Physical AI Is a $40 Trillion Market — S&P Top 10 Already 40% Concentrated

 Sign up to The Lead-Lag Report on Substack and get 30% off the annual subscription today by visiting http://theleadlag.report/leadlaglive.

Transcript

Welcome And CraneShares Overview

SPEAKER_00 0:00

Thank you, Michael. Thank you, everyone, for joining today's uh webinar. Um, so like I'm a senior investment strategist as CraneShares. So Crane Shares is um kind of like a firm dedicated to a lot of emerging technologies uh and a lot of emerging markets. Um firm funded like over like 12 years ago, um, managed by like$10 billion. Um so for investors who don't know us, well, we have a suite of China Sematic and China Core solutions. Uh, and also we have uh a lot of like emerging technologies now within AI, robotics, uh, emerging market technologies. Uh, and also we have a lot of alternatives uh in the carbon markets and other health strategy. Um, so with those solutions, we think um we're trying to provide more innovative, high conviction ideas to our investors globally. Um, but like, I mean, this is like very interesting timing this year. Um, I think like um probably you already know this, but the the market is really concentrated, right? So uh the top 10 names represent like 40% of the the market, the SP 500. Uh I think similar concentration happening, many other broad indexes. Um so that leads to a question. Uh, is this really justified by earnings? Um so far is not, right? So if you think about like the earnings contribution uh compared to the the weight, uh is largely kind of like uh lagging. Um so which means like the market has very high expectation expectation on the potential earnings in the future. So that is something kind of like I think like expectations quite high. Um so even that like those top 10 names can continue to rise, I think that's kind of like um the biggest decision, right? Like if you own Max 7 or you don't own Max 7, your performance is quite different over the past three years. Um, but going forward, right? So how do we really look at this? Um so another trend, right, very structural trend uh over the last three years is really the AI, right? Uh of course everyone kind of like already knows this. Uh AI has been the gross driver uh for the equity markets. Um, and this is something very structural, right? Think about like over the next five years. We're gonna still talk a lot about AI. Uh we have huge capex, we have uh agenda solutions coming, um, there's a new wave of AI development, uh, and uh likely there's be more opportunities. So we think um the given that just like stage, right? So there's a lot of opportunity potentially uh going like beyond Mach 7, right? Going to more broad AI or other sectors, other geographic regions. Um so how should investor really position uh the portfolio, right? So um just given the current like digital AI um perspective, right? You the Max 7's really concentrating around like hyperscalers, those cloud computing, um, but like they're already like trillion dollar companies, right? Like so for them to become like 20 trillion is is is a little bit challenging, but like um I think the growth pace and expectation is given given that like a big question mark, right? Like um people are questioning about like those huge capex, if that's gonna come with like real RI, or if the expectation is really pricing, a lot of upside. Um, so uh with that now I think a lot of investors looking for opportunities within the ecosystem, right? Um, I think like the ecosystem play uh beyond Max 7 add a lot of value at uh just across the different um value chain of the AI, right? You have um chip semis, uh data centers have a lot of like um kind of like shortage uh on the computing side. Uh then of course you have like a lot of model companies now taking a lot of value. Um uh for example, anthropic, right? Every time Anthropic published a new function uh or new like um uh solution, right? So that's lead to a set of software company tank. Uh I think like that's really shown that a lot of value actually can come from AI, but like in a different places that investor may not own. So uh we think uh at this stage, uh AI is really transitioned, right? So if you look at the past three years, um most people talk about AI as ChatGPT and all those like chat box. But now, um starting from I think last year, right? So this transition is really happening uh in the enterprise. The enterprise is accelerating in adopting the AI, right? So that's kind of like why I think investors they're not so familiar with this because they don't work in a big enterprise, right? Um they don't see how the enterprise works using AI. Um but recently I think like the open claw is really showing um some kind of like the um the framework um that enterprise is using. Uh if you don't know open cloud, um that's kind of like the the consumer app, right? Like anyone can install it um and can give instruction, then open cloud can actually do the actual work, right? It's not really just like respond with like like some bunch of like text, but it can uh operate your browser, operate your email, it can just get things done for you. Um it can automatically like reason, plan, and just like get um finished whatever you want to do, right? So that's like a game changer. Uh uh I've I'd like a lot of like AI thought leaders today think this phase of AI development is becoming like early time of computer, right? So the AI becomes like a computer that like can do actual work. Um, then like people tend to pay for that. So that's like I think game changer from a chat GPT. I don't think like people tend to pay a lot of uh money to a chat box, but people tend to pay a lot of money to get the real work done. So with that, um that's literally another question, right? Like software being re-tanking, um, but like um the older like entropics being dominating in in the space, like will that replace a lot of software? The the answer, the short answer is yes, right? So I think a lot of software is gonna be replaced. Um, but like that's mostly the single purpose, uh, like one UI, right? That's kind of like the most um normal uh software companies. But think about like a lot of those big software um companies, like like they own the workflow, they own the uh the data, uh they already integrated with the um the companies. Uh it's really hard to imagine that uh the big enterprise is gonna simply just use AI and not use those software solutions. Um more likely, I think like just recently Jensen Huang mentioned this is you will see um if every enterprise has like tons of AI agents, right? AI agents like kind of like their digital employee. If like keep adding more AI agents in your business, those AI agents eventually still need to use some of the softwares, right? So they become like a lot of those like software is like seed-based. So they come like using um those software more, right? So uh using their like Photoshop, using kind of like the CRM. So that's kind of like one way to rethink about this. Like maybe some of those companies really are undervalued. Um, but of course, over the time, a lot of those software companies they have to change their business model, right? So previously, if the price is purely based on seed, um, that doesn't work because like um you're gonna have like multiple AI agents or hundreds, thousands of AI agents to work on those software, uh, and you're gonna those software companies are gonna integrate with those AI um uh capabilities like APIs. So that means uh they have to change the business model from the seat-based, right, to charge per user or per human, right? To kind of like um more token-based or like result-based. Uh how many how many work actually is actually done using the software, or how many like results delivered, or how many economic value be generated. So those value-oriented pricing uh is mission critical for this company right now. Um so we're gonna see like in the near term, say two years, we're gonna see this transition. It's gonna be paying for, but create a lot of opportunities if those software companies can actually transition their capabilities. Um the the biggest, I think the biggest challenge right now for investor is those two disruptors, right? Say anthropic, uh, choose to like that's a private, right? So I'm like, I mean, like uh for like SpaceX uh or like um people, like people like been like talking about SpaceX forever, but now finally it's it's going to IPO. The rumors coming, uh, anthropic IPO is uh also like talking. But like just generally, um, there's a lot of private companies, right? Like in the in the US, um, they choose to stay longer in the private stage because there's a lot of capital. So that makes um investor miss a lot of the value creation because think about that, like anthropic uh the latest valuation, a latest round G round is 380 billion. If you like your company are like reaching, say 50 billion, 100 billion, you probably should go IPO already, right? Then that's become part of the SP 500 and Nasdaq. But now they can wait until say a trillion dollar does SpaceX, whereas SpaceX now 1.25 trillion, um, parking 1.75 trillion. So that means like investor already missed a major part of the value equation. Um, so just look at the growth right now. Well, this is a little bit late, but like um given the latest growth of Anthropic, um, we have never seen that adding a lot of uh revenue each month, right? Like billions of revenue each month. That's really amazing, right? Um the company is still in like hyper growth mode, while uh I think like a lot of the value is really happening in the private market. Um we we we kind of like own uh XAI, but now it's like merged with SpaceX. We can like see this um this IPOs coming, right? So for SpaceX, um they they own the launching capability, they own the satellite, they own a lot of these uh starting communication. Um but now with the XAI, they have to combine this ecosystem of um large image model and potentially building database in the space. So all that is a big story, but like I think like for investors, um, this is gonna be likely um the top 10 holdings in in your portfolio because, say, like uh NASDAQ or SP included, then it with that with the current uh target market cap, it's gonna be uh a big position. So we think um that's kind of like one one uh trend is a lot of names uh creating value um is really brought out to many like places and is really happening in the private world. Another trend that we notice is the physical AI, right? So think about like um Jensen Huang mentioned a lot about physical AI this year in the GTC conference. Um he managed there's like$40 trillion opportunity because um with just digital AI, right, it's basically generating uh basically large energy models. But like there's a lot of opportunity out there that we can use the digital AI to empower, enable, right? So uh in factories, in warehouses, in robotics, in the defense, in all kinds of things that is kind of like at this very pivotal point that um the AI can bring value, right? So we see, well, humanoids one of the uh probably white eye catchy recently, um, with all those like humanoid coming out, um, the performance uh compared to I think 10 years ago is really dramatic change. Uh if if you know this um like the um Boston dynamic, right? So that's like they published this humanoid, just barely can um normally walk. But now if you look at like a lot of videos of humanoid, right, like they can they can dance, they can jump, they can flip kung fu. It's like a lot of things like that's really advanced in the hardware, right? So that makes people believe uh with now the intelligence part is improving so fast. Um, there's a lot of top air researchers now building the the AI model for humanoid for all kinds of machines and robotics. We're gonna see a similar GPT moment um potentially happen in the physical world, right? So, with those like say FIFA's uh Warlap or Gym Femme Nvidia, um, there's physical intelligence, there's so many like model companies right now. Um, there's like uh in Google DeepMind, there's like World Models. Those models tend to break through, like I think like in the next three to five years, then we could potentially see a huge build out of the physical AI in kind of like AI happening in our physical world. Um a big driver for this, right, so is really our society, right? We have aging population, we have labor shortage. Uh in our factories, right, is a lot of factories highly automated. Uh they already bought a lot of those automation machines, but over the last 10 years, that trend's been really stolen down because they cannot add more machines, um, and they still need a lot of labor. Uh, simply move things around, uh, unload a box, or just like putting the truck, trailer. So those are things that can likely be replaced by humanoid or just other shape of uh AI-enabled robotics. Um, Morgan Stanley believes um by 2050 we're gonna see a billion unit of humanoid. Um, and that's potentially a$5 trillion market just for the humanoid. Um, for physical AI, well, as I said, like Jensen Huang gave you this 40 trillion market estimation, right? So this is a huge market. This is potentially way bigger to our digital AI opportunity, right? So think about digital AI is already like we have like a lot of trillion dollar company already. But like for physical AI, it's still in the early stage. Um, uh like I said, like, well, we have early stage, we can already see a lot of factories start to testing humanoid, right? And all the kinds of like uh AI enabled robots in in their production line, uh in the warehouses. Um loading, moving boxes, uh, those are things like this huge demand, right, given the labor shortage happening in our world today. Um, the service sector is another one, like in restaurants, hospitals, hotels, um, there's service robots. Um, there's companies doing like testing in homes, right? Like OneX from Norway is testing uh$5 to$99 subscription model for humanoid working in your home. So that's that's really upcoming. I would think like when like ordinary investors or consumers, they see this humanoid, right, as we see uh in the in the research lab, but like not in the research lab, but in their real home, in their real like life. That's kind of like a game changer for uh consumers and investors to wake up to these opportunities. Um so another thing I want to mention for the broadening is really the global, right? So uh AI is not really happening in in the US, um, but there's many other uh countries and regions uh developing the AI models. Um China is definitely one of the leaders uh so far in open source models. Um there's a lot of good open source models now coming from China. Uh even a lot of US uh software companies or AI native um uh applications like Cursor starting using um like Kimi's open source model uh as a base, right? So uh I think like a lot of AI researchers, they already mentioned this is very critical because if you don't have open source models, um it's really hard to imagine we have the next 10 years to be the AI era. Um a lot of like startups, right? The like small team of um researchers, a lot of colleges and universities, they need those open source models to really uh build up the their ideas, right? So given cursor, cursor is like a famous AI coding company in the uh application company in the US. Um and but like they trust those open source models from China and they find it very powerful, right? So um, in terms of the performance versus cost, um that solved a lot of problems for the AI applications. So I think with that, um, this is something like potentially investors are missing in their portfolio as well. Um that China is really closing the gap um with new models called scores and like capabilities. Um so um, but like I think like people know potentially know uh one of our um product called China is Internet, but mostly names you you heard, like Alibaba, Tansum, Baidu. Um those companies like really not reflecting the um kind of like performance um in the US peers, right? So if you think about like previously, um their performance is really driven by uh geopolitics and all those like macro narratives regulations, but um they have quietly built out uh quite strong AI capabilities, right? Given their Alibaba's Q and uh Baidu's uh and Tenzin's model. Um but most importantly, they have uh they have a full stack, right? So for Alibaba, Tensor, Baidu, they have um um AI Cloud, they have the chips themselves, they have the uh kind of like models and they have applications, they have distributions. They own the full stack, um similar to say Google, right? So that gives them a lot of advantages compared to um other like competitors like in the startup world. So those incubators in China tend to uh have a long-term mouth uh in the AI world. And for China, I think like then um I already mentioned this open source advantage. Another thing is really um their capabilities in doing the physical AI, right? So um China's been leading in the EVs and drones, those like traditional manufacturing world of AI, right? So if you think about adding the AI intelligence to the humanoid, to the robotics, to all kinds of um um consumer electronics, that could be another kind of way to add value because China's really good at manufacturing things that's very scalable and cheap. Um that's one advantage if you think about like um if you have a bidding unit of humanoid or a bidding unit like quadrupad, those things can happen like uh um with a lot of suppliers, um, make that available in China. Um so I talk already talked a lot about like the um this the thinking, right? But like what is kind of like the portfolio implication, right? So um currently, well, I think like more or less people are exposed to some AI, but um those composure like depends like even your uh allocating the SP or NASDAQ is pretty concentrated, right? So uh really just concentrated to like Mac7 or just like handful names. Um like we think um they could potentially be a lot of uh value add uh from other segments of the AI in the ecosystem, uh, and there's a lot of value add in the private space. Um so adding like selective private AI place uh right now, I think can be helpful for investors to uh harvest some of the value creation um that those like AI disruptor is uh uh is enjoying. Uh then uh a big trend uh right now, I think like still early, uh, is really complement the digital AI with. Physical AI, right? Because physical AI is potentially a bigger market, but uh right now it's really not really uh getting a lot of investors' attention, right? So those humanoid, those automation, those uh physical AI applications uh are really just beginning. Um we've seen this early stage of the uh S-curve. Uh if you think about like uh EVs or drones, they are being really in the late stage right now uh in the S curve. Um their build-up has been years, their adoption is really um quite strong. But for those like robotics and those like physical AI um uh automations, the adoption just began uh with the help of the digital AI capabilities. So we think in the next five years, this could be a big potential growth uh uh area for investors. Um then uh I think one big diversification is really uh finding opportunities uh in other regions, right? Like you have definitely other places um uh contributing a lot of the AI buildup, um, uh for them, like South Korea and Taiwan, but also like there's any many other regions doing model companies. Um they're building their solving AI initiatives. Um, there's a lot of like local technology champions. So like adding a little bit of exposure to really uh be more diversified uh in this portfolio construction can help. Um adding AI exposure but like uh very uncorrelated. So with that, um I just want to introduce a little bit our solution here. Um we have um kind of like a digital AI, but like it's public and private. Um so the AGIX is um uh one of the first uh public and private AI ETF. Um so we invest in anthropic uh quite early when their valuation was only 61.5 billion uh one year ago. Um and uh we invest in XAI um uh uh hook price actually uh through$10 for runs, um, but now XAI automatically converted to uh uh SpaceX. So we're not on SpaceX. Um so those two companies are private, uh then we have a suite of um uh about like 50 companies in the in the public space where they are uh diversified across the AI ecosystem, uh, from the hardware to the applications uh to the uh infrastructures. So those are really a portfolio that's likely to add a lot of uh uh value to the um um uh digital AI world. Then we have this physical AI um ETF called COID, KOID, uh which capture the global uh humanoid supply chain and uh embody intelligence, which is kind of like the uh AI is happening in in our physical world. Um that's really like a basket very diversified, right? That's really one-third from like kind of Asia, one-third from US, one third from uh Europe and other regions. Um, that's gonna uh it's a global portfolio. Um capture the uh kind of like humanoid manufacturers, um the actuators, the the rare earth companies who's the most critical material for um those physical AI development. So this is a uh very diversified portfolio, 50 companies, uh 2% each, um, and uh capture the next wave of the AI deployment. Um then for the for emerging market, we have this emerging market tech, right? So that's um uh plus like say X Hinex, Taiwan semiconductors, right? Those names probably now heard um a lot. Uh and there's about a Southeast Asia and China tech. Um we believe uh good diversifier for investors when they want to find growth outside of the US. Um then for China, we do have this flagship fund called China Internet. Um capture the names I mentioned, Tenson, Arabaga. Um, those names really, they're the uh most well-known companies in China building their AI ecosystem. Um so then should it become like a more value player for investors, uh, given that violation is being really uh getting down by the narrative of macro and geopolitics, geopolitics. And so that's really uh uh kind of like one interesting solution for investors to add AI uh from the China region. So with that, um I'm gonna um kick off this like um discussion, right? Like um just if you have any question, um you can like reach out to us or just um uh we can have like a QA here as well. Um, but like um of course, uh if more follow-up questions, you can email us at info at criticizes.com.

SPEAKER_01 27:38

The uh the first question that comes to mind is are you a better dancer than the humanoids? Uh they can dance.

SPEAKER_00 27:43

They can dance pretty well. Uh I'm not good at dance. Uh actually I tried uh I have uh this um humanoid uh uh uh running with me on Nayzy, um uh which like they are pretty like if you think about the long run, right? Like they run on batteries, so I cannot match.

SPEAKER_01 28:03

Yeah, clearly not. Uh okay. Okay, so so but so I mean uh I want to talk about Ethereum for a little bit. Um because I'm um it seems like every single day, that's at least the joke on on social media is every single day uh Claude is coming out with something else to destroy another job. Uh given given uh and it seems very clear to me that like uh ChatGPT open AI is is is now far behind, right? On um on the AI implementation way that um anthropic claude is. Um are we getting it to a point where it's legitimately gonna be a winner-take-all dynamic win uh the AI side? Like it's gonna be you know, maybe Google and Cloud and then nothing else.

SPEAKER_00 28:44

Well, you I think now you kind of like see this trend. Um, is the the cloud is winning a lot of the enterprise adoption. And also they're kind of like um very useful in the the open source, kind of like the open cloud, those like AI age, a genetic AI um operating system, I would call that, right? So if you have like an operating system become your AI computer, you give instruction to do work, that operating system, that computer is gonna use the cloud API, right? Uh because the cloud API tends to do things well, like in coding. Um, but like they're using other APIs well. They're using Q and Kimi, like a lot of China open source API, because they're cheap and they can do like some work, which is better. Then they can use like Chat GPT for like some like wording, like language related, they can use um XAI for the groc for social media related. So it's like it's becoming like a multiple um API setup already. Um and like you're gonna have this um dynamic cost and output, right? Like um this like um back and forth. Because like I think like now, like I've been using cloud a lot. Now like I reach my limit, I have to upgrade to like thousand dollar, like hundred dollars with like additional API usage. So that's like to me, like, oh, maybe I have to add a little bit other like solutions as well to if some tasks is really not necessary to be that expensive. So I think like going forward, your air computers gonna similarly, right? Like they're gonna allocate resources. Um, if there's computing constraints, um, then you're potentially adding uh API to other models. This is gonna be the next, I would say like those internet companies similar to like electricity companies, or just um um a little bit better, a little bit like different, but like um we're like seeing this infrastructure um built out, um, and we're gonna see a lot more layers built upon that, uh, which can change this dynamic dramatically.

SPEAKER_01 30:57

Uh I am the uh number six power user apparently at Perplexity.

SPEAKER_00 31:01

Yeah.

Valuations IPOs And Measuring Growth

SPEAKER_01 31:02

Uh of like that's how much you talk about tokens I've used a million in the last three weeks. Uh so I've I've spent quite a bit uh using their agent exclusion. And to your point about sort of um the the computer, yeah. Yeah, perplexity computer. To your to your point about the um, you know, the sort of next phase Yeah, it's like you're hiring an AI to determine which AI to use, right? I mean that's really what it is, right? It's like a manager of AI that is AI to allocate breakdown tests. And I think that's that's actually probably you know where it really can go. I mean, true, even why I've gone so ballistic on it. Um let's talk about uh how to think about valuations in the space because yeah, sure, all these types of outside are making a lot of money. Um, but I don't know if you can sort of apply traditional PE metrics and grammar dot analysis to this, right? So, how should how should we know if you know things are overvalued when it comes to digital to physical?

SPEAKER_00 31:49

Yeah, I mean like for different companies, different narratives, right? So I think for those, say like digital AI companies, right? Like the evaluation is mostly um based on expedition, right? Like the the growth, the earnings growth and random growth is like amazing, the margins expanding. Um but like I think like investors become already greedy right now. So they want more, they want higher if if like the company only beat by like 5%, like that's not enough. So at this stage, I think like for digital AI, you have to keep beating the consensus. Uh for a company like Anthropic, right? Like this is something I think a wake-up moment for a lot of masters, right? Just given what I mentioned. But on the valuation perspective, you think about like a company can easily have like billions of revenue each month. That's AR, right? The occurring revenue. So like how do you read value today, right? Given that growth momentum is like you've never seen before. Um, I think that's that's for a lot of private companies. Um, you have to find like comparables uh where that like this is something we don't have a um large model company IPO in the US. So potentially the first IPO is gonna set a validation for a lot of those companies, right? So that's why I think now Anthropic and SpaceX is rushing to IPO as well. Like if follow the today's news, like Anthropic is really targeting a two-four, um, then SpaceX is like potentially filed in the coming weeks. So that's that's kind of like that. This year, we're gonna see a lot of validations that be set. But like I I still like I think this is early stage for a lot of those growth. Just like as you said, you are using a lot of the computer. So I I'm sure like you are like more uh advanced in using AI compared to like I've lost my mind on it.

Macro Stagflation Risks And Barbell

SPEAKER_01 33:56

No, I've exaggerated. It's um yeah, and it's it it's in the productivity. Um even as we're doing this, I'm setting up tasks, right? Like I'm on the screen, and that's how multitasking. Uh but the automation is what's important. But but to that point, so I think it's where um it is interesting. I don't hear too many people talk about sort of the intersection of the AI implementation in what could be, we'll see, maybe a stagflationary environment, especially given the Oran award. If it doesn't last long, we don't have to worry about it because oil probably would drop. But what if we aren't during this period more like the 70s? Um I wonder if there's any sort of macro thoughts on the clean share side as far as a cycle shift that's that dramatic, uh, how that could alter, maybe positively or negatively, you know, some of these investment themes in the AI space.

SPEAKER_00 34:40

Yeah, we think um, just like we published our outlook last year, um, that we are in late stage, right? Like um consumer slowing down, like drop girls slowing down. Um, but AI is kind of like extending this cycle. Um, huge capex, huge buildup on AI genetics, and um there's a kind of new business coming out of AI. So you do have this barbell, right? Um, where where you have we're gonna see the growth still happening in the AI, right? Um those cap packs, the if you think about think about oh, people question the the capex, but now the uh they keep re-rating, they keep revising higher, much higher. Like um we have across the broad shortage of computing. Um, just like given that uh plot is always restricting my usage, as you can see, like we have a shortage in computing. Um that um this build out is gonna tend to be much longer. But as you as you said, Iran, like there's oil price, there's consumers going down. So this macro job is gonna slow down other growth of business. So this could be two things. One, just we have first this huge drawdown, right? Like cross-board. Um, but when you have the fundamentals pick up on the where the growth is happening, does tend to save a lot of those uh stock prices. Um then like if we do say have this huge drawdown, then I don't think that's gonna it's really challenging, right? Because like the oil is really like an input short shock, right? So there's no way for them to really hike the price because like doesn't hike the rates because like doesn't really help uh much. So all they can do when there's like kind of like meaning recession happening is really lower the rates. So that's I think like potentially can favor a lot of those companies still has half growth potential. So for that, maybe like investors should take like a barbell uh allocation, right? Like in in front, betting on some like companies, uh good quality mode can still rather this volatility in macro and keep growing. Um in another barbell is really um more defensive and putting like a lot of um the hash in on energy sector or whatever. So those barbell is what we think potentially can be uh allocation uh this year.

Allocation Sizing And Where To Learn

SPEAKER_01 37:14

Yeah, I think that that makes sense. I think the broader point is like the this like AI has already kind of thrown the investment playbook uh out the window. Uh now Stiflation might do that again, right? If that's the case. And it's um yeah, I think it's a lot of interesting dynamics from an investment perspective. Again, folks, if you're for Credits, I will email you to get your information. I will submit that to the CFP board. Um, any final thoughts here um as far as you know how to think about allocating to the space? You know, I it sounds to me like these are more satellite positions in portfolio, but you know, if somebody's really a believer in the sort of uh digital uh transformation, how's anything about sizing?

SPEAKER_00 37:52

Yeah, so we think like um for something like AGX is more um, even though it's a satellite, but like it for the AI part, um, it's kind of like core AI, right? It's really the core is to read the digital election model and those like build out. Um, we we believe that's likely like five percent uh for investors. Then for like uh KYD, um that's more like the um uh kind of optimistic satellite because it's quite early. Um so that's like um maybe two to three percent allocation for investors uh as more optimistic satellite. Uh for the EM and China side, um I think like compared like that that really depends on your allocation model. We recommend like 5% to like EM China. Um for investors who more prefer like broader EM, the EM tech, uh KMQ, potentially take that 5%. Um, but like for investors who want to add more conviction into like the China region, maybe like a 50-50 of the EM plus like 50% in the KWA. So that's kind of like um like our thoughts on some allocation perspective.

SPEAKER_01 39:03

Uh there, for those who want to learn more about the funds and get more access to crane shares is information research, where do you point them to?

SPEAKER_00 39:10

Oh, yeah, please visit crane shares.com. Um, and for each font specific, if you want to say AGIX, can visit CraneShares.com slash AGIX. Uh we have our presentation, we have fact sheet, we have FAQ, research with white papers, uh, everything you need for the fund. Um, or you if you have more questions, uh you can always email us at info at craneshares.com.

SPEAKER_01 39:36

Those that uh attend this hopefully found it uh worthwhile and interesting. I do these uh webinars uh twice a month with Crane Shares. Keep attending them uh and uh take take a look at all the various funds. And hopefully we don't have stagflation because um I don't know, high oil and humanoids.