Screens · Volatility
Most Volatile Stocks Today
A one-session screen finds instruments that moved unusually far today. It describes an event rather than a character, and by the time a name appears on it, the range that put it there has already traded.
What the screen measures
One session, one number: (high − low) ÷ previous close × 100. Some versions substitute the true range, which takes the larger of that span and the distance from yesterday’s close, so that a session which gapped and then drifted is not recorded as quiet.
The denominator is a real choice rather than a detail. Dividing by the previous close uses a number that was known before the session began and cannot be moved by the event being measured. Dividing by today’s close flatters anything that fell hard, because the same range over a smaller number reads as a larger percentage, a distortion that is negligible on quiet days and material on exactly the days this screen exists to find.
Range is not direction, and neither is news
The figure above is the argument of this page. Three sessions with the same span are three different markets, and the screen that found them cannot tell them apart. Two further readings are needed and both are free.
Where the close sits in the range. At the top, the session ended in the hands of buyers; at the bottom, everything the move gained was given back; in the middle, the day was a disagreement. This is the reading volume spread analysis is built on, and it costs one glance.
What the volume did. A wide range on heavy volume means size was traded against; someone large was on the other side, repeatedly. The same range on the lightest session of the month means an ordinary order walked a thin book, which typically unwinds as soon as anyone real arrives.
Why a same-day list is a starting point, not a finding
Four causes account for most of any day’s list, and the appropriate response to each is completely different. The screen cannot distinguish them; a minute of looking can.
| Cause | How it usually looks |
|---|---|
| Earnings or guidance | A gap at the open, heavy volume all session, and a range that mostly forms in the first hour. |
| Corporate action | A takeover approach or an index change, very heavy volume, and a price that then sits still near a specific level. |
| Sector move amplified | The whole peer group appears on the same list; the instrument moved more, not differently. |
| Thin-book liquidity event | A wide range on volume no heavier than usual, frequently reversed the next session. |
The fourth is the one that most often wastes an afternoon, and it is also the one a liquidity floor removes before the list is ever printed. Filter on median dollar volume — not share volume, which is dominated by price level — and most of the noise never appears.
When the list is worth reading
Given all of that, there is still one use that survives, and it is the day after rather than the day itself. An instrument whose character changes abruptly frequently stays changed for a while. Large moves cluster in time, which is among the better-documented regularities in market data. A same-day list is therefore a reasonable way to find candidates for follow-through work, provided the follow-through is about the size of subsequent moves rather than their direction.
That distinction is worth holding onto because it is where most people go wrong with this screen. Clustering says tomorrow is likely to be eventful too; it says nothing about which way. Position sizing, options pricing and stop placement all take that input directly. A trade does not.
Where it misleads
| Situation | What goes wrong |
|---|---|
| Range read as direction | A session can travel five per cent and close where it opened. Range and the position of the close are separate readings. |
| Run intraday, read as final | Volume and movement cluster at the open and the close, so a midday list is a different universe from an end-of-day one. |
| Today's close as denominator | Flatters instruments that fell hard, the same span over a smaller number reads larger. |
| Gaps ignored | A high-low span records a gap-and-drift session as quiet. Use the true range if gaps matter to you. |
| No liquidity floor | Thin-book moves dominate the list and reverse the next session. |
| Treated as a signal | The move that put a name on the list has already traded. What persists is the size of subsequent moves, not their direction. |
Frequently asked questions
How is a same-day volatility screen calculated?
Usually as the session’s high minus its low, divided by the close or by the previous close, expressed as a percentage. Some versions use the true range instead, which takes the larger of that span and the distance from yesterday’s close. The difference matters on a day that gapped, because a gap-and-drift session has a small high-low range and a large true range. As with every screen on this site, the definition decides the list.
How is this different from a 20-day volatility screen?
It answers a different question. A 20-day screen describes an instrument’s character; this thing typically moves three per cent a day. A same-day screen describes an event, something happened to this instrument in the last few hours. The first is a property you can plan around; the second is news you have already missed the start of, and the two lists have surprisingly few names in common on any given day.
Are these lists just yesterday’s news?
For the move itself, yes, and it is worth being blunt about that. By the time an instrument appears on a most-volatile-today list, the range that put it there has already been traded. What the list is genuinely useful for is the day after: an instrument whose character changed abruptly frequently stays changed for a while, and the screen is a reasonable way to find candidates for that kind of follow-through work.
Does a wide range mean the price ended far from where it started?
No, and this is the single most common misreading. A session can travel five per cent and close exactly where it opened, which describes disagreement rather than direction. The range and the position of the close are two separate readings, and a screen that ranks only on range treats a decisive trend day and a violent round trip as the same event.
Why does the intraday session shape matter?
Because volume and movement are not spread evenly through the day. A disproportionate share of both arrives in the first and last few minutes, around the opening auction and into the close. A screen run at midday therefore reports a different universe from one run after the close, and an instrument that looks unusually active at 10:15 may finish the session unremarkable. If the screen is run intraday, the time it was run belongs beside the list.
Should I use the previous close or today’s close as the denominator?
The previous close is the more defensible choice, because it is known before the session and cannot be moved by the very event being measured. Using today’s close flatters instruments that fell hard, the same range divided by a smaller number reads as a larger percentage. The distortion is small on ordinary days and material on the days a screen like this exists to find.
What causes a stock to appear here?
Four things account for most of the list, and they call for entirely different responses: an earnings report or guidance change, a corporate action such as a takeover approach or an index addition, a sector-wide move that the instrument amplifies, and a liquidity event in a thin book where a modest order walked the price. The screen cannot distinguish them; a look at the volume and the news can, in about a minute.
Does high same-day volatility persist into the next session?
Somewhat, and it is one of the better-documented regularities in market data. Large moves cluster in time rather than arriving independently. That is a statement about the size of moves, not about their direction, so it supports position sizing and options pricing rather than a trade. Any claim that today’s big movers are tomorrow’s winners is a different claim entirely, and one this data does not support.
Is there a live version of this screen?
Not yet, and a stale list here would be worse than none: a same-day screen dated last year is not merely out of date, it is actively misleading in a way a definition page never is. What is on this page is how the screen is built, what the reading means, and what to check next, with enough detail to run it yourself the moment you have a feed.