Michael Zeuner (00:02.727)
Hi everyone, this is Michael Zeuner, one of the managing partners at WE Family Offices. Thanks for listening to the Wealth Enterprise briefing. We’re going to pick up the conversation we had on our last podcast with Sam Sudami, our global head of macro, and we’re going to go a little bit deeper into one of the larger questions that’s out there in the summer of 2026, which is: is there a bubble in AI? Is there a bubble in AI spending? And will that spending pay off and lead to the kinds of increases in profits and returns that the market seems to be anticipating given the significant increases in AI and related stocks over the last six to nine months. Sam, welcome back. I think I’d like to start with just reminding our listeners as to you know, really why this AI capital spending boom has been so important as one of the fundamental factors driving the growth of the US and international equity markets. And then we’ll go deeper into what does it mean? How do you put it into historical context? How likely is it to be sustainable? And what what should investors be taking away from it? But Sam, over to you. Give us a little bit of context, please.
sam (01:25.596)
All right, thank you, Michael. Since last year we’ve been talking about how CapEx is a powerful cylinder of economic growth, and it has been responsible for more than half of the economic growth both this year and last year. And this type of investment has really helped stimulate the stock market. We’ve seen profits rise, for instance, with chipmakers, but we’re also starting to see market skepticism.
About the amount of investment and if companies will get a proper return on it. And these are very reasonable questions. So the is particularly these last few weeks, equity markets have become volatile around the issue of return on that investment and its impact on free cash flow. And credit markets and fixed income have also started to question the amount of leverage needed. The question is how effective will this be?
Michael Zeuner (02:21.443)
And and Sam, you know, is part of that discussion and or questioning that’s going on in the market, is part of that also some of the circular trade, as they call it, of AI, where you have companies investing or lending to each other and then companies striking deals for revenue with each other. How does how does that discussion fit in here?
sam (02:43.238)
So that is part of the worry that the money f just rotates around the sector. But keep in mind that these companies are the real leaders. And the money does have to flow by its very nature. So for instance, NVIDIA is the company that makes the exceptionally high-level chips. It has to flow through them. You do have a handful of these hyperscaler companies, you know, like the Amazons and Googles. And these
five hyperscalers, the investment by them have gone from two hundred and fifty billion to a trillion dollars in just three years, from twenty twenty-four to twenty twenty-seven. And it is the speed of that ramp up which is also making markets quite nervous.
Michael Zeuner (03:30.322)
Well, yeah, Sam, and I think it’s quite incredible. I mean, as we are recording, we’ve seen in the last two or three days, you take a a market which is being driven by AI spending, which is Korea, right? Which is more than doubled. The Korean stock market has more than doubled in the last year. And yet it had swings of up twenty five percent and down, you know, plus or minus twenty percent in the course of forty eight hours, which is extraordinary.
for an equity market which has the r depth and size of of Korea’s capital market. So it seems to me that there’s, you know, you you have the same thing by the way in the US with the chip stocks, right? One day they’re up big, the next day they sell off, and the next day they’re back. And so, you know, a lot of that, you and I talk about this all the time, a lot of that is sentiment and emotion.
You are not a sentiment driven investor, nor are we as a firm, right? We’re fundamental long term investors, but how do you make sense of this volatility through a fundamentalist’s sort of perspective?
sam (04:42.62)
So it is the scale that is hard for people to get their hands around. So for instance, for chipmakers, their earnings went up 100% in a year. And that’s why their stocks went up 100%. You mentioned the two Korean companies. SK Heinex, the actual profit growth in one year was over 550%.
The case of Samsung, their operating profits went up 1100% in a year. It’s not just the sentiment, but they are delivering on their earnings, but it’s the sheer size of these numbers, which are really hard for people to get their hands around. So that adds to it. Of course, in the Korean market, there are other issues about leveraged ETFs. There’s been retail money pouring in, there’s been a massive increase in margin loan.
Which only accelerates the volatility. So you certainly have all of this piling in factor on top of it. But the underlying is that the earnings are growing because the investment spending has been enormous.
Michael Zeuner (05:52.884)
Okay, so look forward now, Sam. Try to look through some of the emotion, the massive amounts of capital spending, right, going in, the circular trades of these companies investing in and financing each other, the leverage that is starting to be put on some of these companies t to sustain this cash flow. What do you see? What are you looking for to guide your view as to whether
all of this will pay off positively for investors.
sam (06:28.998)
So right now the the hyperscalers, and these are the companies that are really doing the investing. They control, it’s estimated about 80% of the computing power that drives AI. And they think it’s important to really get up to scale very quickly and see the risk of underinvesting is more than the risk of overinvesting. But what we a year ago, the market was concerned that it was being that it was not going to be monetized.
But what we’ve seen now, one year later, is extremely high sales growth. You’re seeing, like, you know, some of these major companies that reported this week are sales that have increased 80% year over year in their cloud businesses. And they have seen sales grow very rapidly. They have a large pipeline of clients. That is the important thing to see. They have to invest heavily to convert these prospects.
into actual paying clients. And it’s paid off over this last one year. But the amount of money is incredibly high. You’re talking about companies that have gone from very light assets to more heavy investments. So what that has done is it started to deplete the free cash flow, particularly in 2026 and 2027, because they’re investing so much. But the expectation by the market or consensus expectation
Is that free cash flow will return to a peak in 2028 after they’ve finished scaling up so much? But it is the nature of these types of innovative investments that require a lot of cash.
Michael Zeuner (08:10.011)
So, you know, you’ve talked in the past about other examples through history where we see these, you know, huge cycles of investment which then lead to returns not maybe to every company that participated, but to the sector overall, whether I guess it was railroads or the internet or electricity. I guess you could even go back to the industrialization era and the cotton.
The cotton gin, but but tell us Sam how you put this in historical context.
sam (08:44.544)
So right now we are investing a lot in AI. It is estimated to be 1.5% of GDP. But now let’s put that into a historical context. Railroads that you mentioned in the mid to late 1800s was about 4.5% of GDP at its peak. Railroads were also a 60-year build-out cycle.
And that was instrumental in weaving America together as a country. And then there’s the auto infrastructure. Back in the 1920s, it was 2.1% of GDP. That’s the estimate. Bigger than what you’re seeing in AI today. And then in more recent times, US telecom investment that wired America to the internet, along with the information and communication technology hardware in the 1990s peaked at 3%.
Now we still use all of that technology today. You know, we still use cars, airports.
Michael Zeuner (09:44.606)
So what’s inter what’s interesting, Sam, is that on an absolute dollar perspective, we’ve never seen numbers like are being invested, right, before. But that is because the economy is so much larger than what it was during some of these other peaks. But on a relative basis, the amount of capital being invested relative to GDP is I wouldn’t say modest, but it certainly has not yet approached some of the peaks. I guess the other
The other thing that to keep in mind is that when you look historically at those other investment booms, the payoff was not just to the companies making the investments themselves, but it was also to the economy more broadly, right? Whether it was bringing in efficiency, productivity, margin expansion, the opportunity for new companies to emerge, right, in in a changed environment, e economic environment. Presumably that’s also
what we should expect to see. Maybe we’re in the early days of that because we’re still early in the investment boom, but but do you see that cycle repeating itself more broadly beyond these companies?
sam (10:54.204)
So, a primary benefit in investing in technology is to increase productivity. And in post-World War II, we have seen how investment in innovation, for example, electronics in the 1960s, PCs in the 1980s, and Internet of the 1990s, which was part of the third industrial revolution, sig led to significantly higher US productivity. And the economic
Benefits of increased productivity was higher economic growth with lower inflation. That was positive for both equities and fixed income. You mentioned profit margins. We can see how profit margins today are significantly higher than they were in the past, largely because of those productivity increases. So I would expect to see the same type of payout as AI becomes adapted and adopted by companies.
Where they learn how to use it better to improve their own productivity and lower their costs.
Michael Zeuner (11:56.5)
Okay. Well, Sam, I think that’s all very clear. I appreciate the deeper dive and we’ll look forward to talking again soon.
sam (12:05.178)
Thank you, Michael.
