What is a continuation pattern actually made of?

Strip away the names — flag, pennant, triangle, rectangle — and every "continuation pattern" is the same event: a fast move followed by a pause. The shapes differ only in the geometry of the pause. Learning what causes the pause is worth more than memorizing twenty names for it.

Who is actually in the pause

A price jumps 25% in three weeks. Four groups now have a decision to make:

• Early buyers sitting on a profit. Some take it. Their selling is supply that didn't exist before the move.
• Buyers who missed it. They want in, but not at the extreme. They bid a little lower and wait.
• Holders who bought at the old, much lower price. The move handed them an exit they'd given up on.
• Nobody in particular. Attention fades. The news that caused the move is a week old.

Put those together and you get a narrow band: profit-taking caps the upside, patient bidders cap the downside, and volume falls because most of the people who cared strongly have already acted. The tight range on quiet volume is the pattern. The lines you draw around it are bookkeeping.

A worked example

Rounded and illustrative: a stock trades near $40, then runs to $52 over three weeks on roughly triple its normal volume. For the next two weeks it oscillates between $49 and $52, and daily volume drops to about half its average.

That box is the whole pattern. Read it as a sentence: the crowd repriced this company quickly, and now the buyers and sellers who are left disagree by only about 6%, without much conviction on either side. No forecast in that sentence — just a description of a standoff.

Why the name "continuation" is already a bias

The category is named after what happened next in the examples people remember. A pause is a pause. It resolves upward, downward, or into a longer pause, and only after it resolves does anyone call it a continuation pattern. Draw the same box in real time and you're holding a range with an unknown exit.

This isn't a small quibble — it's the theme of this whole course. Published tests of pattern reliability are genuinely mixed: some find measurable information in certain shapes, others find nothing once you fix the definitions and subtract trading costs. Treat continuation patterns as vocabulary for describing behaviour, not as machinery that predicts direction.

Try it now

1. On the schematic above, measure the run and then the box: the % size of the move, and the box's width as a % of price. Cover the right-hand half and ask yourself, honestly, which way it leaves. Nothing on the page answers that, which is the point of the lesson.
2. Now on prices nobody arranged. In the five years below, find one obviously sharp move and the pause that followed it, and write down the pause's high, low and % width. Then compare participation: the pane under the chart shows each session's volume with its 50-session average drawn over it. Did participation fall the way the story above predicts?

3. Now hunt for a failure: find a pause after a sharp up move that broke downward out of its range. It took you very little searching, didn't it? Keep that experience — you'll need it in Unit 4.