The State as Shareholder: America’s New Economic Order

Read the market analysis and fund positioning

The Pentagon is the largest shareholder in the United States’ only fully integrated rare-earth producer. It holds around 15% of MP Materials and has committed to purchasing the output from the company’s new magnet factory, including a minimum price guarantee.

State ownership, a state purchase guarantee, a state price floor: the three instruments that a free market economy does not provide for, all in a single transaction and under a Republican administration.

That sounds like more state intervention, bureaucracy and regulation. But in fact, 2026 is likely to be one of the least regulated years in American history. The spring regulatory agenda alone provides for the record repeal of 702 existing regulations. And for every new rule, ten are to be repealed. Measured by the volume of regulations, the state is therefore shrinking.
Instead, it is growing in other areas – namely, where it carries greater economic weight: as an owner, tax collector, price-setter and allocator.

The state as an owner

Washington currently holds around 30 corporate stakes worth some $26.7 billion: just under 10% of Intel through converted CHIPS Act grants, as well as stakes in nine quantum computing and commodities companies such as Lithium Americas and Trilogy Metals. When U.S. Steel was sold, the government retained a ‘golden share’ – a permanent right of veto against plant closures and the relocation of production abroad. However, there is no consolidated public register of this portfolio.

The state as a tax collector

The average tariff rate on US imports has risen from around 1.5% (2022) to 11% (2026) – the highest level since 1943. This represents the largest tax increase since 1993 in terms of gross domestic product, amounting to around $1,500 per household in 2026. These tariffs were imposed without a single piece of legislation passed by Congress. Added to this is an unprecedented arrangement: Nvidia is permitted to export to China in return for a 25% levy.

The state as price-setter

The Department of Energy is maintaining more than twenty emergency orders which keep unprofitable coal-fired power stations connected to the grid, against the wishes of the operators and the relevant states. These plants account for less than 1% of US coal consumption. The order forces the continued operation of uneconomical plants and disrupts orderly investment planning. The resulting additional costs, running into millions, are borne by electricity customers.

The state as allocator

Export licences determine, on a case-by-case basis, who is permitted to purchase which computing power. In June, a letter from the Department of Commerce removed two leading AI models from the global market within two hours, without a regulation or justification; the order was revoked after two and a half weeks. It is not the price, but authorisation, that determines access.

The objection that this is nothing new is valid: for the Republicans, the ‘lean state’ has always been more of a mantra than a reality. One need only think of Nixon’s wage and price controls, TARP and Medicare Part D under Bush, or the agricultural subsidies that have outlasted every Republican majority in the post-war era.
The difference compared with today lies not in the intervention itself, but in its form. Those programmes were crisis-driven, enacted by Congress and time-limited. Today’s instruments are executive in nature, crisis-free and have no expiry date. This concentration of power will become permanent because, at the end of June, the Supreme Court lifted the protection against dismissal for members of independent agencies: regulators who previously operated at arm’s length from the White House can now be dismissed at any time – and are thus, in effect, bound by its directives.

What's new isn't the intervention, but its form

One might object that $26.7 billion is nothing compared to a market capitalisation of around sixty trillion, and that most of the shares are non-voting. That is true, but it misses the point. It is not the size of the stake that is decisive, but the breach of competitive neutrality. The OECD links state shareholdings to four conditions: transparent remuneration for public contracts; equal treatment in regulation and procurement; no implicit state guarantees for debt financing; and a market-standard return requirement on the part of the state as owner. Washington fulfils none of these. Intel, AMD and Nvidia all received funding under the CHIPS Act; only in the case of Intel was this converted into state shareholdings. This is selection, not policy. And the market has immediately priced this in: following the announcement of the state stake, Trilogy Metals’ share price rose by around 240%, MP Materials’ by 150%, and Intel’s by 49%. These are not operational improvements, but the capitalised value of state favour, at the expense of the respective competitors.

The fact that Norway and Singapore hold state shareholdings without undermining competition does not refute this but rather confirms it. In those countries, the state, as owner, is bound by a publicly available mandate, return targets, a degree of independence from the government, and an obligation to sell off its shareholdings again. The US government is not subject to any of these constraints but is simultaneously a shareholder in one company and the regulatory authority for its competitors.

What does this mean for investors?

For investors, this is not a question of political persuasion, but of valuation. The US market remains the freest of the major markets. But the state is no longer merely a referee within it; it is also a player.