Generation X

XBRL is revolutionary; the most important advance in business reporting since the spreadsheet; an idea whose time has come; a good idea that has been turned into a monolith; or yet one more confusing IT acronym. It all depends on who you ask, because the eXtensible Business Reporting Language (XBRL) has an identity problem. Although XBRL can make it easier to analyse, publish, use, re-use and exchange finance-related information, from annual reports and financial statements to reference documents and tax filings, most accountants are either unaware it exists or have no idea what it is. “I’d never heard of it until Lord Carter mentioned it recently during a presentation on corporation tax,” says one sole practitioner who wishes to remain anonymous.

“I’ve recommended that by 2010 corporation tax returns from all companies should be filed using XBRL,” says Lord Carter of Coles. “Currently, customers and their agents use 20th century preparation and 19th century delivery,” he adds. But if they want HM Revenue & Customs (HMRC) to provide a 21st century service then they will need to use the latest methods and means.

It is a good job he is prepared to give everyone involved a few years to get used to the idea, because XBRL development tends to follow a long and winding road. “I want to give HMRC and software houses the time to develop and prove their services and products,” explains Lord Carter. “What I have in mind is a fully operational service based on XBRL from 2008 that enables a stepped increase in numbers over the next two years.” Only time will tell how realistic this expectation is; even XBRL evangelists admit the technology is widely misunderstood.

“Raising awareness is an issue,” says Mike Willis, a US partner with PricewaterhouseCoopers and a founding member of XBRL International, the consortia responsible for developing the technology. “XBRL isn’t a software application or a language,” he explains. “It’s an enabler.” Despite its misleading name, XBRL is a standard for the distribution and exchange of financial information that describes financial information using XML data tags (see box, below). The tags work in a similar way to bar codes, but instead of making it easier to handle assets or manage the supply chain, XBRL makes it easier to identify different items of financial information.

By defining an XBRL tag to describe an individual item such as inventory, asset turnover, current liabilities and residual income, the item becomes easier to identify, regardless of national and international differences in reporting standards, languages and terminology. By using XBRL tags to describe the information created or collected by various business applications and stored in disparate and incompatible databases, locations and systems, those items can be identified, exchanged and manipulated more easily. This could enable accountants, analysts, competitors, investors, regulators and other stakeholders to benefit from faster and cheaper access to cleaner, more transparent and usable information.

Complex

Well, that is the theory. So why has XBRL been so slow to catch on? “It may seem conceptually simple,” says Chas Roy-Chowdhury, ACCA’s head of taxation, “but complexity has been an issue.” And developing standards is never easy, as Willis explains: “We spent the first few years just getting the necessary experts together,” he says, “then, in 2003, XBRL International released the commercial specification that software vendors needed to develop XBRL-compliant products, and low-cost implementations became a possibility.”

But creating the tags for thousands upon thousands of fields that make up the dictionaries, or taxonomies, has been time-consuming, and the effort is not yet complete. “The development of the taxonomies has been very slow,” comments Roy-Chowdhury. There are numerous taxonomies each put together by a wide range of interested parties, and the challenge of reaching the necessary consensus in a timely fashion appears to have been too much for some.

Available taxonomies now include IFRS-GP, the world’s first full financial statement taxonomy, developed by the International Accounting Standards Committee Foundation, with the XBRL International developments XBRL FR and XBRL GL. The former of these is a definition for a financial reporting language, the latter a definition for direct data mapping. But for these to have any real impact on the business world, software suppliers must use them as the basis for product development or enhancement.

Cartesis, CaseWare, DecisionSoft, Edgar Online, Fujitsu and Microsoft are among the few that have built XBRL capabilities into existing applications or created new products using the technology. Cartesis offers software that enables companies to self-tag data and generate XBRL documents and pull tagged data into its Cartesis 10 product for activities such as benchmarking. Edgar Online tags companies’ reported data and validates the accuracy of XBRL self-tags for SEC filing, and recently launched I-Metrix, a financial data feed in XBRL format. Some versions of Excel allow users to construct, publish and analyse XBRL data, and Microsoft’s small-business analytics application FRx will convert internal numeric data into XBRL. In general, however, the software industry remains largely underwhelmed.

“As a matter of principle, our members support electronic business reporting,” says Dennis Keeling, chairman of BASDA, the organisation that represents more than 250 of the world’s leading software developers. But it needs the support of those developing XBRL’s specifications and taxonomies. “XBRL GL is still in its infancy and not yet established,” he says, “and although BASDA wants to see XBRL FR adoption increase”, this is unlikely. “The standard XBRL FR taxonomies that define the use of XBRL need to be simplified to enable the participation of a broad range of packaged-software developers,” suggests Keeling.

BASDA has been involved in some pioneering XML implementations, but XBRL is proving a difficult beast to tame. “Unlike standard XML messages, which can be mapped directly to the underlying software data model, XBRL FR requires a user interface so that organisations can map their unique general ledger charts of accounts to the many and varied XBRL FR taxonomies,” he explains.

“The standard XBRL FR taxonomies currently lack the presentational structure required to make simplified financial reports easy to use,” adds Keeling. “Developers of packaged financial software would be able to implement elementary support for XBRL FR as a standard feature in their applications, if the taxonomies were simplified and the presentational structure made more workable.”

It is hoped that all of this will happen in time for Lord Carter’s proposals to be put into place. At the moment, both HMRC and Companies House are advancing towards the use of XBRL in small, manageable steps, giving themselves, software vendors and companies plenty of time to adjust. But XBRL-based applications remain thin on the ground and real live business users even more so. And if history is anything to go by, it is hard to see that situation changing any time soon.

Lesley Meall is a writer on business and technology issues.