1 00:00:00,080 --> 00:00:03,680 SpaceX is currently preparing for what's expected to be the 2 00:00:03,680 --> 00:00:08,119 largest IPO in history, with a target valuation as high as 3 00:00:08,119 --> 00:00:11,960 $1.75 trillion. To put that in perspective, 4 00:00:12,120 --> 00:00:15,960 that's larger than Meta or Tesla, and it would comfortably 5 00:00:15,960 --> 00:00:19,520 unseat Saudi Aramco for the title of the biggest stock 6 00:00:19,520 --> 00:00:23,400 market debut of all time. The offering is expected to see 7 00:00:23,400 --> 00:00:27,760 massive demand from retail investors who have up until now 8 00:00:27,920 --> 00:00:31,320 been unable to invest as the company's private. 9 00:00:31,600 --> 00:00:34,800 Now, in the aerospace world, there's a very important 10 00:00:34,800 --> 00:00:37,800 distinction between astronomy and astrology. 11 00:00:38,040 --> 00:00:40,880 Astronomy is the branch of science that uses math to 12 00:00:40,880 --> 00:00:44,640 calculate orbital mechanics so your multibillion dollar rocket 13 00:00:44,640 --> 00:00:47,240 doesn't turn into a very expensive firework. 14 00:00:47,400 --> 00:00:51,080 Astrology, on the other hand, is the belief that you should stay 15 00:00:51,080 --> 00:00:55,240 home from work next week because Jupiter's feeling a bit spicy. 16 00:00:55,520 --> 00:00:58,760 Usually, the people building rockets and the people casting 17 00:00:58,760 --> 00:01:01,600 horoscopes don't have very much in common. 18 00:01:02,120 --> 00:01:06,160 Elon Musk obviously cares about astronomy, the science. 19 00:01:06,320 --> 00:01:10,120 He has, after all, spent decades talking about the importance of 20 00:01:10,120 --> 00:01:12,240 going to Mars and building rockets. 21 00:01:12,520 --> 00:01:15,840 We know that he also loves a bit of numerology, too. 22 00:01:16,040 --> 00:01:20,880 No one finds the numbers for 20 and 69 funnier than Elon Musk 23 00:01:20,880 --> 00:01:25,080 does with the IPO enhancement. We've learned that he loves a 24 00:01:25,080 --> 00:01:29,960 bit of astrology, too, as SpaceX has reportedly set its sights on 25 00:01:30,040 --> 00:01:34,760 IPO in this June because Jupiter, Venus, and Mercury will 26 00:01:34,760 --> 00:01:37,600 be forming a rare conjunction in the sky. 27 00:01:37,880 --> 00:01:42,200 And I suppose that's auspicious. On top of that, June also 28 00:01:42,200 --> 00:01:45,720 happens to be the month of Mosque's 55th birthday. 29 00:01:46,080 --> 00:01:49,760 And to top it all off, there are rumors that the date being 30 00:01:49,760 --> 00:01:54,880 targeted is June the 9th. Yep, that's right, June the 9th 31 00:01:54,880 --> 00:01:58,360 or 69. If Musk was a bit younger it 32 00:01:58,360 --> 00:02:04,600 would be on 6-7, but he's not. He's a 55 year old man, so 69 it 33 00:02:04,600 --> 00:02:07,720 is. Well, the planetary alignment 34 00:02:07,840 --> 00:02:12,120 makes for a great headline. The timing is possibly driven by 35 00:02:12,120 --> 00:02:16,360 a much more terrestrial concern, the race for limited capital. 36 00:02:16,680 --> 00:02:20,920 As of early 2026, the private markets are starting to look 37 00:02:20,920 --> 00:02:23,920 tapped out. Venture capital funds that have 38 00:02:23,920 --> 00:02:27,400 spent the last few years aggressively funding the AI boom 39 00:02:27,600 --> 00:02:31,520 are now facing a liquidity plateau and are under pressure 40 00:02:31,520 --> 00:02:34,920 to finally return some cash to their own investors. 41 00:02:35,280 --> 00:02:39,160 At the same time, Musk is in a high stakes race to reach the 42 00:02:39,160 --> 00:02:44,400 public markets before his rivals Open AI and Anthropic soak up 43 00:02:44,440 --> 00:02:48,680 all of the available demand. Bankers are quietly warning that 44 00:02:48,680 --> 00:02:52,160 there might not be enough spare change in the public markets to 45 00:02:52,160 --> 00:02:56,440 support three massive capital intensive AI listings at the 46 00:02:56,440 --> 00:02:59,160 same time. And the first mover might get 47 00:02:59,160 --> 00:03:02,960 the biggest slice of the pie, with XAI burning through a 48 00:03:02,960 --> 00:03:06,400 billion dollars a month and massive data center projects 49 00:03:06,400 --> 00:03:08,680 requiring billions more in infrastructure. 50 00:03:08,840 --> 00:03:12,920 Musk may not be just watching the stars, he might be watching 51 00:03:12,920 --> 00:03:16,360 the exits too. The planetary alignment might be 52 00:03:16,520 --> 00:03:19,920 a nice bit of branding, but it's possibly the fear of being 53 00:03:19,920 --> 00:03:24,120 second in line at the ATM that's really fueling the countdown. 54 00:03:24,560 --> 00:03:28,800 This rush to the public markets represents a significant change 55 00:03:28,800 --> 00:03:32,240 of heart for Musk. For over a decade, he's insisted 56 00:03:32,240 --> 00:03:36,800 that SpaceX would not go public until his Mars transport system 57 00:03:37,040 --> 00:03:40,120 was fully in place. He argued that the short term 58 00:03:40,120 --> 00:03:43,800 pressures of being a public company where investors obsess 59 00:03:43,840 --> 00:03:47,600 over quarterly performance and short sellers hit you with large 60 00:03:47,600 --> 00:03:51,480 stakes would be a distraction from the long term goal of 61 00:03:51,480 --> 00:03:55,880 making life multiplanetary. But as we head towards the June 62 00:03:55,880 --> 00:04:01,400 2026 listing, the Mars transport system is still very much a work 63 00:04:01,400 --> 00:04:05,080 in progress, and Musk has recently started sweeping those 64 00:04:05,080 --> 00:04:09,640 Red Planet ambitions under the rug like an embarrassing teenage 65 00:04:09,640 --> 00:04:12,480 phase. The new mission and the core of 66 00:04:12,480 --> 00:04:17,120 the IPO pitch is focused on the Moon, a destination that Musk 67 00:04:17,120 --> 00:04:19,760 once dismissed as a waste of time. 68 00:04:19,959 --> 00:04:23,160 He now describes the Moon as a stepping stone and the 69 00:04:23,160 --> 00:04:27,160 overriding priority for securing the future of civilization. 70 00:04:27,560 --> 00:04:31,080 It turns out that the shift of focus is a lot easier when 71 00:04:31,080 --> 00:04:34,840 you've been foreshadowing it for years, even if accidentally. 72 00:04:35,120 --> 00:04:38,880 Elon Musk has been wearing his famous Occupy Mars T-shirts 73 00:04:38,960 --> 00:04:43,560 since 2019, which a number of people have pointed out actually 74 00:04:43,560 --> 00:04:47,400 features a red tinted picture of the Moon during the lunar 75 00:04:47,400 --> 00:04:49,560 eclipse. It's possible that he hasn't 76 00:04:49,560 --> 00:04:53,320 abandoned his Mars mission. Maybe he just took a close look 77 00:04:53,320 --> 00:04:56,840 at his own chest and realize that he's been accidentally 78 00:04:56,840 --> 00:04:59,080 advertising for the moon all along. 79 00:04:59,720 --> 00:05:03,160 The transition from a high flying rocketry and satellite 80 00:05:03,160 --> 00:05:07,800 firm into what we might now call the Triple X conglomerate, 81 00:05:07,800 --> 00:05:13,120 SpaceX XAI&X has been described by supporters as an act of 82 00:05:13,120 --> 00:05:16,240 genius and by critics as a massive bailout. 83 00:05:16,640 --> 00:05:20,760 This is because just as the SpaceX IPO was being finalized, 84 00:05:20,960 --> 00:05:27,080 Musk decided to fold XAI into SpaceX at a $250 billion price 85 00:05:27,080 --> 00:05:30,120 tag. Merging a rocket company with an 86 00:05:30,200 --> 00:05:34,480 AI start up is strange enough, but XAI comes with some 87 00:05:34,480 --> 00:05:37,360 particularly heavy social media baggage. 88 00:05:37,680 --> 00:05:42,440 Last year, XAI, the AI company, used it's shares to buy Elon 89 00:05:42,440 --> 00:05:46,720 Musk's Everything Up X, the platform formerly and in truth 90 00:05:46,720 --> 00:05:51,040 still known as Twitter. This means SpaceX investors are 91 00:05:51,040 --> 00:05:55,400 now, by extension, the proud owners of the number 17 ranked 92 00:05:55,400 --> 00:05:59,280 social media website, a business that's been struggling with a 93 00:05:59,280 --> 00:06:03,680 persistent exodus of both users and advertisers since Musk 94 00:06:03,680 --> 00:06:07,240 bought it for $44 billion in 2022. 95 00:06:07,680 --> 00:06:12,200 To justify the $250 billion price tag, you have to believe 96 00:06:12,200 --> 00:06:15,600 that XAI is a top tier player in AI. 97 00:06:15,960 --> 00:06:20,240 Currently, ChatGPT and Google's Gemini dominate the field, 98 00:06:20,400 --> 00:06:23,840 holding over 85% of the market between them. 99 00:06:24,080 --> 00:06:29,120 Grok, by comparison, sits at a distant 3.4% market share. 100 00:06:29,480 --> 00:06:33,360 One problem is that Grok is effectively an island. 101 00:06:33,560 --> 00:06:38,000 While xai recently launched a stand alone Grok app, the vast 102 00:06:38,000 --> 00:06:41,560 majority of its usage remains trapped inside Twitter. 103 00:06:41,800 --> 00:06:46,080 Most of Grok's use is just fact checking tweets or generating 104 00:06:46,080 --> 00:06:49,600 the kind of unfiltered memes that have landed the company in 105 00:06:49,600 --> 00:06:53,680 regulatory hot water in Spain, France and the UK. 106 00:06:54,240 --> 00:06:58,160 Well, SpaceX is said to be profitable, generating about $8 107 00:06:58,160 --> 00:07:03,880 billion in Ebeta last year. XAI is a money furnace, burning 108 00:07:03,880 --> 00:07:07,040 through a billion dollars in cash every single month. 109 00:07:07,320 --> 00:07:10,920 In a world where AI could be a winner take all market, much 110 00:07:10,920 --> 00:07:15,200 like Internet search was, being a lower ranked player like XAI 111 00:07:15,400 --> 00:07:17,880 is possibly a dangerous place to be. 112 00:07:18,280 --> 00:07:22,560 That danger is exacerbated by a growing talent drain. 113 00:07:22,760 --> 00:07:26,320 Since the start of the year, a significant number of key staff 114 00:07:26,320 --> 00:07:30,760 have left XAI, including six of the original founders who were 115 00:07:30,760 --> 00:07:32,840 poached from Google and DeepMind. 116 00:07:33,040 --> 00:07:37,400 It's hard to justify a frontier AI premium when the very people 117 00:07:37,400 --> 00:07:39,600 who have built the model have left the building. 118 00:07:40,000 --> 00:07:43,600 Musk's explanation for the merger, delivered in a press 119 00:07:43,600 --> 00:07:46,720 release that reads like the script of a low budget sci-fi 120 00:07:46,720 --> 00:07:51,280 ethic, is that we need to build orbital data centers to extend 121 00:07:51,280 --> 00:07:53,560 the light of consciousness to the stars. 122 00:07:53,800 --> 00:07:57,560 He argued argues that space based AI is the only way to 123 00:07:57,560 --> 00:08:01,000 scale because space is called space for a reason. 124 00:08:01,680 --> 00:08:05,800 It's a move that mirrors the Solar City merger from 2016, 125 00:08:05,960 --> 00:08:09,440 where Musk used Tesla's balance sheet to bail out a struggling 126 00:08:09,440 --> 00:08:11,960 solar company he also happened to own. 127 00:08:12,240 --> 00:08:15,440 As Bloomberg's Matt Levine pointed out, the innovation 128 00:08:15,440 --> 00:08:18,760 engine here isn't just about rockets, it's about a 129 00:08:18,760 --> 00:08:23,240 fundraising ability that allows Musk to negotiate with himself 130 00:08:23,240 --> 00:08:26,440 and set his own value valuations without the nuisance of 131 00:08:26,440 --> 00:08:31,120 independent board oversight. Amusingly, right before I filmed 132 00:08:31,120 --> 00:08:35,919 this video, Elon Musk tweeted that XAI was not built right the 133 00:08:35,919 --> 00:08:39,799 first time around, so it's being rebuilt from the foundations up. 134 00:08:40,000 --> 00:08:43,400 Which might lead SpaceX investors to ask, if this was 135 00:08:43,400 --> 00:08:47,480 the case, why did they just pay $250 billion for it? 136 00:08:47,960 --> 00:08:50,920 For the banks and venture capital firms who were stuck in 137 00:08:50,920 --> 00:08:54,640 the underwater Twitter deal, the merger is fantastic news. 138 00:08:54,800 --> 00:08:58,360 It effectively swaps their struggling social media equity 139 00:08:58,520 --> 00:09:02,840 for shiny new SpaceX stock right before a massive IPO. 140 00:09:03,120 --> 00:09:06,080 And while SpaceX investors might be wondering why they're 141 00:09:06,080 --> 00:09:10,360 suddenly subsidizing an anime girlfriend chatbot, most are 142 00:09:10,360 --> 00:09:13,400 unlikely to complain. In the world of investing 143 00:09:13,400 --> 00:09:17,720 alongside Elon Musk, loyalty is the primary currency. 144 00:09:17,960 --> 00:09:21,240 If you complain too loudly about the corporate governance, you 145 00:09:21,240 --> 00:09:25,440 risk being cut out of the next auspicious planetary alignment 146 00:09:25,800 --> 00:09:29,640 to justify merging a rocket company with a money furnace 147 00:09:29,640 --> 00:09:32,640 chatbot. The IPO narrative relies on a 148 00:09:32,640 --> 00:09:35,920 very specific vision of the future based around the 149 00:09:35,920 --> 00:09:40,520 importance of moving the world's AI computing power off planet. 150 00:09:40,840 --> 00:09:44,560 Musk's argument is the terrestrial data centers are 151 00:09:44,560 --> 00:09:47,880 hitting a wall because of land power and cooling constraint. 152 00:09:48,560 --> 00:09:52,120 His solution is to launch a constellation of a million 153 00:09:52,120 --> 00:09:56,400 satellites to act as orbital data centers, directly 154 00:09:56,400 --> 00:10:00,400 harnessing near constant solar power in the vacuum of space. 155 00:10:01,000 --> 00:10:04,440 At first glance, the pitch might seem to have a certain logic. 156 00:10:04,640 --> 00:10:07,480 It's always sunny in space, and there are no neighbors to 157 00:10:07,480 --> 00:10:11,160 complain about the noise. But in practice, the engineering 158 00:10:11,160 --> 00:10:14,280 hurdles are huge. Take cooling, for example. 159 00:10:14,480 --> 00:10:18,160 On Earth, we keep surface from overheating by blowing cold air 160 00:10:18,160 --> 00:10:21,760 or water over them, a process called convection. 161 00:10:22,000 --> 00:10:25,720 In the vacuum of space, there's no air to move heat away. 162 00:10:25,920 --> 00:10:29,440 The only option is to use massive radiators that eject 163 00:10:29,440 --> 00:10:33,960 heat as infrared radiation. We've already seen a test of 164 00:10:33,960 --> 00:10:36,680 this concept. A start up launched an AI 165 00:10:36,680 --> 00:10:40,160 satellite a few months ago. It was about the size of a small 166 00:10:40,160 --> 00:10:43,960 refrigerator carrying an NVIDIA H-100 chip. 167 00:10:44,200 --> 00:10:47,840 While the chip was able to successfully train a tiny AI 168 00:10:47,840 --> 00:10:50,880 model on the works of Shakespeare, it couldn't run 169 00:10:50,880 --> 00:10:54,440 around the clock because it had to shut down frequently when it 170 00:10:54,440 --> 00:10:58,080 got too hot to function. The company's next version is 171 00:10:58,080 --> 00:11:00,080 scheduled to launch later this year. 172 00:11:00,280 --> 00:11:03,400 And they say that it will carry the second largest cooling 173 00:11:03,400 --> 00:11:07,280 radiator array in space, the only larger one being the 174 00:11:07,280 --> 00:11:10,280 cooling array on the International Space Station. 175 00:11:10,640 --> 00:11:14,640 So to run just a handful of AI chips in orbit, you need 176 00:11:14,640 --> 00:11:18,600 hardware that rivals the most expensive structure ever built 177 00:11:18,600 --> 00:11:21,360 by humanity, the International Space Station. 178 00:11:21,600 --> 00:11:25,280 And then you have to scale that up to the 1,000,000 satellites 179 00:11:25,280 --> 00:11:27,840 mentioned in the SpaceX press release. 180 00:11:28,320 --> 00:11:32,600 To reach the GW scale capacity Musk is talking about, you would 181 00:11:32,600 --> 00:11:36,880 need solar arrays and radiators measuring roughly 4 kilometers 182 00:11:36,880 --> 00:11:39,440 in length. Getting something that size 183 00:11:39,440 --> 00:11:43,000 roughly the size of 40 football fields into orbit and 184 00:11:43,000 --> 00:11:46,160 maintaining it would be a logistical nightmare. 185 00:11:46,440 --> 00:11:50,720 It would be a giant, fragile sail vulnerable to every piece 186 00:11:50,720 --> 00:11:53,040 of space debris in low Earth orbit. 187 00:11:53,800 --> 00:11:55,520 Then there's the issue of Starship. 188 00:11:55,720 --> 00:12:01,320 The entire sentient Sun business model relies on SpaceX reaching 189 00:12:01,320 --> 00:12:04,640 an insane flight rate. We're talking about launches 190 00:12:04,680 --> 00:12:08,400 every hour, carrying 200 tons per flight, according to 191 00:12:08,400 --> 00:12:13,160 Spacex's own press release. Currently, Starship is still in 192 00:12:13,160 --> 00:12:16,000 the frequent explosion phase of development. 193 00:12:16,280 --> 00:12:19,560 While the engineering team has achieved the incredible feat of 194 00:12:19,560 --> 00:12:24,360 catching a 70m booster with giant robot arms, the rocket is 195 00:12:24,360 --> 00:12:28,200 yet to perform a full orbital mission or prove that it can 196 00:12:28,200 --> 00:12:32,960 reliably deploy payloads without what SpaceX calls a Rapid 197 00:12:32,960 --> 00:12:37,760 Unscheduled Disassembly. To round out the sci-fi sales 198 00:12:37,760 --> 00:12:41,920 pitch, Musk has even proposed building factories on the moon. 199 00:12:42,160 --> 00:12:45,200 I'm not making this up. According to the press release, 200 00:12:45,680 --> 00:12:49,560 the plan is to manufacture satellites taking advantage of 201 00:12:49,560 --> 00:12:54,360 lunar resources and then fire them into deep space using an 202 00:12:54,360 --> 00:12:59,480 electromagnetic mass driver. Basically a giant space railgun, 203 00:12:59,720 --> 00:13:04,960 a technology that the US Navy spent over $500 million on over 204 00:13:04,960 --> 00:13:07,880 a 16 year period with no success. 205 00:13:08,040 --> 00:13:11,560 Apparently the railgun prototypes suffered from extreme 206 00:13:11,560 --> 00:13:15,480 heat and physical stresses, the types of problems which would be 207 00:13:15,680 --> 00:13:20,120 only more of an issue in space. It's not clear to me why a 208 00:13:20,120 --> 00:13:23,400 factory on the moon would be more efficient than the factory 209 00:13:23,400 --> 00:13:27,680 in even the most remote places on Earth, or what moon resources 210 00:13:27,680 --> 00:13:30,200 Musk is planning on taking advantage of. 211 00:13:30,360 --> 00:13:33,800 I would imagine some sort of supply chain still needs to be 212 00:13:33,800 --> 00:13:37,600 developed there. It is a bold vision and it 213 00:13:37,600 --> 00:13:41,560 represents a massive U-turn for a man who spent decades calling 214 00:13:41,560 --> 00:13:44,960 lunar missions a distraction from going to Mars. 215 00:13:45,160 --> 00:13:49,240 If you remember, in 2017, he claimed that he would be sending 216 00:13:49,240 --> 00:13:53,520 at least two cargo missions to Mars in 2022 to find water 217 00:13:53,520 --> 00:13:56,840 sources and establish power, mining and life support 218 00:13:56,840 --> 00:14:00,840 infrastructure so that the first crude missions could follow in 219 00:14:00,840 --> 00:14:04,160 2024. He later reaffirmed this target 220 00:14:04,160 --> 00:14:09,280 in 2018 and and again in 2020 before the timeline was delayed. 221 00:14:09,760 --> 00:14:13,640 Ultimately, building a data center in a desert or a 222 00:14:13,640 --> 00:14:17,320 specialized warehouse on Earth will always be cheaper and 223 00:14:17,320 --> 00:14:20,880 easier than trying to build them in space or manufacture and 224 00:14:20,880 --> 00:14:24,760 launch them from the moon. Andrew Mcalep, an engineer who 225 00:14:24,760 --> 00:14:28,600 works for a space startup, has built a web-based calculator 226 00:14:28,600 --> 00:14:31,840 that allows the cost of an orbital data center to be 227 00:14:31,840 --> 00:14:35,440 compared with the cost of an equivalent terrestrial 1. 228 00:14:35,720 --> 00:14:40,120 It estimates that building A1 GW data center and running it for 229 00:14:40,120 --> 00:14:45,640 five years on Earth costs around $16 billion, excluding chips, 230 00:14:45,840 --> 00:14:50,320 and an orbital equivalent would cost over $50 billion, some more 231 00:14:50,320 --> 00:14:54,520 than three times as much. The thing is that building a few 232 00:14:54,520 --> 00:14:58,960 data centers in Texas won't justify a $1.75 trillion 233 00:14:58,960 --> 00:15:02,040 valuation. For that, you need a sentient 234 00:15:02,040 --> 00:15:05,920 sun and a moon catapult. Now, this is possibly a 235 00:15:05,920 --> 00:15:10,200 recurring theme for Elon Musk, the idea that the hardest way of 236 00:15:10,200 --> 00:15:13,200 doing something is actually the most efficient. 237 00:15:13,440 --> 00:15:16,800 This is, after all, the same man who once claimed that digging 238 00:15:16,800 --> 00:15:20,480 miles of tunnels underground would somehow be orders of 239 00:15:20,480 --> 00:15:24,520 magnitude cheaper and faster than just building a road on the 240 00:15:24,520 --> 00:15:27,360 surface. Apparently, if you can convince 241 00:15:27,360 --> 00:15:30,680 investors that the laws of physics and basic accounting are 242 00:15:30,680 --> 00:15:34,680 just optional suggestions, you can put almost any price tag 243 00:15:34,760 --> 00:15:38,160 like on the results. If you can ignore the sentient 244 00:15:38,160 --> 00:15:42,000 Suns and the lunar catapults for a moment, you're still left with 245 00:15:42,000 --> 00:15:46,680 the daunting task of justifying a $1.75 trillion sticker price 246 00:15:46,680 --> 00:15:49,720 for SpaceX. To be clear, that doesn't mean 247 00:15:49,720 --> 00:15:51,680 that Elon Musk won't get that price. 248 00:15:51,840 --> 00:15:55,400 It just means there will be extremely difficult to explain 249 00:15:56,200 --> 00:16:00,480 based on what's being reported. This isn't just a regular IPO, 250 00:16:00,680 --> 00:16:04,120 it's a carefully managed liquidity event for a company 251 00:16:04,120 --> 00:16:07,560 that simply outgrown its capacity to source new funding 252 00:16:07,560 --> 00:16:11,600 in the private markets. For more than 20 years, SpaceX 253 00:16:11,600 --> 00:16:15,880 has existed in a private market bubble where valuations are set 254 00:16:16,040 --> 00:16:19,920 by small, isolated capital raises rather than the cold 255 00:16:19,920 --> 00:16:24,200 reality of daily public trading. But as former head of equity 256 00:16:24,200 --> 00:16:27,400 capital markets from Bank of America America Craig Cobbin 257 00:16:27,400 --> 00:16:31,040 points out in the Financial Times, the laws of financial 258 00:16:31,040 --> 00:16:34,840 gravity are a lot harder to bend than space-time. 259 00:16:35,280 --> 00:16:39,360 The primary metric for any high growth tech company is the 260 00:16:39,560 --> 00:16:44,200 enterprise value to sales ratio. It's particularly favored for 261 00:16:44,200 --> 00:16:48,440 rapidly growing firms that are either not yet profitable or 262 00:16:48,440 --> 00:16:52,240 have inconsistent earnings. This is because sales figures 263 00:16:52,400 --> 00:16:55,920 are often more stable and less susceptible to accounting 264 00:16:56,200 --> 00:16:59,320 portions. The net income, or EBITDA, the 265 00:16:59,320 --> 00:17:03,240 ratio tells you how much are you paying for every dollar of 266 00:17:03,240 --> 00:17:07,319 revenue the company brings. In Facebook, Cobbin points out 267 00:17:07,400 --> 00:17:11,680 when public at 11 times forward sales, this was considered 268 00:17:11,680 --> 00:17:15,560 extremely expensive at the time and it pushed the boundaries of 269 00:17:15,560 --> 00:17:18,240 what public markets were willing to tolerate. 270 00:17:18,599 --> 00:17:22,839 To add more context, most high growth tech companies back then 271 00:17:22,960 --> 00:17:25,400 were trading at much lower multiples. 272 00:17:25,640 --> 00:17:29,160 Google, for instance, was trading around 5 to 6 times 273 00:17:29,160 --> 00:17:32,480 sales. SpaceX is looking to list at 274 00:17:32,480 --> 00:17:36,480 roughly 94 times it's 2025 revenue. 275 00:17:36,800 --> 00:17:40,800 Even if we use the most optimistic projections for 2026, 276 00:17:41,040 --> 00:17:44,520 the ratio still sits at over 60 times sales. 277 00:17:44,760 --> 00:17:47,560 For context, that's significantly higher than 278 00:17:47,560 --> 00:17:51,280 Palantir, which currently has the highest multiple in the S&P 279 00:17:51,280 --> 00:17:54,800 500. Not everyone thinks that that 280 00:17:54,800 --> 00:17:58,000 price is crazy, however. The analysts at Pitch Book 281 00:17:58,200 --> 00:18:02,720 recently made the case that a $1.75 trillion valuation is 282 00:18:02,720 --> 00:18:07,600 expensive but justifiable if you assume that Starlink will reach 283 00:18:07,600 --> 00:18:13,160 1.2 billion users by 2040. The deep flaw in that math is a 284 00:18:13,160 --> 00:18:16,160 concept called the Total Addressable Market. 285 00:18:16,480 --> 00:18:20,240 Based on Hans Rosling's research in the book Factfulness, there 286 00:18:20,240 --> 00:18:24,160 are only about 1 billion people on the planet who earn more than 287 00:18:24,160 --> 00:18:27,880 $32 a day. Most of those people live in the 288 00:18:27,880 --> 00:18:31,520 United States and Western Europe, and most people in those 289 00:18:31,520 --> 00:18:35,680 parts of the world already have access to wired broadband, which 290 00:18:35,680 --> 00:18:38,440 is both faster and cheaper than Starling. 291 00:18:38,840 --> 00:18:43,360 A total addressable market of 1.2 billion users being able to 292 00:18:43,360 --> 00:18:48,400 spend over $100 a month solely on Internet connectivity is next 293 00:18:48,400 --> 00:18:52,000 to impossible. Starlink is a brilliant tool for 294 00:18:52,000 --> 00:18:56,280 people in remote areas, gold miners in Sudan or the military 295 00:18:56,280 --> 00:18:59,720 in Ukraine. But it's I'll suited for cities 296 00:18:59,800 --> 00:19:03,240 where the dish needs a clearview of the sky and because 297 00:19:03,240 --> 00:19:07,360 satellites suffer from network congestion when too many people 298 00:19:07,360 --> 00:19:09,680 connect to the same one all at once. 299 00:19:10,000 --> 00:19:13,880 Realistically, there are only about 32,000,000 households in 300 00:19:13,880 --> 00:19:17,600 the world who can afford the service and don't already have 301 00:19:17,600 --> 00:19:21,920 wired Internet access. Most of that group already have 302 00:19:21,920 --> 00:19:25,600 access to the Internet using mobile phone networks, and less 303 00:19:25,600 --> 00:19:29,040 than half of them might feel that cellular speed and 304 00:19:29,040 --> 00:19:32,440 connectivity is not good enough and are willing to pay up for 305 00:19:32,440 --> 00:19:36,320 Starlink access. Expecting Starling to jump from 306 00:19:36,320 --> 00:19:40,440 its current 10 million subscribers to 1.2 billion is 307 00:19:40,440 --> 00:19:43,920 like expecting everyone on the planet to start drinking Fiji 308 00:19:43,920 --> 00:19:47,560 bottled water even when they already have access to clean 309 00:19:47,560 --> 00:19:50,760 drinking water. We're already seeing the limits 310 00:19:50,760 --> 00:19:53,160 of this growth in the financial data. 311 00:19:53,480 --> 00:19:58,000 Elizabeth Lopato at The Verge recently pointed out that a 2024 312 00:19:58,000 --> 00:20:01,800 Morgan Stanley report projected that Starlink revenue would 313 00:20:01,800 --> 00:20:06,360 reach $19 billion the next year on the strength of 6,000,000 314 00:20:06,360 --> 00:20:10,560 subscribers. SpaceX then managed to reach 9.2 315 00:20:10,560 --> 00:20:14,240 million subscribers that year, more than 50% better than 316 00:20:14,240 --> 00:20:18,880 expected, but its revenue was only $16 billion. 317 00:20:19,200 --> 00:20:23,240 This implies that as Starlink grows it's customer base, it's 318 00:20:23,240 --> 00:20:26,840 margins are falling. Let's talk about the company's 319 00:20:26,840 --> 00:20:31,880 questionable profitability next. SpaceX recently touted $8 320 00:20:31,880 --> 00:20:35,920 billion in EBITDA, but as a number of analysts have noted, 321 00:20:35,960 --> 00:20:39,600 EBITDA, earnings before interest, taxes, depreciation 322 00:20:39,600 --> 00:20:43,400 and amortization is a very generous lens through which to 323 00:20:43,400 --> 00:20:47,640 view a company like SpaceX, which has very high research and 324 00:20:47,640 --> 00:20:51,160 development costs and very high depreciation. 325 00:20:51,480 --> 00:20:55,560 EBITDA is often used to help investors compare very similar 326 00:20:55,560 --> 00:20:58,320 businesses that have made different investment in 327 00:20:58,320 --> 00:21:02,320 accounting decisions over time, but companies that essentially 328 00:21:02,320 --> 00:21:05,080 run the same business. And you're just stripping out 329 00:21:05,080 --> 00:21:08,360 leverage choices, the jurisdictions they're taxed in 330 00:21:08,560 --> 00:21:11,640 and their accounting choices over the timelines they've 331 00:21:11,640 --> 00:21:13,760 chosen to spread out their costs. 332 00:21:14,240 --> 00:21:17,680 This can make sense. It's a great way to make a clean 333 00:21:17,680 --> 00:21:21,000 comparison of the cash flow generation power and business 334 00:21:21,000 --> 00:21:25,160 operations of two very similar widget manufacturing companies. 335 00:21:25,520 --> 00:21:29,760 In the case of high CapEx businesses with enormous cash 336 00:21:29,760 --> 00:21:34,000 burn on early stage research and development operations, EBITDA 337 00:21:34,000 --> 00:21:38,240 becomes very questionable. SpaceX satellites have a useful 338 00:21:38,240 --> 00:21:41,120 life of around 5 years. This means that when you 339 00:21:41,120 --> 00:21:44,560 capitalize the build cost of these satellites and send them 340 00:21:44,560 --> 00:21:48,320 up to space as investments on your balance sheet initially, 341 00:21:48,480 --> 00:21:52,520 you later incurred depreciation expenses where you spread out 342 00:21:52,520 --> 00:21:55,560 the cost of those satellite bills over their five year 343 00:21:55,560 --> 00:22:00,080 useful lifespans. Excluding significant expenses 344 00:22:00,080 --> 00:22:03,640 like this that are core to the business's ongoing operations 345 00:22:03,640 --> 00:22:08,000 and capacity to generate future cash flow is particularly 346 00:22:08,000 --> 00:22:10,320 distorted for a business like SpaceX. 347 00:22:10,560 --> 00:22:13,920 They simply couldn't continue to run their proposed operations 348 00:22:14,080 --> 00:22:17,000 without continuing to replace these satellites. 349 00:22:17,200 --> 00:22:20,360 Excluding the cost of the satellites and the cost of 350 00:22:20,360 --> 00:22:24,280 getting them into low Earth orbit to claim profitability is 351 00:22:24,320 --> 00:22:28,760 somewhat ridiculous if you then account for the billions in cash 352 00:22:28,760 --> 00:22:31,680 used to build Starship. In terms of research and 353 00:22:31,680 --> 00:22:36,240 development, the thing Mask is alluding to as profit starts to 354 00:22:36,240 --> 00:22:38,680 look quite difficult to ever capture. 355 00:22:38,960 --> 00:22:42,760 Very large research and development spend to go to Mars 356 00:22:42,880 --> 00:22:47,360 or to go to the Moon can only be recouped with high volume future 357 00:22:47,360 --> 00:22:51,400 deployments or high payload usage of the tools that they 358 00:22:51,400 --> 00:22:54,920 design and build. In practice, the launch market 359 00:22:54,920 --> 00:22:59,120 of delivering cargo into space has exhibited very little growth 360 00:22:59,200 --> 00:23:03,640 outside of SpaceX delivering its own Starling satellites into 361 00:23:03,640 --> 00:23:07,520 space, leading to the question of whether this is a growth 362 00:23:07,520 --> 00:23:10,800 business at all. When you move away from EBITDA 363 00:23:11,040 --> 00:23:14,800 and start analyzing the actual profitability, the picture is 364 00:23:14,800 --> 00:23:17,920 not nearly as pretty. It's essentially the picture of 365 00:23:17,920 --> 00:23:20,320 Dorian Gray of financial statements. 366 00:23:20,560 --> 00:23:24,400 The IPO pitch looks young and flawless, but the balance sheet 367 00:23:24,400 --> 00:23:27,240 in the attic is looking increasingly gnarled. 368 00:23:27,760 --> 00:23:31,840 Well, SpaceX is a $16 billion revenue business. 369 00:23:32,080 --> 00:23:37,480 Leaked documents show that XAI generated just $120 million in 370 00:23:37,480 --> 00:23:40,840 revenue during the first nine months of 2025. 371 00:23:41,120 --> 00:23:45,240 To put that in perspective, XAI is currently burning through $1 372 00:23:45,240 --> 00:23:48,480 billion a month to build it's data centres. 373 00:23:48,720 --> 00:23:53,760 That means that XAI spends it's entire annual revenue every six 374 00:23:53,760 --> 00:23:56,760 days. By merging the two businesses, 375 00:23:56,960 --> 00:24:01,080 Musk is using the Moat of the rocket business to subsidize the 376 00:24:01,080 --> 00:24:05,280 astronomical costs of the AI business, all while asking the 377 00:24:05,280 --> 00:24:09,040 public to pay a massive premium for the privilege of owning 378 00:24:09,040 --> 00:24:12,640 shares. Ultimately, SpaceX might be an 379 00:24:12,640 --> 00:24:16,040 extraordinary engineering success, but as a public 380 00:24:16,040 --> 00:24:19,560 investment, the price you pay for the expected growth will 381 00:24:19,560 --> 00:24:22,160 matter. If the valuation is based on 382 00:24:22,160 --> 00:24:25,920 comic book heroic assumptions about capturing a billion 383 00:24:25,920 --> 00:24:30,000 customers that either already have better Internet access or 384 00:24:30,160 --> 00:24:33,840 can't afford to pay for the service, the final frontier for 385 00:24:33,840 --> 00:24:37,000 investors might be a very long way down. 386 00:24:37,480 --> 00:24:40,320 This all leads us to the question of how Elon Musk 387 00:24:40,400 --> 00:24:44,480 actually expects to get that $1.75 trillion price tag. 388 00:24:44,800 --> 00:24:48,560 The answer isn't found in the physics of rockets, but instead 389 00:24:48,560 --> 00:24:50,640 in the physics of the stock market. 390 00:24:50,960 --> 00:24:54,520 The most important thing to understand about this IPO is 391 00:24:54,520 --> 00:24:58,120 that SpaceX isn't selling the whole company to the public. 392 00:24:58,320 --> 00:25:01,280 They're likely only releasing a tiny float. 393 00:25:01,520 --> 00:25:05,320 The financial press is suggesting that just five to 10% 394 00:25:05,320 --> 00:25:09,000 of the total shares might be issued by keeping the supply of 395 00:25:09,000 --> 00:25:13,000 stock artificially low while marketing it to every retail 396 00:25:13,000 --> 00:25:16,160 investor on the planet. This could create a supply 397 00:25:16,160 --> 00:25:19,520 squeeze before the opening bell even rings. 398 00:25:19,840 --> 00:25:23,280 When everyone wants a piece of the future but there are only a 399 00:25:23,280 --> 00:25:27,080 few shares to go around, the price has only one way to go, 400 00:25:27,200 --> 00:25:30,200 and that's up. Craig Cobburn wrote about this 401 00:25:30,200 --> 00:25:34,040 low float strategy a few years ago in the Financial Times, 402 00:25:34,240 --> 00:25:38,640 where he pointed out that a lot of recent tech IPOs had floated 403 00:25:38,640 --> 00:25:43,280 only a tiny number of shares, often between 7 and 10%, 404 00:25:43,480 --> 00:25:46,360 compared to the historical average of 20%. 405 00:25:46,640 --> 00:25:50,120 He used the example of Instacart, where only 8 percent 406 00:25:50,120 --> 00:25:53,680 of the shares were floated, and of those, 60% went to 407 00:25:53,680 --> 00:25:57,600 cornerstone investors and 5% to friends and family. 408 00:25:57,840 --> 00:26:01,760 This left a tiny supply of shares to be traded, presumably 409 00:26:01,760 --> 00:26:03,960 to cultivate a sense of scarcity. 410 00:26:04,280 --> 00:26:08,960 The real magic for SpaceX might happen with the index inclusion 411 00:26:08,960 --> 00:26:11,400 trap. This isn't just a lucky break 412 00:26:11,400 --> 00:26:15,920 for Elon Musk, it's reported to be driven by a specific demand 413 00:26:15,920 --> 00:26:19,360 from him. According to Reuters, SpaceX 414 00:26:19,360 --> 00:26:23,760 made early inclusion in the NASDAQ 100 index a necessary 415 00:26:23,760 --> 00:26:26,680 condition for agreeing to list on the exchange. 416 00:26:27,000 --> 00:26:31,040 In response, the exchange is consulting on a new rule that 417 00:26:31,040 --> 00:26:36,160 would allow SpaceX to join the NASDAQ 100 after just 15 trading 418 00:26:36,160 --> 00:26:40,440 days, bypassing the year of price discovery required for 419 00:26:40,440 --> 00:26:43,760 every other company. And it isn't just the NASDAQ, 420 00:26:43,920 --> 00:26:46,720 either. Bloomberg reports that S&P Dow 421 00:26:46,720 --> 00:26:50,280 Jones Indices is also considering historic rule 422 00:26:50,280 --> 00:26:54,560 changes to fast track SpaceX into the S&P 500. 423 00:26:54,880 --> 00:26:59,000 This is significant because there are roughly $24 trillion 424 00:26:59,000 --> 00:27:02,880 tied to that index. Maybe Elon can get on the phone 425 00:27:02,880 --> 00:27:06,360 to the people at Russell to see if he can get SpaceX classified 426 00:27:06,360 --> 00:27:10,720 as a value stock too. If SpaceX is added to the S&P 427 00:27:10,720 --> 00:27:15,240 500 and the NASDAQ 100, every passive investor will find 428 00:27:15,240 --> 00:27:18,280 themselves buying the stock at whatever price the supply 429 00:27:18,280 --> 00:27:22,240 squeeze has manufactured. As veteran fund manager George 430 00:27:22,240 --> 00:27:27,160 Noble points out, the proposed five times float multiplier used 431 00:27:27,160 --> 00:27:30,760 to determine it's NASDAQ index weighting is shameless 432 00:27:30,760 --> 00:27:34,520 structural manipulation. It effectively turns every 433 00:27:34,520 --> 00:27:39,120 pension fund into exit liquidity for SpaceX inside, where your 434 00:27:39,120 --> 00:27:42,600 retirement account is being forced to buy into a bubble 435 00:27:42,600 --> 00:27:46,280 designed by the seller. As Noble argues, the rules are 436 00:27:46,280 --> 00:27:50,160 being rewritten to benefit IPO issuers and early stage 437 00:27:50,160 --> 00:27:54,640 insiders, and your capital is the tool being used to enrich 438 00:27:54,640 --> 00:27:57,400 them. We've seen this movie before. 439 00:27:57,560 --> 00:28:02,600 In December 2020, Tesla was added to the S&P 500 after a 440 00:28:02,600 --> 00:28:06,000 massive run up in price. Since the day of inclusion, 441 00:28:06,160 --> 00:28:10,480 Tesla has underperformed the broader index by more than 20%. 442 00:28:10,720 --> 00:28:13,920 The passive investors who were forced to buy in at a huge 443 00:28:13,920 --> 00:28:18,360 valuation right before sales started drying up became the bag 444 00:28:18,360 --> 00:28:22,040 holders while the early insiders took their victory laps. 445 00:28:22,640 --> 00:28:27,160 SpaceX may be an extraordinary engineering company, but it's 446 00:28:27,160 --> 00:28:30,760 been sold to the public as a financial miracle at a price 447 00:28:30,760 --> 00:28:33,600 that is, to me, very difficult to explain. 448 00:28:33,880 --> 00:28:36,880 That doesn't mean that the IPO will fail or the stock price 449 00:28:36,880 --> 00:28:41,360 will fall, but a lot of things will have to go right for SpaceX 450 00:28:41,480 --> 00:28:43,800 for it to be a good long term investment. 451 00:28:44,200 --> 00:28:46,280 Thanks for tuning into this week's podcast. 452 00:28:46,440 --> 00:28:49,720 The podcast is entirely supported by viewers like you on 453 00:28:49,720 --> 00:28:53,120 Patreon, where you can contribute as little as you feel 454 00:28:53,120 --> 00:28:56,880 like to keep the podcast going. If you'd like to support it, 455 00:28:57,000 --> 00:28:59,040 I'll give a link in the show notes. 456 00:28:59,200 --> 00:29:01,640 Have a great week and talk to you again soon. 457 00:29:01,760 --> 00:29:02,080 Bye.