The Tape: What Actually Traded
The tape is the running record of trades that actually executed: price, size, time and venue, in sequence. It shows what was done rather than what was offered, so a print cannot be cancelled the way a resting order can. It is most useful for timing an entry, not for choosing one.
How it works
The tape is the sequential record of trades that actually executed. Each line carries a price, a size, a timestamp and usually the venue, in the order they printed. The name comes from the paper ticker tape; the modern window is labelled time and sales.
The book shows intent; the tape shows execution. An order resting in the order book can be cancelled in a millisecond, and a print cannot — which is the entire reason anybody watches it.
A trade at the offer is somebody who wanted it now. At the offer the buyer paid up rather than wait; at the bid the seller did. Which side crossed the bid-ask spread is most of what a single line says.
And speed carries more than the prints themselves. Twenty trades in one second and the same twenty spread across a minute are identical in volume and nothing alike as markets.
A large print matters only against what is normal there. There is no absolute threshold: a size that is enormous in one instrument is unremarkable in another, and both shift by the hour.
Why it is harder than it was
Reading it is a skill with a very short shelf life. What a print implied about the participant behind it held while orders were large, human and sent whole. Execution technology removed all three conditions.
Most prints today are one order sliced into pieces. An execution algorithm cuts a large parent order into hundreds of small child orders, precisely so that the visible size distribution reveals nothing.
Some of it never arrives on time, or at all. Trades matched in dark pools reach the record after the fact, and a hidden order shows no size in level 2 before it fills.
The volume bar is the tape, summed and simplified. A candle’s volume figure is only its prints added together, which is why volume analysis and tape reading answer one question at two resolutions.
On a daily chart none of this survives. Aggregation keeps the total and discards the sequence, and sequence is the entire content of a tape.
In practice
And there is no tape at all across a gap. An opening gap is a price change with no executions in between, so there is nothing to read and nobody to attribute it to.
It never tells you where the stop belongs. A print is a fact about the last second; a stop loss is a decision about the next several hours.
And fast reading means fast trading, charged by the round trip. On this site’s shared 576-bar history the median bar spans 0.493 price units and the smallest 0.022, so a 0.0098 round trip is 2% of a typical bar and 45% of the quietest.
It is the one record that cannot be cancelled. Depth of market shows what is offered now and market makers may withdraw it; the tape reports only what has already happened.
Learning what normal looks like
Open a time and sales window beside one instrument and leave it there for several sessions. Do not trade from it. The first task is calibration — learning the ordinary rate, the ordinary size and the ordinary quiet of that instrument at each hour.
Normal is specific to the instrument and to the hour, and nothing substitutes for watching it. The first half hour and the last behave nothing like the middle of the day. A print you would call large at noon is unremarkable at the open, and no general rule will tell you which is which.
Only after that does a single line mean anything. Until you know the baseline, everything unusual-looking is unusual only to you. Most people skip the calibration, decide the tape is unreadable noise, and are half right — it is noise until the reader has a reference.
What the tape is not
- Not an identity. A print carries a price and a size, never a participant.
- Not an intention. The buyer crossing the spread may be opening, closing or hedging.
- Not a headcount. One firm can sit on both sides of a print, and nothing marks it.
- Not a forecast. It describes the past few seconds and claims nothing about the next.
When it fails
In a quiet market the tape is mostly noise. Inside a trading range prints arrive in both directions at similar size and the reader supplies the pattern. On this site’s history, direction runs average 2.01 bars across 286 runs, with a longest of 11.
- When the instrument is thin. A handful of prints a minute means one participant can produce everything you are reading.
- When size is being worked. An algorithm filling a parent order looks exactly like broad participation, by design.
- When the fill is passive. A resting buyer taken at the bid registers as selling pressure, because the seller was the side that crossed.
- When the venue reports late. Off-exchange prints arrive out of sequence, so the order on your screen is not the order that happened.
- When the horizon is wrong. Nothing here informs a position held for weeks; scalping is the horizon on which the attention pays.
- When you are already positioned. Every print then reads as confirmation, and the window turns into a machine for justifying the trade you hold.
The original data
The functional name appears in zero of 31,760 video titles. A scan of
research/search-study-corpus.jsonl, logged in research/broker-coverage.json, finds nothing
titled “time and sales”, while “tape reading” appears six times at a median of 49,427 views. The
romantic name is watched fifty thousand times over; the functional one does not exist.
That is a fact about how the subject is sold, not about how it works. Reading prints is fast
trading, and on this site’s 576-bar history — measured into research/series-measurements.json by
site/measure_series.py — a round trip costs 45% of the smallest bar’s range. Use the tape to
decide how you enter, not whether you enter.
Related
Order flow is the subject this page is a slice of — the tape is its execution record, and the phrase carries 124 videos at a median of 14,001 views.
Level 2 is the other half of the screen, showing resting orders rather than completed ones, across 11 titles at a median of 127,639 views.
Volume analysis is the same information aggregated, and the neighbouring phrase “volume profile” runs to 346 videos at a median of 7,947.
I kept a time and sales window open for most of a year because it felt like being closer to the market. What it actually did was make me trade more, since something is always happening on it and stillness starts to feel like a mistake. I still open it, but only once I have already decided to take the trade and I am looking for a moment to get filled. Watching it before that decision never made the decision better.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money.