What RS measures, where it lies, and today's NSE leaders by sector.

A stock at RS 90 in a market down 20% is down 5%. That is survival, not strength.

RS 90 in March 2020 belonged to NESTLEIND — down eight percent while the Nifty dropped twenty-four. RS 90 in October 2021 belonged to TATAMOTORS — up forty percent while the Nifty rose eighteen. Same reading. Opposite trades. Most retail traders never separate the two. The compact definition: relative strength ranks every stock by how its price moved versus the Nifty 50 over the last 252 sessions, then expresses the result on a 1-to-99 scale. RS 90 means the stock outperformed eighty-nine percent of the NSE universe in that window; RS 30 means it underperformed seventy percent.

The trap sits one inch below that definition. In a bull regime an RS-90 stock is usually compounding capital. In a Stage 4 regime an RS-90 stock can still be losing fifteen percent — it is just losing it slower than the index, which is down twenty. The number reads the same; the outcome for a long position is opposite. RS without a regime read is half a sentence.

Today the NSE universe holds 2,527 actively-tracked stocks. After Stage 2 classification, 824 survive. After applying a sector filter and a high-RS cut, a typical day's candidate set sits in the thirty-to-sixty range. That compression — universe to dozens — is what RS does when used as a secondary filter on top of stage classification. The compression is also where the failure modes start, because filtering by RS alone produces a different list from filtering by RS-inside-a-leading-sector. The rest of this page is about that gap.

The worked example — how an 18% move over 63 days becomes RS 87.

The cleanest way to show what RS computes is one number at a time. Take an NSE stock that rose 18 percent over the measurement window while the Nifty 50 rose 4 percent over the same window. The raw figure is the difference: 18 − 4 = 14 percentage points. That figure is then ranked across every other NSE stock's same-window ratio. If 87 percent of them sit below 1.135, the stock's 63-day RS reading is 87.

The number is straightforward; the design choices behind it are not, and they are ours. Any relative-strength measure trades reactivity against stability — a short lookback moves fast and is easy to game with a single gap, a long one is steady but can flatter a stock for months after leadership has rolled over. Where we sit on that trade-off is proprietary. Read RS here as a slow signal, and use kinetics for the short-horizon question.

The second choice is the benchmark. Compute RS against the Nifty 50 and a midcap leader will look strong against a large-cap-dominated index; compute against the Nifty 500 and the same midcap looks less differentiated. We compute against the Nifty 50 because the index is the headline mental anchor for Indian retail traders, and the question RS answers is “is this stock keeping up with the market most people see?” The window and the weighting are proprietary.

One last thing the formula does not show on its face. Because RS is a percentile rank, the median NSE stock is always at RS 50 by construction. “RS above 50” is a useless filter — it only excludes half the universe. The useful cuts sit at RS 75 (top quartile, roughly the top 632 NSE names today) and RS 90 (top decile, ~253 names). Most setup work happens above the RS 75 line because below it, prior-year drawdowns are too wide to size around a sensible stop. RS 90 is where a 2× ATR stop typically sits inside a base structure rather than hanging in air — at TATAMOTORS' 2023 Stage 2 entry around ₹540, the 14-day ATR was ~₹12, so a 2× ATR stop at ~₹516 lived inside the prior consolidation. Below the RS 75 line, the same stock would have needed a ~₹35 stop to clear noise — a different position size and a different trade.

Stop-distance arguments only matter when the regime allows long entries to work at all. The next chapter is about how an RS-90 reading lies when the absolute frame is pointing the wrong way.

The regime caveat — RS is a relative measure laid on top of an absolute frame.

The strongest argument against RS-as-a-standalone-signal is that it cannot tell a falling market from a rising one. RS 90 in March 2020 — NESTLEIND on NSE that month — meant a stock had fallen eight percent while the Nifty fell twenty-four. RS 90 in October 2021 — TATAMOTORS — meant a stock had risen forty percent while the Nifty rose eighteen. The reading is identical. The capital outcome of buying the first is a loss; the second is a gain.

That asymmetry is why RS belongs underneath the regime read, not beside it. Inside a Bull regime, RS-90 stocks compound at multiples of the index. Inside a Recovery regime, RS-90 stocks recover earliest and are the cleanest 1→2 transitions. Inside a Topping regime, RS-90 stocks are the last to crack — Indian retail's classic example was DMART through Q2-2022, which held up for eleven weeks after the Nifty topped before its own Stage 3 distribution started. Inside a Bear regime, RS-90 stocks are still losing money — they are simply the least bad names in a universe nobody should be buying long-only.

Today 159 NSE stocks closed at a 52-week high and net-new-highs reads -42. When that second number turns negative, RS-90 stocks are leading a falling tide; when it stays strongly positive, they are catching one.

The platform offers a kinetics check alongside the position read. Today, 12.4 percent of the active universe is accelerating. A wide accelerating share means leaders are pulling away from laggards; a narrow share means the universe is moving as one — consolidation or regime change in progress.

Three failure modes that show up after the regime caveat.

Once a trader has internalised that RS reads inside a regime, the next surprise is that RS still fails inside a constructive regime. Three patterns recur on NSE and each leaves the percentile rank looking healthy while the trade structure is rotten.

1. Thin-float pumping in midcaps

A stock with under ₹20 crore of daily delivery volume can be pushed three or four percent on a few hundred lakh of buying. Stack that for fifteen sessions and the 63-day RS reading climbs into the high 80s without any institutional accumulation underneath. The chart looks like leadership; the order book looks like a single concentrated retail pool. Indian retail traders see this most often in ‘multibagger’ YouTube picks where the same name surfaces across three creators inside a week. The prop scan restricts to a ₹500cr–₹1.5 lakh cr market-cap band above ₹30, which strips out most of the micro-cap tail where this happens. Be clear about the limits of that, though: the setup score itself has no liquidity term, and ASM-tagged or T2T-segment names are not excluded from anything.

2. Q-result gaps that inflate the 63-day window for nine weeks

A stock that gapped 18 percent on a single Q-result day looks RS-strong for the next 63 sessions regardless of what it does afterward. A one-year window dilutes any single session, which is the main defence here — but it cuts both ways: a gap that happened ten months ago still counts fully until it rolls out of the window. The defensive check is to look at the chart itself: an RS-90 reading made of one candlestick is a result reaction; an RS-90 reading made of a smooth thirteen-week advance is a re-rating. The percentile cannot tell you which.

3. PSU-and-cement concentration in the top-10 RS list

On 2024-07-23 — the day after the Union Budget capex push — the top-10 RS list across NSE held seven PSU banks, two cement names, and one defence stock. A trader who picked highest-RS without checking sector composition concentrated their entire candidate set into one budget reaction. The fix is structural: rank RS leaders by sector, then pick within each sector, rather than running a flat top-N across the universe. That is also what the stock leadership view does by default.

Time-in-trend is the silent backstop. The median Stage 2 stock on NSE today has been in Stage 2 for 9 weeks; the p25 sits at 4 — younger trends that have not yet earned their structure. RS 90 without a Stage 2 cohort that has cleared the p25 is usually one of the three failure modes above.

The classic teaching pair sits between sector context. BANKBARODA in the 2022–2026 PSU bank re-rating was an RS leader inside a leading sector; the trade compounded for three and a half years from around ₹96 to ~₹325. The full case study lives on market rotation. HINDUNILVR over the same window was a high-quality name with persistently bad RS — multi-year sideways range with FMCG-sector underperformance. The full case study lives on stage analysis. Both were available to a trader running an RS filter alone; only the sector context would have told you which one to size.

Our single-window RS versus the IBD-style and Mansfield conventions.

Three relative-strength conventions show up in trading literature, and the differences matter when you compare numbers across platforms. The platform's NSE measure is one of them; IBD-style RS as published in Investor's Business Daily is another; Mansfield Relative Strength as defined in Weinstein's 1988 book is the third.

IBD's convention is the most familiar to readers who learned RS from US sources. The IBD RS Rating is a percentile rank built from a price-change ratio over four windows, weighted with extra emphasis on the most recent quarter. The benchmark is the S&P 500 for US markets. Ours is also a percentile, and also benchmarked to a headline index — the Nifty 50 — but the construction behind it is proprietary and is not IBD's. The numbers are not exactly comparable between IBD's rating and ours, but a stock at IBD 90 and an NSE stock at our RS 90 are filtering for the same concept of “outperformed nine in ten peers.”

Mansfield is older and structurally different. It is a smoothed ratio line of the stock against the benchmark, plotted as a chart rather than a number, with a zero line marking neutrality. Weinstein used it to spot divergence — when price made a new high but the Mansfield line did not, the leadership was weakening before the price chart said so. On NSE, INFY in August–October 2021 is the textbook example: price marched to ₹1,950 while the Mansfield line versus the Nifty rolled over, and the Stage 3 top arrived seven weeks before the price chart admitted it. The Mansfield reading is on the daily chart of every stock the platform tracks; it complements the RS percentile rather than replacing it. For where Weinstein laid this out in detail, see Stan Weinstein.

RS is position. Acceleration is kinetics.

One distinction matters more than the convention differences. RS is a position read — it tells you where a stock sits in the percentile ladder over a full year. Kinetics — acceleration — is the rate at which the position is changing right now. Two stocks can share an RS reading of 88 while one is steepening and the other is rolling over.04 and the other +0.18. The first is sitting in the top decile because it was strong three months ago; the second is sitting there because it is strong this week. For a Stage 2 trader trying to read the leading edge of leadership, the kinetics matter more than the position.

The next chapter shows both lists side by side — today's RS-position leaders by sector, and today's six fastest-accelerating stocks. A name that appears in both is leading on where-it-is and on how-fast-it's-moving; that overlap is the cleanest structural read the platform produces.

Today's RS leaders, one row per sector — and the kinetics top six.

This is the exhibit the audit of the previous edition flagged as missing. An RS page that does not show what an RS leader looks like today is a page about an idea, not about a market. The table below is the platform's sector-level read on the NSE as of 11 Sep 2026 — the top-ranked stock by RS slope inside each sector with at least one qualifying Stage 2 candidate.

Acceleration top six — different signal, same surface.

The RS table above ranks stocks by where they sit. The next table ranks by how fast they are getting there. Acceleration leaders are the stocks our model ranks as steepening fastest — the leading edge of trend strength rather than the cumulative result of past trend strength. The two lists overlap but never match exactly; the divergence is where new leadership emerges before the RS percentile catches up.

Five years of top-decile-RS rolled up — and the one year it lost.

The empirical question is whether holding an annually-rebalanced basket of the top-decile-RS names actually beats the Nifty over multiple years on NSE. The table below summarises the platform's internal rolling backtest: cap-weighted return of the top decile at each year start, held one calendar year, against the Nifty 50 over the same window.

Two readings off the table. The first: across five years the basket outperforms the Nifty by an average of around 16 percentage points per year, and the median spread is positive in every regime other than the 2022 stealth bear. The second: 2022 underperformed by 12 percentage points — the year the regime read said “reduce long exposure” and the RS read alone said “keep buying leaders.” That single year is the empirical case for using RS underneath a regime filter rather than as a standalone signal. Without the regime check, the basket holds through a bear it had no business participating in.

The third reading sits inside the sample sizes. The top-decile basket holds 215 to 235 names depending on the year — large enough that single-stock blowups (Adani 2023, IDEA across the whole window) drag on returns without breaking the basket. Compounded across the five-year sequence — including the 2022 loss — the basket returns roughly 2.6× the Nifty's cumulative ascent (≈222% vs ≈86%); the absolute path is bumpy but the relative arithmetic is consistent. That is the empirical claim this site rests on, and the year it failed is also the year that names the filter that should sit on top of it.