Seven Years of Signed Demand Will Buy More Interceptors Than Another Order to Move Faster

The Pentagon has called on the U.S. defense industry to accelerate weapons production to replenish its diminished munitions stockpile. The request costs nothing to make. The capacity it asks for costs real capital, takes years to stand up, and will not be financed on the strength of a press statement.

A Pentagon spokesman said the department is focused on increasing munitions acquisitions at the pace the threat demands, and I take the intent seriously. But pace is an output of a production system, and that system responds to exactly one input the department fully controls: demand it can be held to, in writing, across budget cycles. An urgency memo retires no bottleneck. A multi-year minimum-quantity commitment retires several before they ever show up on a program schedule, because it changes what a supplier's board will approve, what a lender will underwrite, and what a machinist will bet a career move on.

Everything that is not bankable demand is weather.

Urgency Is Not A Production Input

Federal News Network put the diagnosis plainly: expanding defense production takes time, money, and confidence that demand will last. Read that list in reverse. Confidence gates the money, and money gates the time, because the people committing capital must answer for stranded assets if the demand goes away. A supplier weighing a rate increase sees mixing capacity, cure ovens, test cells, and a bench of qualified technicians, and every one of those investments pays back over many years, not one.

Against a one-year appropriation and an option year, the rational answer is overtime on existing tooling and a polite surge slide. That is not cynicism; it is fiduciary duty. We then score the surge slide as if it were capacity, and we are surprised, on schedule, when the promised peak turns out to have been a viewgraph. Nobody audits a claimed maximum rate until the moment it is needed, which is the one moment it cannot be fixed.

A request to accelerate carries no consideration a board can book. It cannot be collateralized, cannot be flowed down to a sub-tier supplier as an enforceable commitment, and cannot appear in a financing model except as a footnote about sentiment. The ask is sincere. The mechanism is absent.

What Seven Years Of Credible Demand Buys

Three things become financeable that otherwise are not. Long-lead tooling can be ordered before it is a rate constraint instead of after, which is the entire difference between a ramp and a stall. Labor can be hired and qualified on a timeline that matches the ramp, because a skilled technician will leave a stable job for a factory with a seven-year order book and will not for one living quarter to quarter. And the sub-tiers, where most rate ceilings actually live, will sign their own capacity commitments only when the prime's demand flows down to them in enforceable form.

The primes already behave this way when the demand signal is real. Lockheed Martin agreed to tentatively negotiate purchases of up to 15 tonnes of scandium oxide a year for up to 10 years. Nobody signs a ten-year raw-material commitment against a one-year order book; long input demand is only rational against long output demand. That pattern should run in both directions.

David Norquist at the National Defense Industrial Association says the industrial base needs more and larger firms to sharpen competition. I agree, and demand certainty is how you get them, because a new entrant reads exactly the same signal an incumbent does before pouring a foundation.

The Price Of Pretending, In Public Numbers

The department can commit for the long term when it chooses to. It is preparing to spend as much as $131.23 billion to keep the F-15 in production, more than half a century after the F-15 Eagle first took off. Whatever you think of that particular decision, it is a demand signal a supplier can finance against. Interceptors rarely get one, even while the department publicly asks industry to move faster on the very stockpile those interceptors are supposed to refill.

Capacity that is missing when needed carries an auditable price. The Government Accountability Office found the Navy lost more than 15,000 attack-submarine operational days over ten years to maintenance delays and idle time. These delays produced an estimated $3.4 billion in costs for crews and submarines that provided no operational capability. Different domain, same arithmetic. Paying to sustain capability that cannot be delivered is the most expensive way to save money.

And the demand itself is getting harder, not easier. Militaries now need both inexpensive counter-drone weapons and increasingly sophisticated missiles. That is two product families with different cost structures, both starved by the same annual buying rhythm, and neither one gets a factory built out of a one-year order.

Contract For A Dated Ramp, Not A Promised Peak

Here is the change I would make. Stop scoring surge capacity as a promised peak rate, because a claimed maximum is a free number until the day it is tested. Score and contract a dated, bottleneck-specific capacity ramp instead: name each constraint, whether it is energetics throughput, seeker integration, or test-cell hours; put a retirement date on each one; and back the ramp with a minimum demand guarantee long enough to finance the retirements. Seven years is my number because it covers long-lead tooling deliveries, labor qualification, and sub-tier commitments with margin to spare.

The symmetry matters as much as the duration. If the vendor misses a dated bottleneck retirement, the remedy is specific and verifiable on the factory floor, not in a briefing. If the government cuts the minimum, it pays for the tooling it stranded. Would rounds cost more in year one under this structure? Probably, and the honest comparison is not year-one unit price against year-one unit price; it is the fully loaded cost of capacity that actually exists against the fully loaded cost of a peak rate that was never real.

What Costing A Missile Factory Taught Me

I did not arrive at this position from theory. At Kibernan I produced a complete, costed engineering proposal for an AI Managed Missile Factory, one of six complete, costed programs Kibernan has produced, and every published figure in them carries its maturity: modeled, specified, calculated, or objective. These are proposals. No Kibernan hardware has been built, flown, or fielded, and we say so in writing, because a reader who cannot tell a modeled number from a demonstrated one has been misled.

What the costing exercise showed me is where the rate ceilings actually sit. They were never the headline machines. They were the long-lead tooling deliveries, the labor qualification pipeline, and the sub-tier commitments, and every one of them was financeable years before it became a constraint, provided the demand behind the factory was credible for the life of the payback. Remove the demand certainty and the same spreadsheet collapses to a smaller plant that hits its ceiling early. The engineering did not change between those two cases. The financing did.

So when the Pentagon asks industry to accelerate, I believe the intent and doubt the mechanism. The most valuable input the department can supply to higher interceptor output is not another instruction to move faster. It is seven years of demand a board can approve and a lender can underwrite, attached to a ramp with named bottlenecks and dated retirements. Sign that, and the rate follows.

Demand certainty is the cheapest factory the department will ever buy.