Show helpHide help
What this surface is for
- A stacked, time-aligned history of the regime state. Each dimension is its own horizontal lane sharing one time axis, so you read them as separate strips lined up by true date.
- The structural regime lane is never overwritten by tactical or stress overlays - lane separation is the contract.
Lanes: structural, stance, tactical, stress
- Structural regime and stance / posture are categorical bands. Tactical state is its own band lane.
- Stress is a marker lane: an amber tick marks a governed stress flag. Absence of a marker is never asserted as "calm".
- A per-date value the ledger does not carry is shown as an explicit unknown band, never inferred as stable.
Abstention meta-state lane
- The abstention lane shows the governed meta-state (NORMAL / ABSTAIN) when the backend sources it; ABSTAIN is highlighted distinctly so it stands out from passive bands.
- By default the per-date meta-state is not joined into this read-model, so the lane shows as unavailable with reason SEPARATE_ABSTENTION_SURFACE - this is the abstention STATE lane.
- WHY an abstention occurred (its reasons) lives on the Regime Accountability surface, not here.
Transition & disagreement markers
- Regime transition is a marker lane: a tick marks a date the pipeline classified as TRANSITIONAL or INDETERMINATE (governed regime_quality_daily). DETERMINATE places no tick; a thicker tick marks HIGH severity. It flags label instability - it never reclassifies the structural regime.
- Forecast disagreement is a marker lane: a tick marks an empirical-vs-model forecast divergence (governed sentinel_disagreement_daily). "agreement" places no tick - absence is never asserted as agreement.
- Each marker lane shows its tick meaning in text and a count of marker dates in the window; these are uncertainty / transition cues, not buy/sell signals and not a realized regime flip.
- regime_confidence_fliprisk_daily is per layer (structural / tactical), not per date. Layer rule (2026-06-26, tactical re-included 2026-06-29): the STRUCTURAL layer is surfaced as the "structural confidence (layered)" cue below, and the TACTICAL layer as a SEPARATE "tactical confidence (layered)" cue (the abstention-boundary horizon). Both are descriptive cues, never aggregated across layers; the tactical layer stays excluded from the structural regime BANDS (the "Flip-risk: tactical layer (excluded)" lane is unchanged).
Instability cues (descriptive)
- Flip-risk (0-100 proximity), score disagreement, flip-risk acceleration (signed), structural confidence (layered, 0-1) and tactical confidence (layered, 0-1) are shown as self-scaling governed sparklines - descriptive cues, not signals.
- Score disagreement is an abstention / uncertainty cue (how much the macro scores disagree); it is not a trading signal.
- Structural confidence (layered) is the structural layer of regime_confidence_fliprisk_daily (1 = deep inside the structural regime, 0 = near a boundary); it is distinct from the flip-risk cue (different source and scale).
- Tactical confidence (layered) is a short-horizon diagnostic cue (the tactical / abstention-boundary layer of the same source). It does not redefine the structural regime; structural regime and tactical instability are shown separately. It is a descriptive cue, not a structural band, and is never folded into the regime lane.
- No threshold is drawn (the read-model carries none); a cue with no governed source shows as an explicit unavailable row, never a flat line.
- Flip-risk acceleration crossing zero in the chart simply reflects the governed signed value; it is not a forecast.
Solo-trigger entry block, risk-off & allocation lanes
- Solo-trigger entry block (the solo corroboration / initiation-authority gate) is a governed marker lane: a tick marks a date abstention ENTRY was blocked because a single, uncorroborated trigger fired (corroboration = 2+ trigger dimensions). It is sourced from decision_outcomes_ledger_daily (abstention_entry_solo_blocked); the tooltip carries the governed solo-trigger reason (e.g. solo_flip_risk_accel). A governance cue, not a regime reclassification or buy/sell signal.
- Risk-off has no separate governed boolean at this daily grain: the risk-off condition is already carried inside the structural regime label (e.g. "Risk-Off | Stagflation …") and the risk score, so it stays a visible unavailable lane with reason NO_GOVERNED_SOURCE_COLUMN rather than inferring a boolean from the label text.
- Allocation posture is a portfolio / sleeve concept with effective-date ranges, not a daily regime-history row, so it stays a visible unavailable lane with reason SOURCE_GRAIN_MISMATCH; it is never expanded into this daily history without a governed alignment rule.
- Unavailable lanes are kept visible (not hidden) so a missing or excluded source is explicit; their values are never synthesized from unrelated fields.
Unavailable, degraded & indeterminate
- An unavailable lane or surface stays visible with its reason code (e.g. NO_GOVERNED_SOURCE_COLUMN) - it is never hidden to look cleaner.
- degraded means the surface still renders but carries an annotation explaining the limitation.
- indeterminate means the value is genuinely not determined; it is never zero-filled or shown as a stable state.
Freshness & current-day completeness
- as of / latest available are verbatim governed dates - when they differ, you are seeing the latest available data, not today.
- stale means the data is older than the current period; it is shown as a distinct caution state, never as current.
- current day incomplete (intraday) is shown explicitly: it is not a hard failure - it refreshes after the next governed pipeline run.
- Yesterday's data is never relabelled as current; the freshness badge always reflects the governed truth.
Regime transition · as of latest 2026-06-26stable(DETERMINATE · LOW)last change: Reflation (Growth ↑ / Inflation ↑) → Goldilocks (Growth ↑ / Inflation ↓), 15d ago (2026-06-11)Regime transition status as of the latest governed date: whether the pipeline's own classification reads as in-transition (its TRANSITIONAL / INDETERMINATE state) and how long ago the structural regime last changed. A descriptive cue about the classification state - not a regime reclassification and not a buy/sell signal. The 10-day recency window is a display choice; the days-since fact is always shown.
Structural regime
ReflationGoldilocks
Stance / posture
PRO-CYCLICAL / REAL ASSETSRISK-ON / QUALITY GROWTH
Tactical state
unavailable - LEDGER_UNAVAILABLE
Stress flag
unavailable - LEDGER_UNAVAILABLE
Abstention meta-state
unavailable - LEDGER_UNAVAILABLE
Regime transition
11 markers in window
tick = transition (TRANSITIONAL / INDETERMINATE)thicker tick = HIGH severitya cue, not a regime reclassification
Forecast disagreement
no forecast disagreement markers in the available window
tick = empirical-vs-model disagreementan uncertainty cue, not a buy/sell signal
Flip-risk: tactical layer (excluded)
unavailable - SEPARATE_TACTICAL_HORIZON
Solo-trigger entry block
unavailable - NO_GOVERNED_SOURCE_COLUMN
Risk-off
unavailable - NO_GOVERNED_SOURCE_COLUMN
Allocation posture
unavailable - SOURCE_GRAIN_MISMATCH
2026-05-26
Structural regime, stance, tactical state and stress flags over time - each on its OWN time-aligned lane (descriptive history, not a prediction). The structural lane is never overwritten by a tactical/stress overlay, and the tactical flip-risk layer stays a SEPARATE horizon (shown excluded, not merged in). Abstention meta-state and solo-trigger entry blocks are governed lanes when the read-model sources them. Risk-off and allocation are not sourced at this daily grain (risk-off is carried inside the regime label and the risk score; allocation is a portfolio/sleeve concept) and stay explicit unavailable - visible, never inferred. A per-date value the ledger omits is an explicit unknown, never read as stable. Colours are a neutral categorical palette with no good/bad meaning.
Markers (regime transition, forecast disagreement, and the instability cues below) show uncertainty / transition cues; they do not reclassify the structural regime and are not buy/sell signals.
RiskGrowthInflationLiquidity
ReflationGoldilocks
smoothed: unavailable - NO_GOVERNED_SMOOTHED_SERIES
model: unavailable - NO_GOVERNED_MODEL_SERIES
Macro factor scores (risk / growth / inflation / liquidity) over time on one shared scale, with the structural regime drawn as background bands and regime-change days marked. A descriptive MEASUREMENT of past scores, not a prediction. The zero line and the neutral band are fixed references; a smoothed/model overlay is shown only when the read-model carries one (otherwise it is marked unavailable, never fabricated). Colours are a neutral palette with no good/bad meaning.
flip-riskunavailable - LEDGER_UNAVAILABLE
score disagreementunavailable - LEDGER_UNAVAILABLE
flip-risk accelunavailable - LEDGER_UNAVAILABLE
structural confidence (layered)1.00
structural-layer cue (0-1; 1 = deep inside the regime, 0 = near a boundary), not a signal
tactical confidence (layered)unavailable
Self-scaled governed cues over the available window (2026-05-26 to 2026-06-26): flip-risk proximity, score disagreement, and flip-risk acceleration. These are descriptive CUES, not signals - they do not imply a realized regime flip or any buy/sell action, and no threshold is drawn. Each cue resolves independently from the governed regime-history read model as new runs land.
threshold / tension bands: not available in this read-model (indeterminate, not zero).
| date | regime | stance | risk | growth | infl | liq |
|---|---|---|---|---|---|---|
| 2026-06-26 | Goldilocks (Growth ↑ / Inflation ↓) | RISK-ON / QUALITY GROWTH | 0.97 | 0.92 | -1.14 | -0.44 |
| 2026-06-25 | Goldilocks (Growth ↑ / Inflation ↓) | RISK-ON / QUALITY GROWTH | 0.78 | 0.76 | -1.03 | -0.38 |
| 2026-06-24 | Goldilocks (Growth ↑ / Inflation ↓) | RISK-ON / QUALITY GROWTH | 0.87 | 0.76 | -1.63 | -0.28 |
| 2026-06-23 | Goldilocks (Growth ↑ / Inflation ↓) | RISK-ON / QUALITY GROWTH | 0.81 | 0.90 | -0.98 | -0.04 |
| 2026-06-22 | Goldilocks (Growth ↑ / Inflation ↓) | RISK-ON / QUALITY GROWTH | 1.37 | 1.27 | -0.63 | -0.48 |
| 2026-06-18 | Goldilocks (Growth ↑ / Inflation ↓) | RISK-ON / QUALITY GROWTH | 1.17 | 1.00 | -1.02 | -0.31 |
| 2026-06-17 | Goldilocks (Growth ↑ / Inflation ↓) | RISK-ON / QUALITY GROWTH | 0.95 | 0.81 | -0.44 | -0.74 |
| 2026-06-16 | Goldilocks (Growth ↑ / Inflation ↓) | RISK-ON / QUALITY GROWTH | 0.92 | 0.80 | -0.67 | -0.16 |
| 2026-06-15 | Goldilocks (Growth ↑ / Inflation ↓) | RISK-ON / QUALITY GROWTH | 1.17 | 1.04 | -0.30 | -0.04 |
| 2026-06-12 | Goldilocks (Growth ↑ / Inflation ↓) | RISK-ON / QUALITY GROWTH | 0.99 | 1.01 | -0.12 | -0.33 |
| 2026-06-11 | Goldilocks (Growth ↑ / Inflation ↓) | RISK-ON / QUALITY GROWTH | 0.75 | 0.87 | -0.69 | -0.02 |
| 2026-06-10 | Reflation (Growth ↑ / Inflation ↑) | PRO-CYCLICAL / REAL ASSETS | 0.09 | 0.36 | 0.19 | -0.27 |
| 2026-06-09 | Goldilocks (Growth ↑ / Inflation ↓) | RISK-ON / QUALITY GROWTH | 0.35 | 0.58 | -0.03 | -0.21 |
| 2026-06-08 | Reflation (Growth ↑ / Inflation ↑) | PRO-CYCLICAL / REAL ASSETS | 0.43 | 0.67 | 0.35 | -0.14 |
| 2026-06-05 | Reflation (Growth ↑ / Inflation ↑) | PRO-CYCLICAL / REAL ASSETS | 0.44 | 0.63 | 0.67 | -0.66 |
| 2026-06-04 | Reflation (Growth ↑ / Inflation ↑) | PRO-CYCLICAL / REAL ASSETS | 0.87 | 0.91 | 0.34 | -0.28 |
| 2026-06-03 | Reflation (Growth ↑ / Inflation ↑) | PRO-CYCLICAL / REAL ASSETS | 0.72 | 0.63 | 0.87 | -0.51 |
| 2026-06-02 | Reflation (Growth ↑ / Inflation ↑) | PRO-CYCLICAL / REAL ASSETS | 1.01 | 0.92 | 0.65 | -0.34 |
| 2026-06-01 | Reflation (Growth ↑ / Inflation ↑) | PRO-CYCLICAL / REAL ASSETS | 0.84 | 0.65 | 0.91 | -0.56 |
| 2026-05-29 | Reflation (Growth ↑ / Inflation ↑) | PRO-CYCLICAL / REAL ASSETS | 0.89 | 0.82 | 0.76 | -0.19 |
| 2026-05-28 | Reflation (Growth ↑ / Inflation ↑) | PRO-CYCLICAL / REAL ASSETS | 0.77 | 0.69 | 0.77 | 0.07 |
| 2026-05-27 | Reflation (Growth ↑ / Inflation ↑) | PRO-CYCLICAL / REAL ASSETS | 0.58 | 0.63 | 0.77 | 0.13 |
| 2026-05-26 | Reflation (Growth ↑ / Inflation ↑) | PRO-CYCLICAL / REAL ASSETS | 0.71 | 0.77 | 0.64 | 0.49 |
source: regime_history.parquet