Research · counted from the record
Is the Nifty cheap?
Nifty 50 trades at 19.27 times earnings, lower than on 99% of days since April 2021. Since 2013, months that ended with Nifty in the cheapest fifth of its own P/E history were followed by a higher index a year later 100% of the time, against 78% for all months.
The question
"The market is cheap" is said in every sell-off and "the market is expensive" in every rally. The useful version of the question is narrower: by its own history, where does today's valuation stand, and what has followed readings like it?
How it was counted
For every trading day, Nifty's P/E is ranked against earlier days only, so no day is judged with hindsight.
The ranking is kept inside one earnings basis. NSE worked its index P/E out on standalone profits until March 2021 and on consolidated profits since; the two are not comparable, so a day is ranked only against earlier days on the same basis, and needs a year of such history before it gets a rank.
Each month-end since 2013 is then sorted into the cheapest fifth, the middle three-fifths or the dearest fifth of that ranking, and Nifty's change over the following twelve months is recorded.
What followed
| Month ended in the | Months | With a year after | Higher a year later | Typical change | Middle 80% of outcomes | Stretches | A year apart |
|---|---|---|---|---|---|---|---|
| Cheapest fifth | 22 | 14 | 100% | +22.3% | +12.4% to +37.6% | 7 | 3 |
| Middle three-fifths | 57 | 53 | 79% | +10.5% | −4.4% to +40.5% | 11 | 4 |
| Dearest fifth | 73 | 73 | 73% | +10.2% | −7.2% to +29.4% | 5 | 1 |
| All months | 140 | 140 | 78% | +11.6% | −4.7% to +34.9% |
"Stretches" counts runs of consecutive months as one. "A year apart" counts only stretches that began at least twelve months after the last one ended: the stricter measure of how many separate periods the figure rests on.
How much weight it bears
The result is striking and the sample is small. The 14 cheap months fall into 7 stretches, and only 3 of those began a year or more after the one before had ended. Months close together are not separate evidence: neighbouring months share most of the year that followed them.
So the fair reading is not that a cheap Nifty always rises. It is that, in the few separate periods since 2013 when Nifty was this cheap against its own record, the following year was good each time, and that a handful of periods cannot rule out a different outcome.
Valuation has also said little about timing. Cheap markets have become cheaper before they turned.
Where it stands now
On 9 October 2026 Nifty's P/E sat in the cheapest fifth of its own history on the current basis. Its range since April 2021 runs from 18.92 to 33.61, with 22.14 in the middle.
Questions people ask
Is Nifty overvalued or undervalued right now?
At 19.27 times earnings on 9 October 2026, Nifty's P/E is lower than on 99% of days since April 2021.
What happens after Nifty PE is low?
Since 2013, months that ended with Nifty in the cheapest fifth of its own P/E history were followed by a higher index a year later 100% of the time, with a typical gain of +22.3%. The sample is small: 7 stretches.
Why not use the full P/E history back to 1999?
Because NSE changed how it measures earnings in April 2021. A P/E from the standalone era and one from the consolidated era are different measurements.
Is the Nifty PE ratio a good timing tool?
No. It has been informative about returns over the following year or more, and nearly silent about the following weeks.
Keep reading
Research and information, not investment advice. Every figure describes the past: it is what happened, counted, and not a forecast of what will.