Research / Human review
Review the evidence. Choose the expressions worth confirming.
Positive, negative and uncertain findings belong here. A selection opens a GitHub request for the exact expression; official admission follows confirmation.
Iteration 1 · unassessed
Documented >22 impulse is sparse (~0.8/yr) and knife-edge versus 15.
Iteration 1 · unassessed
Exact Berg rules are untestable; public proxies failed subperiod and overlay gates.
Iteration 1 · unassessed
Investigate-only in prior work; keep as Research Library, not official.
Iteration 1 · unassessed
GARCH-family volatility is useful as a risk state, not a standalone precise bottom oracle.
Iteration 3 · unassessed
Primary local neighborhood stitched OOS CAGR 12.23% with max drawdown -25.45%; fails the preregistered stitched-path -25% floor.
Iteration 1 · unassessed
Imported 53 reconstructed signals in 8 classes from the miltonberg workspace into Other → Indicators. Each card is an independent overlay, not an AllocateSmartly replica and not a claim that we replicate proprietary MB Edge internals.
Iteration 1 · unassessed
Likely redundant with trend; keep as playable negative/weak result.
Iteration 1 · unassessed
Six inferred public-proxy variants were executed on the common 2002-07 through 2026-07 monthly panel; the lag-1 12-month liquidity-level proxy is the declared descriptive baseline.
Iteration 1 · unassessed
The SHA-verified source preserves a documented 3-of-7 CMHI Diffusion-zero cash advisory, but not a playable native series or BUY/re-entry rule.
Iteration 1 · unassessed
Five public proxy variants are reproducible over the longest frozen BTC history, but no candidate establishes native RecessionAlert fidelity.
Iteration 1 · unassessed
Seven inferred/proxy variants execute over the longest checked-in SPY/IEF overlap, while the original source-faithful finding remains insufficient_evidence.
Iteration 1 · unassessed
The 2014 DeMark SP-500 article discloses modified TD-Setup Class-A/B/C buy and sell rules with RA buy>=8/sell>=9 thresholds and a 1-5 bar reaction hint, but no native setup/summation series or portfolio map.
Iteration 1 · unassessed
The archived article documents a daily CMHI bounded from -1 to +1, qualitative regression de-trending, and +0.3/zero pivot concepts without a complete BUY/SELL rule or native observations.
Iteration 1 · unassessed
The source describes a Google recession-search fear gauge but provides no machine-readable observations, exact query, native span, or executable market rule.
Iteration 1 · unassessed
The Improved SP-500 Selling Pressure page adds a seventh DeMark BUY-setup component, retains >=5 component buy hints when diffusion drops below 4, and documents manual breadth +1 via Great Trough/Zweig.
Iteration 1 · unassessed
The checked-in article discloses two international breadth ideas, source-reported 7- and 10-month leads, a 0.57 R-squared claim, and qualitative dip-buying context, but no source-linked observations or complete state rule.
Iteration 1 · unassessed
SHA-verified dashboard discloses multi-pane gauge semantics without observation tables.
Iteration 1 · unassessed
Parallel five-sleeve proxy reduces some drawdown vs always-long SPY in causal trials but is not RA parity.
Iteration 1 · unassessed
useful_as_filter
Lag-1 NEWHI-style INDEX:HIGN z-score proxy shows competitive causal risk-adjusted returns among tested families.
Iteration 1 · unassessed
Pinned offline snapshot sha256:d949a5a1… discloses McClellan VMCOS/STL/MTL formulas and an STL zero-cross buy/sell narrative.
Iteration 1 · unassessed
Measuring Selling Pressure (2013-09-10) discloses a six-component SP-500 selling-pressure diffusion, a numeric 4% peak-to-trough correction trigger, BUY when diffusion rises above 4 then falls below 4, and sell warnings on 0->1 and 1->2.
Iteration 1 · unassessed
Source is a multi-tab monthly chart family with cadence only; native series and thresholds are absent.
Iteration 1 · unassessed
Documented-intent syndrome vote (lag1) Sharpe 0.743, CAGR 0.091, DD -0.428, 14 events, risk-on 0.88 on 1992-01..2026-06; trails SPY buy-hold Sharpe 0.769/CAGR 0.108 but cuts DD from -0.508 to -0.428.
Iteration 1 · unassessed
The SHA-verified article discloses monthly component context, four qualitative probability interpretations, a highest-two preference, and Long-/Short-leading views, but not an executable signal.
Iteration 1 · unassessed
Twenty-five labeled variants executed with lag>=1 causal_primary and lag-0 control_only.
Iteration 1 · unassessed
Twenty-nine labeled variants executed with lag>=1 causal_primary and lag-0/look-ahead control_only.
Iteration 1 · unassessed
The source is a contextual multi-tab family, not one executable indicator; disclosed formulas and thresholds are preserved separately from inferred proxies.
Iteration 1 · unassessed
Pinned /pro-optimum/ sha256:340dfcfa15d9 has zero on-row assets and no executable formulas.
Iteration 1 · unassessed
RFD is documented as a fifteen-model 0–15 recession-warning count with a strict trigger above five, but the checked-in snapshot has no observation-level native series.
Iteration 1 · unassessed
MODELS SAHM Redux article discloses multiple Sahm constructions but no page-local assets or portfolio map.
Iteration 1 · unassessed
The SHA-verified article discloses four monthly cycle families, three ranked characteristics, a composite concept, source-like short/long/2x bands, and optional cash/no-leverage language, but no source-linked signal series, scale conversion, publication clock, or complete execution contract.
Iteration 1 · unassessed
29 labeled variants executed with lag>=1 causal_primary and lag-0/look-ahead control_only.
Iteration 1 · unassessed
The source is a contextual multi-tab family, not one executable indicator; disclosed formulas and thresholds are preserved separately from inferred proxies.
Iteration 1 · unassessed
The source-backed result is an incomplete support/resistance trend-duration and historical-reversal-probability concept; exact native observations, formula, STOP calculation, threshold, release lag, action mapping, and portfolio semantics are absent.
Iteration 1 · unassessed
Twenty-nine labeled variants executed with lag>=1 causal_primary (26) and lag-0/same-close control_only (3).
Iteration 1 · unassessed
Executed 28 labeled variants across seven exploratory families.
Iteration 1 · unassessed
The source-backed result is a documented weekly macro growth-index concept, not a playable native indicator: exact observations, formula, threshold, action mapping, native span, and recurring point-in-time release are absent.
Iteration 1 · unassessed
The source-backed result is a warning concept only: native WLEI/WLEI Diffusion observations, exact historical rank construction, release lag, threshold, state machine, and buy/sell action are not disclosed in the selected snapshot.
Iteration 1 · unassessed
HMA dependence is disclosed; SMA/EMA collapse. Sparse rescue-only overlay.
Iteration 1 · unassessed
Prior sleeve was investigate-only; keep playable in Research Library.
No research matches these filters.
These completed records use an older format. Review the report before preparing an expression for confirmation.