Saying goodbye to broker votes: the immediate impact on director elections may not be meaningful, but potential repercussions loom.

• Author: Walkling, Ralph A.
• Position: EVIDENCE AND PERSPECTIVES

HISTORICALLY, New York Stock Exchange Rule 452 has allowed brokers to vote shares held in street name in uncontested director elections unless shareholders give specific voting directions. On July 1 of this year the SEC voted 3-2 to disallow this practice. The new rule takes place on Jan. 1, 2010.

The change to Rule 452 has serious implications, and it is important to understand and anticipate ways it could affect a particular company. Unfortunately, there is little actual evidence on the subject, and some aspects appear to be misunderstood.

In an article forthcoming in The Journal of Finance, two Drexel University coauthors (Jay Cai, Jacqueline Garner) and I study broker votes in 917 director elections. For small-cap, mid-cap, and large-cap companies, the broker vote averages around 13%. However, because of the way votes are often counted, the actual impact on director elections may be much less. For example, when ISS (now RiskMetrics) calculates the percentage of votes a director receives, broker votes appear in both the numerator and denominator. This process translates a 13% effect into a change of 2.5%.

A simplified example illustrates this measurement. Suppose 90 shares are cast 'for' a particular director, and 10 votes are 'withheld.' Given the way in which votes are often calibrated, the percentage a director receives is equal to the percentage 'for' the director divided by votes cast (the sum of the 'for' and 'withheld' votes). Consequently, the director receives 90% of the votes cast.

Now, suppose broker votes amount to 13% of total votes. After eliminating this quantity from both the numerator and denominator, the new percentage arrived at for the director is 88.5%--calculated as 77/87 x 100.

Thus, the typical change in impact on the election outcome is trivial, and it will not change dramatically even if broker votes are substantially greater. However, company officers should understand the specific procedures for counting votes at their own firms. For example, using total shares outstanding in the denominator (rather than shares voted) produces quite different results.

To put broker votes in director elections in perspective, I remind the reader of some statistics from one of my earlier columns. The typical director receives 94% 'for' votes. Many factors are significantly related to votes for directors, including operating performance and governance. However, only two factors swing votes by more than a few percent: director...