A founder sells a company for $200 million. The term sheet from Series C included a 1x participating preferred stake for investors. Run the waterfall, and that founder's actual payout is $62 million, not the $84 million they would have received under a simple 1x non-participating structure, according to Spectup's liquidation preference calculator. That $22 million gap is not a rounding error. It is the difference between two clauses that sounded interchangeable in a 40-page financing document signed three years before the exit.
This is the arithmetic nobody puts in the pitch deck. The exit number is a headline. What a founder pockets is a residual, and four mechanics decide how large that residual is: how much of the company they still own by the time it sells, where they sit in the payout order, how much the government takes, and whether a handful of tax provisions few founders have heard of trigger on the way out.
Most founders get nothing, and the ones who do get less than they think
Start with the base rate. Across 96,000 US venture-backed startups and 48,000 exits tracked from 1987 to 2021, founders were collectively wiped out in 74 percent of cases, according to the NBER study on capital gains taxation and startup founders. Three out of four times, the company that raised venture capital and eventually exited returned nothing to the people who built it. The remaining quarter is where the real question lives, because even a successful exit runs through a payout order designed to protect investors first.
Liquidation preferences are the mechanism. When a company raises money on preferred stock, investors are typically entitled to their investment back, plus sometimes a multiple of it, before common shareholders (founders, employees, early hires) see a dollar. In a case documented by Sesamers' startup equity guide, a company that raised $50 million in preferred equity with standard 1x liquidation preferences and then exits at $60 million leaves investors with their full $50 million plus a pro-rata share of the remaining $10 million. Common shareholders split what's left over, which in this scenario is close to nothing, despite a headline sale price that sounds like a win.
Participating preferred terms compound the problem. Under a participating structure, investors take their preference and then also participate in whatever remains, alongside common. That is the mechanic behind the $62 million versus $84 million gap above. Founders rarely negotiate this line item hard enough at the term sheet stage, because at Series B or C the number on the page feels theoretical. It stops being theoretical the day the acquisition agreement is signed.
The tax code takes its share regardless of what the waterfall leaves behind
Whatever survives the preference stack then meets the tax system. A founder in a high-tax state can face federal capital gains tax at rates up to 37 percent, an additional state tax of up to 13 percent, and a 3.8 percent net investment income tax on top, according to Ed Lloyd & Associates' exit tax planning guide. Stack those together and more than half of a windfall can be gone before the founder has spent a dollar of it.
There is a real offset, and it is large. Qualified Small Business Stock, under IRC Section 1202, allows founders to exclude up to $10 million, or 10 times their original investment basis, in capital gains from federal taxation, provided the stock is held for at least five years and other conditions are met, per Rooled's founder tax guide. Critically, this exclusion does not disappear if the company later grows past the threshold capitalization typically associated with "small business," according to Acuity's founder stock tax guide: shareholders who acquired stock while the company qualified can still claim the exclusion even after the business becomes, by any normal definition, no longer small. That is a five-year clock founders should be starting on day one of incorporation, not discovering at the term sheet stage of an acquisition.
A less visible trap sits inside M&A agreements themselves. Section 280G, the golden parachute rule, can impose a 20 to 40 percent federal excise tax on compensation triggered by a change of control, such as accelerated vesting, when payouts exceed specific thresholds, according to The Startup Law Blog. Founders who negotiated generous acceleration clauses years earlier, as a hedge against being pushed out post-acquisition, can find that same clause taxed as a penalty the moment the deal closes.
The dilution counterargument, and where it actually holds
None of this means dilution itself is the villain. The standard founder rebuttal is a real one: trading 100 percent of a $5 million company for 25 percent of a $200 million exit is a better outcome in absolute dollars, even after preferences and taxes eat into the smaller slice. Capital that funds real growth, hiring, and market expansion is the mechanism that made the larger number possible in the first place, not a tax on the founder. Most founders who raise institutional money are making exactly that trade deliberately, and for the ones who reach a real exit, it frequently pays off.
The NBER research supports a more precise version of this argument, and it points at policy design rather than dilution mechanics as the actual lever. When founders shift from realization-based to accrual-based capital gains taxation, meaning gains are taxed as they accrue rather than when the founder actually sells and receives cash, average founder payoffs decline by 15 percent, per the same NBER analysis. But the researchers also modeled a fix: if accrual taxation came with fully refundable tax credits, the share of founders receiving any positive payoff would rise from 16 percent to 47 percent. In other words, the system that determines whether a founder ends up with something or nothing is not fixed by nature. It is a set of choices about how and when tax liability attaches, and those choices can be redesigned to protect founders without changing the underlying economics of venture financing.
Tax planning professionals make a parallel point about waterfalls specifically: founders who arrive at closing surprised by their own payout usually skipped a step, not because the preference stack was unfair, but because nobody modeled it against their actual cap table before the deal was signed. That is a process failure, not a structural one, and it is fixable with an afternoon of work well before a term sheet is on the table.
Where the dilution argument breaks down
The trade-off holds only under specific conditions, and founders should know exactly where it stops holding. It fails when preference stacks are participating rather than straight, because participating terms let investors double-dip on both their preference and the common pool, as in the $62 million versus $84 million example above. It fails when a founder's ownership has been diluted below the threshold where even a large exit produces a meaningful personal number, which is the mechanical reality behind the 74 percent wipeout rate. It fails when a founder neglects the QSBS clock and sells before the five-year holding period, forfeiting an exclusion worth up to $10 million for no reason other than timing. And it fails when accelerated vesting clauses, written years earlier to protect the founder, become the very trigger that invites a 20 to 40 percent excise tax under Section 280G.
The people who benefit from founders not running these numbers early are, bluntly, the investors and acquirers on the other side of the table. A preference stack that a founder does not model is a preference stack an investor does not have to defend. A vesting acceleration clause a founder does not stress-test against 280G is a clause the acquiring company's counsel is perfectly happy to leave ambiguous. None of this requires bad faith. It requires only that one side has modeled the outcome and the other has not.
The model that should exist before the term sheet does
The founders who keep the largest share of their own exits are not the ones with the biggest headline valuations. They are the ones who ran the waterfall against every round's actual terms, tracked their QSBS holding period from incorporation, and had counsel stress-test any change-of-control compensation against Section 280G before signing anything. The $1.95 billion headline is the press release. The number that matters is the one sitting three steps downstream of it, after the preference stack, after the excise tax, after the state and federal rates. That number is knowable in advance. Most founders simply never ask for it until it is too late to change.
