Who broke the deal first?
The story you’ve heard is that Iran cheated, so the deal had to go. The inspection record — from the agency whose job it was to catch Iran cheating — says the opposite.
From January 2016, the International Atomic Energy Agency verified Iran’s compliance with the JCPOA in report after report. In April 2018, the U.S. State Department’s own assessment said Iran was “transparently, verifiably, and fully implementing the JCPOA.” Three weeks later, on May 8, 2018, the United States withdrew from it anyway.
Iran remained within the deal’s limits for a full year after the U.S. withdrawal. Its first breaches came in July 2019 — announced in advance, verified openly, and framed as remedial steps under the deal’s own Paragraph 26, which permits scaling back when the other side stops performing.
— Arms Control Association compliance record; Congressional Research Service R40094: "Beginning in July 2019, the IAEA verified that some of Iran's nuclear activities were exceeding JCPOA-mandated limits." [[1]](#where-these-numbers-come-from)[[2]](#where-these-numbers-come-from)The violations did not cause the withdrawal. The withdrawal caused the violations.
The honest footnote. Iran’s record wasn’t spotless: twice in 2016 its heavy-water stock crept marginally over the 130-tonne mark (~130.9t). Both overages were cured within days by shipping the excess to Oman, and neither drew an IAEA non-compliance finding. If someone tells you Iran “never violated anything,” that’s oversold. If someone tells you these rounding-error episodes justified abandoning the deal, that’s oversold far worse. [2]
The breakout clock
Arms controllers measure one thing above all: breakout time — how long Iran would need to produce enough weapons-grade uranium for one bomb. The deal’s whole point was to keep that clock long. Watch what happened to it.
breakout time under the JCPOA — the U.S. government’s own assessment, reaffirmed by DNI Coats, Jan 2019
by June 2020, two years after withdrawal
by late 2024 — enough weapons-grade uranium for 5–6 bombs
Independent hawkish analysts (ISIS) put the deal-era worst case at ~8 months — shorter than 12, still 16× longer than two weeks. [1][3]
The stockpile that became the war’s excuse
Under the deal, Iran could hold 300 kilograms of uranium enriched to 3.67% — reactor fuel. By the eve of the June 2025 Israeli strikes — the 12-day campaign that preceded the U.S.–Israeli war of 2026 by eight months, and the last moment inspectors could count anything — the IAEA’s verified tally showed what eight years of “maximum pressure” had produced:
total enriched uranium — 33× the deal’s cap (IAEA, verified as of June 13, 2025)
enriched to 60% — a short technical step from weapons-grade. No other non-weapon state has ever made it
advanced centrifuges installed by Feb 2025, each 4× the output of the deal-era machines
Figures from IAEA Board report GOV/2026/8, ¶33 & ¶53, checked against the primary document. [4]
Every step of that escalation — 60% enrichment, suspending extra inspections, the six-tonne stockpile — happened after the U.S. left the deal, and was announced by Iran as a response to it.
Concede the steelman — it’s real. By 2025, Iran was genuinely a worse actor than in 2018: it stopped applying the Additional Protocol in Feb 2021, and on June 12, 2025 — one day before Israeli strikes began — the IAEA Board formally found Iran in non-compliance with its safeguards agreement, the first such finding in ~20 years. (Scope matters: that finding concerned a probe into decades-old undeclared material, not the JCPOA’s operating limits — and the extra inspections Iran suspended were ones the deal, which the U.S. had already left, made voluntary.) Also true: the deal had sunset clauses, never covered missiles, and by 2025 even restoring its full limits could only buy back a 4–5 month breakout, because centrifuge know-how is irreversible. The deal was imperfect and partly eroded. The question is whether the fix was worth what it cost. [4][5]
109 days that hit the whole world’s wallet
On February 28, 2026, U.S. and Israeli aircraft opened a war on Iran — aims including the nuclear program and regime change. Supreme Leader Ali Khamenei was killed. Iran retaliated across the region and shut the Strait of Hormuz, the channel that carries a fifth of the world’s oil.
collapse in Hormuz tanker transits in the first week; ~1,000 tankers stuck in the region
the size of the 1973 and 1990 oil shocks — “the largest geopolitical oil supply disruption in history” (Dallas Fed)
WTI crude, late January to the March average — with $110–$167 scenarios if the closure dragged on
Dallas Fed working paper 2609 and IEA oil market reports; shipping data via Lloyd’s List/Kpler. [6][7][12]
The underwriters closed the strait first
Here is the part the “Iran shut Hormuz” shorthand gets backwards: the strait was commercially closed before Iran laid a mine or hit a tanker. Within 48 hours of the February 28 strikes — two days before Iran’s March 2 closure order — war-risk premiums surged fivefold, major marine insurers terminated existing coverage and offered replacements at roughly sixty times pre-crisis rates, Lloyd’s Joint War Committee redesignated the entire Arabian Gulf a conflict zone, and tanker traffic collapsed by more than 80%. As one analysis of the mechanism put it: the commercial shutdown preceded the physical blockade — insurance closed the strait before Iran’s navy did. The −92% figure above is the end of week one, after the closure order deepened it. [34]
The mechanism never let go. By July, Marsh’s global head of marine, cargo and logistics, Marcus Baker, reported war-risk premiums had jumped from 1–3% of hull value to 7.5–10%, with underwriters increasingly reluctant to offer spot terms at all. Do that arithmetic on a $100 million tanker: a passage that cost about $250,000 to insure before the war (~0.25% of hull value) now costs $3–10 million — per voyage. At least nine ships came under attack in the strait in the two weeks after July 6 (IMO data), roughly 6,000 seafarers remained trapped in the region on the IMO’s July count, and the EIA recorded Middle East–to–Asia supertanker rates at their highest since its records began in late 2005. Whoever ends the shooting, the strait reopens only when the underwriters say it does — as Crisis Group’s Chris Newton put it, “There is no military solution to the strait and the wider war.” You can escort a tanker. You cannot shoot down an insurance premium. [34]
Honest footnote. The 48-hour sequence — the fivefold surge, the sixty-times replacement rates, the >80% collapse before any Iranian attack — traces to a single detailed analysis (Dr. John Hatzadony, irregularwarfare.org, March 2026); the direction is corroborated by market data, but those specific numbers are single-source. The July premium figures are the broker’s own (Marsh), and the seafarer and attack counts are IMO tallies in a war zone where every number lags. [34]
It ended — tentatively — on June 17, 2026, with a 14-point U.S.–Iran memorandum signed in stages via Islamabad: military operations halt, the U.S. lifts its naval blockade, Iran reopens Hormuz, and the hard questions go to a 60-day negotiation. As of this writing the ceasefire has already needed one repair, and gas prices are rising again on doubts it holds.
Flag: the fine print isn’t public. No official text of the memorandum has been released. Every published term comes from a U.S.-briefed account that Iran has not confirmed. And it is an agreement to negotiate — sanctions, inspections, and the fate of the uranium are all still open. [8]
What your household paid for it
This war was not free for Americans, and the costs didn’t arrive as a tax. Like the shadow tax, they arrived as prices.
national average gas price: pre-war baseline to the May 21 peak — a 2026 record (AAA)
CPI inflation, year over year, by May — a multiyear high; energy alone drove 40%+ of April’s jump
war cost per typical U.S. household so far — Moody’s Analytics’ “conservative” estimate
Mark Zandi, Moody’s Analytics’ chief economist, itemized that $1,000: roughly $300 in gasoline, $200 in diesel-driven grocery and delivery costs, $250 in military spending per household, $150 in interest-rate impacts, $100 in airfares. His own verdict on the number: “The true cost is likely higher — meaningfully higher. It’s fair to ask whether it was worth it.”
The Federal Reserve stopped hedging, too. Its June 2026 statement attributed elevated inflation to the war’s energy shock by name, raised its 2026 inflation projection from 2.7% to 3.6%, and cut its growth forecast. The taxpayer side keeps climbing: Pentagon and OMB officials testified the operation cost ~$25 billion by late April and ~$30 billion by June 30, and on July 21 Defense Secretary Pete Hegseth put it at $37.5 billion, with a $67.1 billion defense supplemental requested on top — while NBC News reported an internal Pentagon figure of $80–100 billion once damaged-base repairs are counted. [24]
Honest accounting. Not every cent of 2026 inflation is the war — shelter costs were rising on their own, and the Fed’s models say a one-quarter Hormuz closure adds “only” 0.35–0.6 points to inflation (up to ~1.8 in worst-case scenarios). The per-household figure is one firm’s estimate, not a government statistic. The defensible claim is narrower and still damning: a war of choice measurably raised the price of gas, groceries, and money for every American household, and its architects concede the bill is still running. [9][10][11]
What everyone else paid
Wars are pitched in strategy and paid in lives and livelihoods — overwhelmingly by people who had no vote in any of it.
| Who | What the war cost them |
|---|---|
| Iranian civilians | 3,636 documented deaths in the first six weeks — 1,701 of them civilians, including 160+ children in one school strike (HRANA; the monitor stresses these are minimums). Belligerent claims run far higher. Hospitals and schools hit; nationwide internet blackout. |
| Iran’s economy | IMF: GDP −6.1% in 2026 with ~69–89% inflation; food inflation over 100%; the rial hit record lows and the central bank printed its largest banknote ever. Infrastructure damage estimates run $200B+ (single-firm figure — treat with caution). |
| U.S. allies in the Gulf | Iranian retaliation struck Saudi, Qatari, Emirati and Kuwaiti soil (55 killed). Qatar’s Ras Laffan LNG complex lost 17% of export capacity — a 3–5 year repair. IMF April projections: Qatar −14.7%, Kuwait −4.2%, UAE −1.9%. GCC damage estimates ≈ $200B. |
| Asia | 80% of Hormuz oil goes to Asia. IMF cut the region’s 2026 growth from 5.0% to 4.4%; China slowed to 4.6%; India’s inflation projected to more than double. Delivered crude into China briefly cost ~$100/bbl with freight. |
| The world | IMF, April 2026, in a report literally titled “Global Economy in the Shadow of War”: global growth cut from 3.4% to 3.1%, global inflation revised up to 4.7% — hardest on poor, oil-importing countries. |
Sources: HRANA (18 May 2026), IMF WEO April 2026 + July update, World Bank Iran monitor, ADB April 2026, Stimson Center, UN OCHA. Casualty and damage figures inside Iran and party claims on all sides carry real uncertainty — see the flags. [13][14][15][16][17]
What the war did to the security guarantee
The table above counts what the Gulf paid in money. The more durable cost may be strategic: the states whose protection is the entire premise of America’s presence in the Gulf spent 2026 learning what that protection is worth when the shooting actually starts.
Gulf governments given advance notice of the February 28 attack — officials say their warnings that the war would have “devastating consequences for the entire region” were ignored (AP)
Iranian missiles (380+) and drones (1,480+) fired at the five Arab Gulf states in the war’s opening weeks (AP tally, early March)
of all Iranian attacks on the Gulf states were directed at the UAE — “the most attacks of any GCC state” (Congressional Research Service)
By March 6, Gulf officials were telling the Associated Press that Washington had launched the war without telling them and over their objections. One official said the operation appeared “focused on defending Israel and American troops, while leaving Gulf countries to protect themselves” — and that his country’s stock of air-defense interceptors was “rapidly depleting.” The Congressional Research Service later counted thousands of attacks on facilities hosting U.S. forces and on critical infrastructure across the GCC. The bases that were supposed to deter the war became its targets: as the Quincy Institute’s Annelle Sheline put it, the U.S. military presence “had the opposite of a deterrent effect. These military bases became targets.” [32]
By June, the verdict was being said out loud. Firas Maksad, the Eurasia Group’s managing director for the Middle East and North Africa: “(The idea that) America as a strategic ally that can be relied upon is now very much in question in the Gulf states.” CNN’s analysts framed the war as the capstone of years of accumulating Gulf doubts about U.S. security guarantees — that part is their reading, not his quote, and it’s flagged as such. The through-line of this piece is that the war failed to buy nonproliferation; this section’s addition is that it also spent something — the credibility of the guarantee that underwrote fifty years of Gulf security architecture. [32]
Concede the steelman. The simple reading — wounded allies drifting away from Washington — is not what the reporting shows, and the complications cut hard. First: by late March, Saudi Arabia and the UAE were privately urging Trump to continue the war, “arguing that Tehran hasn’t been weakened enough” (AP, via anonymous U.S., Gulf and Israeli officials — flag the sourcing). The Gulf complaint is not that America was too aggressive; it’s that it started a war badly and won’t finish it. Second: these were not passive victims — CRS reports both Saudi Arabia and the UAE “reportedly” conducted their own strikes on Iran, and the UAE “reportedly” welcomed direct Israeli defense aid, including an Iron Dome battery operated by Israeli soldiers. Third: the hedging has limits. Anna Jacobs Khalaf of the Arab Gulf States Institute, on the region’s new partnerships: “This doesn’t mean replacing the US… it means diversifying partnerships.” Diversification is not divorce. But her own bottom line keeps the sting: the war “is making some Gulf states question the value of the US as a security guarantor.” [33]
Flags on this section: the no-notice and keep-fighting accounts are single-outlet AP reporting from anonymous officials (the latter including Israeli officials — a belligerent party with an interest in continuation); the White House’s counterclaim that Iranian retaliation has “decreased by 90%” is a party claim. CRS R48971 is the strongest neutral source, and it hedges the Gulf-strikes items with “reportedly.” [32][33]
So what did the war actually buy?
Be fair to the case for force — and be precise about which force. Two separate campaigns blur together in most coverage: the Israeli strikes of June 2025 (the 12-day war) and the U.S.–Israeli war that began February 28, 2026. The capability destruction below is the 2025 campaign’s work, assessed before the 2026 war closed every window. Then look at what neither campaign touched — or could.
| Bought | Not bought |
|---|---|
| ~22,000 installed centrifuges destroyed or disabled by the June 2025 Israeli campaign — no visible route to weapons-grade uranium at Iran’s declared plants for the first time in 20 years (ISIS assessment) | The uranium. All 440.9 kg of 60% material is unlocated and unverified — “most,” but less than 70%, believed buried in the Esfahan tunnels. Its recovery was deferred to negotiations. |
| A safeguards violator forcibly stripped of declared enrichment capacity | The watchdog. The IAEA “cannot provide any information on the current size, composition or whereabouts of the stockpile… or whether Iran has suspended all enrichment” — its words. After Feb 28, 2026, inspections in Iran effectively ceased. |
| Leverage: Iran signed a memorandum reaffirming it will not build nuclear weapons | Certainty: possible undeclared sites (Pickaxe Mountain, the Natanz tunnels) were never inspected — or hit. The world now knows less about Iran’s program than it did under the deal. |
Left column: ISIS (technically rigorous, hawkish-leaning — this is their case, credited honestly); the centrifuge assessment concerns the June 2025 Israeli strikes, which also ended IAEA site access — the 2026 war re-struck Natanz, Fordow and Esfahan, but with no inspectors in the country its marginal effect on the program is unverifiable. Right column: IAEA GOV/2026/8 ¶35, verified against the primary document. [4][5]
Here is the trade in one sentence: a measured, inspected nuclear program with a 12-month fuse became — via withdrawal, escalation, and war — an unlocated stockpile with no inspectors at all, at the price of the largest oil shock in history.
Diplomacy wasn’t exhausted. It was interrupted.
The last defense of the war is that talking had failed. The IAEA’s own board report — dated the day before the bombs fell — records the opposite.
The Director General attended bilateral negotiations between the United States and Iran on February 17 and 26, 2026. “Technical discussions will take place in Vienna in the week beginning 2 March 2026.”
The war began February 28. The Vienna talks never happened. [\[4\]](#where-these-numbers-come-from)And where did 109 days of war, thousands of deaths, and a global oil shock deliver us? To a negotiation — the 60-day talks now underway, covering enrichment limits, inspections, sanctions relief, and the disposal of enriched uranium. That is the JCPOA’s agenda, renegotiated from scratch: with no verification baseline, no inspectors on the ground, a new Supreme Leader installed by the war, and both sides’ maximalists dug in.
The counterfactual doesn’t require imagination. It requires arithmetic. Keeping the deal held breakout at ~12 months, with daily inspections, for the cost of sanctions relief. Follow-on diplomacy — the “longer and stronger” agreement everyone from the E3 to two U.S. administrations claimed to want — was available every year of the last eight, and was literally scheduled for the week the war began.
Three questions for the next war pitch
1. What does the inspector say? Not the politician — the agency with cameras inside the country. In 2018 it said Iran was complying. It was right.
2. What happens to the clock? Every coercive move since 2018 — withdrawal, pressure, strikes — made the breakout clock shorter, not longer. Demand the before-and-after numbers.
3. What does the war end with? If the answer is “a negotiated agreement,” ask what the same agreement would have cost before the shooting started.
Tearing up a working deal didn’t stop the bomb — it started the clock. The war didn’t secure the uranium — it lost track of it. And the peace, if it holds, will be a deal again. The only parts of this story that can never be negotiated back are the dead, and the eight years.
The hardware and the headcount
A war is also an inventory. Beyond the household bill and the body count, here is what 100-plus days of strikes and counter-strikes have physically destroyed — and who has died operating it. Read the direction of travel as solid and the exact figures as soft: equipment losses are the most propaganda-prone numbers in any war, so every count below is either a documented floor or an attributed estimate.
U.S. and coalition aircraft lost or written off — 42 in the first 40 days alone (congressional list), the majority of them MQ-9 Reaper drones
U.S. aerial equipment destroyed, per CSIS — a figure that pointedly excludes base damage, ground kit, and naval assets
U.S. service members killed and ~687 wounded war-to-date (Pentagon casualty system, Aug 4) — the wounded count jumped by more than 200 in two weeks, after the Pentagon quietly split its books into a second category
| Category | U.S. / coalition side | Iranian side |
|---|---|---|
| Radars & interceptors | At least one THAAD missile-defense radar (Mar 1; $485M–$970M) plus an E-3 AWACS radar plane (Mar 27, Prince Sultan air base; ~$700M); Iran’s late-July drone waves targeted U.S. radar and air-defense systems in Kuwait and Bahrain, damage unassessed. The defensive magazine is the bigger loss: CSIS estimates Patriot interceptors have fallen from ~2,300–2,500 pre-war to 759–827 (about two-thirds expended) and THAAD interceptors from 452 to ~234–278 | Air-defense and coastal-radar sites struck repeatedly — through the eleventh straight night of July strikes on air and coastal defenses; a firm count is not independently verified. Reuters reports Iran may receive up to 400 Chinese air-defense launchers routed via Pakistan — a claim neither Tehran nor Beijing has confirmed |
| Planes | ~70 aircraft lost or written off — MQ-9 Reapers now 36, after another was downed over Babylon Governorate, Iraq (Jul 26); CSIS values the aerial losses alone at $2.3–2.8B | ~43 military aircraft, overwhelmingly obsolete legacy airframes (J-7, F-14, F-4, F-5, C-130, Il-76), plus ~22 Israeli surveillance drones downed by Iran; CENTCOM says the July strikes hit aircraft hangars and drone-storage sites, while Iran’s acting defense minister claims drone production is now triple its pre-war rate |
| Bases | Struck across the Gulf and beyond: Prince Sultan (Saudi Arabia), Al-Udeid (Qatar), Ali Al-Salem (Kuwait), the 5th Fleet’s base and Sheikh Isa (Bahrain), sites in the UAE, Al Azraq and Muwaffaq Salti (Jordan), and Erbil (Iraq); a ballistic-missile salvo at U.S. forces in Jordan (Jul 28) was intercepted in full, per CENTCOM | Air- and coastal-defense bases, missile and drone-storage depots, the port of Bandar Abbas, and the Natanz / Fordow / Esfahan nuclear complexes; U.S. and Saudi forces also struck Iran-backed militias in Iraq (Jul 29) |
| Troops | 18 U.S. service members killed and ~687 wounded (Pentagon’s Defense Casualty Analysis System, Aug 4); independent trackers put the wounded as high as ~900 once casualties the database omits are added; wider coalition military dead ~77, including Israeli soldiers and UN peacekeepers | Documented military dead a floor near 1,220 (HRANA, first six weeks); monitors and belligerents claim 1,800 to 7,650. The Jul 29 strikes in Iraq killed ~20 Popular Mobilization Forces fighters and several IRGC personnel |
Sources: the congressional list of U.S. aircraft lost and OSINT aviation trackers; CSIS equipment valuations via Al Jazeera; a casualty aggregation cross-checked to HRANA, Hengaw, CENTCOM and Gulf-state statements, with the U.S. figures taken from the Pentagon’s own casualty system; interceptor inventories are CSIS estimates, which U.S. officials dispute. [22][23][24][29][30]
Why these numbers wobble — and why they still matter. Equipment kills are the most-inflated statistic in any war: each side advertises the other’s losses and buries its own. The U.S. government restricted satellite imagery of its regional bases, so even CSIS calls base damage the hardest thing to assess — its $2.3–2.8B is aerial losses only. Iran’s materiel losses are the least verifiable of all, pieced together from open-source spotting and adversary claims. Treat the table as a running estimate, not a final invoice. The defensible point survives the uncertainty, and it cuts against the pitch: this has been heavy, sustained hardware and personnel attrition for the United States — dozens of aircraft and billions in kit — not the clean, one-sided operation the war was sold as.
Two things got harder to count in the last two weeks, and both are worth saying plainly. First, on July 26 the Pentagon quietly opened a second casualty category — “Overseas Operations,” backdated to July 7, the date the administration says a new phase began and the War Powers clock reset. Combine the two books and the toll is 18 dead and roughly 687 wounded; read either alone and it looks smaller. Twelve Democratic senators on the Armed Services Committee wrote to Secretary Hegseth asking why, and Republican Sen. Lisa Murkowski warned the clock-reset “circumvents Congress.” The charitable reading is bureaucratic housekeeping for a genuinely new operation; the uncharitable one is that a war whose casualty count keeps climbing became harder to add up in the same week. Second, the deepest U.S. loss isn’t on this table at all — it’s the interceptor magazine. CSIS reckons roughly two-thirds of the Patriot stock and half the THAAD stock are gone, and the exchange rate is the whole argument: a $50,000 Shahed drone is being shot down with a $4 million interceptor. U.S. officials push back on the “dangerously low” framing, and CSIS agrees there are enough missiles to finish this war. The risk it flags is the next one — and rebuilding takes years, not months.
”Massive amounts”
The interceptor arithmetic above has a sequel, and it played out in a single August week. The administration’s words said the magazine is fine. Its purchase orders said something else. Both are documented — read them side by side and weigh them yourself.
The words. On August 6, Trump posted that the U.S. has “massive amounts of ‘munitions,’ especially of certain types,” adding that “large amounts are being manufactured and shipped to the U.S. as needed” — and that the “leakers” behind shortage reports were being hunted down. In the Oval Office the same day: “We have certain types of munitions that are very powerful, that we have unlimited, virtual unlimited supply. We have others where it’s a little bit tighter.” Note that the carve-out was there from the start — his own March 2 post said stockpiles “at the medium and upper medium grade” had “never been higher or better,” while “at the highest end, we have a good supply, but are not where we want to be.” The highest end is where the interceptors live. [35]
The purchase orders — same period:
the munitions-restocking line inside the June 24 supplemental — the single largest item in the $67.1B defense request Hegseth defended in July, and Trump defended in August
the deadline Deputy Defense Secretary Steve Feinberg’s August 5 memo gave industry leaders to submit plans to “drive significantly faster, more aggressive delivery schedules and/or increased production for critical capabilities”
in framework agreements Northrop Grumman signed August 3 with the Pentagon and Lockheed Martin — toward roughly 3× Patriot PAC-3 and 4× THAAD component output, on seven-year timelines
Pressed on why a restocking supplemental was needed if supplies were unlimited, Trump acknowledged more was needed — “we need more all the time” — citing what previous administrations had sent to Ukraine (his “$300 billion” figure for that is his own, unverified). The Feinberg memo, first obtained by The Washington Post, told contractors that “years-long development cycles are not acceptable” and named sixteen programs for accelerated production, including the Next Generation Interceptor. None of this is what an unlimited supply buys more of, faster, on a three-week deadline. [36]
In fairness. The Pentagon’s stated rationale is not invented. Chief spokesman Sean Parnell didn’t merely decline to confirm shortages — he called the reports flatly “false,” and argued that when media amplify them, “the real-world consequence is clear: our enemies are emboldened to provoke – or attack – American troops.” That is a real argument: governments routinely refuse to confirm inventory levels in wartime, for exactly that reason. And this is not a one-administration story. CSIS’s Mark Cancian and Chris Park, in “Six Reasons Why the United States Is Low on Munitions”: “even before the current war, many analyses showed a need for larger munition inventories” — decades of post–Cold War underprocurement left industry “restructured… for efficient production at peacetime rates,” with surge capacity “regarded as unnecessary and wasteful,” and “it takes two to four years for munitions to be delivered once a contract is signed.” Successive administrations of both parties bought too few interceptors; the contractors built no spare capacity because no one paid for it. The war exposed the shortfall. It didn’t create it. [36]
The counter-move: a $15,000 answer to a $4 million problem
The exchange-rate problem — a ~$50,000 Shahed traded against a $4 million interceptor — now has a real counter-move, and it complicates this ledger in an interesting direction. In April testimony, Army Secretary Dan Driscoll told House appropriators that when the war began, “within about eight days, we were able to purchase … 13,000 Merops” — interceptor drones running about $15,000 a unit, projected below $10,000 at scale, against a Shahed’s estimated $30,000–50,000. “That puts us on the right end of the cost curve, and we will make that trade all day long.” The maker, Perennial Autonomy (the Eric Schmidt–founded company previously known as Project Eagle), signed a Pentagon framework worth up to $500 million over three years in May — described in trade coverage as the largest counter-drone deal yet awarded — and the system is in service with Poland and Romania, with Lithuania buying 48. The proof of concept is Ukrainian: interceptor drones accounted for more than 70% of Shahed kills over Kyiv in February, per commander-in-chief Oleksandr Syrskyi, out of roughly 100,000 interceptors Ukraine says it produced in 2025. [37]
Honest footnote. Better economics has not yet meant better protection, and the honest version says so. The Ukraine data that proves the concept also carries the warning: independent analysis of the same theater found Russian launch volumes fell about 30% in June while the share of drones reaching their targets rose month over month — cheaper interception, worse leakage. The fully burdened cost per confirmed kill — launchers, crews, sensors, the interceptors that miss — has never been published, for Merops or anything else. And nearly every number in this story is the seller’s or a belligerent’s: the unit costs are the Army’s own projections, the kill shares are the Ukrainian military’s, and the Shahed price is an analyst estimate. A promising direction on the cost curve is not yet a solved problem. [37]
A “two-week” truce, tracked
The war was sold as a short, decisive operation. The ceasefire meant to end it has been declared, broken, extended, and collapsed so many times it has become a running joke with a body count. Here is the tally — the counter is live, and the log is updated weekly.
War begins — U.S. and Israeli strikes open the 2026 Iran war.
A two-week ceasefire, mediated by Pakistan — violated by both sides almost at once.
The Islamabad talks fail; the U.S. imposes a naval blockade.
Trump extends the ceasefire “indefinitely.”
New 60-day ceasefire terms agreed.
The Islamabad Memorandum is signed — a 60-day window to negotiate a final deal.
Geneva peace talks are abruptly postponed; oil jumps.
Trump threatens fresh strikes and to “complete the job.”
After more clashes, both sides again agree to stop — a truce barely 11 days old and already shaky.
Three ships are attacked in the strait; the U.S. steps up strikes; Trump declares the truce over.
The interim ceasefire collapses after renewed U.S. strikes.
The U.S. hits 80 targets across Iran and reimposes oil sanctions; Iran strikes 85 U.S. targets in Bahrain and Kuwait. Pakistan and Qatar scramble to restart talks.
The U.S. mounts a third straight night of strikes, reinstates its naval blockade and floats a 20% Hormuz transit toll; Iran answers with cruise missiles that disable two Emirati tankers in the strait, and oil jumps more than 9%.
The U.S. widens the war to infrastructure — bombing bridges, energy sites and a port tower to sever Iran’s main port at Bandar Abbas — while Iran strikes a Kuwaiti water-desalination plant and fires missiles at Qatar, both mediators in the war.
Both sides blow past red lines: Iran calls the June memorandum “suspended” and says it has “no plans for negotiations,” while two U.S. personnel are killed in an Iranian strike on a base in Jordan.
A third U.S. service member is killed as the U.S. presses a ninth straight day of strikes; Iran keeps hitting Kuwait, Qatar and Bahrain, home of the Navy’s 5th Fleet.
Mediators from Pakistan, Qatar, Turkey and Egypt hand both capitals a 10-day truce proposal: return to pre-July 9 positions, then negotiate the Hormuz language each side reads its own way. Iran’s foreign ministry confirms receiving “ideas”; a U.S. official says Trump intends to keep striking until he decides otherwise.
An eleventh straight night of U.S. strikes — 75 minutes against operations centers, hangars and drone stores, per CENTCOM — while Iran sends Arash suicide drones into Al Azraq air base in Jordan and Sheikh Isa in Bahrain. Defense Secretary Hegseth tells the Senate the war has now cost $37.5 billion.
Israeli media report Washington has told Israel it will escalate with heavy bombers for the first time, aimed at the buried Pickaxe Mountain nuclear site Trump says will be hit “pretty soon.” Iran’s joint military command answers that every U.S. interest in the region becomes a target if nuclear facilities are struck — and denies any nuclear activity at the site.
The heavy-bomber threat materializes: a twelfth straight night of strikes brings the war’s first B-1B Lancer bombers since it resumed — flown from a U.K. base per U.S. officials and open-source flight-trackers, though CENTCOM won’t confirm the airframe — as Trump widens the target list to “any site where they’re even thinking about nuclear.” A second front opens as Iran-aligned Houthis declare a naval blockade of Saudi Arabia and strike two Saudi oil tankers in the Red Sea, and Brent crude tops $100 for the first time since May 26. [25]
After 13 consecutive nights, the U.S. announces no new strikes for the first time in two weeks, and Iran holds fire too — a senior Iranian official says Tehran will keep its guns quiet as long as the U.S. does, and the U.S. ambassador to the UN says the pause is meant to make room for diplomacy. Trump says the U.S. stopped at Iran’s request; Iran denies asking. [27]
Iran fires ballistic missiles at U.S. forces in Jordan — CENTCOM calls it an attempted surprise attack and says every missile aimed at U.S. troops was intercepted. Trump promises Iran will “get a beating.”
The six-day pause ends: CENTCOM strikes back in the evening, and earlier that day U.S. and Saudi forces hit Iran-backed militias in Iraq, killing roughly 20 Popular Mobilization Forces fighters and several IRGC personnel. Brent jumps back above $90 on Trump’s threat to hit Iran hard.
On day 156, Trump cancels a planned attack, saying Iran and neighboring states asked him to hold off and that the “perimeters of a deal” are agreed: a complete reopening of Hormuz and an end to Iran’s nuclear program. Iran denies making any request, calling it “a new lie.”
Trump calls Iran’s leadership “unbelievably duplicitous” after Tehran denies talks have resumed, says the strikes he called off would have been “the biggest attacks since World War II,” and adds that he wants to give Iran “every last chance before decapitation.”
Secretary of State Rubio reports “progress … but not finality” on reopening the strait. Iran says it is negotiating with Oman, not Washington — and that it made “maximum use” of the lull, its acting defense minister claiming drone production now runs at three times pre-war levels. [28]
An Iran–Oman deal on a “temporary” shipping route is reported to be in its final stages — but on Tehran’s terms: bar U.S. and Israeli vessels, charge hostile states compensation, penalize violators 20% of cargo value, and reopen fully only once the U.S. lifts its blockade. Explosions are reported near Qeshm Island after Iran says it struck “hostile targets” in the strait. Brent settles near $82 — down from above $100 two weeks earlier.
Read this honestly. Both sides have violated every truce — this is not a one-sided ledger, and Iran’s strikes are part of it. The point isn’t who broke which ceasefire; it’s that a conflict pitched as a two-week operation has produced months of on-again/off-again bloodshed and no settled peace. A war plan that ends like this wasn’t a plan.
The cheap gas is borrowed, not free.
There’s a reason the pump price, ugly as it got, never fully matched a war that shut the Strait of Hormuz: since the first week, governments have been draining the world’s emergency oil reserves to hold the line — the largest coordinated release in history, spending a one-time buffer to soften a recurring shock.
barrels released from emergency reserves by the 32 IEA governments in March — the largest coordinated drawdown in the agency’s history, more than double the 2022 response to Russia’s invasion of Ukraine
barrels left in the U.S. Strategic Petroleum Reserve as of early August — a 43-year low, down more than 100 million since February and 43% of capacity, with the draw still running week to week
of forward cover left in OECD inventories, heading for the thinnest since January 2003; member governments’ strategic stocks are already at their lowest since 1990
Strategic reserves exist precisely for a moment like this, and tapping them in a genuine supply shock is textbook policy. But a reserve is a stock, not a tap: every barrel released to cap today’s price is a barrel that isn’t there for the next disruption, and global inventories have been drawing down by roughly 6.3 million barrels a day through the conflict — a pace that turns a strategic cushion into a countdown. The relief at the pump is real. It is also borrowed: the barrels have to be bought back later, competing for the same tight supply, and until they are, the world meets its next shock with the thinnest safety margin in a generation.
In fairness. Releasing reserves during a real disruption is exactly what they are for, and the coordinated action almost certainly prevented a worse spike — the $110–$167 scenarios the Dallas Fed modeled never arrived, partly because of it. The honest claim isn’t that using the reserve was wrong; it’s that it masks the war’s true price. A shock this size should show up at the pump and force the reckoning; instead a finite public asset is being spent to keep the sticker price down — deferring the bill rather than cancelling it, and leaving less cushion each time the ceasefire breaks again.
Sources: IEA collective-action announcements (Mar 11, 2026); U.S. EIA Weekly Petroleum Status Report / DOE SPR levels; EIA Short-Term Energy Outlook and IEA Oil Market Reports on OECD forward cover. [26][31]
Every flip moves the oil price. Someone keeps betting right.
When a war turns on and off by the week, the oil market lurches with it — and a well-timed bet becomes a fortune. Around the Iran flip-flops, the timing of some very large trades has been suspicious enough to draw formal calls for investigation.
bearish oil-futures bet placed hours before the ceasefire was announced
how far oil then fell once the ceasefire went public
in oil trades executed 15 minutes before a Trump Iran post (Mar 23)
The pattern: in the hours before Trump announced the two-week ceasefire, someone placed a roughly $950 million bet that oil would fall — and it fell about 15% when the news broke. Weeks earlier, on March 23, about $580 million in oil-futures trades hit the tape 15 minutes before Trump posted about “productive” talks with Iran, and prices tumbled again.
It has drawn official scrutiny. Rep. Ritchie Torres asked the SEC and CFTC to investigate possible insider trading, market manipulation, or misuse of confidential government information; Senators Warren and Whitehouse opened their own probe of “suspicious oil trades surrounding Trump’s Iran announcements”; and the CFTC reportedly opened an inquiry into the shorts.
The careful version. No one has been charged, and the traders’ identities and motives haven’t been established — large directional bets happen, and some are simply lucky. What’s documented is the timing, the size, and the fact that multiple members of Congress and a federal regulator think it warrants investigation. The deeper point holds regardless: when one man can turn a war on and off with a social-media post, advance knowledge of the next flip is worth a fortune — and that is exactly the conflict of interest a war-by-tweet creates.
Where these numbers come from
Every figure traces to a source below. Primary documents (IAEA, Fed, IMF) were checked directly; single-source and belligerent-party figures are flagged in the text. The full research dossier, with claim-by-claim verification status, lives in this project’s repository.
- Arms Control Association — Iran nuclear program factsheet & Aug 2020 breakout-estimates issue brief (12-month deal-era breakout; 2019 breach timeline; late-2024 <2-week estimate). armscontrol.org
- Congressional Research Service, R40094 — Iran’s nuclear program & JCPOA compliance record, incl. the 2016 heavy-water episodes. congress.gov
- Institute for Science and International Security (ISIS) — “Iranian Breakout Timelines Under JCPOA-Type Limits” (May 2025): 13,555 advanced centrifuges; a restored deal buys only 4–5 months. isis-online.org
- IAEA Board of Governors report GOV/2026/8 (Feb 27, 2026) — stockpile figures (¶33), verification blackout (¶35), the only-60%-producer finding (¶53), and the live U.S.–Iran talks (¶56); plus GOV/2025/38 (June 12, 2025 non-compliance finding). iaea.org
- ISIS — analysis of the June 2026 IAEA reports: the ~22,000-centrifuge destruction assessment (attributed to the June 2025 Israeli campaign — the last strikes inspectors could assess before access ended); HEU location estimates; post-Feb 2026 inspection halt. isis-online.org
- Kilian, Plante, Richter & Zhou (Federal Reserve Bank of Dallas), Working Paper 2609, April 2026 — “The Impact of the 2026 Iran War on U.S. Inflation”: largest disruption in history; WTI $60→$91; closure scenarios. dallasfed.org
- Same authors, CEPR/VoxEU column, May 4, 2026 — updated “cautiously optimistic” scenario (+0.6pp headline inflation). cepr.org
- UK House of Commons Library, research briefing CBP-10637 — war timeline, the June 17 memorandum’s 14 points, and its unreleased text. commonslibrary.parliament.uk
- AAA Newsroom — daily national gas averages: $2.98 pre-war, $4.56 peak (May 21), rising again July 9 on ceasefire doubts. newsroom.aaa.com
- BLS CPI releases via CNBC coverage (Apr 10, May 12, Jun 10, 2026) and the June 17, 2026 FOMC statement & projections. cnbc.com
- Mark Zandi (Moody’s Analytics) — ~$1,000 per household, itemized; “fair to ask whether it was worth it.” Fortune, July 1, 2026. fortune.com · Operation cost testimony: Military Times, Air & Space Forces.
- Lloyd’s List / Kpler — Hormuz transit collapse, stranded tankers, freight rates. lloydslist.com
- HRANA, “Between Missiles and Repression” (May 18, 2026) — documented deaths Feb 28–Apr 8, stated as minimums. en-hrana.org · UN OCHA Iran Humanitarian Update No. 4. unocha.org
- IMF — World Economic Outlook, April 2026 (“Global Economy in the Shadow of War”) and July 8, 2026 update; IMF Asia blog (Apr 16, 2026). imf.org
- World Bank — Iran macro-poverty outlook (GDP −2.7% FY2025/26; rial −44%; food inflation 99%+). worldbank.org · Iran contraction/inflation 2026: Euronews
- Stimson Center — “Iran Conflict Hits Foundations of Gulf Economies” (Gulf losses, Ras Laffan, IMF Gulf projections). stimson.org
- Asian Development Bank, April 2026 outlook — Asia growth & inflation revisions attributable to the conflict. adb.org
- The ceasefire timeline — declared, violated, extended “indefinitely,” and collapsed (Feb–Jul 2026). 2026 Iran war ceasefire (overview) · ABC News timeline · CNN (July 9 collapse)
- The suspicious oil trades — the ~$950M pre-ceasefire short and the ~15% drop; the March 23 $580M trade. Reuters via Investing.com · OilPrice
- The calls for investigation. Rep. Ritchie Torres (SEC/CFTC letter) · Sens. Warren & Whitehouse (Senate Banking)
- The mid-July escalation after the truce collapsed (Jul 13–20, 2026) — resumed strikes, the reinstated blockade and 20% Hormuz toll, the tanker and infrastructure attacks, and three U.S. service members killed. Reuters via Investing.com (Jul 13 strikes & tankers) · PBS NewsHour (Jul 17 bridges & Kuwait desalination plant) · Al Jazeera (Jul 18: MoU “suspended,” ~50 killed) · NPR (Jul 20: 3 U.S. service members killed)
- The materiel tally — U.S./coalition aircraft and equipment losses. The congressional list put U.S. losses at 42 aircraft (~$2.6B) in the first 40 days; open-source trackers carry the running total to ~70; CSIS values destroyed U.S. aerial equipment at $2.3–2.8B (including the THAAD radar and the E-3 AWACS), a figure that excludes base, ground and naval losses. Wikipedia — aviation shootdowns & accidents, 2026 Iran war · Aviation A2Z (congressional 42-aircraft / $2.6B list) · Al Jazeera (CSIS equipment valuation, THAAD & AWACS)
- The casualty and base tally — U.S. and coalition military deaths, Iranian military-death estimates, and the Gulf bases struck. U.S. service-member deaths are cross-checked to CENTCOM statements; Iranian military-death ranges span the documented HRANA floor to monitor (Hengaw, Iran International) and belligerent claims. Wikipedia — Casualties of the 2026 Iran war · NPR (U.S. troops killed in Jordan, Jul 18) · CNN (Jul 19: ninth night of strikes, base attacks)
- The July 20–22 phase — the mediators’ 10-day truce proposal, the eleventh straight night of U.S. strikes, Iran’s drone attacks on Al Azraq and Sheikh Isa, the Pentagon’s own toll as of that week (18 dead and ~482 wounded, DCAS, Jul 22 — since risen to ~687 wounded once the second casualty book is counted; see [30]), the $37.5B war-cost testimony, and the Pickaxe Mountain threat. Reuters via U.S. News (10-day proposal) · The National (mediators, Hormuz language dispute) · Washington Post (U.S. service members killed, where) · CBS News (Pentagon identifications; Jul 21–22 strikes) · Air & Space Forces Magazine and NBC News (Hegseth: $37.5B; internal $80–100B) · Al-Monitor and Times of Israel (Pickaxe Mountain; the heavy-bomber plan, single-origin Israeli reporting) · Washington Times (Iran denies nuclear activity at the site)
- The July 23 escalation — the twelfth straight night of strikes, the first B-1B heavy-bomber use since the war resumed (attributed to U.S. officials and open-source flight-tracking; CENTCOM did not confirm the airframe), Trump’s widened “any site … thinking about nuclear” threat, the Houthi Red Sea blockade and tanker attacks on Saudi Arabia, and Brent crude crossing $100 for the first time since May 26. CNBC (WTI/Brent, $100.05 Brent / $91.08 WTI) · NBC News (oil to $100 on Red Sea attacks; Houthi Saudi blockade) · Axios (B-1 bomber enters the war) · The Jerusalem Post (B-1 deployment, UK base) · The Hill (CENTCOM twelfth-night strikes)
- The reserve drawdown backdrop — the coordinated release masking the pump price. The IEA’s 32 members unanimously agreed on Mar 11, 2026 to release 400M barrels, the largest collective action in its history and more than double the 182.7M released in 2022. The U.S. share (172M) drove the Strategic Petroleum Reserve to ~311M barrels by mid-July — its lowest since 1983, down 100M+ since February and under half its ~680M capacity. OECD forward cover is compressing toward ~50 days, the thinnest since January 2003 (EIA STEO), with government strategic stocks the lowest since 1990 and global inventories drawing ~6.3M bbl/day during the conflict (IEA). IEA — largest-ever member stock release · IEA — member contributions · CNBC (SPR lowest since 1983) · The Hill (SPR / Hormuz crunch) · IEA Oil Market Report, July 2026 · EIA projects record-low OECD stocks (via IER)
- The July 25–August 7 phase — the six-day pause, Iran’s intercepted “surprise attack,” the resumption, and Trump’s cancelled strike. After 13 consecutive nights of U.S. strikes, no new strikes were announced on July 25 and both sides held fire; Iran fired ballistic missiles at U.S. forces in Jordan on July 28 (all intercepted, per CENTCOM); CENTCOM resumed strikes July 29 after U.S. and Saudi forces hit Iran-backed militias in Iraq; on August 2 Trump cancelled a planned attack citing the “perimeters of a deal,” which Iran denied requesting. Trump’s claim that Iran asked for the pause is disputed by Tehran and is reported here as a claim, not a fact. Al Jazeera (Jul 26: both sides pause a second day) · NPR (Jul 26: U.S. and Iran pause) · Wikipedia — July 2026 United States strikes against Iran (day-by-day) · Forbes (Jul 28: intercepted “surprise attack”) · Axios (Jul 28: missiles at U.S. base in Jordan) · Al Jazeera (Aug 2: why Trump halted, and Iran’s denial) · CBS News (Aug 3–4: “decapitation” remark; Rubio on Hormuz talks) · CNBC (Aug 3: Trump insists talks are underway despite Tehran’s denials)
- Iran’s rearmament during the lull, and the Iran–Oman strait talks. Iranian commanders say they made “maximum use” of the pause; the acting defense minister claims drone production is triple pre-war levels, and Reuters reports up to 400 Chinese air-defense launchers may reach Iran via Pakistan — all party claims, unverified independently. U.S. media cited in the same reporting put Iran’s remaining stocks at ~70% of pre-war missiles and ~40% of drones. Al Jazeera (Aug 4: “maximum use” — Iran boosts military strength)
- The interceptor magazine. CSIS (Mark Cancian and Chris Park, July 30, 2026) estimates U.S. Patriot interceptor inventory fell from roughly 2,300–2,500 pre-war to 759–827 — about two-thirds expended — and THAAD interceptors from 452 to ~234–278; an earlier CSIS assessment counted 1,000-plus Tomahawks and ~1,100 JASSM-ERs expended. CSIS’s own conclusion is that the U.S. has enough to finish this war and the risk lies in the next one; U.S. officials have publicly disputed the “dangerously low” framing. The $50,000-drone-versus-$4M-interceptor comparison is CSIS’s. Stars and Stripes (CSIS: ~67% of Patriots consumed) · Military Times (Patriot depletion, readiness) · CSIS — “Last Rounds? Status of Key Munitions” · Al Jazeera (Aug 5: U.S. officials push back)
- The casualty-accounting change. On July 26 the Pentagon’s DCAS database added an “Overseas Operations” category for killed and wounded “starting July 7” — the date the administration says a new phase of the war began, resetting the War Powers Act clock — alongside the existing Operation Epic Fury totals. Combined, 18 dead and 624 wounded as of July 26, rising to ~687 wounded by August 4. Twelve Democratic senators on the Armed Services Committee wrote to Secretary Hegseth seeking clarification, and Republican Sen. Lisa Murkowski objected to the clock reset. Independent trackers put the wounded as high as ~900 including casualties DCAS omits — a higher, less-verifiable figure reported here as a range, not a finding. CNN via KVIA (casualty count eclipses 600; the accounting change) · ABC News (Pentagon quietly updates how it categorizes casualties)
- The reserve drawdown, updated. The U.S. SPR fell a further ~2.85 million barrels in the week to 304.8 million — still the lowest since 1983, and about 43% of capacity (EIA Weekly Petroleum Status Report, August 5, 2026; reading of July 31). The IEA’s 400M-barrel collective action and the ~50-day OECD forward-cover figure are unchanged since the last update. EIA Weekly Petroleum Status Report · SPR level series (304.8M bbl, Jul 31, 2026) · Bloomingbit (SPR falls 2.85M bbl to lowest since 1983)
- The alliance bill. The no-advance-notice account, the “rapidly depleting” interceptor quote, and the 380+ missiles / 1,480+ drones tally: AP (Samy Magdy, Michelle L. Price, Aamer Madhani), March 6, 2026 — two anonymous Gulf officials, single-outlet reporting; the White House’s “decreased by 90%” counterclaim in the same article is a party claim. Maksad: CNN analysis (Abbas Al Lawati & Nic Robertson), June 24, 2026 — the bracketed “(The idea that)” is CNN’s, and the years-of-disappointments framing is the writers’ paraphrase, flagged in-text. Sheline: Al Jazeera (Ali Harb), July 2, 2026. The thousands-of-attacks and UAE-most-attacked findings: CRS report R48971 (Blanchard & Sharp, June 3, 2026), whose attack data partly cites press trackers. AP via PBS NewsHour (Mar 6) · CNN (Jun 24) · Al Jazeera (Jul 2) · CRS R48971
- The alliance steelman. Saudi/UAE-led lobbying to continue the war: AP (Madhani, Magdy, Lee, Mednick), March 31, 2026 — anonymous U.S., Gulf and Israeli officials; the Israeli sourcing is a belligerent party with an interest in continuation. Gulf participation (“Both Saudi Arabia and the UAE reportedly have conducted strikes on Iran, the UAE reportedly has welcomed direct Israeli defense aid” — CRS’s own “reportedly” retained; the Iron Dome detail is Reuters via CRS). Jacobs Khalaf (Arab Gulf States Institute, non-resident fellow): the diversification-not-replacement quote, from the same Al Jazeera piece as [32]‘s Sheline quotes. AP via PBS NewsHour (Mar 31) · CRS R48971 (PDF) · Al Jazeera (Jul 2)
- The underwriters closed the strait. The 48-hour sequence (fivefold premium surge, terminated coverage at ~60× rates, the Lloyd’s Joint War Committee redesignation, >80% traffic collapse before any Iranian attack): Dr. John Hatzadony, “The Insurance Weapon,” irregularwarfare.org, published March 24, 2026 — the specific 48-hour figures are single-sourced to this analysis, flagged in-text; the −92% week-one figure is Kpler’s (Mar 12). July premium levels (1–3% → 7.5–10% of hull value): Marcus Baker, global head of marine, cargo and logistics at Marsh, via S&P Global Platts (Jul 22), corroborated by Al Jazeera (Jul 23). The $250k → $3–10M per-voyage arithmetic and ~0.25% pre-war rate: The National (Jul 17), which also carries the IMO’s ~6,000-trapped-seafarers count (IMO Secretary-General statement, Jul 8). “At least nine ships … since July 6”: IMO data via CNBC (Jul 17). VLCC record: EIA “Today in Energy” (Mar 26) — highest “since at least November 2005, when data were first recorded.” Newton quote: The National (Jun 2). irregularwarfare.org · Kpler (−92%) · S&P Global Platts (Baker) · The National (Jul 17) · IMO SG statement (Jul 8) · CNBC (nine ships) · EIA Today in Energy · The National (Jun 2, Newton)
- The words. Trump, Truth Social, Aug 6, 2026, 12:34 a.m. (“massive amounts of ‘munitions’… large amounts are being manufactured and shipped to the U.S. as needed”) and Mar 2, 2026, 11:56 p.m. (the medium/upper-medium vs. highest-end carve-out) — both checked against the archived posts. Oval Office, Aug 6: “unlimited, virtual unlimited supply… others where it’s a little bit tighter,” and the we-need-more-all-the-time / Ukraine exchange; his “$300 billion” Ukraine figure is his own, unverified, and transcriptions vary by outlet. Truth archive (Aug 6 post) · Truth archive (Mar 2 post) · Washington Times (Aug 6, Oval Office) · CNN live coverage (Aug 6) · UPI (Aug 6)
- The purchase orders — and the fairness flags. The $21B munitions line is the largest item in the $67.1B DOD portion of the $87.6B supplemental submitted June 24, 2026 (Defense Daily; CRS IN12700; CSIS) and defended in Hegseth’s July 21 testimony — it is not an August request. The Feinberg memo (Aug 5, first obtained by The Washington Post; “memo,” per all coverage): Defense News (Aug 11), AP, Fortune (Aug 9). The Northrop framework agreements (Aug 3, >$3B total: ~$2B/7yr PAC-3 MSE components toward ~3× output + ~$1B/7yr THAAD components toward ~4×; one package with Lockheed and the Pentagon, not separate deals): Northrop press release; Bloomberg. Parnell: statement to Just the News (Aug 8) — a single-origin quote; his flat “false” denial and the emboldening rationale are both from it; his everything-it-needs line is from Al Jazeera (Aug 5). The decades-of-underprocurement framing: CSIS, Cancian & Park, “Six Reasons Why the United States Is Low on Munitions” (Jul 31, 2026). Defense Daily ($21B line) · CRS IN12700 · Military Times (Jul 21 testimony) · Defense News (Feinberg memo) · Fortune (Aug 9) · Northrop Grumman (Aug 3 release) · Just the News (Parnell) · Al Jazeera (Aug 5) · CSIS — Six Reasons
- The counter-move. Driscoll’s 13,000-Merops-in-eight-days testimony: House Appropriations defense subcommittee budget hearing, April 16, 2026 (not Armed Services), quoted via Defense News (Apr 20) — the ellipsis in the purchase quote is Defense News’s own; at a separate May 13 Senate hearing he gave a different cut (“23,000 by day 10” including Bumblebees), so the figures are the buyer’s own testimony and vary by telling. Unit costs ($15k now, sub-$10k projected) are Army projections; the Shahed’s $30–50k is an analyst estimate. Perennial Autonomy (Eric Schmidt’s White Stork → Project Eagle → Perennial): $500M ceiling, 3-year IDIQ (May 19); the “largest counter-drone deal” superlative appears in trade-aggregator coverage only. Poland and Romania have fielded Merops; Lithuania bought 48. Ukraine: >70% of Shahed kills over Kyiv in February by interceptor drones per Syrskyi (Ukrinform, Mar 3) and ~100,000 produced in 2025 per Ukraine’s NSDC — both belligerent-party figures. The leakage caveat: ISIS monthly Shahed analysis (Jul 7, 2026) — June launches 5,749 vs May’s 8,161, hit rate up from 6.65% to 8.07% — Russia–Ukraine theater data, not Iran; the unpublished burdened-cost point via drone-warfare.com (Aug 6). Defense News (Apr 20) · Defense News (May 19, contract) · DefenseScoop · Defense News (May 5, NATO fielding) · Ukrinform (Syrskyi) · Ukraine NSDC · ISIS Shahed analysis (Jul 7) · drone-warfare.com