YouTube2Text

The AI Debate is No Longer Boom vs Bust: Why It’s Time to Go Back to Stock Picking | The Weekly Wrap — Transcript

by Steve Eisman · 4,369 words · 658 segments · language en · Watch on YouTube

Full transcript

  1. 0:00Now, it's going to take a while to
  2. 0:01figure out the full implications of the
  3. 0:03resignation of Jensen Wong. [music] We
  4. 0:05know the election, combined with the
  5. 0:07Core Weave debt, the wobbly foundations
  6. 0:09of the Entropic IPO evaluation, and now
  7. 0:12the Jensen drama have all sucked [music]
  8. 0:14trillions of dollars in valuation from
  9. 0:16the stock market and voters' retirement
  10. 0:18accounts. There was some extremely
  11. 0:19negative news about housing that came
  12. 0:21out this week. Inventory is increasing,
  13. 0:23but buyers aren't buying. That is the
  14. 0:26lowest rate in [music] 16 years. But
  15. 0:28it's actually worse than that. Cisco
  16. 0:30reported Thursday night. These are very
  17. 0:32good numbers for Cisco. Charter reported
  18. 0:34last Friday, and the results were not
  19. 0:35good. There will be winners and losers,
  20. 0:37and it's time to return to good
  21. 0:39old-fashioned stock picking, a world I
  22. 0:41know well. Let's start the process now.
  23. 0:47[music]
  24. 0:54Hi, this is Steve Eisman, and this is
  25. 0:55the weekly wrap for the week ending
  26. 0:57Friday, November 13, but recorded
  27. 1:00Thursday night, November 12. If you like
  28. 1:02the work we're doing, please subscribe
  29. 1:04to our YouTube channel and visit us at
  30. 1:06steveeisman.com to subscribe to
  31. 1:09Substack. This past Wednesday, November
  32. 1:1111, on Substack Premium, we posted an
  33. 1:14interview with Melody Wright of Unicus
  34. 1:16Research, who is an expert in housing.
  35. 1:19>> And some would call this, depending on
  36. 1:20which way you're looking at the metrics,
  37. 1:21the most unaffordable in history uh
  38. 1:24right now.
  39. 1:25>> Melody discussed the horrific state of
  40. 1:27US housing and why she thinks housing
  41. 1:29prices are going to decline in most
  42. 1:32parts of the country. This coming
  43. 1:34Wednesday, November 18, we will post an
  44. 1:36interview with Gary Gensler, former head
  45. 1:39of the Commodity Futures Trading
  46. 1:40Commission, the CFTC, and the Securities
  47. 1:43and Exchange Commission, the SEC.
  48. 1:46>> And we tend to build the infrastructure
  49. 1:48before we have sufficient revenues to
  50. 1:50support that infrastructure. Every
  51. 1:52single time plays through to a
  52. 1:55correction. How damaging, when does it
  53. 1:58happen? And I've publicly already said,
  54. 2:00think it's in the next 18 to 30 months.
  55. 2:02>> We will discuss the potential problems
  56. 2:05with the AI trade, regulation of
  57. 2:07prediction markets, and other regulatory
  58. 2:10matters as well. We hope you tune in.
  59. 2:12Now, moving on to the wrap. This week,
  60. 2:14we will discuss one, AI news and
  61. 2:17implications. Two, negative housing data
  62. 2:21and cost of living. Three, earnings
  63. 2:23reports, of course. And four, my fantasy
  64. 2:26novel list. Before we get started, let
  65. 2:29me begin with the conclusion. Until now,
  66. 2:32the AI debate has been between those who
  67. 2:35are extremely bullish and those who
  68. 2:38think it will crash and burn and send
  69. 2:41the US into a horrific recession. For
  70. 2:44the reasons I'm about to go into, I
  71. 2:46think what we now need is a differential
  72. 2:48diagnosis. It's not all good and it's
  73. 2:50not all bad. There will be winners and
  74. 2:53losers and it's time to return to good
  75. 2:55old-fashioned stock picking, a world I
  76. 2:58know well. We will be exploring this
  77. 3:00over the next several weeks, but let's
  78. 3:02start the process now. Now, it's going
  79. 3:04to take a while to figure out the full
  80. 3:06implications of the resignation of
  81. 3:08Jensen Huang. I'm sure we will be
  82. 3:10exploring this over the next few weeks,
  83. 3:13but even prior to that, the AI story
  84. 3:16continues to dominate the news, but with
  85. 3:18a shift in focus. Before last week and
  86. 3:21before the midterm elections, the market
  87. 3:24was focused on the CAPEX and the debt
  88. 3:27needed to fund the CAPEX. Top of mind
  89. 3:29was the increase in the debt costs of
  90. 3:31the neo clouds like CoreWeave. Also, the
  91. 3:34focus was on the growing concern about
  92. 3:36all the capital the hyperscalers were
  93. 3:38raising. Then, last week, when the
  94. 3:41Democrats swept Congress, the AI
  95. 3:44narrative changed again. It's hard to
  96. 3:46keep up. Many elections were won on a
  97. 3:49platform of data center moratoriums. The
  98. 3:53evils of data centers was a major
  99. 3:55election narrative. Investors now are
  100. 3:57naturally worried that Democrats will
  101. 4:00follow through on election promises and
  102. 4:02curtail or even totally halt AI
  103. 4:06construction all over the country. This
  104. 4:08narrative dominated the market last
  105. 4:10Thursday when Nasdaq declined by a
  106. 4:13whopping 5%. The competing narrative,
  107. 4:16which company executives and AI bulls
  108. 4:19are asserting to assure their equity and
  109. 4:21debt audiences, is that nothing could be
  110. 4:24further from the truth. The battle of
  111. 4:25the narratives will continue until we
  112. 4:28have more concrete news. Into the mix is
  113. 4:31of course Anthropic's IPO, but with a
  114. 4:33reduced valuation from the lofty
  115. 4:35pre-election $2 trillion
  116. 4:38that was bandied about. Also into the
  117. 4:40mix goes the resignation of Jensen Wong,
  118. 4:43and if nothing else, we know the
  119. 4:45election combined with the core weave
  120. 4:47debt, the wobbly foundations of the
  121. 4:49Anthropic IPO valuation, and now the
  122. 4:52Jensen drama have all sucked trillions
  123. 4:55of dollars in valuation from the stock
  124. 4:57market and voters' retirement accounts.
  125. 4:59Such is the power of the fear of the
  126. 5:01unknown about data centers, debt
  127. 5:03financing, and the power of AI. It might
  128. 5:05be just me, but publicly asserting that
  129. 5:08your technology is so dangerous it can
  130. 5:11destroy mankind, and saying that on the
  131. 5:15eve of a general election strikes me as
  132. 5:17amongst the strangest behavior I have
  133. 5:19ever seen in my entire investing life.
  134. 5:22Anyway, the voters seem to agree with
  135. 5:24Dario and Sam's assessment of the danger
  136. 5:27and conclusion that AI has to be
  137. 5:29controlled. How controlled is now the
  138. 5:32dominant question. In last week's wrap,
  139. 5:35I pointed out that there could be a
  140. 5:37silver lining in the rollback. If, and I
  141. 5:39want to emphasize it's a big if, AI data
  142. 5:43center construction slows by let's say
  143. 5:46something like 10 to 30%, then we can
  144. 5:49see some really positive implications.
  145. 5:52First, it would mean that AI CapEx would
  146. 5:54peak in 2027 and hyperscalers would
  147. 5:58therefore return to positive cash flow
  148. 6:01by perhaps as soon as the second half of
  149. 6:032027. In my view, that be a big positive
  150. 6:07as investors have grown sick of all the
  151. 6:10debt and equity raised by the
  152. 6:11hyperscalers. The less dependent the AI
  153. 6:14story is on the fixed income markets, in
  154. 6:17my view, the better. I'd also point out
  155. 6:19that the neo clouds would still be in
  156. 6:21trouble because they are still dependent
  157. 6:23on the fixed income markets for all of
  158. 6:25their growth.
  159. 6:26It would also mean that the growth rates
  160. 6:29of the picks and shovel companies, which
  161. 6:30are the equipment and power suppliers to
  162. 6:33AI data centers, would slow. This is a
  163. 6:36long list of companies that would
  164. 6:38include Dell, Cisco, Arista, Eaton, GE
  165. 6:41Vernova, and Quanta, and many others.
  166. 6:43Yes, the growth rates would slow, but
  167. 6:46the cycle might be elongated. On the
  168. 6:49further positive side, slower capex
  169. 6:51spending means less crowding out effects
  170. 6:54on US Treasuries. As I have discussed
  171. 6:56before, US Treasury yields remain
  172. 6:59stubbornly high, and a large part of
  173. 7:01that is because Treasuries compete with
  174. 7:04AI in selling debt. The US government is
  175. 7:07to some degree subject to the same
  176. 7:09economic supply and demand patterns as
  177. 7:11corporate debt issuers. Hence, the
  178. 7:13stubborn yield on Treasuries. Just as a
  179. 7:16quick recap, the US government has to
  180. 7:18fund its deficit by borrowing, which
  181. 7:20means selling US Treasuries. A crowding
  182. 7:23out effect means that there is a
  183. 7:24borrower in the same marketplace chasing
  184. 7:27the same funding dollars. AI debt needs,
  185. 7:31which are $500 billion in 2026, are so
  186. 7:34large that they are having a crowding
  187. 7:36out effect on the Treasury market. It's
  188. 7:38putting pressure on long-term interest
  189. 7:41rates, which in turn puts pressure on
  190. 7:44the US government deficit. Less AI capex
  191. 7:48might mean an easing up on interest
  192. 7:50rates, which would be good for everyone.
  193. 7:52A Substack subscriber pointed out that
  194. 7:56as data center construction projects
  195. 7:58start to get canceled or placed into
  196. 8:01review, the market will react poorly
  197. 8:03because it won't know how far this will
  198. 8:05go. A very valid point, but I think in a
  199. 8:09few months we will have a much better
  200. 8:10idea of how much construction will slow.
  201. 8:12Again, my guess is that a slowdown of
  202. 8:1410% to 30% of projects is where we are
  203. 8:17headed and that will be okay. Also, my
  204. 8:20guess is that the money being thrown at
  205. 8:22local officials to improve the lives of
  206. 8:25the communities hosting these data
  207. 8:27centers will move the needle on
  208. 8:29approvals and permitting. Now that the
  209. 8:31elections are over, the spotlights are
  210. 8:33turned down and focused elsewhere and
  211. 8:36strangers coming to town with bags of
  212. 8:39cash offering to improve schools, lower
  213. 8:42taxes, and generally rain Christmas
  214. 8:45spirit might be enough to get many of
  215. 8:48the data centers built. The market
  216. 8:50reflects all of these options without
  217. 8:52being able to choose one right now. It
  218. 8:55is always difficult to know what the
  219. 8:56market is thinking or why it moves the
  220. 8:58way it does. This week started with a
  221. 9:00strong rally, especially in tech stocks.
  222. 9:03I'll note that the new Congress does not
  223. 9:05sit until January 3rd. On the local
  224. 9:08level, stuff can get done in the new
  225. 9:09year. Until then, there will just be a
  226. 9:12lot of grandstanding about subpoenas and
  227. 9:15legitimate concern about cost of living.
  228. 9:17Moving on to the cost of living, there
  229. 9:19was some extremely negative news about
  230. 9:21housing that came out this week, which
  231. 9:23is that inventory is increasing, but
  232. 9:25buyers aren't buying. Why? Because
  233. 9:28prices are just too high. The housing
  234. 9:30market is locked because buyers can't
  235. 9:33afford the price and sellers don't want
  236. 9:35to lower the price. Most homeowners
  237. 9:37refinanced into 3% to 3 and 1/2%
  238. 9:40mortgages during COVID. If they sell and
  239. 9:44buy somewhere else, their new mortgage
  240. 9:46will cost them 7%. It's a massive
  241. 9:49increase. That's why supply and demand
  242. 9:52can't meet at an agreed price. But it's
  243. 9:55actually worse than that. Despite a
  244. 9:57locked market, housing prices have
  245. 9:58actually gone up all over the country
  246. 10:01and the ancillary costs of owning a
  247. 10:03house have skyrocketed. Property taxes,
  248. 10:06property and casualty insurance, and
  249. 10:08other fees like permits for land and
  250. 10:11construction are part of the equation,
  251. 10:13and those have gone up too. My guess, to
  252. 10:16a guess, is that the cost of buying a
  253. 10:17home and renovate today is at least 50%
  254. 10:20higher than it was only five years ago.
  255. 10:23Remember all those renovation and flip
  256. 10:25shows? Years of television programming
  257. 10:27promised fun, doable renovations leading
  258. 10:30to profitable flips dominated the
  259. 10:32housing markets for years. Those shows
  260. 10:35are gone for a reason. Right now, the
  261. 10:37data indicate that buyers are walking
  262. 10:39away and sellers are not cutting prices.
  263. 10:42Active inventory in October hit 183,500
  264. 10:47homes in the United States, up 7.4%
  265. 10:51versus last year. That is the fastest
  266. 10:53annual growth rate in almost a year.
  267. 10:55Ironically, new listings, new listings
  268. 10:59are falling at the same time, down 7 and
  269. 11:021/2% versus last year. So if new
  270. 11:05listings are falling, but inventory is
  271. 11:08rising, that means buyers are not buying
  272. 11:11at all. It also means that some listings
  273. 11:13are in broker pockets. Anecdotally, we
  274. 11:15hear that for every publicly listed
  275. 11:17listing, there are scores of private
  276. 11:19listings available to any serious buyer.
  277. 11:22Sellers with jobs and low mortgages are
  278. 11:24not motivated to sell at lower prices
  279. 11:27unless they are forced to by change of
  280. 11:29life circumstances. Overall, this isn't
  281. 11:33happening and prices have not fallen
  282. 11:35yet. In a recent interview on our
  283. 11:37Substack premium, Melody Wright of
  284. 11:39Unicus Research predicted that housing
  285. 11:42prices are going to decline all over the
  286. 11:44country over the next few years. If the
  287. 11:46cost of owning a home and mortgage rates
  288. 11:48don't improve soon, I think she's going
  289. 11:50to be right. And in further negative
  290. 11:52housing news on Thursday, it was
  291. 11:54reported that existing home sales
  292. 11:56decreased by 5 and 1/2% to an annualized
  293. 12:00rate of 3.75
  294. 12:02million. That is the lowest rate in 16
  295. 12:05years. Moving on.
  296. 12:07A few weeks ago, a Long Angle member
  297. 12:09posted a thread called let's start being
  298. 12:11concerned, like really concerned. The
  299. 12:13list could have been the rundown for
  300. 12:15this show. Bond yield and AI bubble, the
  301. 12:17national debt, private credit, you name
  302. 12:20it. 93 replies later, members had pushed
  303. 12:23back with their own numbers. One pointed
  304. 12:25out what's sitting in bonds and cash had
  305. 12:27cost since 2024. Others explained how
  306. 12:30they hedge without trying to time the
  307. 12:32market. That's long angle. It's a
  308. 12:34private vetted community of more than
  309. 12:369,000 founders, executives, and
  310. 12:38investors. Most of them passed the
  311. 12:39liquidity event comparing notes on what
  312. 12:41they're doing with their own money.
  313. 12:43Before the SpaceX IPO, over 300 members
  314. 12:46said how they'd approach it, and more
  315. 12:48than four in 10 plan to pass. It isn't
  316. 12:51only online either. There are more than
  317. 12:53a hundred member meetings a month across
  318. 12:5527 cities. Every member is verified.
  319. 12:59Nobody is allowed to sell to anyone and
  320. 13:01it's free to join. Apply at
  321. 13:03longangle.com/eisman.
  322. 13:06That's longangle.com/eisman.
  323. 13:11Charter reported last Friday and the
  324. 13:13results were not good. I bought Charter
  325. 13:15for myself in late January of this year
  326. 13:17because the stock was very inexpensive
  327. 13:19and I thought it was starting to turn
  328. 13:20around. After two bad quarters, I
  329. 13:23admitted I was wrong and I threw in the
  330. 13:24towel. That was a good decision because
  331. 13:26the stock is considerably lower. Charter
  332. 13:30is now combined with Cox and the
  333. 13:32combination has not helped. Revenue was
  334. 13:35down 1 and 1/2% and adjusted EBITDA was
  335. 13:38down four. Worse, the metric investors
  336. 13:41care the most about is broadband
  337. 13:44customers and those were down a whopping
  338. 13:47233,000.
  339. 13:50There is no end in sight for these poor
  340. 13:52fundamentals. The stock is down 44% this
  341. 13:54year. I received comments asking also
  342. 13:57about the housing company Meritage and
  343. 13:59the news obviously is not good giving
  344. 14:01the housing market and unless the
  345. 14:02housing market improves, Meritage will
  346. 14:04continue to disappoint. I threw in that
  347. 14:06towel also. SkyDance reported Monday
  348. 14:09night. This is the combination of
  349. 14:11Paramount and Warner Brothers. The
  350. 14:13problem this company faces is that it is
  351. 14:16down by a great deal of legacy media
  352. 14:18that keeps losing market share too new
  353. 14:20media. Making this into a great company
  354. 14:23is going to be a very difficult
  355. 14:25enterprise because this is not a very
  356. 14:27profitable company. Revenue for the
  357. 14:29third quarter was 7 billion, up 5%
  358. 14:32versus last year. That's not terrible.
  359. 14:35What's terrible is the fact that
  360. 14:36operating income was only 368 million
  361. 14:39for a margin of only 5% and net income
  362. 14:43was a lowly 73 million. Management plans
  363. 14:46on achieving cost savings of 6 billion
  364. 14:48in the next 3 years. Now cost savings
  365. 14:50are one thing, growing revenue is
  366. 14:53another. And how Sky Dance is going to
  367. 14:55do that when it is weighed down with
  368. 14:57legacy media is anyone's guess. Speaking
  369. 15:00of the devil, Core We reported Tuesday
  370. 15:02night. Now before we get to that, let me
  371. 15:04just say that this management team is
  372. 15:07really dicey and that's being kind. The
  373. 15:10company went public on March 27, 2025 at
  374. 15:1439 dollars. The stock is now 56 but it
  375. 15:18got as high as 138 in May of this year.
  376. 15:21Since going public, the three founders
  377. 15:24have sold combined over 3 billion in
  378. 15:28stock. Let me say that again. Combined,
  379. 15:31the three founders of Core Weave have
  380. 15:33sold over 3 billion in stock. How much
  381. 15:36stock have the three founders bought
  382. 15:37back? Zero. Also, Nvidia has invested 2
  383. 15:41billion in both Core Weave and Nebius.
  384. 15:44Investors view Nvidia as a safety net
  385. 15:46for these two neo clouds. That's why the
  386. 15:48scandal around Jensen Huang could be
  387. 15:50important. Now that Jensen Huang has
  388. 15:51been replaced, the next CEO might not be
  389. 15:54so quick to write checks to the neo
  390. 15:56clouds. And now on to Core Weave's third
  391. 15:59quarter earnings. The results were
  392. 16:00pretty good. Revenue reached 3.7
  393. 16:03billion, up 168 %. Despite this strong
  394. 16:06revenue growth, EPS was a large negative
  395. 16:09a dollar 27. The company raised
  396. 16:12guidance. For this company, what is much
  397. 16:14more important than the income statement
  398. 16:16is cash flow and the balance sheet. What
  399. 16:18is the company's free cash flow?
  400. 16:20According to the cash flow statement,
  401. 16:22operating cash flow for the quarter was
  402. 16:24a positive two billion. Cash flow from
  403. 16:27investing activities is negative 9.8
  404. 16:30billion. Almost all of that is from the
  405. 16:32purchase of equipment. So free cash flow
  406. 16:34is two billion minus 9.8 billion which
  407. 16:37equals negative 7.8 billion. To fund its
  408. 16:40negative cash flow, Core Weave has to go
  409. 16:42to the fixed income markets to raise
  410. 16:44capital. Another way to look at it is
  411. 16:46that at the end of this quarter, Core
  412. 16:48Weave had 54 billion in a combination of
  413. 16:50long-term debt and operating lease
  414. 16:52liabilities. That's why these earnings
  415. 16:55results are relevant. It will be a long
  416. 16:57time until Core Weave becomes cash flow
  417. 17:00positive. Until then, its business model
  418. 17:03is to raise debt, take those funds, and
  419. 17:05build AI data centers, and then rent out
  420. 17:08that compute. Right now, Core Weave's
  421. 17:10debt is priced at 15%. That is
  422. 17:12prohibitive. Core Weave says it has
  423. 17:15alternatives for its financing needs.
  424. 17:17Maybe, but I find it difficult to trust
  425. 17:20a management team where the three
  426. 17:22founders have sold over three billion in
  427. 17:24stock. Also, given the change in
  428. 17:26politics, data center build outs may
  429. 17:29slow significantly. The whole point of
  430. 17:31this company is to build data centers
  431. 17:33efficiently and effectively. This is a
  432. 17:35company that is not in control of its
  433. 17:37own destiny. Nevertheless, on Tuesday,
  434. 17:39the stock staged an impressive rally of
  435. 17:4125%.
  436. 17:42Why? Company raised guidance and 20% of
  437. 17:46the float is short. Heavily shorted
  438. 17:48stocks can have big rallies as hedge
  439. 17:51funds cover in a panic. Nebius, one of
  440. 17:54the other neo clouds, reported Thursday
  441. 17:56and the results were not received well
  442. 17:58with the stock down 19%. The company
  443. 18:00reported a loss of 13 cents versus a
  444. 18:04loss of 40 cents last year. Revenue
  445. 18:06reached 981 million, up over 500% versus
  446. 18:11last year. Nothing really wrong with
  447. 18:13these numbers, but here's where the
  448. 18:15numbers get dicey. The company expects
  449. 18:18fiscal year 2026 revenue at the low end
  450. 18:21of the company's 3 billion to 3.4
  451. 18:24billion dollar range. Expectations were
  452. 18:27for 3.35 billion. Nebius also expects
  453. 18:312027 revenue of 12 to 13 billion, which
  454. 18:35is in line with consensus, but I'm sure
  455. 18:38the bulls were hoping for more. But it's
  456. 18:40the Cap Ex numbers that I think some
  457. 18:42investors will be alarmed about or they
  458. 18:45were alarmed about it. The company
  459. 18:46expects 2026 capital expenditures of 27
  460. 18:50billion, higher than its previous Cap Ex
  461. 18:53guidance of 20 billion to 25 billion. As
  462. 18:56for 2027, Nebius predicts Cap Ex of 40
  463. 19:00billion to 45 billion, which is way
  464. 19:04higher than expectations of 29 billion,
  465. 19:07and that 's why the stock was down.
  466. 19:08Finally, Cisco reported Thursday night.
  467. 19:11Cisco is what I would call a picks and
  468. 19:12shovels company in tech. Its stuff goes
  469. 19:15into AI data centers. Its chief
  470. 19:17competitor is Arista. Assuming that AI
  471. 19:20data center construction slows, that
  472. 19:22will hurt Cisco's growth rate to some
  473. 19:23degree. On the other hand, the cycle
  474. 19:26could become elongated and that would be
  475. 19:28positive. This quarter is totally
  476. 19:30backward looking and simply represents
  477. 19:32how strong the AI story currently is,
  478. 19:35which we already know. Cisco has a weird
  479. 19:37fiscal year. It ends July 31st. So it
  480. 19:41just reported its first quarter for
  481. 19:43fiscal 2027.
  482. 19:45EPS of a dollar 40 was up 40% versus
  483. 19:48last year and a beat. Revenue of 18.5
  484. 19:52billion was up 24% versus last year,
  485. 19:55also a beat. These are very good numbers
  486. 19:57for Cisco. Unsurprisingly, it raised EPS
  487. 20:01and fiscal revenue guidance for fiscal
  488. 20:032027. Before I get to my fantasy novel
  489. 20:06list, I'll sum up where we stand in the
  490. 20:08market and economy and housing. Markets
  491. 20:11have always had a hard time digesting
  492. 20:14changes in politics, and we are big time
  493. 20:17in the middle of a political shift.
  494. 20:19While AI continues to be the dominant
  495. 20:21narrative for the markets, there is a
  496. 20:23differential diagnosis, I think, to be
  497. 20:26made for a nuanced bull case which
  498. 20:28favors the hyperscalers with their cash
  499. 20:31flows returning to positive driven by a
  500. 20:34slowdown in AI CapEx combined with their
  501. 20:38highly successful franchises. In this
  502. 20:40debt-driven market, cash flow is king.
  503. 20:43On the other hand, neo clouds are
  504. 20:45subject to the whims of the fixed income
  505. 20:47markets. As for Nvidia, I'll refrain
  506. 20:50from judgment and prediction, but I will
  507. 20:53say that their chip sales will slow
  508. 20:55along with data center construction. The
  509. 20:57picks and shovels companies like Arista,
  510. 20:59Cisco, and Dell will also slow along
  511. 21:02with Nvidia proportionate to the
  512. 21:04slowdown in data center build out. The
  513. 21:08summary of the bull case is that slower
  514. 21:09and consistent data construction will
  515. 21:11support an AI driven narrative for years
  516. 21:14to come. Somewhat slower but elongated
  517. 21:18could be the story and maybe that is a
  518. 21:20better story. I will refrain from
  519. 21:23handicapping local politics other than
  520. 21:25to say that cash is king when local
  521. 21:27bills need to be paid and raising taxes
  522. 21:30is a worse option than negotiating with
  523. 21:32data center contractors. So the bull
  524. 21:34case I just outlined is definitely
  525. 21:37possible. The bear case is extreme and
  526. 21:40ugly. If the whole AI story slows down
  527. 21:43to a crawl and there is little to no AI
  528. 21:45CapEx driving GDP growth, the US economy
  529. 21:49falls into a recession. Massive
  530. 21:51overnight loss in personal wealth and
  531. 21:53jobs will lead to the unlocking of the
  532. 21:55housing market which has been the store
  533. 21:58of value for money. For most Americans,
  534. 22:01the real holder of wealth in this
  535. 22:02country has been housing. In a deep
  536. 22:05recession, job loss and investment
  537. 22:07losses will unlock the housing market in
  538. 22:09a brutal way. The wealth locked up in
  539. 22:12housing is based on multi-generational
  540. 22:14wealth and locked up due to capital
  541. 22:17gains tax overhangs as well as high
  542. 22:19interest rates. If a recession occurs
  543. 22:22and leads to large-scale job loss, this
  544. 22:25hidden wealth will be unlocked. Prices
  545. 22:27will drop fast and furious with demand
  546. 22:30vanishing just as the need to cash out
  547. 22:32on the source of wealth grows.
  548. 22:34Eventually rates will drop, inflation
  549. 22:37will drop, and the economy will recover.
  550. 22:39But without AI, the recovery might be
  551. 22:42anemic. This doomsday scenario seems
  552. 22:44somewhat unlikely, but it is not
  553. 22:46impossible, and that's why the market is
  554. 22:49bouncing between days reflecting relief
  555. 22:51and days reflecting fear. This week on
  556. 22:54Monday, November 9 on our free service,
  557. 22:56we released a special episode. It's a
  558. 22:58masterclass in which I delve into all
  559. 23:01aspects of AI. Think of it as everything
  560. 23:04you wanted to know about AI, but were
  561. 23:06afraid to ask. This coming Monday,
  562. 23:09November 16, we will release an
  563. 23:11incredibly timely interview with tech
  564. 23:13analyst Dan Ives and Gil Luria.
  565. 23:15>> The AI revolution's accelerating. We see
  566. 23:17that in Korea a few weeks ago in terms
  567. 23:19of chips, Taiwan, we see it on the use
  568. 23:21case side in terms of enterprise. But
  569. 23:23then the spending, I think investors
  570. 23:25have said we have to slow this down, not
  571. 23:27realizing that it was going to take a
  572. 23:29blue tsunami to actually slow it down
  573. 23:31which
  574. 23:31>> The interview was conducted just after
  575. 23:33the midterms and we delved into the
  576. 23:36midterm implications as well as all
  577. 23:38aspects of the AI trade. They gave a
  578. 23:41deep dive into the bull case I just
  579. 23:43outlined. Hope you tune in. The best way
  580. 23:45to support the Real Investment Playbook
  581. 23:47is to subscribe to Substack through
  582. 23:50steveeisman.com
  583. 23:51and to subscribe to YouTube. And we
  584. 23:54greatly appreciate your support. And now
  585. 23:56for my fantasy novel list. Number one,
  586. 24:00Wheel of Time series by Robert Jordan.
  587. 24:02Number two, Assassin's Apprentice series
  588. 24:05by Robin Hobb. Liveship Traders series
  589. 24:08by Robin Hobb. Tawny Man series by Robin
  590. 24:12Hobb. Fits and the Fool trilogy by Robin
  591. 24:15Hobb. Game of Thrones series by George
  592. 24:17R. R. Martin. Codex Alera series by Jim
  593. 24:21Butcher.
  594. 24:22Dresden Files series by Jim Butcher. The
  595. 24:25Hobbit by J. R. J. Tolkien. Lord of the
  596. 24:27Rings by R. R. Tolkien, The Name of the
  597. 24:30Wind, The Kingkiller Chronicles by
  598. 24:33Patrick Rothfuss, Chronicles of Narnia,
  599. 24:36the series by C. S. Lewis, his Dark
  600. 24:39Materials series by Philip Pullman,
  601. 24:42Eragon, the Inheritance Cycle series by
  602. 24:45Christopher Paolini, Magician
  603. 24:47Apprentice, the Riftwar Saga series by
  604. 24:51Raymond E. Feist, Daughter of the Empire
  605. 24:54trilogy by Raymond E. Feist and Janny
  606. 24:57Wurts. This is an offshoot of the
  607. 25:00Magician Apprentice series, which I
  608. 25:02think is actually superior.
  609. 25:04Earthsea trilogy by Ursula K. Le Guin,
  610. 25:08The Once and Future King by T. H. White,
  611. 25:10Chronicles of Thomas Covenant series by
  612. 25:13Stephen R. Donaldson, Harry Potter
  613. 25:16series by J. K. Rowling, American Gods
  614. 25:19by Neil Gaiman, Jonathan Strange and Mr
  615. 25:22Norrell by Susanna Clark and The
  616. 25:25Magicians by Lev Grossman. Let me just
  617. 25:27say a bit about some of these novels.
  618. 25:30Everyone has read Lord of the Rings. I
  619. 25:32love the trilogy, but I have to admit
  620. 25:34that Tolkien is not great about delving
  621. 25:37deep into his characters. The more
  622. 25:39recent novels and series are much better
  623. 25:41at that. I love Game of Thrones, but
  624. 25:44George R. R. Martin has one of the worst
  625. 25:46cases of writer's block ever. I doubt he
  626. 25:49will ever finish. The same applies to
  627. 25:51Patrick Rothfuss. The Wheel of Time
  628. 25:54series is off the charts good and I
  629. 25:56mourn the fact that Amazon canceled the
  630. 25:58series on Prime. One caveat, the Wheel
  631. 26:01of Time is composed of 14 books, all of
  632. 26:04them very long. It's a commitment. It
  633. 26:07took so long to write that the author
  634. 26:09Robert Jordan died after book 11.
  635. 26:12Brandon Sanderson finished it and did a
  636. 26:14great job. I highly recommend the Codex
  637. 26:16Alera series by Jim Butcher, which is a
  638. 26:19story basically about Rome with magic.
  639. 26:21It's great, but in the end, my favorite
  640. 26:24is the works of Robin Hobb. Her several
  641. 26:27series are mostly about a character
  642. 26:29called Fitz Chivalry Farseer, who I
  643. 26:32think is one of the greatest characters
  644. 26:34in all of literature. I'm being
  645. 26:36perfectly serious. Most fantasy novels
  646. 26:38are about the hero's journey. Robin Hobb
  647. 26:41wrote a story about a hero's life.
  648. 26:44And that's [music] the wrap.
  649. 26:47This podcast is for informational
  650. 26:49purposes only and does not constitute
  651. 26:51investment advice. The host and guests
  652. 26:53may hold positions in stocks discussed.
  653. 26:55Opinions expressed are their own and not
  654. 26:56recommendations. Please do your own due
  655. 26:59diligence and consult a licensed
  656. 27:00financial advisor before making any
  657. 27:02investment decisions.
  658. 27:04>> [music]

About this transcript

This page contains the full transcript of The AI Debate is No Longer Boom vs Bust: Why It’s Time to Go Back to Stock Picking | The Weekly Wrap by Steve Eisman, generated from the public captions YouTube serves with the video. The transcript has 4,369 words across 658 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.