Bob Schneider is a Partner in 
Wilson So is the Director of Hitachi America, Ltd. A Memo to XBRL-US
Written by Christopher Whalen Posted February 6, 2007
Richard Christopher Whalen is Senior Vice President and a Managing Director of Institutional Risk Analytics (IRA) with responsibility for sales, marketing and business development. Chris is a general securities principal and has worked as an investment banker, research analyst, and journalist for more than two decades.
We are frequently asked why we take a contrarian, sometimes critical view on XBRL, yet remain committed to making this standard for business reporting a reality. The answer is very simple: we can see the value of XBRL in many aspects of data collection, transmission, and delivery, but not all of these business cases have the same strength or urgency.
When we look at the question of achieving the universal adoption of XBRL, we are reminded of a saying in the technology world: The task is possible, but very difficult. This is particularly the case when a well-intentioned group of technologists, who are mostly driven by a transformational vision of how the future of business reporting could or even should look, attempts to impose sudden, comprehensive change on the business professionals who are responsible for the day-to-day workings of the commercial data world.
From a purely theoretical perspective, XBRL makes enormous sense. For example, XBRL organizes and characterizes data, and makes it interoperable. Yet each of these benefits is not necessarily as strong as the other, especially when viewed from the perspective of the business case needs of a decision maker whose main criteria for selecting one technology over another is short-term cost.
In order for XBRL to win broad adoption in the marketplace, we believe that the proponents and vendors alike need to ask themselves some basic questions. The answers to these queries, when applied to the business case needs of real live customers, will help those leading the effort at XBRL-US make efficient choices about where to deploy scarce resources.
First, if you observe the data pipeline and divide the steps that data must travel between the point of creation and end-user consumption, it quickly becomes clear that some parts of the XBRL business case are more compelling than others.
In the case of the FDIC, for example, the data collection and validation tasks are clearly a big winner and have saved that agency enormous amounts of money and man hours in terms of collecting financial statement data from US banks. The accuracy of the data and the reduced time and personnel resources needed to collect and validate it are business case wins that make their own arguments.
Yet in conversations with regulators at the FDIC and other agencies, we find little enthusiasm for pushing the XBRL model down through the entire data delivery pipeline to end users. In fact, most federal regulators seem satisfied with the cost savings and data accuracy wins achieved by using XBRL at the top of the data pipeline.
When asked whether they intend to store or distribute the data in XBRL, the answer is no; the data will be decomposed into its constituent parts, processed and stored using legacy database systems, distributed to examiners and analysts via these same legacy systems, and consumed using existing tools such as Microsoft Excel. Much the same response comes from end users in the investment world who are aware of the XBRL adoption effort.
Indeed, as XBRL-US moves forward with its adoption activities, we predict that the powerful business case arguments in favor of the technology at the start of the data pipeline — namely, organization, collection, and validation — will become more and more prominent in discussions, while the end-use rationales so popular with many downstream vendors will fade. This is not because there are no downstream benefits to XBRL, but because these business case arguments, in relative terms, are not yet mature enough and powerful enough to withstand the critical scrutiny of business case needs.
In fact, if you leave aside the unique example of the SEC’s voluntary filer program and look instead to the numerous potential opportunities for adoption by other public sector entities, the importance of the upstream benefits of XBRL becomes even more pronounced. Thinking about examples like automating financial reporting between the various federal agencies or between Washington and state and local agencies, there are a plentitude of obvious areas where XBRL could be employed to great benefit.
But again, the sales process by proponents of XBRL adoption need go no farther than organizing, collecting, and validating the data. These benefits are very powerful and are more than sufficient to win over those public sector constituencies that XBRL-US needs to convert into soldiers in the campaign for adoption. If XBRL-US takes this lesson and focuses on winning the adoption battle step by step, with the most compelling applications first, then the XBRL adoption process will result in success.
February 7th, 2007 at 6:57 am
It is certainly true that XBRL has had its greatest penetration in markets where the consumer is able to mandate its use to the producer, aka regulators. Where producers and consumers must negotiate to an agrrement on the value of using XBRL, we have so far seen a chicken-and-egg problem preventing adoption, even in cases such as the FFIEC, who are now sitting an ever growing pool of XBRL data.
My own expectation is that this situation will be resolved in one of two ways. Either the dam bursts, which in the case of the FFIEC means that consumers become convinced that the XBRL data, in aggregate, has become of compelling value, or a single savvy consumer attempts to obtain a first-mover advantage (however temporary) and sets off an arms race among consumers which precipitates a demand for XBRL from the producer.
I’m not at all clear that XBRL US should entirely cease its evangelism for XBRL outside the regulatory use case. It would be better off to just tailor investmwents to expected near term payoffs rather than distant future goals.
February 9th, 2007 at 5:52 am
I couldn’t agree with you more. I still use Office 2000 Excel. Until everyone is using a spreadsheet that can both read and write XBRL, out of the box, how much can we expect of XBRL in an end-user environment? As you say, “we find little enthusiasm for pushing the XBRL model down through the entire data delivery pipeline to end users.” The market has spoken.
But you’ve boiled down the features that everyone can benefit from today,
Organization, collection, and validation. OCV all the way!
When I read about XBRL I get the sense they’re trying to build the highways, roads, gas stations and cars. I’m with you (I think). Focus on the Interstate and commercial trucking and let time and market catch up with the rest.