Read executed order flow without confusing it with capital flow
Buy/sell volume describes the side that initiated recorded executions. Every completed trade has a buyer and a seller; the public feed classifies the execution by its aggressive side. A buy-labelled execution normally means a buyer took available sell liquidity. It does not mean that only a buyer participated or that new money entered the exchange.
In this guide
- Read executed order flow without confusing it with capital flow
- Start with one consistent interval
- Three views of the same volume split
- Why this is not net capital entering a coin
- Volume ratio versus execution-count ratio
- VWAP is an average, not an ownership map
- Read range and activity without inventing predictions
- A practical inspection sequence
- Questions readers often ask
Open the MEXC dashboard, select a pair, and read its timestamp and interval before interpreting the colours. The useful question is how the recorded executions were distributed, rather than whether a green number promises the next price move. This guide explains the arithmetic, the limits, and a repeatable way to inspect a session.
The exchange’s spot API documentation defines public trade fields, including trade side. Public executions do not reveal account identities, the trader’s portfolio, or the parent order that may have produced several fills.
Start with one consistent interval
Use the same pair, venue, time boundaries, currency, and data coverage for every comparison. A rolling 24-hour panel differs from midnight-to-now. A recent-trades list can be shorter than the aggregate’s interval. The latest message time and the end of an aggregate window can also differ.
If a connection is stale or an interval is incomplete, treat the displayed result as a partial observation. A page refresh does not prove that the upstream feed is current. Check the visible freshness information and compare an available recent execution with its exchange timestamp.
Do not combine a buy volume from one window with a sell volume from another. If two panels disagree, check their scope before interpreting the disagreement as market behaviour.
Three views of the same volume split
Let B be aggressive-buy quote volume and S aggressive-sell quote volume in the same interval:
| Measure | Formula | Meaning |
|---|---|---|
| Buy/sell volume ratio | B / S | Relative recorded quote volume on the two sides |
| Signed volume difference | B − S | Excess aggressive-buy or aggressive-sell quote volume |
| Buy volume share | 100 × B / (B + S) | Buy-labelled share of recorded quote turnover |
For a hypothetical interval with B = 500,000 USDT and S = 320,000 USDT, the ratio is 1.5625, the difference is +180,000 USDT, and the buy share is approximately 60.98%. These are three transformations of the same two inputs, not three independent confirmations.
For positive S, a ratio above 1 must have a positive difference. A ratio of 1.2 accompanied by a negative difference is inconsistent if both use exactly the same inputs and interval. It can indicate mismatched scopes or an error; large individual sell trades cannot make that arithmetic possible.
If S is zero, the ratio has no finite denominator. If B + S is zero, the share is undefined. Missing values should remain unavailable rather than becoming an artificial neutral market. A “Strong Buy” label is an application’s descriptive threshold, not an independently validated probability of a profitable trade.
Why this is not net capital entering a coin
An execution transfers assets between counterparties. The signed difference measures aggressive execution imbalance, not deposits, withdrawals, wallet balances, or investor accumulation. A positive result does not establish that holders collectively increased their positions: one trader’s purchased units were sold by another.
Quote-volume and base-quantity differences can tell different stories because prices vary within the interval. Large buys at higher prices can outweigh a larger base quantity sold at lower prices. Record both units rather than calling every USDT figure a US-dollar cash flow. USDT is the quote asset; conversion to USD needs its own assumption or price.
Volume ratio versus execution-count ratio
The count ratio divides buy-labelled execution count by sell-labelled execution count. It gives each fill equal weight. The volume ratio weights each fill by its quote value.
A hypothetical session can contain 100 buys of 10 USDT each and two sells of 1,000 USDT each. Its count ratio is 50, but its volume ratio is 0.5. This is mathematically consistent because the measures use different inputs. The result establishes different average execution sizes, not that 100 retail people bought from two institutions.
One order can create multiple fills. Algorithms can split large orders, and individuals can execute large trades. Without account or parent-order information, “small,” “medium,” and “large” are execution-size buckets. They are not verified classifications of retail investors, institutions, or whales. Check whether a distribution shows share of count or share of volume before comparing percentages.
VWAP is an average, not an ownership map
For base quantity q and execution price p, VWAP is Σ(p × q) / Σq. For a consistent quote/base trade dataset, this is total quote turnover divided by total base quantity.
Current price above VWAP means it is above that interval’s quantity-weighted execution average. It does not prove that most volume traded below the current price, that institutions regard the average as fair value, or that the price must return to it. A weighted mean is not a percentile, a median, or a volume-at-price distribution.
VWAP depends on where the interval begins and ends. Compare the same session definition and venue. A short recent window and a full day can have different averages without either being wrong.
Read range and activity without inventing predictions
The dashboard’s high-low range percentage describes the observed price span according to its displayed formula. It is not automatically a standard deviation of returns or an estimate of the next hour’s risk. A wide span can result from a short spike, a sustained trend, or thin liquidity.
Hourly volume identifies when executions occurred. An unfinished current hour should not be compared with a complete prior hour as though both had equal duration. Shorter trade intervals describe more frequent executions; they do not independently establish rising price momentum or an approaching breakout.
The largest recorded execution is useful context for a session. Its relevance depends on venue liquidity and the interval, not a universal 100,000-USDT threshold. A large buy does not prove an institution is accumulating, and a large sell does not identify an owner leaving the asset.
A practical inspection sequence
- Select the venue and exact pair; check quote currency, timestamp, and coverage.
- Compare buy and sell quote volume for one interval. Recalculate the ratio, difference, and share if a value looks inconsistent.
- Compare execution counts and average sizes. Check which denominator the size chart uses.
- Compare current price with the same interval’s VWAP and high-low range.
- Inspect complete hourly bins and recent large executions for concentration in time.
- If assessing execution conditions, examine current spread and order-book depth separately. Historical turnover cannot tell you the price available for your full order now.
This sequence gives you a reproducible reading of the data. It is not a backtested entry rule. The earlier version’s 73% continuation claim, fixed profitable ranges, and specific unexplained trading examples are not supported by a reproducible study and are not used here.
Questions readers often ask
Can all three volume measures agree and still fail to predict price?
Yes. They encode the same historical split. Price also depends on available liquidity, changing orders, other venues, and subsequent activity. Agreement is not three independent predictive tests.
Does positive signed volume mean capital entered the exchange?
No. Deposits, withdrawals, and account balances are different data. Public execution imbalance does not measure them.
Can the count ratio be positive while the volume difference is negative?
Yes. Many small buys can coexist with fewer larger sells. An above-one volume ratio and negative same-window volume difference cannot.
Can the dashboard identify a wash trader or whale account?
Public fills can reveal unusual patterns, but they cannot establish common account ownership or trading intent. Keep an execution-size observation separate from an actor attribution.
How should a predictive claim be evaluated?
It needs a defined dataset, window, signal rule, outcome, sample size, costs, and a separate evaluation period. Selected screenshots or a threshold table are insufficient evidence of a trading edge.
