SAFE Valuation Caps, Discount Rates and Most Favored Nation Terms

What This Means for Founder Dilution

A founder negotiating a Simple Agreement for Future Equity for a 2026 seed round gives up more ownership through the valuation cap than through the discount rate or the most favored nation provision, because the cap fixes the ceiling price at which the entire investment converts to shares. Nationwide 2026 seed data put the median post money valuation near $24,000,000 and typical dilution from a full seed financing near 20 percent, so a cap set below that market band transfers ownership the priced round investors never paid for. An open ended most favored nation provision compounds that exposure by letting every later, more generous SAFE reset the price of the earliest one, and the founder pays that difference alone.

Executive Summary

A 2026 Simple Agreement for Future Equity’s valuation cap controls more founder dilution than its discount rate or most favored nation clause, since the cap alone fixes the conversion price. A lower cap always converts a fixed investment into a larger ownership stake for the investor, regardless of how the priced round eventually values the company. The discount rate matters only when the round prices above the cap or when no cap applies at all; 2026 market data show discounts clustering between 10 percent and 25 percent, with a median near 20 percent when a discount appears at all. The most favored nation clause protects the earliest, least informed investor by letting that investor swap into whatever cap or discount a later investor negotiates, and the cost of that swap falls on the founder alone, since a post money SAFE dilutes only the founders once it converts.

Y Combinator’s post money SAFE, rewritten in 2018, remains the market default for United States pre seed and seed financings, offered in three standard forms: a cap only SAFE, a discount only SAFE, and an uncapped SAFE carrying a most favored nation clause. Ninety three percent of pre seed financings in the first quarter of 2026 used a SAFE rather than a convertible note, the highest share on record. Stacking a cap, a discount, and a most favored nation clause onto one instrument is not the market default; it is closer to a request a founder can often narrow. What remains contested is how hard a founder can push back once a lead investor asks for a most favored nation clause, and how much of the 2026 gap between artificial intelligence sector caps and every other sector’s caps a founder outside that sector can use as a reference point.

Detailed Findings

How does the valuation cap set the ceiling on a 2026 SAFE’s conversion price

The valuation cap operates as a ceiling on the price per share the SAFE investor pays once the instrument converts, not as a statement of what the company is worth on the day the SAFE is signed. Under the post money format Y Combinator adopted in 2018, the calculation runs in one step: divide the investment amount by the cap to find the ownership percentage the investor locks in at conversion. A $500,000 investment on a $5,000,000 post money cap converts to 10 percent ownership; the same $500,000 on a $3,000,000 cap converts to nearly 16.7 percent. That six and two thirds percentage point swing between two caps only $2,000,000 apart shows why the cap, not the discount and not the most favored nation clause, does most of the work in setting a founder’s eventual dilution.

Market data for 2026 place seed stage post money caps for a business to business software company outside the artificial intelligence sector between $10,000,000 and $15,000,000, with pre seed caps for the same category between $5,000,000 and $10,000,000. Financial technology, biotechnology, and life sciences companies command caps several million dollars higher at each stage, and artificial intelligence infrastructure companies command caps that run from $12,000,000 to more than $25,000,000 at pre seed and from $25,000,000 to more than $50,000,000 at seed. A founder building outside artificial intelligence infrastructure who anchors a cap negotiation to that sector’s range invites a lead investor to walk away from the term sheet entirely; the safer negotiating reference is the cap range for the founder’s own sector and stage, not the highest number currently circulating in the market.

Geography moves the cap further. Companies raising in the San Francisco Bay Area and New York command caps several million dollars above companies raising in most other United States hubs at the same stage, and companies raising in continental Europe or South and Southeast Asia see caps several million dollars below the United States hub average. A founder relocating a legal entity or a lead investor relationship to capture a higher regional cap should weigh that gain against the added cost of operating two jurisdictions’ worth of counsel, banking, and tax compliance before treating the cap difference as free money.

How does the discount rate change what a founder gives up when no cap governs

The discount rate reduces the price per share the SAFE investor pays at conversion by a fixed percentage off whatever price the priced round sets, and it changes the outcome only when it produces a lower conversion price than the cap does. Consider a $300,000 SAFE carrying a $6,000,000 cap and a 20 percent discount. If the Series Seed prices the company at a $10,000,000 pre money valuation, the cap implies a conversion price based on $6,000,000 while the discount implies a conversion price based on $8,000,000. Because $6,000,000 is lower, the cap governs and the discount does nothing; the investor already receives a better price through the cap than the discount could produce. The discount takes over only when the eventual priced round values the company above the cap, or when the SAFE carries no cap at all.

2026 market data show cap only SAFEs, the format without any discount, holding roughly half to three fifths of the market, while combined cap and discount instruments hold roughly one fifth to one quarter. Where a discount does appear, it typically runs between 10 percent and 25 percent, with a median near 20 percent. A founder who trades a lower cap for a smaller discount percentage is usually trading a certain, larger dilution cost for an uncertain, smaller one, since the discount only matters if the company outperforms the cap by the time it prices its next round. A founder confident in that outperformance should weigh a lower discount against a higher cap instead of treating the two terms as independent negotiating chips.

How does a most favored nation provision transfer dilution from later investors back onto the founder

A most favored nation provision gives an early SAFE investor the right to swap into a lower cap or a larger discount if the company later issues a SAFE on better terms to somebody else. The clause exists because early investors take the most uncertainty and the least information, and it protects them against being undercut by a later, more sophisticated investor who negotiates a materially better price. The standard mechanics require the company to give the earlier investor written notice once a better term SAFE issues, then allow that investor an election period, typically 10 to 30 days, to choose the improved terms, after which the capitalization table adjusts to reflect the investor’s new conversion price.

The dilution cost of that adjustment falls on the founder alone. Because a post money SAFE fixes the investor’s ownership percentage at the moment it converts and every later SAFE dilutes only the founders’ remaining share, a most favored nation clause that resets an early investor’s price downward does not touch any other investor’s stake; it enlarges the early investor’s slice of the company at the founders’ direct expense. Roughly a third to two fifths of pre seed SAFE rounds now carry a most favored nation provision on at least one instrument, and the exposure compounds when a founder stacks several of these clauses across a raise that closes in pieces over many months rather than in a single signing.

Y Combinator’s own uncapped SAFE template, one of its three standard 2026 forms, carries a built in most favored nation clause precisely because the instrument has no cap of its own; the clause is the only price protection an uncapped investor gets. A founder who accepts an uncapped, most favored nation SAFE from an early check writer and then negotiates a much lower cap with a later lead investor should expect the early check writer to convert at the lower number, even though the later lead investor bore far more diligence cost to reach that price.

How do stacked SAFEs compound dilution before the priced round closes

A seed round rarely closes on a single instrument. Most 2026 companies sign several SAFEs over a period of months, often starting with a smaller, uncapped, most favored nation instrument from an early angel and ending with a larger, capped instrument from the lead investor who prices the round. Each SAFE dilutes the founders independently at conversion, and because post money SAFEs lock in each investor’s percentage at signing, the dilution from an early SAFE cannot be reduced by anything that happens in a later SAFE. Nationwide 2026 data treat 15 to 25 percent cumulative dilution across a full seed stack as the normal range and treat anything above 25 to 30 percent as a warning sign for the priced round that follows.

The practical failure mode is a founder who tracks each SAFE’s terms individually and never models the stack together. A $200,000 uncapped most favored nation SAFE signed in month one, a $400,000 SAFE at an $8,000,000 cap signed in month four, and a $600,000 SAFE at a $12,000,000 cap signed in month seven do not simply add to a blended average cap; the most favored nation clause in the first SAFE can pull its conversion price down to the lowest cap in the stack, compounding the dilution from the second and third instruments onto the first. In thirty years of reviewing capitalization tables for technology companies, the single most common drafting gap I have found is a company that priced its final SAFE correctly and then discovered, only at the Series Seed closing, that its earliest and smallest check had quietly repriced itself down to match.

What risk factors and tax mechanics change the stakes of these three terms

The United States Securities and Exchange Commission warned as early as May 9, 2017, in an investor bulletin published on Investor.gov, that a SAFE despite its name may be neither simple nor safe, because the instrument represents only a conditional promise of future equity that may never convert if the company never raises a priced round or an acquisition. That warning was written for investors, but it cuts the other way for founders too: a company that leans on SAFEs indefinitely without ever pricing a round leaves every SAFE holder, and every dollar of cap, discount, and most favored nation exposure, recorded on the capitalization table without a final price for years.

Tax treatment adds a second layer founders should track when negotiating these terms with sophisticated investors. Internal Revenue Code Section 1202 excludes gain on qualified small business stock, and the One Big Beautiful Bill Act, effective for stock acquired after July 4, 2025, replaced the old five year all or nothing holding period with a tiered schedule: a 50 percent exclusion at three years, a 75 percent exclusion at four years, and a full 100 percent exclusion at five years, while raising the per issuer exclusion cap to $15,000,000 and the qualifying company’s gross asset ceiling to $75,000,000. A SAFE itself is not stock for Section 1202 purposes; the holding period begins only when the instrument actually converts to shares at the priced round. An investor negotiating a lower cap in exchange for accepting a later conversion date is, in effect, negotiating a later start to that holding period clock as well, and a founder who understands that tradeoff can sometimes trade a modestly higher cap for a faster path to a priced round that starts the investor’s holding period sooner.

Want more industry insights? I publish daily analysis on the intersection of deep tech and venture law.

Subscribe to The Innovation Attorney on Substack

Questions Founders Should Still Ask

Whether cap only SAFEs without any discount keep taking share from combined cap and discount instruments through the rest of 2026 matters to a founder deciding how hard to push back on an angel who wants both terms stacked onto a single check; the market share trend is the founder’s best argument for declining the discount entirely.

How a founder should draft the notice and election window inside a most favored nation clause, beyond the 10 to 30 day range now common in the market, matters because a vague or open ended election period lets an early investor wait and see how the whole round prices before deciding whether to invoke the clause, which strips the founder of certainty about the final capitalization table until the last signature.

Whether the Treasury Department will issue guidance clarifying exactly when a SAFE’s Section 1202 holding period begins, at signing, at a later cap amendment, or only at the instrument’s actual conversion, matters to any founder marketing a SAFE to an investor who is counting on the qualified small business stock exclusion as part of the return.

How far a founder outside the artificial intelligence sector can push a 2026 valuation cap negotiation using that sector’s far higher caps as a reference point, without losing credibility with a lead investor, is a question every such founder is testing round by round.

Whether the Securities and Exchange Commission will update its 2017 investor bulletin now that SAFEs make up 93 percent of pre seed financings rather than a narrow crowdfunding instrument matters to any founder whose investor base includes first time angels who read that bulletin, or should have, before wiring a check.

Interested in analysis about the intersection of venture capital, tech, artificial intelligence, public policy and the law? Check out my Substack channel. https://theinnovationattorney.substack.com/