Transcript Presentation

DICO’s Implementation of XBRL
WEB Based Interactive
Regulatory Filings
XBRL Canada Conference
Oct. 17, 2013
Bill Foster
Vice President
Presentation Overview







Who is DICO?
Profile of regulated Institutions
Monitoring & Statistics before XBRL
Decision to go to XBRL
Monitoring & Statistics with XBRL
System Components
Lessons learned
Who is DICO?


Agent of Government of Ontario
Mandate: protect depositors of Ontario
Credit Unions and Caisses Populaires
o
o


Solvency Regulator
Deposit Insurer ($100,000)
Statistics gatherer
French Language Services Act
(everything must be bilingual)
Ontario Sector Profile by Size
Peer Group
2000
2Q012
Assets
% of
# of
Assets
% of
# of
$M
assets
II
$M
assets
II
>$1B
-
-
-
15,072
43
5
>$50M - < $1B
11,100
76
68
18,642
53
84
< $50 M
3,496
24
275
1,283
4
59
Total
14,596
100
343
34,997
100
148
($ millions )
4
Profile of Regulated Institutions
Sector Consolidation: 1977 - 2011
1400
$225
$200
1200
$175
$150
800
$125
600
$100
$ Millions
# of Institutions
1000
$75
400
$50
200
$25
0
$0
1977
1980
1983
1986
1989
1992
# of Institutions
1995
1998
2001
2004
2007
2010
Average Assets
5
Monitoring & Statistics
(before XBRL)

Paper based (1980’s to late 90’s)
o
o
o
o
o
o
o
Over 950 reporting entities in late 80’s
Information quarterly due to high volumes
Up to 300 fields of data – “key punched” by DICO
Data accuracy and data cleansing – a significant
problem
Poor timeliness of report information
Limited reports - pre-programmed by IT Dept.
Limited ability to directly query data
Monitoring & Statistics
(before XBRL)

Electronic – late 1990’s - 2010
o
distributed in-house developed visual basic
application – filed from dedicated computer at
reporting entity (Quarterly, Monthly and Annually)
o
Up to 300 fields of data
o
Ability to upload financial data from banking
system
o
o
100% data accuracy – extensive validation routine
a pre-condition to filing)
Dynamic reporting with user defined peer groups
WHY XBRL?
Existing system lacked flexibility and
scalability needed to address:



Diversity in size and complexity of institutions
Significant ongoing consolidation of
institutions
Monitoring of new business products and
activities
THE XBRL PROMISE

Scalability / Flexibility

Thin client – work on all operating systems
(including dial-up internet access)

Easily change filing frequency & specs at “entity”
and “forms” level (one size does not fit all)

Accommodate DICO’s legacy systems

Do everything existing system does and more!
MAJOR FUNCTIONS

Gather Information

Process and Store Information

Report Information

Support SQL / Lotus Notes legacy
system
DICO’s Current XBRL System

DICO custom taxonomy

Enterprise Application Server - Notes based
user authentication tied with XBRL Reporting
Window specifications

Web forms generator tags and verifies data at
Forms level

XBRL Processing engine (data validation)

Reporting - DICO’s Legacy System
INFORMATION GATHERING
XBRL Reporting Window

Secure log-in / authentication utilizing DICO’s
existing Lotus Notes based Authorized
Contacts database

Institution specific “Reporting Window”
o Reporting period
o Input form(s) specifications
o 1st level data fill-in and verification
INFORMATION GATHERING
(continued)
XBRL Web Based Forms System
o
Tags over 700 fields of data – filed monthly
(or quarterly) and annually from any
computer - individual user authentication
o
Ability to upload financial data from
banking system
o
Scalable through DICO entity specific
specifications and “trigger” fields in forms
(Capability in place but not currently utilized)
INFORMATION GATHERING
(continued)
Data Accuracy
100% accurate – two stage electronic
verification (form level and final validation) - a
pre-condition to filing.
o 573 fields have a rule re required content
o 121 fields are auto totals or sub totals
o 12 fields are auto fill from external “lookup”
o Validation: 263 caution & 73 error rules
INFORMATION PROCESSING
XPE Processing Engine.
XBRL Tagged information is passed from
the Web Based Forms System to the XPE
Processing Engine.
o
o
o
Data is validated before filing can be
completed
Completed filing is “shredded” and passed to
Oracle XDA (proprietary) database
Data posted to XDA is then also posted to
legacy SQL database.
MID PROJECT CHANGE IN DIRECTION
The original project specifications called for the
development of a reporting module by
UBMmatrix, the project contractor.
Meanwhile, SQL Power, (the project subcontractor) developed “XBRL Analytics” a drag
and drop report writer / query tool.
DICO agreed to change the project specs to
utilize the XBRL Analytics reporting capabilities.
MID PROJECT CHANGE IN DIRECTION
(Continued)
Because XBRL Analytics is driven off a
different database than the Oracle XDA, we
are currently loading both the XDA and the
XBRL Analytics databases from the XPE
processing engine.
We are running parallel data transfers to our
legacy SQL database to verify the accuracy
of the data load into the XBRL Analytics
database.
MID PROJECT CHANGE IN DIRECTION
(Continued)
User testing: compare XBRL Analytics reports
against our legacy SQL / Lotus Notes reports
Legacy individual and user designated peer group
financial and performance reports continue to be
available to DICO and its constituents on demand
over DICO’s secure web site
Once user testing of the parallel databases is
complete, we will commence user testing of the
various XBRL Analytics reports
REPORTING INFORMATION
NEXT STEPS

Legacy reports to be replaced by updated
reports based on XBRL database

Longer term: DICO will migrate away from
reliance on its SQL database and
COGNOS by redirecting the Lotus Notes
Corporate Information System data calls to
the XBRL Analytics database.
LESSONS LEARNED
DON’T BE A :
LESSONS LEARNED
REALIZING UPON THE PROMISE OF XBRL
IS FAR MORE DIFFICULT / COMPLICATED
THAN WE WERE FIRST LED TO BELIEVE

Taxonomy: No suitable standard for our
requirements – Finrep Correp > DICO

Complex validation taxonomy requires very
complicated hard coding

Implementation of XBRL for regulatory use is
still largely uncharted territory

Il ne parle pas français
LESSONS LEARNED

Work closely throughout the project with
your integration partner who must:
o
o
o
thoroughly know his business
the products he is integrating
have a very good understanding of your
business.
Acknowledgments
DICO’s XBRL system employs a
combination of “off the shelf” and custom
applications developed by Edgar Online
(UBMatrix) and SQL Power with integration
provided by SQL Power. These companies
are working together to provide DICO with
what we understand is a first of its kind
XBRL, Dynamic, Regulatory Filing and
Reporting Solution
Questions