Volume confirmation and divergence
Whether the bars moving WITH a trend carry more volume than the ones moving against it — a question that only exists once a trend does.
- Assumes:
- The basics of this family
- Markets:
- Crypto · Equities
- Timeframes:
- 1h · 4h · 1d
What it means
A trend is usually described as confirmed by volume when advances are busier than the pullbacks within it, and as diverging when the opposite is true.
The claim is about a relationship between two groups of bars, so it needs both groups to exist and to be large enough to average. It is not a statement about any single bar.
How we detect it
The engine splits the last forty bars into those that moved with the current trend and those that moved against it, averages the volume of each group, and takes the ratio. A ratio at least fifteen per cent above one is confirming; at least fifteen per cent below is diverging; the gap between them is inconclusive.
It runs only when the market state is trending up or down. A range returns inconclusive, because volume confirms a range is not a claim with content.
A feed that publishes zero volume on every bar returns insufficient-data, and that distinction is deliberate: several index feeds carry no volume at all, and they must not look like volume that was examined and found unremarkable. Fewer than three bars in either group returns insufficient-data too.
When the read carries information
- The market state is genuinely trending, which is a stricter test than the labels leaning one way.
- The instrument has real, comparable volume across the window.
- Both groups have enough bars for an average to be an average.
When it misleads
- One enormous bar in either group moves the mean a long way; the read uses a mean rather than a median, and a single session can therefore decide it.
- In markets where volume concentrates in auctions or funding windows, group membership can reflect the clock rather than conviction.
- The window is the same forty bars the state read uses, so a trend that only recently began is being measured partly over the period before it.
How it reaches you
Our surfaces report the confirmation state together with the ratio behind it, and report inconclusive and insufficient-data as different answers, because they are.
Confirmation is not a strength rating and not a reason to act. It is a description of how volume was distributed across a window that has already happened.
Worked example
No worked example is shown here. Display rights over the market data a real example would use have not been established, and an unanswered rights question is treated as a refusal.
Outstanding before an example can be shown
- Display rights over the underlying data have not been established.
- No data source has been selected for it.
- No freshness terms cover the data it would use.
- The example has not been made: no fixture, no figure, no asset review.
- Its caption and image description do not exist in both published languages.
This page stays text only until every item above is settled. Nothing is loading, and there is nothing here to open.