Trendlines and Trend Analysis for CMT Candidates: Draw, Validate, Reassess

Trend first, line second

A trend is a sequence of directional price movement. A trendline is a visual tool that summarizes selected pivots. The line does not create the trend; its interpretation depends on which pivots, time interval, and vertical scale were chosen. The 2026 CMT Program Guide includes trend identification, following, polarity, and scale effects in Level I learning objectives. Level II applies trend and volume analysis to decisions. This page turns those concepts into a reproducible drawing method.

In an uptrend, mark higher swing lows and relevant highs before drawing a line under the lows. In a downtrend, mark lower swing highs and draw above them. A two-point line is possible, but a later reaction provides stronger evidence that participants have noticed the trajectory. There is no universal number of touches that certifies a line; define your pivot rule and use the same one across charts.

Draw a line without hindsight

Suppose daily swing lows occur at day 0, price 50, and day 10, price 60. On an arithmetic chart, the line's rise is (60 − 50)/10 = 1 price unit per day. At day 15 its projected value is 50 + 1×15 = 65. If price reaches 66 on day 15, it is one point above the projected line. This is a geometric description, not proof of support. A later touch near 65 would be a test; a close well below it under a predeclared break rule would challenge the line.

Now imagine a third low at day 15 is 64. A common mistake is to rotate the original line to include 64 and claim the trendline was never broken. A revised line may be useful, but record that the first line failed. Otherwise the method cannot be evaluated. Before revealing later bars, note which pivots established the line and whether you use closes or intraday lows.

A line break should not be equated automatically with a full trend reversal. Price may break a steep line while maintaining higher swing lows. A slower line may still reflect the broader trend. The important question is whether the underlying sequence of highs and lows has changed.

Channels and measured space

A channel pairs the main trendline with a roughly parallel line through a relevant opposite-side pivot. Suppose the lower boundary on day 15 is 65 and the upper boundary is 75. The channel width is ten points on this arithmetic chart. Price at 74 is near the upper edge, but “near” should be defined relative to volatility. If ATR is 4, a one-point gap is only 0.25 ATR; in a quieter market with ATR 0.5, it is two ATRs.

Channels help organize possible continuation and overextension scenarios. They do not demand that price bounce between boundaries. If an asset closes at 78 above the upper line, it may signal acceleration, an irregular bar, or a scale issue. Require independent evidence before labeling it a successful channel breakout. An entry at 78 with invalidation beneath 75 risks at least three points, and a gap may enlarge the realized loss.

Arithmetic versus logarithmic slope

A move from 20 to 30 is +10 points and +50%; a move from 100 to 110 is +10 points and +10%. Arithmetic scale draws these vertical changes equally; log scale gives the first a larger visual rise. Thus, a line that appears to connect points on one scale may miss on the other. For charts spanning large proportional ranges, a log chart often makes percentage trend comparisons more meaningful. For prices near zero, log scaling is not available or is awkward because the logarithm of zero or negative values is undefined.

The interval matters similarly. Ten daily bars and ten weekly bars are not the same elapsed time. A line drawn on a weekly close chart should not be tested against an intraday wick without explanation. Data adjustments for dividends or splits can also alter historical pivot locations.

Three evidence layers

Use swing structure as the first layer: are highs and lows rising, falling, or mixed? Use the line as the second: is price respecting a consistent trajectory? Use volume or a separate participation measure as the third: is the movement accompanied by supportive trading activity? Do not count a moving average and a trendline drawn from the same closes as fully independent confirmation. See Market Breadth for a different view of participation.

A conditional uptrend rule might say: “Keep the working uptrend label while the latest significant swing low remains intact; treat a close more than one ATR below the drawn line as a warning; change the label only after a lower high and lower low form.” The one-ATR buffer is an illustrative convention, not a CMT rule. It separates a line break from structural reversal and states what new evidence is required.

Common mistakes

Do not choose pivots only because they make a desired slope. Do not ignore chart scale. Do not treat a visually steep line as a quantified return rate without converting price and time units. Do not call a line “validated” merely because price touched it once. Be alert to an overfitted fan of lines: if many candidates are drawn, one will appear to work by chance. Keep rejected lines in an analysis log.

Self-check

An arithmetic trendline passes through day 2 at 40 and day 8 at 52. What is its projected value on day 11? Answer: Slope is (52 − 40)/(8 − 2) = 2 points per day. Three days after day 8, projected value is 52 + 3×2 = 58. A close at 57 would be one point below the line; whether that matters depends on your predeclared break rule and volatility.

Next, draw one trendline twice, once on arithmetic and once on log scale, and record any change in the apparent break date before you choose an interpretation.