Cap Table Waterfall Modeling: Why Your Common Stock Payout Depends on Liquidation Preferences You Ignored at Seed

Cap table waterfall modeling is the exit-scenario math that determines who actually gets paid — and how much — when a startup sells, liquidates, or IPOs. It layers every share class's liquidation preference, participation right, and conversion option in priority order, then calculates what remains for common stockholders at each possible exit value. Founders who skip this exercise at seed regularly discover, five rounds later, that a nine-figure acquisition leaves them with nothing.

The FanDuel founders learned this in July 2018 when Paddy Power Betfair (now Flutter) acquired the company for $465 million. According to filings surfaced by Legal Sports Report and coverage in Felix and FutureScot, the preferred stack — anchored by KKR and Shamrock Capital — was entitled to the first $559 million of proceeds. The founders and most employees received $0. That outcome was baked in years earlier, at term sheets nobody stress-tested with a proper cap table waterfall model.

What Cap Table Waterfall Modeling Actually Solves

A cap table lists who owns what. A waterfall model shows what those ownership stakes convert to in dollars at a given exit value. The two are not interchangeable. A founder who owns 40% of the common stock does not receive 40% of a sale; they receive 40% of whatever is left after every senior claim above them is satisfied.

Those senior claims typically include, in priority order:

Secured debt and transaction expenses (banker fees, legal, escrow holdbacks)
Employee bonuses and change-of-control payments contractually senior to equity
Preferred stock liquidation preferences, from most senior (typically the latest round) down to the earliest series
Participation payouts, if any preferred series has participating rights
Common stock (founders, employee-exercised options, angels who took common)
Unexercised options — worth zero unless they're in the money after everything above is paid

A serviceable waterfall model in Excel or Google Sheets takes an exit value as input and returns per-shareholder proceeds as output. A good one runs a sensitivity table across exit values from $10M to $1B so you can see the exact breakpoints where common stockholders start earning anything, and where preferred shareholders convert from preference-taking to as-if-common. Those breakpoints are the numbers founders should have memorized before signing every term sheet.

Takeaway: Before you accept your next preferred round, build the waterfall at three exit values — 1x total invested capital, 3x, and 10x — and confirm you know your personal payout at each.

The FanDuel Cautionary Tale: A $465M Exit That Paid Founders Zero

The FanDuel case, documented in a still-active lawsuit filed by cofounders Nigel Eccles, Lesley Eccles, Thomas Griffiths, Rob Jones, and Chris Stafford, is the cleanest teaching example in modern venture capital.

The mechanics, per court filings summarized by Front Office Sports and Legal Sports Report:

KKR and Shamrock Capital led later rounds with liquidation preferences and drag-along rights.
The combined preference stack totaled approximately $559 million in senior claims.
Paddy Power Betfair offered $465 million — meaning even a full 100% payout to preferred fell $94 million short of clearing the preference stack.
Common stockholders — founders and employees — received nothing.
Drag-along rights meant the common shareholders couldn't block the deal.

The lawsuit — reported by Front Office Sports in 2024 as an expanded complaint — alleges the preferred investors approved a sale price they knew would zero out the founders because that same sale delivered them 100% of the equity in the combined entity, which then went public at a much higher valuation.

The lesson isn't "KKR did something clever." It's that the preference stack, drag-along, and majority-preferred board control were all in the documents the founders signed at each round. A waterfall model at any prior round would have shown that any exit below $559M produced $0 to common. If you can't see that number, you can't negotiate against it.

How Liquidation Preferences Stack Across Rounds

Liquidation preferences are cumulative. Every round adds to the pile that must be cleared before common sees a dollar. The National Venture Capital Association (NVCA) model term sheet, still the reference document for U.S. venture rounds, offers three flavors: 1x non-participating (Alternative 1), full participating (Alternative 2), and capped participating (Alternative 3). What actually gets used has narrowed sharply.

Per Fenwick's Q3 2025 Venture Beacon, liquidation preferences above 1x remained rare in Q3 2025, appearing in under 4% of deals. Cooley's data, cited across 2025 practitioner guides, shows 98% of venture deals using a 1x preference and 95% non-participating. That's the founder-friendly headline. The Pillsbury "Down Round Is Back" update from 2024 tells the other half: in structured down rounds, participating preferred and 2x multiples are back on the table, and senior liquidation preferences (where new money ranks above earlier rounds) climbed from 29.6% of deals in 2022 to 47.0% in 2023.

Here's how a stack builds across four rounds on a company that raised $50M total:

Seed: $2M raised, 1x non-participating → $2M preference
Series A: $8M raised, 1x non-participating → $8M preference
Series B: $15M raised, 1x non-participating → $15M preference
Series C (down round): $25M raised, 2x participating with 3x cap, senior to all prior → $50M base preference

Total preference overhang: $75M. Common stockholders get zero until an exit clears $75M. Above $75M, the Series C keeps participating pro rata up to the cap — meaning even a $150M exit still routes most incremental dollars to Series C, not common. If the founders had modeled this at Series C signing, they might have negotiated the seniority (pari passu instead of senior), the multiple (1x instead of 2x), or the participation cap (2x instead of 3x). Every one of those knobs bends the waterfall.

Takeaway: Track two numbers after every round — total preference overhang and the exit value where your last common share earns $1. Recompute both before signing anything.

Building a Cap Table Waterfall Model: A Step-by-Step Example

The mechanics of a spreadsheet model are straightforward. What follows is a step-by-step recipe for a basic Excel or Google Sheets waterfall. The math itself is arithmetic; the discipline is in modeling the conversion decision that non-participating preferred face at every exit value.

Build the share class ledger. One row per class: seed preferred, Series A, Series B, common, option pool. Columns for shares outstanding, price per share paid, total invested, liquidation multiple, participation type (non / full / capped), participation cap, seniority rank.
Compute the preference amount per class. Multiply invested capital by the liquidation multiple. This is the dollar figure the class receives before junior classes get anything.
Set an exit value input cell. This is your independent variable. You'll flex it later.
Pay preferences in seniority order. Start with the most senior class. Subtract its preference from exit value. If exit value goes negative, junior classes receive nothing.
Model the conversion election for each non-participating class. For each non-participating preferred class, compute two numbers: (a) the preference dollars, and (b) what that class would receive if it converted to common and took its pro rata of remaining proceeds. The class takes the greater. This is where amateur models break — they hardcode the preference path and miss the conversion breakpoint.
Distribute participation for participating classes. Participating classes keep their preference AND take pro rata of leftover proceeds. If capped, stop participating once total proceeds hit the cap multiple of invested capital.
Distribute remaining proceeds to common. Divide the remainder by common shares (plus any converted preferred).
Build a sensitivity table. Copy the model across exit values from $10M to $1B in $10M increments. Chart the payout curves for founders, employees, and each preferred class. The kinks in the curve are your breakpoints — the exact exit values where a preferred class flips from taking preference to converting.

The Breaking Into Wall Street tutorial and Allied Venture Partners' guide both walk through the underlying Excel formulas in more detail; the Waterfalls.app explainer covers the "double-dip" participating math cleanly if you need visuals.

Takeaway: If your waterfall doesn't include the non-participating conversion election in step 5, it's wrong. That single formula is the difference between a model that predicts your exit and one that flatters it.

The Four Terms Founders Ignore at Seed That Break the Waterfall Later

Most founders read a seed term sheet, see "1x non-participating," and assume the liquidation preference is standard and harmless. It usually is — in isolation. The damage happens when four related terms compound at later rounds:

Seniority language. A seed round almost always says preferred ranks above common. Fine. But if the seed docs are silent on how future rounds rank against seed, the default in NVCA-derived templates is often pari passu — which sounds fair until a Series C investor demands "senior to all prior preferred" and your seed protection evaporates. Negotiate explicit language locking future rounds into pari passu absent a supermajority preferred vote.
Automatic conversion thresholds. Preferred stock automatically converts to common on qualified IPOs above a certain size or valuation. If that threshold is set too high, a modest IPO leaves preferred outside common and keeps the waterfall active. Founders benefit from a low threshold; investors want a high one.
Drag-along rights. The FanDuel outcome required drag-along. Without it, common could have blocked the sale. Founders should insist the drag be triggered only by majority of common AND majority of preferred, not preferred alone.
Pay-to-play provisions. Absent pay-to-play, an early investor who skips a follow-on round keeps their full preference. That means dead-money capital continues sitting on top of the waterfall for years. Pay-to-play converts non-participating shirkers to common, cleaning up the stack.

Takeaway: At seed, spend an extra hour on these four terms with a venture-experienced lawyer. Every hour here saves a decade of waterfall drag later.

How to Pressure-Test Your Cap Table Before It's Too Late

The forcing function most founders never implement: run the waterfall as a standing quarterly review. Use these five checks:

Compute your preference overhang. Sum every preferred class's liquidation preference across every round. That's the minimum exit value at which common earns anything.
Identify your breakpoints. At what exit values do non-participating classes convert to common? Those are the values where founders' incremental share of every next dollar jumps.
Model a strategic-buyer scenario at 1x, 2x, and 3x total capital raised. Most acquisitions cluster near 1-2x total raised. If you don't get paid at 2x, you don't get paid.
Model a down-round bridge. If you had to raise $10M more at a flat or down valuation with 2x participating preferred senior to everything, how much preference does that add? What's the new overhang?
Model the IPO conversion path. At what IPO valuation does all preferred convert automatically? What's your effective ownership after conversion vs. after preference?

Founders who run this quarterly stop signing bad term sheets. They also stop turning down good acquisitions that clear the stack, because they can see the math clearly. Investors who see a founder run the numbers in real time on a call negotiate differently, too — it signals sophistication and eliminates the information asymmetry the FanDuel preferred class exploited.

Conclusion: The Model You Should Have Built at Seed

Cap table waterfall modeling is not exotic finance. It is arithmetic layered on term-sheet language. Every founder who has ever raised a priced round is entitled to know, at any moment, exactly what they get at any exit value. Most don't, because building the model from scratch is tedious, and existing tools either abstract the mechanics away or cost enterprise money.

A ready-made waterfall template — one that handles multiple share classes, non-participating conversion elections, participating caps, seniority ranks, and sensitivity tables — pays for itself the first time you use it to renegotiate a single term. If you'd been the FanDuel founders in 2015, you'd have paid seven figures for the two-hour exercise of pointing at your spreadsheet and saying "no."

Build the model. Update it every round. Read the output before you sign. That's the difference between an exit that pays you and an exit that doesn't.