Insights

We write primarily for practitioners: people who care about how a signal is built, how it behaves in different regimes, and what happens when it is exposed to real transaction costs and constraints. Many of these notes come directly from internal research memos—lightly edited, but still close to the models and code that run in production.

Rather than chasing headlines, we focus on a small set of recurring questions: how to size risk through cycles, how to execute efficiently in noisy markets, and how to keep models honest as data and regimes change. Over time, the goal is to build a library that documents what we have learned, including the ideas that did not survive contact with reality.

• Deep dives on specific techniques, with an emphasis on assumptions and failure modes.
• Practical frameworks that can be adapted to different mandates and risk budgets.
• Implementation notes drawn from real pipelines—data, backtesting and live monitoring.