Understanding Volume

9.1 Volume, Turnover Value, Trade Count, and Open Interest

Volume records the quantity traded during a statistical period. Depending on the market, units may be shares, lots, contracts, or units of the underlying asset. Check the platform’s field definition before analysis. Every trade has a buyer and seller, and an exchange usually counts the matched transaction only once. Volume itself therefore has no “net buying” or “net selling” direction. Classifying individual trades as aggressive buys or sells requires another layer of data and algorithms.

Turnover value sums price multiplied by quantity across trades and is closer to notional transaction value. Different price levels, contract multipliers, and currencies prevent direct comparison without conversion. Trade count counts completed transaction records: a large and small trade may each count once. Order splitting, matching rules, and data aggregation alter its relationship with the actual number of decision-makers. More trades therefore establish only a denser sequence of execution records.

Open Interest (OI), used for derivatives, is the total number of contracts outstanding at a given time. Each outstanding contract is counted once, not separately for its long and short sides. A new long matched with a new short normally increases OI; both existing sides closing normally decreases it; one side opening while the other closes may leave it unchanged. Volume measures trading during a period, whereas OI measures outstanding contracts at a point in time. Neither substitutes for the other, and changes alone cannot identify participants.

Their measurement objects make these four fields noninterchangeable. Volume accumulates quantity, turnover value also multiplies by price, trade count counts matching records, and OI counts surviving derivative contracts at a timestamp. Higher prices increase turnover value for the same quantity. Splitting orders raises trade count without necessarily raising quantity. Extended active trading can generate large volume while ending OI hardly changes.

All three opening-and-closing combinations above add to volume. Aggregated OI shows only the net change in outstanding contracts. Volume and OI alone cannot classify every trade as opening longs, opening shorts, closing longs, or closing shorts.

Confirm counting conventions as well. Exchanges normally count a matched long-short contract as one contract of volume and one outstanding contract, rather than adding both sides twice. Providers may display lots, contracts, or underlying units. If turnover value already includes the contract multiplier, do not multiply again. Comparisons over time are meaningful only with consistent field definitions and conversions.

9.2 Actual Trade Records, Tick Volume, and Data Coverage

Actual volume comes from matched trade records, but “actual” does not mean it covers the entire economic market. Central-exchange data can cover that venue, while stocks across venues, cryptoassets, and over-the-counter markets may be represented by only one exchange, broker, or aggregator. Two datasets for the same instrument with different coverage can have different absolute scales and local spikes. Read the data description first.

Tick volume usually counts quote or last-price updates within a bar, with the exact field defined by the platform. It can indirectly indicate activity but is not traded quantity: updates may correspond to different trade sizes, and quotes can change without visible execution. In foreign exchange or CFDs lacking centralized trade data, it may support relative comparisons within the same source and settings, but it cannot represent actual volume across the whole market.

MetaTrader 5’s Volumes indicator can display tick volume and, when exchange transaction data are available, real volume. Whether a particular instrument provides actual volume depends on the server and source. Check field names, historical coverage, session timezone, and missing segments. After changing broker or server, similar chart prices do not guarantee that tick series can be directly joined.

Specify coverage by venue and session. Exchange volume may fully cover that venue while excluding off-exchange or other-venue trading. Consolidated stock data may include several exchanges, but premarket and after-hours trades, delayed dark-pool reports, and corrections depend on the provider. Aggregated crypto volume depends on included venues and quote currencies. Changing sources can alter absolute volume and spikes, so old baselines cannot simply be reused.

Check whether a platform’s tick volume actually counts quote updates, last-trade price changes, or incoming ticks. Added and canceled orders can change Bid and Ask without execution, while multiple trades may be combined into one update. Tick counts therefore have no fixed conversion to quantity traded. They can proxy activity within the same broker, server, period, and settings. Changes in server history, quote precision, or filtering require a new baseline.

Complete price data do not establish complete volume data. A platform may have OHLC without volume or omit bars during a halt. Missing data must remain missing, not be filled with zero. Only explicitly reported absence of trading is zero volume. With partial market coverage, conclusions must concern relative activity in that sample, not actual total market volume.

9.3 Absolute and Relative Volume

Absolute volume is the raw value reported by a source, stating how many units traded during a period. Price-volume analysis more often uses relative volume: whether current activity is above, below, or near normal under consistent coverage, units, session, and aggregation. Relative comparison does not remove flaws in the underlying data.

Relative volume is RV=V_current/V_baseline. Fix the baseline before observation, for example the median historical cumulative volume for the same session period or the average volume of comparable completed swings. RV > 1 means only above that baseline and RV < 1 below it; neither automatically constitutes a volume signal. A zero baseline makes RV undefined. Missing current or baseline volume also makes it missing, not zero. Changing the baseline window creates a different statistic.

Putting current volume in the numerator and baseline volume in the denominator answers, “How many times normal activity is the current activity?” Division cancels units, allowing comparison of relative activity in the same field across dates. It does not automatically make different units or coverage comparable. If one source consolidates trades and another covers one venue, identical RV means only the same multiple of each source’s own baseline, not equal market activity.

Fix the instrument, source, timeframe, and session first. Daily volume cannot be directly compared with hourly volume, nor a completed day with the current unfinished day’s cumulative volume. Intraday, sum volume from session start to the current time, sum the identical interval on each comparable historical day, then form the baseline using the preselected mean or median and calculate RV. Note missing historical segments, early closes, or an unfinished current bar. Comparing half-day volume with completed historical days mistakes ordinary intraday progress for contraction.

Changes in company size, migrating contract liquidity, or trading rules gradually shift long-term volume baselines. Current volume may be high relative to recent history without being unusual over a longer period; state these comparisons separately. Relative volume shows activity, not supply-and-demand direction. Also assess which direction price actually moves, its structural location, and whether subsequent action continues.

RV’s numerator and denominator must cover identical intervals. Compare the current minute bar with historical bars from that minute, or current cumulative volume from the open with historical cumulative volume between the same start and end times. Do not divide half a day’s activity by historical full-day totals. An unfinished bar can be compared only with the same elapsed portion of historical bars or explicitly marked in progress; it cannot be declared low-volume against completed bars.

Different baselines answer different questions. Neighboring bars capture immediate change but are sensitive to session rhythms. The same historical session period measures current intraday activity. Complete same-direction swings support effort-versus-result analysis but also require duration and net-displacement comparisons. The mean is affected by climactic volume; the median is more robust, while percentile rank locates the observation in the historical distribution. Fix the method before observing the current result rather than switching windows to get the desired label.

Retain zero and missing conditions. If V_baseline = 0, RV is undefined. If current volume or the baseline sample is missing, RV is also missing. Truly zero current volume with a positive baseline can yield zero RV, but verify whether trading should have occurred. When company growth, contract migration, or trading-rule changes shift the baseline, restrict comparisons to a comparable regime or use separate segments rather than combine incomparable history into one number.

9.4 Baselines: Previous Averages, Adjacent Swings, and Matching Session Periods

Average historical volume describes the center of recent activity but is affected by window length and extreme events. The arithmetic mean reflects overall activity but may be pulled up by a few exceptional observations. The median is the middle ordered value and more resistant to spikes. A large difference between them indicates skewness. Retain anomalous events alongside the comparison rather than hide them in one average.

Adjacent-bar comparisons identify immediate changes; adjacent same-direction swings better suit effort-versus-result analysis. For two rallies, record cumulative volume, duration, net progress, and intervening volatility. A larger total for the later swing may merely reflect longer duration. Activity per unit of time or displacement per unit of volume may supplement analysis, but these ratios also depend on volatility, liquidity, and price limits.

Intraday markets have stable session rhythms, with naturally greater activity near opens, closes, settlements, and announcements. Compare current cumulative volume after the open with historical cumulative volume at the same session point. Separate night and day sessions according to the actual market’s definitions. Shortened holiday sessions, daylight-saving changes, and temporary halts disrupt comparability and need annotation or exclusion.

Compare both total and per-unit-time activity across completed swings. Longer swings naturally accumulate greater totals. Higher volume per unit of time shows concentration, not necessarily farther directional movement. Also compare net displacement, incremental distance beyond the previous same-direction extreme, internal overlap, and subsequent retracement. A high total with little incremental progress may indicate opposing absorption; a low total with smooth progress may reflect sparse opposing liquidity.

Intraday cumulative baselines must handle early closes, daylight-saving time, and interruptions. A historical sample missing a segment before the current clock time must not be counted as a low-volume day. Announcement and ordinary days may be grouped separately so a few events do not raise every period’s mean. Actual market rules determine whether night and day sessions share a trading day. Changing bar timeframe changes volume aggregation and adjacent comparisons, requiring baseline recalculation.

Express the result specifically as “higher or lower than which comparison.” A current bar can exceed its predecessor while remaining below the historical median for that session period. The first fact shows an immediate increase; the second shows activity still below normal. Select the relevant comparison for the question rather than merge them into an absolute conclusion of “high volume.”

9.5 Recording Expanding, Contracting, and Abnormal Volume

Expanding volume means activity above the selected baseline; contracting volume means below it. Include the comparison, such as “cumulative retest volume is below the preceding breakout segment with similar duration,” rather than label the volume bar alone. One bar can expand relative to its neighbors while remaining below its period’s long-term norm. Both observations can coexist.

Abnormal volume is a substantial departure from normal that requires investigating its cause. Means, medians, percentile ranks, and robust deviations may help identify it, but thresholds depend on the market and purpose and should not be fixed across instruments. First check corporate actions, index changes, concentrated auctions, expirations, announcements, and duplicate data before discussing supply and demand. Retain and explain genuine market events separately. Corrections to confirmed bad data must remain traceable.

Zero volume and missing values differ. Zero means the source explicitly reports no trades during the period. Missing means records were not obtained, the bar was not generated, or the field is unavailable. Filling missing values with zero invents “nobody traded,” while deleting true zeros conceals inactivity. Platforms differ on whether they generate bars without trades; check their rules before assessing continuity.

After finding abnormal volume, first check duplicate records, changed units, and concentrated inclusion of auction or block trades, then corporate actions, index adjustments, announcements, and expirations. Real event volume cannot be deleted merely for departing from the mean. Compare it separately from ordinary periods while observing the actual price movement that day. If bad data are corrected, retain the revision source and timestamp so backtests do not use corrections unavailable at the time.

Also consider whether the current aggregation is complete. An unfinished daily bar contains only the elapsed part of the day and cannot be directly compared with historical full-day totals; use cumulative volume at matching intraday progress instead. On equal-volume charts, a completed bar’s total is preset by construction, so total volume differences between completed bars cannot identify expansion. Compare formation time and price movement for the same volume, then check market activity on equal-time charts.

If a zero-volume bar has changing quotes, the chart may use quote fields. If OHLC are forward-filled, they are not new trades either. Preserve missing bars, zero-volume bars, and closed-session gaps separately. None should enter moving averages unchecked or be interpreted as exhausted supply.

9.6 Combining Volume with Price, Location, and Subsequent Behavior

Volume describes activity, while range, body, close, and displacement show how price changed. Active trading with wide progress, a close at the directional end, and subsequent retention shows greater trading accompanied by proportionate movement. Active trading with limited incremental displacement, increasing overlap, or retreating closes shows meaningful opposition. This may be absorption, two-way turnover, or temporary congestion; observe the subsequent departure direction.

Low activity also has several explanations. A shallow post-breakout retest with reduced volume that cannot return to the old range, followed by resumption, supports limited opposing pressure. A low-volume decline making continued new lows with low closes may reflect withdrawn demand, allowing small trading quantities to move price. “No Supply” and “No Demand” have analytical value only with a clear test object, relative baseline, and subsequent response.

Structural location determines volume’s meaning. A high-volume decline at support unable to make further lows may indicate absorbed selling. A high-volume advance at resistance unable to make further highs may indicate supply restricting buying. The same combination in the range middle generally has weaker directional meaning. A departure crossing local structure may strengthen one explanation; total volume alone cannot establish that “institutions are buying.”

Joint analysis first establishes the historical periods used for the volume comparison, then separately examines price changes. For one bar, inspect full range, body, and closing position; for a swing, net displacement, incremental same-direction extreme distance, and the amount retraced afterward. High volume and wide range with a small body and retreating close is not strong progress simply because range is large. Narrow, low-volume bars closing successively in one direction may still accumulate clear displacement. Volume is effort; retained gains or losses determine how the move is interpreted.

Alternatives for high volume with little result include opposing absorption, two-way turnover, announcement repricing, and auction trades. At support or resistance with a subsequent opposing departure, absorption is more plausible. In the range middle, it shows only concentrated activity. Low volume with a large result may reflect withdrawn opposing quotes or a thin book and differs from a high-volume SOS. A breakout that holds is still a valid result; one fully reclaimed on meeting opposition lacks persistence.

When both price and volume contract, consider the direction. A countertrend reaction with shrinking activity and distance supports limited opposing pressure. Successively reduced volume, net displacement, and new extremes in the trend-direction move itself may indicate exhaustion. Contraction on both sides with more overlap often suggests balance. The next side to break and hold converts a candidate explanation into directional conditions.

Trades may trigger on predefined closing, intrabar transaction, trade-by-trade, or order-book conditions; not every price-volume judgment must await the close. But volume and closing position on an unfinished bar are provisional at that time. Execute protection under order rules when its conditions occur, without waiting for a fuller explanation of price and volume.

9.7 Data Anomalies, Contract Rolls, and Other Checks

Continuous futures join different expiration months, with volume and OI migrating from old to new contracts during rolls. The continuous series alone may mistake migration for an abrupt change in market interest. Specify the active-contract selection rule, roll date, and whether prices are back-adjusted, and inspect individual months around the roll. Price adjustment can improve continuity but does not make two contracts’ volume automatically additive or eliminate calendar spreads.

Stock splits and dividend adjustments alter historical prices. Whether volume is adjusted inversely depends on the provider. Opening and closing auctions, block trades, after-hours trades, and corrections may create volume spikes, with inclusion differing by platform. Cryptoassets also require checks of exchange coverage, quote currency, and wash-trading risk. Over-the-counter and broker data must be described as visible samples only.

For unusual volume, verify source, timestamps, sessions, units, aggregation timeframe, adjustments, known interruptions, and outlier handling. Any change requires a new relative-volume baseline. When data cannot be verified, price structure remains observable, but volume conclusions should receive less weight or be set aside. Smoothing, filling zeros, or joining sources does not remove uncertainty.

At a futures roll, inspect the old and new month’s volume, OI, and executable depth separately. Falling old-month volume with rising new-month volume often reflects migration, not a sudden reversal of overall interest. Volume in the continuous series also depends on the provider’s active-contract rule. Back-adjusting prices handles seam continuity but neither eliminates the calendar spread nor permits unconditional addition of simultaneous volume across contracts. Structure may be observed on the continuous chart, but verify key boundaries and execution against the actual traded month.

For adjusted stocks, check price and volume separately. A split changes per-share price and share-count scales; some providers inversely adjust historical volume and others preserve it. Dividend adjustments usually affect prices only. If prices are adjusted but volume handling is unclear, price-volume efficiency across the split cannot be compared directly. Preserve the status of missing values, corrected trades, and delayed reports rather than smoothing them into false continuity.

After data repair, recalculate historical relative-volume baselines, but real-time judgments may use only the version then available. If an anomaly is confirmed after trading, retain both original decision data and revised data during review to distinguish analytical error from data problems. Uncertain volume can be omitted. Price structure remains observable, with conclusions correspondingly less strong.