The Account, Not the Percentage

Two details in the reporting are worth more than the headline number. The stake is not a cash purchase: the instrument, as described, is penny warrants — an option to take the shares at nominal cost, held without operational control. And the arrangement is overseen by the Pentagon's Office of Strategic Capital.

Not the State Department. Not Treasury, which handles sanctions relief and holds the accounts. Not the Development Finance Corporation, which Congress built in 2018 for the express purpose of taking equity positions in foreign infrastructure and which possesses, on paper, exactly this authority. The Department of Defense, through a financing office created in 2022 to issue loans and loan guarantees for critical-supply-chain technology.

Everyone is arguing about the 35 percent. The 35 percent is a number. What matters is where the money lands, and that question is not being decided in Caracas. It is being decided in a conference committee on a defense authorization bill, in a subtitle almost nobody outside the trade press has read.

The Least Glamorous Wall

Article I, Section 9 of the Constitution says no money shall be drawn from the Treasury but in consequence of appropriations made by law. The clause is usually read as a restraint on spending. Its operative half is the mirror image, codified at 31 U.S.C. § 3302(b): an officer or agent of the government receiving money from any source shall deposit it in the Treasury, as soon as practicable, without deduction for any charge or claim.

That is the Miscellaneous Receipts Statute, and it is the least glamorous load-bearing wall in American government. It says an agency cannot keep what it earns. Everything goes into the general fund and comes back out only through appropriation. The purpose is not accounting hygiene. The purpose is that an agency which funds itself does not need you.

The Senate-reported version of the National Defense Authorization Act for fiscal 2027 contains a new subtitle titled "Equity Investments and Related Matters." Its central provision, Section 1051 — "Authority for equity investments by Office of Strategic Capital" — would grant the OSC explicit authority to take equity positions in private companies: investments up to $500 million, with direct equity limited to 40 percent of the total equity invested in any one entity, in critical minerals, materials, chemicals, and batteries. And it would establish, in the Treasury, a Department of Defense Equity Investment Account, which the OSC director may use to make those investments.

It is tempting to say that the account is itself the carve-out from § 3302, and that is the mistake — the one a piece about plumbing cannot afford to make. An account in the Treasury deposits in the Treasury. The statute is satisfied on its face. The carve-out is never the container; it is the availability language attached to it. Receipts retained. Funds available without fiscal-year limit. Obligation without further appropriation. An account without retention authority is a ledger line. An account with it is a revenue stream. Same words on the letterhead, different animal underneath.

Which leaves one unglamorous question, and it will be resolved by staff: do the returns replenish the account, or do they sweep to the general fund? If they sweep, the Pentagon is an administrator holding assets on behalf of the United States. If they revolve, the Pentagon holds a revenue stream that does not require anyone's vote.

Same 35 percent either way. Different institution.

That binary is cleaner than reality, and the piece should say so before a reader says it back. Even a sweeping account creates an institutional interest: returns get cited in posture hearings, a constituency forms around the portfolio, somebody's career becomes the fund's performance. And a revolving account capped at $500 million is a rounding error against a defense budget near a trillion dollars. The variable that decides whether this is structural may not be revolve-versus-sweep at all. It may be whether the cap reads as a ceiling or as a precedent. The first number is always the hardest one to get.

I want to be fair about the distance here, because the alarming version of this story is available and it is not yet the true one. A passive minority position, taken through warrants, overseen by a financing office, subject to a statutory cap Congress is actively debating, is a very long way from an army that owns pasta factories. Nobody at OSC is drilling anything. The Pentagon's own press office has said flatly that the Office of Strategic Capital does not take equity stakes in private companies — its statutory role being loans, guarantees, and technical assistance — which is either a denial or a description of why the NDAA subtitle exists, and the honest reading is that the legal footing is contested. The caps were written by real authorizers who noticed. The honest description of today is: a financing office with a novel and disputed authority, doing something unusual.

The problem is that "a long way from" describes a distance along a gradient, and gradients have a direction.

We Have Already Run This Experiment

The best evidence for that direction is not Rome or Bengal. It is your county sheriff.

Since the Comprehensive Crime Control Act of 1984 established federal equitable sharing, American law enforcement agencies have been permitted to retain a share of assets they seize. Billions have been distributed. And the documented behavioral result — replicated across decades of study and repeatedly conceded by the agencies themselves — is that enforcement drifts toward seizable value. Departments in states with restrictive forfeiture laws route seizures through the federal program. Interdiction shifts toward the outbound lane, where the cash travels, rather than the inbound lane, where the drugs do.

Nobody in those departments decided to be corrupt. That is the entire point, and it is why moralizing about it has never worked. You do not get a bent institution by hiring bad people. You get one by building an account and letting fifteen years of ordinary budget pressure do the work. Every officer makes a defensible local decision; the gradient integrates them.

This is the mechanism demonstrated rather than asserted, and the terms are favorable to the skeptic: domestic, capped, mundane on arrival, thoroughly documented, subject to elections and a free press and courts. It drifted anyway. The United States ran the experiment on its own police forces, at municipal scale, with data, for four decades. The proposal on the table applies the same design to the institution that holds the aircraft carriers.

The Far End of the Road

The classical cases are what the road leads to, not where anyone currently stands, and the difference is worth stating plainly rather than borrowing menace from it.

In 1765, after Buxar, the Mughal emperor granted the East India Company the diwani — the right to collect the land revenue of Bengal, Bihar, and Orissa. The Company already had an army; armies were an ordinary cost of doing business in a place with no other enforcement. What it acquired in 1765 was a tax base. Within twenty years it fielded more soldiers than the British state, ran courts, and set foreign policy in Asia. Parliament's response — the Regulating Act of 1773, Pitt's India Act of 1784, the Board of Control — is often taught as an anti-corruption story. It was jurisdictional. Parliament had noticed that an armed organization with an independent revenue base is not a company. It is a rival sovereign wearing a company's paperwork, and the only remaining question is which one gives orders to the other.

Rome got there faster and with less paperwork. After Marius opened the legions to the landless around 107 BC, a soldier's retirement came not from the treasury but from a land bill his general pushed through the Senate. Loyalty follows the payer; this is not a moral observation, it is an accounting one. Sulla marched on Rome in 88, Caesar crossed the Rubicon in 49, and by 193 AD the Praetorian Guard — whose pay had become a donative negotiated at each succession — auctioned the empire to Didius Julianus. Not a coup. A transaction.

The living cases are less colorful and better documented. Egypt's armed forces run construction, cement, fuel stations, appliances, and resorts, with credible estimates of their share of GDP ranging from two percent to something near a third — and the width of that range is itself the finding, because the holdings are not auditable by anyone the Egyptian public elects. Pakistan's military welfare foundations, catalogued by Ayesha Siddiqa in Military Inc., ran to roughly $20 billion in assets two decades ago; Siddiqa's term, Milbus, describes capital that is neither fully state nor fully private and accountable to neither. Iran's Khatam al-Anbiya, the IRGC's engineering arm, became the country's largest contractor and now behaves like an interest group with its own foreign policy, because it is one.

None of these institutions decided to become what they became. Each acquired a revenue line for a defensible reason and then optimized, as institutions do, toward the thing that pays. But note the one feature they share and the American case does not: large, self-directed, operating revenue. That is the gap. It is a real gap, and the sheriff is the evidence that gaps of this kind close by accumulation rather than by decision.

The Only Regime That Turned It Off

Here is the case that should make anyone confident the mechanism is real rather than a moral panic: the People's Liberation Army.

Through the 1980s and 90s the PLA ran a sprawling commercial empire — hotels, pharmaceuticals, telecoms, real estate, and a smuggling operation large enough to distort national customs revenue. Tens of thousands of enterprises. In July 1998, Jiang Zemin ordered the military and the People's Armed Police out of business entirely, with the state absorbing the enterprises and compensating with budget increases.

The public rationale was corruption. The operative rationale was control. A one-party state with no meaningful electoral accountability, no free press, and no independent judiciary looked at a self-financing army and concluded it was too dangerous — not to the public, to the Party. An army that funds itself is not reliably anyone's army. Beijing paid a substantial fiscal price to remove that risk, and it is one of the few decisions of that era nobody in the Chinese system has proposed reversing.

Indonesia reached the same conclusion in 2004 and legislated the takeover of military businesses within five years. Implementation was partial, which is its own lesson about how hard the reversal is once the revenue exists.

So the arrangement is not inevitable. It is chosen, and at least one regime with every incentive to keep it judged it not worth the cost. That is the honest counterweight to fatalism. It also means the American question is not whether this ends in Cairo. It is whether the account revolves.

The Part Nobody Is Arguing About

And there is a version of the question that is sharper than the account, which almost nobody has said out loud.

Equitable sharing creates an incentive to seize more. That is budget drift inside an agency's existing mission — the sheriff still polices, he simply polices toward the money. A position in seventeen Venezuelan oil fields is a different animal, and the difference is not one of degree.

Forfeiture rewards an agency for doing more of what it already does. An oil field rewards an agency for the asset performing. And the variables that determine whether 1.5 million barrels a day actually flows are sanctions posture, regional stability, freedom of navigation in the Caribbean, physical security of the fields, and who governs Caracas in 2031. Every one of those is something the Department of Defense advises on, executes, or both.

So the conflict is not that the Pentagon might tilt toward profitable missions. It is that the Pentagon would hold a position whose value depends on the outcome of policies the Pentagon shapes. The sheriff's problem is which lane he parks in. This one is which government stays in power.

Nobody has to intend it — that is, again, the entire point. A twenty-five-year operating agreement and a century of rights outlast every official who will ever be asked about them, and each of those officials will make a defensible local judgment about stability in the Caribbean while the portfolio sits in the background, worth more or less depending on the answer.

There is a second failure mode, quieter, and the Egyptian case names it if you turn it around. The objection to Cairo is not only that the army funds itself. It is that the holdings are not auditable by anyone the Egyptian public elects. An equity portfolio held by the one department in the American government with original classification authority raises that question on its own, independent of where the returns land. Who values the position? Does a number appear in an unclassified budget document? Does the Comptroller General get the books? Self-funding is dangerous because the agency stops needing you. Unauditable holdings are dangerous because you stop being able to see what it is doing with the money whether it needs you or not.

The Senate bill is not naive about the first half of that. It would stand up an Economic Defense Unit, require quarterly briefings, and make the Secretary certify that the Department holds no board seat or other voting representation in anything it owns. That is a real answer to the control question, written by people who thought about it. It is not an answer to the valuation question. And it is not an answer at all to what happens when the certification concerns a field in a country where the Department has ships.

What to Watch

The Caracas announcement will be litigated on the terms everyone finds legible: whether the OSC has authority to hold equity at all, whether warrants in a private venture are a security arrangement or a favor, whether Venezuelan sovereignty survives a century-long grant. Those fights will be loud and mostly beside the point. Deals get unwound. Administrations change. A specific stake in a specific venture is a reversible fact.

An account is not. Statutory plumbing outlives the administration that installed it, and the next occupant inherits the fixture without inheriting the argument. The equitable sharing program was built for cartels and is used for traffic stops.

So the sentence to watch is not in a press release. It is in Section 1051, in conference, and it concerns whether returns to the Department of Defense Equity Investment Account are available without further appropriation. That phrase — without further appropriation — is four words long, it will not be debated on the floor, and it decides whether the Pentagon's budget continues to answer entirely to Congress or begins, at the margin, to answer to a well.

They will not vote on it as that. They never do.