The first thing I check on a new token is not the unlock schedule. Everybody stares at the unlock schedule, and the market has mostly priced it in by the time you are looking, because it is public and dated and easy to model. What almost nobody checks is whether the team can pay salaries next quarter without dumping the token on you. That number is knowable, it is on-chain, and it tells you more about the next twelve months of price action than any vesting cliff.
Here is the mental model that saves you. A treasury denominated 95 percent in the project's own token is not a reserve. It is a promise to sell. The team has payroll, audits, market makers, infra bills, all denominated in dollars, and if the only thing they hold is the thing they issued, every one of those bills becomes a sell order eventually. So the question is never how big is the treasury. The question is how much of it is actually spendable, and how fast are they spending it.
Finding the wallets
You cannot analyze what you cannot locate, and projects are not always eager to make treasury wallets obvious. Start with the ones they publish. Most serious projects list a treasury or foundation address somewhere in their docs, their governance forum, or a transparency page. Take those as a starting point, not the full picture, because published wallets are usually the ones they are comfortable with you watching.
From there I widen out. A few things that work in practice:
- Pull the token contract and look at the top holders. Foundation, treasury, and ecosystem-fund wallets almost always sit in the top twenty, and they are usually labeled or easy to infer from their transaction patterns.
- Trace the initial distribution. When the token launched, the mint or genesis allocation went somewhere. Follow those first hops and you find the DAO multisig, the team allocation, the ecosystem pool.
- Look for the multisig signatures. Treasury wallets are typically Gnosis Safe or an equivalent, and the signer set and threshold are visible. A wallet that moves funds only after collecting several signatures is almost certainly operational.
- Check the outflows. A wallet that regularly sends stablecoins to fresh addresses that then go to an exchange is paying vendors or converting to fund operations. That is your burn wallet.
Label as you go. By the time you are done you want a small map: this is the main treasury, this is the ecosystem grants pool, this is the wallet that actually pays for things. The last one is the one that matters most and the one projects mention least.
Computing runway the honest way
Once you have the wallets, split the holdings into two buckets and be strict about it. Bucket one is what I call hard reserves, meaning stablecoins, blue-chip assets like a major L1 token or wrapped BTC, and anything genuinely liquid that is not the project's own token. Bucket two is the native token itself.
Value bucket one at face value in dollars. Now value bucket two much more carefully, because this is where people fool themselves. Do not multiply the native holdings by the current price and call it a war chest. If the treasury holds a large multiple of the daily traded volume, they cannot actually realize that price. Selling into their own book would crater it. A rough discipline I use is to ask how many days of typical volume the native holdings represent, and if it is more than a handful, I treat most of that stack as roughly unsellable at anything near screen price. It is optionality, not runway.
Runway, then, is hard reserves divided by monthly burn. Burn you estimate from the outflow wallet. Watch a few months of stablecoin and blue-chip outflows that clearly go to operations, average them, and you have a monthly burn figure that is directionally right even if you cannot see the invoices. If hard reserves cover many years of that burn, the project is funded and the native stack is a bonus. If hard reserves cover a few months, the native stack stops being optionality and becomes the funding source, which means it is going to hit the market.
Diversified reserves versus continuous selling
The single most useful thing you can learn is whether operations are funded from a diversified pile that was set aside once, or from an ongoing conversion of native tokens into stablecoins. These two look identical on a balance-sheet snapshot and behave completely differently for you as a buyer.
The tell is in the flow, not the balance. Watch the treasury over time and ask where the operating stablecoins come from. If the stablecoin balance draws down steadily and never gets topped up from token sales, the team raised or set aside real money and is spending it. That is the healthy version, and it puts zero structural sell pressure on the token. If instead you see native tokens moving to a market maker or a DEX, getting swapped for stablecoins, and those stablecoins then flowing out to operations, you are watching a continuous native-token sell program funding the business. It never shows up on an unlock calendar because nothing is unlocking. The tokens were already liquid. They are just being sold into you, quietly, on a schedule set by payroll.
That second pattern is the countdown clock. And it compounds, because the lower the price goes, the more tokens they have to sell to cover the same dollar burn, which pushes price lower, which forces more selling. A project can look perfectly solvent on a snapshot while being structurally committed to selling its own token every single month just to keep the lights on.
A workflow you can actually run
Here is the sequence I go through before sizing any position in a project token:
- Locate and label the treasury wallets. Published addresses first, then top holders, initial distribution, and the operating multisig.
- Split holdings into hard reserves and native token. Value hard reserves at face, and haircut the native stack hard against traded volume.
- Estimate monthly burn from operational outflows over the last several months.
- Compute runway as hard reserves over burn. Years is comforting. Months is a warning.
- Trace the source of operating stablecoins. Drawing down a fixed pile is fine. A native-to-stable conversion pipeline feeding operations is structural sell pressure.
- Only then look at the unlock calendar, and add the two together. Scheduled unlocks plus the hidden operational selling is your real forward supply.
The failure mode I have watched people walk into again and again is treating a nine-figure treasury as a sign of strength when almost all of it was the native token, and the tiny stablecoin sliver was draining fast. The headline number was huge. The actual runway was a couple of quarters, and the only way to extend it was to sell the token that the same people were telling you to buy. None of that was hidden. It was all on-chain. It just required looking at the composition and the flow instead of the total.
You will not get a clean answer every time. Some treasuries are deliberately opaque, some route through custodians you cannot fully trace, and burn estimates are always a bit fuzzy. That is fine. You are not trying to build their books to the dollar. You are trying to answer one question before you buy, which is whether this treasury is funding the project or slowly becoming your exit liquidity. Usually a couple of hours with a block explorer gets you close enough to decide.