The average year myth

A friend laid out his investing plan for me last week. Simple, he said. Park the money in the market, collect the long-run average, roughly 12 percent a year. Nothing wrong with the number. It is close to what the Nifty has actually done over twenty years.

So out of curiosity I pulled two decades of calendar returns and went looking for the average year. Not the average. The year where the market just quietly delivered its 12 and went home.

I found it twice.

Twenty years, and in only two of them did the index land anywhere near its own long-run rate. Every other year it was either sprinting or bleeding. Nine of those twenty years finished up more than 20 percent. Three finished in the red, and 2008 alone gave back about half. The best year was up 76. The worst was down 51. Somehow that violent spread averages out to a calm-sounding 12.

Here is the part most plans quietly ignore. The average is a summary, not a schedule. It does not get paid out in tidy monthly slices. It gets assembled out of a handful of enormous years, and if you are not holding through those years, you do not get the number. Sit out two or three of the sprints and your 12 quietly settles to a 6 without you ever making an obvious mistake.

The trouble is you cannot know in advance which year is the sprint. They arrive unannounced, usually right after the stretch that made you want to step aside. So the honest move is not to guess. It is to stay in the seat for all of them and let the average assemble itself.

That is really the whole case for a rule over a hunch. A rule stays seated. A mood gets up and leaves right before the year that mattered

The same test on recent years

The opening used two decades of calendar years. Here is a shorter window you can check line by line: six financial years of the Nifty 50 price index, from the close of Mar 2019 to the close of Mar 2025, using NSE closing values.

Over those six years the index went from 11,623.90 to 23,519.35. That is a multiple of 2.02, and a CAGR of 2.02 ^ (1 / 6) minus 1 = 12.5 percent a year. Almost exactly the long run number in the opening.

Now look for a year that delivered 12.5. There is none. The closest are 18.9 and 5.3. One year fell 26 percent, the next rose 71. The plain average of the six is 16.2 percent, which is not the number either, because averaging ignores how a fall shrinks the base.

These are price index figures. Dividends, at the 1.22 percent yield NSE showed for the Nifty 50 on 30 Sep 2026, would add about a point a year on top, and they would not change the shape: every year was either well above or well below.

What the one big year was worth

Most of those six years' gain arrived in one of them. Here is the test I find most useful, set at the moment it would have been hardest.

The gap between those two outcomes is one year out of six. No terrible trade, no wrong stock. Just one absence at the worst possible time, which happened to be the time it felt most sensible.

Why a rule helps here

The years that make the average tend to arrive when the news is at its worst. FY 2020-21 began in the middle of a crash. Nobody felt like buying, and nobody could know the next twelve months would be the best in the window. A mood reads the news. A rule reads the calendar.

1. Decide in advanceWrite down when you buy, rebalance and sell, before the market tests you.
2. Act on the dateRebalance on the scheduled day, whatever the headlines say.
3. Size for the bad yearA rule only holds if the position is small enough to sit through a 26 percent year.
4. Let it assembleThe average shows up only for those still holding on the big days.

That third step matters most to me. Staying seated is not courage. It is mostly sizing: an amount you can live with when the year in front of you looks like FY 2019-20.

The free Learn module on why rules beat gut feeling covers how rules handle exactly these moments, with quizzes and a certificate. The single best days version of this lesson is in missing the best days, and the conversion behind every number here is in total return against CAGR.