April 3, 2006
We first thank our colleagues at London-based Global Risk Regulator for their comments regarding IRAâs Bank Monitor analytics benchmarking algorithms for Basel II. GRR observes that the numbers illustrated in IRAâs latest report, âBasel II by the Numbers 2005,â stand in sharp contrast to the results of QIS-4 -- results that were so skewed in favor of the banks as to send the U.S. regulatory system into a year-long tailspin.
Global Risk Regulator opines that IRAâs numbers make intuitive sense and lend weight to our argument that a uniform benchmark computation methodology to estimating credit risk exposures and measure Economic Capital may help the Basel II process succeed. The point isn't that our method for modeling Economic Capital is correct vs. QIS-4, but rather that it is possible for analysts to calculate apples-to-apples Basel II benchmarks for all US banks using regulatory data, all of which is publicly available thanks to U.S. leadership in interactive data.
The Fed restarted the Basel II process this past Thursday with âpreliminaryâ notice of a proposed rule making with respect to the New Basel Capital Accord. As the process moves forward, we hope the virtues of transparency and comparability, driven in part by the public data benchmarks illustrated in our report, are taken into consideration.
IRAâs core mission is to develop and field benchmarking analytics. As a developer of computer enabled data mining tools, we strongly support the advent of publicly available, well-structured or âinteractiveâ data. In the past we have lauded the FDICâs modernization effort, which now has all FDIC-insured depository institutions submitting quarterly financial reports using eXtensible Business Reporting Language or XBRL. The transparency, completeness, consistency and quality of the FDICâs bank information pipeline, which is used in our analysis engines to produce uniform benchmarks for Basel II, enables IRAâs âBasel II by the Numbersâ report series to serve as a canvas upon which to demonstrate the power of âdistillingâ structured data.
As Spring of 2006 arrives, the SEC is intensifying its focus on interactive data. The Congress, as we report in this weekâs issue of our Washington & Wall Street business intelligence service, has just held hearings on promoting enhanced financial reporting. There are a variety of initiatives in the public and private sectors involving structured, machine readable data, all of which lay claim to the powerful label of âinteractive dataâ and all of which employ different types of technology for gathering and disseminating financial information.
We start from our opinion that XBRL now stands as a best of breed option for organizing the gathering and submission of financial reports. It is not the only solution, but it is the one where significant community investment has been made. The FDIC has pioneered solutions worthy of analysis and emulation in this regard. Their use of XBRL to control not only content but the selection of what to collect based on embedding the rules is brilliant. Being systems engineers at heart, this leads to the recognition that pipeline challenges further downstream have now moved up the priority list for treatment. We posit the following questions for the âinteractive dataâ community to consider in 2006.
We rub our worry beads pondering the anthropology of innovation, each component developed piecemeal and each maturing to serve the interactive data space. Not unexpectedly, we see evidence of classic early adoption myopia -- competing solutions ignoring each otherâs value, while pushing, at times aimlessly, in the hope of owning as much of the interactive data real estate as possible. We know from experience that the âone wrench does it allâ approach hurts rather than helps the adoption of interactive data as a resource to the financial community. We believe there needs to be more context as to what functional purpose a technology has to each step in the value pipeline â collection, validation, storage, distillation & dissemination â over which data travels from source to user.
Consider that financial reports submitted to regulatory agencies are legal documents. They have evidentiary standing based on the notion that they are fully encapsulated and independent point-in-time permanent records which satisfy a disclosure duty. But a document that references a controlling element defined externally, an element that may change meaning, could cause the validity of this paramount regulatory assumption to fail, making enforcement even more complicated and costly than it is today.
Conversely, if every document contains a truly independent self-contained point-in-time taxonomy (that's what a printed page is by the way), we potentially create an electronic mad house. Where there is no standardization, the goal of creating better information for individual investors, for example, a major goal of the SEC's current modernization effort, is thwarted. The FDIC and other bank regulatory agencies solved this problem by electing to use agency defined, closed-form, point-in-time taxonomies. The challenge for agencies such as the SEC is to achieve this level of legal document submittal and recording efficiency while not dictating specific accounting treatment.
Do open-form data organizing constructs such as XBRL really have to propagate past the submittal point? And if they do, then what architecture should exist for capturing the infinite extensibility of corporate America while ensuring the legal integrity of the records? How far downstream should an information pipeline carry the echoes of it's front end? Beyond the control gateway where a document is certified the trade offs shift towards downstream transmission technologies that are more attuned to machine-to-machine ("M2M") interoperability.
Following this argument, should regulators and standard making bodies begin to insist that multiple technology solutions (XBRL, XML, SQL, et al) be directed towards working out rules for organizing equivalence of content over a range of M2M outlets? We believe that such a mandate would overcome constraining inertias and result in cooperation to build pathways for migrating structured content seamlessly among âequivalent sets.â How end users employ, or not, the full capabilities of any data transport technology is ultimately driven by economic IT realities like installed base investment, switching cost and capital amortization. Wearing our C-suite hats for a moment, buying into an equivalence architecture design is a lot less risky than committing to a transformational one.
We have seen in this past weekâs Congressional hearings on enhancing financial reporting that "8 of 13 speakers mentioned XBRL," according to one of our colleagues in the consortium. Will regulators take the lead in marshalling resources so that the United States' National Interest need for a comprehensive and inclusive financial reporting pipeline solution will be realized? Our assumption is that taking an inclusive approach to existing downstream technologies will ultimately benefit the cause of wide adoption of interactive data and best serve the disparate constituencies that consume financial information.