Loading summary
Transcript37 lines
- [00:00]
A
Summertime is maybe good. Summertime it may be. It's Macro Mondays. Level up your week. Oh, yeah.
- [00:23]
B
Hello out there. Welcome to Real Vision. Welcome to Macro Mondays. My name is Miguel Rosemal, I'm your usual host and I'm joined by the returning Andreas Steno back from vacation in France. How are you, Andreas?
- [00:37]
A
Good, it's nice to be back. You know, the temperatures here fit my, my weight a little bit better. Let me just put it like that.
- [00:46]
C
Yeah.
- [00:48]
B
And maybe the momentum in markets as well. Andreas looking decently this this morning. Andreas, let's dive straight into the topic on everyone's agenda this weekend, the Kimi scare. I want to get your take on it. You wrote an entire piece on Morning. Has China taken over the, the leadership in the AR race?
- [01:07]
A
No. Next question. I, I mean of, of course it, I, I think it's the first time that we've seen China leading the race in one subcategory in the Arena AI scoring methodology. But they don't lead the race overall, but they have taken the lead and in a subcategory related to front end coding, which I guess is something that we need to consider, especially when we look at how markets should respond to this because I'm not necessarily sure that I agree with the market takeaways from say Thursday, Friday last week, and I basically think the market got this wrong. But we can spend a few minutes on that, Michael.
- [01:52]
B
Absolutely. Andreas, we had these reader school results that, that Kimi K3, the new moonshot model, was essentially the best out there. This caused a whole wave of panic about China taking over the leadership, driving this and a lot of confusion as I saw it on social media and elsewhere around. How do you even play? What does that even mean? If China takes over, who's the losers? Who's the winners from that? So first of all, maybe let's, let's dive. I know, I know you like to dive a little bit into the, the, the actual fabric of this is Kimi the. How much does this matter?
- [02:29]
A
So you know, my preferred third party provider of say artificial intelligence or scoring of artificial intelligence is the provider called artificial analysis. And I think we have a chart on the scoreboard from, from that on page eight. You know, again, yes, if you look at the front end coding from arena AI, which is by the way an old spin out of UC Berkeley, the Kidney K3 model is ahead of Clock Fable, for example. But if you look at it on text, which is basically the broadest use case for many, it is still behind, which is what you would Expect. Right. And if you look at it broadly speaking, from this Artificial Analysis Intelligence index, it is also behind both Club Fable and ChatGPT 5.6 Soul. So are they ahead? Well, in a, you know, in a niche corner, yes. I guess Front end coding is not the main use case for most users of AI, but it's of course an important use case and, you know, top down. I think this is the biggest issue for Anthropic. If you look at the, the average user of Anthropic, they're probably more prone to use Claude Fable for front end coding purposes than the average user of Chat GPT, just to put it in very simple layman terms. Right. So, um, in a niche use case, yes, there is an issue here, potentially in a broader perspective, I, I still see a lot of issues with the notion that China is ahead. Let me just again paint a very simple picture on page nine, Mikkel, because what I did was that I looked at, you know, the overall scoring of the best Chinese model relative to the best US model on all scoring parameters. Right. The Artificial Analysis index again. And nothing has changed. Basically what we typically see is that a new flagship model is shipped by anthropic or OpenAI. The gap opens again. So China is say 5, 10 percentage points behind. Then we get a new flagship model from China catching up, but never getting past the frontier parity goalpost. And, and it's essentially the same that has happened here. The true shocker was when Deepsea got to like 95% of the capabilities of the flagship models in the west in early 2025. They've never really gotten anywhere closer than this, say, 5, 7 percentage points off the frontier model. Right. I think there's a really, really important conclusion hidden in this chart. They're always behind, and I think there's a reason why they're always behind because they obviously use the frontier models to train their models. That would be my best assumption. Watching this, we can get into the weeds of a long technical discussion here. That's not by any means my intention. The point here is that they're always slightly behind. And I actually consider this a pretty fruitful setup in many ways, because if you have someone chasing you like one month, two months after. Right. It keeps you very attentive, it keeps you very focused on shipping the next bottle, et cetera.
- [05:55]
B
Right.
- [05:56]
A
So from a hardware perspective, let's start there. We probably need to see the US administration banning the Chinese competition to really take a beating on that hardware trade, in my opinion, because as long as, you know, they're allowed to, you know, follow the, the, the U.S. models with a lack of one or two months. Everyone needs to invest still, right, to, to, to stay ahead in that race. While if you kind of fence off the competition, I, I, I kind of consider that setup a little bit more, you know, laid back in a sense. Right. You don't need to be as focused on staying ahead if, if you start banning competitors right, left and center. That's for the record, we haven't seen that yet. So what I'm trying to say here is that sure, we probably need to see repricing of Anthropic because of this, due to this exact front end coding niche case. I think we need to price hardware higher as a consequence of this and the very simple reason why it took moonshot maybe 30, 40 hours to explicitly state that sorry guys, we're running out of GPUs, we simply cannot match the demand. And I can assure you that in terms of numbers of users, they are millions and millions of visits after ChatGPT. And those models still, I mean, way, way, way, way behind. I think we have that leaderboard on page 11. It's of course not updated up until Monday. Right. But you know, in terms of users, there is no doubt that the western models have many, many more use cases. And therefore, if we get some sort of a catch up in, in the usage of Chinese models from a user perspective, they need to get hold of some hardware. I know that it's tricky for them to get hold of Nvidia chips or Hynix chips, et cetera. Right, because they're essentially banned. But yeah, who knows how they get them. But they do get them to some extent, at least they have managed to historically.
- [08:02]
B
So the logic here, Andres, as long as the frontier is being pushed, as long as we are getting more productive, more effective, also it might hit the margins and it might hit the market shares of some of the big players like Anthropic and OpenAI, but why on earth should this degrade the memory, the chips story? That's what I don't understand here, Andreas,
- [08:22]
A
at least not for now. And I think the market got that wrong last week, but we'll obviously have to see. I think it's a bigger issue for software than hardware for now. If it's actually true that China is able to provide intelligence cheaper, it should increase the hardware demand from, you know, the standpoint of something that gets cheaper will see more demand. Right. But having said that, I don't think there is, you know, again, watching this leaderboard in terms of users. And you can laugh at meta AI having a, sorry, a thousand million users, but they do have an incredible distribution base via Facebook, WhatsApp, etc Gemini. The same. Right. Most people don't know that they're using Gemini, but they are using Gemini every single day when they use Google search, etc. So this, you know, this distribution embedded in apps, in web pages etc. Is second to none. And the China obviously does not have that distribution. Because you know who's using Bite Dance in the west to some extent we do use TikTok. But still even the US lack of TikTok is now being put under control of the US government, etc. Right. They're at least trying to keep it away from China who's using tencent apps. If you're a Westerner, you don't do that. Right. So all of the consumers are still Westerns and therefore I still struggle with this notion that this is going to know wreak havoc on the AI race because the distribution will eventually turn out to be equally as important as who's got the best model. I mean, who. Let me put it like this, Miggle, do you care whether Gemini three point whatever is the best LLM when you use Google Search and you get some LLM results from that? Probably not. Right? And no one cares really. Right. The average user does not care about the frontier. The average user care about the flexibility, the agility and the distribution. Right. They wanted, they wanted to work where they are and they're still using Western apps, etc. So in that sense distribution will be equally as important. And the final point I'd like to make Mikkel, I know I'm rambling now, but on page 10 again there are probably other ways of showing this, but if you look at what humans actually prefer to use, so there's called an ELO rating on this where you actually ask you when what they prefer to use. Right. Which is more of a, you know, of course you can ask a front end coder. Do you think Kimi AI or Anthropic AI is the best? Right. You probably get a different response from the peer group here. But my point is, you know, having spent countless hours setting up stuff on cowork on my vacation for example, that user experience is also a really, really important parameter. And when I've, you know, I've tried to gain access to keep me over past couple of days for now it's basically impossible. But my first impression was that the user experience was not good and that is probably in my opinion, the most valuable part of Anthropic, it's not Claude Fable's best model characteristics. It is their user experience, which is a lot better than a lot of the competitors. Right, Absolutely.
- [11:49]
B
Andreas, fascinating stuff here. We will dive a little bit more into this and also obviously take a closer look at the situation in Iran after a little intermezzo here. Andreas, did you watch the World cup final yesterday? Andreas?
- [12:06]
A
I tried to watch the whole thing, but I actually fell asleep. And it is the honest answer.
- [12:11]
B
It was horrible. It was horrible. Nevertheless, you asked me to bring along this video if we can get it played here of. Of Donald Trump and obviously the. The victor. And congratulations to Spain. Let's see if we can get the video played here.
- [12:28]
C
It's okay. We don't want anything to do Watch them, watch them come running back. Oh, they'll come running back they're hopeless. I don't want anything to do with Spain. Cut off all trade with Spain, please. Including visits. Okay. We don't want anything to do Watch them, watch them come running back. Oh, they'll come running back they're hopeless. I don't want anything
- [12:50]
B
Watch them come running back. Well, they did and they ran away with everything. So big congratulations to Spain. It's been an incredible tournament. You know, Andres, I've been incredibly well entertained by this especially, especially the latter stages. I quite enjoyed all the show around the match yesterday. It wasn't what we used to from European matches and tournaments, but it was. It was quite enjoyable for me. It was a little bit bizarre at times, but it was quite enjoyable. So much more enjoyable than the game anyway. So, yeah, I hope all this is from the US Enjoyed it.
- [13:20]
A
It wasn't a tough benchmark to beat, but you know what I mean. Did you notice that, you know, through that halftime show, I don't think there wasn't any American artists present in that halftime show. Justin Bieber is Canadian. Chris Martin is from.
- [13:40]
B
Donna is from England.
- [13:41]
A
Yeah, yeah. GGS is from Korea.
- [13:45]
B
Right. Think about that. Yeah, yeah.
- [13:46]
A
It's not a single American. I was a bit surprised to see that. And you know, the Chris Martin stuff at all of that. Oh, I, I do not address.
- [13:58]
B
I have a personal rule of myself, a rule of thumb, the Tom Cruise rule. If you're ever in doubt. About what, which movie to watch, just search for Tom Cruise and you'll. You'll end up with a decent, decently good movie. And yeah, for me, that was a real high point. His speech yesterday, I didn't really understand what his point Was, but it was really, really good as it usually is with, with Tom Cruise. So anyway, that was a bit of a high point for us. Andreas, looking ahead on the week that just started. Andreas. And we'll get back to Iran in a second here. What's the stock that you're watching the most? What's your most. What's the most important stock to track at the moment?
- [14:40]
A
So I don't think we've seen anything like it in momentum stocks than what we've seen over the past two and a half weeks. We're talking about the biggest sell off in the stuff that has worked so far this year ever in, in a single month. And this is probably the stuff that we typically see in times of crisis. Right. So are we in a crisis? Are the wheels coming off in a way. I mean that's basically what I'm watching now and what I'm concerned about. I don't think so. But I need to be attentive and vigilant to the risk of this because it's been happening very, very fast. As the Iran war restarted, as we got several triggers for this sell off of popular trades. Right. It's been incredibly nasty in everything related to the Korean casino. Let me put it like that. As far as Goldman Sachs could count, I think 3 or 4% of all Korean adults received the margin call last week. That's incredible that, that's the most bizarre stat I've ever heard of. I think 300,000 accounts were completely wiped off. So you know, you could put crypto leverage to bed. I mean they're crazy out there. So I've never seen anything like it. And you know, the buildup during May and June of these levered ETFs in Korea was, was absolutely bizarre. But it was a little bit difficult to follow that real time. But, but you know, we basically know why now, why, why is things moved so fast in the other direction once you got a trigger for a sell off? So I considered more technical of nature this whole momentum route, if you, if you will, than fundamental. But we'll have to see because obviously the restart of the war is not good news. And you know, I can also slowly but surely see that you get a little bit concerned with what's ongoing in Iran. So I'll allow you to unpack that a little bit.
- [16:44]
B
Yeah, I have to say so Andres, I mean we, we've had the eight or nine days of escalations step by step, a cycle of escalation. We are essentially back to the beginning of the war more or less. But still, we are at a point where I think the US Is struggling to find stuff to bomb. I know that sounds a little bit stupid because Iran is a huge country. They've claimed, the US have claimed on multiple occasions that they have essentially destroyed the Iranian military and the Iranian offensive capabilities. We can see now that that has either been a lie or a complete misjudgment by the US Military. Iran is very alive and very kicking, we have to admit. So it is, it is a very, very tough situation. And we are also seeing the attacks spreading just, it's no longer focused around the Strait of Hormuz and naval assets. We're seeing renewed attacks on desalination plants, on oil infrastructure, on military infrastructure, and we are seeing increased attacks on the Gulf states, not only US Bases in these states, but also directly on infrastructure in Kuwait for first and foremost, but also threats against Dubai and Abu Dhabi and Qatar potentially as well, even though I think they're slightly better positioned. So just what's happening here, as I discussed last week in my the Drill article, and I don't know how visible this chart is, but what happened after the Memorandum of Understanding was that the Iranian Republican Guard were essentially the ones that the U.S. is fighting right now, the IRGC, they took in, by my estimations, at least $2 billion of oil revenue that they control about 50% of of Iran's oil exports. They took in around 2 billion. Some pundits have, have claimed this is 5 to 6 billion. In any case, they took in a lot of money within two to three weeks of the Memorandum of Understanding. And then things began to, you know, revert to the mean. When we entered four or five weeks after the Memorandum of Understanding, and they probably decided, okay, let's restart the war. We've got our war chest filled up again. We, we can pay wages to our operatives for the next two to three months. Why don't start all over again. They've also managed, obviously, to rebuild some of their inventories on missiles and drones. So where does that leave us at the moment? Well, it's a situation where the Iranians, the irgc, they're not really under a lot of pressure to stop fighting. They want to fight. It gives them leverage in the internal power struggle in Iran. The US is obviously fully committed to fighting this out probably for a few more weeks. And, and the only solution I see right now is, is, is, is still a negotiation. It will probably have to have to involve China at some more or less obvious level to, to, to get this done. They are the only ones who can truly put the screws on the Iranians, as far as I see it. And yeah, yeah, and they've just been
- [19:55]
A
accused of, you know, everything last week and that presser. Right. So, yeah, I mean, it doesn't sound like Trump is, you know, trying to come up with a conciliatory stance versus Beijing to get them involved in this, but I'm, I'm not sure how to read that, to be honest. That whole press conference last week, that's good.
- [20:17]
B
I have quite a bit, quite a different. That was my initial reading as well. Then, then I read into, what's his name, the Commission on Trade in the US Administration.
- [20:25]
A
Peter Jameson.
- [20:26]
B
Clear. Yeah, no, I don't think it was him anyway. One of the trade representatives. Yeah, it doesn't matter. From the, from the Trump administration who said this is not about China at all, this is not a China story. Well, Donald Trump just spent a whole press conference talking about Chinese interference. Apparently this is not about China at all. And I think, I think we need to take that at face value. This is all about getting the Save act passed and this is all about setting up the scene for the midterms and controlling what happens at the polls at the midterms. So I think they've been sending signals to China from back channeling and more or less overtly that I know we're talking about China, but this is not about China. And they're so far keeping the September summit alive and the vibe is okay, they might be able to still work together on oil, hopefully also other issues, but it is very, very delicate balance right now. Go ahead.
- [21:25]
A
Can I just add on, you talked about the RGC seizing that window of opportunity to get some oil revenue in. And for example, JD Vance explicitly referred to the same. Now we have a 60 day window, we can refill our coffers that we'll have to see whether they're cooperating, etc. So I mean, both sides have basically used that. Yeah. Ceasefire to, you know, reclaim some, some, some territory on oil, if you know what I mean. It also means that the first deadline that we had on the drawdowns of inventories in the west has been postponed. You know, I, I, I initially said that, you know, late August, early September would be nasty. Now the strategic floor is probably breached, say between October and November if we keep current flows alive. So right around the midterms, timing ish. Therefore, you know, the US has bought itself another month, month and a half of pumping without too many consequences. Right. So, you know, we, we Also seized that moment to, yeah, take control of the oil flow in a sense. Right. So I guess both, both sides have some patience now due to the revenue on the IRGC side and the oil transported to the west, all of the stock transiting tankers and all of that got to the west during that window, et cetera. Right. And we'll see that over the coming weeks in the inventory numbers, by the way, they will tick up. Having said that, I still think that this war moves, pardon my nerdy language here, moves like a sinus wave, if you know what I mean. You have some extremes in both ends, but there is some mean reversion towards the end of it because both the Iranian side and the US side will be forced to reconsider at some point in a month or two from now because the oil math looks decent right now, but it will not look decent in a month or two from now again. And both sides have that same issue. So they do actually have an issue in common to some degree that they have to face during the autumn. So I'm still of the view that this will be substantially less panicky if you want for oil markets, for equity markets, than round one because we know what we're dealing with now and markets will never discount the same event twice to the same degree. So I consider the risk of oil above $100 very low. I'm not saying that it's completely ruled out, but yeah, let's see. I think we're approaching some sort of climax within say the month of July, mid to mid August, something like that due to this very simple oil math logic. Yeah.
- [24:22]
B
And we are consistently hearing the US telling us, the US administration telling us that Iran wants to negotiate, we hearing stories about mediators. So there is always some kind of contact and we're always just a post on truth social away from, from everything de escalating here. But I agree with you Andres. The one thing that could speed this up, that could change the, the sort of the equation for the US side is if they begin to stack up losses. And we have seen that over the weekend, some of it accidental, but that is the one, one trigger that the Iranians have if they can cause US casualties. Iranian casualties don't really matter. No one keeps track of those. Sorry, it's a bit cynical but that's the way it is. US casualties matter a lot. If you begin to have triple digit US casualties, then the equation changes, then it's no longer about inflation spikes, then it's about boots, then it's about bodybuilding getting Back to the US So that is one thing that can change the equation, the time perspective. Otherwise, I agree with you, Andres. I mean, oil is at $86 or what is it right now. Still we are seeing Greek shippers trying to force their way under fire out of the strait. Even at $86. Imagine what they were willing to do at 105. So that is obviously also the dynamic here, Andreas. Once oils gets close to 100, we saw that last time around, people become very, very, very creative and very, very, very risk tolerant if there's that payout at the end of it. So to sum up a bit here, Andreas, I'm worried about the outlook right now. I'm worried about the US From a strategic political standpoint here. Not as worried about the oil markets. Even though we, obviously, we're not going back to 50, 60 anytime soon, know, and still we're only, we're only a tweet away from this whole thing being dissolved.
- [26:19]
A
And, and that's, you know, I wanted to, to, to say that as well. I, we got out of energy longs during the first week of April with an amazing timing, early July. We said that the risk reward was terrible and being short energy again. So that was okay. We never got long energy. And I'm not really tempted to be either long or short here because, you know, it's too much of a coin to us. Right. Do we really have an edge in terms of predicting Trump's next tweet? And we haven't got the money to buy fast access to Truth Social, which they're now selling. So, you know, insider trading as a service is now a thing in the US it saddens me to read, but having said that, Michael, I agree with you that the risk tolerance is already higher among the shipping owners right now. And I guess you can kind of sense why one of the biggest shipping owners in Greece is laying off five or six coaches a year, not Forest and Premier League. So he's got to get some revenue in there to pay all of those coaches for their services.
- [27:24]
B
Yeah, that's a fair point. Address. Okay. Finally, Andreas. Remember, listeners, this is our weekly free show at Real Vision. We have a lot of great content. I think we can maybe get the weekly plan up here. It might not be too visible because there's so much stuff coming out, but that's a good thing. Obviously, if you're into crypto. I know a lot of our listeners are. We have the monthly panel by Sebastian Purcell hosted by the Solana Foundation. So that's tomorrow at 11:00am Eastern and then address. We host the monthly portfolio update macro meets micro at I think it's 5pm tomorrow Eastern. Which of the three themes in our portfolio we're going to take a deep dive into tomorrow?
- [28:06]
A
Yeah, so we're going to spend a lot of time on various semiconductors and also correlated traits to the whole AI build out. Because I just have to admit to that. Even the traits within fuel cells, solar, etc, they've all taken a beating alongside this AI selloff. Right. So I, I need to, you know, we'll take a deep dive into what is actually truly a standalone theme and what's not tomorrow, if you know what I mean. Because in my opinion, and I've basically double checked across all measurable scores, probably 50% of the market right now is the AI build out trade. So if you have less than 50% exposure to that through a lot of themes, you have less exposure than the market. And do you really want to have that? I'm not sure. So that will be the main discussion for tomorrow and of course we'll try to bring a handful of multi backers. I've spent a lot of time digging into the whole biotech space which is pretty interesting at this stage of the cycle. So let's see whether we have a hidden gem in there.
- [29:18]
B
Absolutely interesting. Of course. Remember guys, we have a lot of content, especially on the pro tier real vision. That's where you get the, get all of it. It is a bit, a little bit expensive. So if you're more into the alpha tier, we publish both a weekly macro and geopolitical article there, a sum up of our writings essentially. And we're now running an ETF portfolio that's really, really interesting to watch. So if you're into that, you can get that on the cheap with the alpha tier real vision. So that's something to check out. For those of you contemplating making the jump into real vision, that's for you. Yeah.
- [29:53]
A
I just have to add a comment from our light feed here. Our good friend David wrote when I said that no, no Americans were present during the halftime show that the Muppets are American. So I just wanted to leave that there. Thanks for that comment, David. We need to have a laugh when we lose money and we've lost money the past few weeks here, but we'll rebound, I promise you that.
- [30:14]
B
Absolutely we will. Andreas, that's all we had for you this week. Thanks to everyone for tuning in wherever you were watching. We'll be back next week.