Cap table

Dilution + liquidation preference: the number copied memos get wrong

This is the single most common hole in a first-time manager's memo. You buy 5% of a company, so you tell your LP "5% at exit." But after two more rounds you don't own 5% anymore, and at a modest exit a preference stack of senior money gets paid before you see a dollar. The honest number can be a fraction of the headline — sometimes a loss where the memo claimed a win. Here's the arithmetic, step by step.

Textbook finance only (MOIC, IRR, dilution, liquidation preference). The formula below is exactly what the kit's model and the landing calculator use.

Two things happen between entry and exit

1 · Dilution. Every later financing issues new shares, so your percentage shrinks. If a round dilutes you 20%, you keep 80% of what you had. Dilution compounds across rounds — you multiply, you don't subtract:

diluted% = entry% × (1 − d₁) × (1 − d₂) × …

2 · The preference stack. At exit, holders of senior preferred typically get their money back first (a "1× liquidation preference," sometimes participating). Only what's left flows to common — and to you, if your shares convert to common. At a big exit this barely matters. At a small one, it can wipe you out.

Worked example

Say you write a $1.0M check for 5.0% at entry. Two later rounds dilute you 20% then 18%. There's a $40M senior preference stack ahead of common. Step through it:

Step 1 — dilution

You entered at 5.0% and you'll exit owning 3.28% — a third less than the headline, before the preference stack even enters.

Step 2 — the exit waterfall

At exit, first subtract the senior preference; whatever remains is the proceeds-to-common, and your slice is your diluted %:

toCommon = max(exit − seniorPreference, 0)
yourProceeds = toCommon × diluted%
MOIC = yourProceeds ÷ yourCheck

^† Two correct numbers sit on this row, and which one you show is a stated modelling choice, not a contradiction. This $250M exit sits above the conversion point, so the honest headline — the figure the sample memo uses — is 8.2×: the senior preferred voluntarily converts to common (converting pays more than taking the $40M preference), the stack falls away, and everyone shares pro-rata on the full $250M. The calculator below is deliberately pinned to the conservative floor — the "preference always paid first" rule — which reads this row as $210M to common → 6.89×. That's not the model disagreeing with itself; it's the downside case by design, so you never over-promise above the conversion point. Below the conversion point — where the stack actually bites and copied memos go wrong — the two are identical. The sample memo shows both on the same waterfall: the 6.89× conservative floor the model computes, and the 8.2× headline once you apply the one-line conversion check (compare "take the $40M preference" vs "convert and take 3.28% of the full exit," and use whichever pays more). Same inputs, both reconciled, nothing hidden.

Look at the $50M row. A memo that skipped this math would say "3.28% of $50M = $1.64M, a 1.6× — fine." The reconciled number is 0.33× — you lose two-thirds of your money — because the $40M preference eats almost the whole exit before common sees anything. That is the number copied templates get wrong.

Run your own numbers

Set your check, ownership, number of dilution rounds, exit value, and preference. It uses the exact formula above — the conservative exit − preference case, so it shows you the downside a copied memo hides.

Illustrative modelling with assumptions you set — not a forecast. The landing page has the same estimate wired to a check-size slider; this one adds the dilution-round and preference controls.

Why this earns credibility with an LP

When an LP asks "what do you actually own after two rounds and the stack," the manager who freezes loses the room. The one who says "3.28% diluted, and here's the waterfall — at a modest exit this position is a loss, which is why I priced my entry the way I did" sounds like an institution. The math isn't hard; it's just usually skipped.