Comments Requested on Newly Named Version of XML for Business Reporting

SPECIAL REPORT

The first specification for a business-reporting variant (XBRL) of extensible markup language (XML) was the subject of an April symposium. The XBRL project committee held the meeting to encourage comments on the specification and to announce its new name. Until then, it had been known as XFRML—for financial reporting. (See Highlights, JofA, Apr.00, page 6.)

Committee representatives said XBRL will enable issuers and users of business information to share it more accurately and conveniently through various electronic applications and media, including the Internet, where an increasing number of companies are disclosing financial information.

Prepared under the leadership of the AICPA, the initial specification pertains to the financial information of only commercial and industrial companies reporting according to U.S. GAAP. Specialists in other industries and national reporting standards are preparing further specifications that the committee expects to issue over the next 18 to 24 months. XBRL conforms with accepted accounting practices and standards and does not require reporting companies to disclose additional financial information.

XBRL uses XML-based “tags” to describe data elements contained in public and private companies’ financial statements. These descriptive attributes make it easier to assemble information in a way that facilitates reissuance in a wide variety of formats (for example, printed financial statements, HTML documents for a company’s Web site, EDGAR documents filed with the SEC and customized credit reports or loan documents), without data re-entry or other time-consuming manual processes. Consequently, companies will find it cheaper to prepare and distribute their financial statements, and investors and analysts will have easier access to reliable information they need to make informed investment decisions. For example, XBRL will make it easier to automate comparisons of financial information, accounting policies, notes to financial statements between companies and other items that today are processed manually.

The XBRL project committee consists of representatives from more than 30 organizations, including the Big Five accounting firms, the IASC, software makers Oracle and Microsoft and financial services provider Morgan Stanley Dean Witter.

At the symposium, AICPA Chairman Robert K. Elliott said that a reported 25% of the $1 trillion spent annually in the United States on medical care is wasted. XBRL, he said, can help the financial sector focus on cost-saving opportunities of the same magnitude. He added that XBRL implementation costs will be minimal for CPA firms and their clients because the specification will be “built into” software and operating procedures.

But Elliott reserved much of his enthusiasm for the project’s other possibilities. “XBRL’s greatest benefit will be in terms of [the financial sector’s] improved ability to create additional value. By reducing information asymmetry,” he said, referring to inappropriate combinations of financial data elements that appear synonymous but aren’t, “we can improve liquidity between suppliers and users of capital and facilitate movement of capital to the new economy.”

Mike Willis, a partner of PricewaterhouseCoopers and chairman of the committee, urged symposium attendees and the financial community in general to submit comments on XBRL’s conceptual structure, or “taxonomy”—a list of data elements the committee selected as the most relevant and most commonly required under U.S. GAAP.

Later, in an interview, Willis told the JofA that XBRL works effectively with other XML-based financial specifications, such as Open Financial Exchange (OFX), which enables financial institutions, businesses and consumers to communicate financial information to each other over the Internet. “OFX describes individual transactions and XBRL reports summaries of them—a highly complementary relationship,” he said.

The conceptual structure also specifies each element’s relationship to others. For example, “current assets” and “assets” are data elements—the former “rolls up” into the latter. The draft version contains more than 1,100 such items, arranged in a hierarchy, that pertain only to commercial and industrial enterprises.

The meticulous detail of XBRL’s structure is essential to ensuring the clarity and accuracy of information, in accordance with existing reporting standards, as it shuttles between the often dissimilar software applications companies, lenders, investors, auditors and regulators use.

The present version of XBRL is in English, but the committee plans to issue future releases in other languages—for example, French and Spanish, in which a significant portion of the world’s financial reporting appears.

Ultimately, XBRL will benefit a wide range of stakeholders—the reporting companies, accountants, regulators, analysts, investment advisers, capital markets and lenders, as well as software developers and data aggregators. Committee representatives said XBRL is designed to balance the needs of these groups, ensuring its convenience and usefulness for each of them.

Since the conceptual structure is expected to expand and change in the coming months, the committee advised interested parties to stay abreast of new developments by registering at the XBRL Web site, www.xbrl.org , where the committee now is accepting comments on the draft version until June 5, 2000. It expects to publish a final version of the release by the end of July.

—Robert Tie