Quantitative Researcher -- Generative Models for Market Microstructure

Boulder, CO · Part-Time Contractor

A small Boulder-based proprietary trading firm is launching a research program in probabilistic forecasting of intraday market dynamics. We're looking for a mathematically strong researcher to help build and validate generative models on real high-frequency market data.

The work sits at the intersection of stochastic analysis, modern generative modeling, and market microstructure. The problems are genuine applied mathematics and physics: building probabilistic models of multivariate financial time series that respect the heavy-tailed, long-memory, and scaling-law structure observed in real order flow. You'll work from the underlying mathematics through to production-grade implementation.

Technical scope includes:

▸ Score-based diffusion models and stochastic interpolants for probabilistic forecasting

▸ Stochastic differential equations, including jump-diffusion and heavy-tailed noise processes

▸ Capturing the empirical structure of order flow, price impact, and metaorder dynamics in our generative models

▸ Calibration, uncertainty quantification, and the geometry of probabilistic forecasts

Who we're looking for:

Strong mathematical maturity is essential. The ideal candidate is comfortable reading and implementing technical papers in stochastic analysis or machine learning, writes clean scientific Python, and enjoys turning hard math into working code. Helpful prior exposure (any subset is meaningful):

▸ Diffusion models, flow matching, score-based generative models

▸ Stochastic calculus or Lévy processes

▸ Time series modeling at high frequency

▸ PyTorch or JAX in a research setting

Finance experience is not required.

To apply:

Send a resume and a brief statement of mathematical background to tw@nickel5.com. If you'd like, include a short paragraph on a paper, technique, or mathematical idea you've recently found compelling -- and why. Selected candidates will be invited for an in-person conversation.

We welcome candidates from physics, applied mathematics, statistics, computer science, or quantitative finance.