What you're actually betting on when you buy an index fund

Boring Investing, part 1.

Money has to sit somewhere. Every place it can sit makes the same trade, priced differently: how much it grows against how much it lurches.

A fixed deposit picks the smoothest line available and pays for it with almost all of the growth. At the top tax slab, a 7% FD nets about 4.8% after tax. Inflation runs at 5–6%. So the balance goes up every year and buys slightly less every year. The safe option isn't safe. It's slow.

The thing that grows is the stock market, and the stock market halves sometimes. That's the tension this whole series exists to resolve: earning meaningfully more than an FD, something like 8–10% a year, while being able to sleep through a 40% fall. Not maximum wealth. Enough, with no drama, from one or two places that don't need thinking about.

The resolution isn't a clever product. It's understanding what you'd be buying well enough that a fall doesn't look like a malfunction. You can hold a rule through a crash. It's very hard to hold a mystery through one. So this first tutorial is about the rule.

By the end, you'll be able to say in one breath what a Nifty index fund is, why the number 11% keeps getting attached to it, what would have to go wrong for it to fail you, and what, exactly, you are betting on when you buy one.

Start with the word.

"Nifty" is a list, not a thing you can buy

The mistake everyone makes on day one is thinking Nifty is a company, or a fund, or a place. It's none of those.

Nifty is a list of companies with a number attached.

That's it. It's a scoreboard. The National Stock Exchange (NSE) keeps a list of the 50 largest companies trading on it, adds up what they're all worth, and publishes that total as a single number every few seconds. When the news says "Nifty is at 25,000," it means the combined value of those 50 companies, expressed on a scale where it started at 1,000.

You cannot buy Nifty, the same way you cannot buy "the Premier League table." What you can buy is a fund that owns every company on the list, in the same proportions as the list. That's an index fund. The fund is the thing you buy. Nifty is the recipe it follows.

A short history. Before 1994, Indian stock trading happened by open outcry on the floor of the Bombay Stock Exchange, brokers shouting at each other. After the Harshad Mehta scam of 1992 exposed how easily that could be gamed, the government set up a brand-new, fully electronic exchange: the NSE. It needed a headline number for people to follow, so in 1996 it launched an index of its 50 biggest stocks. The name is just National + F*ifty*. The starting value was set to 1,000 on 3 November 1995, which is why every long-term chart starts there. Thirty years later that 1,000 is roughly 25,000. Hold that thought.

How the list is built: the rule that does the work

The interesting part isn't that there's a list. It's how a company gets on it, because that rule silently does most of the work in an index fund.

Every listed company has a market capitalisation, or market cap: its share price multiplied by the number of shares that exist. It's what it would cost to buy the entire company at today's price. Reliance is worth around twenty lakh crore rupees. A small chemicals company might be worth two thousand crore. Line every listed company up by market cap, biggest first, and you have a ladder of about 500 rungs.

Nifty 50 is the top 50 rungs. Nifty 100 is the top 100. Nifty 500 is the top 500. Same ladder, three places to cut it.

The ladder is where the diagrams start earning their keep.

Two details of the rule matter more than they look:

1. It's weighted by size. The fund doesn't put 2% into each of 50 companies. It puts money in proportion to market cap. Reliance and HDFC Bank might be 8–10% each; the fiftieth company might be 0.4%. So "owning the Nifty 50" mostly means owning the top fifteen or so, with a long tail behind them.

(One nuance: it's actually free-float market cap, meaning only the shares that are genuinely available to trade. If a promoter family owns 70% of a company and never sells, those shares don't count toward its weight. This stops a company with a tiny public float from dominating the index just because it's nominally huge.)

2. The list rebalances itself. Twice a year, in March and September, NSE re-runs the ranking. Companies that have shrunk below the cutoff get dropped. Companies that have grown into it get added. Nobody decides this; the rule decides it.

This second point is the one to sit with. Think about what it means over twenty years. A company that's dying slides down the ladder and eventually falls off the list, and the fund sells it on the way out. A company that's growing climbs up and gets added, and the fund buys it on the way in. The index is a machine for continuously swapping out losers and swapping in winners, with no human judgment involved.

Of the original 50 companies in 1996, only about a dozen are still in the index. The rest were replaced by companies that, in many cases, didn't exist yet. Nobody had to know in 1996 that Infosys would matter or that a bank called HDFC would become the most valuable in the country. The rule found them.

That's what "boring" means here. Not that nothing happens. That you don't have to do anything.

What "pays 11%" actually means

Now the number. An index fund is often described as the boring engine that pays about 11%. But pays is a slippery word. An FD pays. It sends you interest. An index fund doesn't send you anything.

What actually happens is two things, and it helps to see them separately.

Thing one: the companies get more valuable. Over decades, the 50 largest companies in a growing country earn more profit, and their share prices rise to reflect it. The Nifty went from 1,000 to about 25,000 in thirty years. That works out to roughly 10.9% a year, compounded. This is called the price return, and it's the number the newspapers quote.

Thing two: the companies hand out cash. Most large companies pay a dividend, a slice of profit paid to shareholders each year, typically 1–1.5% of the share price. The index fund receives those dividends and, in a Growth plan, uses them to buy more units. So you own slightly more of the index every year without adding money.

Add the two together and you get the total return. NSE publishes this as a separate index, the Nifty 50 TRI (Total Return Index). Since inception it has compounded at 12.41% a year (as of 30 June 2026). Over the last twenty years, 12.44%.

So the honest expansion of "pays 11%" is: has historically compounded at 11–12% a year, mostly from prices rising and partly from dividends being reinvested, with no year guaranteed.

Here's what that gap between price and total return looks like over thirty years, and it's bigger than it sounds.

The small annual dividend, reinvested, turns a 24x outcome into a 36x outcome. That's why the Growth plan is the one to buy, not the one that pays dividends out to your bank. Paying them out breaks the reinvestment loop and gets them taxed at your slab.

One more thing about that 11–12%. It's an average that already contains every crash. 2008 is in there, when the index fell about 60%. March 2020 is in there, when it fell about 38%. If you removed the crashes, the number would be higher. If you refused to sit through them, your number would be much lower. The average is the reward for staying. (Part 2 of this series is entirely about that sentence.)

Does it really return 10% every year? No. And that "no" is the whole game.

The honest test of "pays 11%" is to stop looking at the average and look at the years. Here is what a Nifty 50 index fund actually did in each calendar year since 2008, price only.

Not one of those bars is 11%. The index cleared 10% in ten of the eighteen years, landed somewhere between zero and 10% in five, and lost money in three. One of those three was a 52% loss. The average is real, but no single year looks like the average. An FD gives you the average every year. An index fund gives you the average only if you stay long enough for the years to add up.

So the right question isn't "what does it return each year." It's "what does it return over the time I'll actually hold it." That's a different chart, and it's the most important one in this series. It takes every possible starting month since 2007 and asks: if you'd bought then and held for one year, or three, or five, or ten, what annual return did you get?

Read it top to bottom. Hold for one year and the outcome is a lottery: anywhere from losing half to gaining 83%, and you lost money one time in four. Hold for five years and the worst case shrinks to roughly breaking even. Hold for ten years and there is no losing case at all. Every ten-year window since 2007, including the one that started at the very top of the 2007 bubble, returned between 5% and 15% a year. Add dividends and every one of those windows clears 6%.

The bar doesn't just move right as you hold longer. It narrows. That narrowing is what "long term" means in practice. It isn't a slogan. It's the time it takes for the lurch to average out and the growth to remain.

The worst-case scenario, with real numbers. Suppose you did the single dumbest thing possible: put ₹10 lakh into the index in December 2007, the exact peak before the worst crash in the index's history. Here's what happened to that money, against the same ₹10 lakh in an FD.

Within a year the ₹10 lakh was ₹4.5 lakh. It spent nearly six years below what was put in. The FD was calmly ahead the entire time. If this were a story about a person, they would have sold in 2009 and told everyone for the rest of their life that the market is a casino.

And yet. By 2026 the worst-timed index investment in living memory is at ₹38.7 lakh on price alone, and about ₹49 lakh with dividends reinvested. The FD is at ₹35.6 lakh before tax and ₹24.1 lakh after it. On price alone, the worst-timed buyer only just edges past the FD before tax. Against the FD they would actually have kept, the after-tax one, they're ahead by about ₹15 lakh, or ₹25 lakh once dividends are counted. And that is the floor. Anyone who started in any other month did better.

What the fall actually looks like from inside. The last piece of honesty: how deep, and how long. This chart shows, month by month, how far the index sat below its previous high. Zero means a new record. Everything below zero is what holding felt like.

Two things to take from it. First, the index is below its previous high most of the time. That's normal, not a warning. Second, the deep falls are rare and the recoveries are complete: 55% in 2008, 29% in 2020, and each time back to a new high. In nineteen years there were exactly three stretches of more than a year underwater, and one of them was the six-year stretch you just saw.

Then why does anyone put money in an FD at all?

Because sometimes it's the right answer, and because the rest of the time nobody shows them the charts above. Four honest reasons.

1. The money has a date on it. Look back at the one-year bar: minus 52% is a possible outcome. If you need the money for a house deposit in eighteen months, or school fees next year, the index is the wrong place, full stop. The five-year bar is roughly break-even at worst; the ten-year bar has no losing case. So the rule is simple. Money with a date within five years goes in an FD. Money without a date goes in the index. That's why this series keeps a year of household spending in an FD and doesn't apologise for it.

2. The habit was formed when the maths was different. For long stretches of the 1990s and 2000s, Indian FDs paid 9% or more, and most people weren't in the top tax slab. At those numbers an FD genuinely was a fine place for surplus money. Rates fell, tax slabs rose, and the habit outlived the arithmetic that justified it.

3. Nobody sees the after-tax number. The bank shows 7%. The 31.2% tax on that interest never appears on the poster. It shows up as a line on a form the following July, mixed in with everything else, and by then it doesn't feel like a cost of the FD. Very few people have ever computed that their FD pays 4.8%.

4. Smoothness is a real product. The FD's straight line is buying something: not having to watch. Never opening the app to see minus 30%. Never having the conversation at dinner. For someone who knows they'd sell at the bottom, that's a legitimate thing to pay for, and the price is roughly 7% a year of growth forgone. The mistake isn't buying smoothness. It's buying it by default, for all your money, without knowing the price.

The whole series comes down to noticing that reason four is the only one that applies to money without a date, and that its price is high enough to be worth understanding the alternative properly. Which is what the rest of this page does.

50 vs 100 vs 500: the same bet, cut three ways

Nifty 50, Nifty 100, Nifty 500. Three names, and the natural assumption is that they're three different products to choose between. Here's the frame that makes it obvious.

Index	Companies	Share of India's total listed value	What it is
Nifty 50	Top 50	roughly 60%	The giants
Nifty 100	Top 100	roughly 70%	Giants + the next tier of large companies
Nifty 500	Top 500	over 90%	Almost the whole market: large, mid and small

The thing to notice: these aren't three different bets. Nifty 50 is inside Nifty 100, which is inside Nifty 500. When you buy a Nifty 500 fund, more than half your money is sitting in the same fifty companies a Nifty 50 fund holds, because they're the biggest and the fund is size-weighted. The other part is the tail: the 450 mid-sized and small companies below them.

Nifty 100 is barely different from Nifty 50. It adds the next fifty large companies, which are still large. Same flavour, slightly more of it. Not worth a separate decision.

Nifty 500 is different in one meaningful way: it includes the mid and small companies, at their natural weight. Historically that tail has grown faster than the giants over long periods (the Nifty 500 total return over the last ten years was 14.85% a year, against the Nifty 50's 11–13%, though a chunk of that gap is the unusually strong 2020–24 run for smaller companies). It also falls harder in crashes, because the tail is where the fragile companies live.

So the honest comparison is:

• Nifty 50: the most stable version of the bet. Slightly lower long-run return, slightly smaller falls.
• Nifty 500: the complete version of the bet. Slightly higher long-run return, slightly bigger falls, and you never have to think about "should I add a midcap fund" because it's already in there at the weight the market assigns it.

Either is fine. For the goal this series is built around, one or two places to put money and then stop thinking, 500 is the one that means you're done. If the app doesn't offer a Nifty 500 index fund with a low expense ratio, the Nifty 50 one is 90% the same position and nothing to lose sleep over.

So what is the bet, precisely? Three things.

When you buy a Nifty 500 index fund and hold it for fifteen years, you are making exactly three bets. Not one. Three.

Bet 1: India's largest companies, as a group, will earn more profit in fifteen years than they do today.

This is the macro bet. Notice it's narrower than "the Indian economy grows." GDP and the stock market are cousins, not twins. GDP includes agriculture, government spending, and the informal economy, none of which you own. The stock market is only the listed, profit-making, large-company slice. That slice can grow faster than GDP (more of the economy formalises, more companies list, margins expand) or slower.

But over long periods the two are tied by a simple chain: a bigger economy means more revenue for large companies, more revenue means more profit, more profit means higher share prices. If you believe India's formal economy will be substantially larger in 2041 than in 2026, and that the companies capturing that growth will mostly be the listed ones, this bet holds. That's the belief you're expressing with your money.

Bet 2: the rebalancing rule will keep doing its job.

This one's structural and it's free. You're betting that the mechanism above, dropping the shrinking companies and adding the growing ones, keeps running. It has run since 1996 without a human decision, and there's no reason to expect it to stop. This is the bet you're not making with an actively managed fund, where you're instead betting on one person's judgment staying sharp for two decades. The index replaces the person with a rule.

Bet 3: you won't sell during a fall.

This is the one nobody puts on the brochure, and it's the one that decides the outcome. The index has never had a ten-year window with a negative total return. The rolling ten-year return has never dropped below about 2.5% and averages 11.3%. But that's the index. Your return is the index's return only if you're still holding at the end. Sell in March 2020 at the bottom and buy back a year later and you've turned a 38% dip into a permanent 38% loss. The fund made 12%. You didn't.

That's why the whole first half of this tutorial was about what the thing is. You can hold a rule. It's much harder to hold a mystery.

What you are NOT betting on

Worth being explicit, because these are the things people worry about that don't apply:

• Not on any single company. Reliance could halve and you'd lose maybe 4% of your money. The list absorbs it.
• Not on a fund manager. There isn't one. A computer copies the list. This is why the expense ratio is 0.1–0.2% instead of 1.5%.
• Not on timing. You're not trying to buy low and sell high. You're buying the whole thing, repeatedly, and letting the rebalancing rule and thirty years do the sorting.
• Not on the fund company. Your units are registered in your name with a custodian. If the AMC or the app disappeared tomorrow, the units are still yours.

The numbers, in one place

All figures are total return (dividends reinvested), annualised. "Since 1995" is the Nifty 50's base date.

Index	Since 1995	20 years	10 years	Worst fall (approx.)
Nifty 50 TRI	12.41%	12.44%	~11–13%	−60% (2008), −38% (Mar 2020)
Nifty 500 TRI	—	—	14.85%	similar, slightly deeper
Nifty Midcap 100 TRI	—	—	17.35%	−65% or worse (2008, 2018–20)

(Nifty 50 figures as of June 2026; Nifty 500 and Midcap ten-year figures as of December 2025.)

Read that table the way a founder reads a metrics dashboard. The right column is not a risk warning. It's the cost of the middle columns. The 17% row costs you a 65% fall. The 12% row costs you a 60% fall. There is no row that costs you nothing, and if someone shows you one, it's an FD wearing a costume.

Which brings the tension from the top of this page back into focus. The FD's 4.8% is not the safe choice and the index's 12% is not the risky one. They're the same trade, growth against lurch, priced at two different points. The rest of this series is about choosing your point on that line deliberately, and then not moving.

A challenge for this week

Open Groww. Search "Nifty 50 index fund" and open the largest one. Then search "Nifty 500 index fund" and open one of those. For each, find:

1. The expense ratio (should be under 0.3% for a Direct plan).
2. The top 10 holdings and their weights.

Notice two things. The top ten in both funds are almost the same companies, in almost the same order. And in the Nifty 50 fund, those ten companies alone are around half your money. That's the size-weighting, made visible.

Then look at one more line on the page: the fund's 1-year return. Whatever it says, remind yourself it means almost nothing. You're not buying a year.

Next in the series: The fall is the price, not the risk. Why a 40% drop sits inside a 12% average, why the return exists because of the fall rather than despite it, and what the gap between the fund's return and the investor's return is made of.

Sources for the historical figures: Bajaj AMC, Nifty 50 historical returns; PrimeInvestor, Nifty 50 20-year data; The Tribune, Nifty 500 historical performance; BMS Money, Nifty performance since 1991. Market-share-of-listed-value figures are approximate and drift with the market.