The IPO share figure counts this year's entrants rather than the installed base, which is why it moves faster than the reincorporation tally. Read from the other side, the non-Delaware column of the IPO market went from 19% in 2024 to roughly 30% in 2025 [3][15]. That is about a seventh off Delaware's own share and a little under three-fifths added to everyone else's [14][15]. The 50-plus departures over two years work out to slightly more than two a month [2][16], a slower series, and the one that will lag whatever the formation decisions are doing now.
The sequencing is the part that binds. Cooley's memo, published on Harvard Law School's corporate governance forum, frames domicile as a strategic governance decision for a founder-led technology company and walks through Delaware, Nevada and Texas [13][1]. Its warning on reversibility is blunt: once a company is public and widely held, reincorporation becomes much trickier, because proxy advisory firms have generally opposed moves to Nevada and Texas and large institutional shareholders have often opposed reincorporations, producing difficult votes at noncontrolled companies [6]. The option to move is therefore priced by the share register, and it decays as the founder's block dilutes.
The numbers also support a narrower reading: very few widely held large-cap companies have left, and Delaware still holds close to 70% of new listings [5][3]. Both figures are accurate, and both describe the wrong population. The movement is concentrated among founder-led companies and companies with large influential shareholders [5], which is precisely the cohort that still has a cheap vote to hold, and the IPO number captures the companies deciding for the first time.
What Delaware is selling is the reduction of variance. The Court of Chancery is a dedicated equity court with constitutionally required political balance among its judges, sitting without juries, with a record of deciding urgent cases in days or weeks [7]. Cooley's own framing concedes the trade: Delaware's case law supports a reputation for predictability that can be worth having even where other jurisdictions offer greater protections in certain areas on paper [8]. Its statute is amended by two-thirds legislative vote on proposals from a bar council seating both management-side and plaintiff-side practitioners [9], and the 2026 amendments signed in June were comparatively measured [10]. Retired Delaware Supreme Court Justice Karen Valihura, in a June 2026 lecture, put the case as 230-plus years of trial and error that cannot be replicated quickly [11], which is a fair argument from an interested witness.
The record here leaves two points unsettled. The memo says prominent companies have chosen Texas and are using the state's statutory tools to tailor stockholder rights, but the excerpt we have does not name those tools, and it breaks off mid-sentence in the Delaware section [4][17]. So the comparison a board actually needs, tool by tool against the cost of a thinner case law record, is not on the page yet.
The board-deck version is that Delaware remains the default and roughly seven in ten new issuers agree [3], but that leaves out that the default is now being chosen rather than inherited, and for a noncontrolled company the choice is close to one-way [6]. A founder-led issuer choosing this quarter is choosing between a court that answers in weeks [7] and a charter written to its own specification [4], and it will operate inside whichever one it picked.
