Latest / Best of Business / Sam Dickie: Fisher Funds expert on whether we should be worried about the AI bubble

Transcript

Sam Dickie: Fisher Funds expert on whether we should be worried about the AI bubble · Best of Business

• 0:09
  Speaker 1: You're listening to a podcast from news Talk zed B.
• 0:12
  Speaker 1: Follow this and our wide range of podcasts now on iHeartRadio.
• 0:17
  Speaker 1: Brian Bridge.
• 0:19
  Speaker 2: Right, it is time to get to what I've been
• 0:21
  Speaker 2: promising you all evening, and that is the markets. Let's
• 0:25
  Speaker 2: go to Sam Dickey, Fisher Fund Sam. Good evening, Good evening,
• 0:28
  Speaker 2: Big week four tech earnings. We've talked a few times
• 0:31
  Speaker 2: before about how it's quite frothy. What are you seeing
• 0:34
  Speaker 2: out there? Should we start with Measure and Microsoft? Yeah,
• 0:38
  Speaker 2: that's right.
• 0:38
  Speaker 3: I mean, in particular, we've been talking about how frothy
• 0:42
  Speaker 3: certain parts of the AI trade have been, in particular Memory,
• 0:46
  Speaker 3: and that's been the primary bottleneck or shortage in the
• 0:48
  Speaker 3: AI trade. And that's where the biggest cracks have just
• 0:51
  Speaker 3: appeared since we last spoke about this. So those rampant
• 0:55
  Speaker 3: kind of animal spirits we talked about a few weeks
• 0:57
  Speaker 3: ago have completely evaporated. So s k Heinez, for example
• 0:59
  Speaker 3: in Korea, which is the poster child for the memory boom,
• 1:02
  Speaker 3: has fallen almost sixty percent since we last.
• 1:05
  Speaker 2: Spoke about it.
• 1:06
  Speaker 3: As retail investors enthusiasm and career unwinds now, AI demand
• 1:11
  Speaker 3: is definitely not vanished overnight. AI heavyweights like Micron, Heinez
• 1:16
  Speaker 3: and Samsung before we talk about Meta and Microsoft all
• 1:19
  Speaker 3: posted monster earnings growth. But it's just given the rampant
• 1:22
  Speaker 3: speculation we've been talking about, the stellar results just simply
• 1:26
  Speaker 3: weren't good enough to satisfy the pretty lofty expectations.
• 1:30
  Speaker 2: What about I mean you mentioned Heine's and Samsung In
• 1:33
  Speaker 2: South Korea, the Korean index heavily reliant on those particular
• 1:39
  Speaker 2: stocks doing well, right, that's right, very much so.
• 1:43
  Speaker 3: I think at the peak they got up to sort
• 1:44
  Speaker 3: of fifty five to sixty percent of the total stock market.
• 1:47
  Speaker 3: That's two stocks. And again, it hasn't that this bubble
• 1:51
  Speaker 3: isn't deflating because demands suddenly evaporated overnight. It's just that
• 1:56
  Speaker 3: the speculation was a bit too rampant, and also there
• 1:58
  Speaker 3: was some genuine real world constraints this trade was running into.
• 2:01
  Speaker 3: But I think Meta and Microsoft critically important because they're
• 2:06
  Speaker 3: the biggest customers of you know, Heynex and Samsung. They're
• 2:10
  Speaker 3: the guides alongside Apple that are buying all this memory
• 2:14
  Speaker 3: to put into the AI servers to run these complex
• 2:18
  Speaker 3: AI models, and both of those businesses, the businesses themselves
• 2:22
  Speaker 3: are doing very well. Meta and Microsoft, They're firing on
• 2:24
  Speaker 3: all cylinders. So you know, take Microsoft's massive as your
• 2:29
  Speaker 3: cloud business that accelerated again driven by AI demand, and
• 2:32
  Speaker 3: Meta showed AIS driving big payback on advertising spin for
• 2:37
  Speaker 3: their customers. It's making their customers get a lot higher
• 2:39
  Speaker 3: return on ad spin, which is stoking demand for their advertising.
• 2:42
  Speaker 3: So the debate is definitely not that AI demands evaporating
• 2:46
  Speaker 3: overnight or that it doesn't work. It's just about what
• 2:49
  Speaker 3: it costs. And again we've spoken about this before. Between them,
• 2:51
  Speaker 3: those two companies are expected to spend hundreds of billions
• 2:55
  Speaker 3: of dollars in CAPEX over the next few years, but
• 2:57
  Speaker 3: they can no longer fund it out a cash flow alone,
• 3:00
  Speaker 3: So they've been taking on massive debt for the first
• 3:04
  Speaker 3: time ever, and the debt capital markets are pushing back
• 3:08
  Speaker 3: by marning higher interest rates. And interestingly, Ryan today, for
• 3:11
  Speaker 3: the first time in two years, Meta didn't raise its
• 3:15
  Speaker 3: kapex guidance. So maybe maybe they're finally it in the message.
• 3:18
  Speaker 2: That they're tapping out. Because I was interested reading something
• 3:21
  Speaker 2: in the Wall Street Journal this morning, Sam, and it
• 3:23
  Speaker 2: was talking It was comparing the data center. You know,
• 3:25
  Speaker 2: you're building all these data centers great comparing that to
• 3:28
  Speaker 2: the railway boom of you know, centuries ago. Now, where
• 3:32
  Speaker 2: you build the railway, once you get you make money
• 3:35
  Speaker 2: off of it for decades. You only have to build
• 3:37
  Speaker 2: it once. With the data center, you will build it.
• 3:40
  Speaker 2: But then maybe in two years, maybe in three years,
• 3:43
  Speaker 2: maybe in four everything's going to need to be replaced
• 3:46
  Speaker 2: to keep up with the technological change. So how you
• 3:49
  Speaker 2: know how much of a long term investment have we
• 3:52
  Speaker 2: got here?
• 3:53
  Speaker 3: Yes, that's been a critical debate is what is the
• 3:57
  Speaker 3: what is the effective useful life of the you know,
• 4:00
  Speaker 3: the Invidia GPUs or in video chips or even Heinez
• 4:05
  Speaker 3: and Samsung's dram and nan memory. So so what is
• 4:08
  Speaker 3: the effective useful life? And most of these big tech
• 4:11
  Speaker 3: companies say that's six years, and then Video will tell
• 4:13
  Speaker 3: you that even its chip versions from five or six,
• 4:19
  Speaker 3: the versions ago are still being fully utilized. So that
• 4:22
  Speaker 3: is a debate, but I agree with you, it's definitely
• 4:25
  Speaker 3: a lot, regardless of whether it's four or six years.
• 4:27
  Speaker 3: It's definitely not sort of twenty five or thirty years,
• 4:29
  Speaker 3: which the average railway lasts for.
• 4:33
  Speaker 2: Yeah, with a bit of maintenance. Of course, Hey, what
• 4:36
  Speaker 2: about the fedsh decision today? Three of the twelve wanted
• 4:41
  Speaker 2: to go up, but Walsh said, hey, just tie home.
• 4:45
  Speaker 3: That's right. He held rates and it was pretty spicy.
• 4:48
  Speaker 3: I think he called it a bit of a family
• 4:50
  Speaker 3: family feudal family debate. I think what's super interesting in
• 4:53
  Speaker 3: the are is, and he's obviously very new, is while
• 4:57
  Speaker 3: short term rates fell because obviously there was something priced
• 4:59
  Speaker 3: in for an immediate rate hike after he singled no
• 5:04
  Speaker 3: immediate hike, longer term yields actually rose quite sharply. It
• 5:08
  Speaker 3: was basically the market calling rubbish on his communications skill,
• 5:12
  Speaker 3: saying if he doesn't stack tackle such stubborn inflation now
• 5:16
  Speaker 3: he'll have to hike harder and faster down the track.
• 5:19
  Speaker 3: So I think that was pretty important.
• 5:21
  Speaker 2: Sam. Good to you on the show. Sam Dicky Fisher funds.
• 5:24
  Speaker 1: For more from News Talks B Listen live on air
• 5:28
  Speaker 1: or online, and keep our shows with you wherever you
• 5:30
  Speaker 1: go with our podcasts on iHeartRadio