Chart pattern
Also called price pattern, chart formation
A recognizable shape that price traces on a chart, such as a head and shoulders or a triangle. Analysts group patterns into reversal and continuation types; spotting them involves judgment.

On this page· 5 sections
Chart pattern analysis has been in wide use since at least the 1920s. Before US markets were regulated by the Securities and Exchange Commission, charts were one of the few reliable sources of information traders had, and many searched them for signs of market manipulation[1].
The shapes they looked for got names: gaps, flags, pennants, wedges, triangles, head-and-shoulders and various tops and bottoms[5]. Formal tests came much later. Levy tested 32 five-point formations on NYSE stocks in 1971[5], and in 1995 a New York Fed study gave the head-and-shoulders pattern a rigorous test[1].
How it works
Analysts sort most patterns into two groups. Reversal patterns point to a change in trend; continuation patterns point to a pause, after which the earlier direction resumes[4]. Some shapes, such as the rectangle, can belong to either group, depending on the price action before them, on volume and on other indicators[4].
The building blocks are swing highs and swing lows joined by lines. StockCharts calls chart patterns more complex versions of trendlines[4].
Patterns come in every size. Some form in a single session, others take many months or years, and they appear on intraday, daily, weekly and monthly charts[4].
Researchers have tried to take the eye out of it. Lo, Mamaysky and Wang smoothed prices and let an algorithm find the patterns, then applied it to a large number of US stocks from 1962 to 1996[2].
Example
Say a stock index makes three highs over two months: 2,000, 2,030 and 2,010. The lows between them are 1,850 and 1,875.
- One analyst draws the top boundary through the highest high, 2,030.
- Another draws it through the average of the three highs: (2,000 + 2,030 + 2,010) ÷ 3 = 2,013.3.
- The index closes at 2,020. For the second analyst that is a breakout, 0.3% above the line. For the first, price is still inside the pattern.
Neither drawing breaks a textbook rule. The two analysts disagree only on where the pattern ends, and that is enough to give opposite readings of the same close.
Watch out
Automation does not remove judgment. The CFA Institute Research Foundation review accepts that some subjectivity remains in pattern analysis, even when the analysis uses automated rules[1].
The evidence points both ways. Levy’s 1971 study of 32 five-point chart formations on NYSE stocks found that none, for any holding period, beat the average opportunity to buy or sell short[5]. Lo, Mamaysky and Wang, with their algorithm, found that several patterns did provide incremental information on US stocks over their sample[2].
Patterns fail. The ASX course says that the patterns classic technical analysts trade do not work all the time, like every tool in trading and investing[3].
Sources
- Professional bodyCFA Institute Research FoundationTechnical Analysis: Modern Perspectives (Literature Review; Scott, Carr & Cremonie, 2016)Checked November 9, 2026
- ResearchNational Bureau of Economic ResearchFoundations of Technical Analysis: Computational Algorithms, Statistical Inference, and Empirical Implementation (Lo, Mamaysky & Wang; NBER WP 7613; Journal of Finance 55(4), 2000)Checked November 9, 2026
- ExchangeASX (Australian Securities Exchange)Shares Online Course 11: Technical analysis (version 5, November 2010)Checked November 9, 2026
- ReferenceStockCharts ChartSchoolIntroduction to Chart PatternsChecked November 9, 2026
- ResearchUniversity of Illinois at Urbana-Champaign (farmdoc)The Profitability of Technical Analysis: A Review (Park & Irwin, AgMAS Project Research Report 2004-04, October 2004)Checked November 9, 2026
