There is a particular flavor of regret that lives in conference rooms during Series A negotiations. A founder who raised a seed round on a stack of "simple" agreements sits down, the lawyers open the spreadsheet, and the founder discovers that the slice of the company she thought she still owned is meaningfully smaller than she imagined. Nobody lied to her. Nobody made a mistake on the documents. The instruments did exactly what they said they would do. The problem is that what they said they would do was hiding in plain sight inside a one-word name: simple.

The Simple Agreement for Future Equity—universally called the SAFE—is the most popular early-stage startup financing instrument in the United States, and arguably the most misunderstood. It was designed to strip the friction out of seed fundraising, and at that it succeeds brilliantly. But "simple to sign" is not the same as "simple to understand," and the gap between those two things is where founders lose equity they did not mean to give away and investors end up with less than they bargained for. This article is a tour of that gap. By the end, you will understand what a SAFE is, how it actually converts into stock, why the difference between a "pre-money" and a "post-money" SAFE is one of the most consequential distinctions in startup finance, how valuation caps and discounts and "most favored nation" clauses work in real numbers, how a SAFE differs from its older cousin the convertible note, why a pile of SAFEs can quietly eat a founder's ownership, how a SAFE can hand its holder a liquidation preference several times what it paid, and why—despite the breezy name—a SAFE is a security subject to federal and state securities laws.

We will define every term of art in plain English the first time it appears, work through the math with labeled hypotheticals (meet Acme Robotics, Inc., a startup we will dilute mercilessly for educational purposes), and flag the places where reasonable lawyers still disagree. If you are a founder, this article will save you from a bad surprise. If you are an investor, it will help you know what you are really buying. And if you are a judge or a lawyer encountering a SAFE for the first time in a dispute, it will give you the vocabulary and the mechanics you need to follow the fight.

What a SAFE Is (and What It Is Not)

A SAFE is a contract. That is the whole of it, and it is worth saying clearly because the name invites confusion. When an investor signs a SAFE and wires money to a startup, the investor does not receive stock. The investor does not become a shareholder. The investor receives a promise: a contractual right to receive shares of stock in the future, when and if a defined triggering event occurs—almost always the company's next priced equity financing round, but sometimes a sale of the company. Until that trigger, the investor holds a piece of paper, not a piece of the company.

The instrument was created by Y Combinator, the influential Silicon Valley startup accelerator, which published the first SAFE in late 2013 and re-released a substantially revised version in 2018. Y Combinator's stated goal was to give founders and angel investors a standardized, off-the-shelf way to do a quick seed investment without the cost, delay, and negotiation overhead of either a priced equity round or a convertible note. The form is freely available, it is short, and—critically—it is meant to be used largely as-is, with only a handful of business terms filled in. That standardization is the source of nearly all of the SAFE's advantages and a good deal of its danger. (Y Combinator coined the name, but it is worth noting that "SAFE" has become a generic label: other players in the startup-finance ecosystem publish near-identical forms under their own names, and the substance is largely the same regardless of which form a company uses.)

It helps to fix the concept of a priced round before we go further, because the SAFE is defined entirely by reference to it. A priced round is a financing in which the company and the investors agree on a valuation, set a price per share, and issue actual stock—typically convertible preferred stock—at that price, right then. A Series A led by an institutional venture-capital fund is the classic example. Pricing a round is expensive and slow: it requires negotiating a valuation, amending the company's certificate of incorporation to authorize a new class of preferred stock, drafting investor-rights and voting agreements, and a stack of related documents. (For a friendly survey of those documents, see our guide to the popular legal documents for startups.) The entire premise of a SAFE is to defer all of that work. The company takes the money now and agrees to figure out the price later, when the priced round happens and a real valuation exists. The SAFE is, in the founders' favorite phrase, "deferred equity."

What a SAFE is not is also instructive. It is not debt. It carries no interest rate, no principal that must be repaid, and—this is the headline feature—no maturity date. A loan comes due; a SAFE does not. It can sit on a company's books indefinitely, waiting for a trigger that may or may not ever arrive. It is not a share of stock, so the holder has none of the rights a stockholder enjoys: no vote, no dividend, no information rights, no seat at the table. And it is not, despite a great deal of casual talk to the contrary, a thing the law treats as outside its reach. It is a security, and we will come back to that.

A Short History: Why the SAFE Was Born

To understand the SAFE, it helps to understand the problem it was built to solve. Before 2013, the dominant bridge instrument for seed-stage startups was the convertible note—a short-term loan that converts into equity at the next priced round instead of being repaid in cash. Convertible notes worked, and they still work, but they carried baggage. As a debt instrument, a note accrued interest, came with a maturity date, and could in some structures be secured against the company's assets, which meant a separate security agreement and the possibility of conflict with the company's other debt obligations. Different investors and law firms had their own pet forms, so every note was a small negotiation.

The deeper problem the SAFE was engineered to solve was the maturity date itself. When a startup that issued convertible notes reaches the maturity date without having closed a priced round—and it almost always has spent the note proceeds by then—the founders have to reengage with their noteholders to negotiate an extension. Noteholders, who could in theory demand repayment and bankrupt the company (a remote outcome, but a real lever), may use that moment to extract better terms: a steeper discount, a lower cap, a sweetener. Even an amicable extension generates legal fees. The SAFE was created precisely to avoid that renegotiation moment. Strip out the interest, strip out the maturity date, strip out the debt characterization, and you are left with something cleaner—and, not incidentally, friendlier to the company.

There was also a jurisdiction-specific spark. At the time, there was uncertainty under California law about whether issuing convertible promissory notes could implicate lender-licensing requirements—an awkward regulatory wrinkle for a startup just trying to take a check from an angel. An instrument that delivered the core economics of a convertible note without being structured as a loan at all sidestepped that thicket.

That history matters for two reasons. First, it explains the SAFE's DNA: it is a company-friendly instrument by design. It removed the very features—interest, maturity, security—that protect investors in a convertible note. Whether that is a bug or a feature depends entirely on which side of the table you sit on. Second, it is a reminder that the SAFE was a solution engineered for a specific moment and a specific jurisdiction, and yet it spread everywhere. Its convenience proved more contagious than its original rationale, and today founders reach for a SAFE reflexively, often without asking whether it actually fits their situation. As we will see, "Is a SAFE even the right tool here?" is a question too few people ask. For the broader strategic picture of when and how to raise early money, see our companion guide on preparing your startup for capital raising and the deeper walkthrough in navigating the capital-raising maze.

The Four Levers: Valuation Cap, Discount, MFN, and Trigger

A standard SAFE has remarkably few negotiable terms. Beyond the investment amount itself, there are really only four levers, and most SAFEs pull only one or two of them. Understanding these four is most of understanding the SAFE.

The valuation cap is the single most important term in early-stage SAFE economics. A valuation cap sets a maximum company valuation that will be used to calculate how many shares the SAFE investor receives when the SAFE converts—regardless of how high the actual priced-round valuation turns out to be. It is, in effect, a price ceiling that protects the early investor's percentage ownership. The drafters' word for the risk it guards against is vivid: valuation whiplash. Picture an angel who writes a small seed check, the company uses that money to build something extraordinary, and the Series A prices at a $50 million valuation. Without a cap, that angel—whose money helped make the high valuation possible—would convert at the lofty Series A price and end up with a trivial sliver of the company, smaller than if she had simply bought equity at a sensible seed valuation in the first place. The cap prevents that perverse result. If Acme Robotics raises seed money on a $5 million cap and then prices its Series A at a $50 million pre-money valuation, the SAFE investor converts at the $5 million number, not the $50 million number—a ten-to-one improvement over what a new Series A investor pays per share. The cap is what makes early SAFE investing worth the risk. In current seed practice, caps commonly run roughly $3–5 million on the low end and $8–10 million on the high end, though this varies enormously by market, sector, and investor leverage.

The discount is the simpler of the two economic terms. A discount gives the SAFE investor a percentage reduction on the price per share paid by the new priced-round investors. A 20% discount means the SAFE converts at 80% of the Series A price. Discounts typically run between 10% and 20% in SAFEs, with 20% the single most common figure, and the broader seed market (including notes) ranging from 10% to 30%. The discount rewards the early investor for time and risk—the theory being that the company has "de-risked" since the seed money went in—but in a modest, linear way; it does not protect against a runaway valuation the way a cap does. If the Series A is priced at a hugely higher valuation than the cap implied, a 20% discount is cold comfort.

A wrinkle worth knowing, more common in convertible notes than in SAFEs but occasionally seen in either: the step-up discount, which increases the longer the instrument stays outstanding before conversion—say, 10% if the priced round closes within six months, 20% within a year, and 30% thereafter. The logic is that an investor whose money sits at risk longer deserves a bigger reward. (In the old days, similar economics were sometimes delivered by issuing the investor warrants rather than a discount; that practice has largely fallen out of favor because of original-issue-discount tax complications.)

Many SAFEs include both a cap and a discount, in which case the investor converts using whichever method gives them more shares—the better of the two outcomes. (Y Combinator eventually retired its standalone "cap and discount" form from its website, in part because the dual-mechanism math confused people and in part because it could over-reward early investors; but cap-and-discount SAFEs still circulate widely and are still requested by investors, so you will encounter them.) We will work the cap-versus-discount comparison in numbers below, because the interaction between the two is genuinely counterintuitive.

The most-favored-nation (MFN) clause is the third lever. An MFN provision lets a SAFE holder elect to upgrade its own SAFE to match more favorable terms granted to any later SAFE investor before a conversion event occurs. Suppose an angel buys an uncapped, no-discount SAFE in January, and in June the company, needing money, sells SAFEs with a generous $4 million cap. An MFN clause lets the January investor reach back and claim that $4 million cap too—or, in some formulations, exchange her instrument outright for the later, better one. (To illustrate with Practical Law's own example: an investor with an MFN and a $5 million cap, who later sees the company grant a new investor a $3 million cap, can demand that $3 million cap for herself.) MFN protects early investors against being leapfrogged by later, better-treated money. Y Combinator publishes an MFN-only SAFE (no cap, no discount, just the MFN right) for investors who want minimal terms but protection against being undercut. Notably, even where there is no contractual MFN, some founders extend better later terms to earlier investors anyway, simply as an investor-relations matter.

The trigger—sometimes embodied in a "financing threshold" or the definition of an "Equity Financing"—sets the dollar size of the future round that will cause the SAFE to convert. The point is to make sure the SAFE converts at a real, bona fide institutional round and not at some tiny issuance that does not reflect a genuine market valuation. Conversion thresholds of around $1 million are common, though watch the drafting: if the threshold is defined to include the seed money already raised, it can be satisfied by very little new money and lose its protective purpose.

Here is the discipline that the SAFE's designers intended and that experienced counsel will press on you: those four levers are essentially all you should touch. The whole value proposition of a SAFE collapses if you start bolting on bespoke terms. If someone tells you there are many other terms to negotiate inside the SAFE itself, that person has likely misunderstood what a SAFE is for. The proper home for anything extra is a separate side letter—and as we will discuss, side letters carry their own hazards.

Pre-Money vs. Post-Money: The 2018 Change That Reallocated Risk

In 2018, Y Combinator did something that looked technical and turned out to be seismic. It replaced its original SAFE with a new default form: the post-money SAFE. The difference between the original "pre-money" SAFE and the post-money SAFE is the most important conceptual point in this entire article, because it quietly moved a large amount of dilution risk from investors onto founders. Most founders who use the post-money SAFE have no idea this happened.

Let us define the terms precisely, because "pre-money" and "post-money" are slippery words even among professionals.

In a pre-money valuation SAFE (the original form), the valuation cap was, loosely, a pre-money number against which the SAFE investor's ownership was calculated, and—crucially—the SAFE holder's final ownership percentage floated depending on how many other convertible securities the company later issued. If the company sold more SAFEs after yours, or issued convertible notes, or expanded the option pool, all of those dilutive issuances diluted you, the early SAFE holder, too. Your percentage was not locked. You shared the dilution pain with everyone.

In a post-money valuation SAFE, the valuation cap is calculated to include the aggregate of all the convertible securities outstanding—all the SAFEs, in particular—and it does so in a way that fixes the SAFE holder's ownership percentage as of conversion against everything that came before the priced round. Y Combinator's own term for the relevant measure is the "post-money" valuation, but it is a slight misnomer that trips up nearly everyone. It does not mean post–Series A money. It means post–all-the-SAFE money but pre–new-equity-financing money. In other words, the "post-money" cap accounts for all the convertible-security money raised, locks the SAFE holder's slice based on that, and lets the new priced-round money come in on top.

Why does that matter so much? Because under the post-money SAFE, the early investor's ownership percentage is fixed and protected against dilution from later SAFEs, notes, and option-pool expansion that occur before the priced round. The investor knows, on day one, the floor percentage of the company they are buying (roughly, the cap divided by the post-money figure). That certainty is wonderful for the investor. But certainty for the investor is uncertainty for the founder—because every dollar of dilution that the post-money SAFE holder is now protected from has to come from somewhere, and that somewhere is the founders' (and any unprotected parties') ownership. Under the old pre-money SAFE, if a founder later sold more SAFEs, the early SAFE holders absorbed part of that dilution. Under the post-money SAFE, the early SAFE holders are shielded, and the founders eat all of it.

The practical lesson is brutal in its simplicity: with post-money SAFEs, dilution stacks, and it stacks onto the founders. Each new post-money SAFE you sell carves out a fixed, protected percentage for that investor, and those protected percentages add up. Sell enough of them, and you can give away far more of your company than you realized, because you were thinking in dollars raised rather than in protected percentages conveyed. We will demonstrate this with a worked cap table in the section on the "SAFE pileup," because numbers make the point in a way prose cannot.

The 2018 change had a second, quieter effect that founders almost never hear about: it shifted the instrument's tax posture. The original pre-money SAFE looked, to a skeptical tax authority, a lot like a variable prepaid forward contract—a derivative whose tax treatment is uncertain and often handled as an "open transaction." The post-money form, by contrast, gives the holder more equity-like features (including the right to share in dividends and to receive the same consideration as shareholders in a liquidity event) and even contains explicit language addressing treatment as common stock for tax purposes. That is not a guarantee—the proper tax classification of any SAFE remains genuinely unsettled, and the characterization can matter for the holding period that drives qualified small business stock and capital-gains treatment—but it tilts the post-money SAFE toward equity treatment from the date of grant. The honest takeaway is that both sides should run the intended SAFE past a tax adviser before signing, because the answer depends on the exact form and any negotiated terms.

The post-money change was, on balance, an investor-friendly move dressed in the language of clarity. It did genuinely make conversion math cleaner and the early investor's position more predictable—both real virtues. But founders should understand that the predictability they are giving the investor is being paid for out of their own future ownership. As you add new investors across rounds, the complete guide to adding new investors after your seed round becomes essential reading for keeping the cumulative dilution picture honest.

How a Valuation Cap Actually Converts: The Math

Let us make this concrete. Meet Acme Robotics, Inc. Acme raises $500,000 on a SAFE with a $5,000,000 valuation cap and no discount. Before any conversion, Acme has 3,000,000 shares of common stock outstanding (held by the founders). Some time later, Acme prices a Series A.

The mechanism of a capped SAFE is that the SAFE investor receives shares as though the company were valued at the lower of (a) the actual priced-round pre-money valuation or (b) the cap. The cap is the safety fence: it stops the SAFE investor's price per share from climbing above the capped level no matter how high the real valuation goes.

Consider what happens at different Series A pre-money valuations. Suppose first that the Series A comes in at a $2,000,000 pre-money valuation—below the cap. The cap is irrelevant here, because the actual valuation is lower; the SAFE simply converts at the real valuation. The SAFE investor's $500,000 buys a slice of a $2,000,000 pre-money company, which is 25%. To find the number of SAFE shares (call it X) that equals 25% of the post-conversion, pre-Series-A share count, we solve X = 25% × (X + 3,000,000), which yields X = 1,000,000 SAFE shares, for a pre-money capitalization of 4,000,000 shares and a SAFE price per share of $500,000 ÷ 1,000,000 = $0.50.

Now suppose the Series A is priced at a $10,000,000 pre-money valuation—well above the cap. Now the cap bites. The SAFE investor converts as though the company were worth only $5,000,000, not $10,000,000. The $500,000 buys 10% of a $5,000,000 company. Solving X = 10% × (X + 3,000,000) gives X ≈ 333,333 SAFE shares and a SAFE price per share of roughly $1.50—far below whatever the new Series A investors are paying. The higher the real valuation climbs, the more valuable that cap becomes to the early investor, because their effective price stays frozen at the capped level while everyone else pays the market price.

The pattern is the key takeaway: as the priced-round valuation rises, the SAFE investor's percentage entitlement shrinks toward the cap-defined floor and then stops shrinking. The cap guarantees the early investor a minimum slice. Below the cap, the SAFE behaves like a plain pre-money investment; at and above the cap, the SAFE investor's protected floor kicks in.

A crucial caveat about which share count goes in the denominator. The price per share in a real conversion is the valuation divided by a "fully diluted" capitalization—and what counts as fully diluted is itself a negotiated choice that materially moves the numbers. Practitioners argue about three things in particular: whether the converting SAFE shares themselves go into the pre-money capitalization (which dilutes only the founders) or are added on top in the post-money capitalization (which dilutes the new Series A investors too); whether the new option pool is carved out of the pre-money valuation (diluting founders and, sometimes, the SAFE holders) or adopted afterward (diluting everyone pro rata); and how to split the difference. None of this is academic. The same SAFE, the same cap, the same Series A valuation can produce visibly different ownership percentages depending on these conventions—which is exactly why you do not run a conversion on a napkin, and why every party should sign off on a single pro forma model before closing.

How a Discount Actually Converts: The Math

The discount is mechanically simpler—until the priced round is expressed as a valuation rather than a per-share price, at which point it becomes quietly recursive. Both versions are worth seeing.

Take Acme again: a $500,000 SAFE, this time with a 20% discount (so the SAFE converts at 80% of the Series A price), and the same 3,000,000 shares outstanding.

Start with the easy case, where the Series A sets an explicit price per share—say $2.00 per share. The SAFE converts at 80% of that, or $1.60 per share. Dividing the $500,000 investment by $1.60 yields 312,500 SAFE shares. That gives a pre-money capitalization of 3,312,500 shares, and the SAFE investor's percentage works out to about 9.4% (312,500 ÷ 3,312,500) immediately before the new money comes in. Clean and direct.

The harder case arises when the Series A is expressed as a pre-money valuation rather than a stated per-share price. Here you cannot simply discount the per-share number, because there is no per-share number yet—the price per share is itself defined as the pre-money valuation divided by the pre-money share count, and the pre-money share count includes the converting SAFE shares, which you cannot compute until you know the price. The definition chases its own tail. In mathematical terms it is a simultaneous-equation problem, not a one-step division, which is exactly the kind of thing the word "simple" should not be hiding.

The workable approach is to apply the discount to the valuation itself. Suppose the Series A pre-money valuation is $2,000,000. Apply the 20% discount to get a discounted valuation of $1,600,000. The SAFE investor's $500,000 then represents 31.25% of that discounted valuation ($500,000 ÷ $1,600,000). Solving X ÷ (X + 3,000,000) = 31.25% gives X ≈ 1,363,636 SAFE shares, for a pre-money capitalization of 4,363,636 shares. You can then back out the new investor's shares and confirm that, when you compute each party's price per share, the SAFE investor's effective price is exactly 20% below the new investor's price—which is the proof that the method is internally consistent. The arithmetic is more involved than the cap case, but, as practitioners like to say, nothing a good spreadsheet cannot resolve.

It is worth pausing to see the full mechanic with a real fully-diluted denominator, the way it appears in practice. Practical Law's worked example assumes 10,000,000 fully diluted shares (9,000,000 outstanding plus a 1,000,000 option pool), a $500,000 instrument with a 25% discount and a $4,000,000 cap, and a $7,000,000 Series A pre-money valuation. The Series A price is $7,000,000 ÷ 10,000,000 = $0.70. The discount price is $0.70 × 75% = $0.53. The cap price is $4,000,000 ÷ 10,000,000 = $0.40. Conversion is at the lesser of the two, so $0.40, and the investor receives $500,000 ÷ $0.40 = 1,250,000 shares. That single example illustrates both the "lower of cap or discount" rule and the central role of the fully diluted denominator in setting every one of these prices.

Cap vs. Discount Together: The Counterintuitive Inflection Point

Now the genuinely fun part. When a SAFE has both a cap and a discount, the investor converts using whichever produces more shares—equivalently, at whichever produces the lower price per share. You compute both outcomes and take the better one. The interesting question is: at what valuation does the better method switch from the discount to the cap?

Most people guess that the discount wins below the cap and the cap wins above it—that the crossover happens right at the valuation-cap number. That guess is wrong, and understanding why is a small but satisfying lesson in how these instruments behave.

Use Acme's numbers: $500,000 SAFE, $5,000,000 cap, 20% discount, 3,000,000 shares outstanding. The crossover does not occur at a $5,000,000 pre-money valuation. It occurs at a higher valuation. Why? Because when you have both terms, you are not comparing the cap to the raw priced-round valuation; you are comparing the cap to the discounted valuation. At a $6,000,000 pre-money valuation, the 20% discount produces a discounted valuation of $4,800,000—which is still below the $5,000,000 cap. So at $6,000,000, the discount is still doing the work; the cap has not yet engaged. The cap only becomes the better deal once the discounted valuation exceeds the cap, which happens at a true valuation of $6,250,000 in this example ($5,000,000 ÷ 0.80). The inflection point sits above the cap, not at it.

This is exactly the sort of subtlety that ambushes founders who negotiated different caps and discounts with different investors and then find themselves, at the Series A, computing conversions for a stack of SAFEs all clustered near their individual inflection points with no clean answer about who converts how. Consistency of mechanics across your SAFE rounds is not a nicety; it is what keeps conversion from becoming a small forensic accounting project.

SAFE vs. Convertible Note: The Family Resemblance and the Key Differences

Because the SAFE was modeled on the convertible note, the two share most of their economics. Both are bridge instruments. Both convert into equity at the next priced round. Both typically carry a valuation cap and/or a discount. Both defer the valuation question. Both usually give their holders priority over common stockholders in a liquidation. To an early investor used to convertible notes, a SAFE feels familiar. The differences, however, are precisely the protections that the SAFE strips away—and they cut in the investor's favor.

A convertible note is debt. That single fact generates three features a SAFE lacks. First, it accrues interest (recently in the 2–12% range, most often 4–8%), which is added to the principal and increases the amount that converts into equity, so the investor ends up with slightly more shares. Second, it has a maturity date—typically one to two years out—a day on which the note becomes due. If the company has not done a priced round by maturity, the noteholder can convert at a formula price, demand repayment (which the company usually cannot manage), or extend—and that last conversation hands the noteholder leverage. Third, a note can in principle be secured against the company's assets, though seed-stage notes are almost always unsecured, because perfecting a security interest against a company with no assets is rarely worth the cost. A SAFE has none of these: no interest, no maturity, no security. The investor's only realistic paths to a return are conversion at a priced round, a payout on a sale of the company, or, in liquidation, a contractual right to (at most) the return of their original investment if anything is left after creditors are paid—which, in a failed startup, usually means nothing.

Practical Law's editors frame the trade-off cleanly: founders frequently consider the absence of maturity and interest to be the SAFE's single biggest advantage over a convertible note, because a maturing note can hand investors leverage to renegotiate, and even an amicable extension costs legal fees. SAFEs are designed to replicate the economics and mechanics of convertible notes minus those two features. Investors, predictably, sometimes see it the other way: no maturity means no deadline that forces the company to do anything, and no interest means no accrual.

There is a related risk that almost no one prices in at signing. Imagine Acme turns out to be quietly profitable, grows organically, and simply never needs to raise a priced round. The SAFE has no maturity, so nothing forces a conversion—the SAFE holder sits in limbo, potentially for years, until a sale or liquidation. Worse, if Acme starts paying dividends to its common stockholders, many older SAFE forms gave the holder no way to share in them. (Y Combinator's 2018 form addressed this by adding a dividend-participation feature, one of the ways the post-money SAFE is more equity-like.) A convertible noteholder, by contrast, could have used the note's maturity-conversion feature to become a common stockholder and collect those dividends. Few technology-startup investors expect this "successful company that never raises again" scenario, which is why they tolerate the gap—but it is a genuine asymmetry that favors the note.

There is also a subtle equity-incentives angle that founders rarely consider. Structuring seed money as a SAFE (or a note) rather than as priced preferred stock tends to keep the fair-market value of the company's common stock—and therefore the strike price of employee stock options under Section 409A of the Internal Revenue Code—lower than a priced round would, which makes equity incentives for early employees more attractive. A SAFE is somewhat less favorable on this dimension than a convertible note, but considerably more favorable than issuing common or priced preferred stock at the seed stage. It is a second-order point, but the kind of thing good startup counsel weighs.

The honest bottom line is that there is no universally correct choice between a SAFE and a convertible note. The right instrument depends on the investor pool, the jurisdiction, the company's existing obligations, and the founders' risk tolerance. Some sophisticated angel groups still strongly prefer notes. Some non-dilutive grant programs forbid debt and therefore steer you toward equity-style instruments. Practical Law maintains an entire checklist devoted to choosing the right seed-financing instrument—comparing convertible notes, SAFEs, Series Seed convertible preferred stock, and plain common stock—precisely because the decision is not obvious. The competing forms multiply, too: beyond the Y Combinator SAFE there is the KISS (Keep It Simple Security) from 500 Startups and various regional variants, each with its own mechanics. The cardinal rule across all of them is consistency. Investors familiar with SAFEs expect reliability; hand them an unfamiliar form and you have negated the "simple" benefit you were chasing.

The "Phantom" Liquidation Preference: The Trap Nobody Warns You About

Here is a consequence of SAFEs and convertible notes so non-obvious that even seasoned founders miss it, and it deserves its own section. When a SAFE or note converts at a discount or a cap, the holder receives Series A preferred stock at a lower price per share than the new Series A investors pay—but the preferred stock they receive usually carries the full Series A liquidation preference, denominated at the full Series A price. The result is a liquidation preference worth more than the holder actually invested. Practitioners call it a "free," "phantom," or "windfall" liquidation preference.

A worked example shows how dramatic this gets. Suppose the Series A price is $1.00 per share and a SAFE converts at a 20% discount, so the holder pays an effective $0.80 per share. Each share still carries a $1.00 liquidation preference. That is a 1.25x liquidation preference ($1.00 of preference for every $0.80 invested)—small in dollar terms at the seed stage, but already off-market for a priced deal. Now let the gap between the cap and the Series A valuation widen. Take a $5,000,000 cap and a $20,000,000 Series A pre-money valuation. On 5,000,000 fully diluted shares, the Series A price is $4.00 per share, but the SAFE converts at the cap-implied price of $1.00 per share ($5M ÷ 5M shares). The holder thus receives shares carrying a $4.00 liquidation preference for every $1.00 invested—a 4x liquidation preference. A 4x preference would be considered wildly off-market if anyone proposed it openly in a term sheet; here it materializes silently, as an artifact of conversion mechanics, whenever there is a big spread between the cap and the priced-round valuation.

Why does this matter? Because liquidation preference is the money preferred stockholders get off the top in a sale before common stockholders (the founders and employees) see a dime. A phantom multiple quietly enlarges the SAFE holders' off-the-top claim at the founders' expense, and it surfaces only in a modest exit—precisely the scenario where every dollar of preference is contested. Sophisticated counsel draft around it in two main ways. The first is shadow preferred: the SAFE or note converts into a separate, otherwise-identical series (call it "Series A-1" or "Series A*") whose liquidation preference is set to match the dollars actually invested, 1:1, rather than the full Series A price. The second is the discount-in-common approach: give the holder enough Series A preferred shares to match their invested dollars at the full preference, then deliver the remaining shares they are owed under the cap or discount as plain common stock (which has no liquidation preference at all). Both fixes add cost and complexity, so they tend to be reserved for rounds where the cap is unrealistically low and the dollars are large enough to justify the drafting. The point for founders is simply to know the trap exists—because the default Y Combinator and Practical Law forms do not solve it for you, and a Series A lawyer who spots a fat phantom preference will not be shy about repricing the deal around it.

A close cousin is the anti-dilution effect: because a low cap converts at a low price, the converting SAFE can behave like an anti-dilution adjustment more aggressive than even a "full ratchet"—again, a term that would be deeply off-market if named openly. The lesson generalizes: SAFEs and notes can smuggle in economics that no one would agree to face-up in a priced round, simply because those economics are buried in the conversion math.

The SAFE Pileup: A Worked Cap Table

Here is the failure mode that hurts founders the most, and it deserves real numbers. Call it the SAFE pileup or stacking dilution. It happens when a company raises across multiple SAFE tranches—often because the seed stage stretches out longer than planned, or because raising one more SAFE always feels easier than pricing a round—and the cumulative effect of all those instruments is not appreciated until conversion day.

Recall the central feature of the post-money SAFE: each one locks in a fixed ownership percentage for its holder, protected against dilution from later convertibles. That fixed percentage is, roughly, the SAFE's investment amount divided by its post-money valuation cap. The trouble is that these fixed percentages add up, and because each is protected, the founders absorb all of the dilution.

Let us watch it happen to Acme Robotics. Acme's founders start owning 100%. They then raise three post-money SAFE tranches as the seed stage drags on:

  • SAFE 1: $500,000 at a $5,000,000 post-money cap → locks in 10.0% ($500k ÷ $5.0M).
  • SAFE 2 (six months later, higher cap): $1,000,000 at a $10,000,000 post-money cap → locks in 10.0% ($1.0M ÷ $10.0M).
  • SAFE 3 (a bridge to get to Series A): $500,000 at an $8,000,000 post-money cap → locks in 6.25% ($500k ÷ $8.0M).

Each percentage is fixed and protected. Stack them: 10.0% + 10.0% + 6.25% = 26.25% of the company, post-conversion and pre–Series A, is now spoken for by SAFE holders. The founders, who started at 100%, are now headed to 73.75% before the Series A investors and any option-pool expansion take their cut—and they raised "only" $2,000,000 to get there.

Now Acme prices a Series A. Suppose the new investors put in money for 20% of the post-money company and the company also creates a new 10% option pool as part of the deal (a near-universal Series A demand). Those two slices—30% combined—come largely out of the pre-existing holders. But the post-money SAFE holders' percentages were fixed as of just before the Series A new money, so within the Series A dilution event their 26.25% is recalculated against the new total. The mechanics are intricate, but the directional result is unambiguous and is the whole point: the founders' final stake gets squeezed from both ends—first by the 26.25% of fixed SAFE ownership, then by the Series A money and option pool diluting whatever is left. A founding team can walk into a Series A expecting to hold a comfortable majority and walk out holding closer to 45–50%, entirely from instruments they signed as "simple" seed paper.

Three things make the pileup worse in practice. First, steep discounts. A discount above 30% can leave SAFE holders over-represented on the post-financing cap table, which is exactly the kind of thing that surfaces in Series A diligence and irritates incoming investors—at best a conversation, at worst a deal-chiller. (There is also a real institutional concern lurking here, articulated by VCs themselves: if the founders' post-financing stake gets too thin, they may no longer be properly incentivized to grind for years, and a lead investor may insist on recutting the conversion to protect founder equity.) Second, mismatched terms. When different SAFE investors have different caps, different discounts, and different mechanics, conversion stops being a formula and becomes a project, and founders routinely end up with less equity than they modeled because they never sat down and ran the combined math. Third, misrepresenting SAFEs on the cap table. Because a SAFE's conversion depends on facts unknown at issuance, there is no single "true" number of shares to list. The professional practice is to model the SAFE using its valuation cap as the stand-in, while explicitly noting that a lower priced-round valuation would convert the SAFE into more shares (greater dilution). What you must not do is drop a vague block of "Friends & Family" onto the cap table at a guessed share count and move on. A clean, footnoted, assumptions-stated cap table is how you avoid being ambushed by your own paperwork. As you approach a term sheet, model the SAFE conversion into the negotiation from the start; the dilution surprise almost always comes from teams that treated conversion as a problem for later. For the structural side of managing layered ownership across entities, our deep dive on corporate structuring and running multiple businesses is a useful companion.

Side Letters: The Escape Hatch That Can Sink You

When an investor wants something a standard SAFE does not offer—pro-rata (preemptive) rights, information rights, a board observer seat, special negotiation rights, or assurances about terms in an eventual stockholders' agreement—the customary fix is a side letter: a separate contract signed alongside the SAFE that carries the bespoke terms. Side letters are so common that Y Combinator's own download package includes a form side letter, which by default grants the investor only pro-rata rights—the right to invest enough in the future priced round to maintain their ownership percentage.

Pro-rata rights (also called preemptive rights or rights of first offer) are a reasonable and frequently appropriate ask, and a clean pro-rata side letter is usually harmless. But here, too, the practitioner instinct runs the other way for founders. Startups generally try to grant preemptive rights to as few parties as possible, because every such right narrows the company's flexibility in structuring future rounds and adds administrative burden—and seed investors tend to be numerous, to invest small amounts, and to be less sophisticated than the institutional VCs who will later expect these rights. When seed investors push for pro-rata rights as a condition to investing, experienced counsel often coaches founders to (1) explain warmly that they will try to make room for the investor in future rounds, while declining a contractual right; and, if the investor persists, (2) limit the right to the Next Equity Financing only, so it does not become a perpetual entitlement across every future round. Board seats and board-observer rights are rarely requested or granted to seed investors at all; information rights (regular financial statements and the like) likewise usually wait for the priced round.

The deeper danger is escalation. Once one investor gets a side letter, the next one wants one too, with slightly different terms, and soon the company is administering a thicket of overlapping, sometimes conflicting promises. Each side letter is a separate obligation that has to be tracked, honored, and disclosed in diligence. A government-backed or heavily regulated investor may genuinely require side-letter terms, and that is fine. But when the terms are not essential to the deal, side letters quietly recreate exactly the negotiation overhead the SAFE was supposed to eliminate. There is a fair question lurking in any deal with a stack of side letters: if you need this many bespoke agreements bolted onto the SAFE, is the SAFE actually the right instrument at all? Sometimes the honest answer is that a convertible note or even a priced Series Seed would have fit better.

Securities Law: The SAFE Is a Security

Now the part that the name actively works against. A SAFE may be "simple," but it is unquestionably a security, and selling one is selling securities. That is not a close question. Section 5 of the Securities Act of 1933 requires that every offer and sale of a security be registered with the U.S. Securities and Exchange Commission unless an exemption applies. Startups never register, so a SAFE round lives or dies on an exemption. The Y Combinator SAFE and the leading practitioner forms are expressly drafted for a private placement to accredited investors in reliance on Rule 506 of Regulation D, which is the SEC's safe harbor under the statutory private-offering exemption of Section 4(a)(2) of the Securities Act.

Section 4(a)(2) exempts transactions "not involving any public offering." What counts as "not public" is a creature of case law: the foundational authority is SEC v. Ralston Purina Co., 346 U.S. 119 (1953), where the Supreme Court held that the exemption turns on whether the offerees can "fend for themselves"—whether, in other words, they are sophisticated enough and well-enough informed not to need the protections of registration. Section 4(a)(2) standing alone is famously fuzzy (there is no fixed cap on the number of investors, only a multi-factor inquiry into suitability, the size and manner of the offering, and the absence of general solicitation), which is why most issuers prefer the bright-line comfort of Regulation D.

Regulation D is the workhorse. Its most-used provision, Rule 506(b) (17 C.F.R. § 230.506(b)), permits a company to raise an unlimited amount from an unlimited number of accredited investors plus up to 35 non-accredited (but sophisticated) investors, provided the company engages in no general solicitation or advertising—no public pitching of the offering, no "we're raising on SAFEs" broadcast to the world. An accredited investor is defined in Rule 501 (17 C.F.R. § 230.501): for natural persons, the classic tests are individual income over $200,000 (or $300,000 jointly with a spouse) in each of the last two years, or net worth over $1 million excluding a primary residence; the SEC has since added categories for individuals holding certain professional certifications and for "knowledgeable employees," among others. Rule 506(c) relaxes the advertising ban and does permit general solicitation—a change Congress directed in the JOBS Act of 2012—but in exchange the company must take reasonable steps to verify that every purchaser is in fact accredited, a meaningfully heavier burden than the self-certification typical under 506(b). A company relying on Regulation D generally must file a Form D notice with the SEC no later than 15 days after the first sale, and must attend to state "blue sky" laws: registration is largely preempted for covered Rule 506 offerings, but states may still require notice filings and fees where investors reside. Bear in mind, too, that securities sold under Section 4(a)(2) or Regulation D are restricted securities—the investor cannot freely resell them without registration or its own exemption.

Two practical consequences follow. First, who you sell to matters. Taking SAFE money from non-accredited investors beyond the limits, or soliciting the general public without satisfying 506(c)'s verification requirements, can blow the exemption—and a blown exemption can give investors a rescission right (the ability to demand their money back) and expose the company to regulatory enforcement. The casual "just download the form and run with it" instinct is genuinely dangerous here. Second, the anti-fraud rules always apply. Even in an exempt private placement, Section 17(a) of the Securities Act, Section 10(b) of the Securities Exchange Act, and SEC Rule 10b-5—the broad prohibition on material misstatements and omissions in the sale of securities—are fully in force. A founder who oversells the company's prospects or hides material risks while raising on SAFEs is not insulated by the offering's exemption from registration.

This is also why doing a SAFE round without competent securities counsel is a false economy. The instrument is short, but the regulatory frame around it is not, and the rules vary by state and (emphatically) by country. The Y Combinator SAFE is a U.S. instrument built around U.S. securities law; a company organized abroad, or selling to investors abroad, cannot assume the default form keeps it compliant with local securities regulation. International SAFE variants exist, but "standard form" never means "compliant everywhere."

Risks for Each Side, Honestly Stated

It is worth closing the analysis by naming, plainly, what each party is really exposed to—because the SAFE's frictionlessness can lull both sides into ignoring real risk.

For the investor, a SAFE is, despite its name, close to the antithesis of a safe investment. There is no interest, no maturity, no security, and no shareholder rights. The investor's entire return depends on a future event—a priced round or a sale—that may never come. A company can hold a SAFE on its books indefinitely with no obligation to act, and can grow profitably forever without ever triggering conversion, leaving the holder in limbo and (under older forms) shut out of dividends. If the company simply stalls out and quietly winds down, the SAFE holder stands behind every creditor, secured and unsecured, and is entitled at most to the return of the original investment if any assets remain—which, in a failed startup, is usually a polite way of saying "nothing." The early SAFE investor is making a high-risk bet that the company will reach a financing event, and is accepting fewer protections than a convertible-note holder would have in exchange for the cap or discount upside. This is appropriate for sophisticated angels and for friends-and-family who genuinely understand the risk; it is a poor fit for anyone who cannot afford to lose the entire investment.

For the founder, the dominant risk is dilution misunderstood. The post-money SAFE's fixed, protected percentages stack, and they stack onto the founders. Steep discounts, mismatched terms across investors, a casual cap table, and the phantom liquidation preference all compound the problem. The excitement of closing a Series A evaporates quickly when a founder who was "counting the money" realizes how much of the company the converting SAFEs are about to claim—and how much comes off the top in a modest exit. The defense is unglamorous: use a consistent, current form; resist the urge to negotiate bespoke terms into the document; keep side letters minimal; model conversion early and often, and get every party to sign off on a single pro forma; and—this cannot be said too many times—get good accounting, tax, and legal advice. The word "simple" describes the signing experience, not the consequences.

There is also a strategic risk that precedes all of these: assuming a SAFE is the right tool at all. A SAFE should fit a deliberate financing strategy, not substitute for one. Some investor pools prefer notes. Some non-dilutive funding programs impose equity-structure requirements or forbid anything "debt-like." Friends and family might be perfectly content with directly issued non-voting shares or a voting trust. A Series Seed priced preferred round—more paperwork up front, but cleaner economics and no conversion surprises—is sometimes the better answer outright. The reflexive reach for a SAFE, without asking whether it serves the company's actual situation, is itself a mistake. Thinking through the instrument as part of a broader plan—the kind laid out in our guides on preparing your startup for capital raising and the capital-raising maze—is what separates a clean cap table from a cautionary tale.

Key Takeaways

A SAFE is a contract that gives an investor the right to receive stock in the future, when a priced round (or sale) triggers conversion. It is not stock, not debt, and not free of the law. Its genius is the removal of friction—standardized form, few negotiable terms, fast and cheap to close. Its peril is that the same simplicity hides mechanics that materially affect who ends up owning what.

Master the four levers—valuation cap, discount, MFN, and trigger—and touch nothing else inside the document; put anything bespoke in a tightly scoped side letter. Understand that the 2018 shift to post-money SAFEs fixed and protected early investors' percentages, which is wonderful for investors and costly for founders, because the protected percentages stack and the dilution lands on the founding team. Watch for the phantom liquidation preference a low cap or steep discount can create, and know that shadow-preferred or discount-in-common drafting can defuse it. Run the conversion math before you sign, not after; insist on a single pro forma everyone has reviewed; keep your forms consistent and current; and treat your cap table as a living document with clearly stated assumptions about how the SAFEs convert. Remember that a SAFE is a security sold under Section 4(a)(2) and Regulation D, that who you sell to and how you solicit them matters, that you must file Form D and respect blue-sky rules, and that the anti-fraud rules apply regardless of any exemption. And never, ever let the word "simple" talk you out of getting qualified legal, tax, and accounting advice. The SAFE rewards the prepared and quietly punishes the careless.

Frequently Asked Questions

Is a SAFE debt or equity? Neither, exactly. A SAFE is a contractual right to receive equity in the future. It is not a loan—there is no interest, no maturity date, and no obligation to repay principal—so it is not debt. But it is also not equity yet; the holder owns no stock and has no shareholder rights until the SAFE converts at a triggering event. For tax purposes the classification is genuinely unsettled: a pre-money SAFE can look like a variable prepaid forward contract (a derivative), while the post-money form leans toward equity treatment. Because the answer affects holding periods and capital-gains treatment, both sides should consult a tax adviser before signing.

What happens to a SAFE if the company never raises a priced round? This is the SAFE investor's core risk. A SAFE has no maturity date, so nothing forces the company to act. If there is never a qualifying equity financing or a sale, the SAFE can sit on the books indefinitely—even if the company is thriving. The holder's last resort is a liquidation or wind-down, in which a SAFE typically entitles the investor to (at most) the return of the original investment, but only after creditors are paid and only if assets remain. In a failed startup, that usually means little or nothing.

What is the difference between a pre-money and a post-money SAFE, in one sentence? A pre-money SAFE lets the early investor's ownership percentage float—diluted along with everyone else by later SAFEs, notes, and option-pool expansion—whereas a post-money SAFE (Y Combinator's default since 2018) fixes and protects the early investor's percentage against that later dilution, which shifts the resulting dilution onto the founders.

What is a "phantom" liquidation preference? When a SAFE converts at a discount or cap, the holder pays a lower effective price per share than the new investors but usually receives preferred stock carrying the full round's liquidation preference. The result is a preference worth more than the holder actually invested—1.25x at a 20% discount, and potentially 4x or more when there is a wide gap between a low cap and a high priced-round valuation. It is "off-market" economics that materialize silently from the conversion math. Counsel can defuse it with "shadow preferred" or "discount-in-common" drafting, but the standard forms do not.

Can I just download the Y Combinator SAFE and use it without a lawyer? You can, but it is a false economy and sometimes a serious mistake. The form is standardized, but selling a SAFE is selling a security, which means complying with Section 4(a)(2)/Regulation D (or another exemption), filing Form D within 15 days of the first sale, observing state blue-sky rules, and respecting the anti-fraud rules. The default form is built for U.S. law and may not fit your jurisdiction, your investor pool, or your existing obligations. Good securities and tax advice up front is far cheaper than untangling a botched round later.

How much of a discount or what valuation cap is "normal"? Discounts in SAFEs typically run 10% to 20%, with 20% the most common figure (the broader seed market, including notes, runs 10%–30%); a discount above 30% can over-reward early investors and create friction in Series A diligence. Valuation caps commonly run roughly $3–5 million on the low end and $8–10 million on the high end at the seed stage, but they vary enormously by company, sector, geography, and investor leverage—there is no single "normal" number, which is why the cap is the most heavily negotiated economic term.

What is a pro-rata side letter, and should I sign one? A pro-rata (preemptive) side letter gives the SAFE investor the right to invest additional money in the future priced round to maintain their ownership percentage and avoid being diluted out. Y Combinator's package includes a form of it. Pro-rata rights are a common ask, and a clean one is often reasonable—but founders usually want to grant them to as few investors as possible, and where pressed, to limit the right to the next equity financing rather than every future round. The bigger caution is against side-letter proliferation: once you grant bespoke terms to one investor, others want their own, and a thicket of overlapping side letters quietly rebuilds the overhead the SAFE was meant to remove.

Is a SAFE the same as a convertible note? No, though they are close cousins with similar economics (both convert at a priced round, both usually carry a cap and/or discount, both sit ahead of common in a liquidation). The key differences are that a convertible note is debt—it accrues interest, has a maturity date, and can in principle be secured—while a SAFE has none of those features. The note's maturity gives the investor leverage and a deadline; the SAFE's lack of maturity is the founders' favorite feature. A note's maturity-conversion right also lets a holder become a common stockholder (and collect dividends) if the company succeeds without ever raising again—a path a SAFE holder lacks. Which is better depends on the deal.

Related Articles

This article provides general information and is not legal advice. SAFEs are securities, and their use, conversion mechanics, tax treatment, and securities-law compliance vary by jurisdiction and by the specific facts of your financing. Consult qualified securities counsel and a startup-experienced accountant before issuing or investing in a SAFE.