Metals Market Intelligence
AI-enhanced macro analysis across 34 structural and behavioral signals.
Overview
A decision-grade intelligence system for metals markets, built within Legacy Quant. The system integrates 34 macroeconomic, structural, and behavioral signals to detect market regimes and identify pattern shifts that traditional analysis misses.
Output is structured for trading decisions, not visualization. The goal is decision infrastructure, not dashboards.
What it does
- Multi-factor analysis across 34 signals
- Market regime detection
- Behavioral pattern recognition
- Signal generation for decision-making
Architecture
- Signal ingestion layer
- Pattern recognition models
- Regime detection module
- Decision output interface
Architecture diagram coming soon
Impact
- Structured macro signals for trading decisions
- Behavioral patterns invisible to traditional analysis
- Decision-grade output, not noise
- Foundation for a broader macro intelligence platform
AI-driven macro intelligence system
Want a system like this?
Book a strategy callWhy 34 signals rather than a model that predicts price
Direct price prediction on liquid markets is where quantitative projects go to die: the target is close to a random walk at short horizons, and any edge that does exist is arbitraged by participants with better data and faster execution. Regime detection is a different and more tractable question — not where the price goes, but which set of relationships is currently governing it.
That is why the input is a wide, heterogeneous panel rather than a deep price history. Macro conditions, market structure, and behavioural indicators fail in different circumstances, and a regime shift shows up as several of them changing character at once, well before it is legible in the price alone.
Output built for a decision, not a dashboard
A dashboard moves the work to the human: it presents thirty-four charts and leaves the synthesis to whoever is looking. That is exactly the step the system exists to perform, and leaving it undone means the tool is used enthusiastically for a month and then not at all.
Structured output — a stated regime, the signals that moved, and their direction — can be checked, logged, and argued with after the fact. It also means the system's record is auditable: what it said, when, and whether it was right, which is the only basis on which anyone should keep using it.