Everyone knows the five-million-dollar missile shooting down the thirty-thousand-dollar drone. Almost nobody has run the other ledger — the one where a one-million-dollar Iranian missile destroyed a four-hundred-and-eighty-five-million-dollar American radar. There are four ledgers in this war. They point in four different directions. Here they are, side by side.
This is not a conspiracy theory. Every figure on this page comes from CSIS, the Congressional Budget Office, the Payne Institute, the Foundation for Defense of Democracies, JINSA, US Army budget documents, CENTCOM briefings, and open reporting. We have simply put the numbers next to each other — which, remarkably, almost nobody has done.
The viral version of this story is that America is bankrupting itself shooting down cheap drones with expensive missiles. That is true, and it is the least important thing that happened. The bigger story is what the cheap stuff killed on the other side of the board, what the fast boats were supposed to do and never got to, and the one constraint that no amount of money can fix.
We will also tell you plainly where the popular version of this story falls apart — because a page that only tells you what you already want to hear isn't information. It's flattery.
This page now links its evidence. Each numbered marker — like this [1] — jumps to the source in the Sources list: the Congressional Budget Office's own letter, CSIS's munitions reports, the Army Secretary's testimony, Defence Minister releases, FDD and the satellite-imagery investigations. Most of the page held up. Five things did not:
1. The aircraft hit on 27 March was a US E-3 Sentry, damaged — not a $700 million E-7A Wedgetail destroyed. The US does not yet fly the E-7A; the only Wedgetail in this war is Australia's, and there is no report of it being hit.[1,2] The "1 : 700" ratio has been withdrawn from the chart.
2. The radars. One US THAAD radar — at Muwaffaq Salti in Jordan — is confirmed destroyed. Three other radar sites were hit, two of them UAE-owned batteries, with damage unconfirmed. We had counted four destroyed US radars worth $2 billion.[3,4] The 1 : 485 ratio stands: it is one missile against one radar.
3. The interceptors. The CBO's figure is "between one-half and two-thirds," counted since June 2025 — so it includes the 2025 Twelve-Day War, not only this one.[5,6]
4. Japan's Tomahawks were delayed by up to two years, not "suspended indefinitely" — our own October update already said so; two older sections did not.[7]
5. Small fixes. Ted Lieu's line came before 3 June (that is the date of the op-ed answering him); the Pentagon's cost comparison is now shown as its actual figures; the supplemental was requested on 24 June; THAAD's pre-war stock was 360. And the Patriot price genuinely varies — $3.9 million in CSIS's figures, about $5.3 million in the Army's own budget — which is why the chart shows a range.[8,9]
None of this changes the argument. The magazine-depth problem is, if anything, better documented than when we wrote it.
The first independent price tag has landed. In September the Congressional Budget Office estimated the war had cost the Department of Defense $38.1 billion as of 1 August[5,10] — munitions, lost equipment (including a THAAD radar destroyed by Iran), extra flying hours and fuel. Defense Secretary Pete Hegseth told Congress in July the war would cost $37.5 billion through September; the Pentagon's inspector general put it at $33.4 billion through 29 June.[11,10] Replacing the munitions alone comes to $21.7 billion: $13.1 billion for interceptors, $7.3 billion for land-attack cruise missiles, and $1.2 billion for everything else. One detail matters more than the total: the CBO notes the Pentagon does not disclose how many interceptors it holds, so its analysts worked from public reporting of what was fired and what had been bought.[10,5] The independent budget office of Congress had to reconstruct the war's cost the same way we did.
The magazine number is worse than our July estimate. The CBO says the war has probably used between one-half and two-thirds of the combined US stock of Patriot, THAAD, SM-3 and SM-6 interceptors — counted since June 2025, so including the 2025 Twelve-Day War. It expects rebuilding to take at least five years.[5,6,12] CSIS's latest counts put Patriot use at 1,060–1,430 from a pre-war stock of 2,330 — up to 61% — and THAAD at 190–290 from 360, up to 81%. No new THAAD interceptors have been delivered since August 2023.[8] Over the past decade the US bought on average about 225 Patriots a year. The CBO's own word for what that means is an "opportunity cost": a reduced interceptor inventory for several years. The Secretary of Defense says the US has "virtually unlimited" mid- to high-grade munitions. The CBO, CSIS and the delivery schedules say otherwise. Believe the delivery schedules.
Production is rising — slowly. Raytheon says it built three times as many Tomahawks in the first half of 2026 as in the same period last year, and the Pentagon has awarded a contract worth nearly $23 billion over seven years. That is real. It is also measured in years, not months, and it serves the US Navy first.
And the queue just got longer — with Australia in it. In September the US warned allies to expect delays of up to five years on key weapons while it rebuilds its own stocks; Germany was told Tomahawk deliveries were unlikely inside that window, and Japan had already been told its 400 Tomahawks would be up to two years late.[13,14] Australia has an order for roughly 220.[15] Defence won't disclose quantities or delivery dates "for operational security reasons." Opposition Senator James Paterson has called announcing orders without delivery dates "performative procurement."[16] CSIS's Australia chair Charles Edel says missile deliveries to allies like Australia are at "a growing risk of delay."[16] This page said in July that our deterrent was built on someone else's magazine. That magazine is now two-thirds empty, and the refill goes to its owner first.
The war itself: major combat is paused but nothing is settled. The 60-day ceasefire lapsed on 17 August, Iran's seven-day plan to reopen the strait was rejected in late September, and October brought the worst run of attacks on shipping of the war — with two cargo ships sunk. US media report the Pentagon has prepared options for heavier strikes if ordered. Every one of those would draw on the same depleted magazines. Fuel side: see the updated Fuel Crisis dossier.
§ 01 — Ledger One
By early June, Congressman Ted Lieu had described American air defence in the Iran war as "throwing Ferraris at Frisbees."[17,18] It is the phrase that took the concept mainstream, and it describes something real. Defence economists call it the cost-exchange ratio: how much the defender must spend to neutralise each dollar the attacker spends.
In this war, that ratio is brutal — and Iran built its entire strategy around it. Fire waves of cheap drones. Force the defender to burn expensive interceptors. Then send the ballistic missiles once the magazine is thin.
The strategy worked exactly as designed. Iran spent roughly $70 million on 2,000 drones and forced adversaries to expend more than $2 billion in interceptors answering them. The Payne Institute counted 943 Patriot interceptors fired in the first four days — equivalent to about eighteen months of combined Lockheed Martin and Boeing output, burned in a long weekend.
§ 02 — Ledger Two
Here is the part that has almost no coverage, and it is the single most important finding on this page.
Iran did not try to saturate American air defences. It went after the eyes.
The AN/TPY-2 is the X-band radar that makes a THAAD battery function. Without it, the launchers are inert tubes. Iranian strikes hit four THAAD radar sites — near Prince Sultan in Saudi Arabia, at Al Ruwais and Al Sader in the UAE (both UAE-owned batteries), and at Muwaffaq Salti in Jordan, where the US radar was destroyed: the one the CBO counts. The others were damaged to an unknown degree. Each AN/TPY-2 costs about half a billion dollars.[3,4,19,1] On day one, an AN/FPS-132 early-warning radar at Al-Udeid in Qatar, valued at $1.1 billion, was also hit.
Now run the arithmetic that nobody ran.
The United States entered this war with three THAAD batteries available to defend the Gulf. Iran did not have to shoot down the interceptors. It just had to blind them.
And it cuts both ways. The US fired 535 Tomahawks in the first sixteen days — about $1.9 billion worth — many of them at mobile launchers and small naval craft worth a small fraction of that. CSIS put it bluntly in its own analysis: a Tomahawk costs about $3.6 million; a JDAM doing the same job on the same target costs $80,000. By day four the Pentagon announced a "point of munitions transition" and started dropping the cheap bombs instead.
Everybody in this war was throwing Ferraris. The Americans were just throwing more of them, faster, and at cheaper Frisbees.
| US / Allied Asset Destroyed | Est. Value | Killed By | Est. Cost of the Weapon That Killed It |
|---|---|---|---|
| 1 × AN/TPY-2 THAAD radar destroyed Muwaffaq Salti, Jordan — plus 3 more sites hit (Saudi Arabia; UAE ×2, UAE-owned) | ~$0.5B destroyed[3,19] | Iranian ballistic missiles / cluster submunitions | ~$1M each |
| AN/FPS-132 early-warning radar Al-Udeid AB, Qatar | $1.1B | Iranian missile, day one | ~$1M |
| E-3 Sentry radar aircraft & tankers damaged Prince Sultan AB, 27 Mar (corrected — not an E-7A) | Not scored — damaged, not confirmed lost[1] | Iranian strike | ~$1M |
| 24 × MQ-9 Reaper drones | ~$720M | Iranian surface-to-air missiles | Low hundreds of thousands each |
| 4 × F-15E Strike Eagle (3 lost to friendly fire, Kuwait) | ~$400M | Kuwaiti air defences — not Iran | — |
| 42 US aircraft total lost or damaged Per CRS report IN12692, 13 May 2026 | Anadolu put US asset losses at $3.84 billion in the first two weeks. CSIS's attrition estimate is $2.3–2.8B. CBS, citing internal US officials, puts full asset losses nearer $11.9B once bases and specialist equipment are counted. The Pentagon comptroller attributed $24B of the $29B total war cost to equipment repair and replacement. | ||
§ 03 — Ledger Three
This is the ledger that should worry you most, because it is the one that didn't happen — and everyone had known for twenty-four years that it could.
Millennium Challenge 2002. The most expensive war game in modern American history. Marine Lieutenant General Paul Van Riper was handed the Red Team and told to play an Iran-like adversary in the Gulf. He did not play by the script. He used motorcycle couriers and calls to prayer from minarets to route his orders around American electronic warfare. Then he launched a saturation cruise-missile barrage, timed with a swarm of civilian speedboats packed with explosives.
The Aegis system was overwhelmed. When the umpires stopped counting, sixteen to nineteen US ships were on the bottom, including an aircraft carrier, and around 20,000 notional American servicemen were dead. It would have been the worst day in the history of the United States Navy.
The exercise was reset. The Red Team's free play was constrained. Blue won. Van Riper submitted a detailed critique and, by his own account, received no substantive response.
The IRGC Navy is not a navy in the sense the Royal Australian Navy is a navy. It is a denial force, purpose-built for the 33-kilometre-wide killing ground of the Strait of Hormuz — where there are only two navigable shipping lanes and nowhere for a capital ship to run.
| Platform | What It Is | Armament | Why It Matters |
|---|---|---|---|
| Peykaap III / Zolfaghar | 17m, ~14 tonnes, crew of 3. Planing hull derived from a British racing-boat design. 50+ knots. | 2 × Nasr-1 or Kowsar anti-ship missiles (35–38km range), 2 × 324mm torpedo tubes, 12.7mm guns. | A three-man boat the size of a large tinnie, carrying missiles that can cripple a merchant tanker. |
| Houdong / Thondar class | Chinese-derived missile boats. ~10 in service. | 4 × C-802 anti-ship missiles each. | The heavy end of the swarm. Outranges most close-in defences. |
| Seraj / Ashura / Ya Mehdi | Small open-cabin craft, Boston-Whaler pattern. Built in hundreds. | 107mm rocket launchers, RPGs, heavy machine guns — or nothing but explosives and a driver. | These are the saturation layer. They don't need to be good. They need to be numerous. |
| Ghadir-class midget submarines | 20+ in service. Shallow-water designed. | Torpedoes and naval mines. | Iran holds the fourth-largest naval mine inventory in the world. Mines don't care how expensive your destroyer is. |
| Sea mines & USVs | Laid by boats, dhows, fishing vessels, submarines. | — | The single cheapest way to close a strait. Demining it afterwards is expensive, slow and dangerous. |
Reporting in the opening days of the war put hundreds of missile-equipped fast boats dispersed across sixteen positions around Larak Island alone — pre-positioned, exactly as the doctrine says, to converge from multiple directions at once.
Because the United States did not fight the swarm. It killed the boats at the pier.
Iranian naval assets were listed as a top-priority target set inside the first 48 hours. Tomahawks launched from destroyers hit the naval bases at Bandar Abbas, Konarak and Chabahar before most of the fleet ever left the wall. Satellite imagery from Planet and Vantor showed fires, sunken vessels and flattened buildings. By mid-March, CENTCOM's Admiral Brad Cooper reported more than 60 Iranian vessels destroyed or disabled. By late March, the Gulf International Forum's count was more than 120 vessels destroyed or incapacitated — from a combined IRIN/IRGCN fleet of roughly 250–300 hulls of all sizes.
Among the losses: 16 minelaying vessels destroyed near the Strait of Hormuz and 10 more killed in port. The drone carrier IRIS Shahid Bagheri, in service barely a year. The IRIS Makran, a converted tanker serving as a floating forward base. And on 4 March, the Moudge-class frigate IRIS Dena was sunk by a US nuclear submarine using a Mark 48 torpedo — the first combat sinking of a major warship by an American submarine since the Second World War.
But here is the catch, and it is a big one. Iran still retains hundreds of small fast patrol craft, coastal anti-ship missile batteries, sea mines, unmanned surface vessels and shore-based radar. Those assets are dispersed, cheap, and impossible to comprehensively destroy. And when the Strait of Hormuz was re-closed on 20 June — three days after a signed peace deal — that is precisely the force doing the closing.
The fleet is gone. The denial capability is not. Which is why our fuel is still not safe. See the Fuel Crisis dossier.
§ 04 — Ledger Four
We are going to be straight with you, because the alternative is propaganda with better graphics.
If you measure the whole war rather than the individual shot, the United States won the cost exchange decisively. Not marginally. Overwhelmingly.
Iran's losses. The Foundation for Defense of Democracies estimates total economic damage to Iran at approximately $144 billion — roughly 40% of pre-war GDP (range: $50B–$300B).[20,21] Of that, replacement cost for military and strategic assets alone is about $46 billion — equivalent to four to six years of Iran's entire pre-war defence budget. Hydrocarbon revenue losses account for another $53 billion.
America's losses. The Pentagon comptroller placed direct war costs at $29 billion as of 12 May. Independent analysis (Stephen Semler, Popular Information, 8 July) puts total direct US costs at $103 billion over 120 days. Take the highest credible figure. Iran still lost more, relative to what it had, by an enormous margin.
In per-capita terms it's not close. JINSA's analysis of the earlier twelve-day war found the conflict cost Iran 93 times more than Israel and 371 times more than the United States measured against GDP per capita. Interceptors, JINSA estimated, saved Israel $13.4–13.5 billion in property damage that would otherwise have landed.
And the strategic argument against the ratio is a serious one. The case made by air-defence advocates is this: an interceptor is not priced against the thing it shoots down. It is priced against the thing it protects. A PAC-3 that stops a missile short of a Gulf port is not trading $4 million for a $50,000 drone — it is trading $4 million for a barracks, a refinery, a port, and the people in them. In January 2024 a single one-way drone got through at Tower 22 in Jordan and three American soldiers died. In 1991, one Scud that got through killed 28 in a single strike. Across hundreds of Iranian air attacks in 2026, US forces suffered three casualties. You do not get to call that a failure of cost efficiency.
So: the 175:1 ratio is real, and it is a per-engagement number wearing a strategic disguise. Anyone selling you "America is bankrupting itself" as the whole story is selling you a third of it.
§ 05 — The Real Constraint
Here is where all four ledgers converge, and where the actual national-security story lives.
Cost was never the binding constraint. Cost is the easy part. The United States can appropriate $87.6 billion in a supplemental — it requested exactly that on 24 June.[5] What it cannot do is manufacture a solid rocket motor faster than a solid rocket motor can be manufactured.
THAAD. Lockheed Martin builds 96 a year. The framework agreement to lift that to 400 was signed in January 2026 — and capacity expansion takes 24 to 36 months before the first extra missile exists. CSIS's Wes Rumbaugh found that no THAAD interceptors were delivered into US inventory at all between July 2023 and 2026. The war fired somewhere between half and four-fifths of the stockpile — CSIS's later count puts it at up to 82%.
Tomahawk. Average procurement over the last decade: 86 missiles a year. Over 1,000 fired in 40 days — more than ten times the annual buy rate. The Navy has requested 785 in the FY2027 budget. Based on DoD's own delivery projections, those missiles start arriving in March 2030, after 34 months of production lead time.
Patriot. 1,060–1,430 fired — up to 62% of a pre-war stock of about 2,300, against average purchases of about 225 a year. Lockheed's seven-year agreement to reach 2,000 PAC-3s per year is real, but it is a seven-year agreement, and 17 other countries plus Ukraine are in the same queue.
The bottlenecks are not financial. They are solid rocket motors, guidance electronics, testing capacity, qualification cycles — and gallium and germanium, critical minerals whose supply China largely controls. You cannot appropriate your way past a mineral you don't have.
As Stimson Center senior fellow Kelly Grieco put it: "magic is still not one of its capabilities." CSIS's Mark Cancian: "it will be three or four years before the additional production arrives."[22]
§ 06 — The Pattern
Compare the podium with the ledger. Both columns are sourced. Neither column is lying. They are just describing different things.
§ 07 — The Australian Ledger
Australia's military contribution to this war was modest and deliberately so. Eighty-five ADF personnel, a Boeing E-7A Wedgetail and a stock of AMRAAM air-to-air missiles, sent to the UAE on 10 March at the UAE's request.[2,23] Three ADF personnel were aboard the US submarine[24] that sank the IRIS Dena. On 16 March, Australia declined a US request to send a warship to the naval coalition protecting shipping through the Strait of Hormuz — partly, per reporting at the time, because of questions about which vessel could actually have gone.
We stayed out. And the bill arrived anyway.
§ 08 — The Honest Part
1. Nobody outside the Pentagon knows what was actually fired. CSIS notes explicitly that in previous Middle East campaigns the Department of Defense issued daily strike updates and regular statistical compendia. It has not done so for this war — and when the Congressional Budget Office asked for information to cost it, the Pentagon did not respond. Every expenditure figure on this page — CSIS, Payne Institute, all of it — is a reconstruction from CENTCOM fact sheets, order-of-battle estimates and historical employment patterns. They are the best figures available. They are not disclosures.
2. The unit costs genuinely disagree, and the disagreement is itself the story. THAAD is quoted at $12M, $15M and $15.5M. SM-3 Block IIA appears as $28.7M in CSIS's figures and $36M in secondary trackers. Patriot PAC-3 MSE runs from $3.7M to the $5.3M in the Army's own FY2027 request. This is not noise. THAAD cost $9.5 million in 2021 and $15.5 million now. SM-3 IB went from about $9 million in FY21 to nearly $24 million in the FY24 supplemental buy. Erratic, stop-start funding is inflating the price of the exact weapon whose price we are complaining about. The cost-exchange ratio is partly a self-inflicted wound.
3. Battle-damage claims are the Pentagon's own, pre-assessment. The 13,000 targets, the 80% of air defences, the 90% of weapons factories — those figures came from a Pentagon podium before comprehensive combat assessments were completed. Analysts asked at the time said the numbers seemed feasible but imperfect. We record them as claims, because that is what they are.
4. Iranian fast-boat unit costs are not published anywhere credible. We have not invented one. What is documented: a Peykaap III displaces about 14 tonnes, is 17 metres long, carries three crew, and is built on a modified racing-boat hull using domestically produced engines because sanctions cut Iran off from German marine diesels. Whatever it costs, it is not in the same universe as a destroyer. That is the entire point of it, and it is the reason a precise figure was never the argument.
5. We are not drawing a conclusion the record doesn't support. The cost-exchange ratio is a real, mainstream concept in defence economics — discussed openly by CSIS, by the US Army's own Low-Cost Interceptor program, and on the floor of Congress. It is not a secret. What is genuinely under-reported is the reverse ratio in § 02, and the magazine-depth problem in § 05. We record the pattern and the question. We go no further than the documents do.
§ 09 — The Fix
The most useful thing on this page is not an accusation. It is a purchase order.
Within eight days of the war starting, the US Army bought 13,000 Merops interceptor drones at roughly $15,000 each[26] — a three-foot fixed-wing aircraft with a two-kilogram fragmentation warhead, from a startup founded by a former Google chief executive, which had already shot down more than 4,000 Russian drones over Ukraine before any American service branch placed an order.
On 23 June, the Army formally launched its Low-Cost Interceptor program[27,9], targeting sub-$1 million kinetic interceptors, government-owned intellectual property, and a live-fire test at White Sands in Q4 2026. The Advanced Precision Kill Weapon System II — a guidance kit bolted onto the ubiquitous Hydra-70 rocket — can knock down a drone or a cheap cruise missile for under $35,000.
When an institution tells you a problem isn't serious, and then spends real money, in a hurry, on the specific fix for that problem — believe the money.
The Pentagon spent twenty-four years not answering Van Riper. It spent eight days buying 13,000 cheap drones. Watch which numbers move, and which ones just get pushed further out.
This is not a call to panic and it is not a call to arms. It is a call to read a balance sheet properly — and to understand that the two things most exposed in this conflict, for Australia, are the two things we import: our fuel, and our firepower.
◆ RELATED FILES ◆
◆ FIND US ◆ FOLLOW US ◆
Numbered in the order they first appear. Official budget, defence and research sources are linked wherever they exist. Checked 24 October 2026.
Still being linked: the Payne Institute's count of 943 Patriots in four days, Iran's ~$70 million drone spend, JINSA's per-capita ratios, Stephen Semler's $103 billion estimate, the Pentagon comptroller's $29 billion (12 May), CRS report IN12692 (42 aircraft), the F-15E friendly-fire losses, the IRIS Dena sinking details, Raytheon's Tomahawk output and contracts, the $35.3B THAAD contract, Hegseth's "virtually unlimited" remark, the Grieco quote, and the Millennium Challenge 2002 details.