Fund Manager Reveals Best Stocks During ‘Bubble’ (Uncut) 04-30-2025

Why Big Tech Is Bleeding: Fund Manager Reveals Best Stocks During ‘Bubble’ | Stephen Yiu

NVIDIA’s product GPU, as everyone would be familiar with, that is very mission critical in terms of where this AI war is heading, right? Who’s going to win this AI race between the US or China, or maybe somewhere in between would be Europe, somewhere there. NVIDIA’s mission critical. From it could potentially weaponize NVIDIA GPU in terms of who is going to get NVIDIA GPU.

It’s a bit like your IO, like who’s your friend. The Nasdaq index is down roughly 12% year-to-date as of the 24th of April. The global market continues to navigate uncertainty.

Tech stocks have let the charge downward. What is next for the tech industry, the tech sector? Are we seeing the beginning of another dot-com bubble burst, or is this the beginning of a V-shaped recovery to come? We’ll find out with our next guest, Steven Yu, who is a CIO, Chief Investment Officer, and co-founder of the Blue Whale Growth Fund, a fund with $1.5 billion in assets under management. The fund has made significant portfolio adjustments recently, divesting from Meta and Microsoft while increasing its stake in NVIDIA.

We’ll find out why. Steven, good to have you back on the show. Welcome back.

Yeah, thank you, David. Like I alluded to in the introduction, you’ve made some recent changes to your portfolio allocation. We’ll talk about that.

First, let’s talk about the broader tech industry and the tech landscape. Like I mentioned, the Nasdaq has not had a very good year. Stocks overall have not had a – the stock sector overall has not had a very good year, year-to-date.

How are you adjusting your overall investment strategy in light of heightened global uncertainty? Let’s put it that way. I think on a very high level, that in the context of a global slowdown and obviously there’s uncertainty with Trump’s tariff, there’s a bit of a reset in valuation in terms of maybe technology stocks, if you’re referring to the Nasdaq. And if you’re looking at the valuation going into this year, I mean, it was trading at a very high level relative to the rest of the world stock market.

So I think at the moment, we’re going through a bit of a journey of an indiscriminate sell-off or risk-off that people just want to maybe move away from American companies, whether you are technology companies or in general, and probably moving some of this cash into European equities or maybe Asian equities in time. But I think what is interesting going forward is we’re probably going to start seeing some differentiation among tech companies throughout this earnings season in the next coming weeks. And obviously Alphabet is going to report tonight and some others to follow as well.

So I think that is probably going to differentiate the performance going forward from here. People, investors overall have been a little more defensive now. Some funds are shifting their stance to a more defensive allocation.

Within the tech space, is there such a thing as more defensive versus more risk on, meaning some stocks would be higher beta than others within the tech space that you would consider? I think that is a fair comment, because on a very high level, when you look at the MAC-7, I think all of them, including NVIDIA that we own, the only stock that we have in the MAC-7, have all gone down quite a lot year to date. But when you look at stocks like, for example, Netflix or maybe some others, I mean, obviously you can consider them as tech stock or maybe consumer discretionary. I mean, they have actually done quite well on the back of the recent kind of uncertainties.

So what we are going to expect is there’s going to be more differentiation among tech companies. And I think companies that probably have attracted quite a lot of capital going into the shares over the last couple of years, especially when you have very big companies in the market space. And we would have a lot of investors who would have invested into ETF.

Let’s say you bought the S&P, you bought the NASDAQ, and now you decide that I want to reduce my exposure to US equities or tech equities, then you’re going to sell stocks. And of course, the big companies at the top would get sold a lot more on a relative basis, just because a lot of money has gone into them so that they become a very big part of the market, which has been one of the concerns over the last 18 months or so. Do you have a preference for size of the company right now? We do still predominantly consider ourselves as bottom up stock pickers.

So we invest into companies rather than a particular sector. And anything that we think is one of a kind company, that we want to go into them. So we don’t actually have a cut off.

I mean, I would say that the smallest company we have in the fund is probably over $10 billion, but we’re very happy to invest into very large companies at the same time. Are there companies within the tech space that you would consider more value versus growth? And how would you make that evaluation? This is going to be very controversial, isn’t it? Because if I try to maybe list out Nvidia, which I think only a few days ago that it was trading at about 20 times earnings, one year forward. And would you consider Nvidia as a value stock based on 20 times earnings? I would certainly think so, right? If you believe that AI spending is going to continue into the coming years, which is our view.

But of course, the market is slightly disagreeing with that. That’s why Nvidia has been derated quite significantly, I would say, considering the level of growth and considering the level of free cash flow they’re going to deliver this year, or even the revenue they’re going to deliver this year. I mean, that number is still very high in absolute terms.

So, of course, I mean, that to us is actually probably much more in the value camp. I think on the other side, which is probably quite accepted that we do equally have a large exposure to is semiconductor equipment companies. So, Lam Research is a company that we owned alongside Applied Materials or ASML, that they are all basically part of this silicon sovereignty that we are going to have more foundries to be built outside of Taiwan to do risk away from Asia-Pacific.

And of course, if you look at the valuation of Lam Research, for example, it’s very cheap. I mean, it’s definitely considered as one of those value names. But the reason being, I think historically, is because the end market is typically a bit more cyclical because it’s exposed to consumer discretionary spending through buying smartphones, PCs, or maybe some autos.

But what we think is going to happen going forward is with the support from governments or institutions that we want to have silicon sovereignty, so that there’s going to be a bit of support from tax subsidy or chipsets or equivalent in the EU to provide a bit of a demand to buy more of these equipments to have the capabilities outside of Taiwan. So, to us, I think the valuation is actually looking very attractive if you think that trend is going to continue. Okay, I want to come back to AI stocks in just a minute, Stephen, but let’s take a look at my screen right now.

This is the Nasdaq index. At its trough first week of April, the Nasdaq declined roughly 25% from its high. Now, when that happened in early April, I remember people were writing about and talking about the possibility of this being a dot-com bubble 2.0, right? I think those fears were in the marketplace at that time.

Did that thought ever occur to you in the beginning of April when Trump announced Liberation Day, reciprocal tariffs? So, from our perspective, we’re fairly selective in terms of what companies that we invest into. And of course, within a tech domain, we are predominantly, I would say, in the AI infrastructure part, which is through NVIDIA, Broadcom, and some semiconductor equipment companies. And if you say to me that, okay, whether we are seeing a bit of a bubble territory for the MAC-6, which is outside of NVIDIA, then I think I probably agree with that just because they are burning a lot of their free cash flow to invest into AI, which then have become the revenue of NVIDIA or Broadcom.

And hence, I think the decision for us to exit some of this stock quite recently would also marry with that. But of course, I think ultimately, I think the difference in today’s context versus a tech kind of the bubble period more than 30 years ago now is that some of this big tech company in the digital world has become very much entrenched into our day-to-day. Like, for example, we’re using Zoom to speak today.

We would be using Microsoft for our day-to-day work, and then we would probably be shopping on Amazon or watching Netflix and doing searching on Google or CheckGBT. We’re just a lot more dependent in terms of where we spend our time and money. So then if you think back about a tech bubble at the time, I mean, most of the technology company, including Amazon, was not profitable.

And at the same time, it was not well entrenched into our day-to-day. So I think that’s a bit of a subtle difference, which is very important too, that if you think that we continue are going to use a lot of these services, then I think they’re here to stay. Of course, then the question is about valuation rather than the business model or the relevance of that offering.

But the tech sector overall, it’s not in such an overvalued territory that we’re about to get some bubble burst. Going back to my earlier question, is it? I think I would agree with you. I just mentioned that NVIDIA is trading at 20 times.

I think if you look at some other Maxxix company, they’re probably on average about 25 times earnings for this coming year, which is, yes, you’re right. In absolute terms, it’s definitely not in the public territory, right? If you’re looking at, I don’t know, the likes of Tesla or maybe Palantir, which we don’t own, I mean, they would be trading at a much higher multiple going forward as well. So then I think when you look at what some of these companies were trading at during the tech bubble, I mean, I would use the example for Cisco just before the tech bubble went bust in March 2000.

A lot of people have compared that versus NVIDIA. Cisco was trading at 130 times earnings in March 2000 before the tech bubble went bust. Of course, if the earnings haven’t come through as quickly and when you’re trading at over 100 times earnings, I mean, it’s a lot easier for that to get derated quite quickly versus a company trading at 20 to 25 times earnings.

Because ultimately, that’s a flaw to that valuation, right? It’s not going to be below a certain level. Are there tech stocks that are more resilient to tariffs than others or types of stocks? Yeah, I think this is a very interesting debate that we are having internally on an ongoing basis. I think from a tariff perspective, because tariffs is imposed on physical goods so far, I mean, it’s not in a digital world yet, you would think that most tech companies are relatively immune, especially if they’re not making physical product, right? If you’re looking at the Mac, maybe Mac 7, with the exception of NVIDIA and Apple.

But I think what is kind of happening, which would relate to the tech company in itself, is what is going to happen on the back of all this tariff or retaliation or the slowdown in global trade. We are going to talk about the scenario of a global slowdown or maybe a US recession. And of course, then when you look at businesses like Alphabet or Matter, which we have recently exited, the entire business is in digital advertising.

So while you can argue that from Matter perspective, which we do like the AI story, that they have 3 billion users through WhatsApp, Instagram and Facebook, that the main business is still in digital advertising. And if we are going into some sort of a recession or slowdown, or when you have the stock market, I think at one point related to the chart that you’ve shown, when the world stock market was down 20% at one point, I think everyone is feeling a bit poorer in terms of where to spend the money. And hence, then it does have implication from the advertiser or company perspective in terms of conversion rate, what sort of rate they would pay in terms of targeting us to spend money.

I think that does have an implication. So even though that you would think like tech is tech, but it’s not really the case just because a lot of these tech companies are actually consumer-facing and you probably want to avoid those. And at the same time, if you’re referring to AI, which obviously we still like, then if you believe that AI is a trend to stay, and if you believe that the reason that some of these big tech companies are spending so much money into AI is because of an existential crisis, rather than they have a choice to do it or not, then that investment journey can continue for longer.

And I think the other thing, which is a stock that I mentioned earlier, it’s about Netflix, which we don’t own. And it’s very interesting in a way that it’s definitely not part of the tariff regime. And if we become a bit poorer and then we may have lost our job, we probably will stay at home and subscribe to Netflix just to spend our spare time to entertain ourselves.

Sure. Yeah. People don’t care about tariffs.

They just want entertainment, especially when things are more expensive, they stay in. Okay. That’s a good point, Stephen.

We’ll come back to some stocks that you like in just a minute, including Nvidia. So I think I just want to talk about tariffs a little bit more right now. The tech companies that I’ve been reading about, a lot of them are implementing what’s called a China plus one strategy, which is to kind of mitigate their higher costs from tariffs by moving to another country for production.

For example, Apple shifted production from China to India, Vietnam, and Target is relocating more of its private label sourcing to countries like Guatemala and Honduras away from China, for example. Now, have you considered what this may do to margins shifting global supply chains? Would that be beneficial for their EBITDA or not so much? Everything is relative, isn’t it, Stephen? So your starting point, because even though you think about other countries, including like India or Vietnam, I mean, they’re still going to be a 10% tariff for this 90 days period, not over 100%. So it’s on a relative basis, it’s better.

But I think the question to ask now, which is very interesting or very important for investors into stocks is like, which company has pricing power? Like when I say pricing power means like, are you able to pass on the cost to your end clients? And of course, someone has to pay for this, right? Either it’s the company who is making the product that they say, oh, the product that we’re making is too commoditized, that consumers are not going to pay for this. So then, and we need to maintain our market share relative of our competitors, then you’re going to have some sort of margin implications, like the company have to stomach this themselves. But at the same time, if you’re referring to some mission critical products, I mean, we have investment into biologics.

We, I mean, we have Danaher, who just reported this week. And of course, I mean, they, according to them, just because they provide mission critical equipments to the biologics or biotech industry, they would think they would be able to pass on the cost of a tariff to these companies. And of course, I mean, within the bioreactors or the diagnostic business that they do is a very small part of the cost space from the end client’s perspective.

So then the question is, of course, I think what is really confusing today is every company is going to say they have pricing power, but then you, it is our job to work out like who actually have the real pricing power, because it’s not everyone can just pass on the cost down to the next leg. And someone has to decide, I’m not going to pay for this. And I’m going to go somewhere else for the alternative product.

And I think that is the really important thing is like, how is that going to translate? Because someone is going to pay for that. So to answer your question about, okay, is that going to be helpful for them to move outside of China? I think it’s definitely helpful on a relative basis, but then that’s still tariff going on. And at the same time, we do notice as a fact that the reason is called China’s plus one in terms of this strategy, which has been going on for a long time now is because China is still a dominant suppliers of most goods or products, because just because they have the infrastructure, they have the labor force, they have a lot of know-how or capabilities and anything else is called plus one.

It’s not light for light. And then you might be able to do a bit in India or in Vietnam and Cambodia and some others, but you still cannot replace China completely. I mean, that could be like a 10, 20 year project.

Well, some companies in the auto sector are considering shifting their manufacturing to the US to avoid tariffs altogether. Do you see that happening in the tech sector? I think that could work, but ultimately it’s about the price that you pay too, right? Like you shift, I mean, this is like a very mathematical exercise. I’m sure every company is working that out, trying to work that out.

It’s like, yeah, sure. You shift your manufacturing back to the US and then you’re not being part of this tariff regime. But then of course the cost base or the operating cost base is a lot higher.

Everything is more expensive, right? I’m sure like we’re based in the UK here, like everything can be shifted back here. And then people are going to work for many fewer hours than people who work in China. And then the minimum wage and then all these bits and pieces is not as straightforward.

Of course, then you have to then work out from a company perspective, whether that’s actually cost saving, right? Like you say, okay, I’m going to double that cost in terms of tariff if you’re going to import it. But then by doing it onshore, like back home, is that going to be a lot cheaper or is that only going to be marginal? Just because, I mean, you can work this out. The average cost of a US, I don’t know, like a blue collar worker is going to be multiple times more expensive than a labor in China, right? So then you probably need to have the same number.

Unit costs are going to go up a lot. And so that goes back to your earlier point, which company has the most, which companies specifically have the most pricing power right now such that if they were to, let’s say, manufacture in the US, they could absorb the cost by raising the price. I think this is what we have been trying to work out is basically companies that you think they are providing some sort of mission critical services.

So not discretionary. Because if I think about Apple, okay, I have no idea how this conversation is going because at the moment it’s still very fluid in terms of whether it’s part of the mix or not. But if you think about Apple, of course, okay, you can argue that it’s mission critical in a way that we probably want to have a new phone if they launch it in time.

But it’s more of a discretionary decision. Like discretionary means that we don’t need a new phone now at this moment, right? We could wait for another six months or maybe 12 months period. Or maybe we could travel to other countries to buy a phone rather than to buy it locally, let’s say, in the US, for example, to pay a very heavy price, assuming that Apple is not going to take a margin hit on the back of their business.

So I think anything to us that’s a bit more commoditized, that may be a bit more discretionary, not mission critical, then I think maybe the kind of users could postpone that decision, or maybe could shift to something else. And the other thing, which is very interesting as well, because a lot of people would try to compare today’s scenario to 2022 about inflation, okay, the Fed might not be cutting interest rate, maybe inflation is going to spike up again on the back of the tariff coming through. But I think from our perspective, it’s a bit different to 2022, just because in 2022, you have the oil price going up to $100.

At the same time, you have us coming out from the pandemic, a lot of people have been stuffed with money, not working from the government, being supported by the government, and then they have been spending a lot of money, of course, a lot of cash saving in the bank, just because they are not able to spend the money. And so by the time we have the reopening, a lot of people just want to spend as much money as possible. But then when you think about what happened today, is even though the tariff is going to impact a certain of our spending, but I would suggest that most of this impact are not part of the necessity consumption, which means that if you go back in 2022, if you have the oil price going up to $100, I mean, it does feed through into your food basket into a lot of things that you’re actually consuming so that you can survive, rather than discretionary, right? Like by buying a few items from Amazon that’s made in China, that is probably less than, I don’t know, $50 of value.

I mean, does that actually change your life? I think my point is, you’re going to expect certain makeshift in a way that people say, okay, it’s a bit too expensive now, it’s a bit discretionary, maybe we don’t need that. So ultimately, I think at the moment, it’s a very confusing messages out there, because there’s so many moving parts, it really depends on which area you’re referring to in terms of companies that can pass on a tariff or not. Well, can we make the argument that all the MAX 7 stocks are mission critical? I think on a very high level, yes.

But of course, you have to look into the business model of different companies, right? Because even if you refer to Amazon, okay, 50% of the business is in consumer facing, selling goods to us, either through themselves or maybe through the third party merchants. But then a lot of some of these things would be coming from overseas, right? Like a lot of these things would be imported from overseas. And then if Amazon is going to make money on the back of that, and if they can’t sell as many of those things than before, then maybe they’re going to make a bit less money.

So when you say mission critical, I think that’s still an element about whether they’re going to survive. I think 100% all of these MAX 7 companies will survive. But then whether they’re going to see some impact in the earnings on the back of a tariff or a slowdown, I think the answer is yes, they’re going to see some impact and some slowdown in spending.

Well, the way I think about it is demand for whatever products are selling are relatively inelastic. But I think the only exception that I can think of, and please correct me if I’m wrong, is Tesla. I mean, nobody needs a Tesla, right? We can always buy something, a different car.

The stock has not been doing well. Have you been looking at the EV space much and whether or not much of this decline was due to either the company itself or just Elon being political? Yeah, we’ve never owned Tesla. By the way, it’s down 50%.

Yeah, just for context. Yes, I’m probably speaking more from an observer perspective. Of course, we have done some research on Tesla over the years, but we have never owned it.

I think the thing that I would probably relate to the recent share price reset is to do more related to the fundamentals of Tesla or maybe the industry landscape. Just because if you look at Tesla many, many years ago, I mean, Tesla was the only company in the world who basically pioneered EV, Elon Musk, to Elon’s credit. But since then, if you look at the landscape today, there’s many more competitors.

And you’re right that not everyone needs to drive a Tesla. Maybe we like an EV car, but it doesn’t need to be Tesla. And of course, when you look at some of the South Korean, Japanese, or even Chinese, the BYD, it’s very competitive.

Very competitive means very good value for money. It’s probably doing the job as much as maybe driving a Tesla, maybe not as a premium. So then when we always, I mean, since the beginning, I think even including today, having seen the share price go down so much, is that there’s always a big premium in Tesla share price on the back of Elon’s magic, whether he’s going to do something very magical.

And of course, there’s always a story about, oh, Tesla is very advanced in terms of autonomous driving, in terms of data they have. So then if they manage to crack autonomous driving, which they haven’t today, then the company should work in trillions, not few hundred million. But then it’s one of those things that you cannot pinpoint to.

And of course, I think where we are today, maybe the journey towards that autonomous driving from Tesla’s product offering perspective, maybe it’s going to take a bit longer, or maybe there’s more competitors. Then when you go back into the fundamentals of Tesla in the context of EV cars, then I think he’s still very much overvalue on the back of that, even though that’s gone down 50%. Yeah.

BYD, by the way, is already the world’s largest EV producer. I wonder if the tariffs are helping Chinese companies more than they’re helping American companies. Keep in mind the tariffs are two way.

So it’s basically a blockade at this point. You can’t ship cars to the U.S. from China where you can just pay a big tariff and vice versa. But do the Chinese really consider that a loss? Right.

Think about that. Yeah, I think it’s very interesting in a way that I think there’s a narrative. We’ll see how this goes over the next couple of months.

Obviously, if assuming that the trade relationship between China and U.S. is not going to be resolved, let’s remain tricky, that maybe China is going to flood the rest of the world with many of their products than before. And it’s going to create a lot of probably a bit of a deflationary force in Europe, for example. I mean, they can easily sell a lot more cars in the U.K. or the Europe just because they can’t ship to the U.S. anymore.

So then ultimately, yeah, who is going to win on the back of that? Is that helping China to conquer the rest of the world, forget about the U.S. for a second and make friends at the same time? I think it’s quite tricky because just because ultimately I think the challenges today, I think for the Trump administration or also going forward is China is here to stay. Like China is here to stay means that they have become very dominant, very rich, relatively powerful in terms of what they are actually manufacturing. And then to remove them from that equation, thinking like, OK, maybe U.S. still number one by a margin, by a small margin.

I think that equation have changed a lot today compared to, let’s say, 10, 20 years ago. So hence, yeah, it’s not clear whether Trump, what Trump is doing is going to actually hinder China’s development, just because I think China is very busy trying to make some new friends on the back of Trump’s administration, especially in the European domain, right, in the European domain, from Europe perspective, like we have lots of friends. Lots of friends means that we can no longer trust U.S. as much as what we did over the last 40 or 50 years.

And then suddenly Chinese comes in. Of course, China might not be that trustworthy, but on a relative basis, maybe they’re offering something that the U.S. is not offering. So then they become a friend.

So let’s go back to AI stocks. NVIDIA, in particular, you mentioned you were increasing your stake from, correct me if I’m wrong, but I believe it was 7%. Now it’s 10% of the holdings.

Now, you mentioned the valuations being an important factor. It’s down, I think, 24%, 25% year to date. So obviously, it looks better from a price perspective.

But fundamentally, any improvements in the stock itself that would make you more bullish? So the story about NVIDIA is at the moment, I would say, at a crossroad in terms of a lot of headline uncertainty, just because NVIDIA’s product, GPUS, everyone would be familiar with, that is very mission critical in terms of where this AI war is heading, right? Who’s going to win this AI race between the U.S. or China, or maybe somewhere in between would be Europe, somewhere there. NVIDIA is mission critical. So what it means is, of course, it being an American company, Trump could potentially weaponize NVIDIA GPU in terms of who is going to get NVIDIA GPU.

It’s a bit like your oil, like who’s your friend. And if you’re not a friend, I’m not going to give you that access. And you would have seen that recently about the restriction of H20 export to China, which they’ve been doing for some time at the lower end of the GPU spectrum.

So I think it’s a bit tricky in a way that when you try to monitor what’s going on, what Trump is trying to say, what is planning in his head, it’s going to cause a bit of volatility. But if you zoom out a little bit, so let’s say maybe just forget about the next couple of months or weeks, then you look at the really fundamental story about AI. I think our beliefs still today, nothing has changed, or we have even got a stronger belief today than maybe six months ago, is the AI is here to stay.

I think we are seeing a lot more AI use cases today than six months ago. And of course, everything that we’re seeing today would have required certain trainings on NVIDIA GPUs. And of course, their dominance in the space through their software platform called CUDA is very much remain as a mode in a way that they are very well positioned.

And I think the one thing I want to point out, which is very interesting, I mean, it’s not coming from NVIDIA, it’s coming from Broadcom, which we equally owned and Broadcom worked with the big tech company to provide in-house GPUs. So you can argue that Broadcom is an indirect competitor to NVIDIA. But what I would say on Broadcom was that on the Monday night, just a few days after liberation date on that Monday, when the market was actually gone down like 20%, the world stock market was falling off as well on the back of the US stock market coming off.

Broadcom announced on that Monday night, just a few days later, they’re going to start a $10 billion buyback program. It was not pre-scheduled. It was not part of the results.

It wasn’t part of a conference. But then they were very, I think the management team probably was looking at the share price. Oh, the share price was down 20 something, maybe 25% or something at a time after a few days.

And they just feel the pipeline of development in AI, of course, they work with Google, Meta and some others, is looking so strong that they don’t think the tariff or even maybe a global slowdown is going to impact the earnings trajectory of the company. So this is what I’m trying to say is in this earnings season that we’re going to hear more from this company and hopefully NVDM, I think from our perspective, hopefully they will come out and say, oh, nothing has changed. We’re still seeing a lot of demand.

It’s in short supply in terms of the GPU. We have a very strong pipeline. I think then you probably then start getting a bit of differentiation or recovery from NVD share price.

But what is not changing though, is the headline uncertainty that is still going on as long as Trump is still having this kind of idea about weaponizing NVIDIA GPU to make friends or to, yeah. Well, speaking of NVIDIA’s costs, this is, you know, going back to what I was saying earlier, one example of a company reshoring back to the U.S., NVIDIA to produce AI tools worth up to $500 billion in the U.S. over the next four years, according to Reuters, with the help and partners such as TSMC. The announcement on Monday, this was last week, includes the production of its Blackwell AI chips at TSMC’s factory at Phoenix, Arizona, as well as supercomputing or supercomputer manufacturing plants in Texas by Foxconn and Wistron.

So it doesn’t look like they’re building their own factories or partnering with other companies with existing factories in the U.S. already, at least for now. Has, do you think this will play a role in their margins? So, so this, I think this, this article is slightly, not this article in itself, but the news, this news announcement is slightly confusing in a way, just because NVIDIA remains or will continue to remain as a capital lights business. And that’s the reason why they are delivering quite a lot of free cash flow on the back of the revenue they make, which is about 50% of that.

And they do all the manufacturing through TSMC. So the, so if you look at the core space of NVIDIA, it’s really about people, about the R&D, about what they, what they, the software engineer that they hire behind the scene, or maybe a bit of distribution. But then in terms of manufacturing, they always partner with TSMC, which is, which is basically the sole supplier.

So this would relate. So the numbers that you mentioned here is, I think is, is in conjunction with TSMC investment in the US to build new foundries as part of the Silicon Sovereignty narrative, that is going to be more GPUs produced in the US through TSMC rather than NVIDIA is going to build their own factory to produce it. But then of course, the next question to ask is, okay, if TSMC can produce NVIDIA GPU in the US, no, not in Taiwan.

Let’s imagine those CPU, GPU would be sold to the big tech companies in the US and it’s going to be more expensive. I think everyone will understand that, more expensive to produce GPU in the US versus Taiwan. Who is going to pay for this? And I think if you, you are of the view that NVIDIA’s product would remain mission critical, means that there’s no, no direct competitor to come in to, to, to do it cheaper or better.

Then I think that he, NVIDIA should be able to pass on the cost to, to the big tech company to see. So, so the hands, hands ultimately is like, who has the pricing power? Well, the stock’s trading at about $105, $106 right now, as we speak, is there a fair value that you’ve assigned? Do you have a new DCF for us? Uh, no, I would say there’s no DCF just because I think, so, so this goes back to about what is, so what’s the most important thing about NVIDIA? So obviously from our perspective, we, we do fundamental research, we build financial model, we forecast up to five years for NVIDIA’s earnings, but a really important number for NVIDIA is probably in the next two years. And of course, then it’s a matter of debate about how real you think AI is going to be, because this year’s number is already in the back in a way that they’re going to make over $200 billion of revenue.

And on the back of that is about $100, $100 billion of free cash flow because it’s capital life. So it’s very profitable. So NVIDIA is going to be very much net cash, very, very big net cash position by the end of this year.

But then the question is not about this year. That’s why the shares trading at just over 20 times earnings about 2026 or 2027. If you link, link it back to the numbers that who is spending all this money this year is the big tech companies, the big six company alongside Apple doing it indirectly through Google, they’re spending over $300 billion investing into AI.

And I would say about half of that have gone to NVIDIA. Okay. So then what is this number going to be in 2026 is that is the key question, right? If you say, that number is going back down to a hundred, then of course there’s no fair value for NVIDIA because even 20 times earnings is too expensive.

But if you think that number is going to be sustained, let’s say at 300 or maybe more, or maybe you have enterprises coming in at the same time, which that journey have not started that we believe is going to come in the enterprises is going to spend more money into AI, not just the big tech. Then yeah, if that you think that number could be sustained, then NVIDIA is very undervalued today. Let’s talk about maybe another area that we can address.

Meta, you previously liked Meta, not so much anymore. Why? So Meta has gone through a very interesting journey for us. So we had Meta in the fund for four years until 2021.

We exited Meta in January, 2022. And then we bought back Meta in October, 2023, mostly on the back of AI. So obviously we were already invested into NVIDIA at that time.

We were looking for new ideas. We want to see which company would be able to benefit from AI. And of course, AI is apart from AI infrastructure, which is where NVIDIA and Broadcom play in, is about data.

Which company have the most data? And of course, Meta with 3 billion users on their platform is probably one of the kind company that knows the most about us. So it’s about data. But the problem with Meta is obviously going into the tariff early this year, it was quite expensive just because they are burning a lot of their free cash flow to invest into AI.

So basically their free cash flow has been converted into the revenue line of NVIDIA or Broadcom. But then we were saying to ourselves, we could stumble this valuation just because they could be the ultimate beneficiary on the back of AI. But I think what have changed on the back of Trump’s tariff is not related to tariff so much, it’s about digital advertising.

Just because between now and then, just imagine if the AI investment journey continues, Meta’s top line is still 100% digital advertising. And between Google and Meta, I think these two companies would control over two thirds of the world advertising market. And everyone would know in a cyclical downturn, or maybe in some sort of slowdown or recessionary scenario, advertising is one of those areas you probably would try to scale back before you do something else.

And so Meta’s business is fairly sensitive to that in terms of the core business itself. So we decided to pull the trigger to avoid those kind of the advertising exposure and exit to the stock. Is there a future for Meta that you see could, like a pivot for the company that you see could revive your interest? Yes, I think it’s all very much dependent on where we are heading from here in terms of Trump’s tariff.

If you say to me in a few weeks time, hopefully not months or quarters, that everything is going back to status quo in a way that there’s not so much of a trade tariff or maybe the trade tariff is already done. And then we have seen some evidence that we are actually not going into a recession. Consumer spending remain very much healthy because the tariff in itself is not impacting our spending.

And I think if that’s the case, then yeah, of course, I think as far as Meta is concerned, it’s all about AI. Then you go back to the AI story, it’s not about a recession or a slowdown. But I think that’s very much subject to the macro environment, subject to what Trump is going to do after 90 days.

We still don’t know what happened in between now and then. So then in terms of trying to forecast what Meta is going to make in terms of earnings, this year is actually quite difficult today. I know that Meta’s revenues primarily come from advertising.

Are you assuming that advertising will slow down this year because of maybe a global recession or just slower growth overall? Definitely will slow down for sure. If we’re talking about a scenario about either global slowdown or US recession, that number is going to come down. I think this is mathematical.

And at the same time, I think this is a bit of a small nuance, but we have to see how this plays out. Previously, I think over the last 12 to 18 months or so, some of the Chinese e-commerce merchants like Timo, which I think is very big in the US, they have been spending aggressively on Meta’s platform or maybe on Google’s platform just to advertise to American consumers to buy the product. And of course, you would be aware of in the US, they had this rule that if you ship goods less than $800, you don’t have to pay the import duty on that, which that rule have now been removed or revoked, which then they have to pay that, which is going to make those products more expensive compared to things that’s already been sold in the US.

So what it means is if that tariff is going to impact some of these Chinese e-commerce merchants who’s going to ship goods from China to US consumers, and then the likes of Timo could actually reduce their spending just because the conversion level through spending on Meta or maybe Google is going to be reduced from here on the back of that reduced demand. Let’s talk about Microsoft now. Microsoft is another company that I think you’ve offloaded recently.

I think there were cost concerns for you, especially regarding their AI division. Can you comment on that? Yeah. So Microsoft has been a very special company in the fund because we have had Microsoft in our top 10 holding for seven years until August 2024.

So it’s a very special holding. It’s one of the longest held stock in the fund. And then we started to reduce our position since September last year in the last six months.

And then recently we sold out of the company completely. The narrative that change is not because Microsoft is going to disappear. I think it will remain with us for a long, long time, just because it’s very entrenched in terms of the Office 365, in terms of Azure, or even including Teams in itself that they developed recently.

The question mark about Microsoft alongside other big tech companies is that they are spending a lot of their money investing into AI. And one of the things that I never believe in is the big tech companies are never good innovators. If you think about WhatsApp, Instagram, and YouTube, they were all acquired businesses.

It’s not like the Google or the Facebook at the time called Facebook or Microsoft developed anything interesting themselves, just because their incumbents, they are able to scale up those businesses once they acquire them. Of course, at the moment, you are going through a bit of antitrust narrative in the US for Meta now. So what I’m trying to say here is Microsoft has invested a lot of money into copilots, which I have used personally.

I’m sure a lot of our audience would have done so too. I mean, the product is fine. Ultimately, it’s what price are we going to pay for this? And I think the price we’re going to pay for that would be significantly lower than what they have invested to develop that.

And ultimately, it’s going to impact their return on investor capital profile and also their margin. So at the same time, I think one thing very important to remember is that over the last two years or so, a lot of this investment is through capital expenditure, which means that it’s not part of your P&L expenses that you would think that in this coming year, maybe more so in next year on from next year onwards, it’s going to hit that P&L because they need to start depreciating this kind of investment. But if you haven’t actually generated additional revenue on the back of that to cover that, then you would expect a bit of a margin compression in the coming years.

I mean, unless if you assume they’re going to make copilot work, and we are very happy users to pay for that, then it’s a very different story. But at the moment, we don’t see any signs of that happening. So we’re a bit concerned about the spending in AI as far as Microsoft.

They could capitalize certain R&D expenses for copilot, can they not? And that would not directly affect their operating costs. So which part? Couldn’t they capitalize their R&D expenses for, I guess, copilot, which is what you’re talking about. So it becomes an asset and it wouldn’t actually… Yeah, I mean, they do.

I think that’s what they have done already. I mean, it’s part of the kind of the capex investment, right? But they still need to depreciate them. I think the question is how quickly do you depreciate them or maybe the life of the asset, whether it’s going to be over three to five years or over 10 years.

Of course, I mean, this is more of an accounting kind of adjustment. You say, okay, actually, maybe three years too soon. Why don’t we just expand that to six years? And of course, then the number would become smaller, but they would need to depreciate them.

So this is a question that I think we’re going to see in the next couple of that they need to make AI work. And the question for us is not about AI, it’s about who is going to make it work. So I think our question mark is about the likelihood of Microsoft making it work based on the investment they have made might not be as high as Meta, which we probably would prefer if we don’t have this global slowdown scenario.

This begs a bigger question, Stephen, which is whether or not these tech companies all need to invest in some sort of AI technology. Is that sort of a necessity that these boardrooms have discussed and decided? Or is this something that they just want because it’s a trend? I think it’s probably a bit of a latter. I would say I would also add on that it’s more of a existential crisis.

So at the moment, it’s a choice in a way, a choice means like Alphabet, Google can decide to stop investing into AI. Their free cash flow would go up significantly, right? Amazon can do that too. Why do they have to develop their own in-house AI tools? Why can’t they just buy it? Exactly.

This is the whole point, right? This is an arms race. It’s a very big decision that I don’t think any of these companies are going to make today that unless they think AI is not as real as what they think it’s going to be, or maybe what we think it’s going to be, the risk of not investing into AI, which I think Apple would be maybe a case study in time, is a big risk to the business. Just because if AI is here to stay, if AI is going to be as game-changing as we think it’s going to be, not having these capabilities, it’s a bit similar to Google Search not having this algo many years ago, because Google Search was the six or seven search engine created.

It wasn’t the first. Where were the other six? I mean, they were all gone, and Google’s entire business was built on Google Search, and then they acquired YouTube later on, and they got other bits and pieces later on, but Google Search was still the most dominant part. So then if you then think about what happened in that landscape many years ago was obviously Microsoft was trying to crack search with Bing.com, which had never been successful, and then at the same time, Google ended up going into Android alongside maybe Apple iOS, and Microsoft missed out on that, which has been very costly to the business in the 2000 era after smartphone was launched.

So then they are still very much part of our day-to-day in PCs or laptops, but then they’ve missed out from the mobile operating system, which is a very big market. And AI could be one of those things, which to us, for any of these companies, of course, I mean, they are quite technical savvy in a way that they have a lot of talent in-house. For them not to be able to crack it, it would be the most costly mistake.

What I’m trying to say here is any of this company at some point, they’re going to drop out from the race. I mean, not everyone is going to make it, right? Just like anything else. But if you ended up dropping out of the race just because you are not very capable, and while your competitors are able to crack AI, especially agentic AI, which is what is coming, AI agents, then maybe you’re no longer relevant.

So maybe in time, we’re not talking about Mac 7, we’re going to talk about Mac 5, Mac 4, or Mac 3. What do you think they should do? Should they spend hundreds of millions, if not billions, of dollars doing exactly what you’re just talking about, keeping themselves in the arms race, or just use some open source software that’s already being developed every single day as we speak? I think there’s a few things going on. I think it’s a combination of a few things going on, right? If you look at like Meta, I mean, they started this open source platform called Lama. So whatever Meta is doing today is on an open source platform, and they are utilizing the open source system to probably improve their software quite quickly.

And DeepSeek, as we all know, have utilized Lama as well. So hence, there’s a bit of a learning in terms of the breakthrough from DeepSeek that’s now been brought back to the Western tech companies. And I think it’s a question about where you want to be.

Because if you think, okay, I’m just going to be a tier two player, of course, you can spend less money, or maybe you can just follow the crowd. But if you want to be a pioneer who’s driving the innovation, who is going to lead the conversation, who is going to maybe create a new space, I think you need to do it from scratch. You need to spend a lot of money.

So that hence, I think if you look at OpenAI, which is not a listed company, is that’s what they were trying to do, or they are still trying to do this. That OpenAI started the GenAI, ChetGBT revolution, and everyone else trying to follow, everyone trying to catch up. And then you’ll be asking the question, why some of this company had been a bit slow to adopt to that before ChetGBT arrived in late 2022.

But OpenAI is now truly established, even though it’s not a listed company yet, but it’s establishing themselves as the pioneer. And of course, then you would have a bit of a first mover advantage. And at the same time, you probably can attract better talent.

And if you’re not making this a priority, like if you say to me, if you are a very brilliant AI engineer, would you rather to work for OpenAI or Microsoft? You probably want to work for OpenAI than Microsoft. Maybe AI, they’re spending a lot of money, but it’s not the entire thing that they’re doing. They have a lot of other things that they’re working on too, which then AI is only part of the equation.

Thank you. Well, that was good. Thank you very much.

Before we close off, where can we find more information about your work? Yes. If you want to see more of our articles or some other content, you can come to our website at bluewell.co.uk. Okay. We’ll follow up in a few months.

Let’s see how your Q2 portfolio is adjusting. So thank you very much. And we’ll speak again soon.

Thank you. Thank you. Thank you for watching.

Don’t forget to like and subscribe.