Buyback & burn · measured

A buyback is a bid.
What is it bidding against?

Every launchpad and exchange in this cycle competes on the same sentence: we spend revenue buying our own token. The figure quoted is always a dollar amount spent — cumulative, gross, and impressive. It is real. It is also the numerator of a fraction whose denominator nobody prints: the supply that exists, does not trade yet, and is still coming.

NetBurn puts the two side by side for every token running a programme, using on-chain supply where the chain can be read and labelling every figure that comes from a vendor.

Loading measured data…

The finding

The same protocol, on the same day, is three different answers

"Years to buy back the overhang" is the natural way to size a buyback against the supply it faces. It is also almost meaningless as a single number, because it depends entirely on which revenue window you annualise — a choice nobody making the claim ever states. Each bar below is one token; the bar's width is the disagreement between annualising the last 30, 90 and 365 days.

Overhang divided by annualised buyback spend. Left dot = fastest window, right dot = slowest. Tokens whose protocol revenue is collapsing get slower as the window widens; tokens on a run get faster.
This is the trap, not a caveat. A metric that moves by this much on an unstated analyst choice is not measuring the protocol — it is measuring the window. The same failure took an earlier project in this family from a "32-minute half-life" to a real answer of three minutes. Wherever NetBurn prints one of these numbers it prints the window beside it.

Every programme in the set

What each protocol spends, and what it is spending against

Sorted by overhang. burn destroys the tokens; hold buys and keeps them, which is a real bid on the market and zero supply reduction; distribute hands them to stakers, which actively returns them to the float. Published tables routinely add these together. They are not the same instrument.

Loading…

The second question

Did any of it show up in the price?

The arithmetic above is about supply. The claim everyone actually cares about is about price: buybacks fuel demand. That is a causal claim about one unit at one moment, which is exactly what synthetic control is for — build a counterfactual token out of a weighted blend of tokens that never ran a buyback, fit it on the period before the programme, and read the gap that opens after.

The estimator runs live in your browser, and you choose the treatment date — then it sweeps every other date in the window as a control, so you can see whether a result at any one of them is worth anything. It was built expecting the method to cry wolf. It does not, which is what makes the answer it gives at the real policy dates worth reading.

Limits

What this does not measure

An unlock is not a sale. The overhang is supply that becomes able to trade, not supply that does trade. NetBurn deliberately does not claim that locked tokens hit the market on schedule; the coverage figure is a bound on what a buyback could absorb, not a prediction of what it must.

Policy shares are published commitments, not measured flows. Where a protocol says it routes 50% of revenue to buybacks, NetBurn uses 50% and links the source. It does not verify that every dollar arrived, and it says so at every point of use.

Vendor circulating supply is an estimate that gets restated. Those series are marked throughout and are used for levels only, never for daily flow. The evidence is on the method page.