What Actually Precedes a Sector Taking Leadership: 18 Years of Indian Market Data

Sector leadership in India changes hands constantly. The group of companies that led returns last quarter is rarely the group leading now. The useful question is not which sector is leading today, which any table can tell you, but whether anything measurable is true about a sector before it takes leadership.

We tested that against eighteen years of NSE index history. This post is the method and the result, including the parts that did not work.

How a rotation turn was defined

Nothing here is judged by eye. A rotation turn is defined mechanically:

• the sector lagged the Nifty by more than 5% over the previous 63 trading sessions, and
• it then beat the Nifty by more than 8% over the following 63 sessions.

That is a laggard becoming a leader, which is what rotation means in practice. Applied to Bank, IT and Pharma from 2008 to 2026, it produces 571 qualifying sector-sessions out of 12,321, or 4.6% of all observations.

Those three sectors are used because NSE index daily history is only available for them. That is a real limitation and it is stated here rather than hidden: the finding is a base rate across three sectors over eighteen years, not a universal law.

What was already true on the day a laggard turned

Each candidate signal was measured using only information available on the turn date, then compared against the same measure taken on every other session in the sample.

Signal on the day	At turns	Every other session
Own PE percentile	27th	52nd
Correlation to the Nifty, 63 sessions	0.59	0.69
Distance from its own 200-day average	-2.0%	+4.8%
Previous 1-month move against the index	-3.6%	+0.4%
Volatility relative to the index	1.38	1.31

Medians on their own can mislead, so the more important test is a hit rate. Of the 2,668 sessions where a sector was lagging, 21.4% went on to lead. That is the number every signal has to beat.

Valuation was the only signal that beat the base rate

Splitting those lagging sessions into thirds by each signal:

Lagging sector, split by its own PE percentile	Went on to lead
Lowest third of its own PE history	34%
Middle third	15%
Highest third	15%

A lagging sector sitting in the lowest third of its own valuation history turned leader more than twice as often as one in the top third. That is the clearest result in the study.

Two results that contradict the obvious reading

Being deeply below the 200-day average made a turn less likely, not more. The instinct that a sector far below its long average is due a bounce is not supported here. Among lagging sectors, those furthest below their 200-day turned 19% of the time, while those nearest or above it turned 27%.

Notice that this contradicts the median table above, which shows lagging sectors sitting slightly below their 200-day at turns. The median describes where turns happened; the hit rate describes where turns were likely. When the two disagree, the hit rate is the one that answers the question.

Falling correlation looks like a signal and is not a reliable one. Correlation to the index was lower at turns, 0.59 against 0.69, which fits the idea that a sector decouples before it moves on its own. But split into thirds, the hit rates run 27%, 14% and 23%. That is U-shaped rather than a trend, so it does not support a rule.

Relative volatility produced no usable separation at all: 22%, 25% and 17% across the three buckets.

What this does not prove

Those 571 turn-sessions are not 571 independent events. They cluster into perhaps twenty to thirty genuine episodes, concentrated in 2009, 2011, 2013 and 2014, because consecutive sessions inside one turn all qualify. Treat the direction as informative and the decimal places as not.

The sample is three sectors. The strongest signal found moves the odds from 21% to 34%, which is an edge over a base rate, not a forecast. And a base rate measured across a population says nothing specific about any one sector today.

Where to see the current picture

The live measurements behind this analysis update daily on the Sector Rotation Tracker: leadership rank across four quarterly snapshots, returns by window, correlation between every sector pair, breadth above the 200-day average, and each sector's PE against its own history.

Two related pages carry the inputs. The index PE tracker holds valuation history for 26 NSE indices, which is where the percentile in this study comes from. The FII and DII flows page shows who has been putting money in, session by session.

Method, in full

Price data is NSE index daily closes via Yahoo Finance, 2008 to 2026, for Nifty Bank, Nifty IT and Nifty Pharma against the Nifty 50. Valuation percentiles come from NSE index-level PE history, which reflects whoever was actually in each index at the time and therefore carries no stock survivorship bias. Correlation is the Pearson correlation of daily returns over a 63-session window. Every figure above was computed from those series directly and is reproducible from the definitions given.

Downstox is not a SEBI-registered investment adviser or research analyst. This is a description of historical data, published for education and information only. It is not a recommendation to deal in any security.