Presentation on theme: "Tutorial on Best Practices Presented at the: IX European Banking Supervisors XBRL Workshop & Tutorial In: Paris On: 29 th September 2008 By: Ignacio Boixo."— Presentation transcript:
Tutorial on Best Practices Presented at the: IX European Banking Supervisors XBRL Workshop & Tutorial In: Paris On: 29 th September 2008 By: Ignacio Boixo Member of CEBS XBRL Network
Why Best Practices?
Interoperability in CEBS environment Basel II, IAS/IFRS … EC 2006/48 & 49 … Country 1 Sup 1 Report 2Report Country 3Country 2Country 27 NCB 2FSA 3Sup 27 Report Report National Regulation Transposition into national Legislation European Law 9X,XX% best practices + EU requirements National Implementation XBRL challenge! Global best practices
Stack of definitions LegislationParliament and International Bodies Business GuidelinesBanking Supervisors Data descriptionCEBS COREP&FINEP Networks TaxonomiesCEBS XBRL Network Information Tech. Best PracticesBest Practices Board FormatXBRL Standards Board Communication and Security W3C, ISO…
Standards, Proprietary and Interoperability A Proprietary IT, used in a Proprietary mode, is not expected to be interoperable with others Minitel (pre-Internet hierarchical web) An Standard IT, when used in a Proprietary mode, is rarely interoperable with others Standard.jpg images when containing business reports A Proprietary IT, even used in a Standard mode, only interoperates with its more or less large perimeter. Proprietary features of MS Internet Explorer A Standard IT, when used in s Standard mode, MAY BE a Standard Implementation HTML pages after passing the W3C validation, as Corollary The Standardization MUST BE applied to ALL the Layers of the Stack, as pre-requisite to preserve the Interoperability.
Cross-fertilization in Best Practices XBRL International Best Practices Board Standards Board Software Interoperability Taxonomy Architecture Reporting Processes Project Management XBRL Europe EU Liason Technical WG Marcomm Bylaws WG CEBS Expert Group on Financial Information Subgroup on Reporting XBRL Network FINREP Network COREP Network
Auditability Audit is over a set of facts PRESENTED in a specific view, not over each individual fact XBRL Instance Document contains facts; the presentations may change Auditor Stamp PDF file XBRL instance Spanish Securities Solution
Traceability Who and when has reviewed each fact? XBRL facts are to be used in different reports. How to track past audits? Auditor Stamp Supervisor Stamp 81 Auditor Stamp 42 Auditor StampTax Stamp 13 Auditor StampTax Stamp Supervisor Stamp Tax Stamp
Precision and rounding threshold: Problem XBRL manages the precision mainly with the decimals attribute, indicating the position of the last rounded/truncated digit, ignoring the right-following digits. XBRL interpretation of the actual sequence 4,444, ,444,000.00Decimals=-3. Thousands 4,444,444.00Decimals=0. Integers 4,444,444.44Decimals= 2. Euros and cents 4,444, Decimals= 4. basic points (percentages) 4,444, Decimals=INF. Absolute precision 4,444, ,444, ,444, ,444, ,444, ,444, ,444, ,444,444.44… 4,444,444.44… 4,444,444.44… 13,332, ,333, ,333, ,333,333.33… Where should be the rounding threshold for validations?
Precision and rounding threshold: Consequences Each Supervisor may choose a proprietary precision and, therefore, a rounding threshold Definitions at European Level MUST be developed for all the possible combinations. Formulae exacerbates this consequence. Each cross border filer must obtain a different set of figures, according the precision required by each Supervisor, and therefore a different DataBase. Valid XBRL Instance Documents for a Supervisor will be invalid for other using different precision Interoperability seriously challenged if not an agreed Best Practice. Two decimals and rounding threshold in thousands as an initial discussion basis? The last but not the least: How to deal with the inequation 1.00 > 1/3 + 1/3 + 1/3 when decimals=2?
Precision and rounding threshold: questionnaire XBRL provides two methods of communicating the precision of a numeric fact: precision or decimals attributes. a) Have you specified for your reporting institutions which method is to be used? b) Have you predefined an accuracy of the data to be reported, i.e. in two decimals or thousand or million or as exact monetary amounts?
Precision and rounding threshold: questionnaire c) Do you handle minor rounding variations with a unique threshold or with variable threshold/s? d) Do you use additional solution/s to handle rounding differences? 2 countries explained that margins on calculation rules are specified (NL, PL). 1 country uses a unique percentage that applies to all calculations (FI).
Other tricky points (from COREP/FINREP lists) Definition of 'Zero' (answered by Walter Hamscher on ) As zero As a Blank As Nil (XML) Whitespace and carriage returns in Instance Documents. (discussion on ) Other points not yet raised?
Dimensional Taxonomies Both scenario and segment are dimensional containers, according XBRL 2.1 Specification. Their simultaneous use may complicate even more the dimensional intricacy. Why not use only scenario for dimensional stuff? Topics related to hypercubes Hypercubes structure, All and notAll combinations… Topics related to dimensions/domains/domain-members Canonical expression of Hypercubes, elements and correlated stuff Topics related to presentation relationships on dimensional artefacts. US-GAAP taxonomy as example
Intellectual Property Police: aspects Security aspects Authentication: XBRL Taxonomy has issued by the competent authority Integrity: XBRL Taxonomy has been not modified. Confidentiality: Not applicable for Taxonomies, only for Instance Docs. Non repudiation: Competent authority issuing the taxonomy should correspondingly accept the Instance Documents. Financial aspects Protecting the significant amount of money, resources and talent has been invested in the development of any XBRL taxonomy The use of XBRL taxonomy work will be royalty free. XBRL consumers MUST be protected against any financial claim (even based on legal subterfuges) for the use of XBRL specifications and XBRL taxonomies.
Intellectual Property Police: Methods Patents: Ideas may be patented in USA, but not in Europe RSA cryptographic algorithm is patented in USA, but free in Europe All the working group meetings in XBRL International start with a patent declaration as a preventative measure. Copyright: International method to protect intellectual property Open Source projects usually requires permission for modifications, but XBRL taxonomies are to be extended, NOT modified Trademark: International method to protect and specify name or acronym XBRL, COREP and FINREP are Trademarks, to protect their fair use. Moral Rights: Legal duty on reputation. Authors MUST be credited. Most software companies claiming against piracy, simply ignore this obligation
Stakeholder Management Categories of Stakeholders Vendors, Commercial End Users, Government Entities, Individual Practitioners, Universities and Academics, Individual Consumers Taxonomy Ownership Legitimacy and authority of the taxonomy's contents Potential and legitimate sources of change Cost of changing reporting standards Project Team Core Team, Extended Team of stakeholders, External stakeholders Transparency Motivating experts, Channelling inputs, Managing expectations, Timing, Accommodating institutional approaches Quality Control, Risk Assessment and Risk Management
Additional data in XBRL Instances Used when adding information for identification or others, as the name of the person authorizing the filing, entity postal address, and so on. The GCD (Global Common Data) Taxonomies are in draft status since The DGI (General Data Identification) Taxonomies are in final status, with the acknowledgment of XBRL International, since DGI taxonomy is in use in Spain, with more than XBRL instance documents received by the Securities Supervisor and published at (XBRL icon). A new version has been developed very recently.www.cnmv.es Topic currently in discussion in the XBRL community
Additional data in XBRL Instances: Questionnaire b) Do you ask for additional data for identifying the reporting institution? Examples: contact details of the preparerer, company name, type of identification code etc. c) How did you extend your national taxonomies to include this requirement? Additional taxonomy to be imported by each instance. Additional hypercube that links to the segment element. Specific tag in an XML structure (XML envelope) wrapped around the XBRL instance.
Identification of the company/entity Revelation 13:17 "so that no one could buy or sell unless he had the mark, which is the name of the beast or the number of his name There is not a Universal Number identifying entities. Each country has its own proprietary system/s. Some extended financial networks (i.e. SWIFT) have an identification for subscriptors, but only useful for banks in practical terms. Europe has a list of Monetary and Financial Institutions, but it is even incomplete for supervisory purposes. ISO is an standard in development for "International Business Entity Identifier". Last news shows that the task is proving to be extremely complex, and the focus would move for something like "International Financial Instrument Issuer Identifier".
Identification of the company/entity: Questionnaire XBRL defines three items for use in identifying who sent the reporting data and for which period: Institution code, Scheme identifier and Reporting period. a) What codification is used to identify the reporting institution (IBAN/BIC/ECB- MFI/Proprietary…)?
Type of report Different instances covering different periods with the same end date (yearly, quarterly, monthly, etc.) Solo / Consolidated / Supervisor Specific Test / Actual data Others… May help here a dimension widely/commonly used as Best Practice, with a set of predefined possibilities?
Type of report (questionnaire results) Data based on COREP can be reported on a solo or consolidated basis. Do you use separate taxonomies, or other solution? SAMESEPARATE Number of countries65 By using the same taxonomy the discrimination of solo and consolidated is being done by adding a dimension in the segment or scenario element or by providing the information in the header of the file.
Non Repudiation of Non Required Data Some Supervisors validates that the XBRL Instance Document contains all the required data, but ONLY the required data. Supervisor A requires data sets P and Q Supervisor B requires only data set P A cross-border entity generates an valid XBRL Instance Document with data sets P and Q The XBRL Instance Document is VALID for Supervisor A The XBRL Instance Document may be INVALID for Supervisor B due to the repudiation of data set Q. Corollary: Non Required data included in an XBRL Instance Document should be NOT a cause of repudiation
COREP/FINREP Versioning as Best Practice
CEBS Questionnaire Some regulators have specified that a reporting institution can only sent a report for one template in one reporting period, but other regulators allow the composition of the data in one instance file for several reporting institutions, templates and periods. Which rules have you defined? Rules:# countries One institution for one template in one reporting period.4 Several institutions with several templates for one reporting period.1 One institution for several templates in one reporting period.4 Several institutions with several templates for several reporting periods.1 One institution with several templates for several reporting periods.1
Delta update on XBRL Instance Documents XBRL Instance NOK OK XBRL Instance NOK OK XBRL Instance Delta update Complete resend
CEBS Questionnaire Do you expect the XBRL instance to be in a character encoding different from UTF-8? Additional accepted character encodings: ISO and ISO (FI, BE)
CEBS Questionnaire Do you ask for (or have created) additional information that the one present in the CEBS (or local regulator) package? If yes, how did you extended the taxonomy? Did you only referred to the CEBS taxonomy or did you look at other countries' taxonomies? Proceeding:# countries Taxonomy extensions by using only CEBS taxonomies5 Additional taxonomy for "general information".3 XML envelope around the XBRL instance.2 No information added.3 Did you delete some parts of the taxonomy? If yes, how did you proceed? What impacts did it had on the validation (calculation linkbase / formulas)? Proceeding:# countries Unused part were disabled, calculation linkbase is not used.5 Unused part were disabled, calculation linkbase is adapted.2 Unused part were disabled.3 No extensions.3
Questionnaires CEBS XBRL adoption and usage questionnaire XBRL Europe questionnaire Best Practices Board questionnaire
Possible Conclusions USA, Japan or India have a central source on Best Practices in Banking Supervision Europe have sources of authority (at least) Only with harmonization at ALL LEVELS the Interoperability may be achieved. Thoughts?