'Big Short' Investor Explains How the AI Bubble Will Burst — Transcript

Full transcript

1. 0:00Let's just imagine that open AI fails.
2. 0:03Could happen.
3. 0:04>> The host of the real Eisman playbook
4. 0:06podcast.
5. 0:06>> I don't know about you, but I know that
6. 0:07I don't have a hundred billion dollars
7. 0:09to spend on building data centers.
8. 0:11[music]
9. 0:12>> You may know our next guest from the Big
10. 0:13Short.
11. 0:14>> Your character.
12. 0:15>> You had to work that in there, didn't
13. 0:16you?
14. 0:17>> I did have to.
15. 0:18>> Do you like being described [music] that
16. 0:19way? I think it's going to be on my
17. 0:20tombstone.
18. 0:21>> The whole United States of America would
19. 0:22go into a recession overnight. Oh,
20. 0:24yikes. Okay,
21. 0:25>> Mr. Anders.
22. 0:26>> Well, I'd have said you're out of your
23. 0:27mind.
24. 0:28>> Yeah, you're insane. to get my
25. 0:30programming to impersonate a DT.
26. 0:31>> This industry, despite all the hundreds
27. 0:33of billions of dollars [music] that's
28. 0:34been spent on it, has
29. 0:37>> join us right now is Steve Eisman.
30. 0:38>> Let's bring in Steve Eisman.
31. 0:41>> Steve, welcome back. You are known for
32. 0:45spotting a bubble before anyone else
33. 0:47does. Michael Lewis wrote a whole book
34. 0:50on it. So, my first question to you is,
35. 0:53where on earth are we in AI right now?
36. 0:57There was a great movie with uh Sean
37. 1:00Conre where he played this I can't
38. 1:01remember the name of Finding Forester.
39. 1:03And I remember the young character asks
40. 1:05him a very complicated question and his
41. 1:08response is as he's eating soup he goes,
42. 1:11"It's not exactly a soup question. It's
43. 1:14a complicated question." [laughter] I
44. 1:16never forgot that line. I thought it was
45. 1:17one of the best lines in in the history
46. 1:18of movies. It's not a soup question.
47. 1:21>> Not exactly a soup question, is it?
48. 1:23>> This is how I look at it. The
49. 1:24concentration
50. 1:26risks here are all inspiring, you know.
51. 1:29So, you take a step back and you and
52. 1:31someone said to me, why don't you
53. 1:33analyze this software company? Forget
54. 1:35about what it does. It's a software
55. 1:36company, an established software
56. 1:38company. And if it turned out that the
57. 1:40company had thousands of customers, that
58. 1:44would be great. If it turned out the
59. 1:46company only had two customers, you'd
60. 1:48say, "I don't want to invest in that
61. 1:49because if something bad happens to one
62. 1:51of those customers, this company is is
63. 1:54dead." There's something of that going
64. 1:57on in the whole AI story. So, let's
65. 1:59start just with Nvidia. So, Nvidia, God
66. 2:02bless them, and I own the stock, okay?
67. 2:05When they reported a few weeks ago, I
68. 2:07think the revenue growth was like it was
69. 2:09like 110%.
70. 2:10>> It's a lot. So, so let's just let's just
71. 2:12take a step back just for a second and
72. 2:14say to ourselves, wait a minute. The
73. 2:17largest company on planet Earth just had
74. 2:19forget about earnings growth, which was
75. 2:21great, too. Just revenue revenue growth
76. 2:24of over 100%. Like, that's insane. So,
77. 2:27that would say the AI story is great
78. 2:30until you read the 10 Q, which came out
79. 2:33that night. And I'm going to impress
80. 2:34your viewers by saying if they look at
81. 2:36it and they go to Note 7.
82. 2:38>> Oh, okay.
83. 2:39>> Okay. Yeah. Note 7 says that 70% of
84. 2:44Nvidia's accounts receivable as of the
85. 2:46end of July came from five customers.
86. 2:50>> Warning flag. Not the end of the world.
87. 2:53Warning flag. Now, let's go to the
88. 2:55hyperscalers. You're talking about
89. 2:57Google, Meta, Amazon, Microsoft, and
90. 3:01let's throw Oracle in for good measure.
91. 3:03massive companies spending massive
92. 3:06amounts of money by buying Nvidia's
93. 3:08chips and everything else under the sun.
94. 3:1070%
95. 3:12of their AI revenue which equals
96. 3:17something like 25 to 30% of their entire
97. 3:21cloud revenue
98. 3:23comes solely from anthropic and open AI.
99. 3:26>> Right?
100. 3:26>> Let me just let me just say it again so
101. 3:28your viewers get it. If you look at
102. 3:30leave out Oracle for a second,
103. 3:32Microsoft, Amazon, [snorts]
104. 3:34Google, 70% of their AI revenue, which
105. 3:38is equivalent to 25 to 30% of their
106. 3:41total cloud revenue is just from
107. 3:44anthropic and open AI. If you go to
108. 3:46Oracle, Oracle puts out a data point
109. 3:50called RPO,
110. 3:53which is basically a form of backlog.
111. 3:56when they reported earnings last year in
112. 4:01October for their August quarter, their
113. 4:04RPO went from like 150 billion to like
114. 4:08400 billion
115. 4:09>> in 3 months.
116. 4:10>> It was a massive jump.
117. 4:11>> It was massive jump. I mean, people went
118. 4:12insane. Yes.
119. 4:13>> And the stock went crazy.
120. 4:15>> The stock went crazy. It went from 230
121. 4:18>> to 330 like in two days.
122. 4:20>> Yeah. And then some of the cellite
123. 4:23analysts who are very good who did some
124. 4:24digging came out with reports that said
125. 4:2750% of that RPO is just from open AI.
126. 4:32>> Now now today
127. 4:34Oracle is is over 600 billion and it's
128. 4:37still like 50% is from open AI.
129. 4:40Basically 50% of future revenue of
130. 4:43Oracle is from a company that loses
131. 4:46money like crazy.
132. 4:47>> Well that's what I was going to ask you.
133. 4:49Maybe you can explain how this works.
134. 4:51But from my perspective, I don't
135. 4:53understand where this money is coming
136. 4:55from.
137. 4:55>> We're coming to that. Okay. Let me let
138. 4:57me just finish the let me just finish
139. 4:58the chain. Yeah.
140. 4:59>> So now we come all the way now to Oric
141. 5:01to Anthropic and Open AI. And my view is
142. 5:05the entire chain from Nvidia to the
143. 5:07hyperscalers all the way down the whole
144. 5:10chain rests on the future health and
145. 5:13growth of enthropic and open AI because
146. 5:15they're creating the commitments to the
147. 5:16hyperscalers. If they don't grow and
148. 5:18have the money to pay for those
149. 5:20commitments, well then the whole chain
150. 5:22slows.
151. 5:23>> Yeah.
152. 5:24>> So that's the risk. I would say between
153. 5:27the two probably Open AAI is the weaker
154. 5:29entity. But it's not clear because
155. 5:31really we really don't have enough
156. 5:32numbers. The only thing we do know
157. 5:34because the Wall Street Journal reported
158. 5:35this, so I'm assuming it's true. Open AI
159. 5:38had I something like six and a half
160. 5:42billion
161. 5:44in revenue
162. 5:46in the second quarter of this year.
163. 5:48>> Okay.
164. 5:49>> Anthropic was at 11 plus
165. 5:52>> but this is just revenue
166. 5:53>> just revenue.
167. 5:54>> Okay.
168. 5:55>> Open AI lost something had cost of
169. 5:58something like 12 billion. So the way
170. 6:00the math worked was in three months Open
171. 6:02AI Open AI's revenue went up a billion
172. 6:05and its cost went up three billion which
173. 6:08which we like to say is upside down.
174. 6:11>> You want the reverse not the former they
175. 6:14want the other way to go.
176. 6:15>> Great business model.
177. 6:16>> And and their revenue grew 18% in 3
178. 6:20months whereas Anthropic's revenue grew
179. 6:23over 100% in three months. So they're
180. 6:25the weaker company at this point. That
181. 6:27could that could change. My my only
182. 6:29point is that this whole this whole
183. 6:31industry makes me nervous because let's
184. 6:33let's just imagine that open AI fails
185. 6:37>> could happen.
186. 6:38>> The whole like the you know the whole
187. 6:40United States of America would go into a
188. 6:42recession overnight if this would and
189. 6:44now eventually there'll be a lot more
190. 6:46diversification and there'll be a lot
191. 6:48more companies but that's going to take
192. 6:51time. So you know within the next year
193. 6:53or so those companies have got to stay
194. 6:55healthy. That's where I think we are. So
195. 6:57there's a concentration risk.
196. 6:58>> There's a massive concentration risk.
197. 7:00>> And that brings me to the to the
198. 7:01question I I was asking before. My
199. 7:04understanding is that well we just spoke
200. 7:06about these uh open AI and anthropic are
201. 7:09not making money. They're losing money.
202. 7:12>> Excuse me. To say that they're losing
203. 7:14money would be kind. They bleed money.
204. 7:18>> What's the nicest way?
205. 7:19>> They lose a lot of money.
206. 7:20>> We're pre Yeah, we're pre-p profofit,
207. 7:22right? Prepit. [laughter]
208. 7:24Something like that. They they make
209. 7:25money if you exclude all costs. Yeah,
210. 7:27exactly. That's how they like to think
211. 7:28about it.
212. 7:28>> That's that's we should start reporting
213. 7:30that metric.
214. 7:30>> Yes, I we [laughter] should.
215. 7:32>> Um so where where does the money come
216. 7:35from?
217. 7:35>> Well, that's actually a very interesting
218. 7:36question. I had thought that most of the
219. 7:40company most of the money was coming
220. 7:41from venture capital and that h happens
221. 7:44to be not true. Most of the company is
222. 7:46coming from Amazon, Google, Microsoft
223. 7:50and Nvidia investing in these companies
224. 7:53and SoftBank.
225. 7:54>> Okay. So, taking equity stakes in
226. 7:57>> taking equity stakes. They raise capital
227. 7:59and those guys have ponyed up money.
228. 8:01Whether they want to continue to pony up
229. 8:03money, I don't know.
230. 8:04>> That sounds like a circle. It sounds
231. 8:06like money or commitments are going one
232. 8:08way and then commitments are coming back
233. 8:10the other way.
234. 8:10>> Yeah, it does have that tendency, does
235. 8:12it? This kind of reminds me of the It's
236. 8:14actually a scene from the big short uh
237. 8:16with your character.
238. 8:18>> You had to work that in there, didn't
239. 8:19you?
240. 8:19>> I I did have to work I remember you
241. 8:22saying, "Oh, it's kind of like CDOA and
242. 8:24then they put parts of that into CDOB
243. 8:26and those two put in CDOC."
244. 8:29>> There's some similarities obviously, but
245. 8:31in their defense, there is a circularity
246. 8:33to the financing. But as long as
247. 8:35anthropic and open AI keep growing very
248. 8:38very rapidly and people keep giving them
249. 8:40money, the chain will hold. It's if one
250. 8:43of those two companies really messes up
251. 8:46and pe people pull money or don't want
252. 8:49to invest it anymore that that's when
253. 8:50the chain doesn't hold. That that's
254. 8:52where the concentration risk problem
255. 8:53comes in. If the industry was much if if
256. 8:56this if if if I had said instead of 70%
257. 9:01of hyperscaler
258. 9:04AI revenue comes from anthropic and open
259. 9:06AI if that number had been 10%.
260. 9:10We'd be having an entirely different
261. 9:12conversation because there clearly there
262. 9:14are a lot more customers out there of
263. 9:16size. M. So that brings me to another
264. 9:19argument that I've heard you make on
265. 9:21your podcast and with your guests is
266. 9:23that if some
267. 9:24>> Let's just plug that podcast for a
268. 9:26second and call it the real Eisman.
269. 9:27>> Real Eman playbook. Yeah, absolutely.
270. 9:29It's very very good and it's gaining a
271. 9:32lot of traction. I think it comes down
272. 9:33to the authenticity of it.
273. 9:35>> Well, I appreciate that. I really enjoy
274. 9:36it. I think it's very very high.
275. 9:38>> My wife and I work on it
276. 9:40>> every day.
277. 9:41>> Both of you guys? I didn't realize. So
278. 9:43my wife is my partner in this and um so
279. 9:46she does god bless her all the editing.
280. 9:50>> Oh really?
281. 9:50>> This gave you an insight into into
282. 9:52things. So I I am a very linear thinker
283. 9:56>> which which the way I would define it is
284. 9:58one two three four five conclusion.
285. 10:02>> Got it?
286. 10:03>> And too often I write that way. So, I'll
287. 10:06write I'll do we have this thing called
288. 10:08the weekly rap which we put out Friday
289. 10:10where I sum up the week
290. 10:12>> and too often when I write it I'll I'll
291. 10:14I'll do one two three four five six and
292. 10:18and Valerie my wife who's who's the
293. 10:21editor will always say you buried the
294. 10:23lead again. Yeah.
295. 10:24>> And she'll flip it.
296. 10:25>> Ah okay.
297. 10:26>> And so because she says you know most
298. 10:29people never get to the bottom. you
299. 10:30know, people get your conclusions at the
300. 10:32bottom and it takes it takes 10 minutes
301. 10:34to get to the bottom. So, she so she
302. 10:37edits it and she runs the business,
303. 10:39>> right? Okay. That's cool. I didn't know
304. 10:41that. That's that's a little behind the
305. 10:42scenes. I like that.
306. 10:43>> Okay. Back back to back to AI train of
307. 10:46thought before I forget my train of
308. 10:47thought again. So going on from what
309. 10:50you're saying, if something happens to
310. 10:52OpenAI or Anthropic,
311. 10:55that's where the problems could be at
312. 10:57this moment in time when we've got the
313. 10:59concentration risk.
314. 11:00>> How do you see things like
315. 11:04Deep Seek's new model or Kimmy, what is
316. 11:07it called now? Kimmy,
317. 11:08>> best name in the biz.
318. 11:10>> Kimmy K3.
319. 11:11>> I like that. It just rolls up the
320. 11:12>> So this is where the the industry has, I
321. 11:15think, real weak business weakness.
322. 11:17Yeah, not not con let leave aside
323. 11:20concentration risk. That's its own risk.
324. 11:23Here's the business risk. There was
325. 11:26something going on for a while which is
326. 11:29called token maxing
327. 11:31which is where for lack of a better term
328. 11:34I work for a company [clears throat] and
329. 11:36I'm the I'm I'm a software engineer and
330. 11:39I've been told by management you're to
331. 11:41use AI 247 whether you need it or not.
332. 11:46>> Just do it.
333. 11:46>> Just do it. Right.
334. 11:47>> What happened was last year Open AI and
335. 11:52Anthropic were dramatically
336. 11:53undercharging for for their services.
337. 11:56Then they raised prices because they
338. 11:58they were so undercharging for the cost
339. 12:00of tokens. It was killing them,
340. 12:01>> right?
341. 12:02>> So they increased the prices that so
342. 12:03that the the customer was bearing more
343. 12:05of the cost of the token.
344. 12:07>> So you get people in once you got people
345. 12:08in raise the price.
346. 12:09>> Raise the price. Uber I think went blew
347. 12:12through its entire AI budget in like
348. 12:14three or four months.
349. 12:15>> Oh yikes. Okay.
350. 12:17>> Okay. And that this was some some other
351. 12:19company that I read about spent $500
352. 12:21million before they even knew they had
353. 12:23spent $500 million. So what's happened
354. 12:26is people have gotten a lot more
355. 12:28costconscious. Token maxing has gone
356. 12:31away and people are using these
357. 12:34openweight models a lot more. You don't
358. 12:37need a Cadillac for everything, you
359. 12:40know. So people will use um anthropic
360. 12:42and open AI as models only for the super
361. 12:46important tasks. Everything else they'll
362. 12:48use Kimmy K3 or or whatever. So what I
363. 12:51like to say about this is that this
364. 12:53industry despite all the hundreds of
365. 12:55billions of dollars that's been spent on
366. 12:56it has no moes.
367. 12:59>> There are no moes. This is you know
368. 13:00Google with its search until very very
369. 13:03recently that was a moat. I mean,
370. 13:06everybody used Google like I mean
371. 13:08>> 90%
372. 13:09>> 90% of planet Earth used Google and no
373. 13:11one even think about it
374. 13:12>> because because it was just better and
375. 13:15nobody could could approach it
376. 13:17>> you know here one day it's Gemini one
377. 13:20day it's chat GPT another day it's
378. 13:22Claude they they just rotate
379. 13:26>> and so well we'll come to it about my
380. 13:29conspiracy theory about the end of the
381. 13:31world
382. 13:31>> conspiracy theory all right so this this
383. 13:34dovetales into my conspiracy Okay. Okay.
384. 13:36>> So, as every as all your viewers know,
385. 13:38the world's going to end
386. 13:40>> some point. It has to
387. 13:40>> at some point it has to maybe it could
388. 13:42be five billion years from now or in a
389. 13:45couple of weeks.
390. 13:46>> Yeah. CNBC and Wall Street.
391. 13:48>> I I think this entire
392. 13:51AI is going to end the world is a
393. 13:54complete subtrauge.
394. 13:55>> Okay.
395. 13:56>> And what I think is really going on is
396. 13:59that these companies are nervous.
397. 14:02They're nervous that token maxing has
398. 14:04ended. They're nervous that there are no
399. 14:06moes. They're nervous that these they
400. 14:08these Chinese openweight models are
401. 14:10taking massive market share. So they're
402. 14:12manufacturing a hysteria which what
403. 14:16they're hoping for is for the government
404. 14:18to come in and regulate the industry and
405. 14:20they think that by manip they could
406. 14:23manipulate that regulation
407. 14:26to create a duopoly
408. 14:28>> so that will that the regulation will
409. 14:30create the moes. the regulation will say
410. 14:32no Chinese AI models. There's too big a
411. 14:34risk.
412. 14:35>> I see.
413. 14:35>> And then all all of a sudden there's a
414. 14:37moat. Yeah.
415. 14:38>> That didn't exist before
416. 14:40>> a legal barrier.
417. 14:41>> That's what I think is is is actually
418. 14:43happening here. So you think that's the
419. 14:44reason behind I cuz I've noticed Elon
420. 14:47he's always talking about it but
421. 14:49recently the uh anthropic CEO Dario
422. 14:53>> and they're also talking about slowing
423. 14:54down
424. 14:55>> slow and that that cannot be that cannot
425. 14:57be taken seriously because if you really
426. 15:00really really if you if I was Dario Modi
427. 15:03>> and I really really really thought
428. 15:06>> that my product is dangerous and I
429. 15:11really need to slow
430. 15:14I'd postpone my IPO.
431. 15:16>> You'd have to postpone your IPO.
432. 15:18>> You could just do take the steps.
433. 15:20>> Take a step back and and you know, I'll
434. 15:22fix fix what we need to fix and we'll
435. 15:24come back. Are they postponing their
436. 15:25IPO? No. You know, Elon Musk had a very
437. 15:28funny quote the other day where I think
438. 15:31he did on X where he said something like
439. 15:33um I'm going to paraphrase. I don't have
440. 15:35the exact This is some messed up 4D
441. 15:39chess where you're saying that the that
442. 15:42your product's going to end. Oh, and by
443. 15:44the way, how much can I allocate to you
444. 15:45for the IPO? [laughter]
445. 15:48>> Yeah.
446. 15:49>> You know, seriously, I I take people
447. 15:51seriously when they put their their
448. 15:53money at risk,
449. 15:54>> right?
450. 15:54>> You know, this this statement about a
451. 15:56slowdown is is just all part of this
452. 15:58hysteria that they're trying to
453. 15:59manufacture,
454. 16:00>> right? So they're aware that there may
455. 16:02not be moes and it kind of for for those
456. 16:04that don't know explain what you what
457. 16:07you mean we're talking moes specifically
458. 16:09in uh not hypers scale in the LLM
459. 16:12providers because the there I guess
460. 16:14there's still
461. 16:15>> the hyperscalers have moes
462. 16:16>> okay hyperscaler so so what's a moat
463. 16:20>> your grocery store doesn't have a moat
464. 16:22>> because somebody could open up a grocery
465. 16:24store across the street tomorrow
466. 16:27>> but there are some businesses that have
467. 16:31real moes real moes around them
468. 16:34>> like a competitive advantage.
469. 16:36>> It's but it's a competitive advantage
470. 16:38that is eternal or or at least very long
471. 16:41lasting like Nvidia makes GPUs.
472. 16:44>> Mhm.
473. 16:45>> Well, nobody else really makes GPUs.
474. 16:48>> That's a moat. certain software
475. 16:50companies have, you know, when if you're
476. 16:54Salesforce or Service Now, which are two
477. 16:56massive software companies, you your
478. 16:58product is embedded in the companies
479. 17:01that you service. Like like
480. 17:03>> those companies that use your your
481. 17:05product, [snorts] they've used it for so
482. 17:07long, they can't function without your
483. 17:09product.
484. 17:10>> Hard to switch away.
485. 17:11>> You try and get to try and switch out of
486. 17:13that is brutal.
487. 17:14>> Yeah,
488. 17:15>> that's a moat. There's no hyperscaler.
489. 17:18What's the moat? Well, I don't know
490. 17:20about you, but I know that I don't have
491. 17:21hundred billion dollars to spend on
492. 17:24building data centers.
493. 17:25>> Okay?
494. 17:26>> They just don't have it.
495. 17:26>> So, there is some moist.
496. 17:28>> So, that so just in terms of size and
497. 17:30money, I mean, there only certain
498. 17:33companies that can actually build data
499. 17:35centers.
500. 17:36>> They're just that expensive. LLMs, the
501. 17:39the creation of the models is expensive,
502. 17:41but there's so much competition
503. 17:44and there's no loyalty. Like, you know,
504. 17:47if you're a if you're a software
505. 17:49developer and you're using Claude, if
506. 17:51tomorrow another LLM shows up that's
507. 17:54better than Claude, you'll switch.
508. 17:56>> Y
509. 17:56>> you're you're switch there's no there's
510. 17:58you're not stuck.
511. 17:59>> That's the problem with the LLM model.
512. 18:02>> Gotcha. So, there's no moes in LLM.
513. 18:04>> Yes. hyperscalers they have I guess they
514. 18:07have diversified business models so
515. 18:08there's
516. 18:08>> and they have modes but they're
517. 18:10dependent upon the LLMs in their cloud
518. 18:12businesses that's their weakness right
519. 18:14now
520. 18:15>> so okay so there's kind of a argument
521. 18:17for hyperscalers for and against having
522. 18:19a mode you kind of in some ways there's
523. 18:21the scale and the cost and the barrier
524. 18:24to entry mode of of the investment and
525. 18:26then there's also the dependency on the
526. 18:28people that are buying
527. 18:30>> that have no modes
528. 18:30>> yes
529. 18:31>> that's the problem
530. 18:33>> interesting Okay.
531. 18:34>> And the other and the other problem with
532. 18:36the hyperscalers this may be temporary
533. 18:39but then again maybe it's not is you
534. 18:42know Microsoft Amazon Google 3 four
535. 18:46years ago and way prior to that these
536. 18:50companies were incredibly profitable but
537. 18:53even more importantly they just threw
538. 18:55off cash like crazy. I mean so much cash
539. 18:59they didn't even know what to do with
540. 19:00it. So they just bought back stock
541. 19:01because they literally didn't have any
542. 19:03enough investments to pour back into
543. 19:05their own businesses. Today, because of
544. 19:08the incredible amount of money that
545. 19:10they're that they're spending on these
546. 19:12data centers, their cash flow is gone.
547. 19:15>> I noticed that.
548. 19:15>> And in some cases, negative
549. 19:17>> negative now. Yeah. Yeah. Very little.
550. 19:19>> I mean, Google raised equity capital 85
551. 19:23billion. I mean, if you had said to me a
552. 19:25year a two years ago that that hey,
553. 19:27Steve, I'm gonna make a prediction.
554. 19:29Google,
555. 19:30which hasn't raised capital since it
556. 19:32went public, is going to raise 85
557. 19:35billion not in debt, in equity capital.
558. 19:38I'd have said, "You're out of your mind.
559. 19:40You're insane." Like like what are you
560. 19:42talking about? They they create 85
561. 19:45billion in cash in like overnight like
562. 19:48why what would what would possess them
563. 19:50to raise equity capital? Well, world
564. 19:52changed.
565. 19:53>> It's a very dramatic shift. Uh I I feel
566. 19:57you know as you know an investor that's
567. 19:59held Google for probably eight years
568. 20:02right the company that I hold now is
569. 20:04very different to the company that
570. 20:07>> and that's an interesting point we've
571. 20:09seen the market get a little bit jittery
572. 20:12with the amount of spending that's
573. 20:14happening there could be a payoff maybe
574. 20:16there's not you
575. 20:17>> by the way let me just jump you for one
576. 20:18second let's go back to Oracle
577. 20:20>> okay
578. 20:21>> because after that I I didn't finish
579. 20:23after Oracle Um every people said that
580. 20:2750% of Oracle's RPO the the backlog is
581. 20:31from open AI. the stock which had gone
582. 20:33from 230 to 330 over the next 2 3 months
583. 20:38went to 200 and today it's 150 and
584. 20:41what's fascinating fascinating
585. 20:44>> is Oracle just reported and the numbers
586. 20:47were pretty good
587. 20:48>> and the stock was up four or five% after
588. 20:51hours
589. 20:53>> and was up 7% at the open and closed
590. 20:56down on day
591. 20:57>> right
592. 20:57>> and I was and and there was no news so I
593. 21:00don't have like I don't have like a news
594. 21:02explanation like nothing happened but
595. 21:06>> clearly people are very nervous about
596. 21:08Oracle because Oracle got downgraded and
597. 21:10its debt rating is like triple B minus
598. 21:12by S&P which is like I think maybe just
599. 21:15one level above junk.
600. 21:18>> So people are nervous about Oracle and
601. 21:19how much debt they have.
602. 21:20>> Well, it seems investors are nervous
603. 21:22about all of these hyperscalers now that
604. 21:25are investing literally hundreds of
605. 21:27billions like
606. 21:28>> hundreds
607. 21:28>> hundreds of billions. It it's just it's
608. 21:31insane. staggering. It's the numbers and
609. 21:33you know the numbers are just so big.
610. 21:36>> I think the number that I heard this
611. 21:37year is that if you just look at the
612. 21:38hyperscalers, they will spend $700
613. 21:41billion on AI capex this year.
614. 21:45>> It's like that's such a huge number.
615. 21:47It's hard to even get your mind around
616. 21:49it.
617. 21:49>> It's so enormous.
618. 21:50>> I'm interested in your perspective on
619. 21:53what Michael Bur has been saying where
620. 21:55he's concerned that the data centers are
621. 21:58taking too long to come online. They're
622. 22:00buying so many chips. He his opinion is
623. 22:03the chips become obsolete way faster
624. 22:05than the depreciation schedules.
625. 22:07>> I I I understand his argument. So So let
626. 22:09me give his it's its due first. Yeah.
627. 22:13>> What he pointed out last year, I think
628. 22:15in November was that the hyperscalers
629. 22:18had changed the depreciation schedule
630. 22:22of the chips from 3 to four years to
631. 22:25like five to six years. And if you did
632. 22:28like a I can't remember exactly what the
633. 22:30calculation but but it's an it's an
634. 22:32enormous increase in profitability just
635. 22:34from the change in that accounting
636. 22:36because by by changing right it's like
637. 22:39click by by um by changing your
638. 22:42depreciation schedule from three 3 to
639. 22:44four years to 5 to 6 years by definition
640. 22:46your depreciation expense which you
641. 22:48report is going to be lower all other
642. 22:50things being equal. He also said that,
643. 22:53you know, there's so many new chips
644. 22:54coming that they become obsolete. Where
645. 22:56I think he's wrong for the moment is
646. 23:00that there is such demand for chips
647. 23:04right now that there's still huge demand
648. 23:06for the older chips whose price has gone
649. 23:08up with all the other chips,
650. 23:10>> right?
651. 23:10>> So I I I think with all due respect to
652. 23:15Michael, I think his argument is too
653. 23:18academic.
654. 23:19>> Okay? Like put this way if AI succeeds
655. 23:23because anthropic and open AI you know
656. 23:25grow like crazy and the hyperscalers do
657. 23:28well etc etc it's not going to matter if
658. 23:31the depreciation schedule changed from 3
659. 23:33to four years to 5 to six years
660. 23:35>> right
661. 23:35>> at the same time if open AI fails and
662. 23:39and the whole chain goes in reverse
663. 23:41we'll have a massive correction which
664. 23:43has nothing to do with the depreciation
665. 23:45schedule I don't I mean I think what
666. 23:47he's deep down what he's is trying to
667. 23:49point out is maybe there's something
668. 23:51wrong here, but I don't think what the
669. 23:54thing that he's pointing to as being
670. 23:55wrong is what's going to is is important
671. 23:57enough,
672. 23:58>> right? There's bigger factors that play
673. 24:00both directions in both
674. 24:01>> much bigger factors.
675. 24:02>> Okay. Interesting. So, we've spoken
676. 24:04about no moes, we've spoken about China
677. 24:08coming in potentially being competition.
678. 24:10Another headwind that I've been trying
679. 24:12to wrap my head around more is is the
680. 24:14power element as well
681. 24:17>> because this is another one of those big
682. 24:19things that we're talking about earnings
683. 24:21and chips and this and that.
684. 24:23>> But when I started to look at power, I I
685. 24:26think it was Elon Musk's interview with
686. 24:28the economist that opened my eyes up to
687. 24:29it. He said China has a chip problem.
688. 24:32The US has a power problem.
689. 24:33>> He's right.
690. 24:34>> However, in his view, China can solve
691. 24:38its chip problem. might take some time,
692. 24:40but a harder one to solve is the power
693. 24:42problem because power is physical
694. 24:43infrastructure. It takes a long time.
695. 24:45>> Correct.
696. 24:46>> I don't know if I have the expertise or
697. 24:48the understanding to know how big of a
698. 24:50restraint or a bottleneck power in the
699. 24:53United States is actually going to be.
700. 24:55>> Get in line. Nobody Nobody knows,
701. 24:57>> right?
702. 24:58>> I mean, I keep looking, you know, I
703. 25:00there are people who say it's a b that
704. 25:01things are slow. There are other people
705. 25:02who say things are fine. I can't I can't
706. 25:06nail it down yet.
707. 25:07>> Right. Okay. I mean, I do know that the
708. 25:09companies that are involved with power
709. 25:12>> are doing great.
710. 25:13>> Like Genova, for example, and that
711. 25:15stock's gone nuts. You know, I'll pat
712. 25:18myself a little bit on the back. I
713. 25:19bought that stock really early. Oh,
714. 25:20really?
715. 25:20>> But but I I bought it because the sell
716. 25:22side analyst I'm very friendly with told
717. 25:23me I should buy it and I just took a
718. 25:24flyer on it.
719. 25:25>> But um
720. 25:26>> for those that don't know Geneva told
721. 25:28me,
722. 25:28>> let me tell you, it's very interesting.
723. 25:31So GE used to be composed basically of
724. 25:34three massive divisions. healthcare,
725. 25:38>> aerospace where they basically make the
726. 25:41jet engines for planes and then they
727. 25:42service them
728. 25:44>> and then call it energy. If you ever saw
729. 25:47a jet engine
730. 25:49>> and looked at at a gas turbine, which is
731. 25:51what goes into a utility that creates
732. 25:53electricity, they look exactly the same.
733. 25:55It's just that the gas turbine is much
734. 25:57bigger.
735. 25:58>> Yes.
736. 25:59>> But it's basically the same technology.
737. 26:01So the energy division of of GE
738. 26:04makes gas turbines. They have all this
739. 26:06electrical equipment that they sell and
740. 26:08then they have a wind division which
741. 26:10does terribly.
742. 26:12>> Now this should show you how like fast
743. 26:14the world can change. The the energy
744. 26:17division was created when I think around
745. 26:212015 or so. GE bought a company in
746. 26:24Europe called Olam. Now Olm did was an
747. 26:27energy company that also created gas
748. 26:29turbines and GE had a business that
749. 26:31created gas turbines. So they mushed
750. 26:33them together just in time for the
751. 26:36entire gas turbine business to fall
752. 26:38apart,
753. 26:39>> right?
754. 26:40>> And this is why IML lost his Jeff Immel
755. 26:43who was the CEO of GE finally lost his
756. 26:45job because that was like enough
757. 26:47already. So eventually all three
758. 26:49divisions got spun out. So there's GE
759. 26:52healthcare. I think its symbol is GE.
760. 26:54>> Okay.
761. 26:55>> And there's the energy division which is
762. 26:57called GE Vernova which is GEV. And then
763. 27:01there's GE which is the aerospace
764. 27:03division.
765. 27:04>> Gotcha.
766. 27:05>> A year before [snorts] GE Vernova got
767. 27:08spun out. So that would have been like
768. 27:112023
769. 27:13maybe or 2022. If you were to read
770. 27:15sellside reports
771. 27:18upon about the industry, the energy
772. 27:20business was so bad
773. 27:23that they ascribed negative value to G
774. 27:26to to Vernova. Negative value that it
775. 27:28was worth negative. I think when one guy
776. 27:30wrote it was worth negative3 billion.
777. 27:32>> Oh my gosh.
778. 27:33>> In terms of a sum of the parts analysis
779. 27:35>> right
780. 27:35>> now what's happened is even prior to the
781. 27:40whole data center thing electrical
782. 27:43production in the United States finally
783. 27:45started to increase for the first time
784. 27:47like in 15 years. Now add on top of that
785. 27:51the
786. 27:52data centers and you're talking about US
787. 27:56electricity growing 3 to 4% per year.
788. 28:00Now that may not sound like such a huge
789. 28:03number but 3 to 4% off of the base of
790. 28:07the United States is the equivalent of
791. 28:10like two large cities.
792. 28:11>> Okay. It's a lot.
793. 28:12>> It's huge.
794. 28:13>> Yeah. You know, a company like GE
795. 28:14Vernova has backlogged like two 20 35
796. 28:18that that's how crazy things are
797. 28:20>> because this is how these data centers
798. 28:23are being powered. It's with these gas
799. 28:25turbines, right?
800. 28:26>> Mostly
801. 28:27>> mostly
802. 28:27>> and there's some alternatives. People
803. 28:29are talking about nuclear and they're
804. 28:30and there's a company called Bloom
805. 28:32Energy which makes its own little
806. 28:34turbine
807. 28:35>> that that you could hook up to a to a
808. 28:37data center, but most of it's going to
809. 28:39be through gas turbines. M well that's
810. 28:41what um Elon had to do with the um
811. 28:44Colossus data center that he built in
812. 28:47Memphis. The grid was too slow. It was
813. 28:50not ready. So he he ended up getting 35
814. 28:52of the portable gas turbines,
815. 28:54>> right? And hooked it up to his hooked
816. 28:56up.
817. 28:56>> Hooked it up. He created his own mini
818. 28:58power center like on site right next to
819. 29:00Okay.
820. 29:01>> So So most most of GNOVA is these um
821. 29:05>> well it's the gas turbines. It's all the
822. 29:07electrical I mean think about it. I mean
823. 29:09there's You know, you're not just when
824. 29:11you're building a new utility plant, it
825. 29:13ain't just turbine. There's all this
826. 29:15other electrical equipment that's got to
827. 29:16get hooked up. They make that too. And
828. 29:18the wind business will always, I think,
829. 29:20be a crappy business. And and that that
830. 29:23so what,
831. 29:23>> right? Okay. Very interesting. Very
832. 29:25interesting. But you would say the thing
833. 29:27to look out for in Genova's case is the
834. 29:30gas turbines. Is that the core of that
835. 29:32business?
836. 29:32>> Yeah. I mean, that's the core. And you
837. 29:33would just want to look at the orders.
838. 29:35>> Yes.
839. 29:35>> Which I think in the last quarter up
840. 29:36like 85%. Something insane. I
841. 29:39>> I mean they I mean there's those gas
842. 29:41turbines like in the room that we're in.
843. 29:43It's like is it's like five of these
844. 29:45rooms combined is how big these things
845. 29:48are. They're huge. They're I mean it
846. 29:50takes years to build them.
847. 29:51>> Yes. Well, that's what I was going to I
848. 29:53think I read something that their
849. 29:54backlog is stretching out to past 2030
850. 29:58or something like it is
851. 29:59>> which is just
852. 30:00>> Well, because people want to they want
853. 30:02to line it up as much as they can. M but
854. 30:04is that even more of a uh an argument
855. 30:08for this power problem if [laughter]
856. 30:11people are making orders now we want
857. 30:12these turbines now and hang on well
858. 30:14we've got 2030 you want
859. 30:16>> I I just don't know I really don't know
860. 30:18I don't have enough information
861. 30:19>> yeah I um I can't remember who it was
862. 30:21that you interviewed the man that knew
863. 30:24uh it was about power but I found that
864. 30:26was a really good interview I might
865. 30:27leave it linked um on screen right now
866. 30:28but I I thought that was a really really
867. 30:30good explanation okay so we've covered a
868. 30:33lot
869. 30:33headwinds when it comes to AI. I think
870. 30:37you you went on the record saying that
871. 30:39if you try and predict what's going to
872. 30:40happen, you're a fool. So, don't don't
873. 30:42try and predict it. I think there are a
874. 30:43lot of people out there that feel
875. 30:45compelled to look at these AI plays to
876. 30:48to look into the realm of AI. If you're
877. 30:52if someone comes to you and says, "Oh,
878. 30:53look, Steve, I really got to get in on
879. 30:55AI somehow." What are what are some of
880. 30:57the maybe safer ways to play AI? And
881. 31:02what's what would you say are the
882. 31:04high-risk ways to play AI? Talking about
883. 31:06just investing in in stuff.
884. 31:07>> I mean, I would play I wouldn't invest
885. 31:10in an LLM because I just think, as I
886. 31:12said, there's no I would not. I would
887. 31:13not
888. 31:14>> because there are no moes.
889. 31:15>> Yeah, that makes sense.
890. 31:16>> I might be a little wary of the
891. 31:18hyperscalers at this point just because
892. 31:20their businesses have they've lost all
893. 31:22their cash flow.
894. 31:23>> But I would be looking at the companies
895. 31:25that are getting that cash flow,
896. 31:27>> right?
897. 31:28>> So, you know, that would be Nvidia. you
898. 31:30know, maybe you you'd want to own
899. 31:32Micron, GE, Verova, um, Arista Network,
900. 31:37Cisco, and then if you want to get into
901. 31:39the industrial side, you could talk
902. 31:40about like an Eaton, which could is
903. 31:43electrification company. That's what I
904. 31:45would
905. 31:45>> So, it's more it's more picks and
906. 31:46shovels.
907. 31:47>> Picks and shovels,
908. 31:48>> right? As opposed to the flashy software
909. 31:50side,
910. 31:50>> right?
911. 31:51>> Okay.
912. 31:51>> The software, you know, the whole
913. 31:53software industry to I mean, I'm sure
914. 31:54you've heard the word SAS apocalypse.
915. 31:56>> I have. Um, [laughter]
916. 31:57and I I I I all I know is there will be
917. 32:02software companies that will have
918. 32:03problems
919. 32:04>> because, you know, take this new Agentic
920. 32:07AI um, Muse, I think it's called, that
921. 32:10Meta put out. You know, if I want to
922. 32:11book
923. 32:13a flight, I say to my muse, oh, that's
924. 32:17good. I say to my muse,
925. 32:19I want to book a flight to Miami on such
926. 32:23and such a date.
927. 32:25book me in the best hotel in Bickl.
928. 32:28>> Okay. And it goes and does it. Well, how
929. 32:30does it do it? It goes on all the travel
930. 32:33sites
931. 32:35and finds the best price and books it.
932. 32:38>> Mhm.
933. 32:39>> Well, that kind of makes the travel
934. 32:41sites worth less because you're not
935. 32:43going to you're not going to bookings or
936. 32:45travel velocity or what whatever
937. 32:47directly anymore.
938. 32:50>> You're not their customer anymore. your
939. 32:52AI is
940. 32:53>> you're AI my AI agent is my customer he
941. 32:56and he does the work
942. 32:57>> and the sidebar ads don't work on that
943. 32:58>> so stuff like that I think stuff in the
944. 33:01payment world could get dicey but on the
945. 33:03other hand you know software that's
946. 33:05deeply embedded in enterprises is
947. 33:06probably okay
948. 33:07>> well that's what I was going to ask you
949. 33:08it sounds like the most important thing
950. 33:09to look at is the switching mode and how
951. 33:11how resilient
952. 33:13>> how resilient is it okay I mean you know
953. 33:15for bookings what's the switching mode
954. 33:16you know I go to I go on the bookings
955. 33:18website and I book a trip so now I don't
956. 33:21go on the bookings website. I my agentic
957. 33:24AI finds just the best deal.
958. 33:26>> So that kind of makes the the travel
959. 33:29online
960. 33:31companies worth less. I think it's a
961. 33:34little early, but I think that's a
962. 33:35possibility.
963. 33:36>> Yeah. Okay. So I I guess another
964. 33:40argument that I've heard and I'm
965. 33:42interested to hear your overarching
966. 33:43thoughts on this around AI is people are
967. 33:46very fast to liken it to 1999,
968. 33:50>> a techbubble 2.0. Oh, no. You know,
969. 33:52that's that's what the media will say,
970. 33:54>> right?
971. 33:54>> I'm very interested in what your
972. 33:57thoughts are on this.
973. 33:59I have my own opinion, but I'm
974. 34:01interested to hear what you think. We're
975. 34:03in the same setup. New technology,
976. 34:05speculation in financial markets,
977. 34:08>> similar setup. Is it different this
978. 34:10time? Is there anything fundamentally
979. 34:12different?
980. 34:12>> I I don't know if it's different or not.
981. 34:14I think it's too early. I mean, if open
982. 34:17AI or anthropic fail, you'll have a real
983. 34:20correction and then the next generation
984. 34:21of people will come up and pick up the
985. 34:23pieces. I don't know if that's going to
986. 34:24happen or not. So, I I just don't know.
987. 34:26>> Fair enough. Going back to the investing
988. 34:28argument, I I'm actually interested
989. 34:30because I didn't ask you last time and I
990. 34:32had some subscribers that are interested
991. 34:34in understanding how you actually go
992. 34:36about your investing, not stocks, not
993. 34:38what stocks you're picking or anything
994. 34:39like that, but when it comes to the Real
995. 34:42Eman playbook, what is the Real Eisman
996. 34:44playbook? How do you analyze companies?
997. 34:47Is it do you stick within a circle of
998. 34:49confidence? Do you go down rabbit holes?
999. 34:50Do you look at certain financial metrics
1000. 34:52that you really love to see or not?
1001. 34:54>> Well, a couple of things. I'm very
1002. 34:56storyoriented.
1003. 34:57>> Story. Okay.
1004. 34:58>> I am not a quote unquote value player, I
1005. 35:01think.
1006. 35:01>> Okay.
1007. 35:02>> You know, stocks are cheap. They're
1008. 35:03probably cheap for a reason. Okay. You
1009. 35:05know,
1010. 35:05>> but I can't see you being someone that
1011. 35:07will grossly overpay for something
1012. 35:08either.
1013. 35:09>> I I see you as personally I see you as
1014. 35:11very rational. You know, I I mean, I
1015. 35:13would have loved to have owned
1016. 35:14Palunteer, but I won't buy it now
1017. 35:15because it's so expensive.
1018. 35:17>> Yeah.
1019. 35:17>> Um,
1020. 35:18>> so story but rational.
1021. 35:19>> Very story. Story but rational.
1022. 35:21[laughter]
1023. 35:21>> Okay. Is there anything uh are there any
1024. 35:25kind of uh metrics on your checklist or
1025. 35:27anything that you love to look at that
1026. 35:29might be a red flag? I'm just interested
1027. 35:31to see like what you really look for. Is
1028. 35:33a moat like a must-have for you or
1029. 35:36>> Not necessarily. I like a moat.
1030. 35:38>> That's why I own Moody's for example. Um
1031. 35:41that's why I own Visa.
1032. 35:43>> Mhm.
1033. 35:44>> But um it's not a complete requirement.
1034. 35:47So no.
1035. 35:48>> Is there anything that you particularly
1036. 35:51hate to see in a company? What's what's
1037. 35:52what are some things that will instantly
1038. 35:54turn you off?
1039. 35:55>> Management selling stock.
1040. 35:57>> Okay.
1041. 35:58>> I generally don't like cyclical
1042. 36:00companies.
1043. 36:01>> Okay.
1044. 36:01>> Because then you're just predicting the
1045. 36:04economy. And I mean there are
1046. 36:06exceptions, but I I I like companies
1047. 36:09that have a real story with real growth
1048. 36:12tailwinds,
1049. 36:13>> right?
1050. 36:13>> That's what I like.
1051. 36:14>> Okay. Interesting. Hey, do you mind if I
1052. 36:16finish off by asking you some questions
1053. 36:18from the audience? Sure. Is that all
1054. 36:19right?
1055. 36:20>> All right. I had a quick screen, but I
1056. 36:23might have forgotten some. [laughter]
1057. 36:25>> All right, let's Oh, this is a really
1058. 36:26interesting one. I did want to get your
1059. 36:27opinion on this. US debt. This is such a
1060. 36:31a an interesting topic and it's very
1061. 36:34very very highly covered. So it's at $40
1062. 36:36trillion now. The average interest rate
1063. 36:39on it has gone from 1.77% in 2020 to
1064. 36:423.45% today. The interest expense has
1065. 36:45risen from 523 billion a year to now
1066. 36:471.22 trillion. Is that something
1067. 36:50investors need to be genuinely worried
1068. 36:52about? Is there a real risk of a debt
1069. 36:55spiral in the future?
1070. 36:57>> I mean all the things be equal. I wish
1071. 36:58the deficit was smaller.
1072. 36:59>> Y
1073. 37:00>> um I I have my doubts about a debt
1074. 37:03spiral. Number one, we are the reserve
1075. 37:06currency of the world. But maybe even
1076. 37:08more importantly, US treasuries are the
1077. 37:12financial system of planet earth. So
1078. 37:14just for example,
1079. 37:16banks all over the world do something
1080. 37:19called repos where they lend to each
1081. 37:20other overnight. They do it through
1082. 37:23overnight treasuries.
1083. 37:25So, as long as the US Treasury is the
1084. 37:30backbone of the financial system of the
1085. 37:32world, I tend not to worry about the
1086. 37:34deficit too much, although I'd like it
1087. 37:36to be smaller. If there was an
1088. 37:38alternative, we'd be in trouble.
1089. 37:40>> Okay?
1090. 37:41>> But there is no alternative at this
1091. 37:43point.
1092. 37:43>> Let me ask you this. The Fed just raised
1093. 37:46rates for the first time in 3 years. Are
1094. 37:49rates uh likely to be a showstopper for
1095. 37:52the market and in particular the AI
1096. 37:54narrative? The rate to look at is the
1097. 37:5510-year.
1098. 37:56>> It's long the long-term rates, not the
1099. 37:58short-term rates that the Fed does
1100. 38:00because that's what people borrow.
1101. 38:01>> Okay?
1102. 38:02>> You know, the Fed is just
1103. 38:03>> the Fed funds rate is the rate at which
1104. 38:05the Fed lends to banks overnight.
1105. 38:07>> Okay?
1106. 38:08>> You don't have access to that. Neither
1107. 38:10do [laughter] I.
1108. 38:12>> Um to the side.
1109. 38:13>> Yeah. So,
1110. 38:15you know, just today, for example, the
1111. 38:17markets rallied because despite the Fed
1112. 38:19raising rates, the 10-year yield went
1113. 38:21down.
1114. 38:23I'm getting the feeling that 5% is sort
1115. 38:25of the Rubicon for the market.
1116. 38:27>> And um as long as we're below that,
1117. 38:30we'll probably be okay. But if something
1118. 38:32were to happen and and we blow through
1119. 38:34that, I think we get a correction.
1120. 38:36>> Okay. Because it's around
1121. 38:37>> that's just my guess. I actually thought
1122. 38:38originally the number was 4 and a half%
1123. 38:40and I was wrong.
1124. 38:41>> Okay.
1125. 38:41>> But five feels more.
1126. 38:43>> Do you know where it is now? It's around
1127. 38:44there.
1128. 38:45>> It's 4.98.
1129. 38:46>> 4.98. Okay.
1130. 38:47>> But it was over 5% yesterday. Yes.
1131. 38:49>> And it's come back down. People are
1132. 38:51watching every tick.
1133. 38:53>> Yes. Interesting. I'm interested to hear
1134. 38:56Steve's thoughts on Here we go. rising
1135. 38:58yields of long-term treasuries. Is Scott
1136. 38:59Bessant's buyback plan really designed
1137. 39:02to increase liquidity in older long-term
1138. 39:04bonds? Or is the government trying to
1139. 39:06manipulate long interest rates to ease
1140. 39:08their interest problem?
1141. 39:10>> The latter.
1142. 39:11He's trying to buy long long-term
1143. 39:14treasuries to bring rates down to ease
1144. 39:17the cost of of money for the United
1145. 39:19States of America. He's failed miserably
1146. 39:22at this point. You know, rates are
1147. 39:24higher than when he made his
1148. 39:25announcement. I have a suspicion that
1149. 39:27he's going to come with something else
1150. 39:29because because well, he he he announced
1151. 39:316 billion. 6 billion is nothing. I mean,
1152. 39:33it's a $40 trillion deficit.
1153. 39:35>> That's that was my thought.
1154. 39:36>> So, I I I don't think he's an idiot. So,
1155. 39:40I think he's going to come with
1156. 39:41something else. What else that is, I
1157. 39:42don't know.
1158. 39:42>> Okay. So, a different plan of attack to
1159. 39:44do the same thing. Yes. Ah, okay. Okay.
1160. 39:47I've always been skeptical about
1161. 39:48precious metals, but Steve's recent
1162. 39:50discussion with Porter Collins and
1163. 39:51Vincent Daniel made me second guess
1164. 39:52that. I would like to know if he has any
1165. 39:55conflicting feelings about precious
1166. 39:56metals, or is he still firmly opposed to
1167. 39:59the asset class?
1168. 40:00>> I'm not opposed, but I don't I've never
1169. 40:02owned it.
1170. 40:02>> Y
1171. 40:02>> I've never owned it. I It's not
1172. 40:05something that has ever really enticed
1173. 40:07me one way or the other. Similar
1174. 40:08thinking to Warren Buffett. It just sits
1175. 40:10there and looks right.
1176. 40:11>> Sits there. Does nothing.
1177. 40:12>> Does nothing. It's not productive. It
1178. 40:14>> It has value because people say it has
1179. 40:15value.
1180. 40:16>> Okay.
1181. 40:16>> That's all. Doesn't pay you an interest
1182. 40:18rate.
1183. 40:19>> Similar argument, I'm guessing, to
1184. 40:20Bitcoin and anything else that sits
1185. 40:22there in
1186. 40:22>> Oh, definitely.
1187. 40:23>> Baseball cards, blah blah blah, whatever
1188. 40:24sits there.
1189. 40:25>> Well, Bitcoin is worse.
1190. 40:26>> Bitcoin is worse. Yeah,
1191. 40:28>> Bitcoin is worse because it trades
1192. 40:30inversely to its own thesis.
1193. 40:32>> Yes, I have noticed that. That is quite
1194. 40:34strange. So, so for those for your
1195. 40:35viewers, what I what I mean by just to
1196. 40:37explain what I mean by that is people
1197. 40:39like if you went to a Bitcoiner
1198. 40:41>> and you said, "Dude,
1199. 40:43>> why do you want Bitcoin?" The answer you
1200. 40:45would invariably get is that that fiat
1201. 40:47currency, which is government
1202. 40:48currencies, has been debased, inflation
1203. 40:51is coming, and you want to hedge against
1204. 40:53this, so buy Bitcoin.
1205. 40:55>> Okay, that sounds reasonable. The
1206. 40:58problem is that if that were the case,
1207. 41:00on days where people worry about
1208. 41:02inflation, rates are going up, and the
1209. 41:05stock market goes is down, Bitcoin
1210. 41:07should be up. And on days where NASDAQ
1211. 41:10is up like crazy and rates are down,
1212. 41:12Bitcoin should be down. But it does the
1213. 41:15opposite.
1214. 41:15>> Yes.
1215. 41:16>> So, so you know, you you say that
1216. 41:18Bitcoin is going to go up when because
1217. 41:20when the sky is blue and then the sky is
1218. 41:23blue and it goes down, you know, why do
1219. 41:25I own it? You have no thesis. It seems
1220. 41:28to be just a speculative asset.
1221. 41:29>> It's just like it's a way to speculate
1222. 41:31about speculating.
1223. 41:32>> Speculate about speculating. I like
1224. 41:34that. All right, let me ask you this.
1225. 41:36Dear Steve, you seem to be very much
1226. 41:38centered on the US stock market. Have
1227. 41:39you ever tried to broaden your investing
1228. 41:40or trading horizon geographically? Do
1229. 41:42you have any interest in companies held
1230. 41:44outside the US?
1231. 41:45>> Great question. When I used to run my
1232. 41:47hedge funds, I used to invest overseas.
1233. 41:49>> Mh.
1234. 41:49>> And I found that there was no night and
1235. 41:51there was no day.
1236. 41:52>> Okay. So for the last many many years,
1237. 41:55all I do is the US and I'm perfectly
1238. 41:57happy. There's plenty to do in the US. I
1239. 41:59don't feel the need to invest overseas.
1240. 42:01>> Fair enough. Plenty of opportunities
1241. 42:02here. I like it. What's his view on when
1242. 42:05all these wars would end? I guess
1243. 42:07particularly the Iran war. Do you see
1244. 42:09inflation coming down anytime soon?
1245. 42:12>> I have no more insight into the war than
1246. 42:13anyone else, so I can't answer the
1247. 42:16question.
1248. 42:16>> Fair enough. Well, I think that is just
1249. 42:19about all we've got. Is there one more?
1250. 42:21Is there one or a set of numerical
1251. 42:23indicators, whether related to interest
1252. 42:24rates or inflation or unemployment or
1253. 42:26otherwise, that Steve could see as being
1254. 42:28the tipping point for the US to take its
1255. 42:30medicine with cutting benefits or wash
1256. 42:32spiking rates up or whatever he thinks
1257. 42:34that medicine might be? I've seen I've
1258. 42:36seen it said that oil will rise until
1259. 42:38stocks fall, meaning that the US won't
1260. 42:40get out of Iran until stocks really take
1261. 42:41a beating. But this is a much broader
1262. 42:43question about getting the debt back
1263. 42:45well under control. I guess we kind of
1264. 42:47touched on that with the kind of touch.
1265. 42:49I mean, all I would say is that's a
1266. 42:51total political question,
1267. 42:52>> and there's no political appetite in
1268. 42:54Washington right now to cut the deficit
1269. 42:57>> a dollar
1270. 42:58>> by either side.
1271. 43:00>> We'll see where that goes.
1272. 43:01>> I think that's just about how we
1273. 43:03finished our last [laughter] our last
1274. 43:04interview. And I was like, on that
1275. 43:06cheery note, so again, on that cheery
1276. 43:08note, Steve, thank you very much for for
1277. 43:10coming on. For those that don't know,
1278. 43:12well, we talked about it earlier, but
1279. 43:13The Real Eisman Playbook is uh is what
1280. 43:15you're currently working on. That's your
1281. 43:16podcast. Can you tell us a little bit
1282. 43:18more? What can people expect from that?
1283. 43:19It's on YouTube. I guess you can get it
1284. 43:21on podcast platforms as well.
1285. 43:22>> Well, we do two two free podcasts a
1286. 43:24week. So, one is an interview. So, and
1287. 43:27then the other one is a market rap where
1288. 43:30on Friday I put out like a summary of
1289. 43:31the whole week and then if you're
1290. 43:33willing to pay for the payw wall on
1291. 43:34Substack, we do an additional podcast
1292. 43:37which is sometimes an interview. This
1293. 43:39week was part one of two-part master
1294. 43:42lecture of how to analyze banks.
1295. 43:45>> All right.
1296. 43:46>> Yeah. I have to check
1297. 43:46>> if anybody who ever But you have to
1298. 43:48subscribe.
1299. 43:49>> Okay, I will subscribe. [laughter] No
1300. 43:51free lunch.
1301. 43:51>> No free lunch there city here.
1302. 43:54>> I will see.
1303. 43:54>> Yeah, but if you want to know how to
1304. 43:55analyze banks,
1305. 43:56>> go there.
1306. 43:57>> That's where you should go.
1307. 43:58>> Awesome, Steve. Thank you very much for
1308. 44:00Thank you very much. Appreciate it.
1309. 44:01>> Great. Bye.
1310. 44:03>> Yes,
1311. 44:05you're the man now, dog.