New York—Bear Stearns Equity Research recently released a report on XBRL (Extensible Business Reporting Language), an electronic coding language designed to facilitate the use of financial and non-financial information from press releases and financial statements. Although XBRL is currently used by only 54 US companies on a voluntary basis, the Bear Stearns report suggests that the SEC will make XBRL reporting mandatory. If so, this will provide a boost to the adoption of RIXML (Research Information Markup Language), which provides standardization for the distribution of investment research.
XBRL is designed to create a consistent set of tags on financial data so that users can select financial and non-financial information from press releases, 10-Qs, and 10-Ks. To work, companies must first adopt XBRL in their financial reporting process and “tag” data in the filings with XBRL. XBRL was first proposed and created in the United States in the late 1990s and is being developed by XBRL International, a not-for-profit consortium of approximately 550 companies and agencies worldwide which are creating an international open standard.
Here’s an example of how XBRL works. Assume that a company released its 10K with XBRL. An analyst builds a cell into Excel or other software that is automatically linked to the XBRL tag for net income (usfr-pte:NetIncome). As a result, the data automatically links to this code and imports the information into the spreadsheet. There is no manual entry and no data vendor intermediaries. The analytic process begins immediately after a filing’s release rather than being delayed by the information collection process.
The benefits to analysts on both the buy and sell side are clear cut. As the Bear Stearns report states, XBRL makes the financial information collection process “faster, cheaper and more accurate.” The view from financial information data providers is more nuanced. Bear Stearns cites two issues with commercial databases:
Some analysts collect financial information directly from financial filings, while others simply download information from fee-based databases such as FactSet or Bloomberg. Whether analysts collect the data themselves or outsource data collection to database providers, this information is prone to the same common types of problems and errors that affect its reliability and relevance. First, there is the collection time delay between when information is released by companies and when it is manually entered into spreadsheets and databases. Further, data collectors often misidentify or mistype information.
The other issue is cost. If XBRL makes the process faster and more accurate, why continue to pay for third party data? For now the answer is coverage and convenience, but over time these advantages dissipate. Analysts get other information from data vendors than just financial statement information, so mandatory adoption of XBRL will not result in a wave of Bloomberg or Factset cancellations. It will have more impact on specific products focused on financial statement information, such as Thomson’s Worldscope, S&P’s Compustat, and Hemscott Data. There will be ripple effects on other financial statement related businesses such as investor relations data providers and earnings estimate distributors.
U.S. Securities and Exchange Commission Chairman Christopher Cox has been a big booster of XBRL throughout his tenure. The SEC has publicly stated that it is “committed to transform its vast database of financial information nicknamed EDGAR, into interactive data format.” On September 25, 2007, Chairman Cox announced the completion of the initial draft of the United States Generally Accepted Accounting Principles (US GAAP) XBRL taxonomy. As a result, there are now approximately 15,000 standardized tags that can be applied to digitalized financial content. Previously there were only about 2,500 tags available, thus there was a greater need for companies to create their own tags for information where standard tags did not exist.
Will the SEC now make XBRL mandatory? John W. White, the Director of the SEC Division of Corporate Finance, the area responsible for managing the filing of financial statements, commented in June 2007:
“While these filers [Voluntary Filing Program members] are using XBRL on a voluntary basis, the future obviously could hold a requirement that tagged data be used in public company filings with the SEC. For my part, the Division of Corporation Finance will continue to be a key player in the SEC’s work with XBRL. If the time becomes appropriate, we stand ready to draft the needed rules requiring tagged filings.”
It is likely that the SEC will make XBRL mandatory (although as with many areas of Cox’s SEC, timing is a question.) What is also unclear is whether XBRL will be required for the entire financial statement or simply the most commonly used portions. XBRL has the ability to standardize the juicier parts of financial statements, notably the footnotes and management discussion. Some of the research firms we track derive significant value by mining the footnotes, particularly the forensic accounting and earnings quality specialists. If XBRL is applied to the full financial statements, some of these providers will be impacted.
As with an increasing number of regulatory issues, the US is falling behind other regulatory domiciles. XBRL is yet another example. Although first developed in the US, adoption of XBRL is growing faster in some non-US jurisdictions. To cite one example, as of September 2005, more than 800 companies in China used the XBRL taxonomies developed by the Shanghai Stock Exchange to report their 2005 half-year reports. Since the US has only 54 companies (out of 10,000) currently reporting, Bear Stearns snarkily calculates that China had approximately 16 times more filers in 2005 than the U.S. currently has.
Once the SEC does mandate XBRL, one of the beneficiaries will be RIXML, the open standard for the distribution of research. RIXML is a cousin to XBRL since it provides similar benefits for the downstream distribution of investment research. Once the buy side analysts are busy adding XBRL tags to their worksheets, they will also be looking to add RIXML tags for earnings estimates and other research items. The RIXML organization, a consortium of buy side and sell side organizations which are creating a common taxonomy for investment research, recently announced a new executive director, Jack Roehrig, who joins from Merrill where he managed the electronic distribution of Merrill’s investment research. The best thing for Roehrig and the RIXML consortium would be for Cox to follow through on his promises…