Lesson 22
Open Interest (OI) & The Four Phases
Read the contract count, choose the comparison window, then interpret the price–OI pattern.
Price and OI. Four combinations.
| OI risesmore contracts outstanding | OI fallsfewer contracts outstanding | |
|---|---|---|
| Price up | ||
| Price down |
Conventional labels for one chosen interval · descriptive, not forecasts · hypothetical NIFTY futures series
Read the count before the story
Let’s separate trades completed today from contracts that remain open, then compare price and OI over one clearly stated window.
By the end you will be able to answer
- What do volume and open interest count?
- How can one trade change open interest?
- What do the four price and OI labels mean?
You are watching NIFTY futures on a trading screen. The price moves up. Volume keeps climbing. Another column shows open interest. All three numbers describe the same market, but each answers a different question. Where did the contract trade? How much traded? And how many contracts are still open?
The familiar names , , and come from comparing the price change with the change in open interest. This lesson follows one made-up NIFTY 50 futures contract. You will see how those names are given, why the comparison window matters, and where an observation ends and an interpretation begins.
Updated
What do volume and open interest count?
Learn what each of the two counts measures.
Mr. Chartist, my screen shows volume of 300 and open interest of 1,000. How can both be true?
They count different things. Volume is contracts traded in a period. Open interest is contracts still open at a moment.
So a busy day can leave open interest unchanged?
Yes. A trade that only hands a position over adds to volume and leaves OI where it was. OI moves when both sides open, or both sides close.
And long buildup means longs were added?
Not quite. It is a name for two measured changes: price up and OI up. It does not tell you who started each trade.
So write down the contract, the price field and both times first. The count comes first, the label second.
Volume is the number of contracts traded during a period. Open interest (OI) is the number of contracts still outstanding at a moment. Volume measures activity. OI measures what remains open. A busy session can end with more, fewer or the same number of open contracts. Lesson 21 showed this on an option series. Here we use it on futures.
Every outstanding futures contract has a long side and a short side, and OI counts the pair once. If 100 contracts are open, OI is 100, even though there are 100 longs and 100 shorts. Adding the two sides would count the same contracts twice. A trade of 100 contracts likewise adds 100 to volume, not 200.
Both figures are counted in contracts, so you need no lot size to read them. Neither is a shareholding, an amount of cash, a notional value or a margin figure. OI alone gives you none of those.
Scope matters too. The OI of one NIFTY expiry is not the OI of several expiries added together. For an option, name the underlying, expiry, strike and CE or PE before you compare. Do not let a count for one strike turn into a count for the whole market halfway through an argument.
Say the screen shows volume of 300 and OI of 1,000. That is perfectly possible: 300 contracts traded today, and 1,000 are open at the time of reading. Many positions were already there before the session began. To know whether the 300 added, removed or only moved positions, you need to know what each side of every trade was doing.
Volume counts the trades, OI counts what stays open
One futures series at one reading: volume 300 and OI 1,000. Both are counted in contracts.
Volume · activity today
- Contracts traded
- 300
Each matched trade adds to volume once.
Open interest · what remains
- Contracts open now
- 1,000
1,000 longs and 1,000 shorts, counted once as 1,000.
Both can be true together. 300 contracts traded today, and 1,000 are open when you read the screen. Most of those 1,000 were already there before the session began.
Made-up numbers. A contract is a long and a short side; OI counts the pair once. No lot size is used.
How can one trade change open interest?
See when a trade adds to the open count.
A buy can open a new long or close an existing short. A sell can open a new short or close an existing long. The direction of the order alone does not say whether a position is created. And "opening" does not mean a first-time trader. It means that the position is new.
If the buyer opens a long and the seller opens a short, one new contract exists. Volume rises by one and OI rises by one. If the buyer closes a short and the seller closes a long, one contract disappears. Volume rises by one and OI falls by one.
The other two combinations only hand a position over. A new long buys from an old long who is leaving. Or an old short buys back from a new short who is arriving. Volume rises by one, and OI does not change. The person leaving does not have to find the original counterparty.
Trades can change the holders without changing OI
One futures series starts with 1,000 open contracts. Each matched trade adds to volume once. OI changes only if both sides open or both sides close. Open + close means one side opens, the other closes. All figures are contracts.
- Prior closeOpen interest the day began with
1,000OI
- Open + close60 contracts · one side opens, the other closes
1,000OI unchanged
- Both open100 contracts · two new positions
1,100OI +100
- Both close100 contracts · two positions end
1,000OI −100
- Open + close40 contracts · one side opens, the other closes
1,000OI unchanged
Traded in total300Net OI change0
Volume is 300, yet OI ends where it started. The 100 new contracts are still open. The 100 closures removed older ones. Flat OI does not mean nothing new opened.
Made-up trades in one series. A contract has a long and a short side, and OI counts it once. No lot size is used.
The ledger below shows this in one made-up series. It starts with 1,000 open contracts and zero volume. First, 60 existing longs hand their positions to new longs: volume 60, OI still 1,000. Then both sides open 100 new contracts: volume 160, OI 1,100. Then both sides close 100 other, older contracts: volume 260, OI back to 1,000.
Last, 40 existing shorts hand their positions to new shorts. Volume ends at 300 and OI ends at 1,000. We assume no expiry, exercise or correction in between.
Look at what flat OI hid. The 100 new contracts are still open. They were simply offset by 100 older ones that closed, and the handovers changed who holds some of the rest. So flat OI does not mean everything opened was closed. It does not mean the volume was all churn either. A different set of trades could end in the same totals.
For ordinary matched trades, the net change in OI equals the contracts where both sides opened, minus the contracts where both sides closed. Handovers add zero. Reported figures can also include expiry events or corrections. And you cannot work out how many positions were opened and closed inside the session just by subtracting the OI change from volume.
What do the four price and OI labels mean?
Meet the four labels and what they do not tell you.
Now compare the change in OI with the change in price. Pick one starting observation and one ending observation, and use the same pair for both. In our example, "price" means the closing price of NIFTY 50 futures, in index points, for one unchanged expiry E. It is not NIFTY spot and not an option premium.
By convention, price up with OI up is called , and price down with OI up is called . Price down with OI down is . Price up with OI down is . These names describe two measured changes. They do not tell you who started each trade, or whose actions moved the price.
Four combinations, not four compulsory stages
Compare the same contract over the same interval. Price uses green for up and red for down; OI colours describe count changes, not profit.
| Futures price changeOpen-interest change | Open-interest changeOI increases | Open-interest changeOI decreases |
|---|---|---|
| Futures price changePrice rises | OI increases A higher price with more outstanding contracts on a net basis. | OI decreases A higher price with fewer outstanding contracts on a net basis. |
| Futures price changePrice falls | OI increases A lower price with more outstanding contracts on a net basis. | OI decreases A lower price with fewer outstanding contracts on a net basis. |
Every outstanding contract has a long side and a short side. OI counts the matched contract once.
These conventional names describe the two observations. They do not identify whose trades caused the price move.
The next interval can repeat, reverse or skip a combination. A covering label does not prove a temporary rally; a buildup label does not guarantee continuation.
If price or OI is unchanged, this exact-change matrix leaves the interval unclassified. A real-data tolerance must be disclosed separately.
Take . What you actually observed is a higher price and more open contracts. Every new contract still has a long and a short. The word "long" does not mean only longs were created, or that buyers outnumbered sellers. In the count of contracts also rises, only the price is lower. The new longs on the other side exist as well.
In and , the observation is fewer open contracts. In ordinary closing trades, a contract that closes removes both its long and its short. The totals do not prove that long exits caused the fall or that short exits caused the rise. New contracts may have opened elsewhere in the same window.
These labels organise what you saw. They do not measure cash entering the market, show conviction or forecast on their own. does not mean a rally must fade. does not guarantee the move continues. Each of those would be another claim, and it needs its own evidence.
"Four phases" does not mean every contract passes through four stages in order. Intervals can repeat, skip or reverse. Up means a positive change and down means a negative change. If either price or OI did not change, keep both numbers and write "no four-cell classification". A flat reading sits outside the four cells.
How do you compare observations consistently?
Compare readings with one fixed start and end.
A label needs a baseline. Follow the same made-up NIFTY 50 futures contract, expiry E, across five consecutive sessions A to E. Each price is the daily closing price in index points. Each OI figure is the end-of-day count of open contracts. These are not live quotes, real dates or current contract terms. Session E is only a label, not an expiry event.
Session A starts at 25,000 points with OI of 1,00,000. B closes at 25,100 with 1,10,000. C closes at 25,020 with 1,18,000. D closes at 24,950 with 1,12,000. E closes at 25,080 with 1,05,000. The figure keeps the price line and the OI table together so each comparison uses matching endpoints.
Choose the window before naming a phase
The same made-up NIFTY futures contract, expiry E: closing price and end-of-day OI over five sessions.
- Expiry E futures closing priceSolid line
NIFTY futures closing price (points)
Session (equal schematic spacing)
The price axis starts at 24,900 points to show the changes clearly. OI is a separate contract count in the ledger; it is not plotted on the price axis. Connecting lines join snapshots, not an intraday price path.
| Series | Session (equal schematic spacing) | NIFTY futures closing price (points) |
|---|---|---|
| Expiry E futures closing price | 0 | 25,000 |
| Expiry E futures closing price | 1 | 25,100 |
| Expiry E futures closing price | 2 | 25,020 |
| Expiry E futures closing price | 3 | 24,950 |
| Expiry E futures closing price | 4 | 25,080 |
| Session | Close · points | OI · contracts |
|---|---|---|
| A | 25,000 | 1,00,000 |
| B | 25,100 | 1,10,000 |
| C | 25,020 | 1,18,000 |
| D | 24,950 | 1,12,000 |
| E | 25,080 | 1,05,000 |
A→E: price +80 points and OI +5,000 contracts gives the conventional label. D→E: price +130 points and OI −7,000 contracts gives the label. Both descriptions end at E; their starting snapshots differ.
Made-up values. A to E are session labels, and E is also the expiry label, not an expiry event. Price is the futures close, not spot or LTP.
Inspect the interval arithmetic
Subtract each starting snapshot from its matching ending snapshot. Use the same endpoints for price and OI.
A→B: price +100 points; OI +10,000 contracts → long buildupB→C: price −80 points; OI +8,000 contracts → short buildupC→D: price −70 points; OI −6,000 contracts → long unwindingD→E: price +130 points; OI −7,000 contracts → short coveringA→E: (25,080 − 25,000) / 25,000 × 100 = +0.32%A→E: (1,05,000 − 1,00,000) / 1,00,000 × 100 = +5%
A to B adds 100 points and 10,000 contracts: . B to C loses 80 points and adds 8,000 contracts: . C to D loses 70 points and 6,000 contracts: . D to E gains 130 points and loses 7,000 contracts: . Four steps, four combinations. That does not make a repeating cycle.
Now compare A to E. Price is up 80 points, or 0.32%, and OI is up 5,000 contracts, or 5%. That window reads as , although D to E reads as . Both are right. They end at the same snapshot but start at different ones. A label without its start and end is incomplete.
The step changes add up: 10,000 + 8,000 − 6,000 − 7,000 = 5,000. Changes measured against A are 10,000, 18,000, 12,000 and 5,000, and adding those gives 45,000, which counts overlapping windows again and again. Check whether a provider's change field is a step or a total against the previous close. Never add repeated refreshes. Volume works the same way: each reading is a running total, not new trades.
On a real screen, record the exchange, the provider, the full contract, the price field and the OI unit. Also record both observation times with timezone, the source's as-of time, any delay, and whether the figure is preliminary or final. A refreshed page can keep a stale last trade. Closing price, LTP and settlement price are different fields, so pick one and stay with it.
Percentage change is the change divided by the starting value, times 100. If the starting OI is zero, show the absolute change and write "percentage change undefined". Classify using unrounded values. If you use a tolerance for "flat", decide it before you look. If data is missing, stale or mismatched, write "insufficient comparable data" rather than force a label.
Does the same reading work for options?
Carry the same reading over to option contracts.
A futures series is fixed by its expiry. An option series needs two more fields: the strike, and whether it is a Call (CE) or a Put (PE). The 25,000 Call for expiry E and the 25,000 Put for expiry E are different contracts. The same Call for expiry F is another one again. Keep their counts apart unless you choose to add them and say so. "NIFTY OI" on its own leaves the scope unclear.
In our core example, the price is the futures price in index points. For an option, the price is the premium of the chosen option, in rupees per unit. Name that field before you give a label. Do not swap in NIFTY spot, another expiry or the premium of a neighbouring strike. Both observations must use the same contract, the same price and the same interval.
An option premium is moved by the index, by implied volatility and by time left. These can pull in different directions. A Call premium can fall while NIFTY rises, if falling IV or lost time outweighs the index move. This is a possibility, not a calculated case. A Put premium rising is also not the same thing as NIFTY falling. The figure holds the other inputs fixed so you can see each pull on its own.
What moves an option premium?
One Call and one Put, same strike and expiry. Each row changes one input and holds the rest fixed. These are tendencies, not calculated premiums.
Underlying spot rises
↑CallTends to rise
↓PutTends to fall
Underlying spot falls
↓CallTends to fall
↑PutTends to rise
Less time remains
↘CallTime value can shrink
↘PutTime value can shrink
Implied volatility rises
↑CallTends to rise
↑PutTends to rise
A premium can fall because the index, the time left or IV changed. So "premium down, OI up" does not tell you who is informed, and it is not a forecast for the index.
A picture of tendencies, not a priced model. Rates, dividends, spreads and quote timing also matter. Blue is a Call and beige is a Put, not profit or loss.
Premium down with OI up tells you two things about that one series: its premium fell and its open contracts rose. It does not say whether the buyers or sellers were informed, whether the positions sit inside a spread, or which input moved the premium. The label cannot become a forecast for the index. Blue marks a Call and beige marks a Put, not who gains or loses.
When you look at several strikes, go back to the checks in lesson 21: the expiry, the Call or Put side, the OI unit and the timestamps. A chain total answers a real question, but a different expiry or strike range is a different question. Write down the filter, and do not set an unlabelled total beside a single-contract reading.
What do the counts leave unknown?
Separate what the counts show from what is unknown.
You can see more open contracts, but not the accounts behind them. The count is real. The explanation is not in it. Price, OI and volume do not show anyone's entry cost, whether a trade opened or closed, the motive, the hedge or the full portfolio. That is a limit of the published feed. It does not mean the exchange keeps no such records.
A big OI number at one strike shows where contracts remain open, inside your chosen scope. It does not prove that sellers will defend the level, that it becomes support or resistance, or that anyone must exit at a given price. A short option may be one leg of a hedge or a spread. Without the linked positions, a story about intentions is only a hypothesis.
What do the snapshots show, and what stays unknown?
Keep what you measured apart from the story you add to it.
| Observation available | Still unknown from price and OI alone |
|---|---|
| Price and OI rose in the chosen interval | Which traders opened; their motives, leverage or future exits |
| OI fell in one selected expiry | How much was rolled, or whether total market exposure fell |
| Call or Put premium moved with OI | Whether the move came from spot, time, IV or changing quotes |
| A strike has high OI | Whether it will act as support, resistance or a guaranteed reversal level |
| A conventional phase label fits the two signs | The next price direction or a trader’s probability of profit |
A phase name is a short description of one contract over one window. Check scope, expiry, times and execution before you form a trading view.
OI counts open contracts, each with a long and a short side. It does not show a full position, a hedge elsewhere or who will gain.
Open positions are also not the orders available right now. High OI does not guarantee a narrow spread, enough quantity for your order, or a fill near the last premium. Check the current quotes, depth and time stamps for the size you need. Lesson 21 covers these fields. OI describes participation. The live order book describes what you can trade.
NSE's participant-wise OI report gives a broader view. The 25 September 2026 file we checked has rows for Client, DII, FII and Pro, with Index and Stock futures, and Call and Put long and short columns, counted in contracts. It has no symbol, expiry, strike, trader ID or entry cost. So it cannot show an FII's position at one NIFTY strike. Keep the label "Client" as published, and do not read it as a retail-only group.
Treat the file as a dated after-the-close report. Note the report date and when it actually reached you. We did not verify the current release time, so we promise none. Category totals do not describe every member. And none of this makes market activity meaningless. It means the four labels alone do not prove a cause, a motive or a forecast.
How should you use OI without over-reading it?
Keep a careful record and avoid over-reading the labels.
Rollover means closing a position in one expiry and opening it in a later expiry. A long sells the near contract and buys the later one. A short buys back the near one and sells the later one. The old contract does not gain a new expiry. After expiry, open positions simply end, so OI that vanishes at expiry does not need a replacement elsewhere.
Here is a separate made-up example. Both snapshots come before the near expiry ends. Near-expiry OI falls from 1,00,000 to 80,000 contracts. Next-expiry OI rises from 40,000 to 60,000. The two together stay at 1,40,000. That fits rollover, but unrelated traders closing near-expiry pairs and opening later pairs would give the same totals. It does not prove the same accounts moved, and it does not measure 20,000 rolled contracts.
An expiry can lose OI while the combined total stays flat
A separate made-up example: two expiries, before and after.
Near expiry · −20,000
Before1,00,000- After80,000
Next expiry · +20,000
Before40,000- After60,000
Both expiries combined · 0
Before1,40,000- After1,40,000
The offsetting changes fit positions moving from the near to the next expiry. A falling near-expiry OI alone does not show people leaving the market.
Equal changes do not prove the same traders rolled. Unrelated closures and openings give the same totals.
A real roll may not change either OI at all. A trader closing a near-expiry long may sell to someone opening a long, so the contract is only handed over. The later-expiry purchase may meet a seller who is closing, another handover. Each leg's OI can stay the same. What the other side is doing decides the count. So even equal and opposite changes cannot audit one trader's roll. Flat combined OI does not mean unchanged exposure either.
For your journal, save the provider, exchange, instrument, underlying and exact expiry. For options add the strike and CE or PE. Note whether OI and volume are contracts or units, both timestamps with timezone, and any delay. Name the price field and its time. Write down the OI baseline and the volume interval. A comparison with the previous close and two snapshots in one day are different windows.
Choose your tolerance for "flat" before you see the result. If a change is flat, missing, stale or under that tolerance, record "unclassified" and do not force a cell. Note expiry changes, contract switches, corporate-action adjustments and revisions separately. Do not add cumulative volume snapshots, repeated refreshes of a day-change field, or overlapping windows. Observations come first, the label second, and your explanation last, as a hypothesis.
If you test whether a rule predicts a later outcome, fix the rule and the horizon in advance. Keep the failures, the exclusions and the unclassified cases. Keep time order, and never let later information leak into an earlier entry. Test on data that did not choose the rule, against a stated baseline. Any claim about trading results also needs real fills, spread, charges and slippage. A journal builds consistency. Four weeks or forty sessions do not guarantee an edge.
The next lesson combines Put and Call activity into the put-call ratio (PCR). Carry this habit forward. First decide whether the ratio uses OI or volume, which underlying, which expiries and strikes, and what time. A ratio inherits the scope and limits of its inputs. Find out what it counts before you ask what it means.
Before you go
Check yourself
5 questions from this lesson. Pick an answer to see why, then slide on to the next.
Question 1 of 5
Questions readers ask
Common questions
What is the difference between open interest and volume?
Volume counts contracts traded over a period. OI counts contracts still open at a moment, once per long and short pair. A trade changes OI depending on whether both sides open or close. Always state the unit, the time and the volume period.
Does rising OI mean more buyers than sellers?
No. Every new open contract has a long and a short side. Rising OI means more open contracts. It does not mean more longs, fresh cash or conviction. A price and OI label is a convention for two measured changes.
Can OI stay flat while new positions remain open?
Yes. New pairs can stay open while other older pairs close, so the net count does not move. Handovers also leave OI unchanged. Equal ending OI does not show what opened and closed along the way.
How is the four-phase OI label calculated?
Compare the price and the same contract’s OI between the same two observations. The signs of the two changes pick the cell. If a change is flat, missing or stale, record it as unclassified under a rule set beforehand. Another start point may give another label.
Does declining near-expiry OI prove rollover?
No. Closing trades and expiry can lower the count with nothing replacing it. A rise in another expiry fits a roll but cannot show the same accounts or the rolled quantity. A real roll can also leave OI unchanged through handovers.
Can an option premium/OI label predict NIFTY direction?
Not by itself. Name the expiry, strike and Call or Put. The price arrow describes that premium. The index, IV and time left can offset each other, so a premium move is not an index move.
Does high OI at a strike guarantee support, resistance or liquidity?
No. OI records open contracts. It is not a duty to defend a price and not an order you can trade against now. Check the bid, ask, spread, depth and time separately.
Can participant-wise OI identify FII positions at a NIFTY strike?
No. The NSE category file groups Client, DII, FII and Pro across broad futures and option columns. It has no symbol, expiry, strike or trader ID. Note its report date and when it reached you. It is not a live strike-level feed.