Buyback Radar

GMX GMX

Payout off peakRank#23▲3Score41

Paid to holders this week: 32% of the best week in 90 days ($73k against $231k), while revenue +15% over 30 days. Also: holders revenue is paid out, not a buyback; the payout to holders is paused.

Derivatives, Arbitrum, Avalanche, MegaETH, 192 days of data

Market cap

$79.6M

price $7.29 · 7d −18.6% · 30d −8.8%

FDV $78.0M (1.0× market cap)

source defillama

Counted per year

$2.2M

holders revenue of the last 30 days, $354k, × 365 ÷ 30, × type factor 0.5 (payouts count 50%)

Distributionbuyback share 73%

Pays out 5.4% of its market cap a year, into the treasury, using 73¢ of every $1 it earns.

Not a buyback: counts for 50% in the multiple (type factor 0.5), so the denominator is $2.2M instead of $4.3M.

Multiple

37.0×

peer median 20.7× (Derivatives, 6 tokens), against 37.0× here

Run-rate

42.0×

last 7 days $73k × 52 × 0.5 (type factor) = $1.9M a year

last week's payout in line with the 30-day pace

What we checked

  1. Valuation

    Multiple 37.0× (type factor 0.5: payouts count 50%), 1.8× more expensive than its Derivatives peers (median 20.7×, 6 tokens). The group's largest earner Hyperliquid trades at 20.2×: 1.8× dearer per dollar returned to holders. On total revenue: 13.5×, in line with the revenue peer median.

  2. Distribution

    Distribution: Revenue is paid out to stakers or holders (often in another asset); no buy pressure on the token. Holders got $354k of $485k revenue in 30 days → buyback share 73%. DefiLlama: “Margin Fees: 27% of margin fees (after referral rebates) and borrowing fees distributed to GMX token stakers”

  3. Sustainability

    Holders were paid on 30 of 30 days (100%); largest single day 7% of the month's holders revenue; paid to holders this week: 32% of the best week in 90 days (more than half below peak). Revenue −23% over 7 days and +15% over 30 days, price −9% over 30 days.

What this does not tell you

The figures come from DefiLlama's adapter for each child protocol. Only a few tokens have an on-chain check (see How the buyback really works); for the rest, whether the tokens were really bought, burned or parked in a treasury wallet is not checked here. Unlocks count where DefiLlama or our research has a vesting schedule; team discretion and liquidity depth only once researched.

Flags

  • Payout >50% below peakthis week's holders revenue is 32% of its best week in the last 90 days ($231k)

How the buyback really works

Researched Nov 1, 2026 · 5 wallets: 3 high, 1 medium, 1 low confidence

  1. The moneyPaid from ongoing revenue
  2. The buyerPartly automatic, partly up to the team
  3. The tokensBought tokens go to the treasury (not burned, not paid out)

The project says 27% of its fees goes to buybacks, in line with DefiLlama (26.6% of gmx-evm's fees). Our research found the payout to holders is paused: the buyback continues, but the bought tokens go to the treasury; see the highlighted red flag.

On-chain: not measurable with our sources (GMX moved into the v2 FeeHandlers (2 transfers, about $4k in 30 days) does not capture the buybacks; the BuybackFees events of the EventEmitter need EventLogData decoding, which this verifier does not do yet).

Method: not measurable: GMX moved into the v2 FeeHandlers (2 transfers, about $4k in 30 days) does not capture the buybacks; the BuybackFees events of the EventEmitter need EventLogData decoding, which this verifier does not do yet. Checked November 12, 2026.

Red flags 6

  • Payout paused, buyback to treasury: Distribution to stakers is suspended (docs: 'GMX staking rewards are suspended', 'Bought-back GMX accumulates in the Treasury and will be distributed to stakers when GMX reaches $90'). There is no timeline, and the tokens are under multisig control.
  • There is no burn. Pending proposals (BBD&B 2026, the CEX supply overhang plan) could change the mechanism again.
  • The second part of the route is discretionary: withdrawFees is onlyFeeKeeper, after which it goes to the FEE_RECEIVER multisig. Only the buyback swap itself is permissionless.
  • The 'gmx' slug also covers GMX-Solana; the 'GT Buyback Season 2' in fees/gmx-sol.ts concerns GT, not GMX. Do not count it.
  • The docs alternately call the 27% 'of fees' and 'of protocol fees'. DefiLlama's calculation applies 27% of all fees (Revenue = 37%).
  • DefiLlama's calculation still labels the 27% as 'distributed to GMX token stakers', ignoring the suspension.
Buyback wallets 5
Sources 9

Open the buyback wallets in their explorer and look for regular purchases of the token.

Check it yourself (technical)

Run eth_getLogs on the EventEmitter (Arbitrum 0xC8ee91A5...; Avalanche 0xDb17B211c34240B014ab6d61d4A31FA0C0e20c26, which is also in the adapter), with topic1 = keccak256('BuybackFees') (EventLog1, eventNameHash). Decode buybackAmount, batchSize, feeToken and buybackToken, and filter on buybackToken = GMX (0xfc5A1A6EB076a2C7aD06eD22C90d7E710E35ad0a). Fetch 30 days in chunks, or via Etherscan v2 chainid=42161/43114. Alternative: GMX Transfers from the FeeHandler to FEE_RECEIVER (DataStore 0xFD70de6b91282D8017aA4E741e9Ae325CAb992d8, key FEE_RECEIVER). According to the researcher this is 0x028220b51343e065d3884780f9a5bd853278365e; the verifier confirmed a proxy (Safe-like) at that address, but the address appears in no published source.

Holders revenue / revenue (30d)
$354k / $485k
last 30 closed days
To holders
Pays out 5.4% of its market cap a year, into the treasury
FDV / dilution
$78.0M / 1.0×
FDV ÷ market cap
Price 7d / 30d
−18.6% / −8.8%
Revenue growth 7d / 30d
−23% / +15%
versus the period before
Divergence
0.23
revenue growth minus price growth, 30d
DEX liquidity
$593k
DexScreener, representative pools
24h volume
$3.5M
all markets, CoinGecko
Unlocks 30d
—
DefiLlama's schedule does not cover the next 30 days while tokens are still locked
Revenue multiple
13.5×
market cap ÷ revenue per year
Fee multiple
4.9×
market cap ÷ fees per year
Distance to peak
32%
this week ÷ best week of 90 days
Data since
May 4
first day with holders revenue

Holders revenue and price

Holders revenue per day7-day averagePrice (right axis)▲ ▼ signal days (history below)

Multiple over time

today 37.0× (scored) · peer median 20.7×

Market cap at close ÷ (holders revenue of the previous 30 days × 365 ÷ 30), with holders revenue capped at revenue and × the type factor: the same effective multiple as the peer median. Values above 100× are drawn at the top. Bands as on the meter; dashed line = peer median today.

Signal history

Signal history of this token: date, type, price at the signal, price after 7 and 30 days, change since the signal and the outcome
DateSignalPrice thenAfter 7dAfter 30dSinceOutcome
November 11, 2026New in top 20closed Nov 11: now ranked #25$7.58——−4%30-day check: day 1
October 31, 2026New in top 20voided Nov 1: it opened on wrong data (revenue of another token was counted)$8.21−3%—−11%not counted
October 31, 2026Multiple below medianvoided Nov 1: it opened on wrong data (revenue of another token was counted)$8.21−3%—−11%not counted
October 31, 2026New protocol with revenuevoided Oct 31: it opened on wrong data (revenue of another token was counted)$8.21−3%—−11%not counted

Versus peers

Derivatives · 6 tokens · median multiple 20.7× · Hyperliquid is valued 212.8× higher. Per dollar that counts for holders, GMX is 1.8× dearer.

Main-ranking peers only. Multiples count holders revenue capped at revenue, × the type factor (buyback 1 · buyback + burn 1 · burn 0.75 · mixed 0.75 · distribution 0.5 · unclear 0.5 · ve-distribution 0.25).

This token against its main-ranking peers: multiple, revenue multiple, buyback share, holders revenue, market cap and revenue growth
TokenDistributionMultipleRevenue multipleBuyback shareTo holders 30dMarket capRev. growth 30d
LighterBuyback14.6×11.8×81%$4.4M$775.1M+14%
edgeXBuyback16.0×7.0×44%$622k$120.8M−43%
HyperliquidleaderBuyback + burn20.2×20.2×100%$68.8M$16.94B−3%
ApeX ProtocolMixed21.2×15.9×100%$145k$28.2M−21%
GMXthis tokenDistribution37.0×13.5×73%$354k$79.6M+15%
dYdXDistribution51.6×25.8×100%$240k$75.2M−15%
Score build-up41 of 100

Total score = 35% valuation + 25% momentum + 20% divergence + 20% quality − penalties, clamped to 0–100. Every number here can be recomputed from the figures on this page.

41of 100#23▲3top 44%
Valuation 12.6Momentum 11.2Divergence 3.0Quality 14.0= 40.9
Valuation35%
36Valuation 36.1: HR half 28.4 (multiple 37.00×; 100 at ≤3×, 0 at ≥100×, log scale), revenue half 43.8 (revenue peer value 0.87×; 0 at ≤0.33×, 50 at 1×, 100 at ≥3×, log scale).Average of two halves. HR half: linear in log(multiple), 100 at ≤ 3×, 0 at ≥ 100×, 0 without effective HR. Revenue half: linear in log(revenue peer value), 0 at ≤ 0.33×, 50 at 1×, 100 at ≥ 3×; neutral 50 when the peer group falls back to the universe.
Momentum25%capped
4545 before the peak capMomentum 44.8: m = 0.5·g7 −0.267 + 0.3·(ln(day jump 1.12) clamped ±1.5 = 0.111) + 0.2·g30 0.136 = −0.073; 0 at ≤−0.7, 50 at 0, 100 at ≥1.1.m = 0.5·g7 + 0.3·ln(day jump, clamped ±1.5) + 0.2·g30; 0 at m ≤ −0.7, 50 at 0, 100 at ≥ 1.1. Neutral 50 when g7 or g30 is missing; a missing day jump counts as 0.
Divergence20%capped
15Divergence 15.2: d = g30 − p30 = 0.228; 0 at ≤0, 100 at ≥1.5.d = max(0, g30 − p30) when g30 > 0; 0 at d ≤ 0, 100 at d ≥ 1.5, linear.
Quality20%
70Quality 70.1: mean of days HR>0 100% → 100.0 (0 at ≤30%, 100 at ≥90%), largest day 7% of HR30 → 100.0 (0 at ≥60%, 100 at ≤15%), distance to peak 0.32 → 10.2 (0 at ≤0.25, 100 at ≥0.9).Average of: share of days with HR > 0 (0 at ≤ 30%, 100 at ≥ 90%), 1 − largest day ÷ HR30 (0 at ≥ 60%, 100 at ≤ 15%) and distance to peak (0 at ≤ 0.25, 100 at ≥ 0.9).

HR7 is below 50% of its 90-day peak: momentum and divergence are capped at 50. This is not a penalty; the distance already counts in quality.

Penaltiesnone

No penalty applies — 10 rules checked, 3 wait for research data

Where the numbers come from
Revenue source
2 active children
DefiLlama adapter per child
Price and market cap
defillama / defillama
latest fetch
Scored on
November 11, 2026
as-of date of the last compute run
On-chain verified
not measurable
checked Nov 12
Protocols behind this token2 at DefiLlama

Per protocol: DefiLlama's methodology text, the distribution type we read from it and the sentence that decided it. The token takes the type of the protocol that returns the most to holders.

  • GMX EVM
    Derivatives · Arbitrum, Avalanche, MegaETH
    type distribution · source label
    “Margin Fees: 27% of margin fees (after referral rebates) and borrowing fees distributed to GMX token stakers”

    27% of position fees (after referral rebates) and of borrowing, liquidation and swap fees goes to GMX token stakers

    Margin Fees
    27% of margin fees (after referral rebates) and borrowing fees distributed to GMX token stakers
    Token Swap Fees
    27% of swap fees distributed to GMX token stakers
    Liquidation Fees
    27% of liquidation fees distributed to GMX token stakers
    adapter on GitHub ↗
  • GMX V1 EVM
    Derivatives · Arbitrum, Avalanche
    type distribution · source label
    “Mint/Redeem Fees: 30% of revenue from mint/redeem fees to GMX stakers”

    30% of all collected fees goes to GMX stakers

    Margin Fees
    30% of revenue from margin fees to GMX stakers
    Token Swap Fees
    30% of revenue from tokens swap fees to GMX stakers
    Mint/Redeem Fees
    30% of revenue from mint/redeem fees to GMX stakers
    adapter on GitHub ↗