You’re a second-time founder. You’ve done this before. You know what a SAFE is. Over fourteen months you raise $12M in pre-seed and seed SAFEs from 47 different investors, on the standard Y Combinator post-money form, with valuation caps from $10M to $35M depending on when each one came in. Six side letters. Information rights, pro rata, one MFN. You don’t track the MFN.
You then raise a $40M Series A at a $200M post-money. Term sheet is clean. Lead is reputable. You sign on a Thursday. Closing is three weeks out.
Monday morning of closing week, your lawyer sends the final pro forma at 1:47 a.m. You and your co-founder, together, own 18 percent of the company. Not the 35 you modeled. Not the 28 your cap-table software said last month. Eighteen.
The option pool didn’t grow. It was reserved pre-money, out of your stake. The SAFEs converted at every cap they were entitled to. The MFN dragged every SAFE down to the lowest cap on the stack. And two of the side letters promised pro rata into the Series A on an as-converted basis that your lead now wants to honor out of the common pool rather than blow up the round.
You call your lawyer. [S: hopefully it’s someone who knows this issue cold.]
This is the story of a meaningful share of the priced rounds closing in 2026. It’s happening right now at yours.
What a SAFE actually is
A Simple Agreement for Future Equity is a contract. Not stock. Not a convertible note. A SAFE gives the investor a contractual right to receive shares of preferred stock at a future priced round, calculated according to the formula in the document. Until that priced round, the investor owns nothing. No equity, no interest, no maturity, no dividend. Only a promise that says: when the next round happens, give me shares on these terms.
The promise is the product. The terms are the ammunition.
The original 2013 SAFE was pre-money. It was designed to be dilutive to the priced-round investors, not to the founders, and it didn’t cleanly stack when a company raised multiple SAFEs at different caps. The 2018 revision moved to post-money by default. That one change rearranged who bore the dilution, and almost no founder has reread their form since.1
The post-money problem in one paragraph
Under the post-money SAFE, the investor’s ownership percentage is fixed at conversion relative to the company’s post-money capitalization at the time of the priced round. That base excludes the shares issued in the priced round itself but includes all other SAFEs, all unallocated option pool (if the pool is set pre-money), and all other convertible instruments. Plain English: every additional SAFE you sell dilutes you, not the earlier SAFE investors. Every expansion of the option pool required by your Series A lead dilutes you, not the SAFE investors. Every instrument that converts into the priced round takes its fixed slice first, from the common stock. The common stock is you.
The post-money SAFE is a wonderful instrument for investors. It is a wonderful instrument for founders too, if the founder fully understands that each SAFE sold is a locked-in percentage of the company, and that every later decision about pool size, additional SAFEs, and priced-round mechanics falls on the common.
Almost no founder fully understands this on the day they sign the second SAFE.
The four mechanics that eat your cap table
Cap stacking and the MFN cascade. When you sell SAFEs over twelve to eighteen months at different caps, the investors with higher caps often ask for a Most Favored Nation clause. It entitles them to the terms of any SAFE you later sell on better terms. Read the MFN. Some are narrow (economic terms only). Some are broad (all material terms). A broad MFN means every SAFE you later sell at a lower cap automatically pulls every MFN-holding SAFE down to that lower cap. You can trigger the cascade by accepting a single $10M check from an angel on generous terms. The cascade closes on the night before your Series A, when your lawyer maps every MFN to every later instrument and the lowest cap wins for everyone who asked. [M: this is really easy to miss.]
The pre-money option pool shuffle. Your Series A lead requires you to expand the employee option pool to, say, 15 percent of the post-closing fully diluted capitalization, pre-money. That phrase, pre-money, is the whole game. Every share added to the pool above the current pool is dilutive to the founders and to the SAFE holders, but not to the Series A. A six-point pool expansion at a $200M post is $12M of economic value transferred from founders and SAFE holders to the Series A. The SAFE holders, by the terms of the post-money SAFE, do not share this dilution if the pool is set pre-money. You do.
Side letters that survive. Six of your 47 SAFE investors signed side letters. Pro rata rights. Information rights. Board observer seats. MFN extensions. A promise to include them in all future financings. Most founders treat side letters as a formality. They are not. They survive the SAFE conversion, bind the company in the priced round, and often contain promises the Series A lead will require you to break, or pay out of the common stock to honor. The negotiation happens on the eve of closing. The founder loses.
Discount vs. cap interactions, and the “most favorable to investor” default. Your SAFE probably has both a discount and a cap. At conversion, the investor takes whichever is more favorable to them. That’s fine. It’s the standard. What’s not fine is how that interacts with the pre-money valuation of the priced round, the new-money MFN side letters that ask for both terms improved, and the no-shop in your term sheet. A creative investor with a strong lawyer can pull three or four percent of your company out of the cracks between the SAFE, the side letter, and the term sheet. There is no clean mechanical defense once the ink is dry. [S: the strong lawyer is not predatory. They are precise, and have read more cap tables in five years than most senior partners read in a career.]
What changed about this problem
Post-money SAFEs have been the default since 2018. The compounding damage has been there the whole time. What changed is the shape of early-stage financings in 2026.
Rounds close faster, at higher numbers, with more investors, over longer SAFE windows, at a wider spread of caps, with more side letters, and with less time for founders to run the pro forma before signing. Pre-seed is now routinely $15M to $30M. Seed can close in a week on two dozen investor names. A founder who started selling SAFEs twelve months ago has 30 counterparties, eight caps, four side letters, two MFNs, and a memory. The math across those instruments isn’t doable in a head. It’s barely doable in a spreadsheet. It’s reliably doable only in the kind of model that runs every SAFE against every scenario before any new piece of paper gets signed.
Almost nobody runs that model before signing. The model runs the night before the Series A closing, and the founder finds out on Monday morning what happened in the nine months they weren’t looking.
If you’re mid-raise and have not modeled the stack against your priced round, that conversation is worth having before the next check clears.
Talk to a Talairis attorney →What this means for you
Three things almost every founder gets wrong.
Every SAFE is a commitment, not an option. The moment it’s signed, a fixed percentage of your company is gone to that investor, calculated off a post-money base you haven’t yet written. You can’t un-sign. You can’t negotiate it down when the numbers embarrass you. The only lever you have is the pre-money valuation of the priced round, and raising that valuation costs you leverage in other terms that also matter.
The side letter is the real contract. Founders focus on the cap and discount. Professional investors focus on the side letter. An MFN plus pro rata plus information rights is often worth more than a 2× better cap. The side letter is where the post-financing friction lives. Read it before you sign it. Have counsel negotiate it, not just review it. Don’t sign one you haven’t mapped against every other side letter in your stack.
The cap table software is not telling you the truth. Every founder-facing cap table tool makes simplifying assumptions.2 Most of them model the MFN cascade incorrectly. Most of them don’t properly handle pre-money option pool expansions against post-money SAFE bases. Most of them don’t surface side letter obligations. The pro forma you show your investors the week before the Series A is not the pro forma your lawyer will produce on closing night. The gap between the two is where your equity goes.
What to do
Four things, in order.
- Pro forma every new SAFE against every existing SAFE, every side letter, every MFN, and three priced-round scenarios, before you sign. Low, expected, and high post-money. Model the option pool being set pre-money at the number your likely Series A lead will demand. The five minutes it takes to find out that the next check costs you six points of the company is the most valuable five minutes in the raise.
- Cap the MFN scope at signing, or don’t agree to one. A broad MFN is a loaded weapon pointed at your future cap table by a counterparty you won’t remember. If you must grant an MFN, narrow it to economic terms and sunset it at the priced round or after a fixed window.
- Standardize side letters, centralize them, and reread them before every new instrument. Keep one binder. Every side letter in it. Every obligation extracted and mapped. No new SAFE, no new investor, no new priced-round term sheet gets signed until someone has read the binder and confirmed you’re not about to breach it.
- Negotiate the option pool in the term sheet, not after. Whether the pool is set pre-money or post-money is worth more than most of the other economic terms in the term sheet combined. If your lead insists on pre-money, price it in by negotiating a higher pre-money valuation. If they won’t move, understand exactly what the pool expansion costs you and your SAFE holders, and decide whether the round is still worth taking.
Get counsel before the next bridge
SAFEs look simple. That’s the design. Two pages, a standard template, the brand-name accelerator’s blessing. The apparent simplicity is what makes founders skip the math.
The math is where the company goes. A founder who raises $12M across 47 SAFEs and then signs a Series A term sheet without re-running the pro forma is making the single most expensive mistake of their career in front of their seed investors, and none of the seed investors have a fiduciary duty to warn them.
Before the next SAFE (not the next round, the next SAFE), have counsel model the stack. Before the next side letter is countersigned, have counsel add it to the binder. Before the term sheet is signed, have counsel run the closing-night pro forma against all three pool-size scenarios. Each of those steps costs less than the dilution it prevents.
A closing thought
Your SAFE is not safe.
It is a contract that fixes someone else’s ownership percentage against a number you have not written yet, in a stack of other contracts you have not reread, with side letters that bind the company and MFNs that cascade on the night of your priced round.
The only thing simple about a Simple Agreement for Future Equity is the paper. The cap table it produces, at the moment you finally raise a real round, is the least simple document in your company. The number at the bottom of the founder row is the one you live with.
Run the math before the next check clears. Nobody else will.
- VHS versus Betamax, and we watched it unfold. SAFEs beat convertible notes in the 2010s on a founder-friendly pitch. Then the 2018 revision moved them to post-money, and the form that won on founder protection became much more investor friendly. ↩
- Every cap-table management solution has its issues. ↩
