CMT Level III Study Guide: Write a Defensible Market Decision

What Level III actually asks you to do

Level III is the work of a technical analyst: integrate charts, indicators, volatility, model output, portfolio weights, and risk into a defensible decision. The March 2026 CMT Program Guide describes grouped cases with three to seven response items, some multiple choice and most short answer. It allows four hours and suggests 120–160 study hours. The official domains are Application of Technical Analysis 95% and Ethics 5%. The 95% application category covers varied methods; it is not one narrow chapter.

An obsolete description of Level III as entirely essay or entirely multiple choice is wrong. The official FAQ says all Level III items are scored and advises using point allocations for time management. It also says responses receive credit for application of concepts; grammar or typographical imperfections alone are not penalized. Write clearly, but prioritize relevant analysis.

Build answers around the decision

A useful short-answer structure is observation → interpretation → action → risk/invalidation. If asked to list two factors, give two distinct factors, not a paragraph of related synonyms. If asked to justify, connect each factor to the proposed action. Use supplied data; do not import a market narrative that the case does not support.

For a trade recommendation, specify instrument, direction, entry condition, size or weight, exit/invalidation, and risk to the existing portfolio. For a model evaluation, assess data quality, sample design, costs, and relevance to mandate. For an ethics item, identify the conduct issue and the appropriate response under the Code and Standards. The current official curriculum and ethics materials are the authority for scoring, not a third-party template.

Worked written-response case

A hypothetical 2,000,000 balanced portfolio has a 60% equity policy weight with a permitted 55–65% band. A technology sector currently represents 12% of the total portfolio, and a 15% sector cap applies. Its absolute price trend is rising, but its relative ratio against broad equities has turned down for two review dates. Volatility has expanded, and market breadth has narrowed.

Question: Should the manager add a 5% portfolio weight to technology? State two reasons and a risk action.

Model response: No. Adding 5 percentage points would raise technology to 17%, breaching the 15% cap; available capacity is only 3 percentage points, or 60,000. Also, deteriorating relative strength and breadth weaken the tactical thesis despite a positive absolute trend. Hold or trim within mandate until relative strength stabilizes. If an approved smaller add is considered, define a relative-strength recovery trigger and an exit if price breaks its identified support; review total equity band and costs.

The 3% capacity is (15% - 12%) × 2,000,000 = 60,000. A concise response earns clarity by tying numerical and qualitative facts to a decision. A generic "bullish because trend up" ignores two pieces of conflicting evidence and a binding cap. If the prompt asks only for two reasons, answer those first before elaborating.

Practice in the response format

Multiple-choice banks help test knowledge, but they cannot replace writing. Turn an official objective into a short case: give yourself a chart description, portfolio limit, and conflicting signal; then write a timed answer without notes. Compare it with a checklist: did you state an action, use case data, quantify any constraint, give an invalidation, and avoid claims unsupported by the exhibit?

Use the points displayed for each item to allocate the four hours proportionally, allowing time to review calculations and incomplete responses. Do not spend ten minutes polishing a low-point response while leaving high-point items blank. Practice typing concise lists and calculations clearly.

A realistic study cycle

Suppose the candidate has 140 hours over 14 weeks, or 10 hours per week. The first five weeks can rebuild current curriculum knowledge and risk calculations; the next five can emphasize integrated cases; the final four can use timed response sets and targeted error repair. This division is illustrative. A candidate weak in portfolio construction may redirect time there rather than following a fixed chapter order.

A useful error log classifies mistakes as missed case fact, wrong calculation, unsupported inference, omitted action, omitted risk control, or ethics misunderstanding. If the same class repeats twice, change the drill. Rewriting a polished model answer from memory is less useful than answering a new case.

Grade a response against the requested task

A self-review rubric should award credit for the action asked, the case evidence used, the calculation if required, and the justification connecting evidence to action. Imagine a prompt asking for two reasons to reduce exposure and one risk control. An answer that lists five indicators without identifying two distinct reasons has not addressed the request. An answer that gives the correct reduction but no case fact may be too thin. Precision matters more than length.

Make numerical work visible. In the worked committee case, write "sector cap 15% - current weight 12% = 3 percentage points of room; 3% ? 2,000,000 = 60,000." This lets a reviewer follow the recommendation and catches a common mistake: confusing 3% of the existing sector position with 3% of the whole portfolio. If the case supplies transaction costs, include them in the decision.

Timed practice should include brief review after each response. Mark omitted facts, unsupported claims, arithmetic errors, and wasted sentences. Re-answer the same question once concisely, then apply the lesson to a new case. Memorizing one polished response is less useful than learning to extract the relevant facts under time pressure.

Common traps and self-check

Avoid reciting every indicator you know, treating all signals as equally important, omitting a binding mandate limit, and giving a target without an invalidation. Do not assume a practice MCQ score predicts written-response performance. Keep recommendations proportional to the case evidence.

Self-check: A case states price is above its long-term average, relative strength is weakening, and a proposed trade breaches a concentration limit. Is "buy because trend is up" sufficient? Answer: No. It ignores conflicting evidence and a binding constraint. A valid response must address the limit and explain whether to hold, reduce, or size within capacity, with a trigger and risk rule.

Next, draft three original case answers from Level III practice topics, then evaluate them against the official learning objectives and the Level III sample-item style in the official guide.