Introduction

ETFs have a long history and they started trading in Canada first in the 1980s. They became popular as passive investment instruments after the 2008 financial crisis. Today, they dominate the global investment scenario. There have been numerous studies in this space related to individual characteristics of ETFs, their regional return comparisons, differentiation between ETFs vs mutual funds, etc., but the space still lacks an effort where all the information can be readily referred to and analyzed. The purpose of this research article is (1) to get an all-encompassing resource for exchange-traded funds, (2) to identify deficiencies in knowledge about exchange-traded funds, (3) to generate innovative research ideas related to exchange-traded funds for future research. The paper's overview subsection gives a brief synopsis of the history of ETFs since their inception, the development of ETFs as commercial products over a period, the rise of the ETFs post-2008 financial crisis. Through this study, we will address the latest trends associated with the ETFs, important issues and themes related to ETFs, and a conceptual framework of investment via ETFs. The article also discourses the existing gaps and future research trends associated with ETFs.

In the recent past, assets under management (AUM) of ETFs have shown tremendous growth in the Western world and are getting popular in emerging markets as the phenomenon of global financial markets integration is rising [102]. The majority of the passive investments globally are made with the help of index funds or ETFs. They track and replicate the performance of associated reference index, e.g., S&P 500, and Nikkei [66]. Low management fees are an intrinsic benefit of exchange-traded funds (ETFs) versus mutual funds, as they just follow and rebalance their portfolio according to the underlying benchmark index, whereas active funds have higher expenses because of additional costs required to maintain research houses, trading costs during the frequent churning of portfolio, which results in higher expense ratio for active funds [27, 118].

ETFs are available for trading on a real-time basis, whereas mutual fund units price is available only at the end of the day [31]. ETFs have favorable tax treatment in many countries, and they provide relatively superior returns as compared to active mutual fund schemes [122]. Some additional advantages of ETFs include low volatility as compared to equities and the opportunity of diversification [98].

It is crucial to measure the returns from various market indices, and ETFs aim to do the same [133]. The ETF's output matches with the increasing or decreasing pattern of an index. ETFs can be used for the production of sophisticated trading strategies, margins, or shortening purposes [17].

When it comes to capital gains taxes and liquidity costs, exchange-traded funds (ETFs) have a clear edge over index mutual funds, although the investing mechanism in both vehicles is similar. Investors making transactions in the ETF are subject to individual transaction costs and are directly exposed to capital gains when the units are sold, but in index funds, the same is borne by all fund investors at the same time [69].

ETFs are categorized in accordance with the methodology and the underlying assets they hold. Physical ETFs comprise underlying assets in the form of stocks or bonds. Synthetic ETFs aim to return the value of the index using total return from swaps [124]. With the above advantages of ETFs over active mutual funds, ETFs grew over the years and became the dominant part of the overall investment space.

Overview of global ETFs

Index funds and ETFs that are widely acknowledged as passive investments have grown internationally over the years. They recently exceeded active investment markets in the USA (Table 1).

Table 1 Evolution of ETF from mutual funds
Full size table

ETFs originated from the idea of a stock collection, like mutual funds or an index portfolio. The Massachusetts Mutual Fund, the first contemporary mutual fund that still exists today, was established in 1924 and went public in 1928 [81].

In 1950, Markowitz introduced the Modern Portfolio Theory of optimal portfolio returns with minimal risk [92]. The application of modern portfolio theory in institutional investments was led by Fama and Sharpe through their efficient market theory [114]. In 1973, Burton Malkiel popularized the idea of "buying the market" by using a portfolio rather than relying on individual stocks through his book "A Random Walk Down Wall Street" [111]. The idea of mimicking the S&P 500 Index by different pension funds, and private portfolios, began by following the advice from the above book.

Vanguard introduced the idea of an index fund managed by John Bogle in 1975 (Bogle, 2014). Present ETFs are the extension of index funds with better tax incentives, readily available on exchanges, and cost-effective. ETFs emerged in the form of program trading in the 1980s, which helped investors purchase shares belonging to an index. One of the root causes of the US market collapse of 1987, when the index plunged by 24 percent on Black Monday, was this intensely programmed trading [48]. After several legal battles and prolonged regulatory support, in 1993, the first ETF known as SPDR was listed in the USA. SPDR was the largest trading ETF fund with assets under management of $280 billion in 2019 [41]. The same fund has an AUM of 490 million dollars today.

Until 1995, most of ETF-related activities were limited to the USA and Canada. Organizations such as Morgan Stanley Capital International and Barclays Global Investors embraced the concept of international ETFs that could be traded even though US stock markets remained closed [96]. Barclays Global Investors introduced first Gold ETF [40], and the introduction of QQQ, a technology shares-based ETF were popular among investors. QQQ alone raised over $18.7 billion, while the overall asset size of the total ETF industry in the previous year was $15.7 billion. It has been found that QQQ ETF stocks with lower weight age have encountered higher liquidity compared to higher-weighted stocks [107]. Powershares introduced smart beta ETFs tracking the smart beta indices in 2003 to outperform the overall markets, and the total asset size of smart beta ETFs has grown to 20% of the total ETF asset size in the USA in 2018 [49].

Following the 2008 financial crisis, many investors wanted to protect their assets from market fluctuations. Hence the minimum ETF volatility [7], fixed income ETF [47], short and leveraged ETF [110] were introduced which were popular among investors.

Figure 1 shows that assets under control for Global ETFs rose from 204 million dollars to approximately 10 trillion dollars by 2021 [100]. 2022 figures are till August 2022. Global ETF assets have grown by over 20 times compared to a 130% growth in mutual fund management assets over the last 15 years [61]. The ETF market started from one ETF in 1993 and developed rapidly. In 2002, there were about 1,000 ETFs, and by December 2021, there were 8522 ETF schemes worldwide [119].

Fig. 1
figure 1

Growth in global ETF investment from 2003 to 2021 (source—[100])

Full size image

Figure 2 depicts the rise of the largest ETF market in the USA, where passive investments surpassed active investments in August 2019 [50]. This phenomenon has created a significant trend that might be followed by capital markets globally, and emerging markets may not be an exception.

Fig. 2
figure 2

Journey of USA Active MFs vs Passive ETF Equity Funds [77]

Full size image

Figure 3 shows the total global fund flows in direct equity investments and ETFs. The trend reveals that the passive market ETFs are attracting new funds, whereas there is negative growth in the case of direct equity funds.

Fig. 3
figure 3

Cumulative positive fund flows for global passive investments [132]

Full size image

If we refer to Fig. 4, according to PWC projections in 2025, the global AUM under passive investments will be well above active investments [45].

Fig. 4
figure 4

Projections of growth in global passive ETF investments [45]

Full size image

With the above information and projected trends, it is very important to understand the current status and prospects of ETFs in developed and developing economies using bibliometric analysis.

Objectives of the study

The primary goal of this study is to review the current state of the research on exchange-traded funds. The queries that come next help to determine the scope of this investigation.:

RQ1: In terms of authors, associated institutions, affiliated nations, affiliated journals, affiliated fields, affiliated research types, and affiliated economies, what are the present trends in exchange-traded funds publication?

RQ2: What are the most important topics and associated themes in this area?

RQ3: What is the conceptual framework of research on exchange-traded funds, how has it changed over time, and what recent advancements have occurred in this field of study?

RQ4: What exactly are gaps in knowledge associated with exchange-traded funds and potential future research areas?

The remainder of the document has been developed as outlined below: Sect. Research methodology is associated with research methodology, Sect. Method of analysis delineates the findings associated with the trends in publications associated with ETFs, Sect. Results deals with citation network, Sect. Citation network analysis details keyword analysis, Sect. Keyword analysis includes co-citation analysis, while Sect. Co-citation analysis covers the conceptual framework of ETF investments, Sect. Conceptual framework of ETF investments details conclusion and discussion, Sect. Results and conclusion contains future trends, Sect. 12 concludes the study with limitations.

Research methodology

Identification of suitable terms for search

The study comprises papers between the last 50 years, ranging from 1973 to the present, connected to the Scopus database. Scopus is a citation and abstract database from Elsevier, which was introduced in 2004. Scopus covers 11,768 publishers, 34,346 peer-reviewed journals, and 36,377 articles as of December 2022 in wide-ranging areas, including management, social sciences, life sciences, engineering, health sciences, etc. According to Jo [65], the scope of the Scopus database is substantially greater than that of the Web-of-Science database, which only contains 12,000 titles and ISI-indexed journals, but it is challenging to locate peer-reviewed journal articles published before 1996 using Scopus. Additionally, authors used free tools like R and Gephi for bibliometric analysis.

Preliminary search results

When searching the Scopus database, authors chose "exchange-traded funds" as their primary keyword in the "keywords, abstract, and title" field. In the initial search, authors found a total of 2056 articles. These were refined with document types “article” and “review.” Authors then sorted the output “by relevance” and focused their studies mostly on articles related to finance, economics, energy, and social sciences. After this filtration process, the authors had 958 articles, which were used for bibliometric analysis in the research paper. The stepwise analysis procedure is shown in Fig. 5.

Fig. 5
figure 5

Stepwise analysis procedure

Full size image

Method of analysis

There are several kinds of systematic review studies, including—structured reviews that concentrate on commonly used techniques, ideas, and frameworks [108], reviews based on theory [131], based on the framework [101], bibliometric review [41], meta-analysis [109], hybrid narration [37], model/framework-based review [22], etc. Bibliometric is a widely used method for tracing the knowledge architecture of a study topic [88]. To summarize the literature, prevent bias, and identify any potential research gaps, systematic literature reviews are utilized [73, 123].

The two bibliometric techniques that are most frequently employed are citation and co-citation analysis which are used for demonstrating commonalities between citing and cited publications [138]. This study uses techniques like citation network analysis and publication trends [93, 120]. Content analysis, clustering, and keyword analysis can be done by using co-citation analysis [135].

To present a comprehensive analysis and fill up a gap in the existing literature on ETFs, the authors conducted a bibliometric study followed by content analysis [103]. Besides acknowledging the current research trends in ETFs through this study, the authors also glimpse potential future research directions. The analysis approach is displayed in Fig. 6.

Fig. 6
figure 6

Analysis approach

Full size image

Results

Descriptive statistics

Table 2 shows the details regarding the number of articles extracted from the Scopus database. A total of 923 articles and 35 review papers were reviewed, summing up the total reviewed documents to 958. These articles were sourced from a total of 358 different sources. The focus of the study was on the finance and economics domain to make sure that authors cover a comprehensive overview of ETFs.

Table 2 Descriptive statistics (created by authors from analysis)
Full size table

A total of 1023 unique keywords were found during the analysis, and the authors used a total of 2248 keywords. Extended keywords connected to the text by the Scopus database are called Keyword Plus (ID). The literature review covers 50 years from 1973 till date, with an average citation per document equal to 9.926. Total of 157 single authors contributed to 199 single-authored documents. There are 1690 authors involved, and they wrote 2251 articles in collaboration and the number of documents per author is 0.563.

Annual publications and growth analysis

ETFs are popular financial instruments, with global assets under management recently hitting $9.37 trillion in August 2022 [29]. There has been substantial research published on the topic over the years, and some of the important contributions include the introduction to ETFs [106], the world’s largest ETF Spider fund [43]¸ analysis of herding behavior in sectoral ETF [51], comparison of mutual funds vs ETFs [3], active vs passive fund management [36], commodity ETFs [76], volatility and ETFs [16], etc. In the last few years, there has been a huge contribution to the body of knowledge associated with ETFs with the publication of hundreds of articles in the last 3 years, which indicates the depth, prospects, and the subject’s potential.

Figure 7 shows the count of research papers published on ETFs yearly from 1973 to 2023. The current analysis includes 951 publications published in finance and economics. Over the previous 50 years, the number of publications has climbed at a stable annual rate of 2.22%.

Fig. 7
figure 7

Number of publications per year

Full size image

In 1973, there was only one publication, and in 2022, there are 127 publications associated with ETFs. The amount of literature published in 2022 (127) shows that the subject has enough depth and the capacity to support additional research, even though despite the fact that the quantity of the research papers published has increased continuously. The figure represents a consistently increasing pattern of article publications over a period with no sluggishness, indicating the overgrowing interest of researchers in ETFs.

Authors with the highest influence

Table 3 displays data of 15 most pertinent and prominent writers who have made notable contributions to the field of ETFs. These authors have helped in testing already established theories by using empirical tests and contributing toward the new theoretical developments occurring in ETFs, which created a platform for further research on the topic. The information in the table is presented in decreasing order, with the author with the greatest contribution mentioned first. Tse Yiuman. has published the most research articles since 1973, with 16 independent publications and 6.67 articles co-authored by others. Xu Liao. is second highest contributor with a score of 15.08 (11 articles as individual authors and 4.08 articles in collaboration with others). The table appropriately credits the efforts of the other writers.

Table 3 Most influential authors
Full size table

Author’s impact

Productivity indices can be used to gauge the impact of the authors on ETFs. Table 4 lists the top 15 author’s contributions in this area. The author’s impact is evaluated by using the three citation impact and productivity measures, viz. h index, g index, and m index. The Hirsch or H index counts the total number of citations linked with each publication and the number of publications. To support the lesser cited articles, the G index examines the total publications with the highest citations to the lowest citations, whereas the M index includes a version associated with h index which describes an author's h index (number of articles published) each year from his or her first publication.

Table 4 Impact of authors by using productivity indices
Full size table

Based on the authors' h index values, Tse Yiuman. has the most influence, followed by Ivanov, Stoyu., and Rompotis, Gerasimos. In case of total number of citations, too, the above sequence of authors is maintained, with Tse Yiuman. being the most cited author followed by Ivanov Stoyu and Rompotis Gerasimos.

Countries represented by corresponding authors

Table 5 contains information about the country associations between the corresponding author and their co-authors. Multi-country publications by multiple authors are denoted by the acronym MCP, while single-country publications are denoted by the term SCP. The MCP ratio is the percentage of all published articles that are multi-county articles. The details below help understand each country’s contribution in terms of research on the topic as well as author’s collaboration with co-authors residing in different countries. According to the table, the USA contributed the most to ETF research with a total publication of 228 articles, out of which 197 articles have been published in the USA alone, and only 31 articles representing 13.6% articles are multi-county articles. Because the MCP ratio in the USA is less, it indicates smaller multi-country cooperation, whereas, in the case of China and the UK, the same is 32.9% and 36.59%, indicating higher multi-country collaboration. China, the UK, and India are among the leading nations contributing to the research topic, but their volume of combined publications is still less than the USA alone. This shows much higher research interest and research concentration on ETFs in the USA.

Table 5 Countries represented by corresponding authors
Full size table

Most prolific journals

Table 6 describes the list of top 15 journals in Economics and Finance that have immensely contributed to the publication of research papers associated with ETFs. In order to learn about the most popular issues in the field both now and in the past, researchers typically review a number of articles published in the most prominent journals related to their chosen research area.

Table 6 Most prolific journals
Full size table

The peer-reviewed journals are arranged according to the quantity of publications, with the Journal of Portfolio Management having the largest number of published papers on the subject of ETFs in the last 15 years, closely followed by International review of Financial Markets and Journal of Futures Markets.

Most influential countries

When it comes to financial products, it has been noted that the product's origin region has undergone the most initial research, with research on the product's operational efficiencies coming next. Subsequently, emerging markets contribute to the product's related research field since the financial product is accessible in both developed and emerging economies. Table 7 describes the number of articles published, total citations, and an average article citations associated with each country. With 701 research publications overall, 4172 citations, with a median publication citation value of 18.72, the USA topped the ranking, followed by emerging market China with 297 publications, 545 citations, and 6.19 average article citations. With the most published articles as well as maximum publication citations, the USA leads the world in contributions to the ETF industry, indicating better quality of research. Other countries like China, the UK, and Canada have also contributed substantially to the research, but they are no way near as compared to the USA.

Table 7 Scientific production countrywise
Full size table

Citation network analysis

Citations indicate the overall strength of the research article, journal, and the authors in the respective research area. Articles with novel research ideas contributing to the existing body of knowledge always attract high citations.

Table 8 lists the top 15 most cited research papers in ETFs. The articles are ranked in descending order, with the highest cited article at the top, followed by other articles as per the total citations. The top cited research article is “Realized GARCH: a joint model for returns and realized measures of volatility” authored by [53]) with 316 citations. In this article, the authors have introduced a newly realized GARCH model combining returns and volatility associated with ETFs. benchmarked against the Dow Jones Index. The second research article is “Intraday Price Formation in U.S. Equity Index Markets,” authored by [54] with 250 citations. The writers of these widely referenced papers are well-known ETF experts.

Table 8 Maximum cited articles
Full size table

Keyword analysis

A study topic can be divided into several subfields, and each subfield can subsequently be examined from diverse perspectives using a variety of perspectives and approaches. The literature generates various relevant keywords that highlight the importance of the field investigation. Thus, it is possible to understand both the breadth and depth of the study field by utilizing keyword analysis.

Most frequently used words

Referring to Fig. 8's word cloud, readers are able to see that "exchange-traded fund" "etfs" and other similar terms are the most frequently occurring terms throughout the investigation. They stand for exchange-traded fund's reliability and domination during the past twenty-five years. Additional well-known terms are “volatility,” “liquidity,” “tracking error,” and “price discovery” and those are all connected to the exchange-traded fund's operating features. In the advanced nations, a great deal of investigation has been conducted in these fields. When the ETFs were new in the markets, the liquidity was very low, and hence, there was a lot of volatility, and price discovery in ETFs was a challenge. As the ETFs are normally benchmarked to certain index, lack of liquidity and difficulty in price discovery resulted in higher tracking errors. All these topics associated with ETFs have been extensively researched. As time progressed, the ETFs moved to emerging markets and ETFs is a popular area of research in these countries.

Fig. 8
figure 8

Word cloud

Full size image

Factorial analysis

The hierarchical clustering relationships between the objects can be explained by using a dendrogram. Figure 9 represents the dendrogram for “exchange-traded funds.” It is categorized into 4 distinct cluster patterns.

Fig. 9
figure 9

Factorial analysis

Full size image

The first cluster represents a commodity-based ETF scheme, which represents keywords associated with commodity ETFs and their analysis. The second cluster represents keywords associated with energy markets, but they represent futuristic time series analysis by using the granger causality method and other forecasting methods. Instead of only focusing on ETFs, it also represents index funds, which are part of the passive investment family.

The third cluster is the largest keyword cluster representing ETF management. It covers various topics related to ETF operations, such as volatility, liquidity, efficiency, cost-effectiveness, benchmarking of the ETFs against different indices, portfolio management, and risk management associated with ETFs. The fourth and the last cluster represents keywords associated with current and future trends in ETF, including the Covid-19 situation and its impact on ETF investments, exchange rate, and spillover effects on various economies. It also focuses on the emerging markets theme, where a lot of contribution to the existing body of knowledge has taken place in the last few years.

Co-occurrence networks for keywords

Keywords are normally available in the research articles after the abstract, and they provide important topics or concepts associated with the research article. Normally, six to eight keywords are mentioned in each research article. Some important keywords always get repeated over a period in different research articles. We call this type of event as keyword co-occurrence. It provides much more analytical meaning for the researchers.

Keyword co-occurrence helps in identifying the most important research sub-areas associated with any specific domain, and they also describe the closeness of one research sub-area with another. Thus, it helps to understand the prominent co-occurred keywords associated with any specific area of interest. Figure 10 describes the keyword co-occurrence associated with ETFs.

Fig. 10
figure 10

Keyword co-occurrence

Full size image

Threefold plot

The threefold plot (Fig. 11) helps us to understand the relationship between researcher (AU), researcher keywords (DE), and source of the publication (SO) associated with ETFs. This information is useful to identify the most prominent authors associated in the research area, essential subtopics that may be found by using the author's keywords, and they can be approached for advice and co-operation. Threefold plot also provides information about the prestigious journals connected to the release of these studies. The journal entries may be useful in identifying the potential journals for future publications in the identical area of research.

Fig. 11
figure 11

Threefold plot

Full size image

Co-citation analysis

Figure 12 shows the co-citation analysis associated with ETFs. Initially, 1690 authors were considered for the analysis. Out of those, disconnected nodes representing standalone articles were removed. The figure represents the top 100 nodes and their co-citation analysis. The higher the node size and the font, the better the author is in terms of co-citation. From the figure, the top 5 authors linked by co-publications are Ben David., Amihud Yakov., Delcoure Natalya., Elton Edwin, and Hasbrouck Joel.

Fig. 12
figure 12

Co-citation analysis

Full size image

Bibliometric coupling

Bibliometric coupling is a completely different process from co-citation analysis, and in this case, papers referenced by the author while writing the article play a major role. This indicates that the most recent 2023 paper is likely to have bibliometric coupling associated with the old research article on a similar subject. This analysis is mainly carried out to find out the related articles and create a relationship with them by using the citation analysis. From Fig. 13, it can be seen that Zhu et al. [137] with an article published in the Journal of Futures Market have referenced earlier authors working in a similar area like Chen Louis., Broman, Lachance, etc.

Fig. 13
figure 13

Bibliometric coupling

Full size image

Data clustering

Different major sets of articles on the research topic can be obtained by using the technique of data clustering. It is observed that the articles belonging to the single set of sectors normally are similar in their characteristics and research area. Therefore, the fact that both articles are part of the same cluster aids in our comprehension of the co-occurrence and related sub-themes of the research issue. Data clustering was employed by the authors in order to determine the recurring themes in the literature related to exchange-traded funds (ETFs), and the details are available in Table 9. Clusters 1,2,3,4, respectively, represented 46,88,81,35 articles as per Table 10, and the top 10 page ranked articles from each cluster have been chosen below.

Table 9 Top research articles per cluster on exchange-traded funds with co-citations
Full size table
Table 10 No of research papers associated with each cluster
Full size table

Cluster 1 investigates several aspects associated with ETFs, such as the effect of increased ownerships of ETFs on pricing and volatility, corporate bond ETFs, liquidity and volatility associated with ETFs, pricing of ETFs and leveraged ETFs.

Cluster 2 discusses various topics such as ETF pricing inefficiency, factors affecting the flows of funds into ETFs, tracking errors, emerging market ETFs, volatility and tracking errors, hedging of ETFs, feedback trading, premiums and discounts associated with ETFs.

Cluster 3 describes research articles associated with ETF arbitrage, the relationship between ETFs and volatility, innovation and informed trading associated with ETFs, return spillovers from USA and Canada markets, and return predictability of ETFs.

Cluster 4 provides invitation results associated with momentum strategies of ETFs, why active mutual funds are taking positions in ETFs, a return comparison of large-cap and small-cap exchange-traded funds, climate change ETFs and ESG ETFs, behavioral finance and ETFs.

Thus, cluster analysis associated with ETFs describes various dimensions associated with the research area in the form of sub-topics in four clusters, although they share a common relationship.

Cluster 1 is associated with the traditional topics associated with the ETFs concerning the operational and demographic aspects of ETFs and covers important parameters of pricing, volatility, investment demographics, and liquidity.

Cluster 2 represents an era of ETFs, where more and more ETFs started listing, trading, and the difficulties they faced during the period in the form of pricing inefficiency and tracking errors associated with ETFs. It also challenged the results of traditional operational aspects as new findings associated with high volumes emerged. It throws light on aspects related to hedging of ETFs, feedback trading, behavioral finance, and passive investments.

Cluster 3 represents the period past 2010, after the 2008 financial crisis, when ETFs became populate investment instruments and discusses the research associated with ETF arbitrage, informed trading in ETFs, and return predictability associated with ETFs. These characteristics indicate stability and growth associated with ETFs, and the risks involved in the investment process.

Cluster 4 is associated with the recent ongoing period, and it represents various topics associated with ETFs, such as ESG investing, which in the recent past has generated tremendous wealth for many investors, behavioral fiancé aspects associated with ETFs, and momentum strategies associated with ETFs. Thus, cluster analysis helps us understand the variety of sub-topics and associations in the area of ETFs.

Table 10 represents the number of articles associated with each cluster, from which the top 10 articles with the highest impact are listed in Table 8 as a part of cluster analysis.

Conceptual framework of ETF investments

Figure 14 depicts a conceptual framework that summarizes a thorough content examination of clusters. This framework illustrates the overall ETF investment process from the retail investor perspective. As of today, 7 out of 10 investors in USA invest in the financial markets by using ETFs or passive investments. The thrust on the ETF investment has increased, as direct mutual fund schemes have not been able to beat the respective benchmark indices during the last few years. Therefore, it is critical to comprehend different factors that are important in ETF investments and the benefits that investors receive out of investing in ETFs.

The impact of social and demographic factors on investment in ETFs is especially important in today’s context, especially after the COVID-19 pandemic [99]. It has been observed that investors of younger age, middle- to high-income investors with better financial literacy, and living in cities tend to invest directly in ETFs by using robo-advisory services [15]. It has also been observed that males tend to be more confident and invest higher amounts via passive routes than females [117]. Some other important factors that investors consider while making passive investment include impact investing, risk tolerance, investment goals, behavioral biases, diversification, tax considerations, region and economy of investment and ESG factors [68, 97]. In a nutshell, social and demographic factors, behavioral factors, structural factors, domestic factors, and traditional factors play a critical part in the individual investment process in ETFs.

Along with the factors discussed above, passive investment risks such as tracking errors [116], liquidity issues, concentration risks [6] are some of the factors along with the financial literacy [127] considered important while investing in ETFs.

Some of the benefits of investing in ETFs include lower maintenance fees, long-term consistent returns, diversification, and low volatility [95]. Thus investors invest in different types of ETFs [9] depending on their risk appetite for favorable returns. Recent themes of ETFs such as ESG ETFs, sustainability ETFs, and green ETFs have provided the best returns to investors in the last few years [121].

Results and conclusion

It is important to interpret the results associated with ETF bibliometric analysis in a systematic manner, and authors have used the technique of sense-making approach for the same [84]. It involves three steps, viz. scanning, sensing, and substantiating. Scanning involves the understanding of the results from important steps of bibliometric analysis like publication and citation metrics, citation and co-citation analysis, coupling analysis, etc. It provides a firm foundation for succeeding phases of sense-making by laying a foundation for the broader analysis, which is done in the sensing phase. Sensing describes the analysis in the form of possible reasons by using “what,” “why,” and “how” approaches. It provides a more complete and comprehensive perspective on bibliometric results, uncovering established, untapped, and unknown aspects. Substantiating helps in the trustworthiness of the overall results by following the process of credibility, dependability, transferability, and conformity. It increases the research’s recognition and dependability, increasing its adoption by scholars as well as business experts.

Scanning

In financial research, exchange-traded fund (ETF) represents a crucial financial resource and is useful in a range of financial and investment contexts. Although it is a heavily explored topic with an increasing number of publications, the current surge (2020, 2021, 2022) in publications highlights the depth and promise of ETFs. In order to establish a distinct backdrop and research boundaries, the article lists the most noteworthy research publications, authors, keywords, and related research clusters of ETFs, by using a bibliometric analysis of 958 research papers published over a 50-year period (1973–2022) that were obtained from the Scopus database after a thorough review of the literature.

According to the findings, the USA is the major contributing nation to ETF research, which is followed by China, the UK, and Canada, and Spain. Barring China, there is hardly any contribution from emerging market economies in the ETF research. Citation-based analysis, articles, and authors reveal that Tse Yiuman., Xu Liao., Ivanov Stoyu., Rompotis Gerasimos., Zhao Yang. were the most influential authors, whereas Tse Yiuman., Ivanov Stoyu., Rompotis Gerasimos., Xu Liao., Liu Qingfu. were the most impactful authors of ETFs research. The majority of these researchers are from established nations, which suggests that emerging markets have not contributed much to the field of research. According to keyword-based analysis, exchange-traded fund, etfs, volatility, liquidity, tracking error, and price discovery are the main keyword-based dimensions of ETF research.

Keyword co-occurrence further reveals that operational aspects of the traditional ETFS, pricing inefficiency, tracking errors associated with ETFs, post-financial crisis ETF characteristics such as ETF arbitrage, informed trading in ETFs, and return predictability associated with ETFs and recent trends in ETFs such as ESG investing, machine learning, behavioral finance, and momentum strategies are most important sub-themes of ETF research revealed from the cluster analysis.

Sensing

ETFs as financial products took time to establish, and one of the obvious reason was, it took time for the people to understand the product and its benefits. During the 1990s people slowly started realizing that active mutual funds despite charging heavy maintenance fees were unable to beat the benchmark returns and people started turning toward ETFs. SPIDER and QQQ ETF schemes in the 1990s were the big success stories of the USA due to the above reasons. It took time for ETFs to establish themselves in the developed world, and they faced issues faced by developing financial products like high volatility, low liquidity, higher tracking error, and difficulty in price discovery. The same problems are currently faced by developing countries, where the passive investment trend is emerging, and these are represented as a part of cluster 1 analysis. 2008 financial crisis turned out to be a major boost for ETFs. People lost substantial money in direct equity and mutual fund investments during the market crash, and they started realizing the power of index investing with minimum maintenance fees and transparency. Over a period, direct investments in equities and mutual funds in developed markets have been declining, and the ETFs are gaining that market share. Important factors such as the use of ETFs for hedging and investor behavior during passive investment were some of the important topics represented in cluster 2. In case of cluster 3, the thrust has been given to risk management associated with the ETFs, whereas cluster 4 talks about new-age ETFs such as green ETFs, ESG ETFs, and sustainable ETFs. From scanning, it can be seen that although ETFs have existed for a number of years, the researcher’s interest is overwhelming and new areas in the ETF space have been continuously explored. It has been found that most of the research has been done in developed countries as ETFs were prospering, whereas authors from developing markets have started making contributions to the body of literature, as ETFs are gaining ground in emerging markets. Overall, it seems that ETFs are here for a long haul, and as the figures indicate, they will surpass the active investments in days to come.

Substantiating

Referring to the most prolific journals and most cited articles, it provides us credibility regarding the importance of ETFs as sustainable financial products and its importance [13]. We can easily cross-validate these articles by considering the number of territories covered and important institutions through which the ETF research is conducted for credibility [128]. Co-citation analysis of ETFs can be easily cross verified using bibliometric coupling [42]. Conformity regarding ETF analysis can be easily established by the thorough analysis of the topic using the proven bibliometric analysis and associated citations [72]. In case of ETFs, credibility and conformity have been already established, which further helps to establish the dependency of the study over a period from existing reviews [86]. These findings from ETF bibliometric analysis are very important and pave the way for the transferability [85] to different close areas like derivatives. ESG and sustainable investing where the ETF products associated with these areas already exist.

These results have significant implications for prospective academics and researchers. Various sub-themes obtained from the results can be interlinked to explore new areas of research associated with ETFs. Policymakers especially from the emerging markets may find this analysis useful to understand the pain points associated with passive investments in emerging markets. If these pain points are properly addressed, it will help in further progress of ETFs in emerging economies. All things considered, research scholars, and senior research fellows alike may find this paper helpful in understanding the breadth and potential of ETF research.

Future trends

The above study highlights that most of the work related to ETFs is done in developed countries, although developing countries like China have shown decent contribution to the existing body of knowledge in recent years. ETFs have been first choice of investments for many retail and institutional investors in developed countries and emerging markets are slowly adopting to this western market changes. However, developing countries are facing issues like tracking errors, liquidity, volatility, limited product offerings, pricing of ETFs, etc. Each market has got unique set of characteristics, and hence although previous studies in this area will help, it has also opened up a wide area of research on above points in emerging markets.

Figure 15 provides a picture regarding trending topics in ETFs. Use of machine learning for passive investments, COVID-19, and its impact on ETF investments are some of the trending and forthcoming research areas. Lot of focus is currently associated with clean energy and associated commodity ETFs, and ESG (environment, social governance) ETFs are of the upcoming research areas. Forecasting ETF returns by using AI, ML are the current trending topics as well as upcoming future areas of research in the ETF space. AI and ML are important in passive investments as these are used to set the algorithms for investment into passive schemes for robo-advisory services. Although some of such experiments have been already successful in the developed world, this is one of the future area of research in emerging economies. ETFs play major role in hedging and risk management. Considering current geopolitical and economic uncertainties due to global inflationary trends, research associated with interest rate ETFs, currency ETFs, commodity ETFs, and their role in overall stability of the markets is an upcoming area of research (Fig. 15).

Fig. 14
figure 14

Conceptual framework

Full size image
Fig. 15
figure 15

Trend topics

Full size image

Limitations

The research was conducted only on Scopus database, which has certain advantages and limitations, which are discussed in details as a part of methodology. The study considers top 958 articles in ETFs, which are highly cited and important. The insignificant articles associated with the topic have been eliminated. Perhaps the omitted articles may come out with slightly different perspective associated with ETFs.