150 likes | 173 Views
ECB-UNRESTRICTED. FINAL. Gemma Agostoni Josep Maria Puigvert ECB Ronald Rühmkorf Plamen Petrunov BNB. Statistical findings on the new Money Market Statistical Reporting. Conference of European Statistics Stakeholders Budapest | 20 - 21 October 2016. Overview.
E N D
ECB-UNRESTRICTED FINAL Gemma Agostoni Josep Maria Puigvert ECB Ronald Rühmkorf Plamen Petrunov BNB Statistical findings on the new Money Market Statistical Reporting Conference of European Statistics Stakeholders Budapest | 20 - 21 October 2016
Overview ECB-UNRESTRICTED FINAL 1 MMSR in a nutshell 2 The data in details: variables and attributes 3 Implementation and Standardisation 4 Validation procedures 5 Reporting since April 2016 6 MMSR additional data quality challenges Conclusion 7 Initial statistical findings on the new MMSR
MMSR in a nutshell (1/3) ECB-UNRESTRICTED FINAL • The money market changed dramatically in the wake of the crisis • modifiedmarket patterns and fragmentation • financial innovations and regulatory changes affecting participants’ behaviourand markets salient features • Consequences • data gaps hinderedto interpret on a very timely basis money market-related elements of the monetary policy transmission mechanism • Proposed solution • collect granular statistical information on money markets to support and better inform monetary policy conduct in euro area • New legal act • MMSR Regulation (ECB/2014/48)adopted on 26.11.2014 Initial statistical findings on the new MMSR
MMSR in a nutshell (2/3) ECB-UNRESTRICTED FINAL • Transaction-by-transaction data on four money market segments: Secured:borrowing/lending of cash secured with a collateral which safeguard the lender against borrower's default Unsecured: borrowing/lending of cash backed only by the borrower’s creditworthiness FX Swaps: purchasing/selling of foreign exchange against EUR OIS: exchange of EONIA rate against a fixed interest rate • Daily frequency: Reporting by 7 a.m. on the day following the trade date and covering all money market trades conducted during the previous day • Timeline: Full reporting as of 1 July 2016, following a 3-month interim period Initial statistical findings on the new MMSR
MMSR in a nutshell (3/3) ECB-UNRESTRICTED FINAL • Reporting population • The biggest 52 euro area MFIs with a balance sheet larger than 0.35 % compared to the total balance sheet of all euro area MFIs • The high correlation between the bank balance sheet and the money market activity ensured achieving coverage of ca. 85% of the money market • Highest money market concentration - around 80% -for reporting agents located in Germany, France, Italy and Spain • Counterparty coverage (with whom reporting agents are trading) • MFIs, i.e. interbank market • Central Banks andGeneral Government • Other FinancialsCorporations, incl. Investment funds, Insurance, Securitisation • “Wholesale” Non-Financial Corp.s classified under Basel 3 LCR framework Initial statistical findings on the new MMSR
Main money market variables • ValuationInterest rates | Basis Point Spread | Price | Swap points • TurnoverTransaction Nominal Amount Counterparty identification • LEI: Legal Entity Identifier (for interbank and CCPs) ISO 17442, or • Sector, via ESA2010 code, andLocation, via ISO 3166 country code Maturity structure • TradeDateSettlementDateMaturityDate Contract creation Payment Termination Collateral identification • ISIN: International Securities Identification Number, via ISO 6166,or • Asset Type : CFI code, via ISO 10962, and Issuer Sector via ESA2010 code The data in details: variables and attributes ECB-UNRESTRICTED FINAL Initial statistical findings on the new MMSR
Implementation and Standardisation (1/2) ECB-UNRESTRICTED FINAL • Documentation • Detailed Reporting Instructions (published a year in advance)explaining the conceptual and technical specifics of the MMSR framework • An extensive Q&A document further elaborating on the reporting requirements • Clear and well-defined XML - Reporting schemas => Ensuring correct and common understanding among reporting agents • Strong interaction with the industry • 3mock-up data exercises conducted in 2015 to test the understanding of the reporting framework and readiness by the 52 Reporting Banks • Several meetings with the Banking industryconcerned In Dec. 2014, Feb. 2015, Dec. 2015, May 2016 and Nov. 2016 • Fully automated processing: from data transmission to dissemination Initial statistical findings on the new MMSR
Implementation and Standardisation (2/2) ECB-UNRESTRICTED FINAL • High level of standardisation • MMSR framework is fully aligned with the ISO 20022 standard • ISO 20022 Universal financial industry message scheme is a single standardisation approach (methodology, process, repository) • used in many fields of financial reporting, such as payments or securities • Strongly supporting high automation of the MMSR data collectionandprocessing for Reporting Agents as well as for the Eurosystem • MMSR Reporting Instructions fully consistent with all ISO definitions and XML-based reporting messages registered by the ISO • making them globally recognised • Usage of the ISO 20022 standard • Bank of England’s Sterling Money Market Daily (SMMD) – equivalent of the MMSR covering the UK money markets • Secured Financed Transactions-Data Store (SFT-DS) under the SFT Regulation Initial statistical findings on the new MMSR
Data Quality Checks • Provided as a detailed document to the reporting agents to ensure transparency and easiness of the implementation on their side • Applied immediately after data reception • Scope • compliance with formats • consistency of the information • plausibility of the data • Follow-up • automatic, system-generated message to the reporting agent presenting a detailed explanation of the breached DQ Check(s) • the Eurosystem message requires response and corrective measures by the reporting agent ASAP, but not later than 10 days Validation procedures ECB-UNRESTRICTED FINAL Initial statistical findings on the new MMSR
Reporting since April 2016 (1/2) • Evolution April–Oct. 2016 of weekly average nbof transactional records • Over 4,500,000 transactional recordsin total, thus far
Reporting since April 2016 (2/2) Evolution of percentage of accepted, of which with warnings, and rejected transactional records: over 99% of the records accepted and <1% rejected!
MMSR additional data quality challenges (1/3) • Further data quality analysis is also being conducted to identify: • Matching of the two legs of transactions (among 52 reporting agents) per segment to identify possible missing transactions • Currently only checked on a monthly basis • Analysis of the counterparty based on Counterparty sector or the LEI • E.g. transactions reported with the LEI, instead of the sector and location (when not a bank or CCP) • Analysis of foreign exchange forward points, basis point spread, haircut and exchange spot rate to identify wrongly reported fields
MMSR additional data quality challenges (2/3) Outlier detection
MMSR additional data quality challenges (3/3) Wrong reporting around “0” for a particular agent Initial statistical findings on the new MMSR
MMSR: an already positive outcome • Timetable and actual delivery fully on time and above expected quality level • New data collection at veryhigh frequency (daily) and timeliness (day + 1 at 7:00) and granularity ensuring high coverage on money market segments • Standardisation, data validation and active preparationguaranteed a high quality dataset and mitigated the reporting burden on the industry • Full automated processing improving the effectiveness and lowering the costs • A very rich source of information serving many purposes among the Eurosystem, the industry and the public at large Remaining challenges ahead • Further improvement of data quality: in particular outlier detection, clustering and multivariate analysis • External and internal dissemination of the data, fully protecting confidentiality • Futureextension of the reporting population Conclusions ECB-UNRESTRICTED FINAL Initial statistical findings on the new MMSR