Should a Token Project Hold Its Treasury in Its Own Token?

Is holding treasury in your own token actually runway?

No. A treasury held in its own token is not runway, because fixed costs fall due in cash whatever the token is worth that month. Salaries, audits, cloud bills and exchange fees arrive on a schedule that has no relationship to price, and the only part of a treasury that can settle them on the day they land is the part already held in stables.

The gap between the two shows up in how teams count. A balance sheet that marks a large native allocation at spot produces a number that looks like years of cover, and that number is paper value rather than spendable liquidity. We separate the two when we look at a project's position, and we count runway only in what can be spent without a sale that depends on market conditions holding.

Timing is what makes the distinction expensive rather than academic. The day money is needed is usually the day the token is down, because the conditions that close funding rounds and slow revenue tend to compress the price at the same time. A native balance cannot be relied on to meet obligations through a drawdown, so the assumption fails at the point it is first tested, and the team discovers the shortfall while it is already trying to cover a payment.

That does not make a native holding worthless on the balance sheet. It makes it a different category of asset, one whose realisable value depends on who is bidding when the team wants out, and on the depth available at that moment across the venues where the token trades. In our view the conversion capacity of a treasury is a separate question from its size, and it is governed by the same liquidity conditions that decide how a launch trades. A team that has never tested how much of its own token it can move without marking the price down still has that measurement ahead of it, and the usual moment it gets taken is the moment the sale has to happen.

Why does concentrating treasury in one exchange or one stablecoin matter?

Concentration matters because the thing that kills a treasury is rarely a price move, it is a single venue or a single issuer going offline with the entire balance inside it. A treasury spread across several custodians survives one of them halting withdrawals, because the payroll run and the next audit invoice get paid out of the parts that still work. A treasury sitting in one account does not survive it at all, and the team discovers this on the day the notice goes up rather than in the planning for it.

The same logic runs through stablecoin exposure. Holding the full operating balance in one issuer means the depeg, the redemption gate or the blocklist action is not a haircut on part of the position, it is a freeze on everything that pays salaries. Splitting across issuers does not remove the risk of any one of them wobbling, it removes the case where one wobble stops the company. We treat that difference as the whole point: diversification here is not about yield or optimisation, it is about whether operations continue during a counterparty event.

Post-2022, our view is that concentration of this kind is an unforced error. The failure modes are documented, the remedies cost almost nothing to implement, and a team that has lived through one exchange collapse already knows what the recovery process looks like. There is no clever argument left for keeping everything in one place, only inertia from whichever account was opened first.

Treasury concentration also tends to travel with market structure concentration, since the exchange holding the balance is often the venue where most of the token trades. That pairing is worth separating deliberately, and the operational cost of trading across several venues is covered in our piece on what fragmented liquidity actually costs. Where the custody sits and where the book sits are two decisions, and collapsing them into one relationship means a single counterparty failure takes both the reserves and the market at the same time.

How should a project diversify out of its native token without spooking the market?

Gradually, through a standing execution program rather than a single announced sale. A sudden disclosed move carries two costs at once: the price impact of the block itself, and the signal that the team has decided its own token is worth less than what it is being swapped for. A program that runs continuously removes the second cost almost entirely, because there is no event for the market to read.

The execution side matters as much as the pacing. Our position is that treasury sales should run only in coordinated windows, with the schedule agreed in advance and the fills structured, through TWAP or VWAP execution or negotiated OTC blocks. What we argue against is treasury flow going out as independent selling through the market-making desk. The desk is quoting both sides of the book; if it is also working down a treasury position on its own clock, the project has no way to tell inventory management from diversification, and neither does anyone watching the tape. Those are different mandates and they belong in different windows. We cover the mechanics of sizing and settling negotiated blocks in more detail in our guide to OTC structures and execution for token projects.

Coordination also means the diversification schedule sits alongside everything else that puts supply into the market in a given month. Unlock tranches, ecosystem grants and liquidity provisioning all draw on the same depth. A treasury window dropped on top of a cliff is the version of this that causes the damage teams are worried about, and it is avoidable with a calendar rather than with discretion.

What that leaves is a reporting question. Once sales are scheduled and structured, the project knows ahead of time roughly when each tranche converts, which lets disclosure follow the program instead of chasing it: a periodic treasury update that states what was executed and over what window, published on a cadence the community has already seen twice before the first tranche of any size runs through it.

What should a published treasury policy actually look like?

A published treasury policy is a short document that assigns each part of the treasury to a named bucket as a fixed percentage, and commits the team to reporting actual balances against those percentages on a stated cadence. The buckets themselves are usually few and plainly named: buyback-and-distribute, liquidity provisioning, strategic reserves. The percentages attached to them are the project's own call, driven by its stage, its revenue and its obligations, and we take no view on what any particular project's numbers ought to be.

As a purely illustrative example of the format rather than a recommendation, a policy might read 40% buyback-and-distribute, 30% liquidity provisioning, 30% strategic reserves. The point of writing it that way is arithmetic rather than strategy. Fixed percentages create a number that a later report either matches or does not, and in our experience that is the only part of treasury communication that holders can check without trusting the team's narrative.

What fixed splits buy, in our assessment, is that deviation becomes visible on a schedule instead of being inferred from on-chain flows weeks later. A buyback executed under a standing published policy reads as the policy operating. The same buyback executed without one invites the reading that the team is defending a price. Reporting against the policy means per-bucket balances, what moved during the period, and an explanation wherever the actual percentage sits away from the stated one. It is the same standard teams already apply to counterparties, which is why the questions worth asking about what a market maker reports back carry over almost unchanged to treasury disclosure. Setting the splits with a team, running execution inside them and producing that reporting is what we do under treasury management.

Amendments belong inside the policy rather than outside it. Raises, listings and changes in protocol revenue all move the sensible shape of a treasury, and a policy that cannot be revised gets ignored instead of updated. The mechanism that keeps the accountability intact is sequencing: the revised percentages and their effective date get published before the first transaction executed under them, so the following period's report is measured against the version that was actually in force when the money moved.

Not investment advice.

Should a raise or treasury balance be framed as covering payroll?

No. A raise buys growth or licences, and we treat payroll as something revenue has to carry, not something a funding event underwrites. The distinction sounds semantic until the first drawdown, at which point a team that sized its headcount against raise proceeds is making staffing decisions under market pressure rather than operating ones.

The mechanism is simple enough. Payroll is a recurring obligation that renews every month regardless of what the treasury is marked at. Raise proceeds are a stock, and in this sector a sizeable part of that stock is often denominated in assets whose value moves. Growth spend and licence applications can be sequenced, deferred or cancelled without severance conversations; salaries cannot. When the two are funded from the same pot and described with the same word, the flexible spending absorbs none of the shock and the inflexible spending takes all of it.

What we look for instead is a fixed cost base capped as a share of revenue, so that hiring responds to what the business actually earns. That cap is what keeps the headcount plan independent of the treasury mark. It also changes what the next round is for: a project whose payroll already clears against revenue raises to enter a new market or fund a licence, while a project whose payroll clears only against the last raise is raising to repeat the same exercise, on worse terms each time.

The framing leaks into documents long before it shows up in a cash position. Board decks that express a round in months of runway, investor updates that lead with burn rather than with what the money bought, hiring plans written the week after a close. In our experience the correction starts in the finance function, where raise proceeds get mapped to named line items such as listings, legal and the ongoing cost of running a token through its full lifecycle, and payroll sits against the revenue line where a shortfall hits the hiring plan first.

Is treasury management the same thing as a yield product?

No. Treasury management is a mandate covering policy, execution, and runway, and we hold that it is separate from any yield product that happens to sit alongside it. The two get merged because they touch the same assets and often the same people, but the questions they answer are different. A mandate asks what the organisation can fund and for how long under rules it has committed to. A yield product asks what a pool of capital can earn.

The cost of merging them shows up in how decisions get justified. Once return becomes a measure of whether the treasury function is doing well, every structural choice starts competing with a yield number, and the conversation stops being about stewardship. We see this most clearly when a deposit, a lending arrangement, or a staking position is defended on the basis of what it pays rather than on whether it leaves the organisation able to meet obligations on schedule. That is a return-seeking argument wearing a treasury label.

Keeping the mandate distinct does not mean a project earns nothing anywhere. It means the policy, the execution process, and the runway calculation are governed on their own terms, and any yield activity is a separate decision measured separately, with its own counterparty and its own exit. Arrangements that put treasury assets in a counterparty's hands belong in that second category whatever they are called, which is part of why we think the difference between a loan structure and a retainer deserves to be read as a treasury question as much as a liquidity one.

In practice the separation is enforced by who signs off and against what. When the same committee approves both under one heading, the yield line tends to absorb the policy line within a few quarters, and the first sign is usually a runway figure that can no longer be stated without naming the return assumption behind it.

Not investment advice.