Immediate Start - Machine Learning Researcher (PhD) – Systematic Commodities Hedge Fund

Moreton Capital Partners is rapidly expanding and seeking a talented Machine Learning Researcher to help design and improve the predictive models that power our systematic commodities trading strategies in our Mexico City Office.

We trade global commodity futures using machine learning, alternative data, and institutional-grade portfolio construction. Our edge comes from research depth, disciplined experimentation, and robust production systems.

This role is for candidates completing or having recently completed a PhD with a strong machine learning, statistics, or applied mathematics focus who want to apply advanced research in a real capital environment. You will work directly with the CIO and sit alongside a world class international quant research team to turn cutting-edge ML ideas into live trading signals. Your research will ship to production and directly impact portfolio returns.

This is not a purely academic role. We're looking for someone ready to hit the ground running and available to start immediately. In return, we offer a competitive salary, substantial performance share, comprehensive benefits, incredible work environment and a relocation package to make the move seamless.

What you will work on

• Designing predictive models for cross-sectional and time-series commodity returns
• Developing new features from price, positioning, options, macro, and alternative datasets
• Improving signal robustness and reducing overfitting through rigorous validation
• Combining and blending multiple models into portfolio-level forecasts
• Regime detection, meta-models, and adaptive allocation frameworks
• Model diagnostics, explainability, and stability analysis
• Translating research ideas into production-ready implementations
• Collaborating with engineers to deploy models into live trading systems
  

Key Responsibilities

• Formulate research hypotheses and test them using clean, time-aware ML pipelines
• Build and evaluate models (tree-based, linear, ensemble, deep learning, etc.)
• Run walk-forward and out-of-sample experiments with realistic costs
• Analyze information coefficients, turnover, drawdowns, and risk-adjusted returns
• Design feature engineering frameworks and reusable research tooling
• Document findings clearly and communicate results to portfolio managers

• Contribute to improving research standards, reproducibility, and processes