Equity · Token cap tables
The two cap tables of a crypto company, and where they disagree.
A company register and a token register, usually in different entities, describing overlapping but non-identical holders on incompatible terms.
On this page
In short
What is a dual cap table?
A dual cap table is the combination of a company cap table and a token cap table run by the same project. One records stockholders and option holders in an operating company; the other records token allocations, often issued by a different entity in a different jurisdiction, to a partly different set of people.
Why there are two in the first place
The second cap table is not an accident of bookkeeping. It usually reflects a deliberate structural choice made early and for several reasons at once, some regulatory, some tax, some governance. A conventional venture-backed company incorporates in a jurisdiction like Delaware because that is what institutional investors underwrite. A token, meanwhile, is generally launched with an aspiration towards being governed by its holders rather than by a board answerable to preferred stockholders. Those two aspirations do not fit inside one legal entity comfortably.
So the pattern that emerged is a split. The operating company — call it Labs — employs the engineers, owns the trademarks, signs the commercial contracts, and raises priced equity rounds or SAFEs. A separate entity — a foundation company, association, or similar vehicle, frequently established in a jurisdiction chosen for its foundation law — issues the token, holds the treasury, and stewards the network. Labs may be a contractor to that entity. The two have different boards or councils, different bank and wallet arrangements, and different books.
The consequence is that a project has two capitalisation records that were created for different reasons, governed by different law, and maintained by different people, describing the economic interests of a group that mostly overlaps. That is the dual cap table problem in one sentence.
The two registers, side by side
| Feature | Company cap table | Token cap table |
|---|---|---|
| Issued by | The operating company, typically a corporation | A separate issuing entity, frequently a foundation or association |
| Legal basis for the recordDelaware requires a stock ledger recording every holder and transfer | Corporate law: charter, board consents, stock ledger, inspection rights | Contract and code. No statutory register, no inspection right |
| Who is on it | Founders, employees with options, angels, funds, sometimes advisors | Contributors, investors who held a token instrument, treasury, ecosystem, airdrop recipients |
| Time mechanic | Vesting with a cliff, and forfeiture of unvested shares on departure | Unlocking on a schedule that commonly continues after departure |
| What a holder is owed | A residual claim on the company, ranked behind a liquidation preference | Whatever the protocol confers — governance, fee access, staking rights, or nothing |
| What an exit looks like | Acquisition or listing of the company, distributed through a waterfall | Market liquidity in the token, with no waterfall and no ranking |
| Who maintains it | Company counsel and finance, usually in cap table software | Whoever the project assigned it to, usually in a spreadsheet |
Describes the common two-entity pattern. Structures vary widely and the right one is a legal question specific to a project, its jurisdictions, and its facts.
The same person, twice, on different terms
The clearest way to feel the problem is to trace individuals across both registers. Consider four people at a token project three years in, and ask a simple question of each: what do they own, and what happens if they leave tomorrow?
| Who | On the company cap table | On the token cap table | If they leave tomorrow |
|---|---|---|---|
| Founding engineer | Common stock, fully vested after four years | A contributor token allocation on a separate four-year unlock that started at token launch, two years later | Keeps the stock. Keeps unlocking tokens for two more years unless the grant says otherwise — and many do not |
| Seed fund from the pre-token round | Preferred stock with a liquidation preference and pro-rata rights | A token allocation only if they signed a token warrant or side letter at the time | Nothing changes; both positions are already fixed |
| Series A fund that invested after launch | Preferred stock, later series, senior in the waterfall | Possibly nothing, if the token was already issued and the entity that issued it is not the one they invested in | Nothing changes, but their equity is a claim on a company that may not capture protocol value |
| Early user who received an airdrop | Not present at all | A meaningful balance, freely transferable, with voting weight in governance | Not applicable — they were never an employee, and can vote against the company |
Where the two tables actively conflict
These are not edge cases. Each of the following is a live disagreement that has to be resolved somewhere, and resolving it in a document is far cheaper than resolving it in a dispute.
| Conflict | How it shows up | What it turns into if unresolved |
|---|---|---|
| Value migration | Protocol fees accrue to a treasury the company does not control, while the company bills for development services | Equity holders funding an entity whose upside lands somewhere they have no claim on |
| Instrument coverage | Some investors hold token warrants or side letters and some do not, depending on when they invested and what they negotiated | A two-tier investor base, and a most-favoured-nation argument at the next round |
| Pro-rata mismatch | Equity pro-rata rights are computed on shares; token entitlements were fixed as a percentage of supply at a different moment | An investor whose equity percentage and token percentage diverge, with no contractual mechanism to true up |
| Two definitions of fully diluted | Company FD includes the option pool and convertibles; token FD includes unissued supply and future emissions | Ownership percentages quoted in a data room that cannot be reconciled to each other |
| Governance divergence | The board controls the company; token holders control protocol parameters and treasury spend | A board decision the token holders can veto, or a governance vote the board is contractually unable to implement |
| Jurisdiction and tax | Equity income arises in the employment jurisdiction; token delivery may come from a different entity in a different country | Withholding and reporting obligations that nobody owns, discovered during diligence |
| Exit asymmetry | A buyer can acquire the company but cannot buy a token held by thousands of independent holders | An acquisition that delivers the team and the trademarks but not the network |
The tension nobody can design away
Underneath all of that sits one irreducible trade-off, and it is worth stating plainly because most treatments of this topic tiptoe around it.
Equity investors want the token to be economically tied to the company they own. Every mechanism that ties it tighter — the company controlling the treasury, the company promising to deliver protocol upgrades, the company marketing token value to purchasers — is a mechanism that makes the token look more dependent on that company. And dependence on the managerial efforts of a promoter is precisely the axis on which securities analysis of a token turns.
The SEC, joined by the CFTC, addressed this directly in a joint interpretation issued on 17 March 2026 (Release Nos. 33-11412 and 34-105020). It describes a non-security crypto asset as becoming subject to an investment contract where an issuer makes representations or promises to undertake essential managerial efforts that lead purchasers reasonably to expect profits, and describes that asset as ceasing to be subject to one where purchasers can no longer reasonably view those representations as remaining attached to the asset — for example on completion or abandonment of the development the promises concerned. The same release set out a five-category taxonomy of crypto assets and addressed airdrops, protocol staking, and mining.
The practical consequence for the dual cap table is that the two registers are pulled apart by design. The company is encouraged to hold the token at arm’s length; the equity holders are compensated for that distance with a separate token instrument — see /equity/token-warrant for how those are usually structured — rather than with company control over token value. That is why the second register exists as a separate document, and why merging the two into a single ownership number is not just difficult but conceptually wrong.
Running both without losing track
A working discipline for two registers
Operational hygiene, not legal advice. What a specific project needs depends on its entities, jurisdictions, and auditors.
Name the entities and what each one issues
Write down, on one page, every entity in the group, what instruments it has issued, and who signs for it. A surprising number of projects cannot produce this page on demand, and every downstream problem starts there.
Map each person to both registers
For every contributor and investor, record whether they appear on the equity side, the token side, or both, and under which instrument. The people who appear on only one side are where the disputes come from.
Reconcile the two vesting clocks
Equity vesting and token unlocks start on different dates, run for different periods, and treat departure differently. Model both against the same calendar so the divergence is visible before someone resigns.
The token vesting calculator models the token side; the vesting calculator handles the equity side.
Keep the two definitions of fully diluted apart
Publish company fully-diluted share count and token fully-diluted supply as separate figures with separate definitions. Never combine them into a single ownership percentage — the units are claims on different entities.
Reconcile the token side against the chain on a fixed cadence
Compare modelled unlocks against actual contract releases, and treasury records against the addresses you believe you control. Treat any gap as an exception with an owner and a due date.
Review instrument coverage before each round
Check which investors hold a token instrument and which do not before you open a new round, because that is when the mismatch becomes a negotiation rather than a surprise.
Glide runs stablecoin treasury and multisig infrastructure, so the wallet side of the token register — who signs, what moved, and from which address — is territory we work in every day. The equity side is a different discipline with its own tooling; the point of this guide is that a project needs both to be legible at the same time, in the same review, against the same calendar.
Frequently asked questions
Why do crypto companies have two cap tables?
Do equity investors automatically get tokens?
Can you combine the two into one cap table?
What happens to token unlocks when someone leaves?
Which cap table matters more in an acquisition?
Who is responsible for keeping the token register?
Sources
External links open in a new tab.
- Application of the Federal Securities Laws to Certain Types of Crypto Assets and Certain Transactions Involving Crypto Assets (Rel. 33-11412; 34-105020, 17 March 2026) — U.S. Securities and Exchange Commission, joined by the CFTC
- Delaware General Corporation Law § 219 — stock ledger and inspection — Delaware Code
- H.R. 3633 — Digital Asset Market Clarity Act, 119th Congress — Congress.gov
- Introducing UNI — an example of a published token allocation — Uniswap Labs
Written by
Glide Research
Payments research
Glide Research maps payment rails, FX corridors, and banking access so travellers, freelancers, and treasury teams can move money without legacy wire tax.
- Published
Glide · Equity
The two cap tables of a crypto company, and where they disagree.
A company register and a token register, usually in different entities, describing overlapping but non-identical holders on incompatible terms.
- Currencies
- 80+
- Spend anywhere
- Visa card
- Regulated legs run by
- Licensed partners