Screens · Volume

Most Traded Stocks of the Month

Sum a month of volume and the total belongs to its two or three biggest sessions. A monthly ranking therefore lists the instruments that had news, not the instruments that are busy, and the two lists have different uses.

What a monthly total is made of

Volume is a skewed quantity: it has a floor at zero, no ceiling, and occasional sessions many times the norm. Summing such a quantity over twenty-one sessions produces a total that is largely a statement about its largest few members.

That is not a subtle effect, and it is easy to demonstrate. The four instruments in the figure below have identical monthly volume by construction — the same total, to the share — and completely different months.

Share of a month's volume carried by its three busiest sessionsA bar chart of four synthetic instruments with identical monthly volume totals, showing what share of each month was carried by its three busiest sessions. The steady instrument's three busiest sessions carry the smallest share; the one with a single index event carries by far the largest.% of the month in its 3 busiest sessionsSteady18 %No event. Volume varies only with ordinary day-to-day noise.Placement38 %A share sale absorbed over three sessions in the first week.Results day45 %One announcement, with the following session still elevated.Index event51 %A single rebalance session near the end of the month.Share of a month's volume carried by its three busiest sessionsA bar chart of four synthetic instruments with identical monthly volume totals, showing what share of each month was carried by its three busiest sessions. The steady instrument's three busiest sessions carry the smallest share; the one with a single index event carries by far the largest.% of the month in its 3 busiest sessionsSteady18 %No event. Volume varies only with ordinary day-to-day noise.Placement38 %A share sale absorbed over three sessions in the first week.Results day45 %One announcement, with the following session still elevated.Index event51 %A single rebalance session near the end of the month.
Fig. 1: synthetic months, computed at build time · identical totalsFour instruments, each generated to the same monthly total of 21 million shares over 21 sessions, differing only in shape. In the steady one, the three busiest sessions carry 19 per cent of the month, roughly what noise alone produces when three sessions out of twenty-one are picked. In the one with a single index event they carry 51 per cent, and that instrument's busiest session is 16.4 times its median session. A monthly ranking sees one number for all four and puts them in the same position on the list.

The instructive comparison is the first row against the last. Both instruments traded 21 million shares in the month. One did it by trading roughly a million shares a day; the other did it mostly in a single closing auction. Only one of them can absorb an order tomorrow, and the monthly list cannot see the difference.

The four events that make a month

What puts an instrument on a monthly list
EventShape it leavesHow to check
Results announcementOne very heavy session, with the next still elevated. Recurs quarterly.The reporting calendar, published in advance, and the single most common cause.
Index addition or deletionA single enormous closing auction, often many multiples of a normal session.The index provider’s rebalance schedule, also published in advance.
Placement or secondary offeringSeveral consecutive heavy sessions as the stock is absorbed.Company announcements. The multi-session shape is its signature.
Options expiryElevated volume on a known date each month, larger on quarterly expiries.The expiry calendar. This one is the easiest to forget precisely because it is routine.

Every row of that table is knowable before the fact, which is what makes the omission remarkable. A monthly volume list is the one screen where the causes of the ranking are almost entirely on a published calendar, and the ranking never mentions them.

What the monthly window is genuinely good for

One thing, and it is worth having: a monthly total is a far steadier liquidity gate than a daily figure. If the question is whether an instrument can absorb a position at all, a month of volume answers it better than a single session that might have been a holiday or an expiry.

Used that way (as a floor that decides who is admitted rather than as an order) it is sound, and it is how professional screens use it. The failure is only in reading the ranking itself as information, and the fix is to publish a shape statistic beside the total.

The median session describes the ordinary state of the instrument. The total excluding its three busiest sessions does the same more bluntly. And the gap between the total and the median × 21 is itself the most useful number of the three: a large gap means the month was one or two events, and a small gap means it was busy throughout. None of those is expensive to compute, and any of them turns a list of news into a list of liquidity.

Survivorship, and the list you are not shown

One further problem belongs to any ranking computed over a period rather than a day, and it is invisible by construction: the list contains only instruments that existed for the whole month.

An instrument that was suspended, delisted, acquired or converted part-way through does not appear, however heavily it traded while it was there, and those are frequently the heaviest-traded instruments of the month, because a takeover or a suspension concentrates enormous volume into a few sessions. So a monthly list systematically omits the most active situations in the market, and it does so silently.

The same effect works in the other direction for new arrivals. A recent listing has a partial month, so its total understates its daily activity, and it is ranked against instruments with a full twenty-one sessions. Either case is easy to correct once noticed, rank by average daily volume over the sessions each instrument actually traded, rather than by the month's total, and neither is corrected in any published list.

The knock-on effect on every quoted average

This is not only a screening problem. The average volume figure printed on quote pages is usually a one- or three-month mean, so it contains whatever event sessions fell inside that window, and keeps containing them until they age out.

The consequence is that after a single large announcement, an instrument’s quoted baseline sits materially above its ordinary level for weeks, and every relative-volume reading taken against that baseline reports the following sessions as quieter than they were. The average volume page works through the fix in detail; the short version is to know what the window contains before dividing by it.

Frequently asked questions

How is a monthly most-traded list built?

By summing each instrument’s daily volume across the month and ranking the totals. That is the obvious construction and it has a consequence nobody states: because volume is a highly skewed quantity, a monthly total is dominated by its largest few sessions. So the list ranks the instruments that had the biggest event days rather than the instruments that trade heavily as a matter of course.

How concentrated is a month’s volume?

Far more than the average suggests. The figure on this page constructs four instruments with identical monthly totals and different shapes: in the steadiest of them the three busiest sessions carry a modest share of the month, and in the one with a single index event they carry a large majority. A monthly total cannot tell those two apart, and they are not the same instrument.

What kinds of event dominate a month?

Four, and all of them are identifiable. A results announcement, which lifts one session and part of the next. An index rebalance, which concentrates enormous volume into a single closing auction. A placement or secondary offering, absorbed over a few sessions. And an expiry, which is scheduled and recurring. None of them is a statement about how liquid the instrument normally is.

Is a monthly list ever better than a daily one?

Yes, for one purpose: it is a much better liquidity gate. If you want to know which instruments can absorb an order at all, a month of volume is a steadier estimate than a single day, which can be quiet or extraordinary for reasons unrelated to the instrument. Used as a filter it is sound; used as a ranking it mostly ranks news.

What would a better monthly measure look like?

The median session rather than the total, or the total with the busiest two or three sessions excluded, either of which describes the instrument’s ordinary state instead of its exceptional one. Publishing both the total and the median alongside each other is more useful still, because the gap between them is itself informative: a large gap means the month was one or two events, and a small gap means it was genuinely busy throughout.

Does the same problem affect monthly price statistics?

It affects anything summed or averaged over a window containing extremes, which is most things. A monthly average price is pulled by the days with the most trading, a monthly return is decided by a few sessions, and a monthly volatility figure is dominated by its largest moves. The general habit worth keeping is to ask what share of a total came from its largest few components before reading the total as a description.

Do index rebalance dates really matter that much?

For an instrument being added to or removed from a widely tracked index, yes. Funds must trade at the reference price, and the volume arrives in one closing auction that can be many multiples of a normal session. The dates are published well in advance, which makes this the easiest of all volume distortions to check for and one of the most frequently missed.

How does this interact with the average volume figure on a quote page?

Directly, and it is worth knowing. A quoted three-month average volume includes whatever event sessions fell inside that window, so it can be materially above the instrument’s ordinary level for weeks after a single announcement. Anyone comparing today against that baseline will read the following weeks as quieter than they are, the same window problem the average volume page works through in detail.