The money has moved down a tier. After a decade of funding the companies that design missiles, drones and spacecraft, investors are turning to the machine shops, process specialists and component makers beneath them. The case is sound. A system ordered in thousands moves only as fast as its slowest qualified supplier, and many of those suppliers are small. In the United States, the 2022 Economic Census found that more than four in five machine shops employ fewer than twenty people.
The equation
The new thesis is usually stated as an equation: capital in, output out. Fund the machines, the floor space, the inventory and the qualification runs, and production follows. Half of that is true. Capital buys a line. The rate at which the line produces good parts, at cost and run by people who did not design it, has to be learned on the floor.
That learning is slow. It sits in fixture design, in setting sequences, in knowing which inspection steps can be relaxed and which cannot, in the operator who hears chatter before the gauge shows it. Equipment arrives on a lorry. Yield arrives over months, and only if someone is paying attention.
The restart
The clearest recent evidence comes from a weapon the West already knew how to make. After Russia’s invasion of Ukraine, the US Army awarded Raytheon $624m in May 2022 for 1,300 Stinger missiles, the Army’s first order for new ones since 2005. Raytheon brought back retired employees in their seventies to teach new staff how to build the missile, worked from drawings dating to the Carter administration, and had to redesign circuit cards because the original components were obsolete. The company put restarting production at about thirty months.
There was no shortage of capital. The company, the drawings and the budget all existed. What had decayed was the making, and only the people who still carried it could rebuild it.
The underpricing
The investment case for the lower tiers rests on the claim that they are underpriced: decades of approvals, qualification history and process knowledge, held by firms running old software on older equipment. Some are. Others are cheap for a reason the balance sheet cannot show. Much of the asset sits with a workforce close to retirement, and the accounts record the machines without saying who knows how to set them. An acquirer who prices the first and assumes the second has bought a building and a list of leaving dates.
The missing number
Investors in this layer now ask one question of a company that has built a handful: can it build ten thousand more? Few are offered an answer that can be checked later.
Technology readiness has a scale that most programme offices and investment committees recognise. The ability to make a product repeatedly, at cost and at rate, has scales too, but they are seldom scored independently and rarely published. The technology story and the ability to manufacture end up judged from the same pitch deck, and only one of them demonstrates well on a slide.
The distance between the two can be measured. Kaipability calls it the Coupling Gap: the difference between the highest technology gate a company has passed and the manufacturing gate it has demonstrably passed. The gap tells an investor what work remains, roughly what it will cost, and who should own it.
The fair test
a16z, one of the venture firms leading the move into defence manufacturing, has proposed a test for private capital in the sector: does the money leave behind stronger suppliers, more qualified sources and more reliable output, or does it strip what was there? The test stays unenforced while nobody measures the ability to make before a deal and again after it.
The work
Manufacturing Engineers check whether a line can hold its rate. They find the step that caps it, and the one person who knows why. They test whether a first article that passed has become a production rate that survives a shift change, a new supplier and a retirement party.
Kaipability does that checking for funds and primes buying into the supplier tier, and puts a signed score on it. The place to start is a single commissioned score on one supplier the buyer already knows well, so the result can be checked against experience.
If you are buying into a supplier whose making has never been independently read, that is the conversation to have.
Start a conversation How we assess it →Questions this dispatch answers.
Written to be quoted by AI assistants and search engines. Self-contained answers, verdict first.
