---
name: acre-lending
description: Deep dive on Acre (ACRE), the session-aware RWA lending protocol on Robinhood Chain, and how its published TVL compares to what is actually on chain. Read whenever ACRE, useacre.xyz, @useAcre or tokenized-equity lending comes up.
sources: [cowork]
aliases: [ACRE, Acre, useAcre, useacre.xyz]
---
Related: [[longbow-bow]], [[crypto-token-research]]

- Deep dive 30 Aug 2026. ACRE 0xd83DE980dB4e1118bF6B14da77B6fC4c15d7D6bA on Robinhood Chain. Ben first pasted the same address ending in "bd" which is empty on 19 chains; the live token ends in "bA". Contract is 3,248 bytes, the stock PonsV2LauncherToken, same as DTF, MAST and BOW: fixed 1B supply, no mint, no deployer privileges. X @useAcre joined Aug 2026, 306 followers, site useacre.xyz
- Market 30 Aug: about $0.000547, mcap $510k to $545k, aggregate liquidity $69k, 24h volume $797k, +419% on 24h. First pool 28 Aug 17:20 UTC so about 36 hours old. ATH $0.000812 = $0.81M mcap at 30 Aug 00:30 UTC, now roughly 33% off. 6h flow 270 buys against 404 sells. 572 holders on Fomo, top 10 hold 33.53%
- THE PRODUCT IDEA IS GENUINELY ORIGINAL and worth remembering independently of the token: tokenized equities trade 24/7 but the underlying securities do not, so for about sixty hours every weekend there is no market to liquidate into. Acre prices that gap. Borrow capacity contracts across Friday and the weekend, tightens around a name's earnings date, and scores a basket of correlated technology names as one exposure rather than six. Liquidation is partial, paced and auctioned, and there is a "deleverage into the close" buffer where you pre-fund a target and anyone can reduce your debt before the market shuts. Nobody else on the chain prices session risk
- BURN CLAIM IS TRUE AND UNDERSTATED. They claimed about 5.71% burnt; 67,945,578 ACRE, 6.795% of supply, is verifiably sitting at 0x000000000000000000000000000000000000dEaD, worth about $37k at spot. The cited tx 0x1f4914a67e46ba4da57f593f06032c505c477f63f74b37ceb889df55c82effb2 is real, 3,394,234 ACRE sent to dead by 0x2bac5d261b6b265db14be80b0179e2da21f9ab48. Note totalSupply still reads exactly 1B because they transfer to dead rather than calling burn, so DexScreener's market cap is computed on the full 1B
- CORRECTION 30 Aug, second pass. I first reported the team as claiming "$40,000 TVL" and called it false by 1,400x. THAT WAS MY ERROR. The full tweet is "Total TVL has crossed $40,000. Equating to ~6.5% of supply burnt forever." They are describing the dollar value of the BURN, not deposits, and 6.5% of 1B at the then price of about $0.000615 is indeed about $40k. Misusing "TVL" to mean value burned is sloppy, but it is not a fabricated deposit figure. They have never published a deposit number, and they publicly disclosed the caps as deliberately small
- REAL LENDING STATE, confirmed twice, by balanceOf and by the pools' own accounting (totalSupplyAssets 0x94877fc4, totalBorrowAssets 0x1aad44c2, utilization 0xea21cd92): CURRENT pool 0x82bD35a34891F1D8B543313101C06CF31b8daD1E holds 2.621794 USDG, LEGACY pool 0x837f9568b7Cdb95D4188122Bb45971Cb28311449 holds 25.000000 USDG. Total supplied 27.62 USDG, total borrowed 0.00, utilization 0%, supplyCap 50,000, borrowCap 30,000, neither paused. My first pass said the current pool was empty and legacy held all 27.62; that split was wrong, caused by an RPC returning silent zeros. Always cross-check balanceOf against the contract's own accounting
- Their own app confirms it independently: useacre.xyz/app reads TOTAL SUPPLIED 0 USDG, headroom 50,000 of 50,000, UTILISATION 0.0%, AVAILABLE IN POOL 0, SUPPLY APR 0.00%, while the live market clock and 1.61% borrow APR show the page is reading the chain fine. So the protocol is deployed, functional and completely unused
- Method note worth reusing: useacre.xyz/acre-config.js publishes every contract address in plain text with comments, including a legacyDeployments array. Reading a project's own config JS is faster than any explorer for finding the real pool
- Contracts are real and substantial (pool 21.5k bytes, plus riskEngine, oracle, calendar, timelock, outcomePool), so this is a built protocol with no users, not vapour. But no established smart money is in it: no unipcs, no bluntz, no frankdegods on Fomo, largest position there is $19,560
- The dominant holder thesis, "first ever lending protocol for tokenized equities" and "AAVE of RH", is FALSE. Longbow shipped the same product on 11 Aug and runs 49 markets; Morpho Blue has $486.97M supplied and $414.66M borrowed on Robinhood Chain; Arrow Finance @arrowfinanceio is a $7.5M mcap lending suite in development. Acre is the fourth entrant, not the first
- Same phishing operator as DTF, MAST and BOW: multiple accounts pushing fake "claim portal" and "eligibility" links on crypto.cryptoloc.lol using the real ACRE contract address
- Sharpest public bear point came from @drakeondigital: "There's not that much TVL on the protocol for that much in buybacks." He is right and the resolution is that the burns are funded by ACRE token trading fees through the Pons launcher, not by lending revenue, so the buyback stops when volume stops
- CORRECTION 2: the ACRE token is NOT source-verified on Blockscout, and neither is the pool. I asserted it was a verified PonsV2LauncherToken. Proved it properly instead by bytecode: ACRE's 3,248-byte runtime is identical to the VERIFIED BOW launcher token except three 20-byte regions (60 bytes, 1.85%), which are the inlined immutables. Those regions reveal deployer 0x2bac5d261b6b265db14be80b0179e2da21f9ab48 and bonding curve 0xe3f915ce2dcb4639b1c3d073e367e356cafea567. Reusable trick: to identify an unverified Pons token, diff its runtime against a verified sibling; only the immutable address slots should differ
- CORRECTION 3: I described "deleverage into the close" and the auto-deleverage buffer as shipped. The team's own post says "Designed and specified. Being built next." The app renders the panel but the feature is not live. Also still pending by their own admission: a live USDG price feed to replace a fixed-dollar adapter, and Chainlink sequencer-uptime checks. And the pool contract is unverified, so the risk engine and liquidation logic cannot be audited by anyone
- BUYBACK IS GENUINE, verified end to end. Burn wallet 0x2bac5d261b6b265db14be80b0179e2da21f9ab48 received 67,944,290 ACRE in 14 transfers and sent 67,944,290 out in 14 transfers, every one to 0xdead. It holds zero ACRE now, so there is no team bag parked there and nothing was sold. Sources of the inbound: 35.97M from the bonding curve, 17.78M from the Uniswap v4 PoolManager, 10.80M from the UniversalRouter, 3.39M from a Relay router. It spent 0 USDG and 0 WETH and holds 6.84 ETH, so the buys were made with native ETH from creator fees. Team states the model openly: "75% of fees are used for buy back and burns, the final 25% for tech improvements", funded by "token fees". Fomo now displays supply as 932M, netting the burn
- EARLY HOLDERS, the new work. In the first 60,000 blocks (about 100 minutes from the first transfer at block 48568046, 28 Aug 20:01:33 UTC), 63 wallets bought 284,070,384 ACRE = 28.41% of supply. They still hold 112,355,360 = 11.24%, so ABOUT 60% OF EARLY TOKENS HAVE BEEN SOLD. 12 still hold, 11 trimmed, 40 sold out completely. The single largest early buyer 0x8f10b468b06c6fd214b65f87778827f7d113f996 took 79.8M (7.98% of supply) and holds zero today, though it is a 13,529-byte contract with nonce 1 so it may itself be an aggregator. The first ten minutes were clean: only 52 transfers and no bundle of equal-sized snipes
- ROUTER TRAP, cost me a wrong answer mid-analysis and matches the standing rule in [[crypto-token-research]]: 0x39b38686a19836ac10162c490e4558e120cbbe5f looked like the biggest early buyer at 84.6M tokens, but it has nonce 1, zero ETH, and cycled 1.88 BILLION ACRE through 1,892 inbound and 1,928 outbound transfers. It is Relay aggregator infrastructure, alongside 0xb92fe925dc43a0ecde6c8b1a2709c170ec4fff4f (RelayRouterV3) and 0x039ec98a76f111092d4751365ff09dd2aec301e8. Filter any wallet whose transfer count dwarfs its balance and whose nonce is 0 or 1 before calling it a buyer
- Holder structure 30 Aug: 647 holders on Blockscout, 566 on Fomo. Top 50 hold 78.6%, but 21.33% of that is infrastructure (V2LaunchLocker 8.163%, dead address 6.795%, Uniswap PoolManager 6.376%). Largest genuine holder is 0x1b3E8118fA81d47d00ecCA997eE161b094a37d9E at 3.586%, which is also an early buyer that grew its position from 6.7M to 35.9M. Unlike DTF, no bonding curve overhang remains
## Full holder-lineage sweep, 30 Aug 07:00 UTC (Ben: "find stealth big players")
- Method: every ACRE Transfer log over the token's whole life, 23,780 events, reconciles exactly to 1,000,000,000 minted. Ledger of acquired / disposed / held per wallet, routers netted out.
- CORRECTION to the earlier note above: the first ACRE Transfer is block 48471183, 28 Aug 17:18:41 UTC, NOT block 48568046 / 20:01:33. The earlier figure was 2h43m late and understated the early cohort. Any "first 60,000 blocks" claim built on it is wrong.
- Cohorts: 2,277 wallets ever touched ACRE. 560 real holders (78.03% of supply). 271 never sold and never sent anything out, holding 41.49%. 289 trimmed but still hold, 36.54%. 1,635 wallets bought and are now flat. Retention is 25.5%.
- Watchlist cross-reference, 30 wallets: only TWO ever touched ACRE. Ghost 0x4123ffc81579e38dfff74b75d8bdf21c41e62237 holds 2,336,795 (0.23%, never sold, cost about $237). DineroDom 0xdecba5d411253f488b49bfc0d581afe6428a0d2a holds 7,044,953 (0.70%, never sold, entered at $300k mcap, about $2,112). Nobody from Capture Board, the swing traders, kioto, Ansem RH or either position-trader archetype is in it. The one clean multi-token winner 0x711281c1b26aaed86e40e4caaf76c1962b45e161 has no ACRE at all.
- NO WHALE EXISTS TO FIND. The largest never-sold position on the chain is 32,542,867 ACRE, about $11,894 at spot. Next best are $6,412, $5,588, $4,599, $4,366. Total mcap is only $365k, so 1% of supply is $3,655. A "big player" in this name would be a $12k position.
- 7702 SIGNATURE, new and reusable. Robinhood Chain user wallets are EIP-7702 delegated EOAs: eth_getCode returns exactly 23 bytes (0xef0100 + implementation), almost all pointing at 0xe6cae83bde06e4c305530e199d7217f42808555b, with nonce 1 and zero native ETH because the app sponsors gas. A 23-byte code length means retail app user, not a contract. A plain 0-byte EOA with a real nonce and real ETH is self-custody or a bot. Every one of ACRE's largest early holders is the 7702 retail type; the funded self-custody wallets all arrived LATE, on 29 and 30 Aug.
- THE SELF-SIGNER TEST BREAKS ON THIS CHAIN. 63.7% of the top-200 holders' inbound is in transactions signed by someone else, which the [[rh-early-cohort-overlap]] rule would flag as airdropped. It is not. Max distinct ACRE recipients in any single tx is 6, so no multisend exists. Sampling the biggest holder's inbound tx proves it: signer 0x56c262027e0de4aea31d2489529cb25d23e58a8b is a Relay solver filling a cross-chain buy, delivering 10,471,000 ACRE to the user. On RH chain "not self-signed" means gas-sponsored or Relay-bridged, both real buys. Use the recipient-count-per-tx test instead.
- COORDINATED CLUSTER A, the only genuine multi-wallet operator found. 0xf17861250703231dacb1119c4a87324ef950c97d, 0x429fa4e0282278348d26ecbc235f0ef475e0e590, 0xf9c13756b50132623f2529da8f4478359c0f6e0e all entered in the SAME block 49390319, sizes 3.61M / 3.58M / 3.45M, nonces 224 / 224 / 225, RH-chain ETH 9.98 / 10.36 / 11.48, Base nonce 9 / 9 / 9, all never sold. Same operator, three wallets. Combined 1.06% of supply, worth about $3,900. Real but tiny.
- CLUSTER B, the launch snipe. Five plain EOAs (0x5ab5362e8fe0599ed209c2b622e06cd2ded266e6, 0x244e5d22e518530395deafcde0932f567159cca3, 0xc658de63f04af9f7846080f858eaa28059ed13a6, 0x9f1cf1c89e15881ca35694ff738b1bbc070f8019, 0xccbd5d905df8d82b58809a5ee2f4d2398084ff4b) each took 10.3M to 10.7M in block 48471354, about 30 seconds after launch, each sold about 8M, each still holds about 2.4M. Uniform sizing and uniform 77% exit = one sniper operator with five bots. Every other same-block launch cluster is now at zero.
- Cross-chain probe of the top 60 holders on Ethereum, Base and Arbitrum: only 5 of 60 are RH-native. The genuinely funded ones, all never-sold, are 0xf281634df5e7a2178f2cdde1bcc344f1dca46295 (21.73 ETH on mainnet), 0x08a16898f9b85308d1316984c1804abc663d86c6 (ETH nonce 53 + Base nonce 182 + RH nonce 1802, the only real multi-chain trader in the book, entered at $101k mcap, up 2.85x), 0x37282d24f576d8ff4be5171c92c87fb193ee217f (16.89 ETH on RH) and 0x1bfe6e6891202302c62ed9a02e19e7920132b0eb (20.0 ETH on RH). Note the biggest holder 0x1b3e8118 has ETH nonce 0 and zero balance everywhere: a throwaway, not a whale.
- Traded-token overlap with the known winners is style overlap, not connection: best is 0x5b0817326f88ad99118590fb0b758196d692c313 at 27 shared names with 0x711281c1, in a universe where both trade 100+ RH memecoins.
- Price at the time of this sweep: $0.0003655, mcap $365,488, deepest pool ACRE/ETH $57,599 liquidity. ATH $0.000812 = $811,965 mcap set 30 Aug 00:35 UTC, so 55% off an ATH that is only SIX HOURS OLD with no base. 1h 20 buys / 34 sells, 6h 165 buys / 283 sells, price -32% on the hour.

- Market 30 Aug 06:00 UTC: $0.000583, mcap $543k, aggregate liquidity $71.2k, 24h volume $797k, +560% on 24h, ATH $0.000812 = $0.81M at 30 Aug 00:00 so about 29% off. 6h flow 263 buys against 397 sells, ratio 0.66. Fomo still shows no unipcs, no bluntz, no frankdegods; largest Fomo position is about $20.9k. Team promised "major updates, reveals and more over the next 24 hours" and had posted nothing new five hours later

## Gem-hunter run, 30 Aug 2026 09:21 UTC (second screen same day, opportunity-side)
- TOOLING GAP, new: robinhoodchain.blockscout.com now returns Cloudflare 403 "Just a moment" to the container (it worked at 07:00). Holder data was rebuilt from raw eth_getLogs Transfer sweep instead. Also note the RPC rejects urllib's default user-agent with 403 but accepts curl's, so set user-agent on any python RPC client.
- Full log sweep: 24,657 Transfer events, first block 48471183 = 28 Aug 17:18:41 UTC, reconciles exactly to 1,000,000,000 minted.
- HOLDER SLOPE, the decisive number. 29 Aug 05:18 = 369 holders / 69.54% of supply held. 29 Aug 17:18 = 540 / 78.08%. 30 Aug 05:18 = 568 / 77.91%. 30 Aug 09:21 = 569 / 76.09%. So +171, then +28, then +1. Growth has flatlined while supply held FELL 1.82 points, and the Uniswap v4 PoolManager balance rose 6.217% to 7.683%, meaning 14.66M ACRE was sold into the pool in that window. Existing holders trimming, nobody replacing them. Retention 24.5% (2,318 wallets ever held, 569 now). Never-sold cohort 282 wallets holding 37.12%, down from 41.49% at 07:00
- CORRECTION to CORRECTION 3 above: auto-deleverage SHIPPED. useacre.xyz/acre-config.js says the 29 August 2026 deployment "adds auto-deleverage, defined outcome pools, the risk tier listing framework with per asset exposure caps, and five more collateral assets". The earlier note calling it "designed and specified, being built next" is stale by one deployment
- NEW CONTRACT, deployed 30 Aug: supplyRebate 0x4Ac5031aE0c91c70eC1310E38573757487e6D0B7, "Returns part of the protocol's reserve factor to suppliers holding ACRE", tiers at 250k, 1M, 5M and 25M ACRE. This is a genuine holder-directed payout mechanism and it is the first one the token has had. It holds 200 USDG, all team-funded, and pays from borrow interest which is currently zero
- Full contract set with code sizes: pool 0x82bD35a34891F1D8B543313101C06CF31b8daD1E 21,506 bytes; riskEngine 0xdd7eF595DEea942D05Fd974E6946931ea0DC28A4 9,651; oracle 0xaB1C64921dE1928a5C3Ee06bc37fAEa7b5C9D703 6,490; calendar 0x351866f5f038d7233d3fb14C7dd822F62C0bC881 4,889; timelock 0xe19f3b52922C126781E0e47028D3d2b2411a271d 2,897; outcomePool 0xdA9E51d56d5F1cBC1d9b3FA69658b9d7F37bA4a1 10,293; supplyRebate 6,475; loanTokenFeed 0x32e20B845A6EC481F919fEDC26FCEa60b803167C 1,803; legacy pool 0x837f9568b7Cdb95D4188122Bb45971Cb28311449 17,194. Config states contracts are immutable by design, a new version means new addresses, "nobody can replace the logic holding your funds, including us"
- THE TVL IS THE TEAM'S OWN MONEY. Current pool supplied went 2.62 to 1,002.62 USDG because the fee/burn wallet 0x2bac5d261b6b265db14be80b0179e2da21f9ab48 deposited exactly 1,000 USDG at 30 Aug 07:48:18 UTC, tx 0x667d14c89b241716dd8c175e47db2efeeb0fe16723cebb56b0badcdb4a56e4d5. Only other deposit is 2.62 USDG from 0xdba091aa2e2054de473d10a6683081b89e000fe6. Real third-party capital in the whole protocol is 27.62 USDG including the 25 in the legacy pool. Borrowed 0.00, utilisation 0%
- ZERO COLLATERAL HAS EVER BEEN POSTED. Checked balances of tokenised NVDA 0xd0601CE157Db5bdC3162BbaC2a2C8aF5320D9EEC, TSLA 0x322F0929c4625eD5bAd873c95208D54E1c003b2d and SPCX 0xC6E696cf4d6590BD844F7aa00d54518e8fF2BE9f across all nine Acre contracts plus the fee wallet: every balance is zero except 0.0001 TSLA in the legacy pool. The "earn tokenised stocks" claim has no on-chain instance
- THE WEEKEND TEST IS RUNNING RIGHT NOW AND IT IS FAILING. US market shut Friday 28 Aug, reopens Monday 31 Aug 13:30 UTC. Acre launched three hours before that close. The sixty-hour session gap Acre was built to price has been live for its entire existence and borrows are zero
- BURN IS DECELERATING, measurable. 71,476,033 ACRE at dead = 7.148% of supply across 19 transactions, but 53.75M of that was hour zero dev-token burn. Organic buyback since = 17.72M over 40 hours. Last three: 30 Aug 06:16:13 = 2,069,308, 07:13:49 = 828,171, 08:09:32 = 632,200. Fee wallet holds 7.2374 ETH (about $17,773, 5.5% of mcap) and zero ACRE, so the war chest is real and nothing is parked, but it just spent 1,000 USDG on seeding its own pool instead of burning
- TICKER COLLISION, narrative-ownership kill. LunarCrush $acre resolves to Ares Commercial Real Estate, a $258M NYSE REIT, not this token. Its entire social series is the stock. @useAcre shows 138 followers there. A token whose ticker is owned by a listed equity cannot own its own search results, so LunarCrush is unusable for this name and only contract-address search works
- Market 30 Aug 09:21: $0.0003492, mcap $324,330, aggregate liquidity $55,441 across 11 pairs but the ETH pair 0xed2a0417e3b96c759ffd3e8cb32fd3eff8295c382bb3be0f1bd7f089e6716d8e is 97.8% of it, quote side 11.0439 ETH = $27,120. 24h volume $851,685, flow 2212 buys / 2458 sells, h1 21/28. 57.0% off an ATH set only nine hours earlier
- SIZING IS THE BINDING CONSTRAINT, not the thesis. Round trip on the quote side: $1,000 = 7.1%, $2,000 = 13.7%, $5,000 = 31.1%, $10,000 = 53.9%. Ben's normal $5k to $15k is mechanically impossible here. Net buy to run: 10x needs $58,642, 20x needs $94,166
- VERDICT: PROBE at $1,000 to $2,000 maximum, or wait. Nothing on the fatal list, genuinely none: no mint, no blacklist, no pause, no deployer privileges, no fake liquidity (turnover is 15x the pool), no narrative migration, no clone farm. But nothing is arriving either. Upgrade trigger, one call: totalBorrowAssets (0x1aad44c2) on the pool goes above zero WITH a non-zero NVDA/TSLA/SPCX balance in the pool
- FALSIFIERS SET: (1) totalBorrowAssets still 0 after the Tuesday 1 Sep US open, (2) third-party supplied capital excluding team deposits still under $5,000 on 6 Sep, (3) holder count fails to pass 700 by 6 Sep (569 now)

## Burn wallet re-check, 30 Aug 2026 10:55 UTC (Ben asked if 0x2baC5D261b6B265Db14be80B0179e2da21F9AB48 is "accumulating ACRE hard" — it is not, it is the burner, third time this wallet has come up)
- Full two-topic eth_getLogs sweep of the wallet: 27 inbound, 25 outbound token transfers, entire chain history. 22 ACRE buy→burn cycles between 28 Aug 17:18 and 30 Aug 10:50 UTC. Every single ACRE inbound is matched by an identical outbound to 0x...dEaD 20 seconds to 2 minutes later. Current ACRE balance is ZERO and has never been non-zero between cycles. It has never sold a token
- Burned to date 86,232,573 ACRE = 8.62% of supply. Total sitting at dEaD is 86,234,637, so this wallet is 99.998% of every ACRE ever burned; the residual 2,064 is someone else
- Cost: 4.9956 ETH of native tx.value across the 22 buys, about $12,287 at ETH $2,459.51. All 22 self-signed by the wallet itself
- CORRECTION to the 09:21 note "53.75M of that was hour zero dev-token burn". WRONG. Every one of those early transfers carried real native tx.value (0.018, 0.193, 0.145, 0.109, 0.082 ETH on the bonding curve, then 0.396 and 0.297 through the UniversalRouter). No dev allocation was burned. All 86.23M was market-bought with ETH
- CORRECTION to "BURN IS DECELERATING". It re-accelerated hard. 14,758,605 ACRE burned in the 89 minutes from 09:24 to 10:50 UTC for 2.53 ETH, against roughly 0.44M/hr before. Two of those buys were 1.1 ETH and 1.0 ETH, the largest single burns of the token's life. War chest fell 7.2374 to 4.8322 ETH ($11,885) in the same window
- FUNDING PROVEN, this was the open question. Every ETH inflow comes from the VERIFIED contract V2FeeEscrow 0xd3AFEB2a57f70eF218Aa82451c51B2fb0416Ac9e — 18 claims totalling 8.7546 ETH. The wallet was seeded with 0.0274 ETH from GasliteDrop 0xE68d0bbc023de3fEBdA04f413db23cE9C5EA1934 at 28 Aug 16:54:47, 24 minutes before its first buy. Zero outbound internal txs, and only two trivial inbound external transfers (0.00205 ETH from 0x7CaE49E99eEF395734f12c5b0AC13b2B6AD3B80D, 0.0000074 ETH from 0x2a82883f0837a2cAD07A3D91f5BB666d15cdF10A). Nothing has ever been skimmed out to a personal wallet
- Classification: plain EOA, eth_getCode returns 0 bytes, nonce 122, holds real ETH, signs every transaction itself. Not a router (routers here have nonce 1 and transfer counts in the thousands), not an EIP-7702 app wallet (those are 23 bytes). It is one scripted operator wallet
- Same wallet holds 99.7% of aUSDG "Acre USDG Supply" 0x82bD35a34891F1D8B543313101C06CF31b8daD1E and posted 0.53 tokenised AAPL, i.e. it is also the protocol's own first depositor. Fee wallet, burn wallet and sole TVL provider are the same key
- SCALE CHECK: the buyback is about $7,886 of buying in the trailing 24h against roughly $425k of buy-side volume, so ~1.9% of the bid. Real, verifiable, and not what is holding the price up
- TOOLING, update the earlier note: robinhoodchain.blockscout.com now Cloudflare-403s the container on BOTH /api (v1) and /api/v2. It is reachable through the desktop browser pane. Its v2 internal-transactions index is incomplete — 8.755 ETH of claims listed against roughly 9.8 ETH implied by the coin-balance-history deltas, about 89% coverage. Use /addresses/{a}/coin-balance-history to catch what internal-transactions misses