What tick volume measures
Tick volume counts price changes. A bar where the quote changed four hundred times shows four hundred; a bar where it changed forty shows forty. It does not know how many ounces changed hands and it does not know who initiated any of it.
The useful property is that price updates track activity. Busy markets tick more. Across a range of studies, including our own, tick volume correlates strongly with futures volume over the same period. For relative comparison, this bar against the last twenty, it is good enough, and relative comparison is all VSA asks.
The limit is direction. Futures volume can be split into trades at the bid and trades at the offer, which is what footprint charts and cumulative delta use; we cover those in footprint charts and CVD basics. Tick volume cannot be split. Absorption, the very thing VSA infers from a heavy volume narrow spread bar, is invisible in tick data as a fact. You infer it from spread and close, and you should hold that inference lightly.
What we measured
We ran the basic test on several years of gold M15 and H1 data: does high or low relative volume, on its own, predict the direction of the next bars? It does not, on any horizon we tried. What it predicts well is volatility: high volume bars are followed by wide bars, low volume bars by narrow ones. The write up is in participation is not direction and volume divergence, calibrated.
This does not kill VSA on gold. It kills the lazy version of it, the one that reads a volume spike and buys. The full method reads volume against spread and close inside a background, and that is a conditional reading that the simple test does not touch. But you should know that the raw material carries no directional signal by itself, because it stops you from trusting a single bar.
The session baseline
Gold volume has three baselines a day. Asia is quiet, London is busy, New York is busiest until the close, then it dies. A bar that is the biggest of the Asian session would be an ordinary New York bar.
Bar 5 is high volume for Asia and would be read as an event by a trader comparing it with its neighbours. It is a third the size of an average New York bar. Bar 14 is ordinary for New York. The rule: compare a bar with the last ten to twenty bars of the same session. Never read an Asian bar against a London average or you will see climaxes that are not there. The session articles on this site, London, New York and Asia, cover how price behaves in each; this is the volume side of the same story.
News bars
A scheduled release, the jobs report, an inflation print, a central bank decision, produces one or two bars of enormous volume and enormous spread. They look like climaxes because every participant updated at once. They are not VSA events; they are the market repricing information, and the bar tells you nothing about who absorbed what.
Skip the release bar and the whipsaw after it. Start reading again from the third bar. If the third and fourth bars are narrow on falling volume and hold the new level, the reprice is accepted. If they are wide on rising volume against it, it is being fought. How to trade the reaction rather than the print is covered in trading gold around news.
Rollover and the weekend
Two more quirks. Around 22:00 UTC the daily rollover widens spreads and thins liquidity, so a single bar can print a wide spread on trivial volume; ignore it. And at the Friday close the last hour often prints a fake climax as positions square. The Monday open is a better read than either.
Practical settings for gold
- Read background on H1, events on M15. M5 only inside London and New York.
- Compare volume within the session, against the last ten to twenty bars.
- Skip the two bars around any high impact release.
- Prefer events that coincide with a level: a zone, a swept high or low, a broken structure point.
- Treat every reading as a hypothesis until the confirmation bar closes.
Course navigation. Previous: Part 8, Accumulation and distribution. Course index: VSA course for gold. Next: Part 10, The trade plan.
FAQ
Should I use futures volume instead?
If you have it, yes, as a cross check. Gold futures volume is real contract volume and it can be split by side. Most retail traders do not have it live, and the good news is that tick volume tracks it closely enough for relative readings. What it will never give you is the split by aggressor.
Does tick volume differ between brokers?
Yes, sometimes by a wide margin, because each broker counts its own price feed. The shape of the histogram is usually similar; the absolute numbers are not. This is one more reason to read volume as a rank against neighbouring bars rather than as a number.
Why is Asian session volume so low?
Because the participants who move gold in size are in London and New York. Asia is inventory building and range trading. Events do happen there, especially tests of a New York climax, but the baseline is a fraction of the day session and must be judged on its own scale.
Is VSA useless on gold if volume has no directional edge?
No, but it is narrower than the books suggest. The edge, such as it is, lives in the conditional readings: a test at a level on low volume after a climax, confirmed. Volume alone is participation, and participation on its own predicts only volatility. Use the method for what it can do.
ⓘ See these ideas on real price: open the free XAUUSD live chart.