Behold the Renaissance: The End of The Dark Ages in Financial Analysis
The relationship between analysts and corporations is changing and data tagging will play a critical role in corporate transparency and analytical research. Here’s how.
November 1, 2007
by Elias-John Kies, CFA
Analysts have long dreamt of the day when they would no longer need to sift through hundreds of pages of Securities Exchange Commission (SEC) filings and other regulatory documents for hidden nuggets of intelligence. Thanks to the SEC’s program for eXtensible Business Reporting Language (XBRL) this dream is becoming a reality.
XBRL attaches “tags” to items in company filings and is poised to become the first global dictionary of financial reporting. While this is a long-term goal, national XBRL taxonomies (or data definitions in plain English) have made considerable headway. As the U.S. advances its XBRL program, countries such as China and the Netherlands have already mandated XBRL filing for all public companies regulators in those countries oversee.
In September, SEC Chairman Christopher Cox took an important step towards this goal by announcing the completion of a US GAAP (Generally Accepted Accounting Principles) XBRL specification containing more than 15,000 tagged financial elements from such reports as 8Ks, 10Qs and 10Ks. Chairman Cox also discussed a potential timeline for mandating XBRL filing by the end of 2008. With a possible mandate in the near future, issuers and users are well advised to become knowledgeable in XBRL as their document filing and analytical workflow is about to drastically change.
Current Methods of Data Retrieval
Today, analysts use two methods of data retrieval: manual extraction and third-party data vendors. The first method, manual extraction, involves locating and retrieving the filing, searching through the filing for the desired information and placing this information in a spreadsheet. Many organizations do this to control quality in their dataset. However, due to time constraints, this method may limit the number of companies analyzed or depth of detail studied. For instance, a sell-side equity analyst, on average, will cover 20 companies. This coverage can be increased when, instead of publishing documents, companies report datasets of their financials through XBRL.
The second method of data retrieval is through a data vendor. Depending on the analysis, the data provided may not be granular, accurate or timely enough. Most data vendors aggregate items (i.e. revenue) rather than provide a breakup of component parts. Figure 1 represents the difference between information from a traditional data vendor and information provided with an XBRL data-download for MGM Mirage.
Whether manual data entry or vendor data is used, both have issues with accuracy. A recent report by Bear Stearns highlighted a significant percentage of erroneous data due to human error, i.e. typing incorrect information into the spreadsheet or database. Before beginning an investment project, a team could spend weeks correcting data provided by a vendor (see Dane Mott, Janet Pegg, Christopher Senyek, Adam Calingasan; XBRL: The Investor’s Path to Better, Faster, & Cheaper Financial Information; October 5, 2007; Bear Stearns & Co. Inc.).
XBRL eliminates the tradeoff between timeliness, granularity and accuracy. As soon as a press release or other filing is available, an equity analyst will have their financial model updated automatically and the analyst will be prepared to ask pointed questions on a subsequent earnings call.
Timely and accurate data is of prime concern with quantitative hedge funds, the majority of whom select investments from a large universe of thousands of securities (for example, any stock contained in the Russell 3000 index). Quant funds (a small family of equity funds with a quantitative investment style) receive data feeds from vendors and run the information through their proprietary trading algorithms to identify opportunities. Many trades are automatically executed, so erroneous data could lead to unwarranted and costly trades. This threat is virtually eliminated through XBRL, as the data feeds are provided directly from the source documents.
Specification vs. Standardization
A key debate in the adoption of XBRL is the amount of specification versus standardization: the need to balance uniform labels and tags of financial concepts versus the need and desire of companies to report unique items. Under the current voluntary filing program with over 60 participants, a company can choose to create new tags when it feels existing tags would lose the unique context of the reported item. Chairman Cox noted that this feature was critical, since a nonflexible structure could quickly face stiff opposition from filers. This concern was addressed in the updated taxonomy because the number of tags was expanded from 2,500 to over 15,000. To capture company-specific uniqueness while maintaining comparability to other firms, XBRL adopts a structure of parent/child relationships. An example of this structure is displayed in Figure 2.
While companies can currently create new tags as a “child” or “parent”, this may change in the future where tag creation is allowed only at the “child” level. Analysts do not want companies forced into arbitrary tags. However, analysts also do not want companies to create tags without a structured system since comparability between companies is critical.
In August 2007, CFA Institute completed an XBRL survey with users of financial reports. Ninety-six percent of respondents believed a structured approach to creating extensions is appropriate, 66 percent believed companies should have limited ability to create new extensions, while 50 percent felt this structured approach should be completed by an independent auditor. It is clear that the market demands assurance services to verify the company has completed its financial reports with appropriate tags, using existing tags when available (see Melissa Looney, XBRL Survey Report; August 27, 2007; CFA Institute).
The Bottom Line …
XBRL represents the most granular, accurate and timely financial data the market has ever accessed. In addition to increasing the number of companies monitored, analysts will be able to spend their time on value-added activities such as completing more in-depth research into companies and market trends. Red flags or anomalies will be identified with greater precision.
To speak a common language, users and issuers need to understand the premise of the XBRL tagging system but by no means need to be experts in the workings of the technology specification. With the high degree of transparency coming from XBRL tags, any attempts to conceal information would be immediately obvious and raise analyst’s questions. Companies will require guidance from their auditors to properly allocate the financial items to XBRL tags. Helping companies make this transition from publishing a document to publishing data represents a significant opportunity for the accounting profession. XBRL is clearly the new definition of “as-reported.” It has the ability to revolutionize financial reporting, the way the Internet revolutionized advertising.
External reporting is one application of XBRL adoption. Taking the common chart of accounts defined by XBRL into the heart of a company’s financial supply chain will provide incredible benefits to companies and the accounting profession. The opportunity within the next few years is exciting for everyone!
Rate this article 5 (excellent) to 1 (poor).
Elias-John Kies, CFA, is a senior analyst for Norwalk, Connecticut-based EDGAR Online Inc. (NASDAQ: EDGR) — a provider of global business and financial information and a leader.