Speaker 1Welcome to the Saxo Market Call. Before we get started, it's important we emphasize that the views and opinions expressed in this podcast are those of the hosts and guests and do not constitute investment advice or recommendations. All information provided is for educational and entertainment purposes only. Hey everyone, it's Friday, 13th of November, 2026, and I just wrapped up a very nice long forum interview with Peter Garnery, CEO and co-founder of Jezda Capital. All things NVIDIA and AI and lots more. I think you're really going to find it a treat, so do stay tuned for that after my brief market wrap, and I was so busy putting together some questions for Peter that I haven't even noted all the basics of what unfolded yesterday, but we saw the market trying to find its even keel here after this NVIDIA bombshell. Those shares were down heavily, as I discussed with Peter. Peter, thank you so much for that coverage there, and we ended the day somewhat with a comeback, some AI names doing well, a lot of poor quality stocks doing terribly, and we see these remarkable market internals once again showing themselves, and a very deep coverage of that as well with Peter later on in our conversation. I just want to briefly do about a five-minute macro wrap here because we have seen some pretty decent drama across into basically all other markets. The focus in equities, I think, narrowly on this whole NVIDIA situation and the AI trade, whereas I think if that wasn't distracting folks, they might be looking over at yields and getting a little bit concerned because we have seen a tremendous repricing now in the two days I was away from the podcast, and it happened just after I recorded the previous one, this huge jump in those yields on the first of the Waller comments, and then he specifically linked, that is the Fed governor, Christopher Waller, linking the CPI core if it hits 0.2%. 0.3% to potentially wanting him to open up for a 0.5%, so a half-point rate hike, pretty unusual language, especially given we've just had a chunky $20 drop in crude oil prices that he's getting so aggressive now, but we've seen some other Fed officials out. The latest was Hammock and Williams chiming in with some hawkish-sounding stuff yesterday, and then, of course, we actually got that hot CPI print. So we've seen about a 25 basis point repricing higher in the U.S. two-year. Suddenly, we're back at. We're back at nipping at 4.9%, only five or six basis points below the recent cycle high, and yesterday, the longer end was a bit more peppy as well. We've seen the 10-year coming up to 5.12% after that long little period or spell below 5% recently. That's a bit of a concern. We saw a soft 10-year auction, and we saw a soft 30-year auction just yesterday. So yields very much feeding into the FX space where we've seen a repricing of the dollar higher as well. It looks like a breakout. The dollar yen is a massive focus here. It's squeezed through that massive overhead resistance area, 158.50 and 159 sort of zone there. And, of course, where is it finding resistance? It's at 160, and the reason it's finding resistance there is not only psychological, but many see this as the line in the sand from the Bank of Japan. There were some previous episodes, even before there was that final run-up into the all-time high. Or not all-time. All-time modern high of 163.99, I think it traded. But before that, there were several episodes of 160 seeing some intervention. There was the big intervention off that high from 163+. And then when it finally found the bottom and started bouncing again, it was in the 160 zone. There was the next big wave of intervention. I think it traded three days sort of back and forth over the 160 level back in late September before the Bank of Japan came in. But again, the driver here is this repricing of the Fed. I just question, you know, look at your dollar. It's breaking down. It's hit new lows, and now it's 19-month lows. We're talking since the beginning of, or early, I should say, 2025. Elsewhere, we're not quite there. We're still within range, even though we've seen quite a sharp move. We're within range in cable. We're within range in dollar Swiss, just barely so there. But we've broken down in euro-dollar. We've broken up in dollar-yen. And if the driver is the FOMC, thoughts that the FOMC is going to be hitting 5% yields sometime next year, that's what we'd have to reprice to to get a further boost in the dollar. I just wonder if we get there. It seems a bit much to ask or to continue to pile on to this projection of yields higher from the current levels. So, you know, a bit of skepticism there. We need to look forward to next week. What could challenge this shift higher in yields? Well, Warsh is out speaking at one of these task, forces he has set down to study different things. This one is the jobs and productivity task force. He has quite prominently mentioned productivity as a maybe a hoped for area that we could bring economic growth, obviously, without, you know, without injection of inflation into the economy. And so observations around that that are particularly pointed could point to him still wanting to wait out this productivity revolution rather than panicking on rates. I don't know. He might want to send the market a message in any case. I sincerely doubt if he'll send a message that can be read as particularly hawkish relative to where we have already gotten won't eliminate anything. Anything can happen. I just suspect that the risks are asymmetric of him being a little bit more pointed on being balanced on the risks, forward risks to inflation relative to the batch of comments we've had this week. So again, he's already up on Monday out speaking. We also have the Tuesday retail sales data out of the U.S. Curious what the mood of the consumer is. There and even more interesting, perhaps, than that read, which is a very big, broad read. We have company earnings coming in next week from. We have Target and the Home Depot up earlier in the week, and I believe Walmart is on Thursday. I need to check my calendar on that one. So, you know, you get some a bit more color in some of these earnings reports. Huge retailers there. So what is the what is the sense that this affordability crisis is feeding into their behavior on the ground? Could be could be interesting. I'm not often a big market mover, but it could be interesting. At the margin, of course, the big earnings event next week is Nvidia's earnings. And, you know, to what degree that company can sort of instill any kind of enthusiasm or sense of solidity after the massive markdown in Nvidia shares from their all time high. Again, more on that in the conversation with Peter. So decent read on the retail space in the U.S. Worse up on Monday as next steps for this big macro move where we've seen rates coming up and seeing the dollar punching higher and also one worrisome note. There to sort of inject is the Germany France spreads. It's edged up a bit here, not flashing red just yet, but bottomed out last week and is back above 150 basis points. I'm sure in part a function of rates rising again in general. So let's see where we go. We've got Aussie dollar, by the way. It fell below 68 for the first time in a long time. That Aussie Kiwi reversal has continued a bit lower. It's interesting that in the case of Australia, I think a lot of that is linked into metals and metals. No big surprise. Very much under fire here with this repricing of yields higher. We've seen silver down below 50 at times and gold hitting those new lows and taking out stops below 3900. I'm guessing those directions or that direction of those trades will be very much hinging on the direction of U.S. yields and the U.S. dollar. And as well, look over at Bitcoin. You're seeing it trading way below that big key level that it took out on the way. Down and even below 69000 at times this morning. I think I lost my train of thought on Aussie. Yes. So Australian rates have not really come down, but there is seeing a lot more noise level on risks to the Australian economy here. So it's not looking good for the Australian dollars. One of the weakest, actually, of the G10 currencies here locally. And I think to revive its fortunes, you probably need to see the metal prices coming in a bit higher. But I'm a little bit more concerned on the economic read. I think these rates are really starting to. To hurt Australian household budgets with those floating rate mortgages, et cetera, there. So I just wanted to throw that into the conversation, too. All right. I'm going to keep it brief here because the main focus in today's podcast is certainly my conversation with Peter Garnery. So just stay tuned for that now. Peter Garnery, CEO and co-founder of Jezda Capital. Thanks so much for joining us today, Peter.
Speaker 2Thank you very much, John. It's always a pleasure.
Speaker 1Well, it's great. It's great to have you with us. Let's dive right into it. NVIDIA. That's what everybody's talking about. Crazy day yesterday. 750 million shares almost were traded. Many multiples of the usual daily trading volume. Ended the day down over 4%. And at one point, it was down more than 14% on top of all of its declines. 32% almost now from the all-time highs where it closed yesterday with Jensen Huang's exit here. What is your take on – well, I don't know where to start, really. I mean, there's so many things here. But look. Let's say – what is your take on his individual culpability in this situation, if you have one, relative to NVIDIA's culpability? There had to have been a lot of people that knew what was going on here. And there was a lot of talk, let's be honest, that it was quite obvious that when – at the time when the Chinese ban on NVIDIA chip imports was implemented, suddenly there was a new category. I think it was Southeast Asia or something like that where there was tens of thousands of chips or something being sent to this new entity. in Southeast Asia. What's your take on the culpability and where to associate that with this situation, if you have one?
Speaker 2Well, you can start by saying this is a perfect illustration of Warren Buffett's old famous saying that you can spend an entire lifetime building up your credibility and throw it out in an afternoon, right? So it's spectacular. I would say on NVIDIA, you and I, back when I worked at Saxo and maybe I think three years ago, you and I, we had some conversations because I already back then saw this very peculiar phenomenon that they didn't highlight Chinese sales and Singapore as a geographical segment in their press release and their earnings presentation slides. But in their 10Q filing with the SEC, you have to be very… strict applying these rules. And if there is a significant geographical segment, you have to show it in your revenue segmentation notes. And you saw Singapore back then. And then there was a little bit of focus on it because there was all these, you know, the trade policy implementation, the restrictions on chips going to China, and suddenly it disappeared again. So it was sort of appearing and then disappearing again. And that has been some of the talk recently. But this is a phenomenon that happened already three years ago. I know, I think NVIDIA already back then. I think NVIDIA already back then knew that this was not a good look. And I think the big question is, OK, so Jensen Huang is out. He was an important CEO. He had 60 people allegedly reporting to him directly. So a very flat organization that will be a CEO vacuum after this. They need to figure out who is going to replace him. Is it internally or an external candidate? Clearly, they need to reorganize the company from a management perspective. The big question in the market is, will they continue to do all this financing? And backstopping and guarantee deals that they have done, they're basically being called the central bank of the AI ecosystem. And we can come back to that in a few seconds, because I think that the big question and I already have people asking me, is this a lifetime opportunity to buy NVIDIA? And I think the big risk that is evolving here is that if Megaspeed, which this company that they sold advanced chips to, was used by Chinese entities to basically use Singapore. As a hub to circumvent US trade restrictions, if that's the case, and if someone in NVIDIA, either Jensen Huang or people close to him knew about this, then it's a very, very serious violation of US laws. And then I think this could go to a completely new level. So I think the only thing I can say with certainty is that there will be a lot of uncertainty around NVIDIA in the coming weeks and months. Of course, if you're a short-term speculator, then there's always mean reversions that you can jump on. Fair enough, people buying the dip. But I think for the longer-term investor, I think there's some serious questions and risks you need to consider now.
Speaker 1Yeah, some great points there. I love that expression, the central banking of the AI trade, or the central bank of the AI trade. There's also, and I brought this into the conversation in the podcast a couple of days back, I think, or maybe it was the last podcast I did here, that there's this very interesting political angle to this in that Jensen Huang was very clever in cozying up. And that apparently you're seeing these articles that Trump knew about this some time back, a decent time ago. And then the questions around the release and from intelligence sources of this story, was it an accident? Was it stupidity or incompetence that saw this story released? Because it is kind of embarrassing, to say the least, for US officialdom. Or did somebody release it as a political hit on Jensen Huang, on the company? And because they felt like, oh, this is going to be a problem. And they felt like it was a national security issue. And they felt like the leadership at the very top was not doing what it needed to be done. And so this was leaked to the press, basically, from intelligence sources. Because all signs point to this. This was an intelligence leak, not something that some clever journalist found his way to. And then there's the angle, can Jensen Huang buy his way out of this trouble? There's him personally and that patronage network leading all the way to Trump. Whereas NVIDIA at large, it's a different kettle of fish when you're talking about if it was not only him personally. But the company itself. But yeah, you don't need to comment to that. Of course, you're welcome to.
Speaker 2But let's. Yeah, I'll just quickly say that the whole political beast of this, I think, will continue to evolve from here. I think the much more important question, the bigger question, is what will happen to these backstops and guarantees that NVIDIA and equity investments. Because they have been ramping up their balance sheet with these equity investments. And guarantees to quite substantial amounts. And the, not rumors, but actually quite credible sources are indicating that there are companies in this ecosystem that cannot get credit financing unless you have guarantees or backstops coming from NVIDIA. And if Jensen Huang was the sole CEO, the mastermind behind these types of deals, is that something that is going to be continued by a new management team? Or will we see a rearrangement of how NVIDIA conducts its business and the risk that they're taking on that? That balance sheet to support all this. That is the fundamental question, I think.
Speaker 1I fully agree. There was this guy, Harris Kupperman, a hedge fund manager on Twitter, with a very cynical take. You make these. You present the question quite politely. He says, "Well, how are you going to replace somebody that has created this massive empire of related parties, special purpose vehicles, self-dealt corporate guarantees, circular transactions, sham sales, oddball financings of partly owned entities, etc.?" You're alluding to all this in a more polite way. And he's saying it almost sounds criminal, the way he's presenting this. So it is not just about a very key chip company, as you say, and its CEO suddenly finding himself on the street. Was he an orchestrator of something a little bit more suspicious, A, malicious, B, and downright manipulative of the whole market, in the worst instance, if all these negative allegations are true? And then I wonder, just going to that, I've seen other critics of the whole. AI trade, or certainly the Nvidia side of it and the silicon side of it, saying there are these shadow inventories out there that also that he was sort of such a cheerleader for. Everybody, you need to lock in your GPUs because there's only a certain amount that would be produced. And so you would lock in your orders. And then the problem is getting these chips plugged into systems, there's the systems integration, and then getting those systems live plugged into the electricity grid. And I think we can segue to the next bit on this, which is. Where is the data center and where is the bottleneck? Is the bottleneck in Nvidia chips, memory chips, other chips, the equipment to make the chips, or simply the capacity to even plug all this infrastructure into the grid and get it operating and feeding out to the high demand there obviously is for the AI compute?
Speaker 2Yeah. So I think the getting data centers built and plug them in and get the GPUs running. That's where the severe bottleneck is. The bottleneck was obviously also in memory chips to a lesser extent in the GPUs. And I think we probably hit peak CapEx cycle growth quarter and quarter and year and year in Q3. So we're past the tailwind period and we'll move into a new phase going forward. I think beyond the memory chip, that constraint will ease. So it is getting the GPUs into the data center. Get the data centers, get the data center, building permits approved, get them into the grid. That's going to be the difficult part. And I think it's a good segue to talk about CoreWeave because part of Jensen Wang's magical financial engineering game here was to try to convince Wall Street that GPUs have collateral value. And I think the story on Microsoft basically, I think call it stabbing a dagger into the back of Oracle and twisting it around. around by pulling out of a big renting deal with the argument that the gpus no longer uh had the the quality and the capabilities needed by microsoft models that's another way in my understanding saying that gpus actually degrade pretty fast in terms of technology and their collateral value which is a bit of a danger for two reasons so first they all extended the depreciation cycle um but if they are degrading so fast and you have as you pointed to these 200 billion dollars of gpu sitting in warehouses cannot get into data centers and the collateral value of those are declining because of you know technology evolving in gpu technology what what what signal and what does the credit market take out from that that is also a very big question especially when we talk about core weave their 31 uh nine percent coupon bonds are trading at 16 yield to maturity currently which is i mean if you think about it you're selling a commodity which is basically just renting out compute your debt financing is at 16 you can't write you can't probably raise a lot of equity right which would sit at 10 12. so if you do the weighted cost of capital for this business it's probably at the margin or incrementally at around 15 ish i don't think you can get 15 return on cap return on capital for a commodity business so i i i don't see the logic in this i i must admit i think these neural clouds are going to to be squeezed very hard in the years to come.
Speaker 1But let's take – if the bottleneck though is – I'm not specifically thinking of Corweave and there was a bit of a relief trade on Corweave earlier this week. I don't know to what degree that was fair. vibe then then it was great uh great news but if the bottleneck is the um is the data center and there is this intense and just almost uh unending demand for ai compute if you do have the capacity the data center capacity and you can provide services and it seems like if the growth or the demand side of this for the compute is there then and you own that data center then then you can charge uh higher prices and you see that you so you see this uh odd i don't know it's odd because obviously some of the compute gets more efficient so we see token pricing uh token prices falling but you see the gpu rentals uh prices have increased what is it 40 percent for the rent on on gpu time at the highest end over the last year so where you know is there anybody that's going to benefit from this obviously you can benefit down the road if if the um if the hardware production that the chip production is going to keep up with whatever the capacity is to install it you're not going to spend all of your budget on the memory and on the the hardware um that that you have recently but is there in the meantime is this good for the hyperscales because they can get more they can extract more from the market for that demand for
Speaker 2the ai compute yeah i think the way to think about the ai value chain right now is that the this branding of computer gpu is like uh refineries in the oil industry and the token prices is basically the price you pay you pay for gasoline or diesel at the pump right so the prices downstream at the end user level is actually declining um demand is accelerating because of these declining token prices but on the refinery side that's why you have the bottleneck bottleneck on this gpu and it's it's correct the prices have risen but they haven't risen to a level where it it really is a staggering business right because if that was the case you said it so there's a little bit of a relief rally in core weave but on the credit side on the bond side it's a little bit of a relief rally in core weave but on the credit side on the bond side it was not like the credit investors really thought that it was this was a game changer so i think the relief story was much more on the equity side and that's obviously also where the first losses of the ultimate loss if the business is not viable uh is taken first so that i obviously understand the cheering there from the numbers but if you look at it they have committed 40 billion dollars net in in after depreciation in in investor capital and they train it generating something like 250 million on their uh on their operating uh side so it's not like the credit it's still very low margin business despite of the rising gpu prices and if you and that's also the signal that we or the information we got from the q3 numbers from spacex which is probably the cleanest measure as well we could get on a on a large scale uh gpu renting business um there is a return but i don't think it's great and i don't think that's where the value will accrue over time so it's this um i think it's a dangerous trade if you if you if you engage yourself from that from the investor
Speaker 1side okay that's so that's super interesting and you talk about a a low margin business now you know there's a huge uh event risk coming up that i think is really critical for the sentiment around the whole ai space i'm not at all liberty to talk about it for for various reasons that will be clear later uh so there's that but there's also just this general um you know what is what is uh why is it a bit of a commodity business and it's part of that as you're seeing the rise uh in the use and then the overall token consumption as far as increasing more rapidly for these open weight models versus the frontier models yeah you know when you're looking at that space or that aspect of all this any observations you have about these these frontier models they're they're needing to run ever faster to keep ahead of the open weights the chinese deep seeks uh etc here
Speaker 2yeah so it it it reminds me a little bit of of the uh of the pc industry in the uh in the um the late 80s and and and throughout the 90s where you had a huge deflation cycle in the early 90s and then you had a huge deflation cycle in the in uh in pc hardware um which made it extremely difficult for uh for these companies to survive dell was one of them but it's it's extremely tough and if you look at the token price depreciation it's somewhere between 70 and 80 percent across uh the different vintages of uh of models and and so that means that you need a very very very high demand uh acceleration constantly to just stand still basically in terms of revenue i think that's the the paul kudrowski argument and i think it just makes it really really tough from economics point of view and i think for me so one thing is renting out compute but on the on the model on the model layer you have the frontier labs they're not going to be a single one dominating one so although it will not be as commoditized as the the vending layer it is still a very fierce not even duopoly uh it will be multiple multiple um models you have keep we may have nvidia you have meta you have the chinese open weight models and it seems like the open weight models are really competing very fiercely a lot some of the frontier labs constantly are slashing their prices on their models to to to keep their market share because it's super important that they can show growth to their investors but the fact is you have these news constantly i think it was the revolution ceo that came out yesterday saying almost bragging hey we spend uh we spend this amount on ford and then you know we switched to these uh kimmy and and other open weight models and we slashed our cost by 95 percent um that's staggering and if a lot of companies do that then it's not really a good thing and i will i will add to that i think that the former uh um the former one of the former chiefs at open ai moratti is a pretty clever woman and she left open ai and the companies that she has built i think is brilliant because she's putting herself into the um almost like the consultancy layer where they they help companies take a foundational model whether it's open or uh obviously it is an open weight because then you can post train it on your own data your own domain data and then you get a much better domain specific model and i think that is a very very interesting path and i think we'll going forward we'll see more local models more open weight models and i think it puts these frontier labs in a very difficult position over time unless they can unless they can sort of branch into the open weight models and i think it puts these frontier labs in a very difficult position over time unless they can unless they can sort of branch into the application layer and that is what into the application layer and that is what we're seeing they they're trying to build applications we're seeing they they're trying to build applications and and go head-to-head competing and and go head-to-head competing and and go head-to-head competing with microsoft and other software vendors
Speaker 1with microsoft and other software vendors with microsoft and other software vendors incredibly valuable observations and that incredibly valuable observations and that incredibly valuable observations and that really takes me to the next point with all really takes me to the next point with all really takes me to the next point with all this i mean we just run through some this i mean we just run through some this i mean we just run through some pretty skeptical observations on on pretty skeptical observations on on pretty skeptical observations on on both the the whole cluster of or mystery both the the whole cluster of or mystery both the the whole cluster of or mystery of what the cluster of these companies of what the cluster of these companies of what the cluster of these companies that nvidia has supported the circular that nvidia has supported the circular that nvidia has supported the circular trade there's the concerns around that trade there's the concerns around that trade there's the concerns around that concerns around the the perspective from concerns around the the perspective from concerns around the the perspective from a chip maker a chip maker a chip maker that can they realize these margins and a that can they realize these margins and a that can they realize these margins and a hard hardware maker in general can they hard hardware maker in general can they hard hardware maker in general can they realize the margins that they realize in realize the margins that they realize in in this cycle. The market is pricing them as cyclical. I mean, let's be honest, some of the forward PEs are staggering on these memory makers. We've stepped through the skepticism around the data center ability to extract the profit margins from the data center that is hoped and the hyperscaler. And yet, when you look at the concentration and global portfolios that are AI linked, it's staggering. I mean, you look at one of these pie charts of the sectors that are up and down, and it just almost blots out the sun, everything that's even just directly related to AI. And there are a lot of things that are downstream of AI as well. So you at Jezda, I mean, you're involved in advisory, so advisory of client portfolios, and as well as running a sort of concentrated, I can't remember how you characterize your fund. I'll let you do that. But you're operating in an environment that is very strange here with this shadow of this AI trade over everything you do. What is your observation? around that, first of all, and how do you just how do you navigate?
Speaker 2Yeah, so I made an observation on LinkedIn, before the AI trade got really wobbling. And what I said was that the most important risk factor, so in the equity industry, we always talk about risk factors, you have the market risk, which is basically the market going up and down, you have your interest rate risk, you have valuation risk, you have all these different risk factors. And I think actually, the AI risk factor, is the most important one right now, because when you look at the market, and you look at the entire value, value chain, the implications from the Capex cycle means that the AI risk have morphed into more almost half of all the industry groups, when you look at the very largest companies. And that means that if you're not really careful, when you construct your portfolios, you will be running a very, very, very high exposure risk to the AI trade, which is an investment cycle. As we all what we know from history, investment booms are exciting, there is a lot of momentum, and you know, people are positive about it. And you know, it is spectacular. And I'm also an optimistic guy by nature. So don't get me wrong. But we also know that investment cycles always come to an end. And there's always some bleeding on the other side of when it ends. So what the way we engage with our clients is that so we really much say, okay, if you if you invest in the MSCI world or in the emerging markets, you get a very high exposure right now to AI. So what we what we are rising on specifically on the US side is actually to go in to go out of the cap weighted indices, because of the concentration risk and the passive flows and go into the equal weight in the sea. And just to give people a little bit of a reflection. So everyone is on the same page here. So the equal weight S&P 500 is basically equal weight of the portfolio in all the 500 stocks in the cap weighted, you obviously weight the the positions to the market size, the free flow. So NVIDIA, Microsoft get a huge as the 10 largest companies are around 40% of the index very concentrated. If you look at the evolution of the equal weight index on S&P 500 from 2003 to around 2014, it actually delivered somewhere around 150 basis points to 200 basis points of excess return over the cap weighted index. Then from 2014 up until now, we've had this monster mega cap trade, the explosion in technology earnings, and you have had the exact opposite. Right now, that spread has been stretched a lot. I really recommend people to think about things like this. Is cap weighted index and passive the right way? You can still be passive, but you can do it in an equal weight. I think it's really worth considering.
Speaker 1There are ETFs that do track the equal weight index, like the S&P 500 equal weight, probably the largest one.
Speaker 2Yeah, and they are super cheap. It's an easy choice to make.
Speaker 1Yeah, and then along the same line, and I often call out that equal weight performance every day relative to the market cap weighted performance because at times, and very often actually of late and in recent months, especially since the massive top in that when the AI trade saw that real blowout into the end of QT, so around the June timeframe, that's really a key demarcation point for when a lot of the speculative froth came out in the most high momentum names. We've seen this day-to-day highly negative correlation, between whole pockets of the market and even the overall equal weight versus the AI trade. And often if the one is up, the other will be down and vice versa from day to day. It really reminds me of the market. I've said this before many times on the podcast, but I'll get your observations in a second. It reminds me of sort of the tech bubble's final phases and the transition phase as before we went into the true bear market back in the tech telecom bubble and bust. I mean, first you rotated out of the dot coms. They were sort of the first to go. The overall NASDAQ composite and NASDAQ 100 peaked in that March run up. And then that was the beginning of a huge and churning market where you saw tech and sort of blue chips and value elsewhere doing different things. Every day if one was up, the other was down and often very violently so. And then later on, as you got into 2001, while the general market went down, value started performing very, very well and value stocks had a spectacular 2001. It was only towards the end of 2001 that finally. Just everybody was leaving the market and just about everything was going down. We've had a lot of sort of crummy and crappy stock rally, if you will, lately as during the phases when the AI trade has stumbled and almost this negative correlation. I'm sure you're picking this up and I'm sure you have some observations as well on all of this.
Speaker 2Yes, the market concentration in itself creates this weird phenomenon almost like the tip of the iceberg. I mean, it's a big analogy where you only see the top 10 percent of the iceberg you see and then everything below it you don't see, but that's really the big mass. And it's kind of the market we have because on the surface, everything looks calm and it looks okay. But underneath, we've had a massive route actually in the U.S. equity market. But the market concentration distorts this. And you're right, there are a lot of similarities and to some extent there are some more extreme things going on this time than you didn't have during the dot com. Then there are also differences that are better than what we had during the dot com. But if you take the concept of a beta, so you can calculate something called a beta where you can, which is basically essentially how much does a single stock fluctuate with the market? So if you have a beta of one, it fluctuates one to one with the market. So the market is up two percent, the stock is up two percent and vice versa. If you do that calculation across the entire 500 stocks in the S&P 500, then at the peak of the dot com, you had 50 stocks, so 10 percent of the market that had a negative beta. So basically the stocks were doing the exact opposite of the overall market, which in itself is extremely weird. This time around, during this AI trade, we have 150 stocks in that category. So three times as many stocks are behaving opposite on every single day, not every single, but on an average trading day, they do the opposite of the market. And I see it myself in our own portfolio. It's a very strange market. I was obviously not. It's not very old during the dot com. I was only 15 years old. So you have a better memory of what happened during the dot com. But you can see from the data points that there is a lot of structurally the same type of dynamics that you had. And I think it's a classic euphoria trade and singular trade where you have this one particular technology or event in the market that everyone is chasing at the same time. And then basically they're throwing out all the other stocks out of the window and they're not really interesting.
Speaker 1Yeah. And I wonder if it's sort of the highest. There's these so-called pod shops across the U.S., which are basically, I guess, sort of actively managed outfits, probably operating on some degree of leverage, as well as just all the general hedge funds and market majors that are market makers that are probably front running a lot of these flows or trying to observe what's going on in an algorithmic sense. But I agree. And I'm just wondering if it's this active overlay of the long shorts. So on any given day, the long short crowd that actively manages, because let's be honest, most of the most retail traders and most are many investors institutionally. They're just sort of passively long in the market. But on any given day, they're either getting barbecued or they're having a really good day because their longs are performing well and their shorts are performing well or vice versa on the opposite days. And I wonder if it just creates some kind of volatility, general volatility risk. And when I said, by the way, that much of poor quality stocks had rallied, that's only very locally here. Your general observation, of course, is spot on. And we've talked about that as well. There was at one point when, oh, gosh, I can't remember the numbers now. Is it 300 and something of the S&P? Stocks were like 10 percent or more off of their after their 52 week highs. You know, some staggering figure when the market was essentially posting market cap weighted was was posting an all time high. So, yeah, super interesting stuff. I think I'm taking a lot of your time here. So I'll get on to the last couple of topics before we close this out. I know you're quite keen on and very knowledgeable about the cybersecurity stocks out there because it's something you're quite active in. Just wondering if you had any. Observations after this recent sort of rash of cybersecurity incidents, I don't know, some of this is sort of hybrid warfare, the supermarket hacks, or if this is some kind of, you know, somebody just cutting loose with a bunch of cheap a driven agents that are doing this. There was this PHP shell incident, which which looked pretty remarkable. Good or bad, Peter, do you think for cybersecurity and any other general observations on what you're seeing in the cybersecurity space?
Speaker 2Yes. Step back a little bit, I would say that the very nature and the evolution we've had in the AI industry in just the past four years is a constant dynamic competitive situation where we started out with open eye, open AI being the front runner, then, you know, anthropic with that cloud idea. So an opening in the market in enterprise sales on the programming side was a genius idea. Then suddenly you had Gemini from from Google taking a lead. Then they had the meta coming on with their own weight model. And lately they're. They're AI agents, the muse. So the leadership is very dynamic. I think it's very, very difficult to predict who who's going to be the winner of all of this. But I would say that there's one particular layer right now, I think, that stands out as a very, very interesting case where I would say, oh, no, you can never be absolutely certain. But cybersecurity has changed forever because of AI and the demand profile and trend from here looks extremely, extremely. It's probably the single biggest risk any company of of a decent size in the world is running right now if you're not protecting yourself, you're also seeing in the the order backlog of a lot of these lot of these cyber security firms and they're growing very fast. I mean, it's it it's it looks like it's an it's an industry where this will be a significant double digit growth business over the next decade. It could be depending on whether from the. From the from the bear case to the bull case, but it could be somewhere between 10 and 20 percent compounded growth rate because the industry is actually not that large given the size of the problem. So I think there's a lot of room here and I will warn a little bit because we do have a position in our fund in a cybersecurity stock and to be transparent, it's Palo Alto. So Palo Alto is up together with all the other cybersecurity stocks massively this year, but they have also taken a beating lately and the valuations are getting quite stretched. are getting quite stretched to say the least in this industry but i think that people are generally willing to underwrite these companies even at very high valuations because the the um the demand profile and the revenue growth over the next decade looks not certain but it looks pretty promising and it looks like it's one of those things where you say it's it's difficult to see how demand would slow down given you mentioned the the phb to show um incident and you will probably see a lot more in denmark you know you and i are located here in denmark denmark as a country has has elevated the cyber security risk to the country to the highest category i i assume that is probably something that is taking place in other countries as well so this is very high on very high on the agenda.
Speaker 1Yeah, it is. And I was actually just to add, by the way, Palo Alto is reporting earnings next week as well. I meant to do that in the, well, I will do that when I preview the week ahead a bit later after our conversation here. But recently I spoke about this F5, what are they called? Just F5 is the name of the company. It's super interesting. The company, they have actually hardware that can some of this packet sniffing it can kind of tell what applications you're using and there's a bit of an ai angle there suddenly this company and it exploded higher there could be a cyber security it's not again it's not a recommendation it's just starting to seep into different interesting areas of companies that may not have been considered the main cyber security players then i just you mentioned you mentioned alphabet there or at least one point you mentioned gemini and i remember you and i were when you were still uh side by side we were side by side on the trading floor you were my colleague we were talking about in the early phases of this um you know the chat the the chat gpt the the models that we were starting to interact with and noticing how much more of our time was spent interacting with ai and doing research on ai rather than uh google search we thought gosh the eyeball time is going to get distracted away from google search and somehow the that that was way too early but we're talking three three years early at least probably on that observation but now suddenly this this most recent earnings call we saw alphabet noticing or noting the weakness of google search do you have any observations uh around that i i think it's a significant concern concern for uh for the company's future at least that that sort of search cash calendar they've they've had for so long
Speaker 2yeah i mean i remember when we talked about it the first time i i always thought it was obvious that this would happen uh to their search business and then i took the clues from my own um behavior with these new uh lms uh doing finding myself doing almost no google searches any longer basically um and maybe i'm an early adopter so so obviously you you have to be a little bit careful there but i always thought it was it was an obvious thing that would happen for google to google and i have some family working at google and i asked about whether they were concerned about this and it seemed like it was a big concern one year ago and i thought it was odd um and but you could also see in the quarterly reporting figures that initially it was not a big deal but you could you could see the numbers and also from from you know conversations and comments from their third-party providers and also where they sent traffic to that they were seeing different traffic patterns and there was some upheaval there and then they rolled out the whole gemini ai search function as part of the core search functionality and the problem that we have seen is that it's very very difficult nobody has cracked this how to place ads in in our chatbots and i i think increasingly it will be a problem for google um and it is their cash cow and they're the they know they basically have depleted all their operating cash flow to to scale up their their ai data centers and compute so it's not a good thing if their cash cow is suddenly uh sputtering to say the least and i am a little bit perplexed why they were so eager and rushed into pulling the ai chatbot feature over their old google search uh domain and business couldn't they have split it to mitigate some of that risk because they but but i think at the end of the day they probably came down to the conclusion that the google search engine had the most distribution of all technologies probably in the technology industry besides the iphone right so they basically say okay if we're going to win and there's only going to be one or two winners in ai this is what we need to do but it could turn out to be the mistake of the century and especially this whole idea of ai on the edge with meaning that you have the the chips on your iphone being powerful enough to run pretty sophisticated and high um high parameter enough models that you can do most of your your ai locally on your on your device on edge then that would create an even bigger severity for a bigger uh threat to the the whole google uh business model on the search side so um i think we're in for a very interesting time in uh in google and i think this whole ai i mean you and i have always thought about this and we know that from from history and capsulism that when you have these big companies it's almost unimaginable how how will they be defeated they're so powerful they have all this they're all this profit and then always some new technology comes by that changes the whole landscape and one of the companies i just want to it's a little bit of a side story but just quickly i i've been reading up a lot we don't have a position in the company but i find the business extremely interesting shopify i've been reading up a lot we extremely interesting shopify this new assistant uh ai assistance and the shopping assistance so just for clarification i mean if you look at the e e-commerce growth picture in the us more than 50 of the growth right now is actually um going to shopify so so some so amazon is actually losing out and one of the things that you're seeing here is back in the days when the internet came everyone thought that i would democratize and and play into the little guy the little business uh that didn't happen so what google's search rank actually did was it catapulted and created enormous market concentration by sending people to the biggest sites including amazon and what we're seeing right now with ai agents is that they are not favoring per se the big guys so you right now you're seeing 2x 3x 4x larger growth in niche category sites versus amazon so i think this again is a very disruptive for some of these very big tech and again i think it's important thing to to think about as an investor especially if you're passive i mean because what you're basically betting on if you if you're a cap weighted investor is that these uh large mega cap technology companies will continue to be the winner and they will never be disrupted but i think we see more and more evidence that that's maybe is what's going to happen that they will be disrupted in very very peculiar ways that we didn't think imaginable just four years ago that's a super
Speaker 1interesting observation it's funny you mentioned that on shopify because the shopping environment here is much worse you know fractured in denmark than it is in a place like the us where you get this you know single market provider like amazon arguably for somebody that wants to do e-commerce because you have lots of little providers here that do a great job and you select based on customer service because everybody or most people have prices that are within uh 10 of one another you choose the one that has uh uh you know well the selection you like but then they they you know that they're gonna deliver on time they deal with returns well this kind of stuff uh very interesting uh but not least i wanted to uh oh by the way you we mentioned this soft private conversation we mentioned the softbank story with a write down on the data centers well i forgot to get to that back in our conversation around data centers i will provide a link to that in today's podcast episode description but i think it just maybe briefly if you mentioned something about you felt like and i haven't read the article yet because i've been too busy putting together my notes this morning uh that this was a sign of of this issue around around the power and the grid being the bottleneck on what's what the story
Speaker 2is yeah so softbank has probably was the um was the initial central bank of the ai infrastructure trade but then as uh as mr sun out of japan not ran out of money but was depleting his watches quite a bit with his big commitments in open ai which is probably his single largest bet ever uh nvidia took over but but what softbank did besides doing a lot of equity investments is that they they help their their companies that they had invested in and and they created these massive forward commitments to um to acquire compute you know acquiring uh you know power equipment and and grid connectors etc etc they actually have a very large entity in the us that they're contemplating uh ipoing or doing ipo on because it's it's a very large uh business and it's all about you know getting the energy getting the grid connectors get the transformers basically the infrastructure around getting the data data center line and the write-down was basically that a lot of these assets are just sitting there idling and and it's so so the softbank write-down was just a very concrete evidence where the bottleneck is and i think if you couple that with the tectonic change we saw in the us and the political landscape with the uh the midterms you saw the meta the meta chief for data center building coming out i think was a day after maybe it was two days after the midterms election basically and so i think it's it's it's been a really very long time since we've seen the data center build out i think it's been a really long time since we've seen the data center build out and i think it's it's been a really long time since we've seen the data center build out and i think it's it's been a really long time since we've seen the data center build out and i think it's it's been a really long time since we've seen the data center build out and i think it's it's been a really long time since we've seen the data center build out and i think it's it's it's been a really long time since we've seen the data center build out and i think it's it's been a really long time since we've seen the data center build out and i think it's it's been a really long time since we've seen the data center build out and i think it's it's been a really long time since we've seen the data center build out and i think it's it's been a really long time since we've seen the data center build out and i think it's it's been a really long time since we've seen the data center build out we need to adhere to the epa rules on on we need to adhere to the epa rules on on we need to adhere to the epa rules on on pollution noise pollution uh water recycling you need to on pollution noise pollution uh water recycling you need to maybe investing in local schools and maybe investing in local schools and maybe investing in local schools and infrastructure infrastructure infrastructure you you know you do the cost for you you know you do the cost for you you know you do the cost for building new roads etc so i think we'll building new roads etc so i think we'll building new roads etc so i think we'll see a different dynamic and that will see a different dynamic and that will see a different dynamic and that will slow down things even further so if slow down things even further so if slow down things even further so if you're a softbank investor i think that you're a softbank investor i think that you're a softbank investor i think that would be one of the key risks you need would be one of the key risks you need would be one of the key risks you need to uh to to reassess because how much how to uh to to reassess because how much how to uh to to reassess because how much how many dollars many dollars many dollars are on the line for softbank if things are on the line for softbank if things are on the line for softbank if things and slow down even more
Speaker 1and slow down even more and slow down even more absolutely cool i'll again i'll put absolutely cool i'll again i'll put absolutely cool i'll again i'll put that link in the podcast that link in the podcast that link in the podcast episode description all right peter episode description all right peter episode description all right peter you've been incredibly generous uh with you've been incredibly generous uh with you've been incredibly generous uh with your time today maybe just close it out your time today maybe just close it out your time today maybe just close it out i often talk about any random book that i often talk about any random book that i often talk about any random book that i'm reading especially of course if it's i'm reading especially of course if it's i'm reading especially of course if it's relevant to financial markets relevant to financial markets relevant to financial markets and uh whenever i get you on you're a and uh whenever i get you on you're a and uh whenever i get you on you're a big reader big reader big reader and i'm curious if uh there's anything and i'm curious if uh there's anything and i'm curious if uh there's anything you'd care to pass on about what you've you'd care to pass on about what you've you'd care to pass on about what you've been reading lately
Speaker 2been reading lately been reading lately and for some art one observation that i i have massively accelerated my my podcast listening um so i i think it's a great media and that's also why i i love coming on on this podcast with you john and to talk i think it's a fantastic media because you can consume it in parallel while you're driving or walking in the in the woods etc but but in terms of the books um i'm still trying to finish up this uh um this uh biography on and abraham lincoln and the the whole u.s civil war and what happened there and then i've picked up a book i i read a long time ago and the reason for that is called titan and it's a biography of john d rockefield and the reason why i did that was because um i i listened to a podcast about charlie munger and basically in a in a conversation said that he was the greatest businessman of all time he was uh you know you can say a lot of negative things about him but there are also some ideas and the way he he solved business that was quite unique so i've actually picked up that book again so that um i'm really i'm really a history uh history nerd there i i find it very inspiring and i always try to when i read these stories try to to reconnect it to things that are happening today or something i can i can do as an individual and uh in the world we're living uh today because there are always learnings uh you can you can learn from history
Speaker 1absolutely and also the how we learn from history and sometimes overlearn from history and that's what i've emphasized because the book that i recently completed uh very rapidly because it's an audiobook and i was just really uh it was very riveting was the book 1873 so it was basically this guy uh liakat ahmed i think his name is who wrote this incredible book um the title which i'm forgetting right now about the uh basically the whole the gold standard and the whole you know need to get off the gold standard around the great depression and really the origins of that being the 1873 uh really the first depression and the first depression and the first depression and the first depression and a lot of it was linked to this gold standard and how just unbelievably crushed the economy was by this orthodoxy of sticking to a you know a metallic currency when there'd been a credit boom and then the horrible hangover from that and then the lessons we've overlearned from all this in our current market cycle since the great financial crisis and the way japan has done qe and everything else i'm going on too long here i'm not trying to entertain you or shouldn't be trying to entertain you but i think i just wanted to i agree very much with these um these deep historical lessons because everybody who's a policymaker reads those and they try to learn the lessons but then they're even sloshed back too much the other way uh and how they apply uh some of those lessons so really critical stuff all right peter awesome to have you on the podcast thank you so much for your time and um how can how can folks get in touch with you or find out what you're pretty active on linkedin i know and i'll put a link uh to your linkedin profile or at least the the jesda profile if not yourself but uh yeah and i'll see you next time anything you care to say about yourself or your organization of course
Speaker 2yeah yeah yeah of course so so i i'm i'm pretty reactive it it comes in cycles um sometimes i have a lot to work so then i'm not there or i and i don't want to just do a comment for the comments sake but i i i do comment on the market and and topics on linkedin so you can watch me there i left twitter long time ago so i'm i'm i'm blue sky um where i i do a lot of commenting on the market and the economy and sometimes also our friends and family and stuff like that so i'll see you next time on linkedin okay thank you so much for having me thank you for having me thank you for having me thank you for having me thank you for having me thank you for having me thank you for having me thank you for my big pain point here is my sub stack or the company's sub stack it is now um i have this long list of things i want to do i'm working on a on a new piece it has been a while because we we're taking in so much business and onboarding clients and doing you know portfolio advice it takes quite a bit of time but i have the next piece that i'm almost done with is called the capital waterfall which i think is uh is going to be quite an interesting piece on the decision making for uh for company uh management and i think it's going to be quite an interesting piece on the decision making for company uh management how you do capital allocation and i i try to do some myth busting in in that piece as well so i hope to get it out soon so sub stack blue sky linkedin um i and i'm always open for uh questions and people can reach out uh i i think i'm pretty generous uh with my my time um to to help people um if they have questions
Speaker 1great stuff you have been generous today and of course do ping us so when you do have that substack piece out and i'll send you the links in the in the podcast episode uh description okay cool um thanks again peter thanks for taking so much of your time today and uh certainly hope to have you back on the podcast again in the future thank you so much john all right take care all right i hope you enjoyed that conversation i certainly did and that is certainly a wrap for today have a wonderful weekend when you get there do realize that next week could provide considerable fireworks stay careful out there and we'll be back next week with the next saxo market call this has been the saxo market call podcast thanks for joining today's episode we're always happy for your feedback and questions of all kinds to reach out you can drop us an email at market call at saxobank.com that's market call at saxobank.com saxo serious trading worldwide [email protected] you 
PodcastTranscript AI