A busy volume bar tells you that a feed recorded many events. It does not tell you how many euros changed hands, who bought, or whether the next candle will rise. This guide rebuilds EUR/USD candles from actual timestamped Bid/Ask quotes, separates two counting conventions and lets you inspect activity alongside price range and spread. The examples explain the measurement before the later VWAP and MFI materials use it.
1. What tick volume measures in Forex
Tick volume measures activity in a particular quote stream. MT4 exposes a bar's tick-volume field through Volume[] and iVolume. It is a count associated with that feed and bar, rather than a record of all currency units traded across the foreign-exchange market. A large count therefore needs a source, timeframe and counting convention before it has a useful meaning.
A quote update is not a signed transaction. Even an uptick does not identify the buyer, seller, order size or initiator. The green or red colour of a volume column may reflect the candle's direction or the platform's display rule. It does not transform that column into measured buying or selling volume.
2. The real EUR/USD data used here
The source is HistData's January 2024 Generic ASCII tick archive: 2 176 113 timestamped EUR/USD Bid/Ask records. Its timestamps use fixed EST, UTC−05:00, without daylight-saving changes. We add five hours. The available archive runs from 2024-01-01 22:00:12.108 UTC to 2024-02-01 04:59:57.721 UTC; the filename's month is not a UTC-month boundary.
The export's final Volume field is always 0. We do not treat those zeros as observed inactivity or replace them with invented trading volume. Instead, we count the original quote records directly. The full archive yields 31 703 observed M1 bars and 6 360 observed M5 bars; 78 M5 groups contain fewer than five observed minutes. Empty intervals receive no fabricated price candle.
3. Two honest counters for the same quote stream
N counts every source quote record in the bar. Q counts a record only when Bid or Ask differs from the preceding record in the archive; the first archive record counts once. The comparison continues across candle boundaries. An unchanged quote at the start of a new minute does not become a price change merely because a new candle starts.
There are no exact duplicate rows in this archive, but 10 602 records leave both Bid and Ask unchanged. They remain in N and are excluded from Q. Different timestamps with the same prices are not automatically duplicate records. Neither counter is claimed to reproduce every broker's native MT4 volume, because feeds and event handling can differ.
4. Close the candle before comparing it
A bar labelled 07:30 UTC contains records from 07:30:00 up to, but not including, 07:31:00 on M1. On M5, 07:30 covers the interval up to 07:35:00. The first observed Bid is Open, the highest is High, the lowest is Low and the last is Close. Ask is preserved for spread calculations; a Bid candle is not an Ask execution history.
A live bar is incomplete. Comparing its first ten seconds with a full previous minute mixes different observation lengths. A large count can arrive early and still change before the close. A close-based trading decision becomes actionable only when the interval is over and the required new information has arrived; it cannot be filled retrospectively at a convenient historical close.
5. Define relative activity using only earlier bars
The viewer divides the selected count by the median of the preceding 10, 20 or 40 complete chronological bars of the same timeframe. The selected bar is excluded from its own reference. If an interval is missing, a required group is incomplete or the median is zero, the ratio is unavailable. We do not silently skip gaps to assemble a more attractive reference.
A ratio of 2 means twice that declared historical median. It is not a probability, a volume of traded money or a universal entry threshold. The local reference can also change as the trading day develops. Comparing the same clock window across earlier sessions is a different study; this short rolling reference does not remove time-of-day effects.
R[t] = K[t] / median(K[t−L], …, K[t−1])
K ∈ {N, Q}; L ∈ {10, 20, 40}6. Follow a real activity spike at 07:30 UTC
On 2024-01-08 at 07:30 UTC, the M1 bar contains 163 quote records and 161 price-changing records. The preceding 20 M1 record counts have a median of 54, so N/median = 163/54 = 3.0185. Its Bid range is 2.8 pips. This is the largest N-to-median ratio inside the displayed 07:00–17:00 UTC window, chosen afterwards to explain the calculation.
That selection does not establish a tradable signal. A strategy would still have to define a direction, recognition time, entry, invalidation and exit, then examine all eligible events. The chart contains no invented entry or winning trade. A conspicuous activity bar alone supplies none of those missing rules.
7. Many records can fit inside a narrow range
At 08:16 UTC on the same day, 134 records fit into a 0.9-pip Bid range. This bar was selected as the narrowest range among the top quarter of record counts in the displayed window. Its preceding-20 median is 147.5, so relative activity is only 0.9085. A high absolute count and a high local ratio are different observations.
Repeated updates can occur near the same prices, while fewer updates can accompany a larger displacement. Tick activity, High−Low range and Close−Open movement measure different features. Inspect the actual candles rather than calling every busy, narrow bar accumulation, absorption or proof of a large institution's intention.
8. Aggregate M1 into M5 without changing the definition
For a complete aligned M5 interval, N and Q equal the sums of their five M1 counts. Open comes from the first minute, Close from the last, and High and Low are the extreme Bid prices across the group. Q remains additive because its preceding-record comparison was calculated once on the continuous tick stream, not restarted for each minute.
The M5 baseline uses earlier M5 bars. It is not the sum or average of five M1 activity ratios. Likewise, the median spread over all quote records is not generally the mean of the five minute medians. The displayed spread median gives each recorded quote one observation; it is not time-weighted.
9. Read observed spread separately from activity
For EUR/USD here, one pip is 0.0001. Each recorded spread is Ask−Bid, converted into pips. The viewer shows its minimum, record-weighted median and maximum within the selected candle. These are observed quote differences from this archive, not a broker's advertised minimum or a promise that your account could trade at those prices.
At 07:00 UTC, the recorded spread ranges from 0.1 to 0.7 pip, with a median of 0.3 pip. The maximum is the largest inside the displayed window, not a universal definition of a wide spread. Commissions, request latency, rejected orders and actual slippage are not supplied by this quote file. Activity cannot replace those separate execution checks.
10. Inspect a real candle and change the activity reference
Choose M1 or M5, the counting convention and the number of earlier bars. Then select a candle or move one candle at a time. Only the displayed measurement changes; these controls do not place a trade or change the historical prices.
Open a worked example · M1, N, 20 bars
The selected candle is highlighted in blue. Prices use Bid; the lower panel shows the chosen count. The reference is the median before that candle, never a median containing it.
On a narrow screen, scroll the chart horizontally to read the full labels.
Values of the selected candle
- UTC interval · end excluded
- —
- Bid Open / High / Low / Close
- —
- Selected count
- —
- Median of earlier bars
- —
- Count / earlier median
- —
- Bid High−Low · pips
- —
- Bid Close−Open · pips
- —
- All records · N
- —
- Price-changing records · Q
- —
- Minimum spread · pips
- —
- Quote-record median spread · pips
- —
- Maximum spread · pips
- —
The ratio describes past quote activity. It supplies no trade direction.
Show the selected calculation
11. Distinguish missing history from quiet trading
A minute with no source record has no observed OHLC. It is left absent rather than filled with the previous Close and a zero count. A gap may reflect a market closure, a quiet feed or unavailable data; these records alone do not identify its cause. The source's own status report is retained alongside the extraction. A complete group here means each required minute has observations; it does not certify that every market update was captured.
The viewer uses a fixed 2024-01-08 07:00–17:00 UTC window, with earlier history retained for its references. It is a teaching slice of one feed and one day. Do not compare its raw counts directly with another broker, pair, month or timestamp convention and call the difference a change in market-wide volume.
12. Check the MT4 field, timeframe and loaded history
In MQL4, iVolume(NULL, PERIOD_M1, 1) requests the tick-volume value of the previous bar on the current symbol's M1 series. Shift 0 refers to the current, developing bar. Confirm the symbol, timeframe, candle boundaries and available history before comparing a terminal reading with an external export.
A returned 0 can occur when local history is not loaded; it does not by itself prove that the market had no activity. Check loading and errors. This article's N and Q are independently declared reconstructions from HistData records, not values exported from a running MT4 terminal. Importing a CSV whose Volume field contains zeros will not recreate the original quote sequence.
iVolume(NULL, PERIOD_M1, 1)
13. Turn the observation into a testable research question
Freeze the source, bar clock, counter, reference length and threshold before evaluating a trading rule. Calculate every reference from information available at the decision time. Compare a price-based rule with and without the activity condition on the same opportunity calendar, preserving skipped events and all costs. Selecting only spectacular spikes would change the question after seeing the result.
A later strategy in this cluster can test whether an activity condition adds useful information. This guide establishes only the measurement. It provides no trading backtest, win rate or evidence that a larger count predicts a profitable scalp. Its retrospective examples and their selection rules are disclosed so that the visual explanation can be checked.
Frequently asked questions
Is Forex tick volume the number of lots traded?
No. Tick volume is a feed-specific activity count. This guide explicitly counts source quote records and, separately, records where Bid or Ask changes. Neither is a market-wide total of currency units or lots traded.
Does a green volume bar prove buying pressure?
No. Its colour depends on the display rule, often a price comparison. These quotes contain no signed transactions, aggressor identity or executed trade size from which buying and selling volume could be established.
Why are there two activity counts?
N includes every quote record. Q excludes records whose Bid and Ask are both unchanged from the preceding archive record. Showing both makes the event convention visible rather than assuming every platform counts identical events.
Can I compare a live M1 count with a closed M5 count?
Not as equivalent observation lengths. Complete aligned M5 counts sum five M1 counts, and their baseline must also use M5 bars. An unfinished bar remains incomplete until its interval has ended.
Does a ratio above 2 make a scalping entry valid?
No. It means the selected count exceeds twice the declared earlier median. It does not define direction, execution, risk or an expected profit. Any proposed trading filter needs a separate fixed-rule evaluation.
Why can MT4 show zero while this guide counts quotes?
The HistData export's Volume field is zero; this guide reconstructs counts from individual quote records instead. MT4 can also return zero when local history is not loaded. Check the source, loading state and field definition before interpreting a zero.
Data, timing and reproducibility
The downloads contain M1/M5 quote-derived OHLC, both counters, observed spreads, prior-only reference medians, source identification and calculation code. Integer prices use a scale of 1 000 000; one EUR/USD pip equals 100 of those units. No missing price candles or trades are generated.
The interactive slice is 2024-01-08 07:00–17:00 UTC. Its three worked examples were selected retrospectively using the disclosed activity, range and spread criteria. The original monthly tick archive remains identified by its download page and SHA-256; the reproducibility package explains the counting method.
- M1 quote activity and prices (CSV)
- M5 quote activity and prices (CSV)
- Data, source details and calculation method (ZIP)