Cap Table Decoded
Cap Table DecodedWhat a $1B Valuation Actually Means for Your 0.1% Stake

What a $1B Valuation Actually Means for Your 0.1% Stake

A billion-dollar valuation doesn't determine what your equity stake is actually worth.

· 8 min read

For an engineer with a small stake, a startup’s headline valuation and the cash they can actually receive are distinct figures. Dilution, investor preferences, exercise costs and the exit itself determine the payout. The useful number is what reaches your shares after those mechanics have done their work.

Why a high valuation is not a share price you own

The latest funding round sets a price for preferred shares. The headline post-money valuation takes that price and multiplies it across all shares as if they were identical. They are not the same financial instrument. Late-stage investors buy preferred stock with rights that employee common stock does not have. Applying the preferred price to every share produces the unicorn number, but it does not give every shareholder the same claim on the company’s money.

Employees generally receive options to buy common stock. Your stake therefore describes an ownership percentage. Investors and employees can receive very different results from the same sale because their shares come with different rights. The funding announcement prices one class of shares and presents the result as a value for the whole business.

How dilution eats your percentage before any exit happens

When a company raises another round, it sells a slice of ownership, reducing the percentage held by existing shareholders. Your grant can stay unchanged while your stake shrinks. The number of options on your paperwork is not the same thing as your percentage of the company. As more shares enter the calculation, the same grant represents less ownership. That happens whether the business is doing well or struggling.

The effect compounds as the company keeps raising. A founding engineer’s pre-seed grant can lose more than half its ownership percentage by Series B through normal dilution alone, even when the company is doing well. A successful funding round can thus give the company a better chance of doing well while cutting your share of the money from an exit. Those two developments are not contradictory. The company has more capital, and you own less of it proportionally.

Grants also tend to start smaller the later an engineer is hired and the further along the company is, with stakes for early engineers much smaller by Series A than at pre-seed and smaller still by Series C. Later hires can begin with less ownership and still face future dilution. AI startups raising at elevated valuations in 2026 are not necessarily finished raising money, so the percentage on an offer cannot be treated as the percentage at exit. None of this requires a disappointing sale.

The liquidation waterfall — why common stock is last in line

Dilution determines how much of the company you own. Liquidation preferences set who gets paid before you. Preferred shareholders have the right to recover their investment, sometimes a multiple of it, before common shareholders receive proceeds. The total claim ahead of common stock is the preference overhang. A company can sell for a lot of money and still leave nothing for the common shares employees hold.

For an illustrative example, suppose a company sells for a large sum and has an equally large preference stack that must be paid ahead of common stock. All the money from the sale goes to pay that stack. Nothing remains for the employees’ common shares. Whether an employee’s stake is unchanged or has been diluted does not change that result. The company achieved a large exit, but its common shareholders did not receive a payout.

With participating preferred shares, investors first collect their preference and then also share in the remaining proceeds with common shareholders. Both the amount of the preference and the participation terms shape what reaches employees. A valuation alone reveals neither.

This explains how vested equity can have negligible value even when a company finds a buyer. Vesting establishes that an employee has earned the options. At companies with large late-stage rounds, the claims ahead of common stock can be substantial. For an engineer evaluating an offer, disclosure of those claims is part of understanding the compensation, rather than a technical discussion to postpone until an acquisition.

How AI-era valuations have made the math worse at the grant stage

AI-native startups in 2026 are raising money at sky-high valuations, sometimes before they ship a product. That raises what people expect of the company before an engineer even gets a grant. A valuation can reflect confidence in a future business without showing that the business has already earned that confidence through its product. For the employee, the entry point is expensive before the outcome is established. There is less room for the eventual company to exceed the promise embedded in the offer.

A small stake offered at a $2B valuation needs a much larger exit to pay out as much as that starting valuation suggests than a grant that is otherwise much the same but offered at a lower valuation. With the same percentage and the same actual exit, the proceeds are the same even if the earlier headline valuation was higher.

That high starting point gets harder to overcome when options cost a lot, the company raises more money and it has a large preference stack. The exit has to support the expectations attached to the grant while leaving money after the claims ahead of common stock. AI/ML specialists command a significant premium over generalist software engineers at startups, but the premium described in the June 2026 compensation notes is being paid in cash. A strong salary premium and a difficult equity calculation can exist in the same offer.

The 90-day exercise window as a silent equity killer

Options are a right to buy shares. To turn vested options into shares you own, you must pay the strike price, and a large later-stage grant with a higher 409A price can be expensive to exercise. Leaving the company starts another clock. Most options expire within a short post-termination window, commonly 90 days, so an engineer who cannot pay can lose vested options even if the company is worth much more later.

An option can be in the money on paper and still expire because its holder cannot afford to exercise it before the deadline. A small number of companies have extended their windows to multiple years. The practical step is to ask explicitly about the post-termination exercise window before signing and negotiate it where possible.

What the distribution of actual outcomes looks like, from Plaid to OpenAI

OpenAI sits at the exceptional end of the private-company opportunity described here, but even that example requires attention to the instrument. For years, OpenAI gave employees Profit Participation Units rather than standard stock options or RSUs because it began as a capped-profit company under nonprofit control. The company restructured into a public benefit corporation in late 2025 and described moving toward a more conventional equity model. The legal form cannot be skipped over. A percentage attached to a Profit Participation Unit is not something an engineer can evaluate simply by assuming it works like a conventional option grant.

Plaid provides a different reference point. Plaid’s most recent priced valuation, set in a February 2026 employee tender offer, remained well below its 2021 peak. Engineers who joined near that peak and whose options were struck near that price saw the valuation premise of their grants collapse and then partly recover. A recognizable company with an employee liquidity event can still leave peak-era hires facing a much weaker equity proposition than the one they joined for.

Engineers mostly hear about well-known private companies, but startups also end in shutdowns and acquihires where founders get value back and employees mostly get nothing. The spectacular result provides an unreliable midpoint for an offer calculation. OpenAI illustrates an exceptional opportunity with an unusual legal structure, while Plaid shows how entry timing can reshape employee outcomes even when a company continues operating and provides liquidity.

The secondary market as a partial but real new variable

Secondary tender offers provide a genuine counterweight to the idea that employees must wait for an acquisition or public listing. At the largest private AI companies, they have become a routine retention tool, and one 2025 tender at a frontier AI company reached $6.6 billion. That creates an earlier route to cash. Share repurchases at this scale were historically associated with companies nearing public trading, but frontier AI companies can now reach mega-cap valuations while remaining private.

OpenAI’s tenders are the most prominent documented examples in the notes, and Anthropic has also completed employee tenders. These transactions help retain engineers and reduce pressure for premature IPOs. For employees who can participate, a private-company grant is therefore not necessarily locked away until a final exit.

The limit is where this mechanism applies. An engineer joining a Series A or B company cannot rely on frontier-scale secondary liquidity as part of an expected payout. Secondary prices can also rest on a small number of large trades in a thin market, and a completed tender is not proof of lasting enterprise value. Neither company has been tested by public markets, and secondary pricing could compress sharply if sentiment shifts. A tender offers cash at a particular moment.

Five things to ask before you treat a valuation number as your financial plan

Five questions turn an equity pitch into a conversation about what the grant can actually deliver. The first is what the total liquidation preference overhang is in dollars. That establishes how much of an exit must go to preferred shareholders before common stock receives proceeds.

The second is whether those preferred shares participate. The dollar overhang alone does not explain the full split if investors can collect their preference and then share in the balance.

The third is what the post-termination exercise window is. A short window can make ownership depend on whether you have enough cash at departure, while a multi-year window gives you more time. The fourth is what instrument you are actually receiving, whether an ISO, NSO, RSU or something non-standard such as a Profit Participation Unit. The label has legal and tax consequences. A percentage and an estimated dollar value do not substitute for knowing the terms of the instrument.

The fifth is what realistic exit would make the grant financially meaningful compared with the current valuation. That calculation needs to apply future dilution and the preference terms rather than simply multiplying today’s stake by the unicorn number. If the company cannot show a path where common equity pays out, that is useful information before signing. These are reasonable questions for a well-run company to answer. Its willingness to explain the numbers is itself a signal about how openly it deals with the people it wants to hire.

Sources

  1. Dilution Modeling for Founder Ownership
Filed underOffer Valuation

More in Offer Valuation