CFA Level II item set format

The current format

CFA Level II is a multiple-choice exam organized around shared vignettes. CFA Institute lists 22 item sets with 88 questions, split into two 132-minute sessions. Each session has 11 sets and 44 questions. Twenty item sets count toward the score and two are trial sets, but the candidate is not told which ones. Treat every set as scored.

How a vignette changes your approach

At Level I, a question generally contains the facts needed for that question on its own. At Level II, the vignette is the shared evidence base. It can describe an analyst’s objective, a company’s reporting choices, a portfolio’s holdings, market assumptions, dates, and numerical inputs. Each linked question may use a different subset of those facts. Read for structure first, then return to specific lines as each question requires.

Avoid two extremes. Reading every vignette as if you must memorize it wastes time. Skimming so quickly that you miss a date, unit, assumption, or reporting convention creates avoidable errors. Build a compact mental map: who is making a decision, what decision is requested, which entity or portfolio is under discussion, and where the relevant facts appear. The questions tell you which facts to retrieve.

The topic label and total point value at the beginning help orient you. They do not give away the method. Topic placement is not fixed by session, and any topic area may appear in either session or both. Read the specific learning task rather than assuming that every “Equities” set tests the same valuation model.

A four-question worked item set

The following original miniature case shows how one vignette can support several distinct questions. It is for practice and is not copied from CFA Institute materials.

Vignette: Lumen Devices

Lumen Devices reports under IFRS. Its analyst is valuing the common equity as of 31 December. The company has 50 million shares outstanding and no preferred shares. This year’s net income is $120 million. Next year’s forecast net income is $150 million. The analyst expects a sustainable forward P/E of 14. The current share price is $38. Ignore taxes on the analyst’s personal holdings. The company also plans a share repurchase next year, but the vignette provides no expected share count after the repurchase.

Before answering, note the valuation date, shares outstanding supplied, current versus forecast earnings, and the explicit forward multiple. The repurchase is mentioned but cannot support an adjusted per-share calculation because no post-repurchase count is given. That is a useful distinction: a vignette can include an interesting fact that is not enough to alter the requested calculation.

What the example teaches

One set can test a valuation calculation, percentage interpretation, ratio construction, and evidence sufficiency. The first question uses forecast figures; the third asks for a trailing ratio. Reusing the first calculation without noticing that the basis changed produces an answer that looks familiar but is wrong. In the second question, the phrase “relative to current price” determines the denominator. The last question tests whether the case supports a conclusion at all.

A useful habit is to label calculations before entering them: “forward value,” “upside over market,” or “trailing P/E.” That small step reduces accidental reuse of the wrong period or denominator. Also keep units visible. Total net income divided by shares gives earnings per share; price divided by earnings per share gives a multiple. A result such as $2.40 can be an EPS figure, while 15.8 is a ratio, not a price.

A repeatable set-solving routine

• Read the set heading and identify the topic and stated point value.
• Read the vignette once for the decision maker, date, objective, and structure.
• Mark facts that can change the answer: reporting basis, period, units, assumptions, constraints, and supplied estimates.
• Read the question stem before selecting a formula. Identify the requested output and reference period.
• Return to the relevant vignette facts rather than relying on memory.
• Calculate with units and signs intact; estimate direction and scale before accepting the result.
• Answer all four questions, then check whether you left any item blank.

Timing across 22 sets

The average pace is 12 minutes per set. A practical session budget is to aim near that average while allowing a difficult vignette to borrow a little time from an easier one. For example, if one set takes 14 minutes, the next two should average about 11 minutes to return to pace. Track time at set boundaries rather than checking the clock after every calculation.

If a question is consuming time, eliminate choices that conflict with the vignette, make the best supported selection, and continue. You can return if the exam interface and remaining time allow. Do not leave a whole set untouched while searching for perfect certainty on one item. The official guide states there is no penalty for incorrect answers, so answer each question.

During practice, record time and error type per set. If the case took too long because you reread every line, practice a concise first-pass map. If you missed a key fact, use a more deliberate scan for dates, units, and qualifiers. If setup was correct but execution slow, drill calculator steps. If the method itself was unclear, return to the relevant lesson before taking another timed set.

Multiple-choice and trial-set details

The questions are multiple choice and are tied to the vignette; they are not independent stand-alone questions. The two trial sets are mixed into the exam experience and are not identified to candidates. Since you cannot distinguish scored and trial content, it would be a mistake to lower effort on any set based on intuition about its style or difficulty.

There is no guessing penalty. If two answers remain plausible, compare each with the requested basis and the exact facts provided. If time expires soon, make the strongest choice. For a calculation, use units or a quick order-of-magnitude estimate to eliminate a distractor before committing.

Calculator use and factual precision

Candidates should be comfortable with an approved calculator and the operations their curriculum requires. In item sets, calculator speed matters most when the setup is already clear. A fast keystroke sequence cannot repair an incorrect choice of earnings period or discount convention. First name the quantity; then choose inputs; then check whether the result makes economic sense.

Watch for percent-versus-decimal inputs, annual-versus-period rates, signs on cash flows, and per-share-versus-total quantities. Keep enough precision through intermediate steps when answer choices are close. If an answer is unexpectedly negative, very large, or in the wrong direction, revisit the setup before recomputing.

How to practice the actual skill

Use question sets that preserve the shared-case structure. A pile of disconnected multiple-choice questions can strengthen recall, but it does not fully train evidence retrieval across linked questions. After solving a set, explain which sentence or number in the vignette supports each answer. Then explain why each distractor is wrong. This review often reveals that the missed concept was not the formula; it was overlooking a qualifier.

CFA Institute provides curriculum questions and mock exams through its candidate resources. Align practice with the exam year and current learning outcomes. For original practice you create yourself, change the company, dates, reporting choice, and numbers while keeping the underlying method. If the only way you can solve a problem is by recognizing a familiar answer, vary the data and ask the same concept in a different direction.

Session and exam-day expectations

The official duration is four hours and 24 minutes, divided into two equal 132-minute sessions, with an optional break between them. Each session contains 11 sets. Plan your stamina for sustained reading and calculations; this is not simply 88 unrelated short questions. Use the break to reset rather than reviewing frantic calculations from the first session.

Follow the current candidate instructions for approved calculator models, identification, test-center procedures, and exam-day rules. These operational policies can change and are separate from the question format. The content structure above is the published Level II format for 2026.

Reading cases with several layers of information

A vignette may combine background narrative, exhibits, analyst commentary, and explicit assumptions. Not every detail has equal importance. A candidate should notice the item’s subject, date, unit, currency, accounting basis, and who is making the decision. Those anchors prevent mixing a company’s fiscal year with the valuation date or applying one entity’s data to another. The question stem then narrows which facts matter.

Use selective annotation. A short note such as “FY1 actual,” “FY2 forecast,” “IFRS,” or “per share” can prevent a later mismatch. Avoid highlighting every number: visual noise makes the important inputs harder to retrieve. Mark a number when it has a clear role, such as a forecast input, a constraint, or a comparison point. When a question asks for a conclusion, locate the fact that supports the conclusion and the fact that could qualify it.

The same figure can support different interpretations depending on the question. A price divided by trailing EPS gives trailing P/E; divided by forecast EPS it gives forward P/E. A return may be arithmetic or geometric, before or after fees, and measured across different periods. A vignette can test whether you noticed that distinction even when the arithmetic is simple. Say the definition aloud in your scratch work before substituting values.

When to calculate and when to reason

Not every question in a quantitative topic needs a long computation. First decide whether the question asks for a number, a direction, a comparison, or an interpretation. If only direction is requested, reason from the relationship and avoid unnecessary calculator work. If a numeric result is requested, identify the formula and units, calculate, then compare the result to the answer choices. This conserves time and reduces opportunities for entry errors.

For multi-step calculations, carry intermediate values carefully and round only near the end. If the question asks “closest to,” do not waste time chasing precision beyond the choices. If a result depends on a convention, such as a compounding frequency or cash-flow timing, make the convention explicit before using a calculator function. A quick plausibility check can expose a missed sign or period.

Review a whole set, not just its score

After practice, review the four questions together. Did one misunderstanding in the vignette affect several answers? Did later questions reveal a fact you should have noticed earlier? Did you make inconsistent assumptions between items? The set format rewards a stable reading of the case, but it also requires you to reset the task at each question. Never carry an earlier answer’s assumption forward unless the case or question supports it.

For each distractor, write the error it represents: wrong period, wrong denominator, sign reversal, unsupported assumption, or correct formula for a different question. This converts distractors into a map of your habits. Over time, you will recognize recurring errors before submitting the answer. Practice sets should vary subject areas so that you learn to identify the relevant method without chapter labels doing the work for you.

The current structure, timing, trial-set count, topic-placement note, and no-penalty guidance come from CFA Institute’s Level II exam guide. The curriculum and topic outlines define the methods and year-specific learning outcomes tested.