Print this article | Return to Article | Return to CFO.com
Even as SEC chairman Cox champions "interactive data," few CFOs seem impressed. Is that because too few of the benefits accrue to them?
Alix Stuart, CFO Magazine
August 1, 2006
It has been heralded as an invention that rivals the bar code and color TV. Internet pioneer Leonard Kleinrock believes we will be "surprised and even amazed" by it. Xerox CEO Anne Mulcahy has drawn a link between it and the French Revolution, invoking Victor Hugo's famous line that "no army can resist an idea whose time has come." Securities and Exchange Commission chairman Christopher Cox has been a tireless booster, mentioning it in more than a third of his public statements since taking office. What is it? It's XBRL, or Extensible Business Reporting Language, a system for encoding the data found on financial statements. Proponents say it will speed reporting, aid analysis, reduce errors, and improve audits. Some claim that it may ultimately streamline the gap between internal and external reporting by uniting disparate information across ERP and other core IT systems.
If this comes as news to you, you're not alone. Few CFOs seem aware of what XBRL can do, where it stands in its development, or how a company might take advantage of it. Few, in fact, seem even to have heard of it, and if you're among the many executives who complain that the world of information technology is riddled with too many "TLAs" (three-letter acronyms), then this four-letter monstrosity will do little to win your interest. It's no wonder that the SEC chairman often uses the more colloquial phrase "interactive data" when trying to sell the benefits of XBRL.
First conceived in the late 1990s, XBRL has been perpetually on the horizon, its promise always a day away. That's been partly due to technological complexity. Conceptually, XBRL is simple: a string of computer code, dubbed a "tag," is assigned to every line item in a financial statement, essentially identifying net income as net income, EBITDA as EBITDA, and so on. Once those figures are properly tagged, they can be reassembled into other reports at the push of a button, or pulled into software for analysis, or imported to or exported from any number of software packages, reports, databases, or other systems.
But creating the tags for thousands upon thousands of different fields has taken years, and the effort is not yet complete. At the same time, procedures have had to be created to allow companies to develop customized tags for numbers that they feel don't map to any existing standardized XBRL tag.
Today, the work is sufficiently advanced to suggest that XBRL may be ready for widespread adoption. And no one says that more often or more fervently than SEC chairman Cox. He has devoted at least four speeches to the topic and presided over an all-day XBRL roundtable in June. XBRL is "truly a revolutionizing and exciting topic," Cox has said, with "the potential to slash hours of waste, cost, and inefficiency — not just for users of financial data but for the companies that prepare it as well."
The SEC has stopped short of mandating XBRL for public filings, but there is a growing sense that the carrot may soon give way to the stick. So far, only about 25 companies have joined an SEC pilot program to file select reports in XBRL, and many of those appear to have some commercial interest in its widespread adoption. Executives at Oracle and Microsoft estimate that fewer than 10 percent of their customers use the XBRL features embedded in the companies' products.
There is a major disconnect between the companies that must generate numbers, few of which see much benefit to bothering with XBRL, and the organizations that analyze or audit those number, which see plenty. "The payback is difficult to quantify," said Comcast controller Lawrence Salva, during the SEC roundtable. "XBRL really doesn't do that much for the company, for the registrant," says Bill Ferko, CFO of Genlyte Group, adding that, to the extent that it enhances transparency, he supports it. At the other end of the spectrum is R. Christopher Whalen, managing director of investment research firm Institutional Risk Analytics, who says that "any company that can't get itself organized to submit tagged documents will soon be worthy of investor skepticism."
The incentives to get organized are increasing, but appear well short of reaching a tipping point. The SEC says that companies that file documents in XBRL format will receive expedited reviews. Adhering to the standard may win a company new fans on Wall Street, as well: with the efficiencies of XBRL, investment analysts should be able to cover more companies, a boon to CFOs in small or niche companies. "Right now, analysts track only 8 to 10 companies, in part because they have to rekey data into Excel spreadsheets. With XBRL, that's all done for them, so they have time for more companies," says Brad Homer, XBRL technical manager at the American Institute of Certified Public Accountants.
Competitive analysis may also get a boost. While there may be no inherent improvement in comparability — your competitors may still obscure their calculations for operating expenses or segment data — the technology makes it faster and easier to work with financial reports. In fact, Edgar Online says about half of its clients for a new XBRL data feed are corporations looking for a tool to analyze competitors and screen potential acquisitions. (On the flip side, such transparency means investors — particularly individual investors — will be more empowered to take CFOs to task. "The interactive data highlights differences, gaps, and inconsistencies that you never would have been able to see in an efficient way before," said Trevor Harris, client-services vice president at Morgan Stanley, during the SEC's June roundtable.)
Most intriguing of all to CFOs is the potential for improved (and perhaps cheaper) audits. Some auditors are already using XBRL-coded data for client risk-assessment purposes. The tagged data "allowed us to do more analysis, do it quicker, and allows our people to spend time looking at results rather than just manually assessing and validating the data," says PricewaterhouseCoopers partner Mike Willis, a founding member of XBRL International, a consortium of hundreds of companies, banks, accounting firms, and other parties developing XBRL standards. PwC began converting documents for its 50 largest clients into XBRL for analysis purposes last fall. Willis says that XBRL provides more-efficient access to and analysis of company data, and "will allow us to do better audits at lower cost," although he says it's too early to comment on how much costs may drop.
Who's in Charge?
But it's premature to ponder the auditing issues posed by XBRL-encoded statements when there are so few of them being produced. If the many potential benefits of XBRL can't win the technology enthusiastic acceptance, what can? Perhaps, some say, a changing of the guard. Currently, responsibility for developing the XBRL taxonomy rests with the private, volunteer-based XBRL-US, which says it has about a year to go to finish the project.
That arrangement may crumble if XBRL is widely adopted and new tags need to emerge concurrent with new accounting standards. "We are actively seeking additional resources," says the AICPA's Homer, who bears the distinction of being the only person to pull down a full-time salary for his work on XBRL. He says the consortium hopes to add up to eight full-time staff positions to the project.
One possible source for those additional resources: the Financial Accounting Foundation (FAF), which oversees the Financial Accounting Standards Board. While no official pronouncements have been made, in May FAF received permission from its board of trustees to conduct a feasibility study on funding and overseeing the development of the XBRL taxonomy, and it has hired a new director of financial-reporting technologies to carry it out. The hope, says Robert DeSantis, FAF's chief operating officer, is to figure out how costly and contentious such an effort would be, and make a recommendation to the trustees by the end of the year. Questions as to whether FAF involvement would complement the work of the voluntary consortium or supersede it remain unanswered.
The specter of having the same organization that writes accounting standards also write the code that complements that effort has some worried that widespread adoption of XBRL will be a backdoor way for FASB to dictate the chart of accounts that companies must report. "The fear is that once you have built out a taxonomy," says research director Robert Kugel of Ventana Research, "it becomes possible to, at a very granular level, mandate that certain kinds of items roll up in a specific hierarchy that would be common across all businesses." But Kugel draws a big distinction between what is possible and what is likely. The current SEC leadership, he says, appears to be "absolutely opposed" to mandating a common chart of accounts. And FAF officials insist that the technology and accounting standards will remain separate efforts with separate staffs. "At the end of the day, we are in business to establish reporting standards," says DeSantis. "We're not going to take our eye off that... [and] the FASB staff won't drive XBRL, because they've got their own work to do."
But that leaves unanswered the question as to whether the SEC will mandate the use of XBRL in the filing of financial reports. Given that FAF and FASB are funded by the SEC, wouldn't their direct involvement be tantamount to an SEC mandate? While the AICPA's Homer says that the SEC has been "very upfront in saying it doesn't want XBRL to be a mandate," others believe it's only a matter of time. Mark Link, chief administrative officer of EMC; John Stantial, director of financial reporting at United Technologies Corp.; and Whalen of Institutional Risk Analytics all predict that the SEC will eventually take a firmer stand. "I think you'll see a rule-making or comment process within the next 12 months," says Whalen.
Getting There from Here
Even amid the uncertainty, companies that have adopted XBRL say the difficulties have been minor. When computer-storage vendor EMC decided to adopt XBRL, its biggest challenges were in shopping for the right software package (it ultimately chose Rivet Software's DragonTag) and tapping the vendor's support team for a fair amount of hand-holding. UTC began its project with the same software but had to upgrade to a more sophisticated (and expensive) product from Fujitsu as it sought to code financial data beyond that found in its 8-K.
That said, embracing XBRL is still simpler and less costly than nearly any other data-standardization effort one could cite. UTC has spent a total of about $35,000 and 500 hours on all of its filings to date, but that included about $10,000 for the Fujitsu tool. EMC's Link estimates the cost at about $2,000 per quarter plus a few days of internal staff time. A brief survey of software vendors finds that costs are generally under $1,000. Companies will likely require a few hours to a few days to set up the first template document for a filing, which can then be reused.
And there's outsourcing, an option that Xerox, Automatic Data Processing, and PepsiCo, among others, have taken. "It was pretty much a nonevent," says Chuck Callan, ADP's senior vice president of regulatory affairs, referring to the company's recent 8-K XBRL filing. With outsourcing, the entire process from decision to filing took about 20 to 25 hours of work time, he says, and future filings are expected to take just 5 to 8 hours. Prices currently range from $6,000 for an earnings release to $19,000 for an entire 10-K or 10-Q.
Xerox got an even better deal: free. It worked with printing giant RR Donnelley & Sons, which is investigating offering XBRL services as an adjunct to its traditional printing of annual reports and related financial forms. "We're trying to work back and forth with Donnelley to make sure what comes back looks like what went in," says Xerox chief accounting officer Gary Kabureck — a common theme in these early days as client and outsourcers ascend the learning curve together. "Our objective right now is to make sure cost isn't an issue for [clients]," says Sherad Cravens, director of business strategy and research at Donnelley. The company has teamed up with Edgar Online to provide the coding service to its clients. Other outsourcers include Rivet and CoreFiling.
In fact, because XBRL offers a way to integrate disparate forms of financial data simply by tagging them with standardized code, companies may find that any number of software companies and related-service providers are only too happy to offer the capability at little or no charge. As the only company, so far, to have filed a 10-K in XBRL, Microsoft may be testing the implications of the technology for its industry-standard Excel spreadsheet software. Payroll processor ADP hopes XBRL features will entice more of the companies it serves to select electronic proxy statements over paper. "Because we think XBRL can offer something to our customers, we want to get our feet wet with it," says ADP's Callan. Xerox hinted that its interest in XBRL may net it more than SEC applause. "Anything that can help documents work better, smarter, and faster is central to our research-and-development strategy," says Kabureck.
XBRL may ultimately satisfy some important company data-integration issues that have little to do with SEC filings. In theory, it could eliminate the need for wrapping every business unit into an ERP system, allowing different units more autonomy in the software they choose as long as the numbers produced by the software are tagged in XBRL so as to move easily from one system to another. PwC's Willis points to Wacoal, a Japanese company that layered XBRL on top of a number of ERP systems instead of converting to a common system. "I think it will work as well in the United States, maybe even better, because so many companies have multiple ERP systems as the result of acquisitions," Willis says.
Using XBRL internally could also make it easier to publish external financials to any other agency lender that required them. "If agencies other than the SEC embrace it, I could tag data right in the database, and then feed it to the SEC, the IRS, and the Bureau of Economic [and Business] Affairs as they need it," says UTC's Stantial. "You take out a ton of manual effort, and the possibility of mistakes."
Which brings us back to where we began: possibilities and promises, the suggestion of some momentum, but still plenty of uncertainty. If XBRL is to prove as revolutionary as its supporters claim it can be, it seems that something, or someone, will have to rally the masses far more effectively. Whether the proper motivation comes in the form of a powerful idea or from the looming shadow of the guillotine remains to be seen.
Alix Nyberg Stuart is senior writer at CFO.
XBRL: A World of Options
From mainstream software companies to newer niche players, XBRL products and services are beginning to proliferate.
Offers an XBRL module that pulls tagged data into the company's Cartesis 10 products for benchmarking and M&A analysis; offers an XBRL Publication module that allows companies to self-tag data and generate XBRL instance documents.
Offers XBRL consulting; tags companies' reported data for SEC filings and press releases (in conjunction with BusinessWire).
Tag companies' reported data and validate accuracy of XBRL self-tags for SEC filing. Also offers I-Metrix, a financial-data feed in XBRL format, for $7,000 per seat.
The Interstage XWand product allows companies to create custom tags, as well as tag basic documents; can also tag internal data and unify data across ERP systems. Starts at $2,000 per seat.
The System 9 Financial-Reporting Module includes an XBRL component that allows users to consolidate, analyze, and report financial results in XBRL formats.
Excel 2003 and 2007 both support custom-defined XML schemas such as XBRL, allowing users to construct, publish, and analyze XBRL data. FRx also allows users to convert internal numeric data into XBRL.
Releases since 11.5.8 (more than five years ago) have supported XBRL and have been updated for most recent taxonomies. Customers can convert internal general-ledger data into XBRL and publish rolled-up data in XBRL for external-reporting purposes.
The $800-per-seat DragonTag product lets you tag Word and Excel documents in XBRL for reporting purposes; Crossfire Analyst (in beta) collects and analyzes XBRL data. Rivet also tags companies' reported data for them for SEC filings.
In conjunction with Edgar Online, tags documents for clients from start to finish or validates self-tagged data from companies.
Risk Intelligence and Financial Intelligence products allow users to analyze XBRL data and publish results with XBRL tags.
Tags company data for reporting purposes. Packages start at $3,900.
XBRL Taxonomy Builder, XBRL Composer, and XBRL Integrator let companies create custom tags, as well as tag basic documents; some tools can also tag internal data and unify data across ERP systems. From $2,500 to $6,500 for single users.
XBRL Taxonomy Designer and UBmatrix Report Builder tools help companies create custom tags, as well as tag basic documents; can also tag internal data and help unify data across ERP, CRM, and other systems.
Source: The companies