Loading page content
Documentation
High-level overview of how Grey Rabbit Markets frames market intelligence. Proprietary formulas, exact weights under active review, and licensed raw series are intentionally not disclosed here. For source categories, see Data Sources.
Detailed explanations for scores and labels shown on Gold and other intelligence pages.
Purpose. Describe the prevailing precious-metals market regime and how evidence clusters across pillars, without prescribing trades.
Inputs. Multi-pillar intelligence snapshots, confidence and freshness metadata, and historical regime context where available.
How scores are generated. Cycle labels and supporting explanations are produced deterministically from stored inputs. Explanations cite contributing evidence categories; exact internal weighting remains proprietary.
Purpose. Assess physical-market conditions that complement futures and ETF signals — inventory, flows, and related stress or strength cues.
Inputs. Licensed and public physical-market datasets (subject to completeness and licensing gates), plus derived physical completeness measures.
How scores are generated. Physical modules contribute only when completeness thresholds are met. Restricted raw feeds are never shown publicly when licensing prohibits display; derived contribution signals may still inform scores.
Purpose. Organize institutional outlooks into comparable forecast views and consensus context for education and research.
Inputs. Published institutional forecasts, dated outlook materials, and structured profile metadata maintained by Grey Rabbit.
How scores are generated. Consensus and track-oriented views summarize available forecasts. They are not predictions by Grey Rabbit and do not constitute investment advice.
Purpose. Describe how participants appear to be thinking, behaving, and crowding — separately from price trend or fundamentals.
Inputs. Scheduled social and search signals where licensed, fear-and-greed style gauges, forecast-consensus context, and narrative crowding diagnostics.
How scores are generated. Psychology modules produce directional sentiment and crowding risk as separate concepts. Unavailable modules are excluded rather than forced to neutral. During beta, psychology may be displayed independently of the overall score.
Purpose. Capture macroeconomic conditions relevant to precious metals and risk assets.
Inputs. Public macroeconomic series such as Federal Reserve and related official datasets, with freshness and revision awareness.
How scores are generated. Macro rules map observed series into directional and confidence contributions. Exact rule thresholds are versioned and not fully published.
Purpose. Measure trend, momentum, and structural price behavior across relevant timeframes.
Inputs. Licensed or permitted market price and OHLC series used for charting and technical modules.
How scores are generated. Technical intelligence aggregates approved sub-modules (for example trend and momentum). Supplementary structure measures may be displayed without scoring until approved.
Purpose. Communicate how much confidence users should place in a given intelligence reading.
Inputs. Module availability, data completeness, source freshness (live / delayed / stale), and evidence quality diagnostics.
How scores are generated. Trust and confidence scores rise when contributing modules are fresh and complete, and fall when data is stale, sparse, or gated. Trust is not a price forecast.
Purpose. Preserve auditable history of scores and regimes so past readings are never silently rewritten after methodology changes.
Inputs. Versioned snapshots including methodology identifiers, timestamps, scores, and confidence metadata.
How scores are generated. Historical records remain frozen to the methodology version that produced them. New methodology versions create new forward records rather than recomputing the past.
Pipeline: Technical Intelligence → Fundamental Intelligence → Psychology Intelligence → Confidence Engine → Overall Intelligence.
Overall score uses the legacy five-module composite (GREY-RABBIT-SCORE-0.1) during Psychology Beta. Psychology is computed and displayed independently where enabled.
Pillar version: INTELLIGENCE-PILLARS-1.0
Pillar version: INTELLIGENCE-PILLARS-1.0
Psychology is experimental during Beta and may be excluded from the production Overall Grey Rabbit Intelligence Score while still displayed for research.
Production social: SOCIAL-SENTIMENT-0.1 via scheduled collection where configured — public pages read stored snapshots.
Minimum scoring completeness thresholds apply. Score states range from Extreme Fear through Euphoria, plus Unavailable.
Pillar confidence derives from module availability, data completeness, and freshness (live / delayed / stale). Overall confidence is a weighted average of contributing pillar confidences.
Market prices, macro series, positioning reports, and physical datasets are used under public-domain, licensed, or approved-import terms. See Data Sources for official versus derived categories, and Financial Disclaimer for advice limitations.