1 00:00:00,040 --> 00:00:03,280 When a company decides to go public, they're required by the 2 00:00:03,280 --> 00:00:06,840 Securities and Exchange Commission to file an S1 3 00:00:06,840 --> 00:00:09,440 prospectus. Historically, these are some of 4 00:00:09,440 --> 00:00:12,280 the driest documents ever committed to paper. 5 00:00:12,520 --> 00:00:15,840 They're written by corporate lawyers for other corporate 6 00:00:15,840 --> 00:00:20,280 lawyers to ensure that no one, anywhere is ever accidentally 7 00:00:20,280 --> 00:00:24,000 entertained. But two days ago, SpaceX filed 8 00:00:24,040 --> 00:00:28,000 its paperwork to go public, and reading through it, it became 9 00:00:28,000 --> 00:00:31,800 immediately clear that this was not going to be your typical 10 00:00:31,800 --> 00:00:34,360 legal disclosure. First start. 11 00:00:34,360 --> 00:00:38,400 It opened with 14 pages of photographs of rockets and 12 00:00:38,400 --> 00:00:42,880 satellites, which reminded me of Adam Newman's I'll fated Wework 13 00:00:42,880 --> 00:00:46,360 prospectus, which famously opened with the line. 14 00:00:46,680 --> 00:00:51,280 We dedicate this to the energy of we greater than any one of 15 00:00:51,280 --> 00:00:55,200 us, but inside each of us. That document was later 16 00:00:55,200 --> 00:00:59,040 described as one of the worst IPO filings in Wall Street 17 00:00:59,040 --> 00:01:02,120 history. The SpaceX prospectus is 18 00:01:02,120 --> 00:01:07,160 similarly unconventional, though I will say the Rockets are more 19 00:01:07,160 --> 00:01:10,520 photogenic than the Wework community photographs were. 20 00:01:11,040 --> 00:01:14,840 The document is also filled with phrases that I'm fairly 21 00:01:14,840 --> 00:01:18,640 confident have never appeared in an SEC filing before. 22 00:01:19,080 --> 00:01:22,160 Under the Business Strategy section, right alongside 23 00:01:22,160 --> 00:01:25,760 standard disclosures about lease agreements and accounting 24 00:01:25,760 --> 00:01:28,560 standards. Management explains that their 25 00:01:28,560 --> 00:01:32,760 primary objective is to extend the light of consciousness to 26 00:01:32,760 --> 00:01:37,120 the stars, a phrase that appears 10 separate times in the filing, 27 00:01:37,320 --> 00:01:39,880 which suggests that it wasn't a typo. 28 00:01:40,480 --> 00:01:44,520 Investors are also warned that we do not want humans to have 29 00:01:44,520 --> 00:01:48,360 the same fade as dinosaurs, which is a fairpoint. 30 00:01:48,360 --> 00:01:50,400 I suppose. It's the kind of sentence that 31 00:01:50,400 --> 00:01:53,760 gets you excited about what the risk disclosure section will 32 00:01:53,760 --> 00:01:56,840 look like. The document then goes on to 33 00:01:56,920 --> 00:02:00,880 outline a string of future business lines that read like 34 00:02:00,880 --> 00:02:04,360 science fiction. They include long haul, point to 35 00:02:04,360 --> 00:02:08,160 point terrestrial travel, which I think means travelling around 36 00:02:08,160 --> 00:02:10,800 the Earth in rockets rather than in airplanes. 37 00:02:11,120 --> 00:02:13,280 In orbit manufacturing? Yes. 38 00:02:13,480 --> 00:02:17,840 In orbit passenger and cargo transportation to the Moon and 39 00:02:17,840 --> 00:02:22,400 Mars, manufacturing and energy production on the Moon and Mars, 40 00:02:22,640 --> 00:02:27,360 and asteroid mining. Because why mine for resources 41 00:02:27,360 --> 00:02:31,000 on Earth when you can chase down an asteroid in the hope that 42 00:02:31,000 --> 00:02:33,040 you'll find useful minerals on it? 43 00:02:33,720 --> 00:02:38,840 The acronym AI appears over 200 times in the document. 44 00:02:39,040 --> 00:02:42,080 The word consciousness appears 10 times. 45 00:02:42,240 --> 00:02:46,040 The word profit appears somewhat less frequently than either of 46 00:02:46,040 --> 00:02:48,760 those. What we're looking at is quite 47 00:02:48,760 --> 00:02:52,320 possibly the most extraordinary document in the history of 48 00:02:52,320 --> 00:02:56,160 American securities regulation. And today we're going to go 49 00:02:56,160 --> 00:02:58,880 through it. Because every single thing I'm 50 00:02:58,880 --> 00:03:02,920 about to tell you, no matter how absurd it sounds, comes directly 51 00:03:02,920 --> 00:03:06,280 from the filing. If you've been following SpaceX 52 00:03:06,280 --> 00:03:10,000 for the last two decades, you probably think of it as a rocket 53 00:03:10,000 --> 00:03:12,600 company. In recent years, with the rapid 54 00:03:12,600 --> 00:03:16,680 expansion of Starlink, you might have updated that to a satellite 55 00:03:16,680 --> 00:03:20,000 Internet company. But if you read the IPO 56 00:03:20,000 --> 00:03:23,280 prospectus, you're in for a surprise. 57 00:03:23,280 --> 00:03:26,080 SpaceX is now apparently an AI company. 58 00:03:26,520 --> 00:03:30,720 According to the document, 93% of the company's total 59 00:03:30,720 --> 00:03:35,640 addressable market is attributed to AI, and based on the capital 60 00:03:35,640 --> 00:03:40,800 expenditure breakdown, roughly 60% of all spending is now going 61 00:03:40,800 --> 00:03:43,680 into artificial intelligence infrastructure. 62 00:03:44,240 --> 00:03:48,880 The prospectus breaks down the business into 3 segments, Space, 63 00:03:49,000 --> 00:03:53,800 connectivity, and AI. In the first quarter of 2026, 64 00:03:54,000 --> 00:03:57,000 the space division, the part that actually builds and 65 00:03:57,000 --> 00:04:01,440 launches rockets, lost $662,000,000. 66 00:04:01,960 --> 00:04:06,120 The connectivity division, which is Starlink, turned a $1.1 67 00:04:06,120 --> 00:04:10,360 billion profit. The AI division lost 2 1/2 68 00:04:10,360 --> 00:04:13,720 billion dollars in just the last three months. 69 00:04:14,120 --> 00:04:18,079 The only division that makes money is the satellite Internet 70 00:04:18,079 --> 00:04:20,760 business. Everything else loses money. 71 00:04:21,360 --> 00:04:25,760 When Musk says that this is an AI company, what he means is 72 00:04:25,760 --> 00:04:29,360 that the one business that actually works is subsidizing a 73 00:04:29,360 --> 00:04:34,040 money losing AI venture while he tries to convince public market 74 00:04:34,040 --> 00:04:38,760 investors that the money losing AI venture is the entire reason 75 00:04:39,000 --> 00:04:42,280 the company's worth $1.75 trillion. 76 00:04:42,880 --> 00:04:46,680 If we look past the talk of extending consciousness to the 77 00:04:46,680 --> 00:04:50,920 stars and focus purely on the income statement, the financial 78 00:04:50,920 --> 00:04:57,000 reality of SpaceX is bracing. In 2025, the company generated 79 00:04:57,080 --> 00:05:03,240 $18.7 billion in revenue. Now $18.7 billion is a lot of 80 00:05:03,240 --> 00:05:08,000 money, but at a proposed valuation of $1.75 trillion, 81 00:05:08,200 --> 00:05:12,960 SpaceX is asking investors to value it at almost 100 times 82 00:05:12,960 --> 00:05:15,880 it's revenue. If SpaceX were dropped into the 83 00:05:15,880 --> 00:05:20,440 S&P 500 today, it would be the 7th largest company in the index 84 00:05:20,440 --> 00:05:24,040 by market cap. But if you ranked it by actual 85 00:05:24,120 --> 00:05:27,160 revenue, it wouldn't even meet 200. 86 00:05:27,440 --> 00:05:30,840 It would be down near Kellogg's, the maker of fruit Luke's 87 00:05:30,840 --> 00:05:33,400 breakfast cereal. You get the point. 88 00:05:33,880 --> 00:05:37,400 Unlike Kellogg's, SpaceX is not profitable. 89 00:05:37,640 --> 00:05:41,720 Space exploration is just a tougher business than breakfast. 90 00:05:42,000 --> 00:05:47,160 Last year, SpaceX posted a net loss of $4.94 billion. 91 00:05:47,440 --> 00:05:50,840 What makes this number remarkable is that Starlink 92 00:05:50,880 --> 00:05:55,280 actually generated $4.4 billion in operating profit. 93 00:05:55,680 --> 00:05:59,400 That means the rest of the business ate up more than $9 94 00:05:59,400 --> 00:06:05,000 billion in a single year. Since its founding, SpaceX has 95 00:06:05,000 --> 00:06:10,400 accumulated over $37 billion in losses, more than any other 96 00:06:10,400 --> 00:06:13,440 company that has ever gone public in history. 97 00:06:13,760 --> 00:06:16,960 And there are no signs of the cash burn slowing down. 98 00:06:17,280 --> 00:06:21,640 Capital expenditure reached $20.7 billion last year, 99 00:06:21,920 --> 00:06:27,640 quintupling from 2023, driven by Starlink expansion, Starship 100 00:06:27,640 --> 00:06:30,760 development, and AI infrastructure spending. 101 00:06:31,080 --> 00:06:36,240 To fund all of this, the company is carrying $29 billion in debt, 102 00:06:36,440 --> 00:06:42,360 including a $20 billion bridge loan taken out in March 20268 103 00:06:42,360 --> 00:06:46,000 weeks ago. So when investors hand over $75 104 00:06:46,000 --> 00:06:51,120 billion in this IPOA, meaningful portion of it may not be going 105 00:06:51,120 --> 00:06:54,040 to orbital data centers or a Mars colony. 106 00:06:54,280 --> 00:06:58,160 It goes straight back to the banks who lent SpaceX money 8 107 00:06:58,160 --> 00:07:01,720 weeks ago. Now, to be fair, the company is 108 00:07:01,720 --> 00:07:04,520 growing. Revenue is up somewhere between 109 00:07:04,520 --> 00:07:09,360 33 and 43% annually, which is genuinely impressive. 110 00:07:09,640 --> 00:07:13,520 But that growth is coming entirely from Starling, not the 111 00:07:13,760 --> 00:07:16,920 AI business. And even that growth comes with 112 00:07:16,920 --> 00:07:19,960 a caveat. Starling's average revenue per 113 00:07:19,960 --> 00:07:24,840 user has fallen by around 18% since 2023, three as the 114 00:07:24,840 --> 00:07:28,040 company's discounting aggressively to juice it's 115 00:07:28,040 --> 00:07:30,600 subscriber numbers before the IPO. 116 00:07:31,120 --> 00:07:35,960 Back in April 2018, Gwynne Shockwell, the president of 117 00:07:35,960 --> 00:07:39,480 SpaceX, stood on a Ted stage and made the promise. 118 00:07:39,600 --> 00:07:43,040 Within a decade, she told the audience, people would be taking 119 00:07:43,040 --> 00:07:46,600 30 minute rocket trips from New York to Shanghai. 120 00:07:46,920 --> 00:07:51,600 These rockets would compete with long haul airlines making up to 121 00:07:51,600 --> 00:07:56,360 10 flights per day. That decade expires in under 2 122 00:07:56,360 --> 00:08:00,320 years now, which is not much time to figure out how to safely 123 00:08:00,320 --> 00:08:04,320 replace the jumbo jet with a reusable ballistic missile that 124 00:08:04,320 --> 00:08:08,280 requires passengers to withstand 3GS at takeoff. 125 00:08:08,920 --> 00:08:12,320 Now if you were paying attention earlier, you may remember the 126 00:08:12,320 --> 00:08:15,720 phrase long haul point to point terrestrial travel. 127 00:08:15,720 --> 00:08:18,640 From the prospectus that is this product. 128 00:08:18,800 --> 00:08:22,280 It is still being pitched to investors as a future revenue 129 00:08:22,280 --> 00:08:24,840 stream. There's no sign of a commercial 130 00:08:24,840 --> 00:08:28,560 commuter rocket and there are under 2 years left on the clock. 131 00:08:28,880 --> 00:08:32,559 In the list of yet to be realized business lines, there's 132 00:08:32,559 --> 00:08:37,120 a section on payment processing. Twitter, Elon Musk's Everything 133 00:08:37,120 --> 00:08:40,760 app, has been talking about becoming a financial services 134 00:08:40,760 --> 00:08:44,240 platform for several years now. They've secured money 135 00:08:44,240 --> 00:08:48,040 transmitter licenses in a handful of U.S. states, but they 136 00:08:48,040 --> 00:08:51,000 don't have a functioning payments product yet. 137 00:08:51,240 --> 00:08:55,520 So SpaceX is now partially in the business of filing paperwork 138 00:08:55,520 --> 00:08:59,320 with the Nevada Department of Business and Industry in pursuit 139 00:08:59,480 --> 00:09:04,400 of the dream of processing an $8 peer-to-peer payment on a rocket 140 00:09:04,400 --> 00:09:09,440 company social media platform. SpaceX claims its total 141 00:09:09,440 --> 00:09:13,840 addressable market is $28.5 trillion. 142 00:09:14,120 --> 00:09:18,720 This is larger than the entire GDP of the United States and 143 00:09:18,720 --> 00:09:22,440 almost entirely built on businesses that do not yet 144 00:09:22,440 --> 00:09:25,320 exist. As the Financial Times Lex 145 00:09:25,320 --> 00:09:28,840 column noted, the market has never before had to price a 146 00:09:28,840 --> 00:09:33,280 stock this speculative at this size, which is probably worth 147 00:09:33,280 --> 00:09:36,360 keeping in mind. According to the prospectus, 148 00:09:36,640 --> 00:09:42,240 enterprise applications of AI make up to 80% of Spacex's total 149 00:09:42,240 --> 00:09:45,520 addressable market. This means they expect large 150 00:09:45,520 --> 00:09:50,280 businesses to pay for masks, AI tools to improve productivity. 151 00:09:50,680 --> 00:09:55,640 If you add in consumer AI, the total AI share of the claimed 152 00:09:55,640 --> 00:10:02,760 addressable market rises to 93%. So 93% of the company's value is 153 00:10:02,960 --> 00:10:06,120 grog. Now I want to be careful here 154 00:10:06,280 --> 00:10:12,120 because 80% of a $28.5 trillion market is an enormous number. 155 00:10:12,120 --> 00:10:15,800 So it's worth asking which enterprises are currently using 156 00:10:15,800 --> 00:10:18,080 Grok. And from what I can tell, the 157 00:10:18,080 --> 00:10:23,040 primary documented use cases for Grok are fact checking tweets 158 00:10:23,080 --> 00:10:27,320 and generating non consensual nude images that have resulted 159 00:10:27,320 --> 00:10:32,080 in regulatory investigations in Spain, France and the UK and at 160 00:10:32,080 --> 00:10:36,840 least one ongoing lawsuit, which is not exactly the enterprise 161 00:10:36,840 --> 00:10:39,240 productivity pitch that I was expecting. 162 00:10:39,600 --> 00:10:44,640 The filing effectively concedes that the enterprise AI market is 163 00:10:44,640 --> 00:10:48,160 dominated by Open AI, Anthropic and Google. 164 00:10:48,520 --> 00:10:52,880 Spacex's own AI product, Grok, holds a market share of roughly 165 00:10:52,880 --> 00:10:58,840 3.4%, according to Bloomberg. Spacex's own engineers have been 166 00:10:58,840 --> 00:11:03,240 slow to adopt Grok for technical work because it underperforms 167 00:11:03,240 --> 00:11:07,080 rival tools. Last month, Musk himself said 168 00:11:07,280 --> 00:11:10,800 that the AI code would need to be rebuilt from scratch. 169 00:11:11,000 --> 00:11:16,480 Shortly after, he sold XAI to SpaceX for $250 billion. 170 00:11:16,920 --> 00:11:20,600 His solution to Grok's weakness in coding was to attempt to 171 00:11:20,600 --> 00:11:25,040 acquire CURSOR, the tool his own engineers were actually using, 172 00:11:25,280 --> 00:11:29,360 for $60 billion with a $10 billion break fee. 173 00:11:29,400 --> 00:11:34,000 If the deal falls through, the company whose AI product 174 00:11:34,000 --> 00:11:38,280 represents 93% of it's claimed addressable market, has a 175 00:11:38,280 --> 00:11:42,240 flagship product it's own engineers won't use, and is 176 00:11:42,240 --> 00:11:46,720 spending $60 billion trying to buy a competitor so that they 177 00:11:46,720 --> 00:11:51,280 can get it to work. Spacex's near term AI revenue 178 00:11:51,280 --> 00:11:56,920 story rests substantially on a $15 billion a year deal to rent 179 00:11:56,920 --> 00:12:01,040 compute capacity to Anthropic, one of the very competitors 180 00:12:01,160 --> 00:12:05,560 whose existence proves the Grok is losing the AI race. 181 00:12:05,880 --> 00:12:09,440 So SpaceX is basically Space Core Weave. 182 00:12:09,840 --> 00:12:14,200 If SpaceX were genuinely going to be a dominant AI company, it 183 00:12:14,200 --> 00:12:18,240 would need that computing power that it owns for its own models. 184 00:12:18,440 --> 00:12:22,040 If Volkswagen were renting out all of its car factories, would 185 00:12:22,040 --> 00:12:25,760 anyone describe Volkswagen as still being in the car business? 186 00:12:26,520 --> 00:12:30,600 The more pressing problem is that this $15 billion a year 187 00:12:30,600 --> 00:12:34,760 contract, roughly 40% of projected near term revenues, 188 00:12:35,000 --> 00:12:38,880 can be cancelled by Entropic with 90 days notice. 189 00:12:39,120 --> 00:12:43,600 Spacex's entire AI business model at the moment is renting 190 00:12:43,600 --> 00:12:47,920 GP us to a competitor on terms that can vanish in a fiscal 191 00:12:47,920 --> 00:12:51,040 quarter. If Spacex's total addressable 192 00:12:51,040 --> 00:12:56,760 market is 28 1/2 trillion dollars and 93% of it is AI I, 193 00:12:57,000 --> 00:13:01,480 then by Spacex's own logic, the leading AI companies must be 194 00:13:01,480 --> 00:13:06,200 worth vastly more than SpaceX and Tropics quarterly revenue is 195 00:13:06,200 --> 00:13:11,000 already roughly double Spacex's entire AI segment revenue. 196 00:13:11,480 --> 00:13:16,080 Both on Tropic and Open AI are reportedly filing for IPOs at 197 00:13:16,200 --> 00:13:21,440 approximately $1 trillion valuations, and yet SpaceX wants 198 00:13:21,440 --> 00:13:27,480 $1.75 trillion on the basis of a market it's direct competitors 199 00:13:27,480 --> 00:13:29,920 are winning. Now, if you've been wondering 200 00:13:29,920 --> 00:13:33,840 exactly who is buying the Tesla Cybertruck, the SpaceX 201 00:13:33,840 --> 00:13:36,360 prospectus finally provides an answer. 202 00:13:36,560 --> 00:13:41,360 It turns out it was SpaceX. According to the filings, SpaceX 203 00:13:41,360 --> 00:13:45,840 purchased $650 million in goods from Tesla last year. 204 00:13:46,160 --> 00:13:51,440 This included 131 million spent on Cybertrucks, all at full 205 00:13:51,440 --> 00:13:54,920 retail price. No volume discount, which is 206 00:13:54,920 --> 00:13:58,200 notable given how hard Tesla has found them to sell. 207 00:13:58,520 --> 00:14:03,800 That works out at roughly 1500 vehicles at one point in the 208 00:14:03,800 --> 00:14:08,320 fourth quarter of 2025. SpaceX single handedly accounted 209 00:14:08,320 --> 00:14:12,360 for 18% of all Cybertruck registrations in the United 210 00:14:12,360 --> 00:14:16,200 States. Using shareholder money from one 211 00:14:16,200 --> 00:14:19,600 company that you control to support the sales figures of 212 00:14:19,600 --> 00:14:23,560 another company you control, right before both companies face 213 00:14:23,560 --> 00:14:28,000 major financial events is generally considered not ideal 214 00:14:28,000 --> 00:14:32,640 corporate governance. There is also a small accounting 215 00:14:32,640 --> 00:14:36,560 discrepancy worth mentioning. SpaceX reports buying roughly 216 00:14:36,560 --> 00:14:41,760 $650 million of goods from Tesla last year, and Tesla's own 217 00:14:41,760 --> 00:14:46,920 filings report selling $573 million of goods to SpaceX. 218 00:14:47,160 --> 00:14:52,480 So somewhere between Austin and Boca Chica, $77 million seems to 219 00:14:52,480 --> 00:14:56,560 have fallen off the back of a truck, presumably a cyber truck. 220 00:14:57,240 --> 00:15:01,200 But a few thousand overpriced pickup trucks are the least of 221 00:15:01,200 --> 00:15:04,200 the related party concerns in this document. 222 00:15:04,520 --> 00:15:09,840 Antonio Gracias is a long term Musk ally and sitting member of 223 00:15:09,840 --> 00:15:13,800 the SpaceX board of directors. He also runs a private 224 00:15:13,800 --> 00:15:17,000 investment firm called Valor Equity Partners. 225 00:15:17,280 --> 00:15:22,440 According to the IPO prospectus, SpaceX and XAI are carrying more 226 00:15:22,440 --> 00:15:26,760 than $20 billion in AI infrastructure lease obligations 227 00:15:27,320 --> 00:15:30,520 tied to agreements with entities connected to valor. 228 00:15:31,000 --> 00:15:34,240 Some of these deals are structured as what accountants 229 00:15:34,240 --> 00:15:39,400 call failed sale leasebacks. Normally, in a sale leaseback, a 230 00:15:39,400 --> 00:15:43,800 company sells an asset to an investor, gets cash upfront, and 231 00:15:43,800 --> 00:15:47,200 leases the asset backed, removing it from the balance 232 00:15:47,200 --> 00:15:49,560 sheet. When the transaction fails to 233 00:15:49,560 --> 00:15:53,560 qualify as a genuine sale under accounting rules, the asset 234 00:15:53,560 --> 00:15:57,160 stays on the books and the associated payment obligations 235 00:15:57,240 --> 00:15:59,440 get recorded as dead. Instead. 236 00:15:59,920 --> 00:16:04,000 The practical effect is that SpaceX is carrying billions in 237 00:16:04,000 --> 00:16:08,720 obligations to a firm run by one of its own board members on its 238 00:16:08,720 --> 00:16:13,000 balance sheet as dead. It's once again considered good 239 00:16:13,000 --> 00:16:17,400 practice to board members who don't have $20 billion worth of 240 00:16:17,400 --> 00:16:20,880 high interest financial loans with the company they're 241 00:16:20,880 --> 00:16:26,800 supposed to be overseeing. But anyhow, back in 2016, Elon 242 00:16:26,800 --> 00:16:31,240 Musk used Tesla shareholder money to acquire Solar City, a 243 00:16:31,240 --> 00:16:35,800 struggling solar company that he also chaired and partly owned. 244 00:16:36,080 --> 00:16:40,080 To help justify the deal, he staged a public demonstration of 245 00:16:40,080 --> 00:16:43,720 Solar City's Solar roof tiles during the acquisition 246 00:16:43,720 --> 00:16:47,040 negotiations. The tiles were later found by a 247 00:16:47,040 --> 00:16:51,200 Delaware court to have been actually inoperable and far from 248 00:16:51,200 --> 00:16:54,760 ready to go to market at the time of the demonstration. 249 00:16:54,960 --> 00:16:58,480 The court reviewed the entire transaction under it's strictest 250 00:16:58,480 --> 00:17:02,960 standard of scrutiny, describing the process as deeply flawed, 251 00:17:03,200 --> 00:17:06,560 noted Musk's apparent inability to acknowledge his clear 252 00:17:06,560 --> 00:17:11,119 conflict of interest, and then found the deal entirely fair 253 00:17:11,119 --> 00:17:15,920 anyway, largely because 85% of the shareholders had a approved 254 00:17:15,920 --> 00:17:19,440 it and Tesla stock subsequently went up. 255 00:17:19,760 --> 00:17:24,359 The judge declined to award Musk his legal costs, which is the 256 00:17:24,640 --> 00:17:28,000 Delaware Court of Chancery's version of a strongly worded 257 00:17:28,000 --> 00:17:31,240 letter. Musk later had a falling out 258 00:17:31,240 --> 00:17:35,320 with Delaware over his Tesla pay package and reincorporated in 259 00:17:35,320 --> 00:17:37,400 Texas. He has now completed a 260 00:17:37,400 --> 00:17:41,040 structurally similar acquisition, this time of XAI, 261 00:17:41,280 --> 00:17:44,440 in a state with far more LAX securities rules. 262 00:17:45,000 --> 00:17:48,960 In Texas, to bring a derivative lawsuit against SpaceX, a 263 00:17:48,960 --> 00:17:52,520 shareholder must own at least 3% of the company. 264 00:17:52,760 --> 00:17:57,200 At the current valuation, that's approximately $52 billion worth 265 00:17:57,200 --> 00:18:00,200 of stock. To win any lawsuit, they must 266 00:18:00,200 --> 00:18:04,480 prove not just a breach of fiduciary duty, but specifically 267 00:18:04,480 --> 00:18:09,600 fraud, intentional misconduct, or a knowing violation of law. 268 00:18:09,920 --> 00:18:13,760 A conflicted transaction, even one with no independent 269 00:18:13,760 --> 00:18:17,800 oversight whatsoever, does not meet that standard on its own. 270 00:18:18,120 --> 00:18:21,280 And if you wanted to go looking for the internal emails that 271 00:18:21,280 --> 00:18:24,400 might help you to prove fraud, as shareholders did so 272 00:18:24,400 --> 00:18:29,080 effectively in Delaware, Texas corporate law now explicitly 273 00:18:29,080 --> 00:18:32,880 excludes emails and text messages from shareholder 274 00:18:32,880 --> 00:18:37,320 document requests. When Tesla reincorporated in 275 00:18:37,320 --> 00:18:42,160 Texas, Governor Greg Abbott posted on Twitter to personally 276 00:18:42,160 --> 00:18:45,720 congratulate Musk and wish him well, which gives you a 277 00:18:45,720 --> 00:18:49,080 reasonable sense of the regulatory environment. 278 00:18:49,720 --> 00:18:53,440 The lesson that Musk appears to have drawn from Solar City is 279 00:18:53,440 --> 00:18:57,200 not don't do conflicted acquisitions, it's do them 280 00:18:57,200 --> 00:19:01,760 somewhere with friendlier corks. The clearest beneficiaries of 281 00:19:01,760 --> 00:19:05,320 this acquisition are not the SpaceX engineers trying to get 282 00:19:05,320 --> 00:19:07,760 to Mars. They're the venture capitalists 283 00:19:07,760 --> 00:19:11,360 and banks, including Morgan Stanley, who were at first stock 284 00:19:11,360 --> 00:19:15,440 holding billions in underwater debt from Musk chaotic buy out 285 00:19:15,440 --> 00:19:19,200 of Twitter. Thanks to the XAI SpaceX merger, 286 00:19:19,440 --> 00:19:22,920 those lenders have been made whole in SpaceX equity, 287 00:19:23,160 --> 00:19:27,280 conveniently timed just before the largest IPO in history. 288 00:19:27,920 --> 00:19:32,400 Now, historically, investment bankers won IPO mandates by 289 00:19:32,400 --> 00:19:36,600 demonstrating deep sector expertise, long term advisory 290 00:19:36,600 --> 00:19:40,520 relationships, and a willingness to work extremely long hours. 291 00:19:40,720 --> 00:19:43,520 It used to be quite a dignified business. 292 00:19:44,000 --> 00:19:47,160 The first sign of slipping standards in investment banking 293 00:19:47,400 --> 00:19:51,840 came in 2007 when the Goldman Sachs bankers who were pitching 294 00:19:51,840 --> 00:19:56,560 for the Lululemon IPO mandate turned up to the meeting led by 295 00:19:56,720 --> 00:20:01,560 David Solomon, also known as DJ Diesel, dressed head to toe in 296 00:20:01,560 --> 00:20:06,120 Lululemon at leisure wear. I want to be clear here, this 297 00:20:06,120 --> 00:20:09,800 slipping standards may well have been the event that triggered 298 00:20:09,800 --> 00:20:12,320 the global financial crisis that year. 299 00:20:12,760 --> 00:20:17,520 DJD Salter, bad boy of corporate finance this time around, 300 00:20:17,760 --> 00:20:21,840 boasted to the press that he secured the lead left position 301 00:20:21,840 --> 00:20:25,920 on the largest IPO in history by having his staffers help him 302 00:20:25,920 --> 00:20:29,160 send Elon Musk a direct message on Twitter. 303 00:20:29,600 --> 00:20:32,960 Morgan Stanley, which had cultivated A Musk relationship 304 00:20:32,960 --> 00:20:36,880 for years through their star tech banker Michael Grimes, is 305 00:20:36,880 --> 00:20:40,840 listed as Co lead. Their name appears second in 306 00:20:40,840 --> 00:20:44,400 alphabetical order, directly after Goldman Sachs. 307 00:20:44,680 --> 00:20:48,400 Morgan Stanley was apparently so furious about this that they 308 00:20:48,400 --> 00:20:52,600 forced SpaceX to explicitly clarify in the term sheet that 309 00:20:52,600 --> 00:20:57,360 the listing is purely alphabetical. 23 banks have 310 00:20:57,360 --> 00:21:00,320 cramped their names onto the cover of this prospectus. 311 00:21:00,560 --> 00:21:04,720 The fees constitute the most lucrative Wall Street IPO payday 312 00:21:04,720 --> 00:21:07,640 ever. The only evidence of enterprise 313 00:21:07,640 --> 00:21:12,800 usage of Musk's Grok AI may well be the millions of dollars of 314 00:21:12,800 --> 00:21:16,920 licenses that Musk is forcing on his lead IPO banks. 315 00:21:17,560 --> 00:21:20,640 All of this brings us to the corporate governance structure 316 00:21:20,640 --> 00:21:24,280 of SpaceX, which is perhaps the most heavily fortified 317 00:21:24,280 --> 00:21:27,920 dictatorship ever brought to US public markets. 318 00:21:28,200 --> 00:21:31,440 The company will have a dual class share structure. 319 00:21:31,720 --> 00:21:35,800 Regular investors will get Class A shares, which carry one vote, 320 00:21:36,000 --> 00:21:39,880 and Elon Musk gets Class B shares, which carry 10 votes 321 00:21:39,880 --> 00:21:43,280 each. While Musk only owns around 41% 322 00:21:43,280 --> 00:21:48,240 of SpaceX, he will control more than 85% of the votes, meaning 323 00:21:48,240 --> 00:21:51,120 that he can't ever be fired by shareholders. 324 00:21:51,680 --> 00:21:55,960 The company has also placed limits on how shareholders can 325 00:21:55,960 --> 00:22:00,040 file legal challenges, as SpaceX writes in the prospectus. 326 00:22:00,280 --> 00:22:04,040 This will limit or preclude your ability to influence corporate 327 00:22:04,040 --> 00:22:06,920 matters and the election of our directors. 328 00:22:07,440 --> 00:22:11,720 On top of that, because Elon must convince NASDAQ to change 329 00:22:11,720 --> 00:22:15,000 their index inclusion methodology in order to win this 330 00:22:15,000 --> 00:22:18,920 listing, SpaceX will be fast tracked for inclusion in the 331 00:22:18,920 --> 00:22:22,960 NASDAQ 100 index. This is a massive deal. 332 00:22:23,200 --> 00:22:27,680 The NASDAQ 100 is tracked by over $600 billion of passive 333 00:22:27,680 --> 00:22:31,400 investor money. By fast tracking SpaceX into the 334 00:22:31,400 --> 00:22:35,320 index, countless pension funds and passive investors will 335 00:22:35,320 --> 00:22:39,880 become forced buyers of the stock almost immediately, long 336 00:22:39,880 --> 00:22:42,640 before there's any serious price discovery. 337 00:22:43,280 --> 00:22:47,000 And they'll own shares in a heavily cash burning company 338 00:22:47,120 --> 00:22:51,160 where they've no voting power and no meaningful ability to sue 339 00:22:51,160 --> 00:22:54,840 management. Musk has ensured that he will 340 00:22:54,840 --> 00:22:57,560 never lose his control over SpaceX. 341 00:22:57,760 --> 00:23:02,320 The filing reveals the creation of a new Class C non voting 342 00:23:02,320 --> 00:23:05,240 stock. This exists so that SpaceX can 343 00:23:05,240 --> 00:23:09,880 acquire other companies in the future using that stock without 344 00:23:09,880 --> 00:23:12,080 ever diluting Musk's voting power. 345 00:23:12,320 --> 00:23:16,440 And if he ever decides to step down, he can place his super 346 00:23:16,440 --> 00:23:19,920 voting shares into trusts, passing those super voting 347 00:23:19,920 --> 00:23:24,560 shares onto his children, ensuring dynastic control. 348 00:23:25,240 --> 00:23:28,120 Finally, we get to his compensation package. 349 00:23:28,440 --> 00:23:32,520 You've most likely read about how Musk is getting 1 billion 350 00:23:32,520 --> 00:23:36,280 performance based restricted shares, and how these shares 351 00:23:36,360 --> 00:23:40,480 only vest when SpaceX establishes a permanent human 352 00:23:40,480 --> 00:23:46,160 colony on Mars and hits a 7 1/2 trillion dollar market cap. 353 00:23:46,160 --> 00:23:49,760 SpaceX explicitly describes the achievement of these milestones 354 00:23:49,760 --> 00:23:52,680 in the prospectus as being improbable. 355 00:23:53,240 --> 00:23:57,280 To be clear, this is not a sudden burst of candor from 356 00:23:57,280 --> 00:23:59,440 management. It's instead a bit of an 357 00:23:59,440 --> 00:24:02,840 accounting trick. Under accounting rules, if the 358 00:24:02,840 --> 00:24:06,160 company admitted that it was probable that Musk would get 359 00:24:06,160 --> 00:24:09,520 these shares, they would have to expense the full value of the 360 00:24:09,520 --> 00:24:13,360 award, potentially hundreds of billions of dollars, on their 361 00:24:13,360 --> 00:24:17,240 income statement today. By categorizing the share grant 362 00:24:17,240 --> 00:24:20,680 as improbable, it gets booked at 0 cost. 363 00:24:21,040 --> 00:24:24,880 But there's a trick. As law professor Anne Lipton 364 00:24:24,880 --> 00:24:28,880 points out, Musk doesn't actually need to achieve any of 365 00:24:28,880 --> 00:24:32,520 those hard to reach goals. He gets the utility of these 366 00:24:32,520 --> 00:24:35,600 shares immediately. He can vote them today. 367 00:24:35,720 --> 00:24:37,920 He can collect dividends on them today. 368 00:24:38,120 --> 00:24:42,240 And with the approval of a board that he entirely controls, he 369 00:24:42,240 --> 00:24:46,200 can pledge them as collateral for loans to get access to tax 370 00:24:46,200 --> 00:24:49,800 free cash. Today he gets all the benefits 371 00:24:49,800 --> 00:24:53,080 of owning the shares immediately, based on a Mars 372 00:24:53,080 --> 00:24:57,240 colony that the company admits he will probably never build. 373 00:24:57,800 --> 00:25:00,960 Now possibly the most extraordinary line in the entire 374 00:25:00,960 --> 00:25:05,000 prospectus is this. SpaceX renounce certain 375 00:25:05,000 --> 00:25:09,440 corporate opportunities that may be presented to Mr. Musk and 376 00:25:09,440 --> 00:25:13,360 certain of our directors and their respective affiliates, and 377 00:25:13,360 --> 00:25:17,600 such persons may have no duty to present such opportunities to 378 00:25:17,600 --> 00:25:20,480 us. What this means is that nothing 379 00:25:20,480 --> 00:25:24,240 stops Elon Musk from setting up a separate private company 380 00:25:24,240 --> 00:25:29,600 tomorrow to pursue whatever the next big tech trend is, rather 381 00:25:29,600 --> 00:25:33,000 than doing it inside SpaceX. Which, if you remember him 382 00:25:33,000 --> 00:25:37,600 claiming Tesla was an AI company and then immediately starting X 383 00:25:37,600 --> 00:25:41,440 AI as a separate competing private entity is a fairly 384 00:25:41,440 --> 00:25:44,880 relevant risk. As I'm Lipton summarized it, 385 00:25:45,120 --> 00:25:49,200 public shareholders in SpaceX will have no votes that matter, 386 00:25:49,360 --> 00:25:52,760 no ability to sell at a fair price discovered through normal 387 00:25:52,760 --> 00:25:56,720 market mechanisms, and no ability to sue the company for 388 00:25:56,720 --> 00:26:00,840 wrongdoing. No votes, no sales, no suits. 389 00:26:01,680 --> 00:26:05,840 There's one technology at SpaceX that everything else depends on. 390 00:26:06,040 --> 00:26:10,520 Not the AI, not the satellite Internet, not the Mars colony. 391 00:26:10,680 --> 00:26:14,400 One rocket. Starship, the largest rocket 392 00:26:14,400 --> 00:26:17,760 ever built, is supposed to be the key to everything. 393 00:26:18,080 --> 00:26:21,480 Cheaper launches, bigger Starlink satellites, orbital 394 00:26:21,480 --> 00:26:25,480 data centers, a city on Mars, the NASA contracts. 395 00:26:25,680 --> 00:26:31,000 Without Starship, the entire $1.75 trillion story falls apart 396 00:26:31,000 --> 00:26:34,560 entirely. Yesterday, Barron's wrote that 397 00:26:34,560 --> 00:26:38,720 the Starship launch was a bigger deal for SpaceX than the IPO 398 00:26:38,720 --> 00:26:41,880 filing itself. The latest version of the rocket 399 00:26:41,880 --> 00:26:46,200 was supposed to launch before the IPO filing, but that launch 400 00:26:46,200 --> 00:26:48,640 was postponed. It was then supposed to launch 401 00:26:48,640 --> 00:26:51,000 yesterday, the day after the filing. 402 00:26:51,280 --> 00:26:55,440 Nicki Minaj even turned up to watch, but there were technical 403 00:26:55,440 --> 00:26:57,680 difficulties, according to the news. 404 00:26:57,840 --> 00:27:01,000 They're going to try again today, so we'll have to see how 405 00:27:01,000 --> 00:27:04,560 that go. The problem so far is that 406 00:27:04,560 --> 00:27:08,160 Starship hasn't yet worked. It's still in the rapid 407 00:27:08,160 --> 00:27:11,320 unscheduled disassembly phase of its development. 408 00:27:11,520 --> 00:27:15,320 It is yet to carry any cargo to orbit other than the banana. 409 00:27:15,560 --> 00:27:19,520 So as Barons rightly states, this launch is really going to 410 00:27:19,520 --> 00:27:22,840 matter. The entire investment pieces is 411 00:27:22,840 --> 00:27:27,000 built around Starship achieving what SpaceX themselves call an 412 00:27:27,000 --> 00:27:30,640 insane flight rate. Hundreds of launches per year, 413 00:27:30,840 --> 00:27:33,880 each carrying 200 tons into orbit. 414 00:27:34,160 --> 00:27:38,200 Starlink version 3 satellites, which are supposed to provide 20 415 00:27:38,200 --> 00:27:41,720 times the capacity of current models, are too large for the 416 00:27:41,720 --> 00:27:44,680 existing rockets and require Starship. 417 00:27:44,920 --> 00:27:49,120 The Orbital Data Center concept requires Starship to launch 418 00:27:49,240 --> 00:27:52,360 1,000,000 satellites. The Mars colony requires 419 00:27:52,360 --> 00:27:56,600 Starship to work reliably enough to transport 1,000,000 people 420 00:27:56,600 --> 00:28:02,320 who I guess want to go there. If Starship fails to deliver, if 421 00:28:02,320 --> 00:28:06,200 the flight rate is half of what is promised, if the reliability 422 00:28:06,200 --> 00:28:10,320 is lower, and if the costs are higher, the entire valuation 423 00:28:10,320 --> 00:28:14,040 collapses. The $1.75 trillion number 424 00:28:14,200 --> 00:28:19,040 assumes Starship works exactly as advertised, on schedule, with 425 00:28:19,040 --> 00:28:23,640 no technical setbacks. History suggests that rockets 426 00:28:23,640 --> 00:28:27,800 rarely work exactly as advertised, so the company 427 00:28:27,800 --> 00:28:32,040 claiming to be worth $1.75 trillion is depending on a 428 00:28:32,040 --> 00:28:36,320 rocket that is yet to figure out how to land without exploding, 429 00:28:36,320 --> 00:28:41,080 which is a risk. Now, IPO prospectuses are 430 00:28:41,080 --> 00:28:44,520 required to tell investors specifically what the company 431 00:28:44,520 --> 00:28:47,640 will do with their money. This is called the use of 432 00:28:47,640 --> 00:28:50,880 proceeds. A piece by Cape Fear Advisors 433 00:28:51,080 --> 00:28:54,440 points out how hard you have to dig, including through the 434 00:28:54,440 --> 00:28:58,320 footnotes, to grasp the full picture of Spacex's use of 435 00:28:58,320 --> 00:29:01,520 proceeds. The IPO is expected to raise 436 00:29:01,520 --> 00:29:05,960 between 50 and $75 billion. When you add up all of the 437 00:29:05,960 --> 00:29:09,520 commitments buried in the document, the company needs 438 00:29:09,560 --> 00:29:14,960 approximately $235 billion over the next few years, between 439 00:29:14,960 --> 00:29:17,800 three and five times the IPO proceeds. 440 00:29:18,120 --> 00:29:22,520 What this means is that the IPO is not the end of fundraising, 441 00:29:22,720 --> 00:29:26,240 it's just the beginning. Matt Levine summed up how 442 00:29:26,240 --> 00:29:29,920 investors should think about this vibes based investment. 443 00:29:29,920 --> 00:29:33,960 In his Money Stuff newsletter, he writes if you're buying 444 00:29:33,960 --> 00:29:37,600 SpaceX stock to tell Elon Musk what to do, you should stop. 445 00:29:37,920 --> 00:29:41,360 If you're buying SpaceX stock because you want exposure to the 446 00:29:41,360 --> 00:29:45,080 space AI business but you have your doubts about current 447 00:29:45,080 --> 00:29:49,000 management, you should stop. If you're buying SpaceX stock 448 00:29:49,160 --> 00:29:52,680 because you like Elon Musk and want to go along for the ride 449 00:29:52,680 --> 00:29:57,000 with him, yes, that's correct. That's the investment thesis 450 00:29:57,000 --> 00:29:59,480 here. If later you're disillusioned 451 00:29:59,480 --> 00:30:02,600 with Elon Musk, if you feel that you were misled that he didn't 452 00:30:02,600 --> 00:30:06,200 do the stuff he promised to do, well, that's kind of his whole 453 00:30:06,200 --> 00:30:08,360 thing. Thanks for tuning into this 454 00:30:08,360 --> 00:30:10,920 week's podcast. If you found it interesting, I'd 455 00:30:10,920 --> 00:30:14,200 really appreciate a five star rating or a short review on 456 00:30:14,200 --> 00:30:18,080 whatever podcast app you use. It really helps the podcast to 457 00:30:18,080 --> 00:30:20,360 grow. Have a great week and talk to 458 00:30:20,360 --> 00:30:21,760 you again soon. Bye.