STRATEGY / ASIA SESSION
XAUUSD Structure research
One continuous hypothetical account. No monthly reset.
00:00–08:00 UTC · 186 observed sessions
ACCOUNT PATH
Equity & drawdown
Original account, date window only. Direction, outcome and search never redraw a fictional portfolio.
High-water mark carries across months. Neither series reveals unobserved intra-gap price paths. Peak not recovered by cutoff.
MONTHLY ATTRIBUTION
The full picture
Click a month to inspect. Returns use its actual starting balance.
Cause analysis pending. Positive and negative months alone do not establish a market regime or a trading edge.
DAILY OBSERVATIONS
Session calendar
Days are not interchangeable.
Gold cells are positive; red cells are negative. Amounts are account USD. Weekends and unavailable dates are not zero-P&L sessions.
01 Jan and 03 Apr: source-empty, holiday-corroborated exclusions. The historical broker calendar is not confirmed.
INSPECT THE EVIDENCE
Trade ledger
FROZEN IMPLEMENTATION
Rulebook
Recognition & entry
This is the tested provisional detector, not a claim that every rule matches discretionary chart reading. Closed candles only. M5 outer wick pivots use 3 neighbours each side; inner pivots use 1. A correction within the outer origin/objective may form an inner attempt. A closed M1 candle must break its trigger while preserving the correction extreme. M15/H1 context is recorded, not a new filter. Parent lifetime is 120 minutes from confirmation, capped at session end; this limits new entry, not holding time.
One position at a time. Occupied opportunities are skipped, not queued. First eligible quote at least 100ms after readiness; no eligible quote within 5 seconds means no entry. Buy at ASK, sell at BID.
Risk, exits & costs
Planned risk ≤1% of current equity including fees; integer ounces, 5× notional cap. Hypothetical 100oz lot, 20:1 margin and 50% stopout. Stop begins at the confirmed correction extreme. After the objective is reached on a completed bar, confirmed M5 pivots may tighten the stop; changes act on a later quote. Exit at the first eligible observed stop quote, or the 07:59:59.999 UTC as-of snapshot (each quote side no older than 5 seconds), not a new 08:00 tick.
$0.035/oz commission each leg; observed BID/ASK spread already included. Primary extra slippage: zero. 47 losses exceeded 1%; worst realized loss 1.13354%. Planned risk is not a loss guarantee.
Extra $0.05/oz adverse each leg plus doubled commission produces −$29,476.53 with fixed quantities. This is an arithmetic sensitivity, not a resized replay.
EVIDENCE & LIMITS
Data quality
822 trades / 1,644 fills reconciled across 186 sessions. 1,395 ready candidates: 822 filled, 572 occupied skips, 1 missing entry quote. Raw integrity, native-candle coherence and bounded re-fetch checks passed.
5,269 occupied quote intervals exceeded 5 seconds across 480 trades; longest 29.397s. Silence is not proof of an outage. Depth, live fills and intra-gap stop paths remain unknown.
2026 is previously studied research, not clean out-of-sample evidence. The full account lost money despite its late recovery. Do not infer live profitability.
Provenance & units
Source: observed demo-provider history. Native engine: . Data frozen through 21 Sep 2026, 08:00 UTC exclusive. All capital is hypothetical. No broker account identifiers or private source paths are included.
Prices / MAE / MFE: gold USD per ounce. Equity / risk / fees: account USD. Net R: each trade’s net P&L divided by its actual planned risk.
Download sanitized data & source hashes