Run an Exit Waterfall at Three Prices, Step by Step
How to run an exit waterfall: list every preference in order, then apply it at a disappointing, a plausible and a good price to see what reaches common holders.
How to run an exit waterfall: list every preference in the order it is satisfied, apply it to three sale prices, and see what reaches common shareholders at each. The arithmetic is not difficult. Not doing it is what makes the outcome a surprise.
This is general education about the method. It is not personal financial, legal or tax advice, and your own documents decide your own answers.
How to run an exit waterfall: the three prices
The guide's central exercise is to run the waterfall at three prices: disappointing, plausible and good. The disappointing price is the one that teaches you something, and it is the one nobody models.
For each price, apply the preference stack in order and see what reaches common shareholders, then what reaches you specifically. Done once, this changes how most founders think about their own cap table permanently, because it makes visible the range in which a sale that looks like a success pays them almost nothing.
| Price | What it shows you |
|---|---|
| Disappointing | Where terms decide who gets paid, and the range in which common shareholders receive effectively nothing |
| Plausible | What your realistic outcome looks like under the stack |
| Good | That almost any structure looks fine when the price is great |
The reason to start with the least comfortable price is that the good outcome tests nothing. It is only at a mediocre price that the structure is tested.
Step one: build the list of preferences
Before any arithmetic, collect the documents and write a complete list. For each preference record four things.
- Its place in the order in which it is satisfied.
- Its multiple.
- Whether it participates.
- Whether it converts to common when that gives the holder a better result.
This is the first group of boxes on the guide's first checklist, and it is the one that cannot be skipped. A model built on a general assumption about how rounds work will be wrong exactly where the surprises live.
Step two: get the order right
The order is not always the one you assume. Later investors frequently sit ahead of earlier ones. Different rounds may carry different multiples or participation rights. Some preferences convert to common when that produces a better result for the holder, and some do not.
So model the actual documents rather than a general assumption. The differences between rounds are only visible if you read what was signed. If a document is missing or unsigned, that is a finding in itself, and it belongs on the readiness list in how to prepare for acquisition due diligence.
Step three: apply the stack to each price
With the list and the order in hand, take the first price and work down the stack. The structure of the calculation is always the same.
- Start with the sale price.
- Satisfy the first preference in the order, up to what the documents entitle it to.
- Continue down the stack, noting for each whether it participates and whether converting would serve the holder better.
- See what is left for common shareholders.
- Apply your own holding to that remainder.
Then repeat for the second and third prices. Keep the three columns side by side. The comparison is the point, because it shows where the outcome changes sharply and where it hardly moves.
| Line in your working | Disappointing | Plausible | Good |
|---|---|---|---|
| Sale price | Your figure | Your figure | Your figure |
| Preferences satisfied | From your list | From your list | From your list |
| Left for common | Result | Result | Result |
| Your share | Result | Result | Result |
The cells are for your own numbers. There is deliberately nothing invented in them, because the right figures are the ones in your documents.
Step four: find the range where common gets nothing
One result from the exercise deserves its own line: the price range in which common shareholders receive effectively nothing. The guide says that range is often wide and often the most probable one.
Knowing where it ends gives you a floor to reason from. A sale above that range starts paying the common pool. A sale below it does not. Once you know the boundary, the question of what price you would want becomes much more concrete. The thinking that follows is in how to decide what a good exit is for you.
Step five: model the option pool as it behaves
Unexercised options do not simply convert into value. Depending on the structure, holders may need to exercise, unvested grants may or may not accelerate, and the treatment of the unallocated pool varies by deal. Work out what employees actually net, not the notional value of their grants.
Do it before anyone asks, because they will ask, and an approximate answer given in a stressful week is worse than none. The people side of the model is expanded in what happens to employee equity when a startup is acquired.
Step six: adjust for headline against received
A waterfall tells you what is distributed. It does not by itself tell you when, or in what form. Add a final pass for the differences between the announced number and the amount that reaches a bank account.
- A portion is typically held in escrow for a period against warranty claims.
- Some may be contingent on future performance through an earnout.
- Some may be paid in acquirer stock rather than cash.
These are negotiable in ways the headline is not. The next post in the series, escrow versus earnout in an acquisition, goes through what each does to timing and certainty.
Step seven: consider secondary as a partial answer
The guide also asks you to consider secondary. Selling some shares in a funding round, where investors and terms permit it, converts part of a paper position into actual money without an exit. It is not always available, and it involves trade-offs: investors may read it as reduced commitment, and it is usually limited in size.
For a founder several years in with a meaningful position and no liquidity, it can substantially change their relationship to risk, which in turn improves their decision making about everything else. It is worth understanding as an option rather than treating an exit as the only route to any liquidity at all.
Step eight: keep the model current
Every round changes the waterfall. A model built once and never updated will be quietly wrong at exactly the moment you rely on it. Refresh it whenever the cap table changes, and put it in the same quarterly review as your cap table hygiene rather than in a separate life.
| Trigger | Action |
|---|---|
| A new round closes | Add its preference, order, multiple and participation to the list |
| A convertible instrument is logged | Record its terms and rerun the three prices |
| Quarterly review | Confirm the model still matches the legal records |
| Before any term sheet is signed | Run the new terms at the disappointing price first |
The last row connects to the main reason for the whole exercise. Once you have seen how preference structures behave at a mediocre price, you negotiate differently. The first reason it matters is in why your ownership percentage is not your payout.
Where tax fits
Tax is a real variable. Structure, jurisdiction, holding period and instrument type can change what you keep by a large margin, and many of the relevant decisions were made years earlier or must be made well before a transaction. This post does not tell you what to do, and nobody offering general tax advice about your specific situation should be relied on.
What can be said is that the cost of speaking to someone qualified early is trivial next to the cost of discovering a structural issue during a sale.
A file to keep beside the model
Numbers without context are easy to misread six months later. Keep a short working file next to the calculation, using the guide's own groupings, so that anyone picking it up can see what was assumed.
| Section of the file | What it records |
|---|---|
| The stack | Every preference, its order, its multiple, whether it participates and whether it converts |
| The three prices | The disappointing, plausible and good figures, and the range where common receives effectively nothing |
| The people | What employees would net, whether options must be exercised, whether unvested equity accelerates and whether any exercise window has expired |
| Headline versus received | How escrow, an earnout and acquirer stock would change what arrives and when |
| Tax | A note that someone qualified has been consulted, early |
The point of the file is to separate what you know from what you assumed. The calculation is only as reliable as the documents behind it, and a line saying that a document was not found is more useful than a number that quietly depends on a guess.
Set aside one afternoon for this, years before you need it. The guide's phrase is that it is the calculation that changes how you read every term sheet, and the effort is small beside what it reveals.
Frequently asked questions
Why three prices and not one?
One price hides the structure. A disappointing price reveals where terms decide who gets paid, a plausible one shows a realistic outcome and a good one shows that nearly any structure looks fine.
Do I need the actual documents?
Yes. Model what was signed, not a general assumption. Order, multiples and participation can differ between rounds, and the surprises live in those differences.
How often should I update the model?
Whenever the cap table changes, and at least in a regular review. A model built once and left alone will be wrong when you most need it.
Does the waterfall tell me what I will receive?
It shows the distribution. Escrow, an earnout, stock consideration, exercise costs and tax then change what arrives and when, so treat the result as a starting point and speak to someone qualified.
The full method, with the checklists, is in The Exit Playbook, a free 32-page playbook.