Oracle Decision Report Guide
SagaHalla Oracle Decision Report Guide
This page explains how to interpret a SagaHalla Oracle Decision Report — a transparent, auditable record of how the Oracle evaluated an asset, resolved uncertainty, and produced a trading stance.
The Oracle Decision Report is not a prediction. It is a decision artifact: an inspectable record of context, constraints, and reasoning that can be reviewed, explained, and audited over time.
1. Signal Snapshot
At the top of every report you will see a Signal Summary:
- Primary Decision: BUY / SELL / HOLD
- Conviction: normalized decision strength (0–1) used to scale trade sizing
- Confidence: overall support for the stance after regime checks, integrity filters, and simulation context (0–1, often displayed as a percentage)
- Exposure: proportion of total account equity allocated to the asset, expressed as a percentage (assumes single-asset account)
This snapshot answers one question:
“Given the Oracle’s current reasoning and constraints, what is the trading stance right now — and how strongly is it supported?”
Conviction governs how much risk is taken; confidence reflects how well supported the stance is.
2. Commentary
The Commentary is a human-readable synthesis of the decision process.
It integrates:
- Recent price behavior
- Trade history context
- Monte Carlo simulation bias
- Integrity diagnostics
- Regime classification
Think of it as the verbal compression of the full Oracle trace.
If you only read one section, read this.
3. Position History
This section grounds the current stance in relation to prior Oracle decisions.
You will see:
- Historical BUY / SELL executions
- Execution prices
- Equity percentage over time
- Position size evolution (derived from conviction)
This ensures:
- No retroactive decision rewriting
- Clear linkage between signals, conviction, and executed actions
The Oracle never hides prior actions.
4. Price Context & Trade Markers
The price chart overlays:
- Candlestick price action
- BUY / SELL markers
- Conviction-weighted intensity
Use this section to visually confirm:
- Entries during favorable structural dynamics
- Exits during rising uncertainty or breakdown conditions
5. Monte Carlo Bias & Sigma
This section exposes the probabilistic context surrounding the decision.
Monte Carlo Bias
- Directional skew of simulated future price paths
- Positive bias supports BUY; negative bias supports SELL
- Interpretable form: Δp = p(buy) − p(sell)
Sigma Percentile (Dispersion)
- Measures dispersion (uncertainty spread) of simulated paths relative to history
- Higher dispersion implies broader outcome uncertainty and typically constrains conviction
- Lower dispersion implies tighter clustering and may support higher conviction (subject to regime and integrity constraints)
Important:
The Oracle does not trade on prediction accuracy — it sizes and filters decisions based on structural uncertainty and admissibility.
6. Integrity
The Oracle evaluates three orthogonal integrity dimensions.
Material (Structural) Integrity
Structural coherence of price action.
Energetic Integrity
Usable price energy given the current structural order. Higher energetic integrity implies the market is active, healthy, and capable of moving in a way that encourages participation
Ethereal (Temporal) Integrity
Memory and trend persistence relative to energetic processes in relation to price history.
Together they answer:
“Is the price structurally sound, energetically alive, and informationally coherent?”
7. Regime Identification
Based on integrity combinations and contextual diagnostics, the Oracle classifies the market environment:
- CALM_TREND
- NEUTRAL
- CHOP
- AGGRESSIVE
- BREAKDOWN
Qualitative factors and combinations determine:
- Allowed trade types
- Conviction limits
- Re-entry behavior
8. Decision Audit
This section shows the mechanical reasoning behind the trading stance.
Typical elements include:
- Ensemble Consensus: net BUY vs SELL vote
- Conflict Resolution: how competing BUY and SELL signals are reconciled
- Statistical Confirmers: z-score, Sharpe, Hurst, ADX, autocorrelation
- Filters & Constraints: integrity thresholds, volatility-of-volatility limits, admissibility checks, and internal governors
Example interpretation:
BUY was selected not because SELL was absent, but because BUY had higher admissible strength under the current regime and constraints.
Appendix: Oracle Trace Parameter Definitions
This appendix provides concise definitions for parameters referenced above to support understanding and auditability. Additional terms will be added as Oracle outputs evolve.
Signal
The Oracle’s discrete action output at the decision timestamp (BUY, SELL, HOLD).
Conviction
A normalized scalar (0–1) representing the strength of the decision and used to scale trade sizing.
Confidence
A normalized measure (0–1) reflecting overall support for the stance after regime checks, integrity filters, and simulation context (often displayed as a percentage).
Ensemble Consensus
The dominant signal direction across multiple internal evaluators.
Sharpe Ratio
Risk-adjusted return per unit of volatility over a rolling window.
Z-Score
Standardized deviation of price or return from its rolling mean.
Hurst Exponent
Measure of long-term memory in the price process.
Autocorrelation
Correlation of current returns with past returns over a defined lag.
Sizing Multiplier
Scalar applied to base position size as a function of conviction and regime stability.
Risk Governor State
Internal constraints limiting sizing based on drawdown, stress, or interpretability limits.
Validation Status
Confirmation that admissibility requirements are satisfied.
Decision Timestamp
Exact time at which the Oracle evaluated the market.
Execution Marker
Visual indicator of trade execution on the chart.
Exposure
Proportion of total account equity currently allocated to the asset, expressed as a percentage (single-asset account context).
Monte Carlo Bias
Directional skew of simulated future price paths.
Sigma (Dispersion)
Statistical spread of Monte Carlo outcomes.
Sigma Percentile
Relative rank of dispersion versus historical levels.
Material (Structural) Integrity
Measure of structural coherence of price action.
Energetic Integrity
Measure of usable market participation energy.
Volatility Percentile
Current volatility relative to historical distribution.
Volatility of Volatility
Stability of volatility itself.
Ethereal (Temporal) Integrity
Measure of price informational order and persistence.
Calmness
Penalty for extreme excitation that reduces interpretability.