XBRL and the US Proxy System

In the wake of the financial meltdown, all US enomic institutions have me under scrutiny, and the proxy system — through which shareholders vote annually for directors and on rporate ernance matters — is no exception. Described as the first serious attempt to overhaul the proxy system in thirty years, the SEC issued its ncept Release on the US Proxy System (CR) last month.

Not surprisingly, the CR is a long document that raises many issues described in briefings by law firms like Baker Hostetler and Winston & Strawn; among them is the potential of XBRL for bringing greater transparency to the process. Before I look at that facet of the CR, however, a little background on the US proxy system is in order.

Schle 14A of the General Rules and Regulations promulgated under the Securities and Exchange Act of 1934 defines what information is required in proxy statements. As one might imagine, the regulations are mplex; among the items the Schle requires in the annual proxy statement is information about nominated directors; mpensation of board members and executives; and background on the audit mmittee as well as fees paid to auditors. Perhaps the best way of learning what mpanies’ proxy statements ntain is reviewing a few: SEC Watch provides links to new proxy filings. As for mutual funds, these generally are required to disclose annually on Form N-PX how they vote proxies relating to securities in their portfolios.

Returning to the CR, of particular interest to the XBRL mmunity is Data-Tagging Proxy-Related Materials (pp. 96-104). It’s an interesting discussion which incidentally ntains nsiderable history on the SEC’s use of XBRL. Here’s what it says on the potential of interactive data for the proxy system:

A significant amount of the textual data in the proxy statement is well-structured and may be suitable for data tagging. If issuers provided reportable items in interactive data format, shareholders may be able to more easily obtain specific information about issuers, mpare information across different issuers, and observe how issuer-specific information changes over time as the same issuer ntinues to file in an interactive data format. This uld both facilitate more informed voting and investment decisions and assist in automating regulatory filings and business information processing.

The CR demonstrates the SEC’s ntinued (or perhaps renewed) interest in adopting XBRL for executive mpensation data. As footnote 221 on page 100 notes, in 2007 the SEC created an Executive mpensation Reader, which was generally well-received. However, in previous mments to SEC, some parties have suggested that “variations among issuers in executive mpensation practices may not lend themselves to the development of standard tags.” Whatever those variations may be, it’s hard to imagine they represent as great a challenge as preparing the already-mandated XBRL financial statements of US mpanies, which operate in a wide range of industries under a host of acunting policies open to differing interpretations.

There’s a footnote on page 99 which I had some problems with: 

We anticipate that any interactive data format version of the information permitted or required would not replace the traditional format version, at least not initially. In general, interactive data currently is machine-readable only. Without the use of software, interactive data is illegible to the human eye. As a result, we expect that any interactive data would be provided in a separate schle or exhibit. It is possible, however, that at some point in the future technology will evolve in a manner that would permit human-readable text and interactive data to appear in the same document.

Two things ncern me. First, I’d like a more forceful statement that, if the XBRL format is adopted, it eventually would replace the traditional version. I understand the need to initially phase XBRL in. But a dual (and perhaps dueling) set of authoritative documentation begs the question of which should ultimately be relied upon. Relegating interactive data exhibits to send-class status only will lessen mpanies’ mmitment to get XBRL data right, and it increases the possibility of nfusion by shareholders.

Send, I’m a bit baffled by the last sentence: “Human readable and interactive data…in the same document” sounds much like the Inline XBRL technology that exists right now. Given Walter Hamscher’s recent statement (see point 11 in this blog post) that the SEC is looking closely at making greater use of Inline XBRL, I’m a little surprised that this aim wasn’t reflected in a discussion of human- and machine-readable formats.

As for the status of an XBRL proxy taxonomy, the CR states:

Currently, there apparently is no standard set of XBRL definitions, or “taxonomy,” available to enable an issuer to provide proxy statement and voting information or any subset of such information in XBRL format. XBRL US, however, is developing a taxonomy for at least some information a proxy statement requires. See http://xbrl.us/Learn/Pages/Initiatives.aspx (“Broadridge Financial Solutions [the former brokerage services division of ADP – ed.] ntributed a proxy taxonomy to XBRL US in Q4 2008. XBRL US will inrporate the taxonomy into a master digital dictionary of terms.”)

The Request for mment section (pp. 101-104) asks for a wide variety of feedback that mprises both the benefits and sts of adopting XBRL for different parts of the proxy system. Interestingly, at a uple of points the discussion raises the possibility of adopting XBRL for executive mpensation data regardless of whether interactive data is used for any other proxy information.

Numerous mments already have been received on various questions posed by the CR. If my searches have been aurate, however, only one thus far addresses XBRL, a negative assessment at the bottom of page 4 of the mment. mments can be submitted at this page of the SEC ; they are due by October 20, 2010.