Objective
Independently replicate the distinct inventory row CHARTS > Dashboard | RecessionAlert from the checked-in data/recessionalert.sqlite snapshot without live-site requests, credentials, private account data, or canonical crawl-database mutation. Because the source omits component observations and a complete executable state, the campaign tests labeled public proxies rather than claiming source parity. A prior insufficient_evidence stop is superseded by enforced best-effort exploration.
Source and provenance
The selected page https://recessionalert.com/dashboard/ is bound to inventory prefix sha256:1ca9a096a and full SHA256 sha256:1ca9a096a70ff56806049e8e678036118b9a0fea8ad4fd27f2344a1da323e7b0, fetched at 2026-08-22T23:38:26Z, HTTP 200, 113,275 bytes. Five same-URL crawl-log rows exist; zero target-page asset rows. Linked PANIC.xlsx and MEGASIGS.xlsx bytes are not captured.
Public proxies were loaded from data/market.sqlite in read-only mode: SPY, IEF, BIL, IWM, ^VIX, BAA10Y, WALCL. The proxy panel is month-end aligned without pre-inception backfill (SPY calendar 1993-01 → 2026-07).
Disclosed rule, native span, and publication lag
Documented fragments include SPX SIGS (-12..+12 step 2), MTDIFF 0..4 with +2/+1 long rules, Net Top-3 Probability (-100..+100), FEAR ≥2-of-six bottoms with a 55 floor under PANIC, and PANIC warning / MEGA entry guidance. Native observation span and underlying publication lag are unknown. Campaign causal policy uses last-available inputs and lag_bars >= 1; same-close is control-only.
Approach and variants
Primitives live in research/research_campaign/recessionalert_1374.py. Families: sigs_proxy, mtdiff_proxy, fear_proxy, panic_mega_proxy, dashboard_composite. Trial ledger: research/findings/campaigns/1374_trial_ledger.json (8 variants). Negative/sparse results (fear, panic-mega) still count.
Results (exploratory)
| Variant | Kind | Sharpe | CAGR | MaxDD |
|---|---|---|---|---|
dash-mtdiff-lag1-lb12-c2 | inferred_best_guess | 1.016 | 0.109 | -0.195 |
dash-mtdiff-lag1-lb12-c1 | inferred_documented_sedate | 0.730 | 0.101 | -0.508 |
dash-mtdiff-lag2-lb6-c3 | inferred_nearby_variant | 0.701 | 0.063 | -0.233 |
dash-sigs-lag1-lb12 | inferred_family_variant | 1.023 | 0.092 | -0.247 |
dash-fear-lag1-e2 | inferred_family_variant | 0.325 | 0.033 | -0.426 |
dash-panic-mega-lag1-dd20 | inferred_family_variant | 0.467 | 0.029 | -0.232 |
dash-composite-lag1-v2 | speculative_extension | 1.021 | 0.094 | -0.187 |
dash-mtdiff-lag0-lb12-c2 | control_only | 1.744 | 0.181 | -0.106 |
Best causal Sharpe in-grid: dash-sigs-lag1-lb12. Lag-0 control is higher and is rejected as primary.
Files and validation
research/findings/1374.md— durable campaign findingresearch/findings/specs/recessionalert-dashboard.json— machine-readable spec (updated disposition)research/research_campaign/recessionalert_1374.py— proxy primitivesresearch/findings/campaigns/1374_trial_ledger.json— trial metricsresearch/experiments/1374.ipynb— executed offline notebooktests/test_recessionalert_1374.py/tests/test_recessionalert_dashboard.py
Validation: focused pytest 16 passed; read-only DBs; no live RA fetch.
Limitations
Public proxies are not official RecessionAlert product series. Exact FEAR thresholds, MEGA factors, and release clocks remain undisclosed. Do not treat exploratory equity paths as a registered Other-indicators signal. Other / Indicators remains a future destination only after native redistributable observations and a complete execution contract exist.