We apply the rigor of the scientific method to financial markets. We are physicists, mathematicians, and statisticians decoding the complex behavior of the global economy.
// Stochastic Volatility Model
dSt = μStdt + √vtStdW1t
dvt = κ(θ - vt)dt + σ√vtdW2t
A signal is replayed before it is trusted.
The historical store is measured in petabytes.
Our researchers don't just "find patterns." They build robust theories about market mechanics and test them to destruction.
Ideas originate from academic literature, market observation, or alternative data anomalies. We encourage intuition, but demand proof.
We replay the strategy against petabytes of historical data using our proprietary "Gauntlet" engine. We account for fees, slippage, and market impact.
Once a model passes peer review, it is deployed to our live trading infrastructure. But the work isn't done; we constantly monitor for alpha decay.
We have deliberately cultivated an academic culture. We don't have "traders" shouting across a floor. We have quiet, collaborative spaces where researchers solve hard problems together.
We publish our own internal research papers. Knowledge is shared, not siloed.
Our research library is a shared resource. You stand on the shoulders of giants.
Walk through our office and you'll see equations scrawled on glass walls, napkins, and desks.
"A signal is replayed before it is trusted. Most ideas die in that replay."
"The data here is cleaner than anything we saw in academia. The engineering team treats data as a first-class citizen, which lets us focus on the modeling, not the munging."
"The work stays with the firm. A signal is replayed before it is trusted."
We are looking for researchers who aren't afraid of hard problems.