Warfare, industry and capital are all evolving at once.

Old playbook: Bigger budgets mean more of the same platforms, bought from the same primes.

The money is moving to attritable, software-defined systems, re-engineered weekly against a live adversary and judged on cost per effect. A $300–600 FPV drone takes out targets priced in millions.

Old playbook: The primes sell hardware.

They sell decades of operating and support attached to it: roughly 70% of a weapon system’s life-cycle cost, per the U.S. GAO. A cheap, attritable system doesn’t just beat the platform; it deletes the annuity behind it.

Old playbook: Incumbents will build the replacement themselves.

They can’t build it without eating their own economics, so acquisition becomes the clearing mechanism. Defense-tech exits hit a record $54.4B in 2025, about 72% of them acquisitions (PitchBook). Strategic M&A is our base-case exit.

Old playbook: The next defense giants will be built at venture valuations.

Raise at 30 times revenue, build at a price the customer will actually pay, return the multiple investors underwrote: you get two. Listed primes trade at two to four times revenue. That’s arithmetic, not sentiment.

Old playbook: Whoever produces the most wins.

It isn’t the bicycle factory that wins; it’s the factory that produces bicycle factories. Russia outproduces Ukraine in drones roughly two to one and is still losing the exchange. The asset is the ability to change what you build inside a week.

Old playbook: Dual-use is enough.

Defense validates the technology and pays first. Civil markets (ports, airports, borders, critical infrastructure) broaden it. Commercial is where the volume, the margin and the exit live. We won’t back a company whose only customer is a defense ministry.

Old playbook: Great technology is the moat.

AI commoditizes code. What’s defensible is proprietary data that compounds with every customer, real IP held where it will be enforced, and integration that’s painful to rip out. All three, not two: two out of three is a feature, not a company.

Old playbook: A U.S. connection is always the advantage.

Europe now writes preference into law: its EDIP program requires 65% European content in a funded product’s components, with design authority in the EU. Every company we back is domiciled outside Ukraine before the first serious conversation.

Old playbook: Valuations, letters of intent, pilots and logos signal traction.

We underwrite booked revenue and repeat orders, adoption inside a real unit, a data or integration position that compounds, and a price we can still exit from. Valuation is not validation.