The missile part the Pentagon can't build fast enough comes from two companies

The war with Iran used up missiles faster than they can be rebuilt. The part the Pentagon's own buyers name first among the bottlenecks is the solid rocket motor, and one of its two makers plans to become a stock of its own.

The United States went to war with Iran on February 28. The Pentagon has not published how many missiles it has fired since, but on September 14 a quarterly report to Congress from the Lead Inspector General relayed the view of the Pentagon's own acquisition office: the munitions spent in the war resulted in strategic inventory shortfalls and revealed industrial base bottlenecks for resupply. The same report ranks those bottlenecks, and the first one it names is not a missile, a warhead or a guidance kit. It is the solid rocket motor, ahead of high-grade explosives and skilled labour.

That ordering is the story. A missile shortage sounds like a problem for the companies whose names are on the missiles. One layer down, it is a problem for the handful of plants that pour and cure rocket propellant, and there are very few of them.

The Pentagon describes the solid rocket motor industry as reliant on a duopoly of two primary manufacturers. The Federal Trade Commission named them in 2022: Aerojet Rocketdyne and Northrop Grumman. Aerojet now belongs to L3Harris, where it operates as the Missile Solutions segment. It is not a small corner of the arsenal. L3Harris says it makes the Mk 135 booster that launches Tomahawk, and the motors in THAAD, Standard Missile, Stinger, Javelin and Patriot.

In January, before the war started, the Pentagon did something it had not done before with a supplier. It committed a billion dollars of convertible preferred equity directly into L3Harris's missile business, the first direct-to-supplier investment of its kind, and the deal closed on April 23. That money is not a contract for missiles. It is a stake in the company that makes the motors, and it comes with terms an investor can read.

A solid rocket motor is a casing filled with solid propellant that, when ignited, pushes hot gas through a nozzle to produce thrust. The chemistry is old. The constraint is capacity: the buildings, the casting lines and the people certified to run them. GAO recorded in 2017 that the industry had consolidated from six US manufacturers to two, and that Pentagon demand at the time could only sustain two. That was an efficient answer to peacetime orders. It is the wrong shape for a war that burns through interceptors and cruise missiles at the same time, because nearly every program queues at the same two sets of plants.

The terms sit in L3Harris's own quarterly filing. The preferred stock converts automatically into common shares of the missile business if it completes a qualifying public listing on or before December 31, 2027, at a 20 percent discount to the offering price. L3Harris expects the government to own roughly 10 percent afterwards, with L3Harris keeping more than 80 percent. If there is no listing by then, a redemption clause takes over instead. In January both the company and the Pentagon said the listing was planned for the second half of 2026. L3Harris filed a confidential draft registration on April 29. On its July 29 call the chief executive said the company expects to revisit the IPO mid-2027, and on August 17 that chief executive left after conduct the company said was inconsistent with its code and unrelated to its financial reporting. The deadline did not move.

Read together, the structure explains the urgency. The discount pays the government for committing before the business is priced by the market, and the deadline gives both sides a reason to get it public while demand is visible in the numbers.

L3Harris has committed to a $3 billion capital build-out to increase solid rocket motor production through 2028, with roughly 60 new facilities and nearly a million square feet across its sites in Alabama, Virginia and Arkansas. In Orange County, Virginia it is spending more than a billion dollars to more than double the manufacturing space, and it says the programme will let it double, triple and quadruple production rates on key programs. It has signed a seven-year framework for THAAD and PAC-3 that it values at about $12 billion of production revenue, and on September 8 it received an undefinitized $4.7 billion, seven-year contract from Lockheed Martin for PAC-3 propulsion. The catch sits in the same release: both new PAC-3 production facilities are expected to be operational in 2027.

Northrop is expanding too. It says it is doubling motor capacity at its Utah facilities, nearly tripling it at the Allegany Ballistics Laboratory in West Virginia and adding 25 percent at Elkton, Maryland, backed by more than $3 billion of framework agreements for PAC-3 and THAAD components signed on August 3. It frames the need in the Army's own terms: PAC-3 MSE production growing from roughly 600 a year today to thousands in the near term.

The backlog and revenue in this issue are reported quarterly. The chart above tracks L3Harris and Northrop Grumman's company-wide reported revenue by quarter on TradingView, and can be saved to an account ahead of Northrop's third-quarter report on October 20. Readers who join through our link receive $15 off a TradingView subscription.

The four statements below come from parties describing their own circumstances, which is why they are worth more than any outside estimate. Note the one that disagrees. Northrop's March petition to the FTC argues the duopoly is over, pointing to new entrants. That is a company arguing its own case in a proceeding where the answer affects it, and the regulator has not ruled. The entrants are real but small by comparison: the Pentagon counted nine investments totalling $191 million in the motor industrial base under one solicitation by May, including $58 million for Anduril, and a $10.98 million award to X-Bow in July. Ursa Major, which holds a Navy award on a Standard Missile motor, has agreed to go public through a merger with a blank-cheque company, with closing expected in the first quarter of 2027. Today it is private, and what trades is the blank-cheque company.

Three things. First, the Pentagon has stopped treating the motor as a procurement problem and started treating it as an ownership problem. A contract pays for output. A convertible stake with a listing deadline pays for capacity and ties the government's return to the business growing. That is a different relationship from the one the defence industry has had for decades, and it is the clearest signal in this story of how seriously the shortfall is taken.

Second, the bottleneck will show up in a company filing before it shows up anywhere else. The watchdog report describes it in words. The segment table in L3Harris's quarterly results describes it in numbers, and the relationship between backlog and revenue is the one to read. Orders that arrive faster than the plants can ship them accumulate as backlog. New plants that open on schedule turn that backlog into revenue. Until 2027, the build-out dates say the first effect should outweigh the second.

Third, the cost of building capacity is not trivial right now. Our macro screen has the ten-year real Treasury yield near the top of its recorded history, which makes long-dated factory investment expensive to finance on ordinary terms. That is part of why a direct government stake is attractive to a supplier today, and it is a reason to expect more financing structures like this one rather than fewer. The duopoly may loosen, as Northrop argues; the question the numbers will answer is whether it loosens before or after the plants the war exposed are built.