Gold XAUUSD volatility: units, position size and execution

A gold-volatility plan needs more than a directional view. Define what the instrument’s price units mean, how an order becomes a cash exposure and which executions the account permits. Then separate the planned trade from the actual fills. No fixed waiting time, ATR multiple or risk percentage guarantees a profitable move or protects a prop-firm limit.

Measure range without treating it as a direction forecast

CME’s gold overview identifies monetary, economic and geopolitical factors, including employment and inflation releases. Those factors do not provide a one-to-one “strong dollar means gold must fall” rule. A chart’s past range and a hypothetical news scenario are different from evidence of a repeatable trading edge. This guide makes no average event-range, pass-rate or expected-return claim.

MetaQuotes defines ATR as a moving average of true ranges. True range compares high−low, absolute high−previous close and absolute low−previous close. In a fictional bar with high 2,040, low 2,028 and previous close 2,025, TR = max(12,15,3) = 15 dollars per ounce. That single bar is not the ATR itself.

Record the symbol, data feed, timeframe, completed-bar convention, period and averaging method. An H1 ATR and H4 ATR measure different intervals. If an observed ATR were 22 and a test used a 1.5 multiple, the stop distance would be 33 dollars per ounce. That arithmetic does not establish that the next move stays inside 33, or that this is the optimal stop.

Resolve price, tick and lot units first

XAUUSD is a symbol label, not a universal contract specification. Read contract size, tick size/value, quote precision, account currency and volume minimum/step on the actual platform. A dollar price movement, a platform point and a trader’s “pip” need not be the same unit. CME Gold futures GC represent 100 troy ounces and MGC 10; those exchange contracts do not define every OTC XAUUSD lot.

Fictional USD-account CFD: assume 1 lot = 100 ounces and a 0.01 dollar-per-ounce tick. One tick is 100 × 0.01 = $1 per lot. An $8 move produces $800 per lot, $200 at 0.25 lot or $40 at 0.05 lot, before costs. On a $5,000 base, $40 is 0.8%, not 8%. These are explicit assumptions, not a claim about your provider.

For a chosen $500 loss budget, $10 stop distance and $10 round-trip commission per lot: loss estimate per lot = 10 × 100 + 10 = $1,010. $500/$1,010 ≈ 0.49505 lot; with a 0.01 step, round down to 0.49. Estimated loss is $494.90. Other costs and slippage still matter. Check margin and remaining daily/total loss headroom separately; do not size from the entire advertised loss allowance.

Check news permission before placing orders

Checked on 16 October 2026, FTMO’s CFD news FAQ restricts execution on affected instruments for two minutes before and after selected releases, with exceptions for Swing and Challenge/Verification accounts. XAUUSD is listed for the selected USD events. The restriction includes Stop Loss, Take Profit and other pending-order executions. Holding an older position does not make an execution within the window permissible. The listed exceptions still require avoiding forbidden practices.

Use the exact programme’s calendar, timezone and current conditions. Do not hard-code “EST all year”, assume every release is at the same time, or treat this FTMO example as another provider’s policy. An EA pausing new entries does not prevent an existing stop or target from executing. Two opposite pending orders may both fill during a reversal; cancellation is not guaranteed to win the race. A pair of orders is not automatically an options straddle or iron condor.

Audit partial exits from actual fills

Separate fictional pre-cost worksheet, not a historical trade: short entry $1,947, stop $1,957, 0.50 lot and the assumed 100 ounces per lot. Initial stop exposure is 0.50 × 100 × $10 = $500 before costs. Assume every exit below actually fills at its stated price.

The closed volume totals 0.50 and gross profit $680: 1.36 times the $500 initial pre-cost risk, not the full 2R shown by one target. A price reaching $1,929 would not prove a fill at $1,927. Slippage, commission and swaps change net results. Moving a remaining stop to entry does not reimburse costs or remove execution risk.

Evaluate the record without a magic trade count

Log event and source timestamp, account permission, spread, bid/ask entry and exits, quantity, rejected/cancelled orders, costs and maximum adverse equity movement. Keep skipped events and losing trades. A backtest must reproduce price units, timezones and execution assumptions; compare performance on data not used for parameter selection. Twenty, fifty or two hundred events alone do not establish significance or future profit.

Fictional expectancy: a 35% win probability, average winner 1.9R and average loser 1R give 0.35 × 1.9 − 0.65 = 0.015R before costs. A cost of 0.02R per trade changes this to − 0.005R. Therefore a reward/risk ratio below 2 does not automatically mean a negative gross result, and a positive gross estimate can still be negative after costs. Estimated probabilities remain uncertain.

Check each documented function of the customer-operated JPTC EA Hub on the exact file and platform; no gold module, latency, news filter or permission is implied beyond its documentation. The separate JPTC Algo is hosted by JPTC: its customer installs or runs nothing and needs no VPS. Company-published broker-statement figures describe the stated Algo period; they are not an independent audit or a test of this event strategy.