Everyone gets a boom. Everyone loses it the same way.

This has already happened to the United States twice, and both times were treated, at the time, as permanent American advantages rather than a phase. American manufacturing dominance built the mid-century boom, then wages rose, and by the 1970s and 1980s that same labor-intensive work was moving to Japan, then Korea, then China. American software and business-process dominance built a second boom in the 1990s, then American wages for that work rose again, and by the 2000s the labor-intensive layers of that industry, coding, testing, customer support, back-office processing, were moving to India. Both waves felt, from inside them, like the natural order of things. Both were, in the language of economics, an entirely ordinary and well-documented lifecycle.

The mechanism has a name. Raymond Vernon described it in 1966 as the product life cycle theory of trade: an innovation is born in a rich country because that is where the capital and skilled labor are, it is produced there while the country still has a cost advantage, and as the country's own wages rise, production migrates to progressively cheaper economies while the innovating country retains only the high-value design and coordination layer. Kaname Akamatsu, writing decades earlier about East Asian industrialization, called the country-to-country version of the same pattern the flying geese paradigm: one economy leads, matures, and hands the labor-intensive work to the next economy in formation behind it, which will eventually do the same to the one behind it.

Now watch India build a real, well-financed deeptech boom, on schedule, right where the theory says a boom should appear next. The question this dossier actually tests is not the one usually asked, whether India's rise threatens America's edge. It is the one the lifecycle model forces you to ask instead: is India's boom the same kind of boom the US and China each got to keep for decades, or does it arrive at the exact moment something new, specifically artificial intelligence, might collapse the runway before India's own wages ever get the chance to rise the old way.

Palantir's numbers are the clearest evidence available of what survives when a boom matures into a design-and-coordination economy. Q1 2026 revenue was $1.63 billion, up 85 percent year over year, with FY2026 guidance of $7.7 billion, and for the first time in the company's history commercial revenue is on pace to overtake government revenue, $3.9 billion projected against $3.8 billion. The market has priced that growth at a roughly $350 billion valuation, a forward price-to-sales ratio of 37.6x. That is not a manufacturing company's multiple. It is a coordination-layer company's multiple, the kind Vernon's theory says a matured economy retains after the labor-intensive work has already left.

Anduril is the more interesting case precisely because it is often described as sitting on top of Palantir, sensors feeding satellites feeding Palantir feeding Anduril, a clean vertical stack. That specific dependency claim does not hold up under a direct check. Financial coverage explicitly describes Anduril as a company that sells "hardware and munitions, not pure software," a distinction drawn specifically to separate it from Palantir's business, and no named public contract or technical integration confirming Anduril's Lattice software ingests Palantir's Gotham or Foundry data streams turned up in this research. not established What is confirmed is Anduril's own trajectory: valuation climbing from $30.5 billion in June 2025 to $61 billion in May 2026, reportedly in talks for $100 billion by July, on 2025 revenue that "more than doubled to $2.2 billion." The two companies are not a stack. They are two separate, adjacent bets on the same fact: that in a matured American deeptech economy, the money concentrates in whichever layer coordinates everyone else's hardware and data, not in the hardware or the data themselves.

The old primes are the fossil record of the layer beneath that one, the layer America has been slowly outsourcing pieces of for decades through subcontracting even while keeping final assembly and systems integration onshore. By federal prime contract dollars committed since October 2023, Lockheed Martin leads at over $150 billion, Boeing at over $68 billion, RTX at over $51 billion, a scale that still dwarfs the new entrants and is a reminder that maturity does not mean an economy stops building. It means the building concentrates in fewer, larger, more coordination-heavy hands, while the parts underneath disperse to wherever is cheapest.

China's manufacturing wages rising over the past two decades is one of the best-documented facts in modern development economics, the entire reason Vietnam, Bangladesh, and increasingly parts of Africa have absorbed the labor-intensive manufacturing China itself once absorbed from Japan and Korea. established economic history, not freshly re-verified in this pass What is more interesting, and squarely inside what this research did verify, is the deliberate, decades-long state project to make sure China's next boom does not depend on cheap labor at all.

In November 1993, the Fourth Plenum of the 14th Party Congress made a twelve-character phrase official policy: zhichi liuxue, guli hui guo, lai qu ziyou, support overseas study, encourage return, freedom to come and go. This is documented in peer-reviewed scholarship by David Zweig, published in the International Labour Review, using China's own statistical yearbooks. Returnees grew from near zero in 1978 to roughly 25,000 a year by 2004, accelerating sharply after 2000, backed by institutions built specifically to receive them, the Hundred Talents Programme in 1994, the Cheung Kong Scholars Programme in 1998. Surveyed scientists at the Chinese Academy of Sciences cited China's own economic development, 58 percent, and good government policy, 47 percent, as their reasons for returning. confirmed, peer-reviewed This is not a company betting on cheap engineers. It is a government spending thirty years building the human capital to skip the cheap-labor boom entirely and go straight to originating one.

The industrial evidence backs it up. China's 2025 launch cadence hit 92 orbital launches, a national record, with a 2026 target near 140 stated by one industry executive rather than confirmed as national policy. industry projection, not a confirmed government target Commercial players like Galactic Energy, which closed the largest funding round ever for a Chinese rocket startup at $330 million, are now pursuing IPOs on the Shanghai STAR Market and Hong Kong exchange, a capital-markets maturity India's sector has not reached. China's actual state space budget, often quoted as a clean $20 billion figure, does not survive a sourcing check; China publishes no itemized space budget, and the real range across serious trackers runs roughly $8 billion to $18 billion. no single reliable figure exists

A widely circulated, more recent claim, that "850 US researchers have left for China since 2011, including 50 tenure-track scholars in the first half of 2025 alone," does not hold up as usually stated. The underlying figure is real but narrower, specifically tenure-track scholars of Chinese descent, and traces to an unnamed "tally by Princeton University researchers" that no outlet links to a named study. refuted as commonly stated Treat the qualitative pattern as real and the specific headline number as unverified.

China is not waiting to lose its boom before starting the next one. That single design choice may be the entire difference between a country that repeats the cycle and one that escapes it.

Europe's capital formation looks nothing like a continent that missed the deeptech era. European deeptech investment reached $20.3 billion in 2025, 32 percent of all European venture capital, a record share, with defense, security and resilience alone accounting for 43 percent of European deeptech funding. Named rounds back it up: Mistral AI at an €11.7 billion valuation, Helsing at €12 billion with a defense-AI alliance between the two, Germany's Quantum Systems reaching unicorn status on a €210 million Bundeswehr contract. The European Launcher Challenge alone has €902 million committed across five companies.

What Europe does not have is output to match that capital, and this is exactly what the lifecycle model predicts a matured, high-wage economy looks like from the outside: rich enough to fund the design and coordination layer, no longer cheap enough to build volume domestically. Europe conducted fewer than ten orbital launches in 2025, against China's 92. The Launcher Challenge's own timeline, framework contracts in 2026, a demonstration deadline in 2027, ESA procurement only by 2030, is itself an admission that the output gap persists for years no matter how fast the funding gap closes. Europe is not behind because it lacks money. It is behind because it is exactly as far into the maturity phase as its wages say it should be, and money alone does not buy back a labor-cost advantage a continent gave up decades ago.

Place all four economies on the same curve and a coherent picture appears. The United States has completed the cycle twice and now holds the coordination layer in both cases, keeping the highest-multiple business (Palantir) while distributing the labor-intensive layers globally, including, until recently, to India. Europe completed an earlier industrial cycle and now holds capital without cheap build capacity, a preview of where every boom eventually lands once wages rise far enough. China is mid-transition, still shedding manufacturing to cheaper economies while deliberately building, since 1993, the human-capital base for its own next boom rather than waiting for the market to force the transition on it. India is where the model predicts the next boom should appear, and the space and defense-tech numbers in this dossier confirm one is genuinely underway.

The part the original hypothesis got wrong, and the part this dossier exists to correct, is the assumption that the interesting question is whether India's boom can out-compete America's mature economy on cost. That was never really the contest. The interesting question, the one this series' own earlier reporting on India's IT-services collapse already answered for one boom and left open for this one, is whether India gets the multi-decade runway the US and China each got before the market forced their transitions, or whether AI has changed the length of that runway for good. The US had roughly three decades before manufacturing wages ended that boom, and two decades before software wages ended the next one. India's most recent boom did not get either. It got three years.

The lifecycle model was never broken. What changed is how much runway a country gets before it has to move to the next act.