Why Does Gold Trading Lose Money Even with a 70% Win Rate? Calculate the Reward-to-Risk Ratio and Costs Together

Author: Thomas | Trading Knowledge | October 29, 2026

"The win rate is 70%, so why is the account still losing?" This is a problem many gold traders encounter during review. The win rate only tells us how many trades ended in profit; it does not tell us how much profitable trades earned on average, how much losing trades lost on average, and it does not include spreads, commissions, or slippage. Reporting only the win rate without other figures makes it impossible to evaluate a method, and it easily leads people to mistakenly believe that the next trade is more likely to be profitable. The purely hypothetical numbers below are for illustration only and are absolutely not actual trading results from this site.

Three Numbers Must Be Considered Together

The first is the win rate, that is, the number of profitable trades divided by the total number of trades. The second is average profit and average loss, which can be calculated by averaging closed trades separately. The third is trading costs, including spreads, commissions, overnight fees, and actual execution deviation; the specific items should be based on the account terms and execution records. Comparing only the "take-profit distance" and "stop-loss distance" cannot represent the final net result that enters the account.

Suppose someone made 10 trades, of which 7 each earned 10 accounting units and 3 each lost 30 units. The win rate is 70%, gross profit totals 70, gross loss totals 90, and before costs are deducted there is already a net loss of 20. Conversely, another method may still have a positive sample result even with fewer profitable trades, as long as the average profit is enough to cover the average loss and costs. The example here only demonstrates arithmetic; it does not mean that any method can replicate this distribution in the future.

Use Expected Value to Check the Sample

A simplified expression that is convenient for review is: average result per trade = win rate × average profit − loss rate × average loss − average cost. If the win rate and profit/loss amounts have already been calculated on a net basis, costs must not be deducted again. Educational materials from the Chicago Mercantile Exchange also discuss profit frequency and average profit/loss in the same formula, rather than treating a high win rate as the whole answer.

Actual records also need attention to sample size. Ten trades may be only coincidence; a one-week win rate cannot be used to label a strategy as "effective." Different market environments will also change results: if small profits during a ranging period encounter one large loss during a trending market, past averages may become distorted. Statistics can be separated by trading environment, time period, and whether execution complied with the rules, rather than mixing all trades into one attractive percentage.

A Higher Reward-to-Risk Ratio Is Not Always Better

Some people hear that "the reward-to-risk ratio must reach 2 to 1" and therefore first place the take-profit far away, then explain the entry backward. Setting a distant target does not increase the probability of reaching it; distance, hit rate, and stop-loss trigger frequency affect one another. It is more useful to first define invalidation levels and reasonable targets based on market structure, then check whether the actual average profit and average loss in historical samples are close to the plan. CME lists the risk-reward ratio as a risk tool in a trading plan, but no fixed ratio can guarantee profit.

For gold CFDs, execution differences also need to be taken seriously. When the quote crosses the stop-loss, the actual execution may deviate from the preset price; during high-volatility news events, spreads and slippage widen, and the paper reward-to-risk ratio may no longer hold. FINRA's investor education on stop orders also reminds that in fast markets the execution price may differ significantly from the stop trigger price. Specific order mechanisms should be checked against the terms of the platform used, and the rules of other markets cannot be directly applied to XAUUSD.

How to Make a Review Table

For each trade, record the entry reason, planned risk, planned target, actual execution price, fees, net result, and the reason for deviation from the plan. At the end of the month, calculate the number of profitable trades, average profit, average loss, and net result per trade separately, and mark unusually large losses or unusually large gains separately. If there is an extreme event in the sample, averages alone should not be used to conceal tail risk; maximum consecutive losses and the account's tolerance should also be examined. The win rate is one column describing past samples, not a promise predicting the next trade's result.