The Map
The three unknowns named in this scan sit at different points on the reducibility spectrum, and conflating them produces bad analysis. Stockpile levels are opaque by design: the Pentagon has consistently declined to release exact interceptor and missile counts, citing operational security, even as multiple outlets independently triangulate similar depletion percentages from leaked internal figures. War cost is a different kind of unknown — not classified but genuinely uncounted, because no single Pentagon accounting captures munitions expended, aircraft lost, base repairs, and personnel costs under one methodology, and outside trackers apply inconsistent scope (military-only vs. total economic impact) to arrive at wildly different headline totals. Taiwan contingency demand is the least reducible of the three: it is a projection built entirely on wargame assumptions about adversary behavior, U.S. force availability, and allied participation that CSIS and other institutions explicitly acknowledge are not transparent even when the wargame itself is published.
On stockpiles, the throughline across CNN, the Wall Street Journal-sourced CSIS commentary, and Al Jazeera's CSIS-based analysis is convergence on direction and rough magnitude — CSIS estimated a pre-war baseline near 2,200 Patriots (two most-modernized variants) and 452 THAAD interceptors, with depletion estimates ranging from roughly half of Patriots and near 80 percent of THAAD by August 2026 reporting, down to CSIS's own count of fewer than 827 Patriots and fewer than 278 THAAD remaining by late July 2026. These are independent-enough sources (CSIS's own published methodology plus separately-sourced CNN reporting citing government officials) to treat the direction and rough scale as corroborated, even though the precise counts diverge by source and date and the underlying classified baseline can only be inferred.
On cost, the spread is far wider and reflects genuine methodological disagreement rather than simple reporting error. The Pentagon's own disclosed figures moved from an initial $11.3 billion for the first six days, to $29 billion per Congressional testimony in May 2026, to a CSIS preliminary estimate near $40 billion for direct military costs alone — while a Pentagon supplemental request reached $80-87.6 billion depending on the source, and outside trackers extended the accounting to $113-132 billion or higher by folding in economic impact, oil-price effects, and household costs that the Pentagon figures explicitly exclude. No single number is 'the' cost; the range reflects at least three distinct accounting boundaries (direct military only, military plus replenishment/repair, and total economic impact) being reported side by side without consistent labeling.
On Taiwan contingency demand, the public record contains no expenditure figure at all — only qualitative wargame findings. CSIS's own published Taiwan blockade and invasion wargames explicitly note that classified Pentagon wargames' assumptions and results are not transparent to the public, and even CSIS's unclassified 26-iteration blockade project and earlier invasion-scenario work report attrition and import-disruption findings rather than munitions-expenditure totals comparable to the Iran-war figures. Any demand estimate for a Taiwan scenario is necessarily an extrapolation from these qualitative findings plus assumptions about intensity, duration, and adversary capability that are not published anywhere in the open literature.
The connective tissue between all three unknowns is that the Iran war has already consumed a meaningful share of exactly the munitions categories — PrSM, JASSM-ER, Tomahawk, Patriot, THAAD — that Indo-Pacific planning has historically assumed would be available for a Taiwan contingency, and CSIS's own commentary states plainly that once Operation Epic Fury ends, forces diverted to the Middle East will return to the Pacific but that restoring depleted stockpiles will take years. That creates a genuinely irreducible near-term unknown: nobody in the open record can currently state with precision how much Taiwan-scenario 'demand' the post-Iran-war stockpile can actually satisfy, because the numerator (remaining stockpile) is imprecisely known and the denominator (wargamed requirement) is not published at all.
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. inventory of Patriot (PAC-3/GEM-T) and THAAD interceptors as of today, September 2026?opaque by design
fact unknown
Turns on: This determines whether the U.S. has adequate air and missile defense stock for a simultaneous or near-simultaneous Indo-Pacific contingency, and whether Gulf allies' own THAAD/Patriot resupply requests can be met without further drawing down U.S. reserves.
How to reduce: Track CSIS Mark Cancian's periodic published commentaries (the 'Last Rounds' series), which the Pentagon has informally validated as directionally accurate per CNN's sourcing — but treat any precise count as an estimate, not a disclosed figure; DoD has stated it will not release exact numbers for operational security reasons.
Best read: Patriot stocks are reduced somewhere between roughly 50 and 65 percent from the ~2,200 pre-war baseline (reconciling CSIS's ~827-remaining figure against CNN's 'roughly half' figure), and THAAD stocks are reduced by 75-80+ percent from the 452 pre-war baseline, based on triangulation of two independently-sourced reports from the same reporting window.reported
Confidence: Medium
What is the true total cost of the 2026 Iran war to the United States?answerable by named party
fact unknown
Turns on: This shapes Congressional appetite for the pending supplemental appropriation, sets the political benchmark against which any future Iran or Taiwan contingency cost will be judged, and determines how much of the FY2026/FY2027 defense topline is effectively already committed to Iran-war replenishment rather than Indo-Pacific modernization.
How to reduce: The Pentagon Comptroller's office and the House and Senate Armed Services Committees hold the authoritative accounting; the concrete next step is tracking the final appropriations conference report on the pending supplemental request, which will fix a Congressionally-legislated dollar figure even if it does not resolve the broader economic-impact estimates.
Resolves: No single documented resolution date; the direct-military-cost figure will firm up once Congress finalizes the supplemental appropriation, but total economic-impact estimates (Moody's Analytics, independent trackers) will remain contested indefinitely.
Best read: Direct military costs plausibly fall in the $29-42 billion range based on the Pentagon's own testimony and CSIS's preliminary estimate, while total supplemental funding requests span $80-87.6 billion and outside total-economic-impact trackers claim $113-132 billion or more — these are not competing answers to the same question but different questions being asked with the same dollar sign.contested
Confidence: Low
How much of the pre-Iran-war U.S. precision-munitions stockpile (Tomahawk, JASSM-ER, PrSM, SM-3, Patriot, THAAD) remains available for a Taiwan Strait contingency today?irreducible for now
outcome unknown
Turns on: This is the single most consequential unknown for Indo-Pacific force planning, since it determines whether current stated intent to prioritize the Pacific is backed by actual inventory or is aspirational given multi-year replenishment timelines.
How to reduce: No single document resolves this because it requires combining (a) the classified true stockpile baseline, (b) the classified actual Iran-war expenditure by munition type, and (c) the classified or wargamed Taiwan-scenario requirement — cross-referencing CSIS's periodic stockpile commentaries against DoD's own production-acceleration announcements is the closest open-source proxy available.
Best read: Given reported Army use of 'virtually all' PrSM/ATACMS stock and reported Tomahawk expenditure exceeding 1,000 missiles against an annual procurement rate roughly a tenth of that pace, near-term Taiwan-scenario stock for at least these two munition types is materially degraded from pre-war levels, but no defensible read exists for the aggregate multi-system picture.inferred
Confidence: Low
What quantity and mix of munitions would a Taiwan Strait contingency actually require, and over what timeframe?opaque by design
structural
Turns on: This is the denominator against which any stockpile-adequacy judgment must be made; without it, statements that the U.S. is 'ready' or 'not ready' for Taiwan given current post-Iran-war stocks are unfalsifiable.
How to reduce: The published CSIS unclassified wargame series (26-iteration blockade project, prior invasion-scenario work) is the best open-source proxy and explicitly states its own assumptions and methodology; classified DoD wargames exist but their assumptions and results are, by CSIS's own account, not transparent to the public and will not be released.
Best read: Open-source wargaming (both CSIS's invasion and blockade projects) consistently finds rapid attrition and 'quick exhaustion of long-range precision-guided munitions' as a qualitative finding, but no public wargame report translates this into a quantified munitions-expenditure total comparable to the dollar or unit figures reported for the Iran war.documented
Confidence: Medium
Are the CSIS-cited depletion percentages (roughly half of Patriots, ~80 percent of THAAD) an accurate reflection of the classified internal Pentagon inventory report, or an independent estimate that merely resembles it?answerable by named party
fact unknown
Turns on: If CSIS's model-derived estimates diverge meaningfully from the true classified baseline, every downstream policy argument built on 'the U.S. has X percent of its interceptors left' is built on an unverified proxy rather than ground truth.
How to reduce: CNN's sourcing states three officials familiar with Pentagon data called the CSIS estimates 'close to' internal figures — the concrete next step for a researcher is watching for any DoD Inspector General or GAO review of munitions accounting practices, which would be the first non-anonymous-sourced corroboration.
Resolves: No documented resolution date; would require either a leak, a declassification, or a GAO/IG report explicitly validating or contradicting the CSIS model.
Best read: Multiple independently-reported figures (CSIS's own count, CNN's percentage-based sourcing, and the earlier April 2026 New York Times report citing classified documents showing JASSM-ER inventory near 1,500 of a 2,600 baseline) converge on the same direction and rough order of magnitude, which is the strongest available corroboration in an opaque data environment.reported
Confidence: Medium
Will accelerated munitions production actually close the Patriot/THAAD/PrSM inventory gap before a Taiwan contingency could plausibly arise?resolves on date
outcome unknown
Turns on: This determines whether current force-posture statements about Pacific prioritization are credible on a multi-year timeline or represent an inventory the U.S. does not yet, and may not soon, possess.
How to reduce: Existing PrSM production contracts specify delivery schedules (54 units in 2026, 208 in 2028, 73 in 2029 per reported contract terms), giving a documented near-term production trajectory that can be checked against actual deliveries as they occur.
Resolves: Contracted delivery milestones in 2026, 2028, and 2029 for PrSM; CSIS's own commentary states THAAD rebuild alone was estimated at three-plus years at pre-surge delivery rates.
Best read: Even with announced production acceleration ('quadruple production' directive) and a $22.9 billion Tomahawk contract reportedly signed, CSIS's own estimate is that restoring depleted stockpiles to pre-war levels will take multiple years at best, meaning any near-term Taiwan contingency would draw on a still-degraded base.reported
Confidence: Medium
Does the disclosed $29-42 billion direct-cost range for the Iran war include or exclude pre-conflict force buildup, base repair, and personnel/readiness costs?checkable now
fact unknown
Turns on: Whether Congress and outside trackers are comparing like-for-like figures when debating supplemental appropriations, and whether the eventual total bill significantly exceeds current public expectations.
How to reduce: Cross-reference the Pentagon Comptroller's House Armed Services Committee testimony transcript (May 12, 2026) against the itemized supplemental request breakdown, which reportedly separates direct war-cost reimbursement from munitions replenishment, aircraft replacement, and naval-vessel repair as distinct line items.
Resolves: Congressional Research Service or GAO cost-accounting review, if commissioned, would be the authoritative checkable source; none is confirmed as underway in current reporting.
Best read: Reporting indicates the $29-42 billion range covers direct military action (munitions, destroyed equipment, base damage) but explicitly excludes standing operational costs already budgeted in the FY2026 defense budget and, per one CNN source, excludes facility repair and regional basing costs — meaning the 'headline' cost figures are narrower in scope than casual comparison to the $80-132 billion figures suggests.reported
Confidence: Medium
Assumed vs Known (4)
What the conventional read treats as settled, against what the record actually shows.
The U.S. can rapidly pivot forces and munitions stocks back to the Indo-Pacific once the Iran war formally ends, restoring the pre-war deterrence posture against a Taiwan contingency.treated as working assumption
The record shows: CSIS's own commentary states that while naval assets will physically return to the Pacific once operations end, restoring depleted munitions stockpiles to desired inventory levels 'will take many years' — the return of platforms and the return of usable stock are not the same event, and reporting does not establish any accelerated timeline that shortens this gap.
If wrong: Force-posture statements asserting Pacific-first prioritization could be materially unsupported by actual available inventory for a multi-year window, undermining deterrence credibility if China perceives the gap.
The Pentagon's disclosed war-cost figures ($29-42 billion range) represent a comprehensive accounting of the conflict's cost to the U.S. government.treated as working assumption
The record shows: Reporting explicitly states these figures exclude standing operational costs already in the FY2026 budget, and one source told CNN the pending $80 billion supplemental request's Iran-specific portion excludes facility repair and regional basing costs — meaning the disclosed 'cost of the war' figures are narrower than the full fiscal exposure.
If wrong: Congressional appropriators and outside analysts debating supplemental funding may be working from an artificially low cost baseline, understating the true fiscal commitment already made.
CSIS's published stockpile estimates are functionally equivalent to the classified Pentagon inventory data.treated as working assumption
The record shows: CNN's sourcing states only that officials familiar with the matter called the CSIS estimates 'close to' internal figures — this is a qualitative endorsement by anonymous sources, not a validated data match, and CSIS's own methodology relies on DOD-projected delivery schedules and estimated (not confirmed) combat expenditure rates.
If wrong: Analysts citing precise CSIS figures (e.g., '827 Patriots remaining') as if they were confirmed government numbers would be overstating the precision of what is, by CSIS's own methodology notes, a modeled estimate.
Publicly available Taiwan wargame findings (CSIS blockade and invasion projects) provide a reasonable proxy for actual U.S. Indo-Pacific Command munitions-requirement planning.treated as working assumption
The record shows: CSIS's own reports explicitly distinguish their unclassified wargames from classified Pentagon wargames, stating the latter's 'assumptions and even results are not transparent to the public' — the unclassified projects are avowedly a substitute for, not a window into, actual classified planning.
If wrong: Public debate over whether current U.S. stocks are 'adequate' for Taiwan contingencies may be anchored to a scenario construct that differs materially from what actual U.S. combatant command planning assumes.
Where Sources Diverge (3)
What percentage of pre-war Patriot and THAAD interceptor stocks the U.S. has expended in the Iran war
CNN's August 4, 2026 reporting (citing two sources familiar with the latest inventory report) states roughly half of Patriot interceptors and nearly 80 percent of THAAD interceptors have been used; CSIS's own July 2026 commentary implies a different percentage by stating fewer than 827 Patriots (of ~2,200 baseline, roughly 62 percent depleted) and fewer than 278 THAAD (of 452 baseline, roughly 38 percent depleted) remained.
Why they diverge: The two figures come from different reporting dates (late July vs. early August 2026) during an active, non-linear conflict where expenditure rates spiked and paused with ceasefire violations, and from different source chains (CSIS's own analytic model vs. CNN sources describing an internal Pentagon inventory report) that may define 'usable interceptors' differently.
The total dollar cost of the 2026 Iran war
Pentagon testimony puts direct costs at $29 billion (May 2026) rising to a CSIS preliminary estimate near $40 billion (June 2026); independent trackers including Moody's Analytics cite $132 billion in combined taxpayer and consumer costs, and other aggregator sites cite figures from $113 billion to $115 billion by folding in economic impact alongside military spending.
Why they diverge: The sources are answering different questions under the same dollar figure: official Pentagon testimony scopes to direct military action only and explicitly excludes budgeted operational costs and facility/basing repairs, while outside trackers deliberately include oil-price shock effects, household energy costs, and broader economic impact that the Pentagon accounting does not attempt to capture.
Whether current U.S. force posture genuinely still prioritizes the Indo-Pacific given Iran-war drawdowns
CSIS's Cancian frames the tradeoff as a deliberate presidential choice — 'accepting this munitions risk...alongside other tradeoffs like the diversion of forces from the Western Pacific' in order to decisively win the current war; former Defense Secretary Leon Panetta characterized the resulting stockpile shortage as 'poor planning' that leaves the U.S. dangerously exposed given 'potential threats from Russia and China.'
Why they diverge: Cancian's framing treats the drawdown as a rational, chosen tradeoff consistent with a 'win the war you're in' theory of resource allocation; Panetta's framing treats the same drawdown as an avoidable planning failure — the disagreement is over intent and foresight, not over the underlying depletion facts, which both accept as real.
Watch Signals (4)
Publication of a new CSIS 'Last Rounds'-series stockpile commentary or equivalent think-tank update citing revised Patriot/THAAD/PrSM inventory figures
Each iteration is the closest available proxy for the classified baseline and shows whether depletion is stabilizing, worsening, or reversing as production accelerates
Where to watch: CSIS Defense and Security Department publications page and Mark Cancian's authored commentaries
A GAO or DoD Inspector General report specifically addressing munitions accounting or Iran-war cost reconciliation
This would be the first non-anonymous-sourced, non-think-tank validation of either the stockpile percentages or the war-cost range
Where to watch: GAO.gov reports database and DoD Inspector General published reports
Congressional Armed Services Committee testimony or hearing transcripts addressing the Iran-war supplemental appropriation's final scope and line-item breakdown
This would fix an authoritative, legislatively-adopted cost figure even if broader economic-impact estimates remain contested
Where to watch: Senate and House Armed Services Committee and Appropriations Committee hearing records
A new unclassified CSIS, RAND, or CNAS Taiwan wargame report published after the Iran war's active phase
This would be the first opportunity to see whether Iran-war depletion has been incorporated into open-source Indo-Pacific wargame assumptions, closing part of the demand-side unknown
Where to watch: CSIS Wargaming and Scenarios research page and equivalent RAND/CNAS publication trackers
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. inventory of Patriot (PAC-3/GEM-T) and THAAD interceptors as of today, September 2026? — Patriot stocks are reduced somewhere between roughly 50 and 65 percent from the ~2,200 pre-war baseline (reconciling CSIS's ~827-remaining figure against CNN's 'roughly half' figure), and THAAD stocks are reduced by 75-80+ percent from the 452 pre-war baseline, based on triangulation of two independently-sourced reports from the same reporting window.
This determines whether the U.S. has adequate air and missile defense stock for a simultaneous or near-simultaneous Indo-Pacific contingency, and whether Gulf allies' own THAAD/Patriot resupply requests can be met without further drawing down U.S. reserves.
○ REPORTEDopaque by design · Medium confidence
What is the true total cost of the 2026 Iran war to the United States? — Direct military costs plausibly fall in the $29-42 billion range based on the Pentagon's own testimony and CSIS's preliminary estimate, while total supplemental funding requests span $80-87.6 billion and outside total-economic-impact trackers claim $113-132 billion or more — these are not competing answers to the same question but different questions being asked with the same dollar sign.
This shapes Congressional appetite for the pending supplemental appropriation, sets the political benchmark against which any future Iran or Taiwan contingency cost will be judged, and determines how much of the FY2026/FY2027 defense topline is effectively already committed to Iran-war replenishment rather than Indo-Pacific modernization.
◑ CONTESTEDanswerable by named party · Low confidence
How much of the pre-Iran-war U.S. precision-munitions stockpile (Tomahawk, JASSM-ER, PrSM, SM-3, Patriot, THAAD) remains available for a Taiwan Strait contingency today? — Given reported Army use of 'virtually all' PrSM/ATACMS stock and reported Tomahawk expenditure exceeding 1,000 missiles against an annual procurement rate roughly a tenth of that pace, near-term Taiwan-scenario stock for at least these two munition types is materially degraded from pre-war levels, but no defensible read exists for the aggregate multi-system picture.
This is the single most consequential unknown for Indo-Pacific force planning, since it determines whether current stated intent to prioritize the Pacific is backed by actual inventory or is aspirational given multi-year replenishment timelines.
— INFERREDirreducible for now · Low confidence
What quantity and mix of munitions would a Taiwan Strait contingency actually require, and over what timeframe? — Open-source wargaming (both CSIS's invasion and blockade projects) consistently finds rapid attrition and 'quick exhaustion of long-range precision-guided munitions' as a qualitative finding, but no public wargame report translates this into a quantified munitions-expenditure total comparable to the dollar or unit figures reported for the Iran war.
This is the denominator against which any stockpile-adequacy judgment must be made; without it, statements that the U.S. is 'ready' or 'not ready' for Taiwan given current post-Iran-war stocks are unfalsifiable.
✓ DOCUMENTEDopaque by design · Medium confidence
Are the CSIS-cited depletion percentages (roughly half of Patriots, ~80 percent of THAAD) an accurate reflection of the classified internal Pentagon inventory report, or an independent estimate that merely resembles it? — Multiple independently-reported figures (CSIS's own count, CNN's percentage-based sourcing, and the earlier April 2026 New York Times report citing classified documents showing JASSM-ER inventory near 1,500 of a 2,600 baseline) converge on the same direction and rough order of magnitude, which is the strongest available corroboration in an opaque data environment.
If CSIS's model-derived estimates diverge meaningfully from the true classified baseline, every downstream policy argument built on 'the U.S. has X percent of its interceptors left' is built on an unverified proxy rather than ground truth.
○ REPORTEDanswerable by named party · Medium confidence
Will accelerated munitions production actually close the Patriot/THAAD/PrSM inventory gap before a Taiwan contingency could plausibly arise? — Even with announced production acceleration ('quadruple production' directive) and a $22.9 billion Tomahawk contract reportedly signed, CSIS's own estimate is that restoring depleted stockpiles to pre-war levels will take multiple years at best, meaning any near-term Taiwan contingency would draw on a still-degraded base.
This determines whether current force-posture statements about Pacific prioritization are credible on a multi-year timeline or represent an inventory the U.S. does not yet, and may not soon, possess.
○ REPORTEDresolves on date · Medium confidence
Does the disclosed $29-42 billion direct-cost range for the Iran war include or exclude pre-conflict force buildup, base repair, and personnel/readiness costs? — Reporting indicates the $29-42 billion range covers direct military action (munitions, destroyed equipment, base damage) but explicitly excludes standing operational costs already budgeted in the FY2026 defense budget and, per one CNN source, excludes facility repair and regional basing costs — meaning the 'headline' cost figures are narrower in scope than casual comparison to the $80-132 billion figures suggests.
Whether Congress and outside trackers are comparing like-for-like figures when debating supplemental appropriations, and whether the eventual total bill significantly exceeds current public expectations.
○ REPORTEDcheckable now · Medium confidence
The U.S. can rapidly pivot forces and munitions stocks back to the Indo-Pacific once the Iran war formally ends, restoring the pre-war deterrence posture against a Taiwan contingency.
CSIS's own commentary states that while naval assets will physically return to the Pacific once operations end, restoring depleted munitions stockpiles to desired inventory levels 'will take many years' — the return of platforms and the return of usable stock are not the same event, and reporting does not establish any accelerated timeline that shortens this gap.
— TREATED AS WORKING ASSUMPTIONif wrong: Force-posture statements asserting Pacific-first prioritization could be materially unsupported by actual available inventory for a multi-year window, undermining deterrence credibility if China perceives the gap.
The Pentagon's disclosed war-cost figures ($29-42 billion range) represent a comprehensive accounting of the conflict's cost to the U.S. government.
Reporting explicitly states these figures exclude standing operational costs already in the FY2026 budget, and one source told CNN the pending $80 billion supplemental request's Iran-specific portion excludes facility repair and regional basing costs — meaning the disclosed 'cost of the war' figures are narrower than the full fiscal exposure.
— TREATED AS WORKING ASSUMPTIONif wrong: Congressional appropriators and outside analysts debating supplemental funding may be working from an artificially low cost baseline, understating the true fiscal commitment already made.
CSIS's published stockpile estimates are functionally equivalent to the classified Pentagon inventory data.
CNN's sourcing states only that officials familiar with the matter called the CSIS estimates 'close to' internal figures — this is a qualitative endorsement by anonymous sources, not a validated data match, and CSIS's own methodology relies on DOD-projected delivery schedules and estimated (not confirmed) combat expenditure rates.
— TREATED AS WORKING ASSUMPTIONif wrong: Analysts citing precise CSIS figures (e.g., '827 Patriots remaining') as if they were confirmed government numbers would be overstating the precision of what is, by CSIS's own methodology notes, a modeled estimate.
Publicly available Taiwan wargame findings (CSIS blockade and invasion projects) provide a reasonable proxy for actual U.S. Indo-Pacific Command munitions-requirement planning.
CSIS's own reports explicitly distinguish their unclassified wargames from classified Pentagon wargames, stating the latter's 'assumptions and even results are not transparent to the public' — the unclassified projects are avowedly a substitute for, not a window into, actual classified planning.
— TREATED AS WORKING ASSUMPTIONif wrong: Public debate over whether current U.S. stocks are 'adequate' for Taiwan contingencies may be anchored to a scenario construct that differs materially from what actual U.S. combatant command planning assumes.
CSIS estimated a pre-Iran-war U.S. stockpile of approximately 2,200 Patriot interceptors (of the platform's two most-modernized variants) and 452 THAAD interceptors, per its published commentary cited by CNN in August 2026.
This is the only published baseline figure against which depletion percentages can be measured, and it comes from a single think tank's estimate rather than an official DoD release.
✓ GROUNDED
A CSIS report dated in the CNN July 29, 2026 article estimated fewer than 827 Patriot interceptors and fewer than 278 THAAD interceptors remained in U.S. stockpiles, with three sources familiar with Pentagon data telling CNN the CSIS estimates were close to internal government figures.
This is the closest the public record comes to independent corroboration of classified stockpile levels, anchoring the depletion range analysts can responsibly cite.
✓ GROUNDED
CNN reported in August 2026, citing multiple sources familiar with internal Pentagon inventory reports, that the U.S. had used roughly half its Patriot interceptors and nearly 80 percent of its THAAD interceptors since the war began, and that the Army had used 'virtually all' of its long-range precision surface-to-air weapons.
This figure conflicts with CSIS's own percentage-implied depletion from f1/f2, illustrating that even well-sourced reporting on the same classified data set produces different depletion percentages depending on date and source chain.
✓ GROUNDED
Pentagon-disclosed war-cost figures moved from $11.3 billion for the first six days (per closed-door Congressional briefing reported in March 2026), to $29 billion per Acting Pentagon Comptroller testimony to the House Armed Services Committee on May 12, 2026, to a CSIS preliminary estimate near $40 billion for direct military costs by June 2026 — while the Pentagon separately submitted supplemental requests ranging from $80 billion to $87.6 billion depending on the source.
This demonstrates the cost figure is not a single classified number awaiting disclosure but a genuinely moving and inconsistently-scoped target even in official testimony.
✓ GROUNDED
CSIS's own published Taiwan wargame reports (the 26-iteration blockade project and prior invasion-scenario work) explicitly state that classified Pentagon wargames' assumptions and results 'are not transparent to the public,' and that even unclassified CSIS wargames report military losses and import-disruption levels rather than a comparable total-munitions-expenditure figure.
This establishes that no public Taiwan-contingency 'demand' figure analogous to the Iran-war cost/expenditure figures exists at all — the demand side of the comparison is qualitative, not quantitative.
✓ GROUNDED
CSIS commentary states that once Operation Epic Fury ends, naval assets sent to the Middle East will return to the Pacific, but that restoring depleted munitions stockpiles to desired inventory levels 'will take many years.'
This directly links the Iran-war depletion unknown to Taiwan-readiness planning, establishing the multi-year recovery timeline as the operative constraint rather than any single stockpile number.
✓ GROUNDED