Behavior and Sustainable Finance: A Bibliometric Approach

Abstract

This paper is intended to highlight the importance of developing knowledge on sustainable finance using bibliometric analysis. The study is based on a review of sources from two of the most prominent scientific databases in the world, Scopus and Web of Science (WOS). A total of 372 and 355 documents were obtained from Scopus and WOS, respectively. The data indicate an exponential increase in the number of publications over the years, suggesting a growing interest in the subject and a need for further research. The research groups appear to have little connection, and the studies are widely scattered both geographically and across different research areas. Sustainable finance is an increasingly interesting field of study, with numerous branches that require further research. One such branch is the analysis of green bonds and sustainable development.

1. Introduction

In recent times, the world has witnessed various global changes, such as global warming and the COVID-19 pandemic, among others. These changes have led to increased concern among companies and investors regarding social and environmental issues, resulting in a greater focus on sustainability (Singhania et al., 2023). Sustainability refers to the ability to meet present needs without compromising the ability of future generations to meet their own needs (Waseem & Kota, 2017). Sustainability involves managing natural resources responsibly, protecting the environment, promoting equitable economic development, and enhancing social welfare. It seeks to balance environmental, social, and economic dimensions to ensure long-term development viability.

The SDGs, or Sustainable Development Goals, are a group of 17 goals adopted by UN member states in 2015 as part of the 2030 Agenda for Sustainable Development (Halisçelik & Soytas, 2019). These goals cover areas such as poverty eradication, gender equality, climate action, quality education, health, and welfare, and aim to promote sustainable development globally. The SDGs are regarded as a global call to eradicate poverty, protect the planet, and ensure that all people can enjoy peace and prosperity.

Regarding sustainable finance, it concerns the investment and funding of projects and companies that consider environmental, social, and governance (ESG) issues in their operations. These investments aim to generate a positive impact on long-term sustainability, both financially and in terms of social and environmental responsibility (Park, 2009). Sustainable finance can encompass various terms such as ethical investment (Lewis & Cullis, 1990), socially responsible investment (Ertz & Sarigöllü, 2022; Hamilton et al., 1993), impact investment, green investment, or eco-friendly investment (Chiţimiea et al., 2021), depending on the perspective from which the concept is being examined.

Sustainable finance can also refer to the management of financial assets that take into account risks and opportunities related to climate change, environmental and social sustainability, and corporate governance (Patterson & McEachern, 2018). These investments aim not only to generate financial returns but also to contribute to the mitigation of environmental and social risks and promote responsible business practices.

Sustainable finance offers advantages at both the corporate and investment levels (Hamilton et al., 1993); however, in this work the focus is mainly on the investment side, particularly investor behavior and preferences in the context of sustainability. At the corporate level, sustainable investments can improve a company’s financial performance by reducing material costs, increasing product differentiation due to a better reputation among customers, and enhancing the ability to raise funds in the investment market. From an investment perspective, sustainable finance can have a positive and stable impact on corporate financial performance over time. In addition, sustainable investment can provide both financial and non-financial benefits to investors, which may include a combination of profitability concerns, social responsibility, and return seeking (Mishra et al., 2023).

Investment decisions are influenced by investors’ heterogeneous preferences regarding sustainability (S. McGregor, 2005; S. L. T. McGregor, 2002) and risk (Lagerkvist et al., 2020). Investors exhibit high variability in their risk tolerance, indicating sensitivity to this factor (de Carlos Fraile et al., 2023; Mirón-Sanguino & Díaz-Caro, 2022; Muñoz-Muñoz et al., 2025). Additionally, it is found that investors are willing to compromise on some profitability to contribute to the achievement of sustainable goals (Bauer et al., 2018). Sustainability and risk are crucial factors that impact investment decisions made by investors (Gutsche & Ziegler, 2019).

The use of choice experiments (CEs) has allowed for the assessment of investor preferences in funds that contribute to the achievement of the ODS (de Carlos Fraile et al., 2023). Experiments have shown that investors have varying patterns of portfolio holdings, indicating the presence of heterogeneous preferences that may influence the creation and design of incentives and policies to meet the stated objectives. In summary, an investor’s behavior can be influenced by his or her preferences for investment funds that contribute to ODS, which may have implications for investment decisions and the design of sustainable investment policies and strategies.

The literature shows that the field of sustainable investment has seen a number of reviews of different aspects and methodologies: exploratory reviews, narrative reviews, structured reviews, bibliometric reviews, etc. (Singhania et al., 2023). From the set of reviews conducted, it is evident that various aspects of sustainable finance have been studied and analyzed, such as: (1) the definition, history, and mapping of the global context of sustainable finances (Abhayawansa & Mooneeapen, 2022; Beisenbina et al., 2023; Dervi et al., 2022; Paul & Benito, 2018; Poyser & Daugaard, 2023); (2) analysis of Islamic perspectives for sustainable financing (Abdul-Muhmin, 2008; Delle Foglie & Keshminder, 2024; Rahman et al., 2020); (3) performance of sustainable investment-based funds and Islamic equities (Friede et al., 2015; Kim, 2019; Masih et al., 2018); (4) analysis of investment impact (Agrawal & Hockerts, 2021; Islam, 2022; Mittal et al., 2021); (5) sustainable finance integrating ESG (Aldowaish et al., 2022; Fan et al., 2022; Hidalgo-Oñate et al., 2023; Rosado-Serrano et al., 2018); and (6) SDG funding (Erasmus & Mathunjwa, 2011; Folqué et al., 2023; MacAskill et al., 2021; Prakash & Sethi, 2022; Ruggeri et al., 2019). Finally, there is another area derived from literature reviews based on consumer behavior with regard to investment products (Berg, 2008; Park, 2009) and investor behavior within sustainable finance, highlighting the work of Mehta et al. (2020), which carries out a review of the conceptual framework based on a conceptual review and a literature review of Dhayal et al. (2023), which shows a research gap in this area in terms of bibliometric analysis.

While previous literature reviews have explored various aspects of sustainable finance, such as ESG integration, green bonds, and the performance of socially responsible investments, few have focused specifically on the intersection of sustainable finance and investor behavior. Moreover, to the best of our knowledge, no bibliometric review has yet combined both VOSviewer 1.6.19 and SciMAT 1.1.06 tools to map the thematic evolution of this field over time. This study addresses this gap by providing a dual bibliometric analysis centered on how investor behavior is approached in sustainable finance literature, offering a structured overview of key research clusters, their development, and future research avenues.

Bibliometric reviews are widely recognized as an effective tool for identifying knowledge gaps and future research directions. They help to identify patterns and clusters and to map the state of the field (Mukherjee et al., 2022). Therefore, with the present literature review, we aimed to highlight the accumulated knowledge in this field and raise future research questions. To achieve this, a search was conducted through the Scopus and Web of Science databases, and VOSviewer 1.6.19 (van Eck & Waltman, 2010) and SciMAT 1.1.06 (Cobo et al., 2012) software was used to analyze the extracted data.

2. Materials and Methods

2.1. Choice of Methodology

The review methodology employed in the present work was bibliometric, enabling analysis of a substantial body of literature on the subject under investigation. Bibliometric analysis is recognized as a well-founded research methodology. Using consistent methods and procedures, it can be employed to discover, choose, and evaluate a field of study. It is designed to minimize bias and provide a comprehensive view of available and relevant research (Xiao et al., 2023; Palmatier et al., 2018). This is achieved by identifying all studies that meet predefined criteria, followed by a careful selection process to extract the information using quantitative methods (Broadus, 1987). However, the goal of a bibliometric study is not only to summarize the existing literature but also to identify gaps in current knowledge and provide directions for future research (Paul & Criado, 2020).

2.2. Source Acquisition

Peer-reviewed articles and reviews from reputable journals were used to ensure greater reliability in the search. Because of their reliability, the Scopus and Web of Science (WOS) databases were used to conduct identical searches. Although we did not follow a formal review protocol such as PRISMA or SPAR-4-SLR, we adopted a clearly defined and reproducible procedure, as detailed in Figure 1, to ensure methodological transparency and replicability.

The search on Scopus was conducted in multiple stages to refine the results until the final outcome was achieved. The initial search used the following keywords in the abstract, keywords, or title: “sustainable invest*” OR “sustainable financ*” OR “socially responsible invest*” OR “socially responsible financ*” OR “green financ*” OR “green invest*” OR “ethical finance” OR “ethical investment”. The “OR” operator was used to include synonymous or related terms. This approach follows standard practice in bibliometric research (Voguel et al., 2021). By using asterisks, more terms with the same root can be included in the search. Applying this first filter yielded a total of 7025 results. Subsequently, a screening process was conducted using the aforementioned results, selecting only peer-reviewed articles and reviews, resulting in a total of 5515 publications. The database search was conducted during the development of this study in 2024, and records from 2024 onwards were excluded due to data incompleteness at the time of analysis.

Next, to identify papers regarding the subject of the present study, the terms “behavi*” and “financ*” were used with the “AND” operator in the keyword search to filter papers that contained both key concepts, yielding 403 documents.

Finally, a filtering by subject was performed, selecting only “business, management and accounting,” “economics, econometrics and finance,” “social sciences,” “environmental science,” “energy,” “mathematics,” “psychology,” “decision sciences,” and “multidisciplinary.” After completing this screening process, the search yielded 372 results, which constituted the final sample used in the present work. A similar procedure was followed with the Web of Science database, yielding results that were highly consistent with those obtained from Scopus in terms of document overlap, key authors, and thematic structure, although the number of documents retrieved was slightly lower (335). Given the high convergence between both datasets and the fact that each was processed using a tool best suited to its format (VOSviewer for Scopus and SciMAT for WOS), the analysis was conducted separately without merging the databases. The similarity in results suggests that using either database independently, or a combination of both with deduplication, would have led to almost identical findings. In addition, we chose to include only peer-reviewed articles and review papers to ensure a minimum standard of academic quality in the sources analyzed. While we acknowledge that gray literature can sometimes offer valuable insights, its inclusion often presents challenges in terms of reliability, replicability, and bibliometric traceability. Furthermore, the number of conference papers and book chapters was comparatively low, so their inclusion would not have significantly altered the overall results.

3. Results and Discussion

3.1. Bibliometric Analysis Using Scopus and VOSviewer

In this section, the results that were obtained using the Scopus database with the VOSviewer program are presented.

The methodology used to analyze the 372 articles involved analytical, statistical, and graphical tools from the software packages Microsoft Excel and VOSviewer (van Eck & Waltman, 2010). These tools made it possible to analyze publications segmented by various variables, including year of publication, the country where the study was conducted, areas of research, specialized journals, keywords, authors, and most cited articles. Citations and co-citations were analyzed using VOSviewer to investigate potential relationships between authors or documents. Furthermore, an examination of the connections between authors, countries, and institutions was carried out, revealing a lack of correlation between them. Lastly, the relationships between keywords were analyzed using VOSviewer software.

3.1.1. Bibliometric Analysis of Growth Rate

Price’s law (De Solla Price, 1965) states that growth in a field follows an exponential pattern until it reaches a linear zone or limit. In bibliometrics, a field is considered of interest when the number of articles published over a series of years fits an exponential curve without signs of linearity or saturation.

Therefore, to visualize the interest in the field under study, a plot of the published articles per year from 1990 to 2023 is shown in Figure 2. The X-axis represents the years in which the studies were conducted, while the Y-axis represents the number of articles published in that year. The correlation coefficient obtained was 0.85, indicating a good fit to an exponential curve. Based on these findings, the present study may be considered part of a growing field of research that is still in its early stages, is currently relevant, and has potential for further exploration through new research. This field is still in its early stages of development, despite being studied for around 30 years. As shown below, the lack of connection between research groups and institutions, as well as the low productivity of authors, are contributing factors to this.

3.1.2. Geographical Distribution of the Research Output

To visualize global research productivity and identify geographical patterns in scientific output, Figure 3 shows the geographical distribution of the complete set of articles retrieved. It can be seen that China is the most productive country, followed by the USA and the United Kingdom. However, China and the USA are the countries of origin of the institutions responsible for only 25% of world production. The remaining 75% is well spread around the world, with 70 countries found in the search output. It is therefore possible to conclude that research in this field is a phenomenon of global scope.

Next, VOSviewer software was used to analyze research productivity, international collaboration networks, and citation impact of the most productive and most cited countries. Figure 4 shows the network visualization of country co-authorships. Each node represents a country that has contributed to scientific publications on the selected topic. The size of the nodes is proportional to the volume of publications affiliated with each country, highlighting the most prolific contributors in the field—notably China, the United States, and the United Kingdom. The connecting lines between nodes indicate instances of international collaboration, with thicker lines representing stronger co-authorship links, i.e., a higher number of joint publications. These connections suggest that collaborations tend to be geographically clustered, with countries more likely to collaborate with their regional neighbors. The color scale, ranging from blue to yellow, reflects the number of citations, with yellow tones associated with more cited research. Figure 4 shows that although China has greater scientific production, research from the United States, the United Kingdom, and the Netherlands has a greater impact, with more citations per article on average.

3.1.3. Subject Areas

Regarding the subject area that encompasses the research described, as seen in Figure 5, it focuses mainly on fields related to economics and business studies, although the analysis of environmental impact also has significant weight, as well as the study of companies related to the energy business. However, it can be seen that “business, management and accounting,” and “economics, econometrics, and finance” account for 40% of the research in this area. This analysis will be detailed in Section 3.2.

The predominance of research areas such as economics, social sciences, environmental science, and energy in the study of sustainable finance is, in our view, to be expected. Given that finance is inherently linked to economic systems, it is natural for economics and the broader social sciences to provide the conceptual and methodological frameworks for its analysis. Furthermore, the inclusion of the term “sustainable” introduces a multidimensional perspective that extends beyond traditional finance, bringing in key themes related to environmental impact, climate change, and energy transition. This interdisciplinary convergence suggests that sustainability in finance is most frequently associated with environmental and energy-related concerns, reflecting global research priorities in this emerging field.

3.1.4. Most Prolific Journals

According to Bradford’s law (Nicolaisen & Hjørland, 2007), the scientific production on a given topic tends to concentrate in a small core of highly productive journals, while the rest of the literature is scattered across a larger number of journals with decreasing relevance. This pattern allows for the identification of the most influential journals within a field. In the context of the present study, this law helps explain why the majority of publications are concentrated in journals related to economics and environmental science, which are the most relevant research areas, as discussed above.

The core zone, the most prolific journals based on the number of publications, are shown in

Table 1. Most significant journals.

Source	Documents	Citations
Sustainability (Switzerland)	43	494
Journal of Cleaner Production	15	413
Journal of Sustainable Finance and Investment	14	289
Journal of Business Ethics	11	500
Resources Policy	8	56
Energy Economics	7	116
Managerial Finance	6	79
Journal of Banking and Finance	6	1005
Critical Studies on Corporate Responsibility, Governance and Sustainability	6	16
Environmental Science and Pollution Research	5	73

Source: Authors’ own elaboration.

. These journals have published at least six articles related to the topic of the present study. The number of citations is also provided.

It is noteworthy that the journal Sustainability, published by MDPI, hosts a high number of publications, being the second-most cited journal in the field. As far as the most cited journal is concerned, it owes its position to an article by Renneboog et al. (2008), discussed below, and which had 874 citations at the time of writing.

3.1.5. Co-Occurrence Analysis

Figure 6, Figure 7 and Figure 8 depict the co-occurrence of keywords and clusters of keywords related to socially responsible behavior of investors and consumers, respectively, in order to identify key research themes and their interconnections. Regarding Figure 6, the most relevant keywords, “socially responsible investing” and “CSR” (corporate social responsibility) stand out, which form the core of the search after analyzing similarities and duplications. Surrounding this core, other clusters emerge that incorporate environmental, behavioral, consumer, and technological dimensions, highlighting the multidisciplinary character of sustainable finance research.

Another relevant area in the co-occurrence network is represented by the green cluster, which includes terms such as green bonds, clean energy, efficiency, and pandemic. This cluster reflects the increasing scholarly attention on the role of green financial instruments and energy transition in the broader sustainability agenda. The prominence of green bonds within this group highlights the emergence of dedicated financial mechanisms aimed at channeling investment toward environmentally beneficial projects. The inclusion of terms like clean energy and efficiency further illustrates the close link between sustainable finance and energy-related outcomes, while the presence of “pandemic” suggests a contextual layer in which environmental and health crises are recognized as catalysts for rethinking financial strategies. This cluster also maintains connections with others—particularly those focused on financial behavior and policy—demonstrating that green finance is not isolated, but instead part of a dynamic and interconnected research field that bridges environmental goals with financial innovation and risk management.

A closer look at the yellow cluster, expanded in Figure 7, reveals a thematic focus on sustainable consumer behavior, along with closely related terms such as digitalization, common-good balance, and regional nodes like China. This cluster suggests a growing body of literature that examines how sustainability principles are incorporated into consumption patterns, individual preferences, and values. Notably, this cluster is directly connected to the adjacent blue cluster, centered on investor behavior, which includes keywords such as financial knowledge, investor biases, and financial literacy. This connection reflects the convergence between the behavior of consumers and investors, emphasizing that both roles are studied through similar theoretical and methodological lenses when it comes to decision-making in the context of sustainability. Moreover, the presence of digitalization points to the increasing importance of technology in shaping and mediating sustainable behavior in both domains.

This area shown in Figure 7 is more diffuse and peripheral, associated with themes from behavioral economics—such as investor and consumer behavior—and ethical dimensions of finance. Unlike the denser and more cohesive thematic clusters, the keywords in this region are only weakly linked, suggesting an emerging or less consolidated research where ethical considerations and behavioral aspects intersect with responsible investment and sustainable consumption.

In addition to this, the central cluster is closely connected to another dense area in the upper-left quadrant of the map. Terms such as behavioral finance, financial behavior, and experimental finance are found, as shown in Figure 8. These keywords are located near the conceptual core and show multiple links to other clusters, indicating their integrative role across the field. Their proximity to central themes like CSR and responsible investing suggests that behavioral and experimental approaches provide valuable insights into how individuals and institutions make decisions in sustainable finance contexts.

3.1.6. Authors’ Analysis

Regarding the number of authors with publications in this field, there are over 1000. However, only 34 of them have published more than one research paper. These authors are grouped into clusters, as shown in Figure 9, to identify the most influential researchers in the field and relationships among their work. The lines in the figure indicate co-authorship of works, while the size of the bubbles indicates the number of articles published. It is evident that there is little connection between clusters, with the elements of each cluster connected only by co-authorship of the same article. Inter-group connections are limited, indicating that cross-cluster interactions are weak compared to the dense intra-group collaborations. However, certain authors stand out for their external connectivity across cluster boundaries. For instance, Ionescu L. appears to serve as a bridge between multiple groups, given their central position and links to authors in distinct clusters. Similarly, Richardson B.J. and Gutsche G. show evidence of connections beyond their immediate cluster, although to a lesser extent. These bridging authors are crucial for facilitating knowledge exchange and promoting interdisciplinary collaboration. Strengthening inter-group collaboration, particularly by encouraging more connections between currently segregated clusters, could significantly enhance the potential of the research community represented in this network.

It should be noted that the preparation of the author study requires considerable manual intervention due to the large number of authors from China, which has the highest number of publications: many of these authors have identical names in the Scopus reference such that manual supervision through the Scopus author number is necessary to differentiate between them.

Two authors, L. Ionescu and F. Paetzold, have published the most articles in the field, with four each. Paetzold’s work focuses on the role of private investors in socially responsible finance, while Ionescu’s research centers on modeling and mathematical analysis of sustainable investment behavior. Considering the limited number of articles per author, both the year of publication and the number of citations per article and author were taken into account to determine the most influential works from the search results. Notably, an article from the Ter Host group (Renneboog et al., 2008) has nearly triple the number of citations compared to the second-most cited article (Riedl & Smeets, 2017).

3.1.7. Most Cited Articles

Next, in Figure 10, those articles with more than 100 citations are represented, showing both the number of citations (through the volume of the bubble) and the year of publication through the color scale. It is generally expected that older articles accumulate more citations over time, simply due to their longer presence in the literature. Therefore, the appearance of recent publications—represented in lighter colors in Figure 10—among the most cited works is a strong indicator of immediate and significant impact. In this regard, the articles by Kumar et al. (2022) and Zhang et al. (2021) stand out remarkably. Despite their recent publication, they have already garnered a substantial number of citations, signaling that these articles have made a notable breakthrough and are rapidly shaping the discourse in the field. On the other hand, an article by Renneboog et al. (2008) clearly stands out due to its consistently high citation count over time. This sustained academic attention suggests that the paper has become a foundational reference in the field.

The diagram also reveals a fragmented collaboration landscape within the field, with no interaction between the most cited authors. This graph also highlights a recent literature review (Kumar et al., 2022), in addition to the two articles previously mentioned (Renneboog et al., 2008; Riedl & Smeets, 2017). A list of the most cited papers is given in

Table 2. Most cited papers.

Authors	Title	Citations	Year	Source
(Renneboog et al., 2008)	Socially responsible investments: Institutional aspects, performance, and investor behavior	874	2008	Journal of Banking and Finance
(Riedl & Smeets, 2017)	Why do investors hold socially responsible mutual funds?	335	2017	Journal of Finance
(Doh et al., 2010)	Does the market respond to an endorsement of social responsibility? The role of institutions, information, and legitimacy	292	2010	Journal of Management
(Zhang et al., 2021)	Fostering green development with green finance: An empirical study on the environmental effect of green credit policy in China	216	2021	Journal of Environmental Management
(Nilsson, 2008)	Investment with a conscience: Examining the impact of pro-social attitudes and perceived financial performance on socially responsible investment behaviour	190	2008	Journal of Business Ethics
(Pham, 2016)	Is it risky to go green? A volatility analysis of the green bond market	147	2016	Journal of Sustainable Finance and Investment
(Rosen et al., 1991)	Social issues and socially responsible investment behaviour: A preliminary empirical investigation	144	1991	Journal of Consumer Affairs
(Trompeter et al., 2013)	A synthesis of fraud-related research	134	2013	Auditing
(Kumar et al., 2022)	Past, present, and future of sustainable finance: insights from big data analytics through machine learning of scholarly research	100	2022	Annals of Operations Research
(Landi & Sciarelli, 2019)	Towards a more ethical market: the impact of ESG rating on corporate financial performance	96	2019	Social Responsibility Journal

Source: Authors’ own elaboration.

, which provides detailed information on these articles, offering insights into their thematic focus and individual contributions. Unsurprisingly, ESG, green bonds, and social investing emerge as the key subjects, while the most relevant journals are primarily related to the economic sciences.

Finally, a cross-analysis was performed to identify the most influential articles and their connections to similar articles in the subject of this search. Figure 11 shows a clear clustering of the most relevant articles, with particular emphasis on the article by Renneboog et al. (2008) due to its high number of citations. However, the cluster containing the literature review by Kumar et al. (2022) is more densely populated.

In summary, the field of study is currently underdeveloped. Research activities are spread among numerous groups from a wide variety of countries, and no established network of cooperation or collaboration has emerged. Although the number of documents is limited, some of them are highly cited, highlighting their validity and relevance.

3.2. Bibliometric Analysis Using WOS and SciMAT

To assess the emergence of the theme of sustainable finance, which is underdeveloped, the analysis was completed using SciMAT 1.1.06 software. This should confirm the importance of analyzing the issue raised, and WOS data are used only to corroborate what was studied with Scopus. Bibliometrics is generally used for quantitative research evaluation of scholarly output, as discussed in the previous points (Callon et al., 1991). In bibliometrics, there are two main procedures: performance analysis and science mapping (Van Raan, 2005). Performance analysis aims to evaluate groups of scientific actors (countries, universities, departments, researchers) and the impact of their activity on the bibliographic database. According to Börner et al. (2003), science mapping aims to show the structural and dynamic aspects of scientific research, and they point out that a science map is used to represent the cognitive structure of a research field. According to De Moya-Anegón et al. (2013), several types of techniques have been developed to construct a science map, the most commonly used being co-citation of documents and co-word analysis.

The aim of this analysis in SciMAT is to provide a general method for analyzing the thematic evolution of sustainable finance. This method (Mirón Sanguino et al., 2024; Romero-Pérez & Pulido-Rojano, 2018), combines performance analysis and science mapping to identify and visualize conceptual subdomains (specific topics or general thematic areas). It also enables us to measure and visualize the thematic evolution of the research field. Strategic diagrams are used to categorize the detected subjects of research and show their evolution. A visualization approach is proposed to graphically display the thematic evolution of the field. Additionally, a performance analysis is conducted using basic bibliometric indicators such as the number of published papers, number of citations received, and the h-index to measure the impact and productivity of these citations. Before delving into the interpretation of the graphs obtained in SciMAT, it is important to clarify some concepts related to the interpretation of results. Two measures can represent the detected networks: centrality and density.

Centrality measures the degree of interaction of a network with other networks, and it can be defined as:

$$c=10\sum e_{kh}$$

(1)

with:

$$e_{kh}={c_{kh}}^2/c_k{c}_h$$

(2)

$c_k$ being the count of publications in which a keyword appears, $c_h$ the occurrences of a different keyword, and $c_{kh}$ the count of publications in which the two keywords simultaneously are present. Therefore, Callon’s centrality measures the degree of interaction among networks and can be used to gauge the importance of a particular keyword in the development of the research area under study (Callon et al., 1991).

Similarly, the internal strength of the network can be measured by Callon’s density (Callon et al., 1991), defined as Equation (3), with keywords i and j belonging to the theme and w the keyword count in the theme.

$$d=100\left(\sum e_{ij}/w\right)$$

(3)

3.2.1. Visualization of Themes and Thematic Networks

By using co-word analysis to perform the mapping, clusters of keywords and their connections to each other are obtained—these are the cluster themes. Each research topic obtained in the process is characterized by the two aforementioned parameters of density and centrality, both the median and the mean values of density and centrality. According to Cahlik (2000) and Callon et al. (1991), there are four kinds of themes. Themes located in the upper-right quadrant are referred to as motor themes due to their strong centrality and high density. These themes are crucial for the structure of a research field. They are strongly linked internally and externally. On the other hand, themes located in the upper-left quadrant are highly specialized and have a peripheral character. Although they have well-developed internal ties, they are considered to be of marginal importance to the field due to their unimportant external ties. Themes in the lower-left quadrant may represent emerging or disappearing themes, as they are weakly developed and show weak internal links. These themes have low density and centrality and have not demonstrated significant research interest, but they cannot be discarded. Themes in the lower-right quadrant represent transversal, general, and basic topics that are important to the research field, but are not well developed.

According to Cobo et al. (2012), a thematic network is created in a theme by the keywords and their links, forming a network graph. The thematic network is labeled with the name of the most significant keyword in the associated theme, usually identified by the most central keyword of the topic. The thematic analysis is described next.

3.2.2. Thematic Networks: Evolution of the Themes

As previously stated, a search was conducted using the WOS database, as outlined in Figure 1, resulting in 355 articles. Figure 12 and Figure 13 illustrate the research areas and corresponding number of documents retrieved. The area of environmental sciences, sustainability, and business and economics stand out. Putting these last two together, they would be the most studied area.

To facilitate diagram analysis, the time period is divided into two sub-periods. The first sub-period includes all years up to 2019, and the second sub-period includes years from 2020 to 2024. The first sub-period includes 116 documents, while the second sub-period includes 219. Thus, the sample comprises 335 articles, containing a total of 1678 keywords.

Table 3. Documents and keywords by time period.

Period	Documents	Keywords
All	335	1678
2020–2024	219	1249
Until 2019	116	643

Source: Authors’ own elaboration.

shows the division into periods with the distribution of documents and keywords.

Using the data from the first sub-period, up to 2019, the first strategy diagram is obtained (Figure 14), whose nodes are estimated with the values of

Table 4. Themes in the strategic diagram until 2019.

Themes	Centrality	Centrality Range	Density	Density Range
Return	1.09	0.50	24.69	0.83
Financial performance	105.41	1.00	17.67	0.67
Attitudes	7.59	0.67	30.45	1.00
Mutual funds	47.81	0.83	9.86	0.17
Market	11.29	0.17	17.20	0.50
Management	11.44	0.33	14.26	0.33

Source: Authors’ own elaboration.

. Therefore, the strategy diagram represents only the first sub-period covering all years up to 2019.

Figure 14 illustrates three driving themes, one highly developed theme, two emerging or disappearing themes, and one basic theme, also known as transversal themes. The size of the circles corresponds to the h-index of each theme.

According to Cobo et al. (2012), the upper-right quadrant of the diagram represents the driving themes that are considered well established and important for the construction of the scientific field. During this initial sub-period, it is evident that financial performance, attitudes, and return have high density and strong centrality, resulting in a significant number of citations. Financial performance is the most centered theme with the highest value, indicating that it is the most closely related theme to the others. Attitudes is the densest theme, meaning that it has the strongest relationship with the aforementioned themes within its node.

Figure 14 also shows management as an emerging theme, although it could also be considered a disappearing theme given its low density and centrality scores. In order to consider whether it is emerging or disappearing, it will be necessary to analyze the following period.

The period 2020–2024 (Figure 15 and

Table 5. Themes in the strategic diagram, period 2020–2024.

Table	Centrality	Centrality Range	Density	Density Range
Attitudes	51.16	0.57	15.26	0.86
Management	41.35	0.43	9.84	0.43
Financial performance	66.21	0.86	15.73	1.00
Performance	69.60	1.00	10.37	0.57
Green finance	35.24	0.29	5.43	0.29
Investment	54.86	0.71	5.17	0.14

Source: Authors’ own elaboration.

) is characterized by three motor themes, two basic themes, three emerging themes, and two peripheral themes.

In this second period, it can be seen that financial performance, attitudes, and performance have high density and a high centrality, so they have a high number of citations, with performance being the most centered and highest-value topic, representing the topic that is most related to all the others. Financial performance is in a similar range of centrality and also of higher density, so it is the theme that is most related to all the themes in its node.

In terms of the basic themes, i.e., those in the lower-right quadrant of the graph, investment is more central than dense. Figure 15 also shows management to be a clearly emerging theme, as discussed when analyzing the previous period. In addition to this, the theme of green finance, which did not appear in the previous period, is also emerging.

With respect to the analysis of the complete period of time (

Table 6. Themes in the strategic diagram, all time periods.

Themes	Centrality	Centrality Range	Density	Density Range
Attitudes	70.62	0.82	14.3	1.00
Return	43.65	0.64	11.9	0.91
Management	54.8	0.73	9.28	0.82
Performance	88.97	1.00	5.94	0.36
Financial performance	72.35	0.91	8.21	0.64
Green finance	30.55	0.45	7.03	0.55
Investors	33.97	0.55	6.55	0.45
Energy	12.62	0.36	8.46	0.73
Green bonds	5.11	0.09	5.83	0.27
Corporate governance	9.66	0.18	4.60	0.18
Sustainable development	10.59	0.27	4.09	0.09

Source: Authors’ own elaboration.

and Figure 16), three emerging themes can be observed: green bonds, corporate governance, and sustainable development. Performance stands out as a basic theme and is very connected to all the others with high centrality, and the same happens with the investors theme, but within its node.

Green finance appears as a theme that starts to develop with medium values of density and centrality, both equally important; therefore, it is a theme that is relatively well connected with all the others and within its own node. The same happens with the energy theme. Although it is not so well connected with the other themes, it is well connected with those of its own node, presenting a high density value.

It is worth highlighting the large number of driving themes that develop the generic theme of sustainable finance. The theme financial performance stands out, with a high centrality score and also as a fairly dense theme. At this level, the return and management themes are more dense than central, and therefore more related to their nodes than to the themes as a whole. Finally, it should be noted that the attitudes theme was a driving theme in all periods.

4. Conclusions

We believe that this analysis is a valuable contribution to this area of research. To the best of our knowledge, no bibliometric analysis has been carried out on sustainable finance and investor behavior, especially considering that two different types of analysis were carried out using VOSviewer and SciMAT. It provides a global synthesis of the data analyzed, although it does not allow an in-depth analysis of the articles used. In any case, citation bias often affects this type of analysis, since certain articles with a longer historical trajectory are more frequently cited and co-cited.

The present bibliometric review identified a set of 372 articles on sustainable finance and behavior using the Scopus database and 355 using WOS published over three decades (1990–2023). The number of publications has increased significantly, especially in the last decade, as shown by both databases. This trend has also been observed in other bibliometric reviews on finance and sustainability (Marín-Rodríguez et al., 2022; Kumar et al., 2022; Dervi et al., 2022) and green consumer behavior (Xiao et al., 2023). It also shows that most research is concentrated in China, the United States, and the United Kingdom, which are also the countries of origin of the most cited papers. This pattern of highly cited and productive countries is also observed in other financial literature reviews (Xiao et al., 2023; Marín-Rodríguez et al., 2022; Kumar et al., 2022), although the United States is sometimes found to be the most productive country ahead of China (Dervi et al., 2022). Therefore, research in this area in countries other than those mentioned above represents an opportunity for research.

The fields of study are mainly focused on economics and business, with about 50% of the production. Thus, the most cited articles were published in business and economics journals. However, the highest number of publications are in journals such as Sustainability and the Journal of Cleaner Production. Socially responsible research and CSR (corporate social responsibility) are the most relevant keywords. In this respect, this highlights opportunities to develop other less developed fields such as sustainable consumer behavior, financial behavior, experimental finance, among others, always applied to the field of sustainable finance.

The analysis of centrality and density shows that there are many motor themes to develop the broad theme of sustainable finance, with the theme of “financial performance” standing out, and as emerging themes, a block made up of three themes stands out: “green bonds,” “corporate governance,” and “sustainable development.” The evolution of these themes over time also shows that green finance is a recent theme with the greatest growth and potential for development, which means new research opportunities, while attitudes and financial performance have remained the driving themes throughout the period analyzed.

The number of authors is relatively high, although there is little specialization, as only 34 have published more than one article in this area, which highlights the need for researchers to specialize in this area of knowledge and to establish a collaborative network.

As a final note, we would like to indicate that one of the limitations raised in this work is the need to merge and identify articles in the two databases, WOS and Scopus, although it is a relative limitation, since the WOS data have only been processed with SciMAT and the Scopus data with VOSviewer. We understand that there are a series of biases that arise from practically all bibliometric reviews, such as the underrepresentation of certain areas and by types of publication, among others that are worth mentioning.

Regarding future work, the bibliometric patterns identified in this study highlight several promising and underexplored avenues for future research in sustainable finance, particularly in relation to investor behavior. Future studies should go beyond descriptive approaches and deepen the understanding of psychological and behavioral mechanisms—such as cognitive biases, emotions, values, and social norms—that influence investors’ sustainability-related decisions. Particular attention should be given to the heterogeneity of investor preferences and the interaction between financial risk perception and ESG motivations.

Second, future research should incorporate experimental and data-driven methods to improve upon traditional bibliometric analyses. Discrete choice experiments, conjoint analysis, behavioral finance experiments, and machine learning techniques (e.g., sentiment analysis on investor communications or ESG fund marketing) can provide robust evidence to capture investor behavior in more realistic scenarios.

Third, there is a need to further develop research on specific and currently less examined areas. These include sustainable consumer behavior within financial contexts, the investor appeal of green bonds and impact instruments, and the application of experimental finance to questions of sustainability. Studies focusing on emerging economies and regional investor behavior would also add valuable geographical diversity to the literature.

Fourth, addressing the observed fragmentation in the co-authorship and citation networks calls for more international and interdisciplinary collaboration. Encouraging the formation of specialized research groups can enhance the field’s cohesion, promote cumulative knowledge development, and facilitate cross-country comparative analyses.

Finally, future studies aimed at bridging the gap between theory and practice are needed. Academic insights need to be transformed into recommendations for policymakers, asset managers, and financial advisors to align investor behavior with sustainability objectives. Researchers can help in designing tools, incentives, and disclosure frameworks that foster the adoption of responsible investment practices.