Time-based exits: the distribution of how long trades last

[Amanhasnoname]

Winners and losers have different time profiles. Many strategies find that winning trades reach the target quickly, while losers linger. The distribution of holding times can be exploited.

Plot the cumulative distribution of the time to target for winners. If 90 percent of winners hit within 40 seconds, then a trade still open at 60 seconds is more likely to be a loser, and the probability of eventual success is low.

A time stop closes the trade at a set duration, cutting the tail of long losers. Test different durations and measure the effect on expectancy.

Interactions with the stop level matter. A time stop may exit at a small loss before the stop would have been hit, or at a small profit that would have reversed.

Time-based rules should be conditioned on volatility, since the speed of price movement varies. A scaled version uses an expected time based on the recent range.

The approach reduces exposure to unfavorable situations and increases capital turnover.

What does the distribution of winner durations look like, and where would a time stop fall?