YouTube2Text

Nvidia’s Troubles Could Become America’s Crisis — Transcript

by dummy finance · 3,883 words · 486 segments · language en · Watch on YouTube

Full transcript

  1. 0:00I think most people are reading Nvidia’s terrible
  2. 0:02month backwards. They look at the falling charts
  3. 0:05and see a story about one company, one
  4. 0:07missing chief executive, and a luxury problem for
  5. 0:10rich tech investors. But in reality, if the
  6. 0:13artificial intelligence boom breaks the way the last
  7. 0:16tech boom did, the biggest holdings in your
  8. 0:18retirement fund do not just take a temporary
  9. 0:21dip, they fall for years. After two thousand,
  10. 0:24the Nasdaq lost more than three quarters of
  11. 0:26its value and took fifteen years to get
  12. 0:29back to even, and in two thousand and
  13. 0:31eight, the average workplace retirement account that people
  14. 0:33held all the way through lost about a
  15. 0:36third of its value in a single year.
  16. 0:38So with Nvidia down about thirty percent and
  17. 0:41the man who ran it gone, does the
  18. 0:43artificial intelligence boom break right here, and how
  19. 0:46much of your four-oh-one-k goes with it? On
  20. 0:48November twelfth, twenty twenty-six, Nvidia announced that its
  21. 0:52chief executive had resigned, an exit that actually
  22. 0:54took effect the day before, and that its
  23. 0:57finance chief, Colette Kress, would step in to
  24. 0:59run the company while the board looks for
  25. 1:01a permanent replacement. Four days before that announcement,
  26. 1:04the company had published an email in which
  27. 1:06he admitted an affair and backed an independent
  28. 1:09board inquiry; that investigation is not finished, and
  29. 1:12nothing has been proven. The official announcement gave
  30. 1:15no reason at all. But investors did not
  31. 1:17wait for reasons. By the close of trading
  32. 1:20on November twelfth, Nvidia’s shares stood about thirty
  33. 1:23percent below their peak of just five weeks
  34. 1:25earlier. Let us invent two of the people
  35. 1:27that kind of drop actually lands on, because
  36. 1:30otherwise this whole thing sounds like abstract digital
  37. 1:33weather. Meet Carol and her son Ben. Carol
  38. 1:36is sixty years old, works an ordinary job,
  39. 1:38and has never bought an individual stock in
  40. 1:40her life. Her four-oh-one-k holds about a hundred
  41. 1:43and seven thousand dollars, which is about the
  42. 1:46median balance at Vanguard for savers aged fifty-five
  43. 1:49to sixty-four. Her money sits entirely in a
  44. 1:52single target-date fund designed for someone retiring around
  45. 1:55twenty thirty, meaning she has five years left
  46. 1:58before she plans to stop working. Her son
  47. 2:00Ben is thirty, has about nineteen thousand dollars
  48. 2:03in the target-date fund for people his age,
  49. 2:05and has thirty-five years of work ahead of
  50. 2:07him. Neither of them ever picked a chip
  51. 2:09stock, but their retirement accounts did it for
  52. 2:12them. My own impression of that five-week slide
  53. 2:15is that the market is finally waking up
  54. 2:17to the real danger, which was never about
  55. 2:19an executive's personal life. To see why one
  56. 2:22chipmaker’s bad month reaches accounts like Carol’s, you
  57. 2:25have to look at where Nvidia’s revenue actually
  58. 2:27comes from. Most people assume a computer hardware
  59. 2:30company operates like an appliance store. You build
  60. 2:33a machine, a customer walks in, pays cash
  61. 2:36for it, carries it away, and whatever happens
  62. 2:38to that customer tomorrow is entirely their own
  63. 2:41business. That is what the word customer usually
  64. 2:43promises. But in the artificial intelligence trade, that
  65. 2:47is not how the room is set up.
  66. 2:48Take CoreWeave, a firm that buys Nvidia’s processors
  67. 2:52and rents out the computing power. In January
  68. 2:54twenty twenty-six, Nvidia paid cash to buy a
  69. 2:57block of CoreWeave’s shares. Then, under an agreement
  70. 3:00signed in September twenty twenty-five, Nvidia legally promised
  71. 3:04that it would buy whatever computing capacity CoreWeave
  72. 3:07reserves and cannot sell to anyone else, all
  73. 3:10the way through April twenty thirty-two. So Nvidia
  74. 3:13is CoreWeave’s supplier, one of its shareholders, and
  75. 3:16the buyer of whatever capacity it cannot rent
  76. 3:19out. Calling a company your customer when you
  77. 3:21own its shares and have contractually agreed to
  78. 3:23buy back whatever it fails to sell is
  79. 3:26a rather charming use of the English language.
  80. 3:28And that is only one room in the
  81. 3:30house. If you add up what Nvidia’s own
  82. 3:32filings say it has promised on behalf of
  83. 3:34its buyers, the total comes to about a
  84. 3:36hundred and sixty-five billion dollars. To be fair
  85. 3:40and exact, that figure is not money already
  86. 3:42lost and it is not corporate debt; it
  87. 3:45is the most those promises could cost. Those
  88. 3:47commitments are made of three very different things:
  89. 3:50promises to buy computing capacity its partners cannot
  90. 3:53sell, leases Nvidia has taken on data centers
  91. 3:56meant for others, and guarantees to pay landlords
  92. 3:59if a tenant defaults. Only that third bucket
  93. 4:02waits for someone to default; the first two
  94. 4:04fall due on their own schedule. The largest
  95. 4:07single piece is a guarantee signed in August
  96. 4:09twenty twenty-six on a cluster of data centers
  97. 4:12in Ohio being built for OpenAI on twenty-year
  98. 4:15leases, where Nvidia has promised to pay the
  99. 4:17rent if the tenant stops paying. On top
  100. 4:19of that, Nvidia holds about a hundred billion
  101. 4:22dollars of shares in other companies, and its
  102. 4:24unpaid customer bills have been stretching out for
  103. 4:27months. In fact, Nvidia’s finance chief stated in
  104. 4:30August that the artificial intelligence labs backed by
  105. 4:33Nvidia’s own balance sheet are expected to bring
  106. 4:35in roughly a quarter of its business next
  107. 4:38year. The cash changing hands is real, but
  108. 4:40so is the loop. When a seller has
  109. 4:42to stand behind its own buyers, it does
  110. 4:45not get to act surprised when a wobble
  111. 4:47on one side of the table shakes the
  112. 4:48other. Which brings us directly to the next
  113. 4:51question: can those customer companies actually afford to
  114. 4:54pay their own debts? When you look at
  115. 4:56the specialist firms doing the borrowing, the accounting
  116. 4:59looks remarkably thin. In the three months to
  117. 5:01September twenty twenty-six, CoreWeave owed about four times
  118. 5:05as much in interest as its underlying business
  119. 5:07generated in operating profit, and it finished the
  120. 5:10quarter with a loss. It owes tens of
  121. 5:12billions of dollars, and some of that debt
  122. 5:15is borrowed directly against the chips themselves. On
  123. 5:18November fifth, twenty twenty-six, some of CoreWeave’s bonds
  124. 5:21changed hands for as little as about seventy-five
  125. 5:24cents on the dollar. That means an investor
  126. 5:26holding a legal promise to receive a dollar
  127. 5:28in twenty thirty-two was willing to hand it
  128. 5:30to a stranger for three quarters of a
  129. 5:32dollar just to be rid of the paperwork.
  130. 5:35That seventy-five-cent price is an open-market trading price,
  131. 5:38a gauge of the lenders' anxiety rather than
  132. 5:40the borrowing rate the company locked in, but
  133. 5:43the anxiety is unmistakable. And it is not
  134. 5:46only CoreWeave: lenders are asking the biggest spenders
  135. 5:49for more too. Vanguard even issued a warning
  136. 5:52in September twenty twenty-six, reminding everyday savers that
  137. 5:55as these tech firms borrow, that corporate debt
  138. 5:58quietly slips into the broad, conservative bond funds
  139. 6:01that people like Carol rely on to preserve
  140. 6:03their savings. So if the weaker links are
  141. 6:06already feeling the strain, what would it actually
  142. 6:08take to bring this boom down? It would
  143. 6:10take a simple change of mind inside a
  144. 6:13very small number of boardrooms. The real weight
  145. 6:15in this market sits with four companies: Microsoft,
  146. 6:19Alphabet, Amazon, and Meta. In twenty twenty-six, those
  147. 6:23four firms planned to spend more than seven
  148. 6:25hundred billion dollars between them on buildings and
  149. 6:28equipment, much of it for artificial intelligence. That
  150. 6:31river of spending is where Nvidia’s sales come
  151. 6:34from, it is the reason the smaller cloud
  152. 6:36providers can borrow, and it is the foundation
  153. 6:38Wall Street is leaning on. But pouring that
  154. 6:41much money into concrete and silicon is beginning
  155. 6:44to stretch even the largest corporate treasuries. Over
  156. 6:47the first nine months of twenty twenty-six, Alphabet
  157. 6:49spent slightly more on buildings and equipment than
  158. 6:52its whole business brought in in cash, and
  159. 6:55it is not alone; some of them are
  160. 6:56now spending more cash than their businesses bring
  161. 6:59in. The comfortable assumption is that all this
  162. 7:01rests on something solid, because people hear the
  163. 7:04phrase order book and picture a vault of
  164. 7:07guaranteed income. But if you open Nvidia’s own
  165. 7:10annual filing and read the plain print, it
  166. 7:12says its customers can generally cancel, change, or
  167. 7:15delay their orders with little notice and without
  168. 7:18penalty. Those are not ironclad sales; they are
  169. 7:21informal reservations. So the whole structure depends on
  170. 7:24four companies choosing to keep spending at this
  171. 7:27historic pace, quarter after quarter, on orders they
  172. 7:30can call off for free. To see what
  173. 7:32happens if they change their minds, one major
  174. 7:34bank ran the arithmetic in November twenty twenty-six:
  175. 7:38if those big spenders reduce their capital budgets
  176. 7:40by twenty percent in twenty twenty-seven, the companies
  177. 7:43that make artificial intelligence chips would lose more
  178. 7:46than a third of their profits, and possibly
  179. 7:48more. That twenty percent cut is a what-if
  180. 7:51scenario, not a forecast that it will happen,
  181. 7:54but it is the clearest way this entire
  182. 7:56market could unwind. When hundreds of billions of
  183. 7:59dollars of revenue depend on four chief executives
  184. 8:02not calling off orders they can cancel at
  185. 8:04no cost, you are leaning entirely on human
  186. 8:07mood. And a spending cut is not the
  187. 8:09only way the boom could stumble, especially now
  188. 8:12that Nvidia’s chief executive is gone. Back in
  189. 8:15July twenty twenty-six, officials at the Federal Reserve
  190. 8:18explicitly warned that if artificial intelligence fails to
  191. 8:21live up to corporate expectations, the disappointment could
  192. 8:24knock share prices down broadly and cause people
  193. 8:27to cut back on their spending. To be
  194. 8:29fair, Jensen Huang’s exit was not the sole
  195. 8:32cause of the stock’s five-week slide; about seven
  196. 8:35tenths of that thirty percent drop had already
  197. 8:37happened before the email was published on November
  198. 8:40eighth. The market was already nervous about something
  199. 8:43wider than one person. What a change at
  200. 8:45the top can alter is the financing arrangements
  201. 8:48Nvidia uses to keep its customers buying. The
  202. 8:50contracts already signed bind the company whoever sits
  203. 8:54in the corner office; the Ohio guarantee, for
  204. 8:56instance, belongs to Nvidia as a corporation. But
  205. 9:00offering new money to buyers is an ongoing
  206. 9:02choice, and on the day of the resignation,
  207. 9:04investors openly worried that a new chief executive
  208. 9:07might be far less willing to bankroll customer
  209. 9:10purchases. Analysts at one bank even raised the
  210. 9:13possibility that about fifty billion dollars of Nvidia's
  211. 9:16funding for other companies could be put on
  212. 9:18hold. No pause has been officially announced, and
  213. 9:21nobody can yet know how a new leader
  214. 9:23will manage those checkbooks. My view of Wall
  215. 9:26Street trying to predict that transition is that
  216. 9:28they are guessing out loud in very confident
  217. 9:31voices, because admitting they do not know is
  218. 9:33the one thing they are not allowed to
  219. 9:35do. Whenever an essential supplier loses thirty percent
  220. 9:39of its value while insisting its technology will
  221. 9:41change civilization, people reach for the history books,
  222. 9:44and the chapter they open is March two
  223. 9:46thousand. Back then, the undisputed centerpiece of the
  224. 9:50digital frontier was Cisco Systems. On March twenty-seventh,
  225. 9:54two thousand, Cisco became the most valuable public
  226. 9:57corporation on earth. It was not an empty
  227. 10:00dot-com startup; it sold the equipment the internet
  228. 10:02was being built with, and it made billions
  229. 10:04of dollars a year in real profit. But
  230. 10:07when its customers stopped building, Cisco’s business stalled.
  231. 10:10In the following year, it was writing off
  232. 10:12billions of dollars of equipment it could not
  233. 10:15sell and reporting a loss, even while cash
  234. 10:17kept flowing through the doors. The expectation every
  235. 10:20investor is taught is that if a company
  236. 10:22is real, essential, and profitable, and if it
  237. 10:25survives the bust, its share price will inevitably
  238. 10:28recover; you only have to wait. Cisco survived,
  239. 10:32it remained an indispensable supplier, and the internet
  240. 10:35turned out to be everything its believers said
  241. 10:37it would be. But its share price did
  242. 10:39not close above that March two thousand high
  243. 10:41until December tenth, twenty twenty-five, more than twenty-five
  244. 10:45years later. A shareholder who was right about
  245. 10:47the technology, right about the company, and right
  246. 10:50about its survival still had to wait a
  247. 10:52quarter of a century just to see the
  248. 10:54share price they paid again. And Cisco was
  249. 10:57not the only hardware giant that learned that
  250. 10:59lesson. Lucent Technologies lent its own customers the
  251. 11:02money to buy its equipment, promising loans and
  252. 11:05guarantees so they could keep buying its products,
  253. 11:07and when those customers could not pay, Lucent
  254. 11:10charged nearly two billion dollars against that bad
  255. 11:13lending in a single year; its shares lost
  256. 11:15nearly all their value. When that broader market
  257. 11:18cleared, the Standard and Poor’s five hundred had
  258. 11:21lost about half its value and took seven
  259. 11:23years to get back to its peak. None
  260. 11:25of that proves Nvidia will walk the same
  261. 11:27path, but it demonstrates that a technology can
  262. 11:30triumph while the people who bought the hardware
  263. 11:32supplier at peak prices get taken apart. That
  264. 11:35was their stock twenty-five years ago. The question
  265. 11:38that matters to Carol is how much of
  266. 11:40this current machine sits inside her own four-oh-one-k
  267. 11:43without her ever choosing it. At Vanguard, more
  268. 11:46than four fifths of savers in workplace plans
  269. 11:48keep their money in a target-date fund. In
  270. 11:51an illustration built from Vanguard’s published asset allocation
  271. 11:54and portfolio holdings from late September twenty twenty-six,
  272. 11:57a typical retirement account holds about two to
  273. 12:00four cents of Nvidia in every single dollar.
  274. 12:03Older workers like Carol sit near the low
  275. 12:05end of that range, while younger workers like
  276. 12:07Ben sit near four cents, because target-date funds
  277. 12:10automatically reduce stock exposure as a worker nears
  278. 12:13retirement. That means Nvidia's thirty percent drop over
  279. 12:17the last five weeks, taken entirely by itself,
  280. 12:20has cost Carol about eight hundred and fifty
  281. 12:22dollars of her hundred and seven thousand. That
  282. 12:24eight hundred and fifty dollars sounds entirely manageable,
  283. 12:27but stopping there is a mistake, because the
  284. 12:30real exposure is not Nvidia alone. An index
  285. 12:33fund tracking the Standard and Poor’s five hundred
  286. 12:36is sold as the ultimate sensible investment because
  287. 12:38it spreads your savings across five hundred distinct
  288. 12:41American companies; that is the promise printed on
  289. 12:44the brochure. But by November eleventh, twenty twenty-six,
  290. 12:47the ten largest corporations in that index made
  291. 12:50up about thirty-eight cents of every single dollar
  292. 12:53in the fund. At the peak of the
  293. 12:55dot-com bubble in March two thousand, the top
  294. 12:57ten companies were only about twenty-seven cents of
  295. 13:00each dollar. Today, eight companies closely tied to
  296. 13:03the artificial intelligence build-out represent more than a
  297. 13:06quarter of the entire fund. In fact, by
  298. 13:09one major fund company’s own calculation, the Standard
  299. 13:12and Poor’s five hundred now spreads money about
  300. 13:14as widely as holding fifty equally sized companies.
  301. 13:18For Carol, the ten largest American corporations represent
  302. 13:21about an eighth of her entire account, bonds
  303. 13:24included, while for Ben they make up about
  304. 13:26a fifth. And neither of them can instruct
  305. 13:28the fund to sell its Nvidia; you own
  306. 13:31the mutual fund, not the companies inside it.
  307. 13:34It is a peculiar kind of diversification that
  308. 13:36promises you five hundred companies and hands you
  309. 13:38fifty, with nearly four dollars in ten riding
  310. 13:41on ten corporate boards. What that concentration does
  311. 13:44to your savings depends heavily on how many
  312. 13:47years you have left on the job, which
  313. 13:49is the deeply unfair part of retirement finance.
  314. 13:52The funds are built to soften blows by
  315. 13:54shifting into bonds, keeping about nine tenths of
  316. 13:57Ben's money in stocks at age thirty and
  317. 13:59about six tenths of Carol's at age sixty.
  318. 14:02But that mechanical shift does not make an
  319. 14:04account immune to a broad collapse. During the
  320. 14:07financial crisis of two thousand and eight, the
  321. 14:09average workplace account held all the way through
  322. 14:12lost about a third of its value in
  323. 14:14a single twelve-month stretch. If a similar broad
  324. 14:17contraction hit the equity markets tomorrow, that average
  325. 14:20fall would take about thirty-seven thousand dollars from
  326. 14:23an account the size of Carol's. That is
  327. 14:25an illustration of the size, not a forecast
  328. 14:28for her fund. We saw this exact generational
  329. 14:31divide during the dot-com bust. Savers who were
  330. 14:34in their sixties when that bubble burst had,
  331. 14:36on average, about a fifth less in their
  332. 14:38accounts three years later, and even after contributing
  333. 14:41fresh money from their paychecks for an extra
  334. 14:43year, they were still short of where they
  335. 14:45started four years into the slump. Meanwhile, savers
  336. 14:49who were in their twenties saw their average
  337. 14:51balances grow through the crash, because the fresh
  338. 14:53contributions they poured in each month were large
  339. 14:56compared to the modest savings they had accumulated.
  340. 14:59That is Ben’s structural cushion: a market crash
  341. 15:02acts as a long-term sale for a thirty-year-old,
  342. 15:05provided he keeps his job and keeps buying.
  343. 15:07But for money already sitting in a Standard
  344. 15:09and Poor’s five hundred fund with nothing new
  345. 15:12added, an investor who held at the end
  346. 15:14of nineteen ninety-nine was not back to even
  347. 15:16until the end of two thousand and six.
  348. 15:19And the damage is far worse for anyone
  349. 15:21who has already retired and must sell shares
  350. 15:23each month to buy groceries, because those depressed
  351. 15:26shares are permanently gone from the ledger when
  352. 15:29the market finally recovers. I dislike the idea
  353. 15:32that a worker five years from retirement is
  354. 15:34told she was reckless, when all she did
  355. 15:36was enroll in the default fund her employer
  356. 15:39selected. Now, before anyone concludes that the entire
  357. 15:42enterprise is falling apart tomorrow, we should look
  358. 15:45fairly at Nvidia’s actual business, because the people
  359. 15:48arguing that the company is wildly profitable are
  360. 15:51describing a real balance sheet. In the six
  361. 15:53months to late July twenty twenty-six, Nvidia’s operations
  362. 15:57brought in approximately seventy-four billion dollars of pure
  363. 16:00cash. It holds considerably more in cash and
  364. 16:03bonds than it owes in debt. Furthermore, issuing
  365. 16:06a financial guarantee is not the same as
  366. 16:09writing a check today: the large Ohio commitment
  367. 16:11is not expected to begin phasing in until
  368. 16:14its twenty twenty-nine financial year. Those are genuine
  369. 16:17strengths, but a sound operating business is not
  370. 16:20the same thing as a safe share price;
  371. 16:22Cisco was generating billions in cash at its
  372. 16:24peak too. The uncomfortable detail is that every
  373. 16:27financial report Nvidia has published covers performance from
  374. 16:31before the chief executive resigned, with the next
  375. 16:33official scorecard not arriving until November seventeenth. Beneath
  376. 16:37the surface, customers are already taking longer to
  377. 16:40pay, with the average collection window stretching from
  378. 16:43forty-five days to sixty in a single quarter.
  379. 16:46And two of Nvidia's largest commitments require no
  380. 16:49customer default to fall due: it has pledged
  381. 16:51to buy about thirty-six billion dollars of computing
  382. 16:54capacity from its partners over roughly six years
  383. 16:57if they cannot rent it out, and it
  384. 16:59has signed leases on data centers that it
  385. 17:01must pay for if nobody else takes them
  386. 17:03over. A healthy income statement is comforting, but
  387. 17:06it does not protect you when your future
  388. 17:08sales depend on propping up the very people
  389. 17:10who buy from you. Nor does the broader
  390. 17:12market’s apparent calm prove that the danger has
  391. 17:15passed. When you look at headline index numbers,
  392. 17:18the system appears remarkably steady. Even as Nvidia
  393. 17:21dropped thirty percent over five weeks, the broader
  394. 17:24Standard and Poor’s five hundred ended November twelfth
  395. 17:27just three percent below its all-time record, sitting
  396. 17:30comfortably in positive territory for the year, while
  397. 17:32corporate profits grew across every sector. Yet that
  398. 17:36calm cracked in plain sight on November fifth,
  399. 17:39when the Standard and Poor’s five hundred fell
  400. 17:41more than four percent in a single day.
  401. 17:43Semiconductor shares as a group ended November eleventh
  402. 17:46down more than a quarter from their June
  403. 17:48peak. The index has held together so far,
  404. 17:51but yesterday’s stability is no guarantee that the
  405. 17:53rest of the market stays contained. Ever since
  406. 17:56two thousand and eight, many savers hold an
  407. 17:58instinctive belief that if things get bad enough,
  408. 18:01a government bailout will step in to make
  409. 18:03them whole, because a rescue was sold back
  410. 18:05then as protecting the broader economy. But under
  411. 18:08rules adopted by the Federal Reserve in twenty
  412. 18:11fifteen, the central bank is explicitly barred from
  413. 18:14lending to rescue one particular failing company. The
  414. 18:17White House’s own adviser on artificial intelligence stated
  415. 18:21publicly a year ago that there will be
  416. 18:23no federal bailout for artificial intelligence corporations. Even
  417. 18:27when bailouts do arrive, the word means something
  418. 18:29very different from protecting shareholders. When the federal
  419. 18:32government rescued General Motors, the factories kept running
  420. 18:36and the business survived, but the old company’s
  421. 18:38shares were canceled in March twenty eleven with
  422. 18:41nothing paid to the people who owned them.
  423. 18:43A rescue protects the business and its operations;
  424. 18:46the shareholders are the ones it is rescued
  425. 18:48from. And common shares are exactly what sit
  426. 18:51inside Carol’s four-oh-one-k. So how real is the
  427. 18:55worst case, has the boom broken, and how
  428. 18:58much of your four-oh-one-k has actually gone with
  429. 19:00it? The answer, based on the documented evidence,
  430. 19:03is not so far, but the risk is
  431. 19:05very real. It has not broken yet because
  432. 19:08while Nvidia fell about thirty percent, the broader
  433. 19:11Standard and Poor’s five hundred dropped only about
  434. 19:14three percent, and an age-balanced fund like Carol’s
  435. 19:16lost less than one percent between Nvidia’s peak
  436. 19:19and the day before the announcement. Ben’s fund,
  437. 19:22with its heavy stock weighting, lost only a
  438. 19:24little more. In funds like theirs, Nvidia itself
  439. 19:27makes up only about two to four cents
  440. 19:29of every dollar, which is too small to
  441. 19:32sink an account on its own. The deciding
  442. 19:34reason for the danger lies elsewhere: the risk
  443. 19:37is that Nvidia, the cloud operators it finances,
  444. 19:40the bondholders funding those clouds, and the ten
  445. 19:42corporate giants that make up thirty-eight cents of
  446. 19:45every index dollar are all leaning on the
  447. 19:47exact same decision. That decision is whether a
  448. 19:50handful of tech giants keep spending more than
  449. 19:53seven hundred billion dollars a year on buildings
  450. 19:55and equipment, much of it for artificial intelligence.
  451. 19:58If they reduce those budgets, the consequences move
  452. 20:01the entire top of the market at once;
  453. 20:04as the bank’s scenario showed, a twenty percent
  454. 20:06cut in spending takes more than a third
  455. 20:08off chipmaker profits. That is the clearest path
  456. 20:11to the worst case, alongside a wider loss
  457. 20:14of faith in artificial intelligence earnings or a
  458. 20:17pullback in financing under Nvidia’s new management. What
  459. 20:20will show which way this goes is not
  460. 20:22the daily gossip, but the numbers: Nvidia’s results
  461. 20:25on November seventeenth, and the capital expenditure budgets
  462. 20:28published by Microsoft, Alphabet, Amazon, and Meta over
  463. 20:32the coming quarters. If those budgets do get
  464. 20:35cut, the cost will not be shared equally.
  465. 20:37A retirement saver does not face the same
  466. 20:40clock at thirty as they do at sixty.
  467. 20:42If the artificial intelligence build-out falters, history shows
  468. 20:46that an indispensable supplier like Cisco can spend
  469. 20:49twenty-five years waiting to see its old share
  470. 20:51price again, while an equity index can spend
  471. 20:54seven years getting back to even. Ben has
  472. 20:57thirty-five years of paychecks ahead of him; he
  473. 20:59will keep buying through any downturn, picking up
  474. 21:02shares at a discount. Carol has five years
  475. 21:04until she plans to retire, with an account
  476. 21:07holding a hundred and seven thousand dollars that
  477. 21:09she will soon need to convert into monthly
  478. 21:11living expenses. And there is no administrative rule
  479. 21:14on the books that will bail out the
  480. 21:16shares inside her account. Nothing has broken yet.
  481. 21:19I don't think that is the same as
  482. 21:21safe: it only means the few companies everybody
  483. 21:24is leaning on are still spending. Thanks for
  484. 21:26watching! Please subscribe, it's free, it genuinely helps,
  485. 21:28and watch this next, it's right there, okay,
  486. 21:29thank you, bye!

About this transcript

This page contains the full transcript of Nvidia’s Troubles Could Become America’s Crisis by dummy finance, generated from the public captions YouTube serves with the video. The transcript has 3,883 words across 486 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.