Excessive Deficit Procedure Statistics Working Group

Transcription

1 Excessive Deficit Procedure Statistics Working Group June 2016 Ampere Beginning at 9:30 Part B Item 5 Eurostat data collection on contingent liabilities and potential obligations overview of the 2015 results and the way forward Joseph BECH Building Tel.: (+352) L Luxembourg Fax: (+352)

2 1 Introduction The aim of this paper is to provide an overview of the 2015 Eurostat data collection on contingent liabilities and potential obligations. It also presents a feedback on the publication of the corresponding data at. In particular, the paper includes the updated Eurostat's assessment of compliance with the Task Force guidance, taking into account the 2015 results. Furthermore, some outstanding issues linked to the next data collection are proposed for discussion. They concern a harmonisation of reporting conventions used for the data transmission, a revision policy and a way forward regarding consolidated data for liabilities of government controlled entities classified outside general government (public corporations). 2 Background and milestones Within the Enhanced Economic Governance package ( Six pack ) adopted in 2011, Council Directive 2011/85/EU 1 of 8 November 2011 on requirements for budgetary frameworks of the Member States ( Directive ) requires the adoption of measures for publication of relevant information on contingent liabilities by Member States. Article 14(3) of Chapter VI states the following: "For all sub-sectors of general government, Member States shall publish relevant information on contingent liabilities with potentially large impacts on public budgets, including government guarantees, non-performing loans, and liabilities stemming from the operation of public corporations, including the extent thereof. Member States shall also publish information on the participation of general government in the capital of private and public corporations in respect of economically significant amounts". In order to assist Member States in the implementation of the new statistical requirements of the Directive, Eurostat set up a Task Force in co-operation with Member States and DG ECFIN 2. Its final report provides, inter alia, a set of templates and related notes indicating the methodology, the scope of compulsory items, the periodicity and the timeliness for national and Eurostat publication of data on contingent liabilities. Most Member States have indicated their willingness to take into account the conclusions of the Task Force when implementing the provisions of the Directive at. The Task Force agreed that the Member States should publish regularly on a national basis the following indicators: government guarantees, liabilities of government controlled entities classified outside general government (public corporations), non-performing loans (government assets) and the data on participation of government in the capital of corporations. The Task Force also agreed that the same indicators (except participation of government in the capital of corporations) would be published by Eurostat. The Eurostat data collection was formalised by the Eurostat Decision of 22 July 2013 on Supplement on contingent liabilities 1 OJ L 306, , p.41 2 Three Task Force meetings took place in 2012: 29 June, 5 September and 6 November

3 and potential obligations to the EDP related questionnaire 3. The Decision also specified that the data on liabilities of government controlled entities classified outside general government (public corporations) would be compiled by Eurostat on the basis of the Questionnaire on government controlled units classified outside general government 4. The Member States were asked to implement the recommendations of the Task Force at a national basis and provide the relevant data to Eurostat by the end of The first round of data publication at national and EU level was reviewed and assessed in the report presented to the EDPS WG and ECFIN in June The second publication of data took place by the end of December 2015 on a national basis and by the end of January 2016 at EU level. This report includes the updated Eurostat assessment of compliance with Task Force guidance taking into account the 2015 results. The Eurostat assessment of the status of the data publication at has been provided to DG ECFIN as a contribution to the Commission overall assessment of Member States' compliance with the Directive. 3 Overview of the 2015 Eurostat data collection on contingent liabilities and potential obligations In the 2015 data collection, all Member States but Cyprus provided data to Eurostat in the 'Supplement on contingent liabilities and potential obligations' around the reporting deadline, i.e. the end December The accompanying metadata files were provided by almost all Member States. The Questionnaire on government controlled entities for the reference year 2014 was provided by all countries but the Czech Republic 5, Germany, France and partly Italy. In these countries the data reported refer to For Cyprus, the available data refer to On the 27 th of January 2016, Eurostat published the figures in millions of national currency and in percentage of GDP 6 in its database 7, together with the related metadata and country specific footnotes 8. An accompanying press release was also issued 9. The main results are presented in Annex Q.pdf/0b35165a-ee53-470a-a15a-7beaa98aac8b 4 The questionnaire received a positive opinion from the CMFB in Data are collected on an annual basis and the first data transmission took place in December At the end of February, respectively at the beginning of May 2016, Czech Republic and Germany submitted to Eurostat an updated questionnaire on government controlled entities with data for The GDP figures provided during the October 2015 Excessive Deficit Procedure notification are used for the calculation of indicators and bb

4 3.1 Completeness and coverage of data The completeness and the coverage of data vary depending on the indicator and on Member States. The detailed analysis of the data completeness and coverage is presented in Annex 2. Government guarantees All Member States but one provided figures for the total government guarantees for general government for the years For Cyprus, data for 2014 was not reported. The data coverage is not exhaustive for Belgium, Ireland, Greece, France and Croatia as the data are not available for the local government subsector. For France, the social security sector is also not covered. In addition, there are some country specific limitations. For Czech Republic, data for enterprises included in the general government sector are not fully complete, however all significant entities are covered. For Germany, data include neither institutional commitments of government bodies in the form of the so-called 'Anstaltslast', provided to public law institutions, nor the so-called 'Gewährträgerhaftung'. 10 For Hungary, data include also guarantees provided to some units classified in the general government sector. For Finland, data are not fully consolidated for local government. For Spain, data are missing for standardised guarantees, however if such guarantees exist the amounts are negligible. For Denmark and Poland, data on standardised guarantees are not available at the level of local government. Off-balance PPPs All countries except Cyprus provided data on off-balance PPPs for Data are not exhaustive for Austria where central government is not covered and for Slovenia and the United Kingdom where data for local government are not available. For Germany, all assets of PPP projects are recorded on the government's balance sheet. For Finland and Sweden, data coverage might be not exhaustive. Non-performing loans For 2014, 23 out of 28 countries provided information on non-performing loans of general government. The missing countries are Belgium, France, Croatia, Cyprus and Slovakia. Data coverage is not exhaustive for Portugal, Italy and Finland. Data does not cover the local government subsector in Spain and the social security sector in Greece; however the amounts are not expected to be significant. Liabilities of government controlled entities classified outside general government For some Member States, there are major differences between the data published last year and the 2015 transmission. The changes between the two reference years are mainly due to significant improvements in data coverage, updated data sources or sector reclassifications. The coverage has been significantly improved for Belgium and Luxembourg including the liabilities of public corporations involved in financial activities, which were not available in 10 Related risks to government budgets have, however, been disclosed as government potential obligations, within the total amount of liabilities of government controlled entities classified outside the general government sector

5 the publication of last year. Besides, for Belgium the data for the local government include more units than the publication of last year. The data coverage was also extended for Spain (compared to 2013 the data for 2014 include the liabilities of nationalised banks), Croatia (the coverage for 2014 was extended compared to 2013 data and more units at local and central governmental level were included) as well as for Netherlands, Poland and Sweden. 4 Publication of data on contingent liabilities and potential obligations at a national basis 4.1 Status of data publication by Member States In February 2016, Eurostat took stock and reviewed the national publication of data on contingent liabilities and potential obligations by Member States. The majority of Member States publish data for all five indicators following the Task Force templates. Two countries (Greece and France) do not publish any data in the context of Article 14.3 of the Directive at. In addition, a number of countries publish data only for some of the indicators and/or not for all the years required. An overview of the publication status by indicator is presented in Table 1 below. The links to the publication of data at are available on Eurostat website. In summary, data on government guarantees are published by all but two countries (Greece and France). For Ireland, the publication does not follow the agreed templates. Data on off-balance PPPs are not published by three countries (Bulgaria, Greece and France). For one country (Cyprus), the data for 2014 are not available. Information on non-performing loans was not made available by five countries (Belgium, Greece, France, Croatia and Cyprus). Four countries (Greece, France, Cyprus and Luxembourg) do not publish data on liabilities of public corporations. In addition, for the Czech Republic, Germany, Italy, Hungary and Austria, data are not available for the reference year Data on the participation of government in the capital of corporations are not published by six countries (Belgium, Greece, France, Croatia, Cyprus and Luxembourg). For Hungary and Austria, data for 2014 are not available. For Italy, data available refer to Almost all publishing countries provided the required breakdown for private and public corporations. A more detailed review of data completeness and coverage is conducted with the analysis of the data provided by the Member States to Eurostat, presented in Annex

6 Table 1. Status of national publication of data on contingent liabilities in the context of Directive 2011/85/EU (status of end February 2016) MS Government guarantees Off-balance PPPs Nonperforming loans Indicators Liabilities of public corporations Participation of government Publishing Authority BE Yes Yes No Yes No Budget Administration BG Yes No Yes Yes Yes NSI/MoF CZ Yes Yes Yes Yes** Yes MoF DK Yes Yes Yes Yes Yes MoF DE Yes Yes Yes Yes** Yes NSI EE Yes Yes Yes Yes Yes MoF IE Partly* Yes Yes Yes Yes MoF EL No No No No No - ES Yes Yes Yes Yes Yes MoF FR No No No No No - HR Yes Yes No Yes No NSI IT Yes Yes Yes Yes** Yes** MoF CY Yes** Yes** No No No NSI LV Yes Yes Yes Yes Yes MoF LT Yes Yes Yes Yes Yes MoF LU Yes Yes Yes No No NSI HU Yes Yes Yes Yes** Yes** State Treasury MT Yes Yes Yes Yes Yes NSI NL Yes Yes Yes Yes Yes NSI AT Yes Yes Yes Yes** Yes** NSI PL Yes Yes Yes Yes Yes MoF PT Yes Yes Yes Yes Yes MoF RO Yes Yes Yes Yes Yes MoF SI Yes Yes Yes Yes Yes NSI SK Yes Yes Yes Yes Yes MoF FI Yes Yes Yes Yes Yes MoF SE Yes Yes Yes Yes Yes NSI UK Yes Yes Yes Yes Yes NSI * For Ireland, the publication does not follow the agreed templates. **Data available for 2013 but not available for For Italy, data on liabilities of public corporations refer to 2013 whereas data on participation of government refer to In a majority of countries, data are published on the websites of the Ministries of Finance. The National Statistical Institutes publish (all or some) data on contingent liabilities in eleven countries. In two countries, data are available to the public on the websites of other government institutions, i.e. the Budget Administration in Belgium and the Treasury in Hungary. An overview of publishing authority is presented in Table 1 above. For transparency purposes, it was agreed by the Task Force members that all indicators required by the Directive (including also the publication of fiscal data related to Article 3.2 of the Directive) would be published together by a national authority on a single dedicated

7 national webpage or website. Almost all publishing countries follow this guidance. Additional progress in this respect is needed in Bulgaria, France and Slovenia. The links to the publication at is available on Eurostat website 4.2 Relation between data published at and data provided to Eurostat Compared to data published at, a higher completeness of data is observed in the data reporting to Eurostat. This mostly concerns the data for the last reference year (2014) and some detailed breakdowns. In a great majority of cases, there are no major differences detected between data published nationally and those provided to Eurostat. Discrepancies are observed for data on non-performing loans in the Czech Republic, Hungary and Slovakia and for data on liabilities of public corporations for the Czech Republic, Denmark, Estonia, Ireland, Lithuania, Hungary and Slovakia. In case of liabilities of public corporations, the discrepancies are due to different methodologies used in the compilation of aggregates. For example, Denmark and Ireland are including in the total amount of liabilities of public corporation also the liabilities of the National Central Bank, which is excluded from the data published by Eurostat. Therefore, the amounts published at are higher than those published at European level. For Slovakia and Estonia, the differences between the figures published by Eurostat and those published by the Ministry of Finance are mainly due to the inclusion of liabilities which are not part of the Maastricht debt in the national publication. As indicated already above, an assessment of the completeness and coverage of data published at a is associated with the corresponding analysis of data provided to Eurostat. The detailed results of this assessment are presented in Annex 2. 5 Way forward Based on the 2015 data collection, Eurostat has identified some issues for further consideration and discussion in the EDPS WG. They concern the reporting conventions used for the data transmission, the revision policy and the way forward regarding consolidated data for liabilities of government controlled entities classified outside general government (public corporations). 5.1 Harmonisation of reporting conventions The following reporting conventions were recommended to the Member States by Eurostat during the data transmission in December 2015: Reporting conventions (used in 2015) "M" => 'not applicable': "0" => 'zero': "L" => 'not available'. It should be limited to the cases when an event is not possible to occur, e.g. a sub-sector does not exist or it is legally restricted for a sub-sector to enter into PPPs or to provide guarantees or loans. It should be reported if an event can occur but has not been yet observed. It should be reported if the event is observed but the data are not available

8 In most cases, they were applied accordingly and allowed for presenting the results in a harmonised and comparable manner. Eurostat highly appreciated the efforts undertaken by Member States to streamline the reporting. However, in some specific cases, there is still a need for further harmonisation of reporting conventions and/or further specification of Eurostat guidelines. In particular, it concerns the distinction between the reporting of "0" => 'zero' and the reporting of "M" => 'not applicable'. The issue is mostly relevant for the data on standardised guarantees and off-balance PPPs. In case of no standardised guarantees observed, most countries apply the reporting of "0" => 'zero', although there are some countries that use the convention "M" => 'not applicable'. A similar situation is observed for reporting of off-balance PPPs: in case there is no off-balance PPPs observed, most countries report "0" => 'zero' whereas one country uses "M" => 'not applicable'. Despite being limited, these differences lead to the difficulties for users in comparing and analysing the results. In the 2015 Eurostat publication, this was clarified in the country specific footnotes accompanying the data. Nevertheless, for the next data collections, Eurostat would like to assure the full harmonisation of reporting convention used by the Member States in the data transmission to Eurostat. Eurostat recommends that the same approach should be used for the publication of data on a national basis. In the context of the above, Eurostat would like to propose streamlining the reporting conventions as follows: Reporting conventions (Eurostat proposal) "M" => 'not applicable': "0" => 'zero': "L" => 'not available'. It should be limited to the cases when the subsector does not exist. It should be reported if an event has not been yet observed. It should be reported if the event is observed but the data are not available. In brief, the reporting of "M" => 'not applicable' would be limited to the cases where a specific sub-sector (usually state or social security funds sub-sector) does not exist in a country. For all other cases, the Member States should report a relevant value, "0" => 'zero' or "L" => 'not available'. In Eurostat's view, this approach is in line with the aim of the data collection on contingent liabilities and potential obligations that is supposed to present the potential (maximum) fiscal exposure of the government resulting from the specified transactions. In practice, it would mean that if there are no standardised guarantees or off-balance PPPs observed in a country, the reporting of "0" => 'zero' should be applied, which is more informative to the users and increases comparability of the results across countries

9 6 Revision policy 6.1 Supplement on contingent liabilities and potential obligations to the EDP related questionnaire To recall, the Supplement is provided annually to Eurostat by the end December of each year (N). It collects the data for government guarantees, off-balance PPPs and non-performing loans for four past years N-1, N-2, N-3 and N-4. In the annual transmission of the Supplement by the end of December of year N, the data for the year N-1 is provided for the first time, whereas the data for the years N-2, N-3 and N-3 can be revised. Supplement (guarantees, offbalance PPPs and nonperforming loans) Data transmission by the end of year N Years covered by the Supplement N-1, N-2, N-3, N-4 Historical years in the Eurostat database N-5 and older Data transmission by the end of 2015 Data transmission by the end of , 2013, 2012, , 2014, 2013, , 2010 Revisions of years covered by the Supplement The revision frequency of the data provided in the Supplement on contingent liabilities and potential obligations to the EDP related questionnaire was discussed in the EDPS WG meeting in December The EDPS WG discussed the issue of more frequent revisions than the annual revision possible during the regular data transmission by the end of the year N. The possibility of additional revisions during the year N+1 (i.e. at the EDP notifications) was explored with the aim of assuring the consistency with the EDP Tables and the EDP related Questionnaire. However, taking into account the reporting burden involved, it was concluded that the approach of annual revisions during the regular data transmission by the end of December N should be followed. Nonetheless, in case of major errors encountered, the dataset could be revised by the Member States between the regular data transmissions. Revisions of historical years During the 2015 data collection, a new issue linked with the revisions was encountered. It concerns the revisions of historical years, i.e. the years not covered by the regular transmission of the Supplement (N-4, N-5, etc.). In its database, Eurostat publishes all the years reported by the Member States with the aim of building up time series. At the moment, the years are made available in the Eurostat database. For the years , the data was provided / revised during the 2015 data collection. The data for the year 2010 (historical year reported in the data transmission - 9 -

10 by the end of 2014) was not revised for most of the countries as this year is not covered by a four-year period reported in the Supplement. Two countries (Germany and the Netherlands) voluntarily provided the revised data for the year Taking the above into account, for most of the counties, Eurostat introduced a flag b-break in the database to accompany the 2010 results. In the context of the above, Eurostat would like to encourage and inform the Member States about the possibility of sending the revised historical data to Eurostat alongside with the regular data transmission by the end of December N. The revised data will be uploaded to the Eurostat database and the flags will be adjusted. 6.2 Questionnaire on government controlled entities classified outside general government Regarding the questionnaire on government controlled entities, Eurostat proposes that as a general rule the data should not be revised. Only in exceptional cases of important backwards revisions due to improvement in data coverage or detection of errors, a revised questionnaire from the previous transmission could be transmitted at the end of the year, at the same time with the current version of the questionnaire. For the liabilities of government entities, during the 2016 transmission, data was revised backwards for Belgium, Luxembourg and the Netherlands. The reasons for the revisions in these countries was that for Belgium and Luxembourg, the coverage has been significantly extended including also the liabilities of public corporations involved in financial activities, which were previously not available. For the Netherlands, the completeness of the list of government controlled entities classified outside the general government sector has also improved. 7 Follow-up on the consolidated data on liabilities of government controlled entities classified outside general government 7.1 Overview of replies to the questionnaire on consolidated group accounts As a result of the discussions held during the EDPS WG in December 2015, Eurostat sent on the 16th of December a questionnaire on consolidated group accounts to the Member States. The aim of the questionnaire was to gather information about the availability of data on consolidated group accounts between public parent companies and their subsidiaries in the individual Member States. The reason for investigating the availability of such data was that, in the context of the first publication of data on liabilities of government controlled entities, some concerns were raised by Member States as well as by users regarding a possible overestimation of the amount of liabilities of public corporations due to non-consolidation. Eurostat launched a questionnaire on consolidated group accounts and proposed at the same time a pilot exercise, on a voluntary basis, for those countries which were able to provide consolidated group accounts in order to capture and analyse the impact of consolidation on the data. 23 countries replied to this questionnaire. The assessment of the questionnaire shows that: 1. Thirteen countries (AT, HR, LU, LT, LV, IE, FI, MT, PT, RO and SE) have such data available and were/are willing to participate to the pilot project initiated by Eurostat in January

11 2. Two countries (UK and NL) have provided consolidated group accounts 3. Eight countries (BG, CZ, DE, DK, FR, PL,SI and SK) do not have data available for consolidated group accounts 4. One country (ES) have such data available but do not have the time and resources to provide data on consolidated group accounts within the framework of the pilot project Five countries (Austria, Ireland, Finland, Lithuania and Sweden) accepted to participate to the first pilot exercise launched in January 2016 and provided, additionally to the nonconsolidated data, also intragroup consolidated figures. The analyses of the two datasets, corresponding to the same reference year (2014), show that the impact of intragroup consolidation on the total amount of liabilities varies among the countries. For Austria, the total amount of liabilities at intragroup consolidated level, expressed in percentage of GDP, is slightly higher than the unconsolidated amount of liabilities (34.5% for the consolidated and 33.8% for the unconsolidated data). This might possibly be explained by different reporting standards used for group accounts' financial statements. For Lithuania, the same pattern as in Austria is observed. The total amount of liabilities at intragroup consolidated level is slightly higher that the unconsolidated amount of liabilities (7.38% for the consolidated and 7.05% for the unconsolidated data). For Ireland, Finland and Sweden, the amount of unconsolidated liabilities is higher than the consolidated amount and, as expected, there is a clear overestimation on the total amount of liabilities. This overestimation is significant for Sweden, where the total amount of liabilities at intragroup consolidated level is much lower that the unconsolidated amount of liabilities (37.8% for the consolidated and 47.11% for the unconsolidated data). For Ireland and Finland the corresponding figures are % for the consolidated and 65.5% for the unconsolidated data for Ireland and 33% for the consolidated and 37.78% for the unconsolidated data for Finland. 7.2 Way forward Eurostat proposes to continue this voluntary exercise for the next data transmission for January 2017, for those countries which already expressed their availability to participate to this project. Eurostat considers publishing these new data on Eurostat's website, in addition to the non-consolidated data. Eurostat would also like to extend the metadata information on consolidated data for those countries providing the information and to elaborate in a publication on the impact on the consolidation of the total amount of liabilities based on the data provided by the Member States. The liabilities of government controlled entities will continue to be reported and published on a non-consolidated basis as agreed with the Member States in the framework of the 2012 Task Force. The pilot exercise and the new consolidated data provided by the Member States should be seen as a voluntary and complementary exercise in order to get a picture of the effect of consolidation on the amount of liabilities. 8 The 2016 round of the Eurostat data collection on contingent liabilities and potential obligations This is to inform the EDPS WG members about the time-schedule of publication of data on contingent liabilities and potential obligations in The Eurostat data collection will take place by the end of December The relevant data and metadata templates will be sent out

12 to the countries by the end of September At a national basis, the Member States should publish the data on guarantees, off-balance PPPs and non-performing loans by the end of October 2016 whereas the data on liabilities of government controlled entities classified outside general government and the participation of government in the capital of corporations should be made available by the end of December Eurostat attaches major importance to the availability of metadata information related to the data on contingent liabilities and potential obligations. To this end, as agreed during the EDPS WG in June 2015, Member States are asked to complete a metadata template in order to provide information in a harmonised way. The information gathered is used, inter alia, for communication with the users via country specific footnotes and Eurostat publications. In order to further improve the availability of information, Eurostat will include additional questions on the undertaken revisions in the metadata template. They will aim to gather the information about the reasons of major revisions. 9 Conclusions Since the first data collection in 2014, significant progress has been achieved by the Member States with regard to the completeness and the coverage of the data. A number of countries reviewed and revised the figures in order to better align to the Task Force guidelines. Eurostat appreciates the efforts undertaken by the Member States and welcomes the results achieved. For the purpose of Eurostat's publication, all countries but one (Cyprus) provided replies to the 'Supplement on contingent liabilities and potential obligations to the EDP related questionnaire'. For a majority of countries, complete and good coverage data are available for most of the indicators. Nevertheless data coverage is not yet fully exhaustive for some Member States. In general, the comparability of data for government guarantees, nonperforming loans and off-balance PPPs can be considered as satisfactory. For the liabilities of public corporations, the coverage has been significantly extended since the last transmission. Nevertheless, there are still very few Member States for which the data reported do not include the liabilities of financial institutions and/or the liabilities of units controlled by local government. In particular, the availability of the metadata accompanying the provided figures has largely improved. It helped Eurostat and the users to better understand the provided figures and the limitation in their analysis and comparisons. Concerning the data publication on a national basis, a majority of the Member States follow the Task Force guidelines and make available relevant information according to the agreed methodology and templates. However, there are two countries (Greece and France) that have not yet published any indicators and a few Member States that provide data only on some indicators. The availability of data for the last reference year (2014) should also be improved in some cases. Eurostat observed that, except for limited cases related mainly to the liabilities of public corporations, there are no major discrepancies between data provided to Eurostat and data published at. The countries where the differences are noticed are asked to investigate the issue and align the data published. In specific cases, bilateral discussion could take place in order to find a way how to better explain to the users the differences in the data

13 In the context of the data collection in 2016, Eurostat encourages the Member States to closely review their current compliance with the Task Force's guidance and, if needed, to continue efforts necessary to improve the completeness and coverage of the data. Moreover, Eurostat proposes solutions to the outstanding issues related to revision policy and reporting conventions. In particular, it proposes to simplify the reporting conventions and to limit the use of "M" => 'not applicable' to the cases where a sub-sector is not identified. Concerning the revision policy, Eurostat re-confirms the approach of the annual revisions for the years covered by the "Supplement on contingent liabilities and potential obligations", and it encourages also the Member States to revise the data for historical years (N-5 and older) for guarantees, off-balance PPPs and non-performing loans. Eurostat has assessed the replies of the questionnaire on consolidated group accounts data for liabilities of government controlled entities and has presented a brief overview of the output of the pilot project. Eurostat proposes to continue the voluntary exercise on the additional data collection on consolidated group accounts for the next data transmission for January 2017, for those 13 countries which already expressed their availability to participate to this project. The EDPS WG members are invited to: - take note and comment on the updated Eurostat assessment of compliance with the Task Force guidance in relation to the publication of data on contingent liabilities and potential obligations, at EU level and on a national basis; - inform Eurostat about any adjustments needed in the assessment; in particular due to new data published or/and incorrect information included in the report. If relevant, the comments should be sent until 30 June 2016 to estat-d1-secretariat@ec.europa.eu with copy to the Desk Officer; - take note that Eurostat encourages the revisions of 'historical years' with regard to the data on guarantees, off-balance PPPs and non-performing loans in order to lengthen the availability of a comparable data series; - express their opinion on the Eurostat proposal for the simplified reporting conventions for the data provided in the 'Supplement on contingent liabilities and potential obligations'; - participate to the pilot project on the collection of data on consolidated group accounts for liabilities of government controlled entities classified outside general government. Eurostat, Unit D.1 Excessive deficit procedure and methodology Eurostat, Unit D.2 Excessive deficit procedure (EDP)

14 Annex 1. Total general government contingent liabilities and non-performing loans in EU Member States, 2014 (% of GDP*) Government guarantees** One-off Standardised Total Liabilities related to off-balance publicprivate partnerships (PPPs) Liabilities of government controlled entities classified outside general government*** Entities involved in financial activities Entities involved in other activities Total Non-performing loans (government assets) Belgium a L Bulgaria Czech Republic a M Denmark a Germany a 16.4 M 16.4 M Estonia a M Ireland a Greece a Spain a 12.8 L France a L Croatia a L Italy a Cyprus a L L Latvia Lithuania Luxembourg Hungary a Malta a 16.8 M Netherlands a Austria a Poland a Portugal a 7.1 M Romania Slovenia a Slovakia L Finland a Sweden a United Kingdom a a See Country specific footnotes available at Eurostat website L: data not available M: not applicable

15 Annex 2. Compliance with the Task Force guidance in relation to the publication of data on contingent liabilities for all sub-sectors of general government (status February 2016) The analysis is based on data and metadata provided by Member States to Eurostat by end December It could be subject to changes once new information is available to Eurostat. The assessment 'partly' concerns the cases where the data is not complete for all subsectors or years and/or the data coverage is not exhaustive. Assessment 'no' relates to the cases where the indicator is not published. The related remarks are provided in the' comments' box; additional information can be found in the Country specific footnotes. The last column of the table indicates whether the assessment also applies to the data published by Member States at national level. Table 1. Government guarantees, MS Data provided to Eurostat Data complies with Task Force guidance in terms of completeness and coverage Central State Local Comments BE yes yes yes partly Data coverage is not exhaustive for yes local government. BG yes yes - yes yes CZ yes partly - partly DK yes yes - partly Data coverage for one-off guarantees is not exhaustive. Data for enterprises included in general government are not fully complete; however, all significant entities are covered. Data for guarantees for local government are available only from 2013 onwards. Data are not available for standardised guarantees of local government. 1) Owing to the unavailability of counterpart information in the source data, only large guarantees issued to other government entities in the context of government interventions during the financial crisis, and guarantees linked to major rerouting cases, have been subject to consolidation. Assessment applies also to publication at yes yes DE yes partly partly partly 2) Data include neither institutional commitments of government bodies in the form of the so-called 'Anstaltslast', provided to public law institutions, nor the so-called 'Gewährträgerhaftung'. Related risks to government budgets have, however, been disclosed as government potential obligations, within the total amount of liabilities of government controlled entities classified outside the general government sector. yes EE yes yes - yes yes IE yes yes - partly yes, however the templates Data not exhaustive for the one-off publication does guarantees of local government. follow the agreed EL yes yes - partly Data coverage is not exhaustive for data not published

16 MS Data provided to Eurostat Data complies with Task Force guidance in terms of completeness and coverage Comments Assessment applies also to publication at Central State Local one-off guarantees of local at government. Data not available for standardised yes ES yes partly partly partly guarantees, however if such guarantees exist the amounts are negligible. Data are not available for local data not published FR yes yes - no government and social security funds at sub-sector. HR yes yes - no Data on one-off guarantees not yes available for local government. Data related to stock of one-off yes guarantees of local government refer to both standardised and one-off IT yes yes - yes guarantees as, at the moment, data source doesn't distinguish between the two different kinds of guarantees. CY partly partly - partly Data for 2014 not available. yes LV yes yes - yes yes LT yes yes - yes yes LU yes yes - yes yes Data include guarantees provided to yes HU yes partly - partly units classified outside budgetary central or local government, but classified inside general government. MT yes yes - - yes NL yes yes - yes yes AT yes yes yes yes yes PL yes yes - partly Data not available for standardised yes guarantees of local government. PT yes yes - yes yes RO yes yes - yes yes SI yes yes - yes yes SK yes yes - yes yes Guarantees provided by yes FI yes yes - partly municipalities, joint municipal authorities and the Åland Government to government units are not consolidated because of lack of data SE yes yes - yes yes UK yes yes - yes yes - : Non applicable

17 Table 2. Off-balance PPPs, MS Data provided to Eurostat Data complies with Task Force guidance in terms of completeness and coverage Central State Local SSF Comments Assessment applies also to publication at national level BE yes yes yes yes yes yes BG yes yes - yes - CZ yes yes - yes yes yes DK yes yes - yes - yes All assets of PPP yes DE projects are recorded on the government's balance sheet. EE yes yes - yes yes yes IE yes yes - yes - yes EL yes yes - yes yes ES yes yes yes yes yes yes FR yes yes - yes yes HR yes yes - yes yes yes IT yes yes - yes yes yes CY partly partly - partly partly Data for 2014 not available. yes LV yes yes - yes yes yes LT yes yes - yes - yes LU yes yes - yes yes yes HU yes yes - yes - yes MT yes yes yes NL yes yes - yes yes yes Data not available for yes AT yes no yes yes yes central government. PL yes yes - yes yes yes PT yes yes - yes yes yes RO yes yes - yes yes yes Data not available for yes SI yes yes - partly yes local government. SK yes yes - yes yes yes FI yes yes - yes - SE yes yes - yes yes UK yes yes - - no - : Non applicable Data coverage can be non-exhaustive. Data coverage can be non-exhaustive. Data coverage is not exhaustive for local government. yes yes yes

18 Table 3. Non-performing loans, MS Data Data complies with Task Force guidance in provided terms of completeness and coverage to Eurostat Central State Local SSF Comments Assessment applies also to publication at BE no no no no no Data not available. BG yes yes - yes yes yes CZ yes yes - yes - yes, however there is discrepancy between data provided to Eurostat and published at DK yes yes - yes - yes DE yes yes yes yes yes yes EE yes yes - yes yes yes IE yes yes - yes - Provisional data for yes 2014 Data are not available for the social security EL yes yes - yes no funds sub-sector; however the amounts are not expected to be significant. Data not available for yes ES yes yes yes no yes local government, however the amounts of the loans granted by local entities are not significant FR no no - no no Data not available. HR no no - no no Data not available. Data not available for yes IT yes yes - no no local government and social security funds. CY no no no no Data not available. LV yes yes - yes yes yes LT yes yes yes LU yes yes - yes yes yes HU yes yes - yes yes yes, however there is discrepancy between data provided to Eurostat and published at MT yes yes - yes - yes NL yes yes - yes yes yes AT yes yes yes yes yes yes PL yes yes - yes yes yes PT yes partly - partly partly Data coverage is not yes exhaustive. RO yes yes - yes yes yes SI yes yes - yes yes yes SK no no - no no Data not available. data published at national level FI yes partly - no no Data coverage is not yes exhaustive for central

19 MS Data Data complies with Task Force guidance in provided terms of completeness and coverage to Eurostat Central State Local SSF Comments government. The stock of nonperforming loans of the central government consists of loans managed by the State Treasury. Data is not available for local government and social security funds. SE yes yes - yes yes yes UK yes yes - no - - : Non applicable Data not available for local government. Assessment applies also to publication at yes

20 Table 4. Liabilities of government controlled entities classified outside general government sector (public corporations), MS Data provided to Data complies with Task Force guidance in terms of completeness and coverage Eurostat Central State Local SSF Comments Assessment applies also to publication at national level Data coverage is not yes BE yes yes yes partly - exhaustive for local government. BG yes yes - yes - yes CZ* yes yes - yes - Yes, however there is a discrepancy between data provided to Eurostat and published at DK yes yes - yes - Yes, however there is a discrepancy between data provided to Eurostat and published at 1) For public deposit-taking yes DE* yes partly partly partly partly monetary financial institutions, the reported data represents the stock of year-end liabilities, at nominal value, as presented in business balance sheets (not group-consolidated) excluding own resources (equity, reserves and provisions). For other public corporations, debt owed to the public and non-public sector has been reported. 2) The aggregates include solely liabilities of resident government controlled entities. Liabilities of lossmaking non-financial corporations might be underestimated owing to the unavailability of the information on operating profit/loss for a number of entities. 3) The ESA 2010 sector delimitation rules have been followed only partially, since the data are mainly based on surveys which are not revised backwards. The amounts include debt of public holdings which were reclassified to the general governments sector in the summer 2015 revision of the German national accounts. EE yes yes - yes - yes, however there is a

21 MS Data provided to Data complies with Task Force guidance in terms of completeness and coverage Eurostat Central State Local SSF IE* yes yes - no Comments Data cover only public corporations controlled by the central government. EL yes partly - partly - Data coverage is not exhaustive for the local government sub-sector. The liabilities of financial institutions that have received significant financial support from the Greek government are not included, as the extent to which government is controlling these entities is being assessed. ES yes yes yes yes - yes Assessment applies also to publication at national level discrepancy between data provided to Eurostat and published at yes, however there is a discrepancy between data provided to Eurostat and published at FR* yes yes - yes - HR yes yes - yes yes yes IT yes yes - yes - yes Data reported refer to 2012 and are not exhaustive. CY* partly partly - partly - Data for some public corporations are not available and the liabilities of public corporations involved in financial activities are not included. LV yes yes - yes - yes LT yes yes - yes - yes, however there is a discrepancy between data provided to Eurostat and published at LU yes partly - partly - Data reported partially include the liabilities of public corporations involved in financial activities. HU yes yes - yes - yes, however there is a discrepancy between data provided to Eurostat and published at Data reported do not yes MT yes yes - no - include public corporations controlled by the local government sector. NL yes yes - yes - yes AT yes yes yes yes - data for 2014 are not published at PL yes yes - yes - yes PT yes yes - yes - yes RO yes yes - yes - yes SI yes yes - yes yes yes

22 MS Data provided to Data complies with Task Force guidance in terms of completeness and coverage Eurostat Central State Local SSF Comments Assessment applies also to publication at national level SK yes yes - yes - yes, however there is a discrepancy between data provided to Eurostat and published at FI yes yes - yes - yes SE yes yes - yes - yes UK yes yes - yes - yes * For Czech Republic, Germany, France and partly Italy, data reported refer to For Cyprus, data reported refer to : Non applicable

23