The Head and Shoulders Trading Strategy (And Inverse H&S)

Consider a heavily-traded semiconductor stock, ticker XYZ, that's rallied hard in the pre-market and continues higher after the open, running from $118.00 to a first peak of $124.60 by 10:05 AM ET on RVOL of 3.1x. That's the left shoulder. It pulls back to $121.80 by 10:22 AM, a trough that will define the left side of the neckline.

Buyers push back in, and XYZ makes a new high of $126.90 at 10:51 AM, this time on RVOL of only 2.2x, noticeably lighter than the left shoulder's push despite the higher price. That's the head. It pulls back again, this time to $121.50 by 11:15 AM, a trough that sits almost exactly level with the first one, defining a roughly horizontal neckline around $121.60 to $121.80.

XYZ attempts one more push higher, reaching $123.90 by 11:48 AM, clearly below the head's $126.90 and only modestly above the left shoulder's $124.60, on RVOL that's fallen further to 1.4x. That's the right shoulder, and the declining volume across all three peaks is the pattern's clearest tell that buying conviction has been fading with every attempt.

At 12:31 PM, a 5-minute candle closes at $121.15, below the neckline's roughly $121.70 level, on volume 1.8x the prior ten-bar average. A trader entering the short at $121.15 sets a stop above the right shoulder's $123.90 high, a $2.75 risk per share, and calculates the target: neckline ($121.70) minus the head-to-neckline distance ($126.90 minus $121.70, or $5.20), projecting toward roughly $116.50.

XYZ initially drops to $119.40 by 1:05 PM, a solid move past the 0.618x extension, where a trader following the plan scales out a portion of the position and moves the stop down to just above the neckline. The stock consolidates through the early afternoon before a broader market wobble accelerates the decline, and XYZ reaches $117.30 by the close, short of the full projection but a well-managed execution of the pattern nonetheless.

The neckline pullback is normal, not a failure signal, provided it doesn't close back above the line. A very common sequence, occurring roughly two-thirds of the time in the broader statistical research on this pattern, is price breaking the neckline, then drifting back up to retest it from underneath before resuming the decline. A trader who panics and covers on the first bounce toward the neckline is often exiting a still-valid trade prematurely; the actual invalidation condition is a close back above the neckline, not a wick or an intraday touch of it.

Volume on the breakdown candle matters more than the size of the first move. A neckline break on light volume is meaningfully less trustworthy than one on a clear volume expansion, even if the initial price move looks similar. This is consistent with the broader theme across this hub's breakout coverage: price without volume is a weaker signal than price with volume, and a three-peak pattern is no exception to that rule.

Trailing the stop at the neckline itself, once price has cleared it by a reasonable margin, is the cleanest way to lock in the trade's core thesis. Since the neckline pullback is such a common event, moving the stop to just above (short) or below (long) the neckline, rather than to the exact entry price, gives the trade room to survive that normal retest while still protecting against the pattern fully failing.

Independent statistical research into this pattern, based on Thomas Bulkowski's large-sample study of confirmed formations on daily charts, found the head and shoulders top reaches its measured-move price target only about half the time, despite a relatively low break-even failure rate, meaning a meaningful share of confirmed breakdowns stall out well short of the full projected decline. The inverse head and shoulders bottom performed notably better in that same research, both in break-even failure rate and in the share of patterns reaching their full target. Both figures come from multi-week daily-chart research, not the compressed intraday version this guide covers, so treat them as evidence the underlying structural logic works, not as an exact intraday probability.

The most common real-world failure is misreading an ordinary pullback-and-continuation sequence in a strong trend as a head and shoulders top. A stock in a genuinely powerful uptrend can produce three peaks with a shallow dip between the second and third simply because trends don't move in a straight line; treating every such wiggle as a reversal pattern leads to shorting into strength repeatedly, a costly and avoidable mistake. The declining-volume signature described earlier in this guide is the best available filter against this specific failure, since a genuine continuation typically doesn't show the same fading conviction across all three peaks that a real reversal does.

A second failure mode is an asymmetric pattern being forced into the head and shoulders label. If the right shoulder is dramatically smaller or larger than the left shoulder, or if the two troughs sit at wildly different levels, the pattern lacks the structural symmetry that makes the measured-move projection meaningful in the first place. A trader tempted to draw a neckline through two very different-looking troughs just to complete a pattern they've already decided is there should treat that temptation as a warning sign, not a green light.

A third failure mode, specific to the intraday version, is a right shoulder that never fully forms because the session simply runs out of time. A head and shoulders that has a clear left shoulder and head by early afternoon, but whose right shoulder is still forming into the final hour of trading, is a pattern still in progress, not a confirmed setup, regardless of how compelling the first two-thirds of the shape already look.

The complex head and shoulders has two left shoulders and two right shoulders instead of one each, essentially a five- or seven-peak version of the same underlying exhaustion story. These take longer to form and are less common intraday, but the same neckline-break logic applies once the full structure is visible.

The sloped-neckline version is more common than the textbook horizontal neckline and simply requires drawing the line through the actual troughs (or peaks, for the inverse) rather than forcing a flat line onto an unequal structure. A rising neckline on a top pattern is actually a stronger bearish signal when broken, since it means sellers overcame not just a static level but a rising one.

The failed head and shoulders, where price breaks the neckline, reverses, and pushes back above the right shoulder's high, is itself a tradeable signal in the opposite direction, similar in spirit to this hub's failed breakout, bull trap, and bear trap guide. A head and shoulders that fails this dramatically often triggers a fast short squeeze as trapped short sellers cover.

Manually tracking peak heights, trough levels, and the volume signature across three separate swings, in real time, across a watchlist, is genuinely difficult without help. A scanner capable of flagging declining relative volume across successive price highs, such as Trade Ideas, can surface candidates showing the fading-conviction signature this pattern depends on well before the third peak and neckline break are visible to the naked eye.

A charting platform with a reliable trendline and measured-move tool is essential here too, since an inaccurately drawn neckline throws off both the entry trigger and the price target. For a broader look at charting platforms suited to pattern work like this, see the tools and reviews section.

The head and shoulders is a reversal-hunting tool, and it belongs in a trading plan alongside other reversal confirmation methods rather than as a lone signal to act on. Pairing it with this hub's reversal trading playbook for the broader context of what makes any reversal attempt credible, and with trendline and channel reading for the underlying skill of drawing an accurate neckline, gives the pattern more support than trading the shape in isolation.

Because the stop sits at a specific structural level, above the right shoulder for the top pattern, below it for the inverse, position sizing should be calculated directly from that distance using standard risk-per-trade math rather than a flat share count applied regardless of how wide or tight that particular pattern's shoulder-to-neckline distance happens to be.