Post-Money SAFE: The Dilution Math Founders Get Wrong

Last month a founder came to us with a cap table showing four SAFEs at different valuations, $1.4M raised, and 17% gone before the Series A even started. He thought he still had 83% of the company. He did. But after modeling the A round he'd just signed a term sheet for, he'd land at 52%—and the lead wanted another 10% set aside for an exec hire. We had to walk him through how he'd already burned most of his negotiating cushion before the priced round ever happened.

The post-money SAFE is the standard instrument for US pre-seed and seed rounds now. Y Combinator introduced it in 2018 to fix the dilution mess the original SAFE created, and it works—if you understand the math before you sign. Most founders don't. They treat the cap as a valuation signal and ignore the ownership percentage that's locked in at signing.

This is the actual math, with examples we've seen go wrong.

How the post-money SAFE actually works

A post-money SAFE is a contract where an investor gives you money now and converts to equity later at a priced round. The "post-money valuation cap" sets the maximum company valuation their stake will be calculated against. The critical difference from the old pre-money SAFE: the investor's ownership percentage is fixed the moment you sign. Any SAFE you issue after that dilutes you, not them.

The formula is straightforward:

Investor ownership % = Investment amount / Post-money valuation cap

If someone invests $200K at a $4M cap, they own 5%. That's locked. If you raise another SAFE at an $8M cap for $800K, that investor owns 10%. Also locked. You now own 85%, and those two investors own 15% combined. All of this happens before any priced round dilution.

According to Y Combinator's SAFE documentation, this structure was designed specifically to make dilution predictable. It does—but only if you model it before you stack three or four SAFEs on top of each other.

A real scenario with two investors

Here's the YC walkthrough example, which mirrors what we see constantly:

Investor A puts in $200K at a $4M post-money cap. They own 5%.
Investor B comes in later with $800K at an $8M cap. They own 10%.

Total raised: $1M. Total ownership transferred: 15%. You still own 85% before the Series A.

Now you raise a Series A. Let's say it's $4M at a $20M post-money valuation. The lead takes 20%, and after syndicate participation the round totals 25% dilution.

Your ownership after conversion:

Investor A: still 5%
Investor B: still 10%
Series A: 25%
You: 60%

But wait. Most term sheets include an option pool expansion, and that typically comes out of the founders' share as a pre-money adjustment. If the A round term sheet specifies a 12% fully-diluted pool and you only have 4% allocated, you're expanding the pool by 8% pre-money. That's another ~6% of dilution on top of the 25% round dilution. You land at 54%.

We've seen founders sign term sheets without catching this. The headline valuation looks fine, but the post-close ownership is a shock.

The cap matters more than the discount

Some SAFEs include a discount—usually 20%—on the priced round price. In practice, the cap almost always controls. Here's why:

If your Series A prices at $25M post-money and your SAFE has a $10M cap with a 20% discount, the cap wins. The discount would give the investor a $20M effective valuation ($25M minus 20%), but the cap is lower, so they convert at the cap.

The discount only matters if the priced round happens below the cap, which is rare for a successful raise. If your A prices at $8M and your SAFE cap is $10M with a 20% discount, the discount gives them a $6.4M effective valuation, which is better for them than the cap. They take the discount.

Most investors assume the cap will control. So do we. If you're negotiating a SAFE, the cap is the number that matters.

Why the post-money version exists

The original SAFE, used from 2013 to 2018, had a pre-money valuation cap. The problem: when you stacked multiple SAFEs, each new SAFE diluted the previous SAFE holders, not just the founders. This made it nearly impossible to predict anyone's actual ownership at conversion. Founders would model their dilution, then discover at the priced round that everyone's math was wrong.

The post-money SAFE fixed this by locking each investor's percentage at signing. Subsequent SAFEs only dilute the founders. It's cleaner. It's also more founder-hostile in one specific way: you absorb all the dilution from a messy, multi-tranche raise.

The stacking trap

Here's a scenario we see every quarter. A founder raises in tranches:

$250K at a $5M cap = 5%
$400K at an $8M cap = 5%
$300K at a $10M cap = 3%
$500K at a $12M cap = 4.17%

Total raised: $1.45M. Total ownership transferred: 17.17%. The founder still owns 82.83%, which feels fine. But then the Series A happens. Another 25% gone to the lead and syndicate. Another 8% effective dilution from the option pool expansion. The founder lands at 49.83%.

At that point, there's no room for the next round. If the Series B takes another 20-25%, the founder is in the low 30s. The company is still fundable, but the founder has no leverage left. We've had to walk founders through bridge scenarios where they're negotiating from 28% ownership because they didn't model the cumulative dilution before they signed the fourth SAFE.

The fix is simple: model the full dilution path before you sign anything. Build a spreadsheet that shows your ownership after the current SAFE, after a hypothetical next SAFE, and after the Series A at a realistic valuation. If you're going to land below 55% after the A, you need to either raise less on SAFEs or raise at higher caps.

Current market caps

We track caps across the deals we work on. Here's what we're seeing in Q1:

Pre-product, strong team: $5-10M caps
Pre-seed with early traction: $8-15M caps
Seed-stage, revenue or strong growth: $12-22M caps
Bridge before Series A: $20-40M caps

These ranges vary by geography and sector. AI companies are seeing caps 30-50% higher than SaaS companies at the same stage. Fintech is somewhere in between. If you're fundraising in India, expect caps to be 20-40% lower than US equivalents, though that gap has been closing.

The cap you can command depends on your team, your traction, and how much FOMO you can generate. If you have two term sheets at different caps, you can usually push the lower one up. If you have one term sheet, you're negotiating from the cap they offered.

The MFN clause

Most SAFEs include a "Most Favored Nation" clause. It says: if you issue a SAFE with better terms to a future investor—lower cap, bigger discount, anything more favorable—the earlier investor can elect to take those terms retroactively.

MFN is standard. It's also a trap if you're not careful. Let's say you raise $500K at a $10M cap from your first investor, then three months later you're desperate and you offer an insider $100K at an $8M cap to close a gap. The first investor can invoke MFN and retroactively convert at the $8M cap. Their ownership just went from 5% to 6.25%. You gave away an extra 1.25% without meaning to.

We've seen this happen. The fix: don't lower the cap mid-round. If you need to offer better terms to close the round, offer a discount or warrants, not a lower cap.

Negotiating the cap

The cap negotiation is usually the only negotiation. Here's how it goes in practice:

You anchor high. If the market for your stage and traction is $10M, you open at $14M. The investor counters at $8M. You meet at $11M or $12M. If you have multiple term sheets, you can push higher. If you don't, you're negotiating against the investor's internal model of what your next round will price at.

The best leverage is the first cheque. Once one investor commits at a specific cap, others tend to anchor there. If you can get a name investor to commit at $12M, the rest of the round usually closes at $12M or within $1-2M of it.

Do not bend on the cap mid-round unless you're genuinely desperate. Lowering the cap triggers MFN issues and signals weakness. If you're not getting traction at your target cap, the problem is probably your story, not the cap itself.

The mistakes we see repeatedly

Founders stack SAFEs without modeling the cumulative dilution. This is the most common mistake, and it's the easiest to avoid. Build the model before you sign the third SAFE.

Founders mix pre-money and post-money SAFEs. Don't. If you issued a pre-money SAFE in 2017 and you're raising now, convert it or buy it out before you issue post-money SAFEs. The math gets messy fast.

Founders issue SAFEs without a lawyer reviewing them. The YC template is clean, but if an investor sends you a modified version, have someone who knows securities law look at it. We've seen SAFEs with liquidation preferences, pro-rata rights, and board observer seats buried in the terms. Those provisions don't belong in a SAFE.

Founders forget about MFN obligations. If you're raising a rolling close and you drop the cap halfway through, you just gave everyone who already signed the lower cap. Plan your cap strategy before you start taking money.

Founders treat the SAFE as "not really an investment." It is. You've sold equity. The fact that it hasn't converted yet doesn't mean it's not real. Model it as if it converted today, because that's the closest approximation of what your cap table will look like after the priced round.

When to use a SAFE instead of priced equity

SAFEs work well for pre-seed and small seed rounds—anything under $1.5M where you're raising from angels and small funds who don't need preferred share rights. They're fast, they're cheap, and they don't require a board seat or complex governance terms.

Priced equity works better for larger seed rounds, especially if you have an institutional lead who wants preferred shares, liquidation preferences, and board representation. Once you're raising $2M or more, most leads will push for a priced round. You can resist, but if they're writing a $1M+ cheque, they usually win that negotiation.

We've also seen hybrid structures: a lead invests on a priced equity term sheet, and the rest of the round closes on SAFEs that convert on the same terms as the lead's preferred shares. This works if the lead is comfortable with it, but it adds complexity.

What founders should do before signing

Build a dilution model that shows your ownership after the current SAFE, after any additional SAFEs you're planning, and after the next priced round at a realistic valuation. Use Carta, Pulley, or Capdesk to track everything. These tools will show you the waterfall at exit and help you avoid the mistakes we've covered here.

If you're about to sign a SAFE and the math feels unclear, stop. Get someone to review it. The cost of fixing a bad SAFE after you've signed it is much higher than the cost of modeling it correctly before you sign.

We help founders structure SAFE rounds so the Series A term sheet doesn't blow up their cap table. <Link href="/book/us">Book a call</Link> if you want to walk through your specific scenario.

The post-money SAFE, explained: Y Combinator's instrument dissected