Open Interest: How Many Contracts Are Actually Outstanding

Open interest is the tally of contracts on one description still standing at a stated instant, opened and not yet finished off by a closing trade, by delivery, or by being carried into a later contract. Two sides are bound inside every one of them and the pair counts as a single item. Creating an obligation lifts the tally; ending one drops it; handing one on leaves it alone.

A contract that has been struck and not yet finished is a promise still standing. Counting the promises still standing at one instant is a different act from counting the trades that made and unmade them during the day, and nearly everything below comes out of refusing to let those two acts blur into one.

What is the tally actually counting, and at which instant?

Start with a lending library on a Saturday morning. The librarian can report two things about her books and they are not the same thing. Forty books are out of the building right now, in somebody's bag or on somebody's bedside table. Separately, a hundred and twenty books crossed the counter this week, going out and coming back. The forty is a photograph. The hundred and twenty is a tally sheet. Asked how many books are out, anyone who hands over the week's counter total instead has answered a different question, and the librarian will say so.

Open interest is the forty: the number of contracts on one description that are outstanding at a stated instant, opened and not yet finished off. Every word in that sentence is doing work. On one description. Contracts of different descriptions are counted separately and never added together. Outstanding. A contract that has been closed, delivered or moved elsewhere has stopped being a promise anybody has to keep. At a stated instant. A level without a moment attached to it is not a reading at all, in the way that a photograph without a date is not evidence of anything.

A contract leaves the tally in exactly three ways, and knowing that there is no fourth way is what stops the number behaving like magic. A contract can be closed out: a position pointing the other way is taken in that very same contract, and the pair cancels. A contract can be delivered. Delivery means the arrangement ran to its end and the referenced thing changed hands. Or the position can be carried out of this contract and into a later one. The carry subtracts it here and adds it there. Nothing else removes a contract from the count. A price move does not. A holder losing interest does not. A quiet day does not.

Underneath sits the timing structure this whole subject runs on. The obligation is struck today. The money moves later. The referenced thing moves later still. Open interest is simply the count of agreements sitting in that gap at the instant it is read. The reference asset used throughout is priced at Rs 2,000.00/- today, and between today and the later date whoever holds it collects nothing from it. The absence of income is what keeps the agreed price a plain financing sum rather than a forecast.

Why does a contract with two bound sides add only one?

A wedding caterer takes a booking for the twelfth of the month. Two parties are locked into it. The household has to pay and the caterer has to turn up with the food, and neither of them may simply decide not to. Ask the caterer how many bookings are on her list for that day and she says one. She does not say two because two people signed it. The booking is the unit, not the signatures on it.

Futures work the same way and it matters more than it sounds. Every contract here binds a long position on one side and a short position on the other. Both are bound from the instant the price is agreed. Neither of them holds a right to decline, to choose, or to walk away. A right to choose belongs to a different kind of arrangement altogether, covered separately. The tally takes the bound pair as one item. A count of forty means forty positions bound to buy and forty positions bound to sell existing at the same instant. Forty outstanding contracts, not eighty.

Now the consequence, and this is the part readers are rarely made to sit with. The two sides are exactly equal at every instant, by construction, in every contract, on every day, in every market where this kind of contract exists. The equality is not something the participants achieved, and it is not evidence that opinion is evenly divided. The equality is arithmetic falling out of what a contract is: nobody can be bound to buy without somebody else being bound to sell the same thing. So the tally can never tell anybody that one side is more crowded than the other. A sentence claiming that open interest shows more buyers than sellers has misunderstood the object it is describing.

What happens to the tally when one contract changes hands?

Hold the tally still at forty and let exactly one contract trade. The base of forty is a working figure, chosen so the four cases have something to move against. Each side of that single trade is doing one of two things: stepping in with no position, or stepping out of a position already held. Two sides, two possibilities each, so four combinations and no fifth.

Case one: a party with no position buys from a party with no position. Neither of them was bound to anything a second ago and now both of them are. A duty has been manufactured out of an agreement, so the tally goes to forty one.

Case two: a party already bound to buy closes against a party already bound to sell, and both of them end their positions. Nothing is passed to anybody. The duty is extinguishedEnded so completely that nothing survives to be performed; the duty is not handed to somebody else, it stops existing., and the tally goes to thirty nine. Notice how different this is from case one even though the screen shows one contract trading in both.

Case three: a party with no position buys from a party who was already bound to buy and is now getting out. Think of a tenancy transferred to somebody new. The flat is still let, the agreement still binds, and the landlord still has exactly one tenancy running. Only the occupant changed. The duty survives intact, it has simply moved to a new holder, and the tally stays at forty.

Case four is case three seen from the other end: a party with no position sells to a party who was already bound to sell and is now getting out. Same reasoning, same result, tally still forty. Cases three and four are the pair readers get wrong most often. A trade unmistakably happened, money moved, a screen updated, and a reader who has been taught that activity means change expects the count to reflect it.

The tally does not respond to a variable that can be slid along a scale. The tally steps by whole contracts according to which of these four things happened, and reading a change between two moments is covered separately.

How is this different from the number of contracts traded?

Line the four cases up and something uncomfortable appears. The number of contracts traded was one in every single case. One contract changed hands, four times over, and the outstanding tally afterwards read forty one, then thirty nine, then forty, then forty again. Same activity. Four different levels. No amount of trading reveals what is standing, and no level reveals how much trading produced it.

The reason is that the two numbers are different kinds of number. A tally of outstanding contracts is a stockA quantity read off at one instant, true for the second at which it was read and silent about every other second.. A stock is read at an instant and has no duration attached. A count of contracts traded is a flowA quantity that piles up while time passes, so it means nothing until the stretch over which it was gathered is stated.. A flow accumulates while time passes and is meaningless until the stretch over which it was gathered is stated. There is no arithmetic that converts either into the other, and there never will be, in the same way that no amount of knowing how many people walked through a shop door reveals how many are inside it now.

The librarian returns to the room. Her forty books out is a stock. Her hundred and twenty crossings this week is a flow. If she reports that a hundred and twenty books were borrowed this week, a reader who concludes that a hundred and twenty are missing from the shelves has made precisely the error this whole distinction exists to prevent, and has made it with a much more familiar object than a futures contract.

One more thing worth holding on to about proportion. Adding a contract to a base of forty moves the level by 2.5 per cent of the tally, and removing one moves it by 2.5 per cent the other way. The symmetry is not a market fact, it is what dividing one by forty does, and it means a reader can talk about how far the level moved without ever knowing how much trading it took to move it.

How much does a tally of forty actually stand on?

Here is where the reporting damage happens. Three rupee figures get spoken about as though they were one, and separating them is the whole job. Take the base of forty contracts, one unit of the reference asset each. One unit a contract is a simplification. The size of one contract, measured in units of the referenced thing, is settled by the regulator, and it moves.

Every quantity here says which of two it is. Take the forty contracts across to the agreed price, Rs 2,130.00/- each, and the product is Rs 85,200.00/- of notional: a face amount, not one paisa of which is anybody's to move. Take the same forty across to the spot price instead, Rs 2,000.00/- a unit, and the product is Rs 80,000.00/- of exposure, meaning the worth of reference asset those obligations sit on.

Where does the agreed price of Rs 2,130.00/- come from? Borrowing Rs 2,000.00/- for a year is not free, and at a cost of 6.50 per cent a year the borrowing charge comes to Rs 130.00/-. Add that charge to what the reference asset costs today and Rs 2,130.00/- is the figure for the later date. The reference asset throws off no income at all, and the carry is the figure that would move if one were bolted on. The agreed price is a cost, worked with a multiplication, and it says nothing whatever about where anybody thinks the reference asset is heading. The whole gap between the notional and the exposure, Rs 5,200.00/-, is simply forty carries of Rs 130.00/- stacked up.

Now the third figure, the only one that involves money anybody has actually put anywhere. Initial margin at 8.0 per cent of the exposure, a working rate rather than a requirement anybody sets, is Rs 160.00/- a unit, so Rs 6,400.00/- has been lodged across the forty. Divide the Rs 80,000.00/- of exposure by the Rs 6,400.00/- lodged behind it and 12.50 times comes back; one unit answers identically, Rs 2,000.00/- over Rs 160.00/-; and growing the tally moves neither answer. Double the base to eighty and the exposure doubles, the collateral doubles, and the multiple sits precisely where it was. The multiple is the reciprocal of the 8.0 per cent margin rate rather than an independent finding. Nothing about the size of the tally can budge a reciprocal.

The same tally, set out as a build

Where does a real figure come from?

An open interest figure is not calculated. The whole answer rests on that one word. An open interest figure cannot be worked out from prices, cannot be inferred from a chart, cannot be derived from how much something moved, and there is no clever route from public information to a number that is fundamentally a headcount of promises. The figure is counted, and it is counted in one place only: the records of the clearing corporation, once every participant's positions in that contract have been netted at the end of the day.

Producing one requires every open position on one contract description, at a single instant, with the offsetting ones already cancelled against each other. Such a list is easy to describe and impossible to obtain. The list lives in a single ledgerA running book of entries kept by one party, which is where a tally of positions physically sits. held by the party standing in the middle of every trade. Nobody outside it has the list. A participantAnybody permitted to hold a position at a venue, permission itself being something the regulator decides. sees their own positions and nobody else's. One row of a table with thousands of rows.

The netting step is worth pausing on because it changes what the figure means. If somebody holds twelve bought and five sold in the same contract, the count that reaches the tally is seven, not seventeen. So the published number is never a grossCounted before anything has been allowed to cancel against anything else. pile of everything anybody entered into; it is an aggregateOne figure made by adding many separate ones, which then hides every one of them. figure arrived at after each holder's own offsets have been cancelled. Two figures that would both be honest descriptions of the same day come out different, and the one that gets published is the netted one.

Where a count came from decides how much weight it can carry. The forty used throughout was chosen so the mechanism has something to move against, and no clearing corporation stands behind it. The provenanceWhere a figure came from and who produced it, which decides how far it can be trusted. of a number is not a detail attached to it, it is half of what the number is, and a count with no source behind it is a teaching device rather than a fact about anything.

What does the tally refuse to disclose?

A number that is genuinely useful is also genuinely narrow, and the narrowness is the thing most worth teaching. Four questions arrive naturally the moment somebody sees an open interest figure, and the figure answers none of them.

The tally does not say who is holding the positions. The tally counts obligations, not people, and one holder may be standing behind forty of them or one, with nothing in the number separating those two worlds. The sides are equal by construction, so the tally cannot say which side is bigger. The tally does not say why anybody took a position: a contract held against something already in hand looks exactly the same in the count as one held entirely on its own, and whether a position counts as a hedge is a determination made under rules set by the regulator, not something a tally reveals. A headcount of standing duties carries no price inside it and points in no direction, so on what the reference asset does next the tally is entirely silent.

None of those four is recoverable from a count of standing obligations, however carefully the count is read.

Who reads this number, and what do they reach for next?

A tally of forty lands in front of four different readers, and every one of them immediately wants a second thing that the tally does not contain. The reader watching a contract into its final week wants to know whether the level is emptying. Emptying needs a second reading at a second moment, and that comparison is covered separately. The reader sizing collateral wants the exposure underneath. Exposure needs the per contract quantity as well as the spot price, and the tally holds neither. The reader writing a note for somebody else wants to say how much is involved, and the honest version of that sentence names three separate figures rather than one. The reader who simply holds a contract wants to know whether the crowd around them is a good sign, and that is the one question here with no data behind it at all.

The tally is worth learning not for the breadth of what it reports but for the precision with which it reports the one thing it knows. A household running its whole budget off one salary understands this instinctively: knowing that exactly one income arrives is a narrow fact, it is completely reliable, and every question worth asking about the household needs at least one more fact beside it. Open interest is that kind of number. Reliable, narrow, and dangerous only when somebody stretches it.

Should a crowded contract change what a holder does?

The number arrives with no instruction attached to it, and the gap between those two is not an oversight. Whether a high level or a low one should change anybody's behaviour is a separate question.

Suppose somebody set out to answer it anyway. Missing material stops them. Why a position was taken in the first place never shows up in a tally. Neither does whatever the same holder is carrying that points the opposite way. An offsetting holding can reverse the answer by itself. Collateral already lodged sits in an account nobody here can see. The amount falling due on the day money genuinely moves sits in a rulebook. Knowing how a tally gets assembled says nothing about whether any particular holder belongs inside it.

Four words also get thrown around loosely, and each of them is worth pinning down. Something ranks as a price when it was agreed between two sides or quoted for all of them: Rs 2,000.00/- today, Rs 2,130.00/- for the later date. A premium is money handed over at the outset to acquire a position. Neither side of a futures contract buys anything on day one, so neither side pays a premium. A payoff is what the position throws off when it finishes. A profit is that payoff with every rupee spent getting there taken back out. Entry at the fair agreed price costs nothing on day one, so a reader who works out the payoff has quietly worked out the profit as well and may not notice they did two jobs in one step.

The reference asset priced at Rs 2,000.00/-, the base of forty contracts standing open and the 8.0 per cent initial margin lodged against them are invented.
Educational material. Not advice on any investment, tax, budget or market position.