Advancing and Declining Volume: Breadth With Activity
Counting stocks and adding their traded volume can tell different stories.
If most stocks rise, I know the move is widespread by count. But where did the trading activity take place? Volume breadth asks a different question by adding the volume associated with advancing and declining stocks.
I think it is useful precisely because it need not agree with stock-count breadth. Agreement is evidence to examine; disagreement is a reason to look more closely.
Define the inputs
Classify each eligible stock by its comparable close versus the previous session’s close. Then take that stock’s entire session trading volume, measured consistently in shares.
Up volume, UV = sum of session volume for advancing stocks.
Down volume, DV = sum of session volume for declining stocks.
Unchanged volume, FV = sum of session volume for unchanged stocks.
The public advancing-volume and declining-volume construction is described in StockCharts’ A/D volume reference. We are studying the raw inputs here, before any cumulative line.
Importantly, a stock’s volume is assigned according to its final price classification. We are not identifying which individual trades were buyer-initiated or seller-initiated.
A hypothetical NSE basket
Assume all five stocks have valid comparable closes and volume:
| Stock | Closing direction | Volume, lakh shares |
|---|---|---|
| A | Up | 2 |
| B | Up | 1 |
| C | Down | 7 |
| D | Down | 1 |
| E | Unchanged | 1 |
Two stocks advanced and two declined. The count is balanced. But UV = 3 lakh, DV = 8 lakh and FV = 1 lakh. More share volume traded in the stocks that finished lower.
Net volume = UV − DV = −5 lakh shares.
Up/down volume ratio = UV/DV = 3/8 = 0.375, provided DV > 0.
Up-volume share of all classified volume = 100 × UV/(UV + DV + FV) = 25%.
If we exclude unchanged volume, the up-volume share among movers is 100 × 3/11 ≈ 27.27%. I would label that denominator explicitly rather than present the two percentages as interchangeable.
How I would use the comparison
Our five-stock count is neutral, while the volume split leans toward declining stocks. Before drawing a conclusion, I would notice that stock C accounts for seven of the twelve lakh shares. The apparent imbalance is concentrated.
In an Indian-market review, I would ask whether a similar pattern appears across several sessions and across the same basket used for the price comparison. One unusually active name can move a volume reading without broad agreement from the rest of the universe.
Volume is not the same as money flow
Adding shares differs from adding rupee turnover. A low-priced stock can trade many shares without representing the same rupee value as a high-priced stock’s smaller share volume. Price, stock splits and changes in trading activity complicate long historical comparisons.
Every executed trade has a buyer and a seller. Up volume should not be described as money entering the exchange, nor down volume as money leaving it. Those phrases imply more than the classification actually measures.
Zero down volume makes the ordinary UV/DV ratio undefined. Zero total classified volume makes a percentage unavailable. Missing volume needs separate handling rather than an assumed zero.
Put the two lenses together
We believe the useful question is whether a move is broad by count and how activity is distributed within it. Start with the basic stock-count lesson, then compare the public measures carefully. On IndexBreadth, use only the visible measures and labels; do not assume an unavailable volume view or infer its internal processing.