A Rule-Based Swing Exit
Specify price, time and exception handling before outcomes are visible.
One Declared Policy
Define the stop, objective, maximum holding clock and response to data failure. If exits are being compared, retain a predeclared research process instead of selecting the best exit separately for each historical trade.
Operational Evidence
A chart level is not an accepted broker order. Confirm session eligibility and order persistence, especially around events. Ambiguous bars need a declared sequencing policy or an unresolved outcome.
Primary Sources
The selected exit is not a daily improvisation
A rule-based swing exit needs its exact version, data availability, quantity ownership and maximum-hold clock. A completed-session routine can advance a holding clock without becoming a daily profit floor or daily stop-management authority. Canonical minute-execution ownership remains the required baseline for actual APEX exit analysis; this publication does not modify it. The new cash lens has a manually declared valuation price, not an exit engine or a broker order. A split-equivalent stop is arithmetic reference only. It cannot establish that an outstanding stop was adjusted, accepted, active or executable under the actual source policy.
A fictional after-action ambiguity
An invented original position has 100 shares at 100 and a planned stop 97. A two-for-one action makes the equivalent reference 48.5 on 200 assumed units. If a broker record still displays 97, the website cannot decide whether it is an old display, cancelled order, auto-adjusted representation or actual dangerous instruction. The next source request is authoritative order state and effective adjustment policy. The tutorial does not send a replacement stop or guess an account owner. Its adverse-price case separately calculates assumed wealth change from 46 and declared cash/costs. Those calculations are complete while the real operational question remains unresolved rather than being answered by a convenient chart number.
Do not reopen an exit tournament after access
Once a real experiment predeclares one exit, later validation should not choose a more flattering alternate exit from the same sample. Holding drift, costs, overlap and exit-adjusted economics can validate or reject the frozen experiment under its authorised contract. They do not redefine the entry objective. Original whole-position native risk and maximum-hold rules remain intact unless explicitly changed in a separate task. A website case, positive arithmetic result or provider narrative grants no exception. Preserve source versions and actual row ownership before a real analysis. The original exit criterion remains, now with concrete adjustment and timing questions rather than a fabricated account test.
A source declaration and an execution differ
A withdrawal message, closing intention, acknowledged request and completed fill are different records. The source-model outcome is not automatically the receiver's funded result. Actual prices, quantities, costs, gaps and trading interruptions need their sources. A stop trigger can lead to an execution price that differs from the reference; a limit adds a different constraint with unresolved execution. The desk has not independently operated a paid provider account or authenticated a complete journal. It retains missing exit evidence instead of inventing a closing price, automatic guarantee or universal service rank to make the criterion appear finished.