What Broadcom's AI-Fueled Rally Means for the Chip Sector

Show Transcript
Bloomberg Jun 13 23:29 · 12.5k Views

Broadcom, a chip supplier for Apple and other big tech companies, saw its shares rise sharply after delivering strong results and an upbeat forecast, lifted by robust demand for artificial intelligence products. Millennia Capital Managing Partner Joe Zhao joins Caroline Hyde and Tim Stenovec to discuss his outlook for the chip sector in the context of AI on "Bloomberg Technology."

Disclaimer: This content is for informational and educational purposes only and does not constitute a recommendation or endorsement of any specific investment or investment strategy. Read more

Transcript

  • 00:00 We used to work over at the Fed.
  • 00:01 So you are the perfect person to sort of
  • 00:03 outline whether in this environment where eventually we might get one, maybe even 2 cuts this year in terms of overall policy, you should be putting yet more money into a Broadcom, into an NVIDIA, into some other AI plays.
  • 00:15 Yeah.
  • 00:15 Look,
  • 00:16 you know,
  • 00:17 by taking a step back, I, I think that there's, we're in this huge AI build out and there's a couple of ways you can play this market.
  • 00:24 You could, you could, you could participate in a, in a semis ETF,
  • 00:27 you could pick the winners.
  • 00:28 But I think what I'm, what I'm thinking I have is
  • 00:32 I think there's some naysayers who are saying,
  • 00:34 well, this AI build out is we're going to see the peak of the curve in the next couple of years.
  • 00:38 I personally think that this hyper growth phase in the
  • 00:41 in the 70s will probably go on for years just because I think what the street is not saying is, is that there's going to be a lot of incremental demand
  • 00:49 from some of the tech lagers that needs to catch up on AI.
  • 00:52 There's going to be foreign
  • 00:53 buyers from let's say Europe and emerging markets.
  • 00:56 And lastly, I think what we're what we're also thinking is there's going to be a lot of
  • 00:59 training and retraining of AIAI is like software every time you know on your iPhone you got to update your software every single,
  • 01:06 every single week.
  • 01:07 So in AI as their new data be becoming available, you have to update the AI, you have to retrain.
  • 01:12 So I personally think that the AI at the infiltrate
  • 01:15 has a has a lot of
  • 01:16 room, room to go.
  • 01:17 Well, Joe,
  • 01:18 I'm sticking with your experience on the Federal Reserve because look, at the end of the day, rates are pretty much everything when we're talking about this stuff.
  • 01:25 I, I'm wondering how you're thinking about these high flying tech companies and what exactly is priced in with regard to rate cuts.
  • 01:31 Are we going to see still moves higher even after the Fed does cut rates if it does cut rates this year, or is that already priced in?
  • 01:40 Look, as a former staffer now in a founder of my our firm,
  • 01:43 I take the view that each
  • 01:45 it's important to look at the long term
  • 01:47 directions of interest rates.
  • 01:49 You know, in economics there's a saying like we're 90% confident, we're 95% confident.
  • 01:54 We're probably 90% confident that there's going to be at least one rate cut sometime the next nine months.
  • 01:59 The weather, it starts in July, September, December, we're
  • 02:02 in January, I think is not as important to me as an investor because, you know, the directions of the rates are coming down.
  • 02:08 So what that means you, what you, what you can do is to put, you know, I, I think there's one more thing that I think the street often gets run out.
  • 02:14 And I'll say this like for the record, is
  • 02:16 I think the street put puts a lot of emphasis on the, that part chart.
  • 02:19 And I think
  • 02:20 what I would do is probably put 60% of your weight on the, that part chart and put 40% of your own research analysis into the thinking.
  • 02:27 So you come up with your own
  • 02:29 framework, what you to, to, to what you think is going to be the, the path of interest rates.
  • 02:32 And,
  • 02:33 and then you put that into your financial model, your DCS model on Excel.
  • 02:36 And then you can sort of figure out, you know, what you think is fair value on, on the S&P, on that
  • 02:41 and on the stocks.
  • 02:42 So for me, I just take, take a step back.
  • 02:45 I, I'm less worried about whether the,
  • 02:47 the cut starts in September or December, but I know I think there's a rate coming, a cut coming,
  • 02:51 so that I'll just put that into my model and then they'll be able to help me make better decisions as an investor.
  • 02:56 You're betting on the trend
  • 02:58 and you're betting actually on private companies.
  • 03:01 I'm thinking of Anthropic.
  • 03:02 You put money into some of the large language models.
  • 03:04 Dissect for us where
  • 03:06 the money there now is allocated in this space.
  • 03:09 Is it the large language models?
  • 03:11 At what point do we start seeing more money allocated to the applications, the use of these large language models?
  • 03:16 Yeah.
  • 03:16 So,
  • 03:17 you know, there's
  • 03:19 sort of six broader themes of this AI value chain in my head.
  • 03:23 If you start from the very bottom right, you have electricity,
  • 03:27 you have data centers, you have,
  • 03:29 you have cooling centers.
  • 03:30 And then on top of that you'll have
  • 03:33 chips to the semis
  • 03:34 and you have the data owners like the, the Max 7.
  • 03:37 And then on the on, on, on the top, you have
  • 03:39 the large language models.
  • 03:41 And then you need human talent, you need smart people to train the AI and work on it.
  • 03:44 And then lastly, you need the, you know, we're going to build the applications.
  • 03:48 I think as an investor, you know, you can obviously play this AI trade in public markets by buying
  • 03:53 public stock, the QQQ.
  • 03:55 But in private markets, what we're doing is sort of like we're sort of an ETF in the private markets in that we're
  • 04:00 invested in Entropic, Lambda, Cohere, because these are the innovations
  • 04:05 companies happening in the private markets.
  • 04:07 But there's also other ways to play this AI trade because you could also be investing in utility, you could
  • 04:11 be buying data centers.
  • 04:13 So I think there's no different ways, but I think the fundamental
  • 04:17 thesis in my head is
  • 04:18 that
  • 04:19 it's a foregone conclusion
  • 04:21 whether AI is real.
  • 04:22 In fact, I think AI is going to be, you know, like software square, the Internet cube, if you will, you know, excuse me, but
  • 04:28 I think that the biggest question in my head is actually AI regulation.
  • 04:30 But that aside,
  • 04:32 I think it, when you take a step back and you look at some of the companies in this AI trade, we're at the very beginning where you know, you can think of it like a 2 year old baby.
  • 04:40 And, and the, and so the saying is being plain words, if this baby at 2 year old
  • 04:44 can do what it can do, imagine when that baby
  • 04:48 now matures as as an adult,
  • 04:49 how much more that person can do.
  • 04:51 OK, well, as a venture capitalist, you're thinking about that baby as an adult.
  • 04:55 So very briefly, paint a picture for us about what AI can do when it's 19 years old.
  • 05:00 Yeah, then you're, you're then we're talking about the, the, the sixth layer, which is the, the application layer, right.
  • 05:05 So there's, you know,
  • 05:06 I guess the, the, the industry phrase here is called
  • 05:09 vertical software, which is you can apply AI in cybersecurity, in automation, in robotics.
  • 05:14 You know, there's a company called Figure AI that's combining robotics with,
  • 05:17 with AI.
  • 05:18 We're invested in a AI cybersecurity company called DB Instinct.
  • 05:22 And but you can also apply AI to farmer research
  • 05:24 to, you know, many other fields that we haven't even thought about.
  • 05:28 There's one thing I got to say that's really, really important.
  • 05:30 When we think back at the major technological innovations of the last
  • 05:34 200 years,
  • 05:35 whether it was electricity, the engine, the Internet,
  • 05:39 how did we as a humanity create that?
  • 05:41 We did it with human intelligence, people working together.
  • 05:44 What happens now when you when you can recreate intelligence, you can recreate many, many times medicines.
  • 05:51 So there's many infinite possibilities that we can create.
  • 05:54 So what what I'm trying to say is
  • 05:55 it's, it's really, I think about how we
  • 05:58 nurture this technology
  • 06:00 in a way that's going to help us
  • 06:02 help help the community move forward
  • 06:04 and grow in a way that's responsible.