TraderX documentation
Yeshiva University's TraderX build extends the FINOS reference application with a sequenced matching engine, risk controls, post-trade processing and a three-member Aeron cluster. It also includes market history, execution algorithms, FIX entry, listed options, OTC contracts and EOD inputs for an external risk engine.
The repository contains 14 upstream numbered learning states and 18 YU states. They describe different lineages. YU18 also contains feature components, including order types, managed run identity, projection recovery and the Trader Desk. A state or component specification can contain unfinished acceptance work.
Choose a starting point
- What's new: the platform, original Demo console and new Trader Desk.
- Testing strategy: local commands, prerequisites and what each check proves.
- Learning guides and learning graph: dependencies and source specifications.
- Getting started: generate a selected state from its source layers.
Read results in context
The matching engine, SQL projections, UI and external risk worker are separate parts of the system. An accepted order may be resting or canceled rather than filled. A stored job result does not prove that its worker is running. Synthetic or assumed pricing inputs remain labelled, and the closed container profile does not provide production portfolio risk.
Source and runtime
Current integration source is on traderX-risk-integration. State generation composes templates and overrides into generated/code/target-generated; full-file overrides apply in lineage order. The state catalog records the learning states and their entrypoints. Branch publication and a running deployment are separate from local generation.
The live-environment registry lists upstream demo URLs; availability is not checked by a local docs build. See the API Explorer guide for runtime API inspection and ADRs for architecture decisions.