Don't trust the demo.

Every claim in Venue0 points at a Robinhood Chain transaction or a frozen campaign file. Here they are, including where Venue0 doesn't help.

Live mainnet proof

Robinhood Chain 4663 · operator proof wallets
3-wallet Stock Token cycleNVDA, AAPL and SPY moved A → C → B → A in one atomic settlement. ReceiptVerified0xfdd1…3035 (opens on Blockscout)
Uniswap residual0.002547 NVDA0.001658 AAPL, CLASSIC, Universal Router 2.1.1Verified0xaec4…84db (opens on Blockscout)
Flash limit orderFilled. Boundary economics: fee 13.6% of a $1.20 residual; the current engine would aggregate it. DetailsFilled0xa30e…5fc6 (opens on Blockscout)
Dynamic agent roundA portfolio agent on a Dynamic server wallet (agent-owned MPC) signed its intent, plan approval and allowance, then settled.Verified0x0f85…1873 (opens on Blockscout)
Settlement contract0x9cf87131…78e86a3dVerifiedBlockscout (opens on Blockscout)

Deployed app

Operator-controlled test accounts through the public app
P1: three-way cycle through the public appVerifier 14/14. Executed via robinhood-mainnet.g.alchemy.com, verified via robinhood-mainnet.g.alchemy.com. Same provider, so not an independent check.Verified0x45c0…ec35 (opens on Blockscout)
P2: three-way cycle through the public appVerifier 14/14. Executed via robinhood-mainnet.g.alchemy.com, verified via robinhood-mainnet.g.alchemy.com. Same provider, so not an independent check.Verified0x17b0…6b1e (opens on Blockscout)
P3: three-way cycle through the public appVerifier 14/14. Executed via robinhood-mainnet.g.alchemy.com, verified via robinhood-mainnet.core.chainstack.com. Independent providers.Verified0xd8d9…edd3 (opens on Blockscout)

Campaign

Synthetic portfolios, not users

400 frozen scenarios, 4 arms

Seed, generator, cost curves and metrics were committed before the run. Each arm saw identical intents, prices and quoted costs. Reference parity 400/400; matcher latency p50 1 ms, p95 2 ms.

ArmCrossed (pooled)MedianExternal orders
Pairwise only (optimal)8.1%8.8%8,778
Market only0.0%0.0%11,865
Venue0 crossing10.2%13.5%8,093
Venue0 + residual engine10.2%13.5%7,328

By cohort

constructed-positive cohorts are not typical behaviour
CohortNaturePairwise onlyVenue0Relative upliftZero-cross
concentrated technatural/randomized14.2%15.9%+12.1%3/40
external route stressstress/control23.8%31.5%+32.1%1/40
high complementarityconstructed-positive9.9%19.7%+99.5%0/40
index rebalancenatural/randomized48.3%51.7%+7.0%0/40
low complementaritystress/control7.5%8.3%+9.8%21/40
partial matchconstructed-positive7.6%14.8%+95.8%0/40
random independentnatural/randomized19.8%25.2%+27.5%4/40
size skewnatural/randomized2.1%2.7%+32.7%0/40
theme circlenatural/randomized34.4%43.3%+25.9%0/40
zero overlapstress/control0.0%0.0%n/a40/40

Where Venue0 doesn't help

  • Below ~10% complementarity, crossing is ~2.4% (129 scenarios).
  • Rounds of 2–3 participants gain little: 1.9% and 3.3% crossed.
  • One-directional flows often cross nothing: 21 of 40 low-complementarity scenarios crossed zero.
  • A large participant in a small crowd crossed only 2.7% of requested notional.
  • Template-like rebalances are already captured well by pairwise matching: 48.3% pairwise vs 51.7% Venue0.
  • In absolute terms multi-party matching added about 2.1 points of requested notional overall (the 26.3% figure is relative to the pairwise optimum).