Investor Forum Autumn 2026

MARK ATKINSON: Good morning, everyone. Thank you very much for coming along to this investor forum. Historically, we've always held this autumn event in London, having had the AGM up here in Scotland in the spring. But now that the AGM is alternating between London and Scotland, we're doing it up here, and we'll follow that pattern. So whenever the AGM is up here, we'll be down in London for this. When it's down in London, we'll be up here.

The agenda today-- I'll just give you a quick run-through. We've got Craig Baker and Stuart Gray from WTW, two of the three people that run the strategy, pick the managers, decide on the allocations that they're given. They'll give you a high-level view of the portfolio as a whole, how it's positioned, how it's performed. And then we'll have presentations from two of the underlying stock pickers-- Sands Capital and GQG, GQG Growth Partners.

After that, we'll do Q&A, so if you wouldn't mind holding your questions to the end. Dean Buckley will join us for the Q&A. Dean's in the room-- together with the rest of the board, in fact. So lots of opportunities to chat to them afterwards, over lunch. With that, I'll hand over to Craig.

CRAIG BAKER: Terrific. Thanks, Mark, and good morning, everyone. It's always a pleasure to be up here and get an opportunity to meet with shareholders-- although I'm not naive. I realize that most of you have probably come to listen to two of our stock pickers who've flown over from the US. So Stu and I will try and keep ours brief, particularly given that we went into quite a lot of detail on this just 145 days ago at the AGM. But we'll try and hit the highlights.

So in my 15 minutes or so, this is the agenda of what I'm going to touch on. So first bit is to give you an update on year-to-date performance, why we've lagged the index. There's been some strong returns, but it has been behind the extraordinary performance of the index. We'll touch on why that is. I'll then put that performance in context with some of the longer-term performance that we did go into some detail on at the AGM, but I'll, again, reiterate what that now looks like.

Spend a fair bit of time on this-- latent value in the portfolio. So why we're genuinely excited about the portfolio today and why we think it's well-placed for strong performance from here. And that links to these reasons for confidence going forward. And then finally, I'll just touch on a review that we alluded to in the interim report that we've been doing on the whole strategy and any conclusions from that.

So on this first slide here, what I've shown are three numbers, on the left-hand side, all relevant to performance this year. So the first number there is the performance in the first eight months, so to the end of August up-to-date. We've got a return of 10.2% on the portfolio-- a pretty strong return for just an eight-month period. So very good in absolute terms. But just to be clear, the MSCI All Country World Index has done better than that. So we've underperformed the index by about 3.5% over that eight-month period.

On the positive side, since the AGM, in each of the last three calendar months, we've actually outperformed-- and cumulatively, over that June, July, and August, we've outperformed by 3.5% relative to the MSCI All Country World Index. So I mentioned that not because I'm saying, oh, well, three months means a lot. It doesn't. It's a very short period of time. What it does show, though, is that when the market isn't just focused on one thing-- and I'm going to talk through this in a bit more detail in a moment.

So AI has been driving returns throughout the market. When it broadens out and looks at other parts of the market, we actually do particularly well, and that's what we're designed to do. This is a balanced portfolio that doesn't want to put all of its eggs in one basket. It should be producing good returns across multiple sectors in the market, and that's why we've continued to produce strong, absolute returns for quite some time now. And we'll talk about that. And it's why we outperformed in those three months.

But on the right-hand side, just some important things to bring out. The first point is that over five years, we have underperformed that index. Again, same story-- strong absolute returns, but below those achieved by the index. And I'll touch on that as well. But what is important is that we've produced positive absolute returns, a rising dividend, a stable discount, and genuinely, we think we've got some really good businesses in the portfolio that will lead to strong share price performance from here.

And just as a reminder, we've done a number of things already. And in particular, I'd note the lower fees. The fees on this, including our fee and the underlying stock pickers' fees, was about 50 basis points, 0.5% per annum. That came down to 46 basis points as of the 1st of April this year, and will go down again to 41 basis points from the 1st of January 2027. Now, a typical fee for a multi-manager approach in the open-ended space is about 125 basis points, to put this in perspective. So 41, we think, offers particularly good value because of the scale that we're able to bring at WTW.

So what's happened in the first eight months of 2026? Well, two letters have pretty much dominated everything in markets-- obviously, in specific AI companies that have produced extraordinary returns. But actually lots of the sectors have been impacted by what's happening in AI across the piece. Now, I sometimes hear people think that we're negative on AI, and that's why we've had that underperformance in recent times. That isn't true. We've actually got a lot of AI in the portfolio.

We've particularly chosen a couple of managers that are going to talk to you a lot about what they're doing in that space. But the issue is that we're just not quite as concentrated in that one bet as the index has become. And we'll come back to that. To put this in perspective, of the five biggest contributors to our return in the eight months to the end of August, four of them were actually AI beneficiaries. So we've got up here Flex, Murata, Samsung, TD SYNNEX are all stocks that we owned and were overweight and really helped our returns.

So this is not a story of us being negative on AI. This is purely the fact that we run a balanced portfolio, and that isn't the only bet we've got, whereas on the right-hand side, we just bring out the fact that the index has actually ended up being quite concentrated in multiple different ways. The first thing being there that the top 10 stocks in the index now make up 25% of the index. That's the most concentrated the index has been in history. So it's about 40% of the US index is made up of the top 10 stocks, and that's the most concentrated that's been since the 1930s.

But perhaps another interesting statistic is that 85% of investors in the market now are not long-term fundamental investors. So that includes passive. That includes short-term traders. That includes retail investors, investing with leverage on meme stocks. And you'll hear a bit more from Stu on that later. But what that means is that the index has historically been the safe place of very diversified, holding a bit of everything in the market. That's actually become pretty concentrated, with one bet dominating it and momentum very much driving it. And we'd posit the fact that the index portfolio is more risky than our portfolio today, which is more balanced.

So to put that eight-month performance in context, we've got up here lots of periods of time since we've been managing the portfolio. So we've been managing it since the 1st of April 2017. So about 9 and 1/2 years. So that's the since inception number on the right-hand side. So if I just take the very top row of the first table, you can see the absolute returns have been very strong-- 10% per annum returns over that 9 and 1/2-year period, 6.5% real returns. That would normally be considered an excellent result, probably better than you might even expect over the very long term.

However, if you take the first row on the second table, you'll see that that is below what the MSCI All Country World Index has produced-- 1.5% per annum behind that index over that full 9 and 1/2-year period, although in reality, that's really a 2 to 2 and 1/2-year issue that's affected those longer-term numbers. The second row in that second table actually shows you that it really has been a concentration issue. So the equal weighted version of that same index-- we've outperformed that by 3% per annum.

It's the index that we've underperformed by 1.5% per annum, and that's because it's been so concentrated in those small number of stocks that have driven the market on one theme. And it's not just the fact that the top 10 stocks are so concentrated in the index. It's that almost all of those are in the technology sector. That is also unusual in this space. Now, we have beaten the average manager. So the average manager has struggled as well over this period. But we haven't beaten that index.

So this slide tries to give you a feel for why, despite that, we're pretty confident about performance going forward. And so what we do is we take the return on both our portfolio and the return on the index and we try and split it into two elements. So the first element is how much of that return has come from the growing value of the businesses that we're invested in or the index is invested in. So the fundamental growth. In this case, we've used book value of those companies. You can do it with earnings. You can do it with cash flows. It doesn't really matter. Book value just tends to be the more stable of the measures and the one that we've used at each of the last two or three AGMs so we've been consistent.

And what we've done is said, well, let's just take that first piece, the how much the businesses have grown. The second piece is how much the multiple has grown. So how much has the market decided each pound of value that a companies producing is worth? And so the price earnings multiple, the price-to-cash flow multiple, or in this case, the price-to-book value multiple has gone up. And that tends to go up and down over time.

What you'll see is that in fundamental terms, our companies have significantly outperformed those that make up the index. What's happened is that the index has, on average, got 5% per annum return just from the multiple expanding-- so not from the businesses doing better, but from the multiples expanding, whereas ours has got a 2.8% per annum return from that, whereas on the fundamentals, in the chart, you can see that we've significantly outperformed in what the businesses are doing. Why do we care about this split? Well, the reason we care is that over the long term, it's the fundamentals of the businesses that drive share prices.

If you look at any long-term period, that is all that drives share prices, on average. Obviously, there's individual stocks where that's not the case, and there's individual shorter periods where that's not the case, like this period. But over the long term, it is the fundamentals that will out. And so we sit here very confident about the portfolio from here, because we can see that that will ultimately end up being reflected in share prices. To be clear, it's not as if we haven't had good share price appreciation. 10% per annum is a good return over 10 years. It's just that the market has done even better through that multiple expansion.

You can see at the bottom there, we've just looked at some statistics on the overall portfolio that we've got today. We've got a price earnings forward, price earnings multiple of 16.7 versus 19.2 for the market-- so a lot cheaper than the market, but also with a higher return on invested capital in the businesses we own. Now it's a simplistic measure, but this implies we've got a much cheaper portfolio that's also higher quality than that of the index.

So what are the reasons for being confident about performance from here? Well, firstly, there is evidence that the market is starting to broaden out, although whether that is a short-term period-- and it goes back to AI driving everything-- only time will tell. But certainly, those three months-- June, July, August-- are evidence of that. And that has helped relative performance in that period.

Secondly, AI spending is under scrutiny. So four of the five big hyperscalers out there gone cash flow-negative. We're not saying that therefore means that they shouldn't be invested in. Far from it. But what we're saying is that's the kind of environment where stock selection and choosing between the winners and the losers will be much more important. And to date, that hasn't really mattered. And that's been the story here. You didn't need to pick the good ones versus the bad ones. You just had to be in the theme or not in the theme, and that's what drove markets.

In the bottom left, we've actually heard a number of the architects of AI in recent times talk about some of the threats to humanity, the reasons why some of that pace needs to be reduced. Again, I'm not saying that therefore means you shouldn't be investing in AI. Again, far from it. We think it's going to be a really important element of our portfolio. But what it does highlight is that there are always risks that you haven't thought fully about, about any theme you run. And you never want to have one theme driving everything in your portfolio, and that's why we continue to run this balanced portfolio.

And finally, those fundamentals that I talked about before ultimately have to reassert themselves. That has to come through in the share price performance, or ultimately, those companies will be taken private, and you'd get a pop in the share price when they are.

Now, one of the things we mentioned in the interim report that was recently published was that we'd been conducting a full review of the approach. This is something that we will always do when performance has been slightly different to expected. Not in absolute terms-- I guess in absolute terms, it's probably been slightly better than you might expect over the long term. But in relative terms, underperformance of the index over a protracted period of time is something we'll naturally look at.

And we've challenged everything we're doing on multiple grounds. Have we've got the right number of managers, the right types of managers, these 20 stocks, the right way of running money, have we got the right approach to risk management-- all of those things, as you would expect, were part of this review. And the board have been challenging us on all of those things, as well as our own internal review. What we've concluded, both WTW and the board, is that we're still very happy with the approach. We will want to continue with exactly the same thing as we've been doing before, but with one or two tweaks.

So the two things that I would just call out is, one, with this hyper-concentration that we've got in markets by stock and by theme, just having a few extra levers in how we manage risk, in particular-- some of those relative risks, not just absolute risks. We've got quite a lot of freedom already in being able to do that with the managers. It hasn't been an issue to date. But if this is going to continue to be a phenomenon that's there for a while, just having a few extra levers will be helpful. And so that's something we expect to tell you about in the next couple of months. That's still something that we're discussing the finer details with the board about.

And then the second piece is, with rates higher than they have been over the majority of the period, we've been managing the portfolio, just looking at whether there are other slightly less costly ways of achieving leverage. To be clear, this is very much at the margin. Most of our leverage is achieved through our private placements. That's about 70%, 80% of our leverage, and those are on very low rates that are there for a very long period of time. So we're only really talking about a couple of percent of the portfolio that's in a revolving credit facility.

So these are all things at the margin. But we wanted to be clear with you that a couple of things have come out of that review that we think might be able to help going forward, but they're not really going to change the way we're approaching things. We're still very confident, for all the reasons that I've gone through. Final slide from me, really, before I summarize and then hand over to Stu is to point out that we're now on course for the 60th increase in dividends in a row. So that 29.32 pence for 2026 would be up 3.5% from the previous year. Reminder, we've got huge reserves in this trust, a few billion. And so it's not as if we're concerned about being able to continue with the progressive dividend policy that's been in place for those 60 years.

So in summary, before I hand over to Stu, we have seen some green shoots of relative performance-- continued, very strong absolute performance, and relative to peer group. But we've started to see green shoots relative to that tough index to beat in a concentrated market. But because of the last couple of years, the long-term numbers don't look as good relative to that index. We're conscious of that. We own that. And we're confident that will change in the not too distant future.

The index has made a bigger bet on AI than we have. We think that brings risks with it, and we want to run a balanced portfolio. But that's clearly been the reason for its extraordinary performance. We think latent value is really building up in the portfolio. That gives us optimism going forward. And we're just going to evolve, a little bit, the risk management and gearing piece of the portfolio. But everything else will be staying as it's been for the last 9 and 1/2 years. At that point, I'll hand over to Stu.

STUART GRAY: Excellent. Thank you very much, Craig. Good morning, everyone. So I've got just a few slides here to talk about the portfolio and more about what WTW has been doing in the last 12 months or so, and then we'll hear from some of our stock pickers about the opportunities they're seeing at the company level. First, just a little bit more on the market context, as Craig touched on earlier, the purpose of this slide is really to try and convey the way the market is behaving at the moment.

So the left-hand chart is showing the stat that Craig mentioned, which is that a significant minority of the market trading today, about 15% or so, is driven by long-term fundamental investors, like Alliance Witan. The vast majority of trading in the market really comes from either retail or algorithmic traders, or quant traders, or ETFs and passive funds. And those are really looking for signals, momentums, narratives, technicals that drive their trading behavior, rather than the long-term thinking about companies.

Why is that important? Well, it means we get some interesting effects happening in the market. And the chart on the right talks about what happened in 2025, but maybe a more up-to-date example is 2026, this year-- you've seen memory stocks really explode in value, up until May this year. So memory stocks, which go into the AI supply chain-- companies like Samsung, SK Hynix, and Micron, which make memory chips. Samsung went up about 500% in the 12 months after May this year, and then fell about 40% in the subsequent few weeks after that.

So there's a very sensible reason as to why Samsung went up in value. There's more demand for memory chips. They're earning more revenue and more profits. That's fine. The problem is with them going up 500% in a very short period of time and then falling 40% immediately after. That's really driven by market trading with leverage, chasing these returns, and then deleveraging afterwards. So this behavior is really driving prices in a very big way, very quickly in either direction. And that's the environment in which people are investing in today.

So what does that mean for Alliance Witan? Well, if you remember, this is the Alliance Witan portfolio. And if you remember, the whole point of what we're trying to do here is to be a core investment for investors to compound their wealth over very long periods of time. We're trying to achieve very high levels of absolute return by investing in our managers who have very concentrated portfolios. We want these maximum 20 stock portfolios to drive high expected returns in the long term.

And as Craig said, we don't want to put all your eggs in one basket. We want diversity. And that's how we get the core element that you can compound wealth over a long period of time. So we get diversity by investing across, typically, 8 to 12 managers. We have 11 in the portfolio today. And how we get that diversity is making sure that we select managers that are excellent at what they do and are very different to one another.

So they're finding different types of opportunities with different drivers, different end markets. And that brings the diversity to the portfolio. So we're trying to pick the best stocks within the growth area-- the value area, large cap area, small cap area, US, non-US stocks. We're just trying to pick the best stocks rather than picking the best part of the market to be in at any one time.

So I'm going to talk a little bit about what actions we've been taking, as WTW, in the last 12 months. So first and foremost, we have to select the managers. As I alluded to, managers get in the portfolio if we think they're the best in the world at what they do and add something to the portfolio. Managers come out. If we are concerned that they are no longer the best in the world at what they do and/or they're not adding something to the portfolio anymore.

In the last 12 months, we haven't made any manager changes. So our typical holding period for a manager is around eight years. With 11 managers, that typically means we have one manager change a year, on average-- sometimes none, sometimes two. It can vary. In the last 12 months, we've changed no managers. That is two key reasons.

One, there's been nothing major enough happening that caused us to change our view about the manager. But secondly, because of the market environment that I was alluding to, we've seen very extreme volatility in the market, extreme price movements. It's not an environment that's very conducive to chopping and changing between managers very aggressively. So we've not wanted to do that. What we have been doing, though, is thinking about the allocations we've made to the managers.

So the chart I'm showing here is I'm showing you the current allocations, which is the dots, and the range in the last three years, which is the bar chart. So the first and hopefully most obvious thing from this chart is that pretty much all of the manager allocations are at an extreme relative to their historical range, or at or near an extreme. And this is really a reflection of the market environment that we're in. This is not Craig and me trying to say, oh, we think this is the best manager for the next 12 months. This is really us trying to balance the overall portfolio in the context of the market environment, which has been very concentrated and very extreme.

So more specifically, a market that's been dominated by AI. It's been a very, very strong market. Market beta has been incredibly strong. The managers in our portfolio, who typically have a higher beta, are the ones with higher allocations today. So managers, like Sands, who you're going to hear from soon-- also, Lyrical; also, Jennison-- these are managers with higher beta portfolios, and they have higher allocations today to help us manage the beta relative to what is a very concentrated market environment.

On the other side of that, managers who focus on higher quality companies, more stable businesses and typically lower beta-- those managers are at the lower end of their range today. So managers like Veritas, Metropolis, or even GQG. They're at the lower end of the range. And that's just a reflection of how we put together the whole portfolio to make sure the characteristics of the combined portfolio really are similar to the market, so that we're not delivering a big factor bet relative to the benchmark. We're really focusing on the stock selection driving those returns.

So with that, I'm going to hand over to our stock pickers, who will tell you much more about what's going on in the portfolio at the company level. Just before I do, I'm going to hand over to Danielle at Sands and then Brian from GQG. But just a quick word on how they fit into the portfolio. Sands Capital is one of our strong growth managers. They're looking for a really high growth businesses that can sustain very high and above-average growth rates for quite long periods of time and invest in those companies with a relatively low turnover, although I'm sure Danielle's going to touch on some of that in her presentation

GQG, on the other hand, is-- well, the clue's in the name, Global Quality Growth. They're looking for quality companies that can deliver growth over the next five years or so-- good, strong growth in cash flow. But they also think about the valuation of companies, and they're a lot more willing to question the risk/reward trade-off they're getting in companies around what price they're paying for it versus the quality of the cash flow growth they're expecting. And that leads to them changing the portfolio a lot more rapidly than some managers in our structure.

So GQG, at times, will be heavily into tech. At times, they might be more concerned about tech valuations and come out of tech and move into energy, health, consumer. More recently, they've rotated back into the technology space. So I'm hoping Brian is going to explain a little bit more about their thinking and how they move that portfolio around for you. But that's how they play a role in the portfolio. They bring something different. We've got other managers who are more focused on value investing on the other side, and that's how we blend them together to create a diversified portfolio. So with that, I'm going to hand over to Danielle from Sands Capital. Thank you very much.

DANIELLE MENICHELLA: Thank you. Good morning. It's great to be here today. Thank you for having me. My name is Danielle Menichella. I'm a portfolio manager at Sands Capital. We're based in Arlington, Virginia, which is near Washington, DC. I'm the portfolio manager on the global focus strategy, and I'm excited to talk to you today about some of the opportunities that we're seeing in the global focus strategy, as well as opportunities in the market.

At Sands, we do one thing. We invest in leading growth businesses globally. We are high-conviction, highly concentrated, low-turnover growth investors. One of the unique things about our investment and research process is we span both private and public, and we think that gives us an advantage because it is interesting to study these new businesses and these new spaces to determine where the world is changing in 10 years down the line, 20 years down the line. At the same time, being able to study established businesses really helps us understand how to take that secular shift and create sustainable, profitable businesses over the long run.

We use our six criteria and we apply it to the market that we are looking at. We are looking for businesses that have sustainable growth, have a market leadership in attractive business space, sustainable competitive advantage, a strong value-added focus, mission statement, good management team, looking for businesses with healthy financials, and pay attractive valuations for that. And our investment philosophy is really steeped in a couple of key observations. Number one, the number one driver of a stock price over the long run is the underlying earnings growth and the sustainability of that for the business. And a lot of times what the market misses is in these truly special, unique businesses that are benefiting from secular shifts, they're missing the duration of that growth as well as the nonlinear nature that that growth could be.

So our portfolio-- as I mentioned, we are looking for innovation. We are looking for disruption. We are looking for strong growers. But we also are looking to combine that with quality market leadership positions, companies that have a compounding nature, have strong free cash flows and strong ROIs as well, so that we create a portfolio with a diversity of earnings drivers, a diversity of sector exposures, a diversity of geography. But we really want to have idiosyncratic drivers in our portfolio. Nonetheless, our current portfolio is highly geared to the theme of AI, and that AI has helped our portfolio over the recent quarters.

So one of the key outperformers for us has been our exposure to AI infrastructure. At the same time, being underweight in sectors like energy and consumer staples has also helped. Now, on the other hand, there have been detractors to performance. For one thing, we saw the market shift, towards the end of last year and the beginning of this year as well, from those perceived players that did well in technology previously-- so the internet platforms, the e-commerce platforms-- and investors took money from that to put it into AI infrastructure, and picks and shovels.

So having exposure to internet platforms did actually hurt us this year. Healthcare names have been tough for us this year. And perhaps the biggest difficulty that we've had is our stock selection in the financial sector-- specifically, we have no exposure to European banks and US banks. And what has happened as rates have stabilized, we've seen the risk go out of the system, and those banks re-rate over the last about 18 months. And that's been a detractor for us.

So I mentioned that we have an overweight in AI infrastructure and picks and shovels. AI is certainly not new. It's been in the news for many years. In fact, when it's in the news recently, it's usually in the realm of, are we in an AI bubble? I'm going to cut to the chase here. Our view is that we are not right now, but we are eyes wide open about this. I mean, there's a couple of things we have to think about. AI is a capital cycle, and not every allocator of capital in this cycle is going to have accretive returns, even though AI-- we see it as one of the most transformative technologies that there are.

So what does that mean? It means we have to be picky. We have to be specific. The idea, the opportunity is not to just go out and buy every shovel that there is in the picks and shovel space, but to pick the few quality businesses that will benefit from being at choke points right now, and even better, those businesses that aren't obviously AI, but they're using AI, and they're using it, thinking about where we're going forward. They're using it to make their businesses better, to have lower costs, to get to their customers faster, to improve customer relationships and entrenchment. Those are the kinds of businesses that we are looking for right now.

When we think about AI in the next 10 years, we certainly think that it's going to be much, much larger than it is today. But that does not mean that the path is straight up and to the right. Very, very possibly, that is going to be a volatile path. We don't think that there is a bubble right now, but there probably will be a bubble at some point. Of course, we're humble about this. When you're looking from the top, things always look pretty good.

But what helps us sleep at night is the fact that there is still significantly more demand for compute than there is supply. Data centers have less than a 1% open capacity. We're seeing companies that have real revenue and real revenue growth associated with AI products, and the companies that are investing in AI are still healthy on a financial basis. They are profitable. They are having operating cash flow. And if they are taking on debt, it isn't very much that at this point.

The next couple of slides really talk about why we're still very optimistic about the AI opportunity, despite all the investment that has gone on so far. I like this slide because it's showing how AI-- we are at the very early stages of it. The growth wall is still very far out, yet the adoption rate is much faster than things that we've seen in the past. We're comparing enterprise AI here to other transformational technologies over past decades-- PC, mobile, and TVs. And you can see that it took decades and decades for these other technologies to reach meaningful penetration levels.

We are a year and a half into enterprise AI, and right now, we're at 5% penetration. And the way we get to that 5% penetration is that we think about the world this way. There are one billion knowledge workers in the world. We believe that, eventually, every one billion knowledge workers will have an agent, an AI agent for use. Now, you can argue maybe those knowledge workers should have more than just one agent. You could also maybe argue that because of agents, there are fewer knowledge workers. But in general, think of a one-for-one ratio.

Right now, there are 50 million agents. That's 5% penetration rate. So that means that we got to that 5% penetration rate in 1 and 1/2 year. In fact, in March, that penetration rate was only 2%. Yet we have 95% left of the market. Another statistic that I really like is The Financial Times talked about Anthropic. In August 2025, Anthropic had $5 billion of revenue. July 2026, 11 months later, it had $65 billion and is expected to go to $200 billion by 2028.

Anthropic has been in existence since 2021-- so in seven years, going to $200 billion in revenue. Now, compare that to some of the software giants that we think about in the past-- Adobe, Salesforce, ServiceNow. It took those companies somewhere between two and six decades to get to less than $50 billion in revenue. So we are moving here very, very quickly.

Now let's talk about median AI spend for corporations. The median corporation spend on AI per employee per month is $11. Sounds pretty good to me, right? You're making $11 a month per subscriber. This could be a lot higher. And the reason why we believe that it can be a lot higher is the top 10% of AI users-- so you can assume that those are users that are getting some real value from deploying AI to their workers-- they are spending $611 per employee per month.

Now, the top 1% is spending $7,500 per employee per month. Look, I'm not saying we get to that level. I think that is an extremely high level, the top 1%. But the median today is $11. The top 10% is $611. That $611 means something to me because-- and this is a number that I shared with my husband recently, and his mind was really kind of quite blown about this-- the median corporate spend for software per employee per month is $740.

Now, that whole SaaSocalypse that we talked about or that we all heard about at the end of last year and the beginning of this year that caused all of those software companies to have their stock prices really beaten up over the long-term fear of disintermediation by AI-- spend on those things is about $750 a month. So to me, that means that AI spend could be much higher than it is today, especially as it proves itself as having real value the way software has, and can move towards those levels.

So obviously, AI is a big theme in the world. It's transformative. It's going to change our lives, and we see it as a big revolution. How do we think about the different layers and the different opportunities that are out there today? As I said, AI is a large overweight in our portfolio. It's a large overweight across all the portfolios at Sands. And we bucket the opportunity set into three main categories.

There are the enablers-- the companies that do the picks and shovels, that do the infrastructure. They're making the chips. They're offering the cloud services. They're providing the models. They're doing the energy infrastructure. Then, we have the builders-- the companies that are building something with AI. They're creating autonomous systems. They're creating robotics, they're creating data and analytics. And then we have the beneficiaries. Now, these are companies that are using AI to make their businesses better. And whether that means that they're making products faster and cheaper, whether they're creating greater entrenchment with their customer base, they're doing better things with customer support, they're providing better products.

These are the three categories. And it's not always that a company that we own may fit in only one of them. There's some overlaps, right? So Amazon, for instance. AWS-- one of the biggest cloud services providers in the world. But they are also using AI to build autonomous systems and robots in their warehouses. They're also benefiting from AI on their e-commerce business by offering AI agents and AI assistants that are helping customers purchase more.

I do want to share our portfolio overall. Our global focus strategy is one of the highest concentrated, highest conviction global portfolios that we offer at Sands. It has 20 businesses that we own in it, and there are a lot of AI businesses in it. We have exposure to the AI enablers. But we're also thinking about all the companies in it in an AI lens.

So we have the AI enablers. We own companies like NVIDIA that make the compute; TSMC that makes the chips for the compute; SK Hynix, who makes the memory so that the AI analysis can be performed; ASML that makes the equipment so that you can make the chips; Quanta, who provides the construction and the maintenance for the data centers to get the energy to them. We also own AI beneficiaries, and we're going to talk a lot more about these. Now, these are companies that, on the surface of it, they're not really AI companies, but they've been using AI already to make better products.

We're going to talk about Axon in a minute. We're going to talk about Cloudflare, and we're also going to talk a little bit about DoorDash. But Intuitive and Keyence are companies that use AI to help make automation and robotics-- Intuitive in the health care environment, Keyence in factories. Mercado Libre and Shopify-- these are e-commerce players and businesses that are using AI agents to help make better sales and to get to more effective offerings for their customers.

And then I do want to focus on this, too. We are not an AI portfolio only. We have AI overweight, but we also like to have idiosyncratic growth drivers. So these are businesses that can do well with or without AI. 3i, who owns action discount retail stores throughout Europe; Galderma, the pure-play dermatology player that does Botox and filler, among other aesthetics; McKesson, a pharmaceutical distribution across different disease types; and Woodward, a beneficiary in aerospace.

Going back to that middle group, these are the companies that are already using AI in a good way to really help drive their sales. What you can maybe see from the charts here is a combination of a couple of different things. We've talked about how the index itself is quite geared and has a heavy weight in AI. But our portfolio, which we also see as being overweight AI, has different types of exposure, and we have a lot of different types of AI businesses that don't necessarily look like the index.

I'm calling out 5 users of AI here-- Cloudflare Axon, DoorDash, MELI, and Shopify. I'm going to point to DoorDash. Just a quick little note on that one, DoorDash-- I'm sure you've all heard of it, but if not, it's a platform that's connecting consumers with restaurants and other retailers, and then drivers to deliver these products. And DoorDash has announced that by just adding an AI assistant to their offering, they've been able to generate between 24% and 35% higher sales. So a very simple thing that they're doing, but using AI already to making their business better.

Now, I'm going to talk about Cloudflare. And I think Cloudflare is an interesting one, because it's a business that a lot of people haven't heard of, yet it's also a business that most of you might, I guess, would be using on a daily basis. It is a company that is providing a service that's sitting between you and a website and an app, and it is making that service fast and cheap and safe. I think about Cloudflare-- and why I say that I think you're probably coming across it on a daily basis is that Cloudflare is used on about 20% of internet sites. And so these are sites that you're going to, you might not even realize that Cloudflare is making it better for you.

We think about Cloudflare as having three key roles. It's acting as a bodyguard, which is blocking bad traffic, preventing hackers from getting in, preventing cybersecurity threats to those websites. It's a traffic cop, making sure that there are fewer crashes for these websites, keeping the service up. And it's a delivery network that's bringing the content as quickly to you as possible. And the way Cloudflare does that that makes it a little bit better than some of the other cloud providers is that it has over 200 servers throughout the world that are sitting closer to the end users, and that makes the service faster and it makes it cheaper.

So companies like using and paying for the service because, let's be honest-- if you go to a website and it crashes, you're probably not making that purchase. If you go to a website and it's really slow, you're probably not going back to visit it to buy something again in the future. Or if you go to a website and you find out that there was some of cyber threat or hack, you've lost all trust, and this website's lost its credibility.

Now, Cloudflare also has one other major product, and it's workers. Workers is an infrastructure that was created for coders, initially-- human ones. And it's morphed, now, more into an area where AI agents can do coding and carry out their workflow. Again, this is preferable to going to an Amazon AWS, because this service is sitting closer to the individual computers that are doing the work, and that's making it faster and cheaper to run. So Cloudflare was a great growing business before AI. Now, with AI, we've seen traffic on the internet, both human and non-human, become much more significant, meaning that Cloudflare is more necessary. And at the same time, it has an infrastructure specifically designed to help AI agents.

So we've gone from the figurative traffic cop/security guard to the physical security guard now, and traffic cop, and real cop, with Axon. Axon's mission is to protect life. Many of you may already it as the monopoly producer of tasers. It's also one of the market leaders in the world for body-worn cameras, and it also is the leader in digital evidence management systems. That's basically a fancy word for a lot of software that helps store and write police evidence reports. What makes Axon special is that its hardware and software are integrated together, which really takes away the friction for police officers, as data being captured by cameras, being stored on the Axon cloud, being held in there indefinitely, and shared with judicial and legislative departments.

What Axon has done in recent years has been, to me, extremely powerful in terms of having a real and monetizable commercial AI product. Axon started offering something called Draft One, which is a software program for its customers. And this program has really saved their customers time and money. I'm using one of the early adopters on this slide to talk about how this provides meaningful payback and ROI for customers. So the data that is being taken via the camera and stored on the Axon cloud is now able to generate automated police reports, evidence reports, and witness reports.

What you may or may not is that in the US, police officers will spend three hours out of their eight-hour shift sitting behind a desk writing up reports-- evidence reports, incidence reports, police reports, witness statements. That's three hours that they cannot be on the ground interacting with the community, stopping crime, solving crime. This is not why people become police officers. Police officers don't like doing this.

What Axon Draft One has allowed is for the amount of time spent behind the desk to fall by almost 70%. So instead of spending three hours, police officers are now only spending one hour. They still have to check the report. They still have to make sure it's correct. They still sign off on it. They still have accountability, but they're able to do a lot less time behind the desk, spend a lot more time on the ground doing their real job.

And maybe you know or you don't know, also, this-- in the US, it's almost impossible to hire police officers. People don't want to do the job. So departments have budgets that they're looking to spend to hire more people, but they can't. So instead, they can spend it on Axon Draft One. And we think that this is a compelling solution.

Speaking of compelling, just a quick slide here on how valuation for this portfolio has become more attractive. The businesses that we own in this portfolio have grown strongly. They've grown in line with our expectations or actually beat our EPS. Our long-term EPS CAGR for global focus is over 30%. Now, you can think about that relative to an index, that the EPS CAGR is probably-- historically, it's been closer to high single-digit or low double-digit, so we're getting significantly more growth. At the same time, the index has probably re-valued or re-rated, whereas this portfolio has de-rated.

And our next 12-month PE for this portfolio is at 18 times. This is one of the lowest valuations that we've seen for it since inception. It's one of the lowest premiums, at about 5% to the index that we've also seen since inception. Yet the strength of the businesses, the fundamentals of the businesses themselves are still quite strong. So we see this as a great opportunity to get into this portfolio.

What do we have to worry about with this portfolio? Well, there's always risks. Obviously, if there was some of AI implosion, that would be bad. To me, the greatest risk is that we have another period like we did in 2025. This is a growth portfolio. And in 2025, in international markets in particular, was the greatest year for value outperformance versus growth in the last 50 years. Other things that we worry about with this portfolio-- higher interest rates for longer, but I think that's being adequately compensated at these valuation levels; any severe geopolitical risk; uncertainty related to war; or macro where the consumer.

But in general, we are looking to own businesses that can make their own weather over time, that have market leadership, that have strong competitive advantage. So in conclusion, we're excited about where the market is. We still want to build diverse portfolios with different earnings drivers. We think that the AI story is meaningful, but it's beginning to shift more to how companies are using AI to make their own products better. The opportunity set is there. The growth is there. Valuations are looking attractive. So now is a great time. Thank you.

BRIAN KERSMANC: Good morning, everybody. My name is Brian Kersmanc. I'm with GQG partners. We are managing two sleeves, essentially, within the Alliance portfolio structure here. We run a concentrated global portfolio within here, which is a 20-stock global portfolio, and we also run emerging markets as well, which is a little bit longer tail in terms of a portfolio, but also a high-conviction portfolio as well.

Just to give you a little bit of high-level brief background-- I don't want to put anybody to sleep here-- we've been around for 10 years. At GQG, we're running about $150 billion in assets under management. The portfolios are run by three individuals-- so myself; the founder, Rajiv Jain; and then also one of our other co-PMs, Sudarshan Murthy. The same team runs our global portfolio, our international-- which in the US, for us, is [? XUS-- ?] emerging market, and US equity portfolio. We believe there's a lot of strength in that, because you can see a lot of the things that are happening in the emerging market side that translate back to the US and vice versa. So a lot of what we do is trying to connect the dots and seeing the forest for the trees in terms of how we look at things and how we operate.

So one of the things I like to start off with is, what are we trying to do? And philosophically, what are we thinking about? And what GQG is trying to do philosophically is we're trying to find the forward quality element of businesses. So we're trying to capture that forward quality of growth. What does that mean, intrinsically? What we believe is that quality is dynamic over time. The definition of quality stays the same, but companies move in and out of those buckets of quality over the course of time. What we're trying to identify is what are those companies that are exhibiting that increasing quality and shying away from the ones that are maybe deteriorating, in that sense, over the course of time.

So let me give you an example of this. Think about an energy company. If I were to ask a room full of people a couple of years ago-- maybe not as much, today. We have the Strait of Hormuz closed down, and everybody's concerned about energy prices. But a couple of years ago, if I would ask a room full of people like this, is Exxon or is Total a high-quality company, I might get one person raising their hand in the air and saying, yes. If I were to go back 10 years ago or 20 or 30 years ago and asked a room full of folks like this, is Exxon a high-quality company, I would get 49 out of 50 people raising their hands in the air.

They would say, extremely high-quality company. This is a blue chip company. It's the largest market cap company in the entire world. They compound through cycles. It is an incredibly high-quality company. So the question that GQG is asking constantly of ourselves is, why was it considered so high-quality back then? Why was it considered so low-quality in the medium term? And are the pieces in place for this to be an exceptionally high-quality company going forward? So this is what we're trying to find, in terms of the companies and the portfolios and how we manage this.

Now, in terms of the returns and what we're looking for from a return perspective, we are a little bit more of an absolute return-focused portfolio. And what we want to do is we want to build a portfolio that drives high single-digit, low double-digit total return. That's what we're trying to do. We also want to try to find a decent amount of the upside when the market is a little bit stronger, but we will inherently lag, because it's going to be harder for us to find the super growthy elements of the markets and things like that. So we will lag a little bit in that sense.

And then on the flip side, when markets get super volatile and incredibly chaotic and things like that, we do protect better to the downside. We are very strong and firm believers in that downside protection and that capital preservation. And one of the things we've highlighted here is if you look at some of these bands going across, these are some of these big downside moments, where the market does get pretty volatile. And you see that outperformance shining across the board.

Now, if I look at over the course of the past year to 18 months or so, we've done a decent job in terms of capturing that total return target or objective that we've talked about. We didn't do enough-- and I'll be the first to admit this-- we didn't do enough to execute on and capture enough of that upside when the markets have been so strong. And we'll talk about where we've been and where we're going at this point in time.

OK. So I don't want to spend too, too much time here, because there's a whole lot of information on this slide. But just to give you a snapshot of what the portfolios look like and how we're viewing the world today as an expression of what we look at today, the one thing that I will caution folks when you look at a holdings report from GQG at any one particular time, is we are very dynamic. And when I say we are very dynamic, I mean we are constantly trying to recalibrate the portfolio. The term that I like to use is a Darwinian approach to deploying capital. The stronger thesis is the one that survives and thrives within our portfolio. It competes away the weaker thesis in the portfolio.

I always joke that there's times where evolution happens a little more slowly and the portfolio stays a little more consistent. The turnover is lower. There's a time and place when the asteroid hits the Earth, it wipes out the dinosaurs, and evolution happens a lot faster. So you see the turnover and the adjustments made in the portfolio changing a little bit faster. I would say that we've gone through a little bit of the latter within the last couple of months. We've changed and adjusted and made some decent rotations in terms of the portfolio.

So you do see we have a decent amount of technology exposure in the portfolio now. That would have not been the case three, six, nine months ago. We would have had a lot more utilities exposure, health exposure, consumer staples exposure. And again, we can talk about some of those things there. Similar dynamic. And again, we can talk about some of these in a little bit across the emerging markets portfolio as well. You can see what we hold right now and where we are from that perspective.

OK. So I know we're talking a lot about AI today. I'm going to give you a little bit different flavor in terms of how we're thinking about AI-- this is coming from somebody in a group that was formerly AI skeptics-- and what our understanding of what the data has evolved over the course of even a year-to-date period and how things have changed, and why we've become a little bit more constructive on what we're seeing more recently. So, for starters, why were we skeptical to begin with? What was the issue? Why were we staying away from it, and why were we positioned very minimally on the tech side of things in AI previously?

Essentially, what we were seeing is heaps of capital, hundreds of billions of dollars of capital that were being invested across the ecosystem through the hyperscalers, through the hardware companies. And essentially, there were two companies that were driving all of this investment. You think about OpenAI and Anthropic that are driving the lion's share of this investment flow. Now, what's the issue here? OpenAI and Anthropic were making, at that time, minimal amounts of revenues, to begin with-- So you're talking about $20 to $30 billion of revenues initially against $700 billion of CapEx spending that was happening-- and very little profits. In fact, negative profits.

And unlike software companies, the more users are actually using this thing, the more the intensity of the cost goes up, so they get even less profitable over time. So this was our issue, is that they were constantly going to the private rounds to raise more cash. And as that cash is being raised, it's going through the hyperscalers and going through the hardware folks. So even though the earnings looked strong for the hyperscalers, to some extent, the growth looked strong and looked really strong for the hardware folks, you couldn't see the sustainability, the headroom for the continued growth going forward.

OK. So what's changed? What's changed on a year-to-date basis? What has changed is that the compute demand has absolutely exploded, and the supply is more constrained than what we thought. And we believe there is actually more durability to these trends than and what we saw before. I like to use the example, actually-- if you look at a year-to-date basis, you have agentic flow, agents and things like that, causing much more demand for compute. And what was previously seen or thought about as token-maxing, where you just try to drive as many tokens and use as many as you possibly can, you're now seeing value-maxing.

Now, what does this all mean? This means that the cost of actually delivering AI is actually coming down. People are using older models for regular mundane tasks, or they're using-- the open source had an absolute explosion in terms of usage on a day-to-day basis going forward. I like to use the analogy of a car and types of cars. So I thought the frontier models are more like a Ferrari or Lamborghini. Right? Super expensive. And if you have a purpose of driving incredibly fast and you want to beat other people around the track, there are people that are willing to spend limitless amounts of money to do that.

But that is not a broad-based global phenomenon that everybody is willing to use. It's not practical. I'm not going to go and pick my kids up in a Lamborghini. Number one, I don't even fit. I'm six-and-a-half feet tall. I don't fit in a Lamborghini anyway, so that's a problem I have beside that. But it's not practical for day-to-day use. Right? So there'll always be somebody that's willing to spend that, but that's not enough to drive everything that we're seeing.

So with this explosion of open source that has come through, with this explosion of use of these older models, you've driven the cost of the vehicle down. And why is that important? That means now, on the same highways, the same toll roads-- which is the cloud-- you have Honda Civics, Fiat 500s, Volkswagen Passats driving. That is actually practical for day-to-day use. And this is why we've seen the demand for compute really accelerate, and why I think it actually is sustained going forward versus what we saw before, because I don't necessarily even need OpenAI and Anthropic to survive for this ecosystem to continue. And that's where things start to get interesting.

Now, how does this translate into some of the data that we're seeing? You see this here. The price per megawatt continues to go higher and driving significantly higher going forward. That's a reflection of the demand coming through. Here's another thing that we've been seeing more recently. This is GPU rental costs. Now, what's interesting about all three charts-- if we would have looked at this last year-- you see that it starts in January 2025, goes to January 2026, and then on to July of '26. If you would have looked last year, you saw declining prices in GPU rentals across the board.

So the B200s-- the ones all the way on the right-- those are the most recent chips that NVIDIA is putting out, the highest, the latest, and greatest. And then you have the H100s, which came out a couple of years ago, and then the A100s came out four to five years ago. And you saw declines across the board in every one of those. So we're saying, OK. Where's the value coming through? Price is declining. There's more capital being spent. We'll talk about margins in a second, but margins were declining. So where is this demand necessarily being reflected in the data itself?

More recently, and this is what becomes interesting-- people ask us, why have we even changed on a year-to-date basis? You've seen an explosion of these prices on a year-to-date basis coming forward. Now, the one I'll call specific attention to-- sorry, I like to walk around when I do this-- what I call specific attention to is the A100 pricing. Now, again, these are four to five-year-old chips and you're seeing a dramatic increase in pricing. What we've heard more recently from some of the hyperscalers or some of the NIO clouds is that these A100 four to five-year-old chips? They're sometimes fetching longer-term contracts out to the end of the decade, to the end of 2030.

So now you're talking about a 10-year monetizable value on the oldest chips that are out there. That's a lot different dynamic than we typically see from chips or hardware networking or data center types of assets, and it's a longer term monetization. What is this turning into? This has also been corroborated by revenue growth on the cloud side of things. So you're seeing this revenue growth coming through a lot faster.

Now, here's the other one that I think is super interesting. If you look at-- this is margins. So AWS cloud margins, OK? And what we were looking at last year was more spending-- remember, the GPU prices were falling across the board-- and then you were also seeing a margin decline of 700 basis points last year, from Q1 to Q2, in terms of AWS margins. So what we are seeing-- and by the way, it wasn't just AWS. We saw Azure's margins decline as well. So the NIO clouds go from 20% margins down to 2%. CoreWeave was in that bucket. So we said, this doesn't look sustainable.

But more recently, you see this margin is expanding. Going back to my car analogy, these are the toll road operators for the car. And what you're seeing is now that more Honda Civics are filling up the highways, you're seeing that the tolls that they're charging, the traffic that they're seeing on those toll roads is going up. And they're monetizing that. The net cost of AI is coming down because the vehicle is a lot cheaper and easier to deploy, but the traffic going across those toll roads, those networks is a lot stronger, and that's translating into the earnings.

So the final thing is, we're individual stock pickers. As we mentioned, we're a little more valuation-sensitive than most in some of these cases. And we're looking for growth, but we also want to get these reasonable price and look at the risk/reward between these. And what we're looking at in our portfolios-- again, as I mentioned, we had a decent amount of health care, staples, decent amount of utilities exposure-- what you can see here is the valuation metrics that we've been seeing more recently for these companies. And what you see is that the Mag 7 names, the hyperscalers, have come down pretty precipitously in terms of valuation.

Last year, we were pricing in Goldilocks, in terms of what the expectation was that they were going to dominate and make money off of AI against those deteriorating trends. More recently, now you're trading at a high-teens multiple-- for example, for Microsoft, for a company that's a high-teens EPS growth story-- with improving fundamentals on the cloud and growth accelerating. And I'm comparing that to J&J. J&J is a fantastic business. Love the company. The problem that it's probably more of a 6% to 8% EPS grower, and the multiple had re-rated from mid-teens up into the 20s.

Same thing with utilities. There is a utility CapEx cycle happening. Utility growth is faster right now-- 7% to 9% growth for AEP, closer to 9%, a couple hundred basis points dividend yield. You get your double-digit return, but now I'm paying into the 20s for that utility as well. Again, I compare that to what I'm getting from Amazon. Amazon is simply a faster growth for a more reasonable valuation.

This is also an interesting one. This goes to the memory side of things. So these are actually the levered ETFs. This speaks to the froth that we were talking about earlier in terms of Stuart's comments. You see how much the froth has been peaked up and taken out of a lot of these memory names, in particular. Samsung went through that 40% to 50% drawdown during the course of July. Very violent downside return. Micron, SK Hynix, and those types of names-- they're much more attractively valued now.

Why? Because in that period where you saw prices dropping by that amount, this is what your earnings estimate trajectory is done in the meantime. In some cases, you see a 30-plus percent earnings growth in terms of the estimates. You're seeing these very aggressive moves to the upside in terms of estimate revisions to the positive side, and you've seen pricing continue to come through on the memory names against these declining valuations on the other side.

Now, I don't think that Goldilocks is necessarily priced into these stocks anymore. Furthermore, if I take a company like Samsung, over the next two years, what they have locked in terms of sales in memory, the free cash flow off of that is going to equate about 2/3 of their enterprise value. So at this point, I'm being paid to take on that amount of risk. So I think it's interesting and rewarding to be on that end of the spectrum from some of these names.

OK. Now I've beat AI like a dead horse. We're talking about AI. I mean, you guys got enough of a dose of this. Let's talk about some of the other things from an outlook perspective in terms of how we're thinking about things, interesting observations that we have overall. I'll start with the negatives first. We'll end on a positive tone here. Obviously, you have stubborn high inflation and higher interest rates. That's going to cause some impacts and deceleration, to some extent, in some areas that are maybe a little bit more capital-intensive or interest rate-intensive. So that's going to be an issue. Have to monitor it.

Then you have the energy spike. Obviously, you have two wars going on. No easy solutions for that. Ukraine keeps bombing Russian refineries. Refining capacity keeps coming offline. You have, obviously, the Middle East, the Strait of Hormuz having issues, and things like that. That's not an easily resolved situation. That's something we need to keep in mind. We still like energy stocks. This is part of the reason that we have some of that allocation there.

The last part is AI is useful. I think we're all finding use. Actually, I'll do this right now. Everybody in the audience-- who actually pays for AI right now? OK. So think about the hundreds, if not trillions that's being spent. The consumer aspect of it hasn't necessarily come through. So it's still a little bit of a show me the money story. The enterprise space is starting to pick up. Kind of to Danielle's point earlier, it is starting to ramp up. But it's still a not-foregone conclusion that it's going to come through.

Now, on the positive side, as much as I just said that nobody's really paying for AI, I don't think it's going anywhere. I don't think we're going back to the place where we're not using it at all anymore, and it's now picked up enough speed with these open-source models and with this utilization and integration to a lot of these enterprises where it seems to be continuing on. Even Anthropic could die or OpenAI could die tomorrow-- it would affect the ecosystem, but AI is going to persist, and I think you're going to still see that investment coming from other places.

Deglobalization. This is an interesting one. Because if you think from a macro perspective, if you would have told me, and we were arguing about this internally as a PM group, if you had told me you have "liberation day"-- I don't feel liberated, by the way-- you have "liberation day," with higher tariffs; you have an energy crisis going on; you have higher interest rates, higher inflation; all these different things going on, trade wars going on. Would you tell me that the economy is still plugging along and doing OK?

But it does seem to be. C&I loan growth, for example, in Europe is still plugging along just fine. US loan growth is actually still coming through. Consumers holding in. OK. So why is this happening? I think these last two points are a part of this. The terms of deglobalization. This started over the course of COVID, and I believe it's accelerated with some of the policies coming out of the US and the White House. I apologize in advance. So you have that happening. Right? What have we seen more recently that pushes this further in terms of this localization of manufacturing and investment and things like that?

You saw 17 pharma companies march through the White House and sign MFN agreements in terms of pharma and having the lowest prices. But also what came attached to that was production guarantees that they would produce more in the US. Do we need more pill production in the US for global demand? No. So you're actually creating more supply here. But what does that mean in the near term? That's more pick and shovel type of work and economic activity that has a big flywheel effect on industrials and banks and all these other types of things that you're seeing, and workers, in terms of wages and things like that.

Defense and utility. Now, we have decent exposure, as I mentioned before, to utilities, because we're constructive that the CapEx was actually picking up and their earnings were actually picking up, just because of the base load demand for power. Forget AI. The base load demand for power has been underinvested for some time. In Europe, there hasn't been enough of the connections, the interconnects, between the different grids. It's not smart enough to handle the variable power supply that renewables have. That is a small portion of the listed index, but a massive impact, from a flywheel effect, on the economy in industrials that need to have work for that, and financials, and things like that.

Defense. I think we can all agree, regardless of interest rate policy or inflation policy, defense spending is higher and it's probably going to stay higher for some time. We're seeing manufacturing actually picking up. That all has, again, small portion of the listed economy, massive part of the economic flywheel. So I do think that we've been talking a lot about AI up until this point. And AI, because of the growth, has out-competed some of these types of things in the portfolio or in the market. But I think it's very important to note as well that there's some really interesting ideas and concepts and names that are benefiting from these other things as well that's outside of just that overall AI CapEx cycle.

So the last thing-- we're not forecasters. Forecasting is hard. I'm not here to predict the future. As we say sometimes internally, we're not here to predict the rain, but we want to build the ark in case the rain does happen, or be prepared for that. We're just going to continue to adapt and adjust for the incoming data points as we see them, and we'll continue to adjust the portfolio accordingly to hopefully give some smoother returns over time. And with that, that's enough for me. Thank you very much.

MARK ATKINSON: Yeah. Q&A. Before we get going on the Q&A, and there are lots of questions have been sent in advance. I'm sure there'll be lots in the room as well. Can I just mentioned to people online that there's a feedback form in your chat function? We'd be very grateful if you could fill that in. For people in the room, there's feedback forms in your bags.

So questions. Let's start with one for Dean, because there are several questions on similar lines. They're all about the dividend. How does the board decide on the dividend? What's a prudent dividend yield? And how is that yield paid when earnings don't cover the dividend?

DEAN BUCKLEY: Yeah. Dividend. When I talk to individual shareholders, I think dividend is the one area where I get the broadest range of views, and I think I'll hear it again when I talk to you outside later. There'll be some of you that tell me that actually, you don't want a dividend. You want all of the return to come through capital growth, because the marginal rates of tax on income are higher than those on gains.

There'll be others of you that say, actually, you'd like the maximum possible dividend, because we rely on investments like this to fund our retirement, and the cost of living is getting more and more challenging, so we want more and more income from the likes of you. And there'll be others that say, actually, I've got all of my investments in allowance in a SIPP or an ISA, and I don't really care whether it's income or capital growth. Now, the board have got to balance all of those things, and we think that we'll appeal to the broadest church of investors by pitching the dividend yield pretty much where we have, which is at a small premium to those that we compete most against.

So in the closed-ended sector, you will probably think that F&C is our closest competitor. Their yield is 1.2. In the ETF world, if you buy a passive ETF, you'll get a yield of about 1.4. So we pitch our dividend at 2.2%, a small premium to those. We think at that level, we can achieve a progressive dividend. It's sustainable way out into the future. Now, in terms of the numbers what does that mean? Well, we've got a portfolio that generates about $100 million in income.

We've got expenses of around about $25 million, so net, our earnings are about 75 and the dividend costs about 110. So we've got a shortfall, if you like, of about $35 million against our earnings. But what we do have is a significant amount of revenue and capital reserves that we can use to support that dividend. So that's why we believe that our dividend is sustainable and progressive. So that's the thinking of the board. I know that some of you will have different views, different preferences, but we think we appeal to the broadest church with that strategy.

AUDIENCE: It sounds, from the figures that you've given, that every time you want to pay a dividend, you've got to go to these portfolio managers and say, sell your high-performing AI stocks and give us the money, and we can give the money back to shareholders-- from the figures you gave us.

DEAN BUCKLEY: Yeah, but it's friction. If you look at it in the context of the overall portfolio, we are talking about very, very small numbers here. And the numbers, by the way, are dwarfed by the numbers on the buyback. So we're spending about 200 million pounds a year buying back shares. So if you think about the $35 million that we're using to supplement the dividend, it's actually a very small number in the context of the overall size of the company.

AUDIENCE: Well, I suppose at the same point would apply that every time you want to buy back shares, you've got to do the same thing. You've got to claw the money back from the portfolio managers who are investing in high-growth, rewarding stocks.

DEAN BUCKLEY: You're right. But again, in the context of the overall portfolio, it's not an enormous number. And WTW will be making changes in the overall portfolio from time to time anyway. So these numbers are in the normal running and execution of the portfolio. There are no significant demands on the portfolio managers that, hey, we need $100 million to pay the dividend. That's not the way that it works. These dividend payments are made from the day-to-day running of the portfolio.

MARK ATKINSON: OK. For Brian and Danielle, I think you've both already touched on this-- certainly, Brian mentioned Amazon-- but what is the future of the Magnificent 7 stocks, which were the stars of the moment just a few months back and now seem to be down in the dumps a bit relative to their recent history?

BRIAN KERSMANC: OK. You want me to take that first?

MARK ATKINSON: Yeah.

BRIAN KERSMANC: OK. Yeah. So in terms of the Mag 7, one of the concerns that we had last year, again, was that the valuation was pricing in sort of perfection that they were going to absolutely win this race, so to speak. We saw that in terms of the 30-plus PE multiples on the businesses. The growth expectations were strong, so the return or the bounce-back, the investment on capital spend-- everybody was assuming this Goldilocks scenario. And we've had some underperformance now. So part of that was legitimate in terms of our staying away from those, because they have underperformed. Unfortunately, things like memory stocks and things like that had actually outperformed well beyond our staying away from those.

What I do see-- and again, this is part of what we're addressing in here, at least from the hyperscaler perspective-- is that the economics and the fundamentals do seem to be getting better for those businesses. The supply is actually tighter than I would have anticipated. The fact that you have moratoriums in New York, you have moratoriums on data center builds for new permits in Texas, the logistical challenges, the NIMBYism-- not in my backyard-- It's just harder.

Now, we wrote five-- we called them Dot-Com on steroids papers comparing what we were seeing to what we saw during the Dot-Com. One of the things that I think is actually different-- and it's always dangerous to say that. I know that full-out. I'm going to regret it already. No, so one of the things that is actually different is that it is harder to build data center capacity than it was to lay fiber back during the Dot-Com. That was much easier. You could roll that out very quickly. It is just harder and more constrained.

So I do think that constraint versus the supply-- that is what you're seeing reflected in terms of what we're seeing in terms of the hyperscalers. And you're seeing, again, the data. The margins have inflected higher. The growth is inflecting higher. The profitability is inflecting higher. Jensen Huang talked about $50 billion data complex projects or centers that were getting less than a one-year payback. So you're seeing that these things getting rapidly paid back. I think you're going to see these things perform, and that's why we, again, reengaged, and what we like where they are.

MARK ATKINSON: Yeah. Danielle?

DANIELLE MENICHELLA: Yeah. I think all of what Brian said is very true and we have similar views on that. The only thing that I would maybe add is a little bit different-- I still think that these businesses are going to do really well. The valuations have come down, but the expectations in terms of growth have not.

They've been meeting their expectations. Their cloud revenue is growing. Their margins are increasing. They're also so big in the indices. It makes it very hard to see a world where they become meaningfully smaller, because the capital is just going to continue to flow to them, especially as you have the outstripping of demand versus supply, filling up capacity for data centers almost as immediately as it's being built, and a greater demand over the next decades for all of these products.

MARK ATKINSON: Great. OK. Question here online, I think, probably, for you, Stu. What evidence are you relying on that you can identify managers who can persistently outperform the MSCI ACWI index or outperform the style benchmarks that they represent?

STUART GRAY: Yeah, sure. So I joined WTW in 2003. So I spent 23 years-- my entire professional career-- trying to pick managers that we think can outperform indices over the long term. When I joined, Craig had been-- well, you'd been running that team for 10 years, I think, already. So there's a lot of experience and a lot of resource dedicated towards understanding asset management businesses and identifying which asset managers we think have a competitive advantage versus the broad universe of asset managers out there.

Now, we know, of course, that also performance is noisy. No one outperforms every single year, year in, year out. The cycles of performance, we know, is noisy in terms of how you interpret it. But over the full decades that we've been doing this, the track record of our preferred managers we've selected-- on average, those have outperformed benchmarks over long periods of time. So I think there's something like 1.4% per annum outperformance from our preferred managers over the last 20 years, or something of that order.

Now, as I say, it doesn't come through every single year. And of course, the last year or two that we've been talking about today has certainly been more challenging for most active managers in the universe but our preferred managers versus the whole peer group have actually done very well. So we're very confident that how we differentiate between managers and those who are skilled at what they do and those who we think do not have a competitive advantage-- we think there's a lot of evidence to suggest we can identify those.

MARK ATKINSON: Any more questions in the room?

AUDIENCE: I think this question has been answered to some extent by the first speaker, but I'll have a go at it. I was wondering what influence-- and I'm not quite sure whether it's the board or whether it's WTW has on investment policy. Or is it left entirely to the now 11 or whatever it is investment managers?

As an example, if WTW and/or the board thought there was far too much-- and this is entirely hypothetical, I have no views on it-- was far too much money invested in low-yielding AI stocks and wanted to get into higher-yielding, what I'm describing as more conventional stocks, what, if anything, do they do about that? Am I right to think you said, the first speaker-- and you use the phrase "rebalancing the portfolio" in your half-year report. Am I right in thinking, from what you said, that you alter the amount of money you give to some managers and add what other managers had in order to achieve your objective, rather than influencing the stock selection by the managers?

CRAIG BAKER: Simple answer is yes. So we don't tell any of the managers how they should be constructing their portfolio. That's for them. We obviously do a lot of research, challenge them, spend a lot of time with them, I'm sure they can attest, to trying to understand why they're comfortable with the companies that are in their portfolio. But we let them run the portfolio. But we see all the trades that are happening every day, and we see what the overall portfolio looks like.

We've got two things open to us, essentially. One is, we can simply replace one of those managers. If we think that we've ended up with three or four or five managers all looking quite similar, we don't need three, four, five that look quite similar, we could take one or two out, even if we think they're high-quality managers that can continue to outperform. So we could do that, and we have done over time-- we change one or two managers a year, as we've said, on average.

But the main thing we're doing all the time is the weightings to those managers. And we start with a principle of rebalancing. And so over time, we are, on average, taking capital from those managers that have done well if the reason they've done well is primarily because of the style that they inherently run, and we're, on average, giving capital to managers if they've underperformed, but primarily for style reasons, and we continue to be positive about that manager. We'd be giving them capital, all other things being equal.

Now, to be clear, it rarely is all other things equal, and so we are actually digging into how much of their outperformance or underperformance has come from them calling the fundamentals of the businesses correctly or incorrectly, versus they've just had multiple expansion or contraction more than the market, and hence their style has been in favor or out of favor. But once we've done the analytics, we're absolutely changing the weightings to the managers.

And we're trying, essentially, to make stock selection drive everything. So very rarely will the portfolio look wildly different to the benchmark index at a country level or a wider industry level, but it will look very different to the benchmark at a stock level. So we've got about 70% active money. So it looks 70% different to the index, to put it in perspective.

MARK ATKINSON: I think there was another aspect to that question about the board. Does the board interfere in--

DEAN BUCKLEY: No. And from the board perspective down, the board sets the strategic framework in which we want to operate. So it will give WTW some sort of very broad guidelines and then allow WTW the full discretion to operate within those guidelines. And at any point, where WTW is pushing the boundaries of those guidelines, or wants to step outside of those guidelines for any reason, it will come to the board, discuss it with the board, and we'll jointly agree how we're going to move forward.

So an example of that would have been a couple of years ago, when we appointed Dalton to the range of managers who were managing a portfolio that was much more focused on Japanese opportunities than the normal manager, which is focused on global equities. At that point, WTW said, we've got a great idea, which we think would be good for the portfolio and good for shareholders. We want to talk to you about it. The board embraced the idea and we executed. But day-to-day, WTW get on and manage the portfolio with full discretion.

AUDIENCE: And they don't have to go to the board if they're making any minor changes in the allocation between managers?

DEAN BUCKLEY: No. Well, they certainly wouldn't come to the board with any allocation changes. If they're appointing a new manager and losing another manager, they will come to the board and advise us of that in advance. And the board might express some views. But ultimately, we respect the fact that this is a discretion WTW need the flexibility and need to be allowed to express themselves as much as possible in the portfolio, and we allow them that discretion.

AUDIENCE: May I ask how many board meetings there are a year?

DEAN BUCKLEY: Oof. There are six formal board meetings. We have four meetings with managers that are outside of board meetings, and we have endless meetings on the phone where I'm speaking to the board about various issues. I have to say that-- and this is not a request for more money, by the way-- but I have to say that the workload on boards has increased enormously over the last few years. But it's something that we all enjoy doing and we get a lot of satisfaction out of doing it.

AUDIENCE: Why has it increased?

DEAN BUCKLEY: I think [? Saba ?] has had a very big impact on the sector. The activity that they've undertaken has been a really big wake-up call to the sector. But I think, also, the competitive environment for the sector has increased enormously. Exchange-traded funds are a much more relevant piece of the UK supply environment now. Passive funds are much more competitive. The market environment is much more challenging, as we've discussed today. So there are a lot more variables for boards to think about today than there were three or five years ago.

MARK ATKINSON: OK. Maybe we can take one of the questions that came in advance. It's a long question, so I won't read it all out, but I think the nub of it is-- and you did touch on this, Craig-- but what does WTW say about performance relative to Alliance Witan's peers? Now, I think this questioner has got his own peer group in mind. But, you've already talked about the global AIC. But what about other peers?

CRAIG BAKER: Yeah. So difficult to comment on each individual's view of what a peer group is. So we look at two primary peer groups. One is the AIC global sector and one is the entire universe of global equity, both open-ended and closed-ended funds. We tend to look at those two extremes. But we're also cognizant of some global managers that are in the global income sector, and we'll look at a couple of those. But the problem with these things are that the more concentrated you make the peer group, the more you have to actually allow for some quite significant differences.

So amongst some of our peer group, there are ones that have a significant amount in private companies, and some of those managers will have got a very different return if they weren't allowed to invest in private companies. A couple of large private companies have made big differences to people's portfolios. So you have to strip that out to compare like for like on a pure public equity portfolio that we manage. And then similarly, some managers are running much lower risk, much closer to index portfolio in how they manage the portfolio.

Well, even though they've generally underperformed the index, they will have underperformed the index by less than us because they've gone closer to the index and run lower risk. So on average, we've beaten the peer groups over five years or more. Had a tougher time over the last two or three years, but beaten the peer groups over any long period. But obviously, you can come up with a peer group that's beaten us. It's just usually ones that have got some that have got private equity in them and/or some ones that have run a lot less risk.

MARK ATKINSON: OK. There's a question here about equity sectors, really, I think-- pharmaceuticals being one and metals being another. Perhaps this is one that maybe everyone can comment on. I mean, hopefully, WTW has made it clear they don't really take sector views, but those sectors do come through in the stock selection. Brian and Danielle can perhaps talk about that. Maybe mention something about defense and the impact of war? This question is saying that war has driven demand for copper and other raw materials stocks. There's also-- perhaps I can throw in another question which is along similar lines, which is about is BAE Systems a buy?

BRIAN KERSMANC: OK.

MARK ATKINSON: Brian, maybe start with you, and then Danielle?

BRIAN KERSMANC: There's a lot in that question. I'll go in reverse order, and if I forget anything, reel me in and bring me back. So in terms of BAE Systems, we do not have it in the global portfolio and not in the concentrated global portfolio. It's more of a risk/reward opportunity, instead of other things that I just think are a little bit more interesting for the type of growth and the valuation that you're getting. That being said, we do own it in our international portfolio. It's a smaller position. I'd say it's about 100 basis points within that portfolio.

So we do like it. I do think they're on some very attractive programs, in terms of the aeronautics programs, the F-35 program, the typhoon fighter program, the electronics programs, the new military vehicles and programs that are coming up for bid. And they're seeing increased exposure. The sea-borne-- so the dreadnought fleet, and things like that. So I think there's some really interesting things there. I think there has been an underinvestment in just basic things like-- this is going to sound bad-- but bullets and missile systems and things like that, so the THAAD system. Those are all going to cause increased demand.

Now, the hard part about a defense company is they are metal-benders at the end of the day. And I think where we got a little bit of trouble with some of the defense businesses is you got these really high expectations that demand was coming through higher, but you need to invest in the CapEx and actually physically grow that CapEx or that capacity faster. And people got a little over their skis in terms of how much of that demand was going to come through and how fast they could grow on the other side. So that's where you need to be careful.

Now, my other comment is BAE Systems buys a lot of stuff from other folks when they increase their capacity. And that's where I think things have gotten really interesting, because European loan growth, Japanese loan growth, Canadian loan growth, US loan growth, industrial demand, rail volumes are actually up. Show of hands. Anybody, the last time they heard that rail volumes were actually up more than GDP-type levels?

But you're actually seeing rail or intermodal up as well, even though diesel prices are actually higher. I think a lot of what you're seeing right now is this increase in demand coming across the board as a result of that-- the utility spending, and things like that. So I do think defense is an interesting area. You just need to be picky and choosy about, number one, opportunistically, what else is out there, and then number two, what's priced in how punchy the expectations are versus what they can deliver overall.

OK. Walk back. We'll go to health care. So health care is another area that we used to have a larger exposure to. We do like the large pharma space. I do think there's some interesting things in there. Very decent longer term growers with headroom for continued growth. A lot of the large pharma companies do have LOE issues-- loss of exclusivity issues-- and certain ones are dealing with that better than others, have more of a portfolio-based approach, and they have other shots on goal. AstraZeneca being one of those that grows a little bit faster and better, versus some of them are facing a little bit more of a problem. They're going to have to either buy their way out of it, and things like that.

So I do think you get consistent growth. But again, this comes back to the opportunity set and the opportunity cost. If AstraZeneca is going to grow low double-digits or a Merck or some of these other pharma companies growing at a low double-digit rate, or high singles, I can go buy some of these other companies that simply are growing faster. So you can get to that high single, low double-digit threshold and have reasonable valuation and good dividend yield, but there's just other things, from an opportunity set, that are just faster. And I think that's the issue.

The last one is just the med device side. And I think you could probably comment more. I think those are fantastic businesses. I think AI is actually enabling some of those, and it was before even the large language model. Things were machine learning in other parts, like hardware/wetware integration. So that's all fine. You just need to be picky and choosy about the multiples that you're paying and what the expectations for growth is. And that's where the higher interest rate environment kind of makes us a little bit nervous. I know there's just fantastic businesses that will compound, and they will be better 10 years from now than they are now.

MARK ATKINSON: Danielle, any views on pharma or precious metals?

[LAUGHTER]

BRIAN KERSMANC: Oh, sorry.

DANIELLE MENICHELLA: Do you want to do precious metals?

BRIAN KERSMANC: I can come back to that, if you want.

DANIELLE MENICHELLA: No, I mean, I think the question is getting at something underlying that's been bubbling up for some time-- that historically, there have been these giant businesses that have bumped along at a normal low single-digit growth rate without having any of sustainable competitive advantage or market leadership. And the markets are really changing. And part of that are structural demand drivers that are being driven by secular shifts. Part of it is some of the uprising we've seen in geopolitics globally. And some of it, I think, is being advanced by AI initiatives that are making these businesses better quality than they were in the past.

We don't have exposure in this portfolio to the defense space. We have waited for a valuation opportunity in some other defense businesses, so we recently bought Kongsberg and Rheinmetall in other portfolios. It hasn't bubbled up to the high conviction, 20-stock version of our greatest conviction in global exposures. In terms of pharmaceuticals, we've recently bought other businesses, other pharmaceuticals in other portfolios. We do believe that GLP-1s are completely changing the world and changing how we get health. We also believe in a lot of the AI changes that will help deliver personalized medicine.

But this, again, is all very early and we think that these will take time to play out. We have not bought any metals or mining businesses. Historically, we would not say that those fit our six criteria as a growth investor, that really focuses on leadership, the ability to make your own weather, and competitive advantage. But I think there are secular shifts that are driving certain of those metals and mining businesses going forward, and it's certainly something that we'll be looking at. But historically, that has not been the place that we think meets our criteria.

MARK ATKINSON: OK. We've got a couple more questions here to get through online. Any more in the room? Gentleman over there, if you mind just waiting for the microphone?

AUDIENCE: At the moment, everyone is very happy and everything is growth, and let's get earnings growth and all the rest of it. I don't want to be a pessimist, but the Wall Street crash occurred in October. I don't know how many years ago it was. And October is a very volatile and difficult month, and you've got the midterm elections in America and all the rest. Now, if there was a huge crash and interest rates were going up and up and up, which they're beginning to do, where would you put your money, and would WTW decide on the liquidity and the gearing?

At the moment you have a little cash and a lot of gearing, and when markets crash, they tend to crash very quickly indeed. It takes 10 years for a market to rise and six months to crash. So if there was a crash, personally-- and this is just me, personally-- I'm actually taking money out of equities and I'm putting it into things like gold and gold shares-- Anglo American, for sake of example. But what I don't understand is while the markets have been very volatile just recently, gold has been going down, silver has been going down, property has been going down. Farmland, of course, is slightly exceptional, but it's difficult to buy. So what is happening? Why is all this happening, and why is gold not going up? Because I've bought it.

[LAUGHTER]

CRAIG BAKER: Might be the answer.

[LAUGHTER]

Well, the guys down here might have a view on why your specific shares are doing as they are.

[LAUGHTER]

You mentioned we've got very little cash and a lot of gearing. Just to be clear, we've got a net gearing of about 6.5% of the portfolio. We've got a beta on the portfolio of about 0.93. So we've actually got a gearing-adjusted beta below that of the market, just to be clear. So we've obviously focused today on two of the managers that have got a lot in AI and technology and the like, because we wanted to make this point that we have got lots of that stuff.

It's just the fact that the index has gotten more, and that's been partly the story. But we've got a number of managers in the portfolio that are far more negative than two of the managers that you've just heard from. But overall, we haven't got a particularly geared portfolio at all relative to the market, is one point that I would make.

AUDIENCE: What's the cash percentage at the moment? What's your cash liquidity?

CRAIG BAKER: It's about 2.5%. So we're about-- yeah.

AUDIENCE: So if there was a crash, how quickly could you press the button to sell, and how quickly would the manager start selling?

CRAIG BAKER: Just to be clear, we would remain fairly fully invested. So that is not designed to be an absolute return fund in the sense that it will move between cash and equities. We don't think many people are very good at that binary decision of equities versus cash. Obviously, we have that as part of our gearing decision and we reflect it. But we size that appropriately.

We want stock selection to drive the returns in this portfolio. So if the equity market crashed, I suspect we would outperform it, but we would still be producing negative returns, because we are a fully invested equity portfolio. Now, cash might creep up above 2.5%, but you're not going to have 20% in cash in the portfolio. The managers can own some cash, but they would typically remain fairly fully invested.

AUDIENCE: Thank you.

MARK ATKINSON: Brian, Danielle?

BRIAN KERSMANC: Gold comments?

MARK ATKINSON: If you got any.

BRIAN KERSMANC: Yeah, sure. Let's go. We're not currently invested in gold. We're active equity managers, and we don't necessarily have the gold producers either right now. There have actually been some decent stocks over the more recent periods. I think one of the things that we always talk about from a GQG perspective is when you go to the party, you want to know who you're going to the party with. And I think one of the issues that gold has had a little bit more recently is a lot of the folks that were buying things like cryptocurrency, a lot of people that are buying a lot of these higher-priced-- gold was meant to be the diversifier for that, but you were buying it at the same time.

So when those types of things come under pressure, so does gold, because they're selling everything at the same time, or reallocating in that way. I think gold has actually gotten a little bit over-bid, in that sense. As an asset, it is a store of value. I think it will continue to be. I think it's fine in a longer-term basis, but it might have just gotten a little bit of ahead of itself. And you have, again, some of the speculative buyers, or just buyers that have exposures to other things that might be a little more volatile also affecting on the gold side as well. So that's what I would say on the gold side of things.

MARK ATKINSON: Great. Well, questions in the room? Yes.

AUDIENCE: Would you ever invest in crypto?

BRIAN KERSMANC: So again, from an equity perspective, there's nothing I can invest in on the crypto side of things-- large cap equity. I do think it's a very interesting technology. We've had several conversations in the last two days about tokenization. You think about what that means for global and emerging markets, how much faster you can do clearing and things like that, and execution and trading. That becomes interesting.

I think it's early for me to make a call in terms of who is going to be the ultimate beneficiary there, but I do think that wave is coming. You've already seen faster transactions, T+1 trading in the US, and that drive for efficiency. There is a margin there and somebody can capture it. And I think, ultimately, blockchain technology will help with that. And I think you're seeing that happen. Sorry, that's my non-answer answer for you.

DANIELLE MENICHELLA: But if there was an area where I was most concerned about regulation and regulatory risk and changes, that would be an area.

MARK ATKINSON: OK. So we've got if we want to give everyone time for a sandwich and a chat, and our US colleagues here have got to get to the airport, so if I'll just ask this one last question, which is coming online. Craig, too, maybe this is one for you. Well, it is one for you. Why do you only have active managers in your lineup rather than using, say, factor-driven ETFs?

CRAIG BAKER: Well, I think if we refer back to Dean's answer to one of the questions earlier around, the board set us specific parameters that we work within, and the parameters we have at the moment are 100% in public equities-- you can have a bit of cash, but 100% in public equities, fully invested portfolio, hence the answer to the crypto point. It's similarly here. The idea is we're investing in active managers that we think can produce significant returns.

We don't generally think it's appropriate for you to be paying an OCR, albeit very low on this. It's still higher than if you went passive. So to fill that with passive management doesn't make sense. All that said, one of the things I refer to is these extra levers in the risk management side for a very small part of the portfolio. Could we, in time, ever consider using ETFs to help with some of that alongside individual stocks? Yes, potentially. But at the moment, we do it all through the active managers.

MARK ATKINSON: OK. Well, unless there's anyone else in the room? But you can always ask your questions next door. I think we'll wrap up there. Say thank you to the speakers and join us for a sandwich next door.

[APPLAUSE]