Valuation Caps and Discounts in SAFEs: What Founders Get Wrong
Valuation caps and discounts are simple in isolation. The mess starts when you offer different terms to different investors — each unique combination creates a separate series of preferred stock at conversion.
SAFEs and convertible notes exist to make early fundraising simple. One-page document. No board seat negotiation. No priced round mechanics. Just money in, equity later, terms everyone can understand.
In theory.
In practice, we regularly see founders turn this elegant instrument into an absolute mess — not because they don't understand SAFEs, but because they don't understand what happens when the SAFEs convert. Specifically, what happens when you've issued SAFEs at five, six, or eight different sets of terms to different investors.
We had a company come to us recently with SAFEs outstanding at eight different valuation caps. Eight. They thought they were being flexible and founder-friendly. What they actually created was a conversion nightmare that would require eight separate series of preferred stock at their next priced round.
Let's back up and explain why.
How caps and discounts work
If you already know this, skip ahead. If not, this is worth understanding clearly before we get to the part where it goes wrong.
The valuation cap
A valuation cap sets the maximum effective valuation at which a SAFE converts into equity. It protects the investor's upside — no matter how high your Series A valuation goes, the SAFE holder converts as if the company were valued at or below the cap.
Example: An investor puts $100K into a SAFE with a $5M post-money cap. You raise a Series A at a $20M pre-money valuation. Without the cap, their $100K converts at the Series A price. With the cap, they convert as if the company were valued at $5M — giving them roughly 4x more shares per dollar than the Series A investors.
The cap is the primary economic term in most SAFEs. It's the number investors care about.
The discount
A discount gives the SAFE holder a percentage reduction on the price per share at conversion. If the Series A price is $1.00 per share and the SAFE has a 20% discount, the SAFE converts at $0.80 per share.
Discounts are most common in convertible notes and less common in post-money SAFEs (where the cap does most of the economic work). When a SAFE has both a cap and a discount, the investor converts at whichever method gives them more shares — the lower of the two calculated prices.
How they interact
Most SAFEs include a cap, a discount, or both. The MFN clause in a SAFE without a cap gives the investor the right to adopt better terms from a later SAFE. In a post-money SAFE (the current Y Combinator standard), the cap defines the investor's ownership stake directly — the math is simpler and more predictable than the pre-money version.
So far, so good. Each of these terms is straightforward on its own.
The problem starts when you give different terms to different investors.
The conversion problem nobody warns you about
Here's the thing founders miss: every unique set of economic terms converts into a separate series of preferred stock.
This isn't a legal technicality. It's structural. When your SAFEs convert in a priced round, the company issues preferred stock to the converting holders. The price per share each investor receives is determined by their specific cap or discount. If two investors have the same cap and the same discount, they convert at the same price into the same series. If they have different terms, they convert at different prices — and different prices means different series.
A typical clean conversion looks like this:
- All SAFEs at $8M cap, no discount → convert into Series Seed at the same price → one series of preferred stock
- Series A investors → Series A preferred stock
- Total: two series
Now look at what happens when the founder got creative:
- Investor A: $5M cap, no discount
- Investor B: $6M cap, 20% discount
- Investor C: $8M cap, no discount
- Investor D: $8M cap, 10% discount
- Investor E: $10M cap, 15% discount
Each of those converts at a different effective price. Each different price requires its own series. That's potentially five series of preferred stock — before the new money even comes in.
Five series means:
- Five sets of conversion calculations in the financing docs
- Five line items on the cap table
- Five sets of shares to track in Carta
- A financing closing that takes longer, costs more in legal fees, and introduces more room for error
- An investor's counsel who opens the data room and immediately asks: what happened here?
The company we saw with eight different caps? Their Series A closing required eight sub-series of converting preferred stock. Their legal bill for the round was meaningfully higher than it needed to be. The investors' counsel flagged it as a governance concern. The founders had no idea this was coming.
Why founders do this
Usually for perfectly understandable reasons:
The round takes longer than expected. You close your first SAFE at an $8M cap. Three months later, you've hit some milestones. The next investor agrees to a $10M cap. Two months after that, a strategic angel wants in but at a $6M cap. Each negotiation felt reasonable in the moment. The result is three different series at conversion.
They want to reward early believers. Giving a lower cap to the first check feels like the right thing to do. And it is — in moderation. One or two cap tiers is manageable. Eight is not.
Nobody told them. This is the most common reason. The standard SAFE is a four-page document with no mention of what happens structurally when it converts alongside SAFEs with different terms. Founders assume it all works out. It does — it just works out expensively and messily.
What to do instead
Pick one cap and stick with it. The simplest approach: every SAFE in a given round converts on the same terms. One cap, one discount (or none), one series at conversion. Communicate this clearly to investors upfront — most will understand and respect the discipline.
If you must offer different terms, limit the tiers. Two cap levels is manageable — an early-bird cap for the first tranche and a standard cap for the rest. Beyond two, you're creating complexity that doesn't serve anyone.
Use the MFN clause properly. If your first SAFE includes an MFN clause, early investors can elect to adopt the terms of any later SAFE. This means if you lower the cap for a later investor, earlier investors can match it — which effectively collapses everyone back to the same terms. This is the intended use of MFN, and it works. But it only works if the MFN is actually in the document and if the company properly notifies investors of subsequent issuances.
Model the conversion before you sign. Before issuing a SAFE at a new cap or discount, model how it converts alongside your existing SAFEs. See how many series it creates. See what the cap table looks like after conversion. If the model shows four series of converting preferred, that's your signal to simplify before signing.
Talk to counsel before the third SAFE. The first two SAFEs are usually straightforward. By the time you're issuing a third at different terms, you're creating structural decisions that will affect your next round. A ten-minute conversation now saves weeks of cleanup later.
The bottom line
SAFEs were designed to be simple. One page, one set of terms, money in the bank. The instrument works beautifully when you respect that simplicity. It breaks down when founders treat each investor negotiation as independent without thinking about how they all interact at conversion.
Every unique set of terms is a future series of preferred stock. Every series adds complexity, cost, and questions from future investors. The founders who raise cleanly aren't necessarily the ones with the best terms — they're the ones who kept it simple enough that the conversion is boring.
Boring conversions are the best conversions.
Need legal guidance for your startup?
Book a free intro call and see how Flux can help.
Book a Free Call