Good Companies, Underwater Cap Tables: Why the SaaS Repricing Is Now a Governance Problem

Public markets repriced software this year. In the first quarter, investors began pricing artificial intelligence (“AI”) as an existential threat to software-as-a-service (“SaaS”). The SaaS Capital Index tracks the median revenue multiple of publicly traded business-to-business SaaS companies. It fell to 3.1x in June, its lowest reading since 2011, from 16.9x at its August 2021 peak.

The private-market reset is harder to observe. PitchBook estimates that U.S. startups whose last round came in 2021 were worth 68% less on average by year-end 2025. PitchBook also estimates that more than 220 former unicorns no longer clear the $1 billion mark. Software is the largest category among them: 75 SaaS companies, double the number of fintech firms.

Those numbers describe the damage. They do not describe what happens next. At Silicon Valley Bank’s State of the Markets launch event in February, Ben Lerer of Lerer Hippeau pointed to a “giant overhang of thousands of SaaS businesses that were really good companies” and asked how it works its way through the system. Lerer was describing the exit backlog: thousands of venture-backed software companies waiting behind a handful of giants for a path to liquidity.

My answer is that many of these companies do not have a business problem. They have a capital structure problem. When the proceeds available to stockholders from a realistic exit fall below the aggregate liquidation preference, the company’s equity securities stop behaving like a single class of equity: each additional exit dollar goes to the preferred, while the common stays at zero. Incentives diverge, and valuation becomes a governance problem.

This article is the second in a series. In August, I argued that venture underwriting can no longer assume a return to near-zero rates, because long-term yields now price higher rates for the foreseeable future. That piece addressed the discount rate. This one addresses the operating assumptions beneath the revenue base. A 2021 valuation can be stale in two places at once: the market has reset, and the durability of the revenue supporting it has weakened.

The valuation is wrong twice

For SaaS companies, annual recurring revenue (“ARR”) times a revenue multiple remains the market’s shorthand, even where formal fair-value work is more involved. Most commentary focuses on the multiple, and with reason. The public median has since partially recovered, to 4.2x at the end of September, but it remains about a quarter of its 2021 peak.

The less-discussed problem is the revenue base the multiple is applied to. Seat-based pricing tied revenue to customer headcount. When software agents do work that seats once did, customers renew with fewer seats.

Price increases can hold headline ARR flat while the seat count shrinks underneath it. SaaS Capital’s data show the index’s median year-over-year revenue growth falling from 31% at the end of 2021 to 13% in September 2026. A 2021 valuation that has absorbed neither the multiple reset nor the weakening revenue base is wrong twice.

That is why I read net revenue retention (“NRR”) before the multiple. If a company has raised prices and still retains only 97% of last year’s revenue from existing customers, the cohort’s revenue is shrinking despite those price increases.

Shrinking seat counts, compressed multiples, and slower growth do not mean the businesses are failing. On the contrary, SaaS Capital also reports that the median public SaaS company now earns an operating profit, for the first time in more than a decade. Public SaaS is growing more slowly and operating more profitably. Many 2021 private-company capital structures still assume the old growth.

Good company, underwater stack

Here is a cap table we see often, shown as a composite rather than any single engagement. The company raised its Series C in 2021 at $20 million of ARR, bringing its total preferred to $185 million. Every series carries a 1x non-participating preference with no accruing dividends, and the company has no management carve-out plan.

Since then, the company has done most things right. ARR has nearly doubled, to $38 million. The company operates near breakeven and holds NRR at 97%.

In 2025, its insiders bridged it with $20 million of senior preferred carrying a 1.5x preference, or $30 million of liquidation preference. The stack now totals $215 million.

Look at the chart below before reading further. The business nearly doubled, yet common does not begin to participate until equity value exceeds $215 million. The 2025 bridge bought time and raised that threshold by $30 million.

For the illustration, ARR stands in for annualized run-rate revenue, the basis SaaS Capital uses. At the public median, the company’s implied equity value is about $160 million. Apply an illustrative 25% private-company discount, and the figure falls to about $120 million. In the sponsorless situations CRAGSI sees, buyer indications for companies like this typically run between one and three times revenue on an enterprise-value basis. With no material debt or excess cash assumed, that implies equity values of roughly $38 million to $114 million.

The common stock receives nothing until equity value exceeds $215 million, roughly 5.7x ARR. On these assumptions, the preferred stock’s right to convert into common does not lower that threshold. A non-participating series converts only when its share of the proceeds as common would exceed its preference. Below $215 million, converting means surrendering a preference for a share of whatever remains after the other series take theirs, and that share is always smaller.

The common is underwater across every valuation case shown.

In August, I also noted that the insider bridges we had reviewed were moving toward participation, seniority, and pay-to-play. For this cohort, those terms buy runway by enlarging the overhang. Every dollar of added senior liquidation preference raises the equity value at which anyone junior begins to participate. A bridge that simply layers new senior preference onto the existing stack compounds the capital structure problem.

The governance trap: who decides between a sale and a recapitalization

Once the stack exceeds the proceeds available to stockholders from any realistic exit, the holders of the company’s securities stop wanting the same thing. Three patterns recur:

That board structure carries a practical cost that syndicates often discover late. Delaware’s 2025 amendments to Section 144 of the Delaware General Corporation Law created new statutory safe harbors for specified conflicted transactions. In February, the Delaware Supreme Court rejected constitutional challenges to those amendments in Rutledge v. Clearway Energy Group LLC. Where Section 144(a) applies, as it can to an insider-led bridge or recapitalization, and a majority of the board is not disinterested with respect to the transaction, the director-approval safe harbor requires a committee of at least two disinterested directors. A board composition set years ago in a financing can narrow the board’s options at the moment it needs them most.

Management is not neutral either

Management incentives complicate the picture in both directions. Where a board has adopted a management carve-out plan, the payout comes off the top of a sale. That converts executives from holders of underwater common into claimants whose incentives resemble the preferred’s. The Delaware Court of Chancery described exactly that conversion in its 2013 Trados decision.

Where there is no carve-out plan, a similar conflict can arise elsewhere. A buyer’s retention arrangements can create a separate, transaction-specific incentive for the executives who receive them.

The talent risk is real, but it is targeted. When option strike prices sit far above the company’s current Section 409A valuation, broad-based option retention weakens. The larger risk is concentrated in a small number of highly portable employees, often the two or three people an acquirer is actually paying for.

The question the cohort turns on

Delaware has addressed this conflict before. In Trados, and again in the Oak Hill litigation over ODN Holding, the Court of Chancery held that directors owe their duties to the common as the residual claimants, even when holders of preferred control the board. In both cases, the directors ultimately prevailed. The Trados court found the deal fair despite an unfair process, because the common had no economic value. The 2020 ODN decision added, after a fact-intensive trial, that a board need not pursue a lottery-like path unlikely to create value for the common.

Those outcomes become much harder to rely on when the common is merely out of the money rather than economically worthless. The Trados court, drawing on venture capital scholarship, noted that the preferred-common conflict bites hardest when a company is neither a complete failure nor a stunning success. That describes the segment of the 2021 SaaS cohort at issue here almost exactly.

The practical lesson is about timing. The amended statute makes board composition and conflict planning matter before a process begins, not after. Boards in this position should involve Delaware counsel early.

The objection worth answering

The strongest counterargument is that patience works. Public multiples have already recovered from their June trough, and a company that doubled its revenue once can do so again. Why recapitalize near the bottom?

It is a fair question, and the arithmetic answers it. To clear its stack at today’s public median, the illustrative company must grow ARR from $38 million to about $51 million. After the illustrative 25% discount, it needs about $68 million. That is roughly five years of growth at the index’s 13% median rate, assuming the multiple holds.

At 97% NRR, the existing-customer revenue base is shrinking 3% annually, so all net growth, and additional ARR to replace that contraction, must come from new customers. A company operating near breakeven must fund that new-customer growth while its lead may not write another check alone. Any further bridge on similar senior terms raises the target again.

Patience may still be the right answer for some companies. But it should be a deliberate choice, made against a modeled recapitalization alternative, not a default that preserves the 2021 valuation for another year.

A five-step playbook for managing partners and boards

For managing partners, navigating this overhang is a governance exercise before it is a transaction. Five steps, taken early, preserve options that disappear once a process starts:

For founders and boards of these companies, the same list works in reverse. If your investors have not raised these questions, raise them yourselves, while the company still has the cash to choose.

None of this requires a view on whether AI ultimately hollows out SaaS. It requires only a clear view of what has already happened.

The market has repriced the software. The cap table still prices 2021. Boards that choose between them should know which one they are protecting.

Managing partners and founders: is your board evaluating a recapitalization as a real alternative to the sale, or only after a sale fails?

Originally published on LinkedIn on October 8, 2026.