THE METHOD BEHIND THE NUMBERS

Transparent by method.

Three distinct layers

Imported master trade records preserve source prices, lots, timestamps, reported P/L and rounded provider Gain. Calibrated follower-equivalent returns describe the workbook's scaling model. Account simulations replay those returns for an illustrative starting balance. These are not interchangeable.

Versioned calibration

Methodology: sonic-calibrated-v1-79a6e5eff07f. Gross follower return equals source period master P/L multiplied by the source calibration factor (0.00018105061968624878, as a return fraction per dollar of master P/L). Positive source period returns are reduced by the 30% performance fee. The calibration factor comes from the source follower/master lot ratio divided by calibration cash; no arbitrary account balance is introduced.

Monthly source attribution

Original monthly returns are preserved. One −$4 trade opened on 20 June 2025 and closed on 20 September 2025 is attributed to June in the workbook's monthly results. Closed-trade statistics place it in September. Both facts are retained and their difference is audited. The original 595-trade control is stale; 604 unique trades reconcile with the monthly source total.

Dates and coverage

Dates are interpreted as Asia/Dubai (UTC+4), provisionally confirmed by the source owner. The final month remains in progress through the source cutoff. Current-week figures are unavailable if that week has not been imported. Date filters use Dubai calendar boundaries. Performance presets end on the latest confirmed trade date: 7D includes that date and the six preceding calendar days. Selected-range returns apply the existing calibration to closed-trade P/L grouped by closing month, deduct the performance fee from positive monthly portions, and compound those portions. Partial months are calculated from their trades, not prorated. This is a follower-equivalent model, not observed account returns. Workbook monthly attribution remains separate and can produce different compounded totals. No after-cutoff results are included.

Compounding and fees

The simulator uses the stored source monthly gross/net return series. The performance page applies the same calibration and fee rules to the trades in its selected date range. The simulator supports 0–100% reinvestment of each positive month’s profit; the remainder is accumulated as withdrawn cash. At 100%, the full profit compounds monthly. Turning reinvestment off withdraws all positive profits. Losses reduce trading capital in full at every setting, and ending value includes withdrawn cash. This illustrates the chosen settings rather than changing a real account’s withdrawal or lock-up terms. The source specifies no high-water mark or loss carryforward. Weekly net returns are shown only for the explicit source snapshot, and fees are never deducted twice.

Liquidity-provider commission

Liquidity-provider commission: $10 per lot of gold traded, per trade. This volume-based charge is distinct from the performance fee. The source workbook does not separately identify this commission. Whether it is already included in its P/L and monthly returns remains unconfirmed; the historical figures are preserved without an additional deduction. A confirmed change to historical fee treatment would require source reconciliation and a documented methodology revision.

Unavailable metrics

The source does not provide a historical account equity series, intratrade equity, leverage assumption, separate commissions or swaps, or a confirmed gross/net basis for master P/L. Actual maximum drawdown remains unavailable. A compounded monthly index is an illustration, not observed account equity. Rounded provider Gain can be zero even when P/L is nonzero. Win/loss classification therefore uses reported P/L.

Source reconciliation and revisions

Every import is checked for required fields, duplicate identity, monthly totals and calibrated returns before it can replace the public snapshot. A file hash identifies the source, and row observations retain provenance. Exact repeat files do not create duplicate trades. The original source data is preserved.

Updates are manual validated imports; there is no scheduled or live platform feed. The coverage date shows how current the results are. Completed months and the latest in-progress month are distinguished. An incomplete month is not a finalized full-month return.

Changed historical trade facts require correction review rather than silent overwriting. Methodology assumptions are versioned separately from raw records. Accepted revisions should retain the previous source, reason and affected periods in the audit trail. A new method is identified explicitly rather than retroactively changing the meaning of an old one.

Trade statistics

Trades are ordered by closing instant and deterministic fingerprint for ties. Win rate is profitable trades divided by all closed trades with known P/L. Profit factor is positive reported P/L divided by absolute negative reported P/L; without losses it is unavailable. Break-even trades interrupt both streaks. Duration uses the source's reported duration, including its retained 91-day trade. Raw amounts use PostgreSQL numeric values and calculations use decimal arithmetic; rounding occurs for display.