In a previous article, I highlighted some uncomfortable truths about proxy voting and how this important shareholder right (and fiduciary duty) has been under-prioritized and under-invested by asset managers for decades. It was prompted by another tirade in the press about proxy advisors, who are a symptom, but not the cause, of this malaise.

Now we read that my alma mater JPMorgan is severing all ties with proxy advisors and instead developing its own AI solution. It’s no secret that Jamie Dimon hates proxy advisors, ever since they had the temerity to suggest shareholders vote against his comp and split the roles of Chair and CEO. But it will be interesting to see how voting patterns at JPMAM change, if at all, as a result of the move.

I talked previously about the endless client questionnaires and due diligence reviews I sat through over the years, where the only question clients asked was: “Did you vote all the proxies?” Usually followed by: “How many times did you vote against management?” Whilst this latter question can be a useful (albeit very basic) barometer of how ‘assertive’ your manager is being with portfolio companies, it’s not actually telling you very much (it’s also very easy to massage, but that’s another story). What follows are seven questions you should really be asking - the questions your asset manager does not want to talk about. I know I didn’t when I was one.

1. How many people in your stewardship team? UK managers and Stewardship Code signatories should be disclosing this in their Stewardship Reports, but definitions can be, shall we say, flexible. Ask how many FTEs are actually involved in proxy voting and stewardship full-time. And don’t be including wider ESG teams or, as I have seen, the entire Equity Research department, just because they look at the odd proxy each year. I have even seen client-facing teams included in this number, because they talk to clients about proxy once a year! When you have that FTE number, divide it by the number of proxies they voted that year, and that will give you a reasonable idea of how much resource is being devoted to proxy. A four person team voting upwards of 20,000 proxies each year tells you something.

2. How much of your proxy process is automated? Either by AI or use of so-called SSIs (Sytematised Standing Instructions). The answer (usually) in reality is: almost all of it. Most managers use a proxy advisor (yes, one of them), codifying and automating what’s called a Custom Policy (basically the advisor’s interpretation of your manager’s public voting policy). This allows voting decisions to be applied automatically, the proxy voting conveyor belt keeps moving and everything gets voted. Depending on your team size, you can then pick a subset of proxies out of the pile for the proverbial hard stare and/or additional scrutiny - or even engagement - depending on things like the size of holding, its profile in the media, or the amount of yelling that’s going on. This is not necessarily a bad way of doing it, by the way, given that 90% of voting is routine and/or uncontentious. But proxy voting is an asset, and managers have a duty to use that asset in the best interests of their clients. And, to my mind, that means real people (and ideally stewardship or investment professionals) looking at the most significant or high-profile votes on a case-by-case basis. Ask for the number. They have it. If they don’t, they should have.

3. (If you have an active manager) What proportion of votes are seen by an investment professional? We hear constantly from CEOs (including my former boss), that relying on proxy advisors is bad (https://www.responsible-investor.com/jd-prox/). But, as I wrote in my previous article: proxy advisors are a Frankenstein of our own making. Everybody wanted a cheap, scalable, solution for a task that nobody really cared about, because nobody ever asked. That’s all different now. So, if you are paying for expensive portfolio managers’ big bonuses, ask them what percentage of votes they actually get consulted on, or make the final voting decision? Better still, if you are an asset owner who gets to do a portfolio review each year, ask your PM to talk you through a proxy vote they got involved in; how they thought about it, where they landed and why. If they are unwilling or unable, that tells you a lot.

4. How many times did you change your vote as a result of engagement? Similar to the previous question, your manager should be tracking this number. Managers will declare in their Stewardship Reports that they have engaged with hundreds, and sometimes thousands, of companies each year on issues including ES and G issues. But how many of those led to a change in voting decision at an AGM or EGM, as a result of the information gleaned? You may be surprised. Some US managers are so terrified of being labelled Activist (which has particularly burdensome legal and regulatory reporting implications that they loathe – ask about Rule 13G), that they are reading out disclaimers at the start of each engagement meeting, basically saying “please don’t think we’re trying to influence you and please don’t listen to anything we say”. This is a whole other story probably worthy of its own article. While we’re on the subject of engagement, ask your manager how they track engagement activity, what systems they use and how they are integrated with the systems used by investment professionals. You’ll be amazed how many boil down to spreadsheets on a server nobody outside of the stewardship team even knows exist, let alone uses. Ask for a demo. If you fancy a laugh, ask a portfolio manager or an equity research analyst to do the presentation.

5. (If you have a passive/index manager) How do you decide how to vote issues with a dollar impact on the portfolio, such as M&A or capital-raising? Or in Activist situations? While the overwhelming majority of voting relates to boring and routine proposals at companies where large voting majorities are guaranteed, a significant subset can have dollar implications on the fund. If an index manager, say, doesn’t have a PM to ask, how do they decide what’s in your best financial interest? Sell-side brokers? The proxy advisor? What the FT said that morning? Again, you may be surprised.

6. How do you deal with conflicts of interest? Every asset manager owns companies in their portfolios where there is the potential for a conflict of interest. Maybe the company is also a client, either managing their corporate pension fund, or of their investment banking division? Maybe the company concerned has a significant business relationship with them, for example providing banking or listed broking services, or a fund distributor? Or maybe a senior executive at the asset manager, or their close relative, also sits on their board, or an NED has a side hustle sitting on the board of a mutual fund or investment trust of theirs? How does your manager identify and track these? And how do they ensure a company where there is a relationship doesn’t get an easier ride when it comes to voting? Heads of Legal hate telling their fund board Chairs that they are voting against them in their day jobs and will go to incredible lengths to try to wriggle out of it. Who makes the final decision in these situations? Is there a Conflicts Committee, for example? Or do they take themselves out of the conflict by outsourcing the decision to a third party, such as…a proxy advisor. How many conflicts did they actually identify in a given year? And how did they deal with them?

7. Can I direct my own voting? Much has been made in the press lately about so-called passthrough voting, or the ability to allow you - the end client - to direct voting for shares in ‘your’ slice of the fund or portfolio. This ability has always existed, it’s just that proxy teams hate doing it because it means stopping the automated voting conveyor belt and putting in a manual split voting instruction to custodians. Most passthrough voting offerings today, when you strip away the layers, just boil down to half a dozen flavours of custom policy from the likes of ISS or Glass Lewis. But that is going to change soon, as AI learns to apply voting decisions in more intuitive ways. All this comes at a cost, of course, and may create its own set of problems. But it’s your money at the end of the day.

These are some of the questions about proxy that your asset manager doesn’t want you to ask. I don’t want this to sound like a tirade against my former colleagues. Far from it. There are many managers out there who take these issues incredibly seriously. I can think of one major US manager where every single proxy goes to the investment teams – laudable, if quaintly inefficient. Before we conclude, I would ask you all to please spare a thought for overworked proxy teams everywhere. Most voting happens in a compressed AGM ‘season’ in Q2, when dozens of votes are happening every day. Proxy teams work long hours to achieve an often-insurmountable task, given how they are never enough of them to do the job properly, and they are never paid enough, because proxy has always been seen as an overhead, not a revenue generator. My former colleagues represent extreme value for money, even if some asset management executives don’t see it that way.

Which brings us back to where we started. I worry that, beneath it all, the real reason for the switch to AI, as with so many things in life, is cost. The same exact reason that managers flocked to proxy advisors in the first place all those years ago. And we’re back to the age-old issue of having to do proxy because regulators tell you to, but nobody really wanting to pay to do it well.

As with so many things, AI is going to revolutionise proxy voting at some point. But, as Jane Sadowsky states in her excellent article, switching to AI doesn’t solve any of the inherent problems we experience today, it just pushes them further upstream to the design stage of the AI algorithm. And at least today you can pick up the phone and call a stewardship team or a proxy advisor to complain. Good luck doing that with the HAL9000. Call me old fashioned, but I’d rather have a human looking at my proxies than an algorithm.

###

Original article can be found here: