Volume and liquidity — how to measure them

The thesis of this lesson: price says what, volume says how

Volume is the number of units that changed hands in an interval. The chart shows price being negotiated; volume shows the commitment behind that negotiation. On its own it does not predict direction, but it can:

• confirm a move — a breakout on rising volume has more fuel than a breakout in an empty market;
• warn of weakness — a pattern on falling volume is often less reliable;
• point to volume divergence — a new price high on lower volume is a caution signal, not a certainty.

The key point: volume is context, not a separate signal. You do not open a position “because volume is big” — you combine it with structure, zones and patterns. The problem is that “big” and “small” mean nothing until you relate them to the history of that particular instrument — which is why volume is measured relatively, and execution costs are measured in the same unit as risk (R).

Where to look at volume — Warsaw Stock Exchange vs CFD

On a CFD volume is often estimated or absent — in that case you rely on price and liquidity (spread, depth). That is another reason to treat volume as context, not an absolute. The differences between a brokerage account and a CFD are covered in the lesson brokerage account vs CFD.

Measurement protocol: five numbers instead of a “big bar”

A measurement is repeatable if you define the units. You use the formulas below for every instrument and every session. The parameter N and the thresholds are illustrative numbers — fit them to the instrument and confirm the definition of volume in the instrument description at your broker.

RVOL            = current_interval_volume / average_volume_over_N_intervals
Spread_in_R     = spread / |entry − SL|
Slippage        = |fill_price − order_price|
round_trip_cost = (spread + commission × 2 + swap) / |entry − SL|
Coverage        = volume_in_top_5_order_book_levels / planned_position_size

Notice two things. RVOL is relative — it divides by the average of the same series, so a value of 3.0 means “three times the typical turnover for this instrument”, not “a lot”. Spread_in_R is dimensionless — a fraction of the distance to the stop, and therefore a direct cost in R (0.05 means 5% of the risk per trade, paid at the moment of entry alone).

The steps, in this order:

1. Set the interval and the N base. The interval must be the one on which you make the decision; N is the length of the averaging window (20 in the examples — an illustrative number). Write both parameters down once, permanently, in your worksheet.
2. Calculate the average volume over N intervals from the window preceding the current interval. Excluding the current bar matters: otherwise a breakout inflates its own baseline and RVOL is understated.
3. Calculate RVOL at the moment of decision. A value below 1.0 means turnover lower than typical; a value above 1.5 (an illustrative number) is only then a candidate for confirmation.
4. Measure the spread and convert it into Spread_in_R, using the distance to the stop from your plan. Read the spread at the same moment you intend to enter — it widens around data releases and at the open.
5. Measure the order book depth on the side you are entering, add up the volume in the top five levels and calculate Coverage. This answers the question of whether your order fits within the visible liquidity.
6. Record the measurement before entering, and add Slippage and the result in R after exiting. Only the complete set of columns lets you compare trades — as in any journal of metrics.

Worked example — every number illustrative. Entry at $100.00, stop at $98.00, a distance of $2.00. Current interval volume 480,000 units, the average over 20 intervals 150,000, so RVOL = 3.2. Spread $0.04, so Spread_in_R = 0.04 / 2.00 = 0.02, that is 2% of the distance to the stop. Adding $0.06 of slippage, the entry cost is $0.10, or 0.05 R. If a total of 800 units is waiting in the top five order book levels while the planned position is 250 shares, Coverage = 3.2. You can calculate each of these numbers from your own order book — and they, not the colour of the bar, decide whether you send the order at all, and whether you do it through a limit or a market order.

Worksheet: measure your instrument

Copy the table into a spreadsheet and fill it in for every instrument you trade. The empty cells are deliberate: the values should come from your own order book and your own history, not from this text. Fill in one row per analysis, including the occasions when no trade followed — rejections are data just as much as entries.

Two columns demand discipline rather than tools: you measure Spread in R at the moment of decision, and you add Slippage after the fill. If after a dozen rows it turns out that Spread in R rises in the same hours as Slippage, you have a concrete reason to shift your trading hours — rather than a general impression that “the market is difficult”.

When volume and liquidity say “do not trade”

Four rules. Meeting any one of them means skipping the setup or clearly reducing size — never “a trial entry with a tighter stop”.

1. RVOL below 0.7 on a breakout from a zone (an illustrative number). Price leaves the level on turnover lower than typical — that is a candidate for a false breakout, not for confirmation. You go back to the pattern and wait for it to close.
2. Spread_in_R above 0.10, that is above 10% of the distance to the stop (an illustrative number). The cost of entering and exiting eats part of R before price has even moved. With a spread that size, calculate the R:R after costs — if it falls below your plan’s threshold, there is no setup.
3. Coverage below 3.0 (an illustrative number). The planned position is more than one third of the visible order book depth, so filling it with a market order risks slippage you will not see in the spread table. Either reduce the size or work with a limit order.
4. No data to measure. An instrument with no volume and no view of the order book (true of some CFDs) does not let you calculate any of the five numbers. In that case you either cap the size at what fits within the visible depth, or you skip the instrument — and you confirm the instrument description with your broker.

A rule that organises the rest: volume rejects weak breakouts, while liquidity rejects entries that are too expensive. These are two different filters, and both have to pass before you calculate position size.

How to read volume in practice

A breakout from a zone. Price moving above resistance on a clear increase in volume means participants accept the new price. A breakout with RVOL below 1.0 is a candidate for a false breakout, not a confirmation — which is why it goes into the worksheet along with the result.

Corrections and volume. A healthy correction in a trend usually carries lower volume than the impulse. Rising volume during a correction can mean distribution — but that is an observation, not a stand-alone signal.

Volume divergence. A new price high on lower volume than the previous high means weakening momentum. In a trend that is a reason to reduce size, not to reverse a position “on a hunch”.

Liquidity — volume is not the whole story

Volume shows how much is happening; liquidity shows at what price you can actually fill your order. The components of liquidity:

• Spread — the bid–ask difference; the narrower it is, the cheaper your entries and exits (related to round-trip costs).
• Order book depth — how many orders wait on each side; in thin Warsaw Stock Exchange equities even a modest order can move the price, which the Coverage column shows directly.
• Slippage — the difference between the order price and the fill price; it grows in thin liquidity and around news. The entry on slippage describes how to limit it.

Practical rule: if you cannot see volume, at least measure the spread and the depth during your session. A thin market plus a market order costs money that the spread table does not show — and you will only see it in the Slippage column after the fact.

Measurable exercise

Choose two instruments: one with heavy turnover and one generally considered thin. For both, measure RVOL, the spread and the depth at three times of the session — at the open, in the middle of the day and before the close. Record the results in the table from this lesson and compare Spread in R and Coverage across those times.

Then do the same for a breakout that has already happened in the past: check what RVOL was in the breakout interval and whether price held the new level. This is not a backtest of a strategy — it is calibration of the thresholds from the “do not trade” section to your own instrument. Without that calibration, the thresholds 0.7 and 3.0 remain numbers from somebody else’s market and mean nothing.

Common questions

Is volume on a CFD reliable?

It depends on the broker: it may be aggregated from liquidity or synthetic. Treat it as context and write down in your worksheet where the number comes from — if volume is unavailable, measure the spread and the order book depth, and confirm the instrument description at the source.

Is volume divergence an entry signal?

Not on its own. It is a warning of weakening momentum — you combine it with structure and zones and base the decision on your plan. In practice it means reducing size or moving the entry closer to confirmation.

Where can I find volume on the Warsaw Stock Exchange?

In your brokerage platform: the volume bar under the chart, tick data and packages in the daily data. For futures contracts, open interest as well. When calculating RVOL, remember that the baseline average must come from the same source and the same interval.

Why calculate RVOL when the volume bar is visible on the chart?

Because a 'big bar' is relative: the same volume can be a record for one company and average for another, and it also changes with the time of the session. RVOL divides current turnover by the average of the same series, reducing the observation to a single number comparable across instruments and days. That lets you apply one threshold instead of judging the chart from scratch every time, and after a series of entries you can check whether the breakouts you rejected really did return to the zone more often. It measures context, not the future — on its own it gives no edge, it only shows whether the condition you set for yourself was met.