CFA Exam Question of the Day

Rationale:
The most appropriate immediate governance action by the Chief Risk Officer (Second Line) for the Board Risk Committee involves a comprehensive review of the limit breach within the broader risk management framework. A VaR limit breach, especially in volatile markets, requires the Board to understand the underlying causes, evaluate the current appropriateness of the limit itself, and consider the potential impact of reactive de-risking on the fund's objectives and overall risk appetite. This approach emphasizes the Second Line's independent oversight role in ensuring the risk framework remains robust and responsive to market conditions, rather than merely enforcing a numerical compliance. The option suggesting immediate de-risking without further governance review represents a First Line operational response rather than a Second Line strategic governance recommendation to the Board, potentially leading to suboptimal decisions such as crystallizing losses. The option advising no urgent intervention because the VaR is below the 1-in-250 year maximum acceptable loss confuses a specific, approved VaR *limit* (a control) with a broader, stress-test derived *risk appetite* (a tolerance level). A limit breach demands a governance response regardless of whether it falls within the wider risk appetite. The option prioritizing immediate VaR model validation, while a crucial risk management activity, is not the *immediate governance response* to an existing breach. The breach itself requires an initial governance reaction, even if subsequent model review is warranted to understand if the model itself contributed to the breach or its measurement.

Rationale:
In the Solow growth model, which is implied by the scenario describing a steady-state equilibrium and parameters like TFP, labor, capital, depreciation, and savings rate, the long-term annual growth rate of total output is determined solely by the sum of the rate of Total Factor Productivity (TFP) growth and the labor force growth rate. The savings rate, depreciation rate, and capital's share of income affect the steady-state *level* of output per capita but do not influence its long-term *growth rate*. Therefore, the long-term annual growth rate of total output is 1.80% (TFP growth) + 0.50% (labor force growth) = 2.30%. The option of 4.30% incorrectly assumes that the 200 basis point increase in the national savings rate directly translates into a 200 basis point increase in the long-term growth rate of total output. While a higher savings rate leads to a higher steady-state level of output and a period of faster growth during the transition, it does not permanently increase the long-term growth rate in the Solow model. The option of 1.80% represents the long-term growth rate of output *per capita*, which is equal to the TFP growth rate, but the question asks for the growth rate of *total output*. The option of 3.10% results from calculating the historical growth rate using the growth accounting equation (gY = gA + α*gK + (1-α)*gL = 1.80% + 0.40*2.50% + 0.60*0.50% = 1.80% + 1.00% + 0.30% = 3.10%). This calculation reflects a past or transitional growth rate, not the new long-term steady-state growth rate as defined by the Solow model's properties.

Rationale:
The fund's objective explicitly states that a false positive (FP) error is twice as detrimental as a false negative (FN) error. To evaluate the models in line with this objective, Emma should calculate a custom weighted error cost for each model. Let's assign a cost of 2 units for each false positive and 1 unit for each false negative. For Model A: False Positives (FP_A) = 20 False Negatives (FN_A) = 50 Weighted Error Cost_A = (2 * FP_A) + (1 * FN_A) = (2 * 20) + (1 * 50) = 40 + 50 = 90 units. For Model B: False Positives (FP_B) = 35 False Negatives (FN_B) = 40 Weighted Error Cost_B = (2 * FP_B) + (1 * FN_B) = (2 * 35) + (1 * 40) = 70 + 40 = 110 units. Since Model A has a lower total weighted error cost (90 units) compared to Model B (110 units), Model A is the most suitable choice according to the fund's specific, asymmetric cost objective. The option stating Model B has superior recall is incorrect. While Model B (Recall = 95 / (95 + 40) = 0.7037) does exhibit higher recall than Model A (Recall = 80 / (80 + 50) = 0.6154), prioritizing recall alone overlooks the stated higher penalty for false positives. The option stating Model A has higher precision is incorrect. Model A (Precision = 80 / (80 + 20) = 0.80) does indeed have higher precision than Model B (Precision = 95 / (95 + 35) = 0.7308). However, focusing solely on precision would ignore the cost associated with false negatives and the overall weighted objective. The option stating Model B has a higher standard F1 score is incorrect. The F1 score is calculated as 2 * (Precision * Recall) / (Precision + Recall). For Model A, F1_A = 2 * (0.80 * 0.6154) / (0.80 + 0.6154) ≈ 0.696. For Model B, F1_B = 2 * (0.7308 * 0.7037) / (0.7308 + 0.7037) ≈ 0.717. While Model B does have a higher standard F1 score, the F1 score implicitly assumes equal importance for precision and recall, which contradicts the fund's explicitly stated asymmetric cost for false positives versus false negatives.

Rationale:
The core principle for calculating uncollateralized exposure under partial netting enforceability requires separating the portion of gross positive exposure that cannot be offset from the portion that can. In this case, 20% of the gross positive MTM ($100 million) is explicitly stated to be irreducible by netting or collateral. This amounts to $20 million (0.20 * $100M). The remaining 80% of the gross positive MTM, which is $80 million (0.80 * $100M), is subject to standard netting and collateralization. This $80 million portion is first netted by the gross negative MTM of $60 million, resulting in a net exposure of $20 million (Max(0, $80M - $60M)). This remaining $20 million is then reduced by the $30 million in cash collateral held. Since the collateral ($30M) exceeds this net exposure ($20M), this portion becomes $0. Therefore, the total uncollateralized exposure is the sum of the irreducible portion ($20 million) and the fully collateralized reducible portion ($0), which equals $20 million. The approach of assuming full netting enforceability, which yields $10 million, incorrectly treats the entire gross positive exposure as reducible by both negative MTM and collateral, ignoring the regulatory legal interpretation about the 20% irreducible portion. The option calculating $8 million likely results from applying a 20% reduction factor to the final net collateralized exposure (i.e., $10 million * 0.80), which misinterprets the nature of the '80% effective' rule as a general discount rather than a specific constraint on offset mechanics for a part of the exposure. The calculation resulting in $70 million incorrectly treats the 80% reducible portion of gross positive MTM as if it's only reduced by collateral, ignoring the benefit of the gross negative MTM, or making an error in the order of operations for netting and collateral.

Rationale:
The core principle here is the duty of loyalty, prudence, and care to clients (Standard III(A)) and the avoidance of misrepresentation (Standard I(C)). James has identified a significant increase in tail risk (CVaR at 280 bps) that breaches a more stringent internal firm threshold (200 bps), even if the client's explicit VaR limit (150 bps) is technically not exceeded. CVaR provides a more comprehensive measure of tail risk than VaR, especially under stressed market conditions or non-normal distributions, and its significant breach represents a material change in the portfolio's risk profile that clients have a right to know. Full and fair disclosure, even if inconvenient or reflecting an internal control issue, is paramount. Delaying communication to first resolve the issue, as in 'Focus solely on restoring the portfolio's CVaR,' would violate the duty of loyalty and potentially misrepresent the actual risk clients were exposed to. Similarly, 'Document the internal CVaR breach and only discuss it with clients if the standard VaR limit is also exceeded' is insufficient because it prioritizes a less comprehensive risk measure (VaR) over a more critical one (CVaR) for tail risk, and delays informing clients of a material risk. The option to 'Adjust the portfolio to reduce the CVaR to 150 bps' demonstrates a misunderstanding of the metrics, as CVaR is inherently higher than VaR for the same confidence level and is not directly comparable to a VaR limit in this manner. Furthermore, even if such an adjustment were made, the initial breach and the client's exposure to that elevated risk still occurred and requires disclosure.