Cross-Asset Validation Report

The Cross-Asset Validation Report is PulseGrid's™ definitive performance artifact. It runs the PG-MIM™ v2.1 scoring algorithm across all 27+ tracked instruments simultaneously and produces a comprehensive analysis suitable for academic publication or book inclusion.

v2.1 Improvements

The v2.1 release introduces three major enhancements:

1. Regime-Aware Divergence Resolution: Resolution horizons are now dynamically adjusted based on the current market regime (crisis compresses by 40%, recovery extends by 30%) and instrument tier (commodities get 25-day windows, crypto gets 7-day). Partial moves in PG's predicted direction now count as correct (e.g., market shifting from hold to buy when PG signaled strong_buy).

2. Historical Accuracy Trending: Each instrument now includes a weekly accuracy history showing cumulative running accuracy over time, rendered as sparklines in the comparison matrix. An overall accuracy trend chart shows the algorithm's aggregate performance evolution.

3. Weekly Report Snapshots: Automated weekly archival (Sunday 00:00 UTC) generates and stores report snapshots with admin-only notifications. Admins can browse historical snapshots and compare week-over-week accuracy changes to track algorithm evolution.

Report Sections

#### 1. Executive Summary

A high-level overview including:

• Overall directional accuracy across all instruments
• Total divergences analyzed and their resolution rates
• Badge distribution (Prophetic / Strong / Neutral / Weak / Insufficient Data)
• A narrative verdict contextualizing results within the current market regime

#### 2. Statistical Rigor

Quantitative validation against a 50% random baseline:

• Binomial Test p-Value: One-tailed test of H₀: accuracy = 50% using normal approximation
• 95% Wilson Score Confidence Interval: More accurate than Wald interval for small samples (Wilson, 1927; Agresti & Coull, 1998)
• Cohen's h Effect Size: Standardized measure for comparing two proportions (Small: 0.2, Medium: 0.5, Large: 0.8)
• Information Ratio: (Mean accuracy − baseline) / standard deviation of per-instrument accuracy
• Z-Score: Number of standard deviations the observed accuracy is from the baseline

#### 3. Side-by-Side Comparison Matrix

A sortable, filterable table of all instruments showing:

• Symbol, name, asset class, and tier classification
• Total divergences, correct/incorrect predictions
• Directional accuracy with color-coded values
• Accuracy sparkline: Mini line chart showing cumulative accuracy over time
• Signal stability (percentage of time the signal stayed consistent)
• Optimal prediction horizon per instrument tier
• Recent trend direction (improving, stable, declining)
• Performance badge assignment

#### 4. Asset Class Breakdown

Grouped analysis panels for each asset class (Equities, Indices, Sector ETFs, Commodities, Forex, Crypto) with:

• Aggregate accuracy per class
• Best and worst performers
• Narrative explanation of why certain asset classes respond better to event-driven scoring
• Expandable detail cards for each instrument within the class

#### 5. Divergence Timeline

Chronological listing of the most significant divergence events where PulseGrid's signal diverged from market consensus, annotated with:

• Date, instrument, and severity level
• PG Signal vs. Market Signal at the time
• Outcome (correct, incorrect, or pending)

#### 6. Overall Accuracy Trend

An area chart showing the algorithm's aggregate accuracy evolution over time, with:

• Weekly cumulative accuracy plotted as a smooth curve
• Total resolved divergences at each point
• Visual indication of whether accuracy is improving, stable, or declining

#### 7. Report Snapshot Archive (Admin Only)

Admins can access historical weekly snapshots showing:

• Snapshot date and overall accuracy at that point
• Total instruments and divergences at snapshot time
• Week-over-week accuracy change with directional indicators
• Manual snapshot generation via "Take Snapshot Now" button

#### 8. Methodology & Limitations

Transparent disclosure of:

• How directional accuracy is measured
• Tier-specific resolution horizons (mega-cap: 10 days, commodity: 25 days, crypto: 7 days)
• Regime-aware horizon adjustment (crisis: 0.6x, high-vol: 0.8x, recovery: 1.3x, steady: 1.0x)
• Partial move acceptance (any shift in PG's predicted direction counts)
• Statistical testing assumptions (independence of predictions)
• Limitations of historical analysis
• Badge classification thresholds

Badge Classification

Tier-Specific Resolution Parameters

Statistical References

• Wilson, E.B. (1927). "Probable Inference, the Law of Succession, and Statistical Inference." Journal of the American Statistical Association, 22(158), 209-212.
• Agresti, A. & Coull, B.A. (1998). "Approximate is Better than 'Exact' for Interval Estimation of Binomial Proportions." The American Statistician, 52(2), 119-126.
• Cohen, J. (1988). Statistical Power Analysis for the Behavioral Sciences. 2nd ed. Hillsdale, NJ: Lawrence Erlbaum.
• Ang, A. & Timmermann, A. (2012). "Regime Changes and Financial Markets." Annual Review of Financial Economics, 4, 313-337.
• Brogaard, J., Hendershott, T. & Riordan, R. (2014). "High-Frequency Trading and Price Discovery." Review of Financial Studies, 27(8), 2267-2306.

PDF Export

The report can be exported as a PDF via the "Export PDF" button, which generates a print-optimized HTML document with all sections, tables, and statistical data formatted for publication.

Limitations

• Accuracy metrics are based on historical signal data and do not guarantee future performance
• The binomial test assumes independence of predictions, which may not hold in correlated markets
• Badge classifications are relative to PulseGrid's own history, not absolute benchmarks
• Results should be interpreted in the context of the current market regime and data accumulation phase