A chart lands in your feed. It is the Philadelphia Semiconductor Index, there are three arcs drawn under the lows, a dashed line running across the top of them, and a projected target sitting slightly above the index's previous record high. It reads like a prediction.
It is not one. Most of what you are looking at is arithmetic, and telling the arithmetic apart from the interpretation is the entire skill. This post walks through both, using that setup as the worked example.
The pattern is four points, not a shape
An inverse head and shoulders is defined by a sequence, not by how symmetrical it looks:
- A first low, the left shoulder.
- A deeper low that follows it, the head.
- A third low, roughly level with the first and clearly above the head, the right shoulder.
- A neckline: the line drawn across the two rallies that separate those lows.
On the chart in question the head prints near 10,430. The neckline runs close to 12,660. The right shoulder bottoms in the 11,800 to 12,100 band, well above the head, which is the condition that makes it a shoulder rather than the start of another leg down.
Symmetry is a nice-to-have, not a requirement. Shoulders are rarely the same depth and almost never the same width. What matters is the sequence: a lower low, then a higher low, with a definable ceiling across the top that the market has already failed at twice.
That ceiling is the point. The pattern is not really about the arcs. It is about a level that has rejected price repeatedly, and a series of lows that stopped getting lower while it did.
The target is subtraction, not prophecy
The projected target comes from a convention called the measured move. You take the vertical distance from the head to the neckline and project the same distance up from the neckline:
- Neckline near 12,660, head near 10,430, so the height is roughly 2,230 points.
- Project that up from the neckline and you land around 14,890.
The index's record high sits near 14,655. So the projection lands roughly one and a half percent above it — which is where "target just above the all-time high" comes from. Nobody forecast that. It is one subtraction and one addition, and any two people drawing the same neckline get the same answer.
Which is also the limitation. The measured move carries no probability and no deadline. It does not say the index will get there, and it does not say when. A pattern that resolves over three weeks and one that resolves over nine months produce identical targets. Treating that output as an expectation rather than as a reference level is the most common way people misuse it.
Confirmation is a close, not a touch
At the time of the chart, spot sits near 12,440 — below the neckline. That matters more than the drawing does.
Until the index closes above the neckline, the pattern is a sketch of something that has not happened. Price tagging the line intraday and falling back is not a breakout; it is the third rejection at the same ceiling, which is what built the pattern in the first place.
The other half of the discipline is knowing what kills it. The higher-low sequence is the structure. A close back below the right shoulder's low, near 11,790 on this chart, breaks that sequence and the setup no longer describes the market. Deciding that level in advance is what separates reading a chart from arguing with one — after the fact, there is always a way to redraw the neckline so the pattern survives.
The prior high is the interesting part, not the pattern
The detail worth noticing here is not the shape. It is that a target of roughly 14,890 and a record high of roughly 14,655 are, for practical purposes, the same place.
Those are two independent things pointing at one area. One is the pattern's own arithmetic, which knows nothing about history. The other is where the last cohort of buyers is sitting underwater and has spent months waiting to get out at breakeven. Supply tends to be heavier there for reasons that have nothing to do with chart geometry.
When a measured move and an old high land in the same neighbourhood, the sensible reading is that the area is likely to be contested, not that it is likely to be reached. Convergence tells you where the argument happens. It does not tell you who wins it.
It also sets up the more useful question, which is not "does it hit the target". It is what happens on the first attempt at the old high: an index that pushes through a record on its first try is telling you something different from one that stalls under it for a month, and both are outcomes the pattern is perfectly compatible with.
What a pattern like this cannot tell you
A chart pattern is a description of what already happened, drawn by someone who knows how it turned out. That is worth remembering when the shape looks obvious in hindsight: you remember the necklines that broke, because those are the ones people post.
It also says nothing about the businesses inside the index. Semiconductor earnings, capital spending cycles, and inventory corrections drive that index over any horizon that matters, and none of them appear in the drawing.
VelaDeck does not trade chart patterns, and the strategies it mirrors do not either. They rebalance when their composition changes, on a monthly or quarterly cadence, and every resulting order traces back to a specific holding entering or leaving a portfolio. That is a deliberate choice rather than a verdict on technical analysis: a rules-based system has to be testable, and "the neckline looked convincing" is not a rule you can backtest or a signal you can hand to an execution engine.
If the underlying theme is what interests you rather than the setup, IT15 (Tech Titans) is the higher-beta technology composition in the catalog — 15 holdings, rebalanced monthly. Its published figures are historical and simulated, and they describe what the composition did, not what it will do.
Frequently asked questions
Is an inverse head and shoulders a bullish signal?
It is conventionally read as a bottoming structure, but it signals nothing until the neckline is closed above. Before that it describes a market that has stopped making lower lows and is still failing at the same ceiling. Both facts are true at once, and the second one is why the pattern fails as often as it does.
How reliable is the measured move target?
Treat it as a reference level, not a forecast. The calculation is deterministic — same neckline, same head, same answer — but it carries no probability, no time frame, and no adjustment for where the price is coming from or what is happening in the underlying companies.
What invalidates the pattern?
A close back below the right shoulder's low. That is the level that keeps the higher-low sequence intact, and once it goes the structure the pattern is built on no longer exists. Picking that level before the trade rather than after is the whole discipline.
Does VelaDeck act on technical setups like this one?
No. VelaDeck mirrors AI-curated equity compositions into your own Alpaca account and generates orders when those compositions change. It does not read charts, it is not an investment advisor, and it never takes custody of your money — the account and the funds stay yours. Live trading is opt-in and off by default.
What to do with this
The useful takeaway is not the target. It is the habit: separate the parts of a chart that are arithmetic from the parts that are a story, and write down what would prove you wrong before you need it.
If you would rather your portfolio moved on published composition changes than on your read of a neckline, that is what VelaDeck automates. Start in paper mode, assign a notional amount to a strategy, and watch a month of signals land before deciding whether to enable anything with real money.