Overview

What Overview is for

Overview is the starting point for one immutable Run Output. It answers four questions before you inspect the detail:

1. What exactly ran?
2. Which data, Strategy or Portfolio version, parameters, and execution model produced it?
3. What happened at a glance?
4. Which limitations or blockers change how the result should be used?

A Run Output is the detailed record of one Strategy or coordinated Portfolio execution. It is not the same as a Study Result, which compares candidates and explains how one was selected.

Find the right record

The catalog at the top of Results includes Strategy and Portfolio records. Use:

• Result to choose a stored record;
• Subject to show All Results, Strategy, or Portfolio;
• Search to find a displayed name or identity;
• the page controls to move through a long catalog; and
• Refresh Results after a run or Study completes.

Run Outputs open in Results. Optimization Results open their Strategy or Portfolio Study Results owner, where candidate sets, leaderboards, parameter surfaces, folds, and selection evidence belong.

Verify identity before performance

The header and Overview must agree on the complete run identity:

• Strategy or Portfolio id and immutable version;
• member Strategy Configurations for a Portfolio;
• parameter vector or Portfolio policy;
• Dataset and Prepared Data lineage;
• time window and market bindings;
• requested and effective execution fidelity;
• execution-model assumptions and fingerprint;
• Summary or Robust analysis depth;
• Compact or Interactive retained-detail profile; and
• parent Study Result and candidate identity when the run was promoted.

Changing any of those inputs creates another record. Results never silently rebases an older Run Output onto the latest Strategy, Portfolio, data, or execution preset.

Read the decision summary

Overview presents the outcome first, followed by the context required to interpret it. Depending on the subject and retained evidence, it includes:

• starting and ending capital;
• net and gross profit or return;
• maximum drawdown;
• trade count, exposure, holding period, and cost summary;
• a compact equity and drawdown path;
• Strategy or Portfolio contribution highlights;
• validation status and important limitations; and
• links to the parent Study, producing job, and retained outputs.

Read the values together. Return without drawdown hides the path. Gross profit without costs overstates the result. A high average based on few trades can be fragile. A smooth Portfolio total can hide concentration in one member or period.

Understand Summary and Robust

Summary answers what happened. It contains the point metrics, scope, lineage, execution assumptions, fidelity, and obvious data limitations needed to understand the run.

Robust includes Summary and asks how fragile the conclusion may be. It can add retained-data resampling, cost tolerance, temporal stability, recovery uncertainty, benchmark-relative evidence, and Portfolio dependence or diversification where applicable. Robust does not silently launch another set of Strategy executions.

The retained chart choice is separate. Interactive Chart adds a detailed Chart tab; it does not strengthen the analysis depth.

Use the tabs in a practical order

1. Overview — verify the record and understand the broad outcome.
2. Validation — check data, execution, fidelity, rerun readiness, and blockers.
3. Performance — reconcile profit, costs, drawdown, distributions, and contribution.
4. Chart — inspect the path and trade locations when Interactive Chart was retained.
5. Parameters for Strategy reuse, or the Portfolio-specific tabs for coordinated policy and attribution.
6. Positions & Trades — inspect positions and individual executions.
7. Monte Carlo & Robustness — review sensitivity and retained Robust evidence.
8. Outputs and Diagnostics — inspect exact files, events, checksums, and operational detail.

Past and simulated performance do not predict future results. A Run Output describes one workflow under recorded data and assumptions; it does not certify profitability or future execution.

If the expected result is missing

• Clear Search and switch Subject to All Results.
• Refresh after the producing job reaches Complete.
• Open System & Jobs -> Jobs for an active, failed, or canceled job.
• Open Data Manager -> Stored Runs for storage and lifecycle actions.
• If an optimization record opens Studies, continue there; the handoff is intentional.
• If a historical schema is unreadable, keep the blocked inventory visible rather than attempting to reinterpret or silently delete it.