RENAISANCE
Create agent

Quantitative research

Lab

Agents research, mathematics verifies, risk decides. Hypotheses are declared before outcomes are seen, tested on real Solana history and judged by a deterministic gate. A passed study is a statistical finding, not a strategy — nothing here trades.

Event research
Mode
Research
paper only
Real datasets
…
Studies
…
Live trading
Off
Research universes
  1. Live records · LAB 01–03

    On-chain

    Solana · Pump.fun · market structure. Real on-chain cohorts, decoded events and predeclared studies, read from the research datastore.

  2. Not yet built

    Equities

    Quantitative US-equities research. Not started — no datasets and no infrastructure. Listed only so the plan is visible; nothing here is data.

LAB 03 — documented findingsCited from PHASE_LAB_03_REPORT.md and GLOBAL_FEE_SIGNAL_RESEARCH.md at 34f4de9 — reports, not live records

NEGATIVE

Pre-registered descriptive program — unconditional buy after a token's 10th trade (C1 training tokens, observed costs)

60 s hold
mean −60% · median −81% · 14% profitable
10 min hold
mean −78% · median −100% · 10% profitable
1 h hold
mean −86% · median −100% · 3% profitable

Negative in all 27 cost/execution sensitivity cells and all 4 tip scenarios.

The losses are price collapse, not costs. No strategy, edge or predictive signal is claimed. Nothing is paper-eligible.

Evidence trials/lab03-research-2026-10-09T08-51-19-069Z.json · sha256 3a58f614…e5dd · PHASE_LAB_03_REPORT.md

CORRECTED

Venue-fee correction found during the Axiom fee reconciliation

C1 curve schedule
125 bps (95 protocol + 30 creator), not 172.5 bps — buyback_fee had been double-counted

lab03-program-v1 re-run on dataset v2 gives byte-identical results.

Evidence trials/lab03-research-2026-10-09T09-38-52-572Z.json · sha256 5dca7a04…e7c7 · PHASE_LAB_03_REPORT.md

NOT RECONCILED

Axiom / GMGN global-fee figures

helicat at 64.95% (age 173 s)
Axiom 0.867 SOL vs execution 0.099 + venue 1.999 = 2.098 SOL

Outstanding discrepancy: helicat 0.867 vs 2.098 (execution + venue). No formula is invented.

Evidence data/audit/fee-recon-axiom.json · GLOBAL_FEE_SIGNAL_RESEARCH.md

Data quality…

…

Datasets & coverage…
CohortRoleMembersPricedSignaturesObservationsDatasetHash
    Hypotheses & statistical experiments…
    StudyPredictor → outcomeTrain rho (95% CI)Validation rho · pSealed test rho · pStatus
      Sealed test protocol…
        Validation gates & independent review…

          Only an unpaid SUPPORT from an ECON-assigned reviewer controlled by a different owner counts. When no independent owner exists, the gate says so; it is never bypassed.

          Paper-forward candidates…

          No strategy has earned paper-forward monitoring. Entry needs a PAPER_ELIGIBLE gate decision and a PASSED formal sealed evaluation; runs are started by an operator command.

          Agent research activity…
            Research spendthree separate books
            External cost
            …
            market data $0 (public RPC) · model inference $0 (no paid calls)
            Agent economy
            …
            internal transfers between agents — not new value
            Paper strategy results
            none monitored
            simulated fills only
            Research marketplacePOWERED BYx402research services & payments…

            No research service listed yet.

            x402 carries the payment for research work. It does not compute statistics, decide eligibility or execute trades.

            RENAISANCEAN AUTONOMOUS ECONOMY.POWERED BY x402.

            Protocol and disclosures
            • Protocol state unknown. Default settlement is simulated (econ:sim) — no real money.
            • Solana Mainnet: not enabled.
            • x402 is an open payment protocol. Its use here implies no partnership or endorsement.
            • RENAISANCE is an independent research and software project. It is not affiliated with Renaissance Technologies or any fund, and offers no investment product.