What a low-volume node actually claims

The theory says price passes quickly through prices the market rejected. That is a statement about the past. Turning it into a trade requires assumptions the theory does not supply — so it is worth being precise about where the reasoning stops.

Where the idea comes from

The vocabulary is older than order-flow software. Peter Steidlmayer and Kevin Koy set it out in Markets and Market Logic (1986), and James Dalton’s Mind Over Markets is where most traders meet it in practice.

The frame is auction-theoretic. A market spends time at prices that both sides accept and moves quickly through prices that one side rejects. Build a distribution of volume by price and you get a shape: fat where trade was accepted, thin where it was not. The thin parts are low-volume nodes.

What the theory actually says

Stated carefully, the claim is descriptive and backward-looking: a thin area is evidence that when price was last there, little business got done. Nothing more.

The usual next step — that price will therefore move quickly through it again, or reject from it again — is an additional assumption. It may be reasonable. It is not what the distribution tells you. The profile is a record of what happened; the forecast is something you are adding.

That distinction matters because both readings are common and they point opposite ways. One says a thin node is a place price accelerates through (continuation). The other says it is a place price gets rejected (reversal). Both are argued from the same picture.

The gap between the theory and a trade

A profile shape does not carry a direction, a timeframe, or an invalidation level. Those have to come from somewhere else, and where they come from is usually the part nobody writes down.

Three questions the theory leaves open:

Which reading applies here? Continuation and rejection cannot both be the default. Something outside the profile has to decide.
Over what window is the node defined? A node measured against one bar is a different object from a node measured over a session. Levels built against a bar denominator move when you change chart — measured on one rule and one dataset, only 69.8% of the levels found at 5m were still flagged at 15m.
When is it wrong? A level with no invalidation is not a level you can trade; it is a level you can rationalise around afterwards.

What this tool does with it

Two deliberate choices follow from the above.

The node is drawn, not interpreted. OrderVane marks where volume was thin and where it was thick, and stops there. It does not label a node bullish or bearish, and it does not fire an alert saying to act. The reading is yours because the profile does not contain it.

The measurement does not use your chart bars. A level defined by your bar size disagrees with itself the moment you change timeframe, which makes it useless as a shared reference. Volume shelves are measured against a fixed source and sit at the same price on every chart. Range profiles are computed over the box you draw, not over the bars inside it.

Both choices are conservative rather than clever. They exist because the theory supports the description and not the conclusion.

What it does not claim

No direction. Nothing here says price will bounce from a node or run through it. Both happen; the profile does not tell you which.
No verification of the underlying idea. Auction theory is a framework, not a measured result. This page describes what the framework asserts and what a tool can honestly draw from it — it is not evidence that trading low-volume nodes works.
No numbers attached to outcomes. The one measurement quoted above is about level stability across timeframes, not about what price does at a level.