CMT Level II common mistakes
CMT Level II · Understand the traps and how to avoid them.
Answering a chart-pattern or indicator question from memory of a textbook picture instead of applying the stated rule to the data given in the item, so the chosen label ('head and shoulders', 'overbought') contradicts the numbers or sequence described.
Level II questions are built around applying method to supplied evidence; candidates who memorized pattern silhouettes skip the step of checking the actual swing sequence, volume behaviour, or threshold values, and pattern recognition becomes confirmation bias.
What to do instead
Before selecting an answer, restate the defining criteria in writing on scratch paper (e.g., three swing highs with the middle highest, neckline break, volume expansion on the break) and tick each criterion against the data in the stem. Reject any option that fails one necessary condition, even if the shape looks familiar.
Treating an indicator's default parameters and standard thresholds as universal, and failing to adjust the interpretation when the item specifies a different lookback, smoothing, or market (e.g., a 5-period RSI versus 14, or a fast versus slow stochastic).
Candidates learn one canonical setting and one canonical threshold, then ignore the parameters printed in the stem, producing answers that would be right for the default case but wrong for the case asked.
What to do instead
Highlight every numeric parameter in the stem before reading the options. Ask explicitly: with this lookback, how many observations are needed, what is the neutral level, and how does shorter or longer smoothing change sensitivity and whipsaw frequency? Practice computing the indicator by hand on a short series so parameter effects become mechanical rather than verbal.
Confusing trend-following and mean-reversion tools, for example using an oscillator crossover as a standalone trend signal or using a moving-average crossover to call a top in a range-bound market.
Indicators are often taught as separate chapters, so candidates do not internalize which regime each tool assumes; oscillators give frequent signals in trends and moving averages give frequent false signals in ranges, and the exam tests exactly that mismatch.
What to do instead
Build a two-column map: tools that assume persistence (moving averages, MACD trend use, channel breakouts, ADX direction) versus tools that assume reversion (RSI, stochastic, Bollinger band tags, %B extremes). For each practice item, first classify the described market as trending or ranging, then eliminate options that use the wrong family.
Reading volume as a standalone bullish or bearish signal rather than as confirmation relative to price action and to its own recent baseline.
Volume is intuitively 'more is better', so candidates label any expansion as bullish and any contraction as bearish, ignoring whether price advanced, stalled, or reversed, and ignoring that volume must be judged against average volume and against the stage of the move.
What to do instead
Use an if-then rule set: rising price with rising volume confirms; rising price with falling volume warns; falling price with rising volume confirms the decline; high volume with little price progress signals absorption or distribution. Always compute or estimate relative volume against the recent average stated in the item before choosing.
Ignoring the multi-timeframe relationship, so a daily oversold reading is treated as a buy signal even when the weekly trend is down and the item describes a primary downtrend.
Candidates analyze each chart in isolation and forget that Level II emphasizes the interaction of timeframes; the same oscillator value means different things depending on the direction of the larger-degree trend.
What to do instead
Adopt a fixed hierarchy: establish the higher-timeframe trend first, then interpret the lower-timeframe signal only as alignment or countertrend. In items, write 'weekly = down' before evaluating the daily signal, and prefer answers that describe a countertrend bounce rather than a trend reversal.
Confusing confirmation with prediction: choosing an answer that says a signal 'will' produce a move rather than that it 'raises the probability of' or 'confirms' a condition.
Everyday market commentary is predictive, and candidates absorb that phrasing; technical analysis as tested is probabilistic, so overconfident wording is usually the distractor.
What to do instead
Scan options for absolute verbs (will, guarantees, must, always) and treat them as red flags unless the item is asking for a definitional identity. Prefer options phrased in terms of probability, confirmation, invalidation levels, and risk placement.
Skipping the invalidation or stop-placement part of a setup and answering only about entry direction, so risk-management sub-questions are missed.
Candidates focus on pattern identification and treat the trade-management portion as an afterthought, yet Level II items frequently ask where a pattern is invalidated or how position size follows from the stop distance.
What to do instead
For every pattern, memorize its invalidation point as part of the pattern itself: beyond the right shoulder for a head and shoulders, back inside the channel for a breakout, below the breakout candle for a flag. Practice stating entry, invalidation, and measured target as a three-part sentence before looking at options.
Mishandling breadth and sentiment data by comparing raw levels across markets or time periods instead of using the same denominator, and by treating a single sentiment reading as a contrarian extreme.
Breadth statistics (advance-decline, percent above moving average) and sentiment surveys are scale-dependent and noisy; candidates compare a 2,000-issue advance-decline line with a 500-issue one, or call a single high reading a top without confirmation.
What to do instead
Always ask what the denominator and universe are, and normalize before comparing. For sentiment, require a multi-reading extreme (several consecutive high or low readings) plus a price confirmation before treating it as contrarian. Practice converting raw counts to ratios or percentages in the calculation items.
Rushing the calculation items, leaving no time to check units, sign conventions, and rounding, so a correct method yields a wrong option.
Level II includes quantitative items under time pressure, and candidates either skip the arithmetic or make scale errors (points versus percent, log versus arithmetic, annualized versus per-period) that match a distractor exactly.
What to do instead
Budget time explicitly and do calculations in a fixed template: write the formula, substitute values with units, compute twice, then compare magnitude and sign against the plausible range. Where an option is suspiciously close to a common slip (forgetting to annualize, using n instead of n-1), recheck that step first.
Selecting the longest or most technical-sounding option when uncertain, rather than eliminating options against the specific condition the stem asks about.
Under time pressure, candidates default to recognition heuristics; well-written distractors are plausible and often longer, and the correct answer is frequently the more conditional, narrower statement.
What to do instead
Force a two-pass method: on the first pass mark the one or two options that directly address the asked condition and strike the rest with a one-word reason. If still stuck, identify which option would be true under a different market regime and eliminate it, since the exam usually tests the regime stated in the stem.