April 6, 2023. University of Chicago. Jon Stewart is onstage with Kathleen Hicks, the Deputy Secretary of Defense, and he’s doing the thing he does — asking a question so obvious it embarrasses everyone in the room.
He points out that every major federal agency is required by law to pass a financial audit. Then he notes that there’s exactly one that hasn’t. Not one that’s struggling. Not one that’s a few years behind. One that has never, in its entire audit history, produced a clean set of books.
Hicks confirms it without blinking: “That would be ours.”
Stewart asks why.
“It’s probably a 10-year process,” she says.
He pushes back. If you give someone a billion dollars and they can’t tell you what happened to it, he says, most people would call that waste. Hicks pushes back harder. A failed audit, she says, “is not suggestive of waste, fraud, and abuse. That is completely false.”
Stewart’s response: “If I give you a billion dollars and you can’t tell me what happened to it, that to me is wasteful. That means you are not responsible.”
The audience laughs. The clip goes viral. Nothing changes.
That exchange was three years ago. Since then, the Pentagon has failed two more audits — the seventh and eighth in a row. The department’s own comptroller wrote in December 2025 that it “will not reach its goal of achieving a clean financial statement audit without a significant acceleration of its efforts.” The GAO has said the Pentagon will likely continue failing through 2028. Among the 24 major federal agencies covered by the Chief Financial Officers Act, the Department of Defense is the only one that has never obtained a clean audit opinion. Every single one of the others has passed. Every one except this one.
Four months after the eighth failure, the administration proposed a $1.5 trillion defense budget — a 44 percent increase over current spending, the largest single-year defense buildup since the Korean War.
The Pentagon can’t tell us what it did with $1 trillion. The ask is to give it $500 billion more.
The case for the buildup
Before getting to what’s wrong here, the buildup deserves its best argument. Because there is one.
The Ukraine war demonstrated, in real time, that modern warfare runs on mass-produced munitions, cheap drones, and industrial base capacity — not on exquisite, decade-in-development weapons platforms. The United States military-industrial base has spent thirty years optimizing in exactly the wrong direction. The F-35 program, designed to be the future of air combat, has a projected lifetime cost north of $2 trillion and has still not delivered a fully operational variant. Procurement is slow, expensive, and contractor-captured. If a serious adversary challenged the U.S. today in a sustained conventional conflict, the munitions inventory problem alone would be existential within weeks.
A massive procurement surge — if it goes toward the right things — is a legitimate national security argument. More Tomahawks, more PAC-3 interceptors, more industrial capacity to build what a real war actually consumes. Defense hawks on both sides of the aisle have been making this case for years, and the arithmetic behind it is sound.
The 44 percent increase is also, in isolation, less alarming than it sounds. Defense spending as a share of GDP has fallen from roughly 10 percent at the height of the Vietnam War to around 3 percent today. The Reagan buildup — which is the historical benchmark everyone is reaching for — topped out at approximately 15 percent in a single year. This would nearly triple that, but the U.S. is also running concurrent wars in Iran and facing a peer competitor in China that has been building at a sustained pace for a decade.
That’s the strongest version of the case. It’s a real case.
Here’s what it misses.
The enforcement gap
The CFO Act of 1990 requires every major federal agency to submit to an annual independent financial audit. It was a bipartisan reform, born out of post-Cold War accountability concerns. The law is clear. The requirement is not optional.
What the CFO Act does not do is attach any consequence to failure.
There is no funding withheld for a failed audit. No penalty assessed. No mechanism by which Congress’s explicit mandate — that federal agencies maintain auditable books — is actually enforced against the one agency that has never complied. The Pentagon fails its audit, issues a press release about “making progress,” and receives the next year’s appropriation.
For eight consecutive years, this is what has happened.
Two bipartisan bills — the Audit the Pentagon Act, introduced by Bernie Sanders and Chuck Grassley in the Senate, Barbara Lee and Michael Burgess in the House — have tried to fix this. The mechanism is simple: any Pentagon component that fails a clean audit returns 0.5 to 1 percent of its budget to the Treasury. Real money, but not ruinous. A genuine incentive where currently there is none.
Neither bill has passed. In two consecutive sessions of Congress, a bipartisan accountability measure that would apply a modest financial consequence for failing a legally required audit has not moved. The Senate and House Armed Services Committees, which have jurisdiction over this, have not advanced it.
The working hypothesis: The $1.5 trillion defense budget is only possible because the enforcement gap in the CFO Act of 1990 has never been closed. Congress has the legal authority and the political will — bipartisan, documented — to require accountability as a condition of appropriations. It has not exercised that authority. The $1.5 trillion proposal is the largest single-year exploitation of that gap in the gap’s 35-year history.
What’s being cut while we do this
To understand the scale of this moment, it helps to know where the money sits in the budget.
The total federal budget is roughly $7 trillion annually. About two-thirds of that is mandatory spending — Social Security, Medicare, Medicaid, programs that run on autopilot by statute. The remaining one-third, about $1.6 trillion in discretionary spending, is what Congress actually negotiates each year. Defense gets roughly half of all discretionary spending. With the $1.5 trillion proposal, defense would be consuming nearly the entirety of discretionary spending on its own — before a single dollar goes to anything else.
If enacted, the Committee for a Responsible Federal Budget estimates the proposal would add $5.8 trillion to the national debt over a decade.
While Congress has shown itself unable to find the money to pass the $1.5 trillion defense budget, it has also found itself unable to fund a number of other things. The government shut down for more than 40 days in fall 2025 — the longest shutdown in U.S. history — because the same Congress that unanimously supports giving the Pentagon more money couldn’t agree on funding for the rest of the government. During that shutdown, Head Start programs in Washington closed. Four hundred preschool children lost care. Seventy staff were laid off.
Emergency Housing Vouchers — a program that helped 60,000 households find and maintain stable housing — are running out of money in 2026. The Child Care Development Block Grant, which helps 1.4 million children access care each month, is flat-funded while costs have risen. The FY2026 budget eliminated the only dedicated federal funding stream for afterschool programs, serving 1.4 million children. Non-defense discretionary spending has fallen 14 percent in real terms since 2010, adjusted for inflation and population growth.
There is no money for afterschool programs. There is $1.5 trillion for a department that, three months ago, couldn’t pass its audit.
The oversight removal sequence
Here is what happened in the twelve months before the budget proposal arrived.
In September 2025, Secretary of Defense Pete Hegseth issued a memo directing the secretaries of the Army, Navy, and Air Force to overhaul the Pentagon Inspector General’s complaint intake process. The memo required “credibility assessments” to be completed within seven days of receiving a complaint, investigations closed within 30 days, and established procedures to manage what Hegseth termed “repeat complainants.” Legal experts called it a violation of the IG Reform and Whistleblower Protection Act of 1978, which establishes the Inspector General’s independence. One former senior IG investigator said the new timelines were designed “not to have any meaningful oversight by the DoD IG.”
Simultaneously, OMB withheld funding from CIGIE — the Council of the Inspectors General on Integrity and Efficiency, the body that supports and trains oversight personnel across the federal government.
Hegseth himself had, at that point, declined to sit for an interview with the Pentagon’s own IG, which was investigating his use of the Signal messaging app to share sensitive military strike information. He provided a written statement instead. The IG ultimately found that his messages “created a risk to operational security that could have resulted in failed U.S. mission objectives and potential harm to U.S. pilots.” Hegseth called it a “total exoneration.” The IG report said, in specific language, that he violated Pentagon policy on using personal devices for official business and violated federal recordkeeping law.
None of this is the mechanism. The mechanism is what it sets up.
The oversight infrastructure for the Department of Defense consists of three primary layers: the independent IG, the annual financial audit, and congressional oversight via the Armed Services Committees. In the twelve months before the largest defense budget request in history, two of those three layers were measurably weakened. The IG’s independence was constrained by executive directive. The financial audit remained structurally unenforceable — as it has been for 35 years. The Armed Services Committees, meanwhile, immediately endorsed the $1.5 trillion topline.
That left one layer fully intact: the appropriations process itself, with no strings attached.
What gets built in the dark
The piece flagged on the investor side of this is worth naming explicitly, though a full treatment belongs elsewhere.
In the weeks before the United States launched its initial attack on Iran in February 2026, a financial broker for Secretary Hegseth contacted BlackRock about making a multimillion-dollar investment in its iShares Defense Industrials Active ETF — a fund specifically designed to capture gains from increased government defense spending. The investment did not go through because the fund was not yet available for Morgan Stanley clients. The Pentagon denied the report. The Financial Times, citing three people familiar with the matter, stood by it.
Senate Democrats have since asked Hegseth whether he shared information about pending military action with his broker. He has not answered.
Presidentially appointed Department of Defense officials are explicitly prohibited by ethics rules from owning stock in the top defense contractors — Lockheed Martin, Northrop Grumman, General Dynamics, Boeing, RTX, Huntington Ingalls, L3Harris — all of which are held in the fund in question.
The question of who benefits from a $1.5 trillion opaque procurement budget — who owns which contractors, who knew what when, how the allocation of $480 billion in new procurement and R&D spending will be directed by a department that cannot account for its existing assets — is a different piece. But the structure that makes those questions hard to answer is the same structure this piece has been describing.
You can only hide money in the dark. Eight audit failures is a very long night.
What would change my mind
1. The FY27 reconciliation bill — the vehicle for the $350 billion tranche — passes with a provision tying fund release to a specific, scored audit milestone. This is not hypothetical. In June 2025, Rep. Maggie Goodlander introduced an amendment to do exactly this: the DoD could access $100 billion immediately, but the final $50 billion would be gated behind meeting the existing statutory December 2028 audit deadline. The amendment did not pass. If that structure, or any materially similar structure — a dollar amount withheld, a milestone required, a release condition specified — makes it into the final reconciliation language, the enforcement gap is partially closed. The test is simple: does the bill contain the words “contingent on” or “upon certification of” in proximity to any audit language? If yes, this thesis requires revision.
2. The FY27 audit result — released in December 2027, covering the fiscal year when the new money first flows — shows a reduction in material weaknesses from 26 to 15 or fewer, and at least three major military service components receiving clean sub-opinions for the first time. The GAO and the Pentagon’s own CFO have said that meaningful progress by FY26 is necessary to hit the 2028 mandate. If the FY26 and FY27 audits show that trajectory — not a promise of it, the actual scored result — the department is demonstrating it can handle the accountability burden that comes with the new money. That’s a concrete, dated, scoreable outcome: December 2027 audit report, 15 or fewer material weaknesses.
3. If the DoD Inspector General’s office publishes its spring 2026 report — which it has stated it will release — and that report shows complaint intake and investigation throughput at or above FY25 levels despite Hegseth’s September 2025 memo, the oversight dismantlement concern is more procedural than operational. The report is coming. The numbers will be in it. If they show normal function, I’m wrong about that part of the argument.
None of these outcomes currently seem likely on the evidence available. The Goodlander amendment was rejected. The FY25 audit found 26 material weaknesses — the same count as FY24, with no net reduction. And the IG reform memo has not been rescinded.
The actual question
Stewart was right about the audit, even if Hicks was technically correct that “failed audit” doesn’t legally mean “missing money.” The distinction matters legally. It doesn’t matter practically.
What a failed audit means is that the external auditors — independent accountants paid $178 million annually to do this job — cannot verify that the Pentagon’s financial statements are accurate. They cannot confirm that the assets exist as reported, that the expenditures went where they were directed, or that the internal controls are sufficient to prevent errors and abuse. They issue a “disclaimer of opinion,” which means: we can’t tell you what happened here.
For eight years in a row, the answer to “what did you do with the money” has been: we can’t fully say.
The proposed response to this record is to double the money.
There is a version of the $1.5 trillion buildup that is defensible. It requires auditable books, real oversight, and enforceable accountability standards — the same standards every other federal agency meets. The proposal on the table doesn’t include any of those conditions.
What it does include is $480 billion in new procurement, flowing into a department whose watchdog has just been constrained, whose audit has never been clean, and whose enforcement gap has been open, bipartisan-acknowledged, and structurally unaddressed for 35 years.
The Pentagon doesn’t have a funding problem. It has a problem with accounting for the funding it has.
The working hypothesis is that we are about to make that problem significantly larger.
Related: Eight Failures and a Raise — Investor Companion — who owns the contractors, how procurement allocation works under audit opacity, and what the $1.5T announcement has already done to defense sector positioning.
Related: The Classified Alpha — the structural enforcement gap on prediction market insider trading follows a similar logic: rules exist, consequences don’t, and the architecture rewards insiders.
If you found this useful, the best thing you can do is forward it to one person who would push back on it. I’d rather be wrong in public than right in private.