and ESG ratings.

	(1)	(2)	(3)
Variables	$1 - 4 FAMILY$	$MULTIFAMILY$	$COMM RE$
$E S G$	0.028	−0.052	−0.754 ***
	(0.032)	(0.101)	(0.157)
$R O A$	−0.179	−3.085 **	1.679
	(0.246)	(1.230)	(1.021)
$E Q U I T Y$	−0.065	−0.031	0.525
	(0.118)	(0.353)	(0.352)
$L O A N S$	−0.009	0.295 ***	0.008
	(0.022)	(0.086)	(0.094)
Constant	10.987 ***	−0.861	38.539 ***
	(2.815)	(8.726)	(10.485)
Time fixed effects	Yes	Yes	Yes
N of observations	1080	767	741
R-squared	0.173	0.104	0.167

Notes: The table reports estimates for the regressions in (1). See variable definitions in Table 1. Standard errors in parentheses are clustered at the firm level. ***

p < 0.01

, **

p < 0.05

Table 6. Banks’ loan composition (real estate and construction loans) and separate E/S/G ratings.

	(1)	(2)	(3)	(4)	(5)	(6)
Variables	$RE$	$RE$	$RE$	$CONSTR$	$CONSTR$	$CONSTR$
E	−0.5639 ***			−0.0914 ***
	(0.090)			(0.034)
S		−0.7074 ***			−0.1142 ***
		(0.130)			(0.042)
G			−1.1528 ***			−0.1244 **
			(0.142)			(0.059)
$R O A$	−3.7922 ***	−3.6282 ***	−3.5343 ***	1.0268	1.0541	1.0933
	(1.259)	(1.261)	(1.183)	(0.855)	(0.873)	(0.883)
$E Q U I T Y$	1.6548 ***	1.8082 ***	1.7345 ***	0.7247 ***	0.7502 ***	0.7534 ***
	(0.406)	(0.401)	(0.373)	(0.239)	(0.259)	(0.262)
$L O A N S$	0.3311 ***	0.3340 ***	0.3534 ***	0.0210	0.0198	0.0251
	(0.076)	(0.076)	(0.070)	(0.043)	(0.044)	(0.044)
Constant	21.1263 **	24.7697 **	65.3333 ***	−4.8623	−4.1814	−1.3814
	(9.013)	(9.783)	(11.702)	(4.335)	(5.576)	(7.222)
Time fixed effects	Yes	Yes	Yes	Yes	Yes	Yes
N of observations	1178	1178	1178	940	940	940
R-squared	0.410	0.409	0.468	0.152	0.150	0.150

Notes: The table reports estimates for the regressions in (1), including separate E, S, and G ratings. See variable definitions in Table 1. Standard errors in parentheses are clustered at the firm level. ***

p < 0.01

, **

p < 0.05

Table 7. Banks’ loan composition (commercial and credit cards loans) and separate E/S/G ratings.

	(1)	(2)	(3)	(4)	(5)	(6)
Variables	$COMM$	$COMM$	$COMM$	$CREDIT CARDS$	$CREDIT CARDS$	$CREDIT CARDS$
E	0.380 ***			0.124 ***
	(0.095)			(0.036)
S		0.470 ***			0.149 ***
		(0.118)			(0.041)
G			0.799 ***			0.134 ***
			(0.134)			(0.045)
$R O A$	2.9141 *	2.7869 *	2.7057 *	0.1748	0.1308	0.1008
	(1.574)	(1.564)	(1.515)	(0.127)	(0.138)	(0.175)
$E Q U I T Y$	−1.1991 ***	−1.3078 ***	−1.2398 ***	−0.0765 *	−0.1107 **	−0.1462 **
	(0.419)	(0.421)	(0.410)	(0.041)	(0.050)	(0.058)
$L O A N S$	−0.1491 *	−0.1520 *	−0.1587 **	−0.0528 **	−0.0552 **	−0.0781 **
	(0.090)	(0.086)	(0.080)	(0.021)	(0.022)	(0.031)
Constant	49.319 ***	47.226 ***	17.794	1.897	1.570	0.917
	(10.331)	(10.785)	(12.493)	(1.532)	(1.796)	(2.619)
Time fixed effects	Yes	Yes	Yes	Yes	Yes	Yes
N of observations	1176	1176	1176	1144	1144	1144
R-squared	0.229	0.226	0.275	0.534	0.514	0.406

p < 0.01

, **

p < 0.05

, *

p < 0.1

Table 8. Banks’ loan composition and ESG ratings interacted with an indicator for low/high greenhouse gas emissions states.

	(1)	(2)	(3)	(4)
Variables	$CONSTR$	$RE$	$CONS$	$COMM$
$L O W G A S E M I S S I O N S \times E S G$	−0.067	−0.849 ***	0.323 ***	0.528 ***
	(0.068)	(0.172)	(0.083)	(0.141)
$H I G H G A S E M I S S I O N S \times E S G$	−0.125 **	−0.918 ***	0.350 ***	0.603 ***
	(0.053)	(0.163)	(0.081)	(0.157)
$R O A$	1.124	−3.816 **	0.329	3.187 **
	(0.868)	(1.493)	(0.683)	(1.432)
$E Q U I T Y \times E S G$	0.912 ***	1.876 ***	−0.488 *	−1.444 ***
	(0.316)	(0.486)	(0.281)	(0.512)
$L O A N S$	−0.006	0.265 ***	−0.160 ***	−0.083
	(0.049)	(0.096)	(0.060)	(0.104)
Constant	−3.792	42.799 ***	22.024 ***	35.199 **
	(6.876)	(12.031)	(7.044)	(13.661)
Year fixed effects	564	707	688	705
N of observations	564	707	688	705
R-squared	0.223	0.451	0.364	0.257

Notes: The table reports estimates for the regressions in (1), including the interaction between

E S G

and an indicator for states with low/high greenhouse gas emissions. See variable definitions in Table 1. In every year, the variable

L O W G A S E M I S S I O N S

(

H I G H G A S E M I S S I O N S

) indicates that the bank is located in a state where the total greenhouse gas emissions are below (above) the median greenhouse gas emissions measured across all of the United States in the same year. Standard errors in parentheses are clustered at the firm level. *** p < 0.01, ** p < 0.05, * p < 0.1.

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

© 2024 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).

Share and Cite

MDPI and ACS Style

Bressan, S. Environmental, Social, and Governance Scores and Loan Composition Inside United States Banks. Sustainability 2024, 16, 8075. https://doi.org/10.3390/su16188075

AMA Style

Bressan S. Environmental, Social, and Governance Scores and Loan Composition Inside United States Banks. Sustainability. 2024; 16(18):8075. https://doi.org/10.3390/su16188075

Chicago/Turabian Style

Bressan, Silvia. 2024. "Environmental, Social, and Governance Scores and Loan Composition Inside United States Banks" Sustainability 16, no. 18: 8075. https://doi.org/10.3390/su16188075

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Menu

Environmental, Social, and Governance Scores and Loan Composition Inside United States Banks

Abstract

1. Introduction

2. Review of the Literature

3. Research Design

3.1. Data and Variables

3.2. Model

4. Results

4.1. Regression Outcomes

4.2. Discussion

5. Conclusions

Funding

Institutional Review Board Statement

Informed Consent Statement

Data Availability Statement

Conflicts of Interest

References

Share and Cite

Article Metrics

Article Access Statistics

Further Information

Guidelines

MDPI Initiatives

Follow MDPI