Volume 26 Issue 3 Article 2 September 2024 Mapping Characteristics and Financial Importance of Mapping Characteristics and Financial Importance of Development Banks Across the World Development Banks Across the World Jan Porenta University of Ljubljana, School of Economics and Business, Ljubljana, Slovenia Vasja Rant University of Ljubljana, School of Economics and Business, Ljubljana, Slovenia, vasja.rant@ef.uni-lj.si Follow this and additional works at: https://www.ebrjournal.net/home Part of the Finance Commons, and the Growth and Development Commons Recommended Citation Recommended Citation Porenta, J., & Rant, V. (2024). Mapping Characteristics and Financial Importance of Development Banks Across the World. Economic and Business Review, 26(3), 168-183. https://doi.org/10.15458/ 2335-4216.1340 This Original Article is brought to you for free and open access by Economic and Business Review. It has been accepted for inclusion in Economic and Business Review by an authorized editor of Economic and Business Review. ORIGINAL ARTICLE Mapping Characteristics and Financial Importance of Development Banks Across the World JanPorenta,VasjaRant * University of Ljubljana, School of Economics and Business, Ljubljana, Slovenia Abstract This paper investigates the nancial importance of development banks across regions and income groups. We construct a global banking dataset for the period 1995 to 2021 and analyse the distribution of development banking assets across macro regions and country income groups. We create a composite global and national Development Bank Financial Importance index (DBFI index) that enables us to rank the most nancially important development banks across the globe. Development banks play signicant and diverse roles in the global nancial system, but their nancial importance varies across regions and income groups. The paper offers a broad analysis of the global development banking landscape and advances the area of research further. Keywords: Development banking, Development nance, Financial importance, Socio-economic development, Regional analysis JEL classication: G21, G28, O19 Introduction D evelopment banking and development nance, which stagnated at the beginning of the cen- tury, have regained considerable relevance in recent years. Development banks and development nance institutions that act as lenders (hereinafter, develop- ment banks) play a crucial role in advancing sus- tainable socio-economic development in developing, emerging, and developed countries. These institu- tions are also nancially important as their nancing accounts for 10% of the total global investment (Xu et al., 2021). Despite their developmental and nancial signif- icance, development banks receive less academic attention than conventional banking institutions. While studies on individual development banks or global surveys of the sector exist (e.g., Luna-Martínez & Vicente, 2012), there is a notable gap in the literature regarding a comprehensive, multiannual global ex- amination of the nancial importance of development banks and, more broadly, the development bank- ing sector. The sector exhibits substantial qualitative and quantitative heterogeneity, resulting from differ- ences in asset size, mandates, geographical scope of operation, and ownership levels of individual institu- tions. A global, regional, and national understanding of the development banking landscape is impor- tant for both academic and practical purposes, as it informs regulators and policymakers about the devel- opment ecosystem and the instruments that can be utilized for development. In this paper we address this gap by asking two research questions: How do development banking sectors differ in their nancial importance across regions, country income groups, and mandates? How can we measure and compare the nancial importance of development banks at the global and national level? To answer these questions, we build upon the work done by Xu et al. (2021) and construct a global bank- ing dataset for the period 1995 to 2021. We gather qualitative and quantitative data on development Received 3 January 2024; accepted 15 March 2024. Available online 16 September 2024 * Corresponding author. E-mail address: vasja.rant@ef.uni-lj.si (V . Rant). https://doi.org/10.15458/2335-4216.1340 2335-4216/© 2024 School of Economics and Business University of Ljubljana. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/ licenses/by-nc-nd/4.0/). ECONOMIC AND BUSINESS REVIEW 2024;26:168–183 169 banks from multiple sources to explore the nancial importance of the development banking sector and development banks, presenting the rst comparative study not only of individual development banks but of entire development banking sectors. This paper contributes to the existing body of knowledge in two key areas. Firstly, it conducts a detailed examination of the global distribution of development banks with respect to their inherent heterogeneity. This also includes an analysis of the distribution of development banking assets across macroregions and country income groups, providing valuable insights into the nancial signicance of the development banking landscape within the respec- tive banking systems of these regions and income groups. Secondly, employing principal component analysis, the paper introduces a novel Development Bank Financial Importance (DBFI) index. This index enables us to highlight and rank the most nancially signicant national and multilateral development banks worldwide. The rest of the paper is organized as follows: Sec- tion 1 presents an extensive literature review that examines the theoretical rationale for the existence of development banks, the historical evolution of the development banking concept, contemporary roles within the socio-economic landscape, and the critical issues and trade-offs inherent in their mandates. In Section 2, we discuss the data utilized in this study, offering detailed insights into the steps taken to con- struct the dataset. Section 3 encompasses the analysis and discussion of the results derived from our inves- tigation. Section 4 concludes. 1 Literature review The academic literature provides a nuanced por- trayal of the roles assumed by development banks within the contemporary economic environment. As government-sponsored nancial institutions with a primary dedication to the provision of long- term capital to industries (Chern, 2019; De Aghion, 1999), development banks operate as agents of socio-economic development. Their functions encom- pass addressing market failures, bridging nance gaps (Chandrasekhar, 2016; Culpeper, 2012; Ger- schenkron, 1962), and providing indispensable tech- nical expertise and advisory services to projects of developmental signicance, thereby enhancing the likelihood of project success (Gutierrez et al., 2011; Musacchio et al., 2017). Their expertise serves as a catalyst for private nance mobilization, mitigat- ing project risk and fostering trust (Geddes et al., 2018; Zhang, 2022). In this capacity, development banks function as knowledge hubs, adept not only at addressing recognized market failures but also at identifying and delineating barriers to develop- ment (Fernández-Arias et al., 2020; Grifth-Jones & Ocampo, 2018; Mazzucato & Penna, 2015). In addition, development banks are frequently tasked with the mandate to mitigate and smooth economic cycles. Accordingly, during periods of economic slowdowns, many national development banks proactively increase their lending activities, in- jecting liquidity into the economy and contributing to the recovery process (Brei & Schclarek Curutchet, 2017; Feil & Feijó, 2021; Frigerio & Vandone, 2020; Gong et al., 2023; Luna-Martínez & Vicente, 2012). In times of economic downturns, these institutions have the capacity to strategically reduce lending in- terest rates, channelling their development initiatives to generate employment opportunities, strengthen social safety nets, and support other sustainable de- velopment initiatives. This countercyclical effect in lending patterns during downturns extends to mul- tilateral development banks, particularly in regions such as Latin America and East Asia (Galindo & Panizza, 2018). Due to diverse roles they are mandated to as- sume within the socio-economic environment, devel- opment banks are also diverse in their qualitative attributes. While some development banks provide direct nancing through lending, others offer credit guarantees without direct nancing (Musacchio et al., 2017). Luna-Martínez and Vicente (2012) further dis- tinguish between rst-tier development banks, which lend directly to end customers, and second-tier devel- opment banks, which lend to other private nancial institutions, which subsequently lend to end cus- tomers. They nd that 52% of development banks in their sample lend rst- and second-tier, while 36% engage in rst-tier lending and 12% engage in second-tier lending only. In addition to conven- tional lending models and credit guarantees, devel- opment banks also engage in direct equity investment (Chandrasekhar, 2016; Lazzarini & Musacchio, 2010; Pissarides, 1999). Another important qualitative attribute of develop- ment banks is their mandates, which are legal acts or laws through which development banks are es- tablished and governed by. A mandate can either be sector-specic, targeting sectors such as agriculture, small and medium-sized enterprises, social housing, infrastructure, or local governments, or broad, focus- ing on general socio-economic development. Luna- Martínez and Vicente’s (2012) survey of 90 national development banks revealed that 47% of institutions have a broad mandate while the remaining 53% have a narrow and specic development mandate. Xu et al. (2021) similarly nd that 37% of 526 development 170 ECONOMIC AND BUSINESS REVIEW 2024;26:168–183 banks and nancing institutions follow a mandate of general development. On one hand, Rudolph (2009) suggests that while a broad mandate allows develop- ment banks to diversify their portfolios, it may also result in suboptimal goal focus and potential inef- ciencies. Scott (2012) echoes this idea by advocating for narrow and explicit policy mandates. Yeyati et al. (2004) more generally state that public sector banks (of which development banks are a subset, mostly) with narrow, well-dened mandates are less prone to conicting objectives and tend to be more efcient. On the other hand, Fernández-Arias et al. (2020) argue that narrow mandates may lack the exibility needed to effectively target market failures. Qualitative heterogeneity in development banks extends to ownership levels, with Buiter and Fries (2002) highlighting the distinctive multilateral share- holder structure and subsidized capital base that sets multilateral development banks (MDBs) apart from other nancial institutions and national de- velopment banks. Formed primarily in the 1960s, MDBs have undergone substantial transformations in their mandates, progressing from addressing Millen- nium Development Goals (MDGs) to the subsequent Sustainable Development Goals (SDGs) outlined in Agenda 2030 (Engen & Prizzon, 2018; Faure et al., 2015). This evolution, as articulated by Prizzon et al. (2017), is a direct response to an increasingly diverse client base and the enhancement of MDBs’ expertise. Notably, the International Bank for Reconstruction and Development (IBRD), initially established in 1944 for European post-World-War-II reconstruction, later broadened its mandate to encompass global growth and poverty eradication. It is important to note that the extensive mandates and numerous institutions within the multilateral development banking landscape pose certain chal- lenges. Kellerman (2019) notes a recurring trend of new MDB establishments, averaging one every three years, leading to signicant and inefcient duplica- tion of international institutions with overlapping functions. Moreover, Kharas et al. (2014) argue that MDBs struggle to adequately address the nancing gap in lower-middle-income countries and contend with overnancing in more developed regions. In addition to distinguishing between multilateral and national development banks, Xu et al. (2021) fur- ther categorize national development banks based on their focus, classifying them as nationally, inter- nationally, or subnationally oriented. MDBs are also differentiated into regionally and globally focused entities. 2 Data The database utilized for our analysis is constructed from three distinct data sources. Fitch Connect Funda- mental Data (FCFD) and Xu et al.’s (2021) databases are the source of bank-specic nancial variables and information on qualitative bank characteris- tics. World Bank’s World Development Indicators database covers country-specic characteristics. Ini- tially, we obtained total assets for both development banks and other banks from the FCFD database and removed duplicated statements. To prevent overrepresentation of a particular bank’s assets, we furthermore proceeded with context-based removal of duplicates, as detailed below. Prior to the data cleansing process, we eliminated inherent duplicated statements. Working with the FCFD dataset presented several challenges. One primary challenge is the absence of a unique identier denoting the specic period covered by each nancial statement. Whereas most banks re- lease their statements at year-end, some banks follow non-calendar scal years, posting their statements in January (e.g., most Russian banks) or March (e.g., Japanese and Indian banks) of the following year. While records point to year tC 1 in this case, these statements in fact contain data for year t. To address this, the approach suggested by Duprey and Lé (2016) was applied, marking such statements as belonging to year t. Another challenge arose due to the use of multiple accounting standards in preparing nancial state- ments. Statements conforming to either GAAP or IFRS were retained, with preference for IFRS in cases where both were available 1 . Duplicates within the dataset were addressed further by retaining only the most recent statement for banks with restated records. Additionally, only audited statements were included, with preference for the most recently audited state- ment in cases when multiple auditors were involved. We also prioritized unconsolidated reports in cases when both consolidated and unconsolidated state- ments were present. As the number of banks in our sample is sufciently large, this approach preserves individual entities while concurrently offering a com- prehensive representation of global banking assets. This approach aligns with the methodology of Micco et al. (2007). Following an extensive check for du- plicates, we arrived to a uniquely identiable panel dataset structure with bank i, in country j and time period t. To maintain a balanced representation of years in our sample, we limited the time period to 1995–2021. 1 Several studies have shown GAAP and IFRS to be comparable for analytical purposes (see, e.g., Beuren et al., 2008, and Barth et al., 2012). ECONOMIC AND BUSINESS REVIEW 2024;26:168–183 171 Focusing our attention on development bank char- acteristics, we matched quantitative nancial data from FCFD with the qualitative dataset on devel- opment banks compiled by Xu et al. (2021). Their dataset contains substantive qualitative information on over 500 development nancial institutions, in- cluding their mandates, size, ownership, and geo- graphical scope of operation. Since the bank names in both databases did not match exactly, and there was no other unique bank identier enabling an ac- curate automated merging process, we merged the data manually. We considered bank names in multi- ple languages. For some banks FCFD provides names in local languages and for some in English, whilst Xu et al. (2021) provide bank names in both lan- guages. For instance, a very important development bank—the German KfW—is listed in Xu et al.’s (2021) database under its original name, Kreditanstalt für Wiederaufbau, and translated to English as Credit Company for Reconstruction. However, in FCFD, it is simply referred to as KfW. Moreover, some mul- tilateral banks have ofcial names in more than two languages. Xu et al.’s (2021) database records the name for the Development Bank of Latin America (CAF) in English and in Spanish under Banco de De- sarrollo de América Latina; however, in FCFD, this bank’s statements are reported under its Portuguese name, Corporacion Andina de Fomento. Therefore, if a match was manually identied in either language or otherwise, the corresponding data was merged. Our nal dataset encompasses 319,690 observa- tions from 22,949 individual banks, of which 233 are categorized as development banks. Although the precise global count of banks remains uncertain, other widely recognized banking databases offer sim- ilar or somewhat larger coverage. For instance, the Bankers’ Almanac, provided by LexisNexis, contains information on over 21,500 banks, while the Bank- Focus database, jointly offered by Bureau van Dijk and Moody’s Investors Service, encompasses data on 46,700 banks across the world. In terms of the population of development banks, the database constructed by Xu et al. (2021) identies 526 public development banks and development - nancial institutions on a global scale. Consequently, our sample provides a meaningful representation of both the worldwide population of banks and de- velopment banks concerning their numbers. These gures also extend to offer a close approximation of the global banking system in relation to total assets. As depicted in the upper left panel of Fig. 1, our sample represents approximately 92% of the total as- sets within the global banking system 2 for the period spanning from 1995 to 2021, ensuring a comprehen- sive and extensive representation. Within our sample, development banks account for approximately 4.8% of the total banking system as- sets (upper middle and upper right panels in Fig. 1). Notably, there is a consistent and gradual increase in the relative share of development banks’ assets within the banking system from 2005 onwards. This trend underscores the growing nancial importance of de- velopment nancial institutions—this is particularly true in the case of nationally owned development banks, which are well positioned to better under- stand the complexities of the socio-economic environ- ment of individual countries. The collective share of multilateral development bank assets remains rela- tively consistent throughout the period from 1995 to 2021. Over time, the number of development banks in- cluded in our dataset also increases; however, a slow decrease is observed from 2014 onwards (lower left panel in Fig. 1). Development banks are on average signicantly larger in terms of total assets as a share of the global banking system assets compared with the average bank size of other bank types, encompassing commercial, retail and consumer banks, and invest- ment banks (lower middle panel in Fig. 1). Across the entire period, the average development bank rep- resents approximately 0.035% of the global banking system assets, while the average across all other bank types in our sample is 0.011%. This substantial dis- parity in asset size emphasizes the signicance of development banks in nancial markets as individual institutions since they typically operate on a consid- erably larger scale compared to other bank types in our study (lower middle panel in Fig. 1). However, it is also important to note that there is substantial variation in development bank size. For example, our sample includes multilateral development banks, which tend to be relatively large in terms of asset size—the lower right panel in Fig. 1 shows that the average size of multilateral development banks falls below the average size of national development banks only in the period from 2007 to 2010 and in 2021. Inter- estingly, the average size of multilateral development banks declined substantially in the years preceding the global nancial crisis of 2008. The crisis marks a clear end to their waning importance. 2 The variable representing the assets of the global banking system is derived from the Fitch Connect Sovereigns package. This was achieved by aggregating the assets of the banking systems of all countries included in the database. Given that the precise value of the global banking system’s assets cannot be accurately computed, alternative approximations of this time series could be employed. One such approximation is provided by the Financial Stability Board (2022). A comparison of our calculations with the report from the Financial Stability Board (2022) indicates that the differences between the two are negligible. 172 ECONOMIC AND BUSINESS REVIEW 2024;26:168–183 Fig. 1. Sample composition and data coverage. Note. DBD all development banks; NDBD national development banks; MDBD multilateral development banks; COMD commercial banks; RETD retail and consumer banks; INVD investment banks. 3 Analysis 3.1 Regional and country income group distribution of development banks We gained further insights into geographic and socio-economic focus of development banks in our sample by looking at their distribution (number of institutions and total assets) across different regions and country income groups. For this purpose, each development bank was assigned to a region and income level group based on the country of its res- idence. The assignment was straightforward in the case of national development banks, which consti- tuted most of our sample. However, it was less clear in the case of multilateral development banks because some of them operate globally, whereas others are mostly conned to continental or subcontinental re- gions. At the same time, all multilateral development banks operate in countries with substantial income level differences, so it is difcult to assign them to any single country income group. For these reasons, our approach to the assignment of multilateral develop- ment banks was as follows. Multilateral development banks with a global scope were not included in the re- gional and country income group statistics because it was difcult to allocate their assets to specic regions or country income groups. 3 Nevertheless, they were accounted for in the overall global statistics. Secondly, banks that primarily concentrate their operations within a specic region, such as the AfDB in Sub- Saharan Africa and the EIB in Europe, were included in the regional statistics 4 . We did not include those banks in country income group statistics—AfDB sup- ports development efforts in both low-income and lower-middle-income countries, while EIB operations are similarly not conned to a single income group of countries. Country income-level groups and regional operational scopes of multilateral development banks were based on data from Xu et al. (2021), manually 3 Three such institutions were not included—International Fund for Agricultural Development (IFAD), International Bank for Reconstruction and Develop- ment (IBRD), International Finance Corporation (IFC). 4 The New Development Bank (NDB) with a primary focus on BRICS countries was not geographically attributed to a specic region. ECONOMIC AND BUSINESS REVIEW 2024;26:168–183 173 Table 1. Regional and country income group distribution of development banks. DB sector assets Difference in relative to banking Average DB total assets to Number of obs. Number of banks system assets asset size other bank types (1) (2) (3) (4) (5) (6) (7) (8) (9) N % N % % mil USD % b p Total 3617 1.12 233 1.01 4.75 34,746 0.035 22,237 .000 East Asia and Pacic 523 1.98 34 1.59 5.68 93,409 0.278 60,531 .082 Europe and Central Asia 1199 0.96 69 0.66 4.29 41,557 0.094 26,418 .010 Latin America and Caribbean 672 4.38 45 3.01 13.77 15,103 0.575 8,560 .045 Middle East and North Africa 305 5.20 20 4.44 3.33 4,974 0.291 489 .770 North America 106 0.08 6 0.10 1.10 39,801 0.249 37,933 .122 South Asia 233 5.01 17 4.43 6.10 10,158 0.638 3,260 .316 Sub-Saharan Africa 511 7.01 38 5.19 6.87 2,535 0.344 717 .479 High Income 1238 0.46 70 0.42 2.63 43,126 0.057 36,437 .000 Upper Middle Income 1040 3.17 71 2.03 8.63 40,481 0.228 22,318 .191 Lower Middle Income 575 3.76 45 2.98 4.87 5,464 0.217 4,447 .057 Low Income 93 3.57 8 2.53 3.49 496 0.702 331 .389 Note. Column (2) shows the percentage of observations pertaining to development banks (DBs) relative to total number of observations in the sample. Column (4) shows the percentage of the number of DBs relative to the total number of banks in the sample. Column (7) shows the percentage of assets that the average development bank represents in the respective banking system. In column (8) a random-effects model was used to test the mean difference in total assets between development banks and other bank types, addressing the downward bias from clustering of standard errors in standard t tests in panel data setting. The coefcient estimate (b) for the dummy variable for development banks represents the mean difference of total assets between development banks and other bank types in millions of USD. Robust standard errors were used to obtain p values. Multilateral development banks operating worldwide are excluded from statistics on regional and income group distribution but not from statistics globally. Out of 233 development banks operating worldwide, 229 were classied into the regional distribution and 194 were classied into the income level group distribution. veried using information from ofcial websites of the respective institutions. Although national development banks’ activities are mostly focused on domestic operations, several national development banks in higher-income coun- tries also conduct cross-border activities. Similarly, multilateral development banks, predominantly op- erating in higher-income countries, conduct activities in lower-income countries. The Fitch Connect and Xu et al.’s (2021) databases do not include detailed in- formation about the regional and income-level group allocations of individual development banks’ lend- ing portfolios, which would allow us to paint a more granular picture of the regional and developmental focus of development banks. Our analysis therefore reects singular regional and income-level group as- signment of development banks as institutions, which may differ from the regional and income-level group allocation of their assets. To some extent, we at- tempted to mitigate this by excluding banks operating across multiple income-level groups (in particular, multilateral development banks with a global scope); however, we recognize that these effects persist in the data. Our results are presented in Table 1. Approximately one third of development banks are located in Eu- rope and Central Asia, yet these banks represent a relatively modest 4.3% of the region’s banking system assets. The average development bank in this region represents less than 0.1% of regional banking system assets, which is the smallest proportion among all regions. Nevertheless, a mean difference test revealed that, on average, development banks in this region are signicantly larger than other bank types. In fact, the average development bank is more than four times larger. The development banking sector holds the largest proportion of total banking system assets in Latin America and the Caribbean (13.2% of the region’s banking system assets). In terms of absolute size, the average development bank in this region is signi- cantly smaller compared to their counterparts in East Asia and Pacic (approximately six times smaller), Europe and Central Asia (approximately three times smaller), or North America (approximately three times smaller). However, in terms of relative impor- tance of individual institutions within the region’s banking system, the average development bank in the region represents almost 0.6% of the total banking sys- tem assets. Development banks in this region exhibit larger asset size on average compared to other bank types, with a statistically signicant difference at the 5% level. The development banking sector in North America accounts for the smallest proportion of the regional banking system assets at only 1.1% of the region’s total. Only 6 development banks from our sam- ple (not accounting for MDBs operating worldwide) are located in and focus their operations on this region. 174 ECONOMIC AND BUSINESS REVIEW 2024;26:168–183 Table 2. Distribution of total assets of development banking sector by mandates. Broad % HOUS % AGRI % EXIM % INFRA % INT % LGOV % SME % Total 74.48 4.41 5.51 6.86 0.89 2.20 1.76 3.90 East Asia and Pacic 78.24 3.19 6.88 10.29 0.29 0.00 0.01 1.11 Europe and Central Asia 80.54 0.00 4.71 3.32 0.00 2.56 3.26 5.61 Latin America and Caribbean 73.65 6.22 1.06 4.60 0.44 0.42 5.50 8.11 Middle East and North Africa 62.25 14.44 0.00 15.20 0.00 0.00 0.57 7.54 North America 0.75 69.29 0.00 22.21 0.00 0.00 0.00 7.74 South Asia 0.48 6.90 35.29 10.07 36.83 0.00 0.00 10.44 Sub-Saharan Africa 83.40 0.21 6.37 0.00 3.75 4.79 0.00 1.48 High Income 72.17 7.14 4.40 6.27 0.00 0.65 2.96 6.41 Upper Middle Income 73.30 3.61 8.40 10.53 0.11 0.00 1.44 2.61 Lower Middle Income 13.25 5.22 26.69 16.02 27.81 0.00 0.28 10.73 Low Income 77.55 0.00 22.45 0.00 0.00 0.00 0.00 0.00 Note. BroadD broad mandate; HOUSD social housing; AGRID agriculture and rural development; EXIMD export and import, foreign trade; INFRAD infrastructure; INTD international nancing of private sector development; LGOV D local government; SMED small and medium enterprises. Multilateral development banks operating worldwide are excluded from statistics on regional and income group distribution but not from statistics globally. The global distribution of development banks based on country income levels (data for country in- come levels comes from the WDI database) reveals that the majority of development banks (141 out of 233 total and 194 of those that can be attributed to specic country income groups) are situated in high-income and upper-middle-income countries, consistent with the ndings of Xu et al. (2021). Conversely, a relatively small number of banks and observations originate from low-income countries. Furthermore, the de- velopment banking sector as a whole represents a substantial 8.63% of the upper-middle-income coun- tries’ banking system assets, which is nearly twice the proportion observed in the lower-middle-income countries and more than double compared to low- income countries. Contrary to what one might expect, these results suggest that development banking is a more prominent component of the banking sys- tem in wealthier nations. This is in line with the critique presented by Kharas et al. (2014) regarding the “overnancing” of more developed and the “un- dernancing” of less developed countries. However, it is important to note that the relative size of the average development bank (as a share of banking sys- tem assets within income level groups) is higher in low-income countries than in upper-middle or upper- income countries. In terms of the absolute amount of total assets, the average development bank size increases across income level groups, from low to high income. Specif- ically, the average development bank in high-income countries is almost 90 times larger than its counterpart in low-income countries 5 . This pattern reveals the re- stricted scale of operation due to limited resources for national development banks operating in economies with lower levels of income, which underscores the importance of MDBs (both regional and those with a global scope) supporting development efforts in less developed countries. Interestingly, the mean difference test indicates no statistically signicant disparities between the aver- age size of development banks and other bank types in either low-income, lower-middle-income or upper- middle-income countries (the difference is close to being signicant in the lower-middle-income group). However, in high-income countries, development banks are on average signicantly larger compared to other bank types. The average development bank in high-income countries is approximately six times larger relative to other banks. 3.2 Distribution of development banks across mandates Distinct mandates are a dening characteristic of development banking, warranting our further atten- tion. Mandates can either be broad or specic (in the sense of targeting individual sectors). Table 2 provides additional insights into the distribution of development banking sector assets across different mandates within regional and country income groups during the sample period. Globally, three quarters of the development bank- ing sector assets are held by banks with broad mandates focusing on general socio-economic de- velopment. Asset concentration in broad mandates is observed in most regions, with the exception 5 While development banks primarily operating in developed countries tend to be larger than those in developing countries, it is crucial to recognize that the former can also function with the aim of supporting developing nations. Despite the exclusion of globally operating MDBs from this analysis, residual spillover effects persist in the data, as elaborated in the initial two paragraphs of Section 3.1. Unfortunately, our dataset constrains the examination of such effects. ECONOMIC AND BUSINESS REVIEW 2024;26:168–183 175 Table 3. Distribution of development banks by ownership level and geographic scope. National development banks Multilateral development banks Total Subnational focus International focus National focus Total Regional MDB Global MDB Banks (N) 194 25 42 127 39 34 5 Assets (%) 79.19 5.79 49.63 23.77 20.81 15.67 5.14 Note. The row “Assets (%)” represents the proportion of global development banking assets covered by banks corresponding to specic operational scopes. Columns “Total” represent the asset coverage and number of banks with respect to their level of ownership. of North America and South Asia. North America stands out with a majority of assets concentrated in banks that primarily focus on social housing and in- ternational trade promotion (export–import banks). The non-standard asset distribution in this region is also attributable to a small number of development banks operating in this region (only 6 in our sample). South Asia is another region that exhibits a distinct asset concentration pattern, with the majority of as- sets held by banks with narrow mandates focusing on agriculture, rural development, and infrastructure, particularly in the electrication and power sectors. Looking at the distribution pattern of develop- ment banking assets across mandates within income level groups, a distinct difference can be observed between high-income and upper-middle-income as well as low-income countries on the one hand, where approximately three quarters of assets are associ- ated with broad mandates, and lower-middle-income countries, where the asset share in broad mandates is much lower (13.25 percent). In low-income countries, national development banks with broad mandates such as the Development Bank of Ethiopia, Devel- opment Bank of Rwanda, and Uganda Development Bank represent a substantial proportion of develop- ment banking assets in that income group. However, those banks are relatively small in terms of asset size, which leaves their nancial capacity to inuence de- velopment outcomes comparatively limited. 3.3 Distribution of development banks according to ownership levels and geographic scope of operations Delving into the nuances of development banks’ ownership levels and operational focus, Table 3 pro- vides an overview of the distribution of development banks concerning ownership level and operational focus. In terms of ownership level, we distinguish be- tween national and multilateral development banks. It is noteworthy that the majority of development banking assets, amounting to 79.19%, is concentrated in nationally owned development banks, with the re- maining 20.81% represented by MDBs. Data further reveals that approximately half of global development banking assets are attributed to nationally owned development banks engaged in international operations. Illustrative examples of such institutions include the KfW, the Development Bank of Japan, and several trade-promoting export– import-oriented development banks. In contrast, sub- nationally focused development banks constitute a relatively modest 5.79% of the total assets in the global development banking sector. Notable instances of such banks include the German NRW Bank, which specically operates within the German state of North Rhine–Westphalia (Nordrhein-Westfalen), and the Brazilian Banco do Nordeste do Brasil (BNB), which primarily concentrates its activities within nine states of the Northeast region of Brazil. 3.4 Global and regional concentration of development banking assets The eld of development banking, as a distinct sec- tor within the banking industry, exhibits a notable degree of concentration, with certain banks having signicant inuence, especially at the regional level. To assess the extent of market concentration in the development banking sector, we computed two key concentration ratios: the 3-bank asset concentration ratio (CR 3 ) and the 5-bank asset concentration ratio (CR 5 ) within regional country groups. It is important to understand that by dening ratios in such a man- ner, CR 3 and CR 5 represent only the concentration of assets within the regional development banking sector and are thus not representative of the bank- ing system as a whole. Additionally, to provide a deeper understanding of the signicance of individ- ual institutions, we graphically ranked the ten largest development banks within each region based on their market shares, determined by their total assets. Fig. 2a and Fig. 2b present global and regional mar- ket shares of the ten largest development banks based on total assets. The asset proportions were calculated from average annual assets of individual develop- ment banks across the sample period from 1995 to 2021. We used these shares to calculate the CR 3 and CR 5 concentration ratios. The upper left panel of Fig. 2a depicts the largest development banks globally. By far the largest in terms of total assets is the China Development Bank, which covers 17.1% of global development banking 176 ECONOMIC AND BUSINESS REVIEW 2024;26:168–183 Fig. 2a. Development bank asset concentration globally and by region (% of region’s development banking assets). assets. It holds nearly as much assets as the sec- ond (European Investment Bank—6.9%), third (Japan Housing Finance Agency—6.6%), and fourth (KfW— 6.6%) largest banks combined. In the remainder of this section, we delve into the signicance of the largest banks and their operations across macro regions. Notably, the development banking sector in Europe and Central Asia exhibits the lowest level of asset concentration, with a CR 3 ratio of 53.1%. The three systemically important banks in this region are the EIB (second largest development bank in the world), German KfW (fourth largest in the world), and Ital- ian Cassa Depositi e Prestiti (CDP S.p.A), which is the fth largest development bank worldwide. These institutions, along with the French Caisse des Dépots, are the founding members of the Long-Term Investors Club (LTIC). The primary objective of the LTIC is to advance collaboration among leading development nancing institutions worldwide and foster devel- opment in both emerging and developed countries by providing sustainable long-term and preventing speculative short-term nancing. The relatively mod- est CR 5 ratio of 66.7% in the region in comparison with other regions further indicates a diverse and dispersed development banking sector. However, it is noteworthy that four out of the ten largest devel- opment banks within the region operate primarily in Germany with either national or subnational opera- tional focus. ECONOMIC AND BUSINESS REVIEW 2024;26:168–183 177 Fig. 2b. Development bank asset concentration by region (% of region’s development banking assets). Similarly, the Middle East and North Africa re- gion experiences medium development banking asset concentration based on CR 3 and CR 5 ratios, which amount to 59.6% and 78.3%, respectively. In this re- gion, Bank Maskan (Housing Bank) holds the position of the largest development bank and the sole provider of residential mortgages in Iran. However, it is worth noting that this is not solely due to a market gap but also due to limitations in offering mortgage ser- vices imposed on commercial banks (Gholipour et al., 2020). The second largest development bank in the region is the Islamic Development Bank (IsDB), which operates as an MDB with a focus on Islamic nance. With ownership shared among 57 member coun- tries, the IsDB carries a broad mandate encompassing general socio-economic development objectives. The third largest in the region is the Kuwait Fund for Arab Economic Development (KFAED). Turki (2014) stresses the signicance of KFAED’s activities in fos- tering not only Arab economic development but also Arab relations with non-Arab countries. These three banks can be considered systemically important de- velopment banks within the region, with roles in promoting economic development and regional coop- eration. In the Latin America and Caribbean region, the 3- and 5-bank asset concentration ratios are also within the middle range (at 64.5% and 74.5%, respectively). With respect to individual banks, the Brazlian De- velopment Bank (BNDES) holds a substantial share 178 ECONOMIC AND BUSINESS REVIEW 2024;26:168–183 of almost 32% in the region’s development banking sector. In Brazil, the systemic importance of BNDES is comparable to the combined signicance of the Instituto del Fondo Nacional de la Vivienda para los Trabajadores (INFONAVIT), Banco Nacional de Obras y Servicios Públicos (BANOBRAS), and Na- cional Financiera (Nan) in Mexico. INFONAVIT serves as the largest lender for social housing projects in Latin America; BANOBRAS focuses on nanc- ing local governments and providing project nance, while Nan supports small and medium-sized enter- prises. As the second most systemically important devel- opment bank in the region, the multilateral Inter- American Development Bank (IADB) covers the en- tire region and operates with a broader mandate of promoting socio-economic development. The IADB holds 17.3% of the region’s development banking as- sets. The region’s substantial share of development banking assets in the broader banking system assets and diversity of mandates coupled with the fact that the average development bank in this region is larger than other bank types indicate the signicant impor- tance and diversity of development banking in this region. The East Asia and Pacic region is also character- ized by a high level of concentration of development banking assets in the largest development institu- tion, namely the China Development Bank (CDB). As the largest development bank in the world, CDB covers nearly 40% of the region’s development bank- ing sector assets, which is approximately equivalent to the combined assets of the next ve banks. CDB provides nancial support for infrastructure projects, industrial development, and helps implement gov- ernment policies. The second Chinese policy bank, the Agricultural Development Bank of China (ADBC), contributes to the modernization, efciency, and sus- tainability of China’s agricultural sector through the provision of loans and expertise, while the third pol- icy bank, the Export–Import Bank of China, promotes foreign trade and investment. These three Chinese policy banks account for more than 50% of the re- gion’s development banking sector assets. It is worth noting that Chinese development nance extends beyond development or policy banks. In that respect, Chen (2020) notes that although they can- not be considered development banks, China’s major commercial banks, such as the Industrial and Com- mercial Bank of China, the China Construction Bank, and the Bank of China, also play a signicant role in providing development nance. Chin and Gallagher (2019) further recognize the role of these commercial banks in conancing projects with the three policy banks. Covering the entire Asia–Pacic region, the mul- tilateral Asian Development Bank (ADB) offers its member countries a range of nancial instruments, including loans, grants, technical assistance, and capacity-building support for various development projects. The ADB operates with a broad mandate and engages in conancing arrangements with other de- velopment institutions and the private sector, which helps mobilize additional capital and mitigate project risks. Primarily nancing development in Sub-Saharan Africa, the African Development Bank (AfDB) covers almost 45% of region’s development banking sector, which is more than the next nine banks combined. It is important to note that Sub-Saharan Africa faces a limited presence of development banking institutions originating from within the region itself. In the South Asian region, the ve most impor- tant development banks in terms of asset coverage operate in India. REC Limited specializes in nanc- ing rural electrication projects in India. The Power Finance Corporation also provides nancial support to companies in the power sector, including thermal power plants, hydroelectric projects, renewable en- ergy installations, transmission networks, and other related infrastructure. Both institutions offer loans, nancial assistance, and advisory services to pro- mote the development of India’s power sector, with a particular emphasis on sustainable and inclusive electrication. Another notable bank in this region is the National Bank for Agriculture and Rural Devel- opment (NABARD) in India. NABARD focuses on providing nancial support and resources for agricul- ture and rural development initiatives. Collectively, these three development nancing institutions ac- count for 75.4% of the total development banking assets in the region. The Export–Import Bank of In- dia and the Small Industries Development Bank of India also contribute signicantly, with the ve banks together representing over 90% of the region’s devel- opment banking assets. High concentration of assets in these ve Indian development banks indicates a limited presence of development banking in other South Asian countries, underscoring the need for fur- ther development and diversication in the region. 3.5 Composite global and national systemic nancial importance of development banks The analysis conducted thus far revealed signicant disparities in the asset size of development banks, both in absolute terms and as a share of global and regional banking and development banking systems. However, the nancial importance of development banks cannot be adequately captured by a single ECONOMIC AND BUSINESS REVIEW 2024;26:168–183 179 Table 4. DBFI index creation—PCA results. Eigenvalue Explained Loadings Unexplained Comp l % Variables on Comp1 % Observations Development banks (DFBI) Comp1 2.06 0.69 Total assets 0.49 0.51 3549 Comp2 0.68 0.23 Regional importance— banking system 0.61 0.23 Comp3 0.27 0.08 Regional importance— DB sector 0.62 0.20 National Development Banks (extended DFBI) Comp1 2.27 0.57 Total assets 0.47 0.49 2371 Comp2 0.86 0.21 Regional importance— banking system 0.58 0.23 Comp3 0.61 0.15 Regional importance— DB sector 0.58 0.24 Comp4 0.25 0.06 Country importance— banking system 0.33 0.76 Note. KMO measure of sampling adequacy is higher than .5 for all groups. metric, but rather through consideration of multi- ple metrics concurrently. To achieve this objective, we created a composite Development Bank Finan- cial Importance (DBFI) index by combining three metrics related to the absolute and relative size of development banks using principal component anal- ysis (PCA) for this purpose. The values of this index are derived from scores of the rst principal com- ponent, calculated through a linear combination of the original standardized three variables. The weights assigned to each variable in this combination are determined by the loadings on the rst principal component. The key aim is to create a single linear combination of the original variables while minimiz- ing any loss of information. PCA is a frequently employed method when re- searchers are confronted with data reduction chal- lenges, pioneered by Hotelling (1933). 6 Jan et al. (2019) utilize bank-specic KPIs such as ROAA, ROAE, and Tobin’s Q as measures of nancial performance from the perspectives of management, shareholders, and markets, respectively. To capture nancial per- formance comprehensively, they derived PCA scores from ROAA, ROAE, and Tobin’s Q, and created the Islamic Financial Index. Likewise, Shi and Yu (2021) applied PCA to construct an index that mea- sures Chinese banks’ risk management, drawing from bank-specic KPIs. This allowed the authors to avoid the arbitrary assignment of weights to individual in- dicators. Our rst metric to enter the PCA analysis was absolute development bank size, measured by total assets, signaling the global size ranking of each de- velopment bank. The second one was the share of a development bank’s total assets in the total assets of the regional development banking sector, conveying the bank’s importance among other regional devel- opment institutions. The last metric was the share of a development bank’s total assets in the total assets of the regional banking system, measuring the bank’s overall signicance in relation to all banks operating in that region. Multilateral development banks that could not be accurately or predominantly attributed to a specic region (i.e. those with a global scope) were excluded from this analysis. The scores obtained from the rst principal component serve as a proxy index of the weighted average global and regional nancial importance of development banks (DBFI index). Furthermore, we also considered an extended in- dex of the global, regional, and national nancial importance of national development banks (extended DBFI index). For this purpose, we incorporated an additional metric of relative bank size into the PCA, measuring the asset share of a particular development bank within its country’s banking system. This metric adds a perspective on the overall national signicance of a particular development bank. The PCA is re- peated on the extended set of four variables. Notably, all multilateral development banks are excluded from this particular analysis as their scope of operations cannot be attributed to a single country. Table 4 displays the PCA results. The left side of the table, containing columns Comp, Eigenvalue, and Ex- plained, shows analytical results for all components derived from the original variables, separately for the DFBI and the extended DFBI indices. The Eigenvalue column denotes the eigenvalues of individual com- ponents, which are relevant in chosing the number of 6 See also A et al. (2012). This method is widely applied across numerous disciplines—Lindman and Sellin (2011) criticize the methodology and arbitrary weighting employed in the construction of the Human Development Index (HDI), which incorporates life expectancy, education, and GNI per capita. They utilize PCAscores to create an alternative composite welfare metric that better captures the complexity of environmental issues. PCAscores are also employed by Bergenfeld et al. (2021) to construct a gender equity index for secondary schools, and by Vyas and Kumaranayake (2006) to develop a composite socio-economic household status index using household-specic data. Similarly, Lamichhane et al. (2021) condense 17 indicators of sustainable development goals (SDGs) for OECD countries into a composite sustainability index score, providing a robust alternative to standard United Nations benchmarking tools. 180 ECONOMIC AND BUSINESS REVIEW 2024;26:168–183 components to keep. Kaiser’s rule suggests keeping only principal components with eigenvalues greater than one (Guttman, 1954; Kaiser, 1960). The Explained column represents the proportion of variance in origi- nal variables explained by each principal component. Total variance explained by all components sums to the total variance in the original variables. The right side of Table 4, containing columns Variables, Loadings on Comp1, Unexplained, and Ob- servations, only shows the analytical results for the rst principal component, which is used to construct the DFBI and extended DFBI indices. The column Loadings on Comp1 is particularly relevant. The load- ings are essentially weights used to construct the rst principal component scores as a weighted linear combination of standardized original variables. The Unexplained column indicates the proportion of vari- ation in the original variable not accounted for by the rst principal component. In the rst sample, comprising all development banks except multilateral development banks with a global focus, all three bank size metrics of nancial importance show positive and moderate to strong loadings on the rst principal component (Comp1). The scores attributed to this component serve as the DBFI index. The rationale for exclusively uti- lizing scores from the rst principal component is supported by two distinct perspectives within the literature: rstly, adhering to Kaiser’s rule, which ad- vocates retaining only those principal components with eigenvalues surpassing one (Guttman, 1954; Kaiser, 1960); and secondly, aligning with the pro- portion of variation captured by Comp1, which is comparable to the proportion explicated by the nan- cial performance index formulated by Jan et al. (2019). In the second sample, which includes only national development banks, the rst three metrics of nan- cial importance, which capture global and regional importance, exhibit positive and moderate loadings on the rst component, whereas the last metric, which measures national importance, exhibits a weaker pos- itive correlation with the rst component. The scores of the rst principal component (Comp1), which has an eigenvalue above one, serve as the extended DBFI index for national development banks, considering their global, regional, and (to a lesser extent) national nancial importance within the banking system. It is important to acknowledge that these indicators do not provide a comprehensive measure of a devel- opment bank’s overall signicance, as they do not account for resource allocation and the impact on de- velopment outcomes. Instead, these indicators serve as a strictly nancial metric of importance, high- lighting development banks that possess the greatest nancial capacity to support and facilitate develop- ment initiatives. Fig. 3 presents the results of the weighted nancial importance of development banks based on the aver- age DBFI index during the sample period. The upper panel shows the results based on the DFBI index for all development banks, whereas the lower panel shows the results based on the extended DFBI index for national development banks. As can be seen from the upper panel, CDB emerges as the most nancially signicant development bank overall, representing over 17% of global development banking assets and nearly 40% of the region’s development banking as- sets in East Asia and the Pacic. BNDES holds the second position, despite being ranked 10th in terms of average total assets (please see the upper-left panel in Fig. 2a). Its signicance within the development banking sector of Latin America and the Caribbean, as well as its size relative to the regional banking system, solidies its nancial importance. Similarly, AfDB, although ranking only as the 31st largest development bank in the sample, demon- strates signicant regional nancial importance within Sub-Saharan Africa’s development banking sector and the region’s broader banking system, which positions it third globally. It is important to note that the exclusion of multilateral development banks operating worldwide prevents the inclusion of the IBRD among the rankings in Fig. 3. This exclusion arises from the challenge of accurately assigning their assets to a specic region and, consequently, the inability to compute their regional importance 7 . IBRD is the sixth largest development bank and one of the most signicant ones globally. Interestingly, two of the globally largest multilateral and national development banks (the EIB and the KfW), while being important, only rank 8th and 9th using our composite index of nancial importance. Considering our extended composite index of - nancial importance on the subset of national de- velopment banks, BNDES and CDB retain their positions at the top (lower panel), further empha- sizing their nancial signicance. The Canada Mort- gage and Housing Corporation (CMHC) 8 and the Japan Housing Finance Agency (JHC) follow as the third and fourth most nancially important national 7 For a detailed explanation, please refer to the rst paragraph of Section 3.1. Three such institutions are not included—International Fund for Agricultural Development (IFAD), International Bank for Reconstruction and Development (IBRD), International Finance Corporation (IFC). 8 Originally established in 1946 with the purpose of assisting war veterans in acquiring housing, CMHC has since expanded its mandate to encompass a broader scope of social housing initiatives. ECONOMIC AND BUSINESS REVIEW 2024;26:168–183 181 Fig. 3. Composite nancial importance of development banks. Note. Horizontal axis in the upper panel represents the DBFI index. Horizontal axis in the lower panel represents the extended DBFI index. development banks, serving as the primary lenders for social housing projects and offering loan guaran- tees and housing market research. Among the top 10 nancially important develop- ment banks globally and regionally, six banks have broad mandates, while three have specic mandates related to social housing. On the national level, ve of the top 10 nancially important banks have broad mandates, while three focus on social hous- ing. These ndings demonstrate the notable nancial representation of social housing mandates within the development banking sector. 4 Concluding remarks In this paper, we conducted an analysis of the nancial importance of global banking sectors. Rec- ognizing existing studies on the lending activities and nancial performance of individual development banks, as well as global surveys focusing on develop- ment banks, we identied a knowledge gap regarding a multiannual global analysis of the nancial impor- tance of development banks. To address this gap, we compiled a global banking dataset comprising 22,949 individual banks, including 233 categorized as de- velopment banks, spanning the period from 1995 to 2021. This paper presents the rst comparative study not only of individual development banks but also of entire development banking sectors. Our research aims to address the following questions: How do development banking sectors differ in their nan- cial importance across regions, income groups, and mandates? How can we measure and compare the nancial importance of development banks at the global and national level? Regarding our rst research question, the analysis reveals signicant diversity in development bank- ing sectors across regions and income groups. The majority of development banks are concentrated in the Europe and Central Asia, East Asia and the Pa- cic, and Latin America and the Caribbean regions. Notably, development banks in Europe and Central 182 ECONOMIC AND BUSINESS REVIEW 2024;26:168–183 Asia, as well as Latin America and the Caribbean, exhibit substantially larger asset sizes compared to conventional banks. In the latter region, the devel- opment banking sector constitutes nearly 14% of the total banking system assets, which underscores its considerable signicance. In terms of income group distribution, the majority of development banks, both in terms of numbers and assets, operate in high- income or upper-middle-income countries, indicating a notable underrepresentation in lower-middle- and low-income countries. Regarding mandates, our nd- ings indicate that approximately three-quarters of global development banking assets originate from in- stitutions with broad mandates, rather than narrow ones. Addressing our second research question, we uti- lized PCA to construct the novel Development Bank Financial Importance (DBFI) index. The data included both absolute and relative metrics of development bank size, considering the representation of a bank’s assets in national and regional banking and develop- ment banking sectors. In our multidimensional index of nancial importance, the Chinese CDB and Brazil- ian BNDES emerge as the two nancially most im- portant national development banks. Among MDBs (excluding those with a global scope), African AfDB ranks the highest due to its regional signicance, de- spite being substantially smaller than many of its global peers. In the global and regional context, six out of the top 10 nancially signicant development banks have broad mandates, with three specically dedicated to social housing. At the national level, ve of the top 10 nancially important banks operate with broad mandates, while three emphasize social housing. These results underscore the considerable nancial inuence of social housing mandates within the development banking sectors. Our results have important policy implications. The uncovered underrepresentation of development banks in lower income economies indicates substan- tial untapped development potential. This under- scores a growing divide between countries which can afford sustainable development and those that can- not. In the light of rapid ongoing digital and green technological transformations, this divide could have adverse and lasting effects on the long-term prosper- ity of lower-income economies and the fulllment of Sustainable Development Goals. One possible solu- tion to level the playing eld is to further concentrate the resources of MDBs on sustainable development of lower-income economies. Another possibility is to consider the cross-border activities of national devel- opment banks, as our dataset indicates that almost half of the global development banking sector’s as- sets belong to national development banks with an international focus. However, the downside of us- ing national development banks for the development of lower-income economies is that their activities are primarily aligned with national (as opposed to international) policy considerations, raising fears of neocolonialism. While this paper is informative about the global development banking landscape, it does not seek to answer all questions. An important limitation of our dataset is that it does not allow us to disentangle cross-border assets of national and multilateral de- velopment banks by region and income level group. Such information would be welcome for a more granular assesment of their geographic and devel- opmental impacts. Our analysis is also limited to studying the structural characteristics of the global development banking sector without looking at the relationships between these characteristics and vari- ous parameters of development banks’ perfomance. Based on our understanding of contemporary de- velopment banking, at least three such parameters might be of interest. The rst two parameters are developmental and countercyclical perfomance, with the former being a classical dening feature of de- velopment banks’ mandates and the latter gaining prominence recently in the light of several crises. The third parameter is development banks’ nancial perfomance, which should be considered as a nec- essary precondition and not as a primary objective for fullling development bank mandates. Both the construction of more granular datasets and the study of relationships between development bank charac- teristics and their performance offer new promising avenues for research. Funding statement This work was supported by the research pro- gramme P5-0161 and the research project V5-2261 funded by the Slovenian Research and Innovation Agency. References A, A. A., May, S., & Clark, V . A. (2012). Practical multivariate analysis (5th ed.). CRC Press. Barth, M. E., Landsman, W. R., Lang, M., & Williams, C. (2012). Are IFRS-based and US GAAP-based accounting amounts compa- rable? Journal of Accounting and Economics, 54(1), 68–93. Bergenfeld, I., Jackson, E. C., & Yount, K. M. (2021). Creation of the gender-equitable school index for secondary schools using principal components analysis. International Health, 13(2), 205– 207. Beuren, I. M., Hein, N., & Carlos Klann, R. (2008). Impact of the IFRS and US-GAAP on economic-nancial indicators. Manage- rial Auditing Journal, 23(7), 632–649. Brei, M., & Schclarek Curutchet, A. (2017). The countercyclical behav- ior of national development banks in Latin America and the Caribbean. ECONOMIC AND BUSINESS REVIEW 2024;26:168–183 183 Repositorio digital universitario. Universidad Nacional de Córdoba. Buiter, W. H., & Fries, S. M. (2002). What should the multilateral development banks do? (Working paper No. 74). European Bank for Reconstruction and Development. Chandrasekhar, C. P . (2016). National development banks in a comparative perspective. In A. Calcagno, S. Dullien, A. Márquez-Velázquez, N. Maystre, & J. Priewe (Eds.), Rethink- ing development strategies after the nancial crisis—Vol II: Country studies and international comparisons (pp. 21–30). United Nations Conference on Trade and Development; Hochschule für Tech- nik und Wirtschaft Berlin. Chen, M. (2020). Beyond donation: China’s policy banks and the reshaping of development nance. Studies in Comparative Inter- national Development, 55(4), 436–459. Chern, C. (2019). Chern on Dispute Boards (4th ed.). Routledge. Chin, G. T., & Gallagher, K. P . (2019). Coordinated credit spaces: The globalization of Chinese development nance. Development and Change, 50(1), 245–274. Culpeper, R. (2012). Financial sector policy and development in the wake of the global crisis: The role of national development banks. Third World Quarterly, 33(3), 383–403. De Aghion, B. A. (1999). Development banking. Journal of Develop- ment Economics, 58(1), 83–100. Duprey, T., & Lé, M. (2016). Bankscope dataset: Getting started. SSRN. Engen, L., & Prizzon, A. (2018). A guide to multilateral development banks. Overseas Development Institute. Faure, R., Prizzon, A., & Rogerson, A. (2015). Multilateral develop- ment banks. Overseas Development Institute. Feil, F., & Feijó, C. (2021). Development banks as an arm of economic policy—Promoting sustainable structural change. In- ternational Journal of Political Economy, 50(1), 44–59. Fernández-Arias, E., Hausmann, R., & Panizza, U. (2020). Smart development banks. Journal of Industry, Competition and Trade, 20(2), 395–420. Financial Stability Board. (2022). Global monitoring report on non-bank nancial intermediation . https://www.fsb.org/wp-content/ uploads/P201222.pdf Frigerio, M., & Vandone, D. (2020). European development banks and the political cycle. European Journal of Political Economy, 62, Article 101852. Galindo, A. J., & Panizza, U. (2018). The cyclicality of interna- tional public sector borrowing in developing countries: Does the lender matter? World Development, 112, 119–135. Geddes, A., Schmidt, T. S., & Steffen, B. (2018). The multiple roles of state investment banks in low-carbon energy nance: An analysis of Australia, the UK and Germany. Energy Policy, 115, 158–170. Gerschenkron, A. (1962). Economic backwardness in historical perspec- tive: A book of essays. The Belknap Press of Harvard University Press. Gholipour, H. F., Oikarinen, E., & Tajaddini, R. (2020). Banks’ lend- ing to public and private sectors and house prices: Does bank ownership matter? International Journal of Housing Markets and Analysis, 13(2), 227–249. Gong, D., Xu, J., & Yan, J. (2023). National development banks and loan contract terms: Evidence from syndicated loans. Journal of International Money and Finance, 130, Article 102763. Grifth-Jones, S., & Ocampo, J. A. (Eds.). (2018). The future of national development banks. Oxford University Press. Gutierrez, E., Rudolph, H. P ., Homa, T., & Beneit, E. B. (2011). Development banks: Role and mechanisms to increase their efciency (Policy Research Working Paper 5729). World Bank. Guttman, L. (1954). Some necessary conditions for common-factor analysis. Psychometrika, 19(2), 149–161. Hotelling, H. (1933). Analysis of a complex of statistical variables into principal components. Journal of Educational Psychology, 24(6), 417–441. Jan, A., Marimuthu, M., bin Mohd, M. P ., & Isa, M. (2019). The nexus of sustainability practices and nancial performance: From the perspective of Islamic banking. Journal of Cleaner Production, 228, 703–717. Kaiser, H. F. (1960). The application of electronic computers to factor analysis. Educational and Psychological Measurement, 20(1), 141–151. Kellerman, M. (2019). The proliferation of multilateral devel- opment banks. The Review of International Organizations, 14, 107–145. Kharas, H., Prizzon, A., & Rogerson, A. (2014). Financing the post-2015 sustainable development goals. Overseas Development Institute. Lamichhane, S., E˘ gilmez, G., Gedik, R., Bhutta, M. K. S., & Erenay, B. (2021). Benchmarking OECD countries’ sustainable development performance: A goal-specic principal compo- nent analysis approach. Journal of Cleaner Production, 287, Article 125040. Lazzarini, S. G., & Musacchio, A. (2010). Leviathan as a minority shareholder: A study of equity purchases by the Brazilian National Development Bank (BNDES), 1995-2003. SSRN. Lindman, C., & Sellin, J. (2011). Measuring human development: the use of principal component analysis in creating an environmental index [Bachelor thesis, Uppsala University]. Luna-Martínez, J., & Vicente, C. L. (2012). Global survey of develop- ment banks (Policy Research Working Paper 5969). World Bank. Mazzucato, M., & Penna, C. (2015). The rise of mission-oriented state investment banks: The cases of Germany’s KfW and Brazil’s BNDES. (SPRU Working Paper Series 2015–26). University of Sussex. Micco, A., Panizza, U., & Yanez, M. (2007). Bank ownership and performance. Does politics matter? Journal of Banking & Finance, 31(1), 219–241. Musacchio, A., Lazzarini, S. G., Makhoul, P ., & Simmons, E. (2017). The role and impact of development banks [Working paper]. World Bank. Pissarides, F. (1999). Is lack of funds the main obstacle to growth? EBRD’s experience with small- and medium-sized businesses in Central and Eastern Europe. Journal of Business Venturing, 14(5– 6), 519–539. Prizzon, A., Humphrey, C., Kaul, I., Kodera, K., McKechnie, A., & Rogerson, A. (2017). Six recommendations for reforming multilat- eral development banks: An essay series. ETH Zurich. Rudolph, H. P . (2009). State nancial institutions: mandates, gover- nance, and beyond (Policy Research Working Paper 5141). World Bank. Scott, D. H. (2012). Strengthening the governance and performance of state-owned nancial institutions (Policy Research Working Paper 4321). World Bank. Shi, X., & Yu, W. (2021). Analysis of Chinese commercial banks’ risk management efciency based on the PCA-DEA approach. Mathematical Problems in Engineering, 2021, Article 7306322. Turki, B. (2014). The Kuwait fund for Arab economic development and its activities in African countries, 1961–2010. The Middle East Journal, 68(3), 421–435. Vyas, S., & Kumaranayake, L. (2006). Constructing socio-economic status indices: How to use principal components analysis. Health Policy and Planning, 21(6), 459–468. Xu, J., Marodon, R., Ru, X., Ren, X., & Wu, X. (2021). What are public development banks and development nancing institutions?— Qualication criteria, stylized facts and development trends. China Economic Quarterly International, 1(4), 271–294. Yeyati, E. L., Micco, A., & Panizza, U. (2004). Should the government be in the banking business? The role of state-owned and develop- ment banks (Working Paper #517). Inter-American Development Bank. Zhang, F. (2022). The policy coordinator role of national devel- opment banks in scaling climate nance: Evidence from the renewable energy sector. Climate Policy, 22(6), 754–769.