Skip to main content

Relative Strength vs the Index

Updated Jun 23, 2026 at 8:47pm

Research Draft Medium 1,178 words

Relative strength (RS) measures how a stock performs against a benchmark — usually a broad index like the S&P 500 — rather than in absolute terms. The core construct is the RS ratio line: stock price divided by the index, plotted over time. A rising line means the stock is outperforming (advancing more, or falling less, than the index); a falling line means it is lagging. The closely related percentile RS rating (e.g., IBD's 1–99 scale) ranks a stock's trailing return against the whole market. This is the cross-sectional / relative momentum concept — and, unusually for a technical tool, it is backed by some of the most robust evidence in academic finance. Its central tension is that the edge is real but cyclical: relative-strength leadership tends to persist for months, then occasionally reverses violently in the worst possible moments.

Critical naming caveat (read first): "Relative strength vs the index" is not the Relative Strength Index (RSI). RSI is Wilder's bounded 0–100 oscillator of a single security's internal up/down momentum — a different, weakly-evidenced tool that merely shares the word "strength." Do not let RSI borrow this concept's academic credibility, and vice versa. This doc is about relative (cross-asset) strength; RSI lives under oscillators.

How it's calculated / formed

RS ratio line (Price Relative). Per StockCharts ChartSchool, the formula is simply:

Price Relative = Base Security price / Comparative Security price

So SBUX:$SPX is Starbucks divided by the S&P 500. (StockCharts scales the open/high/low by the comparative security's close.) The IBD-style RS line commonly multiplies by 100 for readability. The absolute level is meaningless — only the slope and trend of the line carry information.

Percentile RS rating. IBD's RS Rating compresses trailing relative return into a 1–99 percentile, where 99 means the stock outperformed 99% of all listed stocks over the lookback. Per IBD's published methodology, it is roughly a trailing-12-month measure that weights the most recent quarter more heavily than the prior three, to surface emerging leadership. The exact proprietary weighting is not public; open-source replications (e.g., the "skyte/relative-strength" project) approximate it with quarterly return blends.

How to read it

  • Rising RS line → relative outperformance; the stock is a leader.
  • Falling RS line → relative underperformance; a laggard.
  • New high in the RS line before price makes a new high is treated by O'Neil-school traders as a bullish "tell" — relative strength leading absolute price.
  • RS divergence: RS line falling while price still rises is relative weakness building under the surface (bearish); RS rising while price falls is relative resilience (bullish).
  • RS rating thresholds (IBD convention): ≥80 = top quintile / "market leader" territory; O'Neil's historical research reported the biggest winners averaged an RS Rating near 87 before their major advances. These are heuristics, not guarantees.

How it's used in practice

The style-agnostic uses are:

1. Leadership selection / stock screening. Buy what's already winning. Filter the universe to high-RS names (e.g., RS rating ≥80–90) and ignore laggards — the foundational idea behind CAN SLIM and most "strong stocks in strong groups" approaches. 2. Sector / group rotation. Run the ratio line of each sector vs the index to see where money is rotating, then concentrate in the leading groups. This is the canonical relative-rotation use (and the sibling node on rotation covers it in depth). 3. Trend / breakout confirmation. A price breakout accompanied by an RS line also breaking out is higher-conviction than price alone. 4. Pair / relative trades. Long the strong, short the weak — the explicit long/short form the academic literature tests.

Exact swing entry/stop/target mechanics on RS leaders are deferred to the Swing Trading branch — do not duplicate them here.

Standing & evidence

This is the rare technical concept with strong positive academic support, as a cross-sectional momentum factor:

  • Jegadeesh & Titman (1993), Journal of Finance, "Returns to Buying Winners and Selling Losers," established the anomaly: buying past 3–12-month winners and selling past losers earned significant excess returns over the next 3–12 months. They reported the winner-minus-loser strategy averaged roughly ~1% per month, with their 12-month-formation / 3-month-hold variant near ~1.31%/month. This put price momentum on the academic map and is one of the most replicated findings in finance.
  • Asness, Moskowitz & Pedersen (2013), "Value and Momentum Everywhere," Journal of Finance, found the momentum premium positive and significant across eight markets/asset classes (US/UK/Europe/Japan equities, country indices, bonds, currencies, commodities) — strong evidence it is not data-mined. (They also document momentum's ~−0.5 to −0.6 correlation with value.)
  • The crash caveat — Daniel & Moskowitz (2016), "Momentum Crashes," Journal of Financial Economics (122:221–247): momentum returns are strongly negatively skewed. Crashes are partly forecastable — they cluster in "panic" states (after large market declines, in high volatility) and coincide with sharp market rebounds, when beaten-down former losers snap back hardest. They document severe crashes after the 1930s and in 2009. So the high average return comes bundled with rare, brutal drawdowns.

Honest scoping of the edge: the documented edge is for systematic, diversified, cross-sectional momentum portfolios. A retail trader buying a handful of high-RS-rating names is taking a far less diversified, higher-idiosyncratic-risk bet than the academic long/short construct — the factor evidence does not transfer one-for-one to concentrated single-name discretionary trades.

Strengths & limitations

  • Strengths: rare TA tool with genuine, multi-decade, multi-market empirical backing; intuitive; works as both a screen and a confirmation overlay; the ratio line is regime-readable in one glance.
  • Limitations / failure modes:
- Crash risk (Daniel-Moskowitz): the worst losses come precisely at panic-bottom rebounds — high-RS leaders can be the prior-regime winners that get sold to fund the rotation. - Regime dependence: momentum underperforms sharply at sharp trend reversals and in choppy, mean-reverting regimes. - Lookback sensitivity: very short windows pick up noise/mean-reversion; the classic edge sits in the 3–12-month band (a 1-month reversal effect runs the opposite way). - #1 misuse: confusing it with RSI — and, secondarily, chasing a high RS rating with no risk control, treating a backward-looking percentile as a forward guarantee.

System relevance

This is a TA-definition node. Operational swing mechanics (entries, stops, targets, hold periods on RS leaders) live in the Swing Trading branch; sector/group application lives in the sibling Relative Strength & Rotation nodes — cross-link, don't duplicate. For Delvantic's analysis pipeline, RS is a leadership/regime input: the Augustus trade-setup agent should weight it as a documented-but-cyclical edge, and explicitly down-weight or flag high-RS setups during the panic/high-volatility/post-decline conditions that Daniel-Moskowitz identify as crash-prone — deferring the live verdict to Augustus + Cairn's measured record, not to this doc.

Sources