Equity is one of the most powerful tools a startup has to attract and reward people. It is also one of the most confusing. Founders and early employees get hit with a wall of jargon - RSA, RSU, ESOP, 409A, FMV, 83(b), QSBS, double-trigger - and most of it is never explained clearly. The result is that critical decisions get made on autopilot, and people sometimes lose huge amounts of money because they ticked the wrong box at the wrong stage.
This article is an attempt to untangle all of it. It is written for both founders designing an equity plan and employees evaluating an offer. We will cover the most popular equity instruments, when each one makes sense, and how the math changes as a company grows. We will focus on US companies from incorporation through Series B - past that the rules change again.
📚 Terms you need to know
409A (also called FMV) 📊 - an independent valuation of the company's shares, required by the IRS. Rule of thumb: FMV is about 20-35% of the company's most recent round valuation. A 409A is considered expired when any of these happen:
you raise a new round
12 months have passed since the last one
something material happens (big revenue jump, major customer, term sheet in hand, M&A talks)
Once it expires you need a fresh one before granting any more options. Typical cost: $2K-$5K for an early-stage company (more for later-stage), turnaround 2-4 weeks. Providers like Carta, Pulley, or AngelList handle it routinely.
Vesting - equity is not given all at once. The standard is 4 years with a 1-year cliff: you get nothing if you leave in year one, then 25% lands at the 1-year mark, and the rest accrues monthly.
83(b) election - an IRS form. Normally you pay tax on equity as it vests, on whatever it is worth at the time. Filing 83(b) within 30 days of grant lets you pay tax now on the full grant instead of later. At a brand-new company where the equity is worth almost nothing, this is essentially free. At a later stage, it can be a big check. Only available for actual stock (RSA or early-exercised options), not RSUs.
Capital gains and QSBS - if you own actual stock and hold it for more than 1 year, gains are taxed as long-term capital gains - roughly 20% federal at the top, vs. ordinary income that can reach ~40%+ once you add federal, state, and surtaxes. (Plus a 3.8% NIIT surtax on both for high earners.) If you hold it longer in a qualifying small business (under $75M in assets at grant, or $50M for grants before July 4, 2025), it qualifies for QSBS: federal-tax-free gain up to $15M or 10x your basis (or $10M for pre-July-2025 grants), whichever is greater. For stock acquired after July 4, 2025 under the OBBBA changes, QSBS also kicks in on a sliding scale: 50% exclusion at 3 years, 75% at 4 years, 100% at 5+ years. Pre-July-2025 stock still needs the full 5 years for any exclusion. This is the single biggest tax advantage in startup equity. Crucially, you only get it if you own real stock - not RSUs. QSBS isn't automatic, though: to qualify, the issuer must be a US domestic C-corporation (not an LLC, S-corp, or partnership), you must have acquired the stock at original issuance straight from the company rather than on a secondary market, the company's gross assets must have been under $75M ($50M for pre-July-2025 grants) when it was issued, at least 80% of assets must be used in an active business (most tech startups qualify; finance, consulting, law, and health do not), and you must hold the shares long enough (the 3/4/5-year tiers above).
Ordinary income - what you pay on a regular salary. Federal goes up to 37%, plus payroll taxes and state. This is what hits you on RSU settlements.
Double-trigger vs single-trigger 🔄 - an RSU-only distinction. Double-trigger = shares only deliver when both time-vesting and an exit (IPO, acquisition, tender) happen - no personal tax until both fire. Single-trigger = shares deliver on time alone, which creates an immediate personal tax bill at vest. Private-company RSUs are almost always double-trigger because of this (SpaceX is the rare exception). RSA and ESOP are effectively single-trigger - they vest on time alone - but they don't create the same tax problem: RSA is taxed once upfront via 83(b), and ESOP isn't taxed until you choose to exercise.
Vesting acceleration ⚡ - clauses that speed up vesting under certain conditions, most often a sale of the company. Single-trigger acceleration = vesting jumps forward on a change of control alone (acquisition, IPO). Double-trigger acceleration = accelerates only if there's a change of control and you're involuntarily terminated without cause (or resign for good reason) within ~12 months after. Applies to RSA, ESOP, and RSU equally - acceleration is about vesting speed, not about how shares deliver. Founders usually negotiate at least double-trigger acceleration; rank-and-file employees usually get nothing. ⚠️ Naming hazard: "single/double-trigger acceleration" is a different concept from "single/double-trigger RSU" above - same word, different mechanic.
🌍
Heads up: this is all US-centric. 83(b), 409A, QSBS, stock option tax treatment - all of it applies only to people taxed as US persons: US citizens (anywhere in the world), green card holders, and anyone who meets the IRS substantial presence test (broadly: physically in the US for ~183 days in a rolling window).
For employees working from overseas, the tax picture can be completely different - their home country may tax equity at grant, at vest, or at exercise depending on local rules, and US elections like 83(b) often don't help them. Sometimes RSUs are actually the cleaner choice abroad simply because settlement is a single, well-defined taxable event the company can withhold against.
If you're granting equity to someone outside the US, get specialized tax advice for their country before you do.
🎁 The three most popular equity instruments
RSA (Restricted Stock Award) 📜 - actual shares given to you on day one, subject to vesting. If you leave before vesting completes, the company buys back the unvested portion. You own real stock from the start.
ESOP (Stock Options) 🎯 - the right to buy shares later at a fixed strike price set today. The strike is almost always set to the current FMV (per the 409A) - companies can't set it lower without serious tax penalties, and there's no reason to set it higher. You don't own anything until you pay to exercise your options and convert them into actual shares. US companies typically grant ISOs (Incentive Stock Options) to employees for the better tax treatment; NSOs (Non-Qualified Stock Options) are used for contractors, advisors, and grants beyond the ISO $100K/year limit. Throughout this article, when we say ESOP we mean ISOs - NSO mechanics are similar but differ on tax timing; see the ISO vs NSO appendix at the end.
RSU (Restricted Stock Unit) 📝 - not simply a promise to give you shares, but a promise to give them once two conditions are both met: you time-vest and the company has an exit. That pairing is the double-trigger. So at a private company the shares are actually delivered only when an exit happens - an acquisition, IPO, or tender offer - and only for the portion you've already time-vested. Private-company RSUs have an expiration date, usually 7-10 years from grant. When shares are eventually delivered, the full value is taxed as ordinary income.
👩💻 Story 1: Jane joins a brand-new company
Jane is the founding engineer at a company that was incorporated two months ago. No funding, no revenue, no 409A. The founder offers her 2%, which is on the low end of normal for a founding engineer at this stage (1-3% is the standard range).
Shares are worth essentially nothing right now - par value, fractions of a cent. Let's walk through her three options.
Option A - RSA
Jane is granted 2% as actual shares. She files 83(b) within 30 days, paying tax on the full grant at today's value, which is effectively zero. From this moment she owns real stock, and the QSBS clock starts ticking.
She leaves after 3 years in good faith with 75% vested. She walks away owning that 75% outright, with essentially no money out of pocket. If the company sells 5+ years after her grant and her stake is worth $15M, the entire gain is federal-tax-free thanks to QSBS - the cap is the greater of $15M or 10x basis, and her grant fits inside it. (If the sale lands at 3 or 4 years from grant, the tiered QSBS exclusion still gets her 50% or 75% of the gain tax-free.)
Option B - ESOP
Jane is granted 2% in ESOPs, subject to the standard 4-year vesting. Since the shares are worth almost nothing right now, the strike price is also almost nothing. The smart move is to early-exercise - that is, exercise the options right away, before they vest. She files 83(b) within 30 days. At this stage there is no real cost.
But ESOPs are not quite the same as RSA. How the QSBS clock runs depends on the specific tax treatment of the options (I won't go into the details here). Best case - with the right setup and early exercise - it's economically equivalent to RSA: same QSBS clock from day one, same long-term capital gains on the upside. Worst case, the QSBS clock effectively starts only as shares vest, which can push part of the gain outside the 5-year window if the exit comes relatively early.
If she does not exercise early and just lets the options vest: after 3 years she has 90 days to exercise the vested portion. Still cheap, still painless. But the clock starts even later, and she's almost certainly missing QSBS on most of her shares unless the exit is far out.
Option C - RSU
Jane is granted 2% in RSUs (double-trigger). Two things matter: when she leaves, and when the company has an exit.
Scenario 1 - early exit, year 2 at $200M. Jane is still employed, 50% time-vested. Both triggers fire, 50% of her grant settles - worth $2M. But the full $2M is taxed as ordinary income (~40%+). If she had RSA, the same outcome would be taxed at long-term capital gains (~20%) - roughly half the tax bill.
Scenario 2 - she leaves at year 3 with 75% time-vested, company exits in year 8 at $1B. Her 75% turns into $15M of ordinary income. No QSBS - RSUs never qualify.
Scenario 3 - company never has an exit within 10 years. Her RSUs expire. She gets nothing.
Verdict
At this stage, RSA is the cleanest choice: the QSBS clock starts immediately on the entire grant, and there's nothing else to manage. ESOPs are a close second - depending on tax treatment, they can be economically equivalent to RSA in the best case, or have the QSBS clock effectively start later (as shares vest) in the worst case. RSUs are strictly worse at this stage - no QSBS, ordinary income on the upside, and a real risk of expiring worthless.
👨💼 Story 2: Josh joins at Series A
Three years later the company raises a Series A at $100M valuation. The 409A now puts the shares at roughly 25% of that, so $25M. Josh joins as a senior engineer and is offered 0.5%, which is around $125K of stock at the current FMV.
Option A - RSA
To take RSAs Josh would have to pay FMV for the shares: $125,000 in cash, up front, and file 83(b) within 30 days to lock in the basis. Most engineers can't or won't write that check. Even if Josh could afford it, he'd be putting $125K of his own money into a single illiquid stock. Not realistic at this stage.
A variant is for the company to simply gift Josh the shares instead of selling them at FMV. The IRS doesn't treat this as a gift - it's compensation. Josh owes ordinary income tax on the full $125K on day one, which works out to roughly $40K out of pocket depending on his bracket and state. On top of that, the company owes employer payroll taxes (~$10K) and has to withhold from his paycheck. So Josh still writes a real check, on day one, for stock he can't sell - just a smaller one than the FMV-purchase version. Most companies don't do this for rank-and-file hires.
Option B - ESOP
Josh gets options to buy 0.5% at a $2.50 strike (FMV at the time of his grant). No upfront cost.
After 2 years he leaves with 50% vested. He has 90 days to exercise or lose them. Cost to exercise: about $62,500 cash. On top of that, if the FMV has since climbed, exercising can trigger additional taxes on the paper gain between strike and FMV - a real tax bill on income he hasn't actually received.
So Josh's choice is: write a check for $60K+ to keep options that might still go to zero, or walk away with nothing. Many people walk. Most who exercise are uncomfortable doing it.
Option C - RSU
Josh gets 0.5% in RSUs with a 10-year expiration. He leaves after 2 years time-vested on 50%, no exit yet. No check to write, no 90-day deadline, no other taxes due. He simply waits.
If the company exits within 10 years of his grant, those shares pay out as ordinary income. If it doesn't, he gets nothing - but he never risked any of his own money.
Verdict
RSA is impractical. ESOPs force an ugly binary choice when leaving. RSUs are imperfect (ordinary income, no QSBS, expiration risk) but they are the only option that doesn't require Josh to gamble his savings or walk away empty-handed.
⚖️ Comparison
Rows are phrased so that ✅ = good outcome for the employee.
Property
RSA
ESOP
RSU
No cash needed at grant
✅ (only when FMV is ~$0)
✅
✅
No upfront cost or tax
⚠️ only at FMV ≈ $0; otherwise pay FMV in cash or owe income tax on the gifted value
✅ (until exercise)
✅ (only for double-trigger; until exit)
Starts QSBS clock
✅ at grant
⚠️ depends on setup (at exercise or as shares vest)
❌
Long-term capital gains on upside
✅
✅ (held 1y after exercise)
❌ - ordinary income
No forced check when leaving
✅
❌ (90-day deadline)
✅
Survives if company never exits
✅
✅ (once exercised)
❌ (expires)
Works at near-zero FMV
✅
✅
❌
Works at high FMV
❌
⚠️
✅
📈 When does the math flip?
As the company grows, two things happen at once: typical grants shrink as a percentage of the company, and the cash needed to convert options into actual stock grows fast. The point where exercise cost gets bigger than an employee can comfortably write a check for is roughly where RSUs start to make sense.
Stage
Round valuation
FMV
Typical grant
Cost to exercise at year 4
Pre-funding
$0
~$0
2.5%
~$25
Seed
$20M
$5M
1.5%
~$75,000
Series A
$100M
$25M
0.5%
~$125,000
Series B
$400M
$100M
0.25%
~$250,000
The bars show typical founding-engineer grant size (%) at each stage. The line shows the cash needed to exercise the full vested grant after 4 years. By Series A you're looking at $100K+; by Series B, often $250K+. Series A or seed stage (depending on the size) is the inflection point for most employees.
Assumes 10M fully-diluted shares and FMV at roughly 25% of round valuation. Additional taxes on the spread between FMV and strike can add meaningfully on top at Series A and later.
🎯 Recommendation matrix
Stage
RSA
ESOP
RSU
Pre-funding (FMV ~$0)
✅ best choice
❌ unnecessary complexity
❌
Seed (FMV under $5M)
⚠️ needs a meaningful check (~$50-100K)
✅ with early exercise + 83(b)
❌
Series A (FMV $5-50M)
❌
⚠️ only with 7-10 year post-termination exercise window (not the default 90 days)
✅
Series B (FMV $50M+)
❌
⚠️ same caveat - exercise costs are now $250K+
✅
🏢 What do large private companies actually do?
The recommendation matrix above is the textbook view. Here's what the market actually looks like at unicorn scale today:
¹ Double-trigger: settles only when both time-vesting and an exit happen.
² Approximate - specific terms vary by company and individual grant.
Vesting acceleration: rank-and-file employees at all of these companies generally get none. Founders and senior executives typically negotiate double-trigger acceleration (vesting jumps forward only if the company is acquired and their role is eliminated within ~12 months). Specific terms are rarely public.
Two patterns dominate:
Double-trigger RSUs are the standard at the unicorn stage. Once a company crosses ~$5B in valuation, the exercise-cost problem becomes too large to push onto employees, and RSUs are the default.
Tender offers have replaced the IPO clock. A tender offer - where the company or an incoming investor buys back vested shares - is itself a liquidity event, so it can serve as the RSU's second trigger. Once an employee has time-vested (first trigger) and the company runs a tender offer (second trigger), the RSUs settle and the employee actually gets paid, with no IPO required. Companies also use this to rescue RSUs that are about to hit their expiration date. Stripe, Databricks, and OpenAI stay private for 10+ years and run these buybacks at regular intervals - Stripe and Databricks alone have done over $15B of it since 2023.
Notably absent: extended (7-10 year) post-termination exercise windows. Coinbase, Pinterest, Quora, Kickstarter and Patreon famously moved to these in the 2015-2018 wave, but the next generation of unicorns mostly skipped this path and went straight to RSUs.
💡 Conclusion
The asymmetry between RSA/ESOP (capital gains + QSBS) and RSUs (ordinary income, no QSBS) is huge.
Default rules of thumb:
Pre-funding (FMV ≈ $0): grant RSA. QSBS clock starts on the entire position from day one, with no moving parts. ESOPs at this stage are unnecessary complexity.
Seed (FMV < $5M): still RSA if the employee can either write the FMV check (~$50-100K) or, if the company gifts the shares, cover the income tax on the imputed compensation (~$15-30K). Otherwise, ESOPs with early exercise and 83(b).
Series A and beyond: most companies should default to double-trigger RSUs. The market has settled here.
One underrated alternative at Series A: ESOPs with an extended 7-10 year post-termination exercise window (instead of the default 90 days). This solves the "leave and lose your options" problem without giving up QSBS or capital gains. Coinbase, Pinterest, Quora, Kickstarter, and Patreon all did this. Worth considering before defaulting to RSUs.
If you want to offer employees a choice between options and RSUs at the same stage, that is possible but requires real tax guidance - most people can't navigate the decision alone and will default to whatever sounds simpler, which is usually not what's best for them.
⚠️ The Bolt cautionary tale: don't lend your team money to exercise
A tempting fix for the exercise-cost problem is to lend employees money to exercise their options. Bolt famously tried this in 2022 under founder Ryan Breslow (TechCrunch, Fortune).
It went badly. Bolt was valued at $11B in January 2022 and lost ~97% of its value over the next two years. Employees who took the loans were left with real debt they still had to repay even though the stock had collapsed - and Bolt laid off ~30% of staff just months later, triggering a 90-day repayment deadline for anyone affected.
The general principle: when a company "helps" employees exercise underwater options with loans, it converts an upside-only gamble (options expire worthless if the company fails) into a downside-bearing one (employees owe real money if it fails). The Bolt employees who declined the offer are materially better off than those who took it.
Cleaner solutions to the exercise-cost problem:
Switch new grants to double-trigger RSUs.
Extend the post-termination exercise window to 7-10 years.
Run tender offers so employees can sell vested shares and use the proceeds to exercise the rest.
📎 Appendix: ISO vs NSO
A quick reference for the difference between the two flavors of stock options. The body of the article assumes ISOs throughout.
Summary: Both ISOs and NSOs give you the right to buy shares at a fixed strike price. The difference is when and how you're taxed. ISOs are gentler on the employee - no ordinary income tax at exercise (just a potential AMT bill on the spread), and capital-gains treatment on the upside if you meet the holding periods. NSOs treat the spread between strike and FMV as ordinary income at exercise, withheld like salary. ISOs are employee-only and capped at $100K of vesting per year; NSOs cover everyone else and any grant beyond the ISO cap.
🧮
What is AMT? The Alternative Minimum Tax is a parallel tax system that runs next to your regular income tax - you compute what you owe both ways and pay whichever is higher. Exercising ISOs creates no regular income tax, but the spread between strike and FMV counts as income under AMT, so a large ISO exercise can leave you owing tax even though you received no cash. The AMT rate is 26-28%. Any AMT you pay on an ISO exercise generally turns into a credit you can recover against regular tax in later years.
Property
ISO (Incentive Stock Option)
NSO (Non-Qualified Stock Option)
Who can receive them
Employees only
Anyone - employees, contractors, advisors, board
Tax at exercise
None for regular income tax; AMT may apply to the spread between FMV and strike
Ordinary income on the spread, withheld at exercise (like salary)
Tax at sale
Long-term capital gains if held 2y from grant and 1y from exercise; otherwise treated like NSO
Long-term capital gains on appreciation above the FMV-at-exercise (if held 1y)
QSBS clock
Starts at exercise with a timely 83(b) (mainstream view); a more cautious minority of practitioners run the clock from vesting
Starts at exercise (with 83(b) for early-exercised)
Annual cap
$100K/year of vesting at strike value; excess automatically converts to NSO
No cap
Best for
Long-term employees who can meet the holding periods
Non-employees and any grant beyond the ISO $100K cap
The practical implication: most early-stage employees get ISOs by default. NSOs show up automatically once a grant exceeds the $100K/year ISO cap (common for senior hires), and for anyone who isn't a W-2 employee.
Worked example: tax at exercise
Suppose you have 10,000 vested options with a $1 strike (the FMV when you joined). You leave the company and decide to exercise when the FMV has risen to $5/share. You buy 10,000 shares for $10,000 in cash. They're now worth $50,000 on paper. The $40,000 "spread" is what matters for taxes.
NSO: the $40K spread is ordinary income at exercise - a ~$14K tax bill on exercise day (at ~35% combined federal + state + payroll). Total cash out: ~$24K ($10K to exercise + $14K tax) for shares you may not be able to sell.
ISO (still within the favorable window): no regular income tax on the $40K spread. It does get reported for AMT purposes - if your overall AMT situation triggers, you might owe roughly $11K (28% of the spread); in many situations it doesn't trigger at all. Total cash out: $10K to ~$21K depending on AMT. If you later sell after holding 2y from grant and 1y from exercise, the gain from $1 strike to sale price is all long-term capital gains (~20%) instead of partly ordinary income.
⚠️ Important catch: under IRS rules, exercising more than 90 days after leaving automatically reclassifies the option from ISO to NSO. So in this "exercise after departure" scenario, you're taxed as an NSO holder - paying that $14K of ordinary tax - regardless of what your employer's contract says.
This is why the extended (7-10 year) post-termination exercise window discussed earlier in the article doesn't preserve ISO tax treatment - the 90-day rule is an IRS rule, not a company rule. What the extended window does preserve is your right to exercise at all: instead of having to scrape together $10K+ within 90 days of leaving, you can wait for a liquidity event and exercise then, paying the tax out of sale proceeds. You lose the ISO benefit, but you don't lose the option itself.
Legal disclaimer: This guide is not intended to and does not constitute legal or tax advice, recommendations, mediation, or counseling under any circumstance. You should seek the advice of a competent attorney or accountant licensed to practice in your jurisdiction for advice on your particular problem.
Subscribe to our newsletter
Practical finance and equity advice for founders, straight to your inbox. No spam, unsubscribe anytime.
