Who We Are
Oak St. began as a small group of researchers and engineers who thought the quantitative approach to markets was still under-built: that with patient validation and careful engineering, modest and well-understood effects could be combined into something durable. The firm was built from the data outward, one validated relationship at a time. The infrastructure has matured since, but the working method has not changed: examine the evidence, distrust the easy answer, and diversify everything that survives.
We prize a culture in which research, engineering, and trading sit at the same table and argue from the same numbers, whichever office they sit in. Analytical rigour, an open exchange of ideas, and a plain account of what we know and what we do not carry us forward.
Core Principles
Five principles sit underneath everything Oak St. does. They describe how we work, how we decide, and what we are willing to give up in order to keep them. They are written down so that they can be held to.
Show Your Work
Every result at Oak St. arrives with the code, the data, and the assumptions that produced it. A number nobody can reproduce is an opinion, and we do not trade on opinions. Reviewers are expected to run the analysis rather than admire it, and authors are expected to make that easy. The habit costs time on the day and saves it for years.
Diversify Everything
No single signal, market, model, or person should be able to decide the firm's outcome. We would rather hold many modest, independent, well-understood effects than one large effect we only half understand. The same rule applies to our infrastructure, our suppliers, and our own attention. Concentration is a decision, and it has to be argued for.
Prefer the Boring Failure Mode
Systems fail; the question is how. We design so that the likely failure is a halted process and a page to an engineer, not a silent drift in the wrong direction. Simple, observable, and reversible beats clever wherever the two conflict. When we are unsure, we do less, and we make it easy to stop.
Hire for Judgement
Credentials tell us where someone has been; judgment tells us what they will do when the data is ambiguous and the deadline is real. We look for people who can hold two hypotheses at once, who change their minds in public, and who would rather be corrected than be wrong for longer. We hire slowly, in small numbers, and we give new colleagues real responsibility early.
Say What the Data Says
We describe our results in plain language, including the ones that did not work and the ones that worked less well than hoped. Effects are reported with their uncertainty, and forecasts come with the conditions under which they fail. We do not round up. Clear, unadorned communication is how a small firm stays honest with itself.
Our People
We spend a great deal of time looking for people who think carefully: those who reason from evidence rather than habit, who can build as well as analyze, and who set ego aside in the interest of getting things right. Then we give them real problems, direct access to the people who set the standards, and room to work in the way that suits them.
Our colleagues, and the range of backgrounds they bring, are the source of the firm's creativity and of the care with which it works.
They have come to us from research groups, from engineering teams, and from trading desks, and each has a different account of why they came and why they have stayed.