Taking Partial Profits on Gold: What 4 Years of Tests Show

Taking partial profits on gold feels safe, but in our tests early partials cost more than they saved. On four years of XAUUSD M15 data, banking profit at 1.5–2R or half-way to target cut hypothetical returns by roughly 20–40%. A single partial at a real structural target, with the rest left to run, did best.

Why traders take partial profits on gold

Gold moves fast. A trade can be +2R one hour and back at entry the next. Taking something off feels professional and makes the next swing easier to sit through.

But a partial exit changes the shape of your returns. It raises your win rate on paper and trims your average winner. Whether that trade-off pays depends on how your profits are distributed. So we measured it.

What we tested: four exit plans on the same entries

All figures below are from a hypothetical backtest on XAUUSD M15 broker data, July 2022 to September 2026, with spread and slippage on every trade. Entries were identical; only the exit changed. Every version moved the stop to breakeven at 1R.

• The best plan beat "all out at the zone" by 92.7 − 87.1 = 5.6R.
• It beat the 2R/4R/6R plan by 92.7 − 70.6 = 22.1R, or 22.1 ÷ 92.7 ≈ 24% less profit for taking the first third off at 2R.
• The 3R/5R/7R ladder landed close, because its first exit is further away.

In separate tests, early partials at 1.5–2R, or half-way to target, cut four-year returns by roughly 20–40%. Letting trades run longer before a time exit added about 10R. The earlier you pay yourself, the less you get paid.

Why cutting winners early hurts: a few big trades do the heavy lifting

The reason sits in the trade list. The model behind these tests took 146 trades in the backtest:

• 63 winners, 32 breakeven or scratch, 51 losses.
• The 10 biggest winners made 42% of the total R. That is roughly 0.42 × 91.2 ≈ 38R from 10 trades, or about 3.8R each on average.
• The biggest single trade made about +4.6R.
• Winners were held a median of 78 hours. Losing trades were closed a median of 12 hours after entry.

Losses resolve quickly. Winners take days. Take half off at 1.5R on the first morning and you cut exactly the trades that pay for the 51 losers.

This is how a 43% win rate still produced a profit factor of 2.71 in the backtest: the average winner is much larger than the average loser. Early partials attack that gap. If you want the maths behind win rate and reward, read our guide to win rate vs risk-reward and expectancy.

Where the exits actually came from

In the backtest, 49 trades hit the stop, 38 took 50% at the zone with a trailed runner, 30 hit the five-day time exit, 27 finished at breakeven and 2 reached the runner target. Thirty trades were still open when the five-day limit closed them: moves that took days to develop, which a fixed early target would have cut short.

The partial that did work: at structure, not at a number

Our best plan still takes a partial. The difference is where. It takes 50% at the opposite higher-timeframe zone, a level the market has a reason to react to. It does not take profit at an arbitrary 1.5R or 2R.

An opposing imbalance is a logical place for a reaction, so banking half there locks in a gain. The trailed runner catches the cases where gold keeps going.

A fixed 2R target does not know where the next zone is. Sometimes it sits right before a big expansion. The market does not care about your R multiple.

Breakeven after a partial: the hidden trade-off

Most traders pair a partial with a stop moved to entry. That combination has a cost. All four plans in our table moved the stop to breakeven at 1R, and 27 trades still finished at breakeven.

Moving to breakeven protects capital. It also means some trades get stopped at entry before they run to target. When you add an early partial on top, you shrink the winner twice: once by selling half early, and again when the remaining half gets scratched.

• Move to breakeven on structure where possible, not on a fixed tick count. Gold's noise has widened in 2025–2026.
• Decide the rule before entry. Changing it mid-trade is how discipline leaks.
• Count scratches honestly. A breakeven trade is not a win.

A trade at breakeven also opens a different option: adding a position instead of taking size off. We tested that in the backtest and it raised results to about +141R. See pyramiding gold trades with a Free Add for how that worked.

When partial profits on gold still make sense

The data is not the whole story. There are sensible reasons to take a partial early.

Psychology

If holding a full position through a 50% retrace makes you close everything, an early partial may be the lesser evil. A plan you can follow beats a better plan you abandon. Just know the price: in our tests, roughly 20–40% of the long-run return.

Prop firm targets

On a prop challenge, the goal is to hit a target without breaching a loss limit. In our 2026 prop simulation, a $100k two-step account had an 8% Phase 1 target and 1% risk per trade. That is 8R to pass. Near the target, locking in part of a winner can be rational, because a pass is worth more than an extra 2R. Our guide to passing a prop firm challenge on gold covers this. Check your firm's current rules before relying on any figure.

How to test partial profits on your own strategy

Do not copy our numbers. Test your own exits on your own entries.

1. Freeze your entries. Change only the exit.
2. Run at least three versions: all out at target, one partial at structure with a runner, and an early fixed-R partial.
3. Compare net R, profit factor and max drawdown, not win rate alone.
4. Sort your winners by size. If a small group makes most of the profit, early partials will hurt.
5. Check holding times. If winners take days and losers take hours, give winners room.
6. Watch for outliers. If one trade makes a version look great, it is not robust.

Our approach

We use the exit that tested best: half at the opposite zone, the rest trailed. The full four-year hypothetical backtest is on our performance page, and every live signal from the model behind Pulse Signals is recorded automatically, losses included, on the live results page.

FAQ

Is it good to take partial profits on gold?

It depends on where you take them. In our hypothetical backtest, early partials at 1.5–2R cut four-year returns by roughly 20–40%. A single partial at the opposite higher-timeframe zone, with a trailed runner, did best.

Why do early partial profits reduce returns?

A small number of large winners make most of the profit. In our backtest the 10 biggest winners made 42% of total R, and winners were held a median of 78 hours. Early partials shrink exactly those trades.

Should I move my stop to breakeven after a partial?

It protects capital but has a cost. Some trades get stopped at entry and never reach target. If you do it, base it on structure and decide the rule before you enter.

Do partial profits help on a prop firm challenge?

They can near a profit target, when passing matters more than extra R. Size risk so a normal losing streak cannot breach the daily or maximum loss limit, and check your firm's current rules.

This article is educational and not financial advice. Trading gold and leveraged products carries a high risk of loss.

Trading involves substantial risk. This is educational content, not financial advice. See the risk disclosure.