Skip to main content

Head-and-Shoulders (& Inverse)

Updated Jun 23, 2026 at 8:47pm

Research Draft Medium 790 words

The head-and-shoulders top is a major reversal pattern that forms after an uptrend: a left shoulder (a peak followed by a pullback), a higher head (a higher peak), then a lower right shoulder roughly level with the left. A "neckline" is drawn connecting the two intervening reaction lows. The pattern signals that buyers can no longer push to new highs and that demand is exhausting. The inverse head-and-shoulders (also called head-and-shoulders bottom) is the mirror image at the end of a downtrend — two troughs flanking a deeper central trough — signaling a potential bottom and trend change to up.

The setup

The pattern is not valid until the neckline breaks — the shape alone is just three bumps. Confirmation is a decisive close below the neckline (top) or above it (inverse), ideally on expanding volume. StockCharts treats the neckline break as the trigger and warns against acting before it.

  • Trigger: close beyond the neckline (down for top, up for inverse).
  • Entry: on the break, or on the frequent pullback/throwback that retests the neckline as new resistance/support — the retest entry gives a tighter stop.
  • Stop: above the right shoulder (top) or below it (inverse). Some traders use above the neckline retest for a tighter stop.
  • Measured-move target: measure the vertical distance from the head to the neckline, then project that distance from the breakout point. For a top with head $538 and neckline $510 ($28), the target is roughly $510 − $28 = $482.
  • Volume: classically highest on the left shoulder, lighter on the head, lightest on the right shoulder (waning demand), then a surge on the neckline break. For the inverse, the volume surge on the upside neckline break is the most important confirmation. StockCharts flags volume as one of the two things to get right (the other being the neckline itself).

Base rates & evidence

Thomas Bulkowski's chart-pattern statistics (from a hand-curated sample of ~2,800 patterns, bull-market conditions) give the head-and-shoulders top a 19% break-even failure rate, an average decline of ~16%, a measured-move target hit rate of only ~51%, and a pullback rate of ~68% — pullbacks are the norm, not the exception. He ranks it among the better-performing reversal patterns (rank 9 of 36 in his current numbering). These are descriptive, hand-selected statistics on idealized patterns, not a tradable backtest net of costs — treat them as a quality benchmark, not an edge.

Academic evidence is genuinely mixed. Lo, Mamaysky & Wang (2000, Journal of Finance) automated pattern recognition with kernel regression over 1962–1996 U.S. stocks and found H&S (and several other patterns) carry some incremental information versus the unconditional return distribution. But Osler's New York Fed work on U.S. equities framed H&S trading more as noise-trader behavior than reliable edge, and Osler & Chang's currency study is the most positive case. Replications (e.g. the Brazilian-market study adapting Osler & Chang) repeatedly find that gross gains shrink or vanish once realistic transaction costs are applied. The honest read: the pattern likely contains real information, but turning it into net profit after costs is far from guaranteed.

Strengths & limitations

  • Counter-trend. You are betting against the prevailing trend, which is inherently lower-probability than trend-continuation setups — wait for confirmation.
  • Subjective identification. What counts as a "shoulder," how level they must be, and where the neckline sits are judgment calls; the same chart yields different patterns to different traders. This is the core academic critique.
  • #1 misuse — anticipating the break. The most common error is entering on the right shoulder before the neckline breaks, assuming the pattern will complete. Many never break; price often resumes the trend. No neckline break = no trade.
  • Modest target reliability. Only about half of confirmed patterns reach the full measured move (Bulkowski), so scaling out or trailing a stop is prudent rather than holding for the projection.

System relevance

Within this RAG, the head-and-shoulders top/bottom belongs to Augustus (the pattern-recognition / setup-classification layer) as a confirmed-on-break reversal candidate. The operational rule for the system mirrors the trading rule: flag the shape as forming, but only treat it as actionable on a volume-confirmed neckline break, and carry the measured-move target as a soft objective (~50% hit rate) rather than a hard expectation.

Sources