Return to current issue.
The XBRL Revolution: Playing Tag with Powerful Information
By Ira Kaplan
XBRL, eXtensible Business Reporting Language, is an XML (eXtensible Markup Language) based digital document markup language that has the ability to greatly improve how we store, retrieve, share, and understand financial information. Markup languages such as XBRL use tags like <company_name>, <StartDate>, and <OtherAdministrativeExpenses> to describe the contents of Web pages, documents, and databases. These tags are the beginning of the Semantic Web, bringing meaning to the World Wide Web, aka the Internet. Tim Berners-Lee, the inventor of the Web and now overall Director of the World Wide Web Consortium, W3C, and a Senior Research Scientist at the MIT Laboratory for Computer Science, envisions the Semantic Web which "understands and satisfies the requests of people and machines to use Web content." The Semantic Web will be an extension of today's Web eliminating the silos hobbling the easy finding and correlating of information people need.
The benefits expected from using XBRL are so great that on May 14, 2008, the Securities and Exchange Commission "voted unanimously to formally propose using new technology to get important information to investors faster, more reliably, and at a lower cost." XBRL is at the center of this new technology initiative announced by SEC Chairman Christopher Cox. The SEC's press release stated "The proposed rule would require all U.S. companies to provide financial information using interactive data beginning next year for the largest companies, and within three years for all public companies." Conrad Hewitt, the SEC's Chief Accountant stated "Accounting is the business language of the world, and interactive data [implemented using XBRL] will become an easy and reliable technology to improve that language worldwide..."
Here's a glimpse into the emerging world of XBRL. Today, using Microsoft Excel, I am able to perform Web data queries upon documents such as financial statements published through EDGAR, the SEC's Electronic Data-Gathering, Analysis, and Retrieval information service (www.sec.gov). I can directly download many types of data tables including financial statements. Any set of data arranged as an HTML (Hypertext Markup Language) table is instantly identified as suitable for downloading. I simply select the table, import it into Excel, and it immediately appears as a properly formatted worksheet. All that's left is some cleaning up and then I'm ready to analyze. The important thing is the absence of data entry, that error prone process that can lead to serious mistakes. Yet, as good as this sounds, this will seem primitive within a few years, thanks mostly to the advances in XBRL.
Imagine you want to extract only three data elements, net property assets, depreciation and capital lease obligations, from the financial statements of seventy-eight companies presented as seventy-eight separate HTML documents on the Web. The result is to be a single, well formatted worksheet assembled without the tedious, onerous, and error prone copy and paste process. And instead of tweaking just one more worksheet cell with a manual calculation to get it just right, you want to implement robust business rules to transform the data.
XBRL provides the answers. XBRL is the descendent or sibling of document markup languages such as GML, SGML, HTML, and XML, created to add structural and semantic information, metadata or data about data, to documents. Metadata are often presented as tags such as <title>, <body>, and <customer>. XBRL itself is defined using XML and many of us are unaware of the presence of XML running in the background of many databases and interactive information systems such as online electronic retailing catalogs.
My current method of downloading HTML tables into Excel eliminates the risks introduced by manual data entry and copying and pasting. But there is nothing that defines the content of each cell value across various horizontal fiscal periods, such as accounts payable at October 31, 2007 and October 31, 2006. HTML's tags describe structure but not meaning. XML dialects such as XBRL add tags defining the document's contents, important value identifiers for each and every cell so I would not have to worry about data integrity. I could change the layout substantially and yet each cell would "know" what type of value it contains.
The key to understanding XBRL comes from an understanding of XML, the eXtensible Markup Language. XML is a general purpose specification language, a language for creating customized document markup languages such as XBRL for business reporting; MathML, the Mathematics Markup Language for displaying mathematical expressions; MusicXML for music notation; and RTML, the Remote Telescope Markup Language for controlling remote and robotic telescopes. In all these cases XML makes information easy to manage and share.
The first widely used markup language for Web pages and documents was and is HTML, used to control how text is displayed in Web browsers such as Internet Explorer, Safari, and Firefox. XML goes beyond HTML and enables the creation of customized tags describing content and not presentation, using tags like <company_name>, <OtherAdministrativeExpenses>, and <StartDate>. Since XML is concerned only with describing the data content and not presentation on Web pages and documents, XML relies upon XSL, the Extensible Stylesheet Language Family, or CSS, Cascading Style Sheets, for specifying the layout of a Web page including text color, typography, and position.
To ensure a Web page or document using XBRL is well formed, correctly uses the XBRL specific tags, an XML Schema is defined. The standards organization for the Web, the World Wide Web Consortium, states "XML Schemas express shared vocabularies and allow machines to carry out rules made by people. They provide a means for defining the structure, content and semantics of XML documents in more detail."
When initially structuring documents using XBRL, public companies would incur higher costs, but there are valuable benefits to many parties, including:
- More powerful and insightful analytics into industry trends, with nearly zero data risks.
- Analysts and researchers would produce and publish better research, improving investment advice and qualitative insights into the industries they cover.
- Shareholders could analyze their holdings with greater flexibility and accuracy.
- Analysts and end-users of research would be able to quickly exchange highly customized analytics, ready for transformation, resulting in larger and more powerful knowledge communities globally.
XBRL's adoption is increasing and we believe as part of the Semantic Web it will bring us a new and exciting phase in the evolution of financial analysis and research.
Resources:
- An Introduction to XBRL: www.xbrl.org/WhatIsXBRL
- Q&A with Tim Berners-Lee. BusinessWeek.com Special Report April 9, 2007: http://www.businessweek.com/technology/content/apr2007/tc20070409_
961951.htm?chan=search
- S.E.C. Moves Toward Requiring Interactive Data Filings. Grant Gross, IDG News Service\Washington Bureau, IDG. May 14, 2008: http://www.infoworld.com/article/08/05/14/SEC-moves-toward-requiring-interactive-data-filings_1.html
- SEC Proposes New Way for Investors to Get Financial Information on Companies: www.sec.gov/news/press/2008/2008-85.htm
- Semantic Web: http://en.wikipedia.org/wiki/Semantic_web
- Spotlight on Tagged Data and XBRL Initiatives for EDGAR Filings: www.sec.gov/spotlight/xbrl.shtml
- XML: http://en.wikipedia.org/wiki/Xml
Return to current issue.
