One Division Turns Every Munitions Capacity Claim Into a Testable Number

The Pentagon's independent watchdog says the United States spent more than $22 billion on munitions between February and June, leading to strategic inventory shortfalls. The Defense Department's inspector general says the war with Iran caused munitions shortfalls and a bottleneck in resupply. Five months. The same reporting puts the cost of Operation Epic Fury at $33.4 billion as of June 30, with 18 American service members dead. That is the state of the magazine as the government's own auditor describes it.

Watch what happens next, because the pattern is reliable. Every institutional response to this finding will be phrased in one unit: rounds per year. Lines will be described as ramping from some thousands to more thousands annually, testimony will cite annualized rates, and expansion announcements will quote units delivered per year. None of it answers the only question the shortfall poses, which is whether production can catch a drawdown that is measured in days.

Rounds per year is a factory metric. It describes what a line can do in the abstract, against no adversary and no clock. The shortfall the inspector general documented is an operational condition: on any given day, against a stated expenditure rate, the stockpile holds some number of days of fighting. Restoring it is a race between deliveries and consumption, and an annual rate says nothing about who is winning that race.

The Unit That Answers The Question

Here is the metric I would use instead. Take a line's monthly deliveries. Divide by the daily expenditure rate of a stated scenario. The result is magazine-days recovered per month: the number of operational days the line puts back into the stockpile for every month it runs. It is one division, and anyone holding a delivery schedule and an expenditure assumption can compute it.

The division does real work because it forces two numbers into the open that annual rates are allowed to hide. The first is the denominator: which fight we are sizing against, and how many rounds a day of that fight consumes. The second is the ramp, because a line that reaches its claimed rate in three years recovers nothing this year, and averaging across the ramp flatters everyone involved. A capacity claim that omits either number is not a capacity claim. It is a press release.

This unit also makes unlike things comparable. A line delivering 2,000 precision rounds a month against a scenario that burns 100 a day recovers 20 magazine-days a month; a cheaper effector at 12,000 a month against a 400-a-day scenario recovers 30. Annual rates cannot put those two lines on the same chart. Magazine-days can, and the comparison is exactly the one a portfolio decision needs.

Run The Arithmetic On A Real Package

Take a concrete case from the same news cycle. The Washington Post, citing a U.S. official, reported that the United States is preparing an approximately $2.8 billion sale of 2,000-pound bombs to Israel. The package would reportedly consist of 40,000 of the bombs, comprising 20,000 MK-84s and 20,000 BLU-117s. Forty thousand is a large number, and presented as an inventory count it sounds decisive. It says nothing about endurance until someone divides it.

I do not know the recipient's expenditure rate for 2,000-pound bombs, and I will not pretend to. The structure of the calculation does not need the classified number; it needs a stated one. At 200 weapons a day, the package is 200 magazine-days. At 1,000 a day, it is 40. Same inventory, a five-fold difference in what it buys, and the entire difference lives in an assumption the headline never states.

The delivery schedule matters just as much. Suppose, purely for illustration, those 40,000 weapons arrive evenly over four years: roughly 830 a month. Against the 1,000-a-day case, a full month of deliveries recovers less than one day of fighting, which means the line is not refilling the magazine in any sense that matters operationally.

An inventory count without an expenditure rate is a number wearing the costume of information.

The Denominator Moves, So State It

The usual objection is that expenditure rates are scenario-dependent, contested, and often classified. All true, and none of it argues for omitting them; it argues for publishing the assumption next to the claim so the assumption can be challenged. A line delivering 3,000 rounds a month recovers 30 days against a 100-round day and three days against a 1,000-round day, and both statements fit on one chart. Hiding the denominator does not make the arithmetic go away. It just moves the surprise to wartime.

That denominator also refuses to sit still. Russia sharply increased its use of jet-powered attack drones, forcing Ukraine to develop interceptors capable of engaging targets moving at up to 500 kilometers per hour, which means the effector a defender needs, and burns, changed underneath the planners. When the threat mix shifts, the consumption of a defended day shifts with it. A metric with the expenditure rate built into its unit is forced to notice. Rounds per year is not.

None of this demands new data. The delivery numbers and the expenditure estimates already exist wherever a scenario has been seriously costed. The metric does not create information; it forbids the existing information from being briefed separately, which is where the obfuscation lives.

Time Is The Dimension We Keep Paying For

The cost side falls out of the same division. More than $22 billion across roughly 150 days is an average burn north of $140 million a day, straight from the inspector general's own figure. Denominate spending in days and the exchange arithmetic surfaces immediately: what a defended day costs us versus what threatening it costs the other side. That question stays invisible as long as everything is quoted annually.

We already know how expensive it is to drop the time axis, because an adjacent domain has handed us the receipts. The Government Accountability Office found that taxpayers paid roughly $4.17 billion over a decade for Navy submarines to sit idle or wait through maintenance delays. That is not a procurement figure; it is money converted directly into days of unavailability. Any metric that drops the time dimension will eventually present its bill in that dimension.

There is a small counterexample worth studying. The U.K. deployed its first passive counter-drone detection system at a British Army base after two domestic companies delivered it in under six months. Project Halo Shield then logged 29 drone alerts during its first 17 days of live monitoring at Warminster Garrison. Notice the units: months to field, alerts per days of watching. Everything is denominated in time, which is precisely why every claim in that sentence can be checked.

What An Acquisition Review Should Demand

Every munitions capacity chart that reaches a review should carry three lines. First, the stated daily expenditure rate and the scenario it comes from. Second, actual monthly deliveries plotted across the ramp, not the objective rate presented as though it exists today. Third, the crossover date at which magazine-days recovered per month exceeds magazine-days consumed per month under that scenario. Below the crossover, higher nominal output is still losing; the magazine is draining more slowly, which is not the same thing as refilling.

Pair the inventory count with the recovery rate and most of the theater evaporates. A briefer claiming tripled output has to say against what expenditure rate, by when, and what the stockpile holds every month in between. Reviews should stop accepting capacity claims that omit the expenditure assumption and the time required to reach the claimed rate, because those omissions are not oversights. They are load-bearing.

I hold my own work to the same rule. Kibernan has produced six complete, costed engineering programs, including an AI Managed Missile Factory, and every published figure in them carries a maturity tag: 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 number without its maturity tag is exactly the kind of number this article argues against.

The independent watchdog has told us what the drawdown cost: more than $22 billion on munitions in five months. The inspector general has also told us the war with Iran left a resupply bottleneck that cannot keep pace. The industrial base will answer in annual rates, because annual rates are the unit its contracts are written in. The magazine does not care. It drains in days and refills in days, and until production is measured in the unit the fight is fought in, we will keep discovering shortfalls in audit reports instead of on planning charts.