What is the Nifty 50?

Start with the thing the usual description leaves out. Almost every explanation of this index tells you that it holds 50 large companies and acts as a barometer of the market, which is true, and which describes it as though its job were finished once the number is published. It is not. Publishing the number is where the Nifty 50 starts being different from every other statistic in Indian finance. Vast quantities of money are contractually attached to it. Funds are obliged to hold it. Contracts settle against it. A volatility index is computed out of its option prices. The number is not only a report on the market, it is a term in a very large number of agreements, and that changes what it is.

The distinction matters because it decides which questions have answers. A barometer can be wrong without consequence; nothing depends on it. A settlement price cannot be wrong in the same way, because being right is not what it is for. It is for being determinate: computed from a published rule, from prices anyone can see, at a time everyone agreed to in advance, so that when it prints, the question of who pays whom is already answered. Everything that looks like pedantry in the specification below is there to serve that. The narrow eligible universe, the liquidity screen, the four weeks of notice before a change, the fixed effective date: none of these make the index a better description of the economy. They make it a harder number to argue with.

So this guide takes the parts that are specific to a benchmark people execute against. It does not re-run the mechanics that every large cap index shares, because those are set out in full elsewhere: free float weighting, the divisor, why a share split moves the number by exactly nothing, why the index can close green on a day most of its own companies closed red, and why quoting a long-run chart as the performance of a group of companies is a category error are all worked through, with the arithmetic done, in the Sensex guide. Take those as read. What follows is the specification first, and then the four things that only happen to an index once the market starts settling against it.

Two entries in that table are worth pausing on, because they are the ones that make this index what it is and they are the two that most descriptions omit entirely. The derivatives gate and the liquidity gate both sit in front of the size test, and neither of them asks how big anything is. Read in order, the rule does not say "take the 50 largest". It says: take the companies that can be traded in size and that already have a derivatives market, and then, among those, take the 50 largest. That is a different instruction, it produces a different list, and the next section does the arithmetic on the boundary to show how different.

"Who decides what goes in the Nifty 50" has an honest answer and a flattering one. The flattering answer is that a committee of experts picks India's best companies. The honest answer is that a published rule decides almost all of it, and a committee applies the rule. Nothing in the cascade asks whether a business is good, whether its accounts are clean, whether its sector has a future, or whether the price is sensible. Those are all questions a person might reasonably expect an index of the country's leading companies to have considered. It has not considered any of them. It has asked how large the tradable portion of the company is, and whether the market can absorb an order in it.

The second of those is the gate that does the real work, and it has a name most readers have never met: impact cost. It measures the price you move against yourself simply by placing a reasonably sized order, expressed as a percentage of the mid-price. A constituent must maintain an average impact cost of 0.50 percent or lower for at least 90 percent of observations over the previous six months. That is a demanding screen, and it is demanding in a direction that has nothing to do with size. A company can be enormous and thinly traded, because most of its shares are locked in a promoter's hands or a government's, and the shares that do trade change hands rarely. It will read as huge and fail the gate. Meanwhile a smaller company whose float is genuinely liquid sails through.

The reason the gate exists is the whole subject of this page. An index that anyone can execute against has to be an index anyone can replicate. A fund tracking the number must be able to buy the constituents in their weights, and a market maker quoting the index must be able to hedge in the underlying. Admit a name that cannot be traded in size and you have written a specification nobody can fill. So the liquidity screen is not a quality filter that happens to be about trading, it is the load-bearing member: it is what makes the number executable, and everything downstream, the derivatives, the passive money, the volatility index, rests on it. The figure below runs the rule across the twelve candidates either side of the cut, which is the only place the rule ever bites.

So the honest description reverses the usual order. The rule does not rank the Nifty 100 by size and take the top 50; it removes everything that cannot be traded and then ranks what survives. Two further provisions protect that list once it exists, and both matter more than they look. The first is the replacement buffer: an incoming stock must be at least about 1.5 times the average free float size of the smallest current constituent. That is a high bar on purpose. Without it the list would flicker every six months as names a fraction of a percent apart traded places, and every flicker would be a forced trade for every fund tracking the index, which is why the boundary in practice is nothing like as tidy as the figure's ranking suggests: by the time a name is actually replaced it has usually fallen well clear of the pack. The second is the annual cap: routine replacements are limited to 10 percent of the index, five companies, in a calendar year. Both are the specification protecting the people who have to follow it, deliberately making the rule stickier than the data, which is a tell about what the index is really for.

Here is the fact that reorders everything else. For most people who encounter the Nifty 50, the index is not a benchmark they are measuring anything against. It is the underlying of a contract they have an open position in. The derivatives written on this index are the most heavily used in the country, and the retail path to them is short: an option costs a fraction of the notional it controls, it is available on a phone, and it expires this week. That is how most Indian retail traders actually meet the Nifty 50. Not as a summary of the economy. As a strike price.

The ecosystem around it has been rewritten hard since 2024, and the changes were not cosmetic. Under a SEBI framework dated 1 October 2024, each exchange may keep weekly expiries on only one benchmark index. The NSE kept the Nifty 50. Every other index on that exchange lost its weekly and moved to monthly and quarterly cycles, which means the weekly expiry risk that used to be spread across several benchmarks now lands, in its entirety, on this one. The same framework raised the minimum contract value so that lot sizes are set to keep a contract in a 15 to 20 lakh rupee band on review, effective for new contracts from 20 November 2024. Then, from 1 September 2025, the NSE moved its whole expiry cycle to Tuesday, ending roughly twenty five years of Thursday. The mechanics of the lot itself, and what a contract value in that band means for the capital behind a position, are worked through in the lot size guide.

Read that table as one object rather than seven rules and it says something specific. Regulation has concentrated the ecosystem onto this index. There is one weekly left on the exchange and it is this one; the contract behind it is larger than it was and must be funded in full; and the day it all resolves has moved. None of that forbids anything. It reprices everything. A larger contract with no borrowed leverage means more of your own capital sits behind every position and there is less slack for ordinary noise, and a single surviving weekly means the event risk that used to be spread across several expiries now arrives in one place. That is the environment in which the SEBI figure belongs, not as a scare statistic but as the base rate: about 93% of individual traders in equity derivatives made net losses over FY22 to FY24, aggregate net losses exceeding ₹1.8 lakh crore (SEBI, September 2024).

Every other day of the week the Nifty 50 is a description. It goes up, it goes down, and the number is a summary of what the free float value of 50 companies did. On the last Tuesday of a series it becomes something else entirely, and the shift is worth naming precisely, because it is the moment the difference between a benchmark and an underlying stops being philosophical. Contracts are cash settled against the index. Nobody delivers 50 baskets of shares. The exchange takes the settlement level, computes each position's value from it, and moves money. At that instant the index is not reporting on the market. It is the term in the contract that decides the payment.

The consequence is a leverage of meaning that people consistently underrate. The index does not have to move much for the number attached to it to move a great deal, because the contract multiplies it: a lot is 65 units, so every index point is 65 rupees per lot, and an option's value at expiry is the whole distance between the strike and the print. A move of a few tenths of a percent, which would be an unremarkable, unreported day in the life of a barometer, is a decisive event in the life of a contract. The figure below settles the same five positions twice, against two prints 200 points apart, and changes nothing else at all.

This is what people mean, or should mean, when they say the Nifty 50 is where risk is transferred. Risk transfer is not a metaphor for trading activity. It is the literal thing in the figure: two parties took opposite views, the index printed, and money moved from one to the other in the exact amount the level dictated. The index is the referee, and it is a referee with no discretion, which is precisely why it can be trusted with the job. It does not decide who deserved to win. It computes a number from a rule and the contract does the rest.

India VIX is quoted constantly and understood rarely, and the misunderstanding is almost always the same one: people take it for a measurement of the index, the way a thermometer measures a room. It is not. It is a price, extracted from what the market paid for this index's own options, and it exists only because that option chain is liquid enough to read. Take the chain away and there is no number. That makes India VIX the clearest possible demonstration of the argument on this page, because it is a piece of national market infrastructure that could not exist on top of an index nobody traded.

The computation is public and it is worth seeing worked, because seeing it worked is what kills the thermometer intuition for good. India VIX uses the best bid-ask quotes of out-of-the-money near and next expiry Nifty options. It locates the forward index level, picks the strike below it, takes the strip of out-of-the-money strikes on either side, and turns each option's price into a contribution to an option-implied variance. Those contributions are summed, corrected for the gap between the forward and the reference strike, computed for both expiries, interpolated to a constant 30-day horizon and annualised into a percentage. The methodology is the one the CBOE pioneered, licensed and adapted to the Nifty order book. Not one step in that sequence looks at the index's price history. The whole number is built out of what people were willing to pay, this morning, for the right to be somewhere else.

Two practical consequences follow, and the guide to India VIX takes both further than there is room for here. The first is that the number is forward-looking by construction, so treating a low print as evidence that the market has been calm gets the tense wrong: it says the market is not currently paying much for protection, which is a statement about expectations and about supply, not about the recent past. The second is subtler and follows directly from the shape in the figure. Because the near-the-money strikes carry most of the weight, India VIX is most sensitive to exactly the part of the chain where the weekly expiry churns hardest. The volatility index and the expiry cycle are not neighbours. They are reading the same book.

Twice a year the rule from the second section runs again and the list changes. For an index that is only a description, that is an administrative event: a name leaves, a name arrives, the divisor absorbs the discontinuity so the published line does not skip, and the chart carries on. For an index that passive money is contractually obliged to hold, it is nothing of the kind. It is a date on which a specific quantity of money must change hands, in a specific stock, in a specific direction, irrespective of anybody's opinion of the price.

The mechanism is not complicated, which is why it is so easy to miss. An index fund or ETF tracking the Nifty 50 promises to hold the index in its published weights. That promise is the product. So when the entrant joins at some weight, every tracking fund must own that weight of it by the effective date, and when the leaver goes, every tracking fund must be rid of it by the same date. The size of the resulting trade is not a decision. It is assets multiplied by weight. And the timing is not a decision either: the change takes effect from the trading day after the March or September expiry, and a fund measured on its tracking cannot sensibly be early or late. The figure runs that arithmetic and then does the thing that makes it bite, which is to express the answer in the only unit that matters, days of the stock's own normal volume.

It is worth being careful here rather than cynical, because the obvious conclusion is slightly wrong. This does not mean the tracking fund is being harmed: it is measured against the index, and the index buys the entrant at the same price on the same date, so the fund's tracking is fine by construction. That is the quiet part. The cost is real but it is invisible in the only metric anybody checks. It sits inside the index's own return, where nobody is looking for it, rather than in the tracking difference, where everybody is. The route by which a person actually holds this exposure, and the costs and tracking behaviour that come with it, are the subject of the guide to ETF investing in India.

A rank changed, and several thousand crore rupees moved. Nobody re-underwrote the business. That is not a scandal; it is what happens when a rule becomes a mandate.

The last property is the one with the widest reach, because unlike the others it does not require you to hold a contract or a fund. It reaches anybody who has ever quoted what the index has done. The Nifty 50 you see on the news is a price return index. It captures price movement and nothing else. When a constituent pays a dividend, its share price typically falls by roughly the dividend on the ex-date, and the index dutifully records the fall. The cash that left the company and arrived in a shareholder's account is not recorded anywhere, because the price index has no mechanism for recording it. The dividend does not go missing by accident. It is not the sort of thing this instrument can see.

NSE Indices computes a total return index as well, which assumes each dividend is reinvested into the index on the ex-date, and it is the version that answers the question people think the headline answers. The gap between them is not a matter of judgement or market conditions or anyone's forecast. It is arithmetic: over a holding period, the ratio of the total return version to the price version is approximately one plus the dividend yield, raised to the number of years. That is a small number annually and a large one eventually, and the fact that it compounds is exactly why quoting the headline over a long horizon goes so badly wrong.

Now the correction that most treatments of this leave out, and that the figure would flatter if left alone. It is true that the price index understates what a holder earned. It is also true that the total return index overstates it, because the TRI assumes every dividend is reinvested into the index instantly, in full, at no cost and with no tax. No human being has ever done that. A real holder pays something to reinvest, loses something to tax, and does it late. So the honest position is not "use the TRI and you have the truth". It is that the headline is a floor and the TRI is a ceiling, the holder's actual outcome sits between them, and the size of the gap you have just seen is the reason it is worth knowing which of the two a source is quoting at you before you compare anything to it.

Everything above converges on the question almost every reader actually arrived with: the Nifty 50 is up, so how am I doing? The honest answer is that the index has no opinion about you, and by now the reasons are all on the table rather than being a matter of taste. It holds 50 large NSE names that cleared a derivatives gate and a liquidity gate. It excludes the mid and small cap market entirely. It refreshes its list twice a year by a rule that systematically drops the faders. And the version you are quoting almost certainly leaves out every dividend those 50 companies paid. A benchmark means something only when it matches what you actually hold, and if it does not, the gap between you and it is telling you about the difference between two rules, not about the quality of your decisions.

What it is genuinely good for is a great deal, once it is read as what it is. It is a fast, reliable, thirty-year-consistent read on how the largest and most liquid slice of the Indian market is being priced. It is the reference against which large cap exposure is sold. It is the underlying that makes hedging possible at all, which is not a small public good. And because so much depends on it, its specification is unusually well written and unusually well published, which means that unlike most numbers you will be handed this year, this one can actually be checked. The table sorts the questions people bring to it into the ones it answers and the ones it does not.

If there is one habit worth taking from all of this, it is the one that produced every section above: before trusting a number, ask what rule made it and what the rule threw away. Applied here it yields the entire page. The selection rule threw away every question about quality and price, so a weight is a claim about availability and not a verdict. The price return rule threw away the dividends, so the headline is a floor. The mandate that ties passive money to the list threw away discretion, so a rank change is a cash flow. None of these are flaws; each one is the specification working exactly as written. That habit of opening a number before trusting it is not a trick specific to indices, it is the whole of the method we teach, and the most quoted figure in the country is an unusually good place to practise it.

• Read it as a rule, not a place. "The Nifty rose" means a weighted sum of 50 chosen numbers went up. It does not mean the market rose, and the two come apart more often than the phrasing admits.
• A weight is availability, not quality. The cascade never once asks whether a business is good or a price is sensible. It asks how much of the company can be bought and whether an order moves it.
• Check which version you are being quoted. Over a decade the price return number understates a holder by roughly a seventh; over three, by nearly half. The direction of that error never changes.
• Treat an effective date as a flow, not a fact. Passive money must trade on it, sized by weight, at whatever price exists, and the market has known it was coming for about four weeks.
• Remember that the level is a settlement term. On the last Tuesday it stops describing anything and starts deciding who pays whom, and the contract multiplies every point it moves.
• Match the benchmark to what you hold, or drop the comparison. A large cap price index is the wrong yardstick for most portfolios, and the wrong yardstick produces confident, wrong conclusions.

It is not a barometer that happens to be tradable. It is a contract term that happens to be readable, and almost everything odd about it follows from that order.

• NSE Indices Limited, Nifty 50 index methodology. Establishes NSE Indices as the maintainer, free float market capitalisation weighting through an investable weight factor, the base of 1,000 on 3 November 1995 against a base market capitalisation of about ₹2.06 trillion, the switch to free float on 26 June 2009, the Nifty 100 selection universe, the futures and options eligibility requirement, the 0.50 percent impact cost screen over six months, the 1.5 times replacement buffer, the annual change cap and the semi-annual review effective after the March and September expiry. Verified as of 17 July 2026; the methodology is revised periodically. niftyindices.com
• SEBI, Measures to Strengthen Equity Index Derivatives Framework. Circular dated 1 October 2024, which raised the minimum contract value so lot sizes keep the contract in the ₹15 to ₹20 lakh band (effective for new contracts from 20 November 2024) and limited each exchange to one weekly-expiry benchmark index, the rule under which the NSE retained the Nifty 50 weekly. sebi.gov.in
• NSE circulars on expiry day and lot sizes. The move of NSE derivatives expiries to Tuesday from 1 September 2025, with monthly and quarterly contracts settling on the last Tuesday and the BSE moving to Thursday over the same window; and the January 2026 index lot revision setting the Nifty 50 lot to 65 units from 75. Dated as of 17 July 2026; both are revised, so confirm at the exchange.
• NSE, India VIX computation methodology. India VIX is computed from the best bid-ask quotes of out-of-the-money near and next expiry Nifty 50 options, using the CBOE VIX methodology adapted to the Nifty order book, and expresses expected 30-day volatility as an annualised percentage. This establishes that the volatility index is derived from this index's own option order book rather than from its price history. nseindia.com
• SEBI, study of profit and loss of individual traders in the equity derivatives segment. The September 2024 study establishing that about 93% of individual traders in equity derivatives made net losses over FY22 to FY24, with aggregate net losses exceeding ₹1.8 lakh crore. sebi.gov.in
• On the figures marked Illustrative. Every option chain, settlement level, impact cost, index weight, passive asset total, traded value and dividend yield on this page is illustrative and authored to make the arithmetic legible. None of it is exchange, index or fund data, and no current index level is asserted anywhere. The rules, gates, dates and thresholds are the published ones and are sourced above.