What 30 years of Nifty 50 history tells us about SIP returns
The Nifty 50 began life on 21 April 1996 with a base value of 1,000 set against November 1995. By the end of that first year it had drifted to 898 — almost exactly its starting level. Thirty years later, on the last trading day of 2025, the same index closes near 24,500. Compounded, that is a price-only return of about 11.6% a year for the index itself. But almost nobody actually realised that headline return, because almost nobody held a lump sum of money in 1996 and let it sit untouched. What real Indian investors did — and continue to do — is contribute small amounts every month through a Systematic Investment Plan. This page is about what that experience actually looked like.
The backtest above answers the question that every personal-finance blog post handwaves: if you had actually run a monthly SIP through three decades of Indian equity history, what would you have? A ₹5,000 monthly SIP started in January 1996 and never interrupted accumulates to roughly ₹1.59 crore by the end of 2025 on a total investment of ₹18 lakh — a multiplier of 8.8x and a price-return CAGR of about 7.5%. Add a conservative 1.7% Total Return Index uplift for reinvested dividends and the same SIP becomes a roughly ₹2.05 crore corpus, an annualised return closer to 9.0–9.5%. Those numbers are well below the 12–15% headline that fund marketers like to quote, and they are exactly why we built this calculator: real backtests anchor expectations.
Three deeper lessons fall out of the data:
Lesson 1: Time in the market beats timing the market — almost. The 30-year start (1996) and the 25-year start (2001) produce broadly similar annualised returns of 7.0–7.5% PR. The 10-year start (2016) lands at 6.15% PR. The variation between start years matters far less than people assume, because monthly contributions average across cycles. Lesson 2: There is no riskless 10-year window. The worst 10-year window in our data (2007–2016) still produced a positive 3.8% CAGR — not great, but not a loss. The best window (1998–2007) produced 14.8%. The spread — about 11 percentage points — is the volatility you are paying for. Lesson 3: Crashes recover surprisingly fast in SIP land. The 51.8% Nifty drawdown in 2008 and the 38% intraday COVID drop in March 2020 are both visible scars on a lump-sum chart. On an SIP chart they look more like accumulation discounts.
The 2008 crash: what happened to a SIP that didn't blink
This is the moment that breaks most retail SIPs. Looking at a statement that says “you invested ₹2.4 lakh and your portfolio is worth ₹2.0 lakh” is psychologically brutal, and AMFI data shows SIP discontinuations spiked through 2008–2009. The investors who stopped locked in roughly a 16% paper loss. The investors who continued got the most valuable batch of units in their entire SIP life: every ₹5,000 contribution in 2008 bought ~30% more units than the same contribution in 2007.
By 31 December 2010 the same uninterrupted SIP had recovered to roughly ₹5.5 lakh on ₹3.6 lakh of contributions — a 52% gain on cost in just two more years. By 31 December 2025 that same SIP, started in 2005 and never paused, is worth approximately ₹45.5 lakh on ₹12.6 lakh of contributions — a 6.31% PR CAGR over 21 years, or about 8.0% with TRI included. The crash that looked like the end of the SIP turned into its most productive phase.
The COVID drawdown: visible on a daily chart, invisible on a yearly one
This is the most counter-intuitive result in the dataset. The Nifty 50 collapsed from 12,167 on 19 February 2020 to 7,610 on 23 March 2020 — a 37.5% drop in 23 trading days. Front pages screamed depression. Mutual fund SIP discontinuations briefly spiked. And yet: the year-end close for 2019 was 12,168 and the year-end close for 2020 was 13,981 — a calendar-year gain of 14.9%. From the perspective of someone who simply did not look at their portfolio in 2020, no crash happened.
This is the dirty secret of long-horizon SIPs: most crashes are loud and short. The investors who panicked in March 2020 missed the fastest recovery in Indian equity history. Those who kept buying through April, May and June got a 3–4 month window of generationally good prices. By December 2020 their cost basis was lower, their unit count higher, and the index was 15% above where 2019 had ended. The behavioural cost of watching the market every day vastly exceeded the financial cost of ignoring it.
The worst 10-year window: 2007–2016
Final value at 31 December 2016: roughly ₹8.75 lakh. Price-only CAGR: about 3.84%. That is below the 10-year fixed-deposit return of the same period (~7.5% pre-tax) and below CPI inflation of roughly 6.5% over those years. This is the only 10-year window in our 30-year history where a Nifty 50 SIP underperformed a vanilla fixed deposit. The takeaway is not that SIPs are unreliable — it is that 10 years is the minimum horizon at which equity SIPs become reasonably reliable, and that an investor who got unlucky with their start year should be prepared to extend their horizon rather than exit at the wrong moment. The same SIP, continued through 2025, ends at roughly ₹28 lakh on ₹11.4 lakh of contributions — a much healthier 6.1% PR CAGR.
The best 10-year window: 1998–2007
Final value at 31 December 2007: roughly ₹23.87 lakh on ₹6 lakh contributed — a 4x multiplier and a price-only CAGR of 14.81%, the highest of any 10-year window in our dataset. Add TRI dividends and the same window would have produced something north of 16.5% CAGR. This is the kind of decade that fund advertisements lean on, and it is real — but it is one outcome out of 21 possible 10-year windows. The median 10-year window in our data produced about 6.2% PR CAGR. Hope for the 1998 window, plan for the median.
Why these numbers are lower than the headline 12–15%
Two reasons. First, every backtest above uses the Nifty 50 price index, not the Total Return Index. Indian index dividends are modest (long-run yield ~1.5–2%), but compounded over 30 years that gap is large. The Nifty 50 TRI CAGR from inception through 2025 is closer to 12.8–13.0%; the price index CAGR is closer to 11.5–11.8%. The TRI uplift toggle on the calculator above adds a conservative 1.7%.
Second, the gap between index CAGR and SIP CAGR is structural. An SIP that runs from 1996 to 2025 has its “average rupee” invested for only 15 years, not 30 — because the first contribution compounds for 30 years but the last contribution compounds for only one month. So a 30-year SIP in an index that does 11.6% annualised will report something in the 7–9% range as its money-weighted CAGR. This is not a bug in mutual funds; it is the arithmetic of contribution-weighted returns. The right way to read SIP CAGR is as the rate at which your money compounded, which is exactly what the IRR / XIRR formula gives you.
What this means in practice: when you use the main SIP calculator with a 12% assumption, you are implicitly assuming a long-running Nifty 50 TRI or a fund that beats the index by a couple of percentage points after costs. That is achievable but not free — you need a fund with a sustained alpha track record or a low-cost index fund and the discipline to never sell. If you are uncertain, stress-test your plan at 9–10% nominal and check whether the goal still gets hit.
Five things to do with this data
(1) Run your own scenario. Try a 15-year start (2010), a 20-year start (2005), and a 25-year start (2000) at your actual monthly SIP amount. Note the corpus and the CAGR. That is your honest baseline. (2) Stress test against the worst window. If a 2007-start SIP would still get you to your goal at 3.84% CAGR, your plan is durable. If it only works at 14%, your plan is fragile. (3) Compare against an FD. Use our SIP vs FD comparison with the same backtested numbers and see when equity actually beats debt. (4) Layer in step-up. The step-up SIP calculator lets you simulate a 10% annual increase — on the 2005–2025 window, that nearly doubles the final corpus. (5) Account for inflation. A ₹1 crore corpus in 2025 buys what about ₹28 lakh bought in 1996. The inflation-adjusted SIP calculator shows the real number.
The raw data: Nifty 50 year-end closing levels
| Year-end | Nifty 50 close | YoY change |
|---|
Source: NSE historical archives. Verify any specific year-end level at nseindia.com › Historical Data.