Home / Long-term investing
The index record
Twenty-five years of the Nifty 50. The surprising part is not that it went up.
Year-end closes, 2000 to 2025, price index only, no dividends. Even then, a ten-year wait never lost money. The worst ten-year stretch compounded at 5.6%, which is what a traditional plan calls a good year.
Rolling 10-year CAGR, year-end to year-end. The gold line is 5%, a typical traditional-plan return. One bar kisses it. The rest leave it behind.
The crash is not the story. Sitting it out is.
Skip the five best years, 2003, 2005, 2006, 2007 and 2009, and park the money at 0% those years. A 12.9% machine becomes 3.4% over 25 years. That is endowment territory. The recoveries after 2002, 2008 and 2020 are the product. They are also the years people pause their SIP.
Most years felt broken.
From 2000 to 2025 the index fell at least 10% from its peak in 22 of 26 calendar years. The median drop inside a year was about 15%. 2008 closed down 52%. COVID cut 38% mid-year, and 2020 still finished up 15%. Calm years are the exception. Staying is the skill.
Perfect timing was worth 2.5 points.
Across daily closes from 2000 to 2025, the luckiest investor who bought every year's low made 14.3% a year. The unluckiest, buying every high, still made 11.8%. The steady monthly investor made 12.6%. The unlucky number still beats a 5% plan by about six points a year. Timing is not the product.
Five years can go red. Ten years did not.
Of 21 five-year windows, one lost money: buy December 2007, sell December 2012, down 0.8% a year. The median five-year was 12.3%. Stretch to ten years and the negative column is zero, the worst 5.6%. Twenty-year windows clustered between 11.7% and 15.1%. Horizon is the product.
Rolling calendar returns, price only
| Hold | Windows | Worst | Median | Best | Negative |
|---|---|---|---|---|---|
| 5 years | 21 | -0.8% | 12.3% | 41.2% | 1 |
| 10 years | 16 | 5.6% | 12.9% | 18.4% | 0 |
| 15 years | 11 | 7.5% | 12.4% | 16.3% | 0 |
| 20 years | 6 | 11.7% | 13.0% | 15.1% | 0 |
| A traditional plan | n/a | ~4% | ~5% | ~6% | n/a |
Five years is the only row with a red cell. One window of 21 went negative, December 2007 to December 2012, down 0.8% a year. You still had 96 paise on the rupee after the crash. If the money is for a wedding in five years, this table is not your argument. If it is for ten, the red column is already zero.
₹1 lakh invested in December 2000, held to December 2025. Price index only, so this understates the equity side.
The crash is not the story. Sitting it out is. Timing is not the product. Horizon is.
The belief behind it A flat SIP is a pay cut you give yourself every year →Nifty 50 price index, not the total-return index. Dividends would add roughly 1 to 1.5 percentage points a year, so this understates equity, it does not flatter it. Year-end closes from NSE: about 1,264 in December 2000, 26,129 in December 2025. Rolling windows, intra-year drawdowns and the luckiest and unluckiest cases use daily and month-end closes over the same period. Past returns are a record, not a promise. 2026 is not finished: in early September the index sat near 24,000, about 8% below the December 2025 close.
More of the thinking behind the tools
New beliefs and guides as they land, plus the occasional money note. Free, no spam.