Bob Schneider is a Partner in 
Wilson So is the Director of Hitachi Consulting Corporation The XBRL Technology Workshop & Summit
Written by Bob Schneider Posted on August 19, 2009
Because Hitachi Data Systems hosted the XBRL Technology Workshop & Summit, my objectivity may be justifiably questioned if I sing its praises. But the event, presented by XBRL US, was much more interesting than I had thought it would be.
My expectation was that most of the conference would be highly technical and interest only developers. Instead, the wide-ranging presentations illustrated both the breadth and depth of XBRL’s potential, as well as the significant challenges ahead for implementations.
Paul Wilkinson did an outstanding job of blogging the conference proceedings on July 28, 29, and 30, and slides of many presentations have now been uploaded to the XBRL US site. Re-reading Paul’s comments and the slides, I experienced the alternating sensations of exhilaration and concern I had while listening to the speakers.
On Day One, Campbell Pryde, Chief Standards Officer of XBRL US, discussed the development of US taxonomies. What enormous progress has been made in the three years since the SEC awarded funding for the first US GAAP taxonomy! (Someone asked Campbell in another venue: With 17,000 elements in the GAAP taxonomies, why would anyone think of adding an extension?)
At the same time, Campbell’s discussion of the use of 2008 versus 2009 GAAP taxonomies as well as FASB Accounting Standards Codification evoked Matt Kelly’s admonition in his post on the SEC mandate: “Adoption not only requires some specific vow of action; it requires maintenance, day in and day out, and that can be hard to deliver.”
Makoto Koizumi’s talk reminded me of the excellent series of posts by Kobayashi Toshinori on the XBRL implementation at EDINET, and the strong participation by Japanese companies of its voluntary filing program. Koizumi noted, however, that there is no immediate plan to introduce XBRL for footnotes. The SEC’s XBRL mandate requires only limited tagging of footnotes, not the deep tagging that had originally been envisioned. And as Amy Pawlinkci has discussed, there are no immediate plans to allow US companies to tag the MD&A either.
In both the US and Japan, therefore, tagging for the foreseeable future will be mostly for the primary financials. Is that sufficiently ambitious to demonstrate the value of XBRL and gain the support of the investment community? This question was put in relief by the presentation of Darrell Heaps of Q4 Systems: he has talked to many Investor Relations Officers (IROs), and few see what XBRL does for them. On the other hand, the transformation of the “investor web” that Darrell described from only static corporate websites to dynamic social media (Twitter, wikinvest, Seeking Alpha, etc.) creates exciting possibilities for a data standard with the capabilities of XBRL.
For me, the key message of Convergence of Taxonomies for Financial Reporting, a group presentation led by Josef McDonald of Ernst & Young, was that XBRL for IFRS is driven by the individual IFRS standards, while US and Japan GAAP taxonomies are hybrids of standard-driven and market-driven elements. This difference accounts for the lack of industry taxonomies in IFRS, and why US GAAP taxonomies are much more highly developed than those of IFRS. The good news is that, looking out toward a time of convergence in accounting standards using XBRL tagging, the world’s leading taxonomy creators are closely cooperating on international taxonomy development.
The presentation on XBRL and corporate actions given by representatives of DTCC, SWIFT, Citi, and XBRL US was extremely exciting. Their slides demonstrated the labyrinth that corporate actions data must navigate to be fully disseminated, a process involving multiple stakeholders including issuers, filing agents, custodians, stock exchanges, and investors. As described in their excellent conference backgrounder, once the Corporate Action XBRL Taxonomy being readied by XBRL US is completed:
Issuers (or their agents) will be able to create their press release, prospectus or letter of transmittal in XBRL format, clearly indicating the market, event type, security type and whether the offer is mandatory or voluntary. All of the information that the issuer has “tagged” or identified within their document will be “machine readable” and can be extracted, searched on and consumed by the stock market, custodian, depository, transfer agent and, ultimately, the investor. Because the taxonomy is developed following the ISO 20022 standard, the data will be available in a familiar format to all users.
In my opinion, this raises a couple of questions: Doesn’t capturing key data points from text-based documents like press releases and prospectuses present the same sort of difficulties that have slowed tagging of financial statement footnotes? And with an overhaul of financial regulation in view in the US and other jurisdictions, won’t corporate action taxonomy maintenance be unusually difficult for the next few years? Regardless of these challenges, corporate actions seems an especially ripe area for the application of XBRL.
Steve Wright of Salesforce.com Foundation spoke about data in the social sector. Steve discussed the inadequacies of using a currency like the dollar for measurement of costs and benefits in the social services marketplace, and described initiatives for deriving better yardsticks for more effective measurement of social impact. Efforts like PULSE/IRIS for creating financial, operational, and social impact metrics can be effectively leveraged through XBRL reporting.
I agree entirely with Steve on the futility of using dollars to evaluate returns on social investment. The challenge, which Steve recognizes entirely, is coming up with better measures. The concerns I have are data quality and assurance.
For example, in this blog post on Change.org, Steve posits an organization in Dakar working to decrease disease by building urban sanitation stations. Assume that the metrics created accurately reflect social impact. What biases do the data collectors have, how are they reflected in the data, and what kind of assurance can be performed on data in such an environment? But regardless of these concerns, Steve’s work is representative of yet another area outside external financial reporting to which creative minds are seeking to exploit the rich potential of XBRL.
It was extremely encouraging to hear from the young Japanese interns at XBRL US. Their work on topics like formulas, extensions, and corporate actions was all the more impressive for having been accomplished in two weeks (some of their slides looked like they would individually take two weeks to prepare).
I was a bit disappointed, however, to hear little mention of accounting in any of their work, and the absence of any American interns engendered the usual worry of whether US schools were producing kids with these kinds of skills (and if they are, where were they?). But I know that XBRL US must have been delighted to have their contribution.
The last session was on Database and Business Intelligence, which included presentations from UBMatrix, Morningstar, Informatica, and Oracle. I wish Morningstar no ill; but if data aggregation (a chunk of Morningstar’s business) is supposed to wither away with the waxing of XBRL, shouldn’t their people be a lot more unhappy than they seem to be? Homi Byranji of Reuters seemed similarly unworried in his presentation at the London Investment Professionals Conference last year. If their demeanor is anything to go by, XBRL doesn’t doom the data aggregators.
Finally, I have often wondered why bloggers reporting from a conference will end with a mundane personal note that could only be meaningful to a small group of readers. I now understand why they feel compelled, as in my case, to say how great it was to match faces to the names I have admired over the past three years, and to have the opportunity to thank guest writers in attendance for all their hard work on behalf of the blog (Neal Hannon and Ashu Bhatanagar deserve special mention). I also know that this summary has only discussed half the presentations, and has done justice to none. If any speaker would like to follow up their talk with a blog post, please contact me.
August 26th, 2009 at 10:09 am
Bob – I have some comments to the questions you pose regarding the Corporate Actions initiative for XBRL.
BOB: “Doesn’t capturing key data points from text-based documents like press releases and prospectuses present the same sort of difficulties that have slowed tagging of financial statement footnotes?”
MAX: I’m not sure I completely understand all the difficulties for footnotes, but yes, the entire problem in corporate actions is just that: “capturing the data points from text-based documents.” I’m sure we’ll have to be flexible and open in designing the solution. This is especially so because announcement of a corporate event is not a singularity, but a stream of updates from the first, early notification to the final terms. If you think of any big merger, you can imagine the complexities from the first news through anti-trust review, through shareholders’ meeting (think of Oracle-Peoplesoft or Microsoft-Yahoo).
The really big point is that the very difficulty you describe — i.e., teasing out the correct data from a 2 paragraph press release or a 257 page prospectus — is currently being done redundantly by dozens of firms, thus introducing significant opportunity for differences. These differences emerge when the standard messages that flow downstream through financial servicers to the investment firms whose holdings are affected. A “fuzzy” announcement, with room for interpretation, guarantees significant effort to manually reconcile differences, at significant cost. So no matter if this approach is fraught with technical challenges, the benefits of the solution make it all worthwhile.
BOB: “And with an overhaul of financial regulation in view in the US and other jurisdictions, won’t corporate action taxonomy maintenance be unusually difficult for the next few years?”
MAX: The corporate action taxonomy is not going to be driven by financial regulation, but rather by the messaging standard for corporate actions, ISO 20022, which is already quite mature. We don’t, at this point in time, see huge changes coming in content of disclosure and we actually hope for a rule to add XBRL to the existing disclosure requirements. We are taking the position that the corporate action information from issuers is generally adequate, but the method (pure text) is inadequate because it injects the opportunity for faulty data systemically into the notifications that reach the investors.
Another way of saying this is: “the data is there, it’s just darn hard find.” This obscurity is due to the legal wrapping that is required. Remember that this is all in order for a corporation to offer or announce actions that will affect their investors’ holdings. The legal framing and conditions and terms and exclusions really cannot be mucked up. All that necessary verbiage is intended to deal with litigation risk. But buried therein is a lot of data that will operationally generate changes to millions of accounts multiplied by hundreds of servicing accounts multiplied by multiples of amendments per announcement multiplied by 400,000 unique events per year (in the US).
So this is a very fundamental problem and is not significantly confounded by regulatory changes. The standard is able to accommodate everything from a huge merger deal that lasts for many months to a simple dividend. Even then, XBRL will be a progressive step forward for the corporate actions community because if a regulatory change does impact an announcement, an extension can immediately capture the metadata of a change and serve to feed back into both the standard and the taxonomy. If we can close the loop in this system, then changes can be integrated much more efficiently.