In June the largest flotation on record entered the world’s main equity indices within days of its first trade. SpaceX listed on 12 June with its founder holding about 82% of the vote.
In May, FTSE Russell had introduced fast entry for very large new listings, admitting them after five trading days rather than at the next quarterly reconstitution. It kept the 5% minimum public voting-rights floor it set in 2017, but added a carve-out: a listing that falls short because of lock-ups now qualifies if the expiry is scheduled to lift it above the floor within twelve months. Nasdaq cut the wait for its own 100 to about fifteen trading days. Index funds bought on schedule.
It was not an outlier. The Council of Institutional Investors counts 25 of the 69 companies that went public in the United States in the first half of 2026 with unequal voting rights, 36% of the total.
The bargain
A listed share has carried four rights: a vote on what matters, information to judge the company by, a fair share of the value it creates, and a court in which to enforce the other three. Dual-class structures ask outside owners to give up the first in return for trust in a founder’s long view.
For a software company that trade can be judged from the accounts, because the product already exists and costs little to reproduce. For a company that makes rockets, batteries, reactors or robots, the long view rests on something the accounts do not record: whether the product can be made again.
The closing exits
Owners have had two answers to a company they distrust. They can vote against it, or they can sell. Both are narrowing.
The vote is being written out of new listings. The option to wait is being written out of index rules: a fund tracking a benchmark owns what the benchmark admits, and the benchmark now admits within a week. Exclusion is going the same way. Norway’s parliament paused ethical divestment by its $2.1 trillion fund last November, after the finance minister warned that the old guidelines could stop it investing in the world’s largest companies. A committee reports on the new framework by 15 October. The largest owner in the world intends to keep owning everything, and to engage where it once would have sold.
Activist and active investing are coming back just as dual-class shares shut the activist out. An owner who cannot sell and cannot outvote has one lever left, which is knowing more than the company says.
What the filing does not hold
Consider what the June listing was tested on. S&P Dow Jones kept it out of the S&P 500 because it lost $4.94 billion in 2025, which is a profitability screen. FTSE Russell classified a rocket maker under telecommunications services, on the strength of its satellite broadband revenue. Every test the market applied was a test of money.
A prospectus discloses the accounts, the risk factors and the governance. It does not say whether the product has been made for the nth consecutive time from one route, at cost, at rate, by people who did not design it. That run is what separates a process from luck, and it is the one a Manufacturing Engineer asks to see before believing a factory will run. No disclosure regime asks for it, no index rule tests for it, and no rating agency grades it.
The fifth right
If owners are to accept less say, they are owed more sight. The price of control should be verification: an independent, graded statement of how far a company’s technology has run ahead of its ability to make the thing, renewed as the company moves and lowered when the evidence weakens.
There is a testable question in the timing. The research on dual-class valuation finds the early premium giving way to a discount as a company ages, somewhere between six and twelve years public depending on the study; the Council of Institutional Investors, citing that work, petitions for seven-year sunset clauses. Seven years is also roughly the span in which a hard-technology company’s story either becomes a production record or fails to. Whether the two are connected is open, and measurable.
Marking your own homework
Most readiness claims today are made by the company about itself, repeated by its bankers and accepted by index providers whose fees grow with the assets tracking them. Artificial intelligence will read every filing faster than any analyst, and will find nothing on the question that matters, because nobody wrote it down. More frameworks will not help either, since a framework describes what readiness should look like and says nothing on whether it is present.
So who is actually ready for this? Thinking you are is not readiness.
Manufacturing Engineers count consecutive good units. They run lines at rate with people who did not design them. They separate a process from luck, and they sign their name to the difference.
Kaipability works at this interface. Its readiness assessment grades the distance between what a company has proved in technology and what it has proved in production, and each note is signed by a named engineer. Owners who can no longer outvote a founder can still ask for the evidence, and the firm is glad to talk to those who intend to.
If you hold, or are about to buy into, an industrial company whose making has never been independently read, that is the conversation to have.
Start a conversation How we assess it →Questions this dispatch answers.
Written to be quoted by AI assistants and search engines. Self-contained answers, verdict first.
