AnaCredit and RIAD. BIS-BI-ECB Regional Seminar 2017 Bali, Indonesia March 2017

Transcription

1 ECB-RESTRICTED Jean-Marc Israël Head of Division Monetary and Financial Statistics BIS-BI-ECB Regional Seminar 2017 Bali, Indonesia March 2017

2 Rubric Contents 1 Analytical Credit Dataset (AnaCredit) 2 Register of Institutions and Affiliates Database (RIAD) 2

3 Rubric Financial crisis new tasks and new data needs Phase A - post Lehman Brothers Sept 2008 banks rescue, government guarantees Who holds GR, IE, PT government debt? How exposed are euro area investors? Phase B May 2010 Sovereign debt crisis and again banking crisis Phase C - May 2012 Banking Union announced How big are EU financial interlinkages? EU Financial Supervision ESAs/ESRB - Nov 2010 ECB s banking supervision Nov 2014 How healthy are euro area banks? Single Resolution Mechanism Jan

4 The Rubric ESCB response: holistic approach to data reporting Integration managing different areas of statistical and supervision information as parts of a single system Long-term harmonisation of practices, methodologies and processes currently followed for data production 4

5 Rubric AnaCredit - multiple ESCB tasks AnaCredit granular data on credit to support ESCB tasks AnaCredit will cover the need for granular data on credit and credit risk, whose current availability, e.g. from Central Credit Registers, is rather incomplete and heterogeneous across EU countries It will bring increased coverage, and especially harmonisation in underlying concepts and definitions standardisation Supporting a wide range of users Monetary policy analysis and operations Risk control and collateral management Financial stability Micro/macro supervision Economic analysis and research Statistical production for multiple analyses and purposes Access to credit Credit supply constraints Financial constraints Analyses of collateral issues Stress-tests Off-site and on-site supervision Exposures to sectoral risks Credit risk analysis Risk control measures Cross-border credit risk Transmission channels of MP New statistics 5

6 Rubric AnaCredit uses Example of Monetary policy ECB s monetary policy requires a vast amount of data Market policy decisions affect economic activity and inflation through several channels the transmission mechanism of monetary policy impact depends on the nature of the business and the size of the enterprise currently, only broad quantification of the effects of market policy available AnaCredit will help identify and asses developments in credit demand and supply AnaCredit data is crucial to assess the extent to which b/s conditions of banks or debtor are influencing credit expansion Help assess implication for price stability and identify adequate policy actions 6

7 Pros Rubric of collecting granular information ECB-UNRESTRICTED FINAL Aggregated data individual institutions combine their positions and transactions into a limited set of statistics/indicators well established, broad coverage, relatively limited costs of reporting long lead time (years) to address ever changing policy needs Granular data Surveys limited coverage, low comparability often run ad hoc to address urgent needs, no/limited time series expensive, for both reporting agents and central banks Aggregated data can be derived from granular data instrument-by-instrument and counterparty-by-counterparty more informative: more sophisticated analysis, address heterogeneity flexible: meet ever-changing user needs in a timely manner cost-effective: high set-up costs more than offset by running savings in the medium-term, also for banks! 7

8 Rubric AnaCredit: why to move towards more granular data? Currently: exchange of data under a Memorandum of Understanding (MoU), at a borrower level, where indebtedness is above 25,000 Purpose: allow reporting institutions to have a wider picture of indebtedness of their current and potential borrowers, and also for supervisory purposes Task Force on Analytical Credit Datasets, incl. National Central Banks with no granular credit data Work re. harmonisation of concepts, definitions and identifiers 88 attributes standardised Establish a long term framework legal: solution for data sharing / access rights / confidentiality organisational: IT, incl. cost aspects, and procedural issues Aim: to collect granular credit and credit risk data to be shared, incl. attributes on lenders, borrowers, credits and related methodological aspects

9 Rubric AnaCredit uses Towards better understanding of the EU banking sector Non-Performing Exposure ratio is particularly high (10%) in the case of non-financial corporations NPE ratios vary depending on the size of the banks data show that smaller banks struggle with higher levels of non-performing loans, 18% of total loans, compared to 9% of total loans in medium banks and 4% of total loans in large banks Source: EBA 2015 EU-wide Transparency Exercise Report 9

10 Rubric AnaCredit uses More data on SMEs AnaCredit will close existing data gaps on SMEs Covering the whole range of enterprises (thanks to the threshold of 25,000), Leveraging on granular data typically produced by banks for their internal risk management processes Enable identification and assessment of developments in credit demand and credit supply across all SMEs in the euro area in a timely manner Enhance monitoring of credit developments at the borrower-lender level Help assess the credit risk that banks associate with firms Assist in assessing the effectiveness of measures to improve SMEs access to credit (Do capital reliefs for banks make it easier for SMEs to get credit?)

11 Rubric Analytical Credit Dataset - AnaCredit Important building block of an integrated approach Content Granular information (individual, micro) Credit and credit risk data Harmonised concepts and definitions Purpose Multipurpose Monetary policy, macroprudential supervision, risk management, research, and statistics Participants Euro area countries Other EU countries also invited to participate Sourcing / Governance ECB regulation ECB and participating countries NCBs to collect information from Reporting agents Shared platform state-of-the-art IT solution 11

12 AnaCredit Rubric benefits & costs for NCBs & reporting agents Overall benefits Cross-country harmonised information on credit instruments/attributes Enhancement of the credit institutions' assessment on the creditworthiness, in particular of cross-border potential borrowers, also via feedback loops Overall minimisation of reporting burden (after set-up) use of granular credit and credit risk information to enhance existing statistics or produce new ones stability of requirements over time, notably avoid running ad hoc surveys Cost considerations High investment for both NCBs and reporting agents, mainly due to accounting and risk-based data, high volumes and required data quality management Medium to neutral running costs, thanks to e.g. less ad hoc requests, possible discontinuation, in part or in full, of existing aggregated reporting Net benefits reached when AnaCredit allows discontinuing existing CCRs 12

13 External Rubric use of AnaCredit The ECB will not offer credit scorings regarding e.g. individual SMEs In general, no granular (loan-/borrower-/lender level) information will be disclosed (whether anonymised, or not) The ECB will not distort competition in the market of credit information AnaCredit will set standards on a number of important variables/dimensions market-led initiatives could build upon these Methodological support could be provided to such initiatives 13

14 Rubric AnaCredit New IT application Develop a shared platform to serve multiple purposes with core set of data; according to harmonised concepts and definitions pull, store, process and disseminate shared analytical credit data on an EU-wide scale, protecting confidentiality as necessary Allowing the availability of comparable key granular credit and credit risk data to a wide range of users for policy-relevant and analytical purposes set-up and maintenance of differentiated levels of access to data across the ESCB, depending on recognised uses connection with other ESCB applications (Business register; data on Securities) as well as serving Single Supervisory Mechanism, e.g. to credit exposures 14

15 Rubric Preparing for implementation of AnaCredit Eurosystem supports banks in the implementation of AnaCredit Manual To clarify AnaCredit requirements via documentation To provide Concrete additional background and guidance cases/examples Cooperation via BIRD Cooperation with the banking industry within Banks Integrated Reporting Dictionary (BIRD) BIRD provides a representation of AnaCredit as set in the Regulation from the business point of view Commercial banks participate on a voluntary basis 15

16 Rubric Conclusions (1/2) The availability of granular credit and credit risk data is very important for the post-crisis central banks policy making Aggregate/survey data with insufficient accuracy and breakdowns; a key issue is to allow rebuilding aggregates ex post! Central bank s non-standard monetary policy measures and the close link between monetary and macroprudential policies increase the importance of such data 16

17 Rubric Conclusions (2/2) Setting-up granular data sources is costly; but allows to drill-down from aggregates to identify distress in markets or institutions, and better analyse developments Accurately signal problems in the monetary policy transmission channel In the case of SMEs, granular data on credit will facilitate evaluation of financing conditions and assist policy-making Data re-used for different purposes, including feedback loops to banks 17

18 Contents Rubric 1 Analytical Credit Dataset (AnaCredit) 2 Register of Institutions and Affiliates Database (RIAD) 18

19 Rubric Register of Institutions and Affiliates Database (RIAD) A register is key to identify the reporting population(s) and stratify it/them and classify counterparts (by respondents) (financial) groups structures and its value added is even greater if it becomes multi-purposes, e.g. Close links for monetary policy operations Banking and Insurance group structures for financial stability analysis Identification of Payment service institutions, and Supporting the ECB Banking Supervision 19

20 Rubric RIAD as reference dataset for institutions The Register of Institutions and Affiliates Database (RIAD) is operated by ECB DG-Statistics and accessible to all ESCB members designed to hold reference data of (financial) organisational units ensuring that information on statistical reporting population is complete, timely, accurate, and homogenously defined The current system went live in May 2013 and serves business needs Statistics Market operations Financial stability Payment systems oversight, and Microprudential supervision 20

21 Rubric RIAD management and service Currently RIAD is used to maintain four lists of financial institutions ( 9,000 Monetary Financial Institutions MFIs Credit Institutions and Money Market Funds 55,000 Investment Funds IFs 3,000 Financial Vehicle Corporations FVCs 3, 500 Insurance Corporations ICs 7,000 Payments Statistics Relevant Institutions PSRIs RIAD provides simple lists management And, far beyond, also offers all functions of an extended business register RIAD will support the AnaCredit dataset in providing relevant information on lenders and all their counterparts, borrowers or guarantors 212

22 Rubric RIAD Main features Comprehensive data model, incl. reference data on units plus relationships among them Strict separation of the transactional and reporting aspect Multiple source management Historisation of data Elaborate access management incl. the distinction of special roles and different levels of confidentiality A functionality added (for SSM) is on handling different types of reporting 22

23 RIAD RubricMain features: main stakeholders Other initiatives 23

24 Rubric RIAD Modelling of group structures RIAD allows processing of attributes describing relationships between units which are the necessary building blocks to construct group structures Operationally this is achieved via a bottom-up approach, i.e. via individually loading the known / necessary members ( nodes ) and their ownership relationships ( edges ) to construct a group ( graph ) In the medium-term it is expected to receive the underlying relationship information from various sources, e.g. from Statistics departments Supervisory authorities on institutional units on supervised entities Market operations departments on various relationships, or Market and commercial sources 24

25 Rubric RIAD Group views Currently RIAD is generating two types of groups on basis of the available ownership/control information Type A - based on control Type B - based on pure ownership (incl. 50% and less) 25

26 Rubric RIAD Lessons learned According to experience gained, the challenge is to Offer a model with a maximum of (technical) flexibility in mapping real phenomena to formal concepts Organise an adequate governance to maintain and update the dataset servicing various user needs, avoiding inconsistencies and conflicts Satisfy different views on group structures simultaneously Identify and classify counterparties, which implies to - monitor any corporate actions within the reporting population - navigate through the sample over time, and - synchronise publication of the authoritative lists 26

27 Rubric Relationships in the data model Comprehensive data model, incl. reference data on units plus relationships among them A. Identification such as identifiers, address, etc. B. Stratification such as industrial activity, geographical allocation, etc. C. Demographic developments such as birth/closure date, split, merger, etc. D. Relationships between units such as ownership, control, management, etc. D. Relationship variables between units 1. Organisational relationships within an enterprise Relationship between the legal unit and the enterprise that it operates. 2. Relationships between legal units forming an enterprise group Link based on the concept of 'control' Link based on the concept of 'capital share', 'voting rights' 3.Other relationships Link between the 'originator' and 'FVC' Link between the 'management company' of 'FVC / Investment Fund' Link between the 'non-resident branch' and 'headquarter' Link between the 'sub-fund' and 'umbrella fund' 27

28 Rubric RIAD High level architecture 28

29 RIAD Rubricmain features - System 29

30 Questions? Rubric Thank you! Questions? 30

31 RESERVE: Rubric Some granular information on credit is already available Credit Reporting Systems in Europe (indicative list) Source: European Credit Research Institute 31