Short answer: far longer than most people expect, and the spread between typical and longest is enormous. Holding time is the part of a strategy nobody asks about until they own it, and then it is the only thing they notice.
Why the average is the wrong number
Trade durations are not symmetrical. Most close reasonably quickly; a few run for months. One 168-day position drags the mean upward until it describes no trade that ever happened.
The median is the honest headline. It answers the real question - what does holding this normally feel like - and it cannot be inflated by one outlier.
The actual distribution
Why long trades are not a malfunction
A position that sits open for weeks is doing something specific: it is holding a level that has not resolved. The alternative - closing early because it feels slow - converts an unresolved trade into a certain loss.
The cost is real though, and it is not the pips. It is capital occupancy. A position held for two months is capital unavailable for the next setup, and that is a genuine expense even when the trade eventually wins.
What this means for reading any track record
Two strategies with identical pip totals can be completely different products. If one turns its capital over weekly and the other holds for a month, they demand different account sizes and different temperaments.
Ask any vendor for the median hold and the longest hold together. One without the other tells you very little. Ours are above, and every leg behind them is on the track record.