A token buyback, a burn, a reserve purchase and a distribution describe different changes. A buyback spends assets to acquire tokens. A burn destroys token units. A reserve holds assets under particular control. A distribution delivers assets to eligible recipients. One transaction chain can contain several of these steps without producing several independent dollar benefits.
Follow one economic chain
Illustrative example: a protocol spends $10 million to buy two million tokens, then burns the same tokens. The cash outflow is $10 million and the supply reduction is two million token units. Adding another $10 million for the burn would count the purchase twice. Valuing those tokens again at a later market price does not create a second cash expense.
The same issue appears when purchased tokens are distributed. Record the purchase cost and the recipient transfer as linked stages. They answer different questions: how much the protocol spent and what the recipient received. Do not sum both valuations into a single “holder benefit” total.
A burn need not involve a market purchase
A treasury can destroy tokens it already owns. That reduces the relevant token supply if the destruction is confirmed, but it does not establish a new buyback or new operating revenue. Likewise, a network can burn part of the transaction fee directly. The user's payment, the destruction of the fee token and a protocol buying its token in a market are distinct mechanisms.
A transfer to a labelled address also needs interpretation. A provably destroyed token, an inaccessible balance, a time lock and a reserve wallet should not share an unqualified “burned” label. Ask whether the assets can later move and which supply definition excludes them.
Uniswap illustrates why intermediate steps matter
The Uniswap protocol-fee documentation, reviewed on 6 October 2026, separates fee collection in TokenJar from release through a Releaser. Firepit exchanges eligible collected assets against a UNI burn. A balance accumulating in TokenJar is therefore not, by itself, proof that the DAO spent dollars purchasing UNI in the market.
Read the arrows as stages with their own evidence. A rule can explain what should happen; a transaction or compatible financial report establishes what happened during the measured interval.
Reserves have owners and future uses
A revenue-funded acquisition may add tokens to a reserve without destroying or distributing them. Check who controls that reserve, whether governance can change its purpose, and whether its assets back another liability. A reserve may support liquidity or future spending. Calling every acquired token “permanently removed” would conceal those possibilities.
The funding source is equally important. Current customer revenue, older accumulated income, token sales and borrowing are different origins. An active buyback funded by old reserves is not evidence that current operations cover it. On the other hand, spending last year's retained revenue need not fit beneath this month's revenue total.
Separate decisions from execution
Track four questions: was a mechanism proposed, approved, activated and executed? A successful vote can precede a deployment; a deployment can precede the first purchase. A historic pause needs a fresh status check before being described as current. If an API fails, the correct state is unavailable or stale, not “the program has paused.”
For a direct distribution, identify the eligible asset and recipient group. Holding a ticker may be insufficient: staking, locking, voting or activating an NFT can be required. The payout asset may also differ from the token that grants access.
Compare the separate fields in the revenue screener and inspect the Uniswap record. The accounting methodology explains why buybacks, supply changes and direct payouts remain separate instead of becoming a synthetic APY.