저자: DanIves/TomLee · 외국인
요약 보기
Dan Ives는 2026년 테슬라 자율주행·로보택시·옵티머스 성장을 근거로 시총 2~3조 달러 도달 가능성을 주장하고, FSD 침투율 15%→50% 상승이 마진을 크게 높인다고 봄. 스페이스X 상장 시점도 AI 8~10년 투자 사이클상 적기라고 평가. Tom Lee는 AI 투자 사이클이 초기 단계이며 엔비디아가 과소평가돼 리레이팅 가능하다고 주장. 비트코인은 신규 지갑·IRA 계좌 확대 여지가 크다고 긍정적 전망. 은행(JP모건·골드만)도 기술기업화되며 저평가라고 봄.
다듬은 글
Boy, are you in for a treat. I've got a big show for you here. Dan Ives just called 2026 a monster year for Tesla with a potential $3 trillion market cap on the cards driven by autonomous driving and robotics. Then Tom Lee explaining why the market still doesn't understand we're at the beginning of a decade long AI buildout, which means Nvidia might have plenty more room to grow from here.
There's also a chance he says that it could hit a 3 trillion market cap by the end of that year. for more. Let's bring him in. He's here on set with me. Dan Ives is Wedbush global head of technology research. Welcome. Great to be here. You could have picked an easier one. You know, maybe Nvidia like why Tesla to you is it still so clear for all the ups and downs for the auto industry, for all of its ambitions, everything they want to do is always delayed.
Granted, the market cap, you know, is still phenomenal. Why do you think 20 why do you feel so strongly that 2026 will be a great year? I think it's about autonomous. I mean all when it comes to physical AI, I think there's nothing better, you know, when when it comes to physical AI than Nvidia and Tesla.
And I think this is going to be a golden chapter for Tesla when it comes when it comes to robo taxis, when it comes to cyber cabs, volume production, true autonomous, and of course, you look at robotics in terms of optimist. I mean, Kelly, that's why I think we're looking at two trillion ultimately, probably a $3 trillion market cap by late 2026, early 2027.
That's why it continues to top there. What's the E? So, you know, you think about a market cap that big. You say, "Okay, it's high multiple. I don't know what exactly the earnings look like." What are you anticipating on that front? And how do how do how does physical AI kind of feed directly into the earnings power it might have?
I mean when I look out the next three four years I think earnings power is 4x 5x but because it's my view that you when you think about not just the FSD piece but in terms of true autonomous it's going to change the whole margin profile it changes the whole growth story and I think just going forward for Tesla the focus it's okay deliveries they matter of course they matter but in terms of next few years especially Musk now wartime CEO in terms of pay package you know as that's already cemented this is just the start and they're also going to have an ownership of XAI look what's happened SpaceX I mean I think we're now talking about this next AI chapter that's happened across the Musk ecosystem
we had Dan Primack on last week and and he said you know I basically asked him or he said the question to ask is why is SpaceX going public now do you think there's anything related to financial pressure on Tesla or Musk or anything like that or is it a political why would 2026 be the year for SpaceX
I think the time's right I mean you could argue valuation, right? Does it approach a trillion dollars? It's the same reason why it is, you know, anthropic or, you know, ultimately when when you look at open AI and others, I mean, we're hitting at a point where it's only year three of an 8 to 10 year buildout in terms of the AI revolution.
And for the first time in 30 years, me and you've talked about this, you know, US is ahead of China when it comes to tech, but Space X is in inflection point. I think it's important that it goes public. It's the right time. And and look, we are in, I think, two more years in this tech bull market and we're going to see more and more of these great private companies going public.
Won't SpaceX be, if you'll pardon the term, sexier than Tesla? In other words, right now, if you are a big fan of Elon Musk and everything he's doing in the future, as described, this is the only public equity you can own for that story. Now, SpaceX could be the newer, younger thing on the block. What if investors go with that and thus remove some of the halo, right?
the the community effect around those who hold shares of Tesla and support its valuation just to have a peace of mind. Yeah, it's a great point. Look, I actually think it expands. I mean, to some extent, the ecosystem from SpaceX to Tesla and ultimately what's going to be the ownership of XAI because the reality is when you think about what's happening here.
I mean, we're talking about the next 5, 10, 15 years. Musk at the centerpiece of it. But I think it speaks to when you when it comes to Tesla that continues to be the best play because I think autonomous is worth a trillion dollars alone. You think about Optimus in terms of true robotics and as this all plays out I think globally I mean they can own 70 80% of the autonomous market you know over the globally over the coming years.
I know you think that full self-driving penetration could be as high as 50%. In other words, half the people who buy a Tesla pay the extra money to have that option, which would be, as you say, a gamecher for margins and for the stock. So that's part of what you're talking about is for those who own Teslas, they use this technology.
The other part is the cyber cab is what is it called? And how would you describe its position in this Visa v Whimo where Whimo also does quite well? The car is really expensive um regulatory and otherwise. What should we expect on the uh like the cyber cap front from Tesla? Well, first like FSD, you're going to go from penetration 15% to 50 and and I think could even be higher.
So from a margin perspective, that's going to be huge. Look, Whimo, Whimo is going to be a rounding error relative to what Tesla's going to do. I think not just in the US, but I think globally. Look, it speaks to just how expensive those cars are, how they could scale. And I think when you think about autonomous and in terms of the road map, I mean, we would expect probably some sort of executive order, you know, going into early next year, what I think is going to clear in terms of the federal road map.
I mean, more and more you're going to go away from the states to federal and that's going to be extremely important for Tesla in terms of where I think they could take robo tax. Tesla's autonomous vehicle fleet generating billions in realworld training miles and Nvidia's complete infrastructure stack from chips to software creates modes in physical AI that competitors can't easily replicate because both require decadelong investments that only show returns once the technology reaches commercial scale.
Physical AI requires solving problems that pure software AI never faces. An LLM making a mistake means generating wrong text. A self-driving car making mistake kills people. A humanoid robot dropping a package is annoying. A robot in a factory dropping heavy machinery creates liability and safety disasters.
The margin for error in physical AI needs to approach zero, which means you need massive amounts of real world training data to handle edge case scenarios that simulations miss. Tesla's moat comes from having millions of vehicles on roads collecting this data daily. Every Tesla with FSD enabled encounters scenarios that improve the neural network.
Whimo operates just thousands of vehicles in liberties. The data collection gap between Tesla and Whimo spans three orders of magnitude and it compounds because more data enables better autonomous performance which justifies higher FSD adoption which generates even more training data. Nvidia's mode in physical AI comes from vertical integration across the entire stack.
They don't just sell chips. They provide CUDA software libraries, simulation environments like Isaac for robots, training frameworks optimized for their hardware and reference architectures for deploying at scale. A competitor can build chips that match Nvidia's raw performance. But matching the complete ecosystem requires years of software development and thousands of engineers building tools developers actually want to use.
The switching costs aren't just about chip performance. They're about rewriting all of your training code, rebuilding simulation environments, and losing access to the largest developer community in AI. Companies invest millions integrating NVIDIA stack into their workloads. Moving to alternatives means writing off those investments and delaying products by years while engineers learn new tools.
The timeline to commercialization creates another barrier competitors can't overcome quickly. Tesla spent a decade and tens of billions developing autonomous driving before reaching the point where removing safety drivers makes sense. Starting today, a competitor needs similar time and capital investment to match current Tesla capability.
But by the time they finish, Tesla will have advanced another decade with more data, better models, and established market position. The same dynamic applies to human mode robots, where Tesla's factory deployments will generate robotic trained data competitors don't have access to. Nvidia's position benefits from being the training platform for everyone's AI, including competitors attempting to build custom AI chips.
Every new AI application gets prototyped on Nvidia's hardware first because it's the path to the least resistance. By the time custom chips reach production, Nvidia already has next generation products and the application developer already optimized their models for Nvidia's architecture. Next up, Tom Lee is making some pretty significant predictions.
Check this out. Joining us now, Tom Lee, Funstrat, head of research. Tom, great to have you with us. Um, there's a lot of things to talk about with you in terms of the skepticism now around the AI trade, skepticism around Bitcoin. Are they in fact related, do you think? Uh, related, but not related.
Yeah. You know, I I think the AI trade, I mean, it's been a juggernaut, and we know that the visibility for the next few years is really good, but now investors are just trying to figure out how much to discount. You know, what's the appropriate price to put um and that could drive some profit taking because they've done so well this year.
On the crypto side, I I think that there has been some concerns about uh you know, quantum and then there was that huge deleveraging event on October 10th, but the crypto fundamental story I mean is exiting on a really high note this year. You know, the US government passed legislation and regulations favorable for crypto and we know Wall Street has expressed a lot of interest in building products on the blockchain.
So, it's almost like they both have great visibility, but the price levels are just what's being uncertain. I feel like what's common though with both of these stories though, Tom, is investors questioning the growth, the trajectory of growth going forward for both of these stories. Not to say that that things will slow will contract, but maybe growth won't be as strong as we've seen.
We've we've hit the peak sort of growth rates in terms of the story. the maximum bullishness happened with Bitcoin when the administration came on board and things started to sort of go Bitcoin's way. Maybe that was peak sort of trajectory higher when it comes to growth. Same thing with AI. Now it's a period where you're sort of wondering, will this actually get funded after we saw this massive expansion?
Yeah. Well, in in crypto, the best years are definitely ahead because today there's only 4 million Bitcoin wallets with $10,000 in it. There's 900 million IRA and brokerage accounts globally that have $10,000. So there's that's 200 times larger market. Um with regard to AI, I I do think that the business models need to evolve.
You know, you can't just charge $20 a month. I mean really AI agentic services and but the the market ends up being huge because in the US it's really making 130 million existing workers into super workers. I mean, that's a several trillion dollar market just in the US alone. So, I I think there's still opportunities.
When do you think that we're going to see the tipping point on that front? I think arguably you're seeing it a little bit maybe on the margins right now. But do you see a step change? 18 months out from now? Is that 5 years out from now? It's because there's also questions, by the way, about some of these services and what how valuable they are to people right now.
And I know some CTO's and others who are actually saying, you know, we're actually not getting the value out of some of these products that we were hoping to. Yeah, it's I think we're in what the phase what I'd call the bleeding edge phase, like where everything feels like the tech is in front of the applications.
that happened in mobile, which I studied for 17 years, and in internet early days, you know, like when color phones first came out on on mobile or even keyboards, people kind of sneered and said, you know, who's going to be using a color screen? And remember famously what happened when the iPhone was launched and and how the one of the biggest competitors dismissed it and saying it's just a toy, right?
I think we're at that phase with AI where the capabilities will blossom and then services will follow. But that as you know the future happens faster than many expect. So if the business models of these players which people commonly call the mag 7 in general um need to change then what does that do to the valuation?
Are we rethinking the valuation? Because I mean one distinct change has definitely already happened and that is asset light to asset heavy asset spending on assets continues etc. Um is that reflected already or does there need to be further adjustment? Uh, I'm I believe that AI stocks, especially stocks like Nvidia, are going to actually rerate higher because if Walmart and Costco trade at 37 to 45 times earnings because they're predictable, then Nvidia shouldn't be 27 times.
I also think the banks are really criminally undervalued right now, like a JP Morgan, because that's becoming a technology company. And in the past people are stuck in the head that they should trade at one time spoker 10 times earnings. I mean they should trade more like a tech stock. So I think tech and
financial out of all the banks or you I mean JP Morgan in particular has invested obviously heavily in tech but if I said JP Morgan and then I mentioned a bank 10 rungs down would you say the same thing? Uh well it depends on how sort of evangelized a bank is towards adopting AI and being tech forward.
So, okay, on this screen right here, like is Wells Fargo on your list of of tech forward banks? It possibly Goldman certainly is and I think Morgan Stanley has shown a lot of evidence. So, I think banks are sort of getting it, but I they are really becoming technology companies. So, even at record highs,
even at record highs, you know, as uh many technology technical strategists will tell you, it's actually sometimes better to buy things at all-time highs. Tom Lee is arguing the market under values AI infrastructure because investors are still thinking in terms of cyclical semiconductors rather than recognizing Nvidia as a predictable platform company that captures revenue every time AI deploys at scale.
Investors remember Intel's dominance in CPUs, giving way to declining growth and margin compression. They assume Nvidia faces similar dynamics where competitors emerge, custom chips fragment the market, and growth inevitably slows. But Nvidia's business model differs fundamentally from historical semiconductor companies because AI deployments create compounding demand rather than replacement cycles.
Every company deploying AI needs compute for training and inference. Every autonomous vehicle needs AI chips. Every robot needs edge computing. Every data center needs accelerated computing for modern workloads. This demand doesn't disappear during economic downturns the way consumer PC sales collapse.
The comparison to mobile phone adoption and the iPhone dismissal by Nokia reveals we're much earlier into AI deployment than the market pricing suggests. When color screens first appeared on phones, critics questioned their utility. When the iPhone launched, BlackBerry and Nokia executives laughed at the lack of keyboard and questioned whether consumers would pay premium prices.
5 years later, the smartphone revolution had destroyed those incumbent companies and transformed how billions of people live and work. AI sits at a similar infection point where the technology exists but mass deployment hasn't happened. Consumers use chat GBT and similar tools but AI hasn't become as essential as smartphones or internet access.
This early stage means Nvidia's revenue growth could sustain for years if adoption and as adoption is expected to accelerate. The progression from bleeding edge to mainstream typically takes 5 to 10 years in technology cycles. Cloud computing took a decade from AWS launch to becoming the default deployment model.
Mobile took similar time from iPhone launch to becoming ubiquitous. AI infrastructure is likely to follow the same trajectory, which means Nvidia's current scale represents the early innings of a decadel long growth opportunity rather than a late cycle peak before full decline. It's just my opinion, not financial advice.
The trillion dollar market from making 130 million US workers into super workers demonstrates the scale of the untapped opportunity. Most workers don't use AI meaningfully yet. Knowledge workers experiment with chat GPT but haven't integrated AI into daily workflows. Manufacturing hasn't deployed AI powered robotics broadly.
Logistics still operates mostly without autonomous systems. Healthcare uses minimal AI despite enormous potential. The gap between current adoption and full deployment spans orders of magnitude. As this gap closes over the next decade, every step function improvement in AI capabilities requires more NVIDIA infrastructure to train better models and run inference at scale.
The revenue compounds because better AI enables more applications which justify more deployment which requires more compute which funds development of even better AI. Picture this. A potential client searches for what your business offers and your YouTube video appears. Before they've even booked a call, they've built trust with you, turning them into a warm lead.
That's why our clients are hitting $100,000 months because YouTube turns attention into authority. If you run a business, book a call. and I'll show you exactly how to make this
Comments 0