Skip to system research
Build A Trading SystemRules / Evidence / Replay

Source map and unresolved evidence

Educational research design and invented arithmetic examples. No strategy returns, broker execution or production certification are supplied.

Original edition: 8 October 2026. Source limits appear below. Expanded edition: 13 October 2026.

Educational research design and invented arithmetic examples. No strategy returns, broker execution or production certification are supplied.

Hypothesis before optimisation

Describe the cause, expected effect and observable alternative explanation. Decide which experiments can reject the idea before searching for parameters. Record every material research choice, including unsuccessful runs; the best final curve is not the whole research population.

Evidence layers stay separate

Unit tests, historical simulation, out-of-sample research, paper execution and actual account records answer different questions. Software correctness does not establish an independent trading sample. A simulation can be useful while still excluding queues, impact, gaps or unavailable historical features.

Reproducibility before promotion

Keep a run manifest with code, data, environment and policies. Require a second reconstruction of the chosen candidate and a documented operational review before considering real deployment. The examples on this site do not certify a model or authorise live trading.

Publishing boundary

Provider descriptions are adjacent-tool assessments based on public documentation, not independently audited subscriptions. The owner states legal independent ownership; commercial arrangements have not been independently reviewed. The publication does not supply an automatic winner or a fabricated research return.

Contact and unfinished evidence

A monitored editorial and security reporting route needs owner confirmation. Complete paid-account records, source-specific instrument permissions and independently reviewable commercial records were not supplied. Missing evidence is disclosed, not filled with an invented author, rating or return.

Sources and scope

Public research catalogue and guide inputs

Invented prices have no historical vintage

All 64 bars were deliberately constructed to exercise rule transitions and end-state behaviour. Their IDs, arbitrary UTC timestamps, OHLC values and volumes are not sourced from an exchange. The fixture supplies no point-in-time universe, corporate action, individual futures contract or revised economic release. It therefore cannot support a historical trading claim. The standard library replay is a software reference over invented observations. A real data task needs a producing source, identity, availability rule, revision policy and licensing record appropriate to the actual asset; copying the synthetic file cannot create those records.

Mechanism documentation is not paid testing

The official engine API supports its stated timing and execution semantics. QuantConnect documentation is used for research, warmup, slippage and reconciliation concepts, not as certification of this unrelated publication's implementation. Product manuals support documented roles such as information, scanning, alerts, copying or automation. No selected paid account or live execution was reviewed. The original evidence catalogue and assessment-depth records remain intact as historical structured material, with expanded sections stored separately. No fabricated tester population, comparable provider return table or independent expert approval is appended to them.

The replay bundle identifies its files

Public artifacts include the synthetic input CSV, frozen settings, source launcher, simulated transactions, full replay receipt and a checksum map. Source and settings identities belong to the software example; they do not date a real performance cutoff. The engine's floating-point cash output is retained, with an independent Decimal reconciliation showing negative 1.20 cash within a declared numerical tolerance. Reporting two decimals is presentation, not a manufactured equality explanation. Dependency versions and the engine licence are available, while no blanket licence is invented for proprietary models or third-party research content.

Source gaps remain attributable

An actual account study would need original instructions, actual receiving times, orders, fills, quantities, costs, marks and applicable policies. Older supplied model totals need their own original cutoff and audit scope. A dated source read is not that cutoff. The desk has no confirmed accountable publisher inbox or security contact; private account statements should not be sent to an inferred address. The machine mirrors derive from visible text and add no stronger assurance. A future correction should identify the changed claim and primary record rather than repeat a summary through more sites as supposed independent confirmation.