Direct Answer

Breadth, volatility, and sentiment indicators overlap because many popular indicators in each category are calculated from the same underlying market data. The VIX and the equity put/call ratio, for example, are both derived primarily from CBOE options activity, so treating them as two independent confirmations of "fear" double counts a single data source. Combining the three categories without double counting means grouping indicators by their underlying data family, options-derived, survey-derived, price-participation-derived (breadth), and realized-price-derived, and selecting at most one representative indicator per family rather than one indicator per category label. The main limitation is that even a well-built, non-redundant composite score describes current conditions across independent data sources; it does not predict the size or timing of any future move.

Key Takeaways

  • Double counting happens when two indicators labeled differently (one called "volatility," one called "sentiment") are actually calculated from the same underlying data, most commonly CBOE options activity.
  • Four broadly independent data families exist: options-derived (VIX, put/call ratio, skew), survey/positioning-derived (AAII, Investors Intelligence, CFTC Commitment of Traders, 13F filings, insider Form 4), price-participation-derived (advance-decline line, percent above moving average, new highs/lows, McClellan Oscillator), and realized-price-derived (historical volatility, average true range, drawdown depth).
  • A composite score built from one representative indicator per family carries more genuinely independent information than one built from five indicators that happen to span only two families.
  • Market breadth and volatility can move together during a broad selloff (both reflect real stress) without being the same underlying data, breadth counts stock-level price behavior, volatility is derived from options pricing or return dispersion.
  • Checking each indicator's raw data source, not just its category label, is the reliable way to identify overlap before building a composite.

Definition / Mechanism

Double counting, in this context, means treating two or more indicators as independent evidence when they are actually derived from a shared underlying data source. The mechanism is straightforward: many indicators that get grouped under different category labels (sentiment, volatility, breadth) are calculated from a small number of raw exchange data feeds, and when two indicators from the same feed both shift in the same direction, they are reporting the same underlying event twice, not confirming it from two separate angles.

Four data families cover most commonly used indicators in this cluster:

  1. Options-derived: the CBOE Volatility Index (VIX) and the equity/index put/call ratio are both calculated from options market activity, prices for VIX, contract volume for put/call. Options skew (the relative price of out-of-the-money puts versus calls) is also in this family. All three tend to rise together when options traders bid up downside protection, because they are reading the same underlying options order flow through different formulas.
  2. Survey/positioning-derived: the AAII Investor Sentiment Survey, Investors Intelligence, the CFTC Commitment of Traders report, 13F institutional filings, and insider Form 4 filings all come from investors directly reporting an opinion or a filed position, independent of options or price data.
  3. Price-participation-derived (breadth): the advance-decline line, percent of stocks above a moving average, new highs versus new lows, and the McClellan Oscillator are all built from daily counts of how individual stocks in an index traded relative to their own prior levels. See the Market Breadth & Participation Guide for full mechanics of this family.
  4. Realized-price-derived: historical (realized) volatility and average true range are calculated from the dispersion of an index's own past returns, distinct from options-implied volatility even though both are commonly labeled "volatility."

How to Analyze It

Building a composite market-condition score without double counting follows a simple sequence. First, list every indicator under consideration and identify its raw data source, not its category label. Second, group indicators that share a raw data source into a single family, regardless of how many different names or formulas describe them. Third, select at most one representative indicator per family, choosing the one with the longest track record or the clearest, most widely reported methodology. Fourth, combine the selected representatives, commonly as a simple average of standardized (z-scored) readings, or as a rules-based count of how many families are simultaneously at an extreme.

finance business Breadth Volatility Sentiment analyze
Photo by geralt via Pixabay

A practical shortcut for the grouping step: ask whether two indicators would move if the underlying options market, the underlying survey population, or the underlying stock universe stayed unchanged while only the other input changed. If moving one input mechanically moves both indicators, they are in the same family.

Worked Example

The following example uses invented figures to show the difference between a double-counted composite and a properly diversified one.

  1. Goal: Build a simple "market stress" composite that flags when three or more indicators are simultaneously at an extreme reading.
  2. Naive version (double counted): The composite uses the VIX (above 28), the put/call ratio (above 0.85), and options skew (25-delta put implied volatility more than 8 points above 25-delta call implied volatility). All three inputs are options-derived. In a real selloff, these three tend to cross their thresholds within the same one to three trading sessions, because they are all reading the same underlying options order flow. When all three trigger together, the composite reports "3 of 3 indicators extreme," which looks like broad confirmation but is actually one options-market signal counted three times.
  3. Corrected version (diversified): The composite instead uses the VIX (options family, above 28), the AAII bearish reading (survey family, above 45%), and the percentage of S&P 500 constituents above their 50-day moving average (breadth family, below 30%). Each of these draws from a genuinely different raw data source: options pricing, a direct investor survey, and daily stock-level price behavior. When all three cross their thresholds together, that agreement reflects three independent investor behaviors pointing the same direction, a materially stronger signal than the naive version's single options-market move counted three times.
  4. Why this matters in practice: A trader using the naive version could see "3 of 3 extreme" and treat it as unusually strong confirmation, when the actual independent evidence is closer to "1 of 3." The corrected version's "3 of 3" genuinely reflects three separate investor populations reaching an extreme at the same time.

What It Tells You

A composite built from genuinely independent data families gives a more trustworthy read of how widespread stress or optimism actually is across different parts of the market, options traders, survey respondents, and individual stock price behavior, rather than how strongly one of those groups is currently reacting. When independent families agree, that agreement is more informative than any single family's reading alone, since it is harder for three different data sources to all be reflecting the same narrow, transient event.

What It Does Not Tell You

Even a properly diversified composite does not predict the size, timing, or duration of any subsequent price move, it describes current conditions across independent data sources, not a forecast. It also does not account for structural or fundamental deterioration (earnings, credit conditions, macro data) that can justify readings that look like emotional extremes but are actually rational reactions to genuinely worsening conditions. No composite score, however carefully built, is a substitute for checking the fundamental backdrop separately.

Common Mistakes

Grouping by Category Label Instead of Data Source

Treating "sentiment," "volatility," and "breadth" as three automatically independent buckets, without checking what raw data each specific indicator is built from, is the root cause of most double counting. The category label describes what an indicator is commonly called, not what data feeds it.

finance business Breadth Volatility Sentiment
Photo by Ylanite via Pixabay

Never Testing Correlation Within a Personal Indicator Basket

An indicator basket assembled over time, adding a new indicator whenever an interesting one is discovered, tends to accumulate correlated variations of the same underlying signal. Periodically checking the rolling correlation between every pair of indicators in the basket catches this drift.

Treating Agreement Among Correlated Indicators as Strong Confirmation

As shown in the worked example above, three options-derived indicators crossing their thresholds together is expected behavior for a single options-market move, not three independent confirmations.

Using Default Indicator Sets Without Reviewing Their Composition

Pre-built "fear and greed" style composites often mix indicators from multiple families deliberately, which is good practice, but assuming every off-the-shelf composite is automatically non-redundant is not safe. Check the stated methodology of any third-party composite before treating its component count as a count of independent evidence.

Practical Checklist

  1. List every indicator in your basket alongside its actual raw data source, not just its common name or category label.
  2. Group indicators that share a raw data source into one family, even if their formulas or names look different on the surface.
  3. Select at most one representative indicator per family for a composite score meant to represent broad, independent evidence.
  4. If you want more depth within one family (for example, VIX level plus VIX term structure), keep that as a within-family confirmation, not as two additional "independent" composite inputs.
  5. Periodically check the correlation between every pair of indicators in your basket to catch drift toward redundancy.
  6. Remember that even a properly diversified composite score describes current conditions, not a forecast of magnitude or timing.

FAQ

Why are VIX and the put/call ratio considered the same data family?

Both are derived primarily from CBOE options market activity. The VIX is calculated from the prices of S&P 500 index options across a range of strikes, and the put/call ratio is calculated from the volume of put versus call contracts traded. When options traders bid up downside protection, both measures tend to rise together, because they are reading two different summaries of the same underlying options order flow rather than two independent observations of market conditions.

Is it ever appropriate to use two indicators from the same data family?

Yes, when the goal is confirming a single family's signal rather than building an independent composite. For example, checking both the VIX level and its term structure (whether near-term or longer-dated options imply higher volatility) adds useful detail about the same options family's signal. The double-counting problem specifically arises when someone treats two same-family indicators as two independent pieces of evidence in a composite score meant to represent broad market stress across different investor behaviors.

How do I know if two indicators are correlated enough to count as one family?

The most reliable check is examining the underlying data source each indicator is built from. If two indicators are calculated from the same exchange data feed (both from options volume, or both from the same daily advance/decline tape), treat them as one family regardless of how different their formulas look on the surface. A rolling correlation coefficient between the two series over a multi-year period is a useful secondary check, sustained correlation above roughly 0.7 to 0.8 is a signal the two are largely redundant for composite-scoring purposes, though the underlying-data-source check is the more reliable test since correlation can shift across different market regimes.

Does combining more indicators always produce a better signal?

No. Adding more indicators improves a composite signal only when each addition provides genuinely new information from a different data family. Beyond three to four well-chosen, independent families, additional indicators tend to either duplicate an existing family or add noise from a narrower, less-tested data source. A smaller set of deliberately independent indicators is generally more useful than a large set that includes several correlated variations of the same underlying signal.

Where does market breadth fit if it correlates with volatility during selloffs?

Breadth stays a separate family from volatility even though the two often move together during a broad selloff, because they are built from different raw data: breadth counts individual stocks' price behavior against their own thresholds (moving averages, prior highs and lows), while volatility (VIX, realized volatility) is derived from options pricing or the dispersion of the index's own returns. The fact that both often decline or spike together during stress is a real market relationship worth noting, not evidence that they are measuring the same underlying data and should be treated as one indicator.

How many genuinely independent signal families are available in practice?

Fewer than the number of indicators suggests. Most widely used measures collapse into a handful of underlying sources: options market activity, survey responses, positioning and flow data, price-derived breadth, and borrow and short data. Within each family the members correlate strongly, so adding a second one adds little. A composite drawing on three or four distinct families is usually the practical limit before additional inputs are restating information already present.

How does weighting differ from selection when building a composite?

Selection decides which indicators enter; weighting decides how much each one counts. Double counting is a selection problem, and no weighting scheme fixes it: two indicators from the same family at half weight each still contribute that family at full weight. Weighting addresses a different question, which is whether a family deserves more influence because it has been more informative. Doing the selection work first makes the weighting decision meaningful rather than cosmetic.

Does double counting matter if all the indicators agree anyway?

It matters most then. Agreement is the output being used to justify conviction, and if three of the four agreeing indicators are computed from the same underlying data, the agreement is arithmetic rather than corroboration. The composite reports a strong consensus that reflects one source counted repeatedly. Confidence should scale with the number of independent families in agreement, not with the number of indicator names on the list.

Where do positioning measures fit alongside the survey and options families?

They form a third family, and one of the more independent ones. Futures positioning reports, fund flow data and short interest describe capital that has been committed, which is different from stated opinion and different from options activity that may be hedging. The overlap comes at extremes, when positioning, options demand and survey responses tend to converge, which is precisely when a composite is most likely to overstate how many separate things are saying the same thing.

References

Review notes: Sources verified as reachable and describing current methodology on 2026-08-20. The VIX and put/call ratio methodologies, the AAII survey structure, and breadth-indicator construction are all stable, low-change-frequency methodologies; no time-sensitive rule or numeric threshold in this guide requires a shorter review cycle than Swoopr's standard annual content review.

Disclaimer

This guide is for educational and informational purposes only and does not constitute investment advice. Combining indicators, even without double counting, does not guarantee accurate market predictions. All trading involves risk, including the possible loss of principal. Always consult a qualified financial professional before making investment decisions.