Tesla and the Case of the Missing Votes

It’s no secret that Tesla has weathered some … ahem … criticism of its governance structure, and in particular, the (lack of) board supervision over Elon Musk.  After Musk’s antics landed him in trouble with the SEC, the company proposed two charter amendments that would make the board more responsive to shareholders: first, to amend the charter so that future charter amendments will require only majority rather than 2/3 vote of the outstanding stock, and second, to amend the charter to reduce director terms from 3 years to 2. 

Management sponsored proposals – especially those that hand more power to shareholders – are usually kind of a done deal.  But this week, Tesla announced that even though the vast majority of voting shareholders favored the proposals, the proposals had failed to pass.  Now, to be sure, the proposals needed a 2/3 vote of the outstanding stock – that was, after all, one of the things sought to be amended! – but it was still a bit of a surprise to see that the threshold hadn’t been met, given the amendments’ popularity.

I know that I wasn’t the only one who suspected some kind of Musk-related shenanigans, like perhaps Musk got cold feet about relinquishing power at the last minute, and withheld votes on his approximately 21% stake.  The Wall Street Journal article reporting on the vote pointedly highlighted Musk’s ability to block changes he disliked.  Plus, when it was originally posted online, the article reported that a spokeswoman for Tesla did not answer a question about how Musk voted, though later the article was revised to say that the company claimed Musk did not withhold his votes.

But still, I was suspicious about how exactly the proposals managed to fail, so I tried to do some back of the envelope calculations once Tesla filed its 8-K disclosing the tally.

And here’s what I figure.  First, based on proxy disclosures and the 8-K, about 80% of the total voting shares made some kind of appearance at the meeting (in many cases, via broker non-votes).  That compares to about 83% last year, when there were about 4 million fewer shares eligible to vote.

Second, broker non-votes this year exceeded last year’s tally by 11 million – but even if all of those shares had voted in favor of the proposals, it wouldn’t have been enough to change the outcome given the 20% of shares that never made an appearance.

Why so many non-votes/nonappearances as compared to last year?  Well, I’m not sure if this means anything, but it also seems there was a lot higher volume of trading after the record date this year as compared to last year.  Now, Tesla sold 3 million new shares in early May, so I only looked at the first two weeks after the record date of 2018 as compared to 2019, but it seems there was double the amount of trading in 2019.  So, if I’m thinking about this correctly – and someone please feel free to weigh in if I’m getting this wrong, and disclaimer: I did not actually do a statistical analysis so I might be overstating the differences, but – it seems (1) the disparity in trading is not surprising given Tesla’s scandals this year and (2) the increased volume could potentially mean that otherwise-eligible voters sold their shares after the record date; so they didn’t bother to vote, and the new buyers weren’t able to do so.

So it’s definitely possible that all of this is just what it looks like: Tesla’s charter requires a high threshold to amend and the vagaries of public company voting made that threshold difficult to meet, even if most voting shareholders would prefer the amendment.  But what we we don’t have is rock solid certainty that there wasn’t some kind of last-minute management change-of-heart.

And that takes me back to my post last year about the need for disclosure of how insiders vote (holders of high vote shares, high ranking officers, directors).  We don’t have that now;  absent actual class voting, the votes are just disclosed as an undifferentiated block.  Which is what sent me on my bout of amateur sleuthing.

Now, not every company is Tesla and most of the time you can probably look at the vote total and, using other disclosures, figure out how insiders voted relative to other shareholders.  But my point is, I shouldn’t have to open my creaky copy of Excel and consult multiple SEC filings to suss this information out.  A clear statement, and separate totals, matter. They matter in terms of how the public understands the vote – reporting like this is downright misleading – and I think they probably matter in terms of how (some) shareholders understand their votes.  Voting isn’t like efficient markets; whether or not information is “priced in,” actual votes have to be cast with knowledge in mind.  Voters either understand the dynamics, or they don’t, and not everyone is going to do the math.  Shareholders should be able to clearly understand the impact of their votes before and after the ballots are cast.  Plus, I believe for certain kinds of close votes, insiders will be sensitive to how their ballots will be reported publicly, and that will affect their behavior.

Bottom line: Insider votes (and votes cast by high vote shares) should be broken out and reported separately.