cnhkmcp 2.1.9__py3-none-any.whl → 2.3.0__py3-none-any.whl

cnhkmcp/untracked/APP/trailSomeAlphas/skills/brain-data-feature-engineering/examples.md

@@ -0,0 +1,244 @@
+# Case Study: BEME Dataset Analysis
+
+## Dataset Overview
+
+**Dataset ID**: BEME (Balance Sheet and Market Data)
+**Description**: Book-to-market ratio and related financial metrics derived from balance sheet data combined with market data
+**Region**: USA
+**Universe**: TOP3000
+**Delay**: 1
+**Fields Analyzed**: 45 data fields
+
+## Step 1: Field Deconstruction
+
+### Key Fields Analyzed:
+
+1. **book_value_per_share**
+   - **What is it?**: Accounting net asset value divided by shares outstanding
+   - **How measured?**: Quarterly financial statements (audited)
+   - **Time dimension**: Quarterly snapshots (lagged)
+   - **Business context**: Represents historical cost-based net worth
+   - **Generation logic**: (Total assets - Total liabilities) / shares_outstanding
+   - **Reliability**: High (audited), but backward-looking and conservative
+
+2. **market_cap**
+   - **What is it?**: Share price × shares outstanding (total market valuation)
+   - **How measured?**: Real-time market data (continuous)
+   - **Time dimension**: Instantaneous, changes continuously
+   - **Business context**: Market participants' collective assessment of value
+   - **Generation logic**: Last traded price × total shares
+   - **Reliability**: Market-based, forward-looking, sentiment-influenced
+
+3. **book_to_market**
+   - **What is it?**: Ratio of book value to market value
+   - **How measured?**: Calculated from book_value and market_cap
+   - **Time dimension**: Compares slow-moving (book) with fast-moving (market)
+   - **Business context**: Compares accounting perspectives with market perspective
+   - **Generation logic**: book_value_per_share / (market_cap / shares)
+   - **Reliability**: Useful but must understand both components
+
+### Relationship Mapping:
+
+**The Story**: BEME tells the story of how market perception relates to accounting reality
+
+**Key Relationships**:
+- book_to_market connects two valuation perspectives
+- book_value changes slowly (quarterly, accountant-determined)
+- market_cap changes quickly (continuously, market-determined)
+- The gap represents market's view of intangible value
+
+**Missing Pieces**:
+- Why does the gap exist? (growth expectations, brand value, competitive position)
+- How persistent is the gap? (temporary vs. structural)
+- What causes gap changes? (earnings surprises, market sentiment, sector rotation)
+
+## Step 2: Question-Driven Feature Generation
+
+### Q1: "What is stable?" (Analyzing Invariance)
+
+**Feature Concept**: "Market re-evaluation stability"
+- **Implementation**: Rolling coefficient of variation of book_to_market over 60 days
+- **Definition**: Stability of the market's valuation vs. book value assessment
+- **Meaning**: Low CV = stable consensus, High CV = disagreement or uncertainty
+- **Interpretation**:
+  - High stability: Market has made up its mind about the company's valuation
+  - Low stability: Market is uncertain or volatile in its assessment
+- **Why it matters**: Stable mispricing (if book_to_market ≠ 1) can indicate structural factors
+
+**Feature Concept**: "Book value reliability"
+- **Implementation**: Autocorrelation of book_value changes over quarters
+- **Definition**: Consistency of book value reporting
+- **Meaning**: High autocorrelation = smooth reporting, Low = volatile changes
+- **Interpretation**: Sudden changes may indicate accounting adjustments or write-downs
+
+### Q2: "What is changing?" (Analyzing Dynamics)
+
+**Feature Concept**: "Valuation gap velocity"
+- **Implementation**: Rate of change of (market_cap - book_value × shares)
+- **Definition**: How quickly is the valuation gap changing?
+- **Meaning**: Fast increase = market becoming more optimistic or accounting write-downs
+- **Interpretation**:
+  - Positive velocity and acceleration: Market optimism increasing (bubble forming?)
+  - Positive velocity, negative acceleration: Optimism plateauing
+- **Why it matters**: Speed of gap change predicts sustainability
+
+**Feature Concept**: "Book vs. market growth decomposition"
+- **Implementation**: Separate book_value growth from market_cap growth
+- **Definition**: book_growth = (BV_t - BV_{t-1}) / BV_{t-1}
+- **Definition**: market_growth = (MC_t - MC_{t-1}) / MC_{t-1}
+- **Meaning**: Which is driving the book_to_market change?
+Interpretation**:
+  - book_growth > market_growth: Company building real value faster than market recognizes
+  - market_growth > book_growth: Market expectations running ahead of actual performance
+  - **Why it matters**: Distinguishes fundamental from sentiment-driven changes
+
+### Q3: "What is anomalous?" (Analyzing Deviation)
+
+**Feature Concept**: "Unusual valuation persistence"
+- **Implementation**: Days since book_to_market crossed 1.0 (either direction)
+- **Definition**: How long has the stock been valued differently from book?
+- **Meaning**: Persistent premium/discount suggests structural factors
+**Interpretation**:
+  - High persistence: Market has structural view (e.g., growth company, asset-light model)
+  - Low persistence: Temporary mispricing that corrects
+- **Why it matters**: Persistence indicates conviction level
+
+**Feature Concept**: "Book value surprise magnitude"
+- **Implementation**: Actual book_value vs. expected (trend-based forecast)
+- **Definition**: Unexpected change in book value
+- **Meaning**: Large surprises may indicate accounting adjustments
+- **Interpretation**: Positive surprise = asset appreciation, Negative = write-downs
+
+### Q4: "What is combined?" (Analyzing Interactions)
+
+**Feature Concept**: "Intangible value proportion"
+- **Implementation**: (market_cap - book_value × shares) / enterprise_value
+- **Definition**: What portion of enterprise value comes from non-book sources?
+- **Meaning**: Quantifies growth expectations, brand, competitive advantages
+**Interpretation**:
+  - High proportion: Value is in intangibles (risky but potentially high-growth)
+  - Low proportion: Value is in tangible assets (safer but limited growth)
+- **Why it matters**: Helps understand the nature of the company's value
+
+**Feature Concept**: "Valuation tug-of-war"
+- **Implementation**: book_momentum × market_momentum (where momentum is rate of change)
+- **Definition**: Are book and market moving in same or opposite directions?
+- **Meaning**: Agreeing signals vs. diverging signals
+**Interpretation**:
+  - Positive × positive: Both growing (healthy expansion)
+  - Positive × negative: Market doubts book value growth (potential concern)
+  - Negative × positive: Market optimistic despite book declines (turnaround story?)
+  - Negative × negative: Both declining (distressed situation)
+
+### Q5: "What is structural?" (Analyzing Composition)
+
+**Feature Concept**: "Value composition stability"
+- **Implementation**: Rolling correlation between book_growth and market_growth
+- **Definition**: How consistent is the relationship between accounting and market value?
+- **Meaning**: Stable correlation = predictable relationship, Unstable = relationship breaking down
+- **Interpretation**: Declining correlation suggests business model change or market re-evaluation
+
+**Feature Concept**: "Asset backing sufficiency"
+- **Implementation**: book_value / (market_cap / shares) when book_to_market > 1
+- **Definition**: How much asset coverage for market valuation?
+- **Meaning**: Mercantile/asset-heavy businesses should have high ratios
+- **Why it matters**: Helps identify when market undervaluation may be justified (e.g., declining industry)
+
+### Q6: "What is cumulative?" (Analyzing Accumulation)
+
+**Feature Concept**: "Accumulated valuation premium/discount"
+- **Implementation**: Time-weighted sum of (market_cap - book_value) over 1 year
+- **Definition**: Cumulative deviation from book value over time
+- **Meaning**: Persistent premium = sustained growth expectations
+**Interpretation**:
+  - High positive accumulation: Market consistently optimistic
+  - Near zero: Market fluctuates around book value
+  - High negative accumulation: Market consistently pessimistic
+
+**Feature Concept**: "Book quality decay"
+- **Implementation**: Age of assets (based on depreciation schedules) weighted by value
+- **Definition**: How old/stale is the book value?
+- **Meaning**: Older book values less reliable (assets may be obsolete)
+- **Why it matters**: Book value quality affects interpretation of book_to_market
+
+### Q7: "What is relative?" (Analyzing Comparison)
+
+**Feature Concept**: "Sector-relative valuation gap"
+- **Implementation**: Company book_to_market - sector median book_to_market
+- **Definition**: How does valuation gap compare to industry peers?
+- **Meaning**: Sector-relative premium or discount
+**Interpretation**:
+  - Premium vs. sector: Justified if company has better prospects
+  - Discount vs. sector: Potential opportunity or justified by worse fundamentals
+
+**Feature Concept**: "Relative book value trend"
+- **Implementation**: Company's book_growth - sector average book_growth
+- **Definition**: Is company building value faster or slower than peers?
+- **Meaning**: Competitive positioning in asset creation
+
+### Q8: "What is essential?" (Analyzing Essence)
+
+**Feature Concept**: "Core asset efficiency"
+- **Implementation**: book_value / total_assets (stripping out intangibles/goodwill)
+- **Definition**: What portion of assets are "hard" vs. "soft"?
+- **Meaning**: Asset-light businesses have lower ratios
+**Interpretation**:
+  - Low ratio: Intangible-based business (software, brands, networks)
+  - High ratio: Asset-heavy business (manufacturing, real estate)
+- **Why it matters**: Affects interpretation of book_to_market (intangibles not on books)
+
+**Feature Concept**: "Fundamental value anchor"
+- **Implementation**: book_value plus time-adjusted retained earnings
+- **Definition**: Book value adjusted for recent profitability
+- **Meaning**: Asset base plus earnings power
+**Why it's essential**: Combines two fundamental value sources
+
+## Step 3: Feature Documentation Table
+
+| Feature Concept | Fields Used | Question Answered | Logical Meaning | Directionality | Boundary Conditions |
+|----------------|-------------|-------------------|-----------------|----------------|---------------------|
+| Market re-evaluation stability | book_to_market | What is stable? | Consensus stability | Low=stable, High=disagreement | Zero=no variation, ∞=unstable |
+| Valuation gap velocity | market_cap, book_value | What is changing? | Gap change rate | Positive=widening, Negative=narrowing | Zero=no change |
+| Unusual valuation persistence | book_to_market | What is anomalous? | Premium/discount persistence | High=persistent belief | Zero=fluctuating |
+| Intangible value proportion | market_cap, book_value | What is combined? | Non-book value share | High=intangible-based | Zero=all tangible |
+| Value composition stability | book_growth, market_growth | What is structural? | Relationship consistency | High=stable relationship | Zero=breaking down |
+| Accumulated premium/discount | market_cap - book_value | What is cumulative? | Time-weighted deviation | High=consensus, Around zero=fluctuation | Negative=persistent pessimism |
+| Sector-relative gap | book_to_market, sector median | What is relative? | Peer comparison | Positive=premium to peers | Zero=sector average |
+| Core asset efficiency | book_value, total_assets | What is essential? | Hard asset proportion | High=asset-heavy, Low=intangible-based | 0-1 range |
+
+## Step 4: Implementation Insights
+
+### Why This Approach Works:
+
+1. **Novel**: Not just "moving averages of book_to_market" but deep conceptual features
+2. **Meaningful**: Each feature answers a specific question about the data
+3. **Testable**: Can validate if features capture what they claim to
+4. **Actionable**: Clear interpretation guides usage
+
+### Key Discoveries from Analysis:
+
+1. **book_to_market alone is incomplete**: Need to understand both components
+2. **Gap dynamics matter**: How the gap changes is more informative than level
+3. **Persistence is informative**: Long-term premium/discount suggests structural views
+4. **Comparative context essential**: Sector-relative measures remove noise
+5. **Asset composition affects interpretation**: Intangible-heavy businesses naturally have low book values
+
+### Suggestions for Further Analysis:
+
+1. **Add earnings data**: Connect book_to_market with profitability metrics
+2. **Add growth data**: Separate growth vs. value stories
+3. **Add sector context**: Industry cycles affect interpretation
+4. **Add sentiment data**: Market mood explains divergences
+5. **Add fundamental data**: ROE, margins, leverage affect valuation
+
+## Conclusion
+
+This analysis demonstrates how questioning data essence and asking fundamental questions generates meaningful features, not just mathematical transformations. Each feature:
+
+- Answers a specific question
+- Has clear logical meaning
+- Is grounded in data reality
+- Avoids conventional patterns
+- Reveals new insights
+
+The book_to_market ratio becomes more than just "value indicator"—it becomes a window into market psychology, accounting reliability, and fundamental vs. sentiment-driven valuation.

cnhkmcp/untracked/APP/trailSomeAlphas/skills/brain-data-feature-engineering/output_report/ASI_delay1_analyst11_ideas.md

@@ -0,0 +1,285 @@
+# ESG Scores Feature Engineering Analysis Report
+
+**Dataset**: analyst11
+**Region**: ASI
+**Delay**: 1
+
+
+**Dataset**: analyst11
+**Category**: Analyst
+**Region**: ASI
+**Analysis Date**: 2024-05-15
+**Fields Analyzed**: 50
+
+---
+
+## Executive Summary
+
+**Primary Question Answered by Dataset**: How do companies rank within their peer groups across various ESG (Environmental, Social, Governance) dimensions, with emphasis on which ESG factors show the strongest correlation to financial returns?
+
+**Key Insights from Analysis**:
+- This dataset provides multi-layered ESG rankings: raw scores, percentile rankings, and correlation-weighted rankings
+- The dataset distinguishes between "correlation-weighted" (any correlation) and "positive-correlation" (only positive correlations) metrics
+- Peer group comparisons are structured hierarchically: sector > industry > subsector
+- The dataset emphasizes which ESG factors matter most for financial performance, not just which companies score highest
+
+**Critical Field Relationships Identified**:
+1. Hierarchy of peer groups: sector (broadest), industry (middle), subsector (most specific)
+2. Three types of ESG metrics: raw scores, correlation-weighted, positive-correlation-weighted
+3. Three pillar structure: Environmental, Social, Governance, plus composite sustainability scores
+
+**Most Promising Feature Concepts**:
+1. **ESG Factor Alignment Gap** - because it measures the disconnect between a company's ESG performance and what the market actually rewards financially
+2. **ESG Consensus Strength** - because it identifies companies where all peer group rankings agree, suggesting clear ESG positioning
+3. **ESG Financial Relevance Score** - because it quantifies how much a company's ESG strengths align with financially material factors
+
+---
+
+## Dataset Deep Understanding
+
+### Dataset Description
+This dataset provides comprehensive ESG (Environmental, Social, Governance) scoring with a unique twist: it doesn't just measure ESG performance, but weights that performance by how strongly each ESG factor correlates with financial returns. The dataset includes percentile rankings within three peer group levels (sector, industry, subsector) and distinguishes between general correlation-weighted metrics and specifically positive-correlation-weighted metrics. This allows researchers to identify not just which companies are "good" at ESG, but which companies excel at the ESG factors that actually matter for financial performance.
+
+### Field Inventory
+| Field ID | Description | Data Type | Update Frequency | Coverage |
+|----------|-------------|-----------|------------------|----------|
+| community_maxcorr_sector_percentile | Percentile ranking within sector peer group for community score weighted by strongest correlation to financial returns. | Float | Quarterly | ~85% |
+| sustainability_sector_rank | Company's rank within its sector peer group for overall sustainability score. | Integer | Quarterly | ~85% |
+| governance_corr_weighted_industry_percentile | Company's percentile within its industry based on governance score weighted by KPIs most correlated to financial returns. | Float | Quarterly | ~85% |
+| workforce_sector_percentile | Company's percentile within its sector based on employee-related score. | Float | Quarterly | ~85% |
+| employee_training_sector_rank | Company's rank within its sector peer group for employee training, safety, and well-being. | Integer | Quarterly | ~85% |
+| sustainability_corr_weighted_industry_percentile | Company's percentile within its industry based on sustainability score weighted by KPIs most correlated to financial returns. | Float | Quarterly | ~85% |
+| disclosure_transparency_sector_percentile | Percentile ranking within sector peer group for disclosure, transparency, and accountability. | Float | Quarterly | ~85% |
+| workforce_corr_weighted_industry_percentile | Company's percentile within its industry based on employee-related score weighted by KPIs most correlated to financial returns. | Float | Quarterly | ~85% |
+| board_independence_industry_rank | Company's rank within its industry peer group for board independence and diversity. | Integer | Quarterly | ~85% |
+| workforce_positive_corr_sector_position | Company's position within its sector based on employee-related score weighted by KPIs most positively correlated to financial returns. | Float | Quarterly | ~85% |
+
+*(Additional fields follow similar patterns)*
+
+### Field Deconstruction Analysis
+
+#### community_maxcorr_sector_percentile: Community Max Correlation Sector Percentile
+- **What is being measured?**: How a company's community engagement performance compares to sector peers, but only considering the aspects of community engagement that show the strongest statistical relationship to financial returns
+- **How is it measured?**: Percentile ranking (0-100) within sector, with community score components weighted by their correlation coefficients to financial metrics
+- **Time dimension**: Point-in-time snapshot, likely quarterly or annually updated
+- **Business context**: Identifies companies that are good at the community engagement activities that actually impact financial performance
+- **Generation logic**: 1) Calculate correlation between community engagement sub-scores and financial returns, 2) Weight community score by these correlations, 3) Rank within sector, 4) Convert to percentile
+- **Reliability considerations**: Correlation stability over time, sector definition consistency, financial metric selection
+
+#### sustainability_sector_rank: Sustainability Sector Rank
+- **What is being measured?**: Absolute ranking position of a company's overall sustainability performance within its sector
+- **How is it measured?**: Integer rank (1 = best) based on composite sustainability score
+- **Time dimension**: Point-in-time ranking, ordinal rather than continuous
+- **Business context**: Shows where a company stands relative to sector competitors on overall sustainability
+- **Generation logic**: 1) Calculate composite sustainability score, 2) Sort all companies in sector by score, 3) Assign rank positions
+- **Reliability considerations**: Rank is sensitive to number of companies in sector, composite score weighting methodology
+
+#### governance_corr_weighted_industry_percentile: Governance Correlation Weighted Industry Percentile
+- **What is being measured?**: Governance performance percentile within industry peer group, weighted by governance factors' correlation to financial returns
+- **How is it measured?**: Percentile (0-100) within industry, with governance components weighted by their financial correlation
+- **Time dimension**: Snapshot with correlation weighting that may change slowly
+- **Business context**: Identifies governance leaders on factors that matter financially within specific industries
+- **Generation logic**: Industry-level version of correlation-weighted percentile calculation
+- **Reliability considerations**: Industry classification consistency, correlation calculation methodology
+
+### Field Relationship Mapping
+
+**The Story This Data Tells**:
+This dataset tells a sophisticated story about ESG performance with a financial lens. Instead of just asking "who's good at ESG?", it asks "who's good at the ESG factors that actually drive financial performance?" The data is structured in three dimensions: 1) ESG pillars (Environmental, Social, Governance, plus composites), 2) Peer group levels (sector, industry, subsector), and 3) Weighting methodologies (raw, correlation-weighted, positive-correlation-weighted). This creates a rich matrix for understanding not just ESG performance, but financially material ESG performance within relevant competitive contexts.
+
+**Key Relationships Identified**:
+1. **Hierarchy consistency**: For each ESG dimension, there are parallel metrics at sector, industry, and subsector levels
+2. **Weighting gradient**: Raw scores → correlation-weighted → positive-correlation-weighted represents increasing focus on financial materiality
+3. **Pillar interdependence**: Social scores include workforce, community, human rights subcomponents; Governance includes board, disclosure, etc.
+4. **Rank vs. Percentile duality**: Some fields provide ranks (absolute position), others percentiles (relative standing)
+
+**Missing Pieces That Would Complete the Picture**:
+- Time series of these metrics to track improvement/decline
+- The actual correlation coefficients used for weighting
+- Breakdown of which specific ESG factors have highest financial correlation
+- Market reaction data to validate the correlation-weighted approach
+
+---
+
+## Feature Concepts by Question Type
+
+### Q1: "What is stable?" (Invariance Features)
+
+**Concept**: ESG Ranking Consistency Score
+- **Sample Fields Used**: sector_percentile, industry_percentile, subsector_percentile
+- **Definition**: Measures how consistent a company's ESG ranking is across different peer group levels
+- **Why This Feature**: Companies with consistent rankings across sector/industry/subsector have more reliable ESG positioning
+- **Logical Meaning**: High values indicate ESG performance is robust regardless of peer group definition
+- **Directionality**: Higher = more consistent ESG positioning across peer groups
+- **Boundary Conditions**: 100 = perfect consistency, 0 = completely inconsistent rankings
+- **Implementation Example**: `abs({sector_percentile} - {industry_percentile}) + abs({industry_percentile} - {subsector_percentile})`
+
+**Concept**: ESG Financial Materiality Stability
+- **Sample Fields Used**: corr_weighted_score, positive_corr_score
+- **Definition**: Difference between general correlation-weighted score and positive-correlation-only score
+- **Why This Feature**: Measures stability of financial materiality signal - whether financially relevant ESG factors are consistently positive
+- **Logical Meaning**: Small difference suggests ESG factors that correlate with returns do so consistently positively
+- **Directionality**: Lower = more stable financial materiality signal
+- **Boundary Conditions**: 0 = all correlated factors are positively correlated, large values = mixed correlation directions
+- **Implementation Example**: `abs({corr_weighted_score} - {positive_corr_score})`
+
+---
+
+### Q2: "What is changing?" (Dynamics Features)
+
+**Concept**: ESG Peer Group Ranking Divergence
+- **Sample Fields Used**: sector_rank, industry_rank, subsector_rank
+- **Definition**: Standard deviation of rankings across different peer group levels
+- **Why This Feature**: Identifies companies whose ESG performance assessment depends heavily on peer group definition
+- **Logical Meaning**: High divergence suggests ESG performance is context-dependent or peer group sensitive
+- **Directionality**: Higher = more peer group dependent ESG assessment
+- **Boundary Conditions**: 0 = identical ranking across all peer groups
+- **Implementation Example**: `ts_std_dev({sector_rank}, 4) - ts_std_dev({industry_rank}, 4)`
+
+**Concept**: Financial Materiality Signal Strength Trend
+- **Sample Fields Used**: corr_weighted_percentile, positive_corr_percentile
+- **Definition**: Rate of change in the gap between correlation-weighted and positive-correlation rankings
+- **Why This Feature**: Tracks whether a company's ESG strengths are becoming more aligned with positively correlated factors
+- **Logical Meaning**: Negative trend = improving alignment with positively correlated ESG factors
+- **Directionality**: Downward trend = improving financial materiality alignment
+- **Boundary Conditions**: Steep negative = rapid improvement in financially material ESG
+- **Implementation Example**: `ts_delta({corr_weighted_percentile} - {positive_corr_percentile}, 90)`
+
+---
+
+### Q3: "What is anomalous?" (Deviation Features)
+
+**Concept**: ESG Factor Alignment Gap
+- **Sample Fields Used**: sector_percentile, corr_weighted_sector_percentile
+- **Definition**: Difference between raw ESG percentile and correlation-weighted percentile
+- **Why This Feature**: Identifies companies that are good at ESG generally but not at the ESG factors that matter financially
+- **Logical Meaning**: Large positive gap = excels at ESG factors that don't correlate with returns
+- **Directionality**: Higher = greater misalignment between ESG performance and financial materiality
+- **Boundary Conditions**: 0 = perfect alignment, >50 = major misalignment
+- **Implementation Example**: `{sector_percentile} - {corr_weighted_sector_percentile}`
+
+**Concept**: Peer Group Ranking Anomaly
+- **Sample Fields Used**: sector_rank, industry_rank
+- **Definition**: Absolute difference between sector rank and industry rank, normalized by peer group size
+- **Why This Feature**: Flags companies whose ESG assessment changes dramatically between sector and industry peer groups
+- **Logical Meaning**: High values suggest either data issues or genuinely context-dependent ESG performance
+- **Directionality**: Higher = more anomalous peer group ranking difference
+- **Boundary Conditions**: >30% difference = significant anomaly worth investigating
+- **Implementation Example**: `abs({sector_rank} - {industry_rank}) / max({sector_rank}, {industry_rank})`
+
+---
+
+### Q4: "What is combined?" (Interaction Features)
+
+**Concept**: ESG Financial Relevance Score
+- **Sample Fields Used**: corr_weighted_score, positive_corr_score, composite_score
+- **Definition**: Weighted average emphasizing correlation-weighted scores over raw scores
+- **Why This Feature**: Creates a single metric prioritizing financially material ESG factors
+- **Logical Meaning**: Higher values indicate strong ESG performance on factors that matter for returns
+- **Directionality**: Higher = better financially material ESG performance
+- **Boundary Conditions**: 100 = perfect on financially material factors, 0 = poor on all dimensions
+- **Implementation Example**: `0.4 * {corr_weighted_score} + 0.4 * {positive_corr_score} + 0.2 * {composite_score}`
+
+**Concept**: ESG Consensus Strength
+- **Sample Fields Used**: sector_percentile, industry_percentile, subsector_percentile
+- **Definition**: Inverse of ranking dispersion across peer group levels
+- **Why This Feature**: Identifies companies with consistent ESG assessment regardless of peer group definition
+- **Logical Meaning**: High consensus suggests robust, unambiguous ESG positioning
+- **Directionality**: Higher = more consistent ESG assessment across peer groups
+- **Boundary Conditions**: 100 = perfect consensus, 0 = completely inconsistent
+- **Implementation Example**: `100 - (abs({sector_percentile} - {industry_percentile}) + abs({industry_percentile} - {subsector_percentile}))`
+
+---
+
+### Q5: "What is structural?" (Composition Features)
+
+**Concept**: ESG Pillar Balance Ratio
+- **Sample Fields Used**: environmental_score, social_score, governance_score
+- **Definition**: Ratio of strongest pillar to weakest pillar performance
+- **Why This Feature**: Measures balance vs. specialization in ESG performance across pillars
+- **Logical Meaning**: Lower ratio = more balanced ESG performance; higher ratio = specialized strength
+- **Directionality**: Closer to 1 = more balanced; higher = more specialized
+- **Boundary Conditions**: 1 = perfectly balanced, >3 = highly specialized
+- **Implementation Example**: `max({environmental_score}, {social_score}, {governance_score}) / min({environmental_score}, {social_score}, {governance_score})`
+
+**Concept**: Financial Materiality Concentration
+- **Sample Fields Used**: corr_weighted_percentile, positive_corr_percentile
+- **Definition**: Proportion of correlation-weighted performance captured by positive-correlation factors
+- **Why This Feature**: Measures whether a company's financially material ESG strengths are in positively correlated areas
+- **Logical Meaning**: Higher = ESG strengths concentrated in factors with positive financial correlation
+- **Directionality**: Higher = better concentration in positively correlated factors
+- **Boundary Conditions**: 1 = all correlated factors are positive, 0.5 = mixed, 0 = all negative correlation
+- **Implementation Example**: `{positive_corr_percentile} / {corr_weighted_percentile}`
+
+---
+
+### Q6: "What is cumulative?" (Accumulation Features)
+
+**Concept**: ESG Improvement Momentum
+- **Sample Fields Used**: sector_percentile, industry_percentile
+- **Definition**: Weighted moving average of percentile improvements across time
+- **Why This Feature**: Captures sustained improvement trajectory in ESG rankings
+- **Logical Meaning**: Positive = improving ESG standing over time
+- **Directionality**: Higher = stronger improvement momentum
+- **Boundary Conditions**: >0 = improving, <0 = deteriorating
+- **Implementation Example**: `ts_sum(ts_delta({sector_percentile}, 30), 180)`
+
+**Concept**: Financial Materiality Alignment Trend
+- **Sample Fields Used**: corr_weighted_score, composite_score
+- **Definition**: Cumulative improvement in alignment between overall ESG and financially material ESG
+- **Why This Feature**: Tracks whether company is shifting ESG focus toward financially relevant factors
+- **Logical Meaning**: Positive = improving alignment with financially material ESG factors
+- **Directionality**: Higher = faster alignment improvement
+- **Boundary Conditions**: Steep positive = rapid strategic shift toward material ESG
+- **Implementation Example**: `ts_sum({corr_weighted_score} - {composite_score}, 360)`
+
+---
+
+### Q7: "What is relative?" (Comparison Features)
+
+**Concept**: ESG Relative Advantage Score
+- **Sample Fields Used**: sector_percentile, industry_percentile
+- **Definition**: Difference between sector and industry percentile rankings
+- **Why This Feature**: Measures whether a company performs better relative to broader or narrower peer groups
+- **Logical Meaning**: Positive = stronger relative to sector than industry; suggests competitive advantage erodes in closer peer comparison
+- **Directionality**: Positive = better in broader peer group; Negative = better in closer peers
+- **Boundary Conditions**: Large positive = "big fish in big pond"; Large negative = "specialist in niche"
+- **Implementation Example**: `{sector_percentile} - {industry_percentile}`
+
+**Concept**: Financial Materiality Premium
+- **Sample Fields Used**: corr_weighted_percentile, sector_percentile
+- **Definition**: Percentage improvement in ranking when considering only financially material factors
+- **Why This Feature**: Quantifies how much a company's ESG standing improves when focusing on what matters financially
+- **Logical Meaning**: Positive = company is better at financially material ESG than ESG overall
+- **Directionality**: Higher = greater financial materiality advantage
+- **Boundary Conditions**: >0 = stronger on material factors; <0 = weaker on material factors
+- **Implementation Example**: `({corr_weighted_percentile} - {sector_percentile}) / {sector_percentile}`
+
+---
+
+### Q8: "What is essential?" (Essence Features)
+
+**Concept**: Core ESG Financial Alignment
+- **Sample Fields Used**: positive_corr_score, composite_score
+- **Definition**: Ratio of positive-correlation-weighted score to overall composite score
+- **Why This Feature**: Distills ESG performance to its financially essential core - what actually matters for returns
+- **Logical Meaning**: Measures proportion of ESG value that is financially material and positively correlated
+- **Directionality**: Higher = greater proportion of ESG value is financially essential
+- **Boundary Conditions**: 1 = all ESG value is financially essential; 0 = none is financially essential
+- **Implementation Example**: `{positive_corr_score} / {composite_score}`
+
+**Concept**: Peer Group Invariant ESG Strength
+- **Sample Fields Used**: sector_percentile, industry_percentile, subsector_percentile
+- **Definition**: Minimum percentile across all peer group levels
+- **Why This Feature**: Conservative measure of ESG strength that holds regardless of peer group definition
+- **Logical Meaning**: Worst-case ESG ranking across all relevant peer comparisons
+- **Directionality**: Higher = stronger minimum guaranteed ESG standing
+- **Boundary Conditions**: 100 = top performer in all peer groups; 0 = bottom in at least one group
+- **Implementation Example**: `min({sector_percentile}, {industry_percentile}, {subsector_percentile})`
+
+---
+
+## Implementation Considerations
+
+### Data Quality Notes
+- **Coverage**: ~85% coverage for ASI