How FST Works: One Accountable Loop from Signal to Review

FST preserves the evidence, policy, risk budget, approval state, broker response, and outcome as one continuous session instead of six disconnected tools.

Key Takeaways

• Screen ranks opportunities; Signal turns evidence into a thesis with confidence and invalidation.
• Trade checks portfolio fit, sizes bounded risk, previews the exact order, and follows approval policy.
• Monitor watches both position state and thesis state rather than price alone.
• Park applies the selected Exit, Cut, Add, or Hold action; unattended sessions use the displayed safe default.
• Learn journals the outcome and may test research candidates in shadow mode without silently promoting them to live execution.

The Six FST States

Screen examines the configured QS signal universe and ranks candidates using current evidence rather than accepting every alert. Signal opens a candidate's market structure, supporting and conflicting inputs, confidence, invalidation, freshness, and provenance.

Trade converts a qualifying thesis into a concrete plan for one exact account. Before submission, the system checks the requested asset and operation against the account's certified capabilities, applies exposure and loss limits, and produces a preview that can be approved or rejected under policy.

Monitor follows the broker-authoritative position, the original thesis, current market evidence, stop, and target. Park resolves the position through an explicit action or the displayed safe default. Learn records the result so research can evaluate process quality without rewriting production policy during a live session.

How It Works

1. Screen — Rank current opportunities across the selected QS signal universe.
2. Signal — Read evidence, disagreement, confidence, freshness, thesis, and invalidation.
3. Trade — Check account capability and portfolio fit, calculate bounded risk, preview the exact order, and apply approval policy.
4. Monitor — Track the authoritative position and whether the thesis remains intact.
5. Park — Exit, cut, add, or hold under policy, with a visible safe default for unattended decisions.
6. Learn — Journal the outcome and evaluate research improvements in a separate shadow loop.

Frequently Asked Questions

Does FST trade every signal it sees?

No. A signal must pass evidence, portfolio-fit, capability, risk, and approval gates before an order can be submitted.

What happens if the selected broker cannot perform the trade?

The operation fails closed before submission. Provider branding or a connected status alone is not sufficient proof that the exact account supports the requested asset and operation.

What does FST monitor after a fill?

FST monitors the broker-authoritative position together with price, thesis, market evidence, stop, target, and invalidation state.

Can research experiments change a live FST session?

Not silently. Candidate improvements belong in a shadow research loop and require review and promotion before they can affect production policy.

Install FST 2.0 (desktop CLI & mobile app) · See QuantSignals V6