SageTrading is a quantitative hedge fund. We treat markets as a scientific problem — every strategy begins as a hypothesis, survives rigorous testing, and is executed by systems, not sentiment.
SageTrading is a quantitative hedge fund built on a simple conviction: markets are among the most complex systems humans have ever created, and understanding them demands the same rigor we bring to the hardest problems in science.
We treat trading as a research discipline. Every strategy we deploy begins as a hypothesis, is tested against evidence, and survives only if the data holds up. We pair deep quantitative modeling with modern machine learning and artificial intelligence to find signal in noise, to adapt as markets evolve, and to make decisions at a speed and scale no human could match alone.
Probability, statistics, and optimization are our first language. Position sizing, risk, and expectancy are derived — never guessed.
Every strategy is a falsifiable hypothesis. We test out-of-sample, hunt for our own errors, and retire ideas the moment the evidence turns.
Modern models — from classical learners to large language models — extend our research bandwidth and adapt as market regimes evolve.
Research only matters if it executes. Deterministic risk controls, automated monitoring, and auditable systems carry every idea to market.
Our flagship futures system — a combined intraday and overnight engine on the CME E-mini Nasdaq-100 — is presented in two studies: the complete seventeen-year record, evaluated against every one-minute bar since April 2009, and a daily close-up of the most recent five and a half years. Not a sample. Not a favorable window.
Every figure is net of exchange and clearing fees and realized slippage, computed with lookahead-free indicators on tick-synchronized data.
The entire continuous record — spanning the post-crisis recovery, the 2018 and 2022 drawdowns, and the AI-era bull market — evaluated on every one-minute bar with the full risk stack in force.
+21.2% in 2022 while the Nasdaq-100 fell 33% and the S&P 500 fell 20%. The system was also up 54.8% in 2018 — a year both indices closed lower.
Fifteen of eighteen calendar years finished positive; the three down years were −11.7%, −14.1% and −5.5%. The overnight engine alone was profitable in 17 of 18 years, with a win rate between 57% and 84% in every single year.
With one contract per signal, the deepest drawdown in seventeen years was −6.65%, against −35% to −40% for every buy-and-hold benchmark. Compounding at 5× leverage scales returns and drawdowns alike: +269,054% with a −43.45% maximum drawdown.
The most recent five and a half years, marked every day: the 2021 melt-up, the 2022 bear market, and the AI-era rally. A separate evaluation run of the same system (VectorBT Pro; CME fees $2.25 per side plus one tick of slippage).
+20.6% in 2022 (+8.7% with a fixed contract) while QQQ fell 33% and SPY fell 20%. Unmanaged NQ futures at the same 5× leverage lost 89%.
Six for six. The worst year was still +20.6% (+8.7% fixed); QQQ's worst was −33.2%, and both benchmarks finished one of the six years in the red.
The longest stretch below a prior peak was 239 days (225 with a fixed contract) — the 2022 drawdown, recovered by April 2023. QQQ spent 754 days beneath its 2021 high.
Hypothetical performance. Results are derived from historical simulations (VectorBT Pro) of a systematic strategy on CME E-mini Nasdaq-100 futures — the full-history study (April 20, 2009 – August 18, 2026) and a separate daily-curve run (January 4, 2021 – August 17, 2026) — and do not represent live trading or any actual account. Simulated results have inherent limitations: they are prepared with the benefit of hindsight, may not reflect all market frictions, and do not reflect the impact of material economic and market factors on real-time decision-making. Past or simulated performance is not indicative of future results. Nothing on this page is an offer to sell, or a solicitation of an offer to buy, any security or interest in any fund.
At SageTrading, we revere the culture of academia and the discipline of rigorous science. We ask hard questions, we challenge our own assumptions, and we let results — not opinions — settle the debate. We believe the best ideas can come from anyone, and that the pursuit of truth is a collaborative one.
Hard questions are the starting point of every strategy. We would rather sit with an open problem than settle for a comfortable answer.
We challenge our own assumptions and let results — not opinions — settle the debate. A hypothesis survives only if the data holds up.
We prize the elegant solution over the expedient one — systems built to be understood, maintained, and trusted.
The same standards that govern great research govern how we build.
SageTrading was founded by Yuan Fang, Liang Tang, and Eric Li — a team united by a shared belief that the frontier of trading lies at the intersection of mathematics, science, and artificial intelligence.
Together, they set out to build a firm where world-class researchers and engineers could do their best work: solving genuinely hard problems, held to the highest standards, and free to follow the science wherever it leads.
We hire for one thing: the ability to find truth in data. Hover over a role to see the full description.
Questions about our research, strategies, or working with us — send a note and we'll get back to you.
Log in to the investor desk to view strategies, research, and live portfolio performance.