Research
A dislocation is not a story that the market is wrong. It is a constraint that prevents an immediate correction: inventory, information, funding, latency, regulation, or a model that treats two clocks as one.
Quantitative research
Markets displace. We measure the displacement.
Dislocation Systems studies temporary, measurable deviations from equilibrium — and builds the systems that decide whether those deviations are real, tradable, or an artifact of the clock you used to observe them.
Systems
We treat data, research, and execution as an evidence problem. Vendor bytes are kept before they are interpreted. Snapshots publish atomically. Failures stay visible. Nothing is inferred from a clean exit code.
Capital
This is a private firm. We do not sell a product, manage outside money, or publish a track record. The work is the research program, not a marketing surface.
Operating principles
How we refuse to fool ourselves
- Evidence before transformation. Store and verify the native record before parsing, typing, or joining it.
- Temporal honesty. Event time, reporting time, knowledge time, revision time, and extraction time remain separate fields.
- Declared versus observed. What the system expects and what a live run actually finds are different objects. Gaps stay explicit.
- Atomic publication. Research sees a complete validated snapshot or the previous one — never a half-written present.
- Failure visibility. Operational success, data validity, and coverage completeness are independent states.
Selected notes