· DOM & TAPE
Counting prints per second during the open
Prints per second jump at the 9:30 New York open. Ten ES seconds show how to count them, why the peak beats the minute average and how fill grouping skews it.
SHORT ANSWER
To count prints per second during the open, export the tape with millisecond timestamps, cut each timestamp to the whole second and count the rows in each second. Add the sizes for contracts per second. Compare the peak second with the median of the same second on earlier sessions, and record how your platform groups fills.
Counting prints per second during the open means counting the trades that print on time and sales in each second around 9:30 New York time. A print is one row on the tape, with a time, price, size and side. The note on tape speed as a measure of activity compared whole minutes across a session. This note narrows the window to ten seconds around the open in ES, the E-mini S&P 500 future, where a one minute average spreads a five second burst across the whole minute.
Why the open needs its own count
ES trades nearly 24 hours a day, so 9:30 is not the start of the futures session. It is the start of the US cash stock session. Stock traders and index arbitrage programs act at that moment, and ES prints per second rise within the first second.
A one minute rate spreads that burst across 60 seconds. If you want to know how fast the tape ran when the cash session opened, you have to count each second.
How to count prints in each second
- Export or copy the time and sales rows from 9:29:55 to 9:30:05. You need timestamps with milliseconds, because dozens of prints share one second.
- Cut each timestamp to the whole second, so 9:30:01.374 becomes 9:30:01.
- Count the rows in each second. That gives prints per second.
- Add the size column for each second. That gives contracts per second.
- Divide contracts by prints for the average print size.
A spreadsheet does all five steps in a few minutes. Do it by hand once, so you know what any speed figure on your platform counts.
Ten seconds around the ES open
All numbers below are invented.
| Second | Prints | Contracts | Average print size |
|---|---|---|---|
| 9:29:55 | 6 | 14 | 2.3 |
| 9:29:56 | 4 | 9 | 2.3 |
| 9:29:57 | 7 | 18 | 2.6 |
| 9:29:58 | 5 | 12 | 2.4 |
| 9:29:59 | 8 | 21 | 2.6 |
| 9:30:00 | 64 | 410 | 6.4 |
| 9:30:01 | 92 | 530 | 5.8 |
| 9:30:02 | 71 | 380 | 5.4 |
| 9:30:03 | 48 | 260 | 5.4 |
| 9:30:04 | 39 | 190 | 4.9 |
The five seconds before 9:30 printed 30 trades and 74 contracts. That is 6.0 prints and 14.8 contracts per second. The five seconds from 9:30:00 printed 314 trades and 1,770 contracts. That is 62.8 prints and 354 contracts per second.
Prints per second rose by a factor of about 10. Contracts per second rose by a factor of about 24, because the average print grew from 2.5 contracts to 5.6. Traders sent larger orders at the open as well as more of them.
The peak second against the minute
Suppose the whole 9:30 minute printed 1,860 trades and 7,800 contracts. That averages 31.0 prints and 130 contracts per second.
The peak second, 9:30:01, printed 92 trades. That is about three times the minute average. The first five seconds held 17 percent of the minute's prints and 23 percent of its contracts, in 8 percent of its time. A trader who reads the one minute figure of 31 prints per second would not know that the tape ran three times faster for a moment.
A baseline for the same second
Today's count needs a reference. Compare it with the same second on earlier sessions. Here are invented peak-second print counts from the last ten sessions at the open.
55, 61, 70, 58, 140, 66, 63, 59, 72, 64
The average is 70.8. The median, the middle value once the ten counts are sorted, is 63.5. The 140 came from one session with a large news move, and it pulls the average up by more than 7 prints.
Today's 92 is 1.45 times the median and 1.30 times the average. Use the median, so one unusual day does not reset what counts as normal.
How fill grouping changes the count
At the open, one aggressive order often fills against many resting orders. Suppose a 60 lot buy at 9:30:01 fills against 12 resting sell orders at two prices.
- A platform that prints every fill shows 12 rows.
- A platform that groups fills by price shows 2 rows.
- A platform that groups the whole order shows 1 row.
The 9:30:01 count could be 92, 82 or 81, and all three describe the same trades. The contract count stays at 530 in every case. The note on why your tape and another platform's tape do not match covers grouping rules. When you compare counts across platforms, compare contracts per second first.
Where to count prints
NinjaTrader has a Time and Sales window built in. ATAS lists Smart Tape on all plans. Bookmap includes Time & Sales in its core desktop product. Jigsaw daytradr builds its ladder workflow around Depth & Sales and a Reconstructed Tape.
Liquivue has no verified standalone time and sales window. Its trade bubbles plot each trade on the heatmap, so the burst at 9:30 shows as a dense cluster of bubbles beside the resting orders. A trader who connects Liquivue through the NinjaTrader bridge can keep the NinjaTrader Time and Sales window open for the exact count.
A check you can run
Export the ES tape from 9:29:55 to 9:30:05 for today and for the last ten sessions. Count prints and contracts in each second. Write down today's peak second, the median peak of the ten earlier sessions, and the ratio between them.
Questions
How do you count prints per second?
Take the time and sales rows for the period, cut each timestamp to the whole second, and count the rows that share each second. Add the size column for the same seconds to get contracts per second.
Why is the tape so fast at 9:30 in ES futures?
ES trades almost around the clock, but 9:30 New York time is when the US cash stock session opens. Many traders and programs act at that time, so the number of trades per second rises several times over.
Why use the median instead of the average for a baseline?
One unusual session can pull an average a long way. The median is the middle value of the sorted sessions, so a single news day moves it very little.
Does Liquivue have a time and sales window for counting prints?
This site has not verified a standalone time and sales window in Liquivue. Its trade bubbles show the cluster of trades at the open on the heatmap, but an exact print count needs a tape from another platform.
Educational content about market data and charting software. It is not trading or investment advice. Trading involves risk.