Selected from three bounded challengers, its Brier is 0.241980, improvement is +0.004739, and AUC is 0.503030. G7 and G8 pass, but the 95% Brier lower bound remains −0.002580.
MODEL 2 · STEP 6Summarize
Summarize
the model.
Record the challenger’s progress without relabeling reused development dates as fresh evidence: keep the prior controller and require a new paper-forward test.
MARKET CONTEXT · OUTSIDE MODEL FREEZEHOOD
These readings update independently. They are not model inputs, deployment evidence, or a trading recommendation.
- Latest price
- $122.11 · fallback
- 20D momentum
- +30.9% Close-to-close · 20 completed sessions
- RSI 14
- 65.9 neutral · Wilder method
- Relative volume
- 1.2× Versus the prior 20 completed sessions
TECHNICALS THROUGH · TIMESTAMPED FALLBACK
Keep the improvement. Do not promote the signal.
Signed Core Ridge records a lower retrospective Brier score and passes calibration-error and fold-consistency gates. Seven gates still fail, and the same 685 dates are development-only. The fold-local prior remains the active controller.
The fold-local prior stays active at 55.745%. The challenger’s latest 60.40% output is experimental and does not tilt the simulation or become a live signal.
COMPLETE RESEARCH CHAIN
Auditable does not mean deployable.
Three build contracts pass. Phase A is preserved, Phase B records measurable progress, and the incomplete gate result still controls the recommendation.
- 01Target contractPASS
Raw HOOD close, five-session horizon.
- 02Market dataPASS
1,255 aligned SIP sessions and explicit clocks.
- 03Feature contractPASS
60 trailing-only features on a common window.
- 04Phase A replayCOMPLETE
Five candidates; Cross Asset Ridge clears 0 of 9 gates.
- 05Phase B challengerPARTIAL
Signed Core Ridge clears G7 and G8: 2 of 9.
- 06Evidence boundaryFROZEN
All 685 dates remain development-only.
- 07RecommendationGUARDED
Fold-local prior controller; no deployment.
PHASE A · ORIGINAL DEVELOPMENT SEARCH
Cross Asset Ridge ranks first, but does not validate.
Selection means lowest observed Brier score among the five inspected candidates. It does not override uncertainty, discrimination, or calibration evidence.
95% block interval −0.002606 to +0.005568.
Below the 0.50 no-skill reference.
Equals the always-up OOS class rate; fold-prior accuracy is 58.102%.
Slope; intercept 0.822 and five-bin ECE 0.111.
Benchmark-only recommendation.
Active prior controller; development candidate is flat at 57.508%.
PHASE B · DEVELOPMENT CHALLENGER
Signed Core Ridge lowers retrospective Brier, not the bar.
The challenger reuses the same 685 dates already examined in Phase A. Its results are development evidence only—not a fresh holdout or deployment test.
Improvement +0.004739; 95% lower bound −0.002580.
Block-bootstrap lower bound 0.4347.
Slope; intercept 0.312902 and five-bin ECE 0.042856.
Gate G8 passes with 72.7% of folds positive.
G7 calibration error and G8 fold consistency pass.
Experimental challenger; active prior controller remains 55.745%.
MODEL RECOMMENDATIONS
Protect the next honest test.
The frozen 685 OOS dates are now development evidence. They cannot become a fresh holdout after Phase B tuning.
- P0
Keep the HOOD directional model switched off and use the fold-local prior only.
- P1
Freeze Signed Core Ridge, its target, folds, nine gates, and the current benchmark before another outcome arrives.
- P2
Treat its 2-of-9 result as Phase B development evidence, never independent validation.
- P3
Score the frozen challenger and prior controller side by side on newly arriving paper-forward outcomes.
- P4
Require the new paper-forward period to clear all nine gates before reconsidering deployment.
Phase B passes G7 and G8, but 2 of 9 gates is insufficient for deployment. Model 2 remains benchmark-only and no-go.
Alpaca consolidated stock bars · Alpaca US crypto bars · Alpaca corporate-action announcements · Robinhood SEC filings · Robinhood investor relations. The Excel file preserves the source ledger and frozen outputs.
NEXTStep 7 · Simulate the range
Continue through the frozen Model 2 workflow.