How to model multiple SAFEs with different valuation caps and MFN clauses

The right way to model multiple SAFEs with different valuation caps and MFN clauses is to treat each SAFE as a separate, independent conversion: build a complete SAFE schedule, resolve all MFN elections first, calculate each conversion price individually, and never aggregate SAFEs before converting them. Getting that sequence wrong produces a materially incorrect cap table.

Most founders raise their first SAFE and think the math is simple enough to track in their head. Then they raise a second. Then a third, with a higher cap, and one of the earlier investors has an MFN clause. By the time a Series A term sheet lands 18 months later, the cap table is a tangle of different caps, a clause nobody fully remembers, and a conversion calculation that gives a different answer every time. This guide walks through the exact process, from building your SAFE schedule to running scenarios before the term sheet closes.

Pre-money vs post-money SAFEs: name the structure first

The post-money SAFE, introduced by YC in 2018, is now the dominant standard in the US. Under a post-money SAFE, each investor's ownership percentage is fixed at signing: their investment divided by their valuation cap. A $500K investment at a $5M post-money cap is exactly 10%, regardless of how many additional SAFEs are raised after that. You will still encounter pre-money SAFEs in older notes, but if you are signing today, it is almost certainly post-money. This guide uses post-money SAFE mechanics throughout, with all three example SAFEs as cap-only (no discount rate).

Building a SAFE schedule: what to track per SAFE

Before any conversion math, founders need a complete SAFE schedule. Missing one field, especially MFN status, will cause errors downstream.

For each SAFE, record:

• Investor name
• Principal amount
• Valuation cap
• Discount rate (if any)
• MFN clause: yes or no
• Date issued
• Pro rata rights

Here is the example that runs through the rest of this guide. Three SAFEs raised across 18 months, at different caps, with one MFN clause:

SAFE 2 carries the MFN clause, and SAFE 3, the last note issued, has a lower cap. That combination triggers the MFN election.

Determine MFN elections before calculating anything

An MFN (most favored nation) clause is forward-looking: it gives a SAFE holder the right to amend their note to match the terms of any future SAFE issued on more favorable terms.

In the example, SAFE 2 was issued at an $8M cap. SAFE 3 arrived 12 months later at a $4M cap, triggering SAFE 2's MFN clause. The SAFE 2 investor can amend their note from $8M to $4M. By issuing SAFE 3 at a lower cap, the founder inadvertently gave an earlier investor the right to a much better deal.

Resolve all MFN elections before calculating any conversion prices. If you calculate SAFE 2's conversion price at its original $8M cap and then apply the MFN, your share counts are wrong from the start.

After applying the MFN election, the updated schedule reads:

Now the conversion math can begin.

The real dilution risk: the cap-to-valuation gap

The real risk is not the MFN clause itself. It is the gap between the valuation cap on an early SAFE and the valuation at which the Series A closes. A $500K SAFE at a $2M cap, with a Series A at $20M, means that investor converts at one-tenth the price the Series A investors pay. MFN amplifies this by pulling additional investors down to the lowest cap in the stack.

Calculate conversion prices per SAFE independently

For each SAFE, calculate the conversion price from the effective cap. The investor converts at whichever of the cap price or discounted round price gives them more shares. Since all three SAFEs are cap-only (no discount rate), the cap price is the only mechanism to evaluate against the round price.

For this example: the company has 8,000,000 fully diluted shares outstanding before the Series A. The Series A is raising $5,000,000 at a $12,000,000 pre-money valuation.

Series A price per share = $12,000,000 / 8,000,000 = $1.50

SAFE 1 ($300K, $4M cap):

• Cap price = $4,000,000 / 8,000,000 = $0.50
• Cap price ($0.50) is below the round price ($1.50), so the cap governs
• Conversion price: $0.50

SAFE 2 ($500K, $4M effective cap after MFN election):

• Cap price = $4,000,000 / 8,000,000 = $0.50
• Original cap price would have been $8,000,000 / 8,000,000 = $1.00
• After MFN election: conversion price $0.50 (not $1.00)

SAFE 3 ($500K, $4M cap):

• Cap price = $4,000,000 / 8,000,000 = $0.50
• Conversion price: $0.50

Shares issued per SAFE:

Without the MFN election, SAFE 2 would have converted at $1.00, producing 500,000 shares. With it, SAFE 2 converts at $0.50, producing 1,000,000 shares — double the count for the same $500,000 investment.

Convert each SAFE independently: the common mistake is aggregating

The temptation is to total up all SAFE principal, pick a blended average cap, and calculate a single conversion price. It is wrong. Averaging hides the actual dilution from lower-cap notes and obscures the MFN election's impact, and your cap table will be incorrect the moment someone runs the real numbers, creating problems for option pool sizing, ownership percentages, and potentially the closing itself. Always convert each SAFE on its own row, with its own conversion price.

The full picture after conversion

Series A shares = $5,000,000 / $1.50 = 3,333,333 shares

Post-conversion ownership (note: this example does not include an option pool, which in a real Series A would typically be established or expanded pre-close and would dilute founders further):

The founders started at 50% each; after three SAFEs and one priced round, they are at 28.7% each.

Compare SAFE 2's outcome with and without the MFN election:

Same investment, same Series A: the MFN clause nearly doubles the investor's ownership.

Solve the circularity problem: how Cake handles this

The conversion price formula uses fully diluted shares outstanding as the denominator. But converting the SAFEs issues new shares, which changes the share count, which changes the conversion price, which changes the share count again. The formula references itself.

In spreadsheet terms, this is a circular reference. Most spreadsheets either return an error or require iterative calculation to be enabled manually, and even then may converge to the wrong answer depending on how the model is structured. The correct approach is to iterate: calculate conversion prices using the initial share count, issue the new shares, recalculate using the updated share count, and repeat until the numbers stabilize.

Cake handles this automatically. For founders modeling in a spreadsheet: using the pre-conversion fully diluted share count as your denominator is a reasonable approximation for early-stage planning, but for anything going into a term sheet or closing document, use purpose-built software.

Run multiple scenarios: down, base, and upside

The dilution picture shifts with the Series A valuation, and modeling the range before you are in the room is the most useful thing a founder can do.

Two things move the needle most in a stacked SAFE scenario. First: the gap between your lowest effective cap and your Series A valuation; the wider that gap, the more shares the lowest-cap investors receive, and MFN pulls every electing investor down to that same price. Second: how many investors hold MFN clauses. One is manageable; multiple MFN investors all electing down to the same low cap is a different conversation.

One more point: if you raise a new SAFE after an MFN investor has signed, and that new SAFE has a lower cap, the MFN investor gets the right to elect again. The right resets every time a more favorable note is issued.

Down scenario ($8M pre-money): The Series A price per share drops to $1.00. All three SAFEs still convert at their $0.50 cap price (cap still governs). The Series A itself is smaller, so SAFE investors represent a larger proportion of the total post-conversion cap table.

Base scenario ($12M pre-money): As modeled above. 2,600,000 shares from SAFE conversion. Each founder at 28.7%.

Upside scenario ($20M pre-money): The Series A price rises to $2.50. All cap prices ($0.50) are still well below the round price, so all three SAFEs still convert at $0.50 and SAFE share counts are identical to the base scenario. The Series A issues more shares at the higher price, diluting everyone including the SAFE investors.

The goal is not to predict the exact outcome, but to give the founding team a range of dilution outcomes before the term sheet is signed so they can negotiate with accurate numbers.

A note on scope: MFN clauses can apply to any SAFE terms, not just the valuation cap. Discount rates and pro-rata rights can also be subject to election if the clause is broadly drafted. Confirm the scope with your lawyer before issuing a subsequent note.

Model it early, negotiate from knowledge

Stacked SAFEs with MFN clauses are manageable when you understand the sequence: build the schedule, resolve the elections, convert independently, and iterate to solve the circularity. The founders who model this before the term sheet arrives negotiate from a position of actual knowledge.