The Map
The public record on U.S. munitions sustainment for a Taiwan Strait contingency sits at an unusual moment: an active, unrelated war has already consumed a documented share of the same interceptor and missile types a Taiwan scenario would require, and the Pentagon and outside analysts are now in open, dated disagreement about how much. CNN reported in early August 2026 that senior officials described THAAD interceptor stocks as depleted by roughly 80 percent and Patriot by roughly half since the Iran war began, citing two sources familiar with an inventory report; the Defense Secretary publicly called the 80 percent figure "not true" the same week. A separate CSIS analysis, built from publicly available expenditure and production data rather than classified access, estimated Patriot inventories had fallen by roughly 65 percent to fewer than 1,000 interceptors and THAAD by roughly 38 percent to an estimated 234-278 interceptors, using a pre-war baseline of about 2,200 Patriots and 452 THAAD missiles. These are independent think-tank estimates, not confirmed government figures, and they do not agree with each other or with the CNN-reported numbers -- the spread itself (38 percent vs. 80 percent for THAAD) is a measured disagreement between named sources, not noise to be averaged away.
What is genuinely unknown is not whether consumption occurred -- that is documented across multiple outlets -- but its true magnitude, because the Pentagon keeps exact munitions inventories classified and every open estimate is a derivation from partial data. CSIS itself frames its Taiwan-scenario missile estimates as wargame outputs subject to how the conflict evolves, not measured facts, and states plainly that open-source assessments suggest the Pentagon lacks the numbers for a protracted China conflict without confirming the classified baseline that would prove or disprove that. The wargame record adds a further layer of honest limitation: the CSIS Taiwan invasion wargame was played 24 times using rules built from historical analogues (Normandy, Okinawa, the Falklands) and theoretical weapons-performance data, explicitly to fill a gap left by classified wargames that are "not transparent to the public" -- meaning the widely cited finding that the 2026 global LRASM inventory of roughly 450 missiles would be exhausted within seven days is a simulation output from an unclassified model, not a demonstrated real-world expenditure rate, and Chinese state commentary has separately disputed the wargame's assumptions as biased toward underestimating PLA capability.
Production surge capacity is the one area where the open record offers dated, checkable commitments rather than pure assessment. Congress enacted funding for critical munitions procurement in the FY2026 Defense Appropriations Act and approved multiyear procurement authority for eight munitions programs including PAC-3, THAAD, Tomahawk, LRASM, and JASSM-ER, while a Lockheed Martin contract awarded in 2026 covers PAC-3 MSE interceptor production through fiscal 2032 and an RTX framework agreement targets doubling to quadrupling SM-3 production rates. But every account of these contracts is bounded by the same physical constraint cited by named analysts: missile production timelines run two to four years from funding to delivery, and current-fiscal-year delivery rates for Tomahawk and Patriot were reported in early August 2026 at roughly 15 and 20 units per month respectively -- rates that, if accurate, would take years to rebuild pre-war stockpiles even before any Taiwan-scenario surge demand is added on top.
The assumption most at risk of being treated as settled when it is not is that Taiwan-scenario planning and Iran-war attrition draw on separate pools. Analysts cited by defense trade press have identified "sizable overlaps" in the munitions types -- long-range strike and air and missile defense systems -- that both theaters would draw on, meaning the Iran-war drawdown is not a contained, separate event but a live constraint on Taiwan-scenario sustainment happening in real time. Whether that overlap is large enough to be strategically decisive, or manageable given the surge contracts now under contract, is not resolvable from the open record because the classified inventory figures that would settle it are not published, and the two named public estimates of the drawdown (CNN's sourced figures and CSIS's open-source derivation) do not agree with each other.
The Uncertainty Ledger (7)
Each material unknown ranked by what turns on it, with an honest reducibility grade — a checkable question is never dressed as an irreducible one.
What is the actual current U.S. THAAD and Patriot interceptor inventory, as opposed to the competing public estimates?opaque by design
fact unknown
Turns on: Whether U.S. missile defense has meaningful reserve capacity for a Taiwan contingency layered on top of Iran-war consumption, or whether the two theaters would compete for a stock already critically thin.
How to reduce: No open-source route exists; the Pentagon classifies exact munitions inventories as a matter of policy. The nearest checkable proxies are CSIS's published open-source derivation and any future DOD Inspector General or GAO report that discloses aggregate (not unit-level) figures.
Best read: Materially depleted from pre-war levels, with named estimates ranging from roughly 38% (CSIS, THAAD) to roughly 80% (CNN-sourced, THAAD) -- the spread itself is the finding.contested
Confidence: Medium
Do U.S. Taiwan-contingency and Iran-war munitions requirements draw on the same or separate stockpiles?checkable now
structural
Turns on: Whether Iran-war attrition is a contained, separate drawdown or a live, ongoing constraint on Taiwan-scenario sustainment happening in real time.
How to reduce: Cross-reference DOD's published Section 804 multiyear-procurement munitions list (PAC-3, SM-6, THAAD, AMRAAM, Tomahawk, LRASM, JASSM-ER, SM-3 IB) against named systems cited in Pacific-theater wargame reporting; the overlap is a named-analyst claim (Atlantic Council) that can be checked against the enacted program list.
Resolves: No single resolution date; resolves incrementally as each program's delivery and allocation data is published in future budget justifications.
Best read: Substantial overlap exists per a named former Pentagon official, since the same interceptor and long-range strike systems (Patriot, THAAD, Tomahawk, JASSM family) appear on both the Iran-war consumption list and the CSIS Taiwan-wargame expenditure list.reported
Confidence: Medium
Will the newly funded production-surge contracts (PAC-3 MSE through FY2032, SM-3 rate increases, Tomahawk multiyear buy) actually deliver on their stated schedules?answerable by named party
outcome unknown
Turns on: Whether the multi-year timeline to rebuild depleted stocks is 2-4 years as analysts state, or longer if the pattern GAO documents across DOD's costliest programs recurs.
How to reduce: Track GAO's annual Weapon Systems Assessment (next edition expected mid-2027) and DOT&E's annual report for named-program schedule slippage findings against the specific contracts (Lockheed PAC-3 MSE, RTX SM-3, Raytheon Tomahawk) already under contract.
Resolves: Incrementally, as GAO and DOT&E publish their next annual reports (historically released in the first half of the calendar year).
Best read: Uncertain; GAO's July 2026 assessment found DOD "struggled to fully enact existing reforms to achieve speed" and flagged delayed key decisions on some of the costliest programs, a pattern that predates and is independent of the current surge contracts.documented
Confidence: Medium
How reliable are CSIS's unclassified Taiwan-wargame expenditure estimates (e.g., 450 LRASM exhausted in seven days) as a predictor of real-world consumption rates?irreducible for now
structural
Turns on: Whether current production-surge targets (860 JASSM and 240 LRASM annual capacity per this user's prior scan) would be adequate or grossly insufficient against actual wartime demand.
How to reduce: No route exists to validate a real combat expenditure rate without an actual conflict; the closest available check is comparing successive CSIS wargame iterations and any future unclassified DOD wargame release for consistency, but no ground truth exists absent war.
Best read: The wargame's own designers state its rules were built from historical amphibious-operation analogues (Normandy, Okinawa, Falklands) and theoretical weapons-performance data rather than classified combat data, and Chinese state commentary has separately characterized the wargame as underestimating PLA capability -- both bound the estimate's reliability without resolving it.contested
Confidence: Low
What is the Pentagon's true classified assessment of Taiwan-scenario sustainment adequacy, as opposed to public statements?opaque by design
intent unknown
Turns on: Whether senior leadership's public reassurances reflect an actual internal assessment of sufficiency or are a public-messaging position distinct from internal planning assumptions.
How to reduce: No open-source route exists to the classified assessment itself; the nearest available signal is Congressional testimony transcripts (e.g., the Defense Secretary's April 2026 testimony that replenishment could take "months and years") and any future declassified excerpts released through Congressional oversight.
Best read: Public statements are internally inconsistent -- the Defense Secretary testified in April 2026 that replenishment could take "months and years" while separately calling later depletion reporting a "manufactured story," and the Acting Navy Secretary's May 2026 Senate testimony linking a Taiwan arms-sale pause to stockpile concerns was itself publicly contradicted by the Defense Secretary at the Shangri-La Dialogue.contested
Confidence: Medium
Does the Taiwan arms-sale pause reported in May 2026 reflect an actual U.S. stockpile constraint or an unrelated diplomatic/negotiating decision?answerable by named party
intent unknown
Turns on: Whether Taiwan's own defensive posture, and therefore first-week wargame assumptions about Taiwanese resistance capacity, should be revised downward from what prior arms-sale schedules assumed.
How to reduce: Monitor the Taiwan Security Monitor's tracked arms-sale backlog status updates and any further Senate Armed Services or Appropriations Committee testimony from the Acting Navy Secretary or Defense Secretary on the specific $14 billion sale's status.
Resolves: No fixed date; resolves as the specific delayed sale is either delivered, formally cancelled, or re-justified in future testimony.
Best read: Contested -- the Acting Navy Secretary told a Senate subcommittee in May 2026 the pause was to preserve munitions stockpiles, while the Defense Secretary stated at the Shangri-La Dialogue the same month that the pause was unrelated to stockpile concerns.contested
Confidence: Medium
What is the true cost of the Iran war to date, and does the officially reported figure capture the full drawdown?answerable by named party
fact unknown
Turns on: Whether Congress's supplemental funding requests (and therefore the pace of any Taiwan-relevant production surge funding) are sized to the actual problem or to an understated official estimate.
How to reduce: Track the Acting Defense Department comptroller's successive cost updates in Congressional testimony (already revised upward once, from roughly $25 billion to $29 billion between April and May 2026 testimony) against outside estimates cited by Congressional critics ranging up to roughly $72 billion.
Resolves: No fixed date; resolves incrementally with each successive comptroller testimony or supplemental funding request.
Best read: Officially estimated at nearly $29 billion as of May 2026 testimony, already revised upward once; Congressional critics and outside estimates cited in the same reporting range considerably higher.contested
Confidence: Medium
Assumed vs Known (4)
What the conventional read treats as settled, against what the record actually shows.
That the CSIS Taiwan-wargame figures (450 LRASM inventory exhausted in seven days; 5,000+ long-range missiles expended in three weeks) represent a validated, near-certain expenditure forecast for an actual Taiwan conflict.treated as working assumption
The record shows: CSIS itself frames these as wargame outputs from an unclassified simulation built on historical amphibious-operation analogues and theoretical weapons-performance data, explicitly created to fill a gap left by inaccessible classified wargames -- not as measured or demonstrated combat data.
If wrong: Production-surge targets calibrated to these figures could be significantly mis-sized in either direction if real combat consumption diverges from the simulation's assumptions.
That Iran-war interceptor consumption and Taiwan-contingency planning draw on functionally separate stockpiles, so current Middle East attrition does not materially affect Pacific readiness.treated as working assumption
The record shows: A named former Pentagon official (Atlantic Council) has stated there are "sizable overlaps" in the systems both theaters would rely on, and the same interceptor and strike-missile types (Patriot, THAAD, Tomahawk, JASSM family) appear on both the Iran-war depletion reporting and the Taiwan-wargame expenditure lists.
If wrong: Sustainment planning for a Taiwan contingency could be built on an inventory baseline that is already substantially lower than assumed, without that being reflected in current public assessments.
That the Pentagon's public reassurances ("our stockpiles are great," "has everything it needs") reflect the same picture as internal classified inventory assessments.treated as settled fact
The record shows: The Defense Secretary's own April 2026 Congressional testimony that replenishment could take "months and years" sits in tension with his later public dismissal of depletion reporting as a "manufactured story," and the Acting Navy Secretary's stockpile-linked justification for the Taiwan arms-sale pause was itself publicly disputed by the Defense Secretary weeks later.
If wrong: Force planners relying on public statements rather than the underlying testimony record could underweight documented internal disagreement about the stockpile picture.
That the enacted FY2026 multiyear procurement authorities and framework agreements guarantee the stated production increases will materialize on schedule.treated as working assumption
The record shows: GAO's July 2026 Weapon Systems Annual Assessment found DOD "struggled to fully enact existing reforms to achieve speed" and flagged delayed key decisions on some of the costliest weapon programs, keeping DOD acquisition on GAO's High-Risk List -- a documented, general pattern of schedule risk that applies across DOD acquisition, not specific to the munitions surge programs but not shown to be absent from them either.
If wrong: Sustainment timelines built on contractor-stated capacity figures (e.g., 860 JASSM/240 LRASM annual capacity, sixteen-fold Tomahawk rate increase) could slip well beyond the 2-4 year horizons analysts currently cite.
Facts & Figures (17)
The claims behind this analysis, each with its verification status — including what is contested, unverified, or could not be established.
What is the actual current U.S. THAAD and Patriot interceptor inventory, as opposed to the competing public estimates? — Materially depleted from pre-war levels, with named estimates ranging from roughly 38% (CSIS, THAAD) to roughly 80% (CNN-sourced, THAAD) -- the spread itself is the finding.
Whether U.S. missile defense has meaningful reserve capacity for a Taiwan contingency layered on top of Iran-war consumption, or whether the two theaters would compete for a stock already critically thin.
◑ CONTESTEDopaque by design · Medium confidence
Do U.S. Taiwan-contingency and Iran-war munitions requirements draw on the same or separate stockpiles? — Substantial overlap exists per a named former Pentagon official, since the same interceptor and long-range strike systems (Patriot, THAAD, Tomahawk, JASSM family) appear on both the Iran-war consumption list and the CSIS Taiwan-wargame expenditure list.
Whether Iran-war attrition is a contained, separate drawdown or a live, ongoing constraint on Taiwan-scenario sustainment happening in real time.
○ REPORTEDcheckable now · Medium confidence
Will the newly funded production-surge contracts (PAC-3 MSE through FY2032, SM-3 rate increases, Tomahawk multiyear buy) actually deliver on their stated schedules? — Uncertain; GAO's July 2026 assessment found DOD "struggled to fully enact existing reforms to achieve speed" and flagged delayed key decisions on some of the costliest programs, a pattern that predates and is independent of the current surge contracts.
Whether the multi-year timeline to rebuild depleted stocks is 2-4 years as analysts state, or longer if the pattern GAO documents across DOD's costliest programs recurs.
✓ DOCUMENTEDanswerable by named party · Medium confidence
How reliable are CSIS's unclassified Taiwan-wargame expenditure estimates (e.g., 450 LRASM exhausted in seven days) as a predictor of real-world consumption rates? — The wargame's own designers state its rules were built from historical amphibious-operation analogues (Normandy, Okinawa, Falklands) and theoretical weapons-performance data rather than classified combat data, and Chinese state commentary has separately characterized the wargame as underestimating PLA capability -- both bound the estimate's reliability without resolving it.
Whether current production-surge targets (860 JASSM and 240 LRASM annual capacity per this user's prior scan) would be adequate or grossly insufficient against actual wartime demand.
◑ CONTESTEDirreducible for now · Low confidence
What is the Pentagon's true classified assessment of Taiwan-scenario sustainment adequacy, as opposed to public statements? — Public statements are internally inconsistent -- the Defense Secretary testified in April 2026 that replenishment could take "months and years" while separately calling later depletion reporting a "manufactured story," and the Acting Navy Secretary's May 2026 Senate testimony linking a Taiwan arms-sale pause to stockpile concerns was itself publicly contradicted by the Defense Secretary at the Shangri-La Dialogue.
Whether senior leadership's public reassurances reflect an actual internal assessment of sufficiency or are a public-messaging position distinct from internal planning assumptions.
◑ CONTESTEDopaque by design · Medium confidence
Does the Taiwan arms-sale pause reported in May 2026 reflect an actual U.S. stockpile constraint or an unrelated diplomatic/negotiating decision? — Contested -- the Acting Navy Secretary told a Senate subcommittee in May 2026 the pause was to preserve munitions stockpiles, while the Defense Secretary stated at the Shangri-La Dialogue the same month that the pause was unrelated to stockpile concerns.
Whether Taiwan's own defensive posture, and therefore first-week wargame assumptions about Taiwanese resistance capacity, should be revised downward from what prior arms-sale schedules assumed.
◑ CONTESTEDanswerable by named party · Medium confidence
What is the true cost of the Iran war to date, and does the officially reported figure capture the full drawdown? — Officially estimated at nearly $29 billion as of May 2026 testimony, already revised upward once; Congressional critics and outside estimates cited in the same reporting range considerably higher.
Whether Congress's supplemental funding requests (and therefore the pace of any Taiwan-relevant production surge funding) are sized to the actual problem or to an understated official estimate.
◑ CONTESTEDanswerable by named party · Medium confidence
That the CSIS Taiwan-wargame figures (450 LRASM inventory exhausted in seven days; 5,000+ long-range missiles expended in three weeks) represent a validated, near-certain expenditure forecast for an actual Taiwan conflict.
CSIS itself frames these as wargame outputs from an unclassified simulation built on historical amphibious-operation analogues and theoretical weapons-performance data, explicitly created to fill a gap left by inaccessible classified wargames -- not as measured or demonstrated combat data.
— TREATED AS WORKING ASSUMPTIONif wrong: Production-surge targets calibrated to these figures could be significantly mis-sized in either direction if real combat consumption diverges from the simulation's assumptions.
That Iran-war interceptor consumption and Taiwan-contingency planning draw on functionally separate stockpiles, so current Middle East attrition does not materially affect Pacific readiness.
A named former Pentagon official (Atlantic Council) has stated there are "sizable overlaps" in the systems both theaters would rely on, and the same interceptor and strike-missile types (Patriot, THAAD, Tomahawk, JASSM family) appear on both the Iran-war depletion reporting and the Taiwan-wargame expenditure lists.
— TREATED AS WORKING ASSUMPTIONif wrong: Sustainment planning for a Taiwan contingency could be built on an inventory baseline that is already substantially lower than assumed, without that being reflected in current public assessments.
That the Pentagon's public reassurances ("our stockpiles are great," "has everything it needs") reflect the same picture as internal classified inventory assessments.
The Defense Secretary's own April 2026 Congressional testimony that replenishment could take "months and years" sits in tension with his later public dismissal of depletion reporting as a "manufactured story," and the Acting Navy Secretary's stockpile-linked justification for the Taiwan arms-sale pause was itself publicly disputed by the Defense Secretary weeks later.
— TREATED AS SETTLED FACTif wrong: Force planners relying on public statements rather than the underlying testimony record could underweight documented internal disagreement about the stockpile picture.
That the enacted FY2026 multiyear procurement authorities and framework agreements guarantee the stated production increases will materialize on schedule.
GAO's July 2026 Weapon Systems Annual Assessment found DOD "struggled to fully enact existing reforms to achieve speed" and flagged delayed key decisions on some of the costliest weapon programs, keeping DOD acquisition on GAO's High-Risk List -- a documented, general pattern of schedule risk that applies across DOD acquisition, not specific to the munitions surge programs but not shown to be absent from them either.
— TREATED AS WORKING ASSUMPTIONif wrong: Sustainment timelines built on contractor-stated capacity figures (e.g., 860 JASSM/240 LRASM annual capacity, sixteen-fold Tomahawk rate increase) could slip well beyond the 2-4 year horizons analysts currently cite.
CNN reported in early August 2026, citing two sources familiar with an inventory report, that the U.S. military had depleted nearly 80% of its THAAD interceptor stockpile and roughly half of its Patriot interceptors since the Iran war began; the Defense Secretary publicly disputed the 80% THAAD figure that week, calling it "not true."
Establishes an unresolved, dated, named-source disagreement between press reporting and the Pentagon on the actual current drawdown -- the single most consequential open dispute for Taiwan sustainment planning.
✓ GROUNDED
A CSIS analysis using publicly available expenditure and production data (not classified access) estimated a pre-war baseline of roughly 2,200 Patriot missiles and 452 THAAD interceptors, with post-Iran-war levels of fewer than 1,000 Patriots (about 65% depleted) and an estimated 234-278 THAAD interceptors (about 38% depleted) as of late July 2026.
Provides the only named, methodologically-described open-source baseline figures against which any Taiwan-scenario sustainment math must be run, while itself disagreeing with the CNN-reported depletion percentages.
✓ GROUNDED
Analysts cited by Defense News in July 2026 stated current Tomahawk delivery rates run about 15 missiles per month and Patriot deliveries about 20 per month, with CSIS estimating three or more years to rebuild THAAD inventories to pre-war levels at current rates.
Sets the reducible, checkable ceiling on how fast any Taiwan-scenario stockpile gap could close, independent of whether the classified baseline is ever disclosed.
✓ GROUNDED
CSIS's unclassified Taiwan-invasion wargame (played 24 times, scenario year 2026) found the assessed 2026 global LRASM inventory of roughly 450 missiles was exhausted within seven days in some iterations, per this user's prior grounded scan; the same wargame series found the U.S. typically expended more than 5,000 long-range missiles within three weeks.
Anchors the scale of a hypothetical Taiwan expenditure rate against which the currently-depleted, Iran-war-drawn interceptor and missile stocks would have to sustain a second, larger contingency.
● YOUR INPUT
Congress approved multiyear procurement authority under Section 804 of the FY2026 NDAA for eight munitions programs -- PAC-3, SM-6, THAAD, AMRAAM, Tomahawk, LRASM, JASSM-ER, and SM-3 IB -- out of 13 the Pentagon requested, alongside $6.4 billion enacted for critical munitions procurement including $2.1 billion for the multi-year procurement ramp.
Establishes which surge-capacity commitments are actually funded in statute versus merely announced, narrowing which production claims are checkable against a named legal instrument.
✓ GROUNDED
GAO's July 2, 2026 Weapon Systems Annual Assessment (GAO-26-108457) found DOD "struggled to fully enact existing reforms to achieve speed" in weapons acquisition and that key decisions are delayed for some of the costliest programs, keeping DOD's weapon systems acquisition on GAO's High-Risk List.
Provides the canonical, dated, named institutional check on whether surge-production schedule claims from contractors and the Pentagon are likely to hold, independent of any single program's marketing.
✓ GROUNDED