---
name: rh-revenue-board
description: The 15 Sep 2026 re-screen of the ChatGPT "RH plays" shortlist, and the revision that followed when Ben and an independent review both broke it. Read before ranking any Robinhood Chain token. Carries the SHROOM double-count retraction, the holder-revenue-vs-protocol-revenue rule, the buyback-over-depth metric, the memestock flow evidence, and the live position calls.
sources: [cowork]
aliases: [green chain board, rh revenue board, shroom retraction, chatgpt rh screen, hashcats trap, buyback over depth]
---

# Robinhood Chain board, 15 Sep 2026
**LIVE ARTIFACT (revision 11, now a DASHBOARD: stance -> chart -> sector bars -> 8 holdings -> opportunities, research folded into details): https://claude.ai/artifact/7wX6o6frpmXGkt9Agcmwiw** . The old address https://claude.ai/artifact/LixbW3ubbgA9FWeP4CM8UQ is retired. **Both addresses serve a PINNED OLD SNAPSHOT to anyone who is not Ben; the owner sees live. This cannot be unpinned from the assistant side and republishing at a fresh URL did not clear it. Ben must re-share from the page's own share menu.** This is why every review Ben commissioned kept describing content two or more revisions old.

- [stated] Ben pasted a ChatGPT share link ("RH plays", two runs, 51 then 55 entries), asked to check everything in it and find the best investments on Robinhood Chain, on the thesis that the chain has support from the top and memestocks are being pushed. He holds MARSCOIN, SCHIFFY (long.xyz) and PAIREX.
- [stated] **Ben's pushback, and he was right:** *"You are purely looking at revenue... stockbroker is making revenue according to your list but has been performing terribly. So your view seems a little one-sided and not really reliable."* He also asked why DELTA and HOOKR were missing and whether AI could be a strong play.

## FOUR RETRACTIONS TO REVISION 1. Three were found by an independent review Ben commissioned, one here.
1. **SHROOM WAS DOUBLE COUNTED.** I summed DefiLlama's `shroom-token` ($21,002/day) and `shroom-treasury` ($21,607/day) into $42,609/day and published 0.50x. **The adapter source says in its own comments: "Do not add the two scopes: treasury dividends are part of the token stream."** Verified at `raw.githubusercontent.com/DefiLlama/dimension-adapters/master/fees/shroom.ts`. Correct figure is $21,002/day, $573,928/30d -> $6.98M annualised -> **1.12x, not 0.50x**. Worse for the thesis: the token scope is fees and tax from the **SHROOM/MU genesis pool**, i.e. from trading SHROOM itself, not from routing stock tokens. The routing business is the website's separately reported $8,916/day of LP fees.
2. **PONS WAS MEASURED ON THE WRONG LINE.** Protocol revenue gives 1.13x. **Holder revenue is ~$12.66M/30d against ~$441M cap = ~2.7-2.9x**, level with pump.fun at 3.0x. Still the best revenue asset on the chain, no longer a bargain.
3. **PRISM WAS THE WRONG COMPANY.** DefiLlama's `prism` is **Prism Network**, GPU rental at prismnetwork.tech, category DePIN. ChatGPT's PRISM is **Prism Assets** at prismassets.shop. The $1.68/30d belongs to the GPU business. I saw the URL and the DePIN tag and published the row anyway.
4. **A MISSING DEFILLAMA ROW IS NOT ZERO REVENUE.** It means no adapter exists. It blocks an earnings case; it does not prove a business earns nothing. Correct status for INJOH, GAGE, GRASS, MONEY, HARMONIC, CEREBRO, MANY, SOUNDER, UFG, HFARM, RECORD, VEX, DELTA and HOOKR is **unverified revenue**.

## SECOND REVIEW, same day: THE LANE CLAIMS WERE FALSE AND THE MECHANICAL FIX ALREADY EXISTED
- **"SPY has $33.8M/day of volume and NO memecoin" was FALSE, and it was the headline of the buy list.** I read SPY's pools through `api.dexscreener.com/latest/dex/tokens/{stock}`, which **returns at most 30 pools**. SPY's top 30 are all other STOCK tokens (QQQ, META, COIN, TSM, GOOGL, PLTR, AMZN, MU, TSLA), so every memecoin sat below the cut. Re-read through `api.geckoterminal.com/api/v2/networks/robinhood/tokens/{addr}/pools?page=N` over multiple pages, the SPY lane has at least eight: **PAIR pool liq $651,657 - INTC $241,092 `0xc72b96e0e48ecd4dc75e1e45396e26300bc39681` - STONKS $214,186 `0x1ec871553be07dc68ec1661c32eca7e2e46e1e18` - INVEST $193,084 `0x4B7720cb7025eceBE616fc6c93185Ae4414c1E18` (cap $528,387, v24 $328,468, 4d old) - MONEY $121,208 - BRAIN, Grady, HAI, PON.**
- **"FLYBRAIN is the only memecoin on GOOGL" was FALSE.** **QUBIT `0xd1E92BA7c7355c2698addD894919Cba198601E18` has the DEEPER GOOGL pool: $446,033 against FLYBRAIN's $191,344**, cap $1,999,577, v24 $1,082,188, and **RSI `0xce4c9418b8e9bcad3e4ba2bd84123a49a6911e18` does more volume ($1.55M, 1 day old, +1044%)**. Also AGI, RSINU, Goblin, PHYSARUM. **FLYBRAIN's own deepest pool is USDG, not GOOGL.** QUBIT context from [[rh-chain-thesis]]: it was **$60.4M on 29-30 Aug in the froth gauge** (launched in a 95-minute window with zero socials) and is now $2.0M, about -97%, with sells leading buys. Not a buy.
- **THE REPEAT: this is the identical trap logged on 8 Sep in [[marscoin-bnb]]** ("DexScreener returns at most 30 pairs per token"; "the category screen must run BOTH directions"; "enumerate the quote-asset family EXHAUSTIVELY before ranking a category"). **The mechanical fix already existed and I did not run it: `/tmp/tools/quote_family_rank.py`, which pulls up to 200 pools per family member from GeckoTerminal precisely because DexScreener caps at 30.** Per Ben's standing rule, a repeated mistake gets code and a gate, not another paragraph. The code exists. **THE GATE: before any "lane is empty / X owns the lane" claim, run quote_family_rank.py or paginate GeckoTerminal's token-pools endpoint, and state the page count reached. A 30-pool read can never support a zero.**
- **AI's backing was overstated.** I wrote "$6,232,957 of NVDA sitting under it". That is the **whole pool, both sides**. The NVDA side is roughly **$2.5M**, under 1% of a $301M cap, and is not redeemable: artificialinu.com denies NVIDIA ownership and denies holder redemption from the vault. GeckoTerminal reports the same pool's `reserve_in_usd` as **$29,243,899** against DexScreener's $6.6M, so neither is a clean read of one side. **Never quote pool liquidity as "backing".**
- **STANDARD: every tracker prints cap = FDV = $37.3M and it is wrong.** Read on chain: **`totalSupply` 98,185,046 against `maxSupply()` 998,805,780**, so today's price is on 9.8% of eventual supply and the fully diluted figure is near **$380M**. Issuance streams ~700,000/day; holding the token alone earns no branch issuance and no governance. The launch surcharge reaches its 2%/3% floor **one hour** after trading opens, so "wait a day for the tax to decay" was stale when written. **Removed from the buy list.** Same float trap as UP (3.9%) and NET (7.9%) in [[price-driver-model]].
- **"PONS burns 3% of its pool per day" WITHDRAWN** until tied to a named pool, window and unit. PONS's own revenue is unaffected: $1,053,317/day, $22,371,995/30d, cap ~$424.6M = **1.56x on trailing 30d protocol revenue vs pump.fun at 2.49x** ($1.675bn cap, $55.26m 30d). DefiLlama's PONS holder-revenue line counts tokens moved from burner wallets to dead and values them; it is not a total of ETH spent.
- **MOO sell call WITHDRAWN.** It rested on 442 buys against 921 sells; **counts do not weigh dollars**, and MOO was +9.1% on the day that produced those counts. Transaction-count skew is not an exit signal on its own.
- **SLVR buyback re-priced and it holds:** 3.01 ETH, about **$7,446** in 24h, buying and burning 116 SLVR against $157,183 of pool = **4.7% of the pool** (I said 5.2%). Against the $1.79M cap it is 0.42%. It is buy pressure, not a yield, and SLVR also issues fresh supply as mining rewards.
- **PAIREX's 52,435,292 / 12 Oct unlock is UNVERIFIED against the locker contract.** It came from the 12 Sep screen, not from a contract read. Do not trade the date until it is tied to the vesting contract.
- **CLARITY cloture vote timing, corrected to Ben's clock: 01:15 Bangkok Wednesday 16 Sep (18:15 UTC / 2:15pm US Eastern Tuesday).** I told him "today" in a Bangkok-time conversation. It is a procedural step needing 60 votes, not enactment.
- **THE FRAMING THAT SURVIVES:** not "find the empty lane" (none are empty) but **"find the lane whose winner is not yet decided and back the one gaining depth"**. On that test NVDA (AI) and HIMS (BONER) are settled, AAPL is being claimed right now by INU, and SPY and GOOGL are genuinely contested with no candidate worth buying today.

## BEN'S TRADING MANDATE, stated 15 Sep 2026 (this reframes every ranking above)
- [stated] **Horizon is A COUPLE OF MONTHS, flexible with the trend.** No minimum hold, no commitment to ride a full cycle. "Catch the main part of the move": notice a trend early, participate as it strengthens, take a substantial part of the advance. Catching the exact bottom or top is explicitly NOT the objective.
- [stated] **A 2x or 3x is an illustrative ambition, not a required result.** A smaller move can be useful.
- [stated] He thinks there may be **another year to eighteen months of bull market**. That is a planning scenario he holds, not a verified forecast, and it does NOT set the holding period of any single coin.
- [stated] Every candidate must carry: a dated reason to look now, evidence the trend is turning, room left for a worthwhile move, tradability at his size, what could interrupt it during the hold, and **a stated exit or invalidation defined before entry**.
- **CONSEQUENCE: business-quality ranking is BACKGROUND, not the trade ranking.** For a meme the thesis can be narrative demand and rotation; recurring revenue is not a prerequisite to appear. This is the third time in one day that ranking on business quality proved to be the wrong primary axis for what he actually does.

## THIRD REVIEW, same day: TWO ON-CHAIN ERRORS OF MINE, BOTH CONFIRMED BY MY OWN RE-READ
- **DELTA'S 7.5% IS DEPLOYED. I read the wrong contract and reported "could not settle it".** Verified myself on the Robinhood RPC: `feeBps()` returns **750 (7.50%)** on BOTH newer managers, **v8 `0xbCb96b15dC2246D242c879316c86e25e846ad5FB`** (code 22,361 bytes) and **v9 `0x83551A233f6D7b3cb4dEFBF55AbFD887C6811ce4`** (code 29,284 bytes). The v3 manager `0x5cA6214227D1195c4b7b4B96847b8966c688295D` still returns 100 (1.00%) and is simply the OLD one. **LESSON: a contract read is only evidence about the contract you read. Enumerate the deployed versions before concluding a documented rate is not live.** Open question that remains: how much fee-claim activity actually routes through v8/v9, so do not apply 7.5% to historical positions.
- **THE PARE ADDRESS IN MY BOARD WAS WRONG AND HAS NO CODE.** I printed `0x9b8a4bd8a7a5ba1b1e8b03e7e5c05b7cf0e7d3a7` on a row I gave HIGH ENTRY PRIORITY. `eth_getCode` returns empty. **The real PARE is `0x15d36b6a28d8327abc7afabf0f106ae2c9af5c4d`**, verified: symbol PARE, 18 decimals, 1,000,000,000 supply. **GATE: never print an address on a recommendation row without an eth_getCode plus symbol() read first.**
- Also accepted: my "recovery 2/3 OR daily revenue above $1M" release condition was wrong, because it let business quality bypass the market gate; ASTEROID explains 12.48pp of the fixed-basket move, not the whole 48.62pp rotating-vs-fixed gap; "DELTA had the best resilience" contradicted my own table where BOW and PARE fell less; and my header timestamp said 19:30 Bangkok when the page was readable at about 17:20.

## RHX at 17:00 Bangkok, 15 Sep (fresher than the 16:00 read above)
**Zone moved BUY ZONE -> HOLD ZONE**, heat 18.86 -> **31.14**, recovery still **0 of 3**, RHX 7d **-19.17%**, 24h **+4.97%**, **-39.08%** from peak. Fixed cohorts: MEME basket -18.45% (median -17.60%, win 30%, ex-top -21.74%, top FRONG +44.0%); INFRA -20.63% (median -13.50%, win 11%, ex-top -24.50%, top NET +10.3%); **STOCKMEME -29.91% (median -51.21%, win 25%, ex-top -41.89%, top ASTEROID +197.7%)**. Still no cohort keeps a lead once its best coin is removed.

## THE STANDING RULES THIS PRODUCED
- **ALWAYS pull `dataType=dailyHoldersRevenue` alongside `dailyRevenue`.** They differ enormously: STONKBROKER 24h protocol $26,591 vs holders $6,159; HASHCATS $32,413 vs $24,307; shroom-treasury $21,607 vs **$0**. Protocol revenue is the wrong denominator for a token.
- **NEVER sum two DefiLlama scopes for the same token without reading the adapter.** Scope overlap is documented in the adapter comments, not in the API.
- **A revenue multiple is not a price forecast on this chain.** [[price-driver-model]] already lists "cheapness on a revenue multiple" as a scored-zero anti-signal, verified against SSRN 6901559 (rho 0.04, Fama-MacBeth p=0.98). I ignored my own file.
- **THE METRIC THAT REPLACES IT: daily buyback dollars divided by aggregate pool liquidity.** From the constant-product identity, price moves on net flow over depth; a buyback is only a price force when it is large against the pool it buys into. Measured 15 Sep: **HASHCATS 18.4%/day (decaying), SLVR 5.2%, UP 3.1%, PONS 3.0%, SHROOM 1.0%, The Index 0.6%, STONKBROKER 0.14%.** That last number is the mechanical answer to Ben's exact objection: STONKBROKER earns, but its holder bid is 0.14% of depth per day, which cannot move a price.

## THE EMPIRICAL TEST BEN'S PUSHBACK FORCED, and it went against me
Seven-day price change, deepest pool, hourly candles, 15 Sep.
- **Revenue names, nine of nine DOWN, median -26%:** PAIR -76%, SHROOM -65%, ROUTE -43%, UP -31%, SLVR -21%, PONS -15%, DELTA -15%, HASH -9%, plus STONKBROKER flat on a stale pool.
- **No-revenue names, eight of eleven UP:** MONEY +1,010%, CEREBRO +1,899%, MANY +456%, HARMONIC +117%, PRM +56%, **AI +24%**, GRASS +10%, MANCER +6%; down were OURO -73%, ICOIN -65%, GAGE -39%, MOO -43%.
- **Every mover on the chain this week was a token with no measured earnings.** One week, one chain, small n, but it is the window Ben trades in and it is the opposite of what my board implied.

## THE MEMESTOCK PUSH THESIS: SUPPORTED, with receipts
- **Tokenized stocks ARE the chain's volume.** 24h, GeckoTerminal top pools: NVDA $44,131,968 - SPCX $35,109,069 - SPY $33,837,468 - GOOGL $23,516,813 - MEME (ETF) $10,006,356 - AAPL $10,508,692 - amc $9,638,828 - glxy $8,017,013 - HIMS $7,153,484 - gld $5,850,658 - GME $5,133,314 - CRCL $4,164,648. SHROOM by comparison is $3.4M.
- **The stock-paired memes are the second layer:** AI/NVDA $32.3M, BONER ($41.98M cap, AI- and HIMS-quoted) $7.0M, INU/AAPL ($7.73M cap, 4d old, +187% 24h) $6.7M, amc/MEME ($4.33M cap) $9.6M, JUDE/SPCX ($16.9M cap, launched today) $3.1M, FLYBRAIN $5.4M, CHUMP $3.2M.
- **OPEN LANES with no dominant memecoin: SPY at $33.8M/day and GOOGL at $23.5M/day.** That is where a 10x sits, not in the leader.
- **TIER-ONE NARRATIVE OWNERSHIP ON AI, the strongest single piece of evidence.** @theunipcs (317,875 followers) 14 Sep 22:11 UTC, 130 likes / 29,646 views: *"i have only said positive things about $AI in every group chat i have discussed RH memes over the past month, as i objectively believe it is the best memecoin on RH."* Same thread: *"i am extremely bullish on @longdotxyz and the force they've become on RH... i had bought inu on my alt fomo since last week."* **This is the author of Ben's MARSCOIN trade naming AI.**
- @Natan_benish 9 Sep: *"if you're in AI pivot back to $AI. LONG."* @longdotxyz 7 Sep: added ~$135K of NVDA to the AI/NVDA market, total NVDA in that market $3.5M, "nearly 100% of recent" volume. 15 Sep: LONG past **$1.4B lifetime volume, $14.5M stock TVL, "roughly 9% of all stock TVL onchain"** (implies chain-wide stock TVL ~$161M).
- **AI `0x2E8c31162b855A2ffa90F6F8634643Ad6F111e18`**: cap $276,282,105, **aggregate liquidity $39,800,922 (deepest on the chain outside PONS, 6.9x cover)**, v24 $32,342,347, +24% 7d, +14% 24h, buys 2,753 / sells 2,016. Its own docs say holders cannot redeem from the vault and it is not an investment product, so Frame 1 is empty by design. Ceiling comp set: CASHCAT $152.5M below, TRUMP ~$655M above; **2.4x to 3.6x, not a 10x.**

## MISSED IN REVISION 1 AND MATERIAL
- **STANDARD `0x88ad8DdF1E3898412146a534538d418c6F8A9062` launched 15 Sep ~01:00 UTC.** Cap $41,333,028, **liquidity $18,602,537, v24 $51,200,437, buys 26,746 / sells 15,680 (+41% skew)**, quote ETH. @standard_rsv, 34,363 followers. Dutch auction then seeded liquidity, **launch/sniper tax decaying from 70%**, Genesis Charters with emissions, 100 branches auctioned in STANDARD on a descending price. Mechanism: issuance, reserves, buyback and burn driven automatically by ETH flow through its Uniswap v4 pools, rules in contract rather than DAO vote. **Ben already had a launch watch on this account ([[standard-reserve-watch]]).** Two identical "this is how i made $200k from the $Standard launch" sniper-bot brag posts from different accounts = copypasta bait, not signal. A Solana impostor STANDARDWOJAK exists.
- **DELTA `0xe8ffd7e24187f72afb08d75b1bb13088a989a791`** cap $9,806,623, -15% 7d, buys 1,705 / sells 1,731, no DefiLlama adapter. Docs mention the DELTA token zero times; contract take rate 1% of claimed fees, ~$244,000/yr, so ~40x at this cap. Pass.
- **HOOKR `0x18E674231A58c239Dc7DaeDcffE15Ec3A24cff5c`** cap $8,746,032, liq $871,719, v24 $1,661,212, buys 1,324 / sells 1,206, no adapter. Its burn was traced in [[hookr-token]] to 0xZAPS Clanker fees rather than its own product. Mechanism real, denominator never shown.
- **MOO `0xD9dB30BB0D2b8d2eae3826A1372117E058791e18`** cap $17,019,619, **-43% 7d, buys 441 / sells 920**. Being sold two to one. Routing thesis right, toll booth still absent.
- Other live names worth a look and not screened: **BLORB `0x4d14284afe559b7c6b9e6fad6ebaeaa0f6051818`** $64.3M cap / $6.4M liq / $10.3M v24, 21d old; **AnsemCat `0x67c51c928d5d8383f3a58fc591a4e6f6463949ed`** $100.1M cap on $23.8M liq, 3d old.

## The revenue board as corrected, 15 Sep (holder revenue where it differs)
- PONS `0x39dBED3a2bd333467115dE45665cC57F813C4571` cap ~$416-441M, protocol rev $1,020,298/day, holder rev ~$12.66M/30d, ~2.7x, buyback ~3.0% of depth/day. 30d fees $117,705,112 and 30d protocol revenue $20,774,811, an 8.6x step-up from the $2.4M of 30 Aug, so the [[rh-chain-thesis]] "step change or spike" question resolved to **STEP CHANGE**. Daily revenue halved from $1,983,214 on 5 Sep and has held near $1.0M for a week.
- SLVR `0x791229E3EbD6CFdC3D8157f48722684173C29aD9` cap $1,812,680, holder rev = protocol rev $37,626/day = 0.13x, buyback 5.2% of depth/day. The 31 Aug confirmation trigger from [[slvr-token]] (stakers above $25,000/day through 7 Sep) FIRED and kept firing. 71-bettor concentration not re-measured.
- UP `0x57C0E45cB534413D1C20A4240955d6bB250BB4F1` holder rev $41,595/day, circ cap $6,923,160 vs FDV ~$220M. Still not priceable.
- STONKBROKER `0xe934e36A439C94017B64a3FecE66AF12099aBF50` cap $13,861,608, holder rev $6,159/day = 3.2x, buyback 0.14% of depth/day. **The worked example of why revenue alone is useless.** Ben holds a starter.
- SHROOM `0xab093deF657F15dF31b33922A95e047aDd645B29` cap $7,785,596, holder rev $21,002/day = 1.12x, burn 1.0% of depth/day, **-65% on 7d with 6,878 buys against 8,350 sells**. On-chain burn is real: 14,182,831 to dead = 1.42% of supply. Downgraded from buy to small.
- HASHCATS/HASH `0xca75082b85bb7bec8325d513f615b16bda260020` cap $615,820, holder rev $24,307/day. **Daily mint revenue 12 Sep $430,915 -> 13 Sep $198,858 -> 14 Sep $17,269 -> 15 Sep $32,413, a 92% collapse in three days.** 194.6 ETH spent on buybacks (~$483,000 at ETH $2,483), 1.8M burned = 41.3% of supply, `totalSupply` now 2,576,797 with dead balance 0 (real `_burn`). Mint 9,109 of 16,376, epoch cost doubling, 29 ETH queue left. **The buyback is behind the price, not ahead of it. Standing lesson: on a mint-funded buyback, trailing 7d revenue is the least informative number available; read the daily series first.**
- ROUTE `0x4A72B9702f991b790788f8AFA9e7112541f4E8f8` cap $4,249,810, only 1,100,000 burned (0.11%), the 35% lands in a 2-of-3 Safe, daily revenue falling since 12 Sep ($21,471 -> $8,985 -> $7,114 -> $6,598), -43% 7d.
- INJOH `0x2cC0FAC44B8252f6B10208B091aFf2c94B4da77D` cap ~$1.49-2.01M, ChatGPT's number one pick in both runs, no adapter, 51,987,235 burned (5.20%), X feed is entirely CoinMarketCap vote-farm bots.
- Caps: GRASS `0x16391C40e85FB2246A2C8c17bfA2594C5d3EF84b` $3,779,406 - MONEY `0x0a8B4763C71aC39101b3B8a97e62Da0B81549a4f` $1,726,027 - GAGE `0x7163aE1B5AeA2f09EBc609C52b4dcAc0a7a4bC2d` $1,503,904 - HARMONIC `0xdEe52F2ab639b6942B0d0F0565400b93b7a0fbe5` $1,509,624 - CEREBRO `0xEb24a2663af4Ee979dCbfE39cB95B5F12D369c4B` $1,914,034 - MANY `0xa26992C4268A8a78a4d872FE4BDAD2Ed03aC287d` $1,141,497 - PRM `0xf24f8F6b08fE87CF062E833a732AD7F636064BC8` $7,725,794 - CASHCAT `0x020bfC650A365f8BB26819deAAbF3E21291018b4` $152,546,932 - The Index `0x56910D4409F3a0C78C64DD8D0545FF0705389870` $32,305,490.
- DexScreener search returns single-pool artifacts with v24 = $0 and liquidity at exactly half of cap for VEX, EDGE and NET. Those reads are wrong; do not quote them.

## Chain and macro regime, 15 Sep
DeFi TVL $930,551,491 (+3.6% 7d, against +115% in 33 days through August: growth decelerated hard). DEX volume 24h $1,953,855,665 (+30.4% 1d). **Chain fees $11.1M/day, down 35% from $17,070,900 on 6 Sep.** Stablecoins $1,087,952,392, +8.7% since 8 Sep, still climbing. Money arriving, take rate compressing.
**THE BINARY: the Senate cloture vote on the CLARITY Act is TODAY, 15 Sep, needs 60 votes, had not been held as of 13:37 Bangkok.** Bitcoin $77,400, down ~3% from a $79,530 overnight high. Every tokenized-equity thesis on this chain is downstream of that vote.

## Position calls
- **PAIREX `0xB8Cf4ad387cfd607C66a207CFFFC46498aACd9d6`** cap $468,704, down ~67% from the $1,433,030 launch, buys 1,109 / sells 1,656 on 24h, no adapter, 52,435,292 (5.24%) unlocks 12 Oct 2026 06:29 UTC. Sellers lead at every window and the token captures nothing contractually. **The review's fair objection: I do not know his size or entry, so name the tape and the structure, not the trade.**
- **MARSCOIN `0xFe189E97832DA1573e4e4Ff034F4fFC3a15c7777`** $87,190,394, -62% from the 5 Sep ATH of ~$227M, liq $2,146,214, v24 $4,232,092, 1h buys 80 / sells 11. Listing pump fully unwound; the author (unipcs, >2% of supply) is intact and is now also naming AI.
- **SCHIFFY `0x42aFA2124ca5a2B83898E46B2dA9a190995b1E18`** cap $2,748,436, **-8% 7d, the best relative performer in his book**, buys 1,736 / sells 1,624, supply 990,274,405, **dead balance 0 so no buyback has executed**. Gold pair on long.xyz, @Moku_HQ / @BruceMoku, a16z backing claimed by promoters and UNVERIFIED.

## Sources that failed or were not run
fomo.family thesis layer not pulled. Holder counts and holder slope not measured for any name (still the biggest gap). SLVR's 71-bettor concentration not re-measured. purealpha not checked. No founder interview searched. Prism Assets' actual revenue never checked after the identity correction. STANDARD's contract not read.

## REVISION 10, 15 Sep 18:45 Bangkok: the DELTA fee question answered on chain, and it went against my own call
- **THE MIGRATION IS ALREADY OVER, so it is not a catalyst.** Event-log counts over the same 7 days: **v3 (feeBps 100 = 1.00%) 307 logs = 0.86%; v8 (750 = 7.50%) 22,272 = 62.29%; v9 (750) 13,175 = 36.85%.** Effective take rate `1.00% x 0.0086 + 7.50% x 0.9914` = **7.44%**. In the most recent 14 hours v3 emitted **zero** logs. My r9 sentence "bills at 7.5% instead of 1% **as users migrate**" promised a step-up that does not exist. Whatever revenue is visible already has the full rate applied.
- **All three managers share one treasury `0x266311a44e70363b3d168bcb38cf38cbf2ccc51f`, owner `0xb1c2bbf86e557ecdc1812f75ae3fe973e5ec9e69`.** It holds 0.0118 native and 1,550 DELTA: it does not accumulate, it spends.
- **THE TREASURY IS NOT A CONTRACT.** `eth_getCode` returns `0xef010063c0c19a282a1b52b07dd5a65b58948a07dae32b`, an **EIP-7702 delegation on an ordinary wallet**. The key holder can remove the delegation and stop the burn in one transaction with no governance. **The burn is a policy, not a guarantee. Always read the raw bytecode of a "treasury" before calling it a contract: `0xef0100` prefix = delegated EOA, `363d3d37...` = EIP-1167 minimal proxy (immutable), short code = neither.**
- **THE BUYBACK IS REAL AND TRACEABLE.** tx `0x22ff83b4e28149b05c9235fe8a003e6bf57604ddaa29862e44681a824596e5af`: treasury sent **4.136 native** into an aggregator router, bought **984,358.84 DELTA** across three pools, routed via `0xaa61254627b7392b0bc922097b10eb0587db2be7` (17,562-byte executor, holds nothing between runs), and burned it in the same transaction. Cash leaving, tokens destroyed.
- **COMPLETE BURN HISTORY, reconciles to the dead balance to the token:** 03 Sep 03:02 27.14 and 03 Sep 05:27 26.23 (dust, different sender) - **08 Sep 05:15 2,414,691.72** - **11 Sep 04:59 912,096.61 (+2.99d)** - **13 Sep 08:14 993,449.03 (+2.14d)** = **4,320,290.72 total**. **Cadence is every 2 to 3 days, NOT weekly** (r9 said weekly, wrong). Rate since the first burn: 252,658 DELTA/day = **$2,799/day**.
- **THE CALL THIS KILLED: DELTA's buyback is 0.286% of pool depth per day. STONKBROKER's holder bid is 0.262%.** I dismissed STONKBROKER at that level as mechanically unable to move price, then promoted DELTA on the same metric at the same value. **A metric cannot reject one name and promote another at the same reading.** (The 0.14% recorded above for STONKBROKER was measured against a different depth figure; at 15 Sep depth it is 0.262%.) DELTA demoted to a momentum row. $2,799/day is 0.144% of its $1,938,613 daily volume, one part in six hundred.
- **BOW ROW WITHDRAWN, AND THE CALL WAS WRONG INSIDE A DAY.** I made it highest entry priority on "-6.00% on the week while the index fell 19.17%". It is now **-30.10% in 24h and -25.00% in 3 days**. **LESSON: one week of relative strength on this chain is not evidence and must never carry a priority call on its own.**
- **TIMESTAMP ERROR, SECOND TIME.** r9's header read "built 20:05 Bangkok" when it was about 19:30. Header and footer now carry the real publication minute only.

## RHX at 18:00 Bangkok, 15 Sep (supersedes the 17:00 block above)
Level **1,473.53**, HOLD ZONE, heat **31.40**, recovery **0 of 3**, 7d **-20.97%**, 24h **+6.38%**, **-38.74%** from the 7 Sep peak. Legs: 3d trend -4.36% (fail), participation 30% / 6 of 20 (fail), drawdown -38.74% vs -30% (fail). Event-low variant **+7.38%** off the 14 Sep episode low against the live baseline's misleading **+123.50%** off a 19 Aug low that predates the selloff. Fixed cohorts: MEME basket -16.07% (median -20.22%, win 30%, ex-top -19.80%, top FRONG +54.96%); INFRA -22.36% (median -22.12%, win 11%, ex-top -27.71%, top NET +20.47%); **STOCKMEME -28.68% (median -53.99%, win 20%, ex-top -42.90%, top ASTEROID +241.56%)**.
**ASTEROID ARITHMETIC, both figures correct and different:** direct contribution `+241.56% / 20 = +12.08 points`; removal-and-reweighting `basket -28.68% minus ex_top -42.90% = +14.22 points` (larger because dropping a member reweights the rest from 1/20 to 1/19). The review's 11.98 was right for the 17:00 snapshot; ASTEROID moved from +197.7% to +241.56% in that hour. **Rotating index +19.69% vs fixed basket -28.68% is a 48.37-point gap; ASTEROID explains 14.22 of it and the remaining 34.15 is simply that they are different portfolios.**

## THE r10 TRADE ROWS, each with a stated exit (levels are daily closes, prices 18:37 Bangkok)
- **AI `0x2e8c31162b855a2ffa90f6f8634643ad6f111e18` $0.3029, cap $299,729,278 . THE ONLY BUY.** Only name on the chain with an intact trend: **15.2% below its 32-day high** and higher daily lows every day since 10 Sep (0.1613, 0.1988, 0.2510, 0.2385, 0.2264, 0.2662). 7d +34.0%, 24h +22.5%, v24 $32,043,191. **Entry: daily close above $0.357105** (12 Sep high, +17.9%). **Invalidation: daily close below $0.226352** (14 Sep low). **Take profit: thirds at $0.606 and $0.909.** 2x needs ~$1,931,583 net buy. Token is an **EIP-1167 minimal proxy onto `0x3be8b97fd0e713b5abe0649fa830223b6b4bc599`, so the logic is immutable**. Reassess 29 Sep.
- **PARE `0x15d36b6a28d8327abc7afabf0f106ae2c9af5c4d` $0.01179 . HE HOLDS IT, KEEP WITH A STOP.** 3d +40.6%, 7d +35.8%. No addition here (-22.5% today, 1h 14 buys / 14 sells). **Add only above $0.0156906. Invalidation $0.00909561. Scale $0.0156906 then $0.0272712.** The 7 Sep wick to $0.0272712 closing at $0.0119108 is the overhead supply.
- **DELTA $0.01108 . MOMENTUM ROW ONLY.** 24h +41.2%, 3d +7.0%, 70.9% below its 32-day high. **Entry: close above $0.0165283, or a pullback holding $0.00900981. Invalidation: close below $0.00657609, or the 7702 delegation changing, or a burn window passing empty. Reassess at the next burn, due inside 24 hours.** Impact at $25,000 is 5.87%.
- **PONS $0.6319, cap $435,140,666 . WAIT.** Cleanest bid on the board (1h **161 buys / 8 sells**) but **a 2x needs $5,686,299 and a 3x needs $10,049,549**. Slowest horse for a two-month window. Entry: recovery 2 of 3 **and** a close above $0.841549. Invalidation $0.487714.
- **CME `0xe2324FF2a59F8eCBa8c321c6466e59121C00e795` $0.005337 . WATCH.** 7d +167.9%, 2x needs only $91,308, but eight daily candles is a sample not a trend. Entry above $0.0060625 plus a second week. Invalidation $0.00273494.
- **NO: BOW (withdrawn), STONKBROKER (1h 3 buys / 106 sells, supply 2,388,325,344 so it mints), HOOKR (-49.4% 7d), SHROOM (82.7% below its high; today's +82% started at $0.0036119), PRM (-26.07% on the day after +75.8% on the week).**

## CATALYST CALENDAR, verified against sources 15 Sep, Bangkok time
**Next DELTA burn, inside 24 hours** (2.42 days elapsed on a 2.14-2.99 day cadence) . **16 Sep 01:15 CLARITY Act cloture** (14:15 US Eastern, procedural motion to proceed to H.R. 3633, 60 votes needed against 53 Republican seats, NOT final passage) . **17 Sep 01:00 FOMC** (16 Sep 14:00 ET, with SEP and dot plot) . **29 Oct 01:00 FOMC** (28 Oct 14:00 ET) . US midterms in November are why the CLARITY timeline is tight.

## REVISION 11, 15 Sep 20:20 Bangkok. SEVEN OF MY OWN ERRORS, AND THE r10 FEE ANSWER WAS MEASURED IN THE WRONG UNIT
- **EVENT COUNTS ARE NOT FEE DOLLARS. The "7.44% effective take rate" above is WITHDRAWN.** Measured in USD delivered to the shared treasury over the same 7 days: **v3 $0.00 (0 transfers, 0.00%); v8 $9,654.68 (501 transfers, 98.29%); v9 $167.59 (36 transfers, 1.71%)**. Against event-log shares of 0.86% / 62.29% / 36.85%. **v9 emits 36.85% of the logs and delivers 1.71% of the money.** A realized take rate also needs GROSS CLAIMED FEES as the denominator, which I do not have, so quote no take rate at all. **STANDING RULE: never weight a revenue share by event counts. Group value transfers by sender and price them.**
- Direction survives and is now stated properly: the 1% manager delivered **zero dollars** in seven days. Floor only: 141 v8 and 6 v9 transfers were in unpriceable tokens, and **fees arriving as native currency emit no Transfer log at all**, so an ERC20 scan cannot see them. Only **31.2%** of treasury inbound comes from a manager; the largest single inflow ($20,754) is the buyback executor returning bought tokens.
- **"BOUGHT AND BURNED IN THE SAME TRANSACTION" WAS FALSE.** Two transactions, 25 seconds apart, and the burn was **9,090.19 tokens LARGER** than the buy, so some came from existing inventory. Buy `0x22ff83b4e28149b05c9235fe8a003e6bf57604ddaa29862e44681a824596e5af`, burn `0x746c581214322de764657d9f95e8c2fa3227137c53c1a39d426262486d451862`. Earlier burn tx `0xcde52035af7eeff0165a4d2f1d6bdf19781b51c93133328246b547dacf5ea7d1`, its buy `0x1992e3ee28041a3fa77441aca301482d71c4d6a356d6d47433d80acfc027823b`.
- **THE $2,799/DAY BUYBACK FIGURE WAS AN UNDERSTATEMENT. Real cash: 4.240 native (11 Sep) + 4.136 native (13 Sep) = 8.376 total, about $20,768 at $2,479.49. Over the 2.14 days between them that is ~$4,791/day.** I had valued burned tokens at today's price over a longer window. **The 8 Sep burn of 2,414,691 DELTA has NO purchase transaction I could find; its funding is unreconciled.**
- **THE GAP THAT MATTERS FOR A TWO-MONTH HOLD: fee dollars in ~$1,392/day, buyback cash out ~$4,791/day. The buyback spends ~3.4x the income I can see.** Either fees arrive in native currency, or in unpriceable tokens, or it is drawing a reserve. **A buyback funded from a reserve stops when the reserve does, and I cannot tell which it is.**
- **"NOT ONE COHORT KEEPS A LEAD WITHOUT ITS BEST COIN" WAS FALSE.** At 19:00 MEME ex-top **-20.27% vs RHX -20.97% = relative +0.88%** (it was +1.47% at 18:00). The number was in my own pull. **INFRA is the mirror case: basket -15.96% (rel +6.34%) collapses to ex-top -22.95% (rel -2.50%); NET at +39.9% carries the whole group.** Relative figures are RATIOS `(1+sector)/(1+RHX)-1`, never percentage points subtracted.
- **AI'S "HIGHER DAILY LOWS EVERY DAY" WAS FALSE, contradicted by my own table.** Real sequence `0.161329, 0.198815, 0.251020, 0.238471, 0.226352, 0.266221`: two consecutive LOWER lows on 13 and 14 Sep, then today's is the highest of the run. True claim is "higher-low structure with a two-day pullback inside it". **I also compared AI's 32-day drawdown against the cohort's 7-day return, which are different measurements and not a valid comparison.** Matched: AI +33.0% day, ~+34% week vs RHX -20.97%.
- **RISK WAS MEASURED FROM SPOT INSTEAD OF FROM THE ENTRY TRIGGER, which flattered every row.** AI's stop is **-36.6% from entry**, not -25.3%; the "+100%" target was **+69.7%**. **STANDING RULE: on a trigger entry, price the stop and every target from the TRIGGER, never from the screen price.**
- **TWO ROWS HAD A PROFIT LEVEL AT OR BELOW THEIR OWN ENTRY.** DELTA's breakout trigger $0.0165283 equalled its first target, and its structural stop sat 60.2% below that entry, so **the breakout branch is deleted and only the pullback branch survives**. PARE's add trigger and first target were both $0.0156906.
- **PARE IS NOT PAIREX AND I WROTE "YOU ALREADY HOLD THIS" ON THE WRONG ONE.** Ben holds **PAIREX `0xB8Cf4ad387cfd607C66a207CFFFC46498aACd9d6`, cap $435,925**. **PARE is PareStocks `0x15d36b6a28d8327abc7afabf0f106ae2c9af5c4d`, cap $11,908,069**, a separate project with its own file [[pare-stocks]] vs [[pairex-token]]. **GATE: a ticker that resembles another name is not a match. Confirm chain and contract for every holding before writing an action on it.**
- **"NEXT BURN DUE INSIDE 24 HOURS" WITHDRAWN.** Two gaps (2.99d, 2.14d) are two observations, not a schedule. It is a watch, not a deadline, and nothing gets closed because an inferred date passes. **A bigger burn can mean a cheaper token or a larger batch: compare CASH SPENT, not tokens destroyed.**
- I called Bitget and FedRateCalc primary sources. They are secondary. Primary: dailypress.senate.gov and federalreserve.gov/monetarypolicy/fomccalendars.htm. FOMC also 9 Dec 14:00 ET = 10 Dec 02:00 Bangkok.

## BEN'S EIGHT HOLDINGS, stated 15 Sep 2026 (supersedes the three-name list at the top)
- [stated] He holds **Tibbir, MOO, Marscoin, DELTA, SCHIFFY, Premium/PRM, PAIREX and GASOLINU**. He did NOT give quantities, entries, sizes or exit rules. Do not invent them.
- [stated] **He wants the dashboard minimal and easy to scan: sell/hold/buy first, sector performance clearly, then actions for holdings and possible new buys.** Sector strength SUPPORTS a coin's case; **a weak sector is NOT an exclusion criterion for an individually strong coin.**
- **TIBBIR CHAIN IS UNCONFIRMED and must be asked, not guessed.** Solana `998Vi32HYF7fcC72GTg9nkRGdGf5FcoSaNF8SCBs4GUa` $274,299,018 - Ethereum `0xBEA37662BF8846674Ae12d1b4a2a7c4868c044Ef` $153,370,303 - Robinhood `0xa944c6aEE0aBa6cc7345287F37c7DAA8075c8dCb` $185,364,015 with $92,682,007 liquidity and **ZERO 24h volume on one pair**, which reads as a bridged listing nobody trades.
- **MARSCOIN is on BNB CHAIN, not Robinhood.** Nothing on this board applies to it. `0xFe189E97832DA1573e4e4Ff034F4fFC3a15c7777`.
- **GASOLINU `0x1e6EA1e89151cDc8443968Bf047cfa3177181e18`** (Robinhood), cap $2,396,524, liq $495,911 across 3 pairs. A BSC impostor GASOLINU `0xd6f7413Fe7b15c31c2385aaD3D13393BaF877777` exists at $4,005 cap.

## PRICES 20:05 Bangkok, 15 Sep (chain-wide bounce accelerated through the evening)
SHROOM +78.5% day (1h 225 buys/119 sells) - **DELTA $0.01206 +69.2% (1h 121/77)** - **AI $0.3192 +33.0%, 1h 282 buys/48 sells, the firmest bid it has shown** - MOO $0.01795 +27.5% - **PONS $0.6817 +26.1%, 1h 786 buys/119 sells** - SCHIFFY $0.003459 +20.4% - CME $0.005122 +5.0% - MARSCOIN $0.09407 +4.4% (1h 55/25) - PARE $0.0119 +3.4% - GASOLINU $0.002396 -0.7% (6h +27.4%) - HOOKR -3.1% - **PAIREX $0.0004509 -8.7%, 1h 20 buys/30 sells, the only holding red on every window** - STONKBROKER -15.6% - BOW -18.5% - PRM $0.00839 -27.9%.

## RHX at 19:00 Bangkok (supersedes 18:00). THE LABEL FLIPPED TWICE IN THREE HOURS AND THE EVIDENCE NEVER MOVED
**BUY ZONE**, level 1,475.04, heat **19.88** (was 31.40 an hour earlier), 7d -20.97%, 24h +7.36%, **-38.68%** from peak, +7.49% off the event low. **Recovery still 0 of 3** and **participation FELL from 30% to 25%** while the label was turning friendlier. Journal now 5 rows / 4 observations / 2 builds / **2 zone flips** (BUY -> HOLD -> BUY, 17:00 to 19:00). Legs: 3d trend -4.48% fail, participation 0.25 fail, off_low fails on the -30% drawdown condition either way.
Cohorts at 19:00: **INFRA basket -15.96% (rel +6.34%), ex-top -22.95% (rel -2.50%), win 22%, top NET +39.9%** - **MEME basket -17.23% (rel +4.74%), ex-top -20.27% (rel +0.88%), win 30%, top IF +40.6%** - **STOCKMEME basket -29.87% (rel -11.26%), ex-top -44.91% (rel -30.29%), win 15%, top ASTEROID +255.9%**.
Lane coverage, so lane averages are never read as basket results: **AIPAIR 2 of 7 have a full week**, STOCKMEME 27 of 32, MEME 104 of 113, PAD 5 of 5, INFRA 10 of 10.
Feed now publishes BOTH revision denominators: **52.4% of member-hours revised over all stored history** (median 1.52% of the revised subset, 0.04% across all member-hours, audit window 83 members / 28 Jul to now / 54,226 member-hours) and **43.9% since the last rebalance**. Quote the window with the number.

## THE STRUCTURAL FINDING OF r11
**Every honest stop on this chain right now is 27% to 55% wide**, because the index is 38.68% below its peak and the median stock-meme is down 54%, so the nearest real structure under any candidate is a long way down. **That is what "wait" means mechanically: waiting for recovery evidence is what shrinks the stop, not timidity.** Reward-to-risk stays fine (2.5:1 to 6.3:1); the risk itself is enormous.

## r11 TRADE ROWS, priced FROM ENTRY (20:05 Bangkok)
- **AI, conditional buy.** Now $0.3192. Entry **$0.357105** (+11.9% away). Stop **$0.226352 = -36.6% from entry**. Targets **$0.714210 (+100%, 2.7:1)** and **$1.071315 (+200%, 5.5:1)**. Not "the one buy on the chain": that needs a complete screen and I ran ten names.
- **PARE, candidate NOT a holding.** Now $0.0119. Entry **$0.0156906** (+31.9%). Stop **$0.0111291 = -29.1%**. Target **$0.0272712 (+73.8%, 2.5:1)**.
- **DELTA, pullback branch only.** Entry zone **$0.00901 to $0.01000** (below price now). Stop **$0.00657609 = -27.0%**. Target **$0.0165283 (+83.4%, 3.1:1)**. Breakout branch deleted.
- **PONS, wait for the gate.** Entry **$0.841549 AND recovery 2 of 3**. Stop **$0.582318 = -30.8%**. Target **$1.683098 (+100%, 3.2:1)**.
- **CME, too young.** Entry **$0.0060625 plus a second week of data**. Stop **$0.00388906 = -35.9%**. Target **$0.0196965 (+224.9%, 6.3:1)**.

## THE TOY-MODEL CAVEAT, now stated on the page
`net buy for Nx = (L/2)(sqrt(N)-1)` is exact ONLY for a single equal-value full-range constant-product pool with a fixed-dollar quote, no fees, no other flow and unchanged liquidity. **None of that holds here**: tokenised stock quotes move, ranges are concentrated, hooks and aggregators route around it. **Use it for ORDERING candidates, never as a price target.** Daily volume is not net new money. PONS, SHROOM and AI hit the DexScreener 30-pair cap so their liquidity is a floor and their 2x figures are overstated.