On this page
A grant is a chain, not a line item
On the cap table, a stock option looks like one line: a name, a number of options, an exercise price, a vesting start date. That single line hides a chain of records that has to exist behind it, and the option is only as valid as the weakest link in that chain. There is a plan that had to be adopted. There is a board approval for the grant. There is a signed option agreement. There is a vesting schedule that determines how much is exercisable and when. And eventually there is an exercise that turns the option into an actual share, with its own authorization, ledger entry, and certificate.
This is the same pattern we have traced through the rest of the series. A SAFE is a promise about equity you have not issued yet, and we wrote about where it belongs in where SAFEs and convertible notes belong in your records. An option is a different kind of promise: a right to buy shares later, at a fixed price, if the holder stays and chooses to pay. Like the SAFE, it is not a share on the day it is granted, so it does not sit in the share register. But it is a real obligation of the company, and the records have to prove how it came to exist and what happens when it converts. When that proof is missing, options are one of the most common findings in diligence, precisely because the cap-table line looks so tidy.
The plan comes before the grant
Almost every option is granted under an equity incentive plan, and the plan is the first record that has to exist. The plan is the instrument that authorizes the company to grant options at all, reserves a pool of shares to back them, and sets the rules every grant inherits: who is eligible, how exercise prices are set, what happens on termination, and how many shares the pool holds.
The plan has to be properly adopted, which usually means approved by the board and, in most jurisdictions and for tax-qualified options in the US, ratified by the shareholders. A grant made before the plan is adopted, or under a plan that was never properly approved, rests on nothing. This is the same authorization problem we described for share issuances in who authorized this issuance: the action can happen operationally while the authority behind it is missing from the record. The plan document, and the resolutions adopting it, belong in the minute book alongside everything else, a point we made about the book's contents in what a corporate minute book is.
The common failure is a plan that exists as a template someone downloaded, filled in, and never formally adopted, or a pool size that lives in a cap-table tool but was never actually reserved by a board resolution. The grants that follow inherit that weakness. If the plan is not real, the grants under it are not real either.
Authorizing each grant
Adopting the plan authorizes the company to grant options. It does not authorize any particular grant. Each grant is its own decision that the board, or a committee the board has properly delegated to, has to approve. We wrote about the anatomy of a decision that survives review in how to write a board resolution, and an option grant is a textbook case for it.
A defensible grant record captures the optionholder, the number of options, the exercise price, the grant date, the vesting schedule, and the plan the grant is made under. The exercise price matters more than founders expect, because setting it below fair market value at the grant date creates tax problems for the recipient and a discrepancy a reviewer will find. In the US that means the exercise price is set against a contemporaneous 409A valuation, and the grant should be approved on or before the date it is priced. Backdating a grant to a date when the fair market value was lower is exactly the kind of thing diligence is built to catch, in the same way we described for resolutions generally: the date has to be true.
The signed option agreement is the instrument that ties it all together. A grant approved by the board but never documented in a signed agreement, or an agreement signed but never approved, is a half-formed record. Both halves have to exist, and both have to say the same thing.
Tracking the pool so you don't over-grant
The pool is where option records most often drift out of alignment, because it is a running balance that changes with every grant, forfeiture, and exercise, and nobody notices when it stops reconciling. The plan reserves a fixed number of shares. Every grant draws that number down. Every option that is forfeited when someone leaves before vesting returns to the pool, or does not, depending on how the plan is written. Every exercise moves options out of the pool and into issued shares.
The failure is over-granting: promising more options than the pool actually holds. It happens gradually, when grants are tracked in one place and the reserve in another, and the two are never reconciled against each other. A company can find at diligence that it has granted 1.2 million options against a pool of 1 million, which means some of those grants are not backed by authorized shares. Fixing it after the fact means either enlarging the pool, which requires board and shareholder approval and dilutes everyone, or unwinding grants people were promised. Neither is a conversation you want to have during a financing.
The pool only stays honest if the reserve, the grants, the forfeitures, and the exercises are all tracked in one place and reconcile to each other continuously. This is the option-pool version of the same drift we described for ownership generally in your cap table is not your corporate record: two records that were supposed to agree, quietly diverging.
Vesting is a record, not a calculation
Vesting determines how much of a grant the holder has actually earned and can exercise at any point in time. The standard schedule is time-based, often four years with a one-year cliff, but plans use milestones and other structures too. What matters for the record is that vesting is not a formula you rerun from memory when someone asks. It is a schedule anchored to a grant date, adjusted for real events: a termination that stops vesting, a leave that pauses it, an acceleration triggered by a change of control.
When vesting lives only as an assumed schedule and the events that alter it are tracked loosely, the record of how much is vested becomes an estimate rather than a fact. That surfaces at exactly the wrong moments: an employee leaves and there is a dispute about how many options were vested on their last day, or a buyer asks for a vested-versus-unvested breakdown and the numbers do not tie out. Ownership that lacks a defensible timeline is fragile, which is the argument we made in why ownership breaks without a timeline, and vesting is where that fragility shows up first for options.
What changes at exercise: options become shares
Exercise is the moment the option stops being a promise and becomes ownership, and it is where the two halves of the record have to connect. The holder pays the exercise price for some or all of their vested options, and those options convert into issued shares. Everything we have written about share issuances now applies, all at once.
The exercise has to be authorized and recorded as an issuance. The new shares have to be entered in the share register. A certificate, or its verified digital equivalent, has to be issued for them, and the options that were exercised have to be retired from the pool so they are not still counted as outstanding. We covered the issuance side in issuing share certificates properly, and the retirement-and-reissuance discipline that keeps the certificate chain intact in cancelling and reissuing share certificates. Exercise is where an option grant finally reaches the share register and the stock ledger we described in what goes in a stock ledger.
The failure here is an exercise that updates the cap table's option count but never produces a share issuance, a ledger entry, or a certificate. The holder believes they own shares, the cap table half-agrees, and the corporate record shows no issuance at all. For US employees who early-exercise unvested options, there is an added deadline that belongs in the record: the 83(b) election has to be filed within 30 days, and whether it was filed is a question diligence will ask.
The reconciliation test
The test for options is the same one that runs through the whole series, applied along the chain. For any option on the cap table, can you produce the plan that authorized it, the board approval and signed agreement for the grant, the vesting record as of any given date, and, if it has been exercised, the issuance, ledger entry, and certificate that resulted? If every grant traces cleanly from plan to share, the option section of diligence is answered before it is asked.
Where it breaks is predictable: a plan that was never formally adopted; grants approved by email with no resolution; exercise prices set without a contemporaneous valuation; a pool that does not reconcile to the sum of its grants; vesting that cannot be reconstructed for a departed employee; exercises that never became issued shares. Each of these is a small gap in a tidy-looking line, and together they are one of the most common reasons the equity section of a deal slows down, in the same family as the findings we catalogued in the records findings that delay closings. A diligence lawyer reads the option file the same methodical way they read the rest of the book, which we walked through in what a diligence lawyer actually reads in your minute book.
Keeping options defensible is not about tracking them more carefully in a spreadsheet. It is about keeping the plan, the grants, the pool, the vesting, and the resulting share issuances in one system where each links to the next and the whole chain reconciles as you go. That is what our cap table and equity tooling is built to do, sitting in the same record as the corporate records that authorize every grant.
The bottom line
A stock option is a chain that runs from an adopted plan, through an authorized grant and a signed agreement, along a vesting schedule, to an exercise that issues a real share. The cap-table line is only the visible end of it. Options break when part of that chain is missing: a plan nobody adopted, a grant nobody approved, a pool that no longer adds up, a vesting record that cannot be reconstructed, or an exercise that never reached the share register.
Keep the whole chain in the record, linked and reconciled at each step, and options stop being the surprise finding in diligence. Founders rarely lose deals because they granted options. They lose time because they could not prove, link by link, how those options came to exist and what they became.
