Trading Volume: Relative Volume and VWAP Explained

Volume is the one number on a chart that is not derived from price. It counts how many shares changed hands, and it answers a question candles cannot: how many people cared about this move? A 3% rise on a quiet day and a 3% rise on the busiest day of the quarter look identical as candles. They are not the same event.

Raw volume versus relative volume

Raw volume is awkward to read. Apple routinely trades 40–60 million shares a day, Netflix (before its 2025 split) around 3 million. Even within one stock, "normal" drifts over months. The fix is relative volume, often shortened to RVOL:

Relative volume = today's volume ÷ average daily volume over the previous 20 sessions

On October 18, 2024, the day after its third-quarter earnings, Netflix traded 15,974,100 shares. Its average over the previous 20 sessions was 2,989,540. So RVOL was 15,974,100 ÷ 2,989,540 = 5.34. Anything above about 2 stands out on a daily chart; 5 is the kind of day that shows up on market-wide scanners.

Three big-volume days, three different endings

High relative volume tells you something important happened and a lot of people acted on it. It does not tell you whether the move will last. Three 2024 examples, with the close 20 sessions later:

Netflix: volume that marked a new level

Netflix gapped from 687.65 to an open of 737.64 and kept climbing to close at 763.89, near the day's high of 766.28. Buyers were still in control at the bell. The stock never traded back to the prior close in the following weeks; its lowest low over the next month was 744.26 on October 23.

Uber: volume from a story about someone else

Uber jumped 10.81% on 57.3 million shares the day after Tesla's robotaxi presentation, which the market read as less threatening to ride-hailing than feared. There was no change in Uber's own business that day. The next session it closed lower, and 20 sessions later it was at 72.04, below where the spike started. Heavy volume on a reaction to someone else's news can unwind as fast as it arrived.

Apple: volume with nothing behind it

On September 20, 2024 Apple traded 318.7 million shares, 6.78 times normal, and closed down 0.29%. That Friday was a quarterly options expiration, when index funds also rebalance at the close. Enormous volume, almost no price change: mechanical trading, not conviction. Before reading anything into a spike, check the calendar.

VWAP on a daily chart: anchor it to an event

VWAP, the volume-weighted average price, is the average price paid per share, with each trade weighted by its size. Day traders use a version that resets every morning. On daily charts a more useful variant is the anchored VWAP: start the calculation on a day that mattered, such as an earnings gap, and let it run. The line then shows the average cost of everyone who has traded since that event.

With daily bars you approximate each day's price with the typical price, (high + low + close) ÷ 3, and weight it by that day's volume.

Notice how little the second day moved the line. The earnings day carried more than twice the volume, so it dominates the average. That is the point of anchoring on a heavy day: it pins the line to where the most shares actually changed hands.

For three weeks after the gap Netflix went sideways around its anchored VWAP. Between October 23 and November 4 it closed below the line on eight of nine sessions, but never by more than about 1.2%. The average post-earnings buyer was roughly break-even and nobody was forced to act. On November 6 and 7 the stock closed at 780.21 and 796.54, well above a VWAP near 760, and it did not come back. By December 20 price was 909.05 against a VWAP of 835.17.

This is how anchored VWAP is usually read. Price holding above it means the people who bought since the event are, on average, in profit and less likely to sell in a hurry. Price stuck around it means the market has not decided. Price falling well below it means most recent buyers are underwater, which often adds supply on rallies.

Volume patterns worth noticing

• Breakouts on high relative volume are more trustworthy than breakouts on thin volume. A move through resistance at 0.8× average deserves suspicion.
• Pullbacks on shrinking volume inside an uptrend are normal. Pullbacks on rising volume deserve attention.
• A very large volume bar after a long decline can mark capitulation, but only the following sessions confirm it.
• Gaps come with volume by nature. For how earnings gaps behave afterwards, see the guide to gap trading.

What volume will not tell you

Every share traded has a buyer and a seller, so volume cannot say who "won". Only the close does that. Volume also differs by venue and data vendor, and crypto volume varies between exchanges far more than stock volume does. Use it as a measure of how much the market cared, and let price tell you which way it leaned. The simulator shows volume under each historical chart; try guessing what a spike means before the next candle opens.