How the Post-Money SAFE Actually Works for Founders

The safe (simple agreement for future equity) is the default instrument for pre-seed and seed rounds in the United States, and since 2018 the standard form has been the post-money safe. The switch was not cosmetic. It moved a specific block of dilution off the safe investors and onto the founders, and in exchange it made the one question founders most often get wrong answerable on the day the money lands: how much of the company did we just sell?

What “post-money” actually means here

Y Combinator, which wrote and open-sourced the safe, publishes three US variants: a safe with a valuation cap and no discount, a safe with a discount and no valuation cap, and an “uncapped MFN” safe with neither. Non-US founders get Canadian, Cayman Islands and Singapore versions, currently in the valuation-cap-only format. All of them are post-money.

The confusing part is that “post-money” here does not mean what it means on a priced round. YC’s own primer for the post-money safe is blunt about it:

“The Post-Money Valuation Cap is ‘post’ all of the safe money. It is NOT also ‘post’ the Equity Financing.”

So a $10,000,000 post-money cap means the cap is measured after every safe in the round has converted, but before the Series A cheque arrives. That is what makes the arithmetic close. If you invest $500,000 on a $10,000,000 post-money cap, you own 5% of the company as of conversion, full stop, regardless of how many other safes the founders signed.

The mechanics

Two defined terms do the work. Company Capitalization is the denominator: issued and outstanding shares, all converting securities including the other safes, granted and promised options, and the existing unissued option pool. It deliberately excludes any new or increased option pool adopted as part of the equity financing. Safe Price is then simply:

Safe Price = Post-Money Valuation Cap / Company Capitalization

Divide the investment by the Safe Price and you get the share count. Because Company Capitalization includes the other safes, the equation is self-referential and resolves to a fixed percentage.

Cap, discount, or MFN

Founders mostly negotiate one number, the cap, which is part of the instrument’s appeal. But the three variants behave differently, and the uncapped MFN version in particular is doing more than it looks like.

The MFN clause matters more than founders expect because a single later safe at an aggressive cap retroactively re-prices every MFN safe above it. YC’s own standard deal is built on exactly this: as of 2026 it invests $500,000, of which $125,000 buys a fixed 7% on a post-money safe and $375,000 sits on an uncapped MFN safe. YC’s worked example on that page: at a $15,000,000 post-money cap, that $375,000 becomes 2.5% of the company. Version 1.2 of the MFN safe clarified that side letters covering non-economic rights, such as board observer seats or information rights, do not trigger the MFN.

Why pre-money safes broke

Under the original pre-money safe, introduced in 2013 as a replacement for convertible notes, the cap was a pre-money number and Company Capitalization excluded the other converting safes. Every safe therefore priced itself off the same pre-safe share count, which meant the safes diluted each other, and nobody could say what had been sold. YC described the result as a “recursive loop.”

Take a company with 10,000,000 shares on its cap table before any safes: 9,000,000 held by two founders and a 1,000,000-share option pool. It raises $2,000,000 across three safes at caps of $10m, $12.5m and $15m.

Under pre-money safes, each converts off the 10,000,000-share denominator:

1. $10m cap gives a price of $1.00 per share, so $500,000 buys 500,000 shares.
2. $12.5m cap gives $1.25 per share, so $500,000 buys 400,000 shares.
3. $15m cap gives $1.50 per share, so $1,000,000 buys 666,667 shares.

Total shares after conversion: 11,566,667. The first investor thought she was buying 5% of the company. She owns 4.32%. The founders’ 9,000,000 shares are 77.8% of the total. Every additional safe the founders signed quietly shrank the first investor’s stake, and neither side had modelled it.

Under post-money safes with the same three caps, the percentages are fixed at signing: 5.0%, 4.0% and 6.67%, or 15.67% in total. Existing holders keep 84.33%. Company Capitalization resolves to 10,000,000 / 0.8433, or 11,857,708 shares, so the first safe converts at roughly $0.84 per share into 592,885 shares. The founders’ 9,000,000 shares are now 75.9%.

Note what happened: at nominally identical cap numbers, the founders ended up with less under the post-money form, 75.9% against 77.8%. That is the trade. The post-money safe is more dilutive at the same headline cap, because it makes the safes stop diluting each other and puts that dilution back on the common stock. What it buys the founder is a number that is knowable before the Series A.

Carrying the dilution into a priced round

The safes’ 15.67% is a pre-Series A figure. Continue the example: the company raises $6,000,000 at a $30,000,000 post-money valuation, so new investors take 20%, and the term sheet requires a new 8% post-closing option pool. Because Company Capitalization excludes the new pool, both the safes and the founders are diluted by it.

Existing holders retain 72% of the company, so every pre-round percentage is multiplied by 0.72. The safe investors’ 15.67% becomes 11.28%. This is the design intent YC states directly: safes are not diluted by options granted or pools created between the safe round and the Series A, because the safe money paid for that hiring, but they are diluted by the new pool created as part of the Series A, because otherwise founders would absorb two rounds of hiring dilution on one round of capital.

Pro rata is now a separate negotiation

The original safe carried a pro rata right by default. The post-money safe stripped it out into an optional pro rata side letter. That is a meaningful change and one worth being deliberate about: granting pro rata to a long list of small cheques can make a Series A hard to allocate, because a large share of the round is pre-committed to existing holders before the lead has taken its position. YC’s guidance is that there is not much to the instrument beyond asking for the letter and negotiating the ask, which understates how much modelling founders should do before handing several of them out.

Three things to do before you sign

1. Add up the percentages, not the dollars. Every post-money safe you sign is investment divided by cap. Keep a running total, and compare it to the market: Carta’s Founder Ownership Report 2026 puts median founder ownership at 56% fully diluted after a seed round and 36% after a Series A. If your safes alone have already sold a fifth of the company, a Series A lead will do that arithmetic for you and may not like the answer.
2. Treat an uncapped MFN as a blank cheque on price. It will adopt the best terms you issue later, so the discipline you apply to the next safe’s cap is the discipline you are applying to this one.
3. Model the Series A pool now. The new pool comes out of pre-round holders, and at 8% to 10% it is often the second-largest line of dilution in the round after the money itself.

Use the actual forms. They are free, they are the market standard, and the defined terms are the whole point of the instrument. A redlined bespoke safe from a founder’s cousin’s law firm reintroduces exactly the ambiguity the post-money form was written to remove.

Sources

• Y Combinator — The SAFE: the open standard for startup fundraising (safe forms and user guide)
• Y Combinator — Primer for the post-money safe, v1.1 (PDF)
• Y Combinator — The YC Deal
• Y Combinator — Announcing the Safe, a Replacement for Convertible Notes (6 December 2013)
• Y Combinator — Safe User Guide, February 2023 (PDF)
• Carta — Founder Ownership Report 2026