Why most founder equity splits quietly destroy value

I price private companies for a living. More than a thousand of them, on deals up to roughly $26 billion, and in a large share of those files there is a founder equity split sitting at the bottom of the cap table doing quiet damage. Not because anybody got cheated. Because the split was set in an afternoon by two people who wanted to get back to building, using information neither of them had yet. The founder equity split is the first price your company ever sets, and it is the only one you get to set for free. It is also, reliably, the price they spend the least time on.

Most founders will face this decision. Carta's 2026 founder ownership data has two-person teams making up 36 percent of the startups that closed a round on its platform, rising to 40 percent in SaaS, even as solo founding has doubled over the past decade to 36 percent of new companies formed on Carta in 2025. Two founders is still the most fundable shape there is. So the split gets made, and it usually gets made badly. Here is what actually goes wrong, in the order the damage compounds.

The split is a price, and the market can read it

Start with the evidence, because it is better than the folklore. Thomas Hellmann and Noam Wasserman studied 1,476 founders across 511 ventures for a paper called The First Deal, published in Management Science. Roughly a third of the teams divided their equity equally. And equal splitting was associated with lower pre-money valuations at the first financing round. They put a number on what is at stake in the decision itself: about 10 percent of firm equity, roughly 25 percent of the average founder's stake, or something on the order of $450,000 in net present value per venture.

Read that carefully, because the causal claim is not the one people want to make. The authors' own econometric tests suggest the effect is driven by unobservable heterogeneity, and it is most pronounced in the teams that settled the split fastest, in less than a day. In plain terms: the equal split is probably not what lowers your valuation. The equal split is a readout of something else that does. A team that could not sit through one hard conversation about equity is telling you how it will handle the next fifty. That is the thing getting priced.

I have sat on the reading side of this many times. When I open a cap table and find a clean 50/50 with no vesting and no documented reasoning, I am not offended by the fairness of it. I make a note about the team, and the note is not flattering. It says these two have never tested each other. Everything else in diligence then gets read through that note, which is a bad way to be read.

A price set at the moment of least information

Now the timing, which is the actual mechanical failure. Wasserman's work across roughly 10,000 founders found that 73 percent of teams split equity within a month of founding, that most of those splits are static, and that most are equal. So the modal founding team prices a four-year commitment using about three weeks of evidence, and then never reopens it.

Put that next to how anything else in private markets gets priced. I do not value a company on one day of data. I normalize a few years of financials, build a size-appropriate comp set, triangulate across methods, and produce a range I can defend line by line, and that is for a minority stake somebody buys with money they can diversify. The founder split is a larger position, held with no diversification at all and no liquidity for a decade, and it gets decided over dinner.

What you cannot know in week three is close to everything that matters. You do not know who quits their job first, or when. You do not know which of you can actually sell, which is usually the difference between a company and a project. You do not know who can recruit, whose network converts, who holds up in month fourteen when nothing is working and the money is going out anyway. And you do not know who will still be there. You are pricing four years of contribution with three weeks of enthusiasm.

The past-contribution trap

When teams do negotiate, they negotiate over the wrong inputs. The conversation almost always turns on what has already happened: whose idea it was, who registered the company, who built the prototype over the holidays, who put in the first $20,000. Hellmann and Wasserman found the three founder characteristics that drive a premium in unequal splits are idea generation, prior entrepreneurial experience, and capital contributions. Notice that two of those three point backwards.

Sunk contribution deserves something. It does not deserve the majority of a company that has not been built yet. Equity is not a receipt for work performed, it is payment for work not yet performed, and it vests across exactly the period when that work either happens or does not. The idea is worth a premium, not a controlling stake. If you want a frame you can defend, ask a valuation question instead of a credit question: what would it cost this company to replace each of us for the next twenty-four months? Price the roles, not the origin story. That question has answers you can look up, and it moves the conversation off whose feelings are in the room.

Dead equity, the version you can actually measure

Everything above is soft. This part is arithmetic. Carta looked at 22,352 founders in two and three person venture-backed teams and found that about 23 percent of cofounders had left by year three, more than 30 percent by year five, and close to 40 percent by year seven, with recent cohorts departing faster than earlier ones. That is not a tail risk you can wave off. On a long enough horizon it is closer to a coin flip than to an exception.

So run the two branches. If the departing founder's stock is vesting, the unvested portion goes back to the company, per the standard mechanic Cooley describes: a four-year vest with a one-year cliff and a company right to repurchase unvested shares at the original purchase price. The equity returns to a place where it can do work, which usually means hiring the person who now has to cover that role. If there is no vesting, the departing founder keeps all of it, and you have created dead equity: a block of ownership that does no work, cannot be re-granted, and takes its full share of dilution in every round you will ever raise. It also dilutes the people who stayed. Every point of every future round gets shared with somebody who left.

This is why I think the usual framing of vesting is backwards. Founders treat it as an insult, a signal that somebody does not trust somebody. It is the opposite. Vesting is the only thing that makes the split conditional on the thing you were paying for. Without it a split is not a split, it is a gift with a story attached. Investors understand this, which is why they will impose it if you have not: the convention Cooley describes has founders no more than roughly 40 percent vested by the Series A. If you do not put vesting in place at formation, you will accept it later, at a worse moment, on somebody else's terms.

The window closes, and tax law is what closes it

The most common answer to all of this is that you will fix it later if it turns out wrong. That is not a plan. It is a decision to pay for the repair, and the price is set by the tax code rather than by you.

At formation the company's stock is worth approximately nothing, so moving equity around costs approximately nothing, and a timely 83(b) election locks the tax at that grant-date value. Once the company has a real fair market value the picture changes completely. Under Section 83 of the tax code, stock issued in connection with services is ordinary income to the recipient to the extent its fair market value exceeds what they paid for it. So the correction that was free in month one becomes compensation income, with withholding, at a valuation that is now real, for a founder who has no cash to pay it with. The other direction is no easier: vested shares belong to the person holding them, and the routes back are a purchase or a negotiation in which you have no leverage.

Wasserman's canonical example is Zipcar, where the founders shook hands on a 50/50 quickly and then spent roughly the next year and a half in angst over it. The angst is the cheap part. The structural version is that the free window closes after a few months, and most teams discover this at exactly the moment the split has started to hurt.

The 50/50 with no tie-breaker

One more failure mode, and it is a governance problem rather than an ownership one. An even split between two people with no casting vote, no domain-level decision rights, and no designated lead is a deadlock waiting for a disagreement large enough to trigger it. Ownership was never the risk. Unbreakable symmetry is.

Investors underwrite this directly. Cooley's venture financing data has protective provisions, meaning investor veto rights, in more than 90 percent of deals as of the second quarter of 2025. The people funding you are already buying control rights carefully, and they are not inclined to add an untouchable two-person tie to the structure. And team is not a soft factor in that decision: in the survey of 885 venture capitalists at 681 firms by Gompers, Gornall, Kaplan and Strebulaev, VCs rated the management team as more important than the product or the technology, and attributed ultimate success or failure more to the team than to the business. Your team structure is part of the asset being priced. A deadlock is a defect in the asset.

Equal ownership is fine. Equal authority over everything is not. Split the equity however the roles justify, then write down who decides what and who breaks a tie. It takes a page, and it is the cheapest page in the company.

How I would set a split

If I were doing this from scratch, in this order. Start with the next twenty-four months. Write down what each person owns operationally and what replacing them would cost. That is the base of the split. Then adjust for what is genuinely at risk. Who goes full time, on what date, giving up what salary, and who is writing checks into the company. Opportunity cost and capital are real, and unlike the origin story they are forward-looking. Then add an idea premium, out loud and small. Name it as a number, say what it is for, and move on. Then vest everything. Four years, a one-year cliff, a company repurchase right on unvested shares, 83(b) filed on time, no exception for whoever feels most senior. Then write the decision rights. Domains, plus a tie-breaker. Then set a revisit trigger tied to an event, the first employee, the first revenue, the first financing, rather than to whenever somebody finally gets resentful enough to raise it. A split you have agreed in advance to reopen at a defined moment is a different instrument from one that can only be reopened by a fight.

None of that requires the conversation to be comfortable. It requires it to happen once, on the record, while the stakes are still theoretical and the tax cost is still zero.

The takeaway from someone who prices them

Here is the frame I would leave you with. Your split is a multiplier on a number that only gets smaller. Carta's data on rounds raised from 2021 through 2025 has the median founding team holding about 56 percent of the company on a fully diluted basis by the seed round, and about 36 percent by the Series A. Whatever fraction of the founding team you negotiated on day seven, that is the fraction you keep of a shrinking share, through every round, through every option pool top-up, and behind every liquidation preference in the stack. A couple of points argued over at formation compound quietly for a decade.

So the destruction is never dramatic. Nobody gets robbed. A split gets made too fast, on backward-looking inputs, with no vesting and no tie-breaker, and then it simply sits there: signaling something unflattering to every investor who reads the cap table, holding dead equity that dilutes the founders still working, and getting more expensive to fix every month that passes. The equity split is the cheapest valuation decision you will ever make, and you make it when you know the least. Spend a week on it. It is the best-priced week of work available to you.

Common questions

Is a 50/50 founder equity split always a mistake?

No, but a fast 50/50 usually is. Hellmann and Wasserman's study of 1,476 founders across 511 ventures found roughly a third of teams split equally, and that equal splitting was associated with lower pre-money valuations at the first financing round, with the effect strongest among teams that settled the question in less than a day. Their own tests suggest the split is not the direct cause but a signal of something investors price. An equal split reached deliberately, with the reasoning written down, vesting in place, and decision rights assigned, is defensible. An equal split reached in an afternoon to avoid an awkward conversation is a diligence flag, and it is the avoidance rather than the arithmetic that costs you.

What actually destroys value in a founder equity split?

Four things, and none of them is unfairness. First, timing: 73 percent of teams split equity within a month of founding, so a four-year commitment gets priced with three weeks of evidence. Second, backward-looking inputs: the negotiation turns on whose idea it was and who put in the first money, when equity is payment for work not yet done. Third, no vesting, which turns any departure into permanent dead equity that dilutes the founders still working. Fourth, no tie-breaker, which leaves a two-person deadlock sitting in the governance structure that investors have to underwrite. Every one of the four is fixable at formation for nothing.

Should founders have vesting on their own shares?

Yes, always, and it is protection rather than an insult. Carta's data on 22,352 founders in two and three person venture-backed teams shows about 23 percent of cofounders gone by year three, more than 30 percent by year five, and close to 40 percent by year seven. Vesting is what decides whether a departure returns equity to the company or leaves it stranded on the cap table. The standard structure Cooley describes is a four-year vest with a one-year cliff plus a company right to repurchase unvested shares at the original purchase price, with the 83(b) election filed on time. If you do not put it in place yourself, an investor will require it later on worse terms, since the convention is that founders are no more than roughly 40 percent vested by the Series A.

Can we just fix the equity split later?

Only cheaply at the very beginning. At formation the stock is worth almost nothing, so reallocating costs almost nothing, and an 83(b) election fixes the tax at that grant-date value. Once the company has a real fair market value, stock issued in connection with services is ordinary income to the recipient to the extent fair market value exceeds what they paid for it, so the same correction becomes taxable compensation, with withholding, for a founder who has no cash. Vested shares already belong to whoever holds them, and getting them back means a purchase or a negotiation with no leverage. The free window is months, not years, and it closes well before the problem becomes obvious.

What should a founder equity split actually be based on?

Forward contribution, priced the way you would price a hire rather than credited like an origin story. Ask what each founder owns operationally for the next twenty-four months and what it would cost the company to replace them, then adjust for what each is genuinely putting at risk: the full-time start date, the salary given up, and cash going into the company. Add an explicit and modest premium for the idea and the early work, name it as a number, and stop there. Then vest everything, write down who decides what and who breaks a tie, and set a revisit trigger tied to an event such as the first employee or the first financing. Unequal is fine. Undiscussed is what costs you.

Related reading

• The cofounder is the most expensive line on your cap table.
• Find a co-founder with no network, then split equity without regret.
• How to read a cap table before it reads you.
• The option pool: the dilution that is only yours.

Go deeper

If you are setting a split now, or trying to unwind one that was set too fast, let's price the roles and model what the vesting and the dilution actually do to each of you before anything gets signed.

Tomasz Felpel is an investor, founder, and advisor in private markets and healthcare, based in New York. He is a three-time founder of Value Alpha, an AI-powered private-markets valuation platform, Sonnerie VC, an early-stage healthcare venture firm, and Pond. Previously he led corporate development and M&A at Fortune 500 scale, pricing more than 1,000 private companies. Columbia Business School EMBA. Read the full story.