ETF Spotlight

Speakers discuss Jensen Huang's departure from Nvidia and its impact on AI infrastructure investments and valuations

50 min Listen to the original Episode page
NvidiaAI InfrastructureSemiconductorsValuationLeadership TransitionOptical ChipsHedge FundsHyperscaler Capex

Summary

The speakers analyze the market reaction to Jensen Huang's resignation as CEO of Nvidia, noting initial volatility in competitor stocks like AMD and Broadcom. They argue that Nvidia remains undervalued at 10x forward earnings for fiscal 2028, presenting a buying opportunity near $150. The discussion highlights the risk to neo-cloud providers like Nebius due to reduced aggressive financing from Nvidia. Speakers also examine the broader AI infrastructure build-out, framing hyperscaler capital expenditure as a survival strategy against obsolescence. Finally, they review the recent crash in the Philadelphia Semiconductor Index and the investment thesis for optical chip companies like Coherent.

Episode sections

  1. 0:07 – 9:25

    Jensen Huang's Departure and Nvidia's Internal Impact

    The speakers discuss Jensen Huang's resignation as CEO of Nvidia, describing him as a visionary founder whose departure creates uncertainty regarding succession, potential talent loss, and a scaling back of aggressive financing deals for neo-cloud companies.

  2. 9:25 – 18:00

    Nvidia Stock Valuation and Buying Opportunity

    The guest argues Nvidia is undervalued at 10x forward earnings for fiscal 2028, suggesting it is a buy near $150, despite analyst downgrades and potential investigations into espionage or supply chain issues.

  3. 18:00 – 21:45

    Impact on Neo-Cloud Companies and Financing

    The departure is expected to reduce Nvidia's aggressive financing of neo-cloud providers like Nebius, leading to stock declines for these companies as the market prices in reduced support and a shift toward shareholder returns.

  4. 21:45 – 28:00

    Competition and Alternative AI Chips

    The speakers analyze competitors like AMD, Google, Amazon, and Meta, noting that hyperscalers are accelerating their own chip development (e.g., OpenAI's Jalapeno, Cerebras) to avoid vendor lock-in, creating a diversified but competitive ecosystem.

  5. 28:00 – 32:18

    Post-Election Data Center Expansion and Jobs

    Despite the Democratic 'blue wave' in midterm elections potentially signaling stricter regulations, the guest argues data center expansion will continue due to the massive job creation in construction and the economic necessity of the 'fifth industrial revolution'.

  6. 32:18 – 35:14

    Valuation Reset and Bull Case for AI Stocks

    Speaker 2 argues that hyperscalers are safe havens due to massive existing compute capacity and cash flow, citing Andy Jassy's aggressive build-out strategy. The speaker defines the investment thesis as avoiding obsolescence (RIO) rather than just ROI, driven by fear of disruption from companies like Anthropic.

  7. 35:14 – 36:36

    Semiconductor Index Crash and Hedge Fund Woes

    Speaker 2 highlights a 12% single-day drop in the Philadelphia Semiconductor Index (SOX) on November 5th, closing at 9,000. He notes that many hedge funds lost 20-30% in October and were down 30-35% for the year, leading to forced selling and a valuation reset.

  8. 36:36 – 40:19

    Coherent (COHR) Investment Thesis and Optical Tech

    Speaker 2 owns Coherent (COHR), citing Jensen Huang's $2 billion investment in optical technologies like silicon photonics. He explains the technical shift from copper to fiber/lasers for connecting GPU pods and notes Coherent's valuation has dropped to 3x sales with strong growth estimates.

  9. 40:19 – 42:37

    Lumentum Acquisition and Supply Chain Dynamics

    Speaker 2 mentions Marvell's all-stock acquisition of Lumentum, which caused Lumentum's stock to decline. He contrasts this with the broader trend of hyperscaler capex exceeding $1 trillion, noting that memory and optical supply chains are sold out through 2028.

  10. 42:37 – 45:15

    NVIDIA Leadership Transition and Organizational Risk

    Speaker 1 discusses Jensen Huang's management style of instilling fear, referencing a book by Kim. Speaker 2 argues that without Huang, NVIDIA risks becoming bureaucratic, as he maintained a granular grasp of the business through 60 direct reports and one-on-one meetings.

  11. 45:15 – 46:55

    Succession Planning and GTC Catalyst

    Speaker 2 predicts a formal succession plan will emerge, potentially drafted by Colette Kruse, as she may not become CEO due to her non-engineering background. He identifies the upcoming GTC conference on November 30th as a key event for observing leadership dynamics.

  12. 46:55 – 50:05

    ETF Exposure to NVIDIA

    Speaker 1 recaps discussed stocks (NVDA, AMD, AVGO, TSM, COHR) and details NVIDIA's weight in major ETFs: ~6.1% in SPY, ~6.5% in QQQ, ~17% in SMH, ~6.6% in SOXX, and ~13% in XLK.

Assets discussed

AssetKindStanceContext
NVIDIANVDAequitymixedSo NVIDIA is super cheap. And
MicrosoftMSFTequityneutraluh you know that's another way to compete. And and they they just reported last night, and they've got new business with Microsoft
AmazonAMZNequityneutralBut Amazon and Google have done quite well
AlphabetGOOGLequityneutralBut Amazon and Google have done quite well
Invesco QQQ TrustQQQetfneutralAnd the NASDAQ 100 EDFs like Qs, QQQ, they have about 6.5% weight in NVIDIA.

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

Claims worth verifying

Jensen Huang resigned as CEO of Nvidia, a company he co-founded 33 years ago.

Unidentified speaker (speaker_1) ·1:03

How to verify: Check Nvidia's official press release or SEC Form 8-K for the resignation announcement and the date of the filing. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Primary source

Jensen Huang managed more than sixty people directly at Nvidia and personally dealt with heads of governments and key clients.

Unidentified speaker (speaker_1) ·1:43

How to verify: Review Nvidia's latest annual proxy statement (DEF 14A) for executive management details and public records of government meetings. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Primary source

An internal poll of 1,600 Nvidia respondents showed 90% were in favor of keeping Jensen Huang as their leader.

Unidentified speaker (speaker_2) ·5:13

How to verify: Request the specific internal poll data from Nvidia's investor relations or look for the source of the leak cited by the guest. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Primary source

Nvidia stock is down about 35% from its 52-week high at the time of recording.

Unidentified speaker (speaker_1) ·9:29

How to verify: Calculate the percentage decline from the 52-week high to the closing price on the recording date (Friday prior to Nov 16, 2026) using Yahoo Finance or Bloomberg. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Primary source

Nvidia is trading at a forward P/E multiple of 17.9% (likely meaning 17.9x) based on current guidance.

Unidentified speaker (speaker_1) ·9:48

How to verify: Calculate the current stock price divided by the consensus forward EPS estimate from financial data providers like Zacks or Bloomberg. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Primary source

Nvidia's EPS consensus for fiscal 2027 is $9.28 and for fiscal 2028 is $15.60.

Unidentified speaker (speaker_2) ·11:15

How to verify: Check Zacks Earnings Estimates or Bloomberg Consensus for Nvidia (NVDA) for the specified fiscal years. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Primary source

Nvidia is trading at 10 times forward earnings for fiscal 2028.

Unidentified speaker (speaker_2) ·11:23

How to verify: Divide the stock price near $150 by the fiscal 2026 EPS consensus of $15.60 to verify the multiple. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Primary source

Half a dozen investment banks lowered their price targets to around $200.

Unidentified speaker (speaker_2) ·11:54

How to verify: Search financial news archives for analyst reports from major banks (e.g., Morgan Stanley, Goldman Sachs) issued after the resignation news. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Nebius (MBIS) stock gapped up 15% on Wednesday, November 11th, and closed higher, trading up to $160.

Unidentified speaker (speaker_2) ·19:47

How to verify: Check historical stock data for Nebius (MBIS) for November 11, 2026, to verify the gap up and closing price. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Nebius stock fell 7.17% after Jensen's resignation and traded as low as $109.

Unidentified speaker (UNKNOWN) ·20:01

How to verify: Check historical stock data for Nebius (MBIS) for the days following the resignation announcement to verify the intraday low and percentage drop. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Hyperscaler capital expenditure is projected to rise from $500 billion in 2025 to $750 billion in 2026 and $1.25 trillion in 2027.

Unidentified speaker (UNKNOWN) ·33:53

How to verify: Check capital expenditure guidance and investor presentations from major hyperscalers (Microsoft, Amazon, Google, Meta) for fiscal years 2026-2026. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

The Philadelphia Semiconductor Index (SOX) dropped 12% in a single day on November 5th, closing at 9,000.

Unidentified speaker (UNKNOWN) ·35:32

How to verify: Review historical price data for the SOX index (PHLX Semiconductor Sector Index) for November 5, 2026. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Jensen Huang invested $2 billion in Coherent and Lumentum.

Unidentified speaker (speaker_2) ·36:57

How to verify: Search SEC EDGAR for NVIDIA (NVDA) Form 13F filings or press releases regarding investments in Coherent (COHR) and Lumentum (LITE) around the time of the announcement. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Primary source

Coherent (COHR) trades at 3x sales with a $46 billion market cap.

Unidentified speaker (speaker_2) ·38:51

How to verify: Check current market data for Coherent Inc. (COHR) for market capitalization and trailing twelve-month (TTM) sales figures. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Primary source

Marvell proposed an all-stock acquisition of Lumentum at a ratio of 4.5 Marvell shares for 1 Lumentum share.

Unidentified speaker (speaker_2) ·40:27

How to verify: Review the definitive proxy statement or merger agreement filed by Marvell Technology (MRVL) and Lumentum (LITE) with the SEC. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Primary source

NVIDIA accounts for 17% of the portfolio in the VanEck Semiconductor ETF (SMH).

Unidentified speaker (speaker_3) ·48:05

How to verify: Check the current holdings and weightings on the VanEck ETF website or financial data providers for SMH. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Primary source

NVIDIA accounts for 6.6% of the portfolio in the iShares Semiconductor ETF (SOXX).

Unidentified speaker (speaker_3) ·48:45

How to verify: Check the current holdings and weightings on the iShares ETF website or financial data providers for SOXX. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Primary source

NVIDIA accounts for 13% of the portfolio in the Technology Select Sector SPDR Fund (XLK).

Unidentified speaker (speaker_3) ·49:00

How to verify: Check the current holdings and weightings on the State Street Global Advisors website or financial data providers for XLK. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Primary source

Coherent's consensus revenue estimate for next year is $15 billion.

Unidentified speaker (speaker_2) ·39:40

How to verify: Check consensus analyst estimates for Coherent Inc. (COHR) revenue on financial data platforms like Bloomberg or FactSet. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Coherent is on a June fiscal year.

Unidentified speaker (speaker_2) ·39:47

How to verify: Review Coherent Inc.'s (COHR) most recent 10-K or 10-Q filing with the SEC to confirm fiscal year end. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Primary source

Do your own research

  • Verify the accuracy of Jensen Huang's resignation and the official succession plan via Nvidia's SEC filings (Form 8-K) and press releases.
  • Check Nvidia's current stock price and forward P/E ratio against the fiscal 2028 EPS consensus of $15.60 cited in the episode.
  • Review Nebius (MBIS) stock performance and recent earnings reports to assess the impact of reduced Nvidia financing.
  • Analyze the Philadelphia Semiconductor Index (SOX) historical data for the November 5th drop and subsequent trading volume.
  • Examine Coherent's (COHR) financial statements and consensus analyst estimates for revenue growth and EPS projections.
  • Confirm the terms of the Marvell-Lumentum acquisition proposal and Lumentum's subsequent stock performance.
  • Monitor the upcoming GTC conference on November 30th for official announcements regarding Nvidia's leadership structure.
  • Assess hyperscaler capital expenditure guidance from major cloud providers (Amazon, Microsoft, Google, Meta) to validate the $1 trillion annual capex projection.
Full transcriptexpand
0:07
Unidentified speaker (speaker_0): Here to dive into industry trends with leading ETF experts, this is ETF Spotlight with Nina Mishra.
0:15
Unidentified speaker (speaker_1): Hello and welcome to ETF Spotlight.
0:18
Unidentified speaker (UNKNOWN): I'm
0:18
Unidentified speaker (speaker_1): your host, Nina Mishra. My guest today is Zach's Zax, senior stock strategist, Kevin Cook, and we are talking about the future of the AI trade with Nvidia without Jensen Huang at the hem.
0:35
Unidentified speaker (speaker_1): And uh especially against the backdrop of the blue wave. So really excited about this timely discussion. Kevin, welcome back. Uh great to have you with us today.
0:48
Unidentified speaker (speaker_2): Hey, thanks, Nina. Yeah, we haven't done this in a while. So uh I think you wanted to talk about this last week before uh Jensen had to leave. And now it's just become so much more complicated.
1:03
Unidentified speaker (speaker_1): Right, right. Uh so this is the biggest story of the week, of the month, maybe the year two. Yeah. Jensen resigning as CEO of N. NVIDIA, a company he co-founded 33 years back and built into an AI powerhouse. Now Nvidia
1:23
Unidentified speaker (speaker_1): is no ordinary company. Jensen was no ordinary CEO. So Nvidia's cutting-edge GPUs have powered the AI revolution, and Jensen was called the godfather of AI. And uh he was not. only brilliant and charismatic, he was also a great
1:43
Unidentified speaker (speaker_1): visionary and strategist. And uh we know that he managed more than sixty people directly at Nvidia and he personally dealt with heads of governments and key clients. So I'm curious to hear your thoughts on what Jensen's departure means for Nvidia.
2:02
Unidentified speaker (speaker_2): It's uh it's definitely the biggest question uh of this month and uh and this quarter.
2:10
Unidentified speaker (UNKNOWN): Uh
2:10
Unidentified speaker (speaker_2): I did a a live stream earlier with with clients um where I framed the discussion that we're losing the VIP. And VIP IP stands for visionary, influential, passionate.
2:27
Unidentified speaker (speaker_2): aggressive founder. Um
2:30
Unidentified speaker (speaker_2): yeah, you you you
2:32
Unidentified speaker (UNKNOWN): you
2:32
Unidentified speaker (speaker_2): called Jensen brilliant and charismatic. He certainly was. And and he was the founder. And you and there's an interesting thing about founder-led companies. Um
2:42
Unidentified speaker (UNKNOWN): uh we
2:42
Unidentified speaker (speaker_2): all know Mark Andreessen as uh Um, the founder of Netscape, and um you know a a brilliant. brilliant Silicon Valley venture capitalist in his own right. He did a I saw him in a video last couple weeks ago where he talked about how
2:59
Unidentified speaker (UNKNOWN): the
3:00
Unidentified speaker (speaker_2): how corporate America runs it's built on this model of the managerial class that that Harvard Business School and Stanford Business School that they were sort of built around, that a
3:13
Unidentified speaker (UNKNOWN): uh a
3:14
Unidentified speaker (speaker_2): manager. manager of a business, you know, the MBA could run any company because they knew how to run operations and they knew how to, you know, take care of the numbers.
3:26
Unidentified speaker (UNKNOWN): Um,
3:26
Unidentified speaker (speaker_2): and he said, you know, that
3:29
Unidentified speaker (speaker_2): Andreessen looked at
3:30
Unidentified speaker (UNKNOWN): it
3:30
Unidentified speaker (speaker_2): like that. That's actually uh been a detriment for America to think that any manager can run any business versus the concept of the founder-led business.
3:39
Unidentified speaker (speaker_2): Um,
3:40
Unidentified speaker (UNKNOWN): you
3:41
Unidentified speaker (speaker_2): know, you go back to Ford, Rockefeller, Carnegie, but you know, we have uh we have Elon now as a prime example. And Jensen is in that class. And so visionary, influential, passionate. And the the influential part comes after the vision and the passion that Jensen And had for the business. And and once he saw, you know, he made
4:01
Unidentified speaker (speaker_2): critical,
4:02
Unidentified speaker (UNKNOWN): um,
4:03
Unidentified speaker (speaker_2): innovations like CUDA, the software platform, compute unified device architecture, that became the foundation. And this is before he was thinking of anything to do with AI. He was still just building how do we make our GPUs, our graphics processing units, more powerful Um, and and better at what they do
4:23
Unidentified speaker (speaker_2): and more available to developers
4:26
Unidentified speaker (UNKNOWN): to
4:26
Unidentified speaker (speaker_2): innovate. off of that. So
4:29
Unidentified speaker (speaker_2): um,
4:30
Unidentified speaker (UNKNOWN): so
4:31
Unidentified speaker (speaker_2): what I talked about this morning with my people was we've got two realms of a concern here. What happens inside NVIDIA,
4:39
Unidentified speaker (speaker_2): and what happens outside NVIDIA that could impact the entire AI trade. Now, what happens inside NVIDIA? We're looking at you know succession plans, and um, they made a smart move by making Collect Press, the CFO.
4:51
Unidentified speaker (UNKNOWN): um,
4:52
Unidentified speaker (speaker_2): an interim CEO, because she knows the business inside and out. She's been there since 2013. If you listen to her on conference calls, uh, got a lot of confidence in her. But um, you
5:03
Unidentified speaker (UNKNOWN): know,
5:03
Unidentified speaker (speaker_2): the company just lost a leader that they would follow anywhere.
5:07
Unidentified speaker (UNKNOWN): And
5:07
Unidentified speaker (speaker_2): there was an internal poll that was just released, like 1,600 respondents, um
5:13
Unidentified speaker (speaker_2): 90% of them. were in favor of keeping Jensen um as their leader. And so other things that can happen inside Nvidia is there could be a scale back. You've lost the visionary, passionate, aggressive founder. So there's going to be a scale back in all the stuff that he was aggressive on. He was aggressive
5:33
Unidentified speaker (speaker_2): with outside deals, financing,
5:36
Unidentified speaker (UNKNOWN): uh,
5:37
Unidentified speaker (speaker_2): taking stakes and momentum and coherent and uh corning and um you know you know the list goes on, and then financing the neo clouds and expansion into multiple tech industries. That's gonna get scaled back probably pretty quickly. We're losing a key influencer in
5:56
Unidentified speaker (speaker_2): Taiwan, right? He is Taiwanese, so his ability to connect and communicate. communicate with
6:05
Unidentified speaker (UNKNOWN): uh
6:05
Unidentified speaker (speaker_2): Morris Chang of of Taiwan Semi and have that relationship and the and the vision that they shared to build. Uh, and then Jensen's uh influence in Washington, D.C. That's gone.
6:17
Unidentified speaker (UNKNOWN): Um,
6:17
Unidentified speaker (speaker_2): there could be talent loss. Think of all the brilliant engineers who may have wanted to start their own company for a while,
6:26
Unidentified speaker (UNKNOWN): and
6:26
Unidentified speaker (speaker_2): now
6:28
Unidentified speaker (speaker_2): they're thinking, well, maybe now's the time. Maybe this is my sign to go start my own company. Um,
6:33
Unidentified speaker (UNKNOWN): you
6:34
Unidentified speaker (speaker_2): know, and many of them are rich already, so
6:37
Unidentified speaker (UNKNOWN): they
6:37
Unidentified speaker (speaker_2): could do it. Um,
6:39
Unidentified speaker (UNKNOWN): and
6:40
Unidentified speaker (speaker_2): then the competition.
6:41
Unidentified speaker (UNKNOWN): Uh,
6:42
Unidentified speaker (speaker_2): so we're we're talking here on Friday, on Thursday, after Jensen resigned, we saw. stocks like Broadcom and AMD pop 5%. Now they're giving that back today, but the the message from the market was hey, Broadcom and
7:02
Unidentified speaker (speaker_2): AMD can um now have more leverage in the the compute market because AMD is already Lisa Sue is a is a master with her own GPU architectures, but now the AMD ASICs and the TPUs and other platforms from Google, Amazon's training, there can be
7:23
Unidentified speaker (speaker_2): if if there's any loss of faith or
7:26
Unidentified speaker (UNKNOWN): uh
7:26
Unidentified speaker (speaker_2): questions about NVIDIA's business and their ability to deliver, then those companies will do well.
7:33
Unidentified speaker (UNKNOWN): And
7:33
Unidentified speaker (speaker_2): then you so you then you look at what outside NVIDIA,
7:36
Unidentified speaker (UNKNOWN): which
7:37
Unidentified speaker (speaker_2): I talked about TSMC. you know, they've lost their key co-architect. There's nobody, I mean, there there are talented people within NVIDIA. I don't want to say that they're not, like Ian Buck. Well, we'll oh, that's a big question we have to answer. I I did a a blog on it this week is we've got GTC coming up November 30th, the GPU Tech
7:56
Unidentified speaker (speaker_2): Conference in Washington, D.C., November 30th that whole week. Who's gonna going do to the keynote? Well, it's not going to be Jensen's the master of that keynote. My guess is it'll be Ian Buck. And then we'll start to get signals of like who who could replace Elon. I mean, I'm sorry, Jensen, as that, as that
8:16
Unidentified speaker (speaker_2): visionary, passionate, aggressive founder leader. Um, well, Ian Buck's not a founder, but he's been there a while. He's been there his whole career. Like when he when he got his PhD in the early 2000s, he's been there since then. So, you know, uh NVIDIA DNA is in his blood. Um, he could definitely take over.
8:36
Unidentified speaker (speaker_2): I think he's like 49 years old. What would be cool, I think, is to make him and Colette co CEOs. You know, she's got the numbers, and he and he knows the technology, and let them sort of balance it out. Like
8:51
Unidentified speaker (UNKNOWN): it
8:52
Unidentified speaker (speaker_2): would sort of say, yeah, it takes two people to fill Jensen's shoes. Um, so we can talk more about um some of the out some of the impacts, you know, outside of NVIDIA in the larger AI industry. But I'll I'll uh I'll let you take back over here.
9:09
Unidentified speaker (speaker_1): Right. There's there's so much uncertainty as of now, and then there's the geopolitical. angle too, right? There would be congressional hearings, investigations, and we'll find out more in the coming months. Who knew what, when, so and how that impacts Nvidia. Now, uh,
9:29
Unidentified speaker (speaker_1): you mentioned that we are recording it on Friday, and the company reports next week. Uh and uh the stock is down about 35%. from its 52-week high. And it is trading at a forward P multiple of 17.9% now. And according
9:48
Unidentified speaker (speaker_1): to the company's guidance, revenue for the fiscal year ending in early 2028 is still expected to grow at around 70%.
9:59
Unidentified speaker (speaker_1): And the company expects supply to remain a bottleneck. for at least through the end of fiscal 2028. So looking at this stock, do you think this is a buying opportunity, or is it just a situation where investors should wait for the CEO announcement? Uh, you mentioned you suggested
10:18
Unidentified speaker (speaker_1): that probably having two co-CEOs, one with the technical background and the other uh who is interim CEO, Colette Kraus. Uh and we also know that board is uh currently looking at they are approaching many external candidates too. Uh Financial Times just reported that Nvidia's
10:38
Unidentified speaker (speaker_1): board has approached uh ARM CEO Rene has uh because they are searching for a potential successor successes. so we'll find out more in the coming months what But but what is your um uh view on the stock as of now? Would you recommend buying it here?
10:55
Unidentified speaker (speaker_2): Yeah, I mean, when I look at the forward estimates, you know, I'm uh we're in fiscal 27, which ends in January. And um, and so that that EPS consensus, according to our data, is $9.28. But if you look at what I call next year, which is
11:15
Unidentified speaker (speaker_2): fiscal 28, which begins in February,
11:18
Unidentified speaker (UNKNOWN): um,
11:18
Unidentified speaker (speaker_2): the consensus is $15.60.
11:23
Unidentified speaker (speaker_2): So that means NVIDIA is actually trading at 10 times forward earnings.
11:29
Unidentified speaker (UNKNOWN): You know,
11:29
Unidentified speaker (speaker_2): the way I look at it, because fiscal twenty fiscal 28 starts in February. So NVIDIA is super cheap. And
11:37
Unidentified speaker (speaker_2): um
11:39
Unidentified speaker (speaker_2): you know, a lot of those orders are pretty locked in for next year. So despite the disruption to the company and the reverberations throughout the supply chain, um, I think those estimates are pretty safe.
11:54
Unidentified speaker (speaker_2): There were a lot of investment banks that lowered their price targets. Let's let's say half a dozen banks came out and lowered their price. targets to around 200.
12:06
Unidentified speaker (UNKNOWN): I
12:07
Unidentified speaker (speaker_2): don't care about that. You know, that they're just, you know, they're uh those analysts are fighting to save their own career.
12:15
Unidentified speaker (UNKNOWN): Um,
12:15
Unidentified speaker (speaker_2): but I I think the stock is very attractive down here near 150. And
12:21
Unidentified speaker (speaker_2): and that it is a juggernaut that will not crumble
12:26
Unidentified speaker (UNKNOWN): just
12:26
Unidentified speaker (speaker_2): because. we lost the visionary, influential, and passionate uh aggressive founder. So
12:35
Unidentified speaker (speaker_2): I do think the stock has incredible value here, and I would be a buyer
12:41
Unidentified speaker (UNKNOWN): uh
12:41
Unidentified speaker (speaker_2): in the 150s. Now,
12:43
Unidentified speaker (UNKNOWN): uh,
12:44
Unidentified speaker (speaker_2): you brought up uh espionage and investigations. That's another thing like
12:49
Unidentified speaker (UNKNOWN): that
12:50
Unidentified speaker (speaker_2): any any. leader would have to deal with. So for instance, I uh you know how Tim Cook, Tim Cook was sort of floated, right? Like would Tim Cook do it? I mean, I think he could. Um
13:02
Unidentified speaker (UNKNOWN): he
13:02
Unidentified speaker (speaker_2): because he could definitely get up to speed uh on the GPU stuff. And he he's he's been to the mountaintop, so to speak, right? He he's led the other biggest tech company. He's seen it all
13:13
Unidentified speaker (UNKNOWN): and
13:13
Unidentified speaker (speaker_2): he knows how to lead. And he's he's very measured and deliberate and good communicator. But at 66, when he was just he's ready to ride on his Apple yacht into the sunset, I don't think he wants that because look what he would face. Not only the daily violence of managing NVIDIA, but then the ongoing investigations about was
13:33
Unidentified speaker (speaker_2): there espionage, whatever, and how many different uh government agencies or are going to be investigating. Like, even if they start an investigation of what did the White House do, they'll probably want
13:47
Unidentified speaker (UNKNOWN): uh
13:47
Unidentified speaker (speaker_2): access to NVIDIA internal documents, you know, two years worth of uh emails and internal docs to give them,
13:55
Unidentified speaker (UNKNOWN): you
13:55
Unidentified speaker (speaker_2): know, like who knew what when. And so
13:59
Unidentified speaker (UNKNOWN): I
13:59
Unidentified speaker (speaker_2): was even concerned that did Colette Cress know anything. about
14:04
Unidentified speaker (UNKNOWN): what
14:04
Unidentified speaker (speaker_2): was going on with GPUs going through Singapore to China? It looks like maybe she didn't, because we saw that Wall Street Journal article where,
14:14
Unidentified speaker (UNKNOWN): you
14:14
Unidentified speaker (speaker_2): know, somebody leaked something from a meeting recently where Jensen was supposedly berating Colette for being a deer in the headlights because she was pushing back on
14:26
Unidentified speaker (UNKNOWN): all
14:26
Unidentified speaker (speaker_2): these aggressive. deals. And um,
14:29
Unidentified speaker (UNKNOWN): so
14:29
Unidentified speaker (speaker_2): it's possible there was already a rift there between them and she didn't know, but we're gonna find stuff out. So um I did a
14:38
Unidentified speaker (UNKNOWN): uh
14:38
Unidentified speaker (speaker_2): I put together an article of like who are the top dozen candidates
14:44
Unidentified speaker (UNKNOWN): uh
14:44
Unidentified speaker (speaker_2): besides Colette Sides Press to run the company? You know, Ian Buck is obviously number one. You have Gilead Shaner, who is uh a senior guy there in uh networking. Uh Brian Cantanzaro is a VP in applied deep learning research.
15:01
Unidentified speaker (UNKNOWN): Uh
15:02
Unidentified speaker (speaker_2): Bill Daly is our chief scientist. He's brilliant and and uh been there a long time. I think he's been there since 2009. Uh Michael Kagan, uh TTO, T T O, he's in Israel, so uh he's he he got that job because of the Mellanox acquisition in 2020. Um
15:21
Unidentified speaker (speaker_2): you've got uh Jan Kautz is the VP of learning and perception research. Um
15:28
Unidentified speaker (UNKNOWN): uh uh
15:30
Unidentified speaker (speaker_2): Rev, I'm not sure how to pronounce his name. Uh
15:33
Unidentified speaker (UNKNOWN): um
15:34
Unidentified speaker (speaker_2): L E B A R E D I A N. He's focused on omniverse and simulation technologies. And he's spoken to G GTC before, and he's a good presenter. We'll probably see him on stage at the end of the month here.
15:48
Unidentified speaker (UNKNOWN): Uh
15:48
Unidentified speaker (speaker_2): Deeputala is in robotics and edge AI. Kimberly Powell, her name came to light because supposedly
15:57
Unidentified speaker (speaker_1): Sure,
15:57
Unidentified speaker (UNKNOWN): she was uh
15:58
Unidentified speaker (speaker_2): yeah, romantically involved with uh Jensen. She's she's on she's in healthcare, but uh she's a bright leader. Mike Houston, and then one other one that uh was on my list. Um, oh Jonah Albin is is uh
16:13
Unidentified speaker (UNKNOWN): is
16:14
Unidentified speaker (speaker_2): in GPU engineering. And so he's been deeply engaged. He was with been with the company since 1997. Um, and he's got thirty-two 32 patents.
16:25
Unidentified speaker (UNKNOWN): Uh
16:25
Unidentified speaker (speaker_2): so that there's a a deep bench here of top leadership that could definitely carry this company forward, and that I'm confident in when I say I'll buy NVIDIA at 150 bucks a share. Um, we just have to see the the the so my internal doubts
16:45
Unidentified speaker (speaker_2): are this company can recover and do well. Um given that talent pool, we just have to see what's going to go on the outside, what happens in Taiwan, um, what happens with uh the the supply chain. And because there were all this is what we don't know. There was
17:05
Unidentified speaker (speaker_2): some talk that Rubin, Vera Rubin, the next generation, here after the uh Grace Blackwell
17:12
Unidentified speaker (UNKNOWN): uh
17:13
Unidentified speaker (speaker_2): GPU rack scale system.
17:15
Unidentified speaker (speaker_2): that it is not that it may be behind, that there may be some delays with with Vera Rubin next year. Um and so if that was confirmed, then the stock would have trouble recovering right here. But it's still an incredible value. I mean, if if there were if there was some news uh spike and
17:35
Unidentified speaker (speaker_2): it sent NVIDIA down to 140, I would buy it at both hands.
17:39
Unidentified speaker (speaker_1): Interesting. So you mentioned that meeting where uh Colette Kress had questioned some of these risky deals. Um, that is according to a Wall Street uh journal report. Uh and Jensen reportedly called her a deer in the headlights. So he was pursuing these very risky deals, uh
17:59
Unidentified speaker (speaker_1): very creative financing arrangements, also. So that is why. Nvidia was often called the central bank of AI, because they were not only selling GPUs, they were also using their massive cash flow to support the demand for their own chips. And they were financing these new cloud
18:19
Unidentified speaker (speaker_1): providers and data center developers. And they were hoping that these companies will turn into long-term long term customers.
18:27
Unidentified speaker (speaker_1): So now it is quite possible that under the new leadership, at least Tel Collette, Chris remains the CEO, acting CEO, that Nvidia may scale back some of those financing arrangements. And maybe they will focus more on returning cash to shareholders. So
18:47
Unidentified speaker (speaker_1): what does Jensen's departure mean for the these neo cloud companies and uh financing deals in general?
18:56
Unidentified speaker (speaker_2): Well, you're definitely you're definitely seeing it already show up in their stock prices. And that's that's you know, that's why I said that that collectress is not going to be as aggressive with deals or financing. And um, for instance, uh Nebius, uh the stock symbol is MBIS, is the
19:14
Unidentified speaker (UNKNOWN): is
19:14
Unidentified speaker (speaker_2): considered. the premier neo cloud. You know, has the best balance sheet, um, the best growth, et cetera, et cetera. And they just delivered a a really great report. But
19:27
Unidentified speaker (speaker_2): uh the stock is really struggling. So I think this is getting priced in that that the environment for the neo clouds Nebius, Nebbius, Core Weave, Iron is not as optimistic anymore. For instance, on Wednesday, November 11th, after Nebbius's
19:47
Unidentified speaker (speaker_2): strong report, the stock gapped
19:49
Unidentified speaker (UNKNOWN): up
19:50
Unidentified speaker (speaker_2): and closed 15% higher that day.
19:53
Unidentified speaker (UNKNOWN): Uh,
19:54
Unidentified speaker (speaker_2): traded up to 160.
19:56
Unidentified speaker (speaker_2): Today,
19:59
Unidentified speaker (speaker_2): and you know, and then yesterday,
20:01
Unidentified speaker (UNKNOWN): it
20:01
Unidentified speaker (speaker_2): gave back seven. 17% after Jensen resigns. And today it's it's playing with new, I think it's playing with new 52-week lows, if I'm not mistaken. It's trading um it's traded as low as 109, trading 118 while I look at
20:17
Unidentified speaker (UNKNOWN): it.
20:18
Unidentified speaker (speaker_2): And um not quite a 52-week low, but but 109,
20:23
Unidentified speaker (UNKNOWN): um
20:23
Unidentified speaker (speaker_2): that takes you back to
20:27
Unidentified speaker (UNKNOWN): um,
20:27
Unidentified speaker (speaker_2): you know, March levels when we were in when we were in sort of a soft correction back in March. So yeah, it's already showing up for them. Um,
20:36
Unidentified speaker (UNKNOWN): you know,
20:36
Unidentified speaker (speaker_2): Jensen looked at it like not only is he gonna sell to the hyperscalers, but he wants to he wants the ecosystem to be as broad and diverse as possible. That's why he welcomes open source LLM models. And then the NeoClouds can provide compute to whoever wants it. Obviously,
20:56
Unidentified speaker (speaker_2): he's he was always talking his book because it allows him to sell more GPUs, but it there's also an argument to be made of
21:05
Unidentified speaker (UNKNOWN): uh
21:06
Unidentified speaker (speaker_2): avoiding concentration, right? What why let the hyperscalers concentrate the wealth and the models and the compute power?
21:15
Unidentified speaker (UNKNOWN): and
21:15
Unidentified speaker (speaker_2): the energy just because they have war chests and cash flow? Why, you know, we can't let them dominate the industry. So I think I think it was uh his his democratic approach to it, if that makes sense.
21:29
Unidentified speaker (speaker_1): Right. So you also mentioned uh Broadcom and AMD rising after the scandal broke out. Um And investors probably believe that. they could gain market shares. And then there are tech giants also like Google and Amazon, Microsoft also, to a certain extent, they are developing
21:49
Unidentified speaker (speaker_1): their own custom AI chips. So do you think Nvidia's big customers will finally switch to these alternatives? Who, in your view, stands to benefit the most from? Nvidia's troubles?
22:05
Unidentified speaker (speaker_2): Well, that second question is a tough one. Um, ironically, I wrote a uh an article last week before we lost Jensen titled Can NVIDIA's dynasty withstand the competition? You know, and and I was just speculating, like I wanted to survey what's out there.
22:24
Unidentified speaker (speaker_2): What's out there from AMD, which is you know the number two in the uh GPU duopoly. And then
22:31
Unidentified speaker (UNKNOWN): you
22:31
Unidentified speaker (speaker_2): know, how is Google innovating with TPUs? Amazon Training, uh OpenAI comes up with their jalapeno chip. And then uh Meta uh has a platform they're they're calling Meta Training and An Inference Accelerator. Uh kind of wordy, and the the acronym isn't hard is har hard. to remember
22:51
Unidentified speaker (speaker_2): too. It's MTIA. But if you ever see MTIA, that's that's what Meta is up to. Um so it's wild that I wrote this uh last last week, and and I was just speculating. And my conclusion was NVIDIA's not in trouble. I'm not selling NVIDIA. This is just
23:11
Unidentified speaker (speaker_2): this is just natural
23:13
Unidentified speaker (UNKNOWN): uh
23:14
Unidentified speaker (speaker_2): diversity. in the ecosystem. Because if you think about it, if you're Andy Jassy at Amazon
23:18
Unidentified speaker (UNKNOWN): or
23:19
Unidentified speaker (speaker_2): you're Mark Zuckerberg at Meta, um um Sundar Pakai at Google, of course you want to have your own chips, right? Because you don't want to be told nobody wants to be totally dependent on one vendor GPU GP uh for GPU for for compute. So of course they're innovating and they're doing
23:39
Unidentified speaker (speaker_2): it internally for their own uses,
23:43
Unidentified speaker (UNKNOWN): you know,
23:43
Unidentified speaker (speaker_2): their own compute demand as a business internally. And then also if if they're
23:49
Unidentified speaker (UNKNOWN): if
23:50
Unidentified speaker (speaker_2): they can build it to provide cloud services for others that they can that they can lease the compute and and make a profit on it. Even Mark Zuckerberg realized, oh my gosh, we've built so much compute here. Um
24:02
Unidentified speaker (UNKNOWN): when we
24:02
Unidentified speaker (speaker_2): have spare compute, we'll lease it out to somebody else. So
24:07
Unidentified speaker (speaker_2): um there's uh a VC named Vino Kosla, who was uh he was a co-founder of Sun Microsystems back in the early 80s. And um he was an early investor in OpenAI. He was a um uh investor in DoorDash, really smart guy.
24:23
Unidentified speaker (UNKNOWN): Um,
24:23
Unidentified speaker (speaker_2): I don't know if he's an anthropic or not, but he said a couple of weeks ago, he's like,
24:28
Unidentified speaker (UNKNOWN): you know,
24:29
Unidentified speaker (speaker_2): there will be competition, um, but it'll take five years.
24:33
Unidentified speaker (UNKNOWN): And
24:33
Unidentified speaker (speaker_2): I actually think he's wrong there that it's gonna take five years to compete with NVIDIA because look how quickly OpenAI came up with the jalapeno chip and Amazon's training uh platform. They're on already on like generation three.
24:50
Unidentified speaker (UNKNOWN): Uh
24:50
Unidentified speaker (speaker_2): Um,
24:52
Unidentified speaker (speaker_2): and
24:53
Unidentified speaker (speaker_2): just making a ton of progress, and then um, and then meta with their new chips, and also working with AMD with with the new AI tools, they can actually design chips faster, and they can still use the design uh platforms like uh from Synopsis or Cadence, but
25:13
Unidentified speaker (speaker_2): I think that the uh the ability to build a new chip to compete with nvidia has been accelerated greatly just in the past six months and then i also own cerebrus with their wafer scale engine you know a chip that has four trillion
25:32
Unidentified speaker (speaker_2): transistors on
25:33
Unidentified speaker (UNKNOWN): it
25:34
Unidentified speaker (speaker_2): so um
25:36
Unidentified speaker (UNKNOWN): that
25:37
Unidentified speaker (speaker_2): uh you know that's another way to compete. And and they they just reported last night, and they've got new business with Microsoft
25:47
Unidentified speaker (speaker_2): and another hyperscaler. I forget now if it's uh uh maybe Google, but uh but the oh they were basically I said, Do I do you buy Cerebrus under 15 time sales with one big customer? And their one big customer was OpenAI. wanted to use the Cerebrus chips. And I said, Well, we don't like, even though this the
26:07
Unidentified speaker (speaker_2): stock got hammered from 200 down to below 100 because of all the IPO supply coming out, that's over with now. Strong report and new customers. And that's always what I said. Like Cerebrus can only go up now in terms of business growth because.
26:27
Unidentified speaker (UNKNOWN): they
26:27
Unidentified speaker (speaker_2): can only go from one customer to two, three, four. And so
26:32
Unidentified speaker (speaker_2): there's plenty of competition and diversity out there. And in my article that I wrote last week, I call it dissimilar redundancy. Dissimilar redundancy. What that means is any
26:43
Unidentified speaker (speaker_2): uh cloud provider, model builder, um
26:49
Unidentified speaker (UNKNOWN): uh
26:49
Unidentified speaker (speaker_2): enterprise. that's using these advanced compute models, they don't want to be dependent on a single platform. Even though NVIDIA CUDA is great, they'd like to be able to experiment with other chips and see what they do on different workloads. So that's uh that's my take on the diversity. I'm trying to think.
27:09
Unidentified speaker (speaker_2): I wrote this article um
27:12
Unidentified speaker (UNKNOWN): last
27:13
Unidentified speaker (speaker_2): week, and it it it comes out through a new uh Zach's newsletter called Zaxx Tech Trends.
27:19
Unidentified speaker (UNKNOWN): But
27:19
Unidentified speaker (speaker_2): I don't know if we have the archive up yet.
27:21
Unidentified speaker (UNKNOWN): But um but
27:23
Unidentified speaker (UNKNOWN): if
27:23
Unidentified speaker (speaker_1): possible, if possible, we'll include a link to that article in the article version of this podcast. So yeah, our listeners can read that. Excellent. Now we have to talk about the future of the AI trade after the blue
27:39
Unidentified speaker (speaker_3): wave. So we know that uh AI related stocks, except for Nvidia for reasons that we discussed, they most other stocks have staged an impressive rebound this week. And uh after last Thursday's uh steep sell-off in the aftermath of the midterm elections, um
27:59
Unidentified speaker (speaker_3): with Democrats taking control of not only both. houses of congress they have also won governorships in some critical states where most of the data centers are located Texas Ohio uh are located and are being built Texas Ohio Georgia so data center
28:19
Unidentified speaker (speaker_3): expansion now faces As a new reality what are your thoughts
28:24
Unidentified speaker (speaker_2): yeah it's uh it's really hard to pin down um because the
28:30
Unidentified speaker (UNKNOWN): the
28:31
Unidentified speaker (speaker_2): blue tsunami seemed like it was a a validation of ideas
28:37
Unidentified speaker (speaker_2): that uh you know
28:39
Unidentified speaker (UNKNOWN): the
28:39
Unidentified speaker (speaker_2): data center wave should be slowed down. But
28:44
Unidentified speaker (speaker_2): um, and and there are reports that yeah, regulations are gonna going to get stiffer,
28:50
Unidentified speaker (UNKNOWN): but I
28:50
Unidentified speaker (speaker_2): don't know if
28:52
Unidentified speaker (UNKNOWN): you
28:52
Unidentified speaker (speaker_2): know, if uh if any politician wants to step in the way of the job growth that we've seen because of this,
29:02
Unidentified speaker (UNKNOWN): you
29:02
Unidentified speaker (speaker_2): know, revolutionary technological transformation. Um I I call the I say what we're in right now is the fifth industrial revolution.
29:11
Unidentified speaker (UNKNOWN): The
29:11
Unidentified speaker (speaker_2): fourth industrial revolution started in 1995 with the internet.
29:14
Unidentified speaker (UNKNOWN): And
29:14
Unidentified speaker (speaker_2): we got things like genomics
29:17
Unidentified speaker (UNKNOWN): and
29:18
Unidentified speaker (speaker_2): then the cloud, mobile,
29:21
Unidentified speaker (speaker_2): um, SaaS, and all these things were,
29:25
Unidentified speaker (UNKNOWN): you
29:26
Unidentified speaker (speaker_2): know, really what the what the growth in tech was up until 2020. But slowly we we've transitioned into this acceleration, this exponential acceleration with technology because of AI.
29:39
Unidentified speaker (UNKNOWN): and
29:39
Unidentified speaker (speaker_2): because of NVIDIA and OpenAI. Um, so that's why I call it the fifth industrial revolution because the the convergence of technologies is making things happen faster.
29:48
Unidentified speaker (UNKNOWN): Um,
29:49
Unidentified speaker (speaker_2): you know, now we're talking about um intelligent robotics, where you have an embedded intelligence that's not even connected to the cloud. And uh and that takes a lot of compute training. So
30:03
Unidentified speaker (speaker_2): that that's why the the AI build out of data centers is so important, because you need to support that with training, inference, and um, and just the massive amount of data that is being processed. You know, they talk about the the token explosion, where
30:21
Unidentified speaker (UNKNOWN): we're
30:22
Unidentified speaker (speaker_2): we're headed towards quadrillions of tokens being produced. Um you know, uh uh on an annual or quarterly basis. So
30:33
Unidentified speaker (speaker_2): that's creating jobs. And you and the extreme examples you hear about are electricians making two to three times their normal salaries because
30:43
Unidentified speaker (speaker_2): the uh they just can't find enough electricians, and they'll pay, they'll pay electricians, you know, um
30:51
Unidentified speaker (speaker_2): multiples of their current salary, especially if they'll travel and go work on a project for three to six months.
30:58
Unidentified speaker (UNKNOWN): Um,
30:59
Unidentified speaker (speaker_2): pipe fitters,
31:00
Unidentified speaker (UNKNOWN): uh,
31:01
Unidentified speaker (speaker_2): carpenters, heavy equipment operators. Um, those are real jobs for
31:08
Unidentified speaker (UNKNOWN): the
31:09
Unidentified speaker (speaker_2): middle class in America that I don't think that uh any politician wants to get in the way of that. So they think I think they have to be very careful. And I think if they came in, if if they made campaign promises about, you know, um, we need to control the AI threat and we need to
31:29
Unidentified speaker (speaker_2): slow down the data centers, that that will soften as they realize how important it is to the economy and to the jobs. jobs.
31:38
Unidentified speaker (speaker_1): That's true. And in fact, we have seen that shares of some of the biggest hyperscalers have, in fact, arisen since the elections. Now, Microsoft is down because of the other woes, because of the outage that we experienced yesterday and to an extent today also. But Amazon and Google have done quite well
31:58
Unidentified speaker (speaker_1): since the election. And maybe investors believe that. this build out is not going to slow. Or otherwise, uh, even if it slows down, maybe existing data center operators will become much more valuable because their assets will become more valuable if uh the
32:18
Unidentified speaker (speaker_1): build-out slows down. The existing assets will become more valuable. Now, uh, could you walk us through the your bull? and bear case for AI stocks in 2027? Because we are almost at the ear end now.
32:34
Unidentified speaker (speaker_2): Yeah, and and first I'll comment on what you talked about the hyperscalers. They're benefiting because they're like the safest harbors in the storm, right?
32:43
Unidentified speaker (UNKNOWN): Right. Um,
32:44
Unidentified speaker (speaker_2): I mean, not not only do they have
32:47
Unidentified speaker (UNKNOWN): you
32:47
Unidentified speaker (speaker_2): know they'll say, oh, we don't have all the compute we want. they've certainly got enough compute already um they still have cash flow even though it's gone negative and they're borrowing but the these these are these are still sort of fortresses even if the even if the fortress balance sheet is is
33:07
Unidentified speaker (speaker_2): not as stellar as it once was these are fortresses where they can withstand um a lot and they already have the compute. So it's almost like they get they go, Oh wow, I'm so glad we were so aggressive in building. Like I like I I always quote Andy Jassy. Andy Jassy throughout this year, every
33:27
Unidentified speaker (speaker_2): quarter from Amazon, he'd be like, We would be fools not to be building like crazy, you know, because that's the way they look at the future. I always say, uh, during the the big uh ROI debates of the past year, I always flipped it on its head and I said, it's not about ROI return on investment, it's about RIO
33:47
Unidentified speaker (speaker_2): Rio. And RIL RIO stands for relevant instead of obsolete.
33:53
Unidentified speaker (UNKNOWN): And
33:53
Unidentified speaker (speaker_2): that's the way I looked at, you know, why are the hyperscalers going from spending 500 billion in 2025 to
34:04
Unidentified speaker (speaker_2): 750 billion this year to
34:08
Unidentified speaker (speaker_2): um
34:09
Unidentified speaker (speaker_2): projected one and a quarter trillion. Why are they spending that CapEx? Because those CEOs go to bed every night thinking about the hundred different ways that they could be disrupted, that their businesses, that their core businesses could be disrupted by stuff that Anthropic does.
34:26
Unidentified speaker (UNKNOWN): You know,
34:27
Unidentified speaker (speaker_2): anthropic has just uh made everybody paranoid about their business model. Remember uh the uh the old uh in Intel CEO, he wrote a book called uh Only the Paranoid Survive. Um and I mean how true is that
34:47
Unidentified speaker (speaker_2): still 30 years later now, especially with the anthropic threat.
34:51
Unidentified speaker (speaker_2): And they're going to IPO soon and have more cash to disrupt businesses. So
34:57
Unidentified speaker (speaker_2): Microsoft stock has actually held held up really well. Um, even though Anthropic just rolled out another threat to Excel.
35:06
Unidentified speaker (speaker_0): So
35:07
Unidentified speaker (UNKNOWN): um
35:07
Unidentified speaker (speaker_2): bull and bear case.
35:10
Unidentified speaker (UNKNOWN): I'm
35:10
Unidentified speaker (speaker_2): I'm bullish and look at the and
35:13
Unidentified speaker (UNKNOWN): look at
35:14
Unidentified speaker (speaker_2): the valuation. reset that we just got. So
35:17
Unidentified speaker (UNKNOWN): I
35:18
Unidentified speaker (speaker_2): I profiled that a lot of hedge funds got taken to the woodshed last week. When I say last week, I'm talking about the SOX pocalypse of November 5th. On November 5th,
35:32
Unidentified speaker (UNKNOWN): the
35:32
Unidentified speaker (speaker_2): semiconductor, Conductor Philadelphia Semiconductor Index, the SOX, was down 12% in a single day.
35:40
Unidentified speaker (UNKNOWN): Um,
35:40
Unidentified speaker (speaker_2): and closed on the lows at 9,000. That's a big deal. And what was going on was
35:48
Unidentified speaker (UNKNOWN): a
35:48
Unidentified speaker (speaker_2): lot of hedge funds just getting washed out. Um,
35:53
Unidentified speaker (UNKNOWN): the
35:53
Unidentified speaker (speaker_2): I I follow some of the people who track hedge funds on Bloomberg, and they're looking at us uh
36:01
Unidentified speaker (UNKNOWN): at the
36:02
Unidentified speaker (speaker_2): end of October.
36:03
Unidentified speaker (speaker_2): there were funds that lost 20 to 30 percent in October and were down
36:11
Unidentified speaker (speaker_2): 30 percent, 35 percent for the year.
36:15
Unidentified speaker (UNKNOWN): Well,
36:15
Unidentified speaker (speaker_2): that was just October, and we come into November,
36:19
Unidentified speaker (UNKNOWN): and
36:19
Unidentified speaker (speaker_2): you can imagine so you can imagine what was going on. If a hedge fund was already down 30 to 40 percent and getting uh calls for redemptions, um
36:29
Unidentified speaker (UNKNOWN): they
36:29
Unidentified speaker (speaker_2): got they got blown out last Thursday, November 5th.
36:34
Unidentified speaker (UNKNOWN): Now
36:34
Unidentified speaker (speaker_2): we bounced from
36:35
Unidentified speaker (UNKNOWN): that.
36:36
Unidentified speaker (speaker_2): The
36:37
Unidentified speaker (UNKNOWN): the
36:37
Unidentified speaker (speaker_2): key insight here is that uh valuations have become attractive again. Uh I own a stock stock called Coherent, which is uh involved in the optical technologies. Jensen invested $2 billion in Coherent and Lumentum, who are peers in this space
36:57
Unidentified speaker (speaker_2): of lasers and what Jensen calls silicon photonics, because he wants to integrate more optical hardware into the rack. Now, traditionally, when you think about the like Like the the the the current generation of NVIDIA rack scale systems. it's eight feet tall and has
37:17
Unidentified speaker (speaker_2): 72 GPUs in it.
37:19
Unidentified speaker (UNKNOWN): And
37:19
Unidentified speaker (speaker_2): those GPUs are connected by copper wire, which is traditional. And copper at that close distance works really well for conductivity and low heat. But um when you start when you connect um what's called a pod, which is eight of those racks together and 572 GPUs.
37:40
Unidentified speaker (speaker_2): you need fiber, lasers, and optics to connect them. Copper, you know, past a meter or two, um
37:50
Unidentified speaker (speaker_2): doesn't have the same signal strength. And so that's where the optical part comes in. And and Jensen's vision is we're gonna build more optical right into the rack, in right next to the GPU chip. It's either gonna going be to built next to it or integrated into it. It's what he calls extreme co-design. That that vision will go on. I I
38:10
Unidentified speaker (speaker_2): keep saying Jensen, but his engineers are working on this. And and they will continue to work on it and with these investments. So coherent. Let's talk about valuation. Oh it's the um
38:23
Unidentified speaker (UNKNOWN): uh
38:24
Unidentified speaker (speaker_2): your listeners are probably hearing this uh um on Monday, but I just wrote the bull of the day article on Zaxx.com about coherent. And so it's on the it's on the homepage all weekend, but the symbol is C-O-H-R.
38:40
Unidentified speaker (speaker_2): And I own it, so I'm talking my book, but just to give you an idea of the valuation, Situation, it's come down to three times sales, which is
38:51
Unidentified speaker (speaker_2): um pretty fantastic for a company in their position. I'll just take a look at it here. Yeah. 46 billion market cap
39:00
Unidentified speaker (speaker_2): estimates strongly up and to the right. Um, and 56%
39:08
Unidentified speaker (speaker_2): top line Upline growth this year, 82%, 83% EPS growth this year. And the the thing about all these estimates is everything
39:17
Unidentified speaker (UNKNOWN): will
39:18
Unidentified speaker (speaker_2): be hard to comp, right? Like 2027 estimates, it'll be hard to comp the high if you had high double digit growth in 2026, it's hard to comp that for next year. So just looking at um,
39:33
Unidentified speaker (UNKNOWN): I'm
39:33
Unidentified speaker (speaker_2): gonna going go to to the detailed estimates here while I'm on zax.com.
39:37
Unidentified speaker (speaker_2): And
39:38
Unidentified speaker (UNKNOWN): so
39:40
Unidentified speaker (speaker_2): 15 billion is the consensus for next year. And um
39:46
Unidentified speaker (UNKNOWN): uh
39:47
Unidentified speaker (speaker_2): coherent is on a June fiscal year. So
39:50
Unidentified speaker (UNKNOWN): uh,
39:50
Unidentified speaker (speaker_2): but 15 billion, that's three times sales, and they're still gonna grow 37%.
39:55
Unidentified speaker (speaker_2): And um and profitability, they're gonna going to do 15 bucks in EPS.
40:02
Unidentified speaker (UNKNOWN): I
40:03
Unidentified speaker (speaker_2): mean, they're they're uh
40:05
Unidentified speaker (speaker_2): they're trading a little bit more expensive than NVIDIA. So this is just one example of the valuations. And since I mentioned Lumentum, I should add that that created a wrinkle in the um
40:19
Unidentified speaker (speaker_2): in the semiconductor trade because Marvell came out to buy Lumentum
40:25
Unidentified speaker (UNKNOWN): in
40:25
Unidentified speaker (speaker_2): an all all-stock stock deal.
40:27
Unidentified speaker (speaker_2): 4.5 shares of Marvell for Lumentum. Well, thank gosh I'd sold Lumentum. We had Lumentum in my port my Taser portfolio, and we sold that
40:40
Unidentified speaker (speaker_2): um like around eleven hundred dollars back in October on that new high spike. I actually thought, you know, we were we we we we had the correction into July. We bounce out of that, we get a hike from the Fed, and and then we're back off to the bullish races. Nvidia,
40:59
Unidentified speaker (speaker_2): AMD,
41:01
Unidentified speaker (speaker_2): Lumentum,
41:03
Unidentified speaker (UNKNOWN): and
41:03
Unidentified speaker (speaker_2): one other stuff chip stock I've I'm I'm forgetting right now, all made new highs.
41:08
Unidentified speaker (UNKNOWN): And I
41:08
Unidentified speaker (speaker_2): thought, oh, this is great. But I still took profits on Lumentum. Well, it's been straight down since then, especially after the Marvell. deal, because it's an all-stock deal. And so,
41:19
Unidentified speaker (speaker_2): but it's what I'm saying is that all these stocks have been taken to the woodshed. We're coming into a valuation place where the the bullish case is that
41:29
Unidentified speaker (speaker_2): the hyperscalers are still going to invest over a trillion next year in building capacity, regardless of what happens with NVIDIA. And so uh all the stocks in the supply chain ecosystem, like Lumentum and Coherent,
41:44
Unidentified speaker (UNKNOWN): I
41:44
Unidentified speaker (speaker_2): mean, the the optical guys are sold out, the memory guys are sold out. Samsung, SK Heinex,
41:50
Unidentified speaker (speaker_2): um,
41:52
Unidentified speaker (UNKNOWN): uh
41:52
Unidentified speaker (speaker_2): Sandisk and Flash, they're sold out into 2028. So they're building capacity, and so I I don't foresee a bearish case unless, you know, the thing, the thing we obviously have to watch out for is if one of the hyperscalers says, um,
42:12
Unidentified speaker (speaker_2): we're actually backing off on some of our capex for 2027.
42:18
Unidentified speaker (UNKNOWN): Now,
42:18
Unidentified speaker (speaker_2): if we get a if we get a headline like that or an earnings call like that, then that's gonna ripple throughout. the ecosystem.
42:28
Unidentified speaker (speaker_1): Excellent. Now going back to your comment about only the paranoid survive, uh, which you said that uh
42:37
Unidentified speaker (speaker_1): yeah. So uh we were talking about uh Jensen's comment uh on Colette Crest and I saw that take Kim the journalist who wrote the book right uh the Nvidia way Jensen Huan Huang, and uh the making of a tech giant. Uh, he posted on social
42:57
Unidentified speaker (speaker_1): media, and I quote, for anyone worried about the quotes in the latest WSJ article, know this is a standard operating procedure for Jensen's management style. Bereit Bid the presenter or executive, instill fair fear with hyperbolic. world is ending statements
43:17
Unidentified speaker (speaker_1): that Nvidia is just 30 days away from going out of business and demand miracles from employees. So maybe this is the secret how Jensen built NVIDIA into an AI powerhouse. And this probably this was Steve Jobs style too, to write. right?
43:35
Unidentified speaker (speaker_2): Yeah. Uh Yeah, the the there is no nothing comparison to the founder-led company.
43:41
Unidentified speaker (UNKNOWN): Yeah. When
43:41
Unidentified speaker (speaker_2): it comes to their their vision and their tenacity. And yeah, if he was, if he,
43:49
Unidentified speaker (UNKNOWN): if
43:49
Unidentified speaker (speaker_2): there's evidence that he was rough, I mean take him backs it up. Yeah, I like take him. Um and he's he's studied NVIDIA.
43:56
Unidentified speaker (speaker_1): He
43:56
Unidentified speaker (UNKNOWN): knows the company
43:57
Unidentified speaker (speaker_1): very well. Yep.
43:59
Unidentified speaker (speaker_2): Yeah. So
44:01
Unidentified speaker (speaker_2): um
44:02
Unidentified speaker (UNKNOWN): and
44:02
Unidentified speaker (speaker_2): now that's gone, right? Now it's gonna become now that now the risk for NVIDIA is does it become more bureaucratic? We've lost the visionary founder. Um, we've lost that passion and intensity, uh at least at the top. And here's the thing about the organizational structure and why NVIDIA had no succession plan.
44:22
Unidentified speaker (speaker_2): Because Jensen had 60 direct reports,
44:26
Unidentified speaker (UNKNOWN): he
44:26
Unidentified speaker (speaker_2): wanted to. talk individually, one on one, to his most important leaders.
44:32
Unidentified speaker (UNKNOWN): And
44:32
Unidentified speaker (speaker_2): he so he didn't do big group meetings where he would embarrass anybody. He it was a lot of one-on-ones. And um,
44:40
Unidentified speaker (UNKNOWN): you
44:41
Unidentified speaker (speaker_2): know, so he so he had a grasp of the business at that granular granular level, and um,
44:48
Unidentified speaker (UNKNOWN): you know,
44:48
Unidentified speaker (speaker_2): so that's gone now, too. So there will be more.
44:51
Unidentified speaker (speaker_2): bureaucracy. That's why I I'm not predicting that it it's going to be a co CEO, Colette Cress and Ian Buck, but I that's just sort of my wish. Like that would be kind of cool. And it would it would say, Hey, we are going to try to fill Jensen's shoes, but it it's probably going to take at least two of us. And we're we're going gonna to carry on the vision.
45:11
Unidentified speaker (speaker_2): And
45:12
Unidentified speaker (UNKNOWN): we're
45:12
Unidentified speaker (speaker_2): going gonna to learn so much in the next few weeks.
45:15
Unidentified speaker (speaker_2): not about the investigations or espionage, but leading up to GTC. Um November 30th. So um
45:24
Unidentified speaker (speaker_2): the the other the other thing that happens here now with looking at the and the reason I profile like the top 12 people that could lead the company is that now we'll get a succession plan. I bet you Colette Cross Cress is a already drafting like this is what a succession plan should look like, because she's
45:44
Unidentified speaker (speaker_2): assuming, hey, I could stay on as CEO, but but maybe not, because I'm not a technology person. I'm not an engineer. So she's probably drafting a succession plan for the board to approve. And and so that's a positive. Um
46:00
Unidentified speaker (speaker_2): something
46:00
Unidentified speaker (UNKNOWN): which they should
46:01
Unidentified speaker (speaker_1): have done earlier, Matthew. but
46:03
Unidentified speaker (UNKNOWN): exactly,
46:04
Unidentified speaker (speaker_2): exactly. Yeah. Jensen would say, you know, why why should I do that?
46:08
Unidentified speaker (UNKNOWN): But
46:09
Unidentified speaker (speaker_2): uh, you know, the one of the biggest impacts is I I asked the question, what if Elon quit SpaceX?
46:16
Unidentified speaker (UNKNOWN): I
46:16
Unidentified speaker (speaker_2): mean, what would happen to SpaceX, right? I mean, so it's it's it's we're in that same kind of moment where
46:22
Unidentified speaker (speaker_1): yeah,
46:23
Unidentified speaker (speaker_2): everything has to be re recalibrated and re-evaluated.
46:26
Unidentified speaker (speaker_1): Yeah, we'll have to leave it there, Kevin. But uh thank you so much for joining us today and uh for sharing your insights.
46:33
Unidentified speaker (UNKNOWN): All
46:33
Unidentified speaker (speaker_2): right, great discussion. Thanks, Nina.
46:36
Unidentified speaker (speaker_1): That was my colleague, Kevin Cook, senior stock strategist at ZACs. Now let's uh recap uh some of the stock tickers that we talked about, and then I'll also talk about some of the ET. EDFs that have a lot of exposure to NVIDIA. So with
46:55
Unidentified speaker (speaker_1): Kevin, we discussed Nvidia. We all know the ticker symbol is NVDA. We talked about AMD, the ticker symbol is AMD. Broadcom, the ticker symbol is AVGO.
47:10
Unidentified speaker (UNKNOWN): We
47:10
Unidentified speaker (speaker_3): talked
47:11
Unidentified speaker (UNKNOWN): about
47:11
Unidentified speaker (speaker_1): Taiwan semiconductor, the ticker
47:13
Unidentified speaker (speaker_3): symbol
47:13
Unidentified speaker (UNKNOWN): is
47:14
Unidentified speaker (speaker_1): T
47:14
Unidentified speaker (UNKNOWN): TSM,
47:14
Unidentified speaker (speaker_1): S.
47:15
Unidentified speaker (speaker_1): and we discussed coherent the ticker
47:18
Unidentified speaker (speaker_3): symbol
47:18
Unidentified speaker (speaker_1): is COHR.
47:21
Unidentified speaker (UNKNOWN): Now
47:22
Unidentified speaker (speaker_1): talking about the ETFs,
47:24
Unidentified speaker (speaker_1): Nvidia is uh among the largest weights in many
47:29
Unidentified speaker (speaker_3): broad
47:29
Unidentified speaker (speaker_1): market ETFs. For
47:31
Unidentified speaker (speaker_3): example,
47:31
Unidentified speaker (speaker_1): S P
47:32
Unidentified speaker (speaker_3): 500 index
47:33
Unidentified speaker (speaker_1): like SPY, they
47:35
Unidentified speaker (speaker_3): have about 6.1% weight. in Nvidia. And the NASDAQ 100 EDFs like Qs, QQQ,
47:46
Unidentified speaker (speaker_3): they have about 6.5% weight in NVIDIA. And then NVIDIA is the almost in almost all AI or semiconductor ETFs in particular, it is among the top holdings. So if we look at the Venac semiconductor EDF
48:05
Unidentified speaker (speaker_3): SM SMH, Nvidia accounts for 17%
48:12
Unidentified speaker (speaker_3): of the portfolio. And uh the other popular semiconductor EDF SOXX, which is by iShares, and uh I own it in the EDF investor portfolio that I manage. This is a bit more diversified. So if we look at the top
48:31
Unidentified speaker (speaker_3): holdings, AMD is at 11%, and then Broadcom at about 9%, Micron at 7%. Nvidia is the fourth largest holding with about 6.6%
48:45
Unidentified speaker (speaker_3): weight. Then Nvidia is also a top holding in the technology sector, EDF. ETFs like XLK. So if I look at XLK, NVIDIA accounts for almost 13%
49:00
Unidentified speaker (speaker_3): of the portfolio. Thanks for listening. If you like our show, please leave us a rating on iTunes or wherever you get your podcast. Make sure to subscribe so that you do not miss any episode. If you have any comments, questions, or suggestions. please email podcast at zax.com.
49:19
Unidentified speaker (UNKNOWN): This
49:19
Unidentified speaker (speaker_0): material is being provided for informational purposes only, and nothing herein constitutes investment, legal, accounting, or tax advice, or a recommendation to buy, sell, or hold a security. Do not act or rely upon the information and advice given in this podcast without seeking the services of competent and professional legal tax or accounting counsel. Publication and distribution of this podcast is not intended to create in and the information contained here. herein does not constitute
49:39
Unidentified speaker (speaker_0): an attorney-client relationship. No recommendation or advice is being given as to whether any investment or strategy is suitable for a particular investor. It should not be assumed that any investments in securities, company, sectors, or markets identified and described were or will be profitable. All information is current as of the date herein and is subject to change without notice. Any views or opinions expressed may not reflect those of Zach's investment research as a whole.
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.

More from ETF Spotlight

All insights