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.
Topic notes and detailed research exercises.
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.
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.
Reconcile earlier reference windows, original calls, unresolved positions and costs before accepting a mean-reversion performance claim.
Define the strategy, reproduce conditional expectancy and distinguish statistical recovery from executable account economics.
Compare frozen contraction and continuation rules on the same opportunity population without relabelling a chart after the outcome.
Define what returning toward a mean actually measures before treating a directional move as recovery or a statistical pattern as a trade.
Define the unchanged reference, eligible deviation and later distance before calling a price move recovery. Statistical behavior and trade economics remain separate.
The new guides begin with frozen versus moving references, then distinguish crossing, distance, band and first passage. Another path follows observation clocks, clusters and scale collapse. Process identity, structural changes, threshold selection and pairs complete the source and evaluation questions. The original z-score, half-life, stationarity, record-verification and profitability guides remain available. Each addition owns a different contract and an authored counterexample. This is not a set of near-identical vendor rankings or a hidden list of proprietary model outcomes.
Seven paths retain all declared initial and future observations, with markers for later unused rows. Three cards cover price, spread and revised-reference contracts without actual source or master. The new lens calculates bounded metrics locally. The older lab, worksheet and ten-field workbench remain separate working tools. A reader can inspect their different object scopes rather than assume all numbers have one financial denominator. Downloading complete fictional examples does not authorize a live strategy, authenticate a source or establish independent model evaluation.
If a destination changes, read reference identity. If a directed move overshoots, read endpoint ownership. If missing bars disappear, read observation time. If many events describe one excursion, read clustering. If a large standardized score comes from tiny scale, read scale collapse. If an attractive result comes from retuning an inspected sample, read threshold selection. If an equilibrium combines two legs, read spread identity. These routes let readers isolate a specific missing artifact without turning every evidence problem into one generic claim about the best trading provider.