Cumulative Voting

Our Best Practices view:

Cumulative voting should be adopted on case-by-case basis.  Generally, it is useful in protecting minority shareholders in companies with a clear dominant shareholder block.

Content:

Introduction

Description of Cumulative Voting

Calculation of Shares Needed

Implication

Conclusion

Introduction

In companies with a controlling shareholder, cumulative voting can be very important in establishing minority rights and is often a requirement in the IPO process.  For that reason cumulative voting is widely practiced in Europe, where many companies have a dominant or controlling shareholder (either state or private).  It was also popular in the United States when proverbial “rubber barons” controlled the railroads, but also needed to raise capital by selling minority positions.  Cumulative voting is also used in national and local elections where rights of minority population are seen as critical.  For example, in some Southern counties in the 1970’s, cumulative voting was implemented to increase the representation of African Americans on the county councils.

Description of Cumulative Voting

Under Cumulative Voting, a shareholder gets as many votes as there are directors and can concentrate them on one candidate.  For example, if there are 16 directors, each share would get 16 director votes, and the maximum required percentage of voting shares to definitely elect one director comes out to only 5.9%+ 1 share.  Here is the math if you are interested:

Adding-up the Votes

Suppose there are 1,000 shares.

Multiply 5.9% by 1,600 votes and you get 944 votes. 

All cumulative votes available represent 16 director posts *1000 equal to 16,000 votes. 

16,000-944 votes are the remaining votes or 15,056 votes. 

If 15,056 votes where to be perfectly distributed among the other directors, each would get 15,056/16=941 votes, so the minority candidate needs just one more vote to top the rest.  Note that if one of the majority candidates gets more than her ideal vote of 941 (93.75% of votes to the number of shares) that means that at least one candidate gets less than 941, allowing the minority candidate to pass with even fewer votes.

A general equation for the number of shares needed to definitely elect a director is:

X=SN/(D+1)+1,

where:

X is the number of shares needed. 

S = number of shares actually voting at the meeting. 

N = quantity of directors you want to elect (so if a caucus wants to elect 2 directors, plug in 2). 

D = total quantity of director posts up for election.

This assumes that all other votes are perfectly stacked on the other 16 candidates.  If the remaining votes are not perfectly distributed among the other 16 candidates or if there is a 17^th candidates who attracts votes, the required percentage of voting shares to elect one candidate will be less than 5.9%. 

Implications

Take the {Cumulative Voting Proposal at GE} for example.  GE does not have a 50% shareholder, and Lord willing it never will.  The largest GE shareholder as of March 31 was Barclays Global Investors UK Holding Ltd with 4% of all shares outstanding.  In the real World, that will be more than enough to guarantee a Barclays only candidate because:

1. Not all outstanding shares will vote at the meeting.  (It is still a practice with many brokers to award proxies to management without beneficiary’s express consent.  However, NYSE is forbidding this practice for its member, and many brokers are already using proportional voting instead of just giving away votes to the management).
2. There will likely be more than sixteen candidates in a contested election,
3. At least some percentage of non-Barklay shareholders will vote for Barclays candidate, and 
4. Since there is no majority shareholder, and there are stringent restrictions under US SEC Section 14, governing shareholder communication, it will be impossible to caucus the thousands of the remaining shareholders to stack votes exactly evenly among the incumbents. 

This means that all shareholder-appointed directors will be in the hands of 3-4 institutional investors and put a lot of power in the hands of the large shareholders.  Note also that in order to elect more minority directors, institutional shareholders will have to caucus to allocate votes strategically among the directors they want to elect.  It is a certainty that small shareholders will not be invited or have any say in these caucuses.

One could argue that small shareholders would have the power to select which of the incumbent directors will be sacrificed by allocating their votes from the directors they conceive to be weaker to the directors they conceive to be stronger.  However, the opposite can be true.  If the management wants to support a weak director who would nonetheless help them entrench, they can use the vast proxies awarded to them to concentrate votes on the weak directors.  

Is all this good on balance?  This depends on the specific company and your perspective.

On one hand, in case of GE, Barclays can throw a lot of resources behind their director to counterbalance the management power.  On another, there is no guarantee that these institutional investors will not have governance problems of their own.  It is also bad when the board is polarized to the point of assigning specific directors to specific shareholders.

There is also no guarantee that at least some caucus will not be among special interest groups and only have a special interest in mind, which may be an important social agenda, but not oriented toward shareholder value.

Conclusion

Cumulative voting should be evaluated individually for each corporation.  In case of GE (our example), the virtual inability of most US shareholders to caucus will bring unpredictable and possibly strange results out of cumulative voting.  On balance, cumulative voting will likely add confusion, but neither add nor take away from the corporate governance of most large and diversely-held companies.  However, in cases where there is a dominant shareholder, cumulative voting can be an important tool to balance minority interests.