A range-reentry case study with a shared adverse-excursion stop

What the archive specified

The archive records a range-reentry workflow on a bond-futures sample. Range-trading is a signal that treats a near-flat short-horizon slope plus a close near a short average as permission to enter back toward that average for the system holding period.

Maximum adverse excursion is then used as a filter on that same trade set. A stop-loss is taken from the excursion split and capped as a share of account equity.

The range-reentry signal

The range-reentry signal was specified as follows. When a 10-day trendline-slope sits inside a narrow band around zero, and the close sits inside a defined band around the 10-day simple average, the rule enters toward that average.

Trendline-slope is a short-horizon regression slope used as the flatness test. A narrow band around zero marks a nontrending state. The second condition is the entry-range: the maximum distance from the short simple average at which a mean-directed range entry is allowed.

Scope of the test

The evaluated procedure covered only range reentry. Reverse-on-breakout and add-on rules were left out of the test.

Unstopped range-reentry signals clustered in nontrending stretches of the bond-futures sample. Adverse-excursion stops were then applied to that same trade set.

Reading adverse excursion

Maximum adverse excursion records how far price ran against each closed trade so winning-trade and losing-trade pain can be compared before a stop is set.

Adverse-excursion plots of the unstopped range trades showed a sharp split between winning-trade and losing-trade distributions. That split was used as the basis for placing stops.

Stop-loss size and capital

A stop-loss is a pre-trade and in-trade bound on loss size. Here it is taken from the adverse-excursion split and capped as a share of account equity.

Stops were examined in a one to one-and-a-half point band (32 to 48 ticks). That loss was required to stay under 2% of trading capital, which implied $75,000 of capital for a 48-tick, $1,500 stop.

The same distance as a reverse reference

The same adverse-excursion stop distance, about $1,500 in this case, was proposed as the reference both for exiting a failed range trade and for reversing if price moved beyond the range. The reverse-on-breakout rule itself was not part of the evaluated procedure.

Volatility-dependent gates

How flat the slope must be, and how far from the average an entry may occur, were treated as volatility-dependent and market-specific. Arbitrary values were used to demonstrate the analysis sequence rather than a fitted optimum.

The analysis sequence

The prescribed analysis sequence was to define entry and exit, measure results and adverse movement, set stops and capital requirements, then inspect whether the bounded result met a pre-set adequacy test.