CFA Level 2 Portfolio Management: notes and free practice

1. Question 1 of 10Level II topic sample · Case 1 of 5Portfolio ManagementMultiple choice

   Case facts

   A manager’s annual return is 11% versus an appropriate benchmark’s 8.5%. Annualized tracking error is 4%; total portfolio volatility is 18% and the risk-free return is 2%. All measures cover the same horizon and use consistent return definitions.

   The manager’s annual information ratio is:

   A.0.500B.2.750C.0.625
   Answer and explanations

   (11% - 8.5%) / 4% = 0.625, using active return and active risk.

   Review the method

   • A: 0.500 is the Sharpe ratio from (11 - 2) / 18, not the information ratio.
   • B: 2.750 divides total return by tracking error without subtracting the benchmark.
   • C - correct: (11% - 8.5%) / 4% = 0.625, using active return and active risk.
2. Question 2 of 10Level II topic sample · Case 1 of 5Portfolio ManagementMultiple choice

   Case facts

   A manager’s annual return is 11% versus an appropriate benchmark’s 8.5%. Annualized tracking error is 4%; total portfolio volatility is 18% and the risk-free return is 2%. All measures cover the same horizon and use consistent return definitions.

   Why is tracking error appropriate for the manager’s information ratio?

   A.It measures the volatility of benchmark-relative returnsB.It equals total portfolio volatility in every caseC.It measures only the benchmark’s return level
   Answer and explanations

   The information ratio evaluates active return per unit of active-return volatility, matching its numerator and denominator.

   Review the method

   • A - correct: The information ratio evaluates active return per unit of active-return volatility, matching its numerator and denominator.
   • B: Tracking error and total volatility are distinct except in special circumstances.
   • C: A return level is not the volatility of the return difference.
3. Question 3 of 10Level II topic sample · Case 2 of 5Portfolio ManagementMultiple choice

   Case facts

   An active strategy has information coefficient 0.08, 100 independent decisions per year and transfer coefficient 0.75. Use the generalized fundamental law: expected information ratio = TC × IC × square root of breadth. Target annual active risk is 5%; assume these inputs are stable.

   The strategy’s expected information ratio under the generalized fundamental law is:

   A.0.80B.0.60C.6.00
   Answer and explanations

   0.75 × 0.08 × √100 = 0.60 under the independence and stability assumptions.

   Review the method

   • A: 0.80 omits the 0.75 transfer coefficient.
   • B - correct: 0.75 × 0.08 × √100 = 0.60 under the independence and stability assumptions.
   • C: 6.00 uses breadth instead of its square root.
4. Question 4 of 10Level II topic sample · Case 2 of 5Portfolio ManagementMultiple choice

   Case facts

   An active strategy has information coefficient 0.08, 100 independent decisions per year and transfer coefficient 0.75. Use the generalized fundamental law: expected information ratio = TC × IC × square root of breadth. Target annual active risk is 5%; assume these inputs are stable.

   At 5% target active risk, the strategy’s expected annual active return is:

   A.5.00%B.0.12%C.3.00%
   Answer and explanations

   Expected active return = information ratio 0.60 × active risk 5% = 3.00%.

   Review the method

   • A: 5% is the chosen active-risk level, not the expected active return.
   • B: 0.12% does not multiply expected information ratio by active risk.
   • C - correct: Expected active return = information ratio 0.60 × active risk 5% = 3.00%.
5. Question 5 of 10Level II topic sample · Case 3 of 5Portfolio ManagementMultiple choice

   Case facts

   An ETF trades at $51 while its per-share NAV is $50. An authorized participant can deliver the underlying basket to create ETF shares, then sell the shares. For the simple arbitrage illustration, ignore transaction costs, financing, delays and basket-pricing uncertainty.

   The ETF’s market-price premium to NAV is:

   A.2.00%B.1.96%C.1.00%
   Answer and explanations

   51 / 50 - 1 = 2.00%; NAV is the denominator for this premium.

   Review the method

   • A - correct: 51 / 50 - 1 = 2.00%; NAV is the denominator for this premium.
   • B: 1.96% uses market price rather than NAV in the denominator.
   • C: The dollar difference is one, but the percentage premium is two.
6. Question 6 of 10Level II topic sample · Case 3 of 5Portfolio ManagementMultiple choice

   Case facts

   An ETF trades at $51 while its per-share NAV is $50. An authorized participant can deliver the underlying basket to create ETF shares, then sell the shares. For the simple arbitrage illustration, ignore transaction costs, financing, delays and basket-pricing uncertainty.

   Which transaction exploits the supplied frictionless ETF premium?

   A.Buy ETF shares and redeem for the cheaper basketB.Buy the basket, create ETF shares and sell themC.Buy the ETF and hold it regardless of NAV
   Answer and explanations

   Buying $50 of basket exposure and creating shares sold at $51 captures the stated premium before the excluded frictions.

   Review the method

   • A: That direction pays the premium instead of collecting it.
   • B - correct: Buying $50 of basket exposure and creating shares sold at $51 captures the stated premium before the excluded frictions.
   • C: Holding an overpriced share does not lock the illustrated creation arbitrage.
7. Question 7 of 10Level II topic sample · Case 4 of 5Portfolio ManagementMultiple choice

   Case facts

   A $100m portfolio has daily return volatility of 2% and zero expected daily return. Use a normal-distribution model with a one-sided 95% critical value of 1.645. No cash flows occur. The risk committee wants both the modeled one-day VaR amount and its proper interpretation.

   The normal-model one-day 95% VaR is:

   A.$2.00mB.$1.645mC.$3.29m
   Answer and explanations

   $100m × 1.645 × 0.02 = $3.29m modeled one-day VaR.

   Review the method

   • A: 2m is one daily standard deviation, not the supplied 95% quantile.
   • B: 1.645m fails to apply the 2% volatility correctly to $100m.
   • C - correct: $100m × 1.645 × 0.02 = $3.29m modeled one-day VaR.
8. Question 8 of 10Level II topic sample · Case 4 of 5Portfolio ManagementMultiple choice

   Case facts

   A $100m portfolio has daily return volatility of 2% and zero expected daily return. Use a normal-distribution model with a one-sided 95% critical value of 1.645. No cash flows occur. The risk committee wants both the modeled one-day VaR amount and its proper interpretation.

   Which interpretation of the committee’s computed VaR is correct?

   A.Under the model, losses exceed the cutoff on about 5% of daysB.The cutoff is the maximum possible lossC.Every fifth day must lose more than the cutoff
   Answer and explanations

   A one-sided 95% loss quantile leaves approximately 5% modeled probability beyond it; it is not a loss ceiling.

   Review the method

   • A - correct: A one-sided 95% loss quantile leaves approximately 5% modeled probability beyond it; it is not a loss ceiling.
   • B: VaR does not cap the size of tail losses.
   • C: Five percent is a probability, not a deterministic sequence or one-in-five frequency.
9. Question 9 of 10Level II topic sample · Case 5 of 5Portfolio ManagementMultiple choice

   Case facts

   A researcher backtests a 2020-2025 strategy using only companies that remain in the index today and selects the best of hundreds of parameter settings on that same period. She has separate later data not used in this search, with point-in-time constituents available. Costs have not yet been included.

   Using only today’s surviving index companies in the historical test most directly risks:

   A.A fully investable point-in-time universeB.Survivorship and universe-selection biasC.Automatic correction for trading costs
   Answer and explanations

   Selecting current survivors omits historical failures or departures and can overstate simulated performance.

   Review the method

   • A: Today’s membership can exclude firms that were eligible in the historical period.
   • B - correct: Selecting current survivors omits historical failures or departures and can overstate simulated performance.
   • C: Membership selection does not include commissions, spreads or market impact.
10. Question 10 of 10Level II topic sample · Case 5 of 5Portfolio ManagementMultiple choice

    Case facts

    A researcher backtests a 2020-2025 strategy using only companies that remain in the index today and selects the best of hundreds of parameter settings on that same period. She has separate later data not used in this search, with point-in-time constituents available. Costs have not yet been included.

    What is the strongest next validation step for the researcher’s selected strategy?

    A.Report only the best in-sample resultB.Continue retuning on the unused sample indefinitelyC.Freeze the rule and test untouched point-in-time data with realistic costs
    Answer and explanations

    A fixed rule evaluated out of sample with an investable universe and costs tests generalization more credibly.

    Review the method

    • A: The best result among many settings is exposed to selection and overfitting.
    • B: Repeated retuning would make the unused data part of model selection.
    • C - correct: A fixed rule evaluated out of sample with an investable universe and costs tests generalization more credibly.

Four ideas to put to work

Find the method you need

Start with: Benchmark returns, standard deviation, portfolio risk, ETF shares and historical testing.

Questions about this topic

What does an information ratio measure?

Under matching definitions, it compares active return with active-return volatility relative to the selected benchmark. The benchmark and measurement period matter to the interpretation.

Does high active share guarantee high active risk?

No. Holdings differences and return variation measure different aspects of the portfolio. Similar return exposures can arise from different holdings, so evaluate the relevant risk model and benchmark.

Does one positive active return prove skill?

No. Assess risk, benchmark fit, decision evidence and results over time. One period can reflect market conditions or chance as well as a useful decision process.