Who Actually Benefits From Crypto Buybacks?

A crypto project can collect genuine fees, repurchase its token and still fail to create durable value for the people holding it. That possibility is not a technicality. It is the missing bridge in one of the market’s most seductive narratives: revenue is growing, buybacks are happening, therefore the token must be undervalued.

Consider Definitive’s EDGE and StonkFun’s STONK. Their reported economics could hardly look more different. EDGE has a comparatively measured revenue profile and published rules linking protocol activity to token purchases and staking rewards. STONK’s reported buybacks are vastly larger, propelled by a striking surge in launchpad-related activity. Both invite a familiar shortcut: divide a headline dollar figure by market capitalization, call it “yield,” and rank the winner.

I think that shortcut gets the order of operations wrong. Before asking whether a token is cheap, I want to know what its holder is actually entitled to, what was actually purchased, what supply entered the market, and how much could be sold without surrendering the theoretical benefit to price impact. These are different questions. A credible valuation has to answer them in sequence.

Follow the dollar, not the slogan

The first analytical mistake is treating a protocol’s income statement as though it were a shareholder distribution statement. Customer fees may be paid to a protocol, routed to service providers, retained in a treasury, or allocated to token operations. A tokenholder may have a contractual claim on none of it. Even a rule-based buyback program is not necessarily an enforceable right, and a repurchased token is not necessarily a retired token.

1. Customer feesWho paid, for what service, and was payment economically organic?
2. Protocol revenueWhat remains after rebates, incentives, operators and other costs?
3. Eligible allocationMandatory obligation, programmatic rule or discretionary policy?
4. Executed purchaseMatched swap signatures, actual fills, costs and timing.
5. Token destinationBurned, locked, held by treasury, staked or recycled?
6. Net holder effectSupply change, rights, market valuation and realistic execution costs.

An analytical pathway, not a claim that either project has independently verified every stage.

EDGE’s official materials describe a policy allocating 20% of protocol revenue to weekly token buybacks and 10% to qualified staking rewards, with staking eligibility and unstaking conditions. That is a more concrete mechanism than a vague promise to “support the token.” But the research record does not establish whether holders have legally enforceable rights to those allocations, who can amend the rules, or whether purchases are burned, retained or redistributed. STONK’s analytics classify substantial value as buybacks, and a public Solana explorer labels a related buyback wallet. Neither a provider classification nor a wallet label, by itself, proves the final economic destination of the tokens. [1, 2]

The comparison is not a finding that the two tokens have equivalent rights. It is a warning that their rights have not been established to a comparable standard. Verified contract permissions, governance proposals, treasury signers and effective dates belong ahead of any valuation multiple.

The numbers are real as reported—and incomplete as economics

DefiLlama’s published analytics, as captured in the October 21, 2026 research record, report the following figures for August, September and October-to-date. They are provider observations, not a reconstructed ledger of customer invoices or swap executions. The precise requested audit window—July 24 through October 21, 90 inclusive UTC dates—starts eight days before this table.

October figures are month-to-date, not a full month. Totals and ratios are calculated from rounded provider series; they are not exact 90-day sums. “Holder revenue” is an analytics category, not a verified cash yield or universal tokenholder entitlement. [3]

STONK’s September revenue was about 21.7 times its August level. That is the most dramatic fact in the comparison—and one of the least safe to annualize. Launchpad fee income may be economically legitimate while remaining highly sensitive to market attention, token-launch volumes and incentives. A single exceptional month cannot tell me how many customers returned, whether fees were concentrated among a few participants, or what would happen in a prolonged quiet market.

EDGE has a different accounting problem. The research record allocates roughly $132,000 each to buybacks and staking rewards across the reported August–October period. Yet the cited official policy specifies 20% of revenue for buybacks and 10% for staking rewards. Those are not matching splits. Perhaps the provider uses different categories, the policy changed, or effective dates differ. None of those explanations has been established. Until historical rules and transfers reconcile, the apparent precision of the holder-revenue figure should not be mistaken for a completed audit.

The economically relevant unit is not a dollar of reported buybacks. It is the verified, durable benefit per token after new supply and execution costs.

EDGE’s unlock problem is bigger than a ratio

Third-party vesting information in the research record described a roughly 12.97 million EDGE unlock scheduled for November 2, 2026. Assuming that same amount for August, September and October produces an illustrative 38.9 million tokens, worth roughly $3.83 million at the October 21 reference price near $0.0985. Against approximately $132,000 of reported buybacks, that is about 29 times as much potential unlock value. [4]

That comparison is informative, but it is not net selling pressure. A token becoming eligible to trade does not mean it was transferred to an exchange, sold, or even included in the same circulating-supply definition used by a market-data provider. The three historical unlock quantities were extrapolated from a later schedule, not verified against historical vesting events. An honest analysis cannot quietly promote the 29× ratio into an observed economic loss.

Even the starting float is contested. The October 21 CoinMarketCap snapshot estimated approximately 273.14 million circulating EDGE, while a separate tokenomics provider reported about 305.36 million. That is a material disagreement, not a rounding error. It affects any market-cap denominator, dilution rate and per-token interpretation. I would not average the two: the right response is to reconcile their definitions against supply contracts, vesting accounts and dated wallet states. [4, 5]

The stronger EDGE thesis would be that sustained customer fees finance verifiable purchases and externally funded rewards faster than newly liquid supply erodes the benefit. The opposing thesis is that recurring unlocks and discretionary token management overwhelm modest buybacks. At present, neither side has the daily float ledger needed to win that argument.

STONK’s buyback story needs an ordinary month

STONK’s reported holder revenue of approximately $21.33 million over the August–October partial period is striking. Its October 21 market-cap estimate was approximately $107.56 million. On paper, the numerator looks large relative to the denominator. But the numerator is a provider classification rather than an independently matched list of purchases, and the denominator is a single-date estimate rather than the period-average market cap. Treating the ratio as an investor yield would compound both problems.

The relevant transaction audit would start with the publicly labeled StonkFun buyback address, then follow incoming funding, DEX swaps, execution prices, token receipts and subsequent outgoing transfers. It would check whether the same economic event appears more than once through routers or intermediary wallets. It would separate tokens burned from those retained for possible later spending. The identified wallet is a lead, not a substitute for that ledger. [2]

The supply side remains just as important. CoinMarketCap estimated about 787.6 million STONK circulating against a one-billion maximum on October 21. The approximately 212.4 million difference is not a verified insider-unlock calendar. There is no basis here to assume those tokens will all be sold—or that they cannot affect future float. [5]

What would make STONK more convincing? A daily fee series showing paying-user retention, a breakdown of launch-related versus recurring usage, and executed buybacks that continue during low-activity periods. What would weaken it? Revenue dominated by a handful of customers, incentives that effectively subsidize the fees, or repurchased tokens returning to circulation. The point is not to dismiss launchpad revenue. It is to value it at a durability level the evidence can support.

What would the buybacks be worth?

A disciplined comparison needs a valuation denominator and a supply bridge, but it also needs to resist inventing precision. The October 21 CoinMarketCap snapshots estimated market caps of approximately $26.90 million for EDGE and $107.56 million for STONK. Their respective reported fully diluted valuations were approximately $98.48 million and $136.57 million. Those are historical provider estimates, not chain-reconciled valuations. [5]

Neither asset currently has the verified numerator and daily market-cap series needed to compute that metric. The reported holder-revenue ratios above should not be substituted for it. Nor would a high intensity automatically mean cheapness: a token with no enforceable cash-flow claim and a highly discretionary buyback program deserves a different valuation treatment from one with durable rights and transparent governance.

Liquidity is where accounting meets the market

One third-party study cited in the research record sampled STONK centralized-exchange books beginning October 2 and reported median combined resting depth of about $14,900 within 2% of midprice, across five exchanges, alongside a median quoted spread near 36 basis points. The archive was not independently reproduced. More importantly, combined bid-and-ask depth is not the amount a seller can hit. A sell-capacity calculation needs the bid side only, by venue and timestamp. [6]

The original comparison to hypothetical $25,000 and $100,000 sales can illustrate how small the quoted depth is, but it cannot predict slippage. It would be wrong to divide an order size by combined depth and call the result executable capacity. Realistic analysis must integrate bid ladders, fees, venue access, cancellation risk and the contemporaneous state of DEX pools. EDGE lacks an equivalent validated historical book in the available record, so neither token can claim a comparable, tested exit-cost advantage.

That missing symmetry matters. Reported trading volume is a measure of past activity, not a promise that an incremental seller can transact at the displayed midprice. Buybacks may occur over many days; an investor may need liquidity at a particular moment. Those are fundamentally different execution problems.

Three futures, none of them a forecast

Rather than extrapolate September or assign invented probabilities, I would frame the economics as conditions that can be tested. These scenarios are deliberately qualitative. They describe what must be true, not where either token’s price will go.

There is also a market-wide alternative explanation that no token-specific ledger can settle alone. A token may rally despite dilution because risk appetite improves; it may fall despite genuine buybacks because sector valuations compress. Establishing a buyback price effect would require an event-time analysis against appropriate market and sector controls, with attention to confounding announcements, liquidity changes and the timing of actual purchases. No such causal test is established here.

The missing 90 days

The audit period runs from July 24 through October 21, 2026, inclusive: 90 UTC dates per asset, or 180 asset-days. The available research defines that calendar but does not populate a verified daily series of fees, swap signatures, circulating supply, vesting-wallet flows or synchronized CEX/DEX bid-and-ask books. A missing cell must remain missing. It cannot be treated as a zero buyback, zero unlock or zero spread.

This is a completeness statement, not a claim that no fees, purchases or unlocks occurred. The reported August–October aggregate analytics remain separate observations; the July 24–31 interval is not covered by those aggregates.

For reproducibility, a serious published audit would attach the raw transaction signatures and wallet-attribution evidence; the historical token-supply definitions and block heights; provider query parameters and retrieval times; a source-versioned daily fee table; and simultaneous order-book timestamps with venue, instrument identity, bids, asks and fees. Where a provider disagrees with a chain-reconciled figure, the discrepancy should be preserved and explained rather than averaged away.

Without that record, a comparison can be useful journalism and a useful research agenda. It cannot yet be an independently reproducible measurement of net tokenholder accretion.

The observation that would change the verdict

For EDGE, the single most consequential result would be a daily reconciliation showing that verified, sustainably funded buybacks and externally funded distributions produce a durable per-token benefit after actual circulating releases, with the published policy and governance controls matched to the relevant dates. If that fails, a clean-looking revenue chart will not rescue the economic thesis.

For STONK, I would put the highest weight on a verified buyback-and-destination ledger paired with customer-retention evidence through an ordinary, lower-activity launchpad period. That would test both whether the reported purchases happened and whether the cash engine can keep funding them when attention fades. A bid-side exit study would then determine whether the benefit is accessible in the market rather than merely visible on a dashboard.

The larger lesson reaches beyond these two tokens. Crypto has become better at displaying protocol revenue than at explaining who owns its economics. A buyback can create demand without creating a claim. An unlock can increase sell eligibility without creating a sale. A liquid-looking volume figure can hide an empty bid book. And a low market cap can be a bargain, a warning, or simply a denominator attached to the wrong numerator.

My conclusion is not that EDGE or STONK fails the test. It is that neither has passed it. The right question is no longer whether buybacks outrun dilution in a headline comparison. It is how much sustainable, enforceable—or at least reliably executed—economic value reaches each token after supply changes and trading costs, and how much of that outcome the market has already priced in. Until those bridges are measured, I would rather preserve uncertainty than manufacture a winner.