The exit mechanism only one side thought to negotiate

British readers have an instinctive picture of the politically stranded shareholding, and it tends to involve a special share. Rolls-Royce and BAE Systems both carry one, held by the Secretary of State, sitting alongside articles that cap foreign ownership, and the arrangement has been in place since privatisation without anybody pretending it is anything other than what it is. That is precisely why it is the wrong picture. A golden share is disclosed, described in the annual report, argued about in select committees when somebody proposes to remove it. It is a constraint you can see, price and lobby against. The constraints that actually strand capital are the ones drafted between private parties, symmetrical on the page, and asymmetrical only in the moment somebody needs to get out.

The clearest current illustration is not British at all. In November 2023 Renault and Nissan released their statements on the same day, one from Boulogne-Billancourt and one from Yokohama, and reading them side by side is what shows you what had actually been signed. The headline was identical from both directions: equalised cross-shareholdings, fifteen per cent each way, closing a quarter century of imbalance that dated back to the 1999 rescue. Yokohama had been asking for that parity for years, with a persistence that had become the only real negotiating stake left in the alliance, and it got it. Anybody who read to the end of both texts, however, found that the twin holdings were not the same object, and were not the same object by construction.

Renault took its excess above the parity threshold, the twenty-eight point four per cent of Nissan’s capital, and transferred it into a French trust with an independent manager, where the votes are neutralised while the economic rights, dividends and disposal proceeds alike, remain wholly with the French group until the day of sale. The vehicle is not a parking space, it is a disposal mechanism: the mandate is to place the shares progressively, in coordination with the counterparty, which retains a right of first offer. Nissan holds its fifteen per cent of Renault directly, as it has since 2002, with no trustee, no mandate, no timetable, a contractual cap on its voting rights and a lock-up undertaking. One side left the table with a purpose-built release mechanism, the other with the optical symmetry of a number.

Two and a half years later the design is being tested, because Yokohama needs cash and has started looking through its monetisable assets. The Renault holding is worth a little over a billion euros at this summer’s prices, against a French market capitalisation that has fallen below eight billion after a year that burned more than a third of the value. On paper it is a cheque. In practice the March 2025 amendment, which lowered the mutual lock-up from fifteen to ten per cent, makes a third of it saleable, some three hundred and seventy million rather than the billion, with the remainder staying where it is by contract. And that saleable third does not go to market the way an ordinary block would, because every disposal has to be coordinated with the counterparty, which can exercise pre-emption directly or nominate a third party of its own choosing.

Which is where this stops being an automotive story and becomes a structuring problem for anybody who has ever signed a shareholders’ agreement.

Liquidity in a jointly held structure is almost never lost to a prohibition. Prohibitions are visible, they can be challenged, they have an expiry date and a tribunal. Liquidity is lost to pre-emption, to a clause that does not decide whether you may sell but decides to whom, and which therefore takes hold of the one variable that genuinely determines price in a strategic minority. A restricted minority stake is not worth the underlying divided by the share count, it is worth whatever the narrow set of permitted buyers will pay for it. If that set consists of the counterparty and the people the counterparty finds acceptable, the price is not set by the market, it is set by whoever holds the right to say no. Signing a pre-emption right does not surrender the right to leave, it surrenders the right to choose the door, and with it goes the portion of value that depends on the identity of whoever walks in.

I have watched this work in settings that have nothing in common with Yokohama except the structure. A Midlands manufacturing family, four branches by the third generation, a holding company whose articles were drafted in the mid-eighties and never revisited, where the pre-emption provision had been inserted for a reason that seemed obvious at the time and consisted of keeping outsiders out of the family’s capital. Thirty years on, the branch that needs to monetise, because of a divorce, because of an investment outside the perimeter, for the simple reason that lives diverge, discovers that the outsiders have been kept out with complete success and that the only buyer the articles permit is the branch that is in no hurry whatsoever and has no incentive to pay anything close to full. A clause conceived as collective defence ends up operating as a free call option in favour of whoever stays, exercisable at precisely the moment the other party is weakest. Nobody designed it that way. That is how it works.

The question clients ask at that point is always why anyone would sign such a thing. The answer is that at the moment of signature the clause costs nothing and protects against a risk that is immediate and legible, the arrival of a hostile third party, while its cost only materialises when the parties’ needs diverge, which is to say in a future none of the signatories is looking at because all of them are looking at the transaction in front of them. Signing a lock-up means buying stability today with liquidity tomorrow, and the exchange rate is generous until tomorrow arrives. Nissan in 2023 was buying its way out of two decades of subordination and was willing to pay for it; it paid with its own optionality, in a year when nobody in Yokohama was contemplating a six hundred and seventy billion yen loss and a plan that closes seven plants.

The institutional layer then gets added on top, and it is real, but in my view it is overweighted by most of the people describing it. The French state holds fifteen per cent of the capital and roughly double the voting rights thanks to the double vote on registered shares, it has a documented history of intervening in the control of national champions, and it has no intention of watching a meaningful block of the manufacturer migrate to a non-European buyer. All true. But that is the part of the constraint that is visible, commented upon, even politically contestable, and it is not the part that binds. The part that binds was drafted between private parties, is symmetrical in the text and asymmetrical in effect, because only one of the two sides came away with the release vehicle. Tokyo, after all, would screen a significant disposal of Nissan to a foreign buyer just as readily, and anybody building the reading on the nationality of the issuer has to explain why the Japanese filter does not count.

What separates the two positions is not the passport. It is that in 2023 one party negotiated its own exit mechanism and the other negotiated its own entry into Ampere, which has since been wound up.

In the agreements that cross my desk, the release mechanism is the section that is usually missing, or present in twenty generic lines while the blocking provisions run to four pages. Correcting it does not take much: a valuation formula that does not depend on the counterparty’s consent, a window after which an unexercised pre-emption right lapses, a purchase obligation mirroring the veto. These are things you obtain at the opening table, when the other side needs your signature, and that you no longer obtain on the day you telephone to say you need money. Anybody sitting down now to sign a lock-up without demanding the equivalent of the trust would do well to consider that the value of their holding is not the figure on the statement, but whatever the counterparty decides it is at the worst possible moment for them.