Overlapping lending by the European Investment Bank and the European Bank for Reconstruction and Development

ABSTRACT

One of the main criticisms of contemporary lending practice by different multilateral public banks in Europe made by scholars and policymakers focuses on overlapping loans, which can create duplications and inefficiencies. However, no database exists on the lending overlaps between these banks. With the aim of analysing these overlaps, all lending by the two major multilateral public banks in Europe, the European Investment Bank and the European Bank for Reconstruction and Development, were merged into a single database. This new database allows researchers to assess lending overlap, regarding its size, evolution, and geography, facilitating an inquiry into the extent to which banks coordinate this overlapping lending.

KEYWORDS:

• Multilateral development banks
• European Investment Bank
• European Bank for Reconstruction and Development
• development
• European integration
• lending overlap

Objectives

One of the main criticisms of contemporary lending operations in Europe made by scholars and policymakers focuses on overlapping loans by different multilateral public banks which – it is often assumed – is an inefficient practice. The overlap may lead to a lack of coordination among banks, which could result in duplication, fragmentation, and inconsistency of lending policies and practices, and it may create potential financial instability, as the overlapping of loans may expose the borrowers and the banks to higher risks of default (Asatryan and Havlik 2020). This criticism, for example, was made of the European Investment Bank (EIB) and the European Bank for Reconstruction and Development (EBRD) in an influential, independent report by the High-level group of wise persons on the European Financial architecture for development (Council of the EU 2019). Despite this high-profile criticism, there is no publicly available dataset for policymakers and researchers to clearly and easily observe, compare and analyse this lending overlap, in order to understand how it is evolving over time and space. Whilst the EIB website provides an open link to its data (EIB Official Database, 1959–2023), as does the EBRD (EBRD Investments, 1991–2022), the databases are not available in a combined form, to assess the criticism mentioned.

To address this, a single dataset was created, merging lending signed by the EIB and the EBRD, with the aim of facilitating analysis by policymakers and scholars in order to better understand its consequences, whilst proposing policy recommendations in the direction of improving the consistency and coordination of the European financial architecture. The combined database facilitates at-a-glance observation of lending overlap, as regards the evolution of the magnitude and geography of these loans. This research is part of an FNR Intermobility-In grant “Realigning the European Financial Architecture for the C21” (REFA) Grant number INTER/MOBILITY/2022/LE/17567353/REFA-21.

Data description

Although literature on lending operations at the EIB and EBRD has been historically rather scarce, in recent years, scholars have increased their interest in this topic (Clifton et al., 2018; Clifton et al., 2021; Howarth and Kavvadia, 2023). Data on EIB and EBRD lending signatures are publicly available on the banks’ respective websites. However, data is only available at the individual level for each bank. Given the critique of lending overlap, it is of interest to publish an open-access database for policymakers and researchers that merges lending operations to facilitate an analysis of the evolution of the magnitude and geography of lending overlaps.

To do so, we present an originally compiled database of European Investment Bank lending signatures (data file 1) constructed using information extracted from the official database of the EIB (EIB Official Database, 1959–2023) and European Bank for Reconstruction and Development lending signatures (data file 2), derived from the EBRD database (EBRD Investments, 1991–2022), covering all 38 countries that received lending from the EBRD from its start of operations in 1991 up to 2021. The data includes all (100%) of European Bank for Reconstruction and Development loan signatures and nearly one quarter (24.7%) of European Investment Bank lending signatures during this period (1991–2021).¹

When constructing the combined database, we grouped borrowing countries into seven, common, categories, by adapting and combining the current classification systems established by the European Union (2023), as regards EU Member States (EU MS), EU Candidate countries, and EU Potential Candidate countries, together with the EBRD’s (2023) categorization of Transition and Non-transition countries. The seven, combined, categories are provided below, in addition to those countries that are included in these categories, which have received both EIB and EBRD lending in the period under study.

1). “EU Transition MS” (Czechia, Estonia, Hungary, Latvia, Lithuania, Poland, Slovak Republic and Slovenia, which joined in 2004; Bulgaria and Romania, in 2007, and Croatia, in 2013); 2). “EU Non-transition MS” (including New MS, Cyprus, and “Existing” MS, Greece; 3). “EU Transition Candidates” (North Macedonia, Montenegro, Serbia, Albania, Bosnia and Herzegovina, Moldova, and Ukraine); 4). “EU Non-transition Candidates” (Turkey); 5). “EU Transition Potential Candidates” (Georgia, Kosovo); 6). “Eastern Transition Countries” (Armenia, Azerbaijan, Belarus, Kazakhstan, Kyrgyz Republic, Mongolia, Russian Federation, Tajikistan, Turkmenistan and Uzbekistan); and 7) “South-East Mediterranean” countries (Egypt, Jordan, Lebanon, Morocco, Tunisia).

The data includes detailed information (classified by sector) of all lending signatures by the EIB (data file 1) to EBRD recipient countries and all lending signatures by the EBRD (data file 2) between 1991 and 2021. Details included are the year the loan was signed, loan description, country, region, sector, and amount. To compare findings in the international context and, to ensure an unbiased rate of inflation, we report all figures in millions 2015 euro. Quantities are updated to constant terms with information from the World Bank (2023) from 1991 to 1996 and from Eurostat (2023) from 1996 to 2021. When compared to the official databases, this new dataset provides a more detailed description that facilitates assessing lending overlap as regards the evolution of its magnitude and geography.

However, loan signatures are not the only indicator of the banks’ lending activity and impact. The overlap of loans between the two banks has both positive and normative dimensions that need to be considered. On the one hand, the overlap may indicate a competitive or cooperative relationship between the banks, which could have implications for the quantity, quality, and diversity of loans, as well as the pricing and the due diligence costs for the borrowers. On the other hand, the overlap may raise some questions about the efficiency, coherence, and additionality of the European financial architecture for development, as well as the challenges and opportunities for the borrowers in dealing with multiple lenders. Therefore, to analyse the overlap in a comprehensive manner, it is necessary to complement the data of this dataset with other sources of information, such as the banks’ statutes, policies, guidelines, and monitoring practices, which can provide insights into the rationale, objectives, and modalities of their lending activities.

Limitations

Although our compilation of information on lending overlap by two of Europe’smain public multilateral banks, the EIB and the EBRD is robust, there are some limitations to the dataset. First, our data was extracted from the official databases of each bank, and a secondary search for further data on additional projects that may have been implemented was not performed since we assumed the official databases are comprehensive. Second, the database uses loan signatures as an indicator of the EIBs’ and the EBRD’s lending activity. However, this measure has some drawbacks, as it does not capture the actual economic impact of the loans, nor the timing and conditions of their disbursements. Furthermore, loan signatures may be subject to strategic behavior by the banks, which may adjust the number and size of signatures to meet their annual targets, regardless of the demand and supply of loans. A more reliable measure of lending activity would be loan disbursements, which reflect the actual transfer of funds to the borrowers and are linked to specific project outcomes. However, data on loan disbursements are not publicly disclosed by both banks, and thus loan signatures are used as a second-best option, with the awareness that they may overstate or understate the real lending volume. Third, the database may not capture the full extent of the overlap of loans between the EIB and the EBRD, due to the differences in the naming and classification of the projects in the two banks’ databases. Moreover, some of the EIB’s loans are provided through Global or Framework loans, which are flexible instruments that finance multiple smaller investments, thenames and details of which are not publicly available. Therefore, it is possible that some of these smaller investments have received funding from both the EIB and the EBRD, without being identified as such in the database. Fourth, the EIB or the EBRD do not have a comprehensive record of the projects that they co-finance with other regional development banks, which makes it difficult to assess the degree and nature of the cooperation between the banks. Finally, our dataset ends in December 2021 to align with the EBRD’s available data.

Acknowledgments

We acknowledge support from the FNR, as well as staff from and experts on the European Investment Bank and the European Bank for Reconstruction and Development who kindly gave their time for interviews between May and December 2023.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data described in this Data Note can be freely and openly accessed on www.econpoint.unican.es/databases. Please see Table 1 for details and links to the data.

Table 1. Overview of data files/data sets.

Label	Name of data file/data set	File types (file extension)	Data repository (URL)
Data file 1	EIB loan signatures	MS Excel file (.xlsx)	https://www.econpoint.unican.es/databases/
Data file 2	EBRD loan signatures	MS Excel file (.xlsx)	https://www.econpoint.unican.es/databases/
Data file 3	EIB/EBRD overlapping loans	MX Excel file (.xlsx)	https://www.econpoint.unican.es/databases/

Additional information

Funding

Funding for this work has been generously provided by the FNR Intermobility-In Grant, awarded to David Howarth, University of Luxembourg, and Judith Clifton, Universidad de Cantabria and seconded to University of Luxembourg, called “Realigning the European Financial Architecture for the C21-REFA” grant number [INTER/MOBILITY/2022/LE/17567353/REFA-21].

Notes

1. As it focuses on overlapping loans, the dataset does not include EIB loans to countries that are not recipient of EBRD loans.