Quant gesucht in Berlin

• Price models for the intraday market, built from weather model output, fundamentals and our own order-book history.
• Trading strategies for the continuous intraday market, together with the sizing and the risk limits that go with them.
• Running what you build in production, and comparing what it did with what the backtest said it would do.

You take each of these from the first idea to live trading, so we don’t split the work much along the value chain.

How we work

The team has backgrounds in machine learning, particle physics, information systems and economics, and together over two decades of experience in energy markets. Everyone works across research, trading systems, data and infrastructure. We think people do their best work when it is truly theirs.

Everyone also does the ordinary work, from scheduling to the occasional invoice, because we want to stay small, and a small team only works if everyone chips in. Live systems need looking after, so monitoring on weekends and holidays rotates through the team.

Our stack is open source and our own code from top to bottom, and we run it ourselves, so anyone here can read and change every part of it. We write most of that code with AI coding assistants, and the repository, the documentation and the data are organised so that both people and models can work with them.

• You have a strong academic background in physics, statistics or another quantitative field, typically a PhD, but a master’s degree and real work with real data count just as much. We care more about how you think than about the title.
• You want to understand the energy system itself, not only the methods you apply to it. The domain matters to us as much as the modelling. While you’ll find plenty of support and knowledge in the team, you are comfortable taking ownership for your learning journey.
• You have a very strong understanding of statistics and genuinely view the world probabilistically. We value that more than familiarity with any particular class of models.
• You know what clean code and a clean architecture look like, and you hold the code to that standard whether you or an assistant wrote it. Writing Python by hand matters less than it used to; knowing when to do it anyway, for example when you are iterating on a statistical model, matters more. You do not need to be a software engineer.
• You would rather be corrected than be wrong, and you can take an idea all the way to production yourself., * Experience in power markets or as a quant trader.
• Time series modelling. R, where it is the better tool.
• You are at home on Linux.
• Some familiarity with Rust or C., English is our working language; German is helpful but not required. We welcome applications regardless of gender, background or where you studied, and we particularly welcome applications from women.

px dynamics is a small quantitative trading firm in Berlin and Cologne. We are four people. We trade European short-term power markets systematically, on our own account, and we sell trading signals to a few trading desks. We build and run everything ourselves: the data, the models, the execution and the servers. We are looking for one more person to research and run trading strategies with us.