Download a PDF copy
[00:00] Rob Campbell: Today on the Art of Boring, Karan Phadke, Portfolio Manager on our global small cap strategy, discusses whether small caps have been spared some of the worst impacts associated with trade policy announcements, what he's looking to see from management teams through a more uncertain period, and stick around to hear Karan talk through how we approach innovation at Mawer in general and some of the specific process improvements emanating from our global small cap team.
Disclaimer: This podcast is for informational purposes only. Information relating to investment approaches or individual investments should not be construed as advice or endorsement. Any views expressed in this podcast are based upon the information available at the time and are subject to change.
[00:00:26] Rob Campbell: For those listeners who tune into every episode, first of all, thank you. You'll know that we've had a number of folks from our research team on the podcast over the past few months to talk about tariffs, geopolitics, big market movements that have come along with that. But we haven't yet had a small cap perspective, which is why I'm so excited that you're here. So welcome.
[00:00:46] Karan Phadke: Glad to be here.
[00:00:49] Rob Campbell: Very good. Well, what has the small cap experience been like so far in 2025?
[00:00:54] Karan Phadke: It's been pretty positive, I would say. I think one of the big things that is important to keep in mind with smaller companies is that they're typically more local in terms of their exposure. So ironically in a more volatile environment like now, it's not a bad place to be. We obviously look at risk-reward from a bottom-up standpoint, and then we also look at risk top-down. But on a bottom-up basis, it's been fairly limited exposure, at least on the international side to some of these tariffs because production is local, sales are local, as is often the case. Which is not the case with large caps that are global in nature.
[00:01:38] Rob Campbell: Okay. So there's been a measure of installation just given the nature of the types of businesses and where they operate. But coming into the conversation, I would've assumed maybe that being smaller companies, they're more at the whims of some of the more second or third order impacts of trade policy and their impacts on the economy. Has that been the case too, or really it's been what you talked about that's really shined so far?
[00:02:04] Karan Phadke: In my experience, I think the initial phases when there's news or volatility, the stocks trade on more of the first order impact, and then over time as things play out, then the market starts to assess the second and third order impacts. So it could be the case where we're yet to see what some of these unintended implications are. That being said, it is something that we do think about. So on a bottom-up basis, we'll look at companies—we'll see where their revenue is coming from, whether production is local. And then we'll also think about who their customers are, or where is the customer's revenue coming from as you go across the value chain. That's sometimes where you see second order effects as the customers pull back. If they're global automotive companies as an example, you will eventually feel that. So that's a little bit of the work we try to do—considering how this might go through the value chain.
But it's difficult to tell, I think, because it seems like every day there's a reversal in the policy. But I think one thing that is certain is that it does create a lot of uncertainty when it comes to business decisions. And that translates to maybe deferrals or delays in large investment programs or CapEx programs and maybe more sluggish M&A because the companies and the management teams are sort of unsure of how this is going to play out, which leads to almost a reflexive nature and that volatility leads to a pullback in investment and uncertainty, which in turn leads to more volatility. So it feeds on itself almost.
[00:03:52] Rob Campbell: Okay. And you mentioned looking internationally, but I'm curious whether if you're looking at U.S. companies, or companies outside of North America. This is a global portfolio. What has the experience been different for those U.S. companies or largely the same?
[00:04:09] Karan Phadke: Well, I think if we take a step back, one of the most notable forces in the market over the past decade, but especially in the last five years since COVID has been the exceptional outperformance of U.S. companies versus European ones, and it's even more pronounced in the small cap space. So if we look at through to the end of 2024, at least for U.S. small caps, I think they had beaten international small caps as a group by over 800 basis points over a five-year period, which is a massive outperformance.
[00:04:42] Rob Campbell: Wow, you're talking in an annualized number too? Not just...
[00:04:45] Karan Phadke: It is annualized, and that spread was still there. US. .small caps, even if you go back pre-COVID, had outperformed international, but not by nearly as much in terms of margin. So that was a huge tailwind for the U.S. companies just because the economic environment growth was good.
Taxes were low—great for domestic companies.
Since the Tariff News and “Liberation Day”, I guess, there's been a lot more volatility, and at least year to date that relative gap amongst the regions has tightened. So it's not to say that we've reversed all of it and international small caps are trading at the same valuations as U.S., or have reverted to the mean—not by a long shot, but it's just that persistent headwind of U.S. companies constantly outperforming international ones, at least on stock price basis, that has petered out and there have been more dispersion between them because of the tariff noise.
And a lot of it, as I mentioned, is because of this tariff noise. But again, that could reverse tomorrow if Donald Trump lowers taxes and says, "Oh, the tariffs were just a misnomer," and that could be pretty positive.
But the counterpoint to that is the market, and in particular, companies, need to believe that that's the new policy that will last for several years before they start making investment decisions again and have better visibility on what the administration's going to do.
[00:06:23] Rob Campbell: Some similarities to podcast hosts and production teams. I know we've really tightened the time to record versus time to release the podcast based on things changing out there in the world. You talked about those reversals to try and minimize that. Hey, we might've said something on a Tuesday and it's different by Thursday.
Can you go back though to just the work that you and the global small cap team have done looking across the portfolio? Have there been pockets of the portfolio you've been more worried about, or that you felt more strongly about any adjustments that you've made in that regard?
[00:07:01] Karan Phadke: I think when we look across the portfolio, we would guess about maybe 10 to 15% in weight is exposed in some form to tariffs. Ultimately, everything is exposed indirectly, but when we talk about more direct exposure, it's perhaps 10 to 15%. And the way we've tackled analyzing this is really company by company.
So as an example, we owned and we still do own Hikma, which is an injectable generic company. So they make complex generics that are administered through injection in hospital settings. This company's biggest market is in the U.S., and that's where reimbursement is favorable for firms like this. And as Donald Trump started talking about some of this pharmaceutical stuff, which was several weeks or months ago, we started looking more closely before any tariffs or any price changes were announced in that market. And we really did a deep dive to understand, well, one, what portion of their revenue comes from the U.S. How much of it is imported versus produced locally because there's no tariffs on products that are produced locally. And then for the stuff that's imported, in this case, the active pharmaceutical ingredient, where is it imported from? What proportion comes from India, for example, where most APIs or pharmaceutical ingredients are coming from versus China? What they can do about that, whether they can shift supply one way or the other.
And then also what the total cost impact would be based on the input prices of this item, as well as how long it would take if those tariffs were to kick in on that specific cost item, how long it would take to pass that through in prices. So what would be the sort of lag effect in the supply chain from doing something like that?
So that's sort of an example of the analysis we would be doing, both using customer conversations, competitive conversations, but also speaking with management and keeping in mind that a lot of this is also understanding what management is doing to mitigate this. So are they thinking ahead in terms of what are the alternatives and the mitigation strategies and then weighing that against the valuation in the portfolio and the weight in the portfolio.
And in this case, given some of these factors, we trimmed that position. We pulled it back a little bit because we felt that the risk-reward was not there relative to what it was before. This is proactive trading ahead of—I think just two days ago or three days ago, Donald Trump talked about drug pricing again, which came under the spotlight. So this is how we are trying to be proactive.
And one of the benefits perhaps of all this macro noise since 2020 has been that this “muscle” has been practiced a lot. So as you can remember, there was sort of duration and valuation issues at one point. Then there was supply chain and inflation—a lot of macro push and pull. And for each of those, we've worked through it and I think we're getting better and better at just exercising that muscle and trying to be ahead of the curve in terms of analyzing exposures in the portfolio.
[00:10:11] Rob Campbell: I want to ask you about two, or what I thought I heard were two contrasting statements, which was, hey, we're speaking to management to understand what they're doing during this more difficult environment versus what I thought you said earlier, that most management teams are doing kind of nothing. They're deferring, they're waiting. Do you view those as different things or are there specific actions that you're looking to see from management teams that you like?
[00:10:38] Karan Phadke: That is a good point. I think it comes back to what separates maybe the best- in-class managers versus more of the average managers. So in my experience, the best-in-class managers are very proactive in their actions. So they'll be looking ahead and being very active about thinking about resilience in the supply chain, dual sourcing, having conversations early with customers to perhaps change pricing terms and structures. More reactive managements will see some of these costs and noise hit the P&L and based on that, they will then address it. So a little bit of that contrast, I guess, is really trying to evaluate managers as well and their execution. Because what you really want is those best-in- class managers who are mitigating some of that volatility and managing the P&L before it hits the numbers.
[00:11:39] Rob Campbell: Okay. Got it. Shifting gears Karan, I'm always interested to hear what's new process-wise within the team. Anything exciting to share with our listeners?
[00:11:51] Karan Phadke: There's been a bunch of tweaks, around the process that we've been experimenting with. If we take a step back, I think two important elements around process come to mind. So one is that a third of our compensation as an example, is driven by improvement to the process, and then a third is related to collaboration, and a third is performance. So improving the process is a really important part of adapting and evolving the investment philosophy.
So that's one piece. And then the second piece is the model we usually follow here at Mawer on improving the process is what we call the evolutionary model, the ant model. And what that means is that there's a bunch of investment classes or teams within the firm and the platform, but each one has some autonomy on how they implement the process. And if they find something really helpful, they will put it in place and other teams will see that, and then they will copy it if it works, and not copy if it doesn't work.
So it's a very bottom-up organic process in which things evolve and it's based on teams experimenting and coming up with new ways of doing things. So that's just an overview.
[00:12:59] Rob Campbell: Okay, so there's feedback on how good an idea might be based on how quickly it propagates or not across the platform.
[00:13:05] Karan Phadke: Exactly—feedback from other people like, "Hey, this is a great idea. We're going to do it too." Or "This is not such a good idea for us, so we're not going to do it." And the good ideas then propagate throughout the team and the bad ideas die. It's almost a market-based mechanism for determining which improvements around process make sense.
So that's sort of an overview of how the process evolves over time. In terms of some of the experiments we've done specifically in global small cap, I'd say three come to mind.
So one would be just systematically tracking our investment theses every quarter. So, we're long-term investors. We're partnering with these management teams often for eight to ten years, but we also need to make sure that along the way the investment thesis is still on track, and the management team's executing. So we always did monitor quarterly earnings and operating metrics, but we're just being a bit more systematic around it.
We have a dashboard for companies, and every quarter we update that, not just with financial figures, but also with operating metrics. And we use that just to get a lay of the land more systematically in a structured way on how business momentum is progressing around these companies. And that has helped us understand if a thesis is playing out and also just around timing.
As you would know, 2024 was a tough year because it was a very momentum-driven market. So that doesn't mean that you ignore decisions in the short run, but you do have to be mindful of just the momentum in the market. So that's one.
[00:14:36] The second one would be doing biweekly case studies. These are every two weeks. The team sits down over lunch and we go over past mistakes and wins, and not just in global small cap, but we have other portfolio managers join and present too. So we'll have Paul Moroz, for example, who does private equity. He'll join and go over some of his wins and losses and really that's helping both build the culture and collaborate with other teams but also broaden our horizon as it relates to thinking about what can work and what can't work. Using the case study method is the second one.
[00:15:14] Rob Campbell: I love these case studies because my understanding of how this developed was you guys used to get together just for half an hour a week or something for more social time. Like a coffee time to connect and just catch up. My understanding from John Wilson, who's your co-manager, is that you guys just wanted to talk about stocks and so it just naturally gravitated to going through these case studies during the meeting instead.
[00:15:42] Karan Phadke: Yeah, you find that like a lot of these quote-unquote "social events"—they're great, but everyone here loves stocks, so that just ends up being stock talk. So instead of forcing a social event, it's like, "Hey, we all want to talk about stocks, so might as well just make it a meeting where we talk about stocks."
So that was the second one. Related to that, I guess the third one is one of the things we've really been pushing in global small cap is collaborating a lot more with our teammates in other portfolios. So U.S. mid cap in particular, and then emerging markets, really making sure we speak to our colleagues and look at the ideas that they have that could fit in global small cap.
And it's been a really helpful way to round out the portfolio and add chess pieces to the table in an expedient fashion where there is volatility. So as an example with the U.S. volatility, there's a couple of names that they owned in U.S. mid cap, and we're able to get up to speed really quickly on those and then add them to the portfolio.
So that collaboration has really helped us both balance the portfolio and also round it out and add pieces of optionality in a very efficient manner.
[00:16:59] Rob Campbell: Karan, when you talked about the thesis tracker and just the dashboard, I'm curious on that or elsewhere, where has AI played a role in some of those process improvements?
Whether it's with global small cap or across the research team?
[00:17:17] Karan Phadke: Yeah, I think the AI questions started to come up a lot more in terms of how we're actually using it. As opposed to maybe a year or two ago, clients often wondered what our exposure was to AI, for example, through data centers and semiconductors and things of that nature. Whereas now the questions have shifted more to, well, how are we using these tools, if at all.
And I would say it is a very exciting time. I would say the Deep Research launch—so Deep Research is an agent released by OpenAI a couple of months ago that was a game changer, in my opinion, because it really showcased how AI can be used in the research process.
And fundamentally, if you just look at the investment process, if you take a step back, it's really three pieces in it. So one is collecting relevant information. Two is analyzing and synthesizing it into insights. And then three is to make a decision based on these insights.
So here we follow the "leader decides with input" model. So John would be the lead manager, but to make repeatable value-added decisions, he needs good input and insight. So those are the first two pieces: collecting the information and then synthesizing it into insights. And what these tools allow you to do really on those first two pieces is to just cover more and go deeper. So now you can just look at a lot more companies and you can go a lot deeper with them because you have these agents and tools that are sort of your copilot, helping you do it.
[00:18:36] That being said, it's a bit of a red queen race in that everyone else also improves. So it's not enough for you to use it because everyone will use it eventually. So you do need to evolve and adapt just to stay in the same place.
That being said, I would say there's a couple of unique things that excite me about some of these AI tools that we have at Mawer that hopefully will translate to us learning faster than the market over time.
One is that we have a proprietary database. So we've been collecting management interviews and conducting management interviews for over 20 years and recording all of that in a common database that we call M42. So this is data that is manually collected with senior decision makers, CEOs, chairpeople, CFOs for companies over many years. So there's a lot of insight in there. And we can plug in language models on top.
You can go back 20, 30 years and have a conversation with Mark Leonard at Constellation Software and see how his thoughts have evolved. Like that's not something you can just replicate. So putting a language model on top of that is pretty promising. It can yield a lot of fresh insights.
[00:19:49] Then the second one is we have a strong culture where people are encouraged to push on and improve the process. So several years ago, we put in a lab team that works with the research staff and it's staffed with dedicated developers and statisticians.
As an example, there's five people on that team now who sit side by side with research analysts to build useful products. So this close collaboration, I think is going to be very useful for building some of these tools. Luckily we had the infrastructure in place. We'd already been doing things with the lab, but now it can really be kickstarted.
And I think the analogy that gets used is, when Excel came out many years ago, you could have sort of had just the Excel team and you could have been like, "Hey, I'm going to have a separate team that just does Excel and every time I have a spreadsheet to do, I'll farm it over to them." Or you could have integrated it into your day-to-day process and learned side by side. And we clearly know which model works better.
[00:20:44] So I suspect with AI it'll be the same thing where you're not going to have a separate AI team, it's going to be your team and you're going to be using these tools and partnering with people with different skill sets to make them super effective. It's not going to be a silo on its own. So I think that integration and close collaboration will be a pretty big element of how AI gets used across the organization and white collar work more generally, I would say.
[00:21:15] Rob Campbell: You had mentioned something the other day though that I thought quite struck me. Just on the value of not outsourcing too much of the work.
[00:21:23] Karan Phadke: I think it relates to this point where you can't outsource thinking and you can't outsource decision making. So with that Excel example, sometimes the process—so this thesis tracker is a good example. You could probably automate that and you could have AI go and scrape all of this information and compile it for you. And that's certainly a good idea. But for me, I found in some of these things, I like actually doing it yourself manually, just helps you get closer to the company and understand the numbers in more detail because you spot the discrepancies a lot more readily because you're typing it in yourself. It's deeply inefficient, but at the same time it yields long-term fruit.
So I think that's a big part—figuring out which elements of the process and the tasks that you do are inefficient and not very useful long term. And those should be outsourced and automated. And then figuring which other tests are inefficient but do yield long-term benefit and maybe those you don't outsource. And finding the right balance between those because you don't want to outsource everything because you do lose some knowledge and flavor for what the companies are doing if you're not actually doing the work yourself.
Again, we're doing all the work ourselves. It's just AI is one more tool to help us gather information and synthesize parts of it at scale.
[00:22:53] Rob Campbell: Okay. I want to ask you about one last thing before you leave, which is tell us about a new stock in the portfolio. Something that your process has unearthed and suggested is what we think is a great idea.
[00:23:11] Karan Phadke: Yeah, so one of the companies that we used to own and we followed for some time is a Swedish firm called Avanza, and they're the leading online broker for stocks. So if you're a regular investor, you can go online on their platform and you can buy stocks and other assets. And we quite like this business model. It's a very scalable business model because it's online based. And it also takes years to build a local brand. So once you're in the market, it takes many years to build that reputation and brand with the local consumer. And what this means is that if you go to many different countries, what you find is usually there's only like one to three leading online brokerage platforms.
So it ends up being a fairly consolidated market, and the scale really helps you reduce your costs and pass those savings onto the customer, which is what Avanza has done for many years.
That being said, what's really interesting about it is that despite being a consolidated market and they're a market leader, there's still a pretty long runway because most of the market is still served by the traditional legacy banks. So the banks tend to be more expensive. They tend to have a slimmer offering and not as good of a customer experience. So the online pure player brokers over time just continue to take market share from the banks. So that provides it with a pretty good growth runway, which we quite like.
And then for this company specifically, the founder remains involved on the board, and it has a pretty decent valuation. And we use some of the volatility in the stock market of late to build a position.
And it allowed us to collaborate more with other teams too. So back to my earlier point on that collaboration item, global equity was able to follow us into the name and we also used the opportunity to learn more about Swiss Quote, which is the same thing, but in Switzerland and our colleagues in international equity own that. So we were able to get up to speed pretty quickly and initiate a small position in that company as well.
It's just a small example of the types of companies that we like and the way that we use the whole platform to cross collaborate and share ideas.
[00:25:20] Rob Campbell: Very good. Well, whether it's process improvements or ideas, yeah, keep going. It's been a pretty great start to the year for the global small cap portfolio and Karan, love having you back on the podcast. Thank you.
