An Insider’s Perspective: How To Avoid A Yahoo-Like Tabulation Nightmare
This interview first appeared in the Fall edition of Broc Romanek’s new publication, InvestorRelationships.com, which also covered another important “hot topic” in its inaugural issue, with a “Roadmap to Regulation FD under the SEC’s new website guidance”.
Recently, Broc Romanek caught up with long-time independent inspector of elections, Carl Hagberg (who also edits the Shareholder Service Optimizer), to learn more about how the recent vote tabulation snafu at Yahoo’s annual shareholders’ meeting could possibly have been avoided – and how you can protect yourself from facing similar embarrassing mistakes.
As you may recall, Yahoo quickly issued a press release after its annual meeting that announced voting results – and then had to correct them soon afterwards when a large institutional shareholder complained that not all of its votes had been counted correctly. One of the reasons for the problem was that Broadridge’s voting system didn’t have a sufficient number of digits available to record the large number of votes withheld. Carl explains how Broadridge shouldn’t be viewed as the sole culprit – and how you can take steps to avoid a similar situation happening to you.
ROMANEK: Carl, what was your first reaction when you saw the Yahoo news?
HAGBERG: When I opened the newspaper, the first thing that hit me was “this is exactly what I have been warning people about for seven years; I’ve been saying ‘hey, the old days where management got 99.9% of the votes in their favor are fast disappearing - and to the contrary, the level of “no” votes are getting higher, higher, higher with every passing year. You really need to check with extra care before releasing results whenever there’s a close or closely contested matter.”
I’ve felt all along that people weren’t really getting it – that more and more often, the results will be so close - and so closely watched - that you need to exercise more due diligence than ever before - that you really need to have inspectors who know what they’re doing and who actually do inspections.
I’ve also been warning people that they need to prepare the meeting chair for the day that someone jumps up to ask “who inspected these results… and what exactly did they do… and how do we know that they got it right?” And now, what I’d been warning about has become front page news - exactly as I also predicted would happen when the inevitable snafu occurred.
ROMANEK: What did you think about the focus on Broadridge as the cause of the problem?
HAGBERG: I guess it was predictable that all of the emphasis was on the tabulation, since most lay people kind of get that part of the process. But as I’ve been trying to remind people all along, there are two parts to this: one is the tabulation process, but the second - and the far more important part, I think - is the inspection process.
Proxy tabulation is very highly automated these days. Over 90% of the vote typically is tabulated via Broadridge’s online ProxyEdge platform for institutional voters. Most of the remaining votes - coming from individual investors - are tabulated via telephonic and Internet voting systems and via high-speed scanning equipment.
Thus, virtually all the votes are being tabulated untouched by a human hand - and in a way that’s basically unobservable to the human eye. And yes, things can - and do - go wrong mechanically. Or the tabulating systems sometimes get set up incorrectly from the outset.
But in most cases, instances like these are caught right away by implementing routine spot-checking procedures – and Broadridge has very robust ones, by the way – or because the results suddenly veer away from the norm or seem counter-intuitive to a savvy observer.
The Yahoo meeting was a little bit unusual in that Broadridge was tabulating the results in a sub-system that was designed for proxy fights, which the circumstances leading up to this particular meeting started out to be. And that system - I’m sure because no one ever imagined that more than 99,999,999 votes would ever be cast against a director when it was built - could only accommodate eight digits.
But as Broadridge suddenly discovered, they really need nine digits…and maybe ten or eleven to be really prudent in this day and age. It reminded me of the Y2K problem that almost every computer owner in the world suddenly discovered as we prepared to enter the 21st century. I also recalled with a chuckle - being an ex banker - of the way we had to scramble the first time we had to issue a dividend check for more than 99,999,999 dollars.
And, as an Inspector, I still get frustrated that most pocket calculators - and many desk-top models too - can’t carry the 11 digits that are needed to compute percentages for companies with 10 billion shares or more. In any event, this was clearly a “first” – and for Broadridge it was a fairly quick and easy thing to fix once the shortcoming was discovered.
So that brings us to the second – and to what I say is the major part of the vote reporting and certification process - and that’s the inspection. Somebody’s got to say “gee, was all the machinery clicking along in a proper manner? Are the outcomes consistent with what a savvy observer would expect? Is some additional due diligence and double checking required before we certify the results?”
ROMANEK: So if you were the Inspector at the Yahoo meeting, would you have missed the truncation error?
HAGBERG: That’s the very first question I asked myself – and of course, I’d like to say no, I wouldn’t
have. But to be perfectly truthful, you can’t really know for sure. We don’t know exactly what happened behind the scenes. And - very important to note - we all make mistakes; nobody is perfect. But for sure, it’s every Inspector’s nightmare scenario, even at a “routine meeting.”
The fact that we all make mistakes is the main reason I’ve been saying for ten years now that the tabulation should be conducted by one set of people and the inspection should be conducted by another set - so that somebody who is totally independent is always watching your back and doing nothing else. An inspector often needs a second inspector or somebody else to watch their own back as well as the company’s back, now that more meetings are having close or potentially contested results.
ROMANEK: To get a little technical, what exactly does an Inspector review?
HAGBERG: Most people don’t know this but Inspectors are not responsible for inspecting the votes that are cast by street-name holders. The Voting Instruction Forms (otherwise known as VIFs) and other forms and platforms that street-name holders use are not the legal equivalent to proxies, so they’re not technically subject to inspection.
So we have a rather anomalous situation, where the inspector is certifying the final vote - but hasn’t inspected the vast majority of votes that were actually cast.
ROMANEK: With that situation as background, what types of scenarios should raise red flags for Inspectors?
HAGBERG: There are quite a few situations that should lead Inspectors to conduct some extra due diligence before certifying the vote. For example, whenever a director starts obtaining double-digit levels of “against” or “withheld” votes.
For shareholder proposals, anything above 15-20 percentage points on one side or the other of passing or failing is a situation that probably calls for more diligence. And any proposal that comes within 5 percentage points of passing or failing requires a much higher level of due diligence and double checking.
Another major red flag is when a high percentage of the votes are cast in the last 24 hours before the meeting – particularly if there are a lot of last minute revocations and reversals – and most especially if any of the outcomes suddenly change direction.
We served as Inspector for a meeting this year for a Fortune Ten company and there was a shareholder proposal that was failing by a fairly healthy margin until 10 minutes or so before the meeting began, when it suddenly shot ahead by almost one percent, then fell to 1/4th of one percent when we asked the tabulator to take a closer look at the last minute stuff, some of which was sent to them twice. Naturally, before we certified the final numbers we wanted to go back and look over every single vote that could conceivably have made a difference one way or the other.
Last, and most germane to the Yahoo situation I think, is whenever the results are unexpected, or even slightly counter-intuitive to a savvy observer. Interestingly, all the initial press coverage I saw remarked on the surprising “vote of confidence” that the CEO Jerry Yang got in terms of the votes withheld against him compared to the votes withheld from three other directors who had been targets of vote-no campaigns (with only 15% reportedly withheld from Yang vs. the others’ 20%+).
I’d like to think that this would not have passed my own ‘sniff test’ and would have caused me to look a little closer before certifying the results - but second guessing is easy and one can never know for sure.
ROMANEK: What practice pointers do you have for those that employ Inspectors?
HAGBERG: I have six practice pointers to share. The first one is ensure you don’t ever rush your Inspectors to report the final results.
The Yahoo situation is actually a very good argument for my second pointer: our long-recommended practice of only announcing “preliminary results” at the meeting or shortly thereafter. After all, the “system” worked here, when a big voter said ‘whoa, hold on, this doesn’t look reported, there would have been “no news” to report for this meeting.
A third - and related – point: in today’s environment - and particularly if there are close or contested matters on the ballot (or if you’ve been receiving a lot of press coverage about anything) you really need to ask yourself - “Is there a chance we may have missed something? And what would be the consequences if we have to put out a correction later?” Make sure that all the parties involved – the tabulator, the Inspector and company representatives too - are all asking these two questions and watching each other’s backs before the final report is issued. There is no need whatsoever to report the final resultsthe same day.
Fourth - and, as we learned from Yahoo, I hope - there is a major downside to appear to crow about a meeting result - or to otherwise “spin” the results – even if you’re 100% sure they are correct, I’d say. It’s an open invitation to second-guessing and “spinning back”…and, as the saying goes: ‘pride goeth before a fall.’
Fifth, you need to be sure that the Inspector actually inspects – and is prepared to due some extra doublechecking if results are close or counter-intuitive…and has a good “sniffer” to begin with. I’ll share a few trade secrets: In close situations we always ask Broadridge – and any other tabulators too – to do a last-minute “sweep” of their workplaces – especially their fax machines. In close situations, we also conduct our own review of the street-name votes - even though we are not officially ‘inspecting’ the street-vote – to see if all the voters we would expect to vote have indeed voted, and to see if the larger voters are voting the way we would expect them to vote.
The sixth pointer, which is actually the first commandment of tabulation and inspection - and I’ve got to believe that everybody failed to do this at the Yahoo meeting: you must always prove your numbers to the quorum. What exactly does this mean? If there is a majority vote standard, the Inspector needs to add all the “for” “against” and “abstain” votes together and see if they equal the reported quorum (or in other words, the total number of shares that are represented at the meeting). If there is a plurality election standard, the Inspector needs to add the “for” and the “withheld” votes together and ensure they “prove to the quorum.” If someone did this at the Yahoo meeting, I’ve got to believe that the “missing” 200,000,000 withheld votes would have become immediately apparent.
ROMANEK: Do you want to wrap up by talking about overvoting? Is this an issue for tabulators and Inspectors?
HAGBERG: Another one of our favorite subjects. With close votes or high profile matters, you need to really raise the level of due diligence on overvotes. You need to record every time there is an apparent overvote – and exactly what was done to resolve it. Although it does not appear to have been an issue at this year’s Yahoo meeting, it’s an issue for many meetings because improperly handled overvotes can distort the intended outcomes significantly.
ROMANEK: Any last words?
HAGBERG: One final word for the institutional investors out there. They wait and wait until the very last second to vote. I don’t think this was necessarily a big issue at Yahoo, although it probably was a factor to some degree. If you know how you’re going to cast your vote, cast it when you make up your mind. Don’t wait for the evening before the meeting – or even worse, the morning of the meeting. It increases the likelihood of an innocent mistake, as tabulators and Inspectors try to rush around at the last minute. And it greatly increases the chances that your vote won’t get recorded at all.
ROMANEK: As always Carl, this was great. Thanks for sharing your wealth of experience with us