This post uses hypothetical scenarios for illustrative purposes only. It does not describe any actual client, transaction, or representation, and is not legal advice.
Here is the post-termination exercise window story engineers almost never hear before signing an offer letter: an option that has fully vested over four years can disappear ninety days after the engineer’s last day, and the only thing standing between the holder and the exercise price is cash they may not have. The plan calls this an exercise window. In practice it is a tax cliff, sometimes a liquidity cliff, and almost always shorter than the holder thinks. This piece is a companion to the parent post on drafting equity and IP in startup hires, focused entirely on the window between “vested option” and “expired option.”
The three-month default is not a custom — it is a tax rule
Almost every standard stock plan provides that vested options expire three months after termination of service. Holders sometimes assume that number is industry convention, or a holdover from old plans, or something a sympathetic board might extend in their case. It is something more durable than that. It is a feature of the Internal Revenue Code.
Under IRC § 422, the statutory regime for incentive stock options, an option qualifies as an ISO only if it is exercised by the holder while the holder is an employee or within three months after the holder ceases to be an employee. The twelve-month variant in the same provision applies to termination on account of death or disability. If the holder waits longer than three months (or twelve, in the death-or-disability case) after termination to exercise, the option loses ISO status and is treated as a nonqualified stock option for tax purposes. Plan drafters did not invent the three-month window. They imported it from the tax code because anything longer broke the ISO classification the company had been marketing the grants under.
That is why the window appears in plan after plan with metronomic consistency. The company could, as a matter of contract, extend the exercise window — but doing so causes the option to lose ISO treatment for any exercise occurring after the statutory period. The plan default optimizes for tax classification at the price of holder flexibility.
What the holder actually loses when ISO treatment evaporates
The conventional pitch for ISOs is that, if a holder satisfies both the “two years from grant” and “one year from exercise” holding periods, the entire gain from sale is long-term capital gain. There is no ordinary-income event on exercise (for regular tax purposes), and the bargain element does not appear on the W-2. That is the headline.
The fine print is that exercise of an ISO is an alternative minimum tax preference item. The spread between the exercise price and the fair market value on the exercise date counts as income for AMT purposes, even though it is invisible for regular tax. At a high-growth startup with a fast-rising 409A valuation, that spread can be substantial. An engineer who waits to exercise until the fair market value is many multiples of the exercise price, and then exercises a large block at once, can generate a federal AMT bill that is real cash owed to the Treasury — for shares that are not liquid and cannot be sold to fund the tax.
The post-termination context makes this worse. The holder has just left the company. They likely do not have a fresh payroll income stream against which to plan the AMT. The shares they would acquire on exercise have no public market. They may not have the exercise price in cash, let alone the exercise price plus the AMT. The three-month window forces a decision under all of that pressure. And if the company has been increasing its 409A valuation steadily, the spread is at its widest just when the window is shortest.
The for-cause exception eats the window entirely
The three-month window is the friendly default. Many plans go further and provide that on a termination for cause — defined however the plan defines it — vested options expire immediately on the termination date. Not three months. Not the end of the calendar quarter. The day of termination.
That outcome runs the cause definition straight into the option economics, which is why the companion piece on how to negotiate the cause definition matters as much as it does. A broad plan-form cause definition gives the board a one-step path from disputed conduct to extinguished equity. A narrow, negotiated cause definition with notice and cure provisions converts the same dispute into a process, during which the holder retains the ability to exercise vested options under the standard window.
Death and disability go the other direction. Most plans extend the post-termination exercise window to twelve months in those cases, consistent with the ISO statute. That is the most generous treatment built into the default. Nothing in the standard form mirrors it for an ordinary involuntary departure, which is exactly the situation a holder is most likely to face.
The extended-exercise trend, and why it is not yet the default
A handful of well-known startups have, over the past decade, restructured their stock plans to extend the post-termination exercise window for vested options to something much longer than three months. Pinterest in 2015 announced a seven-year post-termination exercise window for employees who had been at the company at least two years. Quora followed with a similar policy. Coinbase publicly extended its window to roughly the remaining option term (often around ten years from grant) for departing employees meeting tenure conditions. Each of those companies framed the move as an alignment-with-employees policy, on the theory that the three-month window penalized loyal early hires who could not afford to exercise on their way out the door.
The extended-exercise model carries trade-offs. Any exercise outside the three-month (or twelve-month death-or-disability) statutory window converts the option from an ISO to a nonqualified option for that exercise — meaning the bargain element is ordinary income on exercise and is reportable to the IRS. Companies that go down this path generally communicate openly that the conversion is the price of the flexibility, and they often provide tax-planning resources to departing employees so the conversion is not a surprise. The model has not become the default for the industry, in part because the conversion is a real loss of the ISO benefit and in part because the standard four-year, twenty-five-percent-cliff vesting schedule already builds in much of the retention the company wants. Extended exercise is best understood as a discretionary benefit a particular company has chosen to offer, not as the new normal.
For an engineer evaluating offers, the question to ask is direct: what is the post-termination exercise window for vested options, in writing, and is it the plan default or something extended? The answer is in the plan and the award agreement, not in the recruiter’s pitch. If extended exercise is offered, the holder should also confirm whether the company will deliver Section 6039 information statements and the relevant W-2 or 1099 reporting in the year of exercise — these are the inputs the holder’s tax preparer needs to handle a post-conversion nonqualified exercise correctly.
The 83(b) early-exercise interaction
Some plans permit early exercise — the holder may exercise the option before vesting, acquiring restricted shares that are subject to the company’s right to repurchase at the original exercise price (cost) until the underlying vesting schedule runs. Early exercise pairs with an IRC § 83(b) election filed within thirty days of the exercise, which accelerates the income recognition event to the date of exercise (when the spread is typically zero or near-zero on a fresh grant) and starts the long-term-capital-gain clock immediately on the shares.
The post-termination angle is that early exercise sidesteps the three-month problem entirely for the shares that have been early-exercised — they are already exercised, the cash is already out the door, and the holder owns the shares (subject to the company’s repurchase right on the unvested portion). On termination, the company can repurchase the unvested shares at cost; the holder retains the vested shares without any further exercise event. The expiration cliff goes away because there is nothing left to expire.
The trade-off is that the holder fronts the exercise price in cash, on shares that may never be worth more than that, and bears the risk that the company never reaches a liquidity event. Early exercise makes sense when the exercise price is low (early stage, low 409A value), the holder has confidence in the company, and the holder has the cash to spare. It makes less sense when the exercise price is meaningful or the holder is not in a position to take the risk.
The cash mismatch is the underrated problem
Even setting AMT aside, the post-termination window forces the holder to confront a cash question they may never have stress-tested. Exercising a block of vested options requires actual money — the exercise price multiplied by the number of shares. For an early hire with a low exercise price, that can be modest. For an engineer who joined a company that is now five years older and has watched its 409A rise, the exercise price on the later-vested tranches can still be low, but the share count can be large, and the total exercise cost can be six figures.
Three months is rarely long enough to assemble that capital under stress. The holder may be between jobs. They may not qualify for a margin loan against illiquid private shares. The company is generally not in a position to lend the exercise price (and any such loan raises securities-law and accounting issues of its own). The result is that some meaningful percentage of the options simply expire — vested, in the money, and lost — because the holder could not write the check fast enough.
This is the reality the headline equity number does not capture. A four-year option grant for a million shares at a ten-cent exercise price has a notional exercise cost of one hundred thousand dollars to capture all of it. If the 409A by the time the holder leaves has risen to two dollars, the AMT preference item on a full exercise is roughly one million nine hundred thousand dollars of spread, which can produce an AMT liability in the six figures, on top of the hundred thousand of cash for the exercise itself. None of that is bad faith on the company’s side. It is the structural consequence of stacking a tax cliff and a cash cliff inside the same ninety-day window.
What to negotiate, and what to plan for
Holders who treat the post-termination window as a fixed feature of the universe often miss the things that are genuinely negotiable. First, the window itself for senior hires. An extension from three months to twelve, or to longer, is a real benefit and is sometimes available on negotiation, especially for executives where the board is already used to bespoke equity treatment (the parent post and the compliant first hire piece walk through the documents this lives in). Holders should know that the extension typically converts the ISO to a nonqualified option for any exercise after the statutory three-month period, and price that conversion into the negotiation rather than be surprised by it.
Second, early-exercise rights. If the company permits early exercise, the offer letter or award agreement should say so plainly, and the holder should consider whether the 83(b) election strategy fits their cash position and risk tolerance. Early exercise can collapse the entire post-termination window problem to a non-issue for the shares that are early-exercised.
Third, acceleration triggers. Single-trigger acceleration on a change of control, or double-trigger acceleration on termination without cause within a window around a change of control, can convert unvested options to vested options at exactly the moment the post-termination window also becomes relevant. The mechanics of single-trigger and double-trigger acceleration are covered separately; the relevant point here is that any acceleration that pulls equity into vested status also subjects it to the same post-termination exercise window unless the agreement provides otherwise.
Fourth, the cause definition that controls which window applies in the first place. The companion piece on the cause definition covers the override mechanics. For the holder negotiating exercise-window protections, the cause definition is upstream of everything else: if cause is broadly defined and the for-cause window is immediate expiration, the negotiated extension to the ordinary window does nothing if the company can plausibly invoke cause.
The mental model worth carrying into the offer
The cleanest way to think about a vested startup option is as a contractual right that has two expiration dates: the contractual expiration of the option itself (typically ten years from grant), and the post-termination exercise window that begins on the holder’s last day of service (typically three months). Whichever comes first is what governs. The vesting schedule determines how much of the option is exercisable at any given time. The post-termination window determines how long “at any given time” lasts after the relationship ends.
Treating the option as “vested equity” without thinking through the exercise mechanics is the mistake that produces the worst post-termination outcomes. The discipline is to assume, at the moment of the offer, that one day the relationship will end on terms the holder did not choose, and to draft and plan accordingly. The parent post walks through the full set of drafting decisions that surround the equity grant. This piece is the one decision an engineer joining a startup should resolve before saying yes.
If you are evaluating a startup offer with a stock option component, or revisiting an old grant with a termination coming, feel free to reach out to my firm manager, Magda, at Magda@montague.law, or fill out our contact form. Mention you read this post.
— John
