The day you get into Y Combinator, $500,000 lands in your bank account against SAFEs you barely had time to read. Six months later you sign seed SAFEs with a valuation cap and a discount, and your bookkeeper records them exactly the same way — one more credit, one more line. That is the mistake. The accelerator money and the seed money are different instruments with different conversion mechanics, and if your books treat them as interchangeable, your cap table will lie to you about who owns your company until the priced round forces the truth out.
This post is about that difference: what the two halves of the standard accelerator deal actually promise, why each books differently from the capped SAFEs you will sign later, and how to set up your ledger so every instrument converts cleanly when the priced round arrives.
The standard deal is two instruments, not one check
Y Combinator's standard deal invests $500,000 in every accepted company, but it arrives as two separate SAFEs with nothing in common except the investor. The first is $125,000 on a post-money SAFE that converts into a fixed 7 percent of the company. The second is $375,000 on an uncapped SAFE with a Most Favored Nation (MFN) provision — no cap, no discount, just the right to adopt the best terms you later give someone else.
Founders routinely book this as "$500,000 SAFE from YC" in a single row. That single row is wrong in two directions at once: it understates how much is already decided (the 7 percent is fixed on day one) and overstates how much is known (the MFN tranche has no conversion price yet, and will not have one until you sign your next SAFE). The rest of this post untangles the two.
The $125,000 tranche: the easy one, because the dilution is already decided
The fixed-percentage tranche is the simplest SAFE you will ever hold, and the reason is the post-money mechanic. On a post-money SAFE, the ownership sold equals the investment divided by the valuation cap — so a $125,000 investment for a fixed 7 percent implies a post-money cap of about $1.79 million, and that percentage does not move when you raise more SAFEs afterward. Later SAFEs dilute you and the earlier SAFE holders alike; they do not dilute the 7 percent away.
That certainty is the whole point of the post-money form, which YC introduced in 2018 precisely because the original pre-money SAFE left both sides guessing about the final ownership split. For your books, it means this tranche can be modeled as a known, fixed claim from the day the wire arrives. You know exactly how many fully diluted percentage points it represents. There is no scenario analysis to maintain and no re-measurement when the next SAFE signs. Record the cash, record the fixed claim, and move on.
One practical consequence: this tranche sets the floor of your seed-stage dilution before you have talked to a single angel. Every cap and discount you negotiate afterward stacks on top of the 7 percent, not instead of it. Founders who model their seed round as "we'll sell 15 percent" while holding the standard deal are usually selling 15 percent plus the 7 they already sold. Put the fixed claim in the cap table on day one so the stacking math stays honest.
The $375,000 MFN tranche: a floating claim on your best future terms
The second tranche works in the opposite direction. An uncapped MFN SAFE has no valuation cap and no discount of its own. Instead, it automatically adopts the most favorable terms of any SAFE you issue afterward — in practice, the lowest valuation cap (or the steepest discount) among the SAFEs signed between the MFN start date and the priced round. Sign a single seed SAFE at a $6 million cap six months from now, and the entire $375,000 converts on that $6 million cap too.
This is why the MFN tranche books differently from everything else you hold. Its conversion price is not merely unknown — it is a function of fundraising you have not done yet. Every new SAFE you sign potentially rewrites what the MFN tranche is worth in shares. Where the fixed tranche is a constant, the MFN tranche is a formula with one input still blank.
Three bookkeeping implications follow:
Track it as its own instrument, never merged with the fixed tranche. The two halves of the standard deal share an investor and a wire date and nothing else. A combined "$500,000 YC SAFE" row forces you to pick one conversion mechanic for money that converts under two. Give each tranche its own sub-account and its own cap-table row from the start.
Re-model it every time you sign a new SAFE. Each capped SAFE you issue sets a new candidate for "most favorable terms." Your cap-table model should answer, at any moment, which outstanding SAFE currently defines the MFN conversion — and what the MFN tranche's share count would be if the priced round happened tomorrow. Founders who model the MFN once, at signing, are modeling a number that stops being true with the next signature.
Watch the MFN start date, not the signature date. The MFN provision measures "subsequent SAFEs" from the later of signing the SAFE or receiving the wire. Close and fully fund your lower-cap SAFEs before issuing higher-cap ones, or a SAFE you thought was out of range can still hand its terms backward to the MFN holder. Your ledger dates should reflect funded dates, because funded dates are what the contract reads.
Then come the seed SAFEs — and the stacking problem
Six months after the accelerator wire, you start signing seed SAFEs: a $500,000 angel check at an $8 million post-money cap here, a $250,000 check with a 20 percent discount there. Each one is individually simple. Together they create the stacking problem that quietly costs founders single-digit percentages of their company.
The mechanics are straightforward but unforgiving. Every post-money SAFE you sign sells a fixed slice — investment divided by cap — and all of those slices, plus the accelerator's fixed 7 percent, plus whatever the MFN tranche claims at the lowest cap, convert simultaneously at the priced round. None of them dilutes the others; all of them dilute you. A founder who signs three "small" SAFEs thinking in pre-money terms ("we raised $1.25 million at roughly an $8 million valuation, so we sold about 13 percent") can discover at conversion that the fixed slices add up to 20 percent or more once the accelerator tranches are included.
The defense is modeling discipline, applied per instrument:
- One row per SAFE, with its own cap, discount, and form. SAFEs with different caps, discounts, or MFN provisions are different securities. Grouping them into a single "seed SAFEs" line destroys the per-instrument conversion math your priced round needs.
- Model conversion at multiple priced-round valuations. A cap only binds when the priced round exceeds it. If the round prices at or below a SAFE's cap, that SAFE converts at the round price like any other money — and your dilution is worse than the capped scenario you modeled. Run the conversion at round prices above, at, and below every cap you hold.
- Never mix pre-money and post-money SAFEs in one formula. The two forms measure the cap against different share counts, so the same dollar cap means different ownership. If an old pre-money SAFE is still outstanding, convert its economics to post-money-equivalent terms in the model and label which is which.
The question your accountant will ask: liability or equity?
Sooner or later — usually during diligence for the priced round — someone will ask how the SAFEs are classified on your balance sheet, and "we put them in equity" will need a better reason than habit. US GAAP has no standard written specifically for SAFEs, so each agreement is analyzed under the general rules for distinguishing liabilities from equity, ASC 480, and the contracts-in-own-equity rules in ASC 815-40. The classification depends on the instrument's specific terms, which is one more reason each SAFE needs its own row.
The broad pattern, confirmed by startup accounting firms that classify hundreds of these: a discount-only SAFE often lands in liability territory, because it settles in a variable number of shares worth a fixed monetary amount — the textbook ASC 480 liability shape. A plain cap-only post-money SAFE more often qualifies for equity classification. And terms that look like boilerplate can flip the analysis — a provision letting the holder take cash on a change of control, an event the accounting rules treat as outside the company's control, pushes toward liability treatment even on an otherwise equity-like SAFE.
In practice, many venture-backed startups present standard SAFEs in the equity section and disclose the terms, while their auditors run the ASC 480 and ASC 815-40 analysis instrument by instrument. The booked classification matters less than the discipline behind it: identical SAFEs get identical treatment, different SAFEs get analyzed separately, and the MFN tranche — whose settlement terms float until you sign the next SAFE — gets its own analysis rather than inheriting its sibling tranche's. If your accountant cannot tell from your ledger which SAFE is which, the classification exercise starts with reconstructing the paperwork.
A ledger setup that survives diligence
The bookkeeping itself is simple once the instruments are separated. Each SAFE gets its own sub-account on receipt of cash, the MFN adoption gets documented when a new best-terms SAFE signs, and everything reclasses into preferred equity at the priced round. In a plain-text ledger, the structure looks like this:
2026-03-01 open Assets:Cash:Operating USD
2026-03-01 open Equity:SAFE:YC-Fixed-Seven-Pct
2026-03-01 open Equity:SAFE:YC-MFN-Uncapped
2026-03-01 open Equity:SAFE:Seed-Angels
2026-03-01 open Equity:Preferred:Series-Seed
2026-03-02 * "YC" "Standard deal tranche 1 — post-money SAFE for fixed 7 percent"
Assets:Cash:Operating 125000.00 USD
Equity:SAFE:YC-Fixed-Seven-Pct -125000.00 USD
2026-03-02 * "YC" "Standard deal tranche 2 — uncapped MFN SAFE, terms float"
Assets:Cash:Operating 375000.00 USD
Equity:SAFE:YC-MFN-Uncapped -375000.00 USD
2026-09-10 * "Angel investors" "Seed SAFEs at 8M post-money cap — new MFN best terms"
Assets:Cash:Operating 750000.00 USD
Equity:SAFE:Seed-Angels -750000.00 USD
2027-04-01 * "Seed priced round" "All SAFEs convert into Series Seed preferred"
Equity:SAFE:YC-Fixed-Seven-Pct 125000.00 USD
Equity:SAFE:YC-MFN-Uncapped 375000.00 USD
Equity:SAFE:Seed-Angels 750000.00 USD
Equity:Preferred:Series-Seed -1250000.00 USDThree things to notice. First, the two accelerator tranches are separate accounts despite arriving together — that separation is what lets each convert under its own mechanic. Second, the seed SAFE narration records that it sets new MFN best terms; when the MFN adoption is a documented event rather than a remembered one, the conversion math is auditable. Third, the conversion entry moves every balance into the preferred series the priced round creates, leaving no orphaned SAFE balances behind — the single most common post-round cleanup failure.
If you keep your books in Beancount, this structure also gives you the cap-table inputs for free: each account balance plus its documented terms is exactly what a conversion model needs. The docs cover account setup and balance assertions if you want to extend the pattern with automated checks.
Mistakes that cost founders real ownership
Every failure mode in this post traces back to treating different instruments as one:
Tracking SAFEs off the cap table until conversion. A SAFE that lives only in a folder of PDFs is invisible to every dilution decision you make after signing it. Add each SAFE to the live, fully diluted cap table the day it funds — YC's own guidance says the same — including the MFN tranche at its current best-terms conversion.
Grouping convertibles into one line. SAFEs with different caps, discounts, or MFN terms modeled as a single instrument produce a single blended conversion price that matches none of them. The priced-round lawyers will unblend them anyway; do it yourself first, while the correction is still cheap.
Forgetting the MFN ratchet when pricing the next SAFE. That friendly $6 million cap you offer a small angel does not just price their check — it reprices the entire $375,000 MFN tranche to the same cap. Price every new SAFE knowing it may become every MFN holder's terms, and model the MFN impact before you sign, not after.
Booking the accelerator wire as revenue or a loan. SAFEs are neither income nor debt with a maturity date. Money that arrives as a SAFE credit goes to the SAFE account, not to revenue and not to a note payable — and the MFN tranche, whose economics float, deserves the most careful labeling of all, not the least.
Your next step
Pull up your cap table tonight and check for three rows: the accelerator's fixed-percentage tranche at its known ownership, the MFN tranche at its current best-terms conversion, and each seed SAFE with its own cap and discount. If any two of those share a row, split them before you sign anything else — because the next SAFE you sign rewrites the MFN math, and you want to watch it happen rather than discover it at the priced round.
