Evaluating maximum-adverse-excursion stop reversals with short time stops

What the evaluation changed

The evaluation tested reversing the position automatically when price reached the maximum-adverse-excursion stop, instead of using that stop only to flatten. Maximum adverse excursion is the largest unfavorable move recorded against a live position. It was used both to site the original stop and to inspect how far a reversed trade then ran against the new side.

The recorded maximum-adverse-excursion values were described as likely overstated because intraday prices were not available.

Adverse movement after the reverse

A reversing trade is a new position taken the opposite way when the original trade hits its maximum-adverse-excursion stop. Reversing at that stop can still produce a large adverse move when price recrosses the original breakeven line and then turns back against the new position inside the same trade horizon.

The breakeven line is the original trade's zero result. Recrossing it after a reverse can create a large adverse excursion on the new side. The trade horizon is the planned holding window of the first setup. The evaluation asked whether the reverse should inherit that window or be closed much sooner.

Price stops on the reversing trade

A stop-loss is a pre-set price limit that ends or flips a position once adverse movement reaches a chosen distance. Here it was examined on the reverse at 32, 48, or 64 points.

On reversing trades in Treasury bond futures, the maximum-adverse-excursion profile of winners and losers supported considering price stops at 32, 48, or 64 points. The write-up stated that reversing-trade losses could be cut anywhere above 32 points while still keeping most eventual winners.

A short time stop for the reverse

A time stop is a rule that closes the reversed exposure after a fixed number of sessions instead of waiting for the original trade horizon to expire. Reversing trades were observed to become profitable quickly, usually within two days.

In the time-stop test, an entry-day exit was only marginally profitable. Waiting one extra day built a clearer result. Waiting two days improved overall profitability by another 50 percent, and further days added little.

Reversals still open after 20 to 25 days were not observed to become large winners, and beyond about 10 days a later winner could not be distinguished from a later loser in advance.