A SAFE Postpones the Valuation Argument Instead of Settling It

The Instrument

a simple agreement for future equity It gives an investor the right to receive shares in a financing with a future price in exchange for money now. It is not debt. There is no interest or maturity date

It exists because valuing a company with no income and three employees is almost arbitrary and negotiation consumes time that neither party has

The SAFE defers that negotiation until the point where there is enough information to carry it out letting the money arrive immediately. What it does not do and what its name slightly obscures is eliminate the negotiation. The price is still being set. It is being set on terms agreed upon today and applied to a number that no one knows yet

The Two Terms That Matter

a valuation limit sets a maximum valuation at which the SAFE converts. If the company then raises forty million with a SAFE capped at ten million the SAFE holder converts as if the valuation were ten million receiving four times as many shares per dollar as new investors get

a discount offers conversion at a percentage lower than the new round price typically between ten and twenty-five percent

When both exist the investor receives the one that produces the most shares

Which Term Actually Binds

Both terms appear in almost every SAFE and one of them is almost always irrelevant. Knowing which one and why changes what is worth arguing about

Let's consider an illustrative SAFE of five hundred thousand dollars with a postal money limit of eight million dollars and a discount of twenty percent. The path of the cap is fixed: five hundred thousand over eight million is 6.25 percent of the company no matter what happens next. The path of the discount depends entirely on where the next round is quoted

The crossover costs exactly ten million dollars and the rule behind it is general. The discount is linked only when the prices of the next round are below the limit divided by one minus the discount. With a twenty percent discount which is 1.25 times the limit

Therefore a company that increases more than 1.25 times its own limit has a SAFE where the discount does nothing at all. Since a SAFE is usually issued precisely because the company expects to be worth much more in the next round the discount is decorative in every outcome that the parties expect. It only matters in the disappointing case when the round comes in flat or just above the limit

The practical consequence is that negotiating the discount from ten percent to twenty is negotiating the scenario that no one wants while a two million dollar move at the limit changes the answer in every scenario that goes well

Pre Money and Post Money Versions

This distinction is the most important and the least understood

under the original pre money SAFE multiple SAFEs dilute each other and also the founders.Under the later post money In the version the SAFE holder is guaranteed a fixed percentage of the company after all SAFEs are converted so the additional SAFEs dilute the founders and not the previous SAFE holders

The post-money version became the standard and it did so because it allowed the investor's result to be known at the time of signing. Under the pre-money version an investor could not indicate what he had purchased until each subsequent SAFE was known which is an awkward position for someone writing a check today

What the Post Money Shift Is Really Worth

The change is usually described as decisive and it is worth attributing numbers to it rather than accepting the adjective

Take for example a company with eight million founder shares. It issues two SAFEs: five hundred thousand dollars with a limit of eight million then one million dollars with a limit of twelve million

In the post-money version the first holder is guaranteed 6.25 percent and the second 8.33 percent so the SAFEs keep 14.58 percent and the founders keep 85.42 percent

In the pre-currency version the same limits translate into stock prices of one dollar and one dollar fifty against the founder's eight million shares yielding 500,000 and 666,667 shares out of a total of 9,166,667. The first holder ends up with 5.45 percent instead of 6.25 percent having been diluted by the second SAFE and the founders keep the87.27 percent

The difference with the founders is 1.86 percentage points over the million and a half dollars raised. That is real and not dramatic

It now manages four SAFEs for a total of three million dollars with limits of five eight twelve and fifteen million. The post-money version leaves the founders with 72.50 percent. The pre-money version leaves them at 78.43 percent. The gap has widened to 5.93 points

The honest reading is that the distinction between pre- and post-money is second order to the question that the founders really control which is how much they raise in SAFE in total and with what limits. The standard changed the burden and the burden only becomes heavy when the founders make it heavy

The Stacking Problem

Because SAFEs are quick and do not require board approval or a new class of stock companies issue them repeatedly. Each one is a small decision and the whole is not

A company that raises multiple SAFEs with different limits over eighteen months may have committed a substantial portion of its capital without ever conducting a pricing round or seeing a full table showing the result

The reckoning comes in the first round of pricing when everyone converts at once. Founders regularly discover at that point that their ownership is much smaller than they thought and by then there's nothing they can do about it

Stacking, With the Arithmetic

The shape of the problem is clearer from the numbers. An illustrative company raises four SAFEs over a year and a half as it progresses with the limit increasing each time to reflect this

Two million dollars had cost 20.83 percent of the company. Every individual decision was defensible and each limit was higher than the last which seemed like progress every time it happened

The comparison that matters is what the same two million raised would have cost once at the last cap the company hit. Two million at a post-money valuation of fifteen million is 13.33 percent. Raising in chunks cost 7.50 percentage points more than raising the same amount at the price the company ultimately proved it was worth

That gap is the true price of the instrument and is not a defect. It is payment for money arriving sooner than a round of payment could have delivered at a stage when the alternative might have been to not receive money at all. The mistake is not using SAFE. The mistake is not knowing the running total

Let's take the example one step further. A listed round then sells twenty percent of the company. The founders and employees who had already dropped to 79.17 percent are diluted to 63.33 percent the SAFE holders reach 16.67 percent and the new investor keeps twenty

The discipline is simple and often omitted: model the fully diluted table after the conversion before signing each SAFE not after the last one. The relevant number is never the limit of the document in front of you. It is the sum of each percentage committed so far

The Most Favoured Nation Variant

A third way does not carry a limit or discount. most favored nation SAFE entitles the holder to the best terms granted to any SAFE issued before the next listed round at its election

This seems generous to the founder because it completely postpones the conversation about the limit. What it really does is make each future limit retroactive

Suppose that three first friendly investors each contribute one hundred thousand dollars under those conditions. A few months later the company in need of money and without leverage accepts five hundred thousand dollars with a limit of six million dollars. That single decision reprices the three previous controls at the same limit. Each one becomes 1.67 percent so together they keep five percent the new SAFE keeps 8.33 percentand 13.33 percent of the company has been committed for eight hundred thousand dollars

The founder who signed the first three thought they were putting off a decision. They were issuing an option on their own future weakness and the option was exercised exactly at the time when they could least afford it

What the SAFE Converts Into

The conversion does not convert a SAFE into the same stock that the new investor is purchasing. It creates a parallel class usually described as a shadow series identical to the new preferred in all respects except price

That exception matters more than it seems. A liquidation preference is typically one time the price paid per share and the SAFE holder paid the maximum price rather than the round price

Go back to a SAFE of five hundred thousand dollars converted to eighty cents per share while the round price is two dollars. The SAFE holder receives 625,000 shares with a preference of eighty cents each five hundred thousand dollars in total. A new investor investing the same five hundred thousand at two dollars receives 250,000 shares with a preference of two dollars each also five hundred thousand dollars in total

Both have the same protection against losses and the SAFE holder has two and a half times as many shares

That is the clearest statement available about what a cap buys. It is a bullish instrument and nothing more. Both investors get back the same money in a liquidation that liquidates the preference pool and only one of them owns two and a half times as much of the company if it works. The anticipated risk is compensated by ownership and not by security which is more or less the opposite of what the name of the instrument implies

What Investors Give Up

SAFE holders have no shareholder rights until conversion. No board seat no voting no protective provisions and no information rights beyond what is contractually promised

They also carry the risk that a priced round will never occur and this is the structural difference from any alternative. A company that becomes modestly profitable and never raises funds again can leave SAFEs in circulation indefinitely since there is no expiration to force issuance and no mechanism that the holder can activate. Most modern forms turn into a change of control or a dissolution which handles the worst version of this but a company that simply continues funding itself from the proceeds and never selling is a scenario in whichwhich the SAFE holder expects without recourse

That outcome is rare and not remote. It's the normal fate of a company that performs but doesn't recover quickly enough to interest a growth investor

Versus Convertible Notes

Convertible notes do the same job of being structured as debt with accrued interest and a maturity date. That maturity is real leverage: If no round has occurred the note becomes callable which most startups can't do

From the founder's point of view the comparison is simple. SAFE eliminates a deadline that would otherwise come at the worst possible time since a note matures precisely when a company has failed to collect which is when it can least pay. Eliminating that deadline is worth more than the interest rate and seniority combined

The instrument a market uses tends to determine which side the leverage is at that time

The Bottom Line

A SAFE exchanges money for future shares at a capped or discounted price deferring the valuation trade without eliminating it. The discount is decorative in any round of prices above 1.25 times the cap so the cap is the term worth arguing about. The post-money standard transfers dilution to the founders in proportion to the number of SAFEs that follow the first making the stack size the actual variable rather than the document version. Four SAFEs thatRaising two million dollars can commit twenty percent of a company when the same money raised once at the final limit would have cost thirteen. Model the outcome fully diluted before you sign each one because that's the only time the terms are still negotiable