The Securities and Exchange Commission can mandate use of XBRL technology for financial reporting all it wants. Ultimately, however, the investing public must take one key action to make companies’ investments in XBRL worth all that time and money.
They must trust that XBRL-tagged numbers work.
That’s where validation—the art of confirming that, yes, all the data in a company’s financial statement really are what XBRL says they are—enters the picture. Without it, a company’s XBRL software might tag inventory numbers incorrectly and they appear on investors’ computer screens as property, plant, and equipment. Or deferred revenue is described as depreciation, or income as expense. You get the picture.
In theory, validation will ensure such mistakes are caught and corrected as they happen. In practice, however, validation is one part binary logic of software and one part “know it when you see it” human judgment. The only thing certain about XBRL validation with respect to SEC filings is that it will evolve over time.
In its proposed rule to mandate XBRL, the SEC says it expects validation software to do several things. Some are the simple checks and calculations computers are good at, such as looking for syntax errors; tallying and identifying non-standard XBRL labels and tags and checking for mathematical errors; and “comparison of interactive data with disclosure in the traditional-format filing.”
Other validation the SEC expects is less straightforward, such as confirming that the XBRL tags a company uses do accurately explain how the company’s financial facts relate to each other.
If that last part sounds a bit elusive, don’t panic: That task, like most other validation chores, can be handled automatically. XBRL U.S., the non-profit consortium that developed the taxonomy of XBRL tags for U.S. Generally Accepted Accounting Principles, has 21 different tagging and validation tools on its Website. That number is sure to increase as XBRL filing goes mainstream in coming years.
Such software can spot an array of problems before it lets someone create an XBRL document, from putting text in a numeric field to creating company-specific tags that already exist in the U.S. GAAP taxonomy. The tools can also monitor a filer’s arithmetic, to ensure that assets still equal liability plus shareholder’s equity.
Rohan
Still, arithmetic isn’t always so simple. For example, automated summation of numbers (which is just simple addition, really) can be a challenge for XBRL software developers considering how diverse American accounting rules are, says Mike Rohan, president of Rivet Software. Rivet was an early XBRL software vendor and has worked with about 40 XBRL filers in the SEC’s voluntary filing program.
TAXONOMY LESSONS
The following excerpt from the XBRL U.S. GAAP Taxonomy Preparers Guide discusses Taxonomies.
A taxonomy is a description and classification system for concepts; an XBRL taxonomy is a particular way to describe and classify reporting concepts. XBRL represents each concept as an element with a name. The XBRL U.S. GAAP Taxonomy describes and classifies elements representing U.S. GAAP reporting concepts. XBRL taxonomies are electronic, machine-readable “dictionaries” consisting of many linked files containing thousands of elements linked to each other. The taxonomy contains human-readable labels such as financial statement line item captions, definitions, and applicable authoritative references for each element. The taxonomy specifies other attributes for elements, as described in greater detail later in this section.
The XBRL U.S. GAAP Taxonomy is based on generally accepted accounting principles in the United States (U.S. GAAP) and is intended for SEC registrants who file financial reports prepared in accordance with US GAAP. The XBRL U.S. GAAP Taxonomy has broad and deep coverage of financial reporting concepts (over ten thousand elements) including all U.S. GAAP and SEC financial statement disclosure requirements and many elements for commonly followed reporting practices. The taxonomy’s size and breadth minimizes the effort required to extend the taxonomies to meet specific reporting needs.
However, even if every line item on the face of the financial statements of a company had a corresponding element in the XBRL US GAAP Taxonomy, the preparer would still need to create an extension. This is because an extension taxonomy has more in it than just new elements to represent company-specific line items. It also specifies the form of statements (direct vs. indirect cash flow statement, for example), the ordering of line items, the disclosures that are relevant, the names of segments, and other reporting details.
Although the XBRL U.S. GAAP Taxonomies v1.0 will be maintained and updated to reflect changing U.S. GAAP and SEC financial statement disclosure requirements, as well as changes in common reporting practice, providing for every reporting possibility would be impossible. Preparers should understand that:
Meeting reporting requirements remains the registrant's responsibility.
The taxonomies should not be considered authoritative.
The taxonomies do not create any authoritative guidance.
The taxonomies should not be used as a GAAP disclosure checklist.
3.1.1 XBRL U.S. GAAP Taxonomies v1.0
This guide focuses on the XBRL U.S. GAAP Taxonomy, the main taxonomy that will be used for financial statement filings in the United States. The current release of the taxonomy was created through the organized efforts of XBRL U.S., Inc. with resources and input from members of XBRL U.S., public accounting firms, participants in the SEC’s Voluntary Financial Reporting Program, securities analysts, and others.
The XBRL U.S. GAAP Taxonomy will be maintained and updated and will be available to the public on the XBRL U.S., Inc. Web site (www.xbrl.us). Annual revisions of the taxonomy are expected, with releases during the year for reasons such as the issuance of new FASB guidance or technical corrections. Although preparers are not responsible for maintaining and updating the XBRL U.S. GAAP Taxonomy, they are responsible for using the latest version of the taxonomy.
The XBRL U.S. GAAP Taxonomy is written in computer language, but preparers may view and read it by using publicly available software. In addition, instance documents of registrants participating in the SEC Voluntary Financial Reporting Program appear on the SEC Website and may be viewed using one of the SEC’s free Interactive Financial Report Viewers, available along with other XBRL-enabled tools at www.sec.gov/xbrl.
Source
XBRL U.S. GAAP Taxonomy Preparers Guide.
He notes that XBRL has tags for allowance for doubtful accounts, net accounts receivable, gross accounts receivable, total accounts receivable, and current assets, just as companies would want to use. But not every company reports a number in every one of those fields, he says—and without some customization, an XBRL validation tool won’t know whether a blank field is an oversight or standard practice for a given company.
“It’s not just U.S. GAAP fluidity; it’s inherent in accounting,” Rohan says. “No matter how much work the SEC puts into it, no matter how much validation effort goes into it, the only thing that’s going to work is have 1,000 people file and have it shake itself out.”
Blaszkowsky
The SEC understands as much, says David Blaszkowsky, director of the SEC’s Office of Interactive Disclosure.
“We’re not necessarily looking to define upfront all the automatic validation criteria,” he says. “In many cases the marketplace will decide if a tag is appropriate or not. Something may be correct accounting-wise but may impede comparability. Let companies and investors negotiate that one.”
John Turner, CEO of XBRL consulting firm CoreFiling, says some accounting tasks just can’t be automated easily. “Knowing whether you’ve selected the right tags is extremely difficult to automate,” he contends. “Tools are being automated—making it easier to select and review tags—but to some extent that is an exercise of professional judgments by accountants.”
Andrew Neblett, CEO of EDGARfilings, agrees. “The bottom line is you need people and you need software for validation,” he says.
Turner
Both Rohan and Turner say today’s network of XBRL software and consulting companies would burst at the seams if it had to handle thousands of first-time XBRL filers all at once. But the field is growing and adding more players all the time, they say, and should be able to meet full demand by the time all U.S. public registrants must file in XBRL three years from now. (Blaszkowsky says the SEC proposed a three-year adoption period starting with 500 large filers with that industry dynamic in mind.)
Life Outside the Realm
Anyone curious to see how XBRL validation has worked elsewhere can start with the Federal Deposit Insurance Corp.; for nearly three years, it has required 8,000 banks to submit their FDIC filings tagged in a fixed 2,000-tag taxonomy. Commercial software pre-validates submissions at banks, using the same formulas the agency will later apply, says Jon Wisnieski, a senior FDIC information systems specialist.
For example, if a bank has real estate loans on the books, the software checks to see if it has reported income from real estate loans elsewhere. If the bank has fiduciary powers, it checks to see whether trust reports have been filed. If deposits suddenly skyrocket—often the result of a bank acquisition—the FDIC system automatically asks for a footnote, Wisnieski says. The program has created transparency and consistency of reporting and has let the FDIC get the information to the public more quickly, he says.
Of course, terms in U.S. GAAP are often anything but fixed, so such automated validation may be harder to achieve in SEC filings. Mark Bolgiano, president of the XBRL U.S. consortium, says validation can be viewed as a stack. At the bottom are the technical validation chores that software excels at; at the top are accounting professionals, making judgments and interpretations about how financial reality gets reflected in data.
Bolgiano’s colleague Campbell Pryde, XBRL U.S.’s chief compliance officer, says he imagines that “the stack will continue to develop over time, and different vendors will provide different solutions to improve quality and compatibility between instances.”
One document that may help bring order to the admittedly blurry situation is the XBRL U.S. GAAP Taxonomy Preparers Guide, released in May. In addition to the basics of XBRL document preparation, the book provides validation guidelines. But, Bolgiano says, the focus is data consistency rather than interpretation of GAAP rules.
“Just like desktop publishing never automated good writing, we don’t see XBRL automating good accounting,” he warns.
No comments yet