Market Breadth: Why the Eligible Stock Count Matters
A participation percentage is only as clear as the basket behind it.
When someone tells me that 70% of stocks are participating, my first question is: 70% of which stocks?
It sounds like a small detail. I think it is one of the most important questions in breadth analysis. The numerator tells us how many stocks met a condition. The denominator tells us what that count is being compared with.
After learning about advances, declines and unchanged stocks, we can put their denominator under the lens.
Following Breadth Analysis 101, this lesson looks at the raw stock counts behind a percentage. All figures below are hypothetical; they are not readings from IndexBreadth or a historical NSE session.
Total stocks and eligible stocks are different ideas
Imagine we choose a basket of 100 NSE stocks.
Total membership, T, is the number of stocks in that chosen basket at the time we are studying it: 100 in this example.
Eligible stocks, E, are the members with the observations required for a particular calculation. Eligibility is specific to the measure. A daily advance/decline classification needs comparable prices for two sessions. A 200-session moving-average calculation needs a sufficiently long price history under the stated convention.
Qualifying stocks, Q, are the eligible members that meet the condition: for example, stocks whose close is above their 200-session simple moving average.
A stock can belong to the basket without having enough history for every indicator. Missing observations should not silently become bearish observations.
The public percentage calculation
For a deliberately defined eligible-stock measure:
Participation percentage = Q ÷ E × 100, provided E > 0.
Alongside it, I would report:
Coverage percentage = E ÷ T × 100, provided T > 0.
The first percentage describes the eligible stocks. The second describes how much of the intended basket we could actually measure. If E is zero, participation is unavailable, not zero percent.
The use of stocks above a moving average as a participation measure is part of the public breadth literature; Greg Morris discusses that concept here. The coverage examples in this lesson are illustrations of denominator choice, not a description of a proprietary indicator.
Example: 54 stocks above their moving average
Suppose our basket contains 100 stocks, but only 90 have the required observations. Of those 90 eligible stocks, 54 are above the selected moving average.
| Item | Count or result |
|---|---|
| Total basket membership, T | 100 |
| Eligible stocks, E | 90 |
| Eligible stocks above the average, Q | 54 |
| Participation, 54 ÷ 90 × 100 | 60% |
| Coverage, 90 ÷ 100 × 100 | 90% |
The accurate statement is: 60% of eligible stocks were above the moving average, with 90% coverage of the basket.
Dividing 54 by all 100 members produces 54%. That can be reported as the known qualifying share of total membership, but it is not the same eligible-stock participation measure. We do not know the missing ten stocks’ classifications from this example.
Likewise, the remaining 36 eligible stocks are not necessarily all below the average. Some could be exactly equal to it. “Above” and “not above” are complements; “above” and “below” require an explicit equality rule.
Why a percentage can improve without a recovery
Here is the part I would pay particular attention to.
On one session, 60 of 100 eligible stocks meet a condition: participation is 60%. On the next session, observations for 20 non-qualifying stocks are unavailable. Assume the other classifications remain unchanged.
Now the same 60 qualifying stocks are divided by 80 eligible stocks: 60 ÷ 80 × 100 = 75%.
No additional stock qualified. Coverage fell from 100% to 80%. Calling that a broadening recovery would confuse a data change with a market change.
This is why I would check the underlying counts when a breadth percentage moves sharply. The chart may be describing a real improvement, but the percentage alone does not establish that.
Practical questions for an Indian-market comparison
Before comparing Nifty 50 participation with a broader NSE basket, I would check that the condition and observation time match. I would also check how much of each basket is measurable.
A smaller basket gives each stock a larger influence on an equal-count percentage. One stock is two percentage points in a fully eligible 50-stock basket, compared with 0.2 percentage points in a fully eligible 500-stock basket.
For historical work, we also need to distinguish constituents at the historical date from today’s constituents examined backwards. Those are different samples. A neat denominator does not remove survivorship bias or make the two histories interchangeable.
Limits and what to check on IndexBreadth
A high coverage percentage does not guarantee clean prices. Stale observations and inconsistent corporate-action adjustments can still distort classifications. Eligibility rules should be stated consistently rather than changed to make a chart look better.
On IndexBreadth, use the visible universe, timeframe and indicator labels as the starting point. If a denominator or coverage figure is not shown, do not infer it from a rounded percentage. This lesson explains a public calculation convention; it does not claim that every platform uses the same one.
We believe a useful breadth reading should make the comparison easier to understand. “60% of eligible stocks, covering 90% of the basket” is a little longer to say, but it tells us much more than “breadth is 60%.”