Quantitative Analyst

Quantitative Analyst - Power Markets / Battery Storage

Energy Quant Engineer - electricity price forecasting for grid-scale battery storage

Paris or Bordeaux

\| Permanent (CDI)

The work

Electricity price forecasting, end to end, for a business that operates flexible assets on the French power markets.

You will own the price models: the ones already running, and the ones that don't exist yet. They are not decision support. They are wired into the optimisation engine that decides when the assets charge, discharge and bid, so a bad forecast costs money the same day. You will see that feedback loop directly, which is unusual and is the main reason to take this job.

Alongside the forecasting you will work on the optimisation layer itself - the MILP formulations behind the dispatch decisions - with the traders and engineers who run it.

Day to day

• Maintaining and improving the live forecasting models, and designing new approaches where the current ones fall short
• Machine learning and deep learning models, and fundamental models, depending on what the horizon calls for
• Forecasting across horizons - from a few days out, used to operate the assets, through to the long-range views that sit behind investment decisions
• Building and improving the scenario generators, and putting proper uncertainty around the central case
• Turning market fundamentals and the physical limits of the assets into features the models can actually use
• Working out what really drives price and volatility, and quantifying it
• Improving the MILP dispatch modules on economic performance and on speed
• Tracking forecast error in production and acting on it
• General hardening of the platform - this is a small team and everyone ships

You'll work closely with a market analyst who covers the fundamentals side, so you are not expected to be the resident expert on everything.

Background

• Around three to five years spent forecasting, modelling or doing quantitative analysis in power markets
• Excellent Python
• Statistics, applied stochastic modelling, and mixed-integer optimisation
• Serious time series and applied ML / deep learning experience
• A working understanding of how the power markets actually function - ancillary services, day-ahead, intraday, balancing
• Comfortable with Git, code review and CI/CD, and used to taking your own work into production rather than handing it off
• Master's or engineering degree in energy, applied maths, data science, econometrics, computer science or similar

Helpful but not required: exposure to dispatch optimisation, systematic trading, or battery storage. Experience of power markets outside France.