Everybody's got a favorite cheap token. Under $300 million market cap, "still early," one announcement away. Fine. Let's run the only two questions this channel ever asks:
- Where does the money come from? Not the story — the money. Who pays, how much, and does any of it reach the token?
- Who's standing behind you in line to sell? If half the supply is still locked, you're not early. You're the exit.
Ten tokens under $300M, including one I happen to know very well. Ranked best to worst on those two questions. No mercy round at the end.
1. Maple (SYRUP) — ~$201M — the boring one that passes
SYRUP · Official website
SYRUP price, last 12 months in USD (weekly closes). One-year change: -55%. Chart data: CoinGecko, Oct 25, 2026.
Maple lends money to real institutional borrowers, on-chain, and charges them interest like a bank that never closes. Assets under management: $4.7 billion. Loans outstanding: about $1.7 billion. September protocol revenue: $1.57 million — from borrowers paying interest, full stop. That's where the money comes from, and you can watch it happen.
Then it gets better. Under a rules-based system (MIP-021), Maple spends a fixed slice of monthly revenue buying SYRUP back off the open market: 10% of revenue in a slow month, 20% once revenue crosses $1.5M, 30% above $2M. September crossed into the 20% tier, so the fund bought back $314,116 worth — 1.37 million SYRUP at an average of $0.229 — confirmed completed on October 9, with every purchase posted on a public transparency page. The program has bought back 2.9 million SYRUP so far. Not a promise. A receipt.
Supply check: 92% unlocked, fully diluted value $219M against a $201M market cap. There's no army of locked tokens behind you.
Two honest catches. First, lending is credit risk — borrowers can default, and a bad credit cycle hits Maple first and asks questions later. Second, the buyback is not a price floor: SYRUP slid about 24% in a single week in mid-October while the buybacks were doubling. A rules-based buyback is a real pipe — it's just a garden hose next to a market tide. It's also not doing a 110% week, ever, in either direction. It's a business.
Score: Legit. The yield has a return address.
2. Orca (ORCA) — ~$104M — passes on paper, and the paper is being rewritten
ORCA · Official website
ORCA price, last 12 months in USD (weekly closes). One-year change: +16%. Chart data: CoinGecko, Oct 25, 2026.
Orca is Solana's old reliable DEX — the place a huge share of Solana swaps still route through. Traders pay swap fees; Orca keeps a 12% protocol share of those fees; and 40% of that share is used to buy ORCA off the market and drop it into the xORCA staking vault. Real trades, real fees, real buybacks. Holders' revenue is tracked publicly at roughly $9.5K a day lately. Supply: about 81% unlocked, FDV around $129M — modest overhang, fine.
Now the fine print, and since the first draft of this post it has become a whole second story. On September 29, a governance proposal landed that would rewrite the deal: cut the xORCA buyback share from 40% to 10%, hand another 10% to a team-run buyback wallet, route 80% of fees to team operations, move about 14.2 million ORCA and 70,000 SOL from community accounts to team-controlled ones, and dissolve the council. The unnamed "acquisition target" turned out to be a full merger — on October 7, Orca announced it had merged with lending protocol Loopscale to form a combined company called Formation, run by Loopscale's CEO. No new token, no new mint; ORCA now represents the combined outfit.
Then the vote fell apart in public. The first vote was cancelled on October 8 — because xORCA stakers, the very people whose payout was being cut, couldn't vote without unstaking for seven days first. A re-run was held October 11–16. And as of the day this post goes up, I can't find an officially announced result — so whether the new split is even live is an open question, while the price has already taken the full ride: ORCA ran up more than 90% to about $3 during the vote mania, then gave nearly all of it back. It's $1.72 today.
The team's argument — "we're a startup, we need the money to grow" — might even be right. Doesn't change the lesson: a buyback written in governance can be unwritten by governance, and sometimes you can't even find the minutes. What a token pays you is only as solid as the vote that approved it — assuming you can find out how the vote went.
Score: Legit mechanism, live renegotiation — and the renegotiation is a mess. Check the governance forum before you check the price.
3. Meteora (MET) — ~$166M — a money machine with a drip
MET · Official website
MET price, last 12 months in USD (weekly closes). One-year change: -44%. Chart data: CoinGecko, Oct 25, 2026.
Meteora is the liquidity plumbing under Solana — its pools sit beneath a giant share of Solana swaps and token launches. The first half of 2026, in its own published holder report: $32.1 billion in volume, $164.3 million in fees, $18.2 million in protocol revenue. Lately it's been running around $34M in fees and $4.1M in protocol revenue per 30 days. Some of that money reaches holders — buybacks funded from protocol revenue, plus staking rewards paid out of pool fees. The business is not imaginary.
The drip is. Only about 56% of MET is unlocked. Team and reserve tokens vest at roughly 7.2 million MET a month — about $2.1 million of fresh supply at today's price — landing on the 23rd of every month until 2031. The most recent one hit two days before this post, on October 23, worth roughly $2.5M. This isn't a theoretical overhang; it arrived this week. And the machine's fuel is Solana trading volume, which cooled hard in Q2 (down 36% quarter over quarter). When Solana sneezes, Meteora's fees catch pneumonia.
Score: Real business, real drip. You're paid to wait out the vesting — know that going in.
4. Akash (AKT) — ~$193M — the receipts don't match the story (yet)
AKT · Official website
AKT price, last 12 months in USD (weekly closes). One-year change: +2%. Chart data: CoinGecko, Oct 25, 2026.
Akash is a marketplace where anyone can rent out GPUs and anyone can rent them, usually far cheaper than the big clouds. In March 2026 it switched on a burn mechanism: network usage now burns AKT, tying the token to actual compute demand instead of vibes. On paper, that's exactly the design this channel applauds.
Then you read the independent numbers. Messari's Q1 2026 report tracked on-chain lease revenue of about $253,000 for the whole quarter, GPU utilization down to 33.7%, and active providers at a record low of about 58. Akash's own reporting for the same period claims roughly $5 million in compute spend. Both numbers can't be the headline. One is the protocol's own dashboard; the other is an outside firm counting what's actually on-chain — and nobody has publicly reconciled the gap.
Also note: "100% unlocked" doesn't mean no new supply. AKT inflates at roughly 9% a year in staking rewards. The dilution isn't locked tokens — it's the printer, running politely.
Score: The story is ahead of the receipts. Interesting, unproven, and the perfect example of why you check the second source.
5. Across (ACX) — ~$31M — great bridge, token gets nothing
ACX · Official website
ACX price, last 12 months in USD (weekly closes). One-year change: -40%. Chart data: CoinGecko, Oct 25, 2026.
Here's where it gets uncomfortable, because Across is an excellent product. It's the third-biggest bridge in crypto by volume — about $169 million moved in a day at last check, behind only Circle and LayerZero. Fast, cheap, capital-efficient. The protocol earns fees on every transfer.
And according to DeFiLlama's holder-revenue tracking, the amount of that money that reaches ACX holders is zero dollars. Not small. Zero. No buybacks, no distribution, no fee share. ACX is a governance token — you can vote, and that's the whole list. Meanwhile 30% of the supply is still locked.
Sound familiar? It should. This is Wormhole's disease in miniature: a bridge doing real volume while the token lives on a promise that someday the money will be shared. Maybe it will — governance can switch it on, same as it can switch Orca's off. Today, it hasn't.
Score: The business is real and the token is a spectator. Same warning label as #6, one-third the size.
6. Wormhole (W) — ~$79M — the control group
W · Official website
W price, last 12 months in USD (weekly closes). One-year change: -84%. Chart data: CoinGecko, Oct 25, 2026.
You knew this one was on the list. Wormhole is genuine infrastructure — BlackRock's BUIDL, Hyperliquid's HYPE, Ripple's RLUSD all move across chains on Wormhole tech, and the protocol sits on about $1.76 billion in total value locked. Nobody's arguing the product is fake.
The token is the problem, and it's the same problem as Across, just louder. W 2.0 promises a staking yield around 4%, backed by a "Wormhole Reserve" funded by protocol revenue — revenue Wormhole has not publicly disclosed. The yield's source is, officially, "trust us." Meanwhile only about 67% of supply is unlocked, the rest drips out bi-weekly through 2028, and this Friday, October 30, another unlock lands for core contributors.
Two weeks ago W doubled in days, kissing $0.03 intraday on rotation and leverage — perpetuals volume at 15 times spot, open interest bigger than the market cap — and it's given back about 60% since, to $0.0119 today. That's what a token does when its price is a story instead of a share of a business.
Score: Bogus until the revenue is shown. Great trade setup, bad investment — and the difference is whether there's a number you can point to.
Round 2 — four more that fit the same rules
7. Geodnet (GEOD) — ~$118M — the DePIN that actually burns
GEOD · Official website
GEOD price, last 12 months in USD (weekly closes). One-year change: +91%. Chart data: CoinGecko, Oct 25, 2026.
Geodnet sells precision GPS correction data — the stuff farms, drones, and surveyors pay for when regular GPS being off by a few feet isn't good enough. Real customers, paying real subscription money. And the fee design is the most aggressive on this whole list: 80% of the fees Geodnet receives are used to buy GEOD back on the open market and burn it. Not "up to." Not "at the DAO's discretion." Eighty percent, with the other 20% running the foundation. Holders' revenue tracked so far is about $558K — small in absolute terms, but the pipe is wide open and pointed at the token.
The catches: only 50% of supply is unlocked (FDV ~$236M, double the market cap — the drip exists here too), the absolute revenue is still modest, and it trades thin — a couple million dollars a day on a good day. But on the two questions? Money comes from customers, and 80 cents of every fee dollar goes to destroying supply.
Score: Legit. The widest pipe between customers and token on this list.
8. Railgun (RAIL) — ~$149M — privacy pays its stakers, carefully
RAIL · Official website
RAIL price, last 12 months in USD (weekly closes). One-year change: +12%. Chart data: CoinGecko, Oct 25, 2026.
Railgun is privacy plumbing for Ethereum — it lets people shield assets and use DeFi without broadcasting every move. Users pay 0.25% to shield and unshield. That fee switch is on: protocol revenue is distributed to stakers from the treasury every two weeks. Last 30 days: about $370K in fees, about $212K of it reaching holders, $3.06M cumulative. Roughly 70%+ of the supply is staked, so the float is thin and jumpy. Supply is effectively fully diluted at ~$149M — no unlock cliff.
Now the fine print, because privacy comes with two kinds. Regulatory: privacy tools are the first thing regulators reach for, and EU rules tightening on privacy coins from 2027 are part of why money is rotating here — that cuts both ways. And operational: in July 2026 a critical flaw in Railgun's rewards contract let anyone permanently freeze about $2.6M of distributable treasury funds. Nobody stole anything, it was fixed by an emergency upgrade — but the machine that pays you was freezable by a stranger for a month. That's not a dealbreaker. It's a reminder that "the code pays me" also means "the code can stop paying me."
Score: Legit mechanism, hazard pay. The yield is real because the risks are.
9. Helium (HNT) — ~$77M — passed, then changed the deal (see also: Orca)
HNT · Official website
HNT price, last 12 months in USD (weekly closes). One-year change: -80%. Chart data: CoinGecko, Oct 25, 2026.
Helium runs a crowdsourced wireless network, and the money is gloriously real: AT&T, T-Mobile, and Telefónica pay to offload their subscribers' data onto Helium hotspots in supermarkets, hospitals, and stadiums. Usage burns HNT — about $50K a day at recent rates — and in Q1 2026 the burn ran about $3.65M for the quarter. Supply is 100% unlocked, capped, with emissions halving on a schedule. Textbook pass.
Except. Helium Mobile used to take 100% of subscriber revenue and buy HNT on the open market to burn — $2.25M of buybacks in Q4 2025 alone. On January 2, 2026, they paused it and pointed the money at "network growth" instead. Holder revenue since: zero, per DeFiLlama. The usage burn continues, but even that has an asterisk — the independent trackers note carriers actually pay around $0.10/GB against a $0.50 protocol peg, and the burned HNT comes from treasury wallets, so how much of the "burn" is genuine market demand is genuinely debatable.
Pair this one with Orca in your head. Two different protocols, same lesson: the deal you bought can be re-dealt. A pause button is a policy, and policies change at meetings you're not invited to.
Score: Real business, real burn, revoked buyback. Grade the current deal, not the old press release.
10. Kamino (KMNO) — ~$250M — the biggest business in the $0 club
KMNO · Official website
KMNO price, last 12 months in USD (weekly closes). One-year change: -31%. Chart data: CoinGecko, Oct 25, 2026.
Kamino is Solana's biggest lending market: $1.34 billion locked, $856M in active loans, $4.9M in fees in the last 30 days, about $633K of it protocol revenue. By business size, it dwarfs half this list.
Holders got $0. The fee switch is off. There are no rules-based buybacks — fee capture goes to the treasury, and anything beyond that is "at the DAO's discretion," which is a phrase that has never once appeared on a receipt. A teardown this month added the ugly details: what Kamino keeps has fallen for six straight quarters, and Jupiter Lend just passed it in deposits and active loans. Meanwhile only 58% of KMNO is unlocked, with insiders holding the big tranches on vesting schedules through 2028.
So Kamino joins Across and Wormhole in the club: real business, real fees, token as spectator. The difference is scale — this is the most expensive spectator ticket on the list at $250M.
Score: Great protocol, governance-only token. You're buying a maybe about a future fee switch, priced like a yes.
Rapid fire — three more that fit the rules
Fluid (FLUID) — ~$183M. Lending protocol with treasury buybacks and holder revenue on paper. Fails the third, unwritten test: it's up ~50% in 30 days. Right church, wrong Sunday — you don't chase the one green chart in a red month. — Website
Marinade (MNDE) — ~$14M. Solana staking pool that has sent 50% of revenue to MNDE buybacks since September 2025. The mechanism is real; the token is a microcap that trades almost nothing some days. A screenshotted curiosity, not a position. — Website
Beefy (BIFI) — ~$6M. Yield aggregator sending 36% of revenue to stakers, fully unlocked, ancient by DeFi standards. Same story as Marinade: honest pipe, thimble-sized bucket. — Website
FLUID · Official website
FLUID price, last 12 months in USD (weekly closes). One-year change: -52%. Chart data: CoinGecko, Oct 25, 2026.
MNDE · Official website
MNDE price, last 12 months in USD (weekly closes). One-year change: -77%. Chart data: CoinGecko, Oct 25, 2026.
BIFI · Official website
BIFI price, last 12 months in USD (weekly closes). One-year change: -49%. Chart data: CoinGecko, Oct 25, 2026.
The scoreboard
- Money reaches the token, on paper you can check: Maple, Geodnet, Railgun (hazard pay)
- Deal changed or changing — read the current one: Orca (vote result unannounced), Helium (buyback paused)
- Real money, real drip: Meteora
- Story ahead of receipts: Akash
- Token gets nothing today: Kamino, Across, Wormhole
Under $300M is where the bodies are buried — almost everything down there is down 90%+ for a reason. But "down a lot" isn't the test. The test is whether somebody, somewhere, is paying for something real, and whether any of that payment has your name on it. Three of these ten pass clean. That's a better hit rate than most people expect — and a worse one than most people assume.
Not financial advice. Prices and figures checked October 25, 2026 — re-verify before you repeat them, including me.
Comments
Post a Comment