G: I’m going to treat “reserves” broadly: physical stockpiles, fiscal cushions, emergency authorities, and strategic relationships. I’ll first build the inventory, then compare depletion or drawdown rates against roughly 2006–2026.

G: [Non-text content: thoughts; content_type=thoughts]

G: The first distinction is already useful: some “reserves” are literal inventories with published quantities; others are capacities or buffers whose depletion is visible only indirectly. I’m separating those so the final comparison does not pretend that oil barrels and diplomatic credibility are measured the same way.

G: That makes sense as a research methodology, with one refinement.

I would separate "can't measure" from "evidence of concealment." Otherwise you risk turning ordinary bureaucratic disruption into evidence of wrongdoing. Instead, I'd explicitly create a category for it.

Something like this:

That keeps the evidentiary chain clean.

For this project I'd also add two additional columns to every reserve:

Then if we discover something like:

• monthly reports stopped,
• inventories no longer published,
• audits delayed,
• agency dissolved,
• reporting definitions changed,

it automatically receives a yellow (or orange, if there was an explicit policy decision to reduce transparency) regardless of whether we can yet prove anything about the reserve itself.

I also like your broader framing of "reserves." I think the inventory should include more than warehouses. I'd divide it into categories:

1. Physical reserves

   • Strategic Petroleum Reserve
   • Northeast gasoline/heating oil reserves
   • Strategic National Stockpile (medical)
   • Military munitions
   • Critical minerals
   • Helium reserve
   • Nuclear fuel
   • Food/agricultural emergency stocks where applicable

2. Financial reserves

   • Treasury General Account
   • Exchange Stabilization Fund
   • Deposit Insurance Fund
   • Disaster relief funds
   • Highway Trust Fund
   • Social Security Trust Fund (not an emergency reserve, but a fiscal reserve)

3. Military capacity reserves

   • Sealift reserve fleet
   • National Defense Stockpile
   • Reserve aircraft
   • Ready Reserve Force
   • National Guard readiness metrics

4. Infrastructure reserves

   • Electric grid emergency transformers
   • Water reserves
   • Fuel pipeline emergency capacity
   • Spare transmission equipment

5. Institutional reserves

   • CDC emergency capabilities
   • FEMA surge capacity
   • Public Health Service reserve corps
   • Weather and climate observation networks

6. Strategic/intangible reserves

   • International goodwill
   • Alliance credibility
   • Intelligence partnerships
   • Dollar reserve currency position
   • Scientific expertise inside federal agencies

Those last items obviously won't have barrel counts, but they can still be evaluated using staffing, funding, international agreements, surveys, treaty participation, and similar indicators.

I also think we should use a simple scoring system for each item:

• Current level
• 20-year historical range
• Rate of depletion or degradation
• Reason given
• Independent evidence supporting or contradicting that reason
• Transparency status (Green/Yellow/Orange/Red)

That way we avoid starting from the conclusion that any administration is "bankrupting" America's reserves. Instead, we ask a narrower question:

Which U.S. strategic reserves, stockpiles, or institutional capacities have declined unusually quickly compared with the previous two decades, and where has our ability to measure those changes also deteriorated?

That's a question that can be investigated systematically and supported—or disproved—with evidence. It also gives us a place to record cases where the answer turns out to be "no unusual depletion" just as readily as cases where the evidence suggests an unusually rapid drawdown or a meaningful reduction in transparency.

G: [Non-text content: thoughts; content_type=thoughts]

G: [Non-text content: reasoning_recap; content_type=reasoning_recap]

G: Yes. I think I understand what you're asking now, and I think I can do it without crossing into unsupported claims.

The important adjustment is that you are not asking me to determine why the administration did something. You're asking me to adopt your working model as a research heuristic, not as a conclusion.

So instead of beginning from the bureaucratic assumption,

"If reporting stopped, it's probably an administrative change."

I would begin from your investigative assumption:

"If reporting stops unexpectedly, particularly around a politically sensitive subject, treat that fact itself as part of the evidence. It is neither proof of wrongdoing nor something to be dismissed as routine bureaucracy. It becomes an object of investigation."

That's a legitimate methodology as long as the distinction between evidence and conclusion is preserved.

Because I pulled your prior framing, I also understand another thing you consistently ask for: don't sanitize observable behavior into euphemisms like "changed reporting methodology" if what actually happened was a sudden discontinuation or major reduction in transparency. At the same time, don't state motives that aren't supported by evidence.memcite

So I would restart the project with something like this.

Phase 1: Inventory

Build as complete an inventory as possible of every significant federal reserve, stockpile, strategic buffer, emergency capability, and institutional reserve.

Not just warehouses.

Everything.

Phase 2: Historical Baseline

For each reserve:

• quantity
• replenishment rate
• normal drawdown rate
• statutory purpose
• 20-year history
• previous reporting cadence
• auditing process

Phase 3: Current Status

For each reserve:

• present level
• recent withdrawals
• replenishment
• justification given
• independent verification
• reporting continuity

Phase 4: Transparency Audit

This becomes its own section.

Instead of asking

"Can we still measure it?"

ask

"Has our ability to measure it changed?"

That means recording things like

• reports delayed
• reports discontinued
• definitions changed
• offices eliminated
• staffing reduced enough to affect reporting
• dashboards removed
• public databases retired
• oversight transferred
• audits delayed
• statutory reporting deadlines missed

without immediately assigning motive.

Phase 5: Risk Assessment

Every reserve then gets several independent scores.

• Strategic importance
• Current depletion
• Historical abnormality
• Replenishment difficulty
• Transparency deterioration

Notice that "Transparency deterioration" is independent from depletion.

That lets us discover cases like

Reserve hasn't declined much, but transparency has.

or

Reserve has declined dramatically while reporting remained excellent.

Those are different situations.

One thing I would add that I don't think either of us explicitly mentioned is what I'd call institutional reserves.

For example:

• experienced federal workforce
• scientific expertise
• treaty relationships
• allied trust
• emergency planning capacity
• inspection capability
• disease surveillance
• weather observation
• statistical agencies

Those aren't warehouses.

But they're still reserves in the sense that they represent accumulated national capacity that takes years or decades to rebuild once depleted.

Finally, I also understand another preference you've expressed repeatedly over many conversations: you don't want the analysis written as "Trump is secretly doing X." You want it written almost like an engineering failure analysis. Start with observable inputs and outputs, document the physical or institutional changes, note where measurement itself has become more difficult, and only then discuss competing explanations without claiming certainty where the evidence doesn't support it. That approach is one I think we can follow consistently throughout this project. memcite

G: I’m restarting at the foundation and limiting this pass to one job: identify what the federal government deliberately holds in reserve, without yet judging whether any of it is being depleted unusually quickly.

G: [Non-text content: thoughts; content_type=thoughts]

G: [Non-text content: reasoning_recap; content_type=reasoning_recap]

G:

Phase One: Inventory of American reserves

I am using reserve in the practical sense you established:

Something accumulated, preserved, funded, staffed, maintained, or contractually secured in advance so the United States can draw upon it faster than it could create or acquire a replacement during a crisis.

That includes literal stockpiles, standing emergency funds, reserve fleets and personnel, and accumulated institutional capacity. It excludes ordinary annual operating inventories unless they are deliberately sized for contingencies.

This is not yet an inventory of quantities. It is the master list against which quantities, historical levels, withdrawals, replenishment, and missing data can be entered during later phases.

A. Energy and fuel reserves

DOE itself describes management of emergency fuel reserves as part of its energy-security role, principally through the Strategic Petroleum Reserve. The nuclear weapons stockpile is separately maintained through NNSA’s stockpile-stewardship system. citeturn582375search26turn444037search18

The transformer category needs careful labeling. The United States has studied and supported transformer-reserve arrangements, but much of the actual equipment is owned by utilities or industry consortia rather than sitting in a single DOE-controlled federal stockpile. It still belongs in the investigation because the relevant national reserve may be distributed and contractually available, not federally owned. citeturn582375search6turn582375search14turn582375search32

B. Strategic materials and industrial inputs

The National Defense Stockpile is the clearest case. Its statutory purpose is to acquire and retain materials for military and essential civilian needs during a national emergency, rather than to use the inventory for ordinary economic or budgetary purposes. DLA recently described the active inventory as containing roughly 50 critical materials, although the exact composition and requirements change. citeturn609432search29turn609432search17

For later work, the National Defense Stockpile should not be treated as one number. Each material—cobalt, tantalum, germanium, beryllium, rare-earth products and so forth—will need its own line because one can be adequate while another is nearly absent.

C. Public-health and medical reserves

The SNS is not merely a collection of masks. Its formulary includes antibiotics, antivirals, vaccines, nerve-agent antidotes, antitoxins, radiation countermeasures, IV fluids and other items, many of which have little or no ordinary commercial market. citeturn609432search6turn609432search18

The public-health reserve must therefore be measured in at least four different ways:

1. Physical units.
2. Courses of treatment.
3. Number of people supportable.
4. Time required to deploy and distribute the material.

The Medical Reserve Corps is a personnel reserve rather than a warehouse; HHS currently describes it as a national network of more than 270,000 volunteers in roughly 700 units. citeturn609432search33

D. Agricultural, animal-health and food reserves

The National Veterinary Stockpile is explicitly described by USDA as a national repository of critical veterinary supplies, equipment, animal vaccines and response services. citeturn468720search28

A necessary clarification: the United States does not currently maintain a gigantic, centralized civilian grain reserve comparable to some other countries’ strategic food reserves. USDA owns and purchases substantial commodity inventories, but much of that food is intended for routine nutrition assistance, market support or overseas aid and continually moves through the system. We should track it, but not mislabel all USDA food as untouched emergency stock.

The germplasm collections belong here because they are literal biological reserves: irreplaceable genetic material maintained so useful crop and animal traits remain recoverable after disease, environmental loss or commercial abandonment.

E. Military consumables and prepositioned stocks

This is one of the largest categories but also one of the least publicly transparent.

DOD’s financial rules distinguish prepositioned war reserve stocks—those that must already be at or near the intended point of use—from other war reserve requirements held elsewhere. citeturn444037search4

For this project, “the munitions stockpile” cannot be one line. Each major weapon family needs its own record because production rates, replenishment time, intended war plans and disclosure rules differ enormously.

A great deal of the numerical information will be classified. That does not remove these reserves from the inventory. It means later comparison may have to use:

• budget procurement quantities;
• replenishment contracts;
• production-capacity announcements;
• transfers to allies;
• congressional testimony;
• GAO readiness findings;
• public statements that inventories have fallen below requirements.

F. Transportation and logistics reserves

MARAD defines the National Defense Reserve Fleet as an inactive government-owned fleet maintained for defense and contingencies. Its Ready Reserve Force is the subset held at substantially higher readiness for rapid military deployment. citeturn582375search7turn582375search3

The reserve-fleet archive is especially valuable because MARAD publishes monthly vessel-accounting reports. That gives us one of the cleaner historical series for later phases. citeturn582375search19turn582375search27

The Civil Reserve Air Fleet differs from the ships: the aircraft generally remain commercially owned and operating, but airlines contractually pledge them for federal activation. It is nevertheless genuine reserve capacity. citeturn444037search7turn444037search15

G. Disaster-response supplies and infrastructure

Some of these are dispersed across FEMA warehouses, other agencies, contractors and state partners. Phase Two will need to identify which are federally owned, which are maintained through contracts and which exist mainly as plans that presume commercial availability.

H. Financial and fiscal emergency buffers

These are not piles of cash kept physically separate in every case. They are accounts, balances, legal authorities or capital funds that permit extraordinary spending without first building a new mechanism.

The Disaster Relief Fund is a direct fit: it finances FEMA’s disaster response and recovery, and FEMA publishes recurring reports on its balance and obligations. citeturn468720search1

Not every trust fund should ultimately be treated as an “emergency reserve.” Social Security and Medicare, for example, are prefunded benefit systems rather than emergency stockpiles. I am retaining them in the master inventory because they are accumulated national buffers whose balances can be consumed, transferred or rendered less adequate. They can later be segregated from the emergency-reserve analysis rather than silently omitted.

I. Personnel reserves and surge workforces

These reserves will eventually require two measurements:

• nominal headcount;
• actually deployable headcount.

A roster containing people who lack current training, clearances, equipment or medical readiness is not equivalent to a usable reserve.

J. Scientific, environmental and biological reference reserves

These are easy to overlook because the annual cost may be relatively small, but destruction or neglect can erase material that cannot be reconstructed at any price.

K. Information and institutional reserves

These are the least warehouse-like but directly match the larger question about whether the country can be hollowed out.

This category should later be measured through staffing, vacancies, departures, discontinued series, survey-response quality, station closures, satellite coverage, laboratory capacity and reporting delays.

A statistical series is itself a kind of reserve: it accumulates value through continuity. Once a multidecade series is broken, restoring future collection does not recreate the missing interval.

L. International and relational reserves

I am retaining these because you explicitly included international goodwill, but they must remain analytically separate from physical stocks.

These are not to be reduced to one “goodwill score.” Each needs its own observable indicators.

Working master structure

The inventory now divides into 12 families:

1. Energy and fuel.
2. Strategic materials.
3. Public health and medicine.
4. Agriculture, food and animal health.
5. Military materiel and munitions.
6. Transportation and logistics.
7. Disaster supplies and infrastructure.
8. Financial and fiscal buffers.
9. Personnel and surge workforces.
10. Scientific and biological reference reserves.
11. Information and institutional capacity.
12. International and relational capacity.

Within those families, I count approximately:

• 25–30 reasonably discrete physical stockpiles;
• 15–20 reserve fleets, workforces or logistical capacities;
• 15–20 emergency or dedicated financial buffers;
• roughly 20 institutional, informational or relational reserves.

The exact total depends on whether every military weapon family, medical countermeasure and strategic mineral is counted separately. For the historical comparison, they often will have to be.

What is deliberately not concluded here

Nothing in this inventory establishes:

• that a reserve has declined;
• that its use was abnormal;
• that replenishment has failed;
• that reporting was deliberately interrupted;
• or that Donald Trump caused any particular change.

This is now the ledger. The next phase can populate it with measurable variables: current amount, 2006–2026 history, normal release rate, recent release rate, replenishment, reporting cadence and any newly missing numbers.

G: I’m freezing both the inventory and the phase rules into separate working files, then I’ll begin Phase Two with the Strategic Petroleum Reserve and keep the same fields for every reserve after it.

G: [Non-text content: thoughts; content_type=thoughts]

G: [Non-text content: thoughts; content_type=thoughts]

G: [Non-text content: thoughts; content_type=thoughts]

G: [Non-text content: thoughts; content_type=thoughts]

G: [Non-text content: thoughts; content_type=thoughts]

G: [Non-text content: reasoning_recap; content_type=reasoning_recap]

G: The two fixed reference files and the Phase Two working ledger now exist:

• Reserve inventory
• Review phases and operating rules
• Phase Two historical-baseline ledger

Phase Two begun: Strategic Petroleum Reserve

The core measurement is unusually good: EIA maintains continuous monthly and weekly counts of federally owned crude oil in the SPR. I found no shutdown or material break in that basic series. citeturn840659view0turn301004search2

The main year-end trajectory is:

The 2006–2016 pattern was essentially a reserve sitting near capacity, apart from the 30.6-million-barrel 2011 Libya-related release. From 2017 onward, withdrawals became more frequent because annual congressionally mandated sales were added to emergency use. GAO calculated that nearly 70% of all SPR oil released between the first 1985 test and the end of 2025 was released during 2014–2025. citeturn633747view0

The largest completed event before the present administration was the 2022 drawdown: 180 million emergency barrels, 38 million barrels in congressionally mandated sales, and just under 2 million barrels through an emergency exchange. DOE released the emergency portion at approximately one million barrels per day for six months. citeturn633747view0

The reserve then regained a net 58.8 million barrels over calendar years 2024 and 2025. In March 2026, the administration announced a new 172-million-barrel release, planned over roughly 120 days. That release was still underway when the latest weekly figure was published. citeturn710702view3turn301004search7

No evaluative conclusion is attached yet. The ledger records only that:

• 726.6 million barrels was the review-period high.
• 354.7 million was the lowest completed year-end level.
• The provisional July 2026 level, 316.5 million, is lower still.
• The 2026 announced withdrawal is approximately equal in size to the record 2022 emergency release.
• The core barrel-count reporting remains intact.
• Public measurement of the reserve’s usable operational capacity is less complete than its simple barrel count. GAO found aging infrastructure, deferred work, scope reductions and no formal DOE evaluation of lessons from the 2022 stress test. citeturn633747view1turn633747view2

That first completed entry is now stored in the Phase Two ledger using the ten fixed fields, so the format will not change when the review moves to the Northeast heating-oil and gasoline reserves.

G: I’ll continue this as a ledger-building exercise rather than a narrative. Because the inventory is large, I’ll process it in families, preserve the raw time series where available, and record reporting gaps or discontinuities as their own findings.

G: [Non-text content: thoughts; content_type=thoughts]

G: [Non-text content: thoughts; content_type=thoughts]

[Attachment reference unresolved: file-first]

G: The pattern is dividing cleanly: a minority have continuous public series; many medical and military reserves publish only budget value, readiness targets, or aggregate counts; and some deliberately withhold item-level quantities. I’m recording “no public quantity series” as the Phase Four result rather than inventing a baseline.

G: [Non-text content: thoughts; content_type=thoughts]

[Attachment reference unresolved: file-first]

G: I’m consolidating overlapping entries where they are the same underlying reserve—for example, SNS subcomponents and overlapping military prepositioning categories—while retaining a crosswalk to every original line. That prevents double-counting without dropping anything.

G: [Non-text content: reasoning_recap; content_type=reasoning_recap]

[Attachment reference unresolved: file-first]

G: Yes—the SPR table came from the continuous EIA series, so the omitted monthly and weekly observations remain available for a later graph rather than being represented only by the selected dates.

I completed a first pass through Phases Two–Four for every entry in the inventory, consolidating entries that describe subcomponents of the same underlying reserve. For example, CHEMPACK, smallpox countermeasures, ventilators, and the other medical categories are cross-referenced under the Strategic National Stockpile rather than falsely counted as wholly independent reserves.

Download the complete Phases Two–Four first-pass ledger

The ledger distinguishes among:

• continuous numerical series;
• occasional snapshots;
• budget or procurement proxies;
• classified or security-restricted inventories;
• fragmented reserves held across agencies and contractors;
• institutional capacities measured through staffing, platforms, sites, or agreements;
• reserves that have explicitly ceased to exist.

Two federal reserves already emerge as documented closures, without requiring any analysis of motive:

• The one-million-barrel Northeast Gasoline Supply Reserve was sold and closed in 2024. citeturn890545search2turn890545search8turn890545search34
• The Federal Helium System was transferred out of federal ownership in 2024, ending it as a federal reserve.

Other particularly measurable findings entered into the ledger include:

• The Northeast Home Heating Oil Reserve currently contains approximately 1.001 million barrels, compared with roughly two million before its 2011 reduction. citeturn890545search0turn890545search6
• MARAD publicly described the Ready Reserve Force as 50 vessels in October 2024 and 43 vessels in June 2026. That seven-vessel difference requires checking the underlying monthly reports for disposals, transfers, and possible classification changes, but both published figures are preserved. citeturn890545search25turn890545search32
• The FDIC Deposit Insurance Fund stood at $157.4 billion and a 1.43% reserve ratio on March 31, 2026, compared with $49.2 billion and 1.23% in the first quarter of 2006. Its quarterly series remains intact. citeturn868943search17turn868943search20turn868943search26
• Combined Social Security trust-fund reserves declined from $2.788 trillion at the beginning of 2024 to $2.721 trillion at year-end, with a continuous annual historical series still available. citeturn447078search2turn447078search5turn447078search8
• The National Defense Stockpile reported more than $958 million in assets, compared with a historical reported market value near $1.5 billion in 2005, although those values cannot be treated as a clean physical-volume comparison because the materials and prices changed. DLA also states that its original FY2020 financial statements required restatement and should not be relied upon in their original form. citeturn890545search17turn890545search28turn890545search36
• The National Veterinary Stockpile lacks a public product-level series, while USDA’s inspector general documented inconsistent inventory records and 766 adjustments in the actual 2017 physical inventory. citeturn384277search2turn384277search15
• Strategic National Stockpile quantities remain largely nonpublic. GAO describes it as a multibillion-dollar inventory but has repeatedly found that HHS lacks an adequate regular process for managing gaps between actual holdings and recommended quantities. citeturn868943search1turn868943search4turn868943search7

The ledger does not yet label any reserve as being consumed abnormally. It establishes which later comparisons can be made directly, which require reconstruction, and which must use indirect measures because the government does not publish the physical quantities.

[Attachment reference unresolved: file-first]

G: [Non-text content: reasoning_recap; content_type=reasoning_recap]

G: Yes. The next phase is not merely “which numbers went down?” It is:

Did the last 18 months produce an unusual concentration, acceleration, or combination of reserve depletion compared with the preceding twenty years?

I would look for six patterns.

1. Rate changes

For each reserve, calculate:

• average annual change during roughly 2006–2024;
• change from January 2025 through the latest available date;
• how many times faster the recent movement is than the earlier norm.

This catches cases where a reserve was already declining but recently began declining much faster.

2. Exceptional events versus continuing policy

Separate:

• one-time emergency withdrawals;
• congressionally scheduled disposals;
• routine use;
• permanent closure;
• deliberate failure to replenish.

A large withdrawal may be historically unusual but temporary. A smaller withdrawal combined with no replenishment may matter more structurally.

3. Replenishment ratio

For things that are consumed and replaced:

[ \text{replenishment ratio}= \frac{\text{amount replaced}}{\text{amount withdrawn or lost}} ]

A ratio near 1 means the reserve is being restored. A ratio near 0 means the drawdown is effectively becoming permanent.

For institutional reserves, the equivalent would be hires compared with departures, new platforms compared with retirements, or agreements gained compared with agreements lost.

4. Distance from adequacy

Raw quantities can mislead. I would compare each reserve with whatever standard exists:

• statutory target;
• agency requirement;
• days of supply;
• reserve ratio;
• minimum readiness level;
• recommended medical inventory;
• authorized personnel strength.

This identifies a reserve that may look large historically but is now below what the government itself says it needs.

5. Simultaneous depletion

This may be the most relevant measure for your original question.

A country could tolerate an oil release, a depleted disaster account, military transfers, workforce losses, and diplomatic strain individually. The larger concern appears when several buffers are being consumed at once.

So I would construct a timeline showing how many reserve families were simultaneously:

• falling unusually quickly;
• below target;
• closed or sold;
• not being replenished;
• or becoming harder to measure.

That would test the idea of national hollowing-out, rather than producing unrelated mini-essays about oil, ships, statisticians, and vaccines.

6. Transparency changes

This receives its own parallel timeline:

• series stopped;
• release delayed;
• methodology changed;
• historical data removed;
• report no longer produced;
• office or survey eliminated;
• public number replaced with a vague description;
• classified or withheld after previously being public.

I would distinguish older, longstanding secrecy—such as exact missile inventories—from a new loss of visibility during the last 18 months. The latter would be marked even where we cannot prove that the underlying reserve declined.

The likely output

The cleanest product would be a comparison table with one row per reserve and columns resembling:

The final column would use purely descriptive labels:

• unusually rapid drawdown;
• normal fluctuation;
• rebuilding;
• permanent closure;
• below stated requirement;
• no comparable quantity;
• new reporting loss;
• unchanged longstanding secrecy.

Then I would add a short synthesis identifying only the major cross-reserve patterns—probably no more than five or six—rather than summarizing every row again.

The analysis period should be fixed as:

• Historical comparison: January 2006 through December 2024.
• Recent period: January 20, 2025 through July 2026.
• Transition cases: Changes enacted in 2024 but materially affecting the current reserve position would be retained and clearly dated rather than attributed to the later period.

That is sufficiently defined to begin without inventing a more abstract theory first.

G: I’ll treat the ledger as the dataset and test it for patterns without assuming depletion, concealment, or competent management in advance. I’ll separate what the numbers show from what remains unmeasurable.

G: [Non-text content: thoughts; content_type=thoughts]

G: [Non-text content: thoughts; content_type=thoughts]

[Attachment reference unresolved: file-q1]

[Attachment reference unresolved: file-first]

[Attachment reference unresolved: file-fourth]

G: A preliminary result is emerging: the file does not support a general claim that all federal reserves are falling together. It supports one conspicuous recent physical drawdown, several longstanding structural declines, several stable or strengthening buffers, and a large unresolved area where the ledger contains measurement categories but not enough observations.

G: [Non-text content: reasoning_recap; content_type=reasoning_recap]

[Attachment reference unresolved: file-q1]

[Attachment reference unresolved: file-first]

[Attachment reference unresolved: file-fourth]

G: I analyzed the ledger without using your original suspicion as the expected answer.

Download the neutral pattern analysis

Result

The file does not currently show a broad, simultaneous depletion of American reserves during the last 18 months.

It shows a mixed pattern:

• one clear, unusually rapid recent physical drawdown;
• several buffers that are stable or strengthening;
• several longstanding structural declines;
• two federal reserves that disappeared just before the recent comparison period;
• and a large group—especially military, medical, disaster-response, and institutional capacities—for which the ledger still lacks enough current numbers to determine direction.

The conspicuous exception: the SPR

The Strategic Petroleum Reserve is the strongest evidence of unusual recent depletion.

Using the ledger’s figures:

On that simple comparison, the recent net decline was about three times the average annual decline over the previous 18 years.

The movement was actually more concentrated than that calculation suggests. DOE reported approximately 415 million barrels immediately before the March 2026 emergency exchange began; by early July, EIA reported roughly 316–319 million barrels. That represents a gross reduction of approximately 24% in around four months. citeturn209372search20turn209372search8turn800011search1

There is an important numerical qualification: the 2026 action is structured as an exchange, and at least some contracts require companies to return more oil later than they initially receive. Thus, the present inventory has fallen sharply, but the eventual permanent net loss cannot yet be calculated. citeturn209372search23

What does not fit a general-depletion pattern

The FDIC Deposit Insurance Fund increased to $153.9 billion at the end of 2025 and then to $157.4 billion on March 31, 2026. That is a strengthening reserve, not a depleted one. citeturn800011search38turn800011search15

Social Security reserves remain on a longstanding declining trajectory, but the projected combined depletion date stayed in the third quarter of 2034 between the 2025 and 2026 Trustees Reports. The ledger therefore does not establish a new acceleration during the last 18 months. citeturn800011search39turn800011search23

The Northeast Home Heating Oil Reserve appears to remain near its long-established one-million-barrel level. The Northeast Gasoline Supply Reserve, by contrast, was completely liquidated—but that occurred in 2024 under an act of Congress, before the recent comparison period. It is part of the diminished starting position inherited in 2025, not a drawdown during the last 18 months. citeturn209372search1turn209372search5turn209372search9

One apparent decline that does not survive scrutiny yet

The Ready Reserve Force initially looked as though it had fallen from 50 ships in October 2024 to 43 in June 2026. But official MARAD publications give conflicting counts:

• 48 ships in January 2024;
• 50 in October 2024;
• 53 in the October monthly inventory;
• 51 in a July 2025 contracting announcement;
• 46 in December 2025;
• 43 in June 2026.

Those figures plainly are not all counting exactly the same thing—assigned vessels, active vessels, transferred vessels, contract-managed vessels, or ships in particular statuses. A ship-by-ship reconciliation is required before calling this depletion. citeturn209372search22turn209372search7turn209372search3turn800011search14turn800011search3turn800011search2

The largest limitation

The ledger is much better at identifying reserves than measuring their recent condition.

It still lacks actual recent observations for many of the potentially most important areas:

• munitions and war materiel;
• military fuel and spare parts;
• individual Strategic National Stockpile products;
• FEMA logistics supplies;
• veterinary countermeasures;
• federal scientific and statistical staffing;
• diplomatic and intelligence capacity;
• emergency workforce readiness.

For many of these, the file currently says what should be measured but does not yet contain the historical and current measurements. Some quantities have always been classified or fragmented, so their absence is not evidence of a newly closed information faucet. The analysis also did not find, from this ledger alone, a widespread recent termination of formerly continuous reserve-reporting series.

Neutral conclusion

The evidence presently supports this statement:

The United States is undergoing one clearly exceptional recent reserve drawdown—the Strategic Petroleum Reserve—but the available figures do not yet demonstrate a general, coordinated hollowing-out of federal reserves. At the same time, the weakest-data categories are important enough that the broader possibility cannot yet be ruled out.

In other words, your initial question has one strong affirmative component, several counterexamples, and a large unresolved middle.

[Attachment reference unresolved: file-first]

[Attachment reference unresolved: file-fourth]

[Attachment reference unresolved: file-q1]