Lesson 0.3. Open Interest
Video version (in Russian) · 16 minutesIn the previous lesson we read the 1.17 row and saw two numbers: in the Friday contract of week four, 2,363 calls and 3,574 puts were added.
Both numbers carry a plus sign. Both mean 'there is more mass now'. But different events stand behind them — and today we learn to see that difference.
One agreement first, about the scale. Every strike in this lesson is a futures strike, as in the previous one: 1.17 here means the strike of an option on the euro future, not the quote on your spot chart. Everything we count and check lives on one scale — the exchange's.
Two counters that constantly get confused
Two different numbers are tracked for every exchange instrument.
Trading volume — how many contracts changed hands during the day. An activity counter. Every morning it resets to zero and starts counting again.
Open interest — how many contracts exist at this moment. Not 'how much was traded' but how many are alive right now. This counter never resets: it grows while positions are opened and falls when they are closed.
A picture to remember it by: volume is the turnstile at the hotel entrance, counting how many times people went through during the day. Open interest is the list of guests actually staying. A thousand passes through the turnstile in a day, and there are still the same three hundred guests.
How a contract is born and how it dies
This is where the whole mechanism hides. Every trade has two sides, and each of them either opens a new position or closes an old one. That gives exactly three combinations.
Both open. One buys a right for the first time, the other takes on an obligation for the first time. The contract did not exist before — now it does. Open interest rises by one.
Both close. One had a bought position, the other a sold one, and they closed against each other. The contract is gone. Open interest falls by one.
One opens, the other closes. The one holding the position left; a new participant took their place. The contract neither appeared nor disappeared — it changed hands. Open interest does not change.
And now the key point: in all three cases volume grew identically — a trade did happen. But open interest rose in one case, fell in another, and stayed the same in the third.
Back to the 1.17 row
Let's take those same two cells from the previous lesson and put volume next to them — how many contracts actually changed hands that day.
| volume for the day | change in OI | before | after | |
|---|---|---|---|---|
| puts 1.17 | 3,833 | +3,574 | 19 | 3,593 |
| calls 1.17 | 2,950 | +2,363 | 2,352 | 4,715 |
Puts. Before Friday there was practically nothing on this strike — nineteen contracts. During the day 3,833 contracts of volume went through, in sixty-five trades, and open interest grew by 3,574. In other words, ninety-three percent of the volume created new contracts. It could not have been otherwise: there was nothing there to close.
Calls. Here 2,352 contracts were already standing. Volume was 2,950, but the increase was only 2,363. The difference of roughly six hundred contracts is trades where one participant left and another took their place.
In the change table both cells look identical — 'mass arrived'. Volume shows that different events stand behind that arrival.
The same strike, the next trading day
The clearest thing is to look at the same cell on Monday.
On the calls of strike 1.17 in the Friday contract, 161 contracts of volume went through, and open interest grew by two: from 4,715 to 4,717.
A hundred and sixty-one contracts changed hands — and left practically no trace. Positions moved from hand to hand all day: someone left, someone took their place, and the total number of contracts on the strike stayed the same.
The same strike, two adjacent days. On Friday money arrived, on Monday only the owners changed. In the change table Monday looks like an empty cell — 'plus two'. Without volume you would decide nothing happened there.
An outflow that took three minutes
Monday also gave the opposite example. In the September monthly contract on strike 1.16, open interest fell by 480: from 4,509 to 4,029.
Volume was 1,009 contracts, and almost all of it fitted into three minutes, from 16:00 to 16:03 GMT, at practically the same price: first a trade of 150 contracts, then 132, 65, 50, then two of 250.
What is clearly visible: out of 1,009 contracts of volume, roughly 480 removed positions from the market, and the other five hundred-odd changed hands. Half the activity did not change the number of contracts at all.
The scale: the premium on that strike was around 0.0089 per contract. With a contract size of 125,000 euros that is on the order of 1,100 dollars each, and across the strike's whole daily volume about a million and a bit in premium alone.
What is not visible: where the mass went and who took it away. There are no comparable inflows in the same contract. The data shows the fact of the departure, not its reason.
Why the table shows a minus
Now the red cells read by themselves. A plus means there were more openings than closings. A minus means more closings.
The thing not to substitute here is the familiar 'they sold'. Selling is present in any trade — every purchase has a seller. A minus does not mean direction; it means the participants walked away: those who held positions on that strike closed them.
Three ways mass leaves
Open interest does not fall only through trades. There are three ways.
Closing by trade. A participant changed their mind or locked in a result. The most frequent case and the most informative one: someone made a decision right now.
Expiration. The contract reached its date, the right was not needed, the option burned out. Positions leave the strike all at once — not because someone changed their mind, but because time ran out.
Exercise. The right ended up in the money and was used. The option disappears and the holder gets a futures position instead.
The difference is visible in the numbers. On that same Monday, the Monday contract of week four expired. Across the whole day 44 contracts of volume went through it — five trades. And open interest fell from 6,677 to zero.
Forty-four against six and a half thousand. No trades could have done that: the mass left not by anyone's decision but on schedule.
The practical conclusion: a minus on an expiration day and a minus on an ordinary day are different news. That is why in our reports it always matters which contract expires today.
Where to look at this yourself
We only see the change in open interest the next morning. Volume is visible immediately — trades reach the tape as they happen. Both are shown openly by the exchange to any registered user.
Cumulative open interest — the Open Interest Heatmap tool: https://www.cmegroup.com/tools-information/quikstrike/open-interest-heatmap.html
It shows how many contracts sit on each strike, separately for calls and puts, for each series, on any past date.
And there is a practical subtlety here worth knowing in advance. The change mode in that tool compares the latest report with one of the previous days. So you could look at what changed over Monday on Tuesday morning — but by Wednesday that option is gone: the latest report is now Tuesday, and Monday can no longer be compared with Friday.
The conclusion is simple: look at changes on the same day. Cumulative interest can wait, it is always available, but the exchange's daily change lives exactly twenty-four hours.
We do not have that limitation: we take the report every morning and file it, so our reports can show the change for any past date — even the Tuesday before last. This whole lesson, in fact, is built on data from two days, 21 and 24 August, and on the exchange's site they can no longer be compared today.
That is not a trick or a privilege: anyone can do the same if they start saving the reports. What matters is understanding that the daily change is not stored by itself. Miss it, and that day's picture is lost for good.
Volume and the trades themselves — CME Globex Trade Browser: https://www.cmegroup.com/tools-information/quikstrike/cme-globex-trade-browser.html
An intraday tape and a week of history: which strikes traded, in what size, at what price. That is where the trades we went through above came from — the burst on strike 1.16.
Both tools work on the futures scale of strikes — the same one every number in this lesson lives on. That is the source: the market's open interest can only be seen at the exchange, and so can the trades.
A caveat about our own screen
We have our own option flow matrix: it shows newly detected volume in three windows — for the day, for the hour and for the last fifteen minutes. The windows are there to separate 'they built it up slowly all day' from 'it arrived all at once in a quarter of an hour'.
But it must not be read interchangeably with the exchange tables, and here is why: its rows are converted spot price levels, not futures strikes. The same option sits on one number in the exchange table and on a different one in our matrix. Whether those numbers match or differ means nothing: they are different scales.
The rule is simple. When you are counting, checking or analysing open interest, you work on the exchange site, in futures strikes. Our matrix is for something else: to see when inside the day the activity arrived, while the exchange's totals are not yet in. Mixing the two scales in one line of reasoning is a reliable way to reach a wrong conclusion.
Assumption and fact
There is a simple way to guess in advance what today's volume will turn out to be: compare it with what is already standing on the strike.
In the Friday contract, strike 1.16 held 1,929 puts, and over Monday 598 contracts of volume came through — there was something to close, so an outflow was possible. On the neighbouring strike 1.1575 only 81 were standing, while volume was 174 — closing 174 contracts when 81 exist is physically impossible, so most of the trades had to be openings.
The first is an assumption, the second an arithmetic necessity. Let's see how it ended the next morning.
On strike 1.16 open interest grew by 536: of the 598 contracts of volume, almost ninety percent turned out to be new positions. The mass was not being closed but added to — from 1,929 to 2,465. On strike 1.1575 the increase was exactly 174 — every contract of volume went into new positions.
The possibility of closing existed, and closing did not happen. In the evening you see the assumption, in the morning the fact, and they do not have to agree. Arithmetic tells you what cannot happen; it does not tell you what will.
Stock and flow
Open interest has two sides, and they should not be confused.
Accumulated mass — how many contracts stand on the strike in total. A slow quantity: it holds for weeks. It answers the question 'how much is hanging there'.
The change — how much arrived or left over the day. It answers a different question: where the money was working yesterday.
The difference shows up well in numbers: in the September contract, strike 1.18 holds almost twenty thousand calls — and over Monday two were added there. A huge stock, no flow.
When these numbers appear
Open interest is not counted in real time.
During the day the exchange knows the volume — trades are visible immediately. How many positions were left open is only established after the trading day closes, when every trade has been allocated to accounts. That is why we see yesterday's change in open interest the next morning.
Hence something important: the table we read in the morning describes yesterday. It is a map of footprints, not a live broadcast. For working with levels that is fine: positions opened yesterday have not gone anywhere overnight — they live until their expiration.
An honest caveat about the arithmetic
It follows from the mechanics that volume cannot be smaller than the change in open interest: every new contract is born from a trade, so each unit of increase needs at least one trade.
In practice, published data sometimes does not add up that way: there are strikes where the increase in open interest exceeds the volume shown. The reason is not the mechanics but the publication — the exchange does not print all volume, and some trades never reach the public tapes. We account for that in our calculations too: premiums, for example, we take from settlement prices rather than from the trade tape.
Such discrepancies are rare and small, but they need to be known: any data has a limit of accuracy, and pretending it doesn't is the worst thing you can do with numbers.
What open interest does not tell you
It does not tell you who is on which side. Increased mass means new contracts appeared, not that 'everyone bought'. Every call bought has a seller with the opposite interest.
Nor does it tell you why a position was opened. The same plus can be a bet on a move, a hedge against a position already held, or part of a construction across several strikes at once.
What it does tell you precisely: on this strike, in this expiry, over this day there are more positions or fewer. That is enough to see where the interest is concentrated — and it is all we take from it.
Four misconceptions
'Large volume means a position was built.' Not necessarily. We saw a cell where a hundred and sixty-one contracts of volume went through and open interest grew by two.
'Open interest is rising, so price will go there.' Growth only means there are more positions. Direction does not follow — every position has two sides.
'A minus is a bearish signal.' A minus is mass leaving a strike. It can mean taking profit, removing a hedge, or simply expiration — like the contract that lost 6,677 contracts in a day without trading even fifty.
'The data is from yesterday, so it's stale.' Positions opened yesterday have not gone anywhere overnight. It is not the mass that goes stale — it is the price relative to it.
What's next
We have covered what an option is, how the table is built and what the number in the cell means. One topic of stage 0 remains — the most uncomfortable and the most important: what this data fundamentally cannot do. Where the line runs between 'visible' and 'assumed', and why our reports publish the misses as well as the hits.