Vish Hindocha: Hello, and welcome to another episode of the All Angles podcast. This is our season finale, our closer, where we look over the top moments of Season Three of All Angles. And in order to do that, we've done something a little bit different this time. So we've asked the producers of the show and the listeners of the show to highlight their top moments of the show, and we're going to play those. And my colleague George and I are going to react to them. George Beasley, as known to All Angles listeners. He's a member of the sustainability strategy team here at MFS. George, welcome back to the All Angles podcast.

George Beesley: Thanks for having me, Vish. Looking forward to this one.

Vish Hindocha: Yeah, that's great. So we've had a good season where we've talked about things ranging from AI to systems thinking to natural capital to all sorts of topics in between. So it'll be really fun to understand: What did the listeners actually really engage with, what were the top moments and top clips from their perspective, and what do we need to think about them going forward? So why don't we roll the first section that was selected by our audience.

Any thoughts there in terms of what the team has been focused on in terms of risk and return or engagement, or how you think about questioning corporate management around some of these themes?

Genevieve Gilroy: Yeah, absolutely. I think you hit the nail on the head in terms of this space is moving at breakneck speeds, and if we were to plan out what is the ultimate market size for all of these chips that are supporting AI, if I look out five years and I think through what are all of the assumptions that go into what that would look like? It is incredible just thinking through from a units perspective, from a price perspective, from a market share perspective, from a gross margin perspective, the range of outcomes across so many companies that I cover, particularly for Nvidia, is wider than it ever has been in history.

So really what we're trying to do as a team is talk to as many companies as possible, talk to as many experts and talk to each other and really just think through range of outcomes. How wide could that really be across all of these businesses and respect within valuation that just given the range of outcomes across every variable at this incredibly involving period of time is so wide, respecting that within valuation, and respecting the difference in range of outcomes across companies. So thinking through, again, there are many businesses that are benefiting within semiconductors from AI, one example is Nvidia, but there's many, many questions around thinking through.

Right now, all of this GPU spend going to Nvidia has been driven by the training of large language models. Eventually, we're going to move to a place where the querying of the large language models actually typing in your question to ChatGPT and having it send back an answer. All of that is what's going to drive semiconductor spend. So we don't know at this stage just how large that will be relative to how many GPUs have been needed to train all these huge models. And again, competition is entering. So AMD, Intel, even companies like Microsoft OpenAI are designing their own chips. So the confidence that we can have long term in market share, the range of outcomes is just very wide. And that has implications for prices and for gross margins. So as I think about a company like Nvidia, again, the range of outcomes is so wide.

Vish Hindocha: So that was Genevieve Gilroy, who leads our sector team looking after semiconductors as well as consumer staples as well. Thinking back to that conversation, Genevieve talked about range of outcomes multiple times, and just thinking that through, we continue to be in a bit of a hype phase around AI, large language models, ChatGPT, version 4.0, 1.0, and all of the plethora of new tools coming our way, which all sound and are very exciting, but we're still exploring what the boundaries and edges are of those.

I think it's interesting how she talks about we have to respect that there's a range of outcomes. There's still up and downside. We don't quite know yet what new industries will be born or where the demand will flow to. And actually as maybe an update, there is some work that the sector teams have been doing around that and the adjacencies on energy demand, for example, coming from semiconductors. So I thought that was a really fascinating conversation. There was so much talk at that time about Nvidia, about their outsized earnings, about their continued momentum and growth, their importance in indices and benchmarks for our clients as well. George, what was your thoughts on the AI conversation?

George Beesley: Yeah. A lot of different directions that we could take it in. I think it sparked a lot of thinking, but I mean it's interesting that that was the first clip we were standing outside Bloomberg TV was on and they were rebasing the performance of Nvidia from 2016 to now. And it is astronomical, and it took me back to thinking about the 100 years project, lessons learned from 100 years of investing. So MFS turned 100 years this year. We have done some work on what have been some of the big fundamental lessons that we've learned. And one of them was fundamentals are your friend. We'll stick to the fundamentals.

So we charted out a number of different bubbles that have happened in the past and it's so easy to get caught up in the market exuberance. And using fundamentals helps us bring it back to, well, let's just start from first principles and do the hard work around what do we actually think is going on in this market? All of the things that Genevieve mentioned, and that helps you to stop getting caught up in the market hype and helps you to think back to what is the value that we really place on whatever issue it is that we're looking at? And it's really difficult to not get caught up in that market hype now with performance just absolutely storming for a number these funds, and a number of very concentrated players in the asset manager industry.

So I think it was a good reminder that a lot of these companies need perfect performance in order to be able to justify those multiples. And the range of outcomes is very wide. So we need to be disciplined and bring it back to fundamentals and think, yes, they've had an incredible bull run, but can we take the temperature down a little bit and think about what do we really think is going to happen in the future? And there are changes coming from, for example, training large language models to querying.

Another thing that I was reading just a couple of days ago actually was on Neo Clouds, which I hadn't really heard of before, which is there's been a lot of interest from Wall Street on this type of company that I wasn't even familiar with, but they are cloud-computing solutions for companies, and they own a huge amount of Nvidia GPUs. So they already bought all of them before, and that seems to be like another hot area and how can we think about where is the industry going now as we move on from just the training level, just adding more parameters to how is this going to develop into the future, and where can we maybe find some more stable cash flows and narrow that range of outcomes because it's not just going to be Nvidia and others that are going to win from that trend?

Vish Hindocha: Yeah, absolutely. So I think that's right. And as the semiconductor AI technology world continues to develop, you'll see it ripple out across different sectors, different types of businesses, different industries. I think it's really interesting. We were at the Global Investment Roundtable very recently, so we're recording this podcast in November 2024, and in October we had the Global Investment Roundtable. So we get all the investors from all over the world in one location in New Hampshire not far from Boston. Just walking around and talking, it wasn't even a main stage session but talking to some of the energy analysts and technology analysts.

Following on from the conversations that we'd been having with Genevieve and the work that that team had been doing, the energy team and the technology sector teams had come together to really answer the question of, what does the growth in the semiconductor industry and the data farms that we need in order to service AI large language models, and even some of the big social media companies, where does the energy come from and what does that doing to the power grid? What is that doing to energy supply? What does that mean for different things? And actually it's a fascinating piece of research that they've done. We're hopeful to bring some of that to life for our clients in the early new year because it's still a very live stream of work for our research analysts.

So this is how you need to start looking at the intersections of where some of these themes can go. It's really deep, and as Genevieve described multiple times, the team is working incredibly hard talking to all players in those specific industries. But then we also have to think, how do we think about this as an investment platform? How do we think about this across the different sectors and the implications it could have because often that might be where the magic is in terms of where there's a mispriced risk or opportunity. As you said, that range of outcomes is much narrower, but they can still benefit from the tailwind of a large theme.

George Beesley: Yeah. And then from there with our sustainability hats on, it's really interesting to see how technology companies made big commitments around net zero. And Microsoft said that they were going to be carbon negative by 2030 or something?

Vish Hindocha: Over the life of their organization.

George Beesley: Yeah. Over the life of all of the emissions that they've ever been responsible for. And that was very doable before AI. But now we have just huge amounts of energy that's required for the whole industry of AI, whether it's querying or training. And that's going to cause a big shift now. And we've seen big tech companies now look to small modular reactors or other types of nuclear energy. And I just think five years ago, you never would've thought that that was going to be an issue, which is probably why they were so happy to commit to those goals. But I think Meta is even opening some small modular reactor at Three Mile Island, which is the site of the biggest nuclear disaster in the US. And it just shows how dynamic industries can be, and you just don't have a good insight into what the future will be.

And I think that ties back to this range of outcomes thing. Maybe five years ago, the range of outcomes on emissions for technology companies was pretty low. So not thinking about valuations, but instead how confident can we be that we can commit to net zero five years ago they were probably pretty confident, but then you saw this complete step change with AI and all of a sudden they're huge emitters and we're seeing their emissions go up each year, and that's going to be a major development for those companies and they're going to have to really think about how they're going to battle with that.

Vish Hindocha: Again, as you said, fascinating how this will play out particularly in the next 5, 10 years. And maybe we just as a thought for our audience, many of whom are asset owners, many of whom are thinking about this, that are suffering from thinking about the risk loaded up in highly concentrated indices, that themselves have made commitments to net zero. And now as you said, a large portion of your passive assets and possibly even your active assets are likely to be contributing quite a lot of emissions that they weren't maybe before. So how you reconcile those two things is going to be really, really interesting. And how you make sure that you stay on path with alignment to some of your goals, commitments, your fiduciary responsibility is likely to come to the fore even more. And we see that every day in our role.

So much more to come on semiconductors, range of outcomes, what that means across the industry. Why don't we move to the second moment that the team picked out?

What's some of your thoughts on how in the intersection of investment and human rights needs to work together? Where are we in that journey?

Rob: People when they think about human rights, they often think about that in, let's say, an emerging markets context. And for me, that's one thing I've always really tried to avoid because there are human rights issues across developed, emerging, all markets really today, and they differ in severity. So I'm not going to try to pretend that increasing a minimum wage in the United States has the same impact as improving access to water and sanitation in a frontier market. So I want to be clear that that's not what I'm saying, but I do believe that we need to think about human rights a little bit more broadly, and we need to bring in conversations on how developed market employees are being treated. And I think if we start to do that, then you really start to see the breadth of what the human rights discussion can be, and you have a lot more angles and areas on which you can really start to get into it.

So just this morning, we had a call with Tesco regarding the way that they manage their employee base, some of the legal issues that they have going on right now, some of the data that they don't disclose, that we feel like we need to really truly understand, like the firm's culture and that experience as a frontline employee. We've had a lot of those kinds of conversations in the US and in developed markets because they are very, very relevant to the way that we need to model the cost base for these companies looking forward. Because I think one of the key things that I want to make sure is clear is that none of us on the sustainability team are ever doing an engagement on our own. The analyst is always there, usually leading it, often a portfolio manager as well in the discussion, maybe multiple. So that's an important distinction that I think again is unique for MFS.

Vish Hindocha: So a lot there. I mean just to start off with maybe a couple of flags to plant, it's interesting how in a three-minute clip you can have so many discussion points, isn't it? So the nuance of, and maybe some of the bias that we sometimes come into these topics with, whether it's... So he talked about when you think about human rights, your mind immediately goes to emerging market issues he was talking about or slavery issues. But actually human rights covers a wide gamut things. And one of the things I always appreciate about Rob, is how he thinks in very nuanced terms. So again, making it clear that we're not arguing that the severity or the importance or the tragedy of some of those human rights issues is exactly the same in develop versus emerging markets, but just recognizing that they are still issues and there are things that we need to get after.

There were some things later in that episode that Rob really drove home for me, which was even though some of these things are softer and a little squishier, and sometimes again, intuitively we might think harder to measure, as investors, as fiduciaries, as stewards of other people's capital, wherever we sit in the capital market, actually the data behind it's incredibly important. So what he's talking about with Tesco and what he went on to talk about with that Tesco example, for example, was actually we need you to disclose because actually we need to have some accountability measures. And we know that the measures aren't perfect, and that's why people will still rely on skilled active investors to exercise judgment where it is qualitative. But we do need data as a bedrock in order to start moving forward. And I think that balance is really important as we think about some of the other frontiers of sustainability that haven't maybe had as much time and attention and maturity as, say for example, climate change or net zero. What are your thoughts?

George Beesley: Yeah. Rob made the specific distinction. So he said more disclosures are essential, but there is a real difference between setting a policy internally and disclosing hard data. And disclosing a policy, it's somewhat useful, but it doesn't really tell you anything. It's not very rich. It doesn't really tell you about how that's being implemented necessarily at the company unless there's some kind of flag. But having the data, he specifically mentioned a couple of data points, I think LTIR and TRIR, so lost time instance rate and total recorded instance rate. And when you can monitor that over time, it gives you that additional lens to be able to hold issuers accountable or help them work through things and say, "Well, actually, if we look at the numbers, we can see over the last three years that this tends to be trending up. So you've set this policy, but actually is that really having the intended consequence?" So having that data is incredibly useful.

I think the other bit that really stuck out to me that Rob was talking about was the differentiated approach at MFS where we have this sustainability team, but it's embedded within the investment team. But the role of the sustainability team is not to be the leaders of all of the integration engagement and proxy voting work. It's instead a collaborative effort. And the analysts are often the ones leading the engagements, which isn't normally or isn't often the case within the asset management industry, but they have so much ability. They know these issuers better than anyone else, and they can apply that nuance. So having the dual partnership of a sustainability specialist who can really go very deep on a few different areas, along with the analysts who knows the issuers inside and out, I think is a very powerful combination that isn't that common to find in asset management now.

Vish Hindocha: Yeah, I agree. And it's the manifesto. If you really believe in the ownership mindset, i.e., all of our investors are long-term owners of the businesses that we want to take a stake in or make a loan to, depending on equity or fixed income, then you have to be leading the charge on those discussions. You can't outsource the things that you think are fundamentally important to your thesis. Maybe just as a flag for future work that the team is doing, so most of the themes that we talk about are always continuous. We're really excited about our partnership with, and there's many streams to ongoing work, but we're really excited about our partnership with the Oxford University in the UK. We've been working with them for a little while on a few issues, but one that we are kicking off now is really around how do we measure culture and some of the S-factors, social factors in work, what are some of the more material ways of thinking about that and working with some of the lead researchers there?

So perhaps a future episode for season four of All Angles could be bringing some of that research to life that we are funding and that we are working very, very closely with Oxford, where they're coming into MFS to work with some of our lead investors, some of our lead engagements to understand how do we do that? How could we do that better? What could we learn collectively around how to analyze, assess what is in the toolkit of the investors when they're thinking about some of those really important societal and stakeholder factors?

George Beesley: Yeah. Another piece for those who are interested is the work that we partly collaborated with the Thinking Ahead Institute on right-sizing stewardship. So it could be a difficult question to answer, how much resource should we be putting into stewardship? And the TAI did some great work on that. So just to plant a flag in that as well, for those who want to read a little bit more and get some guidance on just how much resource best practice would suggest that you should be putting into those engagement and proxy voting efforts.

Vish Hindocha: That's great. Maybe we can put a link in the show notes. Okay. Let's roll. The third moment that mattered.

Roberto Rigobon: The way I entered the ESG world was through the human trafficking. I said, "Well, somebody had to be measuring this." So I discovered by accident that yes, MSCI, Sustainalytics at the time, Vigil, they were asking questions, and they were asking the question about human trafficking in the supply chain. I said, "That's so interesting!" So I started talking to all of them and I said, "Can you tell me how you do this?" And then I realized that they were all measuring different aspects of human trafficking. And I said, "But wait a second. You should be sharing the data. You're all measuring different things, like, how you can call this human trafficking.” So I am not going to be very precise about human trafficking, but let me give you an example that I did uncover on water, which is the same spirit.

So they were calling this water management, and I said, "Okay. So what are you measuring on water management?" I said, "Well, we are measuring whether or not you follow the law." I said, "That's an important aspect." Then the second one is, "Well, I am just measuring how much water you are wasting." I said, "Well, that might be related or not to the job. It's just completely different." And the other one was actually measuring about pollution, and all three of these items had the same name, water management. He said, "You have to be crazy. These are truly not the same thing."

Vish Hindocha: The irrepressible Roberto Rigobon, professor at MIT and lead researcher of the aggregate confusion project, which is something that MFS alongside a few others, have been founding members of to fund some research into essentially how do we improve data collection, aggregation, measurement and thinking around some of these issues that are harder to measure? And he was obviously talking there about both human rights and where he got in as well as water. George, what were your thoughts on Roberto and his missive on how do we do a better job of measuring this and what are some of the mental models or mental accounting traps that we seem to find ourselves in our industry?

George Beesley: Well, I always love Roberto's energy, but I think the thing that I took away was that there are many layers to the issues that you encounter as an investor with ESG ratings. So aggregate confusion 1.0 is mainly focused on we're measuring different things but calling them the same. That's confusing to people who want to use those ratings. So I think I can't remember the stats exactly off the top of my head, but it's something like there was a naught 0.6 correlation between the ratings of different ESG ratings providers, and in credit ratings it's something like 0.96. So it was very, very close. Whereas you have this real dispersion in ESG ratings, and one of the reasons is because they're measuring different things. That was something like 50 or 60% of the variance was due to that.

But you also have a number of other issues as you start to go a bit deeper. It's not just that they're measuring different things. It's also that if you are trying to collapse a lot of different factors into E, S and G, and then you're trying to roll that up into a single risk rating or ESG rating. You've had to make some trade-offs between how important do you think E is versus S versus G? And we often give the example of Tesla, and you say it has a very strong E story. But then on the S side, they've been accused of union busting and lots of other issues in the labor force. And then on the G side, some people have concerns around the very brilliant but somewhat erratic Elon Musk. So how do you trade off those things? That's always going to be subjective.

And if that is the case, you're going to have this dispersion between ratings, and if there's this dispersion, then how do you know who to use? Should you be using ratings provider A versus B? And in some ways, I think it's okay that they don't always have to be the same, but it requires the user to then have a very high level of knowledge as to ratings provider A is excellent on these different parts, whereas ratings provider B is very good on these aspects of social.

And then at that point — then the idea is that it's supposed to take a lot of data and collapse it down into something very easy to consume, but then you need a very higher level of understanding as to the processes behind the rating agencies. So then it doesn't always serve the purpose that you need. A couple of other issues, like it's backward looking often point in time and there's many, many different elements to it. So I think we're really just at the start of digging into this whole aggregate confusion project, and I know it has some future aspirations to change over time as well to look at the various different aspects, challenges and usages of ESG ratings.

Vish Hindocha: Yeah, absolutely. Roberto's fantastic at bringing this to life. And maybe just to use a different analogy that he used, I'm largely lifting from him, and we'll give him some royalties afterwards. But we were with some clients in the US earlier this year, and they had asked the question. They said, "I understand everything, but it is really hard for me. I have many other things to think about. I need to be able to collapse some of this into a single score." So we were going after the single scoring of ESG is loaded with subjectivity error rate. You can't aggregate how some of these things work. Even if you improve data quality, you're still going to have to have some of that.

So he used a fantastic example. So he said, there's a company that could have a great ESG score. It treats — and this is his example, not mine — but it treats us women horribly, but it does really, really well on carbon emissions. And basically, the CEO could come to you and say, "Can I treat my women slightly worse if I emit less carbon?" And of course you'd say, "Well, that's an insane question. No one should be really be thinking like that." But that is essentially the tradeoff that is happening when you're thinking about the single scoring of ESG.

And when we set mandates or we set requirements around what we want the ESG rating to be. So I think people just don't understand that it's not quite the same as looking at things like financial ratios or balance sheets accounts, some of which still has some subjectivity in it and still can have some creativity behind some of those numbers. No question. But it's not quite the same as an apples to apples because we've had US GAAP, IFRS for a long time that's tried to make sure that we've got a standardized way of looking at things. We're still at a relatively early and immature phase of where we are with ESG data.

If I remember correctly, one of the things that Roberto brought up in that episode, when I asked the question about favorite books, was Clayton Christensen, which has actually been a recurring recommendation. I read the Clayton Christensen was the Harvard commencement speech that he gave, and Roberto was actually in the audience for the original session ended up at MIT. So it clearly had a huge impact on him.

But Clayton Christensen also is credited with a lot of thinking around client-centricity and the jobs to be done framework and things that we talk about a lot at MFS. I saw him, I just thought actually, we've been talking a lot about some of the connected work, whether it's systems thinking, data, how we help clients understand, educate and empower them to really understand what it is we are doing underneath the hood to make sure that these risks are well attributed and accounted for, but also how we as an organization try and become more client centric.

George Beesley: The other bit that really stuck out to me from that episode was where he was talking about over a long enough time frame, the different aspects of sustainability all converge into one. So you might think that ESG issues are not financially material over the short term, and that often is the case depending on what the issue is and who the issue is. But we think about things in a compartmentalized way to make them easier to understand. So we make mental models of the world, and we can often think of sustainability in terms of its impact on society or the environment as something over there. And maybe the company itself, and the way that it treats employees, for example, which I think was the example that he gave in the episode.

And then the financial returns of the company is something very different. And you can take a very narrow view of ESG and just say, "All we're trying to do is manage these ESG risks and opportunities." And often that makes sense, especially if you're a short-term investor. And maybe even you don't even have to think about any of those because over the next year or so it's probably not going to have a material impact on the valuation of the issuer. But as you extend that timeline out, and MFS being a very long-term investor, you tend to find that these issues become much more material and you then have to think much more about systemic impacts, for example.

So we need a stable system in order to be able to earn the returns that we want from the issuers that we invest. And if you disrupt that system because you are not thinking about planetary boundaries or the impacts on biodiversity or climate change, or even on the social side as well, then you are going to really struggle, even if your goal is only to maximize financial returns. As over a long enough time frame, you have to start thinking about more bits of the puzzle.

Vish Hindocha: Yeah, agree. And we did cover that a little bit with Alex Edmonds, Roger Owen. We had quite a few guests on the podcast to connect some of that thinking around how you have to, we are reliant on the financial system, but the overall economy in order to generate the types of outcomes results that we, our clients and all investors and savers want. Great. Any closing thoughts, George, on third season? Anything on your mind? Any questions that you would have for our audience?

George Beesley: I think I would love some suggestions on what people would like to see for Season Four. So I think that the podcast has evolved massively from its inception, with much more of a focus just on sustainability related issues, to then broadening out, meeting some of the MFS investors, but also some leading academics. And we're thinking a lot about what is the best way to develop the podcast going forward for the next season? So it'd be brilliant to hear what's on listeners' minds and hear from them what they would like for us to cover, and what direction we should take it to be able to provide the most value.

Vish Hindocha: That's great. They can reach out to us at All Angles at mfs.com. That's great. Okay. Well, thank you very much, George, for joining us today. Thank you for listening to the podcast. I hope that was a useful summary of the key highlights, and moments from Season Three, and you'll see us back for Season Four. Thanks very much.

Speaker 3: The views expressed are those of the speaker and are subject to change at any time. These views are for informational purposes only and should not be relied upon as a recommendation to purchase any security, or as an offer of securities or investment advice. No forecast can be guaranteed. Past performance is no guarantee of future results.

60293.1