← All poststax · cap-table

QSBS: The $10M Tax Decision Most Founders Discover Two Years Too Late

qualified small business stock can shield ten million per founder from federal capital gains. you have to qualify on the day shares were issued, not on exit day.

2026-07-225 min readZift

the call goes the same way every time. founder is in diligence on a $40m secondary alongside their series c primary. someone asks when the c-corp was formed. founder says "we converted from the llc in march 2023 when we raised the seed." there's a pause. the lawyer asks the founder to repeat the date. founder repeats it. the lawyer says "we should probably talk offline."

what just happened: the founder is on track to liquidate $20m of common stock, and the qsbs five-year holding clock started in march 2023, not in 2021 when they incorporated as an llc. exit is june 2027. the math doesn't work. somewhere between $3m and $4m of federal capital gains tax just appeared on the diligence call.

qsbs — qualified small business stock under irc section 1202 — is the single biggest tax instrument available to founders, and it's also the one most likely to be discovered three years after the decision that disqualified you. you have to qualify on the day the shares were originally issued. not on exit day.

what qsbs actually does

section 1202 lets a founder exclude up to $10m of capital gains per issuer, or 10x the basis in the stock, whichever is greater, from federal capital gains tax. that's a $0 federal tax bill on $10m of common stock proceeds if you hold for five years and the issuing company meets the eligibility tests. on a $20m exit, that's a savings of roughly $4.76m at the current 23.8% federal rate back in the founder's pocket.

anomaly · this week2 things moved
incorporation · day 1
form as a delaware c-corp on day one or lose the qsbs clock.

every month you operate as an llc is a month the five-year holding period hasn't started — converting at series a means a 2030 exit barely clears the window.

the eligibility tests are checked as of the date the shares were issued. the issuer must be a domestic c-corp, gross assets must be $50m or less at issuance, and the company must be in a qualified trade or business (which excludes most professional services, hotels, restaurants, and finance). the founder must have held the stock for five years from the date of original issuance. every one of those clauses is a trap door for a founder who incorporated under an llc and "fixed it later."

the llc-to-c-corp conversion trap

founder starts as an llc because their first lawyer told them llcs are simpler. that's true. then the seed lead wires $2m and says "convert to a delaware c-corp before close." the conversion happens. founder owns the same percentage of the cap table. on the surface, nothing changed — except the qsbs clock.

stock issued in the conversion is treated as newly-issued c-corp stock as of the conversion date. the years the founder operated as an llc don't count. so the founder who incorporated in january 2021 as an llc and converted in march 2023 starts their qsbs five-year holding period in march 2023, not january 2021. an exit in 2027 — four years and three months from the conversion — disqualifies the entire $10m exclusion.

the founder didn't know. their seed-stage lawyer probably mentioned qsbs once in a doc that was never read. by the time it comes up in diligence, it's two years too late to fix.

what the math looks like on a real exit

founder owns 5m shares of common, issued at $0.0001 in march 2023 at conversion. company sells in december 2027 for $400m. founder's share, post-dilution, is $20m. holding period at sale: 4 years 9 months.

without qsbs: federal cap gains at 23.8% on $20m of gain = $4.76m federal tax. plus california state tax (california doesn't conform — another $2.66m at 13.3%, paid no matter what). total: roughly $7.4m on a $20m sale.

with qsbs satisfied: federal tax on the first $10m of gain = $0. on the remaining $10m at 23.8% = $2.38m. california still wants $2.66m. total: roughly $5m on the same sale. delta: $2.38m kept versus the same exit without the federal shield, on the back of a single entity-wrapper decision their first lawyer made in 2021.

the section 1045 rollover most founders don't know about

there's a partial rescue. irc section 1045 lets a founder who sells qsbs before the five-year holding period roll the gain into new qsbs within 60 days, preserving the holding period from the original stock. sell a slice of common, roll the proceeds into stock of another qualifying small business within 60 days, and the clock keeps ticking from the original 2023 date.

most lawyers don't bring it up unprompted. founders who roll into a single qualifying startup investment preserve the exclusion. founders who take the secondary and put it in a brokerage account lose it.

what to actually do, in priority order

incorporate as a delaware c-corp on day one. not after the term sheet. not after the friends-and-family round. day one. the few thousand dollars of franchise tax and accounting overhead in years 0-1 is cheap insurance against the $2-5m number above.

document the issuance date of every founder grant. the qsbs clock attaches to the certificate, not the cap table line. keep the stock purchase agreement, the board resolution, and the 83(b) election in one place. carta and pulley track the date but don't verify the supporting docs.

check the $50m gross-assets test before any large primary round. if total assets on the closing balance sheet cross $50m on the day of the next issuance, the new shares issued that day fail. the old shares still qualify.

talk to a tax attorney before any secondary, not after. founders sell secondary into a series b at the round price and discover six months later that they've blown the holding period for the rest of their position. the 60-day section 1045 window expires while you're celebrating the wire.

how zift handles this

zift tracks your qsbs eligibility clock from incorporation date and flags exits or secondaries that approach the five-year threshold in the monday briefing. when gross assets cross $40m and the $50m issuance limit is in sight, the briefing names the next planned round and the date the test gets checked.

if you're a finance lead at a series a or b team managing qsbs eligibility across multiple founders with different grant dates and 1045 rollover plans, zift handles that too.

every founder thinks qsbs is something the lawyer is handling. on exit day, the lawyer's job is to read out the result. the decision was made on day one.

Related

More on this topic.

2026-08-30 · 5 min read

Marketplace Take Rate Is the Only Number That Matters. GMV Is Vanity.

founders pitching marketplaces lead with gmv. investors discount it instantly. take rate × retention is the only revenue signal that survives diligence.

Read →
2026-08-26 · 5 min read

The Investor Reference Call That Passes on You

investors do back-channel references on every term sheet. founders rarely hear the bad ones. the language is consistent and worth recognizing.

Read →
2026-08-23 · 5 min read

Eval Cost Is the AI Line Item Nobody Budgets For

every ai startup runs evals. nobody puts them in gross margin. at scale, eval cost is 12-20% of inference and it sits in the wrong line.

Read →

Finance reports to you, not the other way.

Cash, burn, MRR, runway. Three things to look at this week. Three minutes to read. Reply with a question and Zift answers from your real numbers.

Join the waitlist →Keep reading →