Causal z-scores and the overshoot problem
Define a prior-window reference, measure absolute recovery and avoid calling every correctly directed move a successful reversion.
Observation-count, wall-clock and completed-session horizons are different policies. Freeze the chosen clock and maximum hold before evaluation. A half-life estimate is not permission to extend a losing trade or redefine the record after seeing its exit.
Educational diagnostics, not a signal service. A standardized price displacement is not a calibrated probability; invented examples and user-supplied observations are not verified market performance.
Observation-count, wall-clock and completed-session horizons are different policies. Freeze the chosen clock and maximum hold before evaluation. A half-life estimate is not permission to extend a losing trade or redefine the record after seeing its exit.
For a reversion claim, request the frozen reference definition, observation availability, complete eligible event population, treatment of crossing and overshooting, exit clock and cost model. An alert tool can express a condition without proving that a seller has tested it. Separate a model summary from a subscriber account record.
Write the price series, adjustment policy, prior lookback, standard-deviation convention, reference update clock, event threshold and recovery definition. Save a version before inspecting later outcomes. Count every eligible observation, including zero-variance exclusions, missing bars and events that never return.
Documentation establishes the stated mechanism, not tested trading performance. Checked 9 October 2026; a later version can differ.
Define a prior-window reference, measure absolute recovery and avoid calling every correctly directed move a successful reversion.
Understand positive AR(1) decay, observation clocks, coefficient sensitivity and why half-life is not a trade deadline.
Design a bounded stationarity experiment without turning a unit-root test, fitted spread or chosen chart into proof of trading profit.
The path lens inspects a declared number of closes. It measures no actual minutes and does not decide stop exits. A native strategy can have venue sessions, incomplete intervals and a separately owned maximum-hold clock. Preserve those identities before comparing a diagnostic endpoint with a real exit. The existing positive AR expectation half-life is also not an exit deadline. No daily stop management is added to APEX or any actual model through this article.
If the complete inspected path has no hit, the endpoint is not observed within its horizon. If prices are missing, the lens rejects incomplete input rather than inventing a calendar label. A genuine study must keep source availability and label completion separate from order and fill timing. Advancing a completed-session hold count cannot authorize daily profit floors or retrospective reference changes. Native minute, whole-position and maximum-hold rules remain the actual research policy, not assumptions inferred from a teaching scalar path.
The initial point can already satisfy a band or reference condition. That describes the declared observation, not an order execution at zero elapsed time. Native fills require their own permitted clocks and evidence; the worksheet cannot supply them.