Nvidia's Jensen Huang Scandal & Anthropic's IPO Deflates to $1.4 Trillion — Transcript

Full transcript

1. 0:00Welcome back to the Market Maker
2. 0:02podcast. And for two straight years, the
3. 0:05institutional playbook was pretty
4. 0:06simple. The S&P 500 was essentially
5. 0:09Nvidia in an oversized trench coat where
6. 0:13Jensen Wang went global liquidity would
7. 0:16follow until the day we're recording
8. 0:18this, that was yesterday. Last week,
9. 0:21Wall Street stared straight into the
10. 0:23macro abyss. The NASDAQ 100 took a
11. 0:25historic 6.4% 4% single day beating
12. 0:28wiping out I love this stat Pier $2.3
13. 0:31trillion dollars in market cap as
14. 0:34post-election jitters and data center
15. 0:36skeptimism triggered maximum panic but
16. 0:39yesterday brought a violent counter
17. 0:41punch the S&P 500 actually rallied 2.4%
18. 0:444% its best session since April and chip
19. 0:47indices jumped 7%. Yet in the middle of
20. 0:50all the green there was a glaring red
21. 0:53anomaly and that was Nvidia. Uh they
22. 0:56were down heavy at one point I think
23. 0:58like 8 9%. They fell 4.3 at the close.
24. 1:02That kind of divergence has happened
25. 1:04exactly just three times since 2001.
26. 1:08And I love this. I saw a journalist
27. 1:10saying this. The umbilical cord between
28. 1:14the market and its undisputed king has
29. 1:17been snapped. And the catalyst isn't a
30. 1:20missed earnings print. It's an espionage
31. 1:23scandal. An affair indeed. Admitted by
32. 1:27Jensen Huang, an FBI inquiry and an
33. 1:30Imperial CEO presiding over a board with
34. 1:32no chairman and no clear successor.
35. 1:36Meanwhile, on the software side, we've
36. 1:38had Dario and Ananthropic walking
37. 1:40directly into the storm. Their IPO road
38. 1:43show kicks off this week. Whispered
39. 1:45valuations have deflated from $2
40. 1:48trillion down to the latest we've heard
41. 1:51in the FT just an hour ago from
42. 1:53recording. It was 1.4. I heard some
43. 1:55rumors talking about 1.1 earlier in the
44. 1:58week. their defense obviously dropping a
45. 2:01new methos fable model doesn't hurt on
46. 2:04the eve of the tour uh to engineer the
47. 2:07optics and defend against open AI with
48. 2:09Goldman Sachs the lead left also pumping
49. 2:12some impressive revenue forecasts that
50. 2:15they think for the AI giant so is the AI
51. 2:18trade maturing or is the circular
52. 2:20financing machine surely firing showing
53. 2:23some cracks we're going to break down
54. 2:25the decoupling the boardroom crisis and
55. 2:27the trillion dollar road shows So,
56. 2:29appears caught your debate.
57. 2:31The uh the ball bear case.
58. 2:34>> Oh, yeah.
59. 2:35>> I mean, listeners shouldn't even need to
60. 2:37know who side I'm picking.
61. 2:40>> Uh I'm the I'm the perma hyperscaler.
62. 2:44So, there's no holding this back.
63. 2:46>> It was so Stephen sent me a message. I
64. 2:49think there was it's got to be the the
65. 2:51funniest I it's it's obviously I'm I'm
66. 2:54I'm going with the piss taking side of
67. 2:56things here, but there was a comment on
68. 2:58I think it was on YouTube and it was
69. 3:00like cuz yeah we we dropped the poll,
70. 3:02right? Who are you with me or Steven?
71. 3:04And they replied
72. 3:06um it's like the Messi Ronaldo debate,
73. 3:10you know, you can't you you can't choose
74. 3:13is was their uh was their
75. 3:16theme. I think were they talking about
76. 3:18our arguments or uh or or ourselves? I'm
77. 3:21sure
78. 3:22>> your stature maybe. But uh yeah, I'm the
79. 3:24Messi of those two. That's for sure.
80. 3:26>> I thought I thought it was a slam dunk.
81. 3:27I thought there's no Messi running out
82. 3:28of debate here. Messi is the goat,
83. 3:30right?
84. 3:31>> Yeah, but clearly [laughter]
85. 3:34>> going to get some comments on that.
86. 3:36[laughter]
87. 3:38Um but talking about goats or maybe
88. 3:41fallen goats, goats chopped out at the
89. 3:44knees. Yeah, Jensen's had a bit of a
90. 3:46shocker, hasn't he?
91. 3:47>> Well, he he has this week. You know, by
92. 3:49the time this hits people's ears two
93. 3:51weeks down the line, I'm sure it would
94. 3:53just be a forgotten blip.
95. 3:55>> What affair. Well, there is one there is
96. 3:59one way this could I mean, you know, I
97. 4:03Yeah, I think you're right. What you're
98. 4:05implying is this is just going to be
99. 4:06like yesterday's news pretty quickly and
100. 4:09and he's, you know, lines being thrown
101. 4:12out. too big to or too big to fire is is
102. 4:16the latest, isn't it? Which is
103. 4:17absolutely true. Unless
104. 4:21it does go
105. 4:24if the FBI who are investigating this,
106. 4:26of course, and there's obviously a few
107. 4:28different angles, but one of them is
108. 4:29that this this woman was a spy uh
109. 4:32working for the Chinese. If that turns
110. 4:35out to be true and it turns out he was I
111. 4:41would assume inadvertently
112. 4:43uh passing information then that's
113. 4:46that's that is a real problem. uh and
114. 4:50the problem being compounded by the
115. 4:54incredibly unusual structure of the
116. 4:57board and um he's the Nvidia, they're
117. 5:01the only
118. 5:03company in the S&P 100,
119. 5:06right? So that's the 100 biggest
120. 5:07companies by market cap in the US.
121. 5:09They're the only one that doesn't have a
122. 5:10chairman.
123. 5:12And then on top of that, yeah, you've
124. 5:14obviously got this situation where the
125. 5:17the directors on the board, well,
126. 5:18they've been on that board for like
127. 5:22since the '9s. Um, the average tenure, I
128. 5:25think if you cross go across the whole
129. 5:27board, the average tenure is 15 years.
130. 5:29That's double the usual average across
131. 5:33the rest of the S&P 500. So, it's a bit
132. 5:35unusual. and he's, you know, he he's
133. 5:37he's got quite a flat sort of um I would
134. 5:40say management structure and so he's got
135. 5:4260 people that directly reporting to
136. 5:45him. Um so he Yeah. In terms of that
137. 5:48business and operating that business, he
138. 5:50is too big to fire unless he gets
139. 5:53arrested.
140. 5:55>> Yeah. So yeah, what I mean one of the
141. 5:58ways I'd read it is that
142. 6:00>> you talk about keyman risk and I was
143. 6:02looking at how much of Nvidia does he
144. 6:04own because you would assume kind of
145. 6:06like Zuckerberg, Elon, others you go
146. 6:09right something super meaningful
147. 6:11>> and they have super voting shares and
148. 6:13all these sorts of things meaning that
149. 6:15they get diluted they still have overall
150. 6:17control. He's actually the opposite owns
151. 6:203.6% 6% stake worth a measily
152. 6:25$150 billion dollars. He's been short
153. 6:27changed there. Uh I would say
154. 6:30>> but I guess it's so that's one thing. So
155. 6:33he actually is even though he's he's
156. 6:35kind of outsized in his role in terms of
157. 6:37his personality but because he is the
158. 6:40pinup boy
159. 6:41>> of this whole thing. And third point of
160. 6:44terminology I really
161. 6:46>> kind of I quite enjoyed reading this
162. 6:48morning on the commute was
163. 6:50>> Nvidia has acted as a de facto central
164. 6:54bank of AI.
165. 6:55>> Yeah.
166. 6:55>> And I thought that was great in the fact
167. 6:58that it's directly financing customers
168. 7:00extending loans including Mega Speed uh
169. 7:03I think two years ago which is this
170. 7:05Chinese firm. They've taken equity
171. 7:07stakes in other firms. They're
172. 7:09backstopping data center leases. So that
173. 7:12whole thing about too big to fire, I
174. 7:15mean that came out of you've always got
175. 7:17to have context here. Who is saying
176. 7:19these things and it's Sequoa's Sha
177. 7:23Maguire who I'm sure you've got your
178. 7:25little table of investors in the early
179. 7:28days of a lot of these firms. I'm sure
180. 7:30Sequoa's got their, you know, a little
181. 7:33bit of capital capital on the table here
182. 7:35with some exposure.
183. 7:37So
184. 7:38is that the thing though? Is it because
185. 7:41you know the setup of the company could
186. 7:43go on and survive without him. However,
187. 7:47this isn't just an Nvidia story. Now,
188. 7:49Nvidia is the center of a universe that
189. 7:53is driving our economic universe,
190. 7:55>> right?
191. 7:55>> And so therefore, he's almost
192. 7:57irreplaceable from a technological
193. 7:59leadership standpoint.
194. 8:01>> Yeah.
195. 8:02>> Uh his capacity to navigate Washington
196. 8:04diplomacy.
197. 8:06I guess that's going to be the
198. 8:07interesting one to watch. And how close
199. 8:09is he now
200. 8:10>> to this stuff?
201. 8:12>> Yeah.
202. 8:12>> But Trump is still around, by the way.
203. 8:15>> And Trump wasn't exactly being super
204. 8:17skeptical of some of Jensen's
205. 8:20activities. Yeah. No comment. But they
206. 8:24are they are too big to fail. Like
207. 8:25forget Jensen for a second. They are too
208. 8:27big to fail the business. It is so
209. 8:31it's got its fingers in every single
210. 8:33little pie of this AI boom. It's the
211. 8:37absolute foundation of the whole thing.
212. 8:40And if you remove the foundation,
213. 8:43obviously the whole thing collapses. So
214. 8:45that and and that's not just the AI
215. 8:47boom, right? That that's the de facto
216. 8:49economic growth story right now. So it's
217. 8:52too big to fail. Um I think yeah, you
218. 8:55mentioned that Jensen's role in that
219. 8:58kind of being the sort of diplomatic.
220. 9:01It's almost like the go-between between
221. 9:03the Chinese and the US to be honest.
222. 9:05um you know he's got to be yanked out of
223. 9:08that sort of type of role immediately.
224. 9:11So this will have a negative impact you
225. 9:14know on obviously on the man himself but
226. 9:18I would say on the company I'm just
227. 9:19looking now terms of negative impact on
228. 9:22the company uh it's down 5 a half% today
229. 9:25when we started this podcast which was
230. 9:28about what 3 4 minutes ago it was down
231. 9:324%. So, it's now down five and a half.
232. 9:35>> They've heard that we've gone live, so
233. 9:37they quick.
234. 9:39>> There's I'll point it out now. I do love
235. 9:41a bit of technicals, but there is a
236. 9:43monster technical support level in
237. 9:46Nvidia's share price, which is the March
238. 9:4830th low. It's the low of 2026.
239. 9:51Um, and that was that's down at 165.
240. 9:55We're trading 169, just broken down
241. 9:57below 170. 165 is massive, huge level.
242. 10:03Um, so yeah, we've we've obviously had a
243. 10:05really really really steep reversal, a
244. 10:08second big step to the downside after
245. 10:10the one that we had obviously back in
246. 10:12the last week.
247. 10:14>> Well, look, let's talk about the context
248. 10:16of this week and last week.
249. 10:18>> Yeah,
250. 10:18>> because the NASDAQ 100 6.4% one day
251. 10:22fall.
252. 10:24I mean, that saw the volatility index,
253. 10:26the VIX to levels not seen since April
254. 10:292025.
255. 10:31think people have got a pretty good
256. 10:32sense check of that even though there
257. 10:34was we were there straight on it saying
258. 10:37people are overlooking the data center
259. 10:39you know legislation in certain key
260. 10:41concentrated areas geographically in the
261. 10:44US but now that's become the kind of
262. 10:46what's being pinned on as the kind of
263. 10:49catalyst if you like what about this
264. 10:52reversal and decoupling that we've had
265. 10:55you just mentioned Nvidia's down
266. 10:57>> they were down yesterday but the NASDAQ
267. 10:59had one of his best days in a long time.
268. 11:01>> Yeah,
269. 11:02>> that's not normally normal when an S&P
270. 11:04has uh index weighted components of
271. 11:08which Nvidia is one of the largest
272. 11:10>> companies in the world.
273. 11:11>> So the science or maths being if they
274. 11:15fall, everyone falls.
275. 11:16>> Yeah.
276. 11:17>> And that's been highly correlated, but
277. 11:18it wasn't the case and hasn't been in
278. 11:20the last few sessions. Why is that?
279. 11:22>> Yeah. It's it's you know what is the
280. 11:24driving factor of the move? Um and
281. 11:26actually if we switch to the S&P versus
282. 11:28Nvidia rather than the NASDAQ, um the
283. 11:31S&P rally 2%. Which was its best day,
284. 11:34one of its best days. What did you say
285. 11:35for anyway? For a long time, right? And
286. 11:37Nvidia dropped 4.3% on the same day. So
287. 11:40actually outside of May 2009, this
288. 11:45divergence has almost never happened
289. 11:48since Nvidia joined the S&P 500 back in
290. 11:502001. So you've you've had a 25 year run
291. 11:54where basically
292. 11:56the correlation between Nvidia and the
293. 11:59index is very strong and positive and
294. 12:02it's become ever stronger and more
295. 12:04positive as Nvidia has climbed up
296. 12:06through the rankings to become the
297. 12:08absolute top dog biggest component in
298. 12:10the whole index right um so this is a
299. 12:14fantastically unusual event because the
300. 12:17reason for Nvidia's downside is this
301. 12:20very specific story about its CEO and
302. 12:23whilst yes that might have implications
303. 12:25for right how's the business going to be
304. 12:28run um what's Jensen's role at a more
305. 12:31diplomatic level um you know is there
306. 12:34going to be any sort of consumer
307. 12:36backlash against Nvidia well we'll talk
308. 12:38about that maybe in a minute no how can
309. 12:40you backlash against Nvidia who what
310. 12:43consumers buying GPUs no one right so
311. 12:46but because of this very unique thing
312. 12:48about an individual person This isn't a
313. 12:52um this isn't a trade related to AI.
314. 12:55Well, actually the reversal of the whole
315. 12:57index was a rebound from the big step
316. 13:00lower that we had on Friday. So, you got
317. 13:02the whole market snapping and rebounding
318. 13:04back from the heavy selloff from Friday.
319. 13:07So the macro underlying theme is rebound
320. 13:11risk on and you've then got a singular
321. 13:14negative story that just so happens to
322. 13:16be very negatively impacting the biggest
323. 13:18company in the index which just for that
324. 13:21singular reason is just gone in exactly
325. 13:24the opposite direction. So some some
326. 13:26stats to support what you're describing
327. 13:28there. So Nvidia, well this has probably
328. 13:31changed but at last check just under 7%
329. 13:33of the S&P 500 and historically whenever
330. 13:36Nvidia sheds more than 4% I falls more
331. 13:41than 4% the S&P 500 declines by an
332. 13:43average of 1.6.
333. 13:45That's the normal relationship that it
334. 13:47would have.
335. 13:48>> But the divergence on Monday of this
336. 13:50week the S&P was up 2.35 whereas Nvidia
337. 13:54was down 4.3.
338. 13:56So is this where that terminology people
339. 13:59often talk about which is idiosyncratic
340. 14:02>> in this case governance right
341. 14:04>> themes yeah
342. 14:05>> being that there is an explainable
343. 14:07thing. So does this make you feel
344. 14:10given the context of the debate you had
345. 14:12with Steven and the
346. 14:15>> frothliness of the market at the moment
347. 14:18with anthropic about to list and all the
348. 14:20rest of it and the some of the market
349. 14:21pullback we've had. Is this actually a
350. 14:24little bit reassuring what's happened
351. 14:26with the decoupling?
352. 14:28>> Yes, that actually that's an interesting
353. 14:30spin on it which I hadn't quite thought
354. 14:33about. But you you're right. Um when you
355. 14:35got the biggest player of all having a
356. 14:38monster selloff and that can't stop the
357. 14:41broader market having a really solid
358. 14:43strong rebound then that is that is a
359. 14:46really good sign of the strength.
360. 14:49do I say the durability of the rebound
361. 14:53um obviously we'll see I mean the market
362. 14:54has been the whole AI trade
363. 14:58uh and I would say media
364. 15:01driven to a degree the media are so
365. 15:05anti- the AI boom it is ridiculous and
366. 15:10any any any any negative story it gets
367. 15:14ramped and pumped and it's all over the
368. 15:16headlines right so obviously this is
369. 15:19just another event where they're like,
370. 15:20"Oh my god, this is amazing for us." The
371. 15:22foder, you know, bang and they're all
372. 15:25over Jensen. Um, but the fact that
373. 15:27didn't cause another leg lower, I
374. 15:31think's a really interesting Yeah.
375. 15:33signal about where the market is at with
376. 15:35all of this.
377. 15:36>> Yeah, I did see an interesting stat. I
378. 15:38don't have it here, but I read it
379. 15:39earlier today and I think it was like a
380. 15:41macro
381. 15:43Goldman survey or snaps pulse survey
382. 15:46that they did and it was looking at
383. 15:48actually any of the holders of the index
384. 15:50weren't selling in that move. if it was
385. 15:52all the retail punters and so forth uh
386. 15:56getting out and it it's exactly to that
387. 15:59intro and I kind of had the sensational
388. 16:01intro a little bit on purpose
389. 16:03>> for me you know working in and around
390. 16:05the news for so many well two decades
391. 16:08now
392. 16:10it's funny it's like the invested
393. 16:12interest of the size of capital
394. 16:15and the influence behind the scenes
395. 16:17involved
396. 16:18it's a big thing to fail let's call it
397. 16:21whereas is what sells papers and
398. 16:24generates clicks. I tell you what does.
399. 16:26An espionage scandal, an affair by the
400. 16:29number one person who represents the
401. 16:32poster boy of AI, an FBI inquiry, an
402. 16:36Imperial CEO as they're calling him
403. 16:38because there is no chair.
404. 16:40>> Yeah.
405. 16:40>> Um and he has got no clue. It's just
406. 16:42it's almost like a script that's been
407. 16:44written for journalists to go to town
408. 16:46on.
409. 16:47>> It's it's a Netflix script.
410. 16:51I refuse to believe any other um reason.
411. 16:54But look, as I said, people will forget
412. 16:56it. Although, as I've as I have said,
413. 16:58the share price is down again sharply
414. 17:01today. But as long as he doesn't get
415. 17:04arrested,
416. 17:06and look, that that could happen. I have
417. 17:08got no idea, right? We'll see what the
418. 17:09FBI dig up. As long as he doesn't get
419. 17:12arrested and he's in big uh obviously
420. 17:15legal problems. As long as that doesn't
421. 17:18happen, I no doubt this will, you know,
422. 17:22be a storm in Jensen's teacup and he
423. 17:24will rise above it all. Um, I'm not
424. 17:27condoning his actions, by the way.
425. 17:28That's just life and the way this is
426. 17:30going to work. What will happen for the
427. 17:32long-term benefit of Nvidia. This, I'm
428. 17:35sure, will trigger um, you know, what's
429. 17:38going to go on the board meeting, you
430. 17:40know, agenda item number one at the next
431. 17:43board will be, well, hang on. This has
432. 17:45indicated to us that our keyman risk is
433. 17:47way too high. We need to start now a
434. 17:50process of reducing that key man risk.
435. 17:53So I think this can help Nvidia um over
436. 17:58the medium term. But yeah, as long as he
437. 18:00get doesn't get arrested, this is this
438. 18:03is
439. 18:04>> maybe Jensen's hedged himself by
440. 18:06becoming very close to Trump at this
441. 18:09point. I was just having a quick look
442. 18:11and legally according to article 2
443. 18:13section two of clause one of the US
444. 18:15constitution
445. 18:17might I add
446. 18:19uh he can be pardoned
447. 18:22>> well yeah but can he at any point I
448. 18:25think presidents have that thing when
449. 18:27they leave office
450. 18:29>> they do a whole load of pardons don't
451. 18:31they not for two years so can he just do
452. 18:34an off-the- cuff pardon at any moment
453. 18:37>> well it's I think it doesn't shield him
454. 18:40from other investigations,
455. 18:43>> state civil lawsuits, removal by his own
456. 18:45board, all that sort of stuff. But if
457. 18:47the FBI investigation to mega speed
458. 18:49allegations,
459. 18:51>> uh those export control violations they
460. 18:54had cuz you know they were saying they
461. 18:55were sneaking in the Nvidia chips, which
462. 18:58would be in violation of that.
463. 19:00>> Um oral espionage leads to potential
464. 19:03federal criminal exposure. Trump can
465. 19:06issue a full preemptive pardon.
466. 19:08Okay,
467. 19:11it's kind of cover. He's hedged himself
468. 19:12here. He's a clever guy. Trump Trump's
469. 19:15not Trump Trump will turn on a dime. If
470. 19:18if if there's any sniff that Jensen has
471. 19:21been involved in illegal, you know,
472. 19:24facilitating the legal smuggling of
473. 19:26Nvidia GPUs to China or if there's any
474. 19:29hint he's been involved in any kind of
475. 19:31espionage scenario, Trump will I don't
476. 19:35he won't he will turn on Jensen in a
477. 19:38heartbeat. But if Nvidia fails
478. 19:42>> Yeah. And the center, the plug gets
479. 19:44pulled from the AI bathtub.
480. 19:48>> Yeah.
481. 19:48>> History will remember Donald Trump as
482. 19:52the worst president that led to
483. 19:54catastrophic
484. 19:56>> loss. Donald's probably he's probably in
485. 19:58his bedroom at the moment just buying up
486. 20:00Nvidia stock here before he then
487. 20:02announces the uh the pardon.
488. 20:05Well, look, let's um let's move on and
489. 20:08let's talk about some of the other major
490. 20:09headlines here because the other
491. 20:10interesting one of course is whilst all
492. 20:12of this has been going on anthropic.
493. 20:15>> Yeah, not the best timing that one would
494. 20:17have hoped for. I mean, it feels like
495. 20:20they were probably 6 months too late. If
496. 20:23they just got this away,
497. 20:25>> maybe just ahead of SpaceX or just right
498. 20:27the two weeks after,
499. 20:29>> they could have absolutely hit top of
500. 20:31market. But it looks like compared to
501. 20:36the two trillion whispers actually the
502. 20:39stat line is that was 5 weeks ago.
503. 20:43>> It was early October.
504. 20:44>> Wow. Is that trillion still being about
505. 20:48>> right
506. 20:49>> now? It's gone down to 1.4 to 1.5
507. 20:52trillion and now some are saying uh
508. 20:55prospective ro show bidders is sat at
509. 20:571.1 to 1.3.
510. 21:00So yeah, there was an FT article out
511. 21:03earlier today. What was that giving us
512. 21:05in terms of some signals of how this
513. 21:07might go?
514. 21:08>> Well, so they're so the FT dropped it.
515. 21:10They're targeting 1.4 trillion is is the
516. 21:13latest number. Um
517. 21:18obviously, well, they're looking to
518. 21:20raise and this has been so this is a bit
519. 21:22more interesting info. They're looking
520. 21:23to raise now apparently 64 uh billion
521. 21:28which can I can I just pause like you
522. 21:30got to love Elon not only for the timing
523. 21:35of his IPO
524. 21:37absolutely
525. 21:40just the timing is genius right and will
526. 21:43he now hang on to his you know largest
527. 21:46ever IPO on record status
528. 21:49um and it looks like it he will right in
529. 21:52terms of both the valuation because I
530. 21:54think correct me if I'm wrong SpaceX was
531. 21:551.75 trillion valuation and he raised
532. 21:59$85 billion. So here we're now saying
533. 22:02anthropic 1.4 trillion valuation raising
534. 22:05$64 billion. So Elon will hang on to his
535. 22:09uh his records but um so what they're
536. 22:13looking to do is plan Anthropic are
537. 22:14looking to to to sell 825 million
538. 22:18shares. They've issued a price range
539. 22:22which actually remember SpaceX didn't
540. 22:25SpaceX didn't even have a range. They
541. 22:26said this is the price.
542. 22:29Take it or leave it. Uh which is very
543. 22:31unusual for an IPO but and and actually
544. 22:34yeah it' be interesting if if Anthropic
545. 22:36did go earlier a month or two back. I'm
546. 22:39sure they would have been in the driving
547. 22:41seat. They're the ones with the power
548. 22:42and maybe they would have gone with a
549. 22:44singular price point. but they've lost
550. 22:47the initiative and they're on the back
551. 22:49foot and the AI trade has absolutely
552. 22:52crumpled and so now they're going the
553. 22:54range between 70 and $78.
554. 22:57So if they do sell the full $825 million
555. 23:00at the top of that range which is $78
556. 23:02per share that's where you get your 64
557. 23:04billion raise from. So just talk to me
558. 23:08then strategically
559. 23:10why not delay it? Why not let the let
560. 23:15the ripples just kind of settle?
561. 23:18>> Is it that it could get worse?
562. 23:21>> You have to go now. It's going to look
563. 23:23bad if you pull out.
564. 23:26>> What am I missing there?
565. 23:28>> I think it's it could get worse
566. 23:32and
567. 23:33you know they need the cash. [laughter]
568. 23:35You know, in the end they they don't
569. 23:38have the luxury to go, "Oh, it's fine.
570. 23:39We'll just wait. Let's give it another
571. 23:4112 months.
572. 23:42That's not how this game works. These
573. 23:45data centers, these GPUs, they ain't
574. 23:48cheap, right? And their whole model is
575. 23:51to build out this infrastructure and and
576. 23:53the capital outlay is just ridiculous.
577. 23:57So, they don't have the luxury.
578. 23:58Although, I mean, what was surprising
579. 24:00when they dropped their S1 was that they
580. 24:03had I think it was 100 billion on their
581. 24:06balance sheet in cash. So, I mean, so
582. 24:09they do have some cash. They could delay
583. 24:11it, but yeah, I I just think what
584. 24:14happens if you delay it and then 6
585. 24:16months later we've got a big massive
586. 24:19crash and then they're like, "Ah, damn
587. 24:23it. Do we delay it again?" You know,
588. 24:24ultimately that runway of cash runs out.
589. 24:26So,
590. 24:28so I think I think they are right to
591. 24:31just plow on and take the hit. And look,
592. 24:34to be honest, it's not like they're
593. 24:35going to be selling much of the business
594. 24:37from a percentage perspective. So whilst
595. 24:40the valuation 1.4 trillion let's just
596. 24:42say it's that whilst that is a lot less
597. 24:45than two they don't I mean do they
598. 24:47really is the the valuation is not the
599. 24:50most important thing for anthropic right
600. 24:53it's just not it's just not the most
601. 24:56important thing so therefore I think
602. 24:59they're right to just crack on even
603. 25:01though they're going to be going in at a
604. 25:02bit of a discount to what we were
605. 25:04thinking a few weeks back. Yeah, there
606. 25:05was a strategist at HSBC chap called Max
607. 25:09Ker and he was talking about this
608. 25:13talking about sub one trillion
609. 25:16or a conservative uh pricing structure
610. 25:19might be a healthier route to prevent a
611. 25:22broken IPO or a SpaceX style dump in the
612. 25:25secondary market to avoid
613. 25:27>> as to your point. They're thinking about
614. 25:30where are we going to be in 2030 and
615. 25:32beyond.
616. 25:33>> Yeah.
617. 25:34>> Momentary, you know, they're still going
618. 25:35to raise tens of billions of dollars to
619. 25:37fund the build out, the acquisition of
620. 25:41the hardware, compose, whatever they
621. 25:42need.
622. 25:43>> Yeah.
623. 25:44>> Like why get obsessive? Elon's won that
624. 25:46race now.
625. 25:47>> Yeah.
626. 25:48>> Quite clearly. But and look like if you
627. 25:50take the Elon SpaceX example where they
628. 25:53did go at 1.75 trillion which was a kind
629. 25:56of crazy valuation their share price in
630. 25:59the secondary market since then has been
631. 26:01mega volatile like huge huge swings and
632. 26:05ultimately right now is trading about
633. 26:0815%
634. 26:09it's sorry it's underperformed the S&P
635. 26:12by 15% since its June IPO day. So, it's
636. 26:18massively underperformed and it is being
637. 26:20crazy volatile. Um, which is not a not
638. 26:25that Elon cares, but you know, for a lot
639. 26:28of businesses that would be really
640. 26:30difficult to deal with. Um, from a kind
641. 26:33of customer or sorry, shareholder
642. 26:34relationship perspective, right? Um,
643. 26:37Elon doesn't care. But, um, one thing
644. 26:40you mentioned there was does it go for
645. 26:42below 1 trillion? Now, there would be a
646. 26:44really big problem with that.
647. 26:47So I was looking back at the valuations
648. 26:50of Anthropic through its history, very
649. 26:54short history given it kind of was
650. 26:57founded in 2021.
651. 26:59But we do have certain valuation
652. 27:01timestamps and that's each time they've
653. 27:03raised capital in the private markets,
654. 27:06you know, directly from investors. Were
655. 27:08able to have a little mark to market at
656. 27:10that point. What was the valuation at
657. 27:12that raise? And you go back down the
658. 27:14list. I'll come on to my main point in a
659. 27:16second. There's some very interesting
660. 27:18people.
661. 27:18>> I know. I just saw the name by the way.
662. 27:20>> You are joking. Yeah. So, they did the
663. 27:22series A in May 21. The and the first
664. 27:25two raises there, series A and series B.
665. 27:28These haven't been entirely
666. 27:31uh verified in terms of the valuation.
667. 27:33So, the estimate series A in May 21 they
668. 27:37raised 124 million at a post money
669. 27:39valuation of 550 million. Um,
670. 27:44uh, Dustin Moskovich is the most famous
671. 27:47person people might have heard of that
672. 27:49was involved in that series B, April
673. 27:512022, they raised $580 million. Can you
674. 27:55believe that? Within the first 12 months
675. 27:58of their life as a business, they'd
676. 28:00already raised $700 million.
677. 28:03Um, anyway, that series B was at a 4
678. 28:06billion estimated valuation. So, it gone
679. 28:088x valuation in 11 months. Who was all
680. 28:12over that series B? It's your man Sam
681. 28:15Bankman Freed, baby. There he is. The
682. 28:18disgraced FTX founder who's now well got
683. 28:24his feet up in jail somewhere. Um, he
684. 28:26was all over the C the
685. 28:28>> What is he in line for then when this
686. 28:30thing goes public?
687. 28:32>> Well, 4 billion, right? if he'd if he'd
688. 28:34have just made it through,
689. 28:38God knows what the valuation of of his
690. 28:40holding is now, but anyway, um, series C
691. 28:43was May 23. I won't go through all of
692. 28:44these. I'm actually series C was
693. 28:46interesting because that was another
694. 28:48well May 23, so 13 months later, but
695. 28:51actually the valuation was flat for that
696. 28:5412-month period, April 22 through to May
697. 28:5723. Sam Bankmanf freed overpaid in 2022.
698. 29:01Um, but then anyway, you you go up the
699. 29:04list. By the time we get to March 25,
700. 29:06we're at a 61.5 billion valuation.
701. 29:10September 25, that was just like what, 6
702. 29:15months later, it tripled to 183 billion.
703. 29:20Um, 2 months after that, November 2025,
704. 29:23it's kind of 2 1/2x up again at 350
705. 29:27bill. Feb 26, it's at 380 billion. And
706. 29:31then the key point here May 2026 series
707. 29:34H round where SEOA were involved it
708. 29:38actually went at 965 billion valuation.
709. 29:42So there's there's nothing worse than
710. 29:44what we call a down round. That's where
711. 29:47you raise capital at a lower valuation
712. 29:50than a previous raise in the past. So
713. 29:53you know they are not going to want a
714. 29:55down round here. So it's definitely
715. 29:57going to need to be north of the 1
716. 29:59trillion. And look, they'll get it away
717. 30:00at north of a trillion. You know, if
718. 30:02they go for 1.4, people are going to buy
719. 30:04it. I know. I know the market has puked
720. 30:08here and a lot of corners of the market
721. 30:10are panicking, but look, you got your
722. 30:12mediumterm
723. 30:14hats on and and people are going to
724. 30:16hoover this up.
725. 30:17>> Two two things there. What I do like is
726. 30:19that it was the very partner of Seoia
727. 30:22who's saying, "Don't worry about Nvidia.
728. 30:25Don't worry about it. Not a bubble."
729. 30:28[laughter]
730. 30:29the same guy is absolutely
731. 30:32deep into invid into anthropic being a
732. 30:36success here getting in right on the
733. 30:38last round
734. 30:40>> uh stumping up some serious change. So
735. 30:43[laughter] this ain't no bubbles
736. 30:44according to uh to Goya which is quite
737. 30:47funny. The other thing is uh you
738. 30:49mentioned there about the only real
739. 30:51anomaly was the flat uh valuation
740. 30:53between April 2022 and May 23. Makes
741. 30:56sense because that was when Chachi PT
742. 30:58went public
743. 31:00>> in November of 2022.
744. 31:03>> Yes.
745. 31:03>> So probably they raised money I'm
746. 31:05assuming 6 months after Chachi PT had
747. 31:08gone wildly successful on it initial
748. 31:10launch.
749. 31:10>> Good point. That's exactly right.
750. 31:12>> And that would have been why I would
751. 31:13have thought.
752. 31:14>> Yeah.
753. 31:15>> Um
754. 31:15>> absolutely. Um yeah and then like other
755. 31:19stuff like Goldman's who are leading
756. 31:20this um lead left on this IPO deal. um
757. 31:25they've been trying to reassure markets
758. 31:28themselves. Obviously for Goldman's,
759. 31:30they're fully invested in getting this
760. 31:33IPO successfully away. You know, selling
761. 31:36all of these 825
762. 31:38million newly issued shares, selling all
763. 31:41of them at the highest price, you know,
764. 31:43within reason that they can. That's 70
765. 31:45to 78 price range. By the way, what
766. 31:48happens? Um so they're on the road show
767. 31:51now, right? So they're out there talking
768. 31:52to investors at the moment and that's
769. 31:54the price range 70 to 78. What they'll
770. 31:57be asked to do the investors by
771. 31:59Goldman's they'll be asked then to
772. 32:01submit indicative bids you know in
773. 32:04principle if you're interested you know
774. 32:06how much would you be interested how
775. 32:08many of the 825 million shares would you
776. 32:11be interested in buying and at what
777. 32:13price. This is then a book building
778. 32:15process which enables the investment
779. 32:17bank to gauge demand and therefore with
780. 32:20that demand data then set a singular IPO
781. 32:24price um normally between the range 70
782. 32:29to 78. That that's not fixed though. It
783. 32:32it could it could end up going above the
784. 32:34range or below the range depending on
785. 32:36how that demand gauge looks like in that
786. 32:39book building process. Right? But
787. 32:41Goldman's will want to be getting this
788. 32:43away at the top of the range, right? So
789. 32:44obviously they're talking their own
790. 32:46book. So they're saying ah offropic
791. 32:48they're expecting to make $600 billion
792. 32:50worth of revenue by annual revenue by
793. 32:532030 and that's up more than 130 times
794. 32:57from its sales and revenue last year. So
795. 33:00they're pumping, you know, really
796. 33:02bullish uh projections here on revenue
797. 33:05growth.
798. 33:05>> Yeah. And in combination with that, the
799. 33:07yeah, the spin doctors are working hard
800. 33:09on their stage crafting, dropping the I
801. 33:13saw the the Mythos and Fable models the
802. 33:16day before the road show begins. It's
803. 33:19like, oh, by the way, we've just
804. 33:20upgraded these uh humanity destroying
805. 33:24models even more because destroying
806. 33:28planet Earth wasn't enough. We're now
807. 33:30going bigger. So, by the way, we do have
808. 33:33this uh even though uh there's going to
809. 33:35be lots of things around it that we
810. 33:37won't have details on, but I guess
811. 33:39that's to serve your point. Uh the sales
812. 33:42desks, the syndicate bankers, it's this
813. 33:45kind of fresh product momentum defense
814. 33:48of the uh the valuation multiple against
815. 33:52some of these other concerns of tech
816. 33:55multiple fatigue and so forth. So, yeah,
817. 33:57they are going guns blazing. This is
818. 33:59what you pay Goldman's for though,
819. 34:00right?
820. 34:01>> That's right. But but unfortunately at
821. 34:04the end of the day, there are things
822. 34:05that are under your control and then
823. 34:08there are things that are not. And what
824. 34:10is under their control? Well, we can
825. 34:12pump out bullish revenue growth
826. 34:14expectations. We can, you know, set the
827. 34:17price range and like let's have an
828. 34:18amazing, you know, investor road show
829. 34:21and they can everything in their control
830. 34:23obviously they're experts at. What they
831. 34:26can't control is the macro market
832. 34:28sentiment. What they can't control is
833. 34:32Jensen, you know, sleeping with a spy.
834. 34:35You know, they can't control those
835. 34:36things. And unfortunately, from a timing
836. 34:38perspective, it's shocking,
837. 34:41but
838. 34:42I don't think they have a choice. I
839. 34:44think, well, as we know, they are
840. 34:46cracking on with the IPO, and I think
841. 34:47that is the right call. And maybe as a
842. 34:49segue then to conclude the other thing
843. 34:52they can't control
844. 34:54even though it's a little bit circular
845. 34:55in financing and so on and so forth but
846. 34:58they can't control competitors and the
847. 35:01only competitors in town that can really
848. 35:05change move the needle at all are the
849. 35:08hyperscalers.
850. 35:10>> Yeah. And one of those that I saw that's
851. 35:13also come out in the last day or two was
852. 35:15Alphabet talking about custom silicon
853. 35:18variants, these TPUs. And I know you've
854. 35:21talked about TPUs a number of times on
855. 35:24the show over probably the last 24
856. 35:27months.
857. 35:28>> Yeah.
858. 35:28>> Talk to me about that. What What is
859. 35:30Google's uh product here that they're
860. 35:33offering? This is Google is amazing cuz
861. 35:36they've got so many little
862. 35:39sort of diamonds that they're sat on and
863. 35:43I think they just don't it doesn't
864. 35:45really get much coverage.
865. 35:47Um whether or not you can argue it's not
866. 35:50priced into their share price. I'm not
867. 35:53I'm not too sure I I would go quite
868. 35:56there, but um they've got these things.
869. 35:59That's what you remember that Nvidia
870. 36:01their chips are called GPUs, right?
871. 36:03Well, Google have something called TPUs,
872. 36:07which is slightly different but
873. 36:08equivalent. Okay, so they they are a
874. 36:12straight up competitor to Nvidia at the
875. 36:14chip level. Um, we are in a world where
876. 36:18Nvidia GPUs are sold out. They're sold
877. 36:21out for till like 2029, right? So,
878. 36:25Alphabet are sat on this very
879. 36:27interesting product. Now, who's going to
880. 36:30get first dibs on Alphabet's TPUs?
881. 36:33Google obviously. So, this does does
882. 36:37give them an incredible sort of moat.
883. 36:40And I've spoken in the past about
884. 36:42Google's, you know, complete uh they
885. 36:44they've got the full vertical stack,
886. 36:46right? in that they've got the chips,
887. 36:49they're building data centers, um they
888. 36:52have the cloud, um which is growing gang
889. 36:56busters, and then they've got the
890. 36:57frontier models as well in Gemini,
891. 36:59right? They they're the only ones at the
892. 37:00moment that are occupying all layers of
893. 37:03the stack. Now, that'll change cuz all
894. 37:06the other big hyperscalers are busy
895. 37:09building products in all layers of the
896. 37:10stack, right? And who knows, in like 10
897. 37:12years time, you're probably just going
898. 37:14to have four or five companies that just
899. 37:17do everything. Like Nvidia, they're
900. 37:19going the other way, right? Yes, they've
901. 37:20got the chips, but they're building
902. 37:22buying hugging face and, you know,
903. 37:24building models themselves now, right?
904. 37:26But but anyway, obviously, you know, for
905. 37:30every um for every player in the game
906. 37:35that stumbles, there's a competitor that
907. 37:38benefits. And so whilst Nvidia and
908. 37:42Jensen, you know, do their thing, well,
909. 37:44Alphabet are nicely placed of course
910. 37:46from a TPU perspective. They've just
911. 37:48launched a couple of new products on
912. 37:49that front. TPU8TX,
913. 37:52that's for like confidential computing.
914. 37:54Um, they paired that up with AMD's CPUs.
915. 37:57And then you got something called TPU 8i
916. 38:00max uh which is a dual chip module
917. 38:03optimized for long context inference uh
918. 38:07things like maths and code generation
919. 38:09and stuff like that. Right. So um
920. 38:12Bloomberg intelligence projects external
921. 38:15TPU hardware software revenues are
922. 38:18basically Bloomberg are saying that that
923. 38:20they're going to cross the $135 billion
924. 38:23mark by 2028. So that's Google selling
925. 38:25chips. Um, you've got companies like
926. 38:29Jump Trading, um, which I'm sure many of
927. 38:32our listeners will be familiar with,
928. 38:34client of ours. Shout out to Jump. Um,
929. 38:37but they've disclosed deploying
930. 38:38thousands of these TPUs. And actually,
931. 38:41Jump is saying that they're about 30%
932. 38:44lower cost versus Nvidia. So, Google
933. 38:47have been placing themselves in the
934. 38:48market at a cheaper price point. But
935. 38:50look, whilst all that sounds good and
936. 38:52wow, Google are going to come in and
937. 38:54soak up all the market share, there's no
938. 38:57market share. It's all sold out. You
939. 39:00know what I mean? So, in the in the
940. 39:03short to medium term, none of this
941. 39:06matters for Nvidia in terms of them
942. 39:09selling GPUs unless Jensen gets
943. 39:13arrested.
944. 39:15Otherwise, fine, Google will crack on.
945. 39:18Um, but are they going to properly
946. 39:19benefit from this Nvidia wobble?
947. 39:23I I don't think you can necessarily
948. 39:25connect those two dots. Yeah, I think
949. 39:27the
950. 39:29looking at politics as a precedence.
951. 39:33It's the the it's the big thing until
952. 39:36the next big thing comes along and
953. 39:38Jensen is the big thing of this week.
954. 39:40And just like politicians always do and
955. 39:43I think co the postco environment which
956. 39:45is very lumpy for many politicians
957. 39:48around the globe.
958. 39:49>> I think the strategy that yielded
959. 39:50probably the best results was just just
960. 39:52to stick it out
961. 39:55and in time the next concern arises and
962. 39:59the previous one is forgotten. And one
963. 40:00of the things I was trying to sort of
964. 40:02put together is in my head there's
965. 40:04always this analogy I have of cops and
966. 40:06robbers
967. 40:08and the problem that you have when it
968. 40:09comes to say in this case what I would
969. 40:11say is like you know these frontier labs
970. 40:14are like the robbers in the sense of
971. 40:16their ability to move at the speed that
972. 40:18they can
973. 40:20>> and to be at the cutting edge of
974. 40:22technology and change
975. 40:24>> and adaptation
976. 40:26>> to make you know to take advantage of
977. 40:28opportunity Whereas regulation moves at
978. 40:32the speed of congressional committees,
979. 40:34>> right?
980. 40:35>> Which is at a snail's pace at best, of
981. 40:39which most of it never comes to fruition
982. 40:41anyway
983. 40:42>> because of the composition changes just
984. 40:45like we've had in the midterms where
985. 40:47things just get stopped, started,
986. 40:49changed, reinvented, and by the time
987. 40:51it's come,
988. 40:54the robbers are already, you know, round
989. 40:55the c round down in the next town. Yeah,
990. 40:58[laughter]
991. 40:58absolutely. And and anyway, thank God
992. 41:02Jens is thinking well, thank God
993. 41:04Goldman's are still advising Anthropic
994. 41:06to go ahead with their IPO because
995. 41:08that'll be the next big story in a
996. 41:09couple of weeks time. That is very very
997. 41:12true, actually.
998. 41:14>> Yeah.
999. 41:14>> And then how long have we got to go? I
1000. 41:16mean, it's thanks just before
1001. 41:17Thanksgiving, isn't it? So,
1002. 41:19>> yeah,
1003. 41:19>> he's just got he's got to hunker down.
1004. 41:22>> Yeah. Less than two weeks.
1005. 41:23>> See it out. And look,
1006. 41:24>> just get get through. By no way am I
1007. 41:26saying what Jensen did was right.
1008. 41:29>> Yeah.
1009. 41:30>> Uh his actual uh thing that he did. By
1010. 41:35no means am I condoning. But my point
1011. 41:36being that yeah, let's just see how this
1012. 41:38plays out and washes out. Is he indeed
1013. 41:41too big to be fired. Let us know what
1014. 41:42you think. Drop us a comment. Uh both
1015. 41:46that and also uh Ronaldo or Messi who is
1016. 41:50the uh greatest of all time. Let us know
1017. 41:52in the comments. Uh but cool. Thank you
1018. 41:54very much, Pier.
1019. 41:56>> See you next week.
1020. 41:57>> Pleasure. See you later.