Meteora has already won something real: it is the liquidity layer Solana’s launchpads build on, and the venue where LPs earn the most per dollar of memecoin volume. Across H1 2026 it facilitated $32.1B of volume and $164.3M of fees and kept $18.2M — an 11.1% take rate, up from 9.6% in H2 2025 — on a trailing year of roughly $48M of revenue against ~$443M of fees. If the protocol wins, MET holders win too; the question is through how narrow a pipe. About 89 cents of every fee dollar never reach Meteora at all: LPs keep ~63.8¢ and launchpads ~25.1¢. Of the ~11¢ that does, only ~0.7–1.4¢ reaches token holders, through three routes that are all policy rather than right — a 20% share of DLMM protocol fees paid to stakers and referrers in USDC since 21 July, discretionary buybacks that have been paused since March and warehoused rather than burned, and an implicit claim on a treasury whose disclosure was retired. Against that, 7.22M MET a month — about $28M a year at spot, five to nine times what holders receive — vests to team wallets and the reserve through October 2031. The result is a token that looks cheap on one lens and expensive on the other: ~3.7× trailing revenue, and ~30–60× the cash that actually arrives.
Born in FTX's wreckage
Meteora is not a 2025 startup. It is the second life of Mercurial Finance, a 2021 Solana stable-swap whose token was sold through FTX — and whose relaunch took almost three years to reach a token of its own. After FTX fell, insiders accepted a 50% haircut on unvested MER and holders were promised a share of a future token; that promise was honoured three years later, with 20% of MET’s supply going to Mercurial holders and a reserve. In between came the work that actually built the position: DLMM shipped, JUP launched through a single-sided DLMM pool in January 2024, and the LP Army accrued points for almost two years. Then January and February 2025 delivered both the peak and the scar — TRUMP and MELANIA ran through Meteora’s rails, and LIBRA collapsed within hours, after which CEO Ben Chow resigned and Zen and Soju took over as co-leads.
The token itself arrived on 23 October 2025 and peaked at ~$0.69 on day one — which is why it trades ~53% below that high today. Q4 2025 was the best quarter the business has had ($17.3M of revenue) and it was also when Meteora bought back the most, ~$12.7M, before $1.5M was lost to an OTC impersonator. The last buyback was in March at $0.1427. Since then the movement has been on the fee-share side rather than the buyback side: the DLMM protocol share went from 5% to 10% in May, and from 21 July stakers and referrers began splitting 20% of DLMM protocol fees in USDC. MET bottomed near $0.094 and rallied ~60% in thirty days to ~$0.32. Read as a whole, the history is a company that keeps its promises to the people it owes — MER holders, LPs — while the token’s own claim has been assembled piece by piece, by choice rather than by covenant.
The rails, not the storefront
Meteora’s product is liquidity machinery that others plug into: aggregators route through it, launchpads issue on it, LPs run strategies on it. DLMM is the core fee engine — bin-based concentrated liquidity with zero slippage inside each bin and fees that rise with volatility, about 87% of Q1 2026 volume — and since July it is also the only line shared with stakers. DAMM v2 is where graduated launches migrate, carrying a ~19.5% protocol take, the highest-margin pool type and not shared with stakers at all. The Dynamic Bonding Curve is the white-label launch rail behind Bags and eleven new launchpads in H1, at the cost of handing ~81% of partner fees to them. Terminal added trading-style LP tooling and on-chain limit orders that pay the placer half the fees earned, booking $64K of revenue by end-June. None of it requires MET, and that is the architecture’s defining fact for the token. The irony worth naming sits in the newest product: Bedrock exists to give other launched tokens an enforceable claim on a real company, through a Cayman foundation holding preference shares and a golden share. MET itself carries no such claim.
Competitively, Meteora is first among the unintegrated — and what is left unintegrated is the part that pays the most. Solana’s DEX market reorganised around vertical integration, where the launchpad that owns the user now owns the AMM too, which is how PumpSwap reached ~34% of Solana DEX volume. Meteora leads the general-purpose AMMs on the measure that matters to the people it needs: LPs earn roughly 1.2% net per dollar traded on DLMM against ~0.20% on PumpSwap, about six times more per dollar of memecoin volume in H1. That economics, not branding, is what pulls liquidity in — its volume share of the four-venue set was ~10.2%, but its fee share was ~26.7%, second only to PumpSwap. The vulnerability is the mirror of the strength: distribution is rented. Bags accounted for 95% of H1 launchpad activity and roughly half of DLMM fees come from pump.fun tokens, so the Pump.fun-to-Raydium precedent — a partner internalising liquidity overnight — is the risk written on the wall.
Which dollar are you buying?
MET is not equity: it has no claim on revenue, on the treasury or on the operating entity. So the valuation has to be run twice, and the two answers disagree. On what the protocol earns, $178M against ~$48M of trailing revenue is ~3.7×, and FDV of $320M is ~6.7× — cheap beside RAY at ~22× and PUMP at ~6.3×. On what the holder receives, the same $178M against ~$3–6M a year is ~30–60×, and the ~$28M a year of new supply is five to nine times that flow. PUMP’s two lenses nearly converge; MET’s diverge by ~8–16×, and that divergence is the entire investment question. The scenarios follow from which way it closes: a bear where revenue reverts to the Q2 trough of ~$20–25M a year, integrators pull the long tail in-house and buybacks stay dormant puts the market cap at $60–100M ($0.10–0.16); a base of $40–55M of revenue with staking scaling on DLMM and a holder share of 20–30% gives $160–275M ($0.26–0.44); a bull with a 2025-style launch cycle at $90–120M of revenue, a formal rule sending at least half to holders and the stack burned reaches $450–840M ($0.70–1.35).
Governance is where the two halves meet, and the honest description is transparent but not accountable. The disclosure is genuinely best-in-class for this series — monthly recaps, quarterly and half-year reports, a wallet directory, a live buyback wallet, 40/40 on an external transparency standard. What is missing is any mechanism that binds the money: we found no on-chain vote behind 2026’s changes, program-upgrade authority is not disclosed for a protocol of this size, and the treasury statement was retired in the H1 report, citing attacks on crypto treasuries. The same small, largely pseudonymous group sets the take rate, sets the staking split, and decides what happens to the 38.8M MET sitting in the buyback wallet — burn, lock, sell or spend. So: if Meteora wins, does MET win? Yes, partially, and at management’s discretion. What MET owns of that win is a USDC stream worth about a cent per fee dollar, a warehoused buyback stack and a hope. Meteora did the hard part — it built liquidity machinery that LPs choose on economics, reports to holders like a listed company, and has started paying them in dollars. The remaining step is the cheap one: writing down the rule. The prize is visible from here; the covenant is pending.
Key findings
- The business is real, cash-generative and unusually well documented. Meteora runs Solana’s most-used concentrated-liquidity AMM and kept $18.2M of $164.3M of H1 2026 fees, lifting its take rate from 9.6% to 11.1% without losing LPs — retention held at 48.5% against a 48.4% baseline through May’s fee change. The last thirty days printed $30.3M of fees and $3.56M of revenue on $6.77B of volume. Everything is on-chain and independently indexed, the team publishes monthly, quarterly and half-year holder reports, and it scored 40/40 on Blockworks’ Token Transparency Framework.
- But ~89 cents of every fee dollar never reach the company, let alone the token. Of H1’s $164.3M, LPs kept ~$104.9M (63.8¢) and launchpads ~25.1¢; Meteora’s own slice is ~11.1¢, and of that only ~0.7–1.4¢ flows to stakers and referrers. This is the book’s thesis in one chart — on trading venues most fee value reaches the LP and whoever owns distribution, not the governance token — and Meteora is its cleanest illustration in this series. The LPs are not being cheated; they are the product. That is precisely why the token’s share is the thing to underwrite.
- Three routes to the holder, and not one of them binds. First, referral staking: since 21 July, 20% of DLMM protocol fees are split between stakers and referrers in USDC — cycle 1 paid $262K, cycle 2 was unconfirmed at the time of writing, and the 20% rule caps a cycle near ~$0.52M, so ~$3–6M a year. Second, buybacks: $13.7M for ~36.0M MET since October 2025, none since March, and the wallet’s 38.8M MET is held, not burned. Third, the treasury: last disclosed at $33.9M on 31 March 2026, with disclosure retired in the H1 report and DefiLlama tracking ~$19.4M in known wallets. None of the three is a contractual right, and we found no on-chain vote behind 2026’s changes.
- Dilution runs five to nine times larger than the payout. MET launched with 48% liquid and a one-month cliff on the rest — a deliberate rejection of the low-float, high-FDV model, and genuinely to the team’s credit. The other 52% vests linearly to team-controlled wallets for six years: 7.22M MET a month, roughly $28M a year at spot, against the ~$3–6M a year holders receive. Counting the team’s 18%, the 34% Meteora Reserve (which has no disclosed purpose), the Mercurial Reserve and the buyback stack, about 61% of supply is or will be team- or protocol-controlled — ~56% excluding the 5% Meteora treats as due to MER holders.
- Cheap on one lens, expensive on the other — and the gap is the report. At ~$178M, MET trades at ~3.7× trailing revenue, a fraction of RAY at ~22× and PUMP at ~6.3×, but at ~30–60× the cash holders actually receive. PUMP’s two lenses nearly converge; MET’s diverge by ~8–16×. Three things sit behind that discount: revenue is violently cyclical (fees fell ~82% from Q3 2025 to Q2 2026 before recovering), distribution is rented (95% of H1 launchpad activity came from Bags, and about half of DLMM fees come from pump.fun tokens), and Hurlock v. Kelsier Ventures in the S.D.N.Y. names Meteora and its former CEO in a RICO-style class action whose motions to dismiss were still pending. The allegations are unproven.