Read pivots before naming a shape
A pattern name can make a chart feel certain when the geometry is still unfinished. Start with observed swing highs and lows, the prior trend, and the level where a trader's thesis would change. Only then ask whether the structure resembles a rectangle, triangle, double top, flag, or head and shoulders. This order reduces the temptation to fit a familiar silhouette to noisy prices.
The current CMT Program Guide includes classical reversal and continuation patterns, gaps, candlesticks, and point-and-figure in Level I learning objectives. Level II extends the topic to price objectives and breakout confirmation. This article teaches the chart-reading workflow. For a broader pattern reference, see Chart Patterns for the CMT Exam.
A five-step chart worksheet
| Step | Question | Evidence to record |
|---|---|---|
| 1. Prior movement | What did price do before the formation? | Trend direction and last meaningful swing |
| 2. Structure | Which highs and lows bound it? | Prices, dates, and chosen scale |
| 3. Trigger | What would complete the pattern? | Close, intraday break, or other declared rule |
| 4. Measurement | What scenario does the height suggest? | Arithmetic or proportional method |
| 5. Failure | What event contradicts the thesis? | Return into range, broken support, time limit |
This is a study tool, not a universal trading system. Different instruments, time horizons, and volatility states can make the same visual form behave differently. The trigger should be chosen before the outcome is known.
Worked range-break example
Suppose an asset rises from 80 to 96, then trades between support near 92 and resistance near 100. The range height is 100 − 92 = 8. A close at 101 clears the resistance by one point. An arithmetic measured objective is 100 + 8 = 108. That objective is a way to describe a possible move, not a forecast or guaranteed target.
To turn this into a conditional plan, an analyst might require a daily close above 100 and a following close that stays above 100. If the second close is 98, the confirmation fails. If the asset holds above 100 but then closes beneath 92, the bullish range-break idea is fully contradicted. Depending on the strategy, a return below 100 may be enough for an earlier exit. State the exact rule and the cost of being wrong before treating the objective as actionable.
Volume adds context. If breakout-day volume is 1.3 million shares and a comparable 20-day daily average is 1.0 million, participation is 30% above that benchmark: (1.3/1.0 − 1) × 100 = 30%. That supports the observation that the break involved more activity than usual, but it does not reveal whether the traders were informed or whether the move will persist. Avoid the statement that a single volume cutoff “confirms” all patterns.
Reversal versus continuation
A double top requires a preceding advance, two test areas, and a neckline or intervening low. Two similar highs alone are not enough. If the peaks are 120 and 119 and the intervening low is 110, a breakdown through 110 completes one common interpretation. The top-to-neckline height is about 10 points, giving an arithmetic scenario of 100. If price recovers above the neckline, revisit the bearish interpretation. If it exceeds the peaks, the reversal thesis loses much of its basis.
A flag is a compact pause after a strong move, so the pre-pattern advance matters. A sideways box after a long decline cannot be called a bullish flag just because it is rectangular. A symmetrical triangle has converging boundaries; its eventual break direction must be observed rather than assumed. A head and shoulders requires three prominent highs and a neckline, yet real charts are imperfect. The right shoulder can differ in width and height. Define tolerances before seeing the breakout.
Gaps need context, too. A price jump over prior resistance can signal urgency, but splits, illiquid quotes, or overnight news can create misleading displays. Check whether price series are adjusted and whether volume is meaningful for the venue.
Scaling and chart construction change the picture
On an arithmetic chart, a rise from 10 to 20 and a rise from 100 to 110 each occupy ten price units. On a log chart, the first is +100% and the second is +10%, so the first is visually much larger. A multi-year chart spanning large percentage changes can tell a different story when rescaled. If a pattern objective is drawn by measuring vertical height on an arithmetic chart, do not silently transfer the same pixel distance to a log chart. The official guide includes scaling as a separate learning objective because it affects apparent trendline slope and measured objectives.
Candlesticks compress open, high, low, and close for an interval. A doji near support might show indecision, but one candle does not establish a completed reversal. Compare the next bar, volume, and the surrounding structure. Point-and-figure removes uniform time spacing and depends strongly on box and reversal settings; a pattern there is not automatically identical to a bar-chart pattern.
Mistakes to avoid
Do not redraw support after the move to make a failed break look successful. Do not call a neckline breach when only an intraday wick crossed your closing-price rule. Do not measure a target before defining the actual neckline or boundary. Do not use a pattern's attractive reward distance without comparing it with the loss at invalidation. And do not generalize from one clean textbook chart to every asset. Examples are illustrations, not estimates of pattern success rates.
Self-check
A rectangular consolidation has resistance at 74 and support at 68. Price closes at 75, then at 73. If your rule requires two closes above resistance, is there a confirmed breakout? What is the arithmetic measured objective? Answer: No; the second close is back below 74. The height is 74 − 68 = 6, so the illustrative upward objective would have been 80. A calculated objective remains a scenario even though the chosen confirmation rule failed.
Next, take two charts with similar outlines but different preceding trends. Complete the five-step worksheet without looking at future bars, then check which observations changed your interpretation.