Method
A right is worth what a holder can execute with it, less what execution costs. Four cases from 2026 put numbers on that discount. $951 moved $8.5M. $20,199 steered $9.4M. One quorum was set to one. The numbers change how rights should be stated.
Most descriptions of governance rights describe existence. The holder can recall an operator. The holder can claim a revenue share. The holder can block a treasury spend. The word “can” carries no price, no delay, and no threshold. It treats a right that costs nothing to exercise and a right that costs more than the harm it prevents as the same kind of thing.
They are not the same kind of thing, and 2026 produced enough live cases to say so with numbers rather than argument.
A right has two measurable components. The first is its effect: the treasury it can redirect, the operator it can remove, the value it can claim. The second is its cost: the voting weight that must be assembled, the time that must elapse, the capital that must be committed, the exposure incurred while holding it. Neither component alone is the useful number. The ratio between them is.
When the cost is small relative to the effect, the right is real, and the mechanism that produced it is doing work. When the cost approaches or exceeds the effect, the right exists in the contract and nowhere else. When the cost is zero for one party specifically, because that party already holds the controlling weight, the right belongs to that party as a matter of arithmetic, whatever the vote counts say.
Four cases from the past year, all verifiable from chain state, make each branch of that statement concrete.
Neutron’s proposal 9 was controlled for $20,199 and steered $9.4M of treasury. The cost of assembling the voting weight that determined the outcome was approximately 0.21% of the value it controlled. The interval between the vote closing and execution was effectively zero minutes. The defensive assumption that “the community will notice and act” had no time budget in which to operate.
The relevant computation for a rights determination is not the loss figure. That belongs to the incident record. It is the ratio: control cost as a percentage of controlled value. That number is a property of the mechanism. Quorum, threshold, execution delay, and the price of the token at the moment of measurement. It changes when any of those change. It is also computable in advance, which is what makes it useful. A mechanism whose control cost is a rounding error of its controlled value has no community at all, as a matter of arithmetic. It has a price list.
A recall right — the holder’s ability to remove an operator who is misusing a position — is the standard example of a right that can be real in form and absent in effect. Its cost is not the transaction fee. Its cost is the voting weight required, the notice period needed to assemble it, and the time during which the misuse continues unchecked while the weight is assembled. If that cost exceeds the harm the misuse causes in the same window, the right never fires. The mechanism’s designers have then built a defence that is only exercisable after the damage it was meant to prevent.
Two governance incidents from this period put sizes on this. An exploit in which roughly $951 of a token drained $8.5M from a position. A protocol failure costing its community roughly $20M. In both, the question for holders was not whether a recall mechanism existed on paper. It was how quickly the required weight could be assembled relative to how fast the value was leaving. That ratio, assembly time against loss rate, is the honest measure of a recall right. It is a property of the mechanism, not of the incident.
Astroport ended token-holder governance by configuration rather than deletion. The Assembly contract’s required quorum reads as 1. That is one hundred percent of voting power. Token emissions are zero. The staked lock unwinds through its normal release. Voting power only decays from here. No quorum can be met again at any price, because the required weight can never be assembled from the remaining supply.
This is a distinct branch of the arithmetic. A right whose cost is merely high is a right with a price. A right whose cost is unreachable is a vestige. The distinguishing test is whether the required weight can still exist in principle. quorum == 1 against a decaying supply answers that in one read. Where the same parameter is set below one, the question becomes how far below, and at what token price the remaining weight is purchasable. That folds case three back into case one.
Orca’s transfer proposal asked the community to ratify moving the treasury to an account that already held upgrade authority over the protocol’s core programs. That position was transferred mid-vote, before the vote concluded. For every holder, the cost of exercising governance on that proposal was the ordinary cost of acquiring and voting weight. For the operator it was zero. The control position the proposal was about to formalize already existed.
When one party’s cost of control is zero, the vote counts describe a ritual. The same structure appeared in a different register in Compound’s proposals 612 and 613. The FOR vote was carried by one concentrated position. Remove that single position from the total and a decisive minority remains. The arithmetic is indifferent to whether the position used its weight well. It states what the mechanism does. A right whose exercise depends on one position is held by that position.
If a right’s value is its effect discounted by its cost, then three things follow for any statement of governance rights.
First, cost is part of the right, and it belongs in the description. “The holder can recall an operator” and “the holder can recall an operator, at a cost of X% of controlled value, within Y hours, requiring Z of voting weight” are different claims about different mechanisms. Only the second is checkable. Only the second tells a holder what the right is worth.
Second, cost is computable in advance from state that is public. Quorum, threshold, execution delay, supply, and price are all readable. A determination that reports the cost of exercising each right is not predicting anything. It is doing arithmetic on published state and committing to the result. Re-run the same computation at a later block and it shows whether the right is appreciating or decaying. That is the difference between a snapshot and a monitor.
Third, the ratio is what a determination should grade, not the effect. Two protocols can both list “treasury control” among their rights. One has a control cost of 0.2% of controlled value. The other has a quorum that cannot be assembled from its remaining supply. Both statements are true. Only one of them is useful to anyone deciding what a token’s governance is worth.
Neutron: the controlling position’s cost and the treasury amount were computed from on-chain reads. The close-to-execution interval was read from proposal timestamps. The two incident figures, roughly $951 and roughly $20M, are reported amounts attributed to the incidents as publicly described. We include them for the ratio they illustrate. The ratios for Astroport, Orca, and Compound are read directly from contract state and vote tallies rather than inferred. Astroport’s quorum was read from the Assembly contract’s configuration. Orca’s account structure and upgrade-authority transfers were read on-chain. Compound’s tallies were read from the governor’s view functions. The arithmetic in each case is reproducible from those reads.
One caution, stated plainly. A single measurement is a snapshot. Control cost changes with price, with supply, with any parameter change. This essay’s figures are as of the blocks cited in the linked records. Our exercisability block records these quantities per token precisely so the snapshot can be re-taken, and so that a right described as exercisable at a cost of X can be shown later to have become exercisable at a cost of Y.
proposalVotes — the concentration counterfactual.Nothing in this essay expresses an opinion on the investment merits of any token or protocol, and nothing here is a recommendation. It describes publicly recorded proposals, votes, contract state, and treasury activity. Verify all information independently before acting on it.