What does volume add to price?

Price tells you WHERE the crowd agreed. Volume — how many shares changed hands — tells you HOW MANY of the crowd showed up to agree. It's the difference between a whisper and a roar saying the same words.

The conviction meter

Two identical 3% up-days can be opposite events:

• On triple the usual volume — broad participation; many holders repriced their views; institutions likely moved size (Course 2: big moves have big feet).
• On a third of the usual volume — a drift; few participants; the price moved because nobody objected, not because many agreed. Thin agreement reverses easily.

Same candle, different weight of evidence. That's the whole idea, and it's why serious chart-readers keep volume bars glued under every price chart.

"Usual" is the key word

Raw volume numbers mean nothing alone — 5 million shares is a dead day for a mega-cap and a once-a-decade storm for a small-cap. What matters is relative volume: today against the stock's own recent average. Every platform (including this one) computes it; your eye should always ask "compared to its normal?"

Where volume spikes come from

Earnings days (Course 2's calendar), index changes (Course 2's rebalance-day flows — the volume is huge even when the price barely reacts), big news, option expirations, panics. A volume spike is the market's way of stamping a day as significant to many — the days worth investigating when you review a chart's history.

In the data

Volume sits in the same daily row as the prices. Here is Apple's row for 28 August 2020, the last session before its four-for-one split:

The volume reads 187,630,000 shares — the count the tape actually printed that day, multiplied by the split that landed the next session. A price history restates every old volume in today's share size, the way the adjusted close restates old prices, and that is exactly what makes a long "compared to its normal" average safe to take across a split.

The trap is the other direction, and it is easy to walk into: multiply that 187,630,000 by the raw close of 499.23 and you get about $94bn of turnover in one session, a figure no company has ever printed. Against the adjusted close the same row comes to roughly $23bn. Pair adjusted with adjusted and raw with raw, or the split you were careful about reappears in your answer anyway.

Try it now

1. Open a one-year chart of volume bars — Apple's is below; for your own anchor company, open it in the Terminal and change the symbol there. Find the three tallest bars of the year.

Open AAPL.US in the EODHD Terminal

2. Date them, then check the calendar: earnings? News? Broad-market storm days? Apple's reporting dates are below, and the market's own diary has a fixed entry worth knowing: the third Friday of March, June, September and December, when options expire and index funds rebalance together. A bar the morning after a report and a bar on one of those Fridays have boring explanations; a bar with neither is the one to look into.

3. Turn "usual" into a number. Below is Apple's latest session and the average volume of the twenty sessions ending on it. Divide the first by the second. Above 2 is a day the crowd showed up; near 1 is an ordinary Tuesday whatever the headline says.

Then do the same by eye for your three tall bars. The chart below draws that twenty-session average over the bars, so each bar's height against the line is its multiple, and you will find how far above normal a real event actually pushes it.

4. Now check that the average survives a split, rather than assuming either way. Apple's splits are listed below, and the second table is the session before its 2020 split beside the first session after it. Look for a step in the volume counts: there is none, because the column is adjusted like the prices. Then multiply volume by the raw close on each day and watch the turnover fall by a factor of about three, most of it the split's four — the phantom is in the pairing, and it lands in your arithmetic rather than in the data.