1. Decision Overview
Business question: After filters and gates, what review labels dominate this month, and should a reviewer act, wait, or monitor?
Filter pass rate (Dec 2025)
61.7%
Top label (Dec)
Hold
Year pass rate
61.9%
Cue families
A–E
Decision labels · December 2025
Filter gate outcomes · December 2025
Cue family activity · December 2025
Insight: December is Hold-heavy with active Warning Long/Short counts.
Only about 62% of filtered reviews pass to a firm action label — the stack is
designed to wait when families conflict, not to force trades.
Recommended monitoring: Track Wait Buy / Wait Sell versus Buy / Sell each month,
and investigate any rise in filter fails before discussing allocation changes.
2. Portfolio & Risk
Supporting question: Is performance broad-based, or does one group dominate risk while decision labels stay cautious?
Indexed return (YTD)
-12.5%
Benchmark (YTD)
-2.6%
Dec net return
-3.2%
Dec volatility
12.8%
Dec max drawdown
-3.2%
Insight: Growth held about 33.6% of the synthetic book
but contributed 59.9% of risk and
-1.4% of December return contribution.
Combined with Hold / Warning labels, the practical next step is concentration review — not automatic buying.
3. Validation & Pipeline
Trust question: Did quality checks pass, and does a prototype model earn a seat at the table?
Quality checks passed
12/12
Baseline balanced acc.
0.500
Logistic balanced acc.
0.455
Model status
Rejected
Public decision-review stages
| # | Stage | Business question | Output |
|---|---|---|---|
| 1 | Frame the review question | What should a reviewer decide this period? | decision label or monitoring item |
| 2 | Validate data readiness | Is the period complete and trustworthy? | quality pass or stop |
| 3 | Score cue families A–E | Which public cue families are active? | family activity map |
| 4 | Apply review filters | Do conflicting families force a hold? | filter_gate_result |
| 5 | Assign decision label | Buy, Sell, Wait, Hold, or Warning — at what intensity? | decision_label + intensity_band |
| 6 | Aggregate for the period | How often did each label fire this month? | monthly decision counts |
| 7 | Recommend and monitor | What should the next review check? | monitoring checklist |
Full-year decision mix (synthetic)
Limitation: Counts are synthetic demonstration aggregates shaped like the owner’s
review language. They are not live trading history and do not disclose private formulas.