Market Intelligence
Swoopr Crowding Index
Swoopr Crowding Index is a phased crowding score for stocks, sectors, themes and asset classes combining momentum, positioning proxies, short interest, ownership concentration and valuation and volume acceleration. It is designed to answer one narrow question: Where are investors making unusually similar bets? The output is educational market context, not a forecast, recommendation, or promise of future returns.
Direct Answer
Direct answer: The Swoopr Crowding Index is a composite score that estimates how crowded a stock, sector, theme, or asset class is by combining momentum, positioning proxies, short interest, ownership concentration, valuation relative to history, and volume signals. High crowding does not cause immediate losses, but it raises exit risk: when sentiment shifts, crowded positions tend to unwind faster and more disorderly than uncrowded ones because more holders need to exit the same door. The index is intended to inform position sizing and timing review, not as a standalone sell signal.
What It Measures
The central question this tool answers: Where are investors making unusually similar bets?
The evidence model draws from the following component families:
- Relative momentum
- Price acceleration
- Volume intensity
- Short-interest level and change
- Margin context
- Institutional and ETF ownership concentration (where licensed)
- Options positioning (where licensed)
- Valuation expansion
- Theme-level correlation
The objective is not to maximize the number of inputs but to capture independent information. Every component needs an independence rationale in the methodology registry explaining what unique information it adds and where it overlaps with other components.
How the Score Works
Phase 1 uses permitted price and volume data, FINRA short-interest and aggregate margin context, plus concentration metrics. Phase 2 adds licensed options and ownership data when available. Each family is normalized and a crowding score is calculated only when minimum coverage is reached. A separate "data depth" badge indicates which phase is active.
Aggregate margin statistics are not combined with single-stock conclusions as if they were security-specific data. The methodology page states which approach is used for each component.
Every reading publishes coverage separately from the score. If data are missing, the affected component is excluded, coverage falls, and the page tells the user. Missing data are never converted to a neutral score. The UI uses terms such as Full coverage, Partial coverage, Stale component and Methodology fallback.
Interpretation
| Score | Label | Interpretation |
|---|---|---|
| 0 to 24 | Low crowding | Participation and positioning are broadly distributed; no unusual concentration detected. |
| 25 to 39 | Moderate | Some concentration signals present; within historical norms. |
| 40 to 59 | Elevated | Multiple crowding signals active; positioning is becoming more concentrated. |
| 60 to 74 | High | Strong crowding signals across multiple families; elevated unwind risk. |
| 75 to 100 | Extreme | Unusually concentrated positioning across all measured families. |
These bands are communication aids, not natural laws. A move from 59 to 60 is not a fundamental break in market reality. The page always shows the numeric value, trend, component contributions and the prior reading so context is visible rather than artificial cliffs.
How to Read It
Beginner: The reading tells you whether many investors appear to be making similar bets in the same assets, sectors or themes. A high reading signals concentration, not necessarily an imminent reversal.
Intermediate: Each component family, its current score, trend and data source are displayed. The data depth badge shows whether Phase 1 (price/volume/short interest) or Phase 2 (licensed positioning data) is active.
Advanced: Raw series identifiers, transformations, lookbacks, normalization method, active weights, timestamps, missing-data rules and methodology version are exposed.
Failure Modes and Guardrails
- Positioning data are incomplete; FINRA short interest is semi-monthly and aggregate margin is not security-specific.
- Crowding can persist longer than expected before unwinding; high scores are not timing signals.
- Licensed ownership and options data may not be available for all instruments or periods.
- A crowded trade that does not unwind is not a methodology failure; the score measures participation, not outcome.
If data freshness exceeds the SLA, the component shows Delayed or Unavailable, preserves the last timestamp, and stops generating "current" language. If a data source changes definition or licensing, the affected component is disabled until it is reviewed.
Frequently Asked Questions
Is Swoopr Crowding Index a buy or sell signal?
No. It describes the condition named by the tool and does not recommend a transaction. A high or low reading can persist, and markets can move against the historical pattern associated with any indicator.
How often should it update?
At the fastest cadence supported by the slowest important component, with each sub-component carrying its own timestamp. FINRA short-interest data is semi-monthly; the score reflects that cadence rather than implying daily precision.
Why use a 0 to 100 scale?
A common scale makes heterogeneous inputs understandable and allows consistent components across Swoopr. The scale does not mean 80 is twice as good as 40, nor does it represent an 80% chance of a market outcome.
This score is educational market context only. It is not investment advice, a forecast, or a promise of future returns.