Nifty 50 SIP Returns: Every 5-, 10- and 20-Year SIP Since 1999, Tested

The 20-year Nifty 50 SIP that ended last month was the worst 20-year SIP since the index's total-return record began. It still turned ₹24 lakh of ₹10,000 monthly instalments into ₹82 lakh, a return of 11.05% a year with dividends reinvested. The same money in a bank fixed deposit would have grown to about ₹50 lakh.

Watch it play outAll 88 twenty-year Nifty SIPs beat a bank FD
0 twenty-year SIPs₹10,000 a month, one for every start month₹24 lakh in, ₹82 lakh to ₹1.45 crore outEvery one beat an FD SIP over the same years11.05% a yearthe worst one, which ended last month

Start a ₹10,000 monthly Nifty 50 SIP in any month since mid-1999 and hold it 20 years. There have been 88 such SIPs.

Each put in ₹24 lakh. They ended with between ₹82 lakh and ₹1.45 crore.

Every single one beat the same money placed in bank fixed deposits.

The worst was the most recent, cut short by this year's fall. It still made 11.05% a year.

Source: NSE Indices Nifty 50 Total Return Index; RBI Handbook of Statistics; Yield Theory calculations

Shorter SIPs are a different story. We took the Nifty 50 Total Return Index from NSE Indices, which starts in mid-1999 and includes dividends, and ran a ₹10,000 SIP from every possible start month: 268 five-year SIPs, 208 ten-year SIPs, 148 fifteen-year SIPs and 88 twenty-year SIPs. Then we put each one against a bank FD and against the S&P 500 bought in rupees.

Make a guess
Of the 208 ten-year Nifty 50 SIPs since 1999, what did the worst one return a year?

Tap your guess

Spot on.Close, a little low.Close, a little high.Way off, too low.Way off, too high. It was 4.7%.

4.7% a year, below the 7.6% an FD SIP made over the same decade. It started in March 2010 and ended in March 2020, the month of the Covid crash.

Worst 10-year SIP since 19994.7%
Median 10-year SIP13.7%
Source: NSE Indices; Yield Theory calculations

The short version

  • Time fixed almost everything. 89% of five-year SIPs beat an FD. At ten years it was 98%, at fifteen 99%, and at twenty years every one.
  • The typical result barely changed with time; the range did. The median SIP made about 13.5% to 14.4% a year at every horizon. What shrank was the spread: from −2.8% to 48.5% at five years, to 11.1% to 15.7% at twenty.
  • The worst SIPs had bad end dates, not bad start dates. The worst 5-, 10- and 15-year SIPs all ended in March 2020. The worst 10-year SIP went from 4.7% to 12.9% a year by staying in 12 more months.
  • Recent SIPs look weak. After this year's fall, the last five years returned 4.68% a year, below an FD. The last ten returned 10.23%.
  • The S&P 500 beat the Nifty in rupees, mostly because of the rupee. In 76% of ten-year windows a rupee SIP in the S&P 500 did better. Take out the rupee's fall and the S&P 500 won only 42% of them.
712 Nifty 50 SIPs, one for every start month
89%5-year SIPs that beat an FD239 of 268
98%10-year SIPs that beat an FD204 of 208
100%20-year SIPs that beat an FD88 of 88
13.7%Median 10-year SIP returnXIRR, dividends reinvested

How we tested it

  • Index: the Nifty 50 Total Return Index, daily closes from NSE Indices, from its first value on June 30, 1999. Total return means dividends are reinvested, which is what an index fund investor actually gets. We used the last close of each month.
  • The SIP: ₹10,000 invested at each month-end close, with the whole holding valued one month after the last instalment. A 10-year SIP is 120 instalments, or ₹12 lakh. Returns are XIRR, the annual rate that turns those instalments into the final value. Every SIP is valued no later than the end of last month.
  • The FD benchmark: the same ₹10,000 a month placed in a deposit earning the five major banks' 1–3 year term deposit rate (the midpoint of RBI's published range), resetting every April to the latest rate. Rates come from the RBI Handbook of Statistics: the current Table 59, the 2018 edition and the 2010 edition for the years before 2003. The midpoint ran from 4.6% to 10% over the period.
  • The S&P 500: the S&P 500 Total Return index in dollars, converted to rupees at the month-end USD/INR rate published by the US Federal Reserve.

Both returns are before tax and fund costs. We cover what that changes in the caveats below.

Result 1: the longer the SIP, the narrower the range

SIP lengthSIPs testedInvestedWorstMedianBestBeat an FDAbove 12% a year
5 years268₹6 lakh−2.76%14.39%48.48%89%66%
10 years208₹12 lakh4.67%13.68%23.23%98%72%
15 years148₹18 lakh7.38%13.47%17.57%99%85%
20 years88₹24 lakh11.05%13.69%15.73%100%91%

XIRR of a ₹10,000 monthly SIP in the Nifty 50 Total Return Index, every start month from June 1999.

The median barely moves. Whether you held five years or twenty, the middle SIP made close to 14% a year. Holding longer didn't raise the typical return; it removed the bad outcomes. Only one of 268 five-year SIPs lost money, but 29 of them trailed an FD. By twenty years, the worst SIP made more than the median FD return at any horizon.

The median stayed near 14%; the worst case climbed
Worst SIPMedian SIPBest SIP
−101020304050%0
−2.8%+14.4%+48.5%
5 yearsWorst SIP−2.8%Median SIP+14.4%Best SIP+48.5%
+4.7%+13.7%+23.2%
10 yearsWorst SIP+4.7%Median SIP+13.7%Best SIP+23.2%
+7.4%+13.5%+17.6%
15 yearsWorst SIP+7.4%Median SIP+13.5%Best SIP+17.6%
+11.1%+13.7%+15.7%
20 yearsWorst SIP+11.1%Median SIP+13.7%Best SIP+15.7%
5 years10 years15 years20 years
XIRR of a ₹10,000 monthly Nifty 50 SIP, dividends reinvested, across every start month since June 1999.Source: NSE Indices; Yield Theory calculations

In rupees, that range is the difference between a comfortable result and a great one:

SIP lengthInvestedWorst ending valueMedian ending valueBest ending value
5 years₹6 lakh₹5.59 lakh₹8.61 lakh₹19.19 lakh
10 years₹12 lakh₹15.24 lakh₹24.52 lakh₹41.01 lakh
15 years₹18 lakh₹32.28 lakh₹54.02 lakh₹77.12 lakh
20 years₹24 lakh₹82.01 lakh₹1.13 crore₹1.45 crore

The best SIPs all started in the early 2000s, near the bottom after the dot-com crash, and rode the 2003–2007 boom. The best five-year SIP started in October 2002 and made 48.5% a year. Nobody planned that; it is what buying through a slump looks like in hindsight.

Every 10-year Nifty 50 SIP since 1999, by start month
0510152025%20012004200620072010201120132016Ended in the Covid crash: 4.7%Best: 23.2%10.2%Nifty 50 SIP, a year6.1%Bank FD SIP, a yearAug 199920.3%7.0%Jan 200020.4%7.0%Jun 200021.4%7.0%Nov 200022.3%7.0%Apr 200120.9%7.0%Sep 200116.9%7.1%Feb 200217.0%7.3%Jul 200215.1%7.4%Dec 200215.4%7.6%May 200313.7%7.8%Oct 200312.9%7.9%Mar 200412.8%8.0%Aug 200414.6%8.2%Jan 200515.0%8.3%Jun 200512.7%8.4%Nov 200510.6%8.4%Apr 20069.5%8.5%Sep 200610.6%8.4%Feb 200710.5%8.4%Jul 200712.4%8.3%Dec 200712.7%8.2%May 200812.7%8.1%Oct 200811.5%8.0%Mar 200912.2%7.8%Aug 200910.4%7.7%Jan 201011.2%7.7%Jun 20107.8%7.5%Nov 201011.7%7.3%Apr 201113.3%7.2%Sep 201116.0%6.9%Feb 201214.2%6.8%Jul 201213.7%6.6%Dec 201213.8%6.4%May 201313.3%6.2%Oct 201312.9%6.1%Mar 201414.9%6.0%Aug 201416.4%6.0%Jan 201514.3%6.0%Jun 201515.1%6.0%Nov 201514.8%6.1%Apr 201612.2%6.1%Sep 201610.2%6.1%
XIRR of ₹10,000 invested at each month-end for 120 months, valued a month after the last instalment. Nifty 50 Total Return Index; FD at the five major banks' 1–3 year deposit rate.Source: NSE Indices; RBI Handbook of Statistics; Yield Theory calculations

The chart shows how rarely a decade-long SIP fell below the FD line: four start months out of 208. One started in February 2006 and ended in the slump of early 2016. The other three started in spring 2010 and ended in the Covid crash.

Result 2: the worst start dates were really bad end dates

Here are the six worst SIPs at each length. Look at the right-hand column.

SIP lengthWorst start monthEndedReturn a year
5 yearsMarch 2015March 2020−2.76%
5 yearsFebruary 2004February 20090.71%
5 yearsMay 2015May 20201.30%
10 yearsMarch 2010March 20204.67%
10 yearsMay 2010May 20206.54%
10 yearsApril 2010April 20207.19%
15 yearsMarch 2005March 20207.38%
15 yearsMay 2005May 20208.37%
15 yearsApril 2005April 20208.88%

Nothing was wrong with March 2010 or March 2005 as start months. Those SIPs look bad because they were measured at the bottom of a crash: the index fell 28% between January and March 2020. A SIP's result depends heavily on the market's level when you stop, because by then most of your money is invested.

The worst 10-year SIP shows how quickly that changes.

The worst 10-year SIP, given a little more time
Stopped in March 2020 (120 instalments)₹15.2 lakh · 4.7% a year
Six months later (126 instalments)₹20.7 lakh · 9.1% a year
A year later (132 instalments)₹27.9 lakh · 12.9% a year

₹10,000 a month into the Nifty 50 Total Return Index from March 2010. The extra instalments add only ₹60,000 and ₹1.2 lakh.

Twelve more instalments, ₹1.2 lakh of new money, took the holding from ₹15.2 lakh to ₹27.9 lakh and the return from 4.7% to 12.9% a year. The investor who stopped in the crash locked in the worst decade in the data. The one who kept going got roughly the median.

That is the practical lesson of the whole study. You can't choose a good start month in advance, and you don't need to. What you can control is not being forced to sell in a crash, which means keeping an emergency fund outside your SIP and starting to move money out gradually a few years before you need it.

Result 3: what recent SIPs returned

The numbers above cover all of history. Most people searching "Nifty 50 SIP returns" want the last few years, and those look weaker, because the index is about 13% below its high from the start of this year.

SIP that ended last monthInvestedValue nowNifty 50 return a yearFD SIP return a year
Last 3 years₹3.6 lakh₹3.53 lakh−1.30%6.74%
Last 5 years₹6 lakh₹6.75 lakh4.68%6.41%
Last 10 years₹12 lakh₹20.40 lakh10.23%6.12%
Last 15 years₹18 lakh₹44.61 lakh11.24%6.51%
Last 20 years₹24 lakh₹82.01 lakh11.05%6.87%

Nifty 50 Total Return Index, valued at the end of last month.

The last 10-year SIP sits below the historical middle
All 208 ten-year SIPs since 19995–23%
FD SIP, last 10 years6.1%
Nifty SIP, last 10 years10.2%
Median 10-year SIP13.7%
XIRR of a ₹10,000 monthly Nifty 50 SIP held 10 years, dividends reinvested.Source: NSE Indices; RBI Handbook of Statistics; Yield Theory calculations

A five-year SIP returning less than an FD is unusual but not rare: it happened in 29 of 268 windows, all of them five-year stretches that ended in or after a fall, in 2008–09, 2011–14, 2016, 2020 and now. The ten-year number, 10.23%, is below the historical median of 13.68% but still about four points a year ahead of an FD.

Dividends, and what Nifty BeES actually gave you

The Nifty 50 figure on most news sites and charts is the price index, which ignores dividends. That understates a SIP's return by a steady amount. Across every ten-year SIP since the price series on Yahoo Finance begins in 2007, the price index came in 1.3 to 1.4 points a year below the total return index. For the last ten years, that is 8.86% against 10.23%, or ₹18.97 lakh against ₹20.40 lakh.

Nifty BeES, the oldest Nifty 50 ETF, tracked the total return index closely. A ₹10,000 monthly SIP in it over the last ten years returned 10.19% a year, against 10.23% for the index. Over the last five years it returned 4.69%. A low-cost Nifty 50 index fund should land within a few tenths of a point of the index figures here; an active large-cap fund has to clear them.

Last 10 years: the price index understates what a SIP earned
24681012%0
8.86%
10.19%
10.23%
Nifty 50 price indexNifty BeES ETFNifty 50 Total Return Index
XIRR of a ₹10,000 monthly SIP held for the 10 years to the end of last month.Source: NSE Indices; Yahoo Finance; Yield Theory calculations

Nifty 50 vs S&P 500 in rupees

The same ₹10,000 a month in the S&P 500, converted to rupees each month, did much better recently:

SIP that ended last monthNifty 50S&P 500 in rupeesS&P 500 in dollars
Last 3 years−1.30%27.10%19.54%
Last 5 years4.68%24.19%18.11%
Last 10 years10.23%20.81%15.82%
Last 15 years11.24%19.50%14.92%
Last 20 years11.05%18.54%13.84%
Since mid-199913.11%15.18%11.52%

Both indices include reinvested dividends. "In dollars" is the same rupee SIP with the exchange rate frozen, which isolates the S&P 500's own return.

Over the last ten years, ₹12 lakh in the S&P 500 became ₹35.97 lakh, against ₹20.40 lakh in the Nifty 50. Over twenty years, ₹24 lakh became ₹2.06 crore against ₹82 lakh. But a large part of that gap is the rupee, not American companies. The dollar went from ₹43.45 in mid-1999 to ₹95.83 at the end of last month, a 2.95% annual fall in the rupee that a rupee investor in US stocks collects automatically.

Last 20 years: where the S&P 500's lead came from
Nifty 50 SIP11.05%
S&P 500's edge in dollars+2.79%= 13.84%
Rupee's fall+4.70%= 18.54%
S&P 500 SIP in rupees18.54%
XIRR of a ₹10,000 monthly SIP for the 20 years to the end of last month. The rupee's fall supplied more of the lead than the US market did.Source: NSE Indices; S&P Dow Jones Indices via Yahoo Finance; Federal Reserve H.10 via FRED; Yield Theory calculations

Across every window, the pattern holds. A rupee SIP in the S&P 500 beat the Nifty 50 in 76% of ten-year windows and 77% of twenty-year windows. With the exchange rate frozen, the S&P 500 won only 42% of ten-year windows and 18% of twenty-year windows. Measured across all windows, the rupee added a median of 3.5 to 4.2 points a year to the S&P 500's return, depending on the SIP length.

In rupees, the S&P 500 won every 10-year SIP started after mid-2003
−50510152025%2001200420062007201020112013201620.8%S&P 500 in rupees10.2%Nifty 5015.8%S&P 500 if the rupee never movedAug 19990.8%20.3%−0.6%Jan 20000.8%20.4%0.6%Jun 20000.3%21.4%0.0%Nov 20003.0%22.3%2.9%Apr 20015.1%20.9%5.7%Sep 20013.5%16.9%2.1%Feb 20027.0%17.0%5.7%Jul 20029.4%15.1%5.7%Dec 20029.4%15.4%6.0%May 200312.0%13.7%8.1%Oct 200314.5%12.9%9.1%Mar 200414.8%12.8%10.0%Aug 200416.0%14.6%10.9%Jan 200515.6%15.0%10.5%Jun 200516.1%12.7%10.8%Nov 200516.6%10.6%10.6%Apr 200615.9%9.5%10.2%Sep 200616.4%10.6%10.8%Feb 200717.6%10.5%12.2%Jul 200717.2%12.4%12.7%Dec 200718.0%12.7%13.9%May 200818.5%12.7%13.8%Oct 200819.4%11.5%13.3%Mar 200917.8%12.2%13.1%Aug 200917.7%10.4%12.7%Jan 201018.4%11.2%13.7%Jun 201017.4%7.8%12.1%Nov 201018.5%11.7%14.1%Apr 201119.9%13.3%15.9%Sep 201119.0%16.0%15.6%Feb 201218.2%14.2%14.8%Jul 201216.8%13.7%12.8%Dec 201214.9%13.8%10.5%May 201315.3%13.3%11.3%Oct 201314.4%12.9%10.5%Mar 201417.7%14.9%14.0%Aug 201418.1%16.4%14.6%Jan 201519.0%14.3%15.1%Jun 201518.3%15.1%14.8%Nov 201519.8%14.8%15.7%Apr 201620.7%12.2%15.7%Sep 201620.8%10.2%15.8%
XIRR of a ₹10,000 monthly SIP held 10 years, by start month. Both indices include reinvested dividends.Source: NSE Indices; S&P Dow Jones Indices via Yahoo Finance; Federal Reserve H.10 via FRED; Yield Theory calculations

The other thing the chart shows is that the S&P 500's lead is recent. Every ten-year SIP started from mid-1999 to mid-2003 did better in the Nifty 50. From mid-1999 to mid-2009, a lump sum in the Nifty 50 compounded at 15.5% a year while the S&P 500 lost 1.3% a year in rupees. Anyone extrapolating the last 15 years of US returns is extrapolating the best stretch for US stocks and a weak one for the rupee at the same time.

Check the claimThree SIP claims, tested

Questions readers ask

What are Nifty 50 SIP returns over the last 10 years?

A ₹10,000 monthly SIP in the Nifty 50 Total Return Index over the 10 years to the end of last month returned 10.23% a year, turning ₹12 lakh into ₹20.40 lakh. The median across every 10-year SIP since 1999 is 13.68%.

What are Nifty 50 SIP returns over the last 20 years?

11.05% a year: ₹24 lakh became ₹82.01 lakh. That was the lowest of all 88 twenty-year SIPs since 1999, which ranged up to 15.73%.

What are Nifty 50 SIP returns over the last 5 years?

4.68% a year, ₹6 lakh to ₹6.75 lakh, below the 6.41% an FD SIP earned. This year's fall in the index dragged it down.

What are Nifty BeES SIP returns?

A ₹10,000 monthly SIP in Nifty BeES returned 10.19% a year over the last 10 years and 4.69% over the last 5, within a few hundredths of a point of the Nifty 50 Total Return Index.

Nifty 50 vs S&P 500 returns: which did better over the last 10 years?

The S&P 500, in rupees. A 10-year rupee SIP made 20.81% a year in the S&P 500 against 10.23% in the Nifty 50; about 5 points of that gap came from the rupee's fall against the dollar.

Is there a Nifty 50 SIP calculator?

Our if-you-invested calculators for Nifty 50 stocks show what a past investment in each company would be worth today. For the index itself, the tables above cover every 5-, 10-, 15- and 20-year window since 1999.

Caveats

  • Tax and costs. All returns are before tax and fees. FD interest is taxed every year at your slab rate, while equity gains are taxed only when you sell and at lower rates, so after tax the gap over FDs is wider than shown. Index funds charge a small annual fee.
  • The FD proxy. We used the midpoint of the five major banks' 1–3 year rate, reset each April. Senior-citizen rates, small finance banks and longer tenors would have paid somewhat more.
  • The S&P 500 is not that easy to buy. The index assumes dividends are reinvested in full; an Indian investor loses 25% of US dividends to withholding tax, which trims roughly 0.3 to 0.5 points a year. Overseas investing also runs through the Liberalised Remittance Scheme or Indian funds that have at times stopped taking new money.
  • Month-end prices. Real SIPs buy on a chosen date, often early in the month. That changes individual results slightly, not the pattern.

Our read

A Nifty 50 SIP is a decade-long commitment, and it has paid for it. Held ten years or more, it beat an FD in almost every window since 1999, and the typical result was close to 14% a year. The bad outcomes in the data didn't come from picking the wrong month to start. They came from stopping in a crash. If this year's fall has you thinking about pausing, the March 2010 SIP is the case study: stopping at the bottom locked in 4.7% a year, and staying one more year delivered 12.9%.

On the S&P 500, the honest reading is that Indian investors in US stocks were paid mostly for owning dollars. That is a real benefit worth having, since your future costs in dollars, from imported goods to a child's education abroad, rise when the rupee falls. But it argues for holding a US slice as a currency hedge, not for abandoning the Nifty because of the last ten years. The decade before that went the other way.

What the data says to do with a Nifty SIP
Supported by 27 years of data
  • Hold for 10 years or more
  • Keep instalments going through a crash
  • Use a low-cost index fund or ETF
  • Hold a US slice as a rupee hedge
Not supported
  • Waiting for a better month to start
  • Pausing after a bad year
  • Judging a SIP on its last 3 to 5 years
  • Swapping India for the US on the last decade alone

Our India markets hub collects our research on Indian stocks, the rupee and the RBI.

Go deeper with member research

Keep reading

Members-only research

Members are reading this

The thesis, the numbers behind it, and what would break it. Full access is $39 a month.

Read the full research

Get a full issue free

Enter your email and we'll send you a complete Yield Theory issue, then the opening of each new one. Unsubscribe anytime.