How to split equity between co-founders, including across countries (2026)

There's no equity split formula that's right for every team. Most credible advice, YC's included, favours an equal or near-equal split for co-founders who start together and both commit full-time, with vesting on everyone's shares so nobody walks away with a large stake after a few months. Whatever you choose, decide it early and put it in writing.

This guide covers what YC advises, what Carta's data shows about real splits, how vesting and leaver terms work, and the extra questions you face when co-founders live in different countries: which company, whose tax, the US 83(b) deadline and IP assignment. Our guide to finding a co-founder promised a fuller equity guide. This is it.

The short answer

Split equally or nearly equally if you're starting together and both going full-time, vest everyone's shares over four years with a one-year cliff, and sign it before you issue shares. In a 2015 post, YC's Michael Seibel says YC almost always recommends equal splits among co-founders, and that equal or near-equal splits should become standard. His blunt test: if you won't give a partner an equal share, perhaps you're choosing the wrong partner.

His reasoning is about time. Building a company takes, in his words, 7 to 10 years, so a few months' head start or the original idea is small next to the work still ahead. He also argues an unequal split can signal to investors that the CEO doesn't value the co-founders.

Vesting is the safety net. Seibel describes the typical setup as four-year vesting with a one-year cliff, so a founder who leaves or is fired within the first year gets nothing.

What the data shows about real splits

Carta's data shows equal splits are getting more common, but still aren't the majority. Carta runs equity management software, so its numbers cover startups that use Carta, not every company. Its January 2025 Founder Ownership Report describes its data as covering the US venture ecosystem.

Dynamic duos (Carta, 18 June 2026). This analysis looks at two-founder teams that formed companies on Carta from 2016 to 2025. It found:

• In 2025, 44.6% of two-founder teams split equity equally, up from about 34.5% of teams formed in 2016. Carta calls 2025 the highest point of the past decade.
• The median two-founder split in 2025 was 51-49, down from 57-43 in 2021.
• In SaaS, the median two-founder team from the past three years split 51-49. In biotech, the lead founder in a typical two-founder team gets 58%, which Carta links to companies built around one founder's research.
• In teams of three, four or five, the lead founder typically gets at least 10 percentage points more than the next co-founder.
• Between 25% and 35% of two-founder teams formed from 2016 to 2021 had parted ways after five years. For teams formed from 2016 to 2018, more than 40% had split up within eight years.

Founder Ownership Report 2026 (Carta, 12 March 2026). Its summary says the rate of equal splits rose to 27.3% for three-founder teams in 2025, up from 21% the year before, and to 16.7% for four-founder teams, up from 10.8%. It also shows how fast founders get diluted: on rounds raised from 2021 to 2025, the median founding team holds about 56% after a seed round and 36% after a Series A.

Founder Ownership Report 2025 (Carta, 21 January 2025). This one used anonymised data from more than 45,000 startups incorporated from 2015 to 2024, and put the 2024 equal-split rate for two-founder teams at 45.9%, against 31.5% in 2015.

Carta's figures move between reports as its data updates. The 2025 report's 45.9% for 2024 sits above the 2026 analysis's 44.6% for 2025, even though the later piece calls 2025 the decade high. Read them as a trend, not exact numbers. An older Carta study of 7,764 companies and 18,228 founders founded from 2019 found 41% of two-founder teams split equally.

The breakup numbers matter most. If a quarter to a third of two-founder teams part ways within five years, the leaver terms you sign on day one will probably get used.

When an unequal split makes sense

An unequal split makes sense when the founders' commitments are genuinely different, not when one person had the idea first. Seibel lists the common reasons founders give for unequal splits, such as coming up with the idea, starting earlier, age or experience, or needing a tie-breaker, and argues they miss the mark over a decade-long build.

Situations that can justify a gap:

• One founder isn't full-time. Someone keeping a day job is taking less risk. YC's 10 questions for co-founders include what needs to happen for each of you to go full-time.
• One founder is putting in money. Consider treating cash as a separate investment on its own terms rather than folding it into the founder split.
• The company is built on one person's prior IP. Carta's biotech figures reflect this pattern.
• A founder joins later. Someone joining after real progress is closer to an early hire than a founding partner.

If one of you started months earlier, there's a cleaner fix than a bigger stake. Stripe Atlas lets each founder set a vesting start date up to six months before incorporation to credit work done before the company existed.

Do equity split calculators work?

Equity split calculators can start the conversation, but they can't settle it. They score inputs such as the idea, hours so far and cash, then turn the scores into percentages. We found no primary source that validates any calculator's weightings, and they mostly measure the past, while YC's argument is that most of the work is still ahead.

A better process:

1. Each founder answers YC's 10 questions in writing, especially on roles, money, time commitment and what happens if you split up.
2. Compare your answers. Big gaps in commitment are the real equity question.
3. Check your proposed split against Carta's data for your team size.
4. If you land on anything other than near-equal, write down why, so it still makes sense in three years.

Vesting: how it works and why everyone needs it

Vesting means founders earn their shares over time, so a founder who leaves keeps only what they've earned. Stripe Atlas's documentation says most founders use a four-year vesting period and a one-year cliff, starting on the date of incorporation.

In practice:

• Year one: nothing vests. Leave before the cliff and you get no shares.
• The cliff date: 25% vests at once.
• After the cliff: the rest vests monthly, 1/48 of the total each month, until year four.

Atlas offers four, five or six-year vesting periods for the whole team, a one-year cliff or no cliff for each founder, and the six-month backdating option above. Its docs note the one-year cliff is the industry standard and generally preferred by investors.

Vesting protects the founder who stays. You can't know on day one which of you that will be, which is why it belongs on everyone's shares, including the CEO's.

Acceleration

Acceleration means some or all unvested shares vest early when a set event happens, usually an acquisition. Stripe's founder equity guide describes two kinds:

• Single-trigger: shares vest after one event, typically a change of control. Stripe says this isn't standard in Silicon Valley.
• Double-trigger: shares vest, commonly 100% of what's unvested, only after two events, usually a change of control and then termination by the new owner. Atlas uses double-trigger by default.

What happens when a co-founder leaves

When a co-founder leaves, they usually keep their vested shares and the company buys back the unvested ones. Stripe's guide says the company typically repurchases unvested shares at the price the founder originally paid, not their current value. Atlas's document set includes a stock assignment that lets the company repurchase shares under the stock purchase agreement, and joint escrow instructions that hold founders' shares with the company secretary.

Your agreement should also say who decides a departure, what notice is needed, and what happens to a leaver's board seat and role. Seibel suggests only the CEO holds a board seat before a significant equity raise, to avoid board disputes if a co-founder has to be let go.

Cross-border: which company, and whose law

The company's country, and in the US its state, sets the company law your shares sit under, so the company question comes before the equity question. It's also why YC publishes separate SAFE versions for companies outside the US (more on that below). Our guide to where to incorporate as a remote founder compares the options, and our guide to setting up a US LLC as a non-US founder covers the US route.

The entity type changes how vesting works, so check it:

• US C-corp. Founder shares, vesting, repurchase and the 83(b) election follow the familiar startup pattern above.
• US LLC. Stripe Atlas says equity in an LLC it creates doesn't vest. If you want vesting or leaver terms in an LLC, ask a lawyer to build them into the operating agreement.
• A company outside the US. Local company law and local templates apply. YC publishes its SAFE in versions for Canada, the Cayman Islands and Singapore and says to consult a lawyer licensed in the relevant country before using them. Founder documents need the same care.

Fundraising also dilutes everyone. YC's post-money SAFE makes the stake you sell easy to calculate: investment divided by the valuation cap, so its page gives US$1 million at a cap of about US$6.7 million as roughly 15%. Carta's medians above show where that usually leads by Series A.

Cross-border: each founder's own tax

Each co-founder may face tax on their shares where they live, regardless of where the company is. Founder shares can count as employment income in some countries, and the timing rules differ. Our guide to tax residency for remote founders explains how countries decide who's resident. Here's what three tax authorities publish.

United States: the 83(b) election

A US taxpayer who receives founder shares that vest can file an 83(b) election to be taxed on their value when they're issued rather than as they vest. The IRS instructions for Form 15620 say:

• The election must be filed no later than 30 days after the date the property was transferred. If day 30 falls on a weekend or legal holiday, a postmark on the next business day counts.
• You can use Form 15620 or a written statement that meets the Treasury regulations. The IRS says to file it with the Service Center where you file your return.
• You must give a copy to the company you're providing services to.
• The election can't be revoked without IRS consent.

For non-US founders, Stripe Atlas's documentation says that if you file US returns, or might become a US taxpayer, the election is important. If you aren't and don't expect to be, it doesn't matter, because it affects only US tax. Atlas says it files 83(b) elections for its C-corp founders, and that if you have no SSN or ITIN it tells the IRS you're a non-US taxpayer.

Don't leave this to the last week. Thirty days passes quickly while you're opening bank accounts.

United Kingdom

HMRC's manual says there's no concept of "founders' shares" in the legislation: a founder who is a director from the start acquires employment-related securities. It adds that if the founder pays full market value when they get the shares, that status alone creates no liability on acquisition, but value passed to them later, other than through normal commercial growth or dividends, is taxed as employment reward.

For shares with restrictions such as forfeiture, HMRC describes a section 431 election that the employer and employee make jointly, in a form HMRC approves, no more than 14 days after the acquisition. GOV.UK's employee share schemes overview says shares given as a reward are subject to tax, and you report it through Self Assessment if your employer doesn't deduct it through payroll.

Australia

The ATO's employee share scheme (ESS) rules cover shares or rights you get in the company you work for at a discount. The ATO says the discount is assessable income, and if you pay full value the ESS rules don't apply, though capital gains tax can.

Most founders won't get the ESS concessions. They require the 10% ownership and voting rights test: after acquiring the interests, you can't hold more than 10% of the company or control more than 10% of the votes, counting your associates. The start-up concession also needs an Australian resident employer company.

For any other country, ask its tax authority or a local adviser the same questions: is founder equity taxed as income, when, and is there an election like the US or UK ones?

IP assignment: the company must own what you build

Everything you build for the company, before and after incorporation, should belong to the company rather than to any founder. Investors check this, and unclear ownership becomes a problem the day a co-founder leaves with code they wrote.

Stripe Atlas's standard documents, prepared with Cooley LLP, show the usual approach:

• A Restricted Stock Purchase Agreement and Assignment Agreement, under which Atlas founders pay for their shares with the IP they've developed on behalf of the company. Atlas warns that if you have significant IP or assets to contribute, it might not be right for you.
• A Confidential Information and Invention Assignment Agreement, which Stripe says is commonly used to ensure relevant IP created both before and after incorporation belongs to the company. If you want to exclude prior inventions, Atlas might not be right for you.

Stripe's equity guide adds that startups often make buying equity conditional on signing IP assignments, which helps avoid IP disputes in investor due diligence.

For remote teams, think about:

• Code in personal repositories and domains in personal names. Move them into company accounts.
• Day-job contracts. Check whether a current employer could claim work you did on its time or equipment.
• Contractors and early helpers. They need their own assignment agreements too. Our guide to hiring remote employees internationally covers contractor setups.

What to put in writing

Put everything below in signed documents, prepared or reviewed by a lawyer in the company's country:

• Equity split: each founder's shares and the reasoning.
• Roles and titles: who's CEO and who owns which decisions.
• Time commitment: full-time or part-time, start dates, and what happens if someone can't keep it up.
• Vesting: period, cliff, start date and acceleration.
• Leaver terms: what happens to vested and unvested shares, the repurchase price, notice, and board seats.
• Decision-making: how you break a deadlock.
• Money: salaries, and whether anyone puts cash in and on what terms.
• IP: assignment of everything built before and after incorporation, and any excluded prior work.
• Tax filings: who files the 83(b), s431 or local equivalent, and by when.
• Location: where the company is based and where each founder lives, as YC's questions put it.

If you'd like to talk this through with other founders who build remotely, request an invite to Remote Builders. It's free, and a person reads every request.

Common mistakes

• Splitting on the idea. Seibel's point: execution over 7 to 10 years matters more than who thought of it.
• No vesting because you trust each other. Carta's breakup data says a quarter to a third of two-founder teams split within five years.
• Agreeing on a handshake. Memories differ in year three. Sign it.
• Missing the 83(b) deadline. It's 30 days from the transfer for US taxpayers, so file in the first week, not the last.
• Assuming home-country tax doesn't apply. The UK treats founder directors' shares as employment-related, and other countries may too.
• Forgetting IP assignment. Pre-incorporation code and personal accounts need to move into the company.
• Picking the entity after the split. An LLC and a C-corp handle vesting differently, so choose the company first.
• No deadlock rule. Fix it before you need it.
• Using a calculator as the answer. It scores the past. Your agreement should cover the future.