03 / PERFORMANCE INTEGRITY
If it cannot be reproduced, it is not a track record.
A win rate without the ledger, costs, sample definition and dates is marketing — not evidence. Our measurement rules start with traceability.
RESEARCH FRAMEWORK
Separate the evidence cohorts
Modeled, simulated and executed trades have very different limitations.
HISTORICAL / 01
Backtests
Depend on data quality, selection bias, fill modeling, assumptions and robustness checks.
SIMULATED / 02
Paper & demo
Useful process tests, but not proof of real-world liquidity, funding or execution.
EXECUTED / 03
Verified fills
Require full broker records, timestamps, fees, partial fills, revisions and a documented sample.
The denominator matters.
Cancelled, expired and unresolved signals must remain in the ledger. Any metric must specify which records are included and excluded. R-multiple compares net P/L to initial planned monetary risk where that baseline is verifiable; partial exits and changed stops require explicit policy.
No verified results are published yet.
We currently have no validated complete public performance ledger. We will not invent returns, trading volume, sample size, drawdowns or success rates.
Trading can result in significant financial loss. These materials are educational and not personalized investment advice. There are no verified live signals or automatic executions available on this site at this stage.
SITEWIDE / RESEARCH CONNECTIONS · MEASUREMENT TOOLS
Understand the mathematics before trusting a claim.
PROOF OVER PRESENTATION
A result is useful only when its limitations are visible
Evaluate the data lineage, optimization history, execution friction and what happens when the assumptions stop being favorable. One profitable plot is not an independently verified track record.
Protect the untouched evaluation window.
Model the difference between chart and fill.
Report samples by the conditions in which they occurred.
Hypothetical examples and local calculations are not independently verified executed performance.
What constitutes meaningful trading-performance evidence?
A working calculator, a backtest and a reconciled live-account track record prove different things. No design screenshot, imported CSV or manually entered journal row should be presented as independently verified returns.
Four evidence tiers that must never be combined silently
| Tier | What it supports | What it cannot prove |
|---|---|---|
| Hypothetical model | Arithmetic under explicitly stated assumptions | Actual fill quality, viability or realized returns |
| Historical backtest | Behavior on a defined dataset and rule set | Out-of-sample durability, future liquidity or execution |
| Forward demo | Observed behavior during a declared non-live period | Real-capital fills, account risk or independent verification |
| Reconciled live evidence | A limited audit of documented broker and account records | Future returns, suitability or absence of hidden risks |
Why sample size and costs matter
Consider two hypothetical trades: one net result is +1.88R, another is −1.00R. Their combined total is +0.88R, the arithmetic mean is +0.44R and the positive-result frequency is 50%. Two observations do not establish a statistically robust advantage. If the sequence order changes, the maximum decline of a cumulative R chart can also change.
Do not mistake a cumulative R sequence for a percentage equity curve. Equity results require position-sizing history, deposits, withdrawals, currencies, fees and timing. Report missing and cancelled setups as part of any complete strategy lifecycle.
A credible methodology would disclose
- Data rights, timestamps, time zones, bid/ask and execution assumptions.
- Backtest and out-of-sample separation, plus revisions to the strategy rules.
- Gross and net outcomes, all material fees, gaps and slippage.
- Sample size, exclusions, cancellations and ambiguous observations.
- Account reconciliation and reviewer identity for any future verification claim.
Explore local, unverified Journal Insights ↗
Research boundary: Examples here are illustrative and educational. They do not contain licensed live quotations, individualized investment recommendations, real-money trade instructions or verified broker results.

