The Capital Cycle Podcast - Barbarous Relics

Episode Date: November 26, 2025

Absolute return potential in Emerging Market stocks. Edward Chancellor talks Laura Fyfe, an Emerging Markets Analyst.For more information, or to access select articles from Marathon’s Global Investment Review publications which accompany this podcast series, please visit www.thecapitalcycle.co.uk Hosted on Acast. See acast.com/privacy for more information.

Hello, this is Edward Chancellor with another episode of the capital cycle podcast. And I've got with me, Laura Fife, who's an analyst on the emerging markets team. Welcome, Laura. Thank you, Edward. I've talked with you and I've talked with your boss, Alex Duffy, about emerging market capital cycles in the past. And what we know is that emerging markets are particularly prone to damaging capital cycles. And I can see from this piece, you've read some research from Sparkline Capital that points out that high-capex firms have unperformed low-cap-ex firms in emerging markets by 5% a year. Well, you might not know is that that research was done by my former analyst at GMO called Kai Wu, who I taught the capital cycle to.

Anyhow, I think, again, we probably have discussed in the past how the disappointing returns from emerging in the decade after 2010, can largely be ascribed to excessive investment in commodities that have a relatively high weighting in emerging that were in turn induced in large part by China's really quite extraordinary real estate investment boom. We also discussed in earlier podcasts how China's chronic investment spending is in the process being curtailed. In this podcast, we're going to discuss some recent addition to the marathon's emerging portfolios and how they appear attractive from a capital cycle perspective relative to the big tech hyperscalers that dominate the US stock market. So first, as you're sort of throat clearing in your piece, you talk about why investors

shouldn't aim to keep up with the Joneses. What do you mean by that? Many listeners probably know the phrase keeping up with the Joneses, which is sort of an exercise in measuring success, not in terms of absolute well-being, but relative to other people and relative to those they see around them. So people might end up spending money or even overreaching to spend money for the sake of status or to avoid feeling left behind. But as capital cycle investors, we actually try to do the opposite of that and identify companies which operate in industry. with decreasing capital intensity and decreasing competitive intensity. So under-spending or under-investment can actually create opportunity for us. And you cite a case of investors getting over-excited about investment prospects.

You cite the Chinese internet giants back in 2020. So the parallel, I suppose, is that companies can be, quote-unquote, rewarded by the market for their spending. So status in this sense would be share price appreciation, which can ultimately lead to bubbles. And the one I think you're alluding to that's in my piece is in 2020 when the Chinese internet giants, Tencent, Alibaba, and May-Tuan were rewarded with 86 yuan of incremental market capitalization for every one they spent on CAPEX. So they got status in return for spending, so to speak. And the market also priced in six-won of market value for every one of sales they generated at the

peak, which implied incremental growth that would replicate or exceed their historically quite high returns, thinking in some quite high expectations. But things didn't turn out well. No, no, they didn't actually. The market's rewarding of that spending actually attracted more spending and more competition and ultimately bid down returns and at negative incremental free cash flow margins, more and more of that growth or continued growth in what the market was focusing on, which was sales, fed the gap sort of between perceived and realized value.

So the kicker is that 60% of their market value had vanished by 2023 versus the 2020 peak. And you think the big tech hypers now playing the same game of keeping up with the Joneses and the market is currently rewarding them for their CAPEX announcement, or at least it has until a couple weeks ago. Perhaps, yes. We are weary of some of the spending commitments we've seen from the hyperscalers. And in my article, I included Alibaba, Alphabet, Amazon, IBM, meta, Microsoft, Oracle, and Tencent, sort of as a group to be analogous to the Joneses in that opening paragraph.

So whereas the Joneses might build mansions, so to speak, the parallel is that these hypers are spending records amount on AI CAPX and, in fact, even spending some of it in space, apparently. Yeah, that's latest announcement. data centers in space. Yes. You think that away from this arms race in U.S. data centers, that there's quite a cohort of emerging companies to more attractive. Yes, and these ones actually enjoy lower or decreasing marginal capital and competitive intensity.

So in my essay, I've compared the three most recent additions to our global emerging markets portfolio as a group to sort of counter the Jones's archetype. And those are China Resources Land, which you spoke to Kai about last time, Milacom International, and Intel Chile, despite these three companies spanning two industries and several countries. What they have in common is that they all enjoy moderating capital requirements as they consolidate what are currently out of favor and asset-heavy industries. So we know what China Resources Land does. And Millicom International and Entel, just very brief descriptions.

So the other two, Milacom and Intel, they're relatively small telecommunications companies, which operate leading market positions in several markets across Latin America. What they have in common is that they're both actually benefiting from the retrenchment of a key player, which is Spain's Telefonica, from their respective mix of markets. So the supply side is shrinking, which is something that often peaks our interest at Marathon. So we've been researching the specifics of this over the past several months. And as the supply side shrinking, you say that the capital cycle for these companies has entered a relatively benign phase. Exactly. So their spending, which is to say their average capital intensity or capital expenditure relative to sales,

has historically been quite high. I think most people would associate high spending with telecom business models. And for those three companies, it's averaged about 20 percent over the three years ended 20. which is twice the hyperscaler Jones's average, which was about 10%. And now this is turning right. Yes, and that was actually my inspiration for the essay. I think it's quite interesting. We're now seeing this more or less invert with the spending that the hypers are putting out. So the hyperscalers capital intensity is actually set to clear the 20% ceiling within a few months as they continue this sort of status seeking exercise.

and their spending is actually set to get to levels that will tower over the dot-com aero Jones is with the big spenders then having only reached sort of high teens levels yeah and so I mean in your footnote you put that the capital spending sales at nortel if anyone ever remembers that but the largest Canadian telecoms equipment company rest in peace lucent norta got to 19% of capex sales lucent got 17% Cisco larger stock briefly, got 17% Intel to 15% capEx sales and IBM to 16%. And people are now making lots of comparisons between what's going on in artificial intelligence and what went on back in the dot-com boom.

And in some ways, obviously the big tech companies are much more grounded than a lot of the dot-coms back in 2000. But our focus is on the capital cycle. and clearly the capital cycle dynamics are probably more negative now than they were in 2000. For the hyperscalers, they're actually projected to clear 24% capital intensity for 2026 and 2027. And the emerging market additions, by contrast, have rationalized their spending to the low double digits. Yes, yes. So their CAPX to sales ratio is set to average about 12% over the next three years. So material improvement from the 20% average we just spoke with.

And actually, Laura, I think this is a case, you've selected three stocks, but this is a general point that one makes about emerging markets now, is that they have entered into a benign phase of the capital cycle. And that actually tends to last for many years when you enter into that phase. Now, go back to your stock picks. You say they're throwing off cash. Yes, so it's actually this lower capital intensity, which helped drive the new additions to our emerging market. markets portfolios, free cash flow margins and yields to double-digit levels on average in the most recent fiscal year, so 2024, which is a huge improvement, like we said, from single-digit or even negative levels in prior years. So this would indicate we're starting to see some of the

green shoots of industry repair. And you also point out that these businesses have rising returns on incremental invested capital. Yes. So return on incremental invested capital provides sort of of a pulse check, if you will, on the efficiency of recent expansionary spending. But something I've looked at in the article is cash return on incremental invested capital. A bit of a mouthful of that. But that takes the idea a bit further by using free cash flow in the numerator. And my idea in looking at that is that cash return on incremental invested capital might actually moderate and improve as capital intensity ramps and eases for the group of hyperscalers and the group of new global emerging markets holdings, respectively.

And, Laurie, in your piece, you show a table that compares the forecast returns on invested capital for your emerging picks compared to the hypers. Can you talk us through that? Sure. So I've compared the two groups across four main metrics, estimated cash return on incremental invest in capital, estimated cost of capital, current free cash flow yields, and implied growth rate into perpetuity. So the average estimated cash return on incremental invested capital over the next three years is 13% for the global emerging markets, abbreviated gem companies, versus just 1%

for the hyperscaler. So sort of validating that hypothesis I just mentioned in response to your last question. Those are at and below my rough estimated cost of capital of. of 13 and 8% for the two groups. And I'll talk with that because actually what's interesting is that having a high cost of capital is actually what discourages firms from throwing the capital away. When you point out that the hyperscale is big tax because their stock prices are so high at the moment, they have a much lower, and also their investment grade. So they have a much lower estimated cost of capital.

Speaking to your point of cost of capital, if we subtract actually the current free cash yield from the average estimated cost of capital for the two groups. We can work out an implied growth rate into perpetuity for the two groups and sort of compare expectations, so to speak. For the gem companies with a 13% cost of capital, their 13% free cash flow yield suggests they won't grow. That's a sum of zero. And the hyperscalers with an 8% cost of capital, less their 2% free cash flow yield suggests they sustain a high single digit growth rate again indefinitely. So that's a profound change for both these groups? Yeah, this would be a material change, especially because historically our new additions haven't earned their cost of

capital. Their average return on invested capital over the past five years has been only 4%, which doesn't comply very favorably to the hypers' world-leading returns averaging almost 20%. So on one hand, the hyperscalers have deep pockets, while the recently initiated companies are more so finding their feet. But I think another interesting point is also that the balance sheets are marginally improving and reducing in strength for the two groups as well. Because if your CAPEX starts exceeding your free cash flow generation, then you either have to issue debt or more equities.

And some of these big debt companies have been Oracle and meta recently been issuing debt. So you say that hypers will have to ramp up revenues if they're going to, earn an unacceptable return. So the market currently is rewarding the hyperscalers for announcing their capital spending. Yep. So today, hyperscalers see about $7 of share price growth for every one of sales. And this is interesting because they've actually reinvested at a negative average cash return on incremental invested capital since 2022. Recall, we're talking at the start of our conversation about the Chinese companies. And they were out. actually less well rewarded with only six won of share price for every one of sales.

The hyperscaders are going to have to ramp up revenues very quickly if they're going to earn an acceptable return. Yes, and they might well meet these high expectations, but it appears that the market's pricing in a degree of certainty that I don't have at the moment. Sparkline, for example, is also skeptical, as we've read. They estimate that AI revenues need to grow about a hundredfold in less than five years to justify the data center buildout, which is ongoing. And this is especially interesting in the context of AI CAPEX having now surpassed the contribution telecom fiber optic, CAPEX had to GDP and the internet boom of 2000. So from investment perspective, sometimes it helps to have low, even no expectations.

Exactly. Investment perspective and in general. That's a stoical point. Yes, yes. I would say that lower implied growth. rates almost mathematically leave more room to be positively surprised. The gem portfolio additions actually attracted less than $2 of market value for every dollar of sales they generated. So in a no improvement scenario, clients would receive even then quite attractive low teens, cash yields. Growth, which I would say is probably easier to achieve from low levels and as supply shrinks, as we've been alluding to, would be essentially a free option.

So let's discuss Antel. Yes, a company that exemplifies these sort of low expectations and low spending idea as well is Antel. It's Chile's telecommunications market leader and the newest addition to the gem portfolio. So as a consequence of operating in a struggling industry structure, it appears price to earn only its sub-cost of capital returns forever. And Marathon's appraisal of the potential for industry repair suggests this. is unlikely. Correct. We believe current industry returns are unsustainably low and as a function of that should improve. Chile is actually one of the world's most fragmented and least profitable mobile markets. It's kind of interesting a quarter century since the internet boom,

capital and competitive intensity appears to be abating across telecoms. So another quite interesting inversion. But what's happening in Chile is that telephonica, which is industry number two after Entel, and which has more than a 25% market share is loss making and accelerating its exit, which was first announced in 2019, but they sort of have a renewed impetus under a new CEO who joined earlier this year. And then on top of that, industry number three, called WOM, is also loss making, and it has private equity owners. So it might not be a permanent participant in the market. By contrast, we have Entel, which is owned by three family business groups. So it should be able to sustain its position for the long term.

And broadly, consolidation to a tighter three, if not two-player market, seems plausible. And then on top of this consolidation potential, 15 years of real pricing declines have eroded industry revenues quite substantially. So to the point that they could grow fourfold just to recover their initial levels, which is a lot less than the 100-fold the hyper-scalers need, but might be more easily achieved. So on balance, we estimate the capital intensity could moderate from a recent 18% to 15% or so and be a key driver of free cash flow margin improvement from low single digit to double digit levels for Intel, having the benefit as well of recent telecom industry repair in Brazil and Guatemala providing precedent. And in conclusion, summarized the attractiveness of these stocks. So at an average 13% current free cash flow yield,

our new holdings are priced to shrink into perpetuity, and they might not. By contrast, the hyperscalers at a 2% free cash flow yield are priced to grow at high single-digit rates against the tides of rising competition and with lower projected cash return on incremental invested capital, again, indefinitely. They are trees which might not grow to the sky, so to speak. So we're staying focused on doing the best we can to generate long-term outperformance for clients just by chipping away at understanding underappreciated capital cycles with moderate, though attractive, absolute expected returns irrespective of the Joneses.

Thanks, Laura. Thank you. Thank you for your time today. I hope you will listen to the next edition of the capital cycle. This communication is provided for information purposes only. Please refer to Marathon's website and the Global Investment Reviews for further information, including important disclosures.