A gold stop loss is useful only when connected to quantity, execution assumptions and an affordable loss budget. Typing a price into a ticket does not establish the cash outcome. This guide works through a margined XAUUSD CFD example, separating the loss in a planning worksheet from the loss that an actual fill can produce.

All prices, contract sizes, costs and exchange rates below are hypothetical. They are neither current quotations nor recommended account settings. No stop distance or risk percentage is suitable for everyone, and controlling a planned loss does not establish that a trading method has positive returns.

1. Three numbers with different jobs

The loss budget is the money used to size a position. The stop price is the level used to activate an instruction under its terms. The execution price is the price actually received. A worksheet saying “risk USD 100” needs an assumed exit price and an explicit list of included costs; it is not a contractual promise.

Ordinary stops can execute worse than their target during slippage or gaps. Keep the planned outcome and adverse execution scenarios in separate columns. Otherwise a precise-looking spreadsheet may conceal an assumption that every order fills at exactly the requested level.

2. Define what invalidates the idea before choosing lots

Write down why the trade is being considered and what would make that reasoning no longer applicable. A range-based idea, for example, needs a defined range and time horizon. A technical level remains a hypothesis, not a barrier that the market must respect. Record the rule before the outcome is known so it can be reviewed consistently.

Then calculate a quantity compatible with the chosen distance and budget. If the smallest permitted position is already too large, skipping that setup may be the workable result. Compressing the stop solely to accommodate a desired lot size changes the strategy. A closer stop may simply respond more often to ordinary movement rather than improve the underlying decision.

3. Match the chart to the relevant side of the quote

Our examples buy at ask and sell to exit at bid; a short sells at bid and buys back at ask. A chart displaying one side or a midpoint may differ from the quote relevant to an exit. Check the actual trigger and execution conditions for your account rather than assuming the chart price activates every order.

If an exit looks inconsistent with the chart, retain the order ID, timestamp and time zone, quote history and account statement. Ask whether the chart and trigger use the same price stream. A widening ask matters to a short even when a bid-only candle looks unchanged. Investigating that distinction is more useful than assuming either a platform error or flawless execution without evidence.

4. Convert a cash budget into ounces

For this simple USD-per-ounce model, ounces = (loss budget − estimated additional costs) ÷ adverse price distance per ounce. Then lots = ounces ÷ ounces per lot. Distance must be positive and the money units must match. If costs consume the budget, the formula provides no remaining allowance for a price loss.

Real costs may depend on quantity, holding time or a minimum charge, so recalculate them after rounding the quantity. If entry and exit prices already reflect actual bid/ask execution, do not subtract the spread again. Include separately charged commission and financing only where they are not already accounted for. This is a planned-price calculation, not a bound on gap losses.

5. Worked example: rounding down leaves unused budget

Assume equity of USD 5,000 and a worksheet budget of USD 100, equal to 2%. That percentage is arithmetic, not a recommended allocation. A long fills at an ask of 3,002 and has a planned stop at 2,982, a distance of USD 20 per ounce. Assume USD 10 of total additional costs, fixed solely for this exercise.

The calculation gives (100 − 10) ÷ 20 = 4.5 ounces. With a hypothetical 100 ounces per lot, this is 0.045 lots. If both the minimum and increment are 0.01 lots, round down to 0.04, representing four ounces. Planned loss becomes 4 × 20 + 10 = USD 90. Rounding up to 0.05 creates a modeled USD 110 loss and exceeds the budget.

Hypothetical sizing at a USD 20 distance with USD 10 additional costs
LotsOuncesModeled lossBudget check
0.044USD 90Within USD 100
0.0454.5USD 100Invalid assumed increment
0.055USD 110Above USD 100

The unused USD 10 is not an instruction to add exposure. Nor is it enough to absorb every possible execution difference. Verify contract size before copying any lot figure: four ounces and 0.04 lots are equivalent only under the stated contract assumption.

6. Stress-test the actual exit, not just the trigger

Keep the four-ounce long entered at 3,002 and stop level of 2,982. Each row is a separate hypothetical sale price. Total additional costs remain USD 10 with no other charges in this exercise.

One stop level, different realized losses
Actual exitPrice lossLoss including costs
2,982(3,002 − 2,982) × 4 = 80USD 90
2,980(3,002 − 2,980) × 4 = 88USD 98
2,972(3,002 − 2,972) × 4 = 120USD 130

The last row exceeds both the USD 90 planned outcome and the USD 100 budget. Reducing quantity moderates exposure but does not guarantee that cap. Test several adverse scenarios across the whole account. The worst row in an illustrative table is not the worst event that the market can produce.

7. Reverse the price logic for a short

Suppose a four-ounce short opens at bid 3,000 and buys back at ask 3,020. Price loss is (3,020 − 3,000) × 4 = USD 80, or USD 90 including the assumed additional USD 10. A higher actual repurchase price increases the loss. Do not carry a long-side formula into the worksheet without changing the direction.

Review buy versus sell, quantity, order type and stop side before submission. A wrong-side price can be rejected or represent a different instruction. Practice helps check the workflow, but simulated fills do not establish what live execution will be like during a fast move.

8. Different order names imply different trade-offs

Where supported, a stop-limit activates a limit order and may remain unfilled. Limiting an acceptable execution price therefore does not ensure an exit. Confirm whether the facility is offered for closing this particular instrument and account rather than assuming that a generic platform description applies.

A guaranteed stop is a separate facility with specific contractual terms and possible premiums. Check eligibility, supported instruments, minimum distance and amendment rules. An ordinary stop field does not establish that a guarantee exists or that such a facility is available in your jurisdiction.

9. Moving the stop changes the budget

With four ounces, a USD 20 distance and USD 10 costs, modeled loss is USD 90. Widening the distance to USD 40 without changing quantity produces 4 × 40 + 10 = USD 170. Editing one field can therefore increase planned loss even without adding another position.

Any amendment needs a fresh review of the rationale, quantity, costs and combined account exposure. Moving a stop to entry is also not the same as guaranteeing a net breakeven exit. Fees and execution differences remain relevant. Label whether “breakeven” means a price level or an after-cost result.

10. Verify how a trailing stop operates

A trailing rule adjusts an exit level as specified conditions develop. Check whether that calculation runs on a server or depends on an active local terminal. Record its activation condition, distance unit and behavior after disconnection; do not assume every platform implements the same mechanism.

Know which accepted instruction remains if your device goes offline. A drawing or an automation setting alone is insufficient evidence. Trailing an ordinary stop does not turn it into a guaranteed fill, and tightening the distance is not proof of a better long-run outcome. Compare consistent rules using recorded trades rather than one attractive historical chart.

11. Review the portfolio and margin separately

Three long gold tickets can be three expressions of the same price exposure. Add their planned losses and consider simultaneous adverse fills. That sum is still not a maximum during a gap. Other instruments can also share a driver, so counting tickets is not a meaningful measure of diversification.

Margin is a collateral requirement, not the stop budget. Four ounces at 3,002 have USD 12,008 notional value. A hypothetical fixed 20:1 ratio gives initial margin of USD 600.40, distinct from the USD 90 planned loss and from any maximum loss. Check account closeout conditions separately; having enough collateral to open does not validate the risk budget.

12. Keep account and household currencies consistent

If savings are in another currency, size in account currency first, then translate the outcome using an explicit rate. At an invented rate of ten local units per USD, a USD 100 budget corresponds to 1,000 local units before conversion charges. That is a unit example, not a current exchange quote.

Do not divide a local-currency budget directly by a USD-per-ounce distance. Also distinguish dollar accounts from cent units and ounces from other gold weight conventions. A familiar app language does not establish contract size, governing jurisdiction or local eligibility.

13. Make the worksheet and journal auditable

  1. Record product, account currency, ounces per lot, minimum and increment.
  2. Write the entry rationale, exit condition and cash budget.
  3. Calculate quantity, round down and recompute actual costs.
  4. Test worse exits and combined exposure across open positions.
  5. Verify acceptance, order ID, quantity, level and any expiry; a chart line is not confirmation.
  6. After closure, retain entry, requested stop, fill, costs, timestamps and amendments.

Separate process mistakes from an ordinary losing outcome under the rule. One loss does not prove that the stop should be wider; one winner does not justify a larger size. If you use a daily loss rule, specify its calculation and enforcement in advance. A written rule does not automatically configure the platform to stop new orders.

14. Common questions

How many points should a gold stop use?

There is no universal number. Identify what a point means on the account, then evaluate distance with the trade hypothesis, costs and quantity. Copying point counts across specifications can change the cash exposure.

Does a rebound after an exit prove the stop was wrong?

No single outcome establishes that. Review a consistent sample, including costs and the correct quote side. Removing the next stop out of frustration changes the entire risk profile.

Does a smaller position eliminate the need for an exit plan?

It reduces price sensitivity per dollar of movement, but does not say when or why to exit. The decision rule still matters.

Does the formula guarantee the cash limit?

No. Its output depends on assumed prices and costs. Execution and account terms must be checked alongside the arithmetic.

Educational information. Sources checked 28 September 2026. Provider examples apply to the stated service regions and are not endorsements or statements of local product eligibility.