Nifty P/E ratio page: how to read the valuation zones

What problem does the Nifty P/E ratio page solve?

A P/E of 19.65 is just a number until you know what the Nifty 50 has usually traded at. The Nifty P/E ratio page answers that by reading today's price-to-earnings, price-to-book and dividend yield against the index's own daily history since 1999, from NSE's own publication. All figures here are for the close of Thursday 15 October 2026.

Step 1: read the three cards

The page opens with three cards: Price to earnings, Price to book and Dividend yield. Each card shows today's value, a zone label, and a line such as "37th percentile of 6,887 days · median 20.88".

The percentile is the share of past closes below today's value. Dividend yield reads backwards: a high yield is the cheap end, so a 75th-percentile yield is on the cheaper side even though the number is high. Note how close P/B sits to the edge. At 2.82 it is just above 2.77, the line below which the page would call it undervalued.

Clicking a card switches the zone table to that metric. P/E is the default.

Step 2: find yourself in the zone table

Under "Where the Nifty 50 P/E ratio sits today" is a four-row table. Its bands are percentiles of the same 6,887 daily closes: the 10th percentile, the median and the 90th percentile.

• Undervalued: below 14.67, "historically cheap".
• Fairly valued: 14.67 to 20.88, "nothing to act on".
• Slightly expensive: 20.88 to 26.62, "above the historical middle".
• Expensive: above 26.62, "historically dear".

The current row carries a "you are here" tag. Percentile bands mean one freak print cannot move the edges, and the note under the table says plainly that the bands describe where the index has traded, not where it should. Our guide how to read an index P/E ratio covers why the absolute number says little on its own.

Step 3: read "what came next" before drawing conclusions

The next table, "When the Nifty 50 P/E was here, what came next", takes every month-end since 29 January 1999, sorts it into a zone by its P/E, and shows the index's annualised return over the following 1, 3 and 5 years: the median, the weakest and strongest start, and the share of starts that lost money.

Two things stand out. Cheap starts returned the most, but only 33 month-ends ever landed in the undervalued zone: 27 between 1999 and 2005, and six in 2008–09. And the expensive zone's five-year median, 13.3% a year, beat the fairly valued zone's 12.1%. From the fairly valued zone, 21% of one-year windows lost money; over five years none did, and the worst start still made 2.8% a year. These are price returns, so they understate what an investor earned by roughly the dividend yield. Does Nifty P/E predict future returns? works through the same question in more depth.

Step 4: look at the month-by-month grids

"All three, since 1999" draws each metric twice: a line chart of daily closes with the bands behind it, and a heat map where each cell is the median of that month's closes. Hovering a cell gives the number of trading days behind it. The grid is the quickest way to see when the index was last where it is now.

Worked example: the Nifty Bank page

Back on the Nifty P/E ratio page, below the grids, "Every NSE index, cheapest first" ranks every index NSE priced that day by P/E: 144 of them on 15 October, from Nifty PSU Bank at 7.22 upwards. The 120 with enough history are links to their own pages, such as Nifty Bank's.

The per-index page carries the same cards, zone table, forward-return table and grids, plus a section headed "How the Nifty Bank's valuation reads today":

• Its P/E was 13.20, fairly valued, the 21st percentile of 5,707 daily closes since 2003, against a median of 16.42.
• It traded at 0.67 times the Nifty 50's P/E, against a usual 0.80 times, which the page calls "a deeper discount than usual".
• Its P/E has ranged from 6.22 (March 2009) to 69.08 (September 2019), and it was 15.68 a year earlier.

The comparison with the Nifty 50 is often the more useful line, because a sector can look cheap on its own history while still being dear against the market, or the reverse. Most indices start well after 1999 (Nifty Bank in 2003, the Nifty Midcap 150 in 2016), so their bands rest on fewer years. Median stock P/E by market cap looks at the same question one level down, stock by stock.

What the page does not tell you

NSE publishes these figures once a day after the close, on a consolidated trailing-twelve-month basis, so there is no intraday P/E. A falling P/E can mean prices fell or earnings rose; the page cannot tell you which without the index level beside it. And none of the zones is a buy or sell signal. The page links is the market expensive? for a reason: valuation has said more about ten-year returns than next year's. For the flow side of the same market, the FII and DII data page tracks who is buying. Nothing here is a recommendation.