Relative Strength

Key takeaways · 8

1. 01Relative strength compares a stock's return to a benchmark or to other stocks. The IBD RS Rating, the RS Line on a chart, and the academic momentum factor are all the same idea presented differently.
2. 02It is not the same indicator as RSI. The Relative Strength Index (Wilder, 1978) is a bounded 0-to-100 oscillator computed from a single stock's own up and down moves. Relative strength compares a stock to other assets.
3. 03The IBD RS Rating is the trader vocabulary: a 1-to-99 percentile rank of a recency-weighted 12-month return (40% most recent quarter, 20% on each of the prior three). William O'Neil built it in the 1960s and it is still the standard.
4. 04Every major swing-trading school uses relative strength as the first universe filter. O'Neil 80+, Minervini 70+ (prefers 80+), Weinstein's Mansfield RS above zero, Kullamägi top 1-2% across 1, 3, and 6-month horizons.
5. 05The academic case is settled. Jegadeesh and Titman (1993) found roughly 1% per month abnormal return on long-winners / short-losers in US stocks. Asness, Moskowitz, and Pedersen (2013) found the same effect in eight asset classes worldwide.
6. 06Reading 1-week, 1-month, 3-month, and 6-month percentiles together separates real leaders from single-horizon flukes. Top decile across all four is a structural leader; top on 1W only is usually a relief bounce.
7. 07Momentum can crash hard. Daniel and Moskowitz (2016) found the past-loser decile rose 163% from March to May 2009 while past winners rose 8%. Anyone short the losers got run over. Pair RS with a trend filter and pull exposure when regime turns.
8. 08RS is the screen, not the trigger. You still need base structure, volume, an earnings date, and a defined stop before you can buy.

Quick stop before going further, because beginners mix these up constantly. Relative Strength and RSI share the word "Relative" by historical accident, and screeners abbreviate both of them "RS." They are not the same indicator.

RSI is the Relative Strength Index, a momentum oscillator J. Welles Wilder published in "New Concepts in Technical Trading Systems" in 1978. The formula is RSI = 100 − 100 ÷ (1 + RS), where RS here is the ratio of average up closes to average down closes over a 14-period lookback. Output is bounded between 0 and 100, with 70 commonly read as overbought and 30 as oversold. Wilder's RSI compares a stock to itself — only that stock's own up and down days feed the calculation.

Relative strength without the "Index" compares a stock to a different asset, usually the S&P 500 or SPY. Output is unbounded (a ratio, a percentile, or an excess return), and it answers a leadership question rather than an overbought one. The two indicators are often used together, never as substitutes for each other.

If you are looking at a screener column and unsure which one it is: values 1 to 99 mean RS Rating; values 0 to 100 with horizontal lines at 30 and 70 mean RSI; an unscaled line on a separate chart pane means the RS Line ratio. Same word, different jobs.

Two indicators that share half a name. Left: relational, unbounded, comparative. Right: reflexive, bounded, oscillator. Same word "RS" abbreviates both, and that is why screener columns deserve a label check before you use them.

Four formulas show up in real screens. None of them are new, and all of them measure the same thing in slightly different ways.

Simple ratio. Stock close divided by benchmark close, plotted as a line. TradingView and StockCharts both call this the "RS Line" and render it with syntax like NVDA/SPY or $NVDA:$SPX. No normalization, no smoothing, no time scaling. Rising line means outperformance, falling means lagging.

Excess-return difference. Stock return over N days minus benchmark return over the same N days. If NVDA is up 32% and SPY is up 4% over the last quarter, NVDA's 90-day excess return is +28%. This is what spreadsheet work and academic momentum factors are built on.

Percentile rank. Take the N-period return for every stock in the eligible universe (typically NYSE / Nasdaq / AMEX names with enough liquidity to be tradable), sort them, assign each a percentile from 1 to 99. The 99th-percentile stock outperformed 99% of the universe. The 50th-percentile stock was exactly average. This is what every "RS Rating" column on a screener is doing under the hood.

IBD's recency-weighted twelve-month percentile. The published form: 0.4 × ROC(close, 63) + 0.2 × ROC(close, 126) + 0.2 × ROC(close, 189) + 0.2 × ROC(close, 252), where ROC(close, N) is the percentage price change over the last N trading days. The weighted sum is computed for every stock in the universe and percentile-ranked into the familiar 1-99 scale. The 40 / 20 / 20 / 20 split tilts toward the most recent quarter — a stock accelerating into leadership today scores higher than a stock that put up the same move six months ago and has since cooled. Open-source replications (the widely-used "skyte" implementation on GitHub being the most cited) match IBD's published numbers closely.

To make the math concrete: a stock returns +10% in the most recent quarter and +5% in each of the prior three. Weighted score = 0.4 × 10 + 0.2 × 5 + 0.2 × 5 + 0.2 × 5 = 7.0. Compute that for every stock in the universe and rank them. Where 7.0 lands depends entirely on what the rest of the market did over the same year. That landing position, not the raw 7.0, is the RS Rating that ends up on the stock's tearsheet.

The IBD RS Rating is older than most of the people who use it. William J. O'Neil founded William O'Neil + Co. in 1963, built one of the first computerized stock databases, and worked the proprietary ratings (RS, EPS, Composite, Industry Group RS) into their current form over the following two decades. The 1-to-99 percentile scale has been published by the company since the 1960s. The formula in the previous section is a published replication; IBD itself only ever describes the methodology, not the exact code.

The single most-cited statistic from O'Neil's research, restated and extended through the editions of "How to Make Money in Stocks" (4th edition, 2009): the best-performing stocks of the period studied (originally 1953-1985, extended to 2008 in the 4th edition) had an average RS Rating of 87 before their major price advance. Not after. Before. The leaders were already showing institutional accumulation in the percentile data while their charts were still building bases.

Each school turns the percentile into its own rule. O'Neil tells readers to focus on RS Rating 80 or higher and to avoid the 40s, 50s, and 60s, on the basis that the historical base rate of those names becoming leaders is too poor to be worth the risk. Minervini's Trend Template — criterion eight in his 2013 book "Trade Like a Stock Market Wizard" — sets a hard floor of 70 and a preferred zone of 80-and-above. Kristjan Kullamägi works in the top 1-2% of the universe across 1, 3, and 6-month horizons, which lands at roughly the 98th percentile. The figure below pins these thresholds against the 1-to-99 scale.

A percentile is only as good as the universe it is computed against. The IBD universe is the major US listings (NYSE, Nasdaq, AMEX) filtered for adequate liquidity. A stock with an RS Rating of 95 against the IBD universe is a top-5% performer in actively traded US equities, not a top-5% performer in the global cross-section. If a vendor publishes "RS rating" against an unstated universe, ask what the universe is before trusting the number.

It is also worth being clear about what the RS Rating does not say. The percentile is a return ranking. It does not know whether the company has earnings, whether the trend is intact, whether the stock is extended above its 50-day moving average, whether the float is liquid, whether earnings are coming up next week, whether there is regulatory news in the pipeline. RS is a first filter, never the only filter. Every published methodology pairs it with chart structure and trend confirmation before any entry is considered.

The 1-to-99 percentile scale with the four decision bands and the four canonical school thresholds annotated at their published levels. The "87" marker is O'Neil's most-cited number — the average RS Rating of the best performers before their major run-up.

The on-chart RS Line is the simplest form of relative strength: a single line below or beside price, plotting the ratio of stock close to benchmark close. No smoothing, no parameters, no thresholds. Direction does the work. A line that points up means the stock is beating the market, a line that points down means it is lagging.

The pattern that matters most to swing traders is the RS Line breaking to a new 52-week high while price is still inside its base. O'Neil flagged this configuration as one of the more reliable tells of institutional accumulation. Institutions are paying up for the stock relative to the benchmark, and that buying pressure shows up in the relative ratio before absolute price has cleared the pivot. The RS-line breakout is the heads-up. The price breakout that follows it is the confirmation. The gap between the two is sometimes days, sometimes a couple of months.

IBD MarketSurge annotates this pattern as the "Blue Dot" — a small marker on the RS Line at the moment it makes a new high relative to price. A larger "leading" Blue Dot fires specifically when the RS Line's new high precedes the price breakout. Several third-party TradingView indicators reproduce the same logic for traders without an IBD subscription. The dot is just a visual aid, not a separate signal.

Stan Weinstein's refinement is the Mansfield Relative Strength line. Mansfield RS = ((stock ÷ index) ÷ 52-week SMA(stock ÷ index) − 1) × 100. The normalization centers the line on zero so a flat-but-strong ratio reads neutral rather than bullish. Mansfield crossing above zero is the third leg of Weinstein's Stage 2 entry rule (the other two being price above a rising 30-week moving average and volume expansion to roughly twice the recent average). In his vocabulary, the crossing separates "great buys" from "merely good buys."

For traders who would rather keep the chart uncluttered, the plain RS Line is enough. The new-52-week-high event is visible without normalization, absolute level on the line almost never matters, and a single ratio chart against SPY answers the leadership question for any given stock.

The configuration O'Neil watched for half a century. RS Line at a new 52-week high while price is still inside its base. Schematic, but the order of events — RS line first, price later — is the institutional-accumulation tell.

A single-horizon RS reading is incomplete by itself. A 1-week leader is sometimes a real accelerating winner and sometimes a one-news-cycle pop on an earnings reaction or an FDA approval. A 6-month leader is sometimes a structural leader with twelve straight weeks of accumulation, and sometimes yesterday's winner that has already topped and is starting to roll. The same number can mean opposite things depending on what the other horizons are doing.

That is the reason TickerStance publishes four of them every trading day: 1-week, 1-month, 3-month, 6-month. Each one answers a different question. The 1-week leaderboard catches news-driven acceleration today. The 1-month catches Stage-1-to-Stage-2 transitions and fresh base breakouts. The 3-month confirms Stage-2 trends already running. The 6-month identifies the structural leadership cohort, the names with multi-month institutional sponsorship behind them.

Reading the four together makes the regime obvious. A name top-decile across all four horizons is a real leader — what Kullamägi calls the strongest 1-2% of the market. A name top-decile on 1-week and 1-month but middling on 6-month is usually a relief bounce inside a downtrend: the recent pop is real, but the structural picture is not yet leadership. A name top-decile on 6-month but cooling on 1-week is yesterday's leader, pausing in a base or starting to roll over. Each configuration carries different odds and asks for different actions.

Disagreement is itself information. When the four horizons disagree, the resolution typically comes within two to four weeks. Either the longer horizons confirm the shorter ones (the bounce turns into a trend, the leader continues) or the shorter horizons capitulate to the longer ones (the bounce fades, the cooling continues). Watching the four together is how you avoid mistaking the second pattern for the first.

In practice, the workflow on the dashboard is short: scan the 6-month leaderboard first, pull the names that also appear top decile on 3-month, then prioritize the ones top decile on 1-month and 1-week as well. The intersection is small — usually thirty to sixty names in a healthy bull market, fewer than ten in a narrow leadership regime — and the intersection is where the best base-rate setups tend to live.

Three illustrative archetypes across the four horizons. Top decile across all four is the only configuration that earns the word "leader." The other two configurations carry different odds and call for different actions.

The academic case for relative strength is one of the strongest in empirical asset pricing, which is unusual for an idea practitioners have been using since the 1960s. Four headline papers, in chronological order, do most of the work.

Jegadeesh and Titman (1993), "Returns to Buying Winners and Selling Losers" (Journal of Finance, vol 48 no 1, March 1993, pp. 65-91). The foundational study. Sample: NYSE/AMEX stocks, January 1965 to December 1989. Strategy: rank stocks by past 3, 6, 9, or 12-month returns; long the top decile, short the bottom decile, hold for 3, 6, 9, or 12 months. Result: roughly 1% per month abnormal return on the long-short portfolio, statistically significant at the 1% level. The "J=K=6" cell (six-month formation, six-month holding) is the most-cited variant.

Carhart (1997), "On Persistence in Mutual Fund Performance" (Journal of Finance, vol 52 no 1, March 1997, pp. 57-82). Adds a momentum factor (PR1YR, later operationalized as UMD/MOM) to the Fama-French three-factor model. The four-factor model became the workhorse of mutual-fund performance attribution and explains away most of the apparent persistence in mutual fund returns. Put differently: momentum was doing the work the rest of finance had been crediting to active managers.

Asness, Moskowitz, and Pedersen (2013), "Value and Momentum Everywhere" (Journal of Finance, vol 68 no 3, June 2013, pp. 929-985). The generalization study. Tested value and momentum across eight markets and asset classes: US stocks, UK stocks, continental European stocks, Japanese stocks, country equity-index futures, government bonds, currencies, and commodity futures. Both factors appear consistently in all eight, and momentum is positively correlated across asset classes. After this paper, momentum was no longer a stock-market quirk; it was a factor cutting across the entire investable universe.

Asness, Frazzini, Israel, and Moskowitz (2014), "Fact, Fiction, and Momentum Investing" (Journal of Portfolio Management, vol 40 no 5, Sept 2014, pp. 75-92). A myth-busting paper that walks through ten common objections to momentum — that it is too small to matter, only works on the short side, only in small caps, does not survive transaction costs, and so on — and refutes each one. The standard reference when somebody attacks momentum on theoretical grounds.

One thing the literature does not tell you, and the retail trader needs to remember anyway: the academic momentum factor is a long-short portfolio rebalanced monthly with no transaction-cost or capacity constraint. Real implementations face turnover costs, liquidity friction, and tax drag that eat into the headline alpha. The retail swing-trading version — long-only, RS-screened, trend-confirmed, position-sized — is a different animal from WML / UMD. It draws on the same regularity, though: stocks that have been winning tend to keep winning over horizons of a few months.

Momentum is durable, global, and well-documented. It also has the worst tail of any commonly-traded factor. The textbook example is March-May 2009.

By March 9, 2009, the S&P 500 had fallen roughly 57% from its October 2007 high and made its bear-market closing low at 676.53. The stocks the index had punished hardest — high-beta financials, deeply cyclical small-caps, a handful of left-for-dead consumer names — sat at the bottom of every momentum percentile in the country. The academic WML strategy was structurally short that exact basket: long the past winners, short the past losers, rebalanced monthly.

From the March 9 low through the end of May, the past-loser decile (the basket WML was short) returned roughly +163% cumulative. The past-winner decile (the long side) returned roughly +8% over the same window. The long-short combination got destroyed. Daniel and Moskowitz, in "Momentum Crashes" (Journal of Financial Economics, vol 122 no 2, Nov 2016, pp. 221-247), document the episode in detail. The 1932 equivalent in the same paper is worse: WML returned −91.59% over July-August 1932 as the bear-market losers ripped 232% off the bottom while the market doubled.

Daniel and Moskowitz formalize what practitioners had already felt. The short side of momentum carries a high, time-varying negative beta. In normal markets the loser portfolio behaves roughly like the market; in bear-market panic states it behaves more like a deep call option on the market itself. When the market rebounds violently off a bottom, the losers — beaten-down high-beta junk that had been crushed — snap back hardest. The long side cannot keep pace with that snapback, and the long-short combination crashes up on the short leg.

The takeaway is not that relative strength is broken. It is that any factor with skewed kurtotic returns will occasionally produce a tail event that wipes out a year or two of alpha in a few weeks. Position sizing, trend filters, and reduced exposure when regime turns unfavorable are not extras when you trade momentum. They are the discipline that lets you survive long enough to compound the average return.

For a long-only retail trader, which is the audience this dashboard is built for, the implication is narrower. Long-only RS-screened books do not have a short leg, so they do not crash up on a bear-market rebound. They do hold the past-winner basket, which lags during the rebound and is itself exposed to mean reversion as the regime turns. Pulling exposure when the broad regime indicators (Trend, Breadth, Leadership, Macro) deteriorate keeps you positioned during the long stretches when momentum compounds steadily, and out of the way when it does not.

The cleanest momentum crash on record. Numbers from Daniel and Moskowitz (JFE 2016). The bracket on the right is what the long-short WML strategy lost: the difference between long winners and short losers.

The opposite of a momentum crash is a multi-year run, and the most recent obvious example is the 2020-2021 leadership of Nvidia (NVDA) and Tesla (TSLA). Both names sat in the top decile of every RS leaderboard for the full eighteen months, screaming through the percentile data months before the consensus narrative caught up.

Tesla returned roughly +743% over calendar 2020 and roughly +50% over calendar 2021. The S&P 500 announced TSLA's inclusion on November 16, 2020, and the stock was added in a single step before the open on December 21, 2020. Index inclusion was a forced-buying event for every passive fund benchmarked to the S&P, and the price action priced it in for weeks before the inclusion date. RS leaderboards had Tesla in the top decile by mid-2020. The S&P inclusion was confirmation, not the catalyst.

Nvidia returned roughly +122% over calendar 2020 and roughly +125% over calendar 2021. The data-center and AI-capex narrative that became the company's entire identity by 2024 was already building around the A100 (Ampere) GPU launch on May 14, 2020 and accelerated through Q3 and Q4 2020 earnings. Nvidia's RS Rating sat at the top of any momentum screener that ran during that window. The consensus equity-research narrative, meanwhile, was still calling Nvidia a "GPU company with a gaming franchise" through most of 2020. The percentile data led the cover stories by twelve to eighteen months.

For the relative-strength reader, this is the pattern to internalize. The leaderboard saw both names long before the financial press did, and traders who trusted the percentile over the prevailing narrative were rewarded. Both eventually became consensus to the point of cliché by 2024, and as that happened, leadership rotated elsewhere. RS does not tell you when to sell, but the same percentile that flagged the leader on the way up flags it cooling on the way down.

Sometimes the rest of the market is in a bear and a narrow band of names with positive RS is all there is. 2022 is the textbook example. The S&P 500 finished the year down 19.4%, the Nasdaq Composite down 33%, the "Magnificent Seven" mega-cap basket down roughly 41%. Energy finished up.

Occidental Petroleum (OXY) returned +119.10% on a total-return basis in 2022, the top performer in the S&P 500 for the year. Marathon Petroleum (MPC), Hess (HES), Valero (VLO), and Diamondback Energy (FANG) all finished in the top ten S&P 500 performers, with energy-sector returns spanning roughly +40% to +120% depending on the name. These were not the stocks that had been leading in 2020 or 2021. They were a separate basket entirely, and the rotation between them was clean.

OXY in particular is a useful illustration. Berkshire Hathaway disclosed a common-stock stake in March 2022 after buying about 136 million shares over roughly two weeks in late February through mid-March, reaching roughly 14% of the float and a position worth more than $7 billion. That buying pressure was visible in the price action well before the 13F disclosure, and OXY's RS Rating sat in the high 90s for most of 2022 as a result. (The position later climbed to roughly 28-29% by 2024.)

For a retail trader watching the leaderboards in 2022, the year required nothing more than reading the percentile honestly. The names beating SPY by 30, 50, or 100 percentage points were all in one sector, and the screener identified them in February. A diversified leadership cohort like the post-2020 norm is the easier regime to trade. A narrow leadership cohort like 2022 is harder emotionally — almost everything else looks broken — but mechanically clearer. You are either in the working basket or you are not.

The 2022 episode also gets a longer treatment on the Sector Rotation read, with the XLE-vs-XLK chart that ran the year. This section is about what happened at the stock level inside the sector that was leading. The sector view tells you where to fish; the stock view tells you which fish to keep.

Relative-strength leaderboards read clearly when leadership is broadly distributed across many names. They get awkward when leadership concentrates into a tiny basket. 2023 was the most concentrated leadership year on record outside the late-1990s.

The S&P 500 returned +26.29% in calendar 2023. The "Magnificent Seven" — NVDA, META, TSLA, AMZN, GOOGL, MSFT, AAPL — contributed roughly 62% of that gain. Individual 2023 returns: NVDA +239%, META +194%, TSLA +102%, AMZN +81%, GOOGL +59%, MSFT +58%, AAPL +49%. Equal-weight S&P 500 (the benchmark that strips out cap-weighting) returned roughly +13.87%. The cap-weight versus equal-weight gap was about 12.4 percentage points, one of the widest in three decades.

For a retail trader watching the leaderboard in 2023, the situation was awkward in both directions. The top of the leaderboard was dominated by names already among the largest stocks in the country, so buying them meant buying what was already large. Skipping them meant skipping the only basket that worked. Two equal and opposite failure modes — buying the basket extended, or missing the only working trade by avoiding it.

2024 partially resolved this. NVDA returned roughly +171% for the year, and was up roughly +150% year-to-date by mid-2024 — a single name carrying disproportionate weight in the index. Breadth widened on July 11, 2024, when a soft CPI print triggered the well-known "great rotation" out of mega-cap tech into small-caps and the S&P 493. The cap-weight versus equal-weight gap closed by roughly half over the back half of 2024 as the rally broadened.

When the leaderboard concentrates like that, the trader needs a few extra checks. Watch what equal-weight is doing — RSP (Invesco S&P 500 Equal Weight) versus SPY tells you whether breadth is widening or narrowing. Watch the cap-weight versus equal-weight performance gap, because a widening gap is a narrowing-leadership warning. And ask whether the names in the working basket are extended or in fresh bases. An RS Rating of 95 on a stock 30% above its 50-day moving average is a very different setup from 95 on a stock breaking out of a six-month base. Pure RS does not distinguish them. Chart structure does.

Four of the most-followed retail swing-trading methodologies all use relative strength as the first universe filter. The implementations differ; the rule below them does not.

William O'Neil and CANSLIM. The "L" in O'Neil's seven-letter framework stands for "Leader or Laggard." The published rule, from "How to Make Money in Stocks" (4th edition, 2009): focus on stocks with RS Rating 80 or higher; do not buy stocks with RS Rating in the 40s, 50s, or 60s. The historical study behind the rule is O'Neil's research on the best-performing stocks from 1953 to 2008, which had an average RS Rating of 87 before their major price advance. The "L" rule sits alongside the rest of CANSLIM (current quarterly earnings, annual earnings growth, new products, supply-and-demand, market direction, institutional sponsorship), but RS gates the universe before any of the others apply.

Mark Minervini and the Trend Template. From "Trade Like a Stock Market Wizard" (McGraw-Hill, 2013), the eight criteria, all of which must pass simultaneously: (1) price above the 150-day and 200-day moving averages; (2) 150-day above 200-day; (3) 200-day trending up at least one month, preferably four to five-plus months; (4) 50-day above 150-day and 200-day; (5) price above 50-day; (6) price at least 30% above the 52-week low; (7) price within 25% of the 52-week high; (8) RS Rating ≥ 70, preferably in the 80s or 90s. Together they describe a Stage 2 advance with a leadership-grade RS reading. Failing any one criterion takes the stock out of the universe.

Stan Weinstein and Mansfield Relative Strength. From "Secrets for Profiting in Bull and Bear Markets" (1988), Weinstein's Stage 2 entry requires three conditions at once: a breakout above the resistance line of the base on volume of at least 2× the recent average, price above a now-rising 30-week (about 150-day) moving average, and Mansfield Relative Strength above zero or rising. Weinstein devotes a full chapter to industry-group stage analysis and is explicit that Stage 2 stocks should only be bought when their industry group is itself in Stage 1 or Stage 2. He flatly disqualifies Stage 2 stocks in Stage 4 sectors no matter how clean the chart.

Kristjan Kullamägi (Qullamaggie). The Estonian trader who compounded a small account into nine figures publishes his approach for free on qullamaggie.com. The universe filter, paraphrased from his repeated phrasing across interviews and the FAQ: the stocks that are up the most over the last 1, 3, and 6 months. He targets the top 1-2% of the universe by momentum. He does not publish a numeric RS-Rating threshold; the "RS rank above 90" / "sector top 3" rule that circulates on Twitter is a community formalization rather than his own published rule. The three setups on his "My 3 Timeless Setups" page: Breakout (tight consolidation after a prior advance, breakout on volume), Episodic Pivot (a 10%+ gap up on news, entered on opening-range breakout above the first 1, 5, or 60-minute bar), and Parabolic Short (the third setup, a counter-trend short after an extreme blow-off run). Opening-range-breakout is the entry mechanic across the long setups, not a fourth setup of its own.

Despite the differences in vocabulary, every school does the same thing: RS screens the universe, and chart structure decides the entry. A defined base, a clear pivot, a sensible stop, supportive volume, and a hospitable regime all have to line up before the position gets written. A high RS Rating in a deteriorating regime is the right stock at the wrong time — and the right stock at the wrong time is a losing trade.

Five common ways relative strength misleads a trader who treats it as a stand-alone signal.

Falling-less is still high relative strength. In a market down 12%, a stock down 4% prints a high RS Rating despite being in absolute Stage 4 territory. The relative read is technically correct — the stock is beating the market — and the absolute read is awful, because the stock is still in a downtrend. Always pair RS with a trend filter (price above a rising 200-day moving average, for example) so you are not buying the best house in a burning neighborhood.

Momentum crashes, the 2009 case discussed above. The Jegadeesh-Titman premium is real, but the return distribution is left-skewed and kurtotic. Daniel and Moskowitz catalogue several episodes (1932, 2002, 2009) in which the long-short academic strategy lost a large fraction of its multi-decade alpha in a few weeks. Long-only retail RS books do not crash the same way, but they do underperform during the rebound from a deep regime low, and they do mean-revert as the regime turns.

Extended stocks. A top-decile RS stock that has already run 200% off the bottom is not the same setup as a top-decile RS stock breaking out of a base. The first is vulnerable to mean reversion if the regime flips; a 7% stop on a stock 25% extended above its 50-day moving average becomes a 25% drawdown to the moving average if the trend rolls. Minervini's rule applies cleanly here: the right stock at the wrong time is a losing trade, and extended is the wrong time.

Liquidity and float distortion. Thin small-caps can post extreme RS off small-dollar moves that institutional capital cannot replicate. A $3 stock running on $5M average daily volume can register an RS Rating of 99 while being uninvestable for any account above six figures. The practitioner consensus is to require a dollar-volume floor (often roughly $20 million average daily) before treating an RS reading as actionable. The RS Rating itself does not enforce this; the screener does.

Universe drift. The RS Rating is always computed against a defined universe at a point in time. When the universe changes — a major sector reshuffle, an index reconstitution, a wave of delistings — the percentile a stock gets is computed against a different set of comparators than it was the month before. Long-window backtests on RS-driven strategies have to control for universe stability or the headline alpha is partly artifact.

A short toolkit for watching RS as it happens. None of these require a paid subscription.

Stock RS leaderboards. TickerStance publishes ranked RS leaderboards for every horizon — 1 week, 1 month, 3 months, 6 months — every trading day after the close. The free dashboard shows the top of each list plus the total qualifying-stock count. Independent free options include Finviz's "performance" tab (sortable by month, quarter, half-year, year-to-date) and the IBD-style RS Rating implementations on GitHub, the "skyte" repository being the most-cited.

RS Line on every chart. Both StockCharts.com and TradingView render the ratio chart with simple syntax: NVDA/SPY in TradingView, $NVDA:$SPX in StockCharts. Drop it on the lower pane of every chart you watch. The new-52-week-high event is visible without any normalization, and the line answers the leadership question on every glance.

IBD Industry Group rankings. Investor's Business Daily ranks 197 industry groups by 6-month price performance and publishes the table daily. The top 20 is the standard hunting ground for CANSLIM-style traders. The reason the table matters is O'Neil's old claim that roughly half of a stock's move comes from its industry group: buy the leading stock in a leading group, and ignore breakouts in laggard groups no matter how clean the individual chart looks.

Equal-weight versus cap-weight breadth check. RSP (Invesco S&P 500 Equal Weight) versus SPY tells you whether broad-market leadership is widening or narrowing. A widening cap-weight outperformance is a narrowing-leadership warning even when the headline indexes are at fresh highs. The same check works at the sector level: RSPT versus XLK for Tech, RSPC versus XLC for Communications, and so on across every sector that has both products.

Relative strength shows up in two places on the dashboard. At the sector level, the relative strength of each S&P 500 sector against SPY feeds into the Leadership subscore, which composes into the daily Stance score. The signal anchoring that subscore is broad sector RS — the share of US sectors beating SPY over a rolling 6-month window. A 70% read maps to broad participation; a 30% read maps to narrow leadership. Full weighting and normalization is on the methodology page.

At the stock level, TickerStance computes ranked RS leaderboards every trading day across four horizons (1 week, 1 month, 3 months, 6 months), against an eligible universe filtered for liquidity (close ≥ $5, dollar volume ≥ $1M average daily). The free dashboard surfaces the top five names per horizon plus the total count of qualifying stocks. Pro will show the full ranked roster, with multi-year history and snapshot replay, once that tier opens.

The lists are watchlist seed material, not buy signals. Liquidity is filtered, but base structure, volume confirmation, earnings calendar, and a defined risk point are still the trader's job. The methodology page documents the exact formula, the universe filter, the weights, and the version history of every signal. Every snapshot is reproducible — given the same inputs the same Stance and leaderboard order come out — and a version field on every record keeps historical reads auditable a year from now.

Honestly, a trader who has read this far does not need TickerStance to read relative strength. Watching the four horizon leaderboards, the IBD industry group table, and the cap-weight versus equal-weight gap for fifteen minutes a day produces the same signal. The dashboard mainly does the bookkeeping, so the regime read is repeatable across days, across users, and across regime turns.

Relative Strength (RS) — A measure of how a stock's return compares to a benchmark or to other stocks over a defined lookback window. Implemented as a price ratio (RS Line), a percentile (RS Rating), or an excess-return number depending on the platform.

IBD RS Rating — Investor's Business Daily's 1-to-99 percentile rank of a stock's 12-month return, weighted 40% on the most recent quarter and 20% on each of the prior three. Introduced by William O'Neil in the 1960s. 99 is the strongest, 1 is the weakest.

RS Line — A chart of stock close ÷ benchmark close (typically SPY or the S&P 500), plotted as a line over time. Direction is what matters; rising means the stock is beating the market.

Mansfield Relative Strength (MRS) — Stan Weinstein's normalized RS line: ((stock ÷ index) ÷ 52-week SMA(stock ÷ index) − 1) × 100. Centered on zero; above zero is bullish leadership, below zero is laggard.

RS Rating ≥ 80 / ≥ 90 — The two practitioner thresholds. O'Neil's working zone is 80 plus. Minervini's preferred zone is 80-90. Kullamägi's top 1-2% maps to 98 plus. The 87 historical-average figure comes from O'Neil's study of the best-performing stocks 1953 to 2008.

RSI (Relative Strength Index) — J. Welles Wilder's 1978 momentum oscillator, formula RSI = 100 − 100 ÷ (1 + RS) where RS = avg gain ÷ avg loss over 14 periods. Bounded 0 to 100. Convention: above 70 overbought, below 30 oversold. Compares a stock to itself, not to other stocks.

Cross-sectional momentum — The Jegadeesh-Titman (1993) factor: rank a universe of stocks on past 3-12 month returns, long the top decile, short the bottom decile. The basis of the WML / UMD factor that lives in the Carhart four-factor model.

Time-series momentum — Moskowitz, Ooi, and Pedersen (2012, JFE) variant: each asset evaluated against its own past return, not against the cross-section. Significant across equity index, bond, currency, and commodity futures.

WML / UMD — "Winners minus losers" / "up minus down" — the long-short academic momentum portfolio. The fourth factor in the Carhart four-factor model, alongside MKT, SMB, HML.

Momentum crash — A short-window collapse of the long-short momentum strategy, typically triggered when a violent rebound from a bear-market bottom causes the past-loser short leg to rip up much faster than the long winners. Documented in Daniel and Moskowitz (JFE 2016). 1932 and 2009 are the canonical episodes.

Multi-horizon RS — Reading 1-week, 1-month, 3-month, and 6-month percentile ranks together to separate true leaders (top decile across all four) from single-horizon flukes (top on one horizon only).

Pre-breakout RS-line tell — The configuration where the RS Line breaks to a new 52-week high while price is still inside its base. O'Neil-recognized institutional-accumulation signal; annotated as the "Blue Dot" in IBD MarketSurge.

Sector-first rule — Stan Weinstein's requirement that a Stage 2 stock be bought only inside a Stage 1 or Stage 2 sector. The sector context overrides individual chart strength regardless of how clean the base looks.

Frequently asked questions

Keep reading

• Stock LeadersHow TickerStance ranks stocks across four RS horizons every trading day.
• Sector RotationHow sector relative strength reveals where capital is rotating right now.
• Market RegimeWhy regime context determines whether RS plays succeed or whipsaw.
• MethodologyEvery weight, normalization rule, and signal that powers the Leadership subscore.