SAFEs, caps and founder dilution

Most first money into a company now arrives on a SAFE. It is a short document, it closes in days, and it postpones the hard conversation about price. That is the appeal, and it is also where founders get hurt: the conversation is postponed, not cancelled, and the terms you agreed in a hurry decide how it goes.

Nothing below is exotic. It is arithmetic that takes ten minutes and that a surprising number of founders do for the first time in the week a Series A term sheet lands.

What a cap is, and what it is not

A valuation cap is a ceiling on the valuation used when the SAFE converts into shares. It is not a valuation, it is not a promise, and it is not a floor on anything.

Say you raise $500,000 on a SAFE with a $10 million post-money cap. Eighteen months later you raise a priced round at $30 million. The SAFE does not convert at $30 million — it converts at the cap, so that investor pays roughly a third of what the new investor pays for the same share. That is the point of the instrument. They took the risk early and the cap is how they are paid for it.

The word doing the work is post-money. Since 2018 the standard document has been the post-money SAFE, and the difference from the older pre-money version is not cosmetic.

Post-money means the investor’s percentage is fixed and yours is not

Under a post-money SAFE, the cap is measured after that SAFE is counted. The consequence is simple and most of the trouble comes from missing it:

The investor’s ownership percentage is locked at signing. Everything that happens afterwards dilutes you, not them.

$500,000 on a $10 million post-money cap is 5% — not approximately, not subject to the next round. Five per cent, guaranteed, and if you later raise more money on more SAFEs, that 5% stays 5% while your share falls.

Under the old pre-money SAFE, later SAFEs diluted earlier ones too, and everyone found out where they stood at conversion. The post-money version moved all of that uncertainty onto the founders. It is cleaner, it is more honest to the investor, and it is more expensive to you than it looks.

Stacking is where it goes wrong

One SAFE is easy to reason about. Three are not, and three is normal.

Run it as percentages, because percentages add up and valuations do not:

$500K at a $10M post-money cap — 5%
$750K at a $15M post-money cap — 5%
$400K at a $20M post-money cap — 2%

That is $1.65 million raised and 12% of the company already committed, before a priced round exists, before the option pool is topped up, and before the new investor takes their share.

Now add a Series A buying 20%, and a pool top-up of 10% that — as is standard — comes out of the pre-money, meaning out of the existing holders. The founders are not giving away 20%. They are giving away 20% of what is left after the SAFEs convert, after the pool, and the compounding is unkind.

Do this sum before you sign the second SAFE, not after the third. It takes a spreadsheet with four rows.

The discount, and why it usually does not matter

Many SAFEs carry a discount — typically a percentage off the round price — alongside the cap. The holder gets whichever gives them more shares.

In practice, if the company has gone well, the cap is lower than the discounted round price and the cap wins. The discount is the term that protects the investor when the company has gone sideways and the round prices below the cap. Founders negotiate hard on the discount and lightly on the cap. It is almost always the wrong way round.

The MFN clause nobody models

A “most favoured nation” provision says that if you later issue a SAFE on better terms, the earlier holder may take those terms instead.

It is reasonable and it is common. It is also a loaded gun pointed at your own runway: it means you cannot quietly give a better cap to the investor you need next month without regrading everyone who came before. Before you agree to an MFN, know which of your existing holders has one, because the cheapest money available to you later may turn out to be the most expensive.

What we tell founders

Count percentages, not dollars. Dollars raised is the number founders quote to each other. Percentage committed is the number that decides whether your Series A is financeable.

Decide the floor you will not go under, in writing, before the first SAFE. How much of the company do the founders need to hold going into a Series A for the round to work and for everyone to stay motivated? Write that number down while nothing is at stake. It is an easy number to abandon one SAFE at a time.

A high cap is not a win on its own. A cap far above what the next round will support does not make the problem go away — it moves it to the term sheet, where a new investor has to reprice in front of you and someone has to be disappointed.

If a round is coming, consider pricing it. SAFEs are a tool for speed when pricing is genuinely impossible. Once you have revenue, comparable companies and real interest, “we will decide later” is sometimes just avoidance, and a priced round settles the cap table instead of deferring it.

The honest summary

The SAFE did what it was designed to do: it made early money fast and cheap to accept. What it did not do is remove the consequence, and the post-money version moved the whole consequence onto the founders’ side of the table.

You do not need a lawyer to see it coming. You need a spreadsheet, four rows and one honest afternoon before you sign the next one.

If you want someone to go through yours line by line, ask. We read cap tables for a living and have no interest in selling you an instrument. If the structure question underneath this is where the company should be incorporated at all, our flip tool and the longer piece behind it deal with that one.

General guidance, not legal or financial advice. The figures above are arithmetic on illustrative numbers, not market data — your terms are whatever your documents say. Have counsel read them before you sign.