Volume Analysis: Confirming Moves and Spotting Exhaustion

Every share sold was bought by somebody. That sentence disposes of most of what gets written about volume. The figure contains no direction at all. Picture 8 million shares trading on a 5% decline. Every one of those shares found a buyer at some price on the way down, and the count of shares sold was identical to the count bought all day.

What the number measures is participation: how many holders and would-be holders were willing to transact while the price was moving, and it is the only thing on the chart not calculated from price, which makes it the one independent input available. That is genuinely useful.

Relative volume, and which window you choose

Raw share counts do not compare across securities. Use the ratio.

relative volume = today's volume / average daily volume over the last 50 days

Stock	Today	50-day average	Relative volume	Reading
A	6,200,000	2,000,000	6.2 / 2.0 = 3.1	Heavy participation, something happened
B	850,000	900,000	0.85 / 0.90 = 0.94	An ordinary session
C	310,000	1,400,000	0.31 / 1.40 = 0.22	Almost nobody traded, so treat the move as noise

The averaging window is a choice that changes the answer. A 50-day average smooths through one earnings report and two monthly expirations. That is what you want for judging today. A 10-day average reacts quickly. It is better for asking whether interest has been building over a fortnight. When the two disagree, that disagreement is the reading: above the 10-day and below the 50-day says activity is picking up from a quiet spell, with no claim about absolute size.

Dollar volume governs whether you can trade the name at all: shares traded x price. Stock C at $18 traded 310,000 x 18 = $5.6 million on the day, so a $200,000 position is a substantial share of a day’s business and exiting in a hurry will move the price against you.

The four combinations

Price	Volume	What the pair describes
Rising	Above average	Broad participation in the advance
Rising	Below average	An advance few people joined, easily given back
Falling	Above average	Broad participation in the decline, or forced selling
Falling	Below average	Drift lower because buyers are absent

The fourth row is misread constantly. A stock that slides 3% over five sessions on 0.6 times normal volume has not been heavily sold, and it drifted because nobody was bidding with any urgency, which is a different condition from distribution and often resolves the moment one buyer arrives.

The second row is the one that costs money. A breakout to a new high on 0.7 times normal volume cleared the level with almost no orders behind it, and the support and resistance guide explains why those breaks so often fall back into the range.

Climax volume, identifiable mainly in hindsight

Occasionally the heaviest volume in months prints at the end of a long move. A stock that has fallen for six weeks drops 9% on five times average volume, makes its low, and closes near the high of the session on a long lower wick. The reading is that capitulation has happened. Less supply is left to come.

That reading needs two supports before it becomes a position. The candle has to show the reversal inside the day. That is the wick arithmetic in the candlestick chart guide. And price has to trade above the high of that bar, which fixes the invalidation at the bar’s low.

On-balance volume, and the flaw the table exposes

OBV is a running total. Add the day’s volume on an up close. Subtract it on a down close.

Day	Close	Volume (m)	Direction	OBV (m)
1	50.00	1.0	starting point	0.0
2	50.80	1.4	up	0.0 + 1.4 = 1.4
3	50.20	1.1	down	1.4 - 1.1 = 0.3
4	51.50	2.2	up	0.3 + 2.2 = 2.5
5	51.40	0.9	down	2.5 - 0.9 = 1.6

The level means nothing. Day 1 was set to zero by choice. Slope is the reading, and the usual application is divergence: price grinding to a higher high while OBV makes a lower high says the newer high came on lighter participation.

Now look at day 5. The close fell by 0.10. The formula subtracted the entire 0.9 million shares. A 5% collapse would have subtracted the same amount. OBV treats a rounding error and a rout identically. It reads only the sign of the change. That is visible in the arithmetic and invisible on the chart, which is why I treat an OBV divergence as a prompt to go and look harder at the price series itself.

VWAP, calculated

VWAP = total dollars traded / total shares traded

Take three trades: 1,000 shares at 50.00, 2,000 at 50.40 and 1,000 at 50.80. Dollars are 50,000 + 100,800 + 50,800 = 201,600, shares are 4,000, so VWAP = 201,600 / 4,000 = 50.40.

Institutional desks are graded against this number, which is why price often reacts around it intraday. It resets at each open. Putting it on a daily chart misunderstands what it measures.

The volume that means nothing

Mechanical volume is common and carries no opinion. Quarterly index rebalances force funds to trade in size on a date everyone knows, and options expiration on the third Friday of each month adds hedging activity. ETF creation and redemption moves shares in the underlying companies for reasons that have nothing to do with those companies.

The averages themselves get distorted by the events they contain. The size of the distortion is easy to compute. Take a stock averaging 2.0 million shares. It trades 16 million on an earnings day. The new 50-day average is ((49 x 2.0) + 16) / 50 = 114 / 50 = 2.28 million. For the next ten weeks every ordinary 2.0 million share session reads 2.0 / 2.28 = 0.88 relative volume, so the stock appears to be losing interest when nothing has changed. If a name reported recently, expect its relative volume figures to run low. Check the raw numbers before concluding anything.

Reported volume is also not the entire market. A substantial share of US equity volume executes away from the lit exchanges, and while the consolidated tape captures the prints, the venue and timing detail available on a retail chart is limited.

The stock screener filters on relative volume alongside RSI and moving average position, which is the fastest way to find heavy participation that is also happening somewhere worth caring about. The candlestick scanner pairs with it for the reversal bars, and market breadth answers whether volume is arriving across the market or in a handful of names.

Read the momentum indicators guide for the price-derived half of the picture and the technical analysis pillar for how the two fit together, then try the chart patterns and indicators quiz.