Why Long-Term Equity Returns Remain Uncertain

Holding shares for longer has narrowed the range of outcomes in India's recorded history. It has not removed the range — and the record we have is much shorter than the confident conclusions people draw from it.

Updated 2 September 2026

Rolling Returns AnalyserOpen

The question, and why the usual answer is wrong

Ask what the stock market returns and somebody will give you a single number. That number is an average between one particular start date and one particular end date, and shifting either date moves the answer — often by more than the number itself.

There is a better way to answer the question. Instead of picking one stretch of time, look at every stretch of a given length. What did someone get who invested for ten years starting in 2004? In 2005? On every single day in between?

Do that and the answer stops being a number and becomes a range. The range is the honest answer, and it is far more useful for planning than an average.

Below is that range for the Nifty 50 — India's index of fifty large companies — across 27.2 years, from 30 June 1999 to 31 August 2026.

What actually happened, over every possible period

If you held forPeriods testedWorstPoorTypicalGoodBestLost money
1 year6,511-55.31%-12.75%+12.41%+48.97%+107.61%23.6%
3 years6,016-15.23%+2.40%+13.40%+34.34%+61.68%6.3%
5 years5,519-1.03%+6.31%+13.60%+27.36%+47.64%0.1%
7 years5,023+4.90%+9.27%+13.71%+24.17%+30.47%0.0%
10 years4,282+5.13%+9.43%+13.87%+19.04%+22.27%0.0%
15 years3,043+8.59%+11.14%+13.46%+17.50%+19.37%0.0%
20 years1,804+9.68%+12.03%+14.78%+16.96%+17.99%0.0%
Every possible holding period of each length, before inflation. Each row starts a new period on every trading day and follows it to the end, so a "10 years" row contains thousands of overlapping ten-year stretches rather than one. "Poor" and "Good" mean one in ten periods was worse or better than that; "Typical" is the middle one. Figures are per year, and include dividends but not inflation.

The figures include dividends as well as price changes, and they are shown per year so that periods of different lengths can be compared.

Three things in that table are worth sitting with.

A single year is close to a gamble. The worst twelve months in the whole record lost 55.31% of the money invested while the best gained +107.61%, and about 23.6% of all one-year periods ended below where they started. If you need your money back within a year you are not really investing in this range — you are taking one draw from it and hoping.

Waiting helps, but slowly, and mainly at the bad end. Stretch to five years and the worst outcome improves to -1.03%, with only 0.1% of periods losing money. Notice that the good end shrinks as well: long holding periods trade away the spectacular results along with the disastrous ones, which is a trade most people would take and few are told they are making.

What barely moves is the typical result. It sits near +13.87% a year at ten years, almost exactly where it sits at three. Holding longer did not raise the return you should expect. It reduced the chance of ending up far away from it — which is the whole of what a long horizon buys you, and it is a great deal less than the phrase "time in the market" usually implies.

The most dangerous line in the table

Look at the "Lost money" column beyond seven years. No period in this record finished below where it started.

That is the single most seductive fact on this page, and it is the one most likely to be quoted back at you by somebody selling something. Before anyone relies on it, three things have to be attached to it.

It describes one country. India, over a stretch of unusually fast economic growth. The same table for a country whose economy disappointed would look very different — and the people living through that period had no way of knowing in advance which kind of country they were in.

It describes one run of history, 27.2 years long. The longest rows in the table are built by sliding a window across that same stretch over and over. That produces thousands of periods but only a handful of genuinely independent ones, so the "periods tested" column makes the evidence look stronger than it is.

It ignores rising prices entirely. Every figure above is in rupees of the day they were earned, which brings us to the next section.

The same returns, after inflation

Money is only worth what it buys. If your investment grows while the price of everything else grows too, you are not as far ahead as the statement suggests.

If you held forPeriods testedOn paperIn what it buys
1 year3,095+12.41%+7.71%
3 years2,597+13.40%+8.38%
5 years2,101+13.60%+8.50%
7 years1,606+13.71%+8.18%
10 years863+13.87%+8.14%
The typical return per year, before and after inflation, across the stretch where both market and inflation figures exist. "On paper" is the number an account statement would show. "In what it buys" is what was left after prices rose — the one that decides whether you are actually better off.

Across the years where inflation figures exist, prices rose 5.11% a year. That takes the typical ten-year result from +13.87% on paper down to +8.14% in terms of what the money could actually buy.

Both numbers are true. Only the second one tells you whether you got richer.

What the numbers hide: the journey

A table of end results says nothing about what happened along the way, and the way is where most people give up.

High pointLow pointFallTime fallingTime to get back
11 February 200021 September 2001-50.2%1.6 years2.2 years
14 January 200417 May 2004-29.8%4 months6 months
10 May 200614 June 2006-29.7%35 days4 months
8 January 200827 October 2008-59.5%10 months1.9 years
5 November 201020 December 2011-27.2%13 months17 months
3 March 201525 February 2016-21.7%12 months6 months
14 January 202023 March 2020-38.3%2 months7 months
Every occasion the market fell 20% or more from a high point before recovering. "Time to get back" runs from the low point to the moment the previous high was reached again — so someone who invested at the high point waited the two periods added together before being level.

The worst fall in the record began on 8 January 2008 and reached bottom on 27 October 2008, by which point the market had lost 59.50%. The longest wait from a low point back to the previous high was 2.2 years.

Someone who held on through that would have ended up with a perfectly respectable ten-year return and spent several years watching their money shrink. Those are the same event described two ways, and only one of them appears in a returns table.

That matters because every figure on this page assumes the money stayed invested for the whole period. That assumption is doing more work than any return in the table.

What to take away

Holding for longer narrows the range of what can happen to you. It does not promise a return, it does not make shares safe, and the evidence that it helps comes from one country's recent history rather than from any law of nature.

Use the range the way it should be used. Look at the worst row that matters to you and ask two questions: could my plan survive that, and could I wait as long as the recovery column says?

If your answer depends on the typical result, you have been reading the wrong column.

Disclaimer

Educational content only. This is not personalised financial, investment or tax advice. Figures quoted are historical or illustrative and are not forecasts. Consult a qualified professional before acting on anything you read here.