Verification standard

How a track record is reviewed

A six-metric reconstruction, a source hierarchy, and an explicit boundary around what the evidence does and does not prove.

Review workflow

  1. Define the claim. Identify the person, account scope, period, currency basis, and whether the claim is personal, competition, model, product, or customer-account performance.
  2. Inspect the record. Check period coverage, equity continuity, deposits and withdrawals, closed and open positions, and data consistency.
  3. Recalculate the metrics. Compute the published figures from the available record rather than relying on a screenshot headline.
  4. Cross-reference external evidence. Link championship, regulatory, or administrator records where they independently support the claim.
  5. Publish with boundaries. State the source type, review date, and limitations beside the result.

Six reported metrics

Total return

Percentage change over the defined period, adjusted for external cash flows where the data permits.

Maximum drawdown

Largest peak-to-trough decline during the covered period.

Sharpe ratio

Risk-adjusted return calculated from the available return series, with frequency disclosed where material.

Win rate

Positive closed trades divided by total closed trades.

Trade count

Number of closed positions included in the covered period.

Exact period

First and last dates represented by the evidence package.

Evidence levels

LevelEvidenceSupports
Direct record reviewAccount or broker recordMetrics for the stated account and period
Cross-referencedIndependent competition, regulatory, or administrator publicationThe result published by that source
Cryptographically publishedHash and trusted timestampExistence of material before a later outcome
Self-reportedMarketing copy, screenshot, or unsupported dashboardNo independent verification

What a review does not prove

A reviewed historical record does not guarantee future performance, strategy reproducibility, liquidity, capacity, or identical customer outcomes. A person's result does not establish a product's result. A championship result does not become a personal calendar-year result merely because the same person produced both.

Corrections and versioning

Material corrections should update the human-readable page, structured dataset, feeds, sitemap, and AI-readable mirrors together. Withdrawals are logged on the corrections page.

Questions about an evidence package can be sent to [email protected]. Acceptance depends on whether the material can support a defensible public claim.