원문 보기
We talk about export controls every day in Washington. You're on the other end of this, of course, looking at Nvidia's balance sheet and forecast, trying to figure out how to factor this in. How significant is this to crack back into the Chinese market? Look, there's one chip in the world fueling the AI revolution, that's Nvidia, and you really need to give them access to the China market.
I think that's the understanding that that's been well understood now in the Beltway. It's important because the reality is we're talking 20 billion a year, and if they don't allow them to sell into China, that just gives Huawei and China competition the ability to narrow the gap, and that's what Jensen continues to talk.
I think that that message is definitely being heard. Mhm. Is it possible for a Dan Ives to do channel checks when it comes to China? Do we know that there are buyers on the other end? Yeah, I mean, like obviously we spent a lot of time in Asia, in China. Look, they they want Nvidia chips. The reality is would you want Peter Luger steak or would you want White Castle?
And the reality is they they want the one chip in the world fueling the AI revolution, and that guy's wearing a black leather jacket, and his name is Jensen. And look, there's competition, Huawei and others, they're narrowing the gap, but big tech in China, there's one chip they want, it's Nvidia. Well, it's really interesting because Michael Burry, I think is ready to meet you at White Castle.
He was out once again with a knock on Nvidia today suggesting that energy efficiencies will make their highest power chips less useful in the immediate term. So, as we throw all these data centers back together with an eye on Broadcom and its TPUs, is there some truth to this? Look, I mean, our last Asia trip, if I go back a month ago, demand to supply is 12 to one for Nvidia chips.
So, look, I it's going to continue to be a prove me in terms of the AI revolution, but I mean, we still believe this is early days. You're talking about three to four trillion it's going to be spent next few years. There's there's Look, we're bullish on Google, bullish on AMD. Others are going to get into this AI arms race on the chip side, but right now there's one chip at the top of that mountain, and it's Nvidia's world, and everyone else is paying rent.
It's Nvidia's world. Is this stock then, in your view, with the earnings growth, revenue growth that we've seen in the most recent forecast, is this stock cheaper than it was a year ago? Yeah, I mean, like 250 to 275 is where I see it ending in 2026. It's a cheap stock. And and you know, I think it's one where investors sometimes, you know, there's a lot of noise in this, and you got to separate out what demand ultimately looks like, and we're in the third inning of a nine-inning game in terms of AI revolution, and it's not just about Nvidia, it's about Microsoft, it's about Google, it's about Palantir.
It it's the, you know, the infrastructure, Nebi is another's, GE Vernova. I mean, that This is where it's all heading. You know, even when you look at things like nuclear, right, in terms of from a from a perspective from an energy, it's just starting. So, um you know, semis, obviously it's been kind of the lifeblood of this bull market leadership for 3 years.
It's changed. I guess we got to be specific in terms of which which chip stocks or themes we're uh we're talking about, but how are you viewing this latest burst of enthusiasm for for custom chips and some of these other uh kind of subsectors? Well, I think if you step back and you think about what's going on with chips, the first really 3 years since chat GPT showed up at the end of 2022, it's all it's been all about training, and then about inference this last year.
What you're seeing this year though with open claw being the big, you know, event at the end of January is that you've moved to this age of agentic AI. And so, what you've seen with tokens production is it's gone from growing about 20% over the last couple of months before open claw at the end of January to now it's up over 130%.
And so, this lull I think you've seen in chip stocks, I think you're going to see them get a lot stronger. You're going to have to be selective, but I think you're going to see them get a lot stronger as you go through the rest of this year because of that move to agentic and the massive increase in compute that that takes.
In that context, you know, you mentioned we kind of transitioned away from the the GPU training focus we were in for so long. Does it make sense to you? Does the market have it right that Nvidia has been just kind of sideways for the better part of a year? Yeah, the market did have it right, but what should encourage people if you own Nvidia is the fact that today you had an event that could have easily knocked Nvidia down, which is Broadcom had that big deal between Anthropic, Google, and themselves.
The stock was up 6%, but Nvidia closed the day up. So, I think from a short-term perspective, you see these digestion phases. I've talked about Cisco in 1997 and '98, how it was down 37 to 38% intra-year both of those years, and it finished those years up 31 and up 150%. And I think Nvidia is sort of in the same category where people have rightfully concerns because of what you're seeing in terms of the market getting more selective around the eventual winners, but this move with agentic requiring a lot more tokens means that I think Nvidia ends the year higher than where it is today, even though I think OpenAI you're going to figure out has a lot more problems than investors currently think.
That's interesting. Now, I I was going to just ask about the companies that are kind of creating all this demand for uh the semis with their CapEx budgets. Uh do you Are you among those who feels like we're at a point where the only thing that's going to make like a Microsoft work better is if we do start getting CapEx cuts, or maybe Meta, or have they gotten cheap enough on their own?
No, well, I think it depends on why you get CapEx cuts. I think with Meta, if you get a CapEx cut and they say, "Hey, we're cutting back on all these science projects we have going or, you know, Reality Labs," I think people will like that because if you look at this most recent quarter that they reported, they were the only ones that had a really strong return on investment.
And meaning that the CapEx went up a lot, but the stock I think the next day was up 10% or so reaction to earnings because the revenues got revised higher as did the earnings. So, I think with Meta, that would be taken positively. If you get a CapEx cut at one of the big cloud providers, so Amazon Web Services, Google, or Azure, I think that's a different problem because investors are giving these companies the multiples they have because they're viewing all of them as winners.
Um and I don't think that's That's why I'm saying you're going to have to get more selective. I mean, and you're seeing it with Anthropic coming out and saying, "Hey, we're raising prices." And at the same time, OpenAI is coming out and saying, "Well, we're having to cut prices because now, to me at least, it looks like Anthropic's got faster or higher run rate revenues at around 30 billion versus OpenAI's last estimate of around 24.
And obviously, OpenAI started this in back in the end of 2022 with chat GPT. Spreadsheets based on price-to-earnings ratios and earnings projections are used by the majority of Wall Street analysts to cover Nvidia. UBS uses a completely different method, and what it says about Nvidia is startling. Nvidia's share price should be 400% higher than it is now, according to UBS Holt, the company's own cash flow-based valuation framework.
The company's market capitalization would then be $22 trillion, up from $4.46 trillion on April 8th. John Talbot, UBS's US head of Holt's technology coverage, stated, "We believe the share price should be 400% higher. For investors, it's a significant amount." He continued, "That's the big pushback I get.
What is UBS Holt? Holt is not a typical model for analysts. Professional investors worldwide utilize this unique valuation framework, which was first created by Credit Suisse, and is now a part of UBS after the 2023 merger. Holt is based on cash flow return on investment, or CFROI, a metric that eliminates accounting distortions and concentrates on the actual economic returns a business earns in relation to its asset base, rather than on earnings multiples or analyst price forecasts.
By all historical standards, Nvidia's measurements inside the Holt framework are exceptional. Currently, the company's CFROI is 73%, while the typical non-financial company in the database has a CFROI of only 6%. This puts Nvidia in the top 0.1% of all businesses that Holt has ever monitored. Its asset growth rate is among the top 0.5% of the database.
Crucially, Nvidia has not experienced the model's anticipated decline in returns, which is the moment at which competition starts to erode a company's advantages. The model's long-term return estimates were revised upward because the company consistently outperformed the fading assumption. Seldom in its history has Holt been so optimistic about any stock.
Investors typically react to the $22 trillion figure with instant disbelief, according to Talbot, but the model's output is a direct result of the underlying facts, rather than a judgment decision. With a $22 trillion market value, Nvidia would be around five times larger than it is now. To put things in perspective, the total value of the S&P 500 is almost $40 trillion.
At $22 trillion, Nvidia alone would be worth more than half of the current index's total value. Talbot recognized the opposition he gets because of that scale. In the conventional understanding of an analyst, the number is not a price goal. It is the result of a formula that determines how much Nvidia would be worth if its present return profile, growth pace, and history of exceeding fade expectations continued.
Every stock investor is tacitly responding to the open question of whether they will. Nvidia's stock has been under significant pressure at the time of the Holt review. Due to investor rotation into energy and infrastructure equities related to the Iran war and worries about the returns on AI capital expenditures, the Magnificent Seven has fallen nearly 11% so far this year.
That pressure has not spared Nvidia. In light of this, a model claiming the stock is five times too cheap stands in stark contrast to the general sentiment. Holt is not a framework with a lot of buzz. It is based on cash flow economics rather than storytelling and has been in use for decades. Regardless of where the stock moves from here, the fact that it is generating this output for Nvidia shows the extent of what the firm has truly achieved in its core financials.
Join our investor community. Hit subscribe and like to get your daily Nvidia stock insights and tutorials delivered. After a protracted period of sluggish trading, Nvidia shares are starting to pick up steam. The stock has advanced more than 10% over the last six sessions, marking its longest winning streak since October.
This latest action suggests that the consolidation phase may be coming to an end after a period of subdued performance from September 2025 until the end of last month. Before falling 0.5% on Thursday, the shares closed Wednesday at $180, getting closer to the $185 level that technical analysts are eagerly watching.
With US stocks rising following President Donald Trump's announcement of a two-week truce in the battle with Iran to allow negotiations to continue and reopen the Strait of Hormuz, the wider market backdrop has also offered some support, allaying fears of a possible disruption to the world economy. Nvidia's 2.2% increase and leading point rise in the S&P 500 contributed to that shift.
Technically speaking, analysts suggest that a persistent rise above $185 might imply that investors are starting to return to the stock, possibly signaling a departure from its recent trading range. The $200 mark is also mentioned by others as an extra barrier that would be necessary to validate a stronger rising trend.
Nvidia's stock is currently trading at about 20 times anticipated profits over the next 12 months compared to a 10-year average that is closer to 36, indicating a significant change in the company's valuation. This puts it among the Magnificent Seven's lower valuations and roughly in line with the S&P 500, which might make the shares seem more appealing after the recent reset.
A breakout could have further ramifications for market sentiment because Nvidia is a bellwether for megacap technology and artificial intelligence spending. The technical setup is still balanced, though, and downside risks are still being considered. Analysts are keeping a careful eye on the $170 mark because a break below it might lead to additional drops toward $150.
The stock has been trading inside a defined range in the short term with support at $165 and resistance at $180. This suggests that the stock may continue to consolidate unless there is a clear breakout or break down. The iShares Expanded Tech Software Sector ETF was down approximately 25% in 2026 as of Wednesday's close compared to a roughly 1.4% fall in the Nasdaq 100.
This uncertainty coincides with pressure on portions of the larger technology sector, especially software names. Given this, Nvidia's future course of action may serve as a key indicator for how investors should handle market exposure to AI. As the idea of orbital infrastructure and data centers becomes a reality, Nvidia's stock may be grounded, but its GPUs are traveling to space.
It's getting difficult to ignore the space competition, especially as businesses look to the skies for profit potential with Artemis the coup taking flight, a SpaceX IPO probably scheduled for the beginning of summer, and several space companies preparing for their own launches in the second half. Elon Musk did make some excellent arguments about keeping part of the AI computation up there, even though his idea of data centers in space first seemed very absurd.
I believe orbital infrastructure is at least worth a try, whether it's because of the stronger solar energy up there, the cooling advantages, or the latency advantages of inference on edge devices. In any event, as businesses turn to satellites as a source of 5G connectivity, a type of corporate space race appears to be developing.
Starlink was incredibly ambitious and the business is doing quite well. Bringing the AI hardware up there would be the next step, of course, and if Musk is given the go-ahead, I don't believe it's too unrealistic to imagine a constellation of 1 million data centers. Anyway, let's get back to Nvidia.
With a platform designed to run the industry's newest and greatest in space, the corporation is also aiming for space. It was simple to overlook the Space One Vera Rubin module, which was showcased at the company's GTC 2026. It literally is, which is why it seemed unreal. The key question is whether space-based data centers may be another significant growth engine for Nvidia as CEO Jensen Huang and the company collaborate with other businesses to enable Vera Rubin chips to function in orbit.
It's reasonable to assume that other tech giants will take note and begin making large bets of their own if Musk spearheads an AI data center space race. The only thing more audacious than sophisticated AI is advanced AI in space, and the Mag Seven Titans do have a lot of money to spend. As of right now, it's difficult to predict where the idea of space computing will take the megacap IT giants.
Although it's undoubtedly thrilling, investors should definitely keep their expectations in check for the time being because the development of AI data terrestrial centers on Earth is still in its early stages. Of course, SpaceX has a significant first-mover advantage, particularly as it looks to new development frontiers while the rest of the IT industry struggles to keep up.
The trillion-dollar question is whether hyperscalers will raise capital expenditures in the future to pursue data centers in space. In any case, Nvidia has the picks and shovels that businesses will undoubtedly grab as their aspirations soar into the stratosphere. I find Nvidia's collaborations with companies like Sophia Space, an orbital computing startup that is committed to the emerging field of space computing, to be very fascinating.
As of right now, I don't think Nvidia stock is priced with any orbital data center excitement. Furthermore, I don't think anything will change in the near future, at least not until the hyperscalers start increasing their investments in orbital computers and satellite constellations. It's difficult to predict when Nvidia stock will start to rise again, particularly as the AI pioneers strive to overcome impending energy and connectivity limitations.
Naturally, AI must demonstrate that it can generate revenue. I would expect Nvidia to continue making investments in order to assist the industry in overcoming these obstacles in the interim. However, the longer Nvidia remains at roughly $180-185 per share, the valuation multiple will only become lower.
It's easy to find an expert who anticipates significant profits for the company in the upcoming year, but in my opinion, Nvidia should be the first to exit its consolidation channel. The longer the base, the higher in space, according to the technicians. Don't forget to share your thoughts and experiences in the comments below.
Your input helps create a more knowledgeable community where we can all learn and grow together in the world of investing. Stay up to date with the latest market trends, insights, and key updates by subscribing to Investing Tutorial. We're here to help you make smarter investment decisions and stay ahead of the curve.
Comments 0