The Securities and Exchange Commission has not had a good year. Bear Stearns’ bailout, an evaporating pool of credit, and investigations from Congress have all compounded to put SEC Chairman Christopher Cox on the hot seat. Cox presumably has but a few months left in his tenure, and this lame duck will not be remembered as one of the SEC’s finest stewards. But there’s one initiative on the horizon that may actually make a difference: a transparency effort called XBRL that could drag Wall Street into today’s Internet age.
It’s a mouthful, but XBRL stands for eXtensible Business Reporting Language. It’s a programming language that essentially functions as a dictionary of terms that helps organize financial data. Just as scientists classify new organisms under certain taxonomies, XBRL allows accountants to classify financial metrics under agreed-upon categories. So, for example, when a company reports a particular figure for annual revenues, XBRL tags it in such a way that any computer can almost immediately find and categorize that figure as “annual revenues.”
For the last several years, about 70 companies have been voluntarily filing their reports using XBRL as part of an SEC pilot program. In May, the agency mandated that companies with a “public float” (that is, shares available on the open market) of more than $5 billion (approximately the 500 largest companies) have to file their 2008 annual returns using XBRL. The rest of the publicly held companies will fall in line over the next two years.
Before we cover how XBRL is going to make corporate America more Web-friendly, take a moment to appreciate the quaint way things are done currently. All publicly traded companies are required to file quarterly and annual financial reports to the SEC. Accountants tabulate metrics like revenues and gross margins and then file them in something called the EDGAR database. Basically, EDGAR is a reservoir of dull, numbers-heavy documents that are digital replicas of the paper-heavy reports destroying the environment.
EDGAR’s reports are all stored in HTML and TXT document formats, so while they’re digital, they aren’t easily manipulated. If investors, journalists, or institutional researchers want to start comparing stats between different competitors or contrast this quarter’s Netflix filing with last year’s annual report, they have to transfer the numbers into a spreadsheet by hand. This is an unwieldy and labor-intensive process—which means that unless you’re a monetary masochist, there’s little reason for you to comb through EDGAR’s innards.
The SEC thinks XBRL is a better way. Here’s how it should work: Netflix and Blockbuster both report their net incomes in annual and quarterly reports. As of now, they do that in a table on their EDGAR-filed quarterly return. Once they start filing under XBRL, their income numbers will be tagged “NetIncomeLoss.” Because all net income numbers are tagged uniformly, a computer can automatically pluck the numbers off the report and place them together into a database. Imagine the XBRL filing as a Rubik’s Cube that you can actually pick up, analyze, and rotate to try and crack the puzzle. The EDGAR filing is merely a photograph of the 3-by-3 cube—in order to shift the squares, you have to build your own cube based on the photo of the original.
All of this means that financial analysis is about to get crowd-sourced. Currently, institutional investors are the primary source of transcribed EDGAR data. Companies like Thomson Reuters, FactSet, and Standard & Poors lease out the data to Web sites, investment banks, and money managers. Once XBRL takes root in the financial community, anyone will be able to access and manipulate the data (through a techy process called “parsing”). The bankers’ analytical tools are no longer going to be the exclusive property of Wall Street; they’ll be the product of the masses on the Web.
A similar transformation has taken place in the political world. This year’s primary gantlet was memorable not only for the high drama of the race but also for the technological innovations that accompanied long election nights. Magic walls, exit-poll aggregators, and delegate calculators filled TV and computer screens, made possible by a technology related to XBRL (called XML, or extensible markup language). We could be nearing an era when an industry’s quarterly reports can be ranked and organized instantaneously according to the average investor’s whims.
At least one site is already preparing for the new XBRL age. Wikinvest, a startup out of San Francisco, applies the Wikipedia crowd-sourced model to business coverage. Users can edit nearly anything on the site: tweak text about a company’s financial past, add annotations to a company’s stock graph, or input data to something called WikiData. Currently, WikiData relies on users to transcribe data from EDGAR reports and provide links back to the files to prove the numbers. This, obviously, is not ideal. The creators of Wikinvest say that once XBRL goes universal, the data will automatically port into the site. When those data are combined with the site’s graphical and editorial content, Wikinvest believes it will offer users a contextual experience that only the bigwigs currently get.
This should make fiscal middlemen sweat. Financial journalists will soon be competing with the masses to find the juiciest contextual stats in the reports, just as their political brethren did during the primary season. Institutional investors’ database transcriptions will no longer be needed if their clients spend the time to build their own parser. That means a loss of revenue for many of the data intermediaries as their business model becomes wholly focused on providing smart analysis, not raw data.
The institutional investors I talked to seemed to welcome that change; XBRL allows them to redirect resources from transcription to analysis. That shift is an advantageous one, according to Suresh Kavan, president of the investment and advisory group at Thomson Reuters. Kavan told me that transcription revenue is so small, “I wouldn’t even call it a pimple; it’s smaller than that. It’s maybe a small, pink dot on the underside of my foot.”
When I asked Kavan whether he felt threatened by opening up financial data to the masses, he was quick to shoot the question down. “I love the idea,” he said, and he added that smart financial companies are preparing for a new financial landscape in which analysis on the Internet was just as integral to making financial decisions as reports from professional analysts.
But Kavan was adamant that XBRL and the added presence of financial tools on the Web aren’t a risk to institutional investors. The folks at Reuters Thomson are paid professionals with years of experience and access to sources of information bloggers could only dream of. He made an analogy to the news industry, where he suggested that the public trusts information created by trained professionals more than the thoughts scribbled by amateurs on the Internet. Logically, he’s right. But practically, the current state of the newspaper and newsmagazine industries shows that institutional investors’ wallets may have reason to be anxious.
The stronger argument is that XBRL is not a fundamental change for highly active traders. Real movement in the markets occurs as companies hint at the kind of profits or losses expected this quarter (as seen in Lehman Brothers’ recent dive). That’s well before quarterly reports are filed, so XBRL won’t provide any added insight until after the markets have made much or all of their moves. The technology is better suited for casual traders looking for more information about the historical performance of a company.
It’s this last point that justifies Kavan’s confidence. High-profile clients aren’t going to turn their backs on institutional investors for the blogosphere anytime soon. But for average independent investors who already rely on the Internet for much of their data, XBRL could provide some fancy new tools to play with. It’s possible that XBRL is a win for nearly everyone. For that, Chris Cox and the SEC should be applauded. Even a lame duck can sometimes lay a golden egg.
