Strota Research

A cheaper Nifty P/E was followed by better years since 2021, but the evidence is one cycle deep

By Strota Newsroom · 2026-10-22 · How Strota reports

Nifty 50P/E ratiovaluationresearchbacktest

In the consolidated-earnings era, the cheapest third of Nifty sessions averaged a 15.6% gain over the next year against 4.4% for the dearest third. The direction is textbook; the sample is far thinner than it looks.

Since the Nifty's price-to-earnings ratio moved onto consolidated earnings, the index has done best over the following year when it looked cheapest. Sessions in the lowest third of P/E readings were followed by an average one-year price gain of 15.6%, while sessions in the most expensive third managed only 4.4%. That is exactly the direction valuation theory points to. The catch matters as much as the headline: the whole result rests on a single market cycle.

Why it is worth a look: P/E is the valuation gauge most often quoted for Indian equities, and the claim that a cheap market tends to reward the next twelve months is repeated endlessly. Strota tested it on the data that actually exists for the current definition, and the answer is a qualified yes with heavy caveats attached.

First, the reason the window is so short. NSE moved the index ratio off standalone earnings and onto consolidated ones on 1 April 2021. Readings from before and after that date measure different things, so mixing them would compare apples with oranges. The study therefore starts on that day and ignores everything earlier.

The method is simple. It uses NSE's end-of-day index file for the daily Nifty 50 closing level and its P/E. For every session that has a full year of later trading behind it, counted as 252 sessions, it records the price return over that following year. Sessions between 1 April 2021 and 9 October 2025 qualify, which gives 1,119 of them. Those sessions were then sorted into three roughly equal groups by the P/E on the day.

The cheap band, a P/E below 21.7, held 367 sessions. It averaged 15.6% over the next year, ended higher 95% of the time, and its worst outcome was a loss of 9.8%. The middle band, from 21.7 to 22.8, held 376 sessions, averaged 8.9%, finished up 73% of the time and saw a worst fall of 10.9%. The dear band, 22.8 and above, also held 376 sessions, averaged 4.4%, rose in 74% of cases and had a worst drop of 6.8%. Across all 1,119 sessions the average was 9.6%, with 81% ending positive.

What survives is the ordering of the averages. Cheap beat middle, and middle beat dear, in a clean step-down. The hit rate in the cheapest band also stands out: almost every one-year window that began there finished above its starting level.

What does not survive is any neat story about risk. The expensive third had the shallowest worst case of the three, and its 74% hit rate was a touch above the middle band's 73%. High readings in this window were not followed by disaster. They were followed by smaller gains, and the dearest sessions still ended a year higher in most windows.

Now the limits, which the entry itself is blunt about. A count of 1,119 sounds like a large sample, but the windows overlap almost completely. A year that starts on a Monday shares nearly every trading day with the year that starts on Tuesday. Spread over roughly four and a half years, that collapses into effectively a handful of independent observations. It is like judging a restaurant from a thousand reviews written by the same five diners.

The bands are also tightly packed. Only about one P/E point separates them, so the test is sorting days that were barely different in valuation terms. A big gap in subsequent returns from such a small gap in valuation, inside one cycle, is a thin base for any general rule.

Two further caveats. The figures are price returns, so dividends are left out; the entry estimates they would add somewhere between one and one and a half percentage points a year to every band, which lifts all three without changing their order. And the P/E uses trailing earnings, which lag. A falling ratio can mean earnings are climbing or the price is sliding, and those are very different situations hiding behind the same number.

The fair reading is narrow. Over this stretch, lower valuations came before stronger years, in line with the textbook. But this is a record of how one cycle unfolded, not a method for timing the market.

Sources and method

This note is based on Strota's own backtest research. Historical results are not forecasts or investment advice; see the editorial standards for methodology and limitations.

A cheaper Nifty P/E was followed by better years since 2021, but the evidence is one cycle deep

Sources and method

More money stories