E188 | Teaching Itself to Build Rockets and Challenge SpaceX: Rocket Lab’s Space Ambitions

Summary

• Rocket Lab is the closest challenger to SpaceX. Excluding Chinese and Russian state programs, it ranked a clear second globally with 16 launches in 2024; its business model is end-to-end, and its stock rose as much as 7x in six months before sliding after being targeted by short sellers. Former rocket scientist 陈亮’s assessment: a compact, elegant technical path that escapes the endless Falcon 9 variants, but still runs a beat behind SpaceX.
• A 2025 first flight for Neutron is far from certain. Industry peers interviewed in a short-seller report called the probability zero; based on the standard development sequence—concept validation, design, prototype, test article—陈亮 says Neutron is still at the prototype stage: no complete vehicle, no propulsion-system hot-fire, and launch-site construction remains early. “Under normal timelines, it still has more than a year to go; year-end would be an ideal outcome.” His experience: “It seems none of the commercial space projects I’ve worked on has ever hit its schedule.”
• Neutron’s key breakthrough is not the “Hungry Hippo” fairing, but the closed loop built around a 7-meter-diameter carbon-fiber airframe and vertically integrated production. The fairing was “archaeological innovation”—陈亮’s team evaluated a similar concept in 2019 and rejected it—but “I thought it couldn’t be done in 2019; it rolled out in 2022.” Electron offers the counterexample: helicopter recovery failed, and the value of cleaning a recovered vehicle from the sea is questionable. “It isn’t worth much—why recover it? Only when rocket costs reach a certain scale can the recovery story go very far.”
• The investment thesis: never build only rockets. SIA data puts the 2024 global space economy at roughly $400B, with launch services accounting for just $7.2B versus $110B for satellite services. 黎竹岩’s analogy: rockets are the family’s Alibaba B2B big brother, earning enough to put the Taobao kid brother and Alipay little sister through school; the real opportunity is “China Mobile plus Huawei in the sky.” The next generation of rockets will be judged on three things: full-flow technology, a core stage at least 4.2 meters across, and payload approaching 3%.
• If SpaceX does not spin out its businesses, its long-term value may come almost entirely from Starlink. Today, on financials, the split is roughly 50/50 or 60/40. Geometric declines in launch cost mean a rapid plunge in revenue per unit; payload growth cannot keep up, so “the financials will look worse and worse”—the dating-site paradox. One proposed structure would make X an unlisted “laying hen,” with Starlink, Dragon and other offspring going public to build value for the family.
• The manufacturing system is the heart of it. Relativity is not a 3D-printed rocket company but a company that “compresses the supply chain to the limit”: suppliers fell from 1,500 to 90, and “adding equipment means adding capacity.” “Which is harder to build, a Ferrari or a Toyota? Anyone from manufacturing will say Toyota.” Starship’s point-to-point transport variant, potentially cutting U.S.-Europe travel to one hour, “will to some extent replace today’s Boeing and Airbus.”
• The biggest event of 2025 is still Starship’s second-stage chopstick catch. “There aren’t many things left, beyond SpaceX, that can get everyone excited together.” Geopolitics is a hidden demand driver: BlackSky cut responsive-launch lead times for Ukraine from months to 45 days; integrated communications, remote sensing and navigation, combined with large-scale constellations, could pull through a huge market. But 陈亮 says plainly: “As an engineer, I find it thrilling, but it seems to betray the original purpose.”

Deep dive

1. Rocket Lab: A Mini-SpaceX Whose Stock Rose as Much as 7x in Six Months

• Host 刘一鸣’s opening thesis: SpaceX is valued at $350B, surpassing ByteDance as the world’s most valuable unicorn, but commercial space needs competitors to become a real industry. In the 2024 ranking of orbital-class launches, excluding Chinese and Russian state programs, Rocket Lab ranked a clear second with 16 launches. Its business model is end-to-end, making it “the company most like SpaceX among the emerging players.”
• The capital-markets backdrop: the stock rose as much as 7x over the past six months before declining after short sellers took aim. Even so, Rocket Lab remains worth watching as a company capable of competing with SpaceX on price. Founder Peter Beck was a self-taught “wild” rocket enthusiast from New Zealand; he founded the company in 2006, and its headquarters moved from New Zealand to California only after its 2013 Series A.

2. Electron: The Smallest Successful Liquid Rocket, “Small but Elegant” in Engineers’ Eyes

• 陈亮’s breakdown: Electron has a liftoff mass of just 13 tons and a diameter of 1.2 meters, making it “probably the smallest liquid-fueled launch vehicle to have successfully launched—or, more precisely, the smallest with real batch-launch capability.” Its core innovation is a carbon-fiber airframe and carbon-fiber propellant tanks. “For those of us who build liquid rockets, carbon-fiber tanks are extremely difficult. China has not actually broken through this technology.” Electron combines them with a small electric-pump-fed engine to achieve lightweighting and faster production through radical simplicity.
• The track record comes with caveats: its 2017 maiden flight failed because of a second-stage problem; since then, it should have completed more than 50 launches, with what should be 4 failures. “Its success rate is not particularly high compared with the Long March rockets, but for a newly developed startup it is basically acceptable.” The launch site sits on a scenic peninsula in New Zealand. “The rocket and the launch site are both small but elegant.”

3. Constellation Demand Forced Neutron Into Existence; Beck Really Ate His Hat Over Recovery

• The demand logic: Starlink changed everything. Falcon 9 launches 15-16 tons of Starlink satellites per flight, while domestic constellation demand is also above 5-6 tons, making medium-lift rockets with 4-20-ton LEO capacity an important segment. Neutron’s LEO payload is about 8 tons. Its first stage and fairing are recovered together, while the second stage “spits itself out of the fairing.” 陈亮’s metaphor: “The first stage is essentially a transport truck; the second stage is an expendable rocket. The truck reaches orbit, opens its cargo bay and throws out the second stage, closes the doors, then flips itself back.”
• Beck once declared that he would never build a reusable rocket and joked that he would eat his hat if he did. After seeing Falcon 9 succeed, he put the company hat into a blender, pulverized it and ate the strips. “I have no idea how much he actually swallowed, but he did eat it in the video.”

4. Neutron’s 2025 First Flight? By Engineering Standards, Year-End Would Already Be Remarkable

• A short-seller report called the probability of a 2025 Neutron launch zero. 陈亮 maps the project against a four-stage development process: key technology work is basically complete—the fairing validation test just “put on a show”—and the concept design must be finished. The vehicle is currently equivalent to the prototype stage. There is no visible data on the duration of Archimedes engine hot-fires, no complete vehicle or propulsion-system test, and launch-site construction is still relatively early. “Under normal development rules, it still has roughly a year to go. Flying by year-end would be an ideal outcome.”
• His experience is blunt: “It seems none of the commercial space systems I’ve worked on has ever hit its schedule.” Starship’s delays have also been severe. Going public has a cost: set a target, miss it, then set another. “Musk was deeply resistant to going public early on. Once you are public, the decision-making power is no longer in your hands.” Capital markets can disrupt the development cadence.

5. The “Hungry Hippo” Wasn’t a New Invention: Archaeological Innovation and a 7-Meter Closed Loop

• At the end of 2019, 陈亮 led a team evaluating a similar concept: open the fairing, eject the second stage, close it again and recover the first stage. They rejected it. A slender Falcon 9-type configuration returning with a fairing is “like adding a section of paper to the top of a pencil—a very fragile structure that could easily break during reentry.” They later found that online users had proposed similar concepts years earlier. “There is nothing new under the sun. If you want to innovate, go dig through NASA’s reports from the 1950s and 1960s and 冯·布劳恩’s wild ideas. That is archaeological innovation.”
• Rocket Lab made it work by making the rocket short and fat. A medium-lift rocket with a diameter of 7 meters would be historic; Long March 5 has a liftoff mass of nearly 900 tons, while the largest domestic rocket diameter is only 5 meters. The shorter and fatter the vehicle, the harder it is to break. A 7-meter diameter only makes economic sense with Rocket Lab’s vertically integrated composite-production capabilities, because aluminum structures require expensive tooling and molds. “It can unify production, R&D and the technical route into one closed loop. My understanding in 2019 and 2020 was that it couldn’t be done. It rolled out in 2022, finally letting us escape the endless Falcon 9 variants.”

6. Helicopter Catch Is an Obsolete Plan; Recovery Economics Have to Work

• 陈亮 knows the history of parachute recovery in detail. Falcon 9 Version 1 tried it; the Space Shuttle and Ariane 5 used it for solid boosters. “The results were mediocre. More often than not, it was recovery for recovery’s sake.” Electron’s helicopter catch was “an ideal that was very full, and a reality that was very thin”: the helicopter caught the parachute, but control was difficult, so the approach was abandoned. The coincidence is striking: Lockheed Martin explored almost the same concept around 2000, using essentially the same helicopters—and Lockheed was an investor in Rocket Lab. “At the time, many people wondered whether Lockheed had handed these ideas to Rocket Lab to generate some hype and lift its market cap.”
• The hard economics are unforgiving: sending ships to salvage a vehicle costs hundreds of thousands or even more than $1M, and cleaning it costs more, while the recovered hardware itself may be worth only a few million dollars. “It isn’t worth much—why recover it? Only when rocket costs reach a certain scale, and recovery does not make the launch particularly expensive, can the recovery story go far enough to genuinely reduce costs.” SpaceX says a fairing costs $6M, making recovery worthwhile. “I do not quite understand why it costs that much, but its business model closes, so I accept it. A comparable domestic fairing costs roughly 7x that amount in renminbi, and nobody is interested in recovering it.”

7. The VC Framework: $7.2B in Launch Services vs. $110B in Satellite Services—Never Build Only Rockets

• 黎竹岩 starts with a rule: the core of venture capital is “betting on something big emerging.” Companies with high standing at the Hannover Messe are not necessarily destined to become large. A robot-reducer company, for example, may have sales with “one or two fewer zeros” than Bosch or Toyota. SIA data shows that the 2024 global space economy was roughly $400B, with launch services accounting for just $7.2B and satellite services $110B.
• His central analogy: Alibaba’s B2B business is the family’s big brother, earning money to support the Taobao kid brother and Alipay little sister through university. Rockets are that big brother. “Without the rocket business, there is nothing else. But if you only do rockets, it is like holding a good card and still failing to capture most of the chips on the table.” His vision: “One of the truly great things in the next era will be China Mobile plus Huawei in the sky.”
• His three-way test for next-generation rockets—he says he is not sure whether others agree—comes down to who first solves full-flow technology, who builds a core stage above 4.2 meters, and who gets payload to approach 3% earliest. “All 3 are extremely difficult, but all 3 are highly valuable and meaningful.”

8. Breaking Down SpaceX’s Valuation: “The Laying Hen” and Financials That Keep Looking Worse

• 黎竹岩 estimates that, on today’s financials, rockets and Starlink are roughly “50/50 or 60/40”; at the end of 2023, projected revenue from the 2 sides was already close. Over the long term, however, “the value will increasingly come almost entirely from Starlink.” The paradox is that geometric declines in launch cost mean a rapid fall in revenue per unit, while payload growth cannot keep up. “Multiply the 2 together and the financials still look worse.” It resembles a dating site: “Every time you satisfy the other side’s core need, you lose the ability to monetize those 2 people, because they get married.”
• One structural proposal is to keep X permanently private as a “laying hen,” with its offspring building value for the family. The first child would be Starlink—the mobile network and Huawei in the sky, with the U.S. and China potentially the 2 main players. The second might be Dragon. The third is “I don’t know”; perhaps it comes from the Artemis program: landers, ports in cislunar space, refueling stations that do not pump gas, or hotels.

9. What Makes a Founder Stand Out: Someone Who Can Fix the 2 Wrong Rules in a Set of 100

• 黎竹岩 favors teams that combine disciplines. Musk’s early business was “Alipay,” later merged into PayPal, with no connection to the space industry. Aerospace has little tolerance for error and a deep archive of painful lessons. “If there are 100 rules, 98 are probably right. If you try to change them, you will be proven wrong most of the time—but they cannot be 100% right. Who can change the 2 unreasonable rules without getting shouted to death or spending all the money?” Senior experts from the state system may find this especially difficult: “A Chinese stomach cannot eat too much white man’s food. It is a habit of thinking.”
• 2 vivid examples make the point. Falcon 9’s antenna was damaged by a rainstorm on its maiden flight; a state team would not have allowed it, but the decision may have preserved the launch window and commercial interests. Born in 1986, he remembers having to change into shoe covers and draw heavy curtains before entering the computer lab in elementary school. A few years later, in an internet café, people wore flip-flops, drank Coke and let the owner’s dog run around. The computers were no different. “Do you want a computer room that is pampered and requires shoe covers, or do you want to play games in an internet café? Musk’s chaotic launch site is that internet café.”

10. 陈亮’s Balance: Falcon 9 Was Actually Conventional; Its Boldness Was Tested Step by Step

• Do not focus only on Starship’s disruptions—its tank was not fully welded before rainwater got inside, yet they hoisted it up, poured out the water and kept using it, while traditional practice calls for expensive deionized-water cleaning. Falcon 9’s production conditions and quality management are “not very different from traditional industry.” SpaceX initially took NASA contracts and operated under detailed requirements and supervision; after winning crewed missions, it became more conservative. During the first crewed Dragon flight, Shotwell particularly emphasized that “Bob and Douglas are fathers, sons and husbands,” so the team would understand the stakes.
• Falcon 1 offers the evidence chain for trial and error. It once removed a tank baffle designed to suppress propellant slosh, concluding from analysis that it was unnecessary. After a failure, the baffle was added back. “You need a certain level of skill and confidence before you can be bold. It keeps probing the boundary, perhaps retreats a little, then takes another step forward.” The priority order differs: 陈亮 ranks quality, schedule and cost; Musk might rank schedule, cost and quality. More important is who makes the call. “Even as a deputy chief designer, I would feel nervous approving a breakthrough and worry about what happens if it fails. Musk is a product-manager-style boss. Once he makes the call, he pushes it down.”

11. A Breakthrough That Never Happened: Fixed Grid Fins Rejected Over “What If Something Falls?”

• 陈亮’s firsthand example involved eliminating grid-fin deployment and keeping the fins fixed during ascent to remove the mechanism. “Anything that moves in the sky is unreliable. Aircraft landing gear often fails to deploy. The best way to improve system reliability is to cut the thing out.” The proposal held up technically, but one concern prevailed: insulation debris had historically fallen off, covered a grid fin and caused an attitude-control failure. “I asked around, and nobody dared say that debris would never fall. So I abandoned it.”
• His candid review: “I thought the probability was genuinely low, and it could have been validated through testing, but I did not have enough time to wait and was not quite willing to take the risk. I cannot say whether that was right, but the folding design had been validated and was easier for people to accept.” Commercial space companies have already explored meaningful cost and schedule reductions within state-team standards; they are simply “not as concrete or as disruptive as SpaceX.” Disruption accumulates from small, non-iconic breakthroughs.

12. 2025 Watchlist No. 1: Starship’s Second-Stage Chopstick Catch and the Misread Relativity

• 陈亮’s candid assessment: “It feels like there are not many things left in the world, beyond SpaceX, that can get everyone’s attention and make everyone excited together.” The first-stage chopstick catch has proven feasible. The focus now is whether the second stage can be recovered. Neutron will matter, but its impact will be smaller. European small private rocket companies are beginning to emerge and may deliver results this year.
• 黎竹岩 is less focused on whether Relativity succeeds or fails. “There are too many variables, but the path it is trying to take is unquestionably needed by industry.” Many people think Relativity is a 3D-printed rocket company; in reality it is a company built around “compressing the supply chain to the limit.” It cut suppliers from 1,500 to 90, and to some extent, adding equipment means adding capacity. Tim Ellis’s narrative goes further: put enough printers into rockets to print entire cities and homes for colonies in near-Earth space and deep space. Space-based industrial goods should be sourced locally wherever possible rather than built on Earth and projected into space, “which is an extremely uneconomic equation.”
• Aerospace is a test bed for frontier technology. Semiconductors benefited from Apollo and the U.S.-Soviet space race, while the speed and efficiency of cellular-network development could also be affected. “Its greater value may be that the technology spills out and becomes part of civilian infrastructure.”

13. The Factory Is the Product: Which Is Harder to Build, a Ferrari or a Toyota? Plus Point-to-Point Transport

• From a manufacturing perspective, today’s rockets and satellites resemble hand-built Mercedes cars before the Ford Model T. Everyone works around one station, then moves to the next after completing one unit. “Which is harder to build, a Ferrari or a Toyota? Anyone from manufacturing will say Toyota. Getting industrial products to flow out of a factory like a tide—at a specified low price, specified high volume and without major errors—is extremely difficult. Whether you treat the factory as the product determines why 2 companies whose products look similar can have completely different values.”
• The third watchpoint is Starship’s point-to-point transport variant: U.S.-Europe in 1 hour or less, and the Middle East in a little over 1 hour. It “will to a considerable extent replace the future business-jet and private-aircraft market.” Both China and the U.S. are working on it; it can also be called a hypersonic vehicle. “I believe it will to some extent replace today’s Boeing and Airbus.”

14. 3D Printing Is Already on the Production Line; Nuclear Propulsion and “Brute Force Works Miracles”

• 陈亮 says domestically delivered engines already use 3D printing extensively, including pump housings and thrust-chamber heads. 3D-printed grid fins have also entered actual service. The value is not only lower cost but faster development: “I want to build a particular structure, and it can be produced quickly—rapid setup, rapid validation, rapid experiments. That opens the door to many new technical routes.” In his view, Relativity is “a rocket defined by its production method.” Whether the final product will be competitive remains unknown, but the concept is highly interesting.
• In a change of perspective, he describes discussing Mars missions with Blue Origin engineers. Both sides agreed on nuclear thermal propulsion with hydrogen as the working fluid and high specific impulse—the same approach used by NASA and the Soviet Union in the Cold War-era world of For All Mankind. “We still think about building things with extremely advanced performance and efficiency. SpaceX’s thinking is that lower efficiency is fine if the quantity is large enough to compensate. As long as costs are low enough, sending 3 tons with a 1,000-ton vehicle—0.3% payload efficiency—is acceptable. Brute force works miracles.” The current pursuit of 3% and 4.2 meters reflects the industry’s stage of development.

15. Geopolitics Will Pull Through a Huge Market, but Engineers Would Rather Not Need the Catalyst

• After the Russia-Ukraine war began, BlackSky asked Rocket Lab to alter an orbit on short notice so a satellite would pass directly over the conflict zone. A change that previously took months took Rocket Lab 45 days. Responsive launch—“like Didi on demand”—could generate a large volume of orders. 陈亮’s extrapolation is more consequential: truly integrated communications, remote sensing and navigation, combined with enough satellites, could let command authorities assess battlefield conditions almost in real time. “That is a frightening future trend.” Add wartime satellite-network destruction and countermeasures, and both launch capacity and response time become critical. “From a security perspective, this will inevitably pull through a huge market.”
• But he remains ambivalent. “This resembles a new space race. As an engineer, of course I find it thrilling, but it seems to betray the original purpose.” He would rather see the commercial market solve the problem on its own. “If the space race is about who reaches Mars first or who goes to the Moon first, I think that is fine.”