En bref : Liquidation waterfalls pay senior preferred shareholders first, so a 25 per cent ownership stake can return nothing — ownership percentage anchors expectations that seniority and preference terms quietly override at exit.
A 25 per cent stake does not guarantee 25 per cent of the proceeds
Ownership percentage feels like a promise, and it is the wrong one. As Ilya Strebulaev of Stanford Graduate School of Business, author of The Venture Mindset, puts it in his analysis of seniority, waterfalls, and pari passu, a 25 per cent ownership stake does not guarantee 25 per cent of the exit proceeds. What determines the payout is the liquidation waterfall — the order in which the proceeds of a sale flow to each class of shareholder before anyone junior sees a dollar.
A waterfall pays from the top down. Senior preferred shares are repaid first, then more junior preferred, and only when every preference has been satisfied does common stock participate. The ownership column on a capitalisation table describes who owns what; the waterfall decides who gets paid what. The two diverge sharply at exit.
Kabbage’s ~$430M exit left $22M for two share classes and $0 for common
The Kabbage exit makes the gap concrete. On proceeds of roughly $430 million, the senior series were repaid in strict order: Series F took $161 million, Series E $166 million, Series D $50 million, and Series C $31 million — $408 million absorbed before the two oldest preferred classes saw anything. That left $22 million for the pari passu bucket shared by Series A and Series B, splitting to roughly $7.33 million for A and $14.67 million for B.
Common stock — the largest single stake at roughly 28 per cent fully diluted — received nothing. Strebulaev’s figures show common would only begin to participate above roughly $435 million in proceeds. The founders and employees who held the biggest ownership share by percentage sat below six preferred series, A through F, with F most senior. Their headline 28 per cent was an accurate measure of ownership and a misleading measure of outcome.
Anchoring on the ownership percentage is the behavioural trap
The error has a name in behavioural science: anchoring. The pioneering work of Daniel Kahneman and Amos Tversky showed that people fix on the first salient number and adjust insufficiently from it. On a cap table, the salient number is the ownership percentage — the clean, large figure printed next to a name. Founders and early employees anchor on it and read it as their share of any exit, when the economically decisive figures are the liquidation preferences buried in the term sheet. The percentage is real; it simply answers a different question than the one that matters at exit.
Once one round wins a 2x preference, the next round starts there
Preference terms compound, and Strebulaev illustrates the ratchet with a hypothetical company he calls SoftMet. Round A invests $10 million at a 1x liquidation preference. Round B invests $15 million. If Round B also takes 1x, the total preference stack is $25 million — the sum returned to preferred holders before common participates. If Round B instead negotiates a 2x preference, that same stack becomes $40 million.
The dynamic is self-reinforcing: once one round secures a 2x preference, the next round starts its negotiation at 2x, because no incoming investor accepts worse terms than the round below it. Each escalation lifts the floor that common stock must clear before it earns anything.
Preference terms live in the term sheet and the certificate of incorporation
These terms sit in named legal instruments. Liquidation-preference provisions are negotiated through the National Venture Capital Association (NVCA) model term sheet, the reference document most United States venture rounds start from, and then fixed in the certificate of incorporation of a Delaware-incorporated company, where most venture-backed startups are formed. The pari passu language that governs how equally ranked series share a bucket is standard certificate drafting — Dropbox’s 2014 certificate of incorporation is a frequently cited example. Because the terms are documentary, they are also auditable: anyone holding equity can read the certificate, map the preference stack, and model the waterfall before signing.
What to do, by seat
Les fondateurs should model the waterfall, not the ownership column, before accepting any term sheet — asking each round’s liquidation multiple and the proceeds level at which common begins to participate, and treating a rising multiple as a direct tax on the founding team’s equity.
Investisseurs should map the full preference stack across every series before underwriting, read the Delaware certificate of incorporation to confirm seniority and pari passu language, and price the position on its waterfall outcome rather than its percentage of the cap table.
Références
- Ilya Strebulaev (Stanford GSB). Who Gets Paid First? Seniority, Waterfalls, and Pari Passu. https://ilyastrebulaev.substack.com/p/who-gets-paid-first-seniority-waterfalls
