Mastering the two most distinctive exam formats in the CFA program.
Each vignette is a 1- to 2-page case story (typically 600 to 1 000 words) introducing a fictitious analyst, fund or company. Following the vignette, four multiple-choice questions with three answer options (A/B/C) test the candidate ability to apply curriculum concepts to that specific scenario.
Across both 2h12 sessions, candidates answer 88 MCQ derived from 22 vignettes. The 10 topic areas appear with broadly similar weights as Level I, but heavily emphasised toward FRA (10-15%), Equity (10-15%) and Fixed Income (10-15%). Ethics still weighs 10-15%.
| Topic area | L2 weight 2026 | Typical vignette count |
|---|---|---|
| Ethics & Professional Standards | 10-15% | 2-3 |
| Quantitative Methods | 5-10% | 1-2 |
| Economics | 5-10% | 1-2 |
| Financial Statement Analysis | 10-15% | 2-3 |
| Corporate Issuers | 5-10% | 1-2 |
| Equity Investments | 10-15% | 2-3 |
| Fixed Income | 10-15% | 2-3 |
| Derivatives | 5-10% | 1-2 |
| Alternative Investments | 5-10% | 1-2 |
| Portfolio Management | 10-15% | 2-3 |
The morning session contains approximately 11 essays (called "constructed response" since 2021 CBT migration). Each question can have multiple parts (A, B, C, D…) with explicit point allocations. Candidates type answers into a text editor. The 1 minute per mark rule applies rigorously: a 15-point essay should consume no more than 15 minutes of pure writing time.
The afternoon mirrors Level II: 11 vignettes × 4 MCQ = 44 MCQ in 2h12. Topics are heavily weighted toward Portfolio Management, IPS, Asset Allocation and Wealth Planning. Ethics receives 10-15% (lower than L2).
"Level III is a culmination of the CFA Program, focused on portfolio management for both individual and institutional investors."
Source: CFA Institute, Level III Curriculum 2026 — cfainstitute.org/programs/cfa/exam/level-iii
CFA Institute publishes a list of command verbs that dictate exactly what the grader expects. Every essay marking guideline starts with these.
| Command word | What to produce | Mistake |
|---|---|---|
| Calculate | Show numerical answer with units | Forgetting to show formula or unit |
| Determine | Choose from a finite set + justify | Writing essay-length narrative |
| Justify | State decision + 2-3 specific reasons | Generic theory recitation |
| Discuss | Pros and cons / two-sided analysis | One-sided argument |
| Recommend | State the action + brief rationale | Listing options without choosing |
| Identify | Name the item — bullet acceptable | Long paragraph |
| State | Simple declarative answer | Lengthy explanation |
"Marie (62) and Jean (64) are retiring in 2026. Combined assets EUR 2.4M, expected lifetime spending EUR 90 000/year (real). Determine their return objective. Justify with one quantitative element."
Model answer: "Return objective = pre-tax nominal 5.2%. Justification: real spending need 3.75% (90/2 400) + assumed inflation 2.0% + tax gross-up 25% on dividends/interest = ~5.2%. This exceeds the 4% sustainable withdrawal rule of thumb, signalling moderate equity allocation 40-55%."
A typical L2 vignette has six sections: (1) the firm/analyst introduction, (2) the financial situation, (3) the analyst's task, (4) supporting data (tables, exhibits), (5) discussion among characters, (6) the 4 follow-up MCQ. The CFA examiners deliberately embed "noise" — irrelevant details that test your ability to extract the material facts.
Successful candidates use the "RED-SHIFT" technique: (R)ead the questions first to identify what to look for, (E)xtract the key numerical and conceptual data from the vignette, (D)istinguish material data from distractors, (S)et up calculation framework, (H)ighlight tables and exhibits relevant to each MCQ, (I)dentify potential traps in answer choices, (F)inalise the answer choice, (T)rack flagged questions.
| Activity | Time | Cumulative |
|---|---|---|
| Read all 4 questions first | 2 min | 2 min |
| Skim vignette (1st pass) | 3 min | 5 min |
| Deep-read with notes | 4 min | 9 min |
| Answer Q1 + Q2 | 4 min | 13 min |
| Answer Q3 + Q4 | 4 min | 17 min |
| Reserve / flag review | 1 min | 18 min |
When asked to draft an IPS section, follow this structured template (memorise it):
When asked to identify a bias and recommend an action:
| Trap | How to spot | Fix |
|---|---|---|
| Mixing nominal & real rates | Read inflation context | Convert to consistent basis |
| Using WACC for FCFE | Source of cash flow | FCFE = cost of equity only |
| Forgetting tax shield on debt | WACC formula | r_D * (1 - t) |
| Mismatch frequency (annual coupon vs semi-annual yield) | Read carefully | Adjust to consistent compounding |
| Day count convention | 30/360 vs actual/365 | Match to instrument standard |
| Stale beta | Use historical only when stable | Adjust for fundamental change |
The following vignette illustrates the typical structure and difficulty of a Level II item set. Read it carefully, then attempt the 4 follow-up MCQ before reviewing the model answers.
Sophie Martin, CFA, is a senior equity analyst at Atlantis Capital Management in Geneva. She is preparing a research report on Helvetia BioPharma SA (HBP), a mid-cap Swiss pharmaceutical company specialising in oncology biologics. Martin has access to the following financial data for HBP (in CHF millions):
| Year | Revenue | EBIT | Net Income | FCFE | Dividend |
|---|---|---|---|---|---|
| 2023 | 1 850 | 320 | 235 | 180 | 85 |
| 2024 | 2 120 | 395 | 295 | 225 | 105 |
| 2025E | 2 480 | 485 | 368 | 285 | 130 |
HBP's market beta is 1.15. The Swiss risk-free rate (10-year Confederation bond) is 1.45%. The equity risk premium for Switzerland is 5.5%. HBP's capital structure is 75% equity, 25% debt; pre-tax cost of debt is 3.2%; corporate tax rate 14.6% (Swiss federal + cantonal average).
Martin estimates that HBP's earnings will grow at 14% per year for the next 4 years (driven by 3 pivotal trial readouts) before settling to a stable growth rate of 3.5% indefinitely. The retention ratio is currently 65%.
Martin's colleague Thomas Weber, CFA, also a senior analyst, has been bullish on HBP for two years. Weber suggests Martin use only the P/E multiple of comparable European biotech firms (current average 22.5x forward earnings) as her primary valuation tool. Weber adds: "DCF gives results too sensitive to growth assumptions for biotech — multiples reflect what the market is actually paying."
Martin's mandate from her portfolio manager is to provide a Buy/Hold/Sell recommendation supported by a fundamental valuation. The current HBP share price is CHF 87.50; shares outstanding 18.6 million.
Solution: rE = Rf + β × ERP = 1.45% + 1.15 × 5.5% = 1.45% + 6.325% = 7.78%. Answer: C.
Use D1 growth model assuming dividend grows at earnings rate. D0 = 130/18.6 = CHF 6.99 per share. Stage 1 (4 years at 14%): D1 = 7.97; D2 = 9.09; D3 = 10.36; D4 = 11.81. Stage 2: D5 = 11.81 × 1.035 = 12.22. Terminal value at end of year 4 = 12.22 / (0.0778 − 0.035) = 12.22 / 0.0428 = CHF 285.5.
PV of dividends: 7.97/1.0778 + 9.09/1.0778² + 10.36/1.0778³ + 11.81/1.0778⁴ = 7.39 + 7.83 + 8.27 + 8.74 = 32.23. PV of terminal: 285.5/1.0778⁴ = 211.5. Intrinsic value = 32.23 + 211.5 = CHF 243.7 per share. Note: this is far above the current CHF 87.50 — sanity-check growth assumptions; the example illustrates the sensitivity Weber warned about.
Solution: D. Recommending a single methodology is a methodological opinion, not a Standard violation. Weber's two-year bullish view, by itself, also is not a violation. The violation would arise if Weber's recommendation conflicted with explicit firm policy on methodology, or if he pressured Martin to ignore evidence-based DCF analysis to support a pre-existing thesis.
EPS 2025E = 368/18.6 = CHF 19.78. P = 22.5 × 19.78 = CHF 445.1 per share. This is also far above current. Compare with DDM (CHF 243.7) and current price (CHF 87.50): the wide dispersion suggests the market is pricing in a much lower growth scenario than Martin's. A reasonable response is to triangulate (DCF, multiples, P/B) and document the sensitivity ranges.
Question (15 marks total):
Vignette facts: Marie (62), Jean (64), retired in 2026. Combined investable assets EUR 2.4M. Annual lifetime spending need EUR 90 000 (real, indexed to inflation). Marginal income tax rate at retirement: 30%. Expected inflation: 2.0%. No mortgage. Adult son with disability requiring potential lifetime support of EUR 15 000/year. Couple has stated risk tolerance as "moderate".
A. Required nominal pre-tax return (3 marks):
B. Ability to take risk (4 marks):
C. Two most binding constraints (4 marks):
D. Recommended SAA (4 marks):
| Asset class | Allocation | Justification |
|---|---|---|
| Global equity | 45% | Growth engine + currency diversification |
| Investment-grade bonds (5y duration) | 35% | Income + stability |
| Real estate (REITs) | 10% | Inflation hedge + income |
| Cash/money market (1y spending) | 5% | Liquidity |
| Disability reserve (separate) | 5% (EUR 120K) | Earmarked, conservative |
Expected return: 0.45 × 7% + 0.35 × 4% + 0.10 × 6% + 0.05 × 3% + 0.05 × 3% = 5.45%. Below 8.21% required — adviser should discuss reducing spending or increasing risk slightly.
CFA L3 essay graders use a standardised rubric. Key insights:
| Skim type | Use case | Time |
|---|---|---|
| Question-first skim | Identify which paragraphs matter most | 30 sec |
| Number/data scan | Locate exhibits, tables, key figures | 30 sec |
| Speaker identification | Distinguish what each analyst/character says | 30 sec |
| Conflict identification | Spot Standards violations or biases | 1 min |
| Deep linear read | Full understanding before answering | 3 min |
| Term | Definition |
|---|---|
| Constructed response | Free-text essay (L3 morning session) |
| Item set | L2 vignette + 4 MCQ unit |
| MPS (Minimum Passing Score) | Threshold set post-exam by CFA Board |
| Forward bias | Empirical UIP failure (carry trade exploit) |
| Liability-matching | Building portfolio to match cash flow obligations |
| Tail-risk | Extreme low-probability/high-severity loss |
| SPRA / SRA | BoC standing repo / reverse repo |
| BL (Black-Litterman) | Bayesian asset allocation model |
| OAS (Option-Adjusted Spread) | Z-spread adjusted for embedded options |
| Suitability vs Fair Dealing | Portfolio fit vs information distribution |
La leçon suivante est également gratuite. Découvrez-la sans inscription.
Leçon 2 — Continuer →Choisis quels cookies tu acceptes — modifiable à tout moment.