Cathie Wood Says Nvidia Is More Than ‘the Jensen Show’ — Transcript

Full transcript

1. 0:00Earlier today, I spoke to legendary tech
2. 0:02investor and Arc Invest CEO Kathy Wood.
3. 0:05It was this morning before we got the
4. 0:06news of Nvidia CEO Jensen Huang stepping
5. 0:09down. Right before that, actually, she
6. 0:11joined Yahoo Finance Invest to discuss
7. 0:13the future of Nvidia and what this news
8. 0:16means for the AI trade.
9. 0:19>> Well, he obviously is a force and a
10. 0:22power behind Nvidia. I've been watching
11. 0:25him uh since he's gone public, since
12. 0:28Nvidia went public.
13. 0:30uh and he had this vision way back then
14. 0:33in the late late 90s. Uh he understood
15. 0:35well before uh most people how profound
16. 0:39this AI revolution was going to be and
17. 0:41how many sectors uh it was going to
18. 0:44touch and transform. Uh so there's no
19. 0:48question he's a force. But I would say
20. 0:50we have another very big force in the
21. 0:53industry. Uh first of all um before I go
22. 0:57there I'll say that Elon Musk said he
23. 0:59would consider it an honor to hire
24. 1:01Nvidia and you remember uh e Elon or or
25. 1:06uh Tesla in in effect has started a
26. 1:10terra fab project for computing power in
27. 1:13space and that is where we're going
28. 1:15given all the difficulties here on
29. 1:16earth. Um but uh so so I don't think the
30. 1:22movement is going to stop. You've got
31. 1:24powerful forces like Elon Musk who uh is
32. 1:28also a visionary and understands where
33. 1:30AI is going. And you have companies like
34. 1:33OpenAI and Anthropic who are pushing
35. 1:37this movement forward uh with such great
36. 1:40force that those of us turn around and
37. 1:43last year at this time uh Anthropics at
38. 1:469 billion annualized run rate. Uh by the
39. 1:50end of this year it could be at a
40. 1:51hundred billion. uh so these companies
41. 1:53are doing now I'll tell you one other
42. 1:56thing just to to become uh a little more
43. 1:59sober I I would say about it from an
44. 2:02analytical point of view we had been
45. 2:04assuming that Nvidia would lose share it
46. 2:07has about 83% share in GPUs right now we
47. 2:10had been assuming uh that it would lose
48. 2:13share to companies like Tesla with
49. 2:15Terapab uh and Broadcom and and others
50. 2:20Cerebrris a new company that doesn't
51. 2:23need high bandwidth memory. Um so we
52. 2:25assumed that share would drop to 49%.
53. 2:30And uh so we had a very very strong
54. 2:33compound annual rate of return
55. 2:35expectation for the next uh five years.
56. 2:38If we drop that to 38%.
57. 2:43Our compound annual rate of return would
58. 2:47drop from in the 50s somewhere to 25%
59. 2:51over the next 5 years. So we have to get
60. 2:53over this shock. We have to see who the
61. 2:55new leader is. The new leader is not
62. 2:58going to be alone in driving Nvidia. It
63. 3:01is going to be competition uh like
64. 3:03Tesla, Terrafab, but it is also going to
65. 3:06be these companies who have become very
66. 3:08reliant on Nvidia pushing the envelope
67. 3:12like Anthropic and OpenAI. And I think
68. 3:17share share losses will be more a
69. 3:20function of the competition we were
70. 3:22expecting anyway. Uh and in fact uh we
71. 3:26tweaked up our market share assumptions
72. 3:28recently to something like 53% in five
73. 3:33years because Elon Musk said he was
74. 3:37committing to Nvidia for SpaceX's
75. 3:41orbital compute.
76. 3:43>> And are you saying that changes now?
77. 3:45You're bringing that forecast down. uh
78. 3:47we're just doing our scenario testing
79. 3:49and saying okay you know Nvidia has a a
80. 3:55roadmap already planned out probably for
81. 3:57the next five years uh and so we think
82. 4:00it's a conservative assumption uh to say
83. 4:04conser highly conservative assumption to
84. 4:06say okay the share drops from 83% to 38%
85. 4:10and we still have a 25% compound now I
86. 4:14will say one other thing in terms of
87. 4:16sizing positions in our portfolios. We
88. 4:19do have a scoring system and the first
89. 4:22variable is people management and
90. 4:25culture. So let's see who they bring in.
91. 4:28Clearly uh there might be something
92. 4:30about the culture we did not know
93. 4:32before. Uh and now video will have to be
94. 4:35holier than thou under the new leader.
95. 4:39>> We're certainly finding out news, you
96. 4:41know, feels like every hour. Do you
97. 4:43think that there is someone that could
98. 4:45take that title lead Nvidia? I know you
99. 4:47talked about some of the other companies
100. 4:48that are growing at extraordinary rates
101. 4:50in the system. SpaceX, Anthropic,
102. 4:52OpenAI. Um, yes, this is the AI trade,
103. 4:55but Nvidia plays a very unique spot in
104. 4:58it, right? I I I don't know that you can
105. 5:01just replace it, right? I mean, it's
106. 5:03going to take time even on the custom AI
107. 5:07chip front. Um, so what does this do in
108. 5:10the shorter term to the AI trade as a
109. 5:12whole?
110. 5:14>> You know, I I've watched uh Jensen over
111. 5:17the years and over the years the way his
112. 5:21presentations have gone would have his
113. 5:24pictures and his you know, so he it was
114. 5:26really a lot about him.
115. 5:28>> The Jensen show,
116. 5:29>> the Jensen show, Jensen's face on all
117. 5:32the robots and and and so forth. Um, I
118. 5:36guarantee you there is a group of people
119. 5:39within Nvidia that he relied on
120. 5:43strongly and that there is a leader.
121. 5:46Will that leader be the the force?
122. 5:49Probably not because those people left
123. 5:53Nvidia if they wanted to be the force
124. 5:55and the power. But my sense is there is
125. 5:58a very strong group of talent otherwise
126. 6:01Nvidia would not be where it is today.
127. 6:03I I'm sure there is. I mean, 60 direct
128. 6:06reports, right? That is just a huge
129. 6:07number. But people I've asked are kind
130. 6:09of skeptical. Also, he's just I mean,
131. 6:11you know him, Kathy. He's a workhorse.
132. 6:13He has his eye on everything. He's a
133. 6:15marketer. He's extremely technical. Um,
134. 6:18he is the face of the company. What are
135. 6:21you doing? What do your Nvidia holdings
136. 6:23look like right now? What are you doing
137. 6:24with them?
138. 6:25>> You know, it's interesting. Uh, we don't
139. 6:27pay attention to benchmarks to be
140. 6:29honest. and uh we just size our
141. 6:31positions where we think they should be
142. 6:34given our outlook for the future. So uh
143. 6:37I believe uh our flagship ARK has
144. 6:41something like a 3.4
145. 6:44percent position something around there.
146. 6:46I I was out of town yesterday. So 3.4
147. 6:49four. If you look at the S&P 500 and the
148. 6:53Q's, they're closer to seven, eight, and
149. 6:56I think in one benchmark 12%. Uh, so
150. 7:00we're nothing near that. This is the
151. 7:02only time I've looked at a benchmark.
152. 7:03Okay, what what are other people going
153. 7:05to be feeling today? Well, those who
154. 7:08really manage their portfolios toward a
155. 7:11benchmark like that probably have a lot
156. 7:14more Nvidia than they really want, but
157. 7:16they manage to a benchmark.
158. 7:19So,
159. 7:19>> so are you, so you've been underweight,
160. 7:22Nvidia. Is this a buying opportunity for
161. 7:24you or do you sell a little more?
162. 7:28>> Well, uh, we've been buying recently
163. 7:30because of the SpaceX deal. Um, in fact,
164. 7:33it it has, you know, because of people
165. 7:35management and culture as one of our
166. 7:37scores. I've always had this little bug
167. 7:39in my brain saying, "Wait a minute, this
168. 7:41cannot be just the Jensen show." Every
169. 7:43time he got on stage, uh, he needed to
170. 7:46he needed to present other people. But I
171. 7:48guarantee you in order to do what Nvidia
172. 7:51has done, he has a very strong group of
173. 7:55engineers behind him. Are they great
174. 7:57marketers? Probably not. That's why
175. 8:00they're happy where they are. He does
176. 8:01all of that for them. Uh but I I I
177. 8:05believe very strongly he has incredible
178. 8:07talent. It was a coveted job.
179. 8:10>> Very interesting. Let's talk a little
180. 8:12bit about the rest of the AI trade. It's
181. 8:14been volatile in recent weeks and a lot
182. 8:16of that has centered around, you know,
183. 8:18Neoclouds and some of the financing
184. 8:20concerns. Of course, we got Coreweave
185. 8:22earlier this week. Um, you know, allayed
186. 8:25some of those concerns. How are you
187. 8:27feeling about the Neocloud space right
188. 8:28now? You know, we um we were able to
189. 8:31speak to the top management team at
190. 8:33Cororeweave yesterday after after its
191. 8:36report and this is a topnotch management
192. 8:41that has because of all of the risks out
193. 8:44there, you know, insisted on
194. 8:47prepayments. Now, one of the key
195. 8:50questions we asked uh about prepayments
196. 8:53is, you know, historically, if a
197. 8:56company's willing to prepay,
198. 8:59>> then uh then it's going to get a
199. 9:02discount. Guess what? There is no
200. 9:04discount. It's gone from discount a few
201. 9:07years ago to no discount. Why? Because
202. 9:11the demand out there is so voracious and
203. 9:14coreweave is that good. So just hearing
204. 9:18that one fact um gives us a lot of
205. 9:21reassurance that you know if data
206. 9:24centers aren't going to be built in
207. 9:26Arizona or Texas uh they will be built
208. 9:30somewhere else. And the other thing core
209. 9:32pointed us to was how global its
210. 9:35strategy is. I am sure that the election
211. 9:38year rhetoric they've had a year to
212. 9:40prepare for this has pushed them faster
213. 9:43into other markets. So here, you know,
214. 9:46we have to be careful. We do not want to
215. 9:48leave lose our innovation edge.
216. 9:52>> What are some of those other markets,
217. 9:54Kathy, if they can't build here in the
218. 9:55US?
219. 9:57>> Um, you know, I probably should have
220. 9:58tuned in uh a little bit more on that
221. 10:01point. I was surprised at how strongly
222. 10:03they made the point though and how many
223. 10:04times they uh repeated it. Uh I asked
224. 10:08them yesterday, I is this an
225. 10:10acceleration in your strategy? They said
226. 10:12no. They they didn't say yes or no on
227. 10:14that one. Uh but they they want to be a
228. 10:17global cloud effectively uh and and
229. 10:21really give our existing global clouds
230. 10:24uh a little bit more of a a run for
231. 10:26their money. But we need all that
232. 10:27capacity.
233. 10:29>> This kind of connects back to the Jensen
234. 10:31question, right? Because he had a lot of
235. 10:32soft power as well and had a hand in
236. 10:36building up these neoclouds, right? he
237. 10:38gets to allocate who gets the GPUs that
238. 10:41will fulfill compute that everyone is
239. 10:43looking for. Um, do you worry that if
240. 10:46he's not there, maybe there's a
241. 10:48different leadership that allocates
242. 10:50differently and you know there's the
243. 10:52potential for the Neoclouds to not get
244. 10:54the GPUs they need.
245. 10:56>> There's no question relationships uh are
246. 10:59important in any business. Uh but
247. 11:02coreweave from a neocloud view there
248. 11:05there are only two neoclouds that have a
249. 11:08platinum level uh um rating I think it's
250. 11:13artificial intelligence I could be wrong
251. 11:15on that um and so th those would be
252. 11:18corweave and nebus so whatever happens
253. 11:21here you know the demand's so strong I
254. 11:23just gave you a very important signal
255. 11:25these huge companies who with voracious
256. 11:28demand do not have to pay a discount
257. 11:31and as they prepay for the next few
258. 11:34years,
259. 11:36>> right? And and certainly those
260. 11:37prepayments help sort of settle some of
261. 11:39the nerves in the market. But Kathy, I
262. 11:41think we talked about this a few weeks
263. 11:42ago, this cone of uncertainty, right?
264. 11:44And that's what we've seen jitters in
265. 11:46the market, especially the credit
266. 11:47markets over the last few weeks. Um,
267. 11:49demand can be enormous and financing can
268. 11:51still go wrong. So essentially, these
269. 11:54companies are spending today huge
270. 11:56amounts of money for revenue um years
271. 11:58from now. How much of a delay or
272. 12:01technological change do you think that
273. 12:03they can withstand? I know that you look
274. 12:04over the longer term.
275. 12:06>> Yeah, I would say and I do believe we
276. 12:08discussed this the last time. Um the the
277. 12:12bigger risk near-term has been the the
278. 12:15credit markets. Now I have always said
279. 12:18that um you know traditional equity
280. 12:22investors really who who are are very uh
281. 12:29uh well they're very uh closely or
282. 12:31highly correlated with their benchmarks.
283. 12:34Um so they'll be looking to the future.
284. 12:38Um but they are spending a lot more time
285. 12:40with you know present- day companies not
286. 12:43looking at investments as as in such a
287. 12:46long term as we are. The bond markets
288. 12:49are even worse. Technology
289. 12:52typically has not relied on bondarks
290. 12:56bond markets. Why?
291. 12:57>> Because it's been software driven. So
292. 13:01>> viral apps don't need a lot of capital
293. 13:05spending. they go viral and and there's
294. 13:07been a huge huge cash flow huge cash
295. 13:09flow. So does that worry you that way?
296. 13:13>> So so we are relying on debt markets who
297. 13:17probably do not have technology analysts
298. 13:21certainly not forwardthinking ones. Uh
299. 13:24that has to change. Uh and and I think
300. 13:27it will uh this is too important in
301. 13:30terms of of our competitive standing. So
302. 13:33I think there's going to be a lot of
303. 13:34pressure by companies who pay these
304. 13:37agencies
305. 13:38uh to to rate their debt. These
306. 13:41companies pay for it. So my goodness,
307. 13:44>> are you talking some of the biggest
308. 13:47companies in the world? Are does your
309. 13:49concern extend to the hyperscalers and
310. 13:51big tech that have you know been tapping
311. 13:53debt markets or are these smaller ones
312. 13:55that um have higher risk in tapping the
313. 13:58debt market and less cash flow? Yeah, I
314. 14:01think uh the debt markets understand the
315. 14:03hyperscalers
316. 14:05although they have been through a shock
317. 14:07in that they've gone from massive free
318. 14:09cash flow not needing the debt markets
319. 14:12to uh to uh you know negative free cash
320. 14:16flow. So the bond market a analytics
321. 14:19have to have to get a lot better and I
322. 14:22think
323. 14:22>> does it change your outlook?
324. 14:25No, I think we learned from Coreweeef
325. 14:27actually. I think the what happened in
326. 14:29the bond market to Coree given
327. 14:31everything we've found out since then um
328. 14:35suggests or confirms that the bond
329. 14:37market uh didn't know what it was doing.
330. 14:39It has a little bit of egg on its face.
331. 14:42By the way, we as equity analysts and
332. 14:44and uh portfolio managers understood
333. 14:47about prepayments, understood
334. 14:49everything. We actually were saying,
335. 14:51"Wait a minute. uh maybe we shouldn't be
336. 14:55ignoring this credit default swap or
337. 14:58this yield that's going way up even
338. 15:00though the bond market's improving a
339. 15:02bit. Uh but uh but we so we took a
340. 15:06little bit of not a lot and I'm happy to
341. 15:08say not a lot a little bit of risk out
342. 15:11of our core position. uh but it it it
343. 15:15changed it marginally because of what
344. 15:18the bond market was signaling and now
345. 15:20we're not going to trust the bond market
346. 15:21signal the way we have in the past.
347. 15:23>> Okay. Um Kathy, finally, it's going to
348. 15:25be a busy week next week. Um looking at
349. 15:27the anthropic IPO, how are you thinking
350. 15:29about, you know, this massive private
351. 15:32company? I know you're you have holdings
352. 15:34um both private and public. How are you
353. 15:36thinking about it coming to market and
354. 15:38your sake?
355. 15:39>> Yes. So Anthropic we own in our venture
356. 15:44uh venture fund ArkVX
357. 15:47uh and it's been the best ride the fund
358. 15:50has had. We got in in early 23 right
359. 15:52after we launched the fund. So we've
360. 15:55been paying very close attention to it.
361. 15:57uh we were e even though we've done a
362. 16:00lot of research and had information
363. 16:02rights to see their gross margin in the
364. 16:05S1 up to 85%
365. 16:08and to learn that 75% of anthropic
366. 16:12spending is for training in AI which
367. 16:15does not generate revenues inference
368. 16:18generates the revenue 75%
369. 16:21um we our our confidence shot up
370. 16:23dramatically with that with the S1 and
371. 16:26you know we actually thought they they
372. 16:28would come to the market at a $2
373. 16:30trillion valuation. It looks like 1.4
374. 16:33four is closer to the mark and so we're
375. 16:39believe it or not we are uh thinking
376. 16:40it's it's a very reasonable valuation
377. 16:44>> you think right what do you think it's
378. 16:46valued at um maybe a year from now what
379. 16:48are you guys modeling
380. 16:50>> well we thought we thought that we could
381. 16:54justify almost $2 trillion uh uh so I'll
382. 16:59just leave it there you know it's been
383. 17:01such a force we do think open AI is
384. 17:04gathering some momentum here. So, we're
385. 17:06watching that very carefully as well.
386. 17:07But the demand for both of them uh both
387. 17:10of their uh their models is enormous.
388. 17:15>> Kathy, thanks so much for making the
389. 17:17time and it's great to see you.
390. 17:19>> Congratulations on Yahoo Finance, your
391. 17:21new show.
392. 17:22>> Thank you so much.
393. 17:23>> Thank you, Kathy.