Reading the tape
Lesson 9 · about 9 min
Time and sales, "the tape", is the oldest order flow tool there is. It is a list of every trade: when it happened, at what price, how many contracts or shares, and whether it traded at the bid or the ask. Everything else in this module (footprint, delta, cumulative delta) is a summary of the tape. Learning to read the raw list first makes the summaries make sense.
What a tape looks like
Time | Price | Size | Side | Note
-------------+---------+------+------+------------------------
10:15:02.114 | 5000.25 | 3 | ask | buyer lifted
10:15:02.118 | 5000.25 | 1 | ask |
10:15:02.340 | 5000.00 | 2 | bid | seller hit
10:15:02.912 | 5000.25 | 45 | ask | large buyer
10:15:02.913 | 5000.25 | 12 | ask |
10:15:02.913 | 5000.50 | 20 | ask | level cleared, price ticked
10:15:03.207 | 5000.50 | 2 | ask |
10:15:03.550 | 5000.25 | 5 | bid | seller hit the new bid
The "side" column is the key. Most platforms colour it: one colour for trades at the ask (buyer-initiated), another for trades at the bid (seller-initiated). Some trades print between the bid and ask, or on a locked market, and get classified as neutral. On centralised futures, classification is reliable; on stocks and crypto it depends on the feed.
What to look for
Size relative to normal. A 45-lot on a product where the median print is 2 is an event. A 45-lot on a product where the median is 40 is nothing. Before reading a tape, watch it for ten minutes to learn what normal size is at this hour.
Speed. Prints arriving in bursts mean urgency. A tape that goes from a print every second to twenty per second is telling you aggression just arrived, before price has necessarily moved.
Repetition at one price. Lesson 4 of Module 1 covered icebergs. On the tape they look like this:
10:22:14.2 | 4998.75 | 10 | bid
10:22:14.6 | 4998.75 | 10 | bid
10:22:15.1 | 4998.75 | 10 | bid
10:22:15.3 | 4998.75 | 10 | bid
10:22:15.9 | 4998.75 | 10 | bid
10:22:16.4 | 4998.75 | 10 | bid
Sixty contracts sold into 4998.75 in two seconds and the price did not tick down. Someone is bidding there in clips of ten and refreshing. The tape shows it; the DOM shows a bid of 10 that never changes.
Which side the large prints are on. If the largest prints of the last few minutes are all at the ask, large participants are buying aggressively. If they are all at the bid, selling. If they are mixed, the large players are on both sides and the tape is not telling you much.
A short worked read
Here is one minute condensed to the prints that matter (size 10 or above), on a product where a normal print is 1 to 5.
| Time | Price | Size | Side |
|---|---|---|---|
| 10:40:05 | 5004.00 | 30 | ask |
| 10:40:11 | 5004.25 | 25 | ask |
| 10:40:19 | 5004.25 | 40 | ask |
| 10:40:24 | 5004.50 | 15 | ask |
| 10:40:31 | 5004.50 | 12 | bid |
| 10:40:38 | 5004.50 | 60 | bid |
| 10:40:44 | 5004.25 | 55 | bid |
| 10:40:52 | 5004.25 | 70 | bid |
Large buying lifted the market from 5004.00 to 5004.50 over 25 seconds. At 5004.50 the large prints switched sides: a 12-lot, then a 60-lot hit the bid, and then two more large sells as price gave back a tick. The buyers who lifted 110 contracts between 5004.00 and 5004.50 are now under water. If they were short-term traders, their exits are further selling. A tape reader would say: "large buyers got sold into at the high and the sellers are now in control."
You do not need anything more than the tape to make that read. What the footprint adds is a way to see the same thing across a whole session without watching every print.
Key idea: The tape is the only complete record. Read it for size relative to normal, speed, repetition at a price, and which side the large prints are on. Every other order flow tool is a summary of it.
Filters
Raw tape on a busy product is a blur. Every platform lets you filter: show only prints of at least N contracts, or highlight them. A common approach is to filter to the largest 5% of prints for the hour, which on a thick index future might be 50 or more contracts and on a thin commodity might be 5. The filtered tape is much easier to read and loses little, since small prints rarely change the picture.
There is a cost: an iceberg working in clips of 5 is invisible on a tape filtered to 50. Keep an unfiltered window somewhere, or rely on the footprint (next lesson), which sums the small prints anyway.
Limits of the tape
- It shows what traded, not what was intended. A 500-lot might be a hedge, a stop being triggered, or a spread leg.
- It shows one venue's trades for stocks and crypto. For futures on a single exchange it is complete.
- It cannot tell you who traded. Large prints could be one participant or an aggregation.
- Reading it by eye does not scale across a session. That is the footprint's job.
Try it: Watch an unfiltered tape for ten minutes and write down what a "normal" print size is. Then set a filter at about five times that size and watch for another ten minutes. Record, for the filtered prints, how many were at the ask and how many at the bid, and note whether the side of the large prints matched the direction price actually moved.
Recap
- Time and sales lists every trade with time, price, size and side; it is the source data for all order flow tools.
- Read the tape for size relative to normal, bursts of speed, repetition at one price, and which side large prints are on.
- Repeated same-size prints at one price with no tick is an iceberg refreshing.
- Filter to large prints to make a busy tape readable, but remember filters hide small-clip icebergs.
- The tape shows what traded, not why, and on fragmented markets it shows only one venue.
See it drawn
Original diagrams for the ideas on this page. Illustrative, not real market data.