Process
How We Work
Systematic. Rigorous. Reproducible.
We build investment processes the same way we build software — with version control, testing, and no magic. Every stage runs continuously; there are no office hours, no shift changes, no cognitive fatigue between sessions.
Data Acquisition
We source structured and alternative data across global markets — price series, macro releases, on-chain activity, and proprietary datasets — and normalise them into a unified research environment.
Signal Research
Researchers develop quantitative hypotheses grounded in economic theory and statistical rigour. Every signal undergoes out-of-sample testing and stress analysis before it reaches the portfolio layer.
Portfolio Construction
Signals are combined through a disciplined allocation framework that balances expected return against tail-risk. Position sizing is systematic; discretion operates only at the strategy level.
Execution
Orders are routed through low-latency infrastructure with transaction-cost models embedded in the loop. Market impact is treated as a first-class constraint, not an afterthought.
Risk & Monitoring
A live risk engine tracks factor exposures, drawdown limits, and liquidity constraints in real time. Automated circuit-breakers enforce pre-agreed risk thresholds without human intervention.
Research Loop
Every closed trade feeds back into the research system. We measure signal decay, attribution, and regime sensitivity continuously — the process improves because the feedback is permanent.
Guiding Principles
Rules over discretion
Decisions follow models. Models follow evidence.
Separation of concerns
Research, execution, and risk run independently.
Reproducibility
Every result must be reconstructable from raw data.