Why holding without a reason is the expensive habit
An open winner costs nothing to hold, which is exactly why traders hold too long. Gold spends large parts of the day going sideways after a move, and giving back half of a clean run while waiting for a target that nobody else is aiming at is a slow and common way to turn a good month into a flat one.
The opposite failure is just as costly. Closing at a small gain because the position is green removes the few outsized results that pay for the losing trades. Both errors come from exiting on feeling rather than on a condition written down in advance.
So the question is not whether to take profit, it is which condition you are waiting for. Once that condition is named, the exit becomes mechanical and the screen becomes much quieter.
Exit one, the opposite liquidity
Price moves between places where orders rest. If you entered long from a swept low, the realistic destination is the cluster above: the prior session high, the untested high inside the range, the level that produced the last reaction. Mark those before entry and the first target picks itself.
On gold these pools are often only a few dollars apart during Asia and much further apart once London is in. That is why the same setup deserves a nearer target overnight and a further one in the first hours of London, when the dollar range per hour expands.
The rule for leaving is simple: when price reaches the pool and the candle that gets there closes poorly, with a long wick and no follow through, the move has done its job. Our Today Levels zones mark the strong daily zones on the live gold chart, and the opposite zone is usually the honest first destination.
Exit two, the measured move
Patterns imply a distance. The height of a range projected from the break, the depth of a head and shoulders measured from the neckline, the length of the impulse repeated from the pullback. These give a number in dollars rather than a vague hope.
Gold overshoots these projections on trend days and stops well short of them in a chop. The projection is therefore a planning figure, not a promise. Our own work on distance estimates showed that the ratio between target distance and stop distance mattered more to results than the exact projection method, which is covered in the note on target versus stop distance.
A workable compromise is to treat the measured distance as the point where you stop expecting more and start watching how price behaves, rather than as a price you must be filled at.
Exit three, the session clock
Gold does most of its directional work in two windows: the first hours of London and the period around the New York morning data releases. After the London close, moves frequently reverse as that liquidity leaves, and late New York is thin enough that a target two dollars away can sit untouched for hours.
So a trade taken for a London expansion has a natural expiry at the London close, whether or not the target printed. A trade taken on a data reaction has a shorter life again, often resolved within the first hour.
This is the exit most retail traders never use, and it is the one that protects the most profit. The move was produced by participation. When the participation leaves, the reason for the trade has left with it.
The time stop for a position going nowhere
If a setup was valid, price should move away from your entry reasonably soon. When it instead drifts for two or three hours inside a two dollar band, the information is that nobody else agreed with the read.
A time stop closes the position at or near breakeven after a fixed period and frees the attention for a better window. It is not a loss, it is an admission that the condition never arrived.
Pick the period from the timeframe you entered on. An entry built on five minute structure that has gone nowhere for ninety minutes is stale. A daily swing entry deserves days, not hours.
What trailing a stop really does
Trailing converts an open profit into a floor and gives up the right to the full target. On gold that trade off is sharper than on slower markets, because normal retracements inside a trend day can be five or six dollars deep and will collect a tight trail before the move continues.
If you trail, trail behind structure rather than behind a fixed dollar amount: under the last higher low in an uptrend, above the last lower high in a downtrend. That way the stop only moves when the chart has actually improved.
A trail placed too close is a hidden exit decision. It usually means you wanted out but preferred the market to make the choice for you.
What the chart cannot tell you about leaving
No chart reading will tell you the high of the day in advance. Whatever the exit rule, some trades will continue for another twenty dollars after you close and others will reverse one tick before your target. That spread of outcomes is permanent and no method removes it.
Scheduled releases make it worse. A target sitting just beyond a level can be reached and exceeded in seconds on a print, or the same level can hold for a week. The honest position is that exit rules manage averages, not individual trades.
What you can control is consistency. Changing the exit logic after every disappointing trade leaves you with no measurable record at all, and nothing to improve from next month.
Deciding the exit before the entry
Write the destination on the chart before the order goes in. One line for the pool you expect to reach, one figure for the measured distance, one time at which the trade expires. If those three disagree wildly, the setup is weaker than it looked.
Then let whichever condition arrives first do the work, and record which one it was. After thirty trades the log will show whether your structural targets or your session expiries are producing the results, and that is a far better guide than any rule read in an article.
Placement of the stop that protects all of this is a separate question, covered in our guide to stop loss and take profit placement on gold.
FAQ
Should you take profit on gold every day?
Only if your trades are intraday by design. A position opened for a London expansion has no business being carried into a thin late New York session, so closing it the same day is correct. A swing position built on daily structure should not be closed because the day ended. Match the holding period to the timeframe the setup came from.
How many dollars should a gold target be?
The distance comes from the chart rather than from a fixed figure. Measure to the next liquidity pool or project the pattern height, then compare that distance with your stop. If the target is not comfortably further than the stop, the trade is not worth taking, no matter how convincing the entry looks on the lower timeframe.
Is it better to use a fixed target or to trail the stop?
Fixed targets produce a cleaner record because every trade ends for a stated reason. Trailing captures more of the rare large trend day but gives back part of every normal one. A practical answer is a fixed target for range conditions and a structural trail only on days where price is clearly expanding in one direction.
What if price reverses right after you take profit?
That is evidence the exit worked, not that it failed. The purpose of a target is to leave while buyers or sellers are still willing to transact at your price. Judging an exit rule by the few trades that continued without you leads to holding everything, which is how open profit disappears during the London close reversal.
ⓘ See these ideas on real price: open the free XAUUSD live chart.