← THE CHARTSPOD LEDGEROPEN TERMINAL →
ORDER FLOW · NORMALISATION

How to Normalize CVD Before Comparing Two Crypto Exchanges

A 500 BTC CVD move on a giant venue and on a small one are not the same event. Normalise before comparing exchanges.

By ChartsPod Editorial Team · Reviewed by ChartsPod Research
Published · Updated · 5 min read

Short answer

To compare CVD across exchanges, normalise it: divide cumulative delta by total volume over the same window (delta as a percentage of volume), or standardise each venue's delta by its own typical range. Use the same units (base or quote), the same reset time and trade-based data where possible. Normalised CVD shows which venue's participants were more one-sided, independent of venue size.

Why raw CVD misleads

Illustrative 1h CVD on two venues (BTC)Illustrative 1h CVD on two venues (BTC): Venue A raw 500 ; Venue B raw 120 ; Venue A % of volume 5 ; Venue B % of volume 12 .ILLUSTRATIVE 1H CVD ON TWO VENUES (BTC)Venue A raw500Venue B raw120Venue A % of volume5Venue B % of volume12A: 500 of 10,000 BTC volume = 5%. B: 120 of 1,000 BTC = 12%. Illustrative.
Venue B was more one-sided despite smaller raw delta. Original ChartsPod editorial graphic.

Normalisation methods

MethodFormulaUse
Delta ratioCVD ÷ total volume (same window)One-sidedness
Z-score(CVD − mean) ÷ std, per venueUnusualness vs venue's history
Share of venue volumeVenue delta ÷ all venues' volumeContribution to total flow
Before comparing venuesBefore comparing venues: Same window and reset; Same units (base or quote); Same market type (spot vs perp); Trade-based delta on both; Normalise before ranking.BEFORE COMPARING VENUESSame window and resetSame units (base or quote)Same market type (spot vs perp)Trade-based delta on bothNormalise before ranking
Raw numbers rank venues by size, not by behaviour. Original ChartsPod editorial graphic.

Worked z-score example

Illustrative: venue A's hourly delta over the last 30 days has a mean of +20 BTC and a standard deviation of 150 BTC; this hour it is +500, a z-score of 3.2. Venue B's mean is +5 and standard deviation 40; this hour it is +120, a z-score of 2.9. Both are unusual for their venue, roughly equally so, even though A's raw delta is four times larger.

Z-scores need enough history to be stable, and they assume today resembles the past window. Recompute periodically and avoid using them on venues or contracts with short histories.

Use

A venue with the highest normalised delta during a breakout may be where informed or aggressive participants are. Combine with funding and OI per venue (see How to Switch Binance, Bybit, OKX and Hyperliquid Context in ChartsPod). For spot vs perp, see Spot CVD vs Perpetual CVD: Interpret Disagreement Before a Breakout.

Switch venues on the same contract and compare cumulative delta.

Open CVD in ChartsPod

Normalise CVD by volume or by each venue's history before comparing exchanges.

Review How to Normalize CVD Before Comparing Two Crypto Exchanges with bar replay before using it live.

Open ChartsPod replay

Questions readers ask

Why normalise CVD?

Venue size dominates raw values.

What is the simplest method?

Divide CVD by total volume in the same window.

Can I compare spot and perp this way?

Yes, with matching units and windows, though participants differ.

Sources and methodology

Order-flow concepts are explained from first principles and from TradingView's documented footprint and CVD methods; ChartsPod features were checked on its public pages on 17 October 2026. All numbers in examples are illustrative, not recorded market data.

  1. TradingView: Cumulative Volume Delta — intrabar estimation, anchor period.
  2. ChartsPod: product overview, pricing and FAQ — free terminal, venues, modules, no order placement; checked 17 Oct 2026.

This article explains research tools and workflows. It is not financial advice, does not recommend buying or selling any asset and does not predict prices. Crypto derivatives carry a risk of rapid and total loss of margin.

CHARTSPOD EDITORIAL EDITION · RESEARCH AND EDUCATION ONLY · NOT FINANCIAL ADVICE.