Prof G Markets

Damodaran analyzes Nvidia's governance risks and Anthropic's high-stakes IPO valuation

1h 2m Listen to the original
NvidiaAnthropicIPO ValuationAI InfrastructureMarket OutlookCorporate GovernanceSemiconductors

Summary

Speaker_3 discusses Nvidia's leadership transition and structural business risks, arguing that regulatory scrutiny in China poses a greater long-term threat than the CEO's departure. He highlights Nvidia's precarious margins and reliance on hyperscalers, warning that AI commoditization could erode its pricing power. Regarding Anthropic's $1.5T IPO, Speaker_3 values the company at ~$1.35T, citing regulatory headwinds and massive compute liabilities. He criticizes Anthropic's governance structure and questions whether its revenue growth can justify its $900B compute commitments. Finally, he notes the S&P 500's resilience is tied to AI-driven earnings growth, warning that a slowdown could trigger a 2027 correction.

Key takeaways

  1. 1
    Speaker_3 calculated a $500B debt value for Anthropic's compute leases, subtracting it from equity to derive the $1.475T equity valuation.

    Unidentified speaker (speaker_3) · 32:25

Episode sections

  1. 0:55 – 3:06

    Nvidia CEO Scandal and Value Impact

    Speaker_0 introduces the Nvidia scandal involving CEO Jensen Huang's departure and the board's refusal to reinstate him. Speaker_3 argues the story has two dimensions: personal foibles and a 'China twist' involving potential regulatory scrutiny. He notes that while Jensen's personal impact is significant, the China issue poses a lasting threat to Nvidia's value regardless of the CEO.

  2. 3:06 – 21:16

    Valuing Jensen Huang and Nvidia's Business Model Risks

    Speaker_3 estimates the value of Jensen Huang to Nvidia, suggesting his replacement would have a small direct effect if the management team is competent, but the loss of his 'opportunistic' market vision is a risk. He highlights Nvidia's precarious business model, characterized by high markups and tensions with suppliers (TSMC) and customers (Meta, Microsoft), which could lead to margin compression.

  3. 21:16 – 26:30

    Anthropic IPO Valuation and Pricing

    Speaker_0 discusses Anthropic's IPO priced at $83/share with a ~$1.5T market cap. Speaker_3 reveals he valued Anthropic at $1.475T on Nov 2, but updated it to ~$1.35T due to regulatory headwinds and higher compute costs. He calculated an intrinsic value of $77/share based on updated share counts, expressing concern that the $83 offering price is too close to the banker's target to generate a pop.

  4. 26:30 – 32:05

    Anthropic Governance and Compute Commitments

    Speaker_3 criticizes Anthropic's governance, specifically the single founder share holding 50.01% voting rights, comparing it to a 'golden share' that could be expropriated by regulators. He also addresses the $900B compute commitment, noting that while the IPO cash balance covers the next 3 years, the company must double revenues annually to justify the spend, violating the principle that young companies shouldn't borrow without earnings.

  5. 32:05 – 34:28

    Anthropic Valuation and Compute Risks

    Speaker_3 discusses Anthropic's strategy to secure compute power commitments ahead of competitors, noting a $500 billion debt value for leases subtracted from equity to arrive at a $1.475 trillion equity valuation. The speaker emphasizes that revenue growth is the critical driver for this valuation, as profits and low cost of capital depend on reaching $1 trillion in revenue.

  6. 34:28 – 39:13

    Anthropic Revenue Growth and Competitive Landscape

    Speaker_0 highlights Anthropic's slowing revenue growth, dropping from enormous levels to 47% and potentially 15-25%, raising concerns about doubling revenues annually. Speaker_3 argues Anthropic must avoid overbuilding frontier models and consider entering businesses like drug discovery, though this creates conflicts of interest with enterprise clients.

  7. 39:13 – 46:36

    Anthropic Business Model and Governance Concerns

    Speaker_0 questions how Anthropic can achieve its revenue targets given current growth and risks. Speaker_3 believes Anthropic will become one of two or three dominant players due to stickiness in enterprise data access, but warns that its governance structure may hinder necessary pivots, comparing potential market punishment to Facebook's metaverse struggles.

  8. 46:36 – 57:24

    Broader Market Resilience and Sector Rotation

    Speaker_0 notes the S&P 500 is up 7% year-to-date despite significant AI stock declines. Speaker_3 describes this as the most impressive market performance in a decade, attributing resilience to the diverse US economy and AI-driven economic boom supporting consumer spending in staples and healthcare.

  9. 57:24 – 1:02:13

    2027 Market Outlook and AI Dependency

    Speaker_0 asks about 2027 expectations amidst capital raising concerns. Speaker_3 predicts AI will remain the economic undercurrent; if AI growth slows substantially, it could cause a recession and market correction. He highlights that S&P 500 earnings growth has been exceptionally high (56%) due to AI CapEx, and the rate of this growth slowing is the key variable for 2027.

Assets discussed

AssetKindStanceContext
NVIDIANVDAequitymixedWe finally heard from Jensen Huang after his departure as Nvidia CEO. He made clear that he was open to returning, but the board said during the earnings call that he would not be coming back.
Anthropicprivate companyneutralBut going back to just looking at the anthropic business and and your valuation, one of the risks that
AlphabetGOOGLequityneutralanthropic and alphabet, not from you know revenues it generates as a standalone LLM.
S&P 500indexneutralSo what do you think will happen? I mean, we're a little bit a little bit early to be making predictions, but what would be your expectations going into 2027 for the broader stock market for the SP?

“Stance” is how the speakers talked about the asset, not a recommendation.

Claims worth verifying

Jensen Huang departed as Nvidia CEO, and the board said he would not be coming back.

Unidentified speaker (speaker_0) ·0:55

How to verify: Nvidia Q3 2026 Earnings Call transcript or press release; check board statements from Nov 2026. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Primary source

Anthropic's IPO was priced at $83 per share with a roughly $1.5 trillion market cap.

Unidentified speaker (speaker_0) ·21:16

How to verify: SEC S-1 filing for Anthropic; market data from Nov 2026. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Primary source

Speaker_3 valued Anthropic at $1.475 trillion on November 2, 2026.

Unidentified speaker (speaker_3) ·21:41

How to verify: Aswath Damodaran's blog 'Musings on the Markets' post dated Nov 2, 2026. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Anthropic plans to spend around $900 billion on compute and infrastructure over the next several years.

Unidentified speaker (speaker_0) ·29:06

How to verify: Anthropic S-1 prospectus or investor presentation; check disclosed capital expenditure commitments. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Anthropic spent around $9 billion in the first half of the year.

Unidentified speaker (UNKNOWN) ·29:37

How to verify: Anthropic S-1 filing; check income statement for H1 2026. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Primary source

Nvidia's market cap dropped from $4.8 trillion to $3.5 trillion.

Unidentified speaker (speaker_3) ·12:05

How to verify: Nvidia stock price data from Nov 2026; calculate market cap using shares outstanding. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Primary source

Nvidia trades at 20x trailing earnings and 11x forward earnings.

Unidentified speaker (speaker_0) ·12:54

How to verify: Financial data providers (Bloomberg, FactSet) for Nvidia P/E ratios as of Nov 2026. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Primary source

Coca-Cola trades at 29x trailing earnings and Walmart at 46x.

Unidentified speaker (speaker_0) ·13:14

How to verify: Financial data providers for KO and WMT P/E ratios as of Nov 2026. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Anthropic's founder holds a single share with 50.01% of voting rights.

Unidentified speaker (UNKNOWN) ·26:37

How to verify: Anthropic S-1 filing; check voting rights structure and share class details. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Primary source

Anthropic has compute power commitments that allow it to get ahead of OpenAI in setting up its lane.

Unidentified speaker (UNKNOWN) ·32:05

How to verify: Check Anthropic's public disclosures or press releases regarding compute partnerships and capacity commitments. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Speaker_3 computed a debt value for Anthropic's leases of close to $500 billion.

Unidentified speaker (speaker_3) ·32:25

How to verify: Review Aswath Damodaran's blog 'Musings on the Markets' or his valuation spreadsheets for Anthropic to find the specific lease liability calculation. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Speaker_3 arrived at a $1.475 trillion value for Anthropic's equity after subtracting the $500 billion debt value.

Unidentified speaker (speaker_3) ·32:45

How to verify: Check Aswath Damodaran's valuation blog or published valuation models for Anthropic. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Speaker_3 used a 9.74% cost of capital for Anthropic.

Unidentified speaker (speaker_3) ·33:02

How to verify: Check Aswath Damodaran's valuation blog or spreadsheets for the specific cost of capital input used in the Anthropic model. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Anthropic's revenue growth slowed from enormous levels to 47% and potentially 15-25%.

Unidentified speaker (speaker_0) ·34:39

How to verify: Check Anthropic's quarterly financial reports or earnings releases for revenue growth rates. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Anthropic plans to enter the drug business, potentially competing with pharma companies.

Unidentified speaker (UNKNOWN) ·36:49

How to verify: Check Anthropic's press releases, blog posts, or news articles regarding their expansion into drug discovery or biotech. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

The S&P 500 is up around 7% year-to-date.

Unidentified speaker (speaker_0) ·46:41

How to verify: Check S&P 500 performance data for the year 2026 up to the podcast date (2026-11-20). For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

The Mag 7 is flat or down year-to-date.

Unidentified speaker (speaker_3) ·47:34

How to verify: Check performance data for the 'Mag 7' stocks (Apple, Microsoft, Alphabet, Amazon, Nvidia, Tesla, Meta) for the year 2026. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Do your own research

  • Verify Nvidia's Q3/Q4 revenue guidance and gross margin trends against TSMC's capacity allocation reports.
  • Check Anthropic's S-1 filing for the exact breakdown of the $900B compute commitment and lease liabilities.
  • Review the S-1 for Anthropic's governance structure, specifically the voting rights attached to the founder share.
  • Compare Anthropic's revenue growth rate (47% vs 15-25%) against peer growth rates (OpenAI, Google, Microsoft).
  • Analyze S&P 500 earnings reports to isolate the portion of earnings growth attributed to AI CapEx pass-throughs.
  • Assess potential regulatory actions in China regarding Nvidia's export restrictions and their impact on revenue.
  • Look for evidence of non-Nvidia chips (e.g., AMD, custom ASICs) gaining market share in hyperscaler data centers.
Full transcriptexpand

0:00Unidentified speaker (speaker_0): Support for the show comes from Xero.

0:03Unidentified speaker (UNKNOWN): The

0:03Unidentified speaker (speaker_0): big companies we analyze on this show don't operate according to different financial criteria than everyone else. The metrics that drive their performance are the same as the ones that drive the performance of your small business. You just need the tools to see them. Xero, spelled with an X, is a simple but powerful platform built for small business owners that makes managing business finances easier from day one. Accounting, payments, payroll, and analytics are available in one connected platform. Join the five million customers worldwide using Xero. Learn more at Zero.com. That's XERO.com.

0:45Unidentified speaker (speaker_1): Listen to me.

0:46Unidentified speaker (speaker_2): Markets are bigger than us.

0:47Unidentified speaker (speaker_1): What you have here is a structural change in the world distribution. Cash

0:50Unidentified speaker (UNKNOWN): is

0:51Unidentified speaker (speaker_1): trash.

0:51Unidentified speaker (speaker_3): Stocks look pretty attractive.

0:53Unidentified speaker (speaker_1): Something's going to break. Forget about it.

0:55Unidentified speaker (speaker_0): Welcome to Prof. Markets. This has been one of the wildest weeks in business in a long time, and there is no shortage of news to unpack. We finally heard from Jensen Huang after his departure as Nvidia CEO. He made clear that he was open to returning, but the board said during the earnings call that he would not be coming back. Meanwhile, Anthropic's highly anticipated IPO adds another major development to an already momentous week. This is also our last chance to catch up with our next guest before the end of the year. So we're going to get his take on all of these developments and look back at some of the biggest lessons from 2026 and what he thinks investors should be watching as we head into 2027. So let's get right into our conversation with Professor Aswath De Motoran, the Kirshner family chair in finance education and professor of finance at NYU's Stern School of Business. Professor DeMotoran, great to see you. Thank you for joining us. I'm going to start with the thing that everyone's talking about, which is NVIDIA. Lots of directions we can take this, but I'm going to keep gonna it keep it open-ended. What do you make of this? What are your thoughts?

2:07Unidentified speaker (speaker_3): It's a story with multiple dimensions to it. If it was just Jensen's personal foibbles, and you were thinking about, you know, the company setting him aside because of he's breaking corporate truth by having affairs with employees. This would be a much smaller story. Because in many ways you can talk about who's going to report replace him and you know what they're going to do that's different and how much he could work with them. I think what makes the story potentially more dangerous for NVIDIA investors is the China twist, which is the other affair. I'm not going to pass judgment, I'm not going to say, you know, but clearly it's broadened something into the story that is not going to go away, even if you got the perfect replacement CEO. In fact, I was just finishing up a piece last Sunday, and I'm glad I didn't post it because so much changed over the last four days on how to value a CEO of a company. How much value does a CEO bring? And I start with the general presumption, which is CEOs matter more at young companies, growing companies, and at mature companies. Because there's far less you can do as a CEO. to change value at a mature company.

3:17Unidentified speaker (UNKNOWN): And

3:18Unidentified speaker (speaker_3): if you start with that premise, the question is where is Nvidia in the lifecycle? It's a huge company. But if you look at its growth trajectory, it's behaving like a young startup in terms of growth. Clearly, CEOs matter at that stage. And especially if a CEO's founded and essentially built a the company in his own image,

3:36Unidentified speaker (UNKNOWN): you

3:36Unidentified speaker (speaker_3): can hold two competing thoughts at the same time. One is that Jensen's been an amazing CEO. I mean, he's built one of the great companies on the face of the earth. And at the same time, you can say, well, clearly he's human and he makes some, he's made some very big mistakes that is costing him and the company at the same time. Because I think there are people who pass who are passing judgment on Jensen and also then saying he wasn't that great of a CEO because of what he's done now. But I think we need to keep the two things separate. So at this point, if you ask me, what is the effect of in of Jensen leaving or potentially because now that's become uncertain as well. Is he actually leaving or is he just um coming back? And I think that adds a layer of uncertainty that can't help investors because you now don't know whether he's coming back. or not coming back. I think the odds are high that he can't come back, that it's too difficult to come back from something like this. But there's still the question of what will it mean for NVIDIA's value, whether he comes back or not, to you know how much damage has been done. And that's where the China story has a lasting impact.

4:47Because I think there are going to be hearings in Congress. It's almost a given. There are going to be potentially more examination of what exactly happened. And I think that at the end of the process, it might very well mean that NVIDIA has to give up some operating value to get back into the good graces. That might mean giving up sales to China. I mean, NVIDIA, even in its existing structure, has sales to China. Will that mean that half or two-thirds of those sales go away? What will that do to NVIDIA's value? It might also affect their relationship with companies in Asia, which can affect their margins and their value. So I think the China story has legs no matter who runs the company. And what we will be seeing in the next few weeks is how much of that story is true and how much is just. you know, people essentially coming to a judgment far too quickly.

5:42Unidentified speaker (speaker_0): I'm so glad to have you on while this has happened, because so much of your work is about evaluating things that are important to a business, but that are largely intangible. I mean, I remember reading one of your great posts on trying to value the risks associated with a company being located near to a volcano. And how do you assign numbers to that? How do you figure out how that is reflected in the price? And that does seem to be exactly what is affecting this company right now. And it just happens to be there's one of the most valuable companies in the world because you've got the value associated with how valuable is Jensen Huang as a CEO. You've got the risks of is this guy gonna be tried as as a criminal? Is he gonna what is going to happen in his congressional investigations? What is Congress going to do about NVIDIA as a company and whatever ties they may or may not have had with China? So I I'll just focus on the CEO question because this

6:42has really animated investors. A lot of people believe that without Jensen, it's over, or at least a significant

6:51Unidentified speaker (UNKNOWN): uh

6:51Unidentified speaker (speaker_0): amount of of the value of the company disappears without him, which is why we've seen literally hundreds of thousands of people signing a petition to bring him back. I'm skeptical of it, but what is your estimation of the value of Jensen Huang to this business today?

7:08Unidentified speaker (speaker_3): I think today the effect of just the CEO. So let's put the China story to the side. If it's just a CEO replacement, the effect is going to be small. Partly because there are three things that are going to happen, assuming he's going to get replaced. A new CEO is going to come in. It might be somebody from NVIDIA, either who works at NVIDIA right now or used to work at NVIDIA. The second is to the extent that this is not a hostile environment, Jensen's still going to be around. He might be able to provide strategic help to that CEO. And the third is, how good of a management team has he built? I mean, if I asked you how critical Jeff Bezos was to Amazon, in 2005, you'd say, hey, there's no way this company can live without Jeff Bezos. He is the company.

7:55Unidentified speaker (UNKNOWN): But

7:55Unidentified speaker (speaker_3): he walked out in what, 2017, 18? You see him in uh wearing speedos on a yacht in France. And the company goes on, and that is what good founders do is they create a second layer of management that can do it. We're going to find out what that second layer looks like at Nvidia. I think, notwithstanding all of the bad stuff that's come out of NVIDIA, the sense you get from NVIDIA at the second layer is there are competent managers in place, and that many of them will be able to carry the operating efficiency. efficiencies, the competencies of Nvidia going forward. What they will miss is Jensen's been that the culture that drives how NVIDIA thinks about big markets. If you think about the last 20 years, I've called Nvidia the opportunistic chip company. It seems to find big markets just as they get started and it's uniquely positioned to take advantage of them as they grow. It happened with gaming, it happened with crypto, and it happened with AI. The first time around, you can say you got lucky. The second time you got really lucky. The third time around, you have to say there's something in the company's DNA that allows it to find these new markets ahead of people. And I think a big part of that comes from Jensen and what he brought to the company. That is essentially what you lose with. somebody

9:15else's CEO. How much will that mean? Less than it did when they were a hundred billion dollar company. When you have 4 trillion, you're not talking about opportunities that are going to quadruple your value. Your value effect is going to be smaller because you're building off bigger scale. So I think you can have a scenario where the effect is more benign. It's short term, the new management comes in, Jensen hangs around, he provides the strategic. direction. You don't have a loss of employees because he's left. And the second layer of management essentially does its job. It's not that there'll be no effect. I, you know, I I created a spreadsheet where people could decide for themselves what the effect is on two dimensions. One is is the effect just short-term or is it long-term? If the effect is just short term, the effect on value is 3%, 5%. If it's long term, then I push. Where is it going to shop? Is it less revenue growth? Is it lower margins? Will they have to reinvest more because with because without its connections, you might have more reinvestment? And I let people turn those switches off and on. And I have a table which essentially says: if this causes revenue growth to drop by 10% and margins to decrease by 5%, you're going to see a much bigger impact on value. But that's the kind of effect that you're trying to trace to. And that's why I said if it were just

10:34personal foibles, even with those breaking corporate rules, the effect would be far smaller. The wild card here is: what does China do? Because that's out of the hands of any CEO or second layer management, or what Jensen decides to do with the new CEO. Because then it goes. well, even if it doesn't end up in the criminal courts, it ends up in regulatory, you know, places where you don't want to be as Nvidia, where there are restrictions and what you cannot do that can affect your value as a company. And that's what I would worry about as an Nvidia investor. And unfortunately, bringing Jensen back, and this is what I would emphasize, is not going to make the China problem go away. In fact, it might make it worse. And that I think is what the board is deciding. And that's why I think it's so critical that you break the scandal into two parts. One is the corporate scandal. You broke corporate rules. The other is potentially the question of did was there some evasion of laws and and restrictions they should have followed that they broke. That's the second is a more uh I think the more damaging.

11:42Unidentified speaker (speaker_0): I mean, this question of what value do investors assign to the loss of Jensen Huang and also the risks associated with an investigation into the extent to which they were doing this outside of megaspeed. Again, nothing's proven in court yet, but this appears to be what happened based on the reporting. Investors said it was. worth one and a half trillion dollars, essentially.

12:05Unidentified speaker (speaker_3): It's a drop in value. Just uh so they've right out of so you I think it today Nvidia's market cap is three and a half trillion. It was four point eight trillion not that long ago.

12:15Unidentified speaker (speaker_0): Right.

12:15Unidentified speaker (speaker_3): Four and a half trillion at the start of the scandal. Investors have clearly decided that this is not something that's going to blow over. And I think they're right. I have a feeling. that if the China, and this is where the China wildcard comes in here, if the China scandal essentially turns out, it fizzles. It turns out to be a nothing burger, then I think that this is an overreaction. So at this point, the Nvidia investment becomes a bet on what do you think about the China scandal? Candle And how much do you think it'll last? Because that is what can explain a trillion dollar drop. in market cap. Just changing and see here doesn't do it for me.

12:54Unidentified speaker (speaker_0): Just looking at at the valuation of NVIDIA at this point, trading at 20 times earnings on a trailing basis, 11 times earnings on a forward basis, it is trading at a significantly lower multiple than Coca-Cola, which trades at 29 times trailing earnings, uh, trading a significantly lower multiple than Walmart at 46 times. I mean, it is quite striking how these traditional legacy, as you say when you talk about the corporate life cycle, these old middle-aged companies that you wouldn't think would be receiving these premiums on their multiples, and and NVIDIA's way behind, despite the fact. that it is at the cutting edge of AI. This, I think, implicitly brings up the question of whether this is a buying opportunity, whether there is some dislocation here. What do you make of the valuation at this point? Aaron

13:52Unidentified speaker (UNKNOWN): Powell

13:52Unidentified speaker (speaker_3): Here's the key difference. Coca-Cola is a boring company, but its business model is set in store. Its margins are stable. So if you told me this year's margins, I can pretty much predict what the margins are going to be for the next 10 years. NVIDIA has an astonishing business model. A business model that if I described it to you before NVIDIA came along, you'd say there's no way a company can pull that off.

14:17Unidentified speaker (UNKNOWN): I mean,

14:17Unidentified speaker (speaker_3): guess what it does, right? It designs the chips, it gets TSMC to make the chips. So it doesn't have any of the manufacturing facilities. It marks those chips. up 800 to 900%.

14:30Unidentified speaker (UNKNOWN): And

14:30Unidentified speaker (speaker_3): it sells it to Meta and Alphabet and Microsoft. This model is rife with tension at both ends, right? If you're TSMC, you're saying, why are we letting Nvidia take all of this upside? And if you're Meta a or Alphabet or Microsoft, you're saying, why am I accepting these marks up? And the answer is right now. you're so desperately in need of building data centers, and you've decided that NVIDIA chips are the only way to do it, you're willing to accept this business model. That's always been the case, even in its glory days. The question is not whether the margins will close, but what are the forces that'll cause it to go close sooner rather than later. So it's a model that's destined to see. a margin decline over time. I'm not saying it's going to collapse. I mean remember, the typical chip company has operating margins of 40%. Nvidia's marg operating margins are 70%. That's amazing. That's why we give them the premium. But the 70 cannot stay 70 because you're going to get tensions on both ends. Med and Microsoft are working on their own chips because that way you don't have to pay the premium. I'm sure TSMC is trying to negotiate a better deal for itself because they know without them the chips will not be made.

15:46So I think that that's where the personal connections might become a factor for NVIDIA because NVIDIA's connections with TSMC go back 30 years. You've read the stories about how NVIDIA is a small company, reached out to TSMC. There is a relationship between TSMC. TSMC's management and not NVIDIA's management, but Jensen Wong that is personal.

16:08Unidentified speaker (UNKNOWN): So

16:08Unidentified speaker (speaker_3): the question is, are all of those things going to start to come under stress sooner because Jensen's no longer there? So I think you're right. It looks cheap, but it's also built on a precarious business model where you keep asking, will I mean I remember there was a an there was a part in the anthropic, um, I don't know whether it was an actual prospectus and whether there was an overstatement, that they'd found a way to get Claude to take traditional non-invidia chips. And with the power of AI, they got the NVIDIA ch the non-invidia chips to do what NVIDIA. That is a terrifying thought. Because if AI can become Becomes self-perpetuating and it can replace Nvidia chips with much cheaper chips, there goes your business model.

16:51Unidentified speaker (UNKNOWN): So

16:52Unidentified speaker (speaker_3): the NVIDIA great as the company looks on paper. And it's an amazing company in terms of growth and margins. You keep holding your breath saying, what'll cause this model to implode? And I think the market right now is holding its breath. It's worrying about the things that can cause it to implode. So 11 times earnings might look cheap unless there's an implosion. So I think you've got to build in some kind of a risk effect here into the earnings, which will make the 11 times earnings go back to 25 times earnings if you build in a breakdown of the business model.

17:21Unidentified speaker (speaker_0): When you look at at all of those questions, and this is part of the problem with the stock right now, is that there are so many questions that are so significant to which no one actually knows the answer because you can't. What do you make of stocks like that as an investor? I mean, is that the kind of situation where actually you can't answer enough questions, so you can't invest?

17:44Unidentified speaker (speaker_3): No, that's not true. I think this is the kind of stock where concentrating your portfolio is deadly. I mean, I've said that about Nvidia through the glory days. Don't have a third of your money, half your money in Nvidia, no matter how great you think the company is. And here's the reason it's rife with uncertainties. You're absolutely right. You can make your best estimate, but you're not you knew you know you're going to be wrong 100% of the time. You could convince me that NVIDIA is cheap, but if I have to put all my money in NVIDIA, there is zero chance I'm going to do it because there are too many unknowns that can come back to bite

18:15Unidentified speaker (UNKNOWN): me.

18:16Unidentified speaker (speaker_3): But if I'm adding NVIDIA to a portfolio of 30 companies,

18:21Unidentified speaker (UNKNOWN): the

18:21Unidentified speaker (speaker_3): uncertainties become a lesser factor because you're using the law of large numbers in your favor. I did put a limit by on NVIDIA at 140. I've sold all my NVIDIA holdings that I bought in 2018 over the last two years. But at 140, I'm willing to live with the uncertainties because I think there's enough power left in the business model for me Be to make my money plus more. But that's because I have a portfolio that's diversified. And I think that if you're going to make a bet on Nvidia, make it from a position where it's not 30 or 40% of your portfolio. You reduce your upside, but you can live with the downside. And you need to be able to live with the downside with the stock, at least in the near term,

19:08Unidentified speaker (UNKNOWN): for

19:08Unidentified speaker (speaker_3): you to be able to make your money

19:10Unidentified speaker (UNKNOWN): back. We'll

19:12Unidentified speaker (speaker_0): be right back after the break. And by the way, we're recording our annual Ask Me Anything episode soon, so be sure to submit your questions for me and Scott. We can leave them in the comments on YouTube or Spotify, or you can email them to markets at profgmedia.com. Support for the show comes from zero. Numbers don't lie, but there are easier and harder ways to pass them. If you're running a small business, you already have enough hard things on your plate, like trying to decipher the napkin scribble that you circled and wrote important above. Make keeping track of your financial numbers easy with Xero. Zero, spelled with an X, is a simple but powerful platform built for small business owners that makes managing business finances easier from day one. It makes accounting, payments, payroll, and analytics available. in one connected platform. Xero also lets you tailor your data for your business's specific needs. Xero's reports, visual dashboards, forecasts, and AI

20:12insights can all help business owners better understand cash flow and make smarter decisions. The information is presented in ways that feel like they were designed for actual humans to read and understand. According to a 2024 customer survey, conducted in the US, the UK, and Australia, 91% of customers agree that Xero is easy to use. We don't have data on what percentage of people think napkin scribbles are easy to use, but I feel comfortable guessing that the answer is lower. Join the 5 million customers worldwide who love using Xero. Learn more at Xero.com, zero.com. that's X x-e-o.com. E-R-O.com.

20:56Unidentified speaker (UNKNOWN): We're

20:56Unidentified speaker (speaker_0): back with Prof. Markets. I'm gonna move us on to anthropic, which just priced its IPO at $83 per share. We're recording this just before the stock actually goes out. So when this episode air, airs. it will be trading. But what we know at this point in time is that it is going to go out at roughly one and a half trillion dollars market cap.

21:21Unidentified speaker (UNKNOWN): Um,

21:22Unidentified speaker (speaker_0): you valued this company earlier and you valued it at just just below that $1.475 trillion.

21:32Unidentified speaker (UNKNOWN): Um,

21:33Unidentified speaker (speaker_0): so uh

21:34Unidentified speaker (UNKNOWN): I

21:34Unidentified speaker (speaker_0): guess walk us through your valuation there and then we'll get into how you think this thing will trade over the next 12 months.

21:41Unidentified speaker (speaker_3): I valued enthropic on November 2nd, the day before the midterm elections at 1.475 trillion. The reason I emphasize that is you're saying that's three weeks ago. That wasn't that long ago. But it's a three-week period, which was extraordinarily eventful from an enthropic anthropic AI standpoint.

22:00Unidentified speaker (speaker_0): Yes.

22:01Unidentified speaker (UNKNOWN): You have

22:01Unidentified speaker (speaker_3): an election that's brought in a Congress, a legislature, that is going to be much more hostile to both data centers

22:10Unidentified speaker (UNKNOWN): and

22:10Unidentified speaker (speaker_3): to AI as a technology.

22:12Unidentified speaker (UNKNOWN): It's

22:13Unidentified speaker (speaker_3): going to get more expensive to build data centers. One of the inputs in my model is how much will Anthropic have to spend as incremental investment in compute power to get its revenues? Because I told a big uplifting story about anthropic. A story where they can get to a trillion dollars in revenues. And just from perspective, that would make them the large if in there's no other company right now on the face of the earth that makes more than 800 billion revenues. You know, the closest you get is Amazon and Walmart. I'm giving them a trillion dollars in revenues. I gave them margins that reflect premium pricing in the AI market, though I adjusted for the fact that there might be a regulatory drag. So end result is a margin of 27.5%. And I did assume that they would be able to keep adding compute power, if not easily, at about the same rate as they do today.

23:04Unidentified speaker (UNKNOWN): That's

23:04Unidentified speaker (speaker_3): a 1.475 trillion. I revalued anthropic about 20 minutes before I came on here.

23:13Unidentified speaker (UNKNOWN): I

23:13Unidentified speaker (speaker_3): made two changes. One is I said it is going to get more expensive, at least in the near term, to build data centers and compute power. So basically you will get less revenue per dollar. And the other is the drag from, I don't know what form the regulatory drag is going to take, but it's going to make your margins lower, higher costs. So slightly higher drag, slightly more reinvestment. That lowers my estimated. aggregate value of equity by about 100 billion. So I'm moving in in the other direction. I'm going towards 1.35 trillion. So difficult to get the word trillion out of my mouth. You know, such an uncommon word. And I did compute the value per share this time. I could not do that on November 2nd because the prospectus This still hadn't fleshed out share cut. count. Now we have the share count, share, you know, class eight. A, class N, class D, class E, and of course the famous founder share.

24:07Unidentified speaker (UNKNOWN): So

24:08Unidentified speaker (speaker_3): I bring in the updated share count. My estimated value, if you bring in the share count and the fact that there were a big chunk of options outstanding, came up with a value of $77. That was five minutes before I came on, is I I got that number. I hadn't seen the $83 because the pricing came out just as I came on. here. It's $83, I think, is the listed price.

24:29Unidentified speaker (speaker_0): Correct.

24:29Unidentified speaker (speaker_3): That terrifies me. It terrifies me that I'm so close to a banker's price. I'd be much more comfortable. But it tells me that, I mean, to to show you the trajectory of that banker pricing, if you go back to the start of last week, they were shooting for like 1.15, 1.25 trillion. It went

24:50Unidentified speaker (speaker_0): from two 2 trillion to 1.1 trillion, and now we're in the middle.

24:55Unidentified speaker (speaker_3): And the question is, what changed? And people think that bankers sit in little alcoves, valuing anthropic with business models. Here's what changed: they put the 1.1 trillion to 1.25 trillion out there on Monday and Tuesday. They heard from a lot of institutional investors, and they heard too much enthusiasm.

25:15Unidentified speaker (UNKNOWN): So

25:15Unidentified speaker (speaker_3): they repriced it. This is a pricing game. And my guess is if they've done just the job right. There's only one job bankers do in this IPO, which is to get the pricing to a level where tomorrow, when it opens, you get a pop.

25:30Unidentified speaker (speaker_0): Yes.

25:31Unidentified speaker (speaker_3): 10%, 15%.

25:33Unidentified speaker (UNKNOWN): So

25:33Unidentified speaker (speaker_3): if they've done the their job, I would expect the price to open up, you know, $90, $95, whether it's it stays there is different because this um anthropic also has unlocking happening right from day one. You're allowed you people are going to be allowed to sell their shares. So there's going to be this pressure from people trying to cash out and people trying to get in. It's going to be really, really busy at that exit door and entrance door. It's the same door. People are coming in, people are leaving. So it'll be interesting to watch what the stock does during the course. of the day, not just at opening.

26:05Unidentified speaker (UNKNOWN): Because

26:05Unidentified speaker (speaker_3): the net effect, what it closes at tomorrow, would be the number that I would be comparing my 77 to. The offering price is irrelevant, the opening price is irrelevant. What it ends up to tomorrow evening. And if it ends up tomorrow evening at 71 or 72, which can happen, that's a really bad outcome for banks. Then it starts to get interesting for

26:29Unidentified speaker (UNKNOWN): me. But

26:30Unidentified speaker (speaker_3): it needs to drop a lot before I buy it because the governance structure is one designed in hell for shareholders.

26:37Unidentified speaker (UNKNOWN): With

26:37Unidentified speaker (speaker_3): uh since Google, we've had voting shares and non-voting shares. We're used to that, right? This is a single share held by seven founders with 50.01% of voting rights. The closest thing I can get to comparing it. to is golden shares that were issued by governments when they privatized companies in the 80s and the 90s. Margaret Thatcher introduced it in some of the British privatizations. And what a golden share gives the government, it's a single share, but it gives the government veto power over big decisions, acquisitions, expansion into new geographies. The anthropic founder share

27:14Unidentified speaker (UNKNOWN): is a

27:14Unidentified speaker (speaker_3): golden share. with even more power. And I don't want to give governments and regulators ideas, but I don't know whether the anthropic founders thought through this. But let's say that you have an activist government, a regulator, and they decide at some point that AI is doing things that are not in the public interest. And there will be scandals, there will be breakdowns of and things. What if that activist regulator? then says we're going to expropriate that one founder share? And we will pay the market price on the class A shares.

27:47Unidentified speaker (UNKNOWN): That

27:47Unidentified speaker (speaker_3): one founder share should be worth a couple of hundred billion.

27:52Unidentified speaker (UNKNOWN): But

27:52Unidentified speaker (speaker_3): the regulator's goodness is not traded. We'll pay the price of a class A share. I know the founders can contest this in court and argue it's expropriation rightly, but this is. I would worry about the target you've given an activist regulator or government that looks at that share saying, we'd like to control. We'd like to, I mean, this talk of making AI a utility, the easiest way to do it is actually buy that founder share.

28:20Unidentified speaker (UNKNOWN): And

28:20Unidentified speaker (speaker_3): the government then essentially controls the big decisions of the company.

28:25Unidentified speaker (UNKNOWN): No,

28:26Unidentified speaker (speaker_3): it's uh it's it's I think this the way it's been structured. I'm not sure what the incentive was to set it up this way. I'm sure there was a reason for it. But what I see from the outside is a target. And that target might effectively make that founder share something that's covered It by regulators and governments.

28:47Unidentified speaker (speaker_0): It's really interesting. I mean, I I I would think that the reason they set it up that way. was to consolidate and protect power. But to your point, they might have actually made themselves more vulnerable to giving up that power. But going back to just looking at the anthropic business and and your valuation, one of the risks that I've been obsessed with

29:09Unidentified speaker (UNKNOWN): Um,

29:09Unidentified speaker (speaker_0): is the fact that they told us that over the next several years they're going gonna to spend spend around $900 billion on compute and infrastructure. And

29:21Unidentified speaker (UNKNOWN): they

29:21Unidentified speaker (speaker_0): said that, you know,

29:23Unidentified speaker (UNKNOWN): uh

29:23Unidentified speaker (speaker_0): by before 2030, so that over the next three years, it's gonna be around $400, $500 billion in spending. And a lot of that spending they cannot get out of, or at least it's it's it's contractually set.

29:37Unidentified speaker (UNKNOWN): So

29:38Unidentified speaker (speaker_0): they're going to have to spend hundreds of billions of dollars, which is a completely different business from the business it is today, where they're spending a couple or tens of billions of dollars. They're something spend around $9 billion in the first half. Maybe it'll be a little higher. So the entire cash flow profile of the business is about to completely transform. And we're going to see essentially an entirely different company, which means that we're going to need to see a company that is either one, going to have to raise hundreds of billions of dollars from the capital markets, which basically opens yourself up to the whims of investors who can be very annoying and moody, or they have to generate, as you said, a trillion dollars in revenue. And I'm not so sure of that, or I'm not convinced. So, how do you think about those risks going forward?

30:30Unidentified speaker (speaker_3): Let me separate the two issues. One is the compute power commitments are an issue. They're an issue because they're a mismatch. Mismatch in what sense young companies, companies that are not making money should not be borrowing money. That's a first corporate finance principle. And in effect, anthropic Enthropic is violating that principle by entering into contractual commitments without the earnings tobacco. That'll put most young companies out of business, right? There are two things that I think will buffer Enthropic from the risk. The first is after this IPO, their cash balance should be about $200 billion. That's going to cover the big charges for the next two or three years are the compute commitments. They're covered for the next three years.

31:14Unidentified speaker (UNKNOWN): That

31:14Unidentified speaker (speaker_3): gives them breathing room.

31:16Unidentified speaker (UNKNOWN): But

31:16Unidentified speaker (speaker_3): it's breathing room only if the revenues very quickly start to ramp up. Because getting to a trillion in 10 years means that you have to be basically doubling revenues pretty much every year, at least in the near

31:29Unidentified speaker (UNKNOWN): term. So

31:30Unidentified speaker (speaker_3): this is the combination you're going to be looking at. If revenues climb, it's not a problem. If revenues don't climb,

31:38Unidentified speaker (UNKNOWN): then

31:39Unidentified speaker (speaker_3): of course they're going to be squeezed. Because it's not that the compute commitments are a problem, it's that they can't get to the revenues they need to justify the commitments.

31:48Unidentified speaker (UNKNOWN): But

31:49Unidentified speaker (speaker_3): I can see why they did it. In this business, the AI business, I think compute power is going to be become the scarce resource. They were way behind OpenAI a year ago in terms of compute power. lined up.

32:05Unidentified speaker (UNKNOWN): It's

32:05Unidentified speaker (speaker_3): about a third of the compute power. They've caught up and they've gone ahead of OpenAI with the commitments. So to them, this is their way of saying we've got our lane set up. So when we grow, we'll have the compute power. Because we don't have the compute power, even if the growth is there, we will not be able to deliver on the growth because compute power will become scarce. So I think it's a reflection. of them seeing, looking down the road and saying, we need to lock up that compute power. But it is a risk in the company. In fact, I computed a debt value for the leases of close to $500 billion and I subtract it from the equity. So when I came up with a $1.475 trillion value for the equity, that was after I subtracted out the debt value of $500 billion.

32:50Unidentified speaker (UNKNOWN): But

32:50Unidentified speaker (speaker_3): you're right, the key assumption in this whole story that because I opened up the when I did my anthropic evaluation, I made it an open source so people could come in and pick their own story.

33:01Unidentified speaker (UNKNOWN): And

33:02Unidentified speaker (speaker_3): I let them say this is the input that I disagree with on the most. And I was surprised that so many people made this about the cost of capital. That I used to 9.74% cost of capital. And be quite honest, that wasn't even my top. 10 list of worries. As around the 90th percentile for US companies, the risky company. But I think people in their gut want to give cost of capital 14, 15, 16% for risky companies. But your gut doesn't equate to real. You've got to live in the world you're in. In the world we're in, at 10% cost of capital is at the 90th percentile, I gave give them a pretty high cost of capital. To me, the central input that drives this valuation. the number that I would watch the most, is the revenue number. If I can get towards that trillion dollars in revenues, everything else gets taken care of. My profits will come, I'll be able to meet my compute commitments, my cost of capital will stay low, default risk will go away. If revenues don't start to come in, everything else falls apart. So this is a company that drives writes almost entirely on will there be enough of a market for AI products and services or clawed,

34:09clawed agents around the world, that I could keep doubling my revenues every year for the next five years so I can get to this trillion dollar end revenue. Of all the numbers I'll be watching in the financial statements going forward, that's a number that I will focus on because that's the a number that drives everything else back. in the valuation.

34:28Unidentified speaker (speaker_0): I a hundred percent agree, which is why I was concerned to see that the revenue growth is slowing down as much as it is. We had, I forget what

34:38Unidentified speaker (UNKNOWN): um

34:39Unidentified speaker (speaker_0): Q1 revenue growth was, but it was enormous. And it came down to 47% growth, which is still obviously huge, but it was a but it was a significant slowdown. And now it appears that. they're gonna it's gonna come down to between 15 to 25%. We'll see what happens

34:55Unidentified speaker (UNKNOWN): um

34:55Unidentified speaker (speaker_0): in Q4. And

34:58Unidentified speaker (UNKNOWN): I

34:58Unidentified speaker (speaker_0): mean, if they can continue that, I guess we'll look we're looking at a doubling next year, which I guess is good, but that assumes that it doesn't continue to slow.

35:08Unidentified speaker (UNKNOWN): I

35:08Unidentified speaker (speaker_0): mean, don't they need to change something quite substantively to get those numbers back up?

35:13Unidentified speaker (UNKNOWN): If

35:13Unidentified speaker (speaker_3): you asked enthropic, Anthropic, they would say. it's because we ran out of compute power. That we were, you know, that was a combination of OpenAI underpricing us and compute power becoming a scarce resource. I think the truth is both can be true. OpenAI is going to continue to be that. They keep underpricing, you're going to have an issue because you're going to have to bring prices down to massive prices. And this isn't even opening the door to the other source of competition, which is open source models that are less powerful but just as effective. Most businesses don't need Claude, even if they decide to use AI in-house to do what they need to. It's overkill. And I think one of the things Anthropic has to learn is not build frontier models that are at the that that. essentially state-of-the-art models where it's like building a car with every single potential add-on. When people just want four wheels and a vehicle that works, I think Anthropic has to learn to create models that are powerful enough, but not too powerful. Because why add that extra power? Put the extra knee. kne for compute power when you don't need to. There is, you know, there is a side business,

36:32and this I think is this fascinating side story. And I'm not even sure how the story fully plays out. Where Anthropic, in addition to providing AI agents, AI tools to other companies, plans to enter businesses as a player.

36:49Unidentified speaker (UNKNOWN): The

36:50Unidentified speaker (speaker_3): business which we're talking about is the drug business, where they say, look, we're going to actually not just provide agents, we're going to be providing, if not the farm doing what pharma companies are doing, do what those lead into pharma companies are doing. I mean, San Diego is full of these young biotech and pharma companies where you build up a drug and then you sell it to Pfizer or Merck or one of the lead players. And Anthropic seems to be convenient. convinced that they can enter businesses like that and essentially be able to do what young pharma companies are doing and make revenues that because now you're no longer a tool. You're not being paid for as a replacement or a tool. You're being paid for some for something you do. The problem with that story is if Anthropic tells its story about the drug business, that they're going to be a player.

37:39Unidentified speaker (UNKNOWN): If

37:40Unidentified speaker (speaker_3): I were a consulting firm, do I really want Claude to be somewhere in my. If I'm McKinsey and I invite Claude in, I use it, I'm worried about hey, what if Claude learns enough about my consulting resources that I am facing anthropic consulting and anthropic banking? And it actually gets in the way of the sales pitch that enthropic. is making to the rest of the world. How they balance that, I don't know. But I think part of the, I mean, in fact, much of the excitement I hear from anthropic bulls is it's usually a story about how how AI is going to give a cure for cancer much more quickly. And that anthropic will reap the benefits of that. Even if the first part of that is true, that enthropic and uh f AI. can find a way to come up with a cure for diseases quicker because if you pull the data together, do what human beings will have a tough time doing. I'm not sure the value created from that accrues to entropic. It might just be great for all of us because it's consumer surplus. But you might, in a sense, take a business and make it less profitable across, it might be bad for pharmaceuticals collec collectively.

38:51because you now have a player in the game with lower costs who's willing to underprice you. But I think people make too much too quick a leap from there's a lot of value that is created, and therefore anthropic will claim the value. There's going to be lots of claims on that value creation. And by the time you're done, I'm not sure what percentage of the value created from AI actually accrues to the AI companies.

39:13Unidentified speaker (speaker_0): I a hundred percent agree with that, which is why I'm interested that you saw a pathway to this company achieving a trillion dollars in revenue. Because when I look at the current growth rate, the risks that you lay out, I mean it's clear that in order to do everything that they want to achieve, given the amount that they're spending on compute, they're gonna going have to to enter new businesses the likes of which we haven't seen before. I mean, they're going to need to clearly dramatically change this business. And they need to get, they can they need, I mean, lots of different businesses that are going to be solving lots of different problems. And to me, so many stars have to align for that to occur. And to be clear, they have to align in order for this business plan to work. which to me is quite a bold pitch. So I guess on your end, how is it that those stars aligned for you?

40:12Unidentified speaker (speaker_3): I think ultimately this business is going to become a business with two or three big players in it. It's going to be a big business with two or three players. And with Anthropic, I see it as one of the players. It's the most focused of the AI company. companies in terms of saying, look, we have a, we know we're we're building things for enterprises. This isn't a neat thing that you're going to have on Zoom or some add-on. It's the most focused. And I have a feeling that once you get your feet into a company, let's say you're selling replacements for bankers, and Goldman Sachs invites you into the company. And the way you train is you give a the agent access to data as a bank and you know access to bankers and it learns from them. It's sticky. It's a business where once you get into companies, so I think being the first player here will give anthropic. That's inherent in my story. I could be very wrong. Maybe this is the kind of business where companies are going to essentially use agents, not give them access to data, proprietary data. But you need some stickiness in this model. And I'm assuming that stickiness will develop. And that stickiness has two side effects. One is it becomes much more difficult for later players to come in and underprice you because you're too immersed in the company.

41:32Unidentified speaker (UNKNOWN): I

41:32Unidentified speaker (speaker_3): mean, that's partly why if you make yourself, you know, where you look at Palinter and the government, it's once you get to build become part of that system, it's very difficult to think of how are you going to switch.

41:43Unidentified speaker (speaker_0): Good mode.

41:44Unidentified speaker (speaker_3): So I have, you know, my gut feeling is you're going to end up with two or you know, two, maybe three companies in this space with lots of stickiness in their models.

41:53Unidentified speaker (UNKNOWN): I

41:53Unidentified speaker (speaker_3): think there are lots of risks, lots of disruption risks that you have to face along the way. So it's uh the way I would describe the 1.475 trillion is it's my optimist hat is firmly on, and you needed to stay on for this to work. Which is the reason I'm also worried about governance, because it's also going to be business, even in its best case scenario where you're going to take a path and it doesn't work. You have to be willing to adapt and change and take a different path. That's where governance comes in, right? People pushing back saying you have to change. Yeah. I think the way anthropic is structured, I worry about whether the people running Anthropic will be willing to change because what's the incentive? You know, how do you make them change when they don't want to? Eventually, the market might be the only mechanism to get them to listen, the market price dropping. But that's a blunt instrument. I mean, you saw that with Facebook how much damage had to be created before Mark Zuckerberg gave on the gave up. up on the metaverse. If you multiply

42:53by that by 10, that's what you would need in the AI space to learn the same lesson. So I think that, you know, it's not that's why I said this is not an investment if you're thinking about five stocks to have in your portfolio.

43:06Unidentified speaker (speaker_0): Right.

43:06Unidentified speaker (speaker_3): But it's an investment I'm willing to consider at a discount Count on that 1.475 trillion as one company in the space, because if AI delivers on each even half of its promise, it is going to be a big business.

43:22Unidentified speaker (UNKNOWN): And

43:22Unidentified speaker (speaker_3): right now I can't think of because the rest of building the factory, right? The rest of the AI space is either companies building the data centers or companies supplying the stuff for the data centers. This is the first really big pure play product and service company. out there. And at the risk of worrying about missing out on this entire space. That's why I think there's an advantage to going first year as well. When OpenAI goes public next year, there will be a repricing.

43:54Unidentified speaker (UNKNOWN): And

43:54Unidentified speaker (speaker_3): there'll be more comparisons. But right now, there's nothing to compare it to. You're saying, what about SpaceX? SpaceX has clearly become a builder. It's not a product because almost all of XAI's revenues come from leasing out compute power space

44:07Unidentified speaker (UNKNOWN): to

44:07Unidentified speaker (speaker_3): anthropic and alphabet, not from you know revenues it generates as a standalone LLM.

44:14Unidentified speaker (speaker_0): We'll be right back. And for even more markets content, sign up for our newsletter at profitymarkets.com/slash subscribe. This is advertiser content from zero. Sometimes the hype cycles can make us forget that a stock represents a real business. But whether you're an investor or a business owner, if you understand how to draw insights from the financial data of the underlying business, the rest is noise. So I'm going to talk about some of the ways I evaluate a business. and I'll show you how Zero can help you apply the same criteria to your own business's finances. Xero, spelled with an X, is a platform that connects your business's accounting, payments, payroll, and analytics under one umbrella. Let's call this exercise Ed's Business Health Checklist. First on the list, nobody can or should know your business better than you do. With Zero, not only are profit and loss reports easy to generate, you can track trends over time. Second, you want a business that understands how its customers behave. With Zero, you can track your average revenue per customer, very important. You can even sort and visualize your top customer accounts across areas like sales, cost of sales, and expenses to help

45:31you figure out where the best opportunities for growth are. Third, you need to be flexible. As Dwight D. Eisenhower once said, plans are worthless, but planning is everything. And you know that's coming from a solid source because one of Dwight Eisenhower's best friends was supposedly a guy named Edward Elson. Staying on top of your finances isn't about sticking rigidly to a plan. It's about having the confidence to respond when things inevitably change. Zero is dynamically linked to your accounts with instant bank reconciliations. When something changes in your accounts, it's immediately reflected in Zero. Those are a few of the ways that I think about the health of a business, and a few of the ways Xero can help you keep your business healthy. Lesson adjourned. Learn more at referrals.xero.com/slash profgmarkets. At the time of recording, you can get 90% off for six months on a zero subscription. Terms and conditions apply.

46:32Unidentified speaker (UNKNOWN): We're

46:33Unidentified speaker (speaker_0): back with ProfG Markets. I'm just gonna zoom out

46:36Unidentified speaker (UNKNOWN): uh

46:37Unidentified speaker (speaker_0): and look at the broader stock market. So

46:40Unidentified speaker (UNKNOWN): um

46:41Unidentified speaker (speaker_0): we're coming up on the end of the year. The SP is up around 7% year to date, which is quite striking. after this incredible sell-off in AI. And part of my view of what's happening here, I mean, we're all wondering: when is this bubble going to pop? When are we going to see a correction? People seem to still be asking that question. I look at what's happened to Into AI. To me, the bubble has popped.

47:12Unidentified speaker (UNKNOWN): I

47:12Unidentified speaker (speaker_0): mean, NVIDIA losing more than a third of its value, neo clouds coming down,

47:17Unidentified speaker (UNKNOWN): uh,

47:18Unidentified speaker (speaker_0): a lot of these AI names just getting crushed. What's just striking though is the fact that the stock market as a whole is doing fine.

47:25Unidentified speaker (speaker_3): It's been the most impressive year I've seen in stock markets in the last decade. And that includes years where markets up 25%. The fact, I mean, at the start of this year, the story was as the Mag 7 goes, so goes the market. Well, this year the Mag 7 is pretty flat. In fact, it's down, the market is up. As tech goes, so goes the market. This year, the market's managed to go up. The fact that the market was has been able to hold its own, and it's a number that's going to be shifting as we're talking, right? It's it was up 10% just two just four days ago. It's now down, it's up 7%. I think that continued selling in AI will spill over elsewhere, right? But I think that the fact that it's held its own is extraordinarily impressive. That investors have been able to rotate out of technology into other businesses. Those other businesses have been able to absorb the pain of losing trillions of dollars on five or six companies and being be able to increase to offset. I'll be quite honest, I didn't think the market could pull it off. If you told me at the start of the year what AI and tech companies have done, you say, what will the market be? I'd say it'd be down 10%.

48:40That's how much the magnitude of the drop has been in AI and tech. The fact it's up 7% is incredibly impressive. And it shows you that the US economy has legs that go beyond AI. We've talked so much about AI for the last five years.

48:58Unidentified speaker (UNKNOWN): That

48:59Unidentified speaker (speaker_3): we forget how diverse the US economy is in terms of delivering value across the board, across different businesses. And I think it's a testimonial. I mean, if you want to look at a market that's been decimated by tech smeltdown, smeltown. take a look at the cost P. You can see what happens when you have a market that's concentrated in tech. And when the tech drops, there's nothing else that can absorb the pay.

49:21Unidentified speaker (UNKNOWN): The

49:21Unidentified speaker (speaker_3): US has been able to. Will it continue to do so? It'll be tough to keep doing this month after month. Tech and AI has to bottom out for the market to kind of find its legs again. But uh that's what we're going to be watching over the the rest of this year. I have a feeling the market's going to get tested as we go through the last few weeks of this year.

49:41Unidentified speaker (UNKNOWN): Well,

49:42Unidentified speaker (speaker_0): just looking at some of the sectors that have outperformed this year, uh consumer staples up 16%, healthcare up 18%. The biggest performer has been energy up nearly 40%, which uh you can make an argument that that is an AI trade, but also maybe some defensive in some ways. But you know, healthcare. consumer staples, I mean, these are it's the opposite. And so it seems as though investors have been rotating out of AI and rotating into some of these more defensive stocks. Um, and I guess that would explain the resilience. I mean, I guess two questions for you. One, how do we account for this extraordinary resilience in the stock market? And then two, now that the defensive stocks have gotten all of this juice in that this rotation, where what's left in terms of defense going into 2027? That seems to be the risk.

50:38Unidentified speaker (speaker_3): When you ask people how many publicly traded companies there are in the US, they often underestimate how many there are by a magnitude of three, four, five. There's about six 6,000 thousand publicly traded companies in the US across sectors. There isn't another market in the world that has the kind of spread across businesses of companies that are publicly traded. And I think that's part of the resilience, is and the resilience comes from the fact that AI, in spite of all of its issues, has created an economic boom in certain sectors in certain parts of the country. And those people have money to spend. They're spending it on consumer staples. And of course, healthcare is something you have to spend. It's not something you do discredit. It's part of a healthcare problem we have. I'm not sure that's good news that healthcare is up 18% because it suggests that there's fraying in the healthcare system. But I think it's a fact The that AI has also kept the economy afloat. It's estimated that. one are you know 1% of real GDP growth each year for the last four years has come from the AI boom.

51:46Unidentified speaker (UNKNOWN): So

51:47Unidentified speaker (speaker_3): we're reaping the benefits of that boom in other businesses because people have money to spend. You can't keep doing that though. I mean, as I said, you know, unless AI has to bottom out, if tech keeps dropping 10% every quarter and AI keeps really, you know, has days like yesterday or two weeks. ago on that Thursday. Markets can't hold up against that much selling and that much money because you know that's a lot of market cap leaving the market. Two trillion, three trillion, four trillion. That's tough to make up on a continuous basis. So what we're looking for as investors is when will this tech AI correction bottom up? And uh I think that's that's the question that's going to drive market discussions for the rest of this year and into next year.

52:33Unidentified speaker (speaker_0): Do you have an answer to that question?

52:35Unidentified speaker (speaker_3): I do think we're getting closer to the bottom than we were to the that than than we were just just a month and a half, two months ago.

52:42Unidentified speaker (speaker_0): Yeah, I agree.

52:43Unidentified speaker (speaker_3): You know, and I think that um I can't tell you exactly when it happens, but waiting for the exact bottom is often one of the is an exercise in futurity. You end up losing everything. So if you've never invested in AI and tech because you've always felt the Mag 7 was too expensive and you've had that view for the last 20 years, your portfolio is probably lost because of that view. Now's your chance to start adding those companies at prices Is much better than you'd have got two years ago, three years ago, four years ago.

53:14Unidentified speaker (UNKNOWN): So I think

53:14Unidentified speaker (speaker_3): people have to be willing to look past the bad news stories and say, I'm willing now to put some of my money. And as I said, don't make this all of your money. Make some of your money into tech, into AI. No, whether it's it might not be anthropic, it might be Palantir, it might be constellation energy, it might be the builders, it might be the suppliers. Pick a piece that you think is reasonably enough priced for you to enter and find ways. to enter that space if you've never been there.

53:41Unidentified speaker (speaker_0): Going into 2027,

53:43Unidentified speaker (UNKNOWN): I

53:44Unidentified speaker (speaker_0): think I mean I my prediction for 2026 is that returns would be meh, like single digit. Okay. I believe I think you had a similar prediction, and it's exceeded expectations. We'll see what happens uh at the towards the end of the year, what happens in December.

54:03Unidentified speaker (speaker_3): And I also said it'd be good to have a May year, is also what I said at the start of the year. It's good to have a single-digit return year

54:12Unidentified speaker (UNKNOWN): for

54:12Unidentified speaker (speaker_3): two reasons. One is it lets investors consolidate, think about what they've done. And second, it teaches people what risk actually means. Risk is a word for most people. The only way you understand what risk does is to live through. a downturn. If you're on an Nvidia four weeks ago, you know what risk is. I don't have to describe it for you because you've seen what it does to your portfolio. And that's good for investors to get that, you know, get that prod saying, hey, you can't overreach on risk because it looks good today, but what will it look like two years from now? So I think the May return. might vary. But with five six weeks left, let's wait and see. It could become double-digit returns or it could become a negative year very quickly because of the way things are moving. But I think it's going to be a year that is definitely not like the years we've had. We've had three, four years of not du you know, 15, 20% percent returns, which historically,

55:13Unidentified speaker (UNKNOWN): you

55:13Unidentified speaker (speaker_3): know, are you know we had. we've had a three year stretch which ranks up there with some of the best three year stretches in history. So I think it was time for an easing, time to take a breath, and I think 2026 the market eased up and took a breath and you got to give it credit that even though AI and tech had that big collapse, it was able to do that and still not implode. implored.

55:35Unidentified speaker (speaker_0): But it still rose quite I mean quite somewhat significantly. And I guess going into 2027, the question becomes I mean, when's the last time we saw an actual correction that was broad-based? I guess 2022. So we're going on four years of pretty good returns.

55:55Unidentified speaker (speaker_3): If you invest in an index fund or if you invest in a diversified portfolio.

55:59Unidentified speaker (speaker_0): Yes.

56:00Unidentified speaker (speaker_3): Eighty 80% percent of the world doesn't do that, right? So what might seem like a painless, a relatively painless year for you if you're diversified, has been an extremely painful year for many people who've invested. Because then that's the other thing about tech doing so well for so long is you have investors in the entire portfolio is composed of tech stocks or the Mag 7. And it's not been a good year for them. In fact, it's been a pretty damaging year. So you can have damage that is deep in big segments of the market. And I'm a feeling a lot of investors, if you look at, even though the market is up 7%, I'd wager 80% of investors this year

56:44Unidentified speaker (UNKNOWN): are

56:45Unidentified speaker (speaker_3): going to underperform, are going to have negative returns for the year. One reason you can tell is even though the market was up, so this was actually a month ago when the market was up 11, 12%, 52% of stocks this year were still down for the year. It's not a market where stocks are all climbing. It's a market where depending on the sector, depending on the stocks you pick, you can get a return very different from the return you're seeing for the overall market.

57:11Unidentified speaker (speaker_0): So what do you think will happen? I mean, we're a little bit a little bit early to be making predictions, but what would be your expectations going into 2027 for the broader stock market for the SP?

57:24Unidentified speaker (UNKNOWN): Um,

57:25Unidentified speaker (speaker_0): you know, some feel very worried. Paul Tudor Jones, for example, has, I forget exactly what his words were, but he said, you know, with all of the all of the capital raising coming down the pike, how hundreds of billions of dollars that are going to be raised in the capital markets by not only the the anthropics and the open AIs of the world, but also the the big tech companies who are going to go out and raise more money to build more data centers. So given all of that happening, um what are your expectations for this market going into 2027?

58:01Unidentified speaker (speaker_3): I think that not what's notwithstanding the AI investor collapse this year, AI continues to be the big story driving the economy and markets. If the AI story really slows down, and I'm talking about the building of data centers slows down, the supplying of the data center slows down, the products and services don't take off, then I think there's an economic hit we're going to take

58:25Unidentified speaker (UNKNOWN): to

58:25Unidentified speaker (speaker_3): the overall and that's going to cause a market correction. So really everything goes back to AI. This is much as we like to run away from AI and say it's not as big as it used to be. We've managed pretty well without AI. AI is the undercurrent that's driving the economy in the market. And I think 2027, the question is: can AI keep doing it? Or will a substantial slowdown in AI result in a recession? and a market correction broadly because people don't have the income?

58:55Unidentified speaker (UNKNOWN): And

58:55Unidentified speaker (speaker_3): that I think is the big question hanging over 2027. What happens? Because it is going to happen. AI is going to slow down. The earnings effect of AI is going to go away. And all of a sudden you're going to see a price earnings ratio rise, even though the market might not be, you know, might not be going up because earnings are giving up. some of those gains. Because a big chunk of the gay, the growth in earnings is this was the, I think, the highest growth rate in earnings for the aggregate earnings for the S P 500 that I've seen in the last 40 years. It's like a 56%. It's an astonishing

59:30Unidentified speaker (UNKNOWN): crazy.

59:31Unidentified speaker (speaker_3): But much of it came from AI, right? People spending the CapEx reporting reported as revenues and earnings.

59:38Unidentified speaker (UNKNOWN): And you

59:39Unidentified speaker (speaker_3): start to slow that process down. It's not that people stop building data centers, but they slow the process down. You're no longer, it starts to catch up. The depreciation on existing AI capex is going to go up. The revenue benefits from new AI capex is going to go down. The earnings effect is going to go away. The question is: what does the 56% fifty six percent growth look like next year? It's definitely not going to be 56%. fifty-six percent, but is it going to be four or five percent, or are we talking about fifteen or twenty percent? That will be the number that ends up driving what the market does, I think, next year.

1:00:13Unidentified speaker (speaker_0): Aswat de Moderan is the Kirshner family chair in finance education and professor of finance at NYU's Stern School of Business, where he teaches corporate finance and valuation. You can read his research on his blog, Musings on the Market. Markets. Aswath, thank you so much for joining us.

1:00:30Unidentified speaker (speaker_3): Thank

1:00:31Unidentified speaker (UNKNOWN): you. This

1:00:35Unidentified speaker (speaker_0): episode was produced by Claire Miller and Alison Weiss and engineered by Benjamin Spencer. Our research team is Dan Shallon, Kristen O'Donohue, and Mir Silverio. Drew Burrows is our technical director, and Catherine Dillon is our executive producer. Thank you for listening to Prophety Prof G Markets from Prophet Prof G Media. If you like liked what you heard, give us a follow and join us for a fresh take on markets on Monday.

1:01:10Unidentified speaker (UNKNOWN): In time we are

1:01:32Unidentified speaker (speaker_0): support for the show comes from zero. Whether you're an investor or you're running a small business, you never want to lose track of the financial. fundamentals. It doesn't matter how great your product or idea is if you don't have a clear picture of how cash is moving through your business. Zero can help. Zero, spelled with an X, is a simple but powerful platform built for small business owners that makes managing business finances easier from day one. It makes accounting, payments, payroll, and analytics available in one connected platform. Join the 5mm 5 million customers worldwide who love using Zero. Learn more at Zero.com. That's X-E-R-O.com.

This page was produced automatically: the audio was transcribed with nvidia/parakeet-tdt-0.6b-v3+sortformer-v2.1 and interpreted by Qwen3.6-35B-A3B-UD-IQ3_S.gguf. Transcription and interpretation can contain errors; always confirm against the original episode. Nothing here is investment advice.
All insights