Partial Profits and Scaling Out of Gold Trades

What a partial exit actually changes

Closing half at a first target locks a gain and leaves a smaller position exposed. If price then runs to the full target, you earn less than you would have holding everything. If price reverses, you keep more than you would have. That is the entire trade off.

Expressed as a distribution, scaling out narrows it. Fewer large wins, fewer round turns to breakeven, and a smoother equity curve that is easier to keep trading. Smoother is not the same as better.

Which matters more depends on the method. A method whose results come from a small number of large trend captures is damaged by scaling out. A method that wins often and modestly is barely affected.

The arithmetic, with round numbers

Take a trade risking one unit, with a first target at one unit of reward and a second at three. Closing half at the first target and the remainder at the second returns two units rather than three. Closing half at the first target and getting stopped at entry on the rest returns half a unit rather than nothing.

So on a winning sequence you earn roughly two thirds of what a full hold would have produced, and on a failed continuation you salvage a small gain. Whether that exchange is worthwhile depends entirely on how often the second target is reached.

Measure it rather than assume it. The relationship between target distance and stop distance is the part our own testing found mattered most, which is covered in our note on target versus stop distance.

When partials genuinely help on gold

Three situations on gold justify the cost. The first is a scheduled release inside your holding period: taking part off before an inflation print or a rate decision removes exposure to a gap that your stop may not catch.

The second is a session boundary. A trade opened for a London expansion that is still open near the London close is holding through the hours when such moves most often reverse, and reducing there is a reasonable response.

The third is a trade that required an unusually wide stop. If volatility forced a stop eight dollars away, the position is small and the first target is far, so banking part of it at an intermediate level keeps the trade manageable.

A two part plan that is easy to follow

Keep the structure simple enough to execute under pressure. Complexity in an exit plan is how traders end up improvising.

• Target one at the nearest liquidity pool, close half there.
• Target two at the structural destination, close the rest there or at the session expiry, whichever arrives first.
• Stop stays where it was until target one is reached, then moves behind the last higher low rather than to entry.

Write the two prices and the expiry time on the chart before the entry. A plan decided after the position is green is not a plan, it is a reaction to an open profit.

Why the stop should not jump straight to entry

Moving the stop to the entry price the moment a partial is taken feels prudent and quietly ruins the second half. Gold retraces several dollars inside perfectly healthy moves, so a stop at entry gets collected by ordinary noise and the remaining position never reaches the far target.

Behind the last structural low is the better place. It respects what the chart is doing rather than what your entry price was, and your entry price is not a level the market knows about.

If the only way you are comfortable holding the remainder is with a stop at entry, that is useful information: the honest version of that trade was all out at target one.

When closing the whole position is better

On a small account, the arithmetic usually settles it. If the calculated size is 0.01 lots, there is nothing to split. At 0.02 you can split once, and the remaining 0.01 earns so little that it is not worth the attention it will consume.

Costs matter too. Two exits mean paying the spread twice on the way out, and on gold the spread is a real fraction of a short target. A method aiming at three or four dollars loses a meaningful part of its result to the extra fill.

Range conditions also favour all out. When price is rotating between known boundaries, the opposite boundary is the destination and there is no trend to leave a runner in.

What scaling out cannot fix

It cannot create an edge. A method with no edge scaled out is a method with no edge and smaller numbers, and the smoother curve makes that harder to notice rather than easier. Expectancy is the measure that answers this, and our note on expectancy sets out the calculation.

It also cannot rescue a bad entry. Taking part off at a small gain to feel better about a position that never should have been opened is a habit that produces a long run of tiny wins and occasional full losses.

And it will not fix the discomfort of holding. If partials are being taken because watching an open position is unpleasant, the position is too large, and the fix is the lot size rather than the exit plan.

How to use this on the chart

Decide before entry whether this trade is a one exit trade or a two exit trade, and let the conditions choose rather than your mood. Wide stop, a release in the window, or a session boundary ahead means two exits. Range conditions, minimum size, or a short target means one.

Then record both exits separately in the log, including what the second half would have returned had you held it all. After thirty trades that column answers the question for your method specifically, which no general rule can do.

If the column shows the runners were consistently worth more than the early exits, stop scaling out. If it shows the opposite, the partial was earning its cost.