Analyst Perspectives

The mandate by the U.S. Securities and Exchange Commission (SEC) that requires its filers to apply eXtensible Business Reporting Language (XBRL) tags to their financial statements has been in effect for several years. (XBRL is a core element of our Office of Finance Research Agenda for 2012.) One of the most important ideas behind this “interactive data” requirement was to make it as simple as possible for investors to be able to consume and analyze corporate financial data filed with the SEC. This intent sets the SEC mandate apart from most other XBRL tagging requirements, which are designed for the needs of regulatory bodies such as the Bank of Japan, the Australian federal and state governments and the U.S. Federal Deposit Insurance Corporation (FDIC). Moreover, I believe the depth and breadth of the SEC’s database and the size of the U.S. equity capital markets make this the most important public-focused use of XBRL in the world. Considerable progress has been made toward the main objective, but considerably more is needed, and the sooner the better.

Over the past decade XBRL has advanced from a concept to a technology that underpins many day-to-day financial communications. In the U.S., we are now past the initial, substantial hurdle of creating a workable taxonomy to define the meaning of XBRL tags. Corporations are beginning to transition from approaching XBRL tagging as a project to making it just another step in the disclosure process. There’s been a tenfold increase in the number of companies doing their tagging in house. As they gain familiarity with what other, similar corporations are doing, companies have been using fewer “extensions” to the taxonomy, which means that there is greater comparability – a vital quality to making the reported data useful for investors.

Yet most of the progress made to date is in capturing data; more progress is needed to create usable tools, especially open source ones that are available to all investors. This has not escaped the notice of the organization XBRL-US, which established an XBRL Challenge to “encourage the development of more tools and build awareness among analysts about the wealth of data available to them.” Participants in the Challenge were asked to contribute an open source analytical applications that would enable investors to use corporate XBRL data. In my opinion, the results are a good start, but they don’t go far enough. Just because software is free does not excuse it from being full-featured and user-friendly. There’s certainly enough sponsor revenue to make it worthwhile for Yahoo and Google (one of the sponsors of the challenge) to offer online analytic tools and certainly enough incentive for online brokers to offer these capabilities as part of their research offerings.

From my perspective, one of the promises of XBRL was enabling individuals to array company data in any format of their choosing and to include a broad array of analytic metrics for the income statement, statement of cash flows and balance sheet. One might want something as straightforward as all flavors of profit margins (for example, operating margin or earnings before interest, taxes, depreciation and amortization – EBITDA) or balance sheet items such as the current ratio or debt-to-equity. Others may want to look for indicators of financial shenanigans such as changes in discount rate assumptions, useful life assumptions for depreciation or a large jump in days sales outstanding. The analytic application supporting this might have some interface that allows users to quickly drag-and-drop different data elements (revenue, tax expense, inventory or dividends paid, for instance). Or it might enable one to automatically populate a spreadsheet with those elements in whatever form and format they wish, drawing data from documents such as a company’s annual (10-K) or quarterly (10-Q) filing. This would represent a real improvement on either manually rekeying data from financial statements or “scraping” the information from electronic documents by performing a series of copy and paste actions. However, we’re not yet to the point where it’s easy enough for an individual to freely do this at a basic level, whether that is an investor or a business analyst in a company that wants to use this data for analysis or performance benchmarking.

I believe the closest we’ve come so far to this is a demonstration application, XBRL to XL, which was a finalist in this year’s XBRL Challenge. I found it offers a good deal of flexibility for working with single or multiple company data, but you have to be a spreadsheet jockey with a reasonable amount of analytic skills to really make use of it. As well, at this writing the data is limited to the annual reports (10-K). Other tools in the Challenge include Calcbench, the winner of the competition, which presents canned reports and allows for comparisons to other companies. It’s probably more useful to a wider audience but I find its capabilities very basic.

To be sure, there are data aggregators such as Edgar Online and Thomson-Reuters that provide readily consumable financial statement information for a fee and can do many of the things I allude to above. As well, SavaNet offers XBRL-enabled publication and analysis software tools designed for institutional investors. It enables “sell-side” research departments (investment banks and brokers) to easily publish their analysts’ company or industry models so they can be integrated with their “buy-side” (those working for, say, investment advisors or mutual funds) portfolio managers and analyst counterparts’ models. Not only do these models incorporate analysts’ forecast earnings models, they also have historical product line and industry data that is not available as tagged items in company reports. (Providing sell-side analysts’ models to clients is an established practice. As early as the 1980s this was done on an individual basis by mailing a floppy disk to a client.)

There are still many holes on the data collection end. For one thing, most companies’ earnings releases are not yet tagged. Just like in the old days, one has to manually enter or scrape the data into some spreadsheet to update a company model or create a table quantifying the differences between actual results and projections. For sell-side analysts, this consumes precious minutes between the initial release and the conference call discussing the results; for buy-side analysts, who cover many more companies, the problem is even worse, especially if the company is not followed by an analyst or just one or two. For another, although the XBRL taxonomy for US-GAAP is well-developed (and in my judgment the most comprehensive in the world), there is still a need for industry-specific measures (such as revenue passenger kilometers for airlines or kilowatt hours produced for utilities), and it would be nice if one could grab company-specific product line data (for instance, revenues of specific drugs for a pharmaceutical company) as a tagged item. This is probably the job for equity analysts and portfolio managers. In the early 1970s, as part of the SEC’s “full disclose” effort, these analysts (often represented by industry-specific “splinter groups”) were heavily involved in creating business segment reporting guidelines. These same analysts (and splinter groups) also could help companies in specific industries maximize commonality in how they report and tag their results to achieve the highest degree of comparability between companies.

So although it’s evident that the SEC’s XBRL mandate has made substantial strides over the past five years, much more needs to be done to make the millions of data elements from more than 9,000 reporting companies more consumable by ordinary investors and business analysts. I’m confident it will be done, and I’m hopeful it will be done relatively soon. The first two major hurdles – developing a workable taxonomy and assembling a substantive database – have been surmounted. With growing investor awareness of how tagged information can benefit them and tangible demonstrations of the many ways this information can be used, I expect increasing investments in solutions designed to unlock the value of this data trove.

Regards,

Robert Kugel CFA – SVP of Research