YouTube2Text

Nvidia-Krise: jetzt kaufen? 5 Crash-Aktien im CHECK, Anthropic-IPO & DAS ist unser Dip der Woche! — Transcript

by Buy The Dip · 12,418 words · 1,976 segments · language en · Watch on YouTube

Full transcript

  1. 0:13Hi and welcome to a new episode of by
  2. 0:14the Dip. Three microphones, three
  3. 0:16opinions, seven exciting financial
  4. 0:18topics with Eriksen, Hell and Bauzus.
  5. 0:20Hello.
  6. 0:21Yes, Lars, Sebastian, there is exactly
  7. 0:23one thought that has been bothering me
  8. 0:25since the beginning of this week, giving
  9. 0:28me sleepless nights. Which of you two is
  10. 0:30the honey trap that was set on me?
  11. 0:32Come on, come on.
  12. 0:35>> You have to judge that based on your own
  13. 0:37preferences.
  14. 0:39>> Yeah, no, I keep wondering which foreign
  15. 0:41secret service would come up with such a
  16. 0:44crazy idea, because yes, two middle-aged
  17. 0:46, middle-looking men, well, not my first
  18. 0:49preference. Okay, Lars, when something
  19. 0:51>> starts with middle, it doesn't get any
  20. 0:53better afterwards, right?
  21. 0:54>> Okay, yes, yes, that's right, that's
  22. 0:57right. No, uh Lars, you can start your
  23. 0:59defense speech right away. It's not you
  24. 1:01, is it?
  25. 1:02>> Uh, no, no, no, it's not me. I'm
  26. 1:04completely perplexed too. I didn't know
  27. 1:06I had to defend myself here. I'm not
  28. 1:08really suitable as a strategically
  29. 1:10relevant person, maybe I'm from the
  30. 1:12Danish secret service.
  31. 1:15But what do they have yours for? For 4
  32. 1:17million and they're the happiest people
  33. 1:18in the world. Sorry, I can only tell you
  34. 1:21, Timo, I'm not.
  35. 1:23But I guess that's the standard answer
  36. 1:25from all spies.
  37. 1:26>> Yeah, okay. And Sebastian, which secret
  38. 1:29service would you work for? I think I'd
  39. 1:31think somewhere in the direction of
  40. 1:33Azerbaijan.
  41. 1:34>> I wanted to say the secret service of
  42. 1:36one of the wine bars we go to. I'm
  43. 1:37always allowed to lure you in there.
  44. 1:39>> Exactly.
  45. 1:42You're the emissary of Alcoholics
  46. 1:44Anonymous.
  47. 1:48Okay, but let's get a little serious
  48. 1:50again at this point.
  49. 1:54Let's discuss our topics.
  50. 1:58Today, we're going to talk about the AI
  51. 2:00trade.
  52. 2:03It may be about to enter the next stage
  53. 2:05.
  54. 2:07What's behind it and what our views are
  55. 2:09, more on that in a moment.
  56. 2:12Then of course we'll also talk about
  57. 2:14Nvidia for reasons.
  58. 2:16I don't think I need to go into them any
  59. 2:18further now, but of course also with
  60. 2:19reference to stocks, so it won't just be
  61. 2:21gossip.
  62. 2:24Then we'll talk about our personal dip
  63. 2:26of the week and I'll also have a nice
  64. 2:27announcement to make later for anyone
  65. 2:29interested.
  66. 2:32So it's definitely worth listening a
  67. 2:34little more closely to topic number
  68. 2:35three. We'll talk about Core Weave, Iron
  69. 2:38and Nebius. Is the big turnaround coming
  70. 2:41now, we ask.
  71. 2:43We've taken a close look at some German
  72. 2:45stocks this week and have come up with a
  73. 2:47theory that's really surprising.
  74. 2:49We'll unravel what exactly. We received
  75. 2:52a listener question from Lukas. Best
  76. 2:54wishes to you, Lukas. Rheinmetall buy
  77. 2:56the dip, he asks.
  78. 2:58Softbank, yeah, we'll also take a look
  79. 3:00at that group.
  80. 3:02A totally crazy bet, question mark, and
  81. 3:05that's actually seven topics already, so
  82. 3:07it's a good program in that respect, I
  83. 3:09think. It's going to be an outstanding
  84. 3:12episode, I have a feeling, and you can
  85. 3:15submit listener questions in our app
  86. 3:17www.bei-der-dip.de. There you can find
  87. 3:20all the information about our project,
  88. 3:22which is also celebrating its two-year
  89. 3:24anniversary this weekend.
  90. 3:28We're very proud of that, so just get
  91. 3:30informed, download the thing and take a
  92. 3:32look.
  93. 3:35Topic number one: What's going on with
  94. 3:37the AI trade or is the next stage being
  95. 3:39ignited?
  96. 3:41That's the initial thesis, and Lars, I
  97. 3:43actually saw a very interesting video on
  98. 3:46your channel this week.
  99. 3:50You have a favorite analyst, if I may
  100. 3:52say so, and he has a very, very special
  101. 3:55thesis on the subject, and yes, let's
  102. 3:57hear what our colleague said.
  103. 4:02>> Exactly. Um,
  104. 4:04>> Aswath Damodaran is his name, and the
  105. 4:06interesting thing about him is the first
  106. 4:08comment, where I thought, my goodness,
  107. 4:10if you're going to talk nonsense, you
  108. 4:12don't have to be the first for it to be
  109. 4:14noticed right away. The first comment
  110. 4:17under the video is, oh, Wall Street, uh
  111. 4:19, uh, bulls or ghosts always only write
  112. 4:22or talk according to their own wallets
  113. 4:24anyway.
  114. 4:27So the man is a lecturer in New York and
  115. 4:29doesn't have his own, he doesn't even
  116. 4:31have a paid blog, nothing.
  117. 4:37But I think anyone who likes fundamental
  118. 4:39analysis and the valuation of companies
  119. 4:41will come across Damodaran at some point
  120. 4:43and I can honestly say that it's a blog
  121. 4:45worth reading.
  122. 4:49I think the reason he doesn't have that
  123. 4:52many followers and views is that he
  124. 4:54doesn't bother to simply say, yes, I'll
  125. 4:56write the conclusion here, but instead
  126. 4:58he sometimes writes many, many pages to
  127. 5:00justify why a company is valued the way
  128. 5:03it is.
  129. 5:07So. And he commented on Entropic and
  130. 5:09this is perhaps a bit, but now I'll give
  131. 5:12a short summary, um, his valuation
  132. 5:14roughly corresponds to what Entropic is
  133. 5:16aiming for as an IPO valuation.
  134. 5:23So I think we'll end up with 1.4 to 1.5
  135. 5:26trillion US dollars and before we had
  136. 5:28the prospectus, and that can still be
  137. 5:30changed, yes, there are rumors that
  138. 5:33demand is so high that they want to go
  139. 5:35towards 1.5, maybe even a bit higher.
  140. 5:41And he already said before that from his
  141. 5:43point of view the justified value at the
  142. 5:45moment would be 1.45 trillion dollars,
  143. 5:48so pretty much in the middle of this
  144. 5:50issue range.
  145. 5:55And now I'm also thinking about what I
  146. 5:57highlighted in the video as particularly
  147. 5:59exciting, but what needs to be said
  148. 6:01clearly and I would like to share this
  149. 6:03with you, and his table illustrates this
  150. 6:06very well, is that you could easily
  151. 6:07value Anthropic 50%higher, but there are
  152. 6:10only a few adjustments missing, because
  153. 6:12we're simply talking about growth here
  154. 6:14and you could also say that Anthropic is
  155. 6:16actually only worth half.
  156. 6:21I'll look up the figures for you in a
  157. 6:22moment.
  158. 6:24There's a very interesting one from the
  159. 6:27very busy analysts at Morningstar, but
  160. 6:29yes, maybe Sebastian or you can start
  161. 6:31first and then I'll give it to you again
  162. 6:33in a moment.
  163. 6:36>> Yes, Sebastian, the stage is yours.
  164. 6:40>> Would you like my assessment of
  165. 6:42Anthropic or whether I would underwrite
  166. 6:44it for an IPO?
  167. 6:46Feel
  168. 6:46>> free to talk about Anthropic first and
  169. 6:48you can follow up on it straight away if
  170. 6:50you like.
  171. 6:51>> So, we talked about Anthropic two weeks
  172. 6:54ago. I can only roughly repeat the
  173. 6:56conclusion.
  174. 6:57I think Anthropic is a highly
  175. 6:58efficiently managed company.
  176. 7:01Also, what I found really fascinating in
  177. 7:03this S1 filing–that's the IPO prospectus
  178. 7:06that you first initiate and file and
  179. 7:08announce–and at that time, the number of
  180. 7:11shares and the valuation weren't
  181. 7:13included, but you could see that
  182. 7:14Anthropic had raised around 140 billion
  183. 7:17dollars in capital in its short time
  184. 7:19since its founding, and they still had
  185. 7:21130 billion.
  186. 7:23So you can see how efficiently the
  187. 7:25company is managed and now the course is
  188. 7:27really being set, as Lars also mentioned
  189. 7:29, how do you view the value? So you see
  190. 7:31it very positively and say, the demand
  191. 7:33for Anthropic and I've also dealt a lot
  192. 7:35with the topic of business and then Lars
  193. 7:37has a point, you keep saying that AI
  194. 7:39will eventually become a commodity. That
  195. 7:41means it will become such a common good
  196. 7:43that you can't make any money with it. I
  197. 7:45read an interesting example of this
  198. 7:47recently, when refrigerators became
  199. 7:49widespread among the masses. That means
  200. 7:51that when people bought refrigerators,
  201. 7:53it wasn't the refrigerator manufacturers
  202. 7:55who made the big money, but Coca-Cola,
  203. 7:56because Coca-Cola was then able to bring
  204. 7:58completely different drinks onto the
  205. 8:00market and make completely different
  206. 8:02commercials thanks to the cooling. So
  207. 8:04that means that you don't necessarily
  208. 8:06have to profit from AI as an AI lab, but
  209. 8:08the question that needs to be answered
  210. 8:10now is: will AI become a common good?
  211. 8:13And I would now say: Classic AI, which
  212. 8:15currently costs nothing, like at Cloud,
  213. 8:17at OpenAI or even at Google, at Gemini,
  214. 8:20where people type something in, where AI
  215. 8:22is not actually used like AI in my
  216. 8:24opinion, but where people ask "When is
  217. 8:26the next football game?", "What's the
  218. 8:28weather like?", I don't need AI for that
  219. 8:30. So no one will make money with that. A
  220. 8:32bit of advertising is played and maybe
  221. 8:34that runs on the side.
  222. 8:36But the AI agents in particular, if
  223. 8:38Anthropic manages to really get into
  224. 8:40companies and bite into everything,
  225. 8:42establish themselves and then sell
  226. 8:44so-called corporate subscriptions,
  227. 8:46that's a different case again.
  228. 8:49So that's something to watch out for. On
  229. 8:51the other hand, of course, when it comes
  230. 8:53to risks, we have these almost 1
  231. 8:54trillion dollars in obligations for data
  232. 8:56centers, although I have to say that I
  233. 8:58think it's always presented that
  234. 8:59Anthropic has leased 900 billion dollars
  235. 9:01in the form of computing capacity. Will
  236. 9:03that work out? Around 480 billion will
  237. 9:05be leased until 2030. That means the
  238. 9:08rest is initially long-term, some
  239. 9:10contracts up to 20 years. That means
  240. 9:12they would have to spend an average of
  241. 9:14100 to 120 billion dollars per year for
  242. 9:16computing capacity over the next four
  243. 9:18years. Yes, it's a tall order, but of
  244. 9:19course it's significantly lower than
  245. 9:21this huge figure that's always being
  246. 9:23hung up in the shop window. But I can
  247. 9:24keep it short, I wouldn't draw.
  248. 9:27>> Okay.
  249. 9:28You can explain that briefly later.
  250. 9:31Let's stick with the two blocks.
  251. 9:33So on the one hand, the thesis that the
  252. 9:36AI trade is changing, the next level, so
  253. 9:38to speak, and then we'll also talk about
  254. 9:41the monetization opportunities.
  255. 9:46If I understood correctly in the video,
  256. 9:48Lars, what you recorded, Damodarane also
  257. 9:51said: The interests in the AI trade of
  258. 9:53the individual players–neoclouds,
  259. 9:55hyperscalers, chip manufacturers, and
  260. 9:58also the lenders who finance the whole
  261. 10:00thing–are now slowly diverging, whereas
  262. 10:03before, it was the case that everyone
  263. 10:05was, so to speak, moving in one
  264. 10:07direction.
  265. 10:09Can you perhaps
  266. 10:11break that down a little bit?
  267. 10:13>> Exactly, for months, the months in which
  268. 10:16everyone was making money with this AI
  269. 10:18trade, there was only one question: Will
  270. 10:20more be spent on capex? If so, everyone
  271. 10:23benefited.
  272. 10:25Those who spent capex, ironically, the
  273. 10:27hyperscalers have also increased, and so
  274. 10:30have all the other groups you just
  275. 10:32mentioned.
  276. 10:34And now we are more in the competitive
  277. 10:36phase.
  278. 10:37In other words, this AI trade is
  279. 10:39maturing, so to speak.
  280. 10:42Because the question is: Who would
  281. 10:43benefit from a slowdown, which is bound
  282. 10:46to happen at some point–it’s easy to
  283. 10:48extrapolate that at this rate…the Capex
  284. 10:50expenditure has multiplied, and of
  285. 10:52course it won’t in the next few years.
  286. 10:55That means growth is high, but there
  287. 10:57will definitely be a slowdown.
  288. 11:00Who will benefit from this?
  289. 11:02I can give you the answer in advance:
  290. 11:04the hyperscalers themselves, because
  291. 11:06they will very likely become cash flow
  292. 11:08positive again in the foreseeable future
  293. 11:10, in addition to Microsoft, who
  294. 11:12currently still are.
  295. 11:14And who could possibly suffer from this
  296. 11:15?
  297. 11:17Of course, you’re thinking of the
  298. 11:19infrastructure providers for this AI
  299. 11:21trade, i.e. the infrastructure in the
  300. 11:23form of data centers, the neo-clouds,
  301. 11:26but also the chip manufacturers.
  302. 11:29And the big question that the market is
  303. 11:31currently dealing with is, okay, there
  304. 11:33will be a slowdown, I think that’s
  305. 11:35already reflected in the current prices
  306. 11:37, but how much will the whole thing slow
  307. 11:39down?
  308. 11:42And I find that really hard to answer,
  309. 11:44because for me it's comparable to the
  310. 11:46fact that AI is like a toaster.
  311. 11:51And Sebastian mentioned that
  312. 11:53refrigerators or who has benefited from
  313. 11:55refrigerators becoming so big?
  314. 11:59And some time ago I
  315. 12:03looked at the adoption curve, the
  316. 12:05adaptation curve of some devices, and
  317. 12:07the toaster, and we all know how
  318. 12:09important a toaster really is in today's
  319. 12:11household.
  320. 12:14The toaster was flat for a long time and
  321. 12:16then took off like no other household
  322. 12:18appliance.
  323. 12:20Why?
  324. 12:22Because of course there were toasters,
  325. 12:24General Electric built them very early
  326. 12:26on, but there was no electricity in most
  327. 12:28homes.
  328. 12:30So what do you want with a toaster if
  329. 12:31there is no electricity? And that's the
  330. 12:33AI question for me.
  331. 12:37We know that the models are there, we
  332. 12:39know that we'll probably build the data
  333. 12:41centers, but when will the demand in the
  334. 12:43world out there be so great that we can
  335. 12:45really say that it was worth it?
  336. 12:48So from one or two trillion in
  337. 12:50infrastructure spending, three, four,
  338. 12:53five trillion in sales will also come
  339. 12:55later.
  340. 12:58And we haven't actually reached that
  341. 13:00point yet, but we're guessing when it
  342. 13:02will be the case that every medium-sized
  343. 13:04company, every large company, will
  344. 13:06introduce AI on a large scale and then
  345. 13:08pay for it.
  346. 13:10And for me, that's
  347. 13:12simply a question of valuation. A few
  348. 13:14months ago, it was perhaps priced in
  349. 13:16that the whole thing would happen very,
  350. 13:18very quickly. In my opinion, the current
  351. 13:21prices
  352. 13:23already reflect that this pace is
  353. 13:24slowing down significantly.
  354. 13:27>> Yes, okay.
  355. 13:30I would slightly disagree with Damodaran
  356. 13:32.
  357. 13:34I also think that the expectations of
  358. 13:36investors, regardless of whether they
  359. 13:38are equity investors or debt investors,
  360. 13:41are now also increasing. As you
  361. 13:42mentioned, Lars, you can see that in the
  362. 13:45share prices.
  363. 13:46But I believe that what Damodaran says
  364. 13:48is also correct, that the hyperscalers
  365. 13:50are the strongest link in the chain,
  366. 13:52because if they were to stop investing,
  367. 13:54they simply still have a core business
  368. 13:56that would tend to continue.
  369. 14:01But in reality, they cannot slow down
  370. 14:03the Capex very quickly, very quickly, or
  371. 14:05even significantly, but this Capex
  372. 14:07system and also the off-balance sheet
  373. 14:09leasing obligations that they have given
  374. 14:12up have to continue, because that is
  375. 14:14what this circular system was created
  376. 14:16for.
  377. 14:18So that means that in my opinion, the
  378. 14:21weakest chain is A) the AI labs, B) the
  379. 14:23neo-clouds. I don't think any of you
  380. 14:25would disagree with that.
  381. 14:28But in order to support them, an
  382. 14:30internal circular system was built,
  383. 14:32which we have also critically examined
  384. 14:34here and there, but in fact everything
  385. 14:36depends on this CapEx.
  386. 14:41And they can't just shut down like that
  387. 14:43, because then the entire system would
  388. 14:45collapse.
  389. 14:48That's point number one. Point number
  390. 14:51two, they might actually benefit in the
  391. 14:53short term, meaning they would fall less
  392. 14:55, the hyperscalers than the others, but
  393. 14:57thirdly, then we'll just be in a
  394. 14:59complete bear market and then it won't
  395. 15:02do the hyperscalers any good either,
  396. 15:04because their company valuation on the
  397. 15:06stock exchange will then, like
  398. 15:08everything else, simply be reduced by 30
  399. 15:10, 40, 50%.
  400. 15:11So in my opinion, if we don't want to
  401. 15:13end up in a bear market, then the
  402. 15:15hyperscalers have to continue investing
  403. 15:17.
  404. 15:19And that's point number one.
  405. 15:21The interests don't diverge that
  406. 15:25much, because they're all interconnected
  407. 15:27anyway.
  408. 15:30And then I would also say that, for
  409. 15:32example, the chip manufacturers
  410. 15:33shouldn't be underestimated in this
  411. 15:35chain.
  412. 15:39They are not as weak as one might think
  413. 15:41, because they have very high visibility
  414. 15:43, i.e. a high backlog with order volumes
  415. 15:45.
  416. 15:50But if one of the hyperscalers stops
  417. 15:51buying chips, then they will be out of
  418. 15:53business anyway, so they will have to
  419. 15:55continue paying for what has been
  420. 15:57produced so far.
  421. 16:01So I googled these contracts to see how
  422. 16:03it actually works.
  423. 16:05And then you will be out of business,
  424. 16:07i.e. if you stop ordering or cancel an
  425. 16:09order, you will have to pay for it
  426. 16:11anyway. Point number one and point
  427. 16:12number two, you will end up at the back
  428. 16:14of the list.
  429. 16:17So you will then be way back at the
  430. 16:18bottom of the list and may not have the
  431. 16:21opportunity to get chips as quickly in a
  432. 16:23later upswing phase, or you will have to
  433. 16:25pay much, much, much more than everyone
  434. 16:27else to get back in.
  435. 16:30And I think that ultimately no one can
  436. 16:31afford that. So these chip manufacturers
  437. 16:33are actually a bit more powerful, in my
  438. 16:36opinion, than the market is currently
  439. 16:38generally pricing them out to be. Um,
  440. 16:40yeah.
  441. 16:41>> I just want to add, I don't have to
  442. 16:43defend him, but he described it like
  443. 16:45this.
  444. 16:47The hyperscalers would only be better
  445. 16:49off than the neo-clouds.
  446. 16:52The hyperscalers would, if the AI trend
  447. 16:54, um, if it is said that everything was
  448. 16:56overinvested, then they would also
  449. 16:58suffer from it, just less than those who
  450. 17:00are at the end of this order chain.
  451. 17:04But, um, without a doubt, this AI trade
  452. 17:07must not be reversed.
  453. 17:08So if one day it is said, you know what
  454. 17:11, it worked, but how likely is that?
  455. 17:14That's why there have been some who have
  456. 17:16said, um, in the end, what could be
  457. 17:18better than a file folder?
  458. 17:21Although that is still part of the
  459. 17:23administration.
  460. 17:24So maybe it's a bad example, but yeah,
  461. 17:27you know what I mean.
  462. 17:29It always depends on the question of
  463. 17:31where you are in this cycle right now.
  464. 17:33Are you really in a major downturn or
  465. 17:35are you simply experiencing a flattening
  466. 17:37of growth? Of course, that makes a huge
  467. 17:40difference.
  468. 17:40>> Hmm.
  469. 17:42Yes, Sebastian, you've already clearly
  470. 17:44outlined your ideas on monetization
  471. 17:47options.
  472. 17:49Maybe you can elaborate on that a bit
  473. 17:51more, but of course you should also
  474. 17:53provide your reasons for not subscribing
  475. 17:55to Anthropic's IPO, even though it's so
  476. 17:58oversubscribed and showing signs of
  477. 18:00success. I'll get to that in a moment.
  478. 18:02>> I also think it's really positive that
  479. 18:04Anthropic is so oversubscribed, because
  480. 18:05that naturally creates a positive mood
  481. 18:07for the AI trade.
  482. 18:10But as far as monetization is concerned
  483. 18:11, I always see half of the article
  484. 18:12saying that money is being made with it
  485. 18:14. And I mean, I could make it easy for
  486. 18:16myself: money is being made on the
  487. 18:17Nvidia side, with construction companies
  488. 18:19, with cloud companies, where sales are
  489. 18:21growing. They invest a lot, but money is
  490. 18:23being made. Where we are now, and this
  491. 18:26is also what Lars said, is the adoption
  492. 18:28in the broader economy. That will take
  493. 18:30some time, because many people are not
  494. 18:32that quick. You need a certain business
  495. 18:34case to convince you that it will work.
  496. 18:36I looked it up, for example Ernst&Young
  497. 18:38, the consulting agency, they are
  498. 18:40already saving millions through AI.
  499. 18:42Klarna, for example, is saving millions
  500. 18:44. Palantir's business is growing
  501. 18:46extremely strongly, because many
  502. 18:48companies are saying, rather than
  503. 18:49everyone else messing around with us,
  504. 18:51we'd rather use a solution from Palantir
  505. 18:53, a kind of ecosystem where AI can be
  506. 18:55implemented in the company. And another
  507. 18:56interesting observation, but I'll give
  508. 18:58you a counter-argument to that in a
  509. 19:00moment.
  510. 19:01I looked at the latest GDP data from the
  511. 19:03US and you can divide the growth in the
  512. 19:05third quarter between agriculture and
  513. 19:07non-agriculture.
  514. 19:10And non-agriculture, for example, grew
  515. 19:13by 3.7%, while at the same time the
  516. 19:16hours worked only grew by 0.1%.
  517. 19:20That means we had a productivity
  518. 19:21increase of 3.6%here, which–and this is
  519. 19:24what we need to find out exactly now–
  520. 19:26may also be attributable to AI, because
  521. 19:28productivity can simply be increased
  522. 19:31through it. Now, it has to be said that
  523. 19:33the Fed published a study two days ago.
  524. 19:36They said they see it too, but they
  525. 19:38can't quantify it yet. Is this simply a
  526. 19:40different efficiency that has emerged,
  527. 19:42or is AI already involved? Well, it
  528. 19:44can't be measured yet, but at least we
  529. 19:46currently have good signs that AI is
  530. 19:48already being monetized, but of course
  531. 19:50not yet on this large scale. And that
  532. 19:52brings us to the hyperscalers and also
  533. 19:54the investments, I wanted to say
  534. 19:55something briefly about that. Of course
  535. 19:56, they also have to invest so that they
  536. 19:58don't get left behind. You can't stop
  537. 20:00now and say, we at Meta are taking a
  538. 20:02year off, then the others will overtake
  539. 20:04you. So you are forced to invest in
  540. 20:06order to stay in the game.
  541. 20:08And this wave of investment won't last
  542. 20:10forever, but will continue until the
  543. 20:12cash flows of others are strong enough
  544. 20:14for them to be able to support their
  545. 20:16companies or businesses.
  546. 20:18So I think that with this Capex, people
  547. 20:20often forget or simply extrapolate that
  548. 20:22it will continue for 10 or 20 years and
  549. 20:24if it were to stop, then the world would
  550. 20:26end. It will continue until the
  551. 20:28companies in the end-and this is not
  552. 20:29just the chip manufacturers, but also
  553. 20:31the neo-clouds, we'll get to that-have
  554. 20:33enough cash flow to stand on their own
  555. 20:35two feet. Then they'll need fewer chips
  556. 20:37, and then the whole situation will ease
  557. 20:39. And why am I not buying Anthropic? I'm
  558. 20:42not a big IPO fan, which means there's a
  559. 20:44long history of IPOs that were simply
  560. 20:46totally overvalued, then canceled and
  561. 20:48then you could buy, and I'm also waiting
  562. 20:50for Anthropic because I think it's
  563. 20:52positive that it's oversubscribed, that
  564. 20:54creates a good mood for the AI trade,
  565. 20:57but I don't have to be there from the
  566. 20:59start.
  567. 21:00So I'll just take the time to watch how
  568. 21:02it goes.
  569. 21:04>> Yes, pure politics. Pure politics. You
  570. 21:07know, he buys Valneva. Dario Amodei is
  571. 21:09too left-wing, he says, no, I'll leave
  572. 21:12him out, but SpaceX, yes, I drew him for
  573. 21:14fun.
  574. 21:14>> Yes, you're right. But that was really
  575. 21:17just for fun with really clean money, if
  576. 21:19I hadn't done it, nothing would have
  577. 21:21changed.
  578. 21:23So I totally agree with your criticism
  579. 21:25Lars, it was a quick IPO pop, but
  580. 21:27actually, when I look at the capital
  581. 21:29that I invested, then it wasn't, as they
  582. 21:31say in Wall Street banking, with
  583. 21:33conviction.
  584. 21:36So it was just: I'm in and I cut myself
  585. 21:38a nice dinner with wine, maybe a little
  586. 21:40more, but I could have left it alone.
  587. 21:43And if there's so little conviction
  588. 21:45behind it, I've decided for myself: Then
  589. 21:46I don't need to draw Anthropic either,
  590. 21:48instead I'll take a look at the company
  591. 21:50.
  592. 21:50>> Drawn. Okay, so now I only have what I
  593. 21:53have,
  594. 21:54>> but I'm not going to invest 50, 60k more
  595. 21:56in it.
  596. 21:57>> Yeah, okay.
  597. 21:59>> Half a week's salary, I mean,
  598. 22:01>> you have to defend your reputation. So I
  599. 22:03subscribed, simply because I don't want
  600. 22:05to miss out, you know, otherwise I'll
  601. 22:07just say, "Oh crap, there was something
  602. 22:09yesterday."
  603. 22:11>> The man said that in one episode, I
  604. 22:13thought it was great, yes, I bought the
  605. 22:15stock, just so I wouldn't forget it, so
  606. 22:17I could check it every now and then. So
  607. 22:19that's also such a beautiful hubris, I
  608. 22:21also bought a Linde there once, because
  609. 22:23I didn't feel like using the watchlist
  610. 22:25on TradingView. I'll just put it in my
  611. 22:26portfolio for a moment. Things are
  612. 22:28really
  613. 22:28>> hot in Switzerland at the moment.
  614. 22:30>> I think it's quite good that you
  615. 22:32subscribed, otherwise it would have been
  616. 22:33oversubscribed by two or three times. So
  617. 22:35it's quite good if you went in with a
  618. 22:37little change.
  619. 22:38No, just a quick word on monetization
  620. 22:41concerns.
  621. 22:44I honestly don't see them either, and I
  622. 22:46honestly can't really understand the
  623. 22:48supposed AI panic on Wall Street,
  624. 22:50because it's pretty obvious.
  625. 22:54So these big commodities, as you always
  626. 22:56say, these everyday goods, these large
  627. 22:58language models that everyone uses for
  628. 23:01perhaps simple queries and prompts, that
  629. 23:03will simply become a classic platform
  630. 23:05business model.
  631. 23:08You have a huge user base on the one
  632. 23:10hand, so you'll onboard all the
  633. 23:12advertisers on the other side and that
  634. 23:14will blow up, it will scale endlessly.
  635. 23:20It will be annoying as a user, but I
  636. 23:22mean, if you go to Facebook or Instagram
  637. 23:24or wherever, YouTube, then yes, you're
  638. 23:27already used to having ads displayed in
  639. 23:29front of your content.
  640. 23:33And in my opinion, these commodities
  641. 23:35will also be able to finance themselves
  642. 23:37. And point number two, what are these
  643. 23:40TPUs that Google and Broadcom are
  644. 23:42building in cooperation
  645. 23:46for?
  646. 23:51In my opinion, they will be used for all
  647. 23:54these special applications, i.e. for
  648. 23:56this application that you then need so
  649. 23:58that you have AI models that you can
  650. 24:00also feed into SAP, for example, so that
  651. 24:02the SAP developers can then make their
  652. 24:05ecosystem even better.
  653. 24:10And you will probably then simply
  654. 24:12receive subscription revenues for that,
  655. 24:14right?
  656. 24:17>> So yes, I think all of this will happen
  657. 24:19, but what is often underestimated is
  658. 24:21that there is a lot of concern about
  659. 24:23whether it will happen.
  660. 24:27And there are no signs that we should
  661. 24:28question whether.
  662. 24:31But the question of when is fundamental
  663. 24:33.
  664. 24:35I looked at the capital market day or
  665. 24:37the information from Oracle, which is
  666. 24:39known to be quite heavily indebted and
  667. 24:41has used a lot of external capital.
  668. 24:46And, um, I have a very rough estimate,
  669. 24:48yeah, that's actually the only thing
  670. 24:51that's still being done today, but
  671. 24:53they're at $ 127.
  672. 24:57And if you say that everything you're
  673. 24:59describing is going to happen, then the
  674. 25:01question is, will it take five or six
  675. 25:03years, which would be nothing in terms
  676. 25:05of technology, then Oracle is probably
  677. 25:07overvalued at its current level.
  678. 25:11Or will it take two years, then Oracle
  679. 25:13might be worth twice as much or more.
  680. 25:17So when that happens is crucial. That's
  681. 25:19also absolutely crucial for the Neo
  682. 25:21Cloud.
  683. 25:22That would be different with zero
  684. 25:24interest rates, because then you don't
  685. 25:25have to discount. But we won't see that
  686. 25:28anymore.
  687. 25:29And that's why you not only have to get
  688. 25:31a feel for whether it's going to happen
  689. 25:33, yes, but I can only repeat it, when it
  690. 25:35will happen.
  691. 25:37And maybe I can also, so that we don't
  692. 25:39push it aside, um, for me it's about I
  693. 25:41don't want to be invested in the model
  694. 25:44manufacturers in the long term.
  695. 25:49It's really about, um, yes, it would be
  696. 25:51a bit, I mean: Of course, it's taxed
  697. 25:53money and it shouldn't get any less and
  698. 25:56so on, but um, I can imagine that
  699. 25:58because the mood is so bad, if it were,
  700. 26:00if it were, then the opening will be
  701. 26:03slightly more positive than many people
  702. 26:05think today.
  703. 26:09Um, that would be different if we came
  704. 26:11out in the phase of euphoria, but then
  705. 26:13it would probably also be valued at, um
  706. 26:15, two trillion. We're not there anymore
  707. 26:17. In short, keep in mind the disclaimer
  708. 26:19.
  709. 26:20I'll probably sell again within the
  710. 26:22first 72 hours of trading, one way or
  711. 26:24another.
  712. 26:26So this is not a recommendation to
  713. 26:28follow my example. That's
  714. 26:30>> then this
  715. 26:31>> 10%drop that comes all of a sudden. Out
  716. 26:33of nowhere.
  717. 26:34>> Yes, unfortunately I don't have a
  718. 26:36holding period, because that would have
  719. 26:38meant I might have gotten in two or
  720. 26:40three years ago.
  721. 26:41Then it wouldn't matter to me at all
  722. 26:43whether they come out here.
  723. 26:44It would be a deal either way.
  724. 26:46>> But there was also a 10%drop with Nvidia
  725. 26:49, Sebastian.
  726. 26:50>> Even more.
  727. 26:52>> Namely, shortly before Jensen Huang's
  728. 26:54resignation actually became apparent,
  729. 26:56because the stock had come under
  730. 26:58pressure again beforehand.
  731. 27:04As everyone knows, yes, several affairs
  732. 27:07: on the one hand with a suspected
  733. 27:09Chinese spy, then almost worse with a
  734. 27:11direct subordinate and then apparently
  735. 27:14with a, uh, third person, uh, also from
  736. 27:17the company.
  737. 27:21And yes, Jensen Huang is no longer the
  738. 27:23CEO of Nvidia, the mastermind of Nvidia
  739. 27:26, the rock star of AI trading and you
  740. 27:28still bought it, right?
  741. 27:31>> Right.
  742. 27:32Jensen Huang, who I have also studied a
  743. 27:34lot, there are also good books about him
  744. 27:36. Now, of course, his departure is a bit
  745. 27:39bitter, I have to say, because he made
  746. 27:41the company big. He also tailored it to
  747. 27:43his own needs.
  748. 27:44I will come to the perhaps positive
  749. 27:46aspects that can still be derived from
  750. 27:48this in a moment, but Nvidia has not
  751. 27:50fundamentally changed anything as a
  752. 27:52group.
  753. 27:53So we don't have any adjusted earnings
  754. 27:55expectations or lower growth or anything
  755. 27:57like that, quite the opposite. Nvidia is
  756. 28:00valued at a level.
  757. 28:03So if I just quickly go into that again
  758. 28:05, we currently have a P/E ratio of 16
  759. 28:07for the current fiscal year and 11 for
  760. 28:10the next fiscal year.
  761. 28:13That's unbelievable, and that's with
  762. 28:15growth rates of 90%in sales this fiscal
  763. 28:18year and almost 70%expected next year.
  764. 28:20So we don't have any adjustments there.
  765. 28:22That means we have an inherent personnel
  766. 28:25problem at the top, but yes, Jensen
  767. 28:27Huang was always the rock star and
  768. 28:29Nvidia made the company big, also
  769. 28:31accelerating the development cycles to
  770. 28:33twice a year.
  771. 28:34That's a problem, but demand for Nvidia
  772. 28:37chips continues to be gigantic and I
  773. 28:39remember, I think it was Warren Buffet,
  774. 28:41who said that he likes to buy companies
  775. 28:43that even a fool could run.
  776. 28:45I would say that in the current climate
  777. 28:47, even a fool–and I don't want to say
  778. 28:50that interim CEO Colette Kress is a fool
  779. 28:52here, but even the most inexperienced
  780. 28:54person could probably manage Nvidia at
  781. 28:56the moment, simply because sales are
  782. 28:58secure. If you don't screw up now and
  783. 29:00the development cycles go back, and
  784. 29:01there is also middle management,
  785. 29:03although I'll come back to that in a
  786. 29:04moment, that means you can already lead
  787. 29:06Nvidia. The question is, is the next CEO
  788. 29:08well-suited to the job after all the
  789. 29:11phases or tearful phases where Jensen
  790. 29:13Wang got his company through? So from
  791. 29:15that perspective, I have complete
  792. 29:16confidence in the company. I think it's
  793. 29:18valued really attractively. I also
  794. 29:20bought more shares this week with the
  795. 29:22Hell Investiert Pro community,
  796. 29:23increasing my position by about a fifth
  797. 29:25, so the long-term. I see this as a good
  798. 29:28opportunity to get into Nvidia. But–and
  799. 29:30as I said, this is important–what
  800. 29:32happens in the future if things don't go
  801. 29:34as well as they are now? And perhaps as
  802. 29:36a side note: Jensen Wang has also
  803. 29:38repeatedly pushed the investments that
  804. 29:40you mentioned earlier, Timo. So they
  805. 29:42will continue to invest in other
  806. 29:44manufacturers, now also in Entropic,
  807. 29:46Nvidia is also supposed to invest 5
  808. 29:48billion, will that happen? However, I
  809. 29:51think that is currently more of a
  810. 29:52problem on a smaller scale. So in short
  811. 29:54, I support Nvidia and the departure of
  812. 29:56Jensen Wang could also improve the
  813. 29:57governance, i.e. the entire corporate
  814. 29:59management and structure, because Nvidia
  815. 30:01was already completely tailored to him.
  816. 30:03So a bit more transparency and a few
  817. 30:05more decision-making levels, so that the
  818. 30:07group becomes more stable and is not
  819. 30:09just dependent on one person, that can
  820. 30:11only be positive for shareholders, even
  821. 30:13if things might get a bit choppy in the
  822. 30:15short term.
  823. 30:18>> Lars, you didn't buy, correct me if I'm
  824. 30:21wrong.
  825. 30:23What are your expectations for the
  826. 30:25figures on Tuesday?
  827. 30:29>> Uh, yes, those are the official figures
  828. 30:31.
  829. 30:33Well, I have to say that you can't say
  830. 30:36from a value that is in focus, well, a
  831. 30:38margin of 75%and 108 billion is expected
  832. 30:41, I don't know, and then I go through it
  833. 30:44and say, oh, I think it will be 109. So
  834. 30:46yes, I'm much more interested in the
  835. 30:49reaction and I would like to, would like
  836. 30:51to contradict Sebastian even more
  837. 30:53vehemently, but I can't.
  838. 30:58What I certainly miss is that we are not
  839. 31:00making anything positive out of the fact
  840. 31:02that Jensen Huang has to step down.
  841. 31:06It is crystal clear that it is negative
  842. 31:08for the company and because his
  843. 31:09influence is enormous, it may be that in
  844. 31:12five years it turns out that a new
  845. 31:14successor is even better.
  846. 31:17Anything could happen. But it could also
  847. 31:19be that by then we will have flying
  848. 31:21shoes on our feet. So in that respect,
  849. 31:23of course, anything is possible, but for
  850. 31:25me it is only logical that Nvidia is
  851. 31:27getting a discount here.
  852. 31:30And by the way, I've read in our app now
  853. 31:32and then, yes, the board is of course
  854. 31:34really stupid, why are they removing him
  855. 31:37and so on.
  856. 31:38So they couldn't act any differently now
  857. 31:40, because otherwise the board would of
  858. 31:42course be committing a crime.
  859. 31:44That's why the other tech giants have
  860. 31:46all built structures where they can no
  861. 31:48longer be thrown out.
  862. 31:49So they are stronger than the board.
  863. 31:52But that's not the healthy way, I have
  864. 31:54to say.
  865. 31:58And maybe it will be possible-but
  866. 32:00certainly only after a certain pause-to
  867. 32:02put Jensen Huang in a position, unless
  868. 32:04he hires himself like at SpaceX (as I
  869. 32:06just read, or I think it was more of a
  870. 32:09rumor), to put him in a position where
  871. 32:11he still has an influence on the company
  872. 32:13, but no longer as CEO.
  873. 32:16So I made a video about it yesterday.
  874. 32:23My base case, probability 50 to 55%, is
  875. 32:25that the stock is too cheap at this
  876. 32:27level over a 12 to 24-month period, and
  877. 32:30this valuation already includes the fact
  878. 32:33that the AI trend is slowing down, and
  879. 32:35that has little to do with the company's
  880. 32:37management.
  881. 32:43It does include the possibility that the
  882. 32:45margin may weaken slightly.
  883. 32:49However, it also includes the
  884. 32:50possibility that there will be a
  885. 32:52successor to this position who would not
  886. 32:54completely disappoint the market.
  887. 32:58But of course everyone knows that, and
  888. 33:01Jensen Huang himself still has 3.5%of
  889. 33:03Nvidia.
  890. 33:05That doesn't sound like much when you
  891. 33:06convert it, but there is still a bit of
  892. 33:08skin in the game.
  893. 33:14So I'm also sure that he's not sitting
  894. 33:16at home with his phone numbers canceled
  895. 33:18, but that he no longer has an official
  896. 33:20position, and I don't see him having any
  897. 33:22influence on the company at all.
  898. 33:27But yeah.
  899. 33:28That's my expectation.
  900. 33:33Yes, let me just add something, because
  901. 33:35-and that's why I can't disagree so
  902. 33:37loudly, even though the negative
  903. 33:39consequences weren't discussed loudly
  904. 33:41enough-but my conclusion was exactly
  905. 33:43this: If I were a long-term investor in
  906. 33:45Nvidia, that is, someone who regularly
  907. 33:47buys a share, then I would (and
  908. 33:49Sebastian and I haven't spoken to each
  909. 33:51other) then I would buy a share right
  910. 33:53now and I would even do it before the
  911. 33:55numbers are released, because I wouldn't
  912. 33:57care about the reaction.
  913. 34:01Sebastian won't sell if they hit exactly
  914. 34:04$ 140.
  915. 34:06He's proven that elsewhere.
  916. 34:09So if they fall to $ 140, then it
  917. 34:11doesn't matter to the long-term investor
  918. 34:13.
  919. 34:14And from that perspective, yes, I can
  920. 34:16understand why long-term investors are
  921. 34:19buying here.
  922. 34:20>> I just want to briefly note on the
  923. 34:22positive side that it doesn't come
  924. 34:23across the wrong way. I would prefer
  925. 34:25that nothing had changed and Jensen
  926. 34:27Huang remained at the top. But of course
  927. 34:29, as a personality cult, it poses a risk
  928. 34:32to shareholders, which we have seen.
  929. 34:34That is why, if this risk is now removed
  930. 34:36–and I completely agree with you, Lars–
  931. 34:38: Jensen Huang will be pulling the
  932. 34:39strings in the background, whether as a
  933. 34:41consultant or somewhere else, we will
  934. 34:43find something. At least that is a risk
  935. 34:45that has now been eliminated and if he
  936. 34:47somehow remains on board, I think that
  937. 34:49is positive and, above all, Nvidia
  938. 34:51simply has the opportunity, I think, to
  939. 34:53become broader and less dependent on
  940. 34:55individuals, similar to Apple. Steve
  941. 34:57Jobs passed away at that time, which
  942. 34:59means he could no longer exert any
  943. 35:01influence, logically. But people also
  944. 35:02thought that this would continue and Tim
  945. 35:04Cook then created values where no one
  946. 35:06would have expected that this was
  947. 35:08possible. So something like that doesn't
  948. 35:10have to be the end.
  949. 35:12>> Hmm.
  950. 35:15Nevertheless, I don't think Peak Nvidia
  951. 35:18is stupid at all, because if you assume
  952. 35:20that this trend is slowing down, i.e.
  953. 35:23the AI trend, the AI trade, excuse me,
  954. 35:25and now the negative sentiment is coming
  955. 35:28in.
  956. 35:32So the company is probably very busy
  957. 35:34with itself right now and trying to get
  958. 35:36itself back on track, which could also
  959. 35:38lead to uncertainty among customers.
  960. 35:44It's not as if there is no competition,
  961. 35:46but there are certainly competitors who
  962. 35:48are simply not as powerful or perhaps
  963. 35:50not as advanced technologically in some
  964. 35:52areas; but it's not as if the players
  965. 35:54can't catch up, as if they are just the
  966. 35:56snobs who are still on the market.
  967. 36:00So peak Nvidia?
  968. 36:02>> Well, I'll keep it short, that will only
  969. 36:04be a problem and that's what I meant. At
  970. 36:06the moment, Nvidia can lead to stupidity
  971. 36:08because everyone wants these products.
  972. 36:10But now of course it depends on whether
  973. 36:12they are only busy with themselves for
  974. 36:14months and years, deliveries decrease,
  975. 36:16development cycles decrease, the chips
  976. 36:19get worse, the others catch up. So now
  977. 36:21it's simply important that the machine
  978. 36:23continues to run as before and that's
  979. 36:25exactly what I said. It was tailored to
  980. 36:27Jensen Wang. That means you don't have a
  981. 36:29governance structure like in other
  982. 36:30companies.
  983. 36:32It has to be proven that it works
  984. 36:33without him.
  985. 36:34Then I don't see any problem with that.
  986. 36:36>> Yes, because exactly.
  987. 36:40He had, according to his own statements
  988. 36:42, thousands of customer contacts per
  989. 36:44year, so he met and spoke to several
  990. 36:46thousand people, high-ranking managers
  991. 36:48from the customer companies every year.
  992. 36:55So that means he was also Mr. Salesman,
  993. 36:58I think, and on the other hand, I think
  994. 37:00he had, the numbers vary between 35 and
  995. 37:0365 direct reports, i.e. employees who
  996. 37:05were directly subordinates and who
  997. 37:07reported directly to him.
  998. 37:12That's pretty crazy, isn't it? There's
  999. 37:14endless power knowledge there.
  1000. 37:17>> Yes, he took full care of the group.
  1001. 37:20You have to say that, but of course you
  1002. 37:22can
  1003. 37:24put it this way, in short, it's
  1004. 37:25speculation. We now have to see how
  1005. 37:28Nvidia reacts, who will be the new CEO,
  1006. 37:30how do you simply set up these
  1007. 37:32intermediate levels? Will there then be
  1008. 37:34a CEO who can continue to do this and is
  1009. 37:35able to do so, or will they create
  1010. 37:37another management level and that will
  1011. 37:38be outsourced or relocated? But at the
  1012. 37:40moment I can't imagine that this would
  1013. 37:42somehow be the end or peak of Nvidia,
  1014. 37:43because they are still far too much of a
  1015. 37:45leader for that. In that case they would
  1016. 37:47really screw it up. And I don't think
  1017. 37:49Jensen Huang will allow that either. He
  1018. 37:51can continue to meet with customers, he
  1019. 37:53can continue to sell. I don't think
  1020. 37:55he'll just sit at home and just look out
  1021. 37:56the window.
  1022. 37:57>> And to quickly mention the risk, all
  1023. 37:59tech companies have that.
  1024. 38:02Why is there-I made a video about it
  1025. 38:04once-name me a major tech company that
  1026. 38:06has dominated for 30, 40, 50 years.
  1027. 38:09It's normal in this sector that you then
  1028. 38:11become the leader, that you are
  1029. 38:13overtaken.
  1030. 38:14Then came the objection of IBM, Apple.
  1031. 38:19Okay, I would call them tech companies,
  1032. 38:21Apple with its service business, but of
  1033. 38:23course there are an incredible number of
  1034. 38:26examples of technological advances that
  1035. 38:28cannot be maintained for decades.
  1036. 38:31But that is not a problem specific to
  1037. 38:33Nvidia.
  1038. 38:34>> Okay, good.
  1039. 38:38So I think that's all that needs to be
  1040. 38:40said at this point and the rest would be
  1041. 38:42just gossip and yes, one more thing: I
  1042. 38:44think the malice that is being poured
  1043. 38:46out on this man is a bit exaggerated.
  1044. 38:56Of course, he is one of the richest
  1045. 38:58people in the world and that the media
  1046. 39:00treats him a bit differently than us
  1047. 39:02normal people, I think that's clear, and
  1048. 39:04I think that's fair given his social
  1049. 39:06relevance, but yeah.
  1050. 39:16I think it's a bit sad, especially the
  1051. 39:18internet culture in some social media.
  1052. 39:24That's not relevant to the topic at hand
  1053. 39:26, but just a little side note.
  1054. 39:29Let's move on to the Dip of the Week.
  1055. 39:33I'd like to say a little bit about that
  1056. 39:35, because the Dip of the Week is
  1057. 39:36actually a format that we came up with,
  1058. 39:38which is then
  1059. 39:43regularly released in the By the Dip app
  1060. 39:45on Monday evenings.
  1061. 39:47In this case, a stock of mine is
  1062. 39:49examined in detail, which is either in a
  1063. 39:51short-term dip, in a medium-term
  1064. 39:53correction, or even in a long-term crash
  1065. 39:56in the bear market.
  1066. 39:59And there I do a stock analysis, we look
  1067. 40:01at the company, fundamentally, but also
  1068. 40:03the price behavior, i.e. the chart
  1069. 40:05technique, and then I also give a
  1070. 40:07relatively clear conclusion.
  1071. 40:11And if you like that, you can of course
  1072. 40:13register directly in our app, completely
  1073. 40:15free of charge, it's a free area.
  1074. 40:21Or, we also have another page in the
  1075. 40:22show notes, where you can take another
  1076. 40:25look at the format.
  1077. 40:28Also, if you're thinking, "No,
  1078. 40:30downloading an app is a bit too far for
  1079. 40:32me," you can also simply
  1080. 40:35sign up for a free email distribution
  1081. 40:37list that we've created specifically for
  1082. 40:39this purpose. In that case, you'll
  1083. 40:41receive this stock analysis delivered to
  1084. 40:44your inbox every Tuesday morning at 7
  1085. 40:46a.m., completely free of charge.
  1086. 40:48Feel free to use it. And by the way,
  1087. 40:50there's no risk involved. If you don't
  1088. 40:52like it anymore, you can simply
  1089. 40:54unsubscribe.
  1090. 40:55And now we're going to go through our
  1091. 40:57dips of the week.
  1092. 40:58I've just talked a lot.
  1093. 41:01I can already announce that I'm going to
  1094. 41:03talk about Caterpillar, but what about
  1095. 41:04you?
  1096. 41:06Sebastian, what did you bring? Let me
  1097. 41:08start
  1098. 41:09.
  1099. 41:09>> I took a look at GE Vernova.
  1100. 41:12We had them in a duel against Siemens
  1101. 41:14Energy a few weeks ago. At that time, my
  1102. 41:16conclusion was that GE Vernova is an
  1103. 41:18established corporation that makes a lot
  1104. 41:19of money with gas turbines, and that's
  1105. 41:21where Siemens Energy wants to go, which
  1106. 41:23means that Siemens Energy still has a
  1107. 41:25lot more potential to catch up. In the
  1108. 41:27classic stock duel, Siemens Energy was
  1109. 41:29the stock with a slightly better
  1110. 41:30risk-reward ratio for me, because it has
  1111. 41:32a lot of upside and there's also a
  1112. 41:34turnaround potential somewhere. I found
  1113. 41:36that interesting, but if I now simply
  1114. 41:38look at stocks that are interesting for
  1115. 41:40follow-up purchases or in general, I
  1116. 41:42find GE Vernova interesting because it
  1117. 41:44has fallen by about 20%in the last four
  1118. 41:46weeks and at the same time, it has been
  1119. 41:48sold off, also with the sell-off in
  1120. 41:50chips and tech and everything else. We
  1121. 41:52are at this point and perhaps we can go
  1122. 41:54into that in more detail in another
  1123. 41:55episode. I assume that next year and
  1124. 41:57beyond will be the year of the "bring
  1125. 41:59your own power plant". This means that
  1126. 42:02when the corporations, the hyperscalers
  1127. 42:04, build, they bring their own power
  1128. 42:05plants with them because, and we are
  1129. 42:07seeing this now, and this has already
  1130. 42:09partly caused this crash, the
  1131. 42:11electricity costs for the local
  1132. 42:12population are simply exploding. So they
  1133. 42:14have to provide the electricity supply
  1134. 42:16on their sites accordingly, on the one
  1135. 42:18hand through solar, on the other hand
  1136. 42:20perhaps through wind, but gas is one of
  1137. 42:22the cheapest energy sources, especially
  1138. 42:24in the US, also to secure a kind of base
  1139. 42:26load and I am sure that the GE Vernova
  1140. 42:28is in a great position. So I think this
  1141. 42:30is the most interesting dip of the week
  1142. 42:31and I find the entire sector really
  1143. 42:33interesting and I am currently looking
  1144. 42:35at some stocks and also ETFs because in
  1145. 42:37my opinion this will be the really big
  1146. 42:39phase, this electricity supply,
  1147. 42:40especially now that the Democrats are
  1148. 42:42putting so much pressure on.
  1149. 42:44>> By the way, because it is called the dip
  1150. 42:46of the week, it is not now, the stock
  1151. 42:47does not necessarily have to have dipped
  1152. 42:49only this week. I think I already
  1153. 42:51indirectly mentioned that earlier, so
  1154. 42:53people don't want to say, "Yes, but
  1155. 42:55Caterpillar didn't dip this week."
  1156. 42:59That's right.
  1157. 43:00They actually bounced a bit.
  1158. 43:03They went from a peak of $ 1,100 to $
  1159. 43:07550 in a relatively short period of time
  1160. 43:10, I have to say, within a few months, so
  1161. 43:13a 50%crash.
  1162. 43:16However, Caterpillar hasn't just been a
  1163. 43:18construction machinery manufacturer for
  1164. 43:20a long time now.
  1165. 43:22You're probably all familiar with those
  1166. 43:24yellow excavators with the Cat logo.
  1167. 43:28That's because they have a division,
  1168. 43:30Power and Energy.
  1169. 43:34They build generators and gas turbines,
  1170. 43:36including for all the AI data centers,
  1171. 43:38and that's now actually the division
  1172. 43:40with the highest sales, the
  1173. 43:42fastest-growing division, and the
  1174. 43:44division that earns the highest margin,
  1175. 43:46at 23%.
  1176. 43:50They also have a very adequate backlog,
  1177. 43:53i.e. an order backlog that in some cases
  1178. 43:55–at least for the large turbines–
  1179. 43:57extends until 2030. So, they are
  1180. 43:59completely booked up.
  1181. 44:02However, they have nevertheless fallen
  1182. 44:04very sharply, because, yes, you have to
  1183. 44:07say, they had a P/E ratio of 40 at times
  1184. 44:10, so they were always priced at around
  1185. 44:1217 on average over the long term.
  1186. 44:16Now, after the 50%crash, they have just
  1187. 44:19about reached this level again, at least
  1188. 44:22with the forward P/E ratio, i.e. at this
  1189. 44:24level of 17 for 2027. So, you can't
  1190. 44:27really talk about this as a value
  1191. 44:29bargain or anything like that.
  1192. 44:33At least if you only look at the
  1193. 44:34long-term average.
  1194. 44:38But you have to see that due to this
  1195. 44:40rapidly growing division, which is also
  1196. 44:42somehow anchored in the AI trade,
  1197. 44:44Caterpillar may have simply been
  1198. 44:46punished a bit too harshly and that this
  1199. 44:48slight premium may even be justified.
  1200. 44:54I've added the conclusion, so as a
  1201. 44:56long-term investor I wouldn't do it.
  1202. 44:59I think you can rather construct the
  1203. 45:00opportunity for a short-term trade here
  1204. 45:03, so with a stop and a target price.
  1205. 45:05um, yes, exactly.
  1206. 45:08The exact conclusion is then included in
  1207. 45:10our app and yes, at this point I'll hand
  1208. 45:13over to you, Lars.
  1209. 45:16>> Yes, thank you very much.
  1210. 45:19So the stock I brought with me, um, held
  1211. 45:21up quite well this week, but has fallen
  1212. 45:24significantly from its high.
  1213. 45:27>> Finally worse than RWE. Thanks, Lars.
  1214. 45:30>> Finally about RWE.
  1215. 45:33>> Your information was right.
  1216. 45:35>> Yes, that's true, that's true, but the
  1217. 45:37stock I'm talking about today will also
  1218. 45:39rise again. But maybe RWE too, that
  1219. 45:42could well be the case.
  1220. 45:44>> Very well countered.
  1221. 45:46>> So, um,
  1222. 45:48>> I brought Prysmian, Nvidia.
  1223. 45:50I brought Nvidia for us.
  1224. 45:51>> Nvidia
  1225. 45:52>> No, no, Prysmian, okay.
  1226. 45:54>> I brought Prysmian. Most people know
  1227. 45:56that I am invested there with the
  1228. 45:58readers of the Renditespezialisten and
  1229. 46:00they may forgive me for also divulging
  1230. 46:02here that this was the purchase in the
  1231. 46:04future portfolio this month.
  1232. 46:06This is the long-term portfolio and
  1233. 46:08accordingly, which is very important for
  1234. 46:11the classification, this is not a
  1235. 46:13statement, here the share will
  1236. 46:15definitely stabilize at 109 euros or in
  1237. 46:17our case it was 105 euros.
  1238. 46:20Um, if it falls to 90 or 80 or 70 many,
  1239. 46:23because the worries about the end of the
  1240. 46:25AI trade, yes, they make the cables,
  1241. 46:28namely the big thick cables, um, because
  1242. 46:30they ensure this fear that the share
  1243. 46:32will fall even further, then I am
  1244. 46:34totally fine with that.
  1245. 46:38I'm not interested in where Prysmian
  1246. 46:40will be in three months, but in three,
  1247. 46:42five, and six years. And, uh, the stock
  1248. 46:44is fast-forward, because I've talked
  1249. 46:46about it a few times. The current P/E
  1250. 46:49ratio of 23 will fall to 14 in the next
  1251. 46:52two years if sales and profits develop
  1252. 46:54as expected.
  1253. 46:57And, uh, we also have to look at where
  1254. 47:00the sales come from.
  1255. 47:02So 40%from industry, 19%from power
  1256. 47:05distribution networks, 15%from
  1257. 47:07high-voltage and transmission, which are
  1258. 47:10essentially these data centers.
  1259. 47:15But you don't just have to go through
  1260. 47:17the USA, you can also go through Italy
  1261. 47:19and our countries and you can see that
  1262. 47:21infrastructure will have to be
  1263. 47:22modernized and expanded anyway.
  1264. 47:27And I like that, this is the best seller
  1265. 47:29of the best sellers, and from that
  1266. 47:31perspective I was happy to buy more here
  1267. 47:33.
  1268. 47:34>> Okay.
  1269. 47:36And now we still have Core Weave, Iron,
  1270. 47:39and Nebius on the table.
  1271. 47:42Who wants to start?
  1272. 47:44Lars, can you just continue?
  1273. 47:48>> Yes, that's the lower end of this order
  1274. 47:51chain and they've
  1275. 47:56lost a lot from the high, and
  1276. 48:01rightly so, we'll tell you about that
  1277. 48:04later.
  1278. 48:05So I'll leave the Core Weave sales to
  1279. 48:07Sebastian, because he prepared them.
  1280. 48:14I recently described Nebius as the value
  1281. 48:16that could perhaps still be described as
  1282. 48:18the most conservative, because we also
  1283. 48:20looked at how they are financed and do
  1284. 48:22they have their own expenses?
  1285. 48:27So they also have to place orders.
  1286. 48:31Do they already have all the financing
  1287. 48:33secured or do they have to react to
  1288. 48:35flexible prices?
  1289. 48:38And from that perspective, an Iron is
  1290. 48:40not like a Core Weave and a Core Weave
  1291. 48:42is not like a Nebius.
  1292. 48:44What they all have in common is that we
  1293. 48:46will only know in retrospect whether
  1294. 48:48they are dirt cheap or horrendously
  1295. 48:50expensive at this level.
  1296. 48:58One, and my expectation is that someone
  1297. 49:00who is aware of the risk and who
  1298. 49:02classifies this as hardcore speculation
  1299. 49:04will get very attractive risk-reward
  1300. 49:06ratios at this level, because some of
  1301. 49:08them have come under so much pressure
  1302. 49:10and most of these companies are now
  1303. 49:12acting as if growth is not leveling off
  1304. 49:14but is almost collapsing.
  1305. 49:21Timing is everything here. I would never
  1306. 49:23buy a stock like this, no matter which
  1307. 49:25one you choose, without a stop.
  1308. 49:29I'm happy to justify it using chart
  1309. 49:31technology, but even if it's hardcore
  1310. 49:33speculation, it's not uninteresting at
  1311. 49:35this level.
  1312. 49:38>> Sebastian Nebius surprised us very
  1313. 49:40positively two quarters ago.
  1314. 49:44Now they've
  1315. 49:46surprised us negatively again. I think
  1316. 49:48they've also made losses. You have a
  1317. 49:50better idea of the figures. Correvio, on
  1318. 49:52the other hand, was very positive, right
  1319. 49:54? So what are the figures?
  1320. 49:56>> I even think that both of them,
  1321. 49:59if I pick up on Lars's point again, are
  1322. 50:01not only being praised as if growth is
  1323. 50:04going away, but also as if there is no
  1324. 50:06hope that they will ever make money.
  1325. 50:08And there is a valid point in that.
  1326. 50:12So if I go into Nebius, there was a real
  1327. 50:14rumble within the share on Thursday,
  1328. 50:16because the forecast was withdrawn
  1329. 50:18somewhat and then there is also the
  1330. 50:20annual revenue run rate, i.e. the
  1331. 50:22annualized sales figure.
  1332. 50:24They wanted to reach 7 billion. That
  1333. 50:27will now be delayed by two months.
  1334. 50:29Personally, I wouldn't care if it were
  1335. 50:31delayed by two months, but the market is
  1336. 50:33currently highly sensitive and emotional
  1337. 50:35and that is why Nebius was severely
  1338. 50:37punished. What is interesting, however,
  1339. 50:39is that Nebius has now managed to ensure
  1340. 50:42that the advance payments from customers
  1341. 50:44cover the capex. That means that they
  1342. 50:46don't have to invest any new money, but
  1343. 50:48can instead make the capex from the
  1344. 50:49payments from customers. Of course, that
  1345. 50:51provides significant relief on the
  1346. 50:53financing side. I think that's something
  1347. 50:55to look at, in addition to all the
  1348. 50:56criticism. And with Correvio, which is
  1349. 50:58of course the most speculative, we had
  1350. 51:00sales growth of almost 70%. We had a
  1351. 51:02profit of two and a half, well, what's
  1352. 51:05better, EBITDA, let's say, from two
  1353. 51:08billion, over two billion, 62%margin,
  1354. 51:10132.5 billion in the backlog, a sales
  1355. 51:13forecast raised to 13 to 13.3 billion.
  1356. 51:16And now with Criteo, we also had a
  1357. 51:19positive cash flow of two billion. The
  1358. 51:21problem on the other side is that all of
  1359. 51:23that is currently being used up for
  1360. 51:24interest payments. At the same time,
  1361. 51:26however, Criteo has said that they will
  1362. 51:28have to raise less capital in the highly
  1363. 51:30speculative bonds, which are also the
  1364. 51:32most expensive, because the financing
  1365. 51:34structure is simply better and it has
  1366. 51:36also improved by 2.5%compared to the
  1367. 51:38previous year, so they are paying less.
  1368. 51:40Now it's simply a matter of finding out
  1369. 51:42, and this is precisely the transition:
  1370. 51:44Are they making money with this now? Can
  1371. 51:46these cash flows be converted into real
  1372. 51:49profits later on? Or will it all
  1373. 51:51continue to simply go up because more
  1374. 51:52and more financing is needed, because
  1375. 51:54they have to buy more and more expensive
  1376. 51:56chips from Nvidia, because they have to
  1377. 51:57write these chips off again and again?
  1378. 51:59So they are exactly at this transition
  1379. 52:01point. But I have to say, if I draw a
  1380. 52:03line under it: What we saw, this
  1381. 52:05massacre with the neoclouds, I found
  1382. 52:07this week's figures to be generally
  1383. 52:09positive and also more pro-AI trade,
  1384. 52:11rather than saying that these are the
  1385. 52:13canaries that have fallen over and now
  1386. 52:15the lights are completely out. So I see
  1387. 52:17more of the positive aspects here, but
  1388. 52:18the risks outweigh the negative ones. I
  1389. 52:20don't own any of the shares, so I think
  1390. 52:22I can look at it relatively objectively
  1391. 52:24.
  1392. 52:25>> Yes, I don't own any of the shares
  1393. 52:27either.
  1394. 52:29Uh, to be honest, I wouldn't want to
  1395. 52:31touch any of them right now because I
  1396. 52:33think there are better value
  1397. 52:34opportunities or at least a bit more
  1398. 52:36fundamentally substantiated ones within
  1399. 52:38this AI trade.
  1400. 52:41Uh, yes, I'm invested elsewhere and yes
  1401. 52:44, I don't miss any of these three
  1402. 52:46players right now.
  1403. 52:47But I can understand anyone who says,
  1404. 52:49yes, I want to be there with a bit of
  1405. 52:51play money.
  1406. 52:53Uh, you should just be aware that these
  1407. 52:55are definitely not buy-and-hold-forever
  1408. 52:57stocks for retirement, at least that's
  1409. 52:59what I would say at the moment. I can
  1410. 53:01understand that. Go a
  1411. 53:02>> little further than Lars and set a stop
  1412. 53:04loss of almost zero. Because these
  1413. 53:07things, I think, if the problems
  1414. 53:08continue to manifest themselves, then
  1415. 53:11they'll crash down. I don't know if
  1416. 53:12you'll get out of there before then. I
  1417. 53:14consider it so speculative that I would
  1418. 53:16say, for example, if I invest 1,000
  1419. 53:19euros in Knorr-Bremse, I risk a tenfold
  1420. 53:21increase or even zero. I would even go
  1421. 53:24that far.
  1422. 53:24>> Hmm, yes, understandable.
  1423. 53:27Okay, let's go over to Germany. There
  1424. 53:30was a lot going on there this week. Uh,
  1425. 53:32Sebastian, can you just continue? What
  1426. 53:35is your German value?
  1427. 53:36>> We have, well, my German value is
  1428. 53:38Siemens. We discussed that a long time
  1429. 53:41ago, and we have the category. That
  1430. 53:43surprised us; we always have such an
  1431. 53:45imbalance in the USA, and I thought we
  1432. 53:47would do this specifically for Germany,
  1433. 53:49and I looked at the figures for Siemens
  1434. 53:51, Siemens, Siemens, and they were
  1435. 53:53positive. So we had sales up 11 percent
  1436. 53:56, incoming orders up 58 percent, a new
  1437. 53:59record, order backlog of 143 billion, of
  1438. 54:02which 7 billion were data center
  1439. 54:04projects alone. What I would like to
  1440. 54:07bring to the table here, and this
  1441. 54:09surprised me, is that Siemens is
  1442. 54:10increasingly delving into the AI trade
  1443. 54:12and profiting from it. Now, of course,
  1444. 54:14for those who say that AI, as we have
  1445. 54:16discussed enough now, will work out well
  1446. 54:18and whether money can be made, then it
  1447. 54:20is of course not necessarily a good idea
  1448. 54:22to go to Siemens. But for those who
  1449. 54:23believe in the AI trade and are looking
  1450. 54:25for a German company, I have to say that
  1451. 54:27Siemens is profiting more and more.
  1452. 54:29There are certainly others, but I found
  1453. 54:31the figures from Siemens this week to be
  1454. 54:33really, really positive, especially
  1455. 54:35because orders in the data center sector
  1456. 54:37have doubled. So there is really strong
  1457. 54:39growth there.
  1458. 54:41Sorry for the short interruption, but I
  1459. 54:43have a small request on my own behalf.
  1460. 54:45This appeal goes out to everyone who has
  1461. 54:47not yet subscribed to this channel on
  1462. 54:49YouTube.
  1463. 54:52We want to reach our target of 35,000
  1464. 54:54subscribers in the next few weeks, and
  1465. 54:56to do that, we need the help of those
  1466. 54:58who are not yet part of the community.
  1467. 55:02Become part of the stock market
  1468. 55:03newsletter community on YouTube, leave a
  1469. 55:06subscription, click on all, activate the
  1470. 55:08bell, and then I would like to say thank
  1471. 55:10you from Lars, Timo, and of course me.
  1472. 55:13And now let's continue.
  1473. 55:15>> Okay. Lars, what have you brought with
  1474. 55:17you?
  1475. 55:18>> Uh, yes, today I have a repetition of
  1476. 55:20the, I don't want to say warning, but a
  1477. 55:23personal, rather skeptical assessment.
  1478. 55:25Uh, we talked about cyber security here
  1479. 55:28, uh, Secunet. That is basically the
  1480. 55:31only relevant German cyber security
  1481. 55:33company. Uh, they have come under
  1482. 55:34considerable pressure. That is mainly
  1483. 55:36because they are investing.
  1484. 55:39Sales weren't all that bad, but my
  1485. 55:40assessment was–and it was based on an
  1486. 55:43expert telling me this at a panel with
  1487. 55:45my colleagues from Aktien Lust–that he
  1488. 55:47said: If they want to grow, then it's
  1489. 55:49basically only because German
  1490. 55:51authorities are ordering security,
  1491. 55:52cybersecurity here.
  1492. 55:56Of course, they don't order from Chinese
  1493. 55:58providers and now they don't order from
  1494. 56:00American providers either, and lo and
  1495. 56:02behold, practically all of their growth
  1496. 56:04is based on that, which also means that
  1497. 56:07German private investors, and I can't
  1498. 56:09even say what the product range is like
  1499. 56:11, prefer foreign providers.
  1500. 56:18And if you compare that growth with the
  1501. 56:20growth we see at American cybersecurity
  1502. 56:22companies, but also at companies like
  1503. 56:24Cloudflare, which are still not
  1504. 56:26profitable but are growing incredibly
  1505. 56:28fast, then I would like to reiterate at
  1506. 56:30this point: Secunet–I mean, it wouldn't
  1507. 56:33be a stock, because who knows how
  1508. 56:35quickly someone will say: "We need to
  1509. 56:37expand cybersecurity."
  1510. 56:43That's
  1511. 56:45not something I want to be in the same
  1512. 56:47boat as on the investment side. I
  1513. 56:48noticed them today, no, this week.
  1514. 56:52>> Yes, I noticed SMA Solar. Attention,
  1515. 56:55disclaimer.
  1516. 56:56I currently have a small active position
  1517. 56:58there.
  1518. 57:00They released figures and this is a
  1519. 57:02stock that is also very heavily shorted
  1520. 57:04.
  1521. 57:06I think Citadel and others had
  1522. 57:08relatively high short positions in it.
  1523. 57:11They are currently being reduced
  1524. 57:13slightly.
  1525. 57:15And the figures they released on
  1526. 57:17Thursday were actually quite good.
  1527. 57:22There was a special effect involved.
  1528. 57:24That's why EBITDA, or operating income
  1529. 57:27before interest, taxes, depreciation,
  1530. 57:30and amortization, increased
  1531. 57:32significantly.
  1532. 57:34If you factor out all these special
  1533. 57:36effects, then they were slightly down in
  1534. 57:38terms of operations, but they did have a
  1535. 57:41very strong increase in their order
  1536. 57:42backlog.
  1537. 57:45It increased by 45%compared to the
  1538. 57:47previous year.
  1539. 57:51And, in particular, there is now
  1540. 57:53probably also, among private investors,
  1541. 57:55the urge for photovoltaics, so to speak
  1542. 57:58, but the sentiment is improving, more
  1543. 58:00orders are coming in.
  1544. 58:06In the large customer business, which in
  1545. 58:08my opinion is the most important
  1546. 58:10business, there was a slight slowdown,
  1547. 58:12but they said that all the orders that
  1548. 58:14are still in the backlog will now be
  1549. 58:16called up and realized in Q4.
  1550. 58:22And ultimately, long story short, in my
  1551. 58:25opinion, SMA Solar would have deserved
  1552. 58:27to show a stronger reaction to these
  1553. 58:29figures than they have shown now.
  1554. 58:35I think Lars, you said this with me in a
  1555. 58:38webinar recently, yes, yes, the truth is
  1556. 58:40in the open.
  1557. 58:44So currently, the share is down by a
  1558. 58:46good 5%on Friday.
  1559. 58:49Um, yeah, I think it's a little, a
  1560. 58:52little too, uh, weak, but I can't find
  1561. 58:54any fundamental reason why any short
  1562. 58:56seller would have to say: "Yeah, I'm
  1563. 58:59going to build up my short position here
  1564. 59:01again very, very clearly," because A,
  1565. 59:04they're overvalued and B, um, they don't
  1566. 59:06bring anything to the, uh, well, uh,
  1567. 59:08they don't bring anything to the, uh, on
  1568. 59:11the street. The opposite is the case.
  1569. 59:14And yes, in that respect, I'll just say
  1570. 59:16, SMA Solar, I was surprised that they
  1571. 59:19didn't just get out of the mess a little
  1572. 59:21stronger.
  1573. 59:22>> Can I just say something about that
  1574. 59:24briefly?
  1575. 59:24>> No.
  1576. 59:27It's not even
  1577. 59:28>> about your trade. It's about educating
  1578. 59:30everyone who's listening and watching us
  1579. 59:32.
  1580. 59:33But you've brushed it aside. Can I do it
  1581. 59:35anyway?
  1582. 59:35>> Yes, of course.
  1583. 59:37>> Also, I just wanted to say, since you
  1584. 59:39mentioned Citadel, these short seller
  1585. 59:41positions, which always have to be
  1586. 59:42evaluated and which many people look at
  1587. 59:44, you always have to ask yourself what
  1588. 59:46the company that holds the short
  1589. 59:48position is doing. And Citadel in
  1590. 59:50particular is one of the largest market
  1591. 59:52makers in the world. That means they
  1592. 59:54have a huge options trading book where
  1593. 59:56they post prices, buy prices, sell
  1594. 59:58prices, they have positions on the book
  1595. 1:00:01. And if they have positions in SMA
  1596. 1:00:03Solar, for example, they then sell them
  1597. 1:00:05via the stock, via a short sale or a
  1598. 1:00:07short position. So, that doesn't
  1599. 1:00:09necessarily mean that they think the
  1600. 1:00:11stock is bad. They may have a completely
  1601. 1:00:13different interest in simply keeping
  1602. 1:00:14their trading book so-called
  1603. 1:00:16delta-neutral. So, I just wanted to say
  1604. 1:00:17that you first have to check who is
  1605. 1:00:19selling these short positions and maybe
  1606. 1:00:21they have a completely different
  1607. 1:00:23business and SMA Solar is not interested
  1608. 1:00:25in them at all, but only their trading
  1609. 1:00:27book. So, that's just a side note for
  1610. 1:00:29anyone who analyzes this data.
  1611. 1:00:32>> Yes, definitely good advice.
  1612. 1:00:34And so, we'll cut that out now anyway.
  1613. 1:00:38>> Fair, fair enough. Do that.
  1614. 1:00:40>> No, we're going over to Rheinmetall now
  1615. 1:00:42.
  1616. 1:00:43It's also down 50%, so it's already the
  1617. 1:00:46second stock that's crashed 50%in the
  1618. 1:00:48course of the year.
  1619. 1:00:52And we want to take a closer look,
  1620. 1:00:54because Lukas sent us the listener
  1621. 1:00:56question on The Dip.
  1622. 1:01:00Um, who has a strong opinion on
  1623. 1:01:01Rheinmetall?
  1624. 1:01:03Probably Sebastian the most, right?
  1625. 1:01:07Sure.
  1626. 1:01:08>> I don't have the stock. We've even had a
  1627. 1:01:1060%slump since the highs of 2025. We
  1628. 1:01:13recently had another weak phase because
  1629. 1:01:15the sales forecast was revised down a
  1630. 1:01:18bit. But that's simply because, and who
  1631. 1:01:20would have thought, Bundeswehr projects
  1632. 1:01:22are being delayed. Sales are simply
  1633. 1:01:24being pushed into next year. The market
  1634. 1:01:26didn't like it in the current weak phase
  1635. 1:01:28. I have to say, I'm not invested in
  1636. 1:01:29Rheinmetall right now. I'm in Lockheed
  1637. 1:01:31Martin for the long term, which has
  1638. 1:01:33performed positively again this year,
  1639. 1:01:35it's up about 26%. But it must also be
  1640. 1:01:37said that Rheinmetall had an incredible
  1641. 1:01:39run before, it was an incredibly hyped
  1642. 1:01:41stock.
  1643. 1:01:42In my opinion, the problem with
  1644. 1:01:44Rheinmetall at the moment is that we are
  1645. 1:01:46facing a complete transformation within
  1646. 1:01:48the entire defense industry. This means
  1647. 1:01:50that drones are becoming more important
  1648. 1:01:52. Rheinmetall does have certain joint
  1649. 1:01:54ventures, agreements and investments,
  1650. 1:01:56but of course they are not as large in
  1651. 1:01:58terms of margins, sales and profits as
  1652. 1:02:00if they were doing it themselves. And
  1653. 1:02:01now we are simply at the point where
  1654. 1:02:03many are asking themselves: is
  1655. 1:02:05everything now moving more towards
  1656. 1:02:06drones? Is everything moving more
  1657. 1:02:08towards space? Do we need tanks or do we
  1658. 1:02:10need tanks protected by drones? So I
  1659. 1:02:12think Rheinmetall in particular is
  1660. 1:02:14currently suffering in this regard. I
  1661. 1:02:16personally do not find the stock all
  1662. 1:02:18that exciting, but I am convinced that
  1663. 1:02:20defense is a megatrend. It will not go
  1664. 1:02:22away. I am currently looking at a few
  1665. 1:02:23ETFs in this sector. So, I am not a big
  1666. 1:02:26Rheinmetall fan, but I am convinced that
  1667. 1:02:28defense will continue to be a boom topic
  1668. 1:02:30. The question now is whether a
  1669. 1:02:32full-range supplier like Rheinmetall can
  1670. 1:02:34keep up or whether it's really moving
  1671. 1:02:35more and more towards drones, because we
  1672. 1:02:37can see the wear and tear in Ukraine.
  1673. 1:02:40Drones cost nothing but can disable a
  1674. 1:02:42tank that costs tens of millions. So
  1675. 1:02:44there's a huge upheaval in armaments at
  1676. 1:02:46the moment.
  1677. 1:02:47>> Yes, yes, yes. Absolutely good point.
  1678. 1:02:51Lars, um, apart from the fact that you,
  1679. 1:02:54as a left-wing green activist, woke
  1680. 1:02:56activist so to speak, don't think much
  1681. 1:02:59of Rheinmetall, so, sorry, that little
  1682. 1:03:02dig.
  1683. 1:03:03>> I'd like to correct that.
  1684. 1:03:05I've never, ever, ever suggested that
  1685. 1:03:08I'm green.
  1686. 1:03:10I'd be happy to be called a left-wing,
  1687. 1:03:12woke, but not green.
  1688. 1:03:14>> Okay, sorry.
  1689. 1:03:17>> That was maybe when the Greens were
  1690. 1:03:19still green, so it was actually worse,
  1691. 1:03:21but um, yeah, that's over.
  1692. 1:03:28Um, because the Greens are, and that
  1693. 1:03:30brings us to the topic, definitely also
  1694. 1:03:32in favor of orders for Rheinmetall and
  1695. 1:03:35yes, I recently read an article or it
  1696. 1:03:37was more or less a blog post that said,
  1697. 1:03:39earn so much money and so on, with sales
  1698. 1:03:42, why is the share price falling?
  1699. 1:03:48That's a rough summary.
  1700. 1:03:50And these are also operational problems
  1701. 1:03:52, right?
  1702. 1:03:54I just checked, the cash conversion was
  1703. 1:03:56at 40%and the cash conversion in
  1704. 1:03:58relation to the first quarters is zero.
  1705. 1:04:03So the cash flow was negative last time
  1706. 1:04:05.
  1707. 1:04:08Now, to be fair, you can't just blame
  1708. 1:04:10Rheinmetall for that.
  1709. 1:04:14It's because they are building up
  1710. 1:04:16considerable inventories in order to be
  1711. 1:04:18able to process orders in the future and
  1712. 1:04:20that is probably in the sector where you
  1713. 1:04:22simply buy a lot of raw materials and so
  1714. 1:04:24on.
  1715. 1:04:27They are not, yes, they are not a drone
  1716. 1:04:29company yet, but they build really heavy
  1717. 1:04:31equipment and of course you have to
  1718. 1:04:33install the raw materials beforehand and
  1719. 1:04:35I mean, I wish it were different.
  1720. 1:04:38Should I say that? No, let's leave it
  1721. 1:04:40out.
  1722. 1:04:46Even though it's true, I don't think
  1723. 1:04:48that defense spending in Europe will
  1724. 1:04:49decrease, not even if there is peace or
  1725. 1:04:51a permanent ceasefire or whatever you
  1726. 1:04:53want to call it between Ukraine and
  1727. 1:04:55Russia, and these investment commitments
  1728. 1:04:57sometimes extend over years.
  1729. 1:05:06So, if Rheinmetall can work through its
  1730. 1:05:08order backlog, then fundamentally it is
  1731. 1:05:11certainly no longer a stock that is too
  1732. 1:05:13expensive, but the rest is also just
  1733. 1:05:15psychology.
  1734. 1:05:18They look very weak in terms of chart
  1735. 1:05:20technology. In the Czech Republic, this
  1736. 1:05:23defense company corrected significantly
  1737. 1:05:25this week. That put the share price
  1738. 1:05:27under pressure again.
  1739. 1:05:29Well, and then the market probably says
  1740. 1:05:32to itself, and every active investor
  1741. 1:05:34would say this, yes, then there will
  1742. 1:05:36definitely be another hit to the
  1743. 1:05:38downside when it says, there is now
  1744. 1:05:40peace, ceasefire, whatever.
  1745. 1:05:44Then I buy and that is kind of a
  1746. 1:05:45wait-and-see for a situation that might
  1747. 1:05:48cause the share price to fall even
  1748. 1:05:49further, but fundamentally they are
  1749. 1:05:51certainly no longer expensive in view of
  1750. 1:05:53the growth expectations.
  1751. 1:05:59>> And at 800 euros, I know they are below
  1752. 1:06:02that, so they are currently at 763 or
  1753. 1:06:05744 euros rounded up in real time.
  1754. 1:06:10So at 800 euros there would have been at
  1755. 1:06:12least some major support.
  1756. 1:06:15Yes, that was right, but okay, this
  1757. 1:06:17short-term undershoot, I don't think it
  1758. 1:06:20matters to the cent, if it goes here now
  1759. 1:06:22, right?
  1760. 1:06:26So that's at least maybe going there
  1761. 1:06:27again and then there's another slight
  1762. 1:06:29fight there.
  1763. 1:06:32So I think if you were interested in it
  1764. 1:06:34in the long term, then maybe now, if you
  1765. 1:06:36can find reasons that Rheinmetall can
  1766. 1:06:38now fundamentally get out of the
  1767. 1:06:40quagmire again.
  1768. 1:06:43So that's what I wanted to say again.
  1769. 1:06:45Otherwise, if they slip through there
  1770. 1:06:46now, then it's capitulation, right? Then
  1771. 1:06:51>> tactically, if you were to do the same,
  1772. 1:06:54you would say: Okay, because Donald
  1773. 1:06:56Trump doesn't have much to do
  1774. 1:06:57domestically anymore, he says, I still
  1775. 1:06:59want to get the medal from Stockholm,
  1776. 1:07:02finish it in the next two weeks or
  1777. 1:07:03ensure some kind of agreement between
  1778. 1:07:06Ukraine and Russia.
  1779. 1:07:10Rheinmetall shoots down again, falls
  1780. 1:07:12into the range between 600 and 650, and
  1781. 1:07:15is therefore technically more oversold
  1782. 1:07:17than it has been for ages, or even never
  1783. 1:07:20before.
  1784. 1:07:22And then you have a great setup, yes, if
  1785. 1:07:25you want to do it, because then it's
  1786. 1:07:27oversold, nobody wants the stock, and
  1787. 1:07:29there's an overshoot in terms of news,
  1788. 1:07:31everything that can come is negative.
  1789. 1:07:34Yes, and from that perspective, that
  1790. 1:07:36might be a setup for the interested
  1791. 1:07:38investor.
  1792. 1:07:40>> Yeah, maybe a tank ban, then they'll be
  1793. 1:07:43quick.
  1794. 1:07:43>> Yeah, my, my, I'm allowed to park my
  1795. 1:07:45hybrid, which is of course a scam, we
  1796. 1:07:48agree on that, I don't need to write
  1797. 1:07:50anything down below, I'm not allowed to
  1798. 1:07:52park for free with my hybrid from
  1799. 1:07:5401.01.27. No, because the city is taking
  1800. 1:07:57that away, that was supposed to be an
  1801. 1:07:58incentive to buy. You know, if you had
  1802. 1:08:00an E, you only had to put the parking
  1803. 1:08:02disc in and now that's being taken away
  1804. 1:08:04from me. Goodness, where are I and all
  1805. 1:08:06the SUV owners supposed to park our cars
  1806. 1:08:08? They're so big, they need two parking
  1807. 1:08:11spaces. Cheeky and not even for free
  1808. 1:08:13anymore. Annoying.
  1809. 1:08:17>> I mean, with the subscription profits
  1810. 1:08:19from the IPO there, you could probably
  1811. 1:08:21just buy the rest of the city, so to
  1812. 1:08:23speak.
  1813. 1:08:26>> Or the remaining Cybertrucks.
  1814. 1:08:28>> Oh yeah, oh yeah. Wasn't
  1815. 1:08:30>> that somehow the best product of all
  1816. 1:08:32time, which is now being scrapped? Yeah
  1817. 1:08:34, it doesn't matter. Let's move on to
  1818. 1:08:36Softbank before we get bogged down in
  1819. 1:08:38chatter.
  1820. 1:08:41Softbank has released figures, has
  1821. 1:08:43slightly increased its sales and profits
  1822. 1:08:46, but Sebastian, enlighten us, there may
  1823. 1:08:48be some cause for concern, right?
  1824. 1:08:54>> What do you mean, this focus on AI and
  1825. 1:08:56the bets?
  1826. 1:08:58>> Yes.
  1827. 1:08:59>> Very good. Yes, I just wanted to
  1828. 1:09:01double-check that we're on the same page
  1829. 1:09:03. Softbank, I almost wanted to say that
  1830. 1:09:05I didn't even include the figures here,
  1831. 1:09:07because I don't think they play any role
  1832. 1:09:09at Softbank.
  1833. 1:09:11Masayoshi Son, the founder of Softbank
  1834. 1:09:13and also, you could say, an investment
  1835. 1:09:16strategist in one person, is not
  1836. 1:09:17necessarily someone who runs traditional
  1837. 1:09:20investment companies with a certain
  1838. 1:09:22degree of diversification, but rather
  1839. 1:09:24really focuses on the future and is
  1840. 1:09:26investing huge amounts of money in them
  1841. 1:09:28.
  1842. 1:09:30And Softbank is a gigantic bet on Open
  1843. 1:09:32AI, so it owns up to 13%of Open AI, and
  1844. 1:09:35then there's Arm Holdings, a chip
  1845. 1:09:37manufacturer and designer from the UK.
  1846. 1:09:40Nvidia wanted to buy it a few years ago
  1847. 1:09:42, but if I remember correctly, it didn't
  1848. 1:09:44quite go through antitrust law. So those
  1849. 1:09:46are the two big chunks. Of course, there
  1850. 1:09:48are a lot of other investments in the
  1851. 1:09:50Vision Fund, but Softbank is actually a
  1852. 1:09:53highly leveraged AI bet.
  1853. 1:09:55And of course, you have to say that if
  1854. 1:09:57you invest in Softbank, you're investing
  1855. 1:09:59in the fact that AI and Agentic AI will
  1856. 1:10:02really become the mega thing, and then
  1857. 1:10:04you'll have a lot of fun with the stock
  1858. 1:10:06, but that's a...I would say that it's a
  1859. 1:10:08highly speculative bet on two companies
  1860. 1:10:10, and especially if Open AI goes public
  1861. 1:10:12sometime next year, then you can also
  1862. 1:10:14buy the stock directly on the stock
  1863. 1:10:16exchange, you don't necessarily need
  1864. 1:10:18Softbank. So that's a bit of a
  1865. 1:10:20make-or-break stock, I think. What's
  1866. 1:10:23>> the connection with Arm Holdings like?
  1867. 1:10:25Could you maybe explain that briefly?
  1868. 1:10:30>> Softbank is involved, has a majority
  1869. 1:10:32stake in Arm Holdings and therefore has
  1870. 1:10:34the opportunity to take it public or
  1871. 1:10:36sell it or whatever, and they're betting
  1872. 1:10:39heavily on AI becoming a huge megatrend
  1873. 1:10:41.
  1874. 1:10:45>> Okay.
  1875. 1:10:47Softbank, Lars, what do you think of the
  1876. 1:10:49company?
  1877. 1:10:51It's
  1878. 1:10:52>> too concentrated for me as a long-term
  1879. 1:10:54investment.
  1880. 1:10:57I don't find it exciting and, above all
  1881. 1:10:59, the man is really, you can see it
  1882. 1:11:02positively and negatively.
  1883. 1:11:05He has a very pronounced all-in
  1884. 1:11:07investment approach.
  1885. 1:11:10To be fair, in the past, not everything
  1886. 1:11:13Softbank has done has gone wrong, and I
  1887. 1:11:15think they're...you know, if you're
  1888. 1:11:17thinking about doing Neo Cloud.
  1889. 1:11:20Okay, it's always a bit more difficult
  1890. 1:11:22to buy a Neo Cloud company from an
  1891. 1:11:24investment company than from one of
  1892. 1:11:26these companies, but the bet is the same
  1893. 1:11:29, except that SoftBank might be a bit
  1894. 1:11:31more diversified.
  1895. 1:11:32So, fully on AI. And one thing is also
  1896. 1:11:35clear: if we look at it from a bullish
  1897. 1:11:37perspective, if the Entropic IPO goes
  1898. 1:11:39really well on the stock exchange and
  1899. 1:11:41the mood improves, then SoftBank's stock
  1900. 1:11:43should benefit massively because the
  1901. 1:11:45whole world will have to price it in.
  1902. 1:11:48Well, then OpenAI's Sam Altman will
  1903. 1:11:50think about it the next day, yeah, they
  1904. 1:11:52were already thinking about it.
  1905. 1:11:54Sam Altman had previously considered
  1906. 1:11:56going public and then gave up because
  1907. 1:11:58the last valuation rounds were simply
  1908. 1:12:00not that good.
  1909. 1:12:02And SoftBank should definitely benefit
  1910. 1:12:04from that.
  1911. 1:12:06Anyone who says that AI is currently
  1912. 1:12:08underrated will find SoftBank a nice
  1913. 1:12:11derivative, if you will, that it will
  1914. 1:12:13recover.
  1915. 1:12:15But please pay attention to the position
  1916. 1:12:17size.
  1917. 1:12:17>> I would even say that at SoftBank you
  1918. 1:12:19have Masayoshi Son as a financial
  1919. 1:12:21gambler. The old phrase "financial
  1920. 1:12:23gambler" came to mind again, I always
  1921. 1:12:24have to think about that. SoftBank was
  1922. 1:12:27previously a real opportunity to invest
  1923. 1:12:29in OpenAI, also significantly, in Arm
  1924. 1:12:31Holdings, which is also listed on the
  1925. 1:12:33stock exchange via IPO, but I think
  1926. 1:12:35almost 90%-maybe you just wanted to get
  1927. 1:12:37out of there, Timo-is in the hands of
  1928. 1:12:39SoftBank. But theoretically you can buy
  1929. 1:12:41Arm shares if you set a limit and
  1930. 1:12:43because it is a bit illiquid in some
  1931. 1:12:45areas, you will soon be able to buy
  1932. 1:12:46OpenAI as well. That means if I now say
  1933. 1:12:49that I want to invest in OpenAI, in Arm
  1934. 1:12:51, in maybe something new and crazy that
  1935. 1:12:53he finds, as a financial gambler via
  1936. 1:12:55Masayoshi Son, then of course this is
  1937. 1:12:57suitable. But if I say that I want to
  1938. 1:12:59have Arm Holdings or I want to have
  1939. 1:13:01OpenAI, for example, then I can also buy
  1940. 1:13:03the shares directly on the stock
  1941. 1:13:04exchange, so soon at OpenAI. I guess in
  1942. 1:13:06the first quarter or second quarter, if
  1943. 1:13:08the AI trade doesn't completely collapse
  1944. 1:13:10, they'll of course quickly go public
  1945. 1:13:12because the old investors will want to
  1946. 1:13:14cash out at some point and maybe
  1947. 1:13:16diversify a bit.
  1948. 1:13:17>> Hmm.
  1949. 1:13:19>> I don't really see any reason to buy
  1950. 1:13:21SoftBank at the moment, to be honest.
  1951. 1:13:24>> Yes.
  1952. 1:13:26I can understand that from my point of
  1953. 1:13:27view.
  1954. 1:13:29Considering the time it's getting late,
  1955. 1:13:31I'd say we're closing the books now.
  1956. 1:13:35The bookbuilding phase is over for the
  1957. 1:13:37three of us this week, and with that,
  1958. 1:13:39we'll send you off into the weekend or
  1959. 1:13:41into the new week, depending on when you
  1960. 1:13:43hear from us.
  1961. 1:13:45Thank you very much for your attention.
  1962. 1:13:49Thank you very much for your loyalty and
  1963. 1:13:51trust in the Buy the Dip app for two
  1964. 1:13:53years, to all those who have downloaded
  1965. 1:13:55it.
  1966. 1:13:57And yes, let Sebastian keep you both
  1967. 1:13:59informed. I'll keep you posted.
  1968. 1:14:01Which of you two is the honey trap? I'm
  1969. 1:14:04suspicious of that. I'm suspicious of
  1970. 1:14:06that.
  1971. 1:14:07>> Yeah.
  1972. 1:14:07>> I don't want to exclude any of you, it's
  1973. 1:14:09probably both of you. So have a nice day
  1974. 1:14:11. Take care. See you
  1975. 1:14:13>> later. Thank you. See you
  1976. 1:14:15>> later. Ciao.

About this transcript

This page contains the full transcript of Nvidia-Krise: jetzt kaufen? 5 Crash-Aktien im CHECK, Anthropic-IPO & DAS ist unser Dip der Woche! by Buy The Dip, generated from the public captions YouTube serves with the video. The transcript has 12,418 words across 1,976 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to choose a caption track available on the video.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, caption language selection, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.