CMT Level II Study Guide: Apply and Defend Your Analysis

The change from Level I

Level II asks you to apply technical-analysis concepts to opportunities and risk, rather than only recognize definitions. The March 2026 Program Guide specifies 170 multiple-choice questions in four hours, of which 150 are scored and 20 are trial items. The official planning range is 100–140 study hours. There is no reason to guess which items are unscored; answer all.

The weighting means Classical and Advanced deserve substantial practice, but it does not identify guaranteed item counts by topic. The official curriculum for the registered cycle controls exam scope. The Association's FAQ says Level II curriculum is sufficient for the exam apart from the separately supplied ethics standards; there is no requirement to reread all Level I material, though it remains available for reference.

Practice decisions, not isolated labels

For each chart, write five lines before looking at answer choices: trend and timeframe; supporting evidence; conflicting evidence; proposed action; invalidation. For model output, ask whether the sample, benchmark, assumptions, and costs make implementation plausible. For a relative-strength signal, inspect absolute trend and concentration. For volatility, distinguish a past return dispersion estimate from option-implied expectations.

A topic drill is useful after a unit, but mixed sets reveal whether you can choose the right tool without a chapter label. Alternate focused work with mixed questions. When reviewing a miss, explain why the tempting wrong option was wrong. A candidate who can repeat an answer but cannot explain that difference still has a fragile understanding.

Worked preparation case

A candidate has 120 hours over 12 weeks, or 10 hours per week. As a starting allocation, the official weights imply about 8.4 hours Theory, 48 Classical, 48 Advanced, 12 Application, and 3.6 Ethics. This is arithmetic, not a directive.

The diagnostic finds the candidate strong in classical chart recognition but weak in volatility and risk applications. A revised plan uses 8 Theory, 42 Classical, 52 Advanced, 13 Application, and 5 Ethics hours, totaling 120. The extra time targets actual mistakes while still covering all domains. Weeks 1–6 may emphasize official curriculum and targeted drills; weeks 7–9 mixed cases and error correction; weeks 10–12 timed rehearsal, model interpretation, and ethics review. Shift the calendar if progress data demand it.

For a worked analytic question, suppose a sector rises 8% while its benchmark rises 5%. That is outperformance in return terms, but does not alone justify a full portfolio weight. Check absolute trend, sector cap, volatility, and transaction costs. If the sector's relative line breaks its defined support, invalidate the tactical overweight. This reasoning is what moves a Level I observation toward a Level II decision.

Time and exam technique

Four hours for 170 questions is about 85 seconds per question on average. Some chart and calculation questions take longer than definitions. In practice, mark a difficult item and move on if it threatens the rest of the set. Rehearse with the same answer review and break conditions permitted in your actual delivery format; check the current exam FAQ before test day.

The Association does not publish a universal pass percentage. A practice score is a diagnostic of your preparation under that practice set's quality and difficulty, not a pass prediction. Track by objective and error type. Do not chase a numeric mock threshold while leaving clear reasoning gaps.

Invalidation rules for a study plan

If two mixed sets show the same weakness, revise the study allocation. If errors are mostly misread exhibits, slow down on chart labeling; if errors are arithmetic, reproduce the formula inputs and units. If a topic bank is exhausted, return to official exhibits and create a fresh case, rather than memorizing question order. Ethics should be practiced as conduct scenarios, not only a list of standards.

Use the Level II practice topics as independent diagnostics and compare each missed concept with the official learning objectives. For higher-level integration, see risk management and intermarket analysis.

Practice model suitability as a decision

A Level II model question is not answered by saying that a backtest return was high. Suppose a trend model reports 12% annualized return against a 7% benchmark, but it was tested on today's surviving stocks, uses the signal close as its fill, and excludes trading costs. The reported five-point excess return is not a reliable implementation estimate. A useful answer identifies the specific biases, asks for a universe as known at each date, shifts fills to an executable time, includes costs, and evaluates a separate period.

Similarly, a volatility calculation must be translated to the case. A larger ATR may justify a wider technical allowance but then requires fewer units under the same money-risk budget. A relative-strength ranking may favor one security while a concentration limit prevents a full weight. These are not exotic exceptions; they are the natural tension in applied analysis.

For each practice case, add a short "why not the alternative?" line. If a question has three plausible tools, explain why the selected tool answers the exact question and why the others do not. That habit improves multiple-choice discrimination without guessing what the test writer intended. Use the official objective to verify that your reasoning stays in scope.

Common traps and self-check

Avoid memorizing pattern names while ignoring price location, interpreting a relative line without a benchmark, treating implied volatility as a return forecast, and applying a risk model without checking execution. A Level II answer should show what evidence changes the conclusion.

Self-check: A model ranks a security first, but a proposed purchase would breach a 10% name cap. What is the correct response? Answer: The model ranking does not override the mandate. Limit the weight to the available capacity, select another eligible security, or refrain from trading. Document the binding constraint and the model's limitations.

Next, take one official applied objective and write a mini-case with two plausible choices, one risk constraint, and one invalidation condition. Solve it a day later without notes.