New COREP & FINREP - Experiences
5 May 2014 | Rome
Owen Jones | CRR Taxonomy Project EBA
© EBA | European Banking Authority
New COREP & FINREP – EBA Experiences
• The chicken and egg problem
• The “two level” problem
• DPM / taxonomy split & the wider process
• Validation rules
• The near future
2
The “Chicken and Egg” Problem
The format and tools
• working at the edge of the XBRL standard
• using the “under development” Table linkbase
• DPM approach and highly dimensional XBRL had had limited exposure.
• How can you define and impose a Europe wide reporting format in the
absence of widespread software to build, test, check & use it ?
• How will the market build such software in the absence of an agreed
format?
Testing
Much larger data model than in most previous XBRL projects, everything changing at
once – the legislation, policy, business model, data gathering and reporting chain
• Limited realistic test data – and unlikely to be any until real reporting imminent
3
The Market Based / Kevin Costner Solution:
“If you build it … they will come”
Field of Dreams - 1989
A wide variety of software vendors and consultancies have
developed solutions – as have many reporting institutions.
A selection can be seen over these two days.
Developed the format whilst the standard was still being worked on
• Inevitably we have ended up using a non-finalised version.
• Expect the EBA to update to the finalised version in a future release.
4
The Reporting Landscape – The “Two Level” Problem
Each CA has choice of mechanism
for first level reporting
EBA XBRL
CA
Reporters
Vendors
How should the EBA
communicate precise data
requirements in the absence of a
common format?
EBA XBRL
Own XBRL
CA
Reporters
EBA
Vendors
Own format
Who is responsible for what?
Reporters
CA
Vendors
5
The Reporting Landscape
EBA XBRL
Reporters
CA
Vendors
XBRL Taxonomy Relevant
EBA XBRL
Reporters
Own XBRL Taxonomy
EBA
CA
EBA XBRL
Vendors
EBA XBRL
Own format
Reporters
CA
Vendors
DPM Relevant
6
The Reporting Landscape - Experience
• Fairly complex landscape
• Fuzzy responsibilities
• Uncertain communication channels
EBA XBRL
CA
Reporters
Vendors
EBA XBRL
Has taken time for all CAs to make
and communicate their decisions
Even when using EBA XBRL, many practical
questions on first level reporting remain the
domain of the CAs.
Own XBRL
CA
Reporters
EBA
Vendors
Own format
Reporters
CA
Vendors
7
DPM / Taxonomy Split
What should the Data Point Model be for?
• Communication? (to precisely define/explain ITS)
• Data collection? (to create a taxonomy)
• Analysis? (to support usage of the data)
Regulations
Transmit to EBA
Templates
DPM
Taxonomy
Instructions
Report to CA
Understand Data Requirements
Validations
Analyse Data?
EBA Business experts
EBA IT experts
Competant Authorities
Credit Institutions
8
DPM / Taxonomy Split - Experience
Good:
• Get good, helpful feedback in Consultations
• Get feedback on DPM from people who would be uninterested in /
unable to understand XBRL
Bad:
• Consultation timescales not long enough (DPM + particularly XBRL)
• Relatively little feedback
• Get significant feedback after the official window
• XBRL savvy are often uninterested in engaging until XBRL available.
• Miss the chance to influence at DPM stage
9
The Process
Fairly linear:
• ITS and templates  DPM 
XBRL Taxonomy
• Each stage consulted on and
finalised before the next.
Regulation
ITS
Experienced typical “waterfall” issues:
• Problems "set in stone" by the time they
are realised
• Little chance to correct errors/improve impact of
decisions of earlier stages
• Slippage of early steps (e.g. regulation, ITS) cuts
time for later ones
• Consultations only pick up views of the main
audience for each individual step (e.g. policy for
ITS text/templates, technical for XBRL)
DPM
• Draft
• Consultation
• Final
XBRL
Taxonomy
• Draft
• Consultation
• Final
• Patches
Implementation
+ Reporting
10
Future Improvements
Since size & technical innovation in future releases will be less
• Can aim to bring the ITS, DPM and Taxonomy releases together
• Maximise time for adjustment/correction
• Engage the complete audience at once
Approach
• Avoid specific consultations with (very) short windows
• Keep an “always open” channel for feedback on DPM and XBRL
• Incorporate later feedback in next version or patches.
11
Validation Rules
• Including and publishing validation rules was ambitious:
• Time pressure, lack of test data, systems, performance concerns etc
• But necessary to trigger development in reporter/CA systems
• EBA recently published a list of rules that should not be applied
10%
83%
17%
7%
One XBRL mapping error
Various content errors
Remaining rules
Expect a patch update to the 2.0.1 taxonomy removing the disabled rules
to be published shortly
12
Near Future Complexity – Outline 2014/15 Timeline
August ‘13
December ‘13
May ‘14
September ‘14
XBRL Dev
ITS & DPM
XBRL Dev
ITS & DPM
ITS & DPM
XRBL Dev
Implementation
Implementation
Refine+Patch
June ‘15
October ‘15
L1 Submission
L1 Submission
L2 Remittance
L2 Remittance
Implementation
XBRL Development
January ‘15
L1
Submission
L2
Remit.
Refine+Patch
Implementation
L1 Submission
L2 Remittance
For the next 6 months
•
Increasing overlap of multiple phases of taxonomy development,
patches, implementation and usage
•
Rapid change
•
Even shorter timescales
After that – gets better, settle into regular (semi)annual update cycle
13
Contact information
Owen Jones
owen.jones@eba.europa.eu
European Banking Authority
Floor 18 | Tower 42 | 25 Old Broad Street
London EC2N 1HQ | United Kingdom
t +44 (0)20 7933 9900
f +44 (0)20 7382 1771
info@eba.europa.eu
www.eba.europa.eu
14