YouTube2Text

Nebius (NBIS): Why it may not recover... — Transcript

by UNRIVALED INVESTING · 3,175 words · 520 segments · language en · Watch on YouTube

Full transcript

  1. 0:00Nebius stock, NBIS,
  2. 0:03an AI investor favorite,
  3. 0:06has lost nearly 20% of its value today,
  4. 0:08following their third quarter earnings
  5. 0:10update. In this video, I'm going to walk
  6. 0:12you through
  7. 0:13why I'm skeptical
  8. 0:16that Nebius stock will be able to
  9. 0:18retrace its highs and reach all new
  10. 0:20highs from here.
  11. 0:23If anything, I think the stock is
  12. 0:24starting to break down. Yes, it's down a
  13. 0:26lot from its highs, but I suspect it
  14. 0:29could actually go a lot lower. And let
  15. 0:31me walk you through why I'm skeptical
  16. 0:35in this video. And this is not I don't
  17. 0:37have a position on Nebius, but I was
  18. 0:38curious if this was potentially a
  19. 0:40bargain that would get me interested.
  20. 0:42So let's go through it. As always, my
  21. 0:44name is Daniel. You're watching
  22. 0:45Unrivaled Investing. If you enjoy
  23. 0:47educational investing content, I
  24. 0:49appreciate you hitting that thumbs up,
  25. 0:50hit that subscribe button. Also, I
  26. 0:52welcome your comments at the end of this
  27. 0:54video because I'm looking for pushback.
  28. 0:56I'm looking for ideas and if you think
  29. 0:58I'm missing the mark on this, let's
  30. 1:00explore it together. So first of all,
  31. 1:02one should recognize that their results
  32. 1:03were actually really solid. At least
  33. 1:06that's the the the initial 30, you know,
  34. 1:09the the initial 5-second impression when
  35. 1:11you look at it cuz you see revenue. Holy
  36. 1:14moly. Wow wow
  37. 1:16revenue up nearly 600%.
  38. 1:20Fantastic results. And then you see
  39. 1:22their EBITDA.
  40. 1:24And, you know, they're going from loss
  41. 1:27to hugely profitable. Doesn't this look
  42. 1:30great?
  43. 1:31And then, you know, management saying,
  44. 1:33look, we have a lot of potential to keep
  45. 1:35growing. We're going to be launching
  46. 1:37more and more sites, a lot more
  47. 1:39gigawatts.
  48. 1:40They're going to be coming online around
  49. 1:42the world, new sites.
  50. 1:44And with this launch of new,
  51. 1:48you know, data centers that we're going
  52. 1:49to be building out,
  53. 1:51our revenue is going to keep going
  54. 1:54higher.
  55. 1:55And they previously said, you know, 530
  56. 1:57million to three 3.4 to to let's say 3
  57. 2:01billion dollars. So hypergrowth.
  58. 2:04And
  59. 2:06you know, the business model being, hey,
  60. 2:07we got a lot of this covered. You know,
  61. 2:10our customers are prepaying most of
  62. 2:11this.
  63. 2:13You know, so and by the way, not only
  64. 2:15are they prepaying it, but the economics
  65. 2:17of it suggests that it gets paid back
  66. 2:19very quickly. So you might say this is a
  67. 2:21great business model. I mean, yeah, it's
  68. 2:24capital intensive as you build out all
  69. 2:25these data centers, but your customers
  70. 2:27are paying for it and the payback time
  71. 2:30is measured in months. You know, this is
  72. 2:33fantastic and it's getting better.
  73. 2:34Arguably it got better from the second
  74. 2:36quarter to the third quarter. This is
  75. 2:37what I'd like to see.
  76. 2:39And the cost or the their pricing is
  77. 2:43getting better. So the annualized um you
  78. 2:46know, contract value per megawatt,
  79. 2:48that's still going higher. So they're
  80. 2:50saying the third quarter deals were even
  81. 2:51better than the second quarter deals. So
  82. 2:53better economics,
  83. 2:55customers that are paying for the
  84. 2:57financing of this build out.
  85. 3:00What's not to like? So why is Nabia
  86. 3:03stock NBIS collapsing?
  87. 3:06So a few things. First of all, I just
  88. 3:08did a video on Nvidia with Jensen Huang
  89. 3:11out.
  90. 3:12This is partly down in sympathy with the
  91. 3:14broader AI trade. Now I think it is
  92. 3:17important to recognize that Nvidia has
  93. 3:19provided a lot of circular financing to
  94. 3:22number of players in the the the
  95. 3:24ecosystem.
  96. 3:26And so they'd say, hey, I'll I'll give
  97. 3:27you this financing. I want warrants in
  98. 3:30your company and they're therefore
  99. 3:32you're going to buy my chips.
  100. 3:35Without Jensen there,
  101. 3:37you know,
  102. 3:38is it still as certain that that sort of
  103. 3:40favorable circular financing deal still
  104. 3:41happen?
  105. 3:42Probably not. I think that is a real
  106. 3:45risk that the market it gets harder to
  107. 3:47get the money. That that's like if you
  108. 3:49want to boil it down to one key aspect
  109. 3:52on why I suspect it'll be tough for
  110. 3:55Nabias to make new highs from here is
  111. 3:58because it's harder to get the money.
  112. 4:01That that core idea. So you it with
  113. 4:03Jensen Wong at Nvidia.
  114. 4:06Another key aspect on why the stock is
  115. 4:08down
  116. 4:09largely actually due to execution. Their
  117. 4:11previous expectations were 2026 revenue
  118. 4:13of around 3 to 3.4 billion. Now they're
  119. 4:16revising it to 3 billion, so a little
  120. 4:18bit of a haircut. Previously talking
  121. 4:20about their annualized recurring revenue
  122. 4:21target by the end of this year of 7 to 9
  123. 4:23billion. Now they're saying it'll be 7
  124. 4:25billion plus by February 2027. So these
  125. 4:28things take time. Sometimes you get push
  126. 4:30back. You know, sometimes the
  127. 4:32communities don't want the data center.
  128. 4:34So these delays happen. But they're
  129. 4:36still saying, look, our hyper growth is
  130. 4:38still on target and we're putting up
  131. 4:39fantastic numbers. Another core aspect
  132. 4:42on why Nvidia stock has been coming
  133. 4:44under pressure is I think the amount of
  134. 4:46money it's going to cost to build this
  135. 4:48out. You know, for example,
  136. 4:50for their 2026 figures, they were
  137. 4:53previously saying, hey, it'll be 20 to
  138. 4:5425 billion. Now it's going to be closer
  139. 4:56to 27 billion.
  140. 4:58Then 2027, their estimates of how much
  141. 5:00it's going to cost to invest in all this
  142. 5:02infrastructure. Previously analysts were
  143. 5:04estimating, you know, somewhere between
  144. 5:0525 and 30 billion dollars. So a little
  145. 5:07bit of a step up from 2026, but not the
  146. 5:11figure that they talked about today of
  147. 5:1340 to 45 billion. That's a huge jump. 50
  148. 5:1650% plus jump
  149. 5:18or 40 to 50% plus jump.
  150. 5:20And also you're seeing this huge jump in
  151. 5:22capital expenditures. Meanwhile
  152. 5:24management saying their revenue is not
  153. 5:27expected to also jump significantly. So
  154. 5:29that is a huge aspect on I think why the
  155. 5:32stock came under pressure. Investors are
  156. 5:33saying, wait a second. You have to put
  157. 5:35up
  158. 5:36tens of billions more in capital
  159. 5:38potentially and to build this out. And
  160. 5:42we we haven't even talked about how
  161. 5:44we're going to finance that.
  162. 5:45And we're not seeing a 40% jump in
  163. 5:47revenue beyond what our expectations
  164. 5:50previously were. So that's the key
  165. 5:52aspect. It's once again, show me the
  166. 5:53money. Show me the money dynamics.
  167. 5:57And so this is the key aspect here,
  168. 6:01which is that investors are worried
  169. 6:03about the
  170. 6:05and financing.
  171. 6:06Because with this $40 to $45 billion in
  172. 6:09expected capital expenditures next year,
  173. 6:11management has said, yeah, we've got
  174. 6:13about 80% coverage, somewhere between 75
  175. 6:15to 85%. That's the, hey, we've got
  176. 6:18partners, we've got customers that are
  177. 6:20willing to cover it. We've got vendor
  178. 6:23financing, we've got secured lending,
  179. 6:27you know, someone someone saying, hey,
  180. 6:28I'll lend directly to these,
  181. 6:31you know, on these GPUs, on these server
  182. 6:33racks, on the land, what have you.
  183. 6:36But there's still a difference
  184. 6:38of about 20 to 25%
  185. 6:42where
  186. 6:44management's going to need to figure out
  187. 6:46where they're going to get that money
  188. 6:48tied to that $40 to $45 billion in
  189. 6:50capex. And that hole is somewhere around
  190. 6:54$6 billion to $11 billion
  191. 6:56just for 2027.
  192. 6:58So
  193. 6:5975 to 85% management said they got
  194. 7:01covered,
  195. 7:02$40 to $45 billion they're spending,
  196. 7:05and they need to find that last 15 to
  197. 7:0725%
  198. 7:09leading to roughly, let's say averages
  199. 7:11out to around $8 billion
  200. 7:13that they need.
  201. 7:15So where's that coming from? Because you
  202. 7:18see CoreWeave
  203. 7:19and their bonds are at 15%.
  204. 7:22There are numerous stories of banks
  205. 7:24looking to sell their AI exposure or
  206. 7:27pull back a bit.
  207. 7:29So where's the where's this $8 billion
  208. 7:31coming? Are you going to issue, you
  209. 7:33know, a lot of stock?
  210. 7:34That's one possibility. Are you going to
  211. 7:36issue, you know, go the CoreWeave route,
  212. 7:38issue a lot of debt?
  213. 7:40So there there is that question, which
  214. 7:42is
  215. 7:43this is the core aspect, the heart of
  216. 7:46why I'm skeptical that Nebius can
  217. 7:48retrace its highs.
  218. 7:51Now let's keep going in just a second.
  219. 7:53First, please recognize none of this is
  220. 7:55financial advice. I'm sharing just some
  221. 7:58some of my thoughts on Nebius NBIS. Also
  222. 8:02a couple members, a few members from
  223. 8:04Unrivaled Investing have called out
  224. 8:05Darshan P just this past week. He said
  225. 8:07thank you for and he called out one of
  226. 8:09the companies I was recently buying. He
  227. 8:11said paid for another 10 years of
  228. 8:13subscription with ample change left
  229. 8:15over. This follows Josh L. He wrote love
  230. 8:18the content, informative, new, and
  231. 8:20meaningful perspective on valuing these
  232. 8:22companies. And I intend on remaining a
  233. 8:24member for as long as you continue
  234. 8:26producing content. If you're looking for
  235. 8:27real money portfolio updates, compelling
  236. 8:29ideas, exclusive videos, come check out
  237. 8:32Unrivaled Investing. Also, sometimes I
  238. 8:34get push back here on this channel where
  239. 8:36folks say Daniel, I want to learn more
  240. 8:38about your journey. Name a stock that
  241. 8:40you've recently been buying. And so in
  242. 8:42this case,
  243. 8:43I did a video recently on Mercado Libre
  244. 8:47and why I bought more. So if you're
  245. 8:49interested, go check that out. That's
  246. 8:52just one example of the type of
  247. 8:54companies that I'm interested in.
  248. 8:55Mercado Libre is, in full disclosure,
  249. 8:57one of my top five holdings. Could go
  250. 9:00lower. There's still very very real
  251. 9:02risks and I you know I might scale it
  252. 9:04up, continue to keep scaling it up.
  253. 9:06We'll see. But
  254. 9:08I I wanted to call this out if you
  255. 9:09wanted to learn more about my journey in
  256. 9:12terms of how I'm thinking about
  257. 9:13companies. When I turn to Flash, my free
  258. 9:15to use AI assistant, see the link below,
  259. 9:17it says yeah, Nubank has a lot of upside
  260. 9:20but also has a lot of downside. I'm more
  261. 9:22skeptical personally. I don't think it
  262. 9:25has as much upside as Flash thinks.
  263. 9:29And it comes down to this funding gap
  264. 9:31which is, you know, we just called out
  265. 9:33for 2027
  266. 9:35there's around $8 billion
  267. 9:37that they'll need. And this assumes
  268. 9:39let's say you have a model of 80%
  269. 9:41covered, 20% that they need to figure
  270. 9:43out where it comes from. By no means
  271. 9:45does that mean the market's always going
  272. 9:47to be like that. The market can change.
  273. 9:50We've just been through this wonderful
  274. 9:52period with lower rates and exuberance
  275. 9:54for AI where folks are saying hey, I've
  276. 9:56got this covered.
  277. 9:58But the second folks start having losses
  278. 10:00or they're looking to sell some of their
  279. 10:02loan exposure, it might not be 80 20. It
  280. 10:06can go 70 30. It can go 60 40. That
  281. 10:09dramatically changes the picture for
  282. 10:12Nimbia's growth profile.
  283. 10:15Super important to reflect on. So, let's
  284. 10:17assume they've got 2027 covered, and
  285. 10:19we'll talk about some of the assets they
  286. 10:21have. They have a sizable, you know,
  287. 10:22cash balance. They do have some
  288. 10:24convertible debt. They have some other
  289. 10:26assets as well. So, let's assume 2027 is
  290. 10:29covered, but then that leaves, let's
  291. 10:31say, you know, tens of billions of
  292. 10:34dollars in the years ahead, and this is
  293. 10:36the story that's required in order to
  294. 10:39really believe that you can, you know,
  295. 10:41let's say go to new highs from here.
  296. 10:44Because you're looking at a cumulative
  297. 10:47capital expenditure figure of $100
  298. 10:49billion plus, of which 20% would
  299. 10:52suggest, hey, we've got a, you know, a
  300. 10:55very sizable $20 to $30 billion hole
  301. 10:57that we're going to need to fill in the
  302. 10:59years ahead.
  303. 11:00Now, investors might say, Daniel, you
  304. 11:02know, you you need to make sure you
  305. 11:03don't overlook the fact that they have a
  306. 11:05lot of assets. They have database
  307. 11:06software, autonomous vehicles, a lot of
  308. 11:08stuff on their balance sheet that, you
  309. 11:10know, maybe it's worth somewhere between
  310. 11:11$5 and $10 billion, these stakes in
  311. 11:13businesses that they own. And so that
  312. 11:15could help pay down. So, let's assume
  313. 11:17they use that. And it's important to
  314. 11:19recognize, yeah, you some folks might
  315. 11:21say, well, aren't they profitable or
  316. 11:22won't they get to profitability soon?
  317. 11:24And management's saying, yeah, we're
  318. 11:26hoping to be pre-tax profitable in the
  319. 11:28fourth quarter of 2027, you know, after
  320. 11:31they sort of whiffed on their recent
  321. 11:32expectations. Maybe that that gets
  322. 11:34punted a bit. But that is important to
  323. 11:36reflect on when you're thinking about,
  324. 11:38well, how are they financing this,
  325. 11:40this dynamic?
  326. 11:42This is
  327. 11:43why I'm I'm sort of struggling. Let's
  328. 11:45assume that 2027 gets financed, no
  329. 11:48problem.
  330. 11:49And that gets them to 2028, and let's
  331. 11:52assume by then the AI capital markets
  332. 11:54tighten up. One shouldn't assume that
  333. 11:56it's always going to be great times with
  334. 11:59fantastic, you know, low rates and easy
  335. 12:02capital access. Let's assume things
  336. 12:05actually get a little tougher.
  337. 12:06Especially, you know, look, we're seeing
  338. 12:08the 10-year Treasury yield
  339. 12:10inching higher now. It's above 5%. What
  340. 12:11if it goes to 5 and 1/2 %?
  341. 12:14You know, you're seeing all these
  342. 12:15hyperscalers talking about issuing lots
  343. 12:18and lots of money. Meta talking about
  344. 12:20needing to raise hundreds of billions of
  345. 12:21dollars. What if they sort of squeeze
  346. 12:23out someone like because they go, wait a
  347. 12:26second, would I rather lend to,
  348. 12:28you know, Meta
  349. 12:29who already has these very lucrative ad
  350. 12:31platform or someone that's pre-profit?
  351. 12:35You have these dynamics with with
  352. 12:36serious customer concentration dynamics.
  353. 12:40And so looking at,
  354. 12:42you know, what this looks like. Let's
  355. 12:44assume the capital markets tighten up so
  356. 12:45that in 2028 they're not investing as
  357. 12:47much, but they're able to do the capital
  358. 12:49expenditures required for 2027. So you
  359. 12:52get to around 22 billion dollars in
  360. 12:54revenue for 2028. So that is still a
  361. 12:57huge uptick, you know, from from where
  362. 12:59they are today and where they they're
  363. 13:01targeting for 20 27. You know, we're
  364. 13:03saying, hey, you go from let's say 12
  365. 13:05billion to 22 billion. Still fantastic
  366. 13:07growth. Let's assume your margins are
  367. 13:09fantastic, around 50%. So you're getting
  368. 13:12around 11 billion dollars in EBITDA.
  369. 13:15Now, this is a capital intensive
  370. 13:16business.
  371. 13:17Management says that they, you know,
  372. 13:19look at the depreciation schedule of
  373. 13:21around five years. And so if you're
  374. 13:23investing, let's say 40 billion dollars
  375. 13:26like management expects and it's a five
  376. 13:28year life, you know, you're you're
  377. 13:30talking about,
  378. 13:32you know,
  379. 13:338 billion dollars roughly.
  380. 13:36You know, it is important to think about
  381. 13:38that dynamic,
  382. 13:40you know, as as you amortize this
  383. 13:43uh over its lifetime.
  384. 13:46Okay, so normalized GPU replacement
  385. 13:48CAPEX. This 8 billion dollars could be
  386. 13:50higher, could be lower. Something to
  387. 13:52think about in terms of sustaining the
  388. 13:55business.
  389. 13:56Okay, so 11 billion dollars, 8 billion
  390. 13:58dollars. How about the interest expense?
  391. 14:01Just to sustain that that's a sustaining
  392. 14:03figure, not to grow this 22 billion
  393. 14:06significantly.
  394. 14:07Okay, interest expense around 500
  395. 14:10million. This might get a little tricky
  396. 14:12because, you know, in terms of interest
  397. 14:14expense,
  398. 14:15once again,
  399. 14:17you know, what do they choose to do for
  400. 14:18the 2027 financing, that 20% hole?
  401. 14:23You know, is it
  402. 14:24closer to a Coreweave scenario where
  403. 14:26they're now going to be issuing, you
  404. 14:27know, let's say $5 billion of bonds at
  405. 14:2910%? Well, that alone would cause this
  406. 14:32interest expense figure to jump
  407. 14:33significantly higher. Okay, so then you
  408. 14:35get to the pre-tax figure, pre-tax
  409. 14:37earnings of around $2.5 billion. Keep in
  410. 14:39mind, management said they expect to get
  411. 14:41to break even by the end of 2027. So I'm
  412. 14:44giving them plenty of credit here for
  413. 14:46what 2028 might look like.
  414. 14:48So then you get some sort of tax expense
  415. 14:50and thinking about normalized earnings
  416. 14:53potentially around $2 billion. There's a
  417. 14:56lot of wiggle room here. There's a lot
  418. 14:59of, you know, ways that this can go
  419. 15:01worse and better.
  420. 15:04One is around interest expense, the
  421. 15:05other is thinking about the replacement
  422. 15:07for CapEx.
  423. 15:09This could be closer to 1 billion, this
  424. 15:11could be closer to 3 billion. A lot of
  425. 15:13wiggle room here. Let's assume around a
  426. 15:1515 times multiple. That gets to around
  427. 15:1730 billion dollars cuz the reason why
  428. 15:20I'm assuming 15 times multiple is once
  429. 15:21again, if we're talking about AI capital
  430. 15:23markets
  431. 15:24being tougher as we're starting to see.
  432. 15:28And once again, going to the Coreweave
  433. 15:30example, going to numerous banks pulling
  434. 15:33back.
  435. 15:35And
  436. 15:37you look at this,
  437. 15:39you know, this is less of a growth sort
  438. 15:41of figure. This is assuming like, okay,
  439. 15:42what if what if this sort of steady
  440. 15:44state getting around a 6 7% yield, that
  441. 15:47suggests a $30 billion market cap using
  442. 15:50their approximate diluted share count
  443. 15:53currently.
  444. 15:55Currently, you know, 330 million, maybe
  445. 15:57it's around 320,
  446. 15:59a little bit messy. Then you're getting
  447. 16:00around $90 share price. And so this is
  448. 16:03the the core aspect of why I'm
  449. 16:05struggling to say like, oh yeah, this is
  450. 16:08going to be a $300 $600 stock in the
  451. 16:11future. Yeah, there's a it's clearly
  452. 16:13delivering a lot of growth right now
  453. 16:15based on how much money they've been
  454. 16:17able to invest. But the reason why the
  455. 16:20market is starting to penalize them is
  456. 16:22because the market always thinks ahead,
  457. 16:24and it's wondering, wait a second,
  458. 16:27how much money are you going to be able
  459. 16:28to invest in the future? Who on whose
  460. 16:30terms?
  461. 16:32And frankly, it reminds me a little bit
  462. 16:34about the Warren Buffett quote about
  463. 16:35trying to find businesses that don't
  464. 16:37need to rely on the kindness of
  465. 16:38strangers.
  466. 16:40And that's effectively the challenge
  467. 16:42here with Nubius. It has to rely on the
  468. 16:45capital of strangers to plug that 20%
  469. 16:48hole. Maybe not for 2027.
  470. 16:51Maybe they're going to be able to do
  471. 16:52that with their balance sheet and the
  472. 16:53existing assets they have. But if you
  473. 16:56believe in the longer-term growth story,
  474. 16:59you're going to need to require that
  475. 17:01kindness of strangers. That's just the
  476. 17:03nature of this business model. It is not
  477. 17:06the well-diversified dynamics like you
  478. 17:09see with the hyperscalers, something
  479. 17:10like a Google where you have, you know,
  480. 17:13GCP and you have this advertising
  481. 17:15business and a bunch of other assets.
  482. 17:17That is a way better competitive
  483. 17:18position
  484. 17:20versus Nubius
  485. 17:22in my opinion.
  486. 17:24So even if Nubius reaches $22 billion of
  487. 17:27revenue, the shareholder value depends
  488. 17:30really heavily on how much it costs to
  489. 17:32finance it
  490. 17:34and maintain
  491. 17:36that revenue, not simply how much EBITDA
  492. 17:39reports cuz folks are the folks get
  493. 17:41super excited. They go, haven't you seen
  494. 17:44the growth and the price-to-sales
  495. 17:46multiple?
  496. 17:48This looks very cheap.
  497. 17:51Not so cheap if you're thinking about
  498. 17:53what their margins could be,
  499. 17:55you know, longer term and that this is a
  500. 17:56capital-intensive business.
  501. 17:58So I think that's the part that, you
  502. 18:00know, investors
  503. 18:03it would be worth thinking about. And
  504. 18:05that's why I'd love to read your
  505. 18:06comments below. If you want to learn
  506. 18:08more about my journey, come check out
  507. 18:11unraveledinvesting.com. I also have a
  508. 18:13free course where I do share, you know,
  509. 18:15example, you know, I I have a case study
  510. 18:17on trying to find a potential
  511. 18:19four-bagger in four years and that it
  512. 18:21does work through, you know, what what
  513. 18:23am I looking for? That's a completely
  514. 18:25free course for those interested. That's
  515. 18:26also at unrivaledinvesting.com.
  516. 18:28I hope this video has been helpful for
  517. 18:30you. If so, please make a point of
  518. 18:31hitting that thumbs up, hit that
  519. 18:32subscribe button. Thanks so much for
  520. 18:33tuning in.

About this transcript

This page contains the full transcript of Nebius (NBIS): Why it may not recover... by UNRIVALED INVESTING, generated from the public captions YouTube serves with the video. The transcript has 3,175 words across 520 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.