Bernstein Research's Rasgon on The Year of Nvidia

Show Transcript
Bloomberg Dec 19, 2023 05:45 · 6970 Views

Bernstein Research MD & Senior Analyst, Stacy Rasgon speak with Katie Greifeld and Scarlet Fu to discuss the year of Nvidia on Bloomberg Markets: The Close.

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 Given that outperform
  • 00:02 rating that you have, it seems like you think this is
  • 00:04 this momentum is going to continue.
  • 00:07 Yeah, look, I think numbers for now we're going up and probably up materially.
  • 00:12 Demand right now is off the charts.
  • 00:15 They don't have the ability to supply like anywhere near where the demand is
  • 00:19 now we can argue about sustainability and and it and it's kind of funny like if if you look at the stock, the stocks like more than tripled this year
  • 00:26 and in fact the multiple has actually come down even more.
  • 00:29 It's actually the cheapest of all the AI stocks out there.
  • 00:32 It's cheaper than A&B, it's cheaper than Marvell, it's even cheaper than Intel.
  • 00:36 And the reason is the numbers are getting so big so quickly.
  • 00:39 People are worried about that sustainability, right.
  • 00:42 And I and I kind of get that we can talk about it.
  • 00:44 But
  • 00:45 in the near to medium term over the next like you know
  • 00:48 probably these 12 months maybe more,
  • 00:50 my suspicion is the numbers are much more likely to go up than down.
  • 00:54 I think the valuation has already come down and I think it's extremely attractive here.
  • 00:57 And I will be honest,
  • 00:58 I am incredibly bullish long term on the opportunity here both around AAI
  • 01:04 as well as
  • 01:05 around the the broader opportunities around accelerated computing.
  • 01:08 And I'm still convinced
  • 01:10 we're looking up five years, 10 years.
  • 01:12 We'll be talking about numbers that are still materially higher than the numbers we're talking about today.
  • 01:16 OK.
  • 01:16 So you make a good case, especially when you point out that it's actually cheaper than some of the other AI stocks.
  • 01:21 But
  • 01:21 people who want to have exposure to AI don't want to put all their
  • 01:25 eggs in one basket and only be exposed to NVIDIA, because what if something happens?
  • 01:30 What if Jensen Huang gets hit by truck, for instance?
  • 01:33 Let's let's
  • 01:34 not, Right.
  • 01:34 Right.
  • 01:34 So what would be the next best alternative?
  • 01:37 I mean if if everything else is expensive, where do you go to just get some
  • 01:41 diversification?
  • 01:43 Well, let's talk about the option.
  • 01:44 I mean the the natural one that people are looking at is AMD, right, because I mean they've they've got a road map
  • 01:49 with products on it and they've given some numbers and they they actually, by the way they, they did this very smartly.
  • 01:55 They gave a a number when they reported earnings, a few, oh, a month or two ago.
  • 01:59 And they said we think we'll do more than $2 billion of of AI revenues next year.
  • 02:04 But it was not a good quarter that they had.
  • 02:06 I suspect if they hadn't given us a number, the stock would, would not be nearly as strong as it has been.
  • 02:10 But they've given people a narrative to latch onto.
  • 02:13 They've actually left room for those numbers to come up because 2 billion is not very much.
  • 02:18 All of it in, in the grand context of things, it's a rounding error, like let's say they do two or three or four or whatever it is,
  • 02:24 like if NVIDIA were to do 70 next year, it'd probably be viewed as disappointing.
  • 02:28 So
  • 02:29 I think there's a view that
  • 02:31 no matter what number AMD puts out, they can hit it,
  • 02:35 right.
  • 02:35 And I think there's a desire to be able to play this, this sort of a theme
  • 02:39 with less perceived cult like air pocket risk.
  • 02:42 Although I'm I'm not actually sure that
  • 02:44 if there really was a real air pocket then anybody would be spared.
  • 02:46 But that that is sort of the perception.
  • 02:48 And and AMD, like I said, to their credit, they've left room I think for that narrative to continue
  • 02:53 and that's what what people would like to see.
  • 02:55 And so clearly people in buying the stock on that theme, OK.
  • 02:58 So there's a lot of messaging here that's taking place.
  • 03:01 I I know we mentioned Jensen Huang getting hit by truck and hopefully that doesn't happen.
  • 03:04 But is there an issue with key man risk here for NVIDIA?
  • 03:08 And look, Jensen's a pretty remarkable individual and he is the heart and soul of NVIDIA.
  • 03:13 He's he's in his early 60s I think he doesn't seem like he's going anywhere though.
  • 03:18 But I mean clearly, you know Jensen and NVIDIA are are sort of viewed as kind of one in the same.
  • 03:24 Yeah.
  • 03:24 I wouldn't want to see Jensen go anywhere.
  • 03:26 Sure.
  • 03:27 OK.
  • 03:27 So maybe
  • 03:28 key man risk,
  • 03:30 not a huge concern.
  • 03:31 What about geopolitics when it comes to NVIDIA, you think about the US
  • 03:35 and China, the relationship there, export controls, how are you modeling that into your very bullish NVIDIA forecast?
  • 03:42 Yeah, you bet.
  • 03:42 So people are worried about China, cuz like as of right now, they can't ship anything really in into China.
  • 03:47 The new export controls are fairly constrictive
  • 03:50 and they effectively block every product that they were shipping.
  • 03:53 And these ones are are more difficult to work around than the than the prior ones that were put into a year a year ago which were fairly easy to work around.
  • 04:00 Now that being said, it does look like NVIDIA is developing new products
  • 04:06 that
  • 04:06 ideally will be legal to sell into China.
  • 04:08 They will, they will be much lower performance than what they're currently selling,
  • 04:11 but they will be software compatible.
  • 04:14 And my view was like like look like right now there's enough demand from everywhere
  • 04:18 that they don't have to really worry about China and and they're able to backfill that demand.
  • 04:22 Like clearly clearly with
  • 04:23 even though it's gone, they they've got other folks who are lining up to buy those parts
  • 04:26 with some of the new products that are developing.
  • 04:28 I would say maybe we don't have to completely write off
  • 04:31 China.
  • 04:32 I would say though over the long term
  • 04:35 I think you do have to view
  • 04:37 the China Tam as potentially being impaired.
  • 04:40 I think it was gonna happen anyways though, whether or not new export controls were put into place.
  • 04:44 Because I think about this,
  • 04:46 even if these performance thresholds were never lowered, they were never going to be raised.
  • 04:51 And so over time the gap between what was legal to sell in China versus elsewhere was going to just going to get bigger and bigger and bigger and bigger anyways
  • 04:58 and so.
  • 05:00 The the what their ability to sell stuff in China was going to get less and less competitive anyways.
  • 05:04 And so I think
  • 05:05 this may pull forward that point
  • 05:08 but over time it was going to happen.
  • 05:09 I still think the overall market, you know China right now is about has has been about 20% of their data center sales.
  • 05:14 I think even if you check what you thought the overall opportunity was in five years or 10 years and and hair cut up by 20%, it's still big.