Taxonomy-alignment and transition risk: A country-level approach

Alessi, Lucia; Battiston, Stefano Working Paper Taxonomy-alignment and transition risk: A country-level approach JRC Working Papers in Economics and Finance, No. 2023/12 Provided in Cooperation with: Joint Research Centre (JRC), European Commission Suggested Citation: Alessi, Lucia; Battiston, Stefano (2023) : Taxonomy-alignment and transition risk: A country-level approach, JRC Working Papers in Economics and Finance, No. 2023/12, European Commission, Ispra This Version is available at: https://hdl.handle.net/10419/283108 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ Taxonomy-alignment and transition risk: a country-level approach Alessi, L. Battiston, S. 2023 JRC Working Papers in Economics and Finance, 2023/12 This publication is a Working Paper by the Joint Research Centre (JRC), the European Commission’s science and knowledge service. It aims to provide evidence-based scientific support to the European policymaking process. Working Papers are pre-publication versions of technical papers, academic articles, book chapters, or reviews. Authors may release working papers to share ideas or to receive feedback on their work. This is done before the author submits the final version of the paper to a peer reviewed journal or conference for publication. Working papers can be cited by other peer-reviewed work. The contents of this publication do not necessarily reflect the position or opinion of the European Commission. Neither the European Commission nor any person acting on behalf of the Commission is responsible for the use that might be made of this publication. For information on the methodology and quality underlying the data used in this publication for which the source is neither Eurostat nor other Commission services, users should contact the referenced source. The designations employed and the presentation of material on the maps do not imply the expression of any opinion whatsoever on the part of the European Union concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. Contact information Name: Lucia Alessi Address: via Enrico Fermi 2749, Ispra (VA), Italy Email: [email protected] EU Science Hub https://joint-research-centre.ec.europa.eu JRC135889 Ispra: European Commission, 2023 © European Union 2023 The reuse policy of the European Commission documents is implemented by the Commission Decision 2011/833/EU of 12 December 2011 on the reuse of Commission documents (OJ L 330, 14.12.2011, p. 39). Unless otherwise noted, the reuse of this document is authorised under the Creative Commons Attribution 4.0 International (CC BY 4.0) licence (https://creativecommons.org/licenses/by/4.0/). This means that reuse is allowed provided appropriate credit is given and any changes are indicated. For any use or reproduction of photos or other material that is not owned by the European Union permission must be sought directly from the copyright holders. The European Union does not own the copyright in relation to the following elements: - Cover page illustration © stock.adobe.com How to cite this report: Alessi, L. and Battiston, S., Taxonomy-alignment and transition risk: a country-level approach, JRC Working Papers in Economics and Finance 2023/12, European Commission – Joint Research Centre, 2023, JRC135889. Executive summary The European Union (EU) has developed the EU Taxonomy for sustainable activities to provide a definition of ‘green’ economic activities, which is used as a basis to assess the greenness of financial investments. However, data on Taxonomy-alignment are only becoming available for larger EU firms, and several challenges remain open to improve the usability of this tool by financial institutions. In parallel, regulators and supervisors, as well as individual financial institutions, have been increasingly paying attention to financial risks stemming from climate change. In particular, a key question regards the exposure of particular investments, portfolios, financial institutions and the financial system as a whole to climate-related transition risk, i.e. the risk linked to certain economic activities which will need to be abandoned in the low-carbon transition, such as those involving fossil fuels. Against this background, in a previous paper (Alessi and Battiston (2022a)) we proposed a methodology to estimate the greenness, or Taxonomy-alignment, and the exposure to transition risk of financial institutions’ investments in the absence of granular information on investee and borrower companies. In particular, we developed Taxonomy-alignment coefficients (TACs) for climate change mitigation and Transition-risk exposure coefficients (TECs) that are specific to each economic sector and largely based on the definitions provided in the Taxonomy. In this paper, we overcome one of the main limitations of TACs and TECs as proposed in Alessi and Battiston (2022a), i.e. their focus on the EU as a whole. In particular, while continuing to focus on climate change mitigation, we develop country-specific coefficients for individual EU Member States and for several non-EU countries. Based on country-level coefficients, we provide an assessment of the Taxonomy-alignment and exposure to transition risk of each economic sector across countries. Moreover, by applying these coefficients to confidential security-by-security data from the European Central Bank on stock and bond holdings of EU investors, we estimate the exposure of each investor category in each country to green and harmful activities. While confirming the findings of the previous paper, i.e. an average Taxonomy-alignment of around 3% and an average exposure to transition risk at around 11%, the empirical application shows that the exposure to transition risk of less regulated financial institutions has more than tripled from 2014 to 2023 to around 18% of total exposure, and 20% of their portfolio holdings. Looking at the cross-section of holders, the levels of Taxonomy alignment and transition risk exposure are largely heterogeneous across sectors and countries, in some cases even within the same sector. In perspective, both estimates at the sector and investor level are needed to assess the speed at which financial markets are moving towards green and away from highly-emitting activities. Taxonomy alignment and transition risk: a country-level approach ∗ Lucia Alessi1,2 and Stefano Battiston3,4 1European Commission - Joint Research Centre 2CefES – Center for European Studies (Universit´a degli Studi di Milano-Bicocca) 3University of Zurich 4Ca’ Foscari University of Venice July 2023 Abstract When firm-level information is not available, the greenness of financial portfolios, in terms of alignment to the EU Taxonomy, and their exposure to climate-related transition risk need to be estimated with a top-down approach. We improve the accuracy of available estimates by providing country-specific coefficients for both dimensions, based on homogeneous definitions of greenness and transition risk across countries. An application on confidential data from the European Central Bank shows that the exposure to transition risk of less regulated financial institutions has more than tripled from 2014 to 2023. Moreover, we show that the levels of Taxonomy alignment and transition risk exposure are largely heterogeneous across countries and sectors. Keywords: greenness, climate-related transition risk, climate-related financial disclosures, EU Taxonomy, green financial flows. J.E.L. classification: G2; G3; Q54 ∗Disclaimer: The content of this article does not necessarily reflect the official opinion of the European Commission. Responsibility for the information and views expressed therein lies entirely with the authors. The authors would also like to gratefully acknowledge research assistance by Carlotta Gianni. E-mail: [email protected], [email protected]. 1 1 Introduction The share of global financial assets under management deemed as ‘green’ under one or more labels has been steadily growing in recent years. This development reflects the increased demand by institutional and private investors for sustainable finance. Drivers for increased demand include compliance with new regulatory standards and changes in preferences. On the other hand, there has been a lack, until recently, of common and science-based approaches to define the greenness of financial investments. This is critical because the achievement of sustainability goals requires methods to measure greenness that are transparent, replicable and widely accepted. The European Union (EU) has developed the EU Taxonomy for sustainable activities to provide a definition of ‘green’ economic activities, which is used as a basis to assess the greenness of financial investments. However, data on Taxonomy-alignment are only becoming available for larger EU firms, and several challenges remain open to improve the usability of this tool by financial institutions. In parallel, regulators and supervisors, as well as individual financial institutions, have been increasingly paying attention to financial risks stemming from climate change. In particular, a question that is asked more and more often regards the exposure of particular investments, portfolios, financial institutions and the financial system as a whole to climate-related transition risk. This is the risk linked to certain economic activities which will need to be abandoned in the low-carbon transition, such as those involving fossil fuels, but also those involving obsolete and high-emitting production processes. It should be stressed that while particular economic activities will need to (almost) disappear, this does not at all mean that particular firms will need to disappear. In fact, companies who are currently carrying out highly-emitting activities can develop credible transition plans and implement them, in order to become progressively less dependent on high-carbon activities and improve their environmental performance, eventually becoming green. Hence, while these companies are in principle exposed to transition risk, this risk may not materialize if the companies themselves transition towards low-carbon. For this reason, investors should carry out a careful assessment of each individual company and its transition plan. In this context, Taxonomy alignment acts as a shield against transition risk, as it shows that the activities carried out by a given firm, which could be in principle exposed to transition risk, are actually compatible with the transition. Also in the case of transition risk, however, firm-level information is often lacking and surely not available on a large scale. At the same time, estimates of financial institutions’ exposures to transition risks are needed, not least for prudential purposes. Against this background, in Alessi and Battiston (2022a) we propose a methodology to estimate the greenness, or Taxonomy-alignment, and the exposure to transition risk of financial institutions’ investments in the absence of granular information on investee and borrower companies. In particular, we develop Taxonomy-alignment coefficients (TACs) for climate change mitigation and Transition-risk exposure coefficients (TECs) that are specific to each economic sector and largely based on the definitions provided in the Taxonomy. As such, TECs reflect a broader definition of transition risk than the one generally used in the literature, as the emphasis is not only on carbon emissions but also on energy inefficiency (for example, when assessing buildings), and are therefore more in 2 line with the definition of transition risk underlying international and European climate-related and sustainability disclosure standards. TACs and TECs can be used to characterize greenness and transition-risk exposure of financial portfolios in the absence of firm-level data on Taxonomy-alignment and exposure to transition risk, as the only information that is needed is the economic sector where the non-financial counterpart is active. In this paper, we overcome one of the main limitations of TAC and TEC as proposed in Alessi and Battiston (2022a), i.e. their focus on the EU as a whole. In particular, while continuing to focus on climate change mitigation, we develop country-specific coefficients for individual EU Member States and for several non-EU countries. This is important because TAC and TEC are meant to estimate the Taxonomy-alignment and transition-risk exposure of economic sectors, which can vary across countries for a given economic sector. For example, the TAC and TEC associated with electricity production are based on the share of renewable and fossil energy, respectively, which vary widely across countries. Hence, country-specific TAC and TEC, to be used based on the location of the investee company, yield more accurate estimates than coefficients based on the average EU level. The contribution of the paper is twofold. First, by developing country-specific coefficients, we provide an assessment of the Taxonomy-alignment and exposure to transition risk of each economic sector across countries. To our knowledge, this information was not available so far in a structured fashion, and for some sectors was not available at all. Second, we apply these coefficients to confidential security-by-security data from the European Central Bank on stock and bond holdings of EU investors. Based on this data, we are able to estimate the exposure of each investor category in each country to green and harmful activities. In perspective, both estimates - at the sector and investor level - are needed to assess the speed at which financial markets are moving towards green and away from highly-emitting activities. The empirical application shows that the exposure to transition risk of less regulated financial institutions has more than tripled from 2014 to 2023 to around 18% of total exposure, and 20% of their portfolio holdings. Looking at the cross-section of holders, the levels of Taxonomy alignment and transition risk exposure are largely heterogeneous across sectors and countries, in some cases even within the same sector. The paper is structured as follows. Section 2 provides some policy background on the EU Taxonomy and on relevant corporate disclosures. In Section 3, we explain how we derive country-specific standardized coefficients for the estimation of Taxonomy-alignment and exposure to transition risk. In Section 4 we describe the data used in the empirical application on EU investor’s holdings, while Section 5 discusses the results. Section 6 concludes. 2 Policy background The EU Taxonomy Regulation1adopted in 2020 establishes a framework to facilitate sustainable investment by providing a clear definition of ‘sustainable’ activities. So far, the Taxonomy has been developed with respect to the environmental dimension, which comprises the following six objectives: i) climate change mitigation; ii) climate 1Regulation (EU) 2020/852 of the European Parliament and of the Council of 18 June 2020 on the establishment of a framework to facilitate sustainable investment, and amending Regulation (EU) 2019/2088 (OJ L 198, 22.6.2020, p. 13-43). 3 change adaptation; iii) sustainable use and protection of water and marine resources; iv) transition to a circular economy; v) pollution prevention and control; and vi) protection and restoration of biodiversity and ecosystems. In particular, a large list of green activities relevant for the first two objectives, has already become EU law.2 An economic activity is defined green in the Taxonomy if it complies with the following requirements: i) it provides a substantial contribution (SC) to at least one of the six objectives mentioned above; ii) it does no significant harm (DNSH) to any of the other objectives; and iii) it complies with a set of minimum social safeguards (MMS). For the SC and DNSH conditions, technical screening criteria are provided, which may take the form of quantitative thresholds (e.g. in terms of maximum CO2 emissions). However, in several cases and especially with respect to DNSH, they make more generic references to existing EU legislation or consist in high-level, qualitative requirements (see Hoepner and Schneider (2022)). Based on the Taxonomy Regulation, all large firms, including financial institutions, need to disclose on the Taxonomy alignment of their business. In particular, as of 2022 it is mandatory to disclose the share of a company’s business that is Taxonomy-eligible, i.e. for which there exist criteria in the Taxonomy. 2022 disclosures refer to FY2021 and to the two Taxonomy climate objectives only. Notice that the Taxonomy-eligible share is only an upper bound for the Taxonomy-aligned share, as it needs to be tested against the SC, DNSH and MMS criteria. As of 2023, non-financial companies need to disclose the shares of their revenues, capital and operational expenditures that are Taxonomy-aligned. One year later, this obligation will extend to financial institutions.3Moreover, the Taxonomy Regulation also amends the Sustainable Finance Disclosure Regulation (SFDR)4by imposing that investment funds marketing themselves as green (the so-called Article 8 and Article 9 products) disclose on the Taxonomy alignment of their investments. The number of firms mandated to disclose their Taxonomy-alignment is bound to increase. For the moment, concerned firms are those in the scope of the Non-Financial Reporting Directive (NFRD)5, which are those with more than 500 employees, i.e. about 11.000 firms in the EU. The NFRD will be replaced by the Corporate Sustainability Reporting Directive (CSRD)6, which will extend the scope of sustainability-related disclosures to all companies (also unlisted) with more than 250 employees and listed companies.7This includes listed SMEs, but with the exception of listed micro-companies, as well as non-EU companies generating a net turnover of EUR150mn in the EU and which have at least one subsidiary or branch in the EU, for a total of around 50.000 firms.8Notably, large financial firms are in the CSRD scope too. 2EU Taxonomy Climate Delegated Act and its Annex 1 and Annex 2 C/2021/2139 (OJ L 442/1, 9.12.2021). Complementary Delegated Act C/2022/0631 amending Delegated Regulation (EU) 2021/2139 as regards economic activities in certain energy sectors and Delegated Regulation (EU) 2021/2178 as regards specific public disclosures for those economic activities. 3Delegated Act supplementing Article 8 of the Taxonomy Regulation C/2021/4987 OJ L 443, 10.12.2021, p. 9–67. 4Regulation (EU) 2019/2088 of the European Parliament and of the Council of 27 November 2019 on sustainability-related disclosures in the financial services sector OJ L 317, 9.12.2019, p. 1–16 5Directive 2013/34/EU of the European Parliament and of the Council of 26 June 2013. 6Directive of the European Parliament and of the Council amending Directive 2013/34/EU, Directive 2004/109/EC, Directive 2006/43/EC and Regulation (EU) No 537/2014, as regards corporate sustainability reporting, 2021/0104 (COD), 30 June 2022. 7To be precise, also companies with more than EUR20mn balance sheet total or more than EUR40mn net turnover will be in scope. 8For listed SMEs an opt-out would be available during a transitional period until 2028. 4 Against this regulatory background, even assuming that the Taxonomy-alignment that will be disclosed by CSRD non-financial companies is precise and reliable (see next section) as these disclosures will be subject to auditing, there is a practical issue that financial institutions will face. Large parts of banks’, insurers’ and investment funds’ exposures are to counterparts which have no obligation to disclose based on the Taxonomy, i.e. (unlisted) SMEs and most non-EU corporates, as well as governments and central banks. In particular, even with the scope enlargement due to the CSRD, there will still be 25 million SMEs in the EU which will have no obligation to report on their Taxonomy-alignment.9They can do that on a voluntary basis, and to this aim, simplified reporting standards currently developed by the European Financial Reporting Advisory group (EFRAG) should become available in the next couple of years. In order to allow banks to also consider these exposures in the assessment of their greenness, the European Banking Authority has developed the so-called Banking book Taxonomy-Alignment Ratio (BTAR), where the use of estimates is allowed. Estimates derived by the present methodology could be used for the calculation of banks’ BTAR. Looking at transition-risk, financial supervisors and central banks in Europe and beyond are paying increasing attention to this dimension, as proved by the publication of a third report on climate risk by the European Systemic Risk Board (ESRB, 2022). With respect to banks, the Banking Supervision arm of the European Central bank has published its first climate stress test on significant institutions (ECB, 2022), while the European Banking Authority has carried out a pilot exercise to investigate how climate risk assessment and classification tools perform (EBA, 2021) and has launched a discussion on the role of environmental risks in the prudential framework (EBA, 2022). The European Insurance and Occupational Pensions Authority (EIOPA) has launched a climate stress test for pension funds, as climate risks are particularly relevant for long-term investors. The European Securities and Markets Authority (ESMA) has started developing a climate risk stress testing framework tailored to the specificities of Central Counterparties (CCPs). In general, there is a growing pressure on financial institutions to identify, measure, manage and monitor climate risk (and broader sustainability risks) in the context of their risk management frameworks. 10 Finally, it is worth mentioning that the EU sustainable finance framework has adopted the so-called ‘double materiality’ perspective, meaning that assessing risks stemming from sustainability factors, including climate, is as important as assessing the impacts of businesses on people and the environment. The double materiality perspective will be central in the CSRD, while the SFDR already asks financial market participants to disclose a list of so-called Principal Adverse Impact (PAI) indicators at entity level. For climate, these include the greenhouse-gas intensity of investee companies and the share of investments in companies active in the fossil fuel sector, among others. 9In terms of number of firms, SMEs represent 99% of all businesses in the EU. Looking at value added, based on Di Bella et al. (2023), in 2022 SMEs accounted for more than 50% in six ‘industrial ecosystems’, namely construction (70%), tourism (65%), textiles (64%), proximity, social economy and civil security (60%), retail (59%) and cultural and creative industries (53%). 10See European Commission, ‘Strategy for Financing the Transition to a Sustainable Economy’, COM/2021/390 final. 5 6. The same procedure is carried out for the TEC. We are of course aware the above procedure is crude and relies on simplistic assumptions. In particular, in reality, the TAC distributions could have fat tails and could have, in general, country specific features. What we present here is best solution we could find to the problem of lack of firm level data, given the objective of estimating at least the order of magnitude of the TAC/TEC. 3.2 Non-ETS Sectors Also for non-ETS sectors, TAC and TEC are derived considering the criteria for SC to climate mitigation and DNSH to mitigation in the EU Taxonomy. For economic activities not included in the Taxonomy, TECs are derived building on the framework of Climate Policy Relevant Sectors (CPRS, see Battiston et al. (2017)), which allows to identify economic activities highly exposed to transition risk. We refer to Alessi and Battiston (2022a) for a more detailed description of the rationale and sources of the coefficients. For example, the share of Taxonomy-alignment in the electricity generation sector is estimated as the share of generation from renewable energy sources (biomass, geothermal, hydro, solar, wind). Similarly, the share of transition-exposure in the electricity generation sector is estimated as the share of generation from fossil fuel sources (coal, oil and gas). To extend the set of TAC and TEC from EU level to individual EU countries, we resort to the statistics that are at the basis of the EU-wide coefficients, mostly from Eurostat. For example, to derive the TAC of the sector NACE H.49.10 (Passenger rail transport, interurban), we have used the ratio of the length of electrified railways over the total length of railways. This number varies across countries and it is available for several EU countries for the years 2011-2019. We estimate the TEC of the same sector as the complement to 1 of the TAC, because it represents the share of non electrified railways (by length). Table 5 in appendix provides an overview of the underlying rationale for the derivation of TAC and TEC for non-ETS sectors, as well as the data sources used for EU countries. Notice that for many sectors TEC are the complement to 1 of TAC, but this is not always the case. In the example above of electricity generation, the sum of generation from renewable and from fossil do not sum up to one, since nuclear is also to be accounted. Despite the inclusion of gas and nuclear among Taxonomy-eligible activities, at this stage we do not to consider these sources of power generation for the development of TAC. We actually take a conservative stand and consider gas-powered electricity generation as an activity which is exposed to transition risk. Indeed, the alignment of electricity generation from gas requires the plant to fulfill a number of criteria that are not specific to the technology used (emission intensity) but are specific to the firm and/or the country of operation. As a result, it is not possible to estimate, based on available data, what is the share of gas-based power plant that would be aligned, and of those that risk to become stranded assets. As for nuclear, there are some requirements that apply at country level and 12 that only a handful of EU member States currently fulfil.14 Within these countries, is currently not possible to estimate the share of existing nuclear facilities fulfilling the plant-level requirements set out in the Taxonomy. Finally, for non-EU countries, we resort to publicly available statistics comparable to those available for EU Member States. All the details on data sources are available in the Excel tool that accompanies the paper. Please note that the TAC/TEC coefficients provided in the accompanying tool are the best estimate we could derive at this stage based on publicly available data. Some specific limitations may apply in terms interpretation of the proxy. Notably, effects due to export are not taken into accounts in the recycling of plastic, or the sales of electric cars. This has been highlighted wherever possible in the tool. 4 Data The analysis is based on yearly data from 2014Q1 to 2023Q1. The main data source is a confidential security-bysecurity database, namely the Eurosystem’s Securities Holding Statistics (SHS) Database - Sector module. The SHS contains information on the holdings of investors aggregated at the level of ESA2010 sectors, and by country. In particular, SHS data cover debt securities, equity instruments and investment fund shares held by investors residing in the Euro Area and several non-Euro Area EU countries (namely Bulgaria, the Czech Republic, Denmark, Hungary, Poland and Romania), as well as non-resident investors’ holdings of Euro Area securities that are deposited with a Euro Area custodian. The SHS database covers around 83% of the total outstanding amount of securities issued by Euro Area residents. The SHS database does not contain information on the NACE codes of the issuers. Thus, we associate the NACE code (4 digits) to each issuer on the basis of the ISIN code of the security using Refinitiv EIKON, which is also used as source for price data. Table 2 reports for every period in the data sample, the number of holding records, the number of issuers and ISIN codes, and the total value in nominal terms prior to any further selection. For example, for 2023Q1, the sample comprises about 1.01 million records of holdings of stocks issued by 37710 distinct issuers, corresponding to a total value in market capitalization of 12985 bn Euros. Table 3 reports descriptive statistics of the coverage of the sample after matching issuers with their NACE codes. The coverage in terms of value is higher than the coverage in terms of number of issuers and always above 85%. The coverage of the sample increases over time and is larger than 95% both in terms of issuers and value for 2023 data. In the analysis we aggregate the monetary values of the holdings along combinations of the following dimensions: 1) ESA2010 sector and country of the holder, 2) ESA2010 sector and country of the issuer, 3) NACE code (4 digits) of the issuer. Finally, to improve the readability of the results, we group the ESA2010 sectors into meta-sectors and exclude from the analysis the sectors that only represent a negligible value of the holdings (see Table 6 in the appendix). 14For example, disposal facilities for low-level waste must be operational already, and Member States should have in place a detailed plan to have in operation, by 2050, a disposal facility for high-level radioactive waste. 13 Period # holdings # issuers # ISIN Value (bn Euro) 2014Q1 540542 35578 38783 7143 2015Q1 545821 33691 36624 8851 2016Q1 549187 33834 36760 7933 2017Q1 569661 32204 35044 9386 2018Q1 599324 33136 36001 9888 2019Q1 645969 32818 35426 9867 2020Q1 670356 32407 35247 8344 2021Q1 781559 34161 37006 12414 2022Q1 967877 36103 39309 13180 2023Q1 1013630 37710 40718 12985 Table 2: Descriptive statistics of the data set. Columns report for every period the following information: number of holding records (with positive value in Euro); tnumber of issuers identified by the internal organization code (unique); number of ISIN codes (unique); total value of the holdings in nominal terms prior to assigning NACE codes and filtering by country. Period % issuers % value Value (bn Euro) 2014Q1 79.75% 86.94% 6210 2015Q1 81.47% 88.52% 7835 2016Q1 82.99% 89.61% 7109 2017Q1 84.09% 92.20% 8654 2018Q1 85.69% 92.99% 9194 2019Q1 87.24% 93.54% 9230 2020Q1 88.47% 94.33% 7871 2021Q1 91.12% 94.99% 11792 2022Q1 93.31% 96.52% 12722 2023Q1 95.90% 97.58% 12671 Table 3: Descriptive statistics of NACE codes coverage. Columns report for every period the following information: percentage of issuers for which the NACE code is available (% issuers); percentage of the total value of holdings that is covered (% value); total value of holdings (Value). 5 Results In this section we apply the country-level TAC and TEC to compute the level of Taxonomy-alignment and transition risk exposure of EU investors’ holdings. We use the following definitions. •Taxonomy Aligned (TA) holdings refers to the value of equity holdings that are Taxonomy aligned. For each individual holding of investor iin a given issuer in NACE sector jfor the amount Xin Euros, the amount that is Taxonomy aligned is T Aij =Xij T ACjwhere T ACjis the Taxonomy alignment coefficient for NACE sector j. •Transition Exposure (TE) holdings are defined analogously. The amount that is exposed to transition risk is T Eij =Xij T ECj, where T ECjis the transition risk coefficient for NACE sector j. In order to compare alignment and exposure of portfolios over time it is useful to separate the effect of changes in the prices of the securities from the effect of changes in the amounts of securities held. The value of TA and TE 14 holdings in real terms are defined as follows: TAij (t) =Xij (t)Pj(tb) Pj(t)TACj(7) TEij(t) =Xij (t)Pj(tb) Pj(t)TECj.(8) In the expression above, tbrepresent the base year, t the current year and Pjthe price of one unit of security of issuer j. For instance, with tb= 2023, for a given year t, T Aij(t) represents the value of the holding in 2023 prices. Changes over time should be interpreted as changes in the value of the holding as if prices were those of 2023. We then aggregate over issuers to obtain the amount of TA holdings for a given investor TAi(t) = X j Xij(t)Pj(tb) Pj(t)TACj(9) TEi(t) = X j Xij(t)Pj(tb) Pj(t)TECj.(10) 5.1 Evolution of TA and TE holdings over time As an illustration of the type of questions that our methodology can address, we first focus on the question whether TA and TE holdings have changed in the recent years, and whether we can detect any particular trends. To this end, we examine the evolution over time of TA and TE holdings in real terms (i.e. expressed in 2023 prices). At the most aggregate level, Table 4 reports for every period in the data sample, the number of combinations of holder countries and sectors, the value of holdings and the value of TA and TE holdings also in percentage of total. Monetary values are expressed in real terms unless indicated otherwise (i.e. “nominal”). The real term value is computed at the individual security level using the last available price (mostly 2023, unless the firm has defaulted or the ISIN code is not held by any holder in the sample) and then aggregated. The total value of the holdings increased from 6556 to 9068 bn Euro from 2014 to 2023. TA grew slightly from 254 bn to 303 bn Euro, while TE increased from 850 bn to 1100 bn Euro. However, in percentage of total holdings, TA and TE holdings decreased slightly from 3.88% to 3.34% and from 12.97% to 12.13%, respectively. It is out of the scope of the present analysis to examine if this decrease is significant and what is its origin. However, we observe that the NACE coverage improves over the period, hence the trend does not seem to be due to data quality issues. As TA and TE estimates are based on the economic sector of the issuer, the trend observed in the data would be consistent with the hypothesis that holders have a tendency to invest a growing part of their portfolios in NACE sectors characterized by zero TAC and TEC, and a decreasing share of their portfolios into real economy sector characterized by larger TAC and TEC15. To investigate whether this is the case, Table 7 in the appendix shows 15Note that if investors tend to invest over time larger fraction of their portfolios into securities issued by financial firms, then the TA share and TE share can both decrease. The fact that TA share decrease does not mean the the TE share has to increase. Indeed, at the portfolio level both can increase or decrease or any combination, depending on the set of sectors on which the investor put more weight. 15 the evolution of the share of aggregate holdings in each NACE sector over time. The percentage of holdings in NACE sector K (Finance) has almost doubled over the considered time span. The financial sector has TAC and TEC equal to 0, as no technical screening criteria exist in the Taxonomy for financial activities as such for climate change mitigation (see Section 3). Hence, investments in equity funds or in any financial entity do not contribute to TA and TE, although a financial institution holds a portfolio of investments in companies, some of which operate in the real economy and have non-zero TAC and TEC. In any case, we are unable to unpack the holdings of financial intermediaries and need to stick to the Taxonomy approach, whereby financial activities are not Taxonomy-eligible. The percentage of holdings in NACE sector J (Information and communication) has almost tripled from 2014 to 2023. While the Taxonomy does include two activities related to this NACE sector, its TAC and TEC are zero. Finally, holdings in manufacturing companies (NACE sector C) decreased from 72% to 45%. While NACE sector C is very large and comprises many subsectors with zero TAC and TEC, it also includes some of the sectors with a non-zero TAC and a relatively large TEC of 0.5. Value (nom.) Value TA TE TA TE Period bnEUR bn EUR (bn EUR) (bn EUR) (% of Value) (% of Value) 2014Q1 4283 6556 254 850 3.88% 12.97% 2015Q1 5366 7065 244 901 3.45% 12.76% 2016Q1 4839 7182 255 978 3.54% 13.62% 2017Q1 5839 7529 286 978 3.80% 12.99% 2018Q1 6154 7500 285 978 3.80% 13.03% 2019Q1 6452 7863 294 983 3.74% 12.51% 2020Q1 5524 7960 304 1013 3.82% 12.73% 2021Q1 8433 8473 282 1031 3.32% 12.16% 2022Q1 9318 8932 287 1065 3.21% 11.93% 2023Q1 9068 9068 303 1100 3.34% 12.13% Table 4: Aggregate holdings over time. Columns report for every period: value of holdings in bn Euro; value of TA and TE holdings in bn Euro and in percentage of total. Values in Euro are in real terms unless specified otherwise (nominal). The area plot in Figure 2 shows the evolution of TA and TE holdings in monetary value (in real terms) by holder sector, while the area plot in Figure 3 shows the evolution of TA and TE holdings as percentage of total holdings, with the breakdown by holder sector. Throughout the sample, “NFCs”, “Investment funds”, and “Households and No profit institutions” have the largest volumes of TA holdings. Looking at transition risk, it is noteworthy the increase of OFI’s TE holdings (in fuchsia), which over time becomes comparable to TE holdings of households and no profit institutions (in green). As percentage of overall TE holdings, OFI’s TE holdings increased by more than three times from 5.5% of total TE in 2014 to 18.3% in 2023 (see Table 9 in the appendix). As we do not observe a generalized in crease in TE holdings across sectors, nor across financial institutions, this finding is consistent with the hypothesis that transition risk might be shifting to less regulated parts of the financial system. This result is in line with Alessi et al. (2021a), who show that after the Paris Agreement, European investors reduced their participation in high-carbon companies at the aggregate level, but OFIs increased it. 16 Figure 2: Evolution over time of TA (left) and TE (right) holdings in monetary values Figure 3: Evolution over time of TA (left) and TE (right) holdings as percentage of total holdings across EU investors. We complement the previous results with the study of the evolution over time of the TA and TE shares with reference to the total holdings of a given sector, as opposed to total holdings across sectors. In other words, we look at the % of Taxonomy alignment and Transition-risk exposure of each sector defined as the ratio of TA or TE holdings over the value of the holdings of the sector. As shown in the left panel of Figure 4, TA holdings as percentage of the portfolio value by holder sector is generally below 6% and at about 3% on average. However, the Government sector exhibits an exceptionally large share of TA compared to the other sectors, exceeding 17% in some years. This might be due to Governments’ holdings into energy and utility companies, active in NACE sectors which are among those with the highest TAC. TA shares remain relatively stable over time, with some exceptions. In particular, as shown in Table 10 in the appendix, banks’ TA share more than halves from 4.4% to 2.1% from 2014 to 2023. Looking at the right panel of Figure 4, the TE portfolio share is for all sectors higher than the TA 17 share, and about 11% on average. The Government sector is associated the largest portfolio TE share, probably again owing to its holdings in NACE sectors linked to energy, which not only have among the highest TACs but also fairly large TECs (and particularly large in some countries). OFIs exhibit the largest variation in the TE share of their portfolio, which was at 5.5% in 2014, peaked to almost 24% in 2020 and is now at around 20%, again pointing towards a tendency of less regulated financial institutions to invest in sectors characterized by high TEC and low or zero TAC, such as those exclusively linked to fossil fuels. Figure 4: Left: Evolution over time of TA portfolio shares by investor class. Right: Evolution over time of TE portfolio shares by investor class. 5.2 TA and TE levels of holder sectors: cross country heterogeneity Next, we address the question of the heterogeneity across countries of TA and TE portfolio shares for the various types of holders. To this end, the box plots in Figure 5 represent for each holder sector the interquartile range (IQR) and the median of TA portfolio shares across countries. In several cases the values of the IQR (i.e. the length of box) are comparable or larger than the value of the median (the red bar), indicating a large dispersion of the values across countries. For instance, looking at TA (left panel), Gov is the sector with the largest IQR while InvFund is one with the smallest IQR. While they have a comparable median, the IQR of the first is about 7 times larger than the one of the second. The comparison is even starker for TE (right panel, note the different scale of the two charts ). This indicates a much larger dispersion in TE portfolio shares across Governments of different EU countries, compared to investment funds and other investor classes across EU countries. One possible explanation is that Governments tend to have a less diversified portfolio of holdings than, e.g., financial institutions. In particular, some governments may have large stakes in domestic energy companies with low TAC, while othergovernments may have large stakes in domestic utilities companies with relatively high TAC. Similar considerations hold for non-financial corporations, which tend to have a strong domestic component in their holdings. 18 Figure 5: Box plots representing for each investor class the interquartile range (IQR) and the mean of TA portfolio shares (left panel) and TE portfolio shares (right panel) across countries of the holder sector in 2023Q1. Note the different scale of the two charts. The tool to compute TA and TE across sectors and countries makes it tempting to compile a ranking of virtuous countries. Caveats should be highlighted before proceeding to such an exercise. Notably, levels can vary across countries upon the following factors: •TA can be high in certain economic sectors, such as electricity, utility and railways. Holder sectors with, relatively speaking, larger shares in these sectors will have, ceteris paribus, higher TA. However, this could reflect institutional factors rather than incentives or decisions to green their investments. •Similarly, TE can be large in sectors of primary energy (e.g. the oil & gas value chain) and energy-intensive manufacturing (e.g. cement and iron & steel), for which the concentration of exposure could again reflect national specialization or institutional factors. For less diversified holder sector/countries, with investments concentrated in particular economic sectors or even individual companies, the above phenomena can lead to particularly high levels of TA and/or TE. A deep dive in an example can be useful to understand the caveats. In 2021Q1, the sector Non-financial investors (S16) of PT had a portfolio weight of 89.4% on a company classified as utility electricity transmission (hence with TAC = 1), plus some additional smaller weights on companies classified as railways and utility electricity, totaling a TA of 92%. In other periods, the weights over these companies vary, but the overall TA of this country-holder sector remains very high on average. While this sector represents a small portion of holdings across holders in PT, when we rank holder sectors in different countries we need to keep in mind that in some countries, some holder sectors may be less diversified that in other countries. 19 5.3 Country-level TAC and TEC vs EU-level TAC and TEC Finally, as a robustness check, we compare the results described above with those we obtain by applying EU-level coefficients as in Alessi and Battiston (2022a), as opposed to the country-level coefficients developed in this study. Figure 7 in the appendix plots the estimates of TA portfolio shares (left panel) and TE portfolio shares (right panel) for the various investor classes at the aggregate EU level, based on EU-level TAC and TEC (orange bars) and country-level TAC and TEC (mycol bars). The differences between the two sets of estimates are relatively small, i.e. in the order of a half a percentage point for TA, and around 4 p.p. for TE, on average. This is, on the one hand, reassuring, as the two approaches do not yield different messages overall while looking at aggregate exposures at EU level. On the other hand, discrepancies between the two sets of estimates indicate that there is indeed value in using country-level coefficients, as the estimates are not exactly the same, and are necessarily more precise. Figure 1 in the appendix looks at differences between the two approaches at a higher level of disaggregation. Each dot in the scatter plots represents a combination of a holder sector, a country and a period. The position of the dot in the quadrant depends on the value of the estimate obtained by using EU-level coefficients (x-axis) and country-level coefficients (y-axis). For those combinations of country-sectors and periods on the 45 degree line, it makes no difference to adopt one approach or the other. However, there are a large number of dots that are not on the 45-degree line, and some are actually quite far from it. Therefore, even if at the aggregate EU level it may not make a big difference to use one approach or the other, there are several cases in which country-level coefficients are clearly preferred. 6 Conclusions and further research The main contribution of this paper is to extend at the country level the methodology previously developed in Alessi and Battiston (2022a), which estimates on the one hand the level of Taxonomy-alignment of financial institutions’ investments and, on the other hand, their exposure of climate-related transition risk. The goal of this methodology is to overcome the problem of limited availability of data for many counterparties of financial institutions, which makes it difficult or impossible to estimate the Taxonomy-alignment and the transition risk of the portfolio as a whole. While the coefficients (TAC and TEC) proposed in Alessi and Battiston (2022a) are estimated on the basis of EU-level statistics, here we develop country-specific coefficients for individual EU Member States and non-EU countries. This extension is crucial to enhance the precision of the estimates. Indeed, TAC and TEC at country level may largely differ from those at EU level. For instance, in the sector of electricity generation, the level of reliance on renewable vs. fossil sources varies substantially across countries. As a result, country-level TAC and TEC bring higher granularity in the estimation. We apply the methodology to a confidential dataset covering equity and bond holdings for investors located in 20 EU27 from 2014 to 2023. In the aggregate, no marked trends are observable. However, some changes over time become visible for specific holder sectors, in particular a substantial increase in the exposure to transition risk of less regulated financial institutions. Looking at cross-country heterogeneity, our results indicate a large dispersion of Taxonomy-alignment and transition-risk exposure across countries, in particular for some investor classes. Our estimates of Taxonomy-alignment and transition-risk exposure for individual investor classes can be used by supervisors as benchmark levels for each sector and country, against which the performance of individual financial institutions can be assessed. From a macro(prudential) perspective, our estimates provide information on where the market stands in terms of greenness and risk exposure. This information can be used, for instance, to identify clusters of country/sectors with similar values of alignment and exposure. It should also be stressed that the methodology developed in this paper does not need confidential or supervisory data, as it can be applied to any portfolio of holdings. As such, it can be used by a financial institution to assess its own exposures, as well as on publicly available data. Looking at Taxonomy-alignment in particular, given the particular features of the relevant regulatory environment, a perfect measure of the overall Taxonomy-alignment of financial institutions will not be available in the foreseeable future. However, our methodology can be used already now to assess how green individual financial institutions and the financial system as a whole are, considering their SME and non-EU exposures too. This information is needed to financial supervisors, as an increase in the Taxonomy alignment of a financial institution can be seen as a mitigating measure towards environmental risks the institution may be exposed to. For the same reason, it is a crucial piece of information for macroprudential supervisors, who are in charge of monitoring risks to the financial system as a whole. Estimates of Taxonomy-alignment are essential not only to policymakers, but also to financial institutions themselves, as they need to design their transition plans and deserve their transition efforts to be recognized against measurable performance indicators. To this aim, financial institutions can use the present methodology for voluntary disclosures and whenever the regulation allows the use of estimates. Finally, this is an information that the market is asking for, to be able to make informed investment decisions. Turning to climate-related transition risk, reliable scenarios and stress-testing exercises can only be based on a reliable assessment of financial institutions’ exposures. Since a legal definition, or Taxonomy, of harmful activities is lacking, for the time being such assessment can only be based on estimates and proxies. Scenarios and stress-testing exercises also need to be carried out at some level of aggregation, since firm-level information is not only often unavailable, but would also be difficult to process in the context of large-scale exercises. However, to increase the reliability of the results, modellers should try to reflect the ‘transition discussion’ in their analysis. In other words, not only no company is ‘doomed’ owing to the low-carbon transition, but only very few and well-defined economic activities are entirely exposed to transition risk, e.g. coal mining. Indeed, even within sectors characterized by high carbon emissions, such as e.g. transport, some manufacturing activities, and buildings, companies can not only improve their environmental performance and reduce their exposure to transition risk, but even become fully green. 21 B Aggregation and selection of ESA2010 holder sectors HS Code Description HS Name Filter U Unallocated Unallocated 0 S 11 Non-financial corporations NFC 1 S 121 Central Banks Central Banks 1 S 122 Deposit taking corporations except central banks Banks 1 S 123 Money market funds (MMF) MMF 0 S 124 Non-MMF Investment funds Inv.Funds 1 S 125W Other financial corporations1 excluding financial vehicle corporations OFI 1 S 125A Financial vehicle corporations OFI 1 S 128 Insurance corporations Ins.&Pens. 1 S 129 Pension funds Ins.&Pens. 1 S 12KU Monetary financial institutions (sub-sector not identified) Monetary financial inst. 0 S 12QU Other insurance corporations and pension funds (sub-sector not identified) (transitional period) Ins.&Pens. 1 S 1311 Central government (voluntary breakdown) Gov. 1 S 1312 State government (voluntary breakdown) Gov. 1 S 1313 Local government (voluntary breakdown) Gov. 1 S 1314 Social security funds (voluntary breakdown) Gov. 1 S 13U Other General Government (sub-sector not identified) Gov. 1 S 14 Households excluding non-profit institutions serving households (voluntary breakdown (for resident investors); mandatory if third party holdings) HH&noP 1 S 15 Non-profit institutions serving households (voluntary breakdown) HH&noP 1 S 1MU Other households and non-profit institutions serving households (sub-sector not identified) HH&noP 1 S 16 Non-financial investors excluding households (to be reported if third party holdings) Other nonFInv 0 S 1KK Central banks and general government (to be reported only for holdings by non-euro area countries) Non-EA 0 S 1KL Investors other than central banks and general government (to be reported only for holdings by non-euro area countries) Non-EA 0 Table 6: Holder sectors codes and legend. The table reports the following information. The holder sector (HS) code identifies sectors according to the ESA2010 classification. The description of the sector is the textual description from the SHS documentation. HS name is the name used in the legends. Filter=1 identifies the sectors used in the analysis. Several HS codes are intentionally aggregated under the same HS name for readability of the charts. 28 C Holdings by issuer’s NACE sector NACE 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 A 0.04 0.03 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 B 0.75 1.16 1.53 1.66 1.62 1.67 1.59 1.60 1.87 1.73 C 72.24 59.64 43.45 44.46 45.25 47.91 45.44 46.72 45.24 45.47 D 2.36 2.83 4.71 5.31 5.35 5.22 5.54 5.06 5.24 5.27 E 0.20 0.26 0.42 0.42 0.46 0.45 0.50 0.52 0.50 0.48 F 0.91 1.13 1.83 1.73 1.80 1.69 1.75 1.86 1.77 1.70 G 1.80 3.15 4.96 5.28 5.27 4.79 5.01 5.06 5.17 5.54 H 1.02 1.44 2.65 2.66 2.65 2.48 2.58 2.70 2.74 2.62 I 0.25 0.33 0.51 0.55 0.58 0.60 0.59 0.61 0.65 0.66 J 5.05 7.15 10.97 11.26 10.91 10.72 11.76 11.48 13.22 14.52 K 7.94 9.16 14.71 14.77 14.85 13.94 14.42 13.86 14.25 13.72 L 0.75 1.06 1.79 1.79 1.88 1.73 2.00 1.95 2.06 2.06 M 0.79 1.04 1.76 1.62 1.66 1.58 1.69 1.78 1.91 1.91 N 5.67 11.25 10.05 7.79 6.98 6.41 6.23 5.69 4.22 3.16 O 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 P 0.01 0.01 0.01 0.01 0.01 0.02 0.03 0.03 0.03 0.04 Q 0.14 0.23 0.37 0.39 0.40 0.45 0.46 0.45 0.47 0.46 R 0.06 0.10 0.19 0.21 0.21 0.21 0.28 0.52 0.53 0.54 S 0.01 0.04 0.05 0.06 0.06 0.06 0.07 0.06 0.06 0.06 U 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Grand Total 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 Table 7: Evolution of the value of holdings in each NACE sector (main section) as percentage of total holdings in the full dataset 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 A B C D E F G H I J K L M N O P Q R S U Figure 6: Area plot (stacked) showing the evolution of the value of holdings in each NACE sector (main section) as percentage of total holdings in the full dataset. 29 D TA and TE - contribution by holder sector Holder sector 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 Banks 6.7 4.1 2.9 2.3 2.1 2.0 2.0 2.6 2.2 2.0 Gov. 12.2 12.0 16.6 15.6 15.5 15.4 14.8 16.0 15.8 15.7 HH&noP 21.7 21.8 22.3 19.7 19.6 18.4 17.7 19.8 20.1 19.9 Ins.&Pens. 3.2 3.4 3.7 3.5 3.8 3.9 4.0 4.2 4.1 3.9 Inv.Funds 18.4 22.1 21.6 22.6 23.0 22.8 22.4 25.8 26.8 25.9 NFC 32.4 25.2 24.1 28.9 28.8 29.9 31.1 24.0 23.9 26.0 OFI 5.4 11.5 8.9 7.3 7.3 7.5 8.0 7.4 7.0 6.6 Grand Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Table 8: Evolution of TA holdings by holder sector as percentage of total across sectors. Holder sector 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 Banks 13.3 13.3 13.5 14.5 16.1 16.6 16.2 16.8 16.9 16.8 Gov. 25.7 23.1 20.0 19.7 19.3 19.4 20.7 20.6 20.9 20.5 HH&noP 14.1 13.0 12.9 13.4 12.8 13.3 13.1 13.9 13.8 14.1 Ins.&Pens. 12.1 11.3 10.6 10.6 10.3 10.4 9.7 10.1 9.4 8.6 Inv.Funds 9.7 8.6 8.2 8.1 7.9 7.7 6.8 7.1 7.0 6.9 NFC 19.5 17.0 16.4 14.9 13.5 14.5 13.3 11.8 12.6 14.7 OFI 5.5 13.8 18.3 18.7 20.0 18.2 20.1 19.6 19.4 18.3 Grand Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Table 9: Evolution of TE holdings by holder sector as percentage of total across sectors. E TA and TE shares by holder sector Holder sector 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 Banks 4.4 2.5 2.2 2.1 2.1 2.1 2.2 2.2 2.0 2.1 Gov. 13.4 12.8 17.3 17.3 17.5 17.3 16.1 15.2 15.1 15.0 HH&noP 5.9 5.5 5.6 5.6 5.6 5.3 5.0 4.9 4.8 4.8 Ins.&Pens. 2.3 2.3 2.2 2.3 2.5 2.5 2.5 2.5 2.4 2.1 Inv.Funds 1.7 1.7 1.7 1.8 1.9 1.8 1.8 1.8 1.8 1.8 NFC 5.9 4.4 4.3 5.5 5.3 5.6 5.9 4.2 4.2 4.6 OFI 2.8 5.8 5.1 4.4 4.3 3.9 4.6 3.5 3.3 3.1 Table 10: Evolution of TA share. i.e. fraction of TA holdings over holdings of each holder sector. Holder sector 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 Banks 13.2 14.0 15.4 16.1 18.6 18.2 19.0 18.9 18.7 18.5 Gov. 25.6 24.4 22.9 21.9 22.1 21.3 24.3 23.2 23.2 22.5 HH&noP 14.1 13.8 14.8 15.0 14.7 14.6 15.4 15.6 15.3 15.4 Ins.&Pens. 12.0 11.9 12.2 11.8 11.9 11.4 11.4 11.4 10.4 9.4 Inv.Funds 9.7 9.1 9.4 9.0 9.1 8.4 8.0 8.0 7.8 7.6 NFC 19.4 18.0 18.8 16.6 15.6 15.9 15.7 13.3 14.0 16.1 OFI 5.5 14.6 20.9 20.8 23.0 20.0 23.6 22.1 21.5 20.1 Table 11: Evolution of TE share. i.e. fraction of TE holdings over holdings of each holder sector. 30 F Country-level vs EU level TAC and TEC 0.00% 5.00% 10.00% 15.00% 20.00% Inv.Funds Banks Ins.&Pens. OFI NFC HH&noP Gov. TA % EU TA % cntr 0.00% 5.00% 10.00% 15.00% 20.00% 25.00% 30.00% Inv.Funds Ins.&Pens. HH&noP NFC Banks OFI Gov. TE % EU TE % cntr Figure 7: Portfolio TA share (left) and TE share (right) across holder sectors, comparing estimates using countrylevel TAC, TEC (mycol bars) and EU-level TAC, TEC (orange bars). Figure 8: Scatter plot of portfolio TA share (left) and TE share (right) computed using EU-level TAC, TEC (x-axis) and country-level TAC, TEC (y-axis), across combinations of holder sectors,countries and periods. 31 GETTING IN TOUCH WITH THE EU In person All over the European Union there are hundreds of Europe Direct centres. You can find the address of the centre nearest you online (european-union.europa.eu/contact-eu/meet-us_en). On the phone or in writing Europe Direct is a service that answers your questions about the European Union. You can contact this service: — by freephone: 00 800 6 7 8 9 10 11 (certain operators may charge for these calls), — at the following standard number: +32 22999696, — via the following form: european-union.europa.eu/contact-eu/write-us_en. FINDING INFORMATION ABOUT THE EU Online Information about the European Union in all the official languages of the EU is available on the Europa website (europeanunion.europa.eu). EU publications You can view or order EU publications at op.europa.eu/en/publications. Multiple copies of free publications can be obtained by contacting Europe Direct or your local documentation centre (european-union.europa.eu/contact-eu/meet-us_en). EU law and related documents For access to legal information from the EU, including all EU law since 1951 in all the official language versions, go to EUR-Lex (eur-lex.europa.eu). Open data from the EU The portal data.europa.eu provides access to open datasets from the EU institutions, bodies and agencies. These can be downloaded and reused for free, for both commercial and non-commercial purposes. The portal also provides access to a wealth of datasets from European countries.