How to audit a JATM production target before you call it capacity

The JATM framework agreement announced in September 2026 is not a contract. It has no appropriated funding behind it. Those two facts set the boundary on what the announcement can be made to mean, and the boundary is worth walking before anyone calls the agreement capacity.

What the Pentagon and Lockheed Martin actually said is that they struck a broad framework agreement intended eventually to result in a multi-year production contract for the Joint Advanced Tactical Missile. The missile itself is still in development, close to entering production but not in it. Those are careful words, chosen by people who understand the difference, and they deserve a careful reading, because the gap between a framework and a running line is where output targets go to die.

A Framework Proves Demand Duration And Nothing Else

I want to be fair about what a multi-year commitment is worth, because it is worth something real. The Defense Department said the agreement establishes a framework for future Multi-Year Procurement of JATM and executes congressional authority to secure the nation's highest-priority missile systems. A supplier who believes the demand will still exist in year four will invest in tooling and hiring in year one, and a supplier who suspects the order book ends at the next budget cycle will not. That is the entire economic logic of a multi-year instrument, and it is sound.

But be precise about what the instrument proves. It proves the customer intends to keep buying for a known duration. It proves nothing about supplier yield on the components that matter, nothing about test throughput, nothing about whether the pacing bottleneck, whichever subsystem it turns out to be, has been identified, let alone cleared. Demand is not supply. A multi-year framework converts uncertainty about order-book duration into confidence, and that is the only conversion it performs.

Context makes the temptation to over-read obvious. The agreement landed as part of a broader push to boost production while the Pentagon rebuilds munitions stockpiles. The war on Iran depleted U.S. and allied missile stocks and spurred a push toward less expensive long-range precision-strike weapons. When magazines are thin and the pressure is public, every announcement gets read as inventory. That is precisely when the discipline matters most.

Where An Output Target Fails Before The First Lot Ships

The Pentagon has already said the quiet part in its own voice. It has raised concerns about whether the U.S. defense industrial base has enough skilled personnel and production capacity to handle simultaneous development and eventual mass production. It has named increased munitions production a procurement priority and is urging companies to invest in capacity and equipment. Northrop Grumman alone was negotiating component-supply expansion for roughly eight weapons programs through ten potential agreements with prime contractors. None of that describes an industrial base with slack in it.

Now add the physics of this particular missile. JATM is designed to fit inside the internal weapons bays of the F-22 and F-35, and a missile packaged for an internal bay is by definition a tightly packaged missile. Tight packaging means precision components, and precision components mean yield curves that nobody has measured at rate, because the weapon is still in development. An output target set before those curves exist is an aspiration with a unit attached. The framework cannot tell you whether it will be met.

Yield is only the first gate. Every round that comes off a line still has to clear acceptance testing, and test capacity is capital equipment with lead times and throughput limits of its own. If schedule risk on a program like this lives anywhere, it lives in a subsystem or a test cell that nobody outside the line is tracking, which is the oldest and least glamorous failure mode in the business.

The Pac-3 Ramp Is The Arithmetic Lesson

We have a live case of what a real, funded ramp costs, and it is instructive. In April, the Department of War signed a $4.7 billion contract intended to raise PAC-3 interceptor output from 600 missiles annually to 2,000 by 2030. Read those numbers slowly. That is roughly a 3.3x increase against a known baseline rate, and it still requires years of runway. The production effort involved 15 locations across 11 states, with 13,000 suppliers in the chain.

Thirteen thousand suppliers. Every one of them is a node where yield, lead time, or quality can slip, and the ramp moves at the speed of the slowest constrained node, not the average one. The genuinely encouraging data point in that story is small and specific: Lockheed Martin received its first batch of Patriot missile parts produced through the new General Motors partnership in 22 days. General Motors is providing precision castings that house the PAC-3 Patriot missiles. That is what capacity progress actually looks like: a dated delivery of a named component from a named source.

Hold JATM against that yardstick. The PAC-3 ramp has a baseline rate, a target rate, a date, and a thirteen-thousand-supplier footprint, and it is still a hard, multi-year problem. The JATM framework, as announced, is not an official contract and does not yet have appropriated funding behind it. If the funded ramp is that hard, the unfunded framework is a press release standing where a production plan will eventually have to stand.

Four Labels That Separate A Number From A Wish

Here is the test I apply whenever an output figure appears under an instrument like this one: ask which of four labels the number carries. Modeled means it came out of a production or engineering model with stated assumptions you can inspect. Specified means it is written into a requirement or a contract with acceptance criteria attached. Calculated means it is arithmetic derived from other figures, and you can check the arithmetic yourself. Objective means it is a goal, which is honorable, provided nobody dresses it up as anything else.

A number with no label cannot be audited, and a production target that cannot be audited is not capacity. It is content. My claim is narrow and I will state it plainly: I would not count any JATM output target created under this multi-year framework as capacity until the target carries a maturity label. Not because the program is unserious, and not because frameworks are useless, but because the instrument that created the target proves demand duration and nothing more. Mixed-maturity documents are fine, normal, even necessary early in a program, as long as every figure wears its label where the reader can see it.

How To Run The Audit In Five Minutes

When the first JATM rate figure appears in public, run a five-minute audit. Ask what label the number carries, and who put it there. Ask which subsystem is the pacing constraint and who is tracking it, because if nobody can answer, the schedule risk is living somewhere unwatched. Then ask what a 22-day-style proof would look like for this program, a dated delivery of a named article from a named source, rather than another agreement about future agreements.

I hold my own work to the same rule, because a rule that only applies to other people is a pose. Kibernan has produced six complete, costed engineering programs, and every published figure in them carries its maturity: modeled, specified, calculated, or objective. They are proposals. No Kibernan hardware has been built, flown, or fielded, and we say so in writing, because a proposal labeled as a proposal is more useful to a serious buyer than a target dressed up as a capability.

The JATM framework may become exactly what the stockpile problem needs, and I hope it does. But hope carries a label too, and the label is objective. Until the output targets under this framework show their maturity in plain sight, count them as intent, ask for the verification that would upgrade them, and keep asking the labeling question every time a new number appears.

Intent fills briefings; it does not fill magazines.