TREE3 is a research-led investment program designed to harvest structure that conventional statistical arbitrage cannot reach — through a disciplined cascade of strategies built on frontier physics, photonic compute, and signature-preserving money management.
“Our mission is to deliver 75+% gross annual returns consistently over the next 100 years — leveraging frontier science and technology, and disciplined money management practices.”
Every candidate strategy must carry a Hodge–Riemann signature on its associated bilinear form. Strategies that pass local backtests but fail the global extension test are discarded — structurally, not by judgment.
For paired venues at sub-microsecond decision horizons, heralded entanglement supplies a measurable, Bell-inequality-violating coordination advantage on every cross-venue decision — a small per-trade edge that compounds across millions of decisions per second.
Standard 12U colocation modules at NY4, Carteret, LD4, FR2, TY3 and beyond. Tick-to-trade in sub-microseconds. Cross-asset Frobenius operations on silicon-photonic tensor cores. Continuous-time market dynamics evolved on Lorenz-attractor reservoirs.
Local profitability is necessary, never sufficient. Each accepted strategy carries a certificate that it composes — under signature-preserving operations — with the entire book across every venue, regime and asset class simultaneously. Linear blending is structurally rejected.
| Grade | Trading meaning | Computational realization | Assets unlocked |
|---|---|---|---|
| 0 |
Single-venue lit-book stat-arb
The Medallion regime · triangle inequality holds
|
FPGA tick-to-trade | US equities |
| 1 |
Reg-NMS-style two-venue arbitrage
Causal coupling propagated · queue-position critical
|
Cross-colo fiber + FPGA | + FX spot G10 |
Darwin barrier · grade-0 obstruction class | |||
| 2 |
Correlated-pair, non-metric hedging
Ding–Jiang quantum edge · q* − c* ≈ 0.002 / decision
|
Quantum telepathy, Type I | + Warrants, crypto spot |
| 3 |
Intercontinental correlated baskets
Light-cone delay dominates · basket coordination is the unit
|
Heralded entanglement memory network | + Govt & corp bonds |
| 4 |
Cross-asset-class macro overlay
Equity ↔ rates ↔ FX coupled via Frobenius eigenmodes
|
Photonic tensor + Ouro loops | + STIR, index futures |
| 5 |
Global vol-surface market making
Single coherent surface across SPX, NDX, VIX, single-name
|
Reservoir computing on FPGA fabric | + Options, VIX |
| 6 |
Regime-conditioned cross-asset
CB cycle, geopolitics, commodity cycle as conditioning context
|
LLM ensemble + retrieval, three-family Ouro | + Commodities, exotics |
Frontier gate · long-horizon R&D program | |||
| 7 |
Game-theoretic adversarial execution
Routing accounts for the response of other adaptive routers
|
Ouro + opponent model in closed loop | All listed |
| 8 |
Endogenous regime · reflexive impact
Trading actions priced against their own consequence
|
Causal simulation + global state coordination | Tokenized RWA |
| 9 |
Single globally coherent book
Every asset, every venue, every regime, one decision surface
|
Full stack synchronized | Frontier objective |
| Asset class | Primary venue(s) | Access model |
|---|---|---|
| US equities | NY4 · Carteret | Colocated matching |
| Listed equity options | NY4 · Cboe (NY5) | Colocated matching |
| Equity-index futures (ES, NQ) | CME · Aurora, IL | Colocated matching |
| STIR futures (SOFR, ESTR) | CME Aurora · Eurex (FR2) | Colocated matching |
| VIX & inverse-VIX | Cboe (NY5 / Chicago) | Colocated matching |
| Commodity futures | CME Aurora · ICE | Colocated matching |
| FX spot & forwards | NY4 · LD4 · TY3 (ECNs) | Colo + OTC ECN |
| Government bonds | BrokerTec · CME futures | Colo + electronic |
| Corporate bonds | Tradeweb · MarketAxess | OTC / RFQ |
| Crypto (spot & perps) | Global CEX / DEX | Distributed |
Every grade above Tier-0 is earned by P&L from the grade beneath it. Capital does not flow into a higher grade until the lower has demonstrated a 12-month rolling Sharpe above its gate threshold, and the Darwin barrier remains intact through the window.
Backtests are not evidence of edge; vendors are not evidence of robustness; intuition is not evidence of anything. The single quantitative object that decides whether the fund is healthy is the realized tail behavior of the cascade against its specified log-concave envelope.
Twelve binary quarter-end gates govern the build. Either every checklist item passes, or the gate fails and the next block does not start. Failure is not a renegotiation; it is a stop. Three pre-approved fail modes — extend, de-scope, wind-down — remove the temptation to drift.
Classical HFT stack. Medallion-class Sharpe on equities, FX, futures. SM-Pod hardware program begins.
Heralded entanglement between paired US venues. Ding–Jiang hedging problem in production. Measured q*−c* edge.
Quantum memory across US/EU/APAC venues via dark fiber. Vacuum-beam-guide construction starts on NY–LD.
Photonic tensor cores live for cross-asset Frobenius operations. Ouro meta-strategy in production. All listed asset classes covered.
VBG ring complete. Grade-6 macro overlay book live. OTC and exotic derivatives covered. Capacity managed via spinout.
As the frontier R&D program matures, grades 7–9 progressively open. The framework remains compounding at Phase 4 throughout.
If you allocate capital at scale or represent investors who do, we'd welcome a conversation. We are particularly interested in introductions from parties who can read the framework as a long-horizon thesis.
This page summarizes the thesis behind a long-horizon research program and is intended for qualified purchasers and professional investors. It is not an offer to sell, or a solicitation of an offer to buy, any security; any such offering will be made only by means of formal offering documents in jurisdictions where lawful.
The target return band represents an aspirational long-horizon objective, not a guarantee. It reflects what the framework anticipates as successive phases of the research program mature. As with any sophisticated investment strategy, outcomes depend on a range of conditions, and no specific result can be assured.
Past performance is not indicative of future results. Prospective investors should consider their own financial circumstances and consult professional advisers before any commitment.
Statements regarding future plans, capabilities, or outcomes are forward-looking and subject to change as the research program advances. Nothing on this page constitutes investment, legal, tax, or accounting advice.