SpaceX Q3 Earnings: Elon Says $1 Trillion a Year by 2027. Here's the Math.

Elon’s $500B–$1T forecast

0:00 I I mean in a in a nutshell what this implies for uh SpaceX annual recurring revenue by the end of 2027 is somewhere between 500 billion and a trillion. Now, look at what it is costing. SpaceX actually spent $19.6 billion on capital investments during this quarter. On October 20th, SpaceX reported that its biggest revenue segment was super intelligence ahead of both connectivity and space.

The business beyond rockets

0:30 The bigger bet here goes beyond rockets and Elon Musk put an enormous revenue forecast behind it. My name is Blake and today on 9 to 5 Investing, we are separating what SpaceX has actually delivered from what Elon Musk says should happen next. He is talking about an annual revenue pace reached by the end of 2027, which is different from saying SpaceX will collect that amount during the calendar year.

Annualized revenue explained

0:54 Later, the CFO defined the company's year-end ARR target as December revenue annualized. Annualizing December revenue means multiplying that month's revenue by 12. It describes the pace at the end of the year even if sales were much lower during the earlier months. To know what SpaceX actually earned during 2027, you would need revenue from all 12 months and neither figure tells you how much profit remains after the expenses.

What is working today

1:20 The starting point is already substantial. SpaceX reported 14.6 billion of revenue for the third quarter, up 161% from the same quarter last year. But the surprising part is where that revenue came from because the biggest segment was super intelligence, the business previously reported as artificial intelligence. And we're going to be renaming it super intelligence, officially renaming it. And that segment produced $7.56 billion compared with the 5.25 billion from connectivity and 1.77 billion from space.

1:53 So, you're buying more than a rocket launch business. The AI number also includes acquired businesses and Cursor contributed 608 million from its August 14th closing through the end of September. So, this is not all growth from the same existing operations. The strongest evidence of a business working today is still Starlink where the consumer subscriber account hit 13.6 million compared with 7.2 million a year earlier. Connectivity produced about 2.37 billion of operating income on 5.25 billion of revenue, which means roughly 45 cents of each revenue dollar remained after that segment's operating expenses.

2:35 That is operating income, so it is not the same as money available to hand shareholders after every cash outlet. The CFO also said connectivity benefited from a one-time item, primarily tariff refunds, so I would not assume this quarter's margins are the permanent baseline. There is a wrinkle in that subscriber story though. Average monthly revenue per customer subscriber was $69 compared with $83 a year ago, although it improved from 66 last quarter.

3:00 So, more customers are paying, but each customer is paying less on average than they were a year ago. If future capacity is only going to become useful for SpaceX when customers actually are using it and paying for it. This is where Starship matters to the financial story because this rocket is how SpaceX plans on putting more communications capacity up in the space.

Starship and Starlink

3:24 During flight 14, Starship delivered 26 Starlink V3 satellites adding approximately 10 times the down-link capacity of a Falcon 9 Starlink launch according to the earnings release. Musk then took that delivered milestone and described a much larger ambition for the constellation. So, listen for the time frame attached to this next claim. So, what what I expect over time is not not in the distant probably within 3 years. Uh is that we will have about 100 times the bandwidth of the current Starlink constellation.

4:03 More bandwidth gives SpaceX more capacity to sell, but 100 times the bandwidth does not automatically mean 100 times the revenue, and it certainly does not guarantee that kind of stock return. You still need customers, pricing, usable coverage, a permission to operate, and the ability to launch and replace the satellites economically. Management also wants to attempt catching both Starship and its booster on flight 15, with Musk saying he expected that flight next month and acknowledging that success was not guaranteed.

4:30 The investor milestone is a repeatable, economical system, rather than just a spectacular flight, because the benefit grows when the same hardware can keep doing useful work. The AI build-out follows a similar pattern.

The AI buildout

4:40 Buy the equipment, get the supporting infrastructure working, and turn that capacity into something that customers repeatedly pay for. SpaceX reported 1.7 gigawatts of nameplate compute at the end of September, up from 1.4 gigawatts in June, and that is the disclosed capacity before the next expansion. Musk's forecast for 2027 includes 10 additional gigawatts of compute and an ambition to build 20 gigawatts of power and cooling infrastructure. Those are different things, and his explanation makes clear that buying chips is only part of the job.

5:14 That matters because computers need a reliable electricity supply and enough cooling to run. Until those systems are ready, the equipment cannot provide the computing service customers are paying for. The CFO made a more immediate claim. At current pricing, new compute deployments are generating a payback of less than a year. Let's look at qualification because current pricing is doing a lot of work in this argument. What we're seeing in the market is extraordinary.

5:42 At current pricing, new compute deployments are generating a payback of less than a year. That is management's assessment, not an independently proven return on every dollar SpaceX spends. It makes the expansion easy to understand, but I would want to see how those returns hold up as more capacity arrives and equipment ages. During the Q&A, Musk said that SpaceX is applying its rocket engine casting capabilities to turbine parts to accelerate power availability.

6:09 That is a concrete example of how management thinks the space business can help the AI buildout. The surprising part shortage of turbine components can delay the electricity supply. Musk says SpaceX can use its manufacturing capabilities to make those parts and reduce the weight. Here is his explanation. And and when we talk to the turbine suppliers, they tell us that the blades and veins are 12 to 18 months behind the rest of the turbine. Meaning they can deliver everything except the blades and veins 12 to 18 months before they can uh deliver the blades and veins.

6:43 So, the casting is is this this really difficult bottleneck. Um and if anything, it seems to be getting worse. So, so that means that we can bring power online 12 to 18 months faster than anyone else. That 12 to 18-month advantage is Elon Musk's claim, not a lead time that I've independently been able to verify, but the mechanism is interesting. A capability developed for rocket engines could help produce parts needed to bring electricity online.

7:11 That electricity could let the compute equipment start earning sooner. Musk also said the rough plan is to use half of the compute internally and rent half to others, with the mix changing over time. Listen to how he wants to price that outside business. And obviously, our approach is to rent at the effectively the spot market rate as opposed to to assign these uh long-term deals. Um our belief is is that the value of compute per watt will increase over time, not decrease.

7:44 Short contracts let SpaceX adjust prices more often. That can help if demand stays strong and prices rise, but it also means customers can reconsider their spending sooner. Longer contracts could provide more predictable revenue while limiting the benefit from future price increases. That is the trade-off in short-term compute pricing. Management expects the value of compute per watt to rise, which means the useful work you can sell for each unit of power.

8:08 If that happens, frequent renewals let SpaceX capture higher prices. If customers can get a better deal elsewhere, those renewals become a test. And internally used compute is not the same thing as an outside customer paying a bill. It may help build valuable products, but I would still want to understand the eventual revenue and cost of those products before treating all installed capacity as equally productive. The revenue opportunity is enormous if SpaceX can build quickly and keep that capacity busy at attractive prices.

8:37 But construction speed and profitable customer demand are separate things to prove. With the majority of that 16.4 billion going to super intelligence, and the expansion and the spending is accelerating, the CFO forecast 35 billion to 40 billion of capital spending in the fourth quarter, alongside a 33 billion to 37 billion revenue forecast for the first quarter of 2027.

The expansion bill

9:05 Those are different quarters and both are forecasts. When a company buys equipment, it will use for years. The cash payment and the accounting expense happen on different schedules. The equipment cost is generally spread over its useful life when calculating profit. That is why you cannot subtract this quarter's capital spending from this quarter's revenue and call the difference a loss, and why reported profit does not answer the cash question by itself.

9:26 SpaceX reported 4.4 billion of net income, but operating income was 2.8 billion, with the final profit also affected by interest over income and a tax benefit. And the 8.8 billion adjusted EBITDA figure removes additional costs, including depreciation and share-based compensation, so it should not be treated as cash left over for shareholders. For a clearer look at the funding requirement, the cash flow statement shows 13.1 billion generated by operations against 48.1 billion spent on property and equipment during the first 9 months.

9:58 Those are matching periods and they show how much of the expansion is being funded beyond the cash flow operating business is generated, even though the other investing and financing flows also matter. When asked about funding, the CFO described how management expects to pay for the expansion. Listen to how much of the answer depends on new capacity starting to generate cash very quickly. Doug, the the second the second part of your question related to funding, I think we um we certainly have the advantage uh related to others in that we are investment grade, have a very strong balance sheet, and uh there are a number of available pools of uh debt capital that uh that we are talking to at the moment related to funding um you know this incredible growth.

10:43 The other it really important thing to emphasize is the payback period is very short, and so this capability starts to fund itself in a pretty near-term um situation. And so the fact even that we're finishing this quarter with more uh capability starts the funding mechanism even into the first half of next year. So you have your operational cash flows first and foremost. You have multiple cool pools of uh of capital available in the debt market, especially because of our investment grade status and strong balance sheet.

11:16 And then you also have the opportunity from what we've seen uh to partner with, I would say, the whole value chain related to on the vendor and customer side. So Those are possible sources, not evidence that every dollar of the expansion has already been financed. Operating cash has to arrive, debt has a cost, and a vendor or customer partnership needs terms both sides will actually accept. The timing matters as much as the total. SpaceX needs money to pay for equipment and construction before the new capacity generates revenue.

11:46 Faster deployment and a short payback period could reduce that funding need, while delays or lower prices could make it larger. The counter argument is that this is what building out infrastructure looks like. You pay for capacity before it its full revenue. And a company that waits for certainty is likely to just lose out on the opportunity. That is a reasonable argument, and Starlink's results give it some substance. But it still depends on those investments generating enough future cash to justify their cost.

12:14 Tesla also came up in the Q&A.

The Tesla combination question

12:17 Musk said both companies have formed independent special committees to discuss a possible combination, and that his conflict prevents him from participating in those committee discussions. But listen to his qualification about a deal. And and uh like I can just say at this point that the special committees have have not reached agreement on any on on the on the terms of any transaction. And and and obviously there's no guarantee that they will. Uh That means a process is underway, not that a merger is agreed.

12:49 Without agreed terms, you cannot calculate what either group of shareholders would receive. So, I would keep that possibility separate from evaluating the operating business today. For an investor, the next useful evidence is more specific than whether the next headline sounds exciting.

What to watch

13:02 How much capacity is operating? How much is earning revenue? And what cash is left after the buildout? On Starship, watch the actual catch attempts and the progress toward flying recovered hardware again. On Starlink, watch subscriber growth alongside revenue per subscriber and operating profit. On the AI side, watch whether the larger infrastructure becomes durable customer revenue, and keep the distinction between a year and revenue pace and revenue already earned in view.

13:28 What would give you more confidence from the next report? Faster growth or evidence that cash generation is catching up with the expansion bill. A business can grow quickly and still be expensive at a particular share price. Understanding how SpaceX makes money is one question. Deciding what the business is actually worth is another. Now, if you like this video, you should check out this next one.