Strategies · Record

a5-11-failed-mean-reversion-continuation

a5 / a5-11-failed-mean-reversion-continuation @ a5-v1 · code 67d8d64

campaign strategyREJECTED

Validation Gate: Validation expectancy 0.0074R is less than 30% of Discovery's 0.0715R -- severe degradation.

Lifecycle
completed
Outcome
fail
Phase
Phase A
Research gate
VALIDATION
Engine
strategy lab
Market
MNQ
Timeframe
MNQ 1-minute OHLC
Side
BOTH

Tracking

Family, lineage and what this record needs next
Mechanism family
Mean reversion
a5-11-failed-mean-reversion-continuation
Mechanism parent
a03-mean-distance-reversion
Phase detail
Scientific validation lane (Phase A).
Next required test
None — this record is complete.

Integrity & certification

Signals as recorded; NOT_PASSED means not passed or not applicable
Engine certified
unknown
Execution certified
unknown
Integrity audit
unknown
Shadow verifier
unknown
Technical invalidation
None recorded.

Artifacts

Everything committed for this record
Campaign record (A5)
research-output/strategy-lab/campaigns/a5/campaign_state.json

Stage ladder

Each cell carries the campaign runner's own classification label — hover for the verdict and its reason
Intake
passed
Pre-reg
passed
Lightning
passed
Robust
passed
Exec/stat
passed
Validation
failed
H-eligible
—
Holdout
—
Audit
—
Closed
failed
INTAKE
Registered into campaign A5 at code version 67d8d64.
PREREGISTRATION
Pre-test passed: distinct, causal, executable, required data available.
LIGHTNING
SURVIVED_LIGHTNING_SCREENAggregate expectancy 0.0581R (PF=1.090) over 6900 resolved trades, 4/4 qualifying years (n>=15) agree with the aggregate sign, top-10 winner concentration 5.0% (below the 50% concentration flag). Survives the Lightning Screen -- NOT validated.
ROBUSTNESS
ROBUSTNESS_SURVIVEDNo single-year dependence, no severe winner concentration (no preregistered neighborhood declared).
EXECUTION STAT
EXECUTION_STAT_SURVIVEDBootstrap CI excludes zero and the edge survives up to 0.5pt adverse slippage.
VALIDATION
VALIDATION_FAIL_SEVERE_DEGRADATIONValidation expectancy 0.0074R is less than 30% of Discovery's 0.0715R -- severe degradation.
CLOSED
REJECTEDValidation Gate: Validation expectancy 0.0074R is less than 30% of Discovery's 0.0715R -- severe degradation.

Evidence recorded

Quoted from the campaign record — never recomputed here
Screen n (FIXED_2R)
6,900
Expectancy
+0.058R
Profit factor
1.09
Win rate
35.3%
Total R
401
Max DD
66R
Longest L streak
22
Trades / year
1926.5
SplitWindownExpectancyPFMax DD
Discovery2022-06-03..2025-03-315,450+0.072R1.1166R
Validation2025-04-01..2025-12-311,450+0.007R1.0144R
Holdout—————
Discovery → Validation
Expectancy degradation +0.064R (positive means Validation came in weaker than Discovery).
Bootstrap CI
95% CI +0.019R .. +0.097R · P(expectancy > 0) = 100%
Winner concentration
Top 5 2.5% · top 10 5.0% · top 20 10.0% of total R.
Leave-one-year-out
No sign flip when any single year is removed.

Read from research-output/strategy-lab/campaigns/a5/campaign_state.json

Holdout position

One-shot, sealed, never retried
Eligible
—
Authorised
Not authorised.
Specification frozen
—
Sealed data consumed
No.
Validation data consumed
Yes.

Lineage

Declared at intake, carrying any prior negative result
Relation to prior work
experimental descendant
Parent
a03-mean-distance-reversion
Note
A0.3 tested the reversion ARM (extreme displacement from a causal intraday mean predicts normalization) and was rejected. A5.11 tests the FAILURE of that same reversion, an arm never previously run. Declared descendant; carries A0.3's negative prior.

Frozen specification

Snapshotted at intake — not re-read from a live module
Mechanism hypothesis
When an extended market attempts to revert and cannot reclaim its own mean, the original displacement is confirmed rather than corrected.
Long rule
z = (close - SMA50) / sd(close - SMA50, 50); z >= 2.5; price then retraces toward the mean but fails to close BELOW it within 15 bars; then closes back above the original extreme: enter LONG next bar open.
Short rule
Mirrored for z <= -2.5.
Entry trigger
Three-stage on closed bars. A close through the mean cancels the setup outright -- reversion succeeded, so the hypothesis does not apply.
Earliest executable entry
The OPEN of the next 1-minute bar after the trigger bar closes. Enforced structurally by A5StrategyBase's pending-entry queue, which is drained at the top of the following ingest() -- a rule cannot fill on its own trigger bar.
Initial stop
The reversion attempt's extreme (the closest approach to the mean).
Exit
FIXED_2R via the shared TradeTracker, max-hold backstop.
Session restriction
RTH only — The z-score population and the SMA both behave differently across the session break; restricting to RTH keeps the statistic homogeneous.
Causality
The mean and its deviation spread end at the trigger bar; each later stage is confirmed on a closed bar. Every statistic is computed from closed 1-minute bars up to and including the trigger bar, against trailing (already-observed) history. Entry is deferred to the next bar's open. No same-bar entry.
Expected frequency
LOW
Required data
MNQ 1-minute OHLC
Campaign data window
MNQ 2022-06-03..2026-09-15
Parameters
maLen = 50 (integer) — Mean length and deviation window.zThreshold = 2.5 (number) — Extension required to arm the setup.reclaimWindow = 15 (integer) — Bars allowed to reclaim the mean.

Read from research-output/strategy-lab/campaigns/a5/campaign_state.json

History

Every dated event this record appears in, newest first
2026-09-17
a5-11-failed-mean-reversion-continuation@a5-v1 registeredREJECTED · VALIDATION
[a5] Validation Gate: Validation expectancy 0.0074R is less than 30% of Discovery's 0.0715R -- severe degradation.
research-output/strategy-lab/experiment_registry.json