Passive Investing's Cognitive Labor Blind Spot¶
September 3, 2025
$13.29 trillion in passive equity assets are allocated based purely on market capitalization, without consideration of which companies sell human judgment that AI can replicate versus which can leverage AI to compress competitors' margins.
This creates a systematic allocation inefficiency affecting the $6.1 trillion global professional services market—firms whose primary product is cognitive labor that faces increasing automation pressure.
The Passive Allocation Machine¶
As of 2024, passive investing has reached historic milestones:
- 50.2% of US equity fund assets now in passive vehicles
- $13.29 trillion in total passive assets under management
- 2.3 percentage points annual growth in passive share since 2013
Passive funds allocate capital based purely on market capitalization without systematic screening for AI automation risk or AI leverage potential.
The rise of passive investing has created systematic allocation blind spots
The Cognitive Labor Economy¶
The professional services sector represents a significant portion of the economy:
- $6.1 trillion global professional services market (Research and Markets, 2025)
- $2.8 trillion US professional services revenue (SelectUSA, 2022)
- 11+ million workers in US professional services (SelectUSA, 2024)
- $58.7 billion in US management consulting (Consultancy.uk, 2017)
Within this broader sector, certain firms face pressure as AI systems reach the "good enough" threshold where cost-benefit analysis favors automation over human expertise.
What Is Cognitive Labor?¶
Cognitive labor firms sell human judgment at scale. Their core product is knowledge work: analysis, recommendations, pattern matching, and synthesis.
The disruption risk comes not from AI matching human quality, but from AI becoming a trusted source at a fraction of the cost for many use cases. These companies occupy a vulnerable middle ground between pure software businesses (where AI enhances existing scale) and physical operations (where hands-on work provides protection from automation).
Examples of high cognitive labor exposure:
- Research firms: Gartner, Forrester selling synthesized opinions
- Consultancies: Accenture's lower tiers doing templated analysis
- Rating agencies: Moody's standardizable credit assessments
- Recruiting firms: Robert Half's pattern matching at scale
- Ad agencies: WPP's media buying optimization decisions
- Financial advisors: LPL Financial's basic portfolio construction
- Real estate brokers: Compass's information arbitrage model
The Disintermediation Framework¶
Instead of counting AI mentions in earnings calls—which can be misleading marketing language—we propose a screening model based on business fundamentals. Management teams often make aspirational claims about AI adoption that don't translate into measurable business impact.
This framework would measure two opposing forces using quantifiable metrics from standard financial statements, focusing on what companies actually do rather than what they say.
Disintermediation Risk Index (DRI, 0-100)¶
Cognitive Labor Exposure (45% weight)¶
- Services revenue as percentage of total revenue
- High exposure: Services > 70% → Score 80-100
- Medium exposure: Services 30-70% → Score 40-80
- Low exposure: Services < 30% → Score 0-40
Labor-Intensive P&L Structure (25% weight)¶
- Formula: (SG&A + R&D) ÷ Revenue
- Analyzes both current level and 3-year trend
- Higher ratios with upward trends increase risk score
Switching Cost Indicators (15% weight)¶
- Uses contract liabilities as percentage of revenue
- Lower ratios indicate weaker customer retention
- Proxy for ease of customer departure
Labor Efficiency Gap (10% weight)¶
- Compares revenue per employee to industry benchmarks
- Below-benchmark efficiency suggests automation potential
- Companies with lower productivity may have more routine, automatable tasks
Margin Compression (10% weight)¶
- Compares current gross margin to 3-year average
- Declining margins indicate commoditization pressure
- Score increases as margins compress
AI Potential Index (API, 0-100)¶
Platform/Workflow Power (40% weight)¶
- Recurring/subscription revenue mix and growth trends
- Workflow ownership indicators (controls supply and demand sides)
- Higher scores for recurring mix > 50% with ecosystem control
Execution Leverage (30% weight)¶
- Revenue growth with expanding EBIT or OCF margins
- Measured over 8-12 quarters for consistency
- Formula: Revenue CAGR × Margin expansion factor
Labor Mix Optimization (20% weight)¶
- Compares revenue growth rate to employee growth rate
- Positive spread indicates software/automation leverage
- Uses 3-year average to smooth temporary effects
Build vs. Hire Investment (10% weight)¶
- R&D spending increasing while G&A spending stable/declining
- Requires stable gross margins to avoid false signals
- Indicates investment in automation over headcount
Scoring Methodology¶
The framework follows a standardized process to ensure comparability across companies and industries:
- Normalization: All factors computed as z-scores within GICS-3 industry groups
- Winsorization: Extreme values capped at 2.5th and 97.5th percentiles
- Scaling: Rescaled to 0-100 range within each industry
- Weighting: Apply the weighted average using percentages above
- Net Tilt: Calculate API minus DRI (range: -100 to +100)
Quadrant Classification¶
Continuous risk-potential spectrum showing gradual transitions between quadrants
| Quadrant | Criteria | Interpretation |
|---|---|---|
| AI Winners | API ≥ 60 & DRI ≤ 40 | Strong AI leverage, low automation risk |
| Tension Plays | API ≥ 60 & DRI > 40 | High potential but also high risk |
| Value Traps | API < 60 & DRI > 60 | High automation risk, limited AI leverage |
| Stable Laggards | API < 60 & DRI ≤ 40 | Low risk but limited upside potential |
Current Market Examples¶
High Risk: Cognitive Labor Targets (Estimates)¶
| Company | Ticker | DRI | API | Net Tilt | Quadrant |
|---|---|---|---|---|---|
| Gartner | IT | ~75 | ~45 | -30 | Value Trap |
| LPL Financial | LPLA | ~70 | ~35 | -35 | Value Trap |
| Robert Half | RHI | ~80 | ~25 | -55 | Value Trap |
Gartner (IT): Magic Quadrants are structured data collection. Once AI recommendations normalize, their "CYA" brand protection transfers to "we used the AI."
LPL Financial (LPLA): Robo-advisors already attacking the low end. Mid-market advice becomes increasingly standardizable.
Robert Half (RHI): AI recruiting tools can screen resumes, match candidates, and assess fit at scale.
Winners: AI Disintermediators (Estimates)¶
| Company | Ticker | DRI | API | Net Tilt | Quadrant |
|---|---|---|---|---|---|
| Amazon | AMZN | ~20 | ~85 | +65 | AI Winner |
| Uber | UBER | ~30 | ~80 | +50 | AI Winner |
| Toast | TOST | ~25 | ~75 | +50 | AI Winner |
Amazon (AMZN): AI agents will favor the platform with best pricing, shipping, and payment integration.
Uber (UBER): Each ride makes the algorithm better. Expanding into autonomous vehicles and delivery.
Toast (TOST): Replacing restaurant consultants and back-office providers with integrated software.
Tension Zone: Can Go Either Way (Estimates)¶
| Company | Ticker | DRI | API | Net Tilt | Quadrant |
|---|---|---|---|---|---|
| WPP | WPP | ~65 | ~55 | -10 | Tension Play |
WPP (WPP): Agency model under pressure, but owns client relationships and campaign data.
The Systemic Mispricing¶
This isn't theoretical. Every month, passive funds mechanically rebalance $13+ trillion without asking fundamental questions:
- Does this company's revenue depend on human pattern matching that AI can automate?
- Can AI agents route around this firm's role in the value chain?
- Is this firm positioned to use AI to compress competitors' margins?
The result: potential overallocation to cognitive labor firms and underallocation to AI disintermediators.
Scale of the opportunity: Active managers can make concentrated bets unavailable to passive funds. A focused strategy allocating 10% to AI Winners (vs. ~5% market weight) while avoiding cognitive labor exposure entirely (vs. ~3% market weight) could generate significant alpha during the disruption cycle. Assuming AI Winners outperform by +10% annually and Value Traps underperform by -15%, this positioning delivers 95 basis points of annual alpha (5% overweight × 10% + 3% avoidance × 15%). Applied across active strategies, this represents meaningful outperformance potential as AI disruption unfolds.
Investment Approach¶
Long Basket: AI Disintermediators¶
Selection Criteria: - Cognitive labor exposure < 20% - Platform characteristics > 50%
Key Indicators: - Workflow ownership with data gravity and network effects - Margin expansion alongside revenue growth - Revenue growth exceeding headcount growth - Rising recurring revenue mix
Avoid/Short Basket: Cognitive Labor Targets¶
Selection Criteria: - Cognitive labor exposure > 60% - Switching costs < 30%
Key Indicators: - Services-heavy P&L structures - Standardizable knowledge work - Low barriers to customer departure - Margin compression trends - Inferior Good Preference
The Bottom Line¶
Passive investing allocates $13+ trillion without systematic consideration of AI automation risk or AI leverage potential across different business models.
Firms selling human judgment at scale receive the same capital allocation treatment as those building AI-powered platforms to automate similar functions. This creates potential opportunities for active strategies that screen for cognitive labor exposure and AI implementation capability.
Methodology Note¶
All analysis based on publicly available financial data including:
- Revenue mix and segment disclosures
- Margin trends and employee productivity metrics
- Contract characteristics and retention indicators
- Capital allocation patterns (R&D, capex, headcount)
No reliance on management guidance or AI marketing claims.