Introduction

Axiom CVD Osc Lite is a lower-pane oscillator for TradingView that estimates cumulative volume delta across up to three independent timeframe slots, normalizes each one to a -100 to +100 range, and optionally blends t...

Written By Axiom Admin

Last updated About 1 month ago

Axiom CVD Osc Lite

What this tool is

Axiom CVD Osc Lite is a lower-pane oscillator for TradingView that estimates cumulative volume delta across up to three independent timeframe slots, normalizes each one to a -100 to +100 range, and optionally blends them into a single composite reading. It gives you a multi-timeframe view of estimated buying versus selling pressure β€” in one pane, from one indicator, with controls you can adjust for different instruments and market conditions.

The key word is estimates. This oscillator does not read order flow. It does not have access to bid/ask data, the order book, or tick-level trade records. It works from standard OHLCV candle data β€” the same open, high, low, close, and volume that every chart already has. What makes it different from a naive volume-split approach is the classification model underneath: instead of simply calling every green bar "buying" and every red bar "selling," it examines each sub-bar's candle structure β€” body direction, where the close sits within the range, and how the wicks behave β€” to produce a more considered estimate of which direction the volume was likely committed to.

That distinction matters because it sets the ceiling on what this tool can tell you. It can give you a structured, tunable read on estimated directional pressure across timeframes. It cannot tell you what actually happened at the order book. Everything else in this manual follows from that boundary.

Why it exists

Most CVD tools on TradingView have one of two problems. Either they depend on tick-level data that is not available for the instruments most traders actually chart, or they use a volume-splitting method so simple that the output barely tells you more than the candle color already did. The first kind does not load β€” you add it, the pane stays empty, and you spend ten minutes wondering whether your data feed is wrong before you realize the indicator just cannot run on your symbol. The second kind loads fine but gives you a CVD line that tracks the candle colors so closely you might as well not have it. Neither helps you answer the question CVD is supposed to answer: is volume committing to a direction, or is it just going along for the ride?

There is also the stacking problem. A trader who wants to see whether 5-minute, 15-minute, and 60-minute volume pressure agree or conflict has to add three separate CVD indicators, configure each one independently, and then visually compare them across three separate panes. That is a lot of screen space and cognitive overhead for a question that should have a compact answer β€” and it gets worse when the indicators do not share a consistent estimation method, so you are comparing readings from different models without knowing it.

We built this because we ran into both problems ourselves. The estimation model exists because we needed CVD that works on standard OHLCV data without pretending it is something it is not. The multi-timeframe stacking exists because comparing pressure across timeframes should not require three copies of the same indicator. And the tuning parameters β€” Pressure Sensitivity and Wick Weight β€” exist because we did not believe one hardcoded heuristic could serve every instrument and every market condition honestly. Different markets have different candle characteristics, and the person reading the chart knows their market better than we do.

Who this is for

You will get the most from this tool if you already think about volume pressure as part of your analysis and want a way to see it across timeframes without the hassle of managing multiple indicators. You are comfortable with the idea that an OHLCV-based estimate is a different thing from exchange-level order flow, and you would rather have a tool that is honest about that distinction than one that quietly pretends otherwise.

You do not need to understand the estimation model to use the oscillator. But you do need to engage with what it means that this is an estimate β€” what it can tell you, what it cannot, and where you need to verify its readings against your own judgment and other tools. This manual is built to help with that.

Who this is not for

If you are looking for a literal replacement for footprint charts, depth-of-market CVD, or tape-reading tools, this is not it. The oscillator estimates directional pressure from candle structure. It does not reconstruct order flow. The distinction is not academic β€” if you treat this like order-flow data, you will make inferences about market microstructure (who is aggressive, where the book is absorbing, when large players are stepping in) that the data simply cannot support. The oscillator can tell you that candle structure favors one side. It cannot tell you why.

If you want a tool that produces buy and sell signals, this is not that either. The oscillator shows you estimated pressure context. What you do with that context β€” whether to act, how to act, what else to check first β€” is entirely your decision.

If the phrase "estimated volume delta" sounds like a downgrade from what you expected, it is worth understanding why before you dismiss the tool or, worse, use it while believing it does something it does not. The For the Geeks page explains what the estimation model considers and why the approach is more capable than a simple volume split β€” without pretending it is something it is not.

The trust boundaries you need to know about

Two things determine how much you can rely on what this oscillator shows you, and both are worth understanding before you start forming opinions about the readings.

It is an estimate, not a measurement. The oscillator classifies sub-bar candle structure to infer directional volume commitment. That classification is more considered than a coin-flip split, but it is still working from price data, not from the actual order book. When the candle structure is clean β€” strong bodies, clear direction, decent volume β€” the estimate tracks well. When the structure is ambiguous β€” dojis, long wicks on both sides, low volume β€” the model is doing its best with less to work with, and the output carries less weight even though the number looks the same. A +60 from a clean trending session and a +60 from a choppy mess are not the same thing, even though they sit at the same point on the scale. The manual names these limits throughout, and the Limitations and Trust Boundaries page maps the full perimeter.

On Bar Close controls whether the chart's history is reliable. This setting is on by default, and the default exists for a reason. When it is on, each slot uses the last confirmed higher-timeframe bar's values β€” what you see on past bars is what you would have seen live. When you turn it off, the oscillator updates faster but historical values show the final state of each bar, which may differ from what was visible while the bar was forming. The danger is subtle: with On Bar Close off, the chart can look like the oscillator caught a reversal cleanly or signaled a move right at the start β€” but that clean read only exists because the historical bar had the benefit of knowing how it ended. In real time, the reading would have looked different mid-bar. If you plan to study historical oscillator behavior, scroll back to compare readings, or use this tool in any kind of replay or backtest context, leave On Bar Close on or read MTF and Repainting before you start drawing conclusions.

What to read next

If you want to...

Go to

Get it on the chart and see it working

Quick Start

Understand every setting and what it actually changes

Settings

Learn what the visual elements mean and how to read them

Visuals and Logic

Set up alerts and understand what they do not tell you

Alerts

See validated use patterns and common mistakes

Workflows

Understand the repaint switch in depth

MTF and Repainting

Know where the tool stops being reliable

Limitations and Trust Boundaries

Understand the estimation model under the hood

For the Geeks

Diagnose something that looks wrong

Troubleshooting