A large reported trading volume does not by itself establish suitable liquidity for a particular order. Identify the source and scope, then inspect spread, depth, timing and distribution across markets. Keep unusual observations separate from accusations and a data score separate from a guarantee about funds.
The example shows why recorded activity and a current book snapshot cannot substitute for each other.
BTKWIN.COM · EDUCATIONAL EXAMPLEDefine the number before judging its quality
Are you reading volume for one pair, an asset across venues, an exchange aggregate or derivatives? Is the period rolling twenty-four hours or a calendar day? Two columns labeled volume can cover different assets, markets and times. Preserve those definitions before comparing their values.
The CoinMarketCap volume methodology explains pair inputs, reference-currency conversion and different aggregation scopes. A processed exchange-reported field is not automatically independent verification of every underlying transaction.
Save the original field name, URL, unit and observation time. A later methodology change can explain why an old worksheet differs from a fresh reading. Without source and timestamp, a copied number becomes difficult to review, especially when paired with a book snapshot taken at another moment.
Use a large-volume example without assuming misconduct
Assume a hypothetical market records daily volume of one hundred million but currently displays only two thousand of nearby liquidity on the side relevant to your operation. Another records twenty million with ten thousand of nearby depth at the comparison time. Daily volume alone cannot establish the cost of a current five-thousand order.
Legitimate timing, coverage or market changes can explain the difference. The observation does not prove that the first market is fake or that the second will always execute better. It shows that a historical activity total and a current directional quote measure different things.
A purchase consumes asks; a sale consumes bids. Adding both sides and comparing the combined figure with the order exaggerates what that direction can use. Start with the opposite side of the book and the intended quantity, then examine the actual price levels.
Check one: spread and depth for the operation’s size
Record the best bid and ask, the quantities available and additional levels in a consistently defined range. A narrow spread can coexist with little quantity. A displayed large quantity can also disappear. State the reference and units if measuring a band such as one percent around a price.
Suppose asks offer one unit at 100 and five at 101. Buying three from this unchanged hypothetical snapshot costs 100 + 202 = 302 and averages approximately 100.6667 before fees. Millions of earlier turnover do not alter the one-unit capacity at the first level.
For stability analysis, compare several dated snapshots with a stated sampling interval. A thin snapshot during a data outage does not describe an entire market, while one deep snapshot does not establish persistence. Describe a sample of displayed liquidity rather than guaranteed future execution.
| Observation | What it can describe | What it does not guarantee |
|---|---|---|
| Spread | Gap between current best prices | Enough size at those prices |
| Directional depth | Displayed quantities in a defined range | Persistence until processing |
| Period volume | Recorded activity under its scope | Liquidity reserved for a new order |
Check two: trade records, frequency and units
Where public trade data is available, determine what each row represents. An interface can aggregate executions or divide a larger instruction into multiple fills. Row count is not participant count, and repeated activity can belong to the same participant.
If a consistent dataset records one million of value over one thousand distinct trades, average trade value is one thousand. That arithmetic check can expose a mismatched period or unit, but it does not reveal the size distribution or number of people. Note any aggregation or rounding.
Recurring similar sizes can justify further investigation without proving economically artificial activity. Automated execution can also split orders into similar portions, and display methods can create patterns. State the narrow observation supported by the sample rather than inventing a percentage of illegitimate volume.
Check three: concentration across accessible markets
An asset aggregate covering ten markets does not imply equal depth or activity in each. Calculate contributions within the provider’s scope and identify the pairs you can actually access. Most activity may occur on another venue or in a pair requiring a different funding route.
For hypothetical total volume of fifty million, forty million in one market means an 80% contribution. The remaining ten million is spread elsewhere. This concentration is relevant context, but it does not itself determine the quality of the largest market.
Avoid building a new aggregate by adding overlapping asset totals. Review the aggregation unit and excluded markets. The volume-to-cap guide explains why differently scoped totals can produce incomparable ratios even when each underlying field is correctly calculated.
Check four: the current score and what it covers
A pair-level score and an exchange-level score describe different scopes. Do not transfer one pair’s color or number to every market on the venue. Methodologies also change, so an old explanation of the weights may not describe today’s calculation.
The CoinGecko Trust Score methodology describes multiple components and a relative comparison among evaluated exchanges. Such a score summarizes a model; it is not a certificate that every reported trade is valid or every balance can be withdrawn without restriction.
The Basilisk update explanation documents changes including moving away from earlier web-traffic normalization toward direct volume and depth assessment. Keep the methodology date with the observation. A rank change can reflect relative standing or other components without proving a new incident occurred.
Check freshness and separate custody questions
Fields shown together can update at different times. A rolling volume can remain visible while a book feed stalls. Compare last-update timestamps before interpreting a mismatch. If a likely gap exists, inspect the direct source or service status and preserve the missing-data interval.
Unavailable data limits verification without automatically proving bad activity. Write that a particular check cannot be completed instead of inventing a score. Also keep custody, account eligibility and withdrawal conditions separate from market-data analysis. The wallet helper and transfer calculator address other parts of the route.
A repeatable comparison records scope, source, time, reported versus adjusted status, directional depth, trade-record limitations, concentration and score version. Its useful result is a specific explanation of what the evidence supports and what remains unknown.
Continue reading
English content and examples reviewed: 27 October 2026. Read the Arabic edition.
