#27 — Zone Recovery trail-lock redesign

Hypothesis — what it tests

Does an aggregate-P&L trailing lock (lock at small profit, trail all legs together, close everything flat) beat the PSAR-escape exit for Zone Recovery?

Description

Zone Recovery (ZR) places alternating recovery legs as price crosses a zone's boundaries; the original design exited via a PSAR 'escape' at target levels. A live post-mortem found the fatal flaw: the PSAR escape fired on ANY target crossing with no awareness of aggregate portfolio P&L, once catastrophically closing a net-long 160-unit position at the lower target (the frozen trade ledger later showed a −5,159.5-pip psar_escape close). The redesign, to the user's specification, ripped out PSAR entirely: the only exit is an aggregate-P&L trailing stop that arms once the whole cycle is up min_lock pips and tightens from trail_base toward min_lock as the peak grows, closing ALL legs flat simultaneously — the cycle can only exit in profit once locked. A 144-combo sweep, then a corrected 168-combo v2 requiring min_lock ≥ spread, validated min_lock=2.0/trail_base=3.0/trail_max=30.0 (8/168 passing all gates) and it was deployed live on accounts 011/012. A companion test confirmed ZR's edge lives in the exit mechanics, not entry timing — directional entry filters actively hurt (H1 SMA10 filter: −4,660 p/d). Note the audit table quotes results in vol-pips (volume-weighted pips), and the broader ZR book was later closed; this row records the design experiment, which was a clean win over the old exit.

Key result

Indicators

Algorithms

Code