Bob Schneider is a Partner in 
Wilson So is the Director of Hitachi Consulting Corporation XBRL: An Interview with Paul Wilkinson (Part 1)
Paul Wilkinson is Chief Strategy Officer of CLOUD, Inc., the Consortium for Local Ownership and Use of Data. He was Senior Adviser to SEC Chairman Christopher Cox from 2005 until 2009, where he oversaw XBRL adoption. From 2001 to 2005 he was Executive Director of the U.S. House of Representatives Majority Policy Committee, handling international and economic policy. He can be reached by email.
This is the first installment of a three-part interview. Part 2 has questions 6 to 10, and Part 3 includes questions 11 through 15.
1. You recently discussed the role XBRL can play in helping to restore the asset-backed securities (ABS) market. Could you explain what benefits XBRL would bring that ASCII and HTML documents do not?
That’s a timely question. The most important near-term benefit would be exploiting XBRL’s off-the-shelf capability to support effective disclosure standardization for ABS. Longer term, I think the “faster, better, cheaper” advantages that SEC Chief Accountant Don Nicolaisen foresaw in the GAAP arena would apply in the ABS arena, as would XBRL’s disclosure maintenance capability.
With respect to the implementation process, it’s true that regulators or an industry consortium could sit down and choose standard ASCII or HTML or Excel column headings or row labels for most of the important information about ABS. In fact, they tried to do so in the run-up to the crisis, but it was a slow process and has yet to result in a fair and efficient market for ABS.
On the other hand, the open process of creating XBRL financial taxonomies in the U.S. proved extraordinarily effective. Once the technical process was underway, I was astonished by the speed, quality, and openness of the XBRL US project. White House adviser Beth Noveck wrote a book called Wiki Government about crowd sourcing patent application research. Well, the effectiveness with which XBRL US crowd sourced the GAAP taxonomy deserves its own book. Your recent interview with the CEO of XBRL Australia reflects my experience in the U.S.: The technology was simple; the politics were the complex part. We couldn’t crowd source the SEC implementation rule, but XBRL US transcended all of the typical politics associated with U.S. GAAP by crowd sourcing the taxonomy.
In addition to the process advantages, there are, of course, myriad technical reasons XBRL is superior to ASCII, HTML, Excel, or any other open or proprietary standard. Extensibility is a big advantage, as is the ability to observe the use of extensions and to create and update standard tags over time – without requiring users to buy new software.
As we’ve seen with ABS, keeping disclosure regulation up-to-date with financial innovation is critical. Over the past decade, one reason capital flowed disproportionately to ABS relative to public companies is because regulators used proven manual systems to keep GAAP up to date. SOX was expensive. It helped prevent more Enron’s and WorldCom’s, but at the same time, it drove capital toward non-GAAP investments. Despite Reg. AB, which was generally a codification of many years of asset-backed securitization legal practice, ABS financial practices continued to evolve, contributing to both the housing bubble and to the growth of multi-layered complex securities on top of basic ABS.
Non-standard ABS reporting made it difficult to see the real risk presented by ABS and by derivatives built on top of ABS. ABS, because they didn’t create human drama or present business strategy questions like public companies do, got less attention than their valuations might have merited were the information available in a larger, more transparent, more liquid market. XBRL won’t bring the excitement of boardroom fights, hostile takeovers, or new consumer products to ABS. But to the extent XBRL for ABS helps investors understand basic facts about the instruments they’re being offered, it’s analogous to GAAP for public companies.
Thanks to computers – and thanks to the fact that ABS are considerably less complex than public companies – technology that took decades to develop for GAAP is already available for ABS. The SEC has plenty of authority to mandate disclosure for securities, which are simply investments of money in common enterprise for profit derived from the efforts of others. The questions are how fast regulators can understand XBRL’s capabilities to enhance transparency and how the SEC can navigate the regulatory minefield set by those whose oxen would be gored by transparency.
2. Part of your work at the SEC was to help mandate the use of XBRL for public companies, mutual funds, and credit ratings agencies. Are you convinced that XBRL implementations at the national level will always require mandates? Or do you see a path for voluntary adoption to be successful?
The "chicken and egg" metaphor is hackneyed but nonetheless applicable: People only want to use a standard when adoption is sufficient to make the standard cost effective. While more than 100 public companies found it cost effective to participate in the SEC’s voluntary filing program, the ultimate information users – investors – required much greater adoption rates to make XBRL worthwhile. I don’t like where this metaphor is going, but effectively we had to force companies to lay eggs. Fortunately, they’ve been AAA extra large so far. Future implementations with lower adoption hurdles may not require mandates.
One reason to support XBRL adoption was to empower independent auditors to provide faster, better, and less expensive assurance by efficiently combining integrated software tools with professional judgment to create more reliable financial statements. If investors in private companies become accustomed to XBRL-quality data from public companies, perhaps they’ll demand similar information from private companies, where more highly-concentrated investors typically have more leverage over management. It certainly makes sense for private companies hoping to go public within a few years to start using XBRL well in advance to get ready for a fast and efficient IPO.
Moreover, with fixed costs for public companies for GAAP XBRL implementation, the cost to expand XBRL to other applications becomes mainly marginal. With the most complex data set – U.S. GAAP – out of the way, implementing the same or similar technology to automate other parts of the business information supply chain is more manageable. For example, XBRL GL looks much more attractive. Someday, a supermarket will scan the UPC bar code on a can of tomatoes and the dollar of revenue will end up in a 10-K without any human intervention except to monitor the systems that ensure straight-through processing.
If the U.S. is to remain competitive in the world economy, we have no choice but to improve efficiency by automating business processes. Much of that work not only can be voluntary – it must be voluntary because it’s the consequence of competition among businesses. XBRL is poised to win its share of competitions. A nice voluntary step, for example, might be XBRL support from Google’s Data Liberation team. With Google and other market participants throwing huge support behind open standards, XBRL will have to repeatedly prove its capabilities in the years to come.
3. An argument has been made that financial regulators like the SEC primarily adopt XBRL to help their own analysts do their job more efficiently, and the needs of investors are a secondary consideration. To what extent do you think that argument has validity?
In my three-and-one-half years at the SEC, I never saw any air at all between the interests of investors and the interests of the Commissioners and staff. People at the SEC might sincerely disagree about the best ways to help investors, but everyone believed their primary mission was to help investors. The nice thing about XBRL is that it can help both government and private analysts do their jobs better.
It’s true that SEC analysts look for civil liability more often, while private analysts mostly look for investment opportunities, but there’s significant overlap. Either type of analyst is likely to consider whether a company has accurately disclosed all material information and either is likely to act on their conclusion. A private analyst decides to invest or not; an SEC analyst could call a company, issue a comment letter, or consult with the Division of Enforcement. While it could be a big mistake to trust a company just because the SEC hasn’t sued it, it’s also very clear that the faster and better the SEC can do its job, the less likely issuers can get away with fudging the facts. The roles of the SEC experts and private analysts are highly complementary to each other – not contentious.
Because so much money is at stake, and because resources for technology at the SEC are limited, I expect the private sector, not the SEC, to continue to lead the way in using XBRL. The SEC has considerable work to do to use XBRL to support more efficient compliance and surveillance. And because the legal environment in which the SEC itself works is itself enormously complex, the constraints won’t be technology. The SEC’s constraints will be legal and policy constraints.
What sort of data discrepancies are legally sufficient to justify imposing costs on public company investors to respond to an SEC investigation? How does one use technology to differentiate among skillful, lucky, and corrupt trading to show cause for a subpoena? Compared to questions like these, diversified investing (perhaps using XBRL to differentiate asset classes and industries and to identify different but potentially correlated investments) is easy.
4. How do you think XBRL adoption by the SEC is proceeding? Do the issues described in the observations of SEC staff on the statements filed thus far signal substantial problems? Or are they about what you had expected, or even relatively minor?
By all accounts, adoption is going well. The staff observations are informative. The first concern on the list relates to rendering, which is somewhat encouraging given that rendering isn’t particularly relevant to XBRL’s main purpose. After all, if you want the benefit of thousands of years of experience of putting ink on paper in a visually appealing way, you can still look at the dead tree format financial statements.
I also find it encouraging that the staff is urging filers to use standard elements. But that’s not a surprise either, since the use of standard elements helps investors make comparisons among companies and of performance over time. I expect the staff comments will be particularly helpful as the U.S. GAAP taxonomy is updated. As the taxonomy and the newly codified edition of U.S. GAAP become fully interoperable, tools using XBRL should make it easier for accountants and investors alike to visualize the importance of high quality accounting.
The really neat thing is that, before XBRL, the SEC found it difficult or impossible to hold third party data taggers accountable for the changes to GAAP in the normalization process. Now the SEC has tools to understand not only the official company reports but to understand how investors use those reports. The line of accountability, from companies to investors and the SEC, is much more direct. So is the feedback loop for GAAP itself. It will be interesting to watch how data about the use of GAAP is used to improve GAAP.
5. At least in public, Mary Schapiro, the current Chairman of the SEC, has not been as outspoken as her predecessor, Chris Cox, on the importance of implementing XBRL for financial reporting. What consequences does this have for XBRL adoption at the agency? At this stage, with a final XBRL rule in place and adoption proceeding according to schedule, does commitment at the top of the agency make much of a difference?
To be fair, Chairman Schapiro has had quite a bit on her plate the past 10 months. And during his final year in office, Chairman Cox was extraordinarily busy dealing with the crisis of opacity via proprietary financial instruments, so he couldn’t spend as much time advocating transparency via open technology as he did early in his tenure. Fortunately, he identified XBRL as a priority very early. His experience with capital markets and technology dating to his service in President Reagan’s White House and earlier made XBRL a natural priority as soon as SEC technology director Corey Booth and the SEC’s first XBRL expert, Jeff Naumann, briefed him. Getting the process started early made it possible to finish with approval of the final public company rule in December 2008.
The most important time for the SEC to speak out for XBRL was during the work to develop the rules. With the roadmap now in place, there’s no reason for the Commission to be a backseat driver. What’s most important is for the Commission to support the staff’s work to make sure the SEC’s XBRL capabilities keep up with innovation in the private sector.
There’s always a place for leadership when it comes to transparency. Fair, fast, and efficient disclosure of asset-backed securities won’t happen without leadership from the top. I hope Chairman Schapiro and the rest of the Commission leverage the success of the GAAP and other XBRL programs to make ABS disclosure fairer, faster, and more efficient.