Free FRM Part II Practice Exam: 80 Applied Risk Questions

How to use this set

This fixed set contains 80 original questions from the reviewed Finance Prep FRM Part II bank. It has 16 questions each in market, credit and operational risk; 12 each in liquidity/treasury and investment risk; and 8 in Current Issues. The allocation follows the six approximate 2026 domain weights as an editorial practice design, not a claim about an official exam form.

Each self-contained case has four choices, one best answer and a rationale for every choice. For a pacing exercise, allow 240 minutes and answer before opening explanations. This page does not reproduce GARP’s testing interface or official proprietary content; it is not an exam dump.

Pacing and score record

Use these checkpoints for an optional four-hour attempt. They describe an average pace, not a deadline for each item.

Score manually: give one point for the keyed answer and zero for an incorrect or unanswered item. Divide by 80, including unanswered questions. For example, 60 correct, 15 incorrect and five unanswered gives 60/80 = 75%. Keep an answer record before opening explanations; the page has no automatic timer or scorer.

Record market, credit and operational results out of 16 each, liquidity/treasury and investment results out of 12 each, and Current Issues out of 8. A strong aggregate can conceal a gap in a smaller domain. No universal fixed FRM passing percentage is established by the official sources used here; this practice result is not a validated pass prediction.

Reserve a separate explanation-review session. Repeating this fixed set can reward memory, so follow correction work with fresh app questions.

Practice questions

Questions 1-25

Question 1

Topic: Market Risk Measurement and Management

An interest-rate risk analyst is reviewing a dealer’s ask of $898.25 for a default-free zero-coupon bond paying $1,000 in three years. The desk uses the following short-rate tree as its valuation benchmark.

Valuation inputs:

• Each node rate is an annual effective short rate for the immediately following one-year interval.
• At every branching node, each successor has a risk-neutral probability of 0.50.
• The desk’s real-world forecast assigns probability 0.75 to the higher-rate successor and 0.25 to the lower-rate successor.
• Ignore transaction costs.

Which conclusion correctly evaluates the dealer’s ask against the desk’s valuation benchmark?

• A. The model value is approximately $873.27, so the dealer’s ask is above model value.
• B. The model value is approximately $898.96, so the dealer’s ask is below model value.
• C. The model value is approximately $915.14, so the dealer’s ask is below model value.
• D. The model value is approximately $897.63, so the dealer’s ask is above model value.

Best answer: B

Explanation: Risk-neutral pricing discounts conditional expected continuation values at the current node’s short rate. For each one-year branching step, \( V_t = [0.5V_{t+1}^{H}+0.5V_{t+1}^{L}]/(1+r_t) \), where \( r_t \) is the node’s annual effective short rate.

At year 2, discounting the $1,000 maturity payment gives values of $925.9259, $961.5385, and $1,000.0000 at the 8%, 4%, and 0% nodes, respectively.

At year 1, the higher-rate node has value \( (925.9259+961.5385)/(2\times1.06)=890.3134 \). The lower-rate node has value \( (961.5385+1{,}000)/(2\times1.02)=961.5385 \).

The initial value is therefore:

\[ V_0=\frac{0.5(890.3134)+0.5(961.5385)}{1.03}\approx898.96 \]

The ask is approximately $0.71 below model value. Real-world probabilities describe anticipated rate movements; they do not replace risk-neutral valuation weights.

• A. This amount uses the 0.75 forecast probability for higher rates rather than the supplied 0.50 risk-neutral probability required for valuation.
• B. Backward induction using the risk-neutral probabilities gives $898.96, exceeding the $898.25 ask by approximately $0.71.
• C. This amount discounts all three years at the current 3% rate, overlooking the future node rates specified in the tree.
• D. This calculation uses the risk-neutral average rate for each year instead of discounting each conditional continuation value at its node’s short rate.

---

Question 2

Topic: Market Risk Measurement and Management

A bank applies a joint rise in bond yields and decline in equity prices to its trading portfolio. Full revaluation produces the following results. All positions remain open through the next-business-day collateral deadline.

Capital and escalation policy:

• Pre-stress loss-absorbing capital is $180 million, and the internal post-stress capital floor is $150 million.
• The net portfolio stress-loss limit is $20 million. Stress P&L changes capital dollar for dollar; ignore taxes and other capital adjustments.
• Every identified stress-loss limit breach, capital-floor breach, and inability to meet a collateral deadline requires escalation.

Scenario funding:

• Repo lenders require $18 million of additional cash collateral by the next business day.
• The bank has $15 million of immediately available cash.
• A $10 million committed facility becomes drawable on the second business day.
• No other funding sources or position cash flows are available before the collateral deadline.

Which escalation recommendation follows from these results and the policy?

• A. Escalate the stress-loss limit breach and next-day funding shortfall; capital remains above its floor.
• B. Escalate the stress-loss limit breach, capital-floor breach, and next-day funding shortfall.
• C. Escalate the stress-loss limit breach only; capital and next-day funding remain sufficient.
• D. Escalate the stress-loss limit and capital-floor breaches; next-day funding remains sufficient.

Best answer: A

Explanation: Stress losses, capital capacity, and funding liquidity measure different constraints. The long positions lose $30 million, offset by the $6 million hedge gain, producing a $24 million net stress loss. This exceeds the $20 million limit and requires escalation.

Capital falls from $180 million to $156 million, remaining above the $150 million internal floor. Cash collateral posting does not create a separate P&L loss and should not be deducted from capital again.

The next-day collateral requirement is $18 million, but only $15 million of cash is available, creating a $3 million funding shortfall. The committed facility arrives after the deadline, and the unrealized hedge gain supplies no cash while the position remains open. Therefore, the stress-loss breach and liquidity shortfall require escalation, despite sufficient post-stress capital.

• A. The $24 million loss breaches the $20 million limit, capital remains $156 million, and cash is $3 million short of the collateral requirement.
• B. Posting collateral does not create an additional trading loss, so post-stress capital remains $156 million and exceeds the $150 million floor.
• C. The facility is unavailable before the collateral deadline, leaving a $3 million shortfall that also requires escalation.
• D. The $24 million net loss leaves capital above its floor, while the delayed facility cannot cover the next-day cash shortfall.

---

Question 3

Topic: Market Risk Measurement and Management

A risk manager reviews a corporate-bond portfolio with interest-rate hedges under a curve-steepening and spread-widening scenario. Government-yield factors represent general risk. The issuer spread, measured over the government curve, represents specific risk.

The table gives net, signed first-order P&L sensitivities after hedging. Each sensitivity measures the dollar P&L change for a +1 bp factor move, holding other factors constant.

Scenario assumptions:

• Full-revaluation P&L equals mapped first-order P&L plus a correction that can take any value from -$150,000 to +$150,000. This scenario-specific correction captures omitted nonlinear and hedge-basis effects; no other corrections apply.
• The maximum permitted full-revaluation scenario loss is $1,000,000.

Which assessment of the portfolio’s scenario loss and limit compliance is supported by these data?

• A. The mapped loss is $1,000,000; the full-revaluation loss ranges from $850,000 to $1,150,000, so limit compliance is unresolved.
• B. The mapped loss is $900,000; the full-revaluation loss ranges from $750,000 to $1,050,000, so limit compliance is unresolved.
• C. The mapped loss is $700,000; the full-revaluation loss ranges from $550,000 to $850,000, so the portfolio meets the limit.
• D. The mapped loss is $1,200,000; the full-revaluation loss ranges from $1,050,000 to $1,350,000, so the portfolio breaches the limit.

Best answer: B

Explanation: First-order factor mapping multiplies each signed dollar sensitivity by its factor change and sums the results. A positive sensitivity produces a gain when its factor rises.

General-risk P&L is \(10,000 \times 5 - 30,000 \times 25 = -700,000\) dollars. Specific-risk P&L is \(-20,000 \times 10 = -200,000\) dollars. The total mapped P&L is therefore -$900,000, equivalent to a $900,000 loss.

Adding the correction gives full-revaluation P&L between -$1,050,000 and -$750,000. Equivalent losses range from $750,000 to $1,050,000. Because this interval straddles the $1,000,000 limit, neither compliance nor a definite breach is established. Further valuation information is needed to resolve the classification. The correction preserves the stated nonlinear and basis uncertainty rather than treating the local sensitivity approximation as exact.

• A. The 2-year exposure produces a $50,000 gain, not a loss; reversing that sign overstates the mapped loss by $100,000.
• B. General-risk losses of $700,000 plus specific-risk losses of $200,000 yield $900,000; the correction permits losses on either side of the limit.
• C. The $700,000 figure captures general-rate risk only and omits the additional $200,000 loss from the 10 bp issuer-spread widening.
• D. Using 25 bp instead of 10 bp for the issuer-spread move overstates the spread loss by $300,000.

---

Question 4

Topic: Credit Risk Measurement and Management

A CCP with six clearing members proposes to reduce member A’s initial margin and increase every member’s prefunded default-fund contribution. The five other members’ initial margins remain unchanged. Aggregate prefunded resources are unchanged.

Resources, in USD millions:

Default and cost assumptions:

• A’s latest $8 million variation margin payment has settled. If A defaults, the CCP’s remaining close-out loss is $43 million, already net of that payment, under either policy.
• Losses consume A’s initial margin, then A’s default-fund contribution, then the committed CCP capital, and finally the surviving members’ default-fund contributions, allocated equally. Surviving members’ initial margins are unavailable.
• For this comparison, the default scenario has a 2% annual probability, and no other members default.
• Each member’s annual economic clearing cost is a 4% funding charge on its full opening initial margin and default-fund contribution, plus its expected default-fund loss. Ignore other costs.

What are the proposal’s effects on a representative surviving member’s annual economic clearing cost and the total loss mutualized if A defaults?

• A. Annual cost increases by $120,000; the total mutualized default loss is $16 million.
• B. Annual cost increases by $144,000; the total mutualized default loss is $16 million.
• C. Annual cost increases by $80,000; the total mutualized default loss is $6 million.
• D. Annual cost increases by $128,000; the total mutualized default loss is $21 million.

Best answer: A

Explanation: Variation margin settles current mark-to-market exposure, so the settled payment must not be deducted again from the stated close-out loss. Initial margin and the defaulting member’s default-fund contribution absorb losses before mutualization.

Currently, the surviving members absorb \(43 - 30 - 3 - 4 = 6\) million dollars. Under the proposal, they absorb \(43 - 18 - 5 - 4 = 16\) million dollars. Equal allocation among five survivors raises each member’s conditional loss from $1.2 million to $3.2 million.

Each survivor posts an additional $2 million to the default fund, increasing annual funding cost by $80,000. Its expected annual loss increases by 2% of the $2 million increase in conditional loss, or $40,000. The total annual cost increase is therefore $120,000.

Although aggregate resources remain unchanged, A’s own loss-absorbing resources fall by $10 million. The proposal lowers A’s collateral funding burden while shifting more default losses to other members, illustrating the allocation and incentive effects of mutualized resources.

• A. Mutualized losses rise from $6 million to $16 million; each survivor incurs $80,000 more funding cost and $40,000 more expected loss.
• B. This includes the proposed expected loss of $64,000 but fails to subtract the current expected loss of $24,000 when measuring the cost increase.
• C. Unchanged aggregate resources do not preserve loss allocation: reducing A’s own loss-absorbing resources increases mutualization and adds expected loss beyond the funding charge.
• D. These amounts omit A’s own default-fund contribution, which must absorb losses before CCP capital and surviving members’ contributions.

---

Question 5

Topic: Operational Risk and Resilience

A bank’s Payments department outsources daily reconciliation of high-value payments. A vendor system failure has left 34 payments unreconciled for more than 48 hours, compared with a board-approved tolerance of five. No financial loss has been confirmed. The Head of Payments proposes using manual controls for six weeks while the vendor repairs its system.

Governance and review:

• The executive operational risk committee coordinates cross-functional remediation. The board risk committee oversees the bank’s operational risk appetite.
• Independent operational risk has challenged the effectiveness of the temporary manual controls.
• Compliance found no current regulatory notification requirement; its review did not assess reconciliation effectiveness.
• Internal audit has not reviewed the temporary controls and is scheduled to provide independent assurance after remediation.

Which allocation of responsibility and reporting is most consistent with Basel operational risk governance expectations during the six-week period?

• A. Assign remediation accountability to Payments, require operational risk to challenge the control changes, and have the executive committee report the open breach to the board risk committee.
• B. Assign remediation accountability to operational risk, require Payments to implement control changes, and have the executive committee report the open breach to the board risk committee.
• C. Assign remediation accountability to Payments, require operational risk to challenge the control changes, and have the executive committee report the breach to the board risk committee when compliance requires regulatory notification.
• D. Assign remediation accountability to Payments, require internal audit to direct the control changes, and have the executive committee report the open breach to the board risk committee.

Best answer: A

Explanation: Basel operational risk governance separates business accountability, independent challenge, and assurance. Outsourcing reconciliation does not transfer the bank’s responsibility for the process. Payments remains accountable for temporary controls, vendor remediation, and residual exposure. The executive operational risk committee coordinates delivery rather than replacing that owner.

Operational risk independently challenges the adequacy of controls and the resulting exposure. Compliance assesses applicable obligations, but its notification conclusion is not an assurance opinion on reconciliation effectiveness. Internal audit provides independent assurance and should not direct the changes it will later assess.

The 34 unresolved payments exceed the board-approved tolerance of five. This open breach should be reported through the board risk committee during remediation, even though no loss is confirmed and audit assurance is pending. Board-level oversight of appetite does not make the board responsible for operating reconciliations.

• A. Payments retains accountability for its outsourced process, operational risk provides independent challenge, and the unresolved tolerance breach requires board-level oversight.
• B. Assigning remediation accountability to operational risk displaces the business’s first-line ownership, even when Payments implements the changes.
• C. The absence of a regulatory notification requirement does not justify withholding a board-approved tolerance breach from the board risk committee.
• D. Directing control changes gives internal audit a management role and creates a self-review conflict when it later provides assurance over remediation.

---

Question 6

Topic: Risk Management and Investment Management

A fund-of-funds risk officer is assessing a proposed allocation to a hedge fund manager. The committee wants to determine whether the reported performance history is credible enough to support the allocation.

Due-diligence findings:

• The manager reports using split-strike conversion: holding liquid large-cap stocks, buying index puts, and selling index calls.
• Returns were positive in 34 of the past 36 months, with annualized volatility of 2%.
• A sampled statement reports purchasing 9,000 put contracts in one series on a particular exchange. That exchange’s official total volume for the same series and day was 4,000 contracts. The manager confirms that all 9,000 contracts were executed there that day.
• The manager’s affiliated broker executes trades, holds securities, and produces account statements. No independent custody reports or execution confirmations have been obtained. The manager declines to authorize direct independent confirmation, citing strategy secrecy.

Which due-diligence focus should the risk officer prioritize?

• A. Prioritize independent position and execution checks to substantiate the reported holdings and option trades.
• B. Prioritize independent stress tests to determine whether the reported strategy understates downside exposure.
• C. Prioritize independent valuation checks to determine whether stale marks account for the smooth returns.
• D. Prioritize independent capacity checks to determine whether option-market depth can accommodate additional capital.

Best answer: A

Explanation: The central warning is unverified investment activity, not merely unusually smooth returns. Split-strike conversion combines an equity basket with long puts and short calls and can reduce volatility, so return stability alone does not prove fraud.

Purchasing 9,000 contracts on an exchange reporting only 4,000 total contracts for the same series and day is inconsistent with the stated execution account. That discrepancy requires independent reconciliation. Statements produced by the manager’s affiliated broker do not independently substantiate the assets or transactions, and strategy secrecy does not resolve this evidence gap.

The Madoff case illustrates why reported strategies and performance must be distinguished from verified investment activity. Independent custody and execution evidence should support reliance on the return history. These findings justify withholding that reliance pending investigation, rather than conclusively declaring fraud.

• A. The inconsistent execution record and absence of independent custody evidence directly undermine whether actual holdings and trades support the reported performance.
• B. Stress testing evaluates losses from an assumed portfolio; it cannot substantiate actual holdings or resolve the inconsistent reported transaction.
• C. Stale pricing could smooth returns, but correcting marks would not reconcile purchases exceeding total exchange volume or independently establish asset ownership.
• D. Capacity matters for additional allocations, but assessing future market depth cannot authenticate reported purchases exceeding the exchange’s entire volume.

---

Question 7

Topic: Operational Risk and Resilience

A bank’s operational risk committee is reviewing renewal of a critical loan-servicing outsourcing agreement. Its priority is to demonstrate service continuity if the bank voluntarily replaces the supplier while the supplier remains solvent.

Review evidence:

• Pre-engagement due diligence covered the supplier’s financial condition, security, and operational capabilities.
• Ongoing control reports show no unresolved exceptions. The latest recovery exercise met the bank’s recovery-time objective by restoring service at the supplier’s secondary site.
• The contract establishes bank ownership of records and requires daily exports of complete transaction histories in a documented format.
• Reconstructing loan balances from those histories requires the supplier’s proprietary calculation engine. An escrow agreement releases the engine only upon supplier insolvency and licenses it solely for the bank’s internal use.

Which additional assurance package most directly addresses the committee’s priority?

• A. Review the replacement provider’s financial strength and operating controls, then reconcile transferred transaction histories against the bank’s retained source records.
• B. Secure engine access and replacement-provider licensing for voluntary exits, then test end-to-end servicing with the exported transaction histories.
• C. Refresh the incumbent’s financial and security due diligence, then repeat its secondary-site recovery exercise using the bank’s current transaction volumes.
• D. Validate the escrowed engine’s completeness and functionality, then test bank-run recovery under the existing escrow-release and internal-use licensing terms.

Best answer: B

Explanation: Third-party controls must address the relevant lifecycle risk. Pre-engagement due diligence assesses supplier suitability. Ongoing monitoring evaluates continuing performance, controls, and changes in risk. Exit planning establishes whether services and records can be transferred while maintaining continuity.

Here, successful incumbent recovery does not demonstrate substitutability. Record ownership and complete exports support information transfer, but operational use still depends on the proprietary calculation engine. The escrow arrangement addresses supplier insolvency and permits only bank-internal use, so it does not support the contemplated voluntary transfer.

The bank should secure engine availability during voluntary exit and usage rights for a replacement provider. An end-to-end transition test should then demonstrate that the replacement can use the exported histories to continue servicing.

• A. Provider suitability and record completeness support a transition, but they do not establish the replacement provider’s access to the required calculation engine.
• B. A voluntary transition requires lawful engine access for the replacement provider and evidence that transferred records support continued loan servicing.
• C. These checks reassess incumbent suitability and recovery capability, rather than establish whether loan servicing can continue through a change of provider.
• D. The existing escrow supports an insolvency contingency, not engine access or replacement-provider use during a voluntary exit from a solvent supplier.

---

Question 8

Topic: Operational Risk and Resilience

A bank’s risk committee is considering renewal of a fallback payment channel. It wants assurance that every payment of at least $2 million receives approval from someone other than its initiator before bank release.

Current evidence:

• Operations attests that the control works across normal and fallback channels. Its dashboard reports full compliance using a table populated only when an approval is recorded. Both channels feed this table.
• Internal audit independently tested 60 records from that table, verifying approver identities and approval timing, with no exceptions.
• Daily general-ledger payment totals reconcile to bank debits. Normal payments are individually itemized, but fallback batches appear as aggregate postings.
• An authenticated file obtained directly from the bank lists every completed payment from both channels, including its identifier, amount and release timestamp.

An independent assurance team will conduct further testing of approver separation and pre-release approval. Which approach would provide the strongest additional assurance for reliance on this control?

• A. Use the bank-release file as the population, selecting completed payments independently and tracing them to their underlying authorization evidence.
• B. Use individually itemized ledger payments as the population, validating its completeness through the reconciled daily bank-debit totals.
• C. Use approval records matched to the bank-release file as the population, resolving unmatched approval records before independent execution testing.
• D. Use the approval table as the population, expanding the independent test sample and stratifying it across both payment channels.

Best answer: A

Explanation: Assurance over operating effectiveness requires reliable evidence drawn from the events the control is intended to govern. Here, those events are completed payments subject to the approval requirement, not recorded approvals.

The existing audit results support operation within the tested approval-record population. They do not establish that released payments lacking approval records would have been detected. Increasing sample size cannot repair that selection limitation.

The bank-release file provides a transaction-level population across both channels. Independent testing can select qualifying releases and inspect evidence that a distinct approver authorized each payment before release. Missing evidence should be investigated rather than excluded from selection. Management’s attestation and reconciled accounting totals remain useful supporting evidence, but neither substitutes for verified control execution. Any reliance decision should reflect the resulting exceptions and testing limitations.

• A. Sampling completed releases can expose missing or late approvals, rather than conditioning selection on the existence of an approval record.
• B. Daily value reconciliation includes aggregated fallback batches but does not make their constituent payments available in an individually itemized ledger population.
• C. Matching from approval records to releases tests recorded approvals, while released payments lacking approval records remain outside the sampling population.
• D. Stratification and a larger sample cannot reveal payments omitted because no approval record exists, leaving population completeness untested.

---

Question 9

Topic: Liquidity and Treasury Risk Measurement and Management

A bank’s liquidity risk manager reviews an uninsured-deposit runoff indicator. Weekly runoff has increased from 2.2% to 3.8% to 5.6%; the latest withdrawal amount was $280 million. The bank’s current liquidity coverage ratio (LCR) is 126%.

Liquidity policy:

• Weekly uninsured-deposit runoff above 5% requires same-day escalation to the liquidity risk committee.
• The initial contingency funding stage must be activated if any projected end-of-day unrestricted cash balance over the next five business days falls below $75 million under the approved stress scenario.
• Funding receipts enter that projection only when contractually committed and available by the relevant day.

Funding evidence:

• Opening unrestricted cash is $200 million.
• A committed $120 million facility can settle at the start of day 4. All drawdown conditions are satisfied, but earlier settlement is unavailable.
• Another lender has provided a nonbinding $80 million credit offer with proposed settlement on day 1.

The approved projection contains the following net outflows, excluding these funding receipts. No other funding sources are available within the horizon.

Which governance recommendation and supporting basis are most consistent with the policy and funding evidence?

• A. Escalate today and continue enhanced monitoring, using the projected day-five cash buffer to determine whether contingency activation is necessary.
• B. Escalate today and continue enhanced monitoring, including the indicative credit offer when determining whether projected daily cash buffers require contingency activation.
• C. Escalate today and continue enhanced monitoring, using the current liquidity coverage ratio to determine whether contingency activation is necessary.
• D. Escalate today and activate the initial contingency stage, using the lowest projected daily cash buffer to determine whether activation is necessary.

Best answer: D

Explanation: An early-warning breach requires governance attention, but it does not by itself predict inevitable funding failure. The 5.6% runoff breaches the bank’s warning threshold and requires same-day escalation. Contingency activation depends separately on the stressed daily cash profile.

Starting with $200 million, projected end-of-day cash balances are $160 million, $60 million, $40 million, $140 million, and $120 million. These balances include the committed $120 million receipt on day 4 but exclude the nonbinding credit offer. The minimum balance of $40 million breaches the $75 million buffer requirement, so the initial contingency stage must be activated now.

Funding arriving on day 4 cannot repair an earlier buffer breach. Likewise, a positive day-five balance and an LCR above its regulatory minimum do not override the bank’s shorter-horizon internal trigger. Activation is a preventive response to projected vulnerability, not a declaration that default is certain.

• A. The day-five balance of $120 million conceals earlier buffer breaches; the policy tests every projected daily balance rather than horizon-end cash.
• B. The $80 million offer is nonbinding and therefore ineligible for the projection; including it would incorrectly conceal the interim cash-buffer breach.
• C. An LCR of 126% does not override the internal five-day cash-buffer trigger, which is breached before committed funding arrives.
• D. Projected cash reaches $40 million on day 3, below $75 million, requiring contingency activation in addition to escalation of the runoff breach.

---

Question 10

Topic: Market Risk Measurement and Management

A market risk manager is reviewing a desk’s one-day 99% VaR against a $4.75 million limit. A breach occurs when VaR exceeds the limit. The approved reporting model uses a non-parametric uniform-kernel density estimate with a $2.0 million bandwidth. A $0.5 million bandwidth is used for comparison.

Historical sample: The 100 equally weighted daily losses are summarized below. All loss amounts in the exhibits are in USD millions; negative values denote gains.

Each observation’s probability mass is spread uniformly from its loss minus the bandwidth to its loss plus the bandwidth.

Tail cumulative probabilities: Both smoothed CDFs are linear between the listed loss levels.

Define VaR as the smallest loss whose cumulative probability is at least 0.99. Which conclusion about limit compliance and bandwidth sensitivity is supported?

• A. Approved VaR is $4.5 million and comparison VaR is $5.0 million; only the comparison estimate breaches the limit.
• B. Approved VaR is $5.0 million and comparison VaR is $4.5 million; only the approved estimate breaches the limit.
• C. Approved VaR is $5.5 million and comparison VaR is $4.5 million; only the approved estimate breaches the limit.
• D. Approved VaR is $4.0 million and comparison VaR is $4.0 million; neither smoothed estimate breaches the limit.

Best answer: B

Explanation: Kernel smoothing replaces discrete probability masses with continuous distributions, so the resulting quantile need not equal an observed loss. Without smoothing, the 99th ordered observation is $4.0 million, making empirical 99% VaR $4.0 million.

For the approved estimator, cumulative probability increases from 0.9875 at $4.5 million to 0.9925 at $5.5 million. Linear interpolation gives VaR, in USD millions:

\[ \mathrm{VaR}_{0.99} = 4.5 + \frac{0.99 - 0.9875}{0.9925 - 0.9875} \times (5.5 - 4.5) = 5.0 \]

The comparison CDF first reaches 0.99 at $4.5 million, even though it remains flat afterward. Thus, the approved estimate breaches the $4.75 million limit, while the narrower-bandwidth estimate does not.

The bandwidth change reverses the limit classification in this sample. Only two observations exceed $1.0 million: smoothing redistributes their probability mass but creates no additional tail evidence, so tail-data scarcity remains.

• A. The approved CDF reaches 0.99 at $5.0 million, while the comparison CDF reaches it at $4.5 million, so these estimates are reversed.
• B. Interpolating the approved CDF gives $5.0 million, while the comparison CDF first reaches 0.99 at $4.5 million, placing them on opposite sides of the limit.
• C. At $5.5 million, the approved CDF already equals 0.9925; interpolation locates the 0.99 quantile at $5.0 million rather than the first listed exceedance.
• D. $4.0 million is the discrete empirical 99% VaR; kernel smoothing moves the comparison and approved quantiles to $4.5 million and $5.0 million, respectively.

---

Question 11

Topic: Operational Risk and Resilience

An operational risk manager reviews an unauthorized change to a bank’s payment-validation application.

Review findings:

• Testing was completed on a package, and an independent release manager approved its release label, pay-42. Developers cannot alter the approval record.
• A compromised developer account then replaced the package stored under pay-42 before deployment.
• At deployment, a trusted automated signing service signs whichever package is currently stored under the label. The production gate verifies that signature and checks that the label is approved.
• The replacement altered payment-validation rules, passed the existing scan for known vulnerabilities, and was deployed successfully. The signing key was not compromised.

The bank’s priority is to prevent unreviewed replacement packages from reaching production. Which control response best addresses the documented weakness?

• A. Require a vulnerability rescan of each retrieved package immediately before deployment and reject packages with critical known vulnerabilities.
• B. Require two independent release approvals for each label and reject deployment unless both approvals are present.
• C. Require the retrieved package’s digest to match the reviewed package’s digest in the protected approval record and reject mismatches.
• D. Require hardware-protected signing keys and reject deployment unless the retrieved package has a valid signature from the trusted signing service.

Best answer: C

Explanation: Preventive change control must bind independent authorization to the exact package deployed. Here, approval attaches to a mutable label, so it remains valid after the underlying package is replaced. Signing the replacement authenticates the signing service and protects integrity from signing onward; it does not establish that the package received independent approval.

The cryptographic digest of the tested package should be stored in the protected approval record. At deployment, the retrieved package’s digest should be compared with that approved value, with any mismatch blocking deployment. Because developers cannot alter the approval record, replacing repository contents cannot authorize a different package. This preserves separation of duties and closes the gap between review and production deployment.

• A. The replacement passed vulnerability scanning, and absence of known vulnerabilities does not establish that its contents were reviewed and approved.
• B. Two approvals of a mutable release label still authorize that label after its associated package has been replaced.
• C. Matching the retrieved package to the independently approved digest prevents a replacement from inheriting authorization through an unchanged release label.
• D. The signing key was not compromised, and the trusted service would still sign the unreviewed replacement package.

---

Question 12

Topic: Liquidity and Treasury Risk Measurement and Management

A euro-area corporation needs one-year USD funding. If it borrows EUR, it will hedge both principal and interest. Its treasurer reviews these executable quotes:

Conventions: All interest rates are simple annual rates, with principal and interest paid at maturity. Ignore default, bid-ask spreads, and costs other than those specified.

Market conditions: Non-US institutions have increased demand to obtain USD through FX swaps. Dealer balance-sheet capacity tightens around reporting dates. An additional dealer trade that borrows USD, invests in EUR, and sells the EUR maturity proceeds forward incurs a balance-sheet and liquidity charge of 120 basis points of initial USD notional, payable at maturity. This charge does not apply to the corporation’s funding arrangements.

Which conclusion correctly identifies the corporation’s lowest-cost funding route and the dealer trade’s net return relative to initial USD notional?

• A. Use hedged EUR borrowing at 3.00%; the dealer trade returns -0.14% after balance-sheet and liquidity charges.
• B. Use direct USD borrowing at 4.00%; the dealer trade returns -0.14% after balance-sheet and liquidity charges.
• C. Use hedged EUR borrowing at 5.06%; the dealer trade returns -0.14% after balance-sheet and liquidity charges.
• D. Use hedged EUR borrowing at 5.06%; the dealer trade returns 1.06% after balance-sheet and liquidity charges.

Best answer: C

Explanation: Covered interest parity equates covered currency returns under frictionless funding and investment conditions. When exchange rates are quoted in USD per EUR, the effective USD cost of EUR borrowing is:

\[ r_{\mathrm{USD,hedged}} = \frac{F}{S}(1+r_{\mathrm{EUR}})-1 = \frac{1.1330}{1.1000}(1.02)-1 = 5.06\%. \]

Here, \(S\) is the spot rate and \(F\) is the forward rate. Hedged EUR funding therefore saves the corporation 14 basis points relative to its 5.20% direct USD quote.

The dealer’s EUR investment, hedged into USD, also earns 5.06%. Subtracting its 4.00% USD funding cost leaves a gross convergence spread of 1.06%. The additional 1.20% balance-sheet and liquidity charge reduces the net return to -0.14%.

The apparent arbitrage is consequently uneconomic for the dealer. Increased demand for USD through FX swaps can raise hedged dollar funding costs, while constrained dealer balance sheets limit the supply of offsetting trades. Such post-crisis deviations can persist, especially around reporting dates, even when exchange-rate exposure is fully hedged.

• A. The 3.00% forward premium excludes EUR interest; converting both principal and interest at the forward rate raises the hedged USD cost to 5.06%.
• B. The 4.00% USD quote belongs to the dealer, not the corporation, whose 5.20% USD borrowing cost exceeds its 5.06% hedged EUR cost.
• C. Hedged EUR borrowing costs less than the corporation’s 5.20% USD quote, while the dealer’s 1.06% gross spread is below its 1.20% charge.
• D. The 1.06% spread is before the 120-basis-point charge; deducting that charge produces a net return of -0.14%.

---

Question 13

Topic: Operational Risk and Resilience

A bank is validating its 2025 annual loss total under the Basel Committee’s consolidated operational-risk standardized approach, using the loss-component dataset threshold convention in OPE25.18 FAQ 3. Its draft ten-year loss history produces an internal loss multiplier below 1, which management interprets as evidence of low operational risk.

The bank applies a €20,000 per-event threshold to cumulative net loss within the ten-year loss-component calculation window. No event exclusions have supervisory approval. The following separate events are the only records potentially relevant to the 2025 total, and all their loss and recovery impacts within the ten-year window are shown.

Event records:

• Payment error: Occurred and was first accounted for in 2025. Gross loss was €60,000, with a €45,000 reimbursement received in 2025.
• External fraud: Occurred in 2024, but was discovered and first accounted for in 2025. Gross loss was €100,000. An insurance receivable of €30,000 was recognized in 2025 but remains unpaid at the review date.
• Conduct settlement: Misconduct occurred in 2023 and was discovered and first accounted for in 2025. Gross loss was €160,000, with a €40,000 recovery received in 2025. The related product was discontinued during 2025.
• Internal fraud: Occurred in December 2025, but was discovered and first accounted for in January 2026. Gross loss was €90,000, with no recovery.

Draft 2025 extract: The payment error and conduct settlement were omitted. External fraud was included at €70,000 and internal fraud at €90,000, producing a total of €160,000.

Which eligible net loss total should replace the draft 2025 total before the bank recalculates its loss component?

• A. Replace the draft 2025 total with €220,000 of eligible net operational losses.
• B. Replace the draft 2025 total with €115,000 of eligible net operational losses.
• C. Replace the draft 2025 total with €205,000 of eligible net operational losses.
• D. Replace the draft 2025 total with €235,000 of eligible net operational losses.

Best answer: A

Explanation: For the specified loss-component dataset, OPE25.18 FAQ 3 applies the event threshold to cumulative net loss within the ten-year window, using the accounting dates of loss and recovery impacts. The payment error’s net impact is €15,000, below €20,000, so it is excluded from this calculation. Only recoveries actually received reduce net loss; the unpaid insurance receivable cannot reduce the external fraud loss.

For these events, annual allocation follows the first accounting date. External fraud belongs in 2025, whereas internal fraud belongs in 2026; the latter was also not discovered until 2026. Discontinuing a product does not automatically justify excluding its losses, and no exclusion has supervisory approval.

The eligible 2025 total, in euros, is:

\[ 100,000 + (160,000 - 40,000) = 220,000 \]

The corrected annual total must feed into the ten-year loss history before recalculating the loss component and multiplier. A multiplier below 1 reflects recorded historical losses relative to the business indicator component, not proof of low underlying exposure. Complete identification and consistent inclusion are prerequisites for interpreting that result.

• A. The payment error’s net loss is €15,000, below the €20,000 event threshold for the ten-year loss-component dataset. The eligible 2025 losses are €100,000 for external fraud and €120,000 for the conduct settlement, totaling €220,000.
• B. This total includes the below-threshold €15,000 payment loss while omitting the €120,000 conduct loss. Discontinuing the product does not authorize exclusion; without supervisory approval, the eligible conduct loss remains in the dataset.
• C. This total includes the below-threshold €15,000 payment loss and deducts an unpaid €30,000 insurance receivable. It equals €15,000 + €70,000 + €120,000; neither treatment is appropriate for the loss-component dataset.
• D. This total incorrectly includes the €15,000 payment-error residual. The event’s cumulative net loss is below the €20,000 loss-component dataset threshold; its €60,000 gross loss does not make it eligible for that calculation.

---

Question 14

Topic: Credit Risk Measurement and Management

A dealer holds an uncollateralized, USD-settled FX forward with a regional bank. The dealer’s positive exposure increases when the local currency depreciates against the USD. The counterparty has predominantly local-currency assets and substantial USD debt.

Risk-neutral pricing assumptions:

• During nondefault periods, local-currency depreciation is associated with widening counterparty CDS spreads.
• A specified joint stress event produces counterparty default and a discrete further currency depreciation. Positive replacement value rises from $4 million immediately before the event to $7 million immediately afterward.
• Contractual close-out uses the post-event replacement value. Recovery is fixed at 40%, and the dealer’s own default risk is ignored.

All candidate models must retain the same CDS-implied marginal survival curve. Which dependence treatment is most defensible for unilateral CVA?

• A. Use a parametric model that includes the event increase in unconditional expected exposure, then combine that profile with independently simulated default times.
• B. Use an FX-linked hazard-rate model with continuous FX paths, calibrating its intensity sensitivity to represent both predefault co-movement and event-time exposure.
• C. Use a correlated structural model with continuous asset-value and FX paths, calibrating its default barrier to represent the event-time exposure increase.
• D. Use an FX-linked hazard-rate model with a default-associated FX jump, measuring close-out exposure after the jump on paths where default occurs.

Best answer: D

Explanation: Wrong-way risk arises when exposure increases as counterparty credit quality deteriorates. Here, currency depreciation raises the dealer’s exposure while weakening a counterparty with USD debt and local-currency assets.

A market-linked hazard-rate model can connect default intensity to FX conditions. A correlated structural model can connect declining counterparty asset value to increasing exposure. Parametric models can also represent dependence when calibrated to exposure conditional on default; fitting unconditional exposure and default distributions separately does not establish that dependence.

The discrete default-associated depreciation introduces an additional mechanism. Continuous paths can select high-exposure states at default, but they do not generate the specified jump in replacement value. Combining FX-linked intensity with a default-associated jump captures both mechanisms while retaining the calibrated survival curve. The jump adds $3 million to default-time exposure and $1.8 million to conditional default loss before discounting. CVA must weight this exposure on default paths rather than assume independence.

• A. An unconditional exposure adjustment combined with independent default times averages the event increase across paths instead of attaching it specifically to default paths.
• B. FX-dependent intensity creates dependence through the predefault FX state, but continuous FX paths cannot generate the stipulated discrete increase in close-out value at default.
• C. Correlated asset and FX diffusions can generate wrong-way risk, but crossing a default barrier does not itself produce a discontinuous move in FX exposure.
• D. Linking intensity to FX captures predefault dependence, while the default-associated jump captures the additional replacement value required by the contractual close-out terms.

---

Question 15

Topic: Risk Management and Investment Management

An investment committee will select one equity fund for its entire risky allocation and hold any remaining capital in risk-free bills. It seeks to maximize expected annual portfolio return subject to annual portfolio volatility of no more than 12%. Investor-level borrowing and short selling are prohibited.

Performance evidence: All estimates use the same 60 monthly observations. Returns are annualized arithmetic means, net of fees and financing costs, and already reflect each fund’s leverage.

The equity benchmark’s annual return is 8%, and risk-free bills yield 2%. Treat the historical estimates as forecasts. Bills have zero volatility, and reallocating between a fund and bills incurs no costs.

Which recommendation best meets the committee’s objective?

• A. Select Alder based on its higher Sharpe ratio.
• B. Select Birch based on its higher Jensen’s alpha.
• C. Select Birch based on its higher information ratio.
• D. Select Birch based on its higher Treynor ratio.

Best answer: A

Explanation: For an allocation combined only with risk-free bills, maximizing expected return at a fixed total-volatility limit requires comparing Sharpe ratios:

\[ S = \frac{E[R]-R_f}{\sigma}. \]

Alder’s Sharpe ratio is \( (10\%-2\%)/12\% = 0.667 \), while Birch’s is \( (14\%-2\%)/20\% = 0.600 \). Mixing a fund with risk-free bills scales its expected excess return and volatility proportionately, leaving its Sharpe ratio unchanged.

Alder can receive a 100% allocation, delivering an expected return of 10% at 12% volatility. Birch requires a 60% allocation, with 40% in bills, to meet the same limit. Its expected portfolio return is \( 0.60\times14\%+0.40\times2\%=9.2\% \).

Alder therefore provides the higher expected return within the risk budget. Existing fund leverage is already reflected in the reported returns and volatility; it should not be adjusted for again.

• A. Alder’s Sharpe ratio of 0.667 exceeds Birch’s 0.600, producing the higher expected return when both allocations satisfy the total-volatility limit.
• B. Birch’s alpha is 7.2%, versus Alder’s 4.1%, but CAPM-adjusted return does not establish which fund maximizes return at the total-volatility limit.
• C. Birch’s information ratio is approximately 0.369, versus Alder’s 0.229, but the committee constrains total volatility rather than benchmark tracking error.
• D. Birch’s Treynor ratio is 15.0%, versus Alder’s 12.31%, but beta does not measure the total volatility constrained by the committee.

---

Question 16

Topic: Credit Risk Measurement and Management

An international bank reviews a publicly listed Indonesian manufacturer using the original Altman Z-score model for publicly traded manufacturers. Its reference bands are distress for \(Z < 1.81\), gray for \(1.81 \le Z \le 2.99\), and safe for \(Z > 2.99\).

Financial data: All amounts are in USD millions.

An analyst proposes upgrading the borrower to an internal grade with a one-year probability of default (PD) of 0.50% if its score falls in the safe zone. The bank’s score-to-PD mapping was calibrated on US manufacturers and has not been validated for Indonesian borrowers.

Which assessment of the score and proposed upgrade is most appropriate?

• A. Z = 2.50, in the gray zone; the proposed 0.50% PD assignment requires validation for Indonesian borrowers.
• B. Z = 3.01, in the safe zone; the proposed 0.50% PD assignment requires validation for Indonesian borrowers.
• C. Z = 2.89, in the gray zone; the proposed 0.50% PD assignment requires validation for Indonesian borrowers.
• D. Z = 3.01, in the safe zone; the proposed 0.50% PD assignment is supported by the existing US mapping.

Best answer: B

Explanation: The original Altman model uses working capital/assets, retained earnings/assets, EBIT/assets, market value of equity/total liabilities, and sales/assets. These ratios are respectively 0.15, 0.20, 0.10, 1.20, and 1.50.

\[ Z = 1.2(0.15) + 1.4(0.20) + 3.3(0.10) + 0.6(1.20) + 1.50 = 3.01. \]

The result exceeds 2.99 and therefore falls in the stated safe zone. Market equity, rather than book equity, is required for this model variant.

A Z-score is a discriminant financial-distress screen, not itself a one-year PD estimate. Its reference bands do not guarantee survival or establish a particular default probability. Differences in accounting practices, bankruptcy regimes, and economic conditions can affect performance across jurisdictions. The bank therefore needs evidence that its US-calibrated score-to-PD mapping applies to Indonesian borrowers before using this result alone to support the proposed rating upgrade.

• A. The 2.50 result uses the private-manufacturer Z-prime coefficients and book equity rather than the prescribed original model for publicly traded manufacturers.
• B. The original model produces 3.01, above the safe-zone cutoff, but the US-calibrated mapping does not establish a 0.50% PD for Indonesian borrowers.
• C. The 2.89 result substitutes book equity for market equity; the original model requires market value of equity in the equity-to-liabilities ratio.
• D. The score is calculated correctly, but a safe-zone classification does not establish that the US score-to-PD relationship transfers to Indonesian borrowers.

---

Question 17

Topic: Credit Risk Measurement and Management

A credit risk manager is validating a simplified 12-month exposure report used to allocate counterparty limits. A bank holds an uncollateralized swap and a purchased cash-settled option with the same counterparty in one legally enforceable close-out netting set.

Simulation: Five equally likely paths produce the following trade values from the bank’s perspective, in $ millions.

Month-4 values are measured immediately before the option’s final settlement. The option settles in full immediately afterward and has zero value at month 12. The swap remains outstanding.

Reporting conventions:

• For time-weighted expected positive exposure (EPE), month-4 expected exposure (EE) represents the first four months, and month-12 EE represents the next eight months.
• At each date, 80% potential future exposure (PFE) is the fourth-smallest of the five path exposures, including zeros, with no interpolation.
• Peak PFE is the higher of the two dated PFEs.

Which EPE and peak PFE should the manager approve?

• A. EPE: $3.73 million; peak PFE: $12 million.
• B. EPE: $3.73 million; peak PFE: $9 million.
• C. EPE: $4.00 million; peak PFE: $9 million.
• D. EPE: $4.30 million; peak PFE: $9 million.

Best answer: B

Explanation: Counterparty exposure is the positive part of the net trade value in each path. Netting must occur before negative portfolio values are replaced by zero.

At month 4, net values are -8, 3, 6, 9, and 12 million. Exposures are therefore 0, 3, 6, 9, and 12 million, giving EE of $6 million. After the option settles, month-12 exposures come only from the swap: 0, 0, 0, 4, and 9 million, giving EE of $2.6 million.

The prescribed time weighting gives \( \mathrm{EPE} = (4 \times 6 + 8 \times 2.6)/12 = 3.7333 \) million. The fourth-smallest exposures are $9 million at month 4 and $4 million at month 12, so peak PFE is $9 million.

The signed expected mark-to-market values, $4.4 million and -$1.2 million, are not EE. Negative values can offset positive trade values within a netting set, but negative exposure in one simulation path cannot offset positive exposure in another.

• A. $12 million is the maximum month-4 exposure, not its specified 80th percentile; the fourth-smallest exposure is $9 million.
• B. Pathwise netting gives dated EEs of $6 million and $2.6 million; time weighting gives $3.73 million, and the larger dated PFE is $9 million.
• C. Taking each trade’s positive value separately gives month-4 EE of $6.8 million and EPE of $4 million, incorrectly ignoring enforceable netting.
• D. A simple average of the two dated EEs gives $4.30 million, but the specified EPE calculation assigns four and eight months to those EEs.

---

Question 18

Topic: Operational Risk and Resilience

A bank is reviewing the design of a critical end-of-day payments batch. For a specified stressed processing window, the fault tree below includes every cause of missing the settlement cutoff. The risk committee approves the design only if the probability of missing the cutoff is no greater than 4.50%.

Probability estimates:

Dependence assumptions: All probabilities refer to the same stressed window. Controller unavailability is independent of the joint availability state of the primary and backup services.

Which estimated top-event probability and approval conclusion are consistent with these facts? Round probability estimates to two decimal places.

• A. The estimated probability is 2.78%; the design meets the approval criterion.
• B. The estimated probability is 4.94%; the design fails the approval criterion.
• C. The estimated probability is 5.00%; the design fails the approval criterion.
• D. The estimated probability is 4.35%; the design meets the approval criterion.

Best answer: B

Explanation: An AND gate requires the probability that both input events occur. The primary and backup failures are dependent: the backup’s conditional failure probability is 30.00%, compared with its marginal probability of 8.00%. Therefore,

\[ P(A \cap B) = P(A)P(B \mid A) = 0.10 \times 0.30 = 0.03. \]

The top OR gate combines controller failure with simultaneous service failure. These events can occur together, so their overlap must be subtracted. Independence of the controller from the joint service state gives:

\[ P(\text{cutoff missed}) = 0.02 + 0.03 - (0.02 \times 0.03) = 0.0494. \]

The estimated probability is 4.94%, above the committee’s 4.50% maximum, so the design fails the approval criterion. Independence permits multiplication when finding the overlap; it does not make OR inputs mutually exclusive.

• A. Multiplying the marginal service-failure probabilities assumes independence, contradicting the stated 30.00% conditional backup-failure probability.
• B. Using the 3.00% joint service-failure probability and subtracting its overlap with controller failure gives 4.94%, exceeding the 4.50% limit.
• C. The OR inputs overlap with probability 0.06%, so adding 2.00% and 3.00% overstates the top-event probability.
• D. Multiplying 8.00% by 30.00% reverses the conditioning; the stated conditional probability must be weighted by the 10.00% primary-failure probability.

---

Question 19

Topic: Credit Risk Measurement and Management

A bank’s credit committee is reviewing a proposal to raise its commercial property lending limit. Management needs an executive summary that reconciles reported impairment with portfolio risk and assesses one-year capital capacity using relevant recovery evidence.

All amounts are in USD millions.

Supporting evidence:

• Downgrade rates cover 12-month cohorts with comparable initial rating distributions.
• Portfolio exits, including repayments and sales of defaulted loans, removed 20 from the allowance. All other allowance movements were remeasurements of loans retained through both reporting dates. The allowance includes both 12-month and lifetime expected credit losses.
• Current portfolio-wide economic capital for one-year unexpected credit losses is 90 using a 65% property recovery rate. A sensitivity using 50% recoveries increases economic capital to 120.
• Comparable completed property workouts produced 50% recoveries, net of costs and discounted on the same basis as the model assumptions.
• Available capital for unexpected credit losses is 105 and already reflects booked impairment expense.

Which executive summary best supports management’s assessment of the proposed limit increase?

• A. Report stable sector concentration and worsening migration, a $10m allowance increase on retained loans, and a $15m deficit under the recovery-sensitive capital estimate.
• B. Report rising sector concentration and worsening migration, a $10m allowance decrease on retained loans, and a $15m deficit under the recovery-sensitive capital estimate.
• C. Report rising sector concentration and worsening migration, a $10m allowance increase on retained loans, and a $15m surplus after counting the allowance as additional capital.
• D. Report rising sector concentration and worsening migration, a $10m allowance increase on retained loans, and a $15m deficit under the recovery-sensitive capital estimate.

Best answer: D

Explanation: Management credit reporting should connect concentration, migration, impairment movements, recovery assumptions and capital capacity. Property exposure remains $2,000m, but its portfolio share increases from 20% to 25%. The comparable downgrade rate also rises from 6% to 11%.

Adjusting the allowance for portfolio exits reveals a retained-loan increase of $10m: the ending $30m exceeds the exit-adjusted opening balance of $20m. The aggregate decline therefore does not indicate improving retained-book credit quality.

The accounting allowance includes different expected-loss horizons, while economic capital measures one-year unexpected losses. These amounts should not be combined as interchangeable capital resources. Baseline economic capital of $90m leaves $15m of headroom, but the recovery sensitivity requires $120m against $105m available, creating a $15m deficit. Management should see both results before considering more concentrated lending. Comparable completed workouts support examining the 50% recovery assumption, although their experience does not guarantee future recoveries.

• A. Unchanged property exposure does not imply stable concentration: its portfolio share rises from 20% to 25% as total credit exposure contracts.
• B. The $10m aggregate allowance decline includes $20m removed through portfolio exits, implying a $10m increase, not a decrease, on retained loans.
• C. The allowance is not additional capital for unexpected losses; adding it to the defined capital capacity incorrectly converts the sensitivity’s $15m deficit into a surplus.
• D. Property concentration and downgrades increase, retained-loan allowances rise after adjusting for exits, and the workout-supported capital sensitivity exceeds available capital by $15m.

---

Question 20

Topic: Credit Risk Measurement and Management

A bank is considering replacing two borrowers from different industries with two borrowers from the same industry. Both the current and proposed credit books contain two $100 million loans, measured on a one-year mark-to-market loss basis.

Single-name inputs: Every borrower has a 2% one-year default probability, an 8% probability of a nondefault downgrade, and 40% loss given default. A nondefault downgrade creates a $5 million loss; otherwise, a surviving loan has no loss.

Portfolio model: The table fully specifies each book’s aggregate one-year loss distribution.

The bank defines economic capital as 99% one-year portfolio VaR minus expected loss. The committee may approve the reallocation only if the proposed book’s economic capital does not exceed $50 million.

Which assessment correctly identifies the capital impact and the approval decision?

• A. Approve the reallocation; economic capital remains $37.6 million in both books.
• B. Reject the reallocation; economic capital rises from $40 million to $80 million.
• C. Reject the reallocation; economic capital remains $77.6 million in both books.
• D. Reject the reallocation; economic capital rises from $37.6 million to $77.6 million.

Best answer: D

Explanation: Single-name analysis describes each borrower’s marginal loss risk. Portfolio measurement additionally requires dependence information to determine how defaults and migrations combine in the loss distribution.

Each loan’s expected loss is \(0.02 \times 40 + 0.08 \times 5 = 1.2\) million, so both books have expected loss of $2.4 million. Dependence changes the distribution, not this sum of expected losses.

For the current book, cumulative probability reaches 96.2% at $10 million and 99.8% at $40 million. Its 99% VaR is therefore $40 million. For the proposed book, cumulative probability at $40 million is only 98.8%, making its 99% VaR $80 million.

Subtracting expected loss produces economic capital of $37.6 million and $77.6 million, respectively. The proposed book breaches the $50 million limit. Greater joint-default probability increases tail risk despite unchanged individual credit inputs. Summing standalone capital cannot identify this change in diversification.

• A. Unchanged single-name default and migration inputs do not imply unchanged portfolio tail risk; the proposed book’s 99% VaR increases to $80 million.
• B. These figures are portfolio VaRs, not economic capital; the bank’s definition requires subtracting $2.4 million of expected loss from each.
• C. Summing standalone capital gives $77.6 million for both books, missing the diversification reflected in the current book’s joint loss distribution.
• D. Portfolio VaR rises from $40 million to $80 million; subtracting $2.4 million of expected loss gives capital requirements that breach the limit after reallocation.

---

Question 21

Topic: Market Risk Measurement and Management

An oil distributor holds 40,000 barrels of a local crude grade and will sell them at tomorrow’s spot price. It can hedge by selling benchmark crude futures today and closing the position tomorrow. Each contract covers 1,000 barrels. Spot price \(S_t\) and futures price \(F_t\) are quoted in dollars per barrel.

Historical estimates: With \(\Delta\) denoting a one-day price change, contemporaneous OLS regressions using the same sample and including intercepts give:

• Price levels: \(\widehat{S}_t = a_L + 1.50F_t\), with \(R^2 = 0.96\).
• Daily changes: \(\widehat{\Delta S}_t = a_C + 1.20\Delta F_t\), with \(R^2 = 0.64\).
• Reverse daily changes: \(\widehat{\Delta F}_t = a_R + 0.533333\Delta S_t\), with \(R^2 = 0.64\); the slope is rounded.

Daily price-change standard deviations are $3 per barrel for spot and $2 per barrel for futures. Use the historical price-change variances and covariance for tomorrow’s forecast. Ignore transaction costs, margin funding, and discounting.

Which hedge minimizes forecast next-day P&L variance, with the corresponding standard deviation correctly stated?

• A. Sell 60 futures contracts and forecast a P&L standard deviation of $24,000.
• B. Sell 32 futures contracts and forecast a P&L standard deviation of $72,000.
• C. Sell 75 futures contracts and forecast a P&L standard deviation of $90,000.
• D. Sell 48 futures contracts and forecast a P&L standard deviation of $72,000.

Best answer: D

Explanation: Minimum-variance hedging uses the price changes that generate P&L. For inventory quantity \(Q\) and hedge ratio \(h\), the hedged P&L is \(Q(\Delta S-h\Delta F)\). The appropriate ratio is the slope from regressing spot changes on futures changes:

\[ h^* = \frac{\mathrm{Cov}(\Delta S,\Delta F)}{\mathrm{Var}(\Delta F)} = 1.20. \]

The distributor therefore sells \(40,000\times1.20/1,000=48\) contracts. With an intercept in this single-regressor OLS model, the change-regression R-squared equals squared correlation. The forecast hedged P&L standard deviation is:

\[ \sigma_{\mathrm{P\&L}} = 40,000\times3\sqrt{1-0.64}=72,000\text{ dollars}. \]

Reversing a regression changes its dependent variable and the residual being minimized. Forward and reverse OLS slopes multiply to squared correlation, here 0.64, rather than 1. Consequently, the reciprocal reverse slope is \(1.20/0.64=1.875\), not the optimal hedge ratio. The higher R-squared for price levels concerns a different risk forecast and does not establish a better hedge for daily P&L changes.

• A. The level-regression slope and R-squared describe prices, not daily changes; applying them to the daily P&L forecast does not produce the minimum-variance hedge.
• B. Using the correlation of 0.80 as the hedge ratio ignores the spot-to-futures volatility ratio; $72,000 applies to the 1.20 hedge ratio.
• C. Inverting the reverse slope gives a hedge ratio of 1.875 and correctly implies $90,000 volatility, but this hedge does not minimize variance.
• D. The change-regression slope gives 48 contracts, and its R-squared of 0.64 implies residual P&L volatility of $72,000.

---

Question 22

Topic: Market Risk Measurement and Management

A bank defines a one-day trading loss of at least $100 million as the failure threshold for its trading portfolio. Historical crisis replays produce losses of at most $80 million. The risk committee wants to identify plausible joint market moves that could reach the failure threshold.

Portfolio stress mapping:

• Each 1-bp credit spread widening produces a $0.5 million loss.
• Each 1% equity index decline produces a $2 million loss.
• Each 1% EUR depreciation against USD produces a $2 million net FX gain.

Assume these sensitivities remain valid over the proposed shocks, their P&L effects are additive, and every proposed joint shock is economically plausible.

Which joint shock qualifies as a failure-inducing reverse stress scenario?

• A. Credit spreads widen by 150 bp, the equity index falls by 15%, and EUR depreciates against USD by 5%.
• B. Credit spreads widen by 120 bp, the equity index falls by 25%, and EUR depreciates against USD by 5%.
• C. Credit spreads widen by 100 bp, the equity index falls by 20%, and EUR depreciates against USD by 5%.
• D. Credit spreads widen by 140 bp, the equity index falls by 20%, and EUR depreciates against USD by 10%.

Best answer: B

Explanation: Reverse stress testing starts with a specified failure outcome and works backward to identify plausible conditions that produce it. Historical replay instead starts with observed market moves and calculates their consequences. A historical maximum below the failure threshold does not establish that the portfolio cannot fail.

Here, a 120-bp spread widening produces a $60 million loss, a 25% equity decline produces a $50 million loss, and 5% EUR depreciation generates a $10 million gain. The resulting $100 million net loss reaches the threshold exactly. Because the joint shock is economically plausible, it identifies a vulnerability without requiring historical precedent. Retaining this scenario does not forecast its occurrence or establish its probability; likelihood assessment is a separate exercise.

• A. The $105 million combined credit and equity loss is reduced by a $10 million FX gain, leaving $95 million, below the failure threshold.
• B. The net loss is $100 million: $60 million in credit losses plus $50 million in equity losses minus a $10 million FX gain.
• C. The net loss is $80 million: $50 million in credit losses plus $40 million in equity losses minus a $10 million FX gain.
• D. The $110 million combined credit and equity loss is reduced by a $20 million FX gain, leaving $90 million, below the failure threshold.

---

Question 23

Topic: Operational Risk and Resilience

A bank is assessing whether its daily automated reconciliation controls the risk that approved FX trades are omitted from the booking ledger, causing understated exposures or missed settlements.

Control and test evidence:

• An interface transfers approved trading-platform trades into a staging table. Both the ledger import and the reconciliation’s expected-trades file use this table. The reconciliation flags expected trade IDs missing from the ledger.
• A reviewer sampled 80 IDs from the expected-trades file and found matching ledger entries. In a test environment, deleting six ledger entries produced six reconciliation alerts.
• Four trades approved before the applicable export cutoff were subsequently found missing from staging and the ledger because the interface had rejected them. The reconciliation produced no exceptions on those incident dates.

Which testing change would most directly address the gap between the stated risk and the assurance provided by the existing test?

• A. Repeat the ledger-deletion tests at peak processing volume using incident-date staging records and check whether each omission triggers an alert.
• B. Reconcile every expected-trades-file ID to ledger records on the incident dates and investigate any unmatched IDs.
• C. Reconcile approved trading-platform IDs to staging and ledger records on the incident dates and investigate any unmatched IDs.
• D. Review all reconciliation alerts on the incident dates and compare their resolution times with the relevant settlement deadlines.

Best answer: C

Explanation: A reconciliation cannot detect trades absent from both sides of its comparison. Here, an interface rejection excludes a trade from staging, which supplies both the expected-trades file and the ledger import. Agreement between those populations therefore does not establish that all approved trades were booked.

The sample supports matching accuracy for the records tested. The deletion tests support detection of ledger omissions when the corresponding trades remain in the expected population. Neither provides evidence about trading-platform-to-staging completeness.

Using approved trading-platform trades as the starting population and tracing their IDs through staging and the ledger directly assesses the missing coverage. The problem is the population boundary, not merely the number of records tested. Successful downstream tests should not be extrapolated to end-to-end capture effectiveness.

• A. Peak-volume testing checks processing capability, but staged records still exclude the upstream-rejected trades responsible for the observed failures.
• B. Increasing coverage within the expected-trades file cannot reveal trades missing from that file because it is generated from the same incomplete staging input.
• C. Starting from approved trades before the interface tests upstream completeness and can reveal omissions absent from both populations used by the automated reconciliation.
• D. The incidents produced no alerts, so reviewing alert closure cannot assess the failures that prevented trades from entering the reconciliation population.

---

Question 24

Topic: Market Risk Measurement and Management

A market risk manager must decide whether to execute one of two fixed-size hedge tickets before the 15:00 trading cutoff. The desk’s stress policy requires the upper end of its estimated stress-loss band to be no more than $4.50 million.

Current report:

• The 99% one-day VaR is $3.20 million against a $4.00 million VaR limit. Its correlation calibration predates recent market repricing and has not been validated for those conditions; it remains approved for routine monitoring only.
• Position-based factor sensitivities have been independently checked. The CS01 range reflects uncertainty from current spread quotes.
• Positive sensitivities indicate losses from a 1 bp increase in the relevant factor.

Stress and hedge assumptions:

• The prescribed scenario combines a 25 bp interest-rate increase with a 20 bp credit-spread widening. Add the first-order losses and ignore higher-order effects.
• The rate hedge reduces net DV01 to $10,000 per bp and leaves CS01 unchanged.
• The spread hedge reduces both CS01 bounds by $25,000 per bp and leaves DV01 unchanged.
• Both tickets are executable before the cutoff. At most one may be executed; ignore transaction costs.

Which recommendation correctly uses the report for this decision?

• A. Retain current positions; the central stress estimate of $4.50 million permits continued exposure.
• B. Retain current positions; the reported VaR of $3.20 million permits continued exposure.
• C. Execute the spread hedge; the upper-bound stressed loss falls to $4.40 million.
• D. Execute the rate hedge; the upper-bound stressed loss falls to $4.25 million.

Best answer: C

Explanation: Clear market risk reporting connects factor exposures, uncertainty, and model-use qualifications to the decision that must be made. Here, checked sensitivities support the prescribed stress calculation even though the VaR correlation calibration remains qualified.

The rate shock contributes $500,000: 25 bp at $20,000 per bp. The spread shock contributes $3.60 million-$4.40 million: 20 bp at the CS01 bounds. Total stressed loss therefore ranges from $4.10 million to $4.90 million. Its upper end exceeds the $4.50 million limit.

The spread hedge lowers CS01 to $155,000-$195,000 per bp, reducing stressed losses by $500,000. The resulting loss band is $3.60 million-$4.40 million, which satisfies policy. Neither the central estimate nor the monitoring VaR substitutes for the required upper-bound assessment. Reporting enables the timely decision; it does not itself reduce spread exposure. That reduction occurs when the hedge is executed.

• A. The central estimate equals the limit, but policy applies to the upper-bound loss of $4.90 million, so unchanged positions fail the requirement.
• B. Compliance with the VaR limit does not establish stress-limit compliance, and the qualified VaR estimate is approved only for routine monitoring.
• C. The hedge reduces upper-bound CS01 to $195,000 per bp, giving $500,000 of rate loss plus $3.90 million of spread loss.
• D. The rate hedge produces an upper-bound loss of $4.65 million; $4.25 million uses the midpoint CS01 rather than its upper bound.

---

Question 25

Topic: Liquidity and Treasury Risk Measurement and Management

A bank reports a 30-day liquidity coverage ratio (LCR) of 135%, against a regulatory minimum of 100%. Treasury wants to determine the smallest additional opening cash buffer needed to meet every payment in a specified five-business-day stress scenario.

Available resources:

• Opening cash: $45 million.
• Unencumbered sovereign bonds: market value of $60 million, monetizable with a 10% haircut. Proceeds settle before day-1 payments.
• Unencumbered corporate bonds: market value of $40 million, monetizable with a 25% haircut. Proceeds settle before day-5 payments; earlier monetization is unavailable.
• Unspent cash carries forward. No other funding sources are available during the scenario.

Stressed cash requirements: These are cumulative net outflows after contractual inflows, excluding opening cash and asset-monetization proceeds. There are no payments or receipts on days 2 and 4.

Which monitoring assessment best supports treasury’s buffer decision?

• A. Use the day-5 stressed cash-flow headroom; no additional cash buffer is required.
• B. Use the 30-day LCR headroom above its minimum; no additional cash buffer is required.
• C. Use day-1 liquidity generation capacity against peak cumulative outflows; a $21 million additional cash buffer is required.
• D. Use the minimum payment-date stressed cash-flow headroom; a $6 million additional cash buffer is required.

Best answer: D

Explanation: Funding-liquidity monitoring must align usable resources with the timing of cash requirements. Asset market values are not equivalent to cash available for payments: haircuts reduce proceeds, and settlement delays determine when those proceeds can be used.

The sovereign bonds generate $54 million, providing $99 million of cumulative resources on days 1 and 3. Corporate-bond proceeds add $30 million before day-5 payments, increasing cumulative resources to $129 million. Subtracting cumulative net outflows gives headroom of $19 million on day 1, a $6 million deficit on day 3, and $9 million on day 5. An additional opening buffer of $6 million therefore covers every payment date.

A positive terminal balance can conceal an earlier funding deficit. Similarly, LCR compliance measures regulatory coverage over a different horizon rather than guaranteeing payment-date liquidity in this scenario.

• A. The $9 million terminal surplus does not resolve the day-3 shortfall, which occurs before corporate-bond proceeds become available.
• B. Meeting the 30-day regulatory ratio does not establish that usable cash is available at every payment date in this five-day scenario.
• C. Comparing $120 million with $99 million ignores $30 million available before day-5 payments and therefore overstates the smallest required buffer.
• D. The bank has $99 million available by day 3 against $105 million of cumulative net outflows, producing the largest payment-date shortfall.

Questions 26-50

Question 26

Topic: Risk Management and Investment Management

A fund’s stress committee will retain the one-day return path with the lower conditional density for dependence-risk analysis. Conditional on the proposed regime, returns of equity indexes X and Y are bivariate normal with zero means.

For a volatility or correlation parameter \(p\), its historical score is \(z_p=(p-\bar p)/s_p\), where \(\bar p\) and \(s_p\) are the historical mean and standard deviation of estimates of that parameter.

The scenario z-score is the Mahalanobis distance \(D\). Standardized returns are \(z_i=r_i/\sigma_i\), using each index’s proposed daily volatility. With proposed correlation \(\rho\),

\[ D=\sqrt{\frac{z_X^2+z_Y^2-2\rho z_Xz_Y}{1-\rho^2}} \]

Supplied scenario calculations:

Which recommendation correctly applies the selection criterion and interprets the historical parameter scores?

• A. Retain co-movement; the negative correlation z-score makes same-direction moves less plausible, while the positive volatility z-score indicates elevated volatility.
• B. Retain either path; their equal sums of absolute return z-scores imply equal conditional density, while the regime has elevated volatility and below-average but positive correlation.
• C. Retain divergence; its larger covariance-adjusted distance implies lower conditional density, while the regime has elevated volatility and below-average but positive correlation.
• D. Retain divergence; its distance represents a standard normal z-score giving an unconditional tail probability, while the regime has elevated volatility and below-average but positive correlation.

Best answer: C

Explanation: For a multivariate normal distribution with a fixed covariance matrix, conditional density decreases as squared Mahalanobis distance increases. Divergence has \(D^2=10\), compared with approximately \(6.67\) for co-movement, so divergence meets the committee’s criterion. Although both paths have individual standardized moves of magnitude 2, positive correlation makes opposite-direction moves more unusual than same-direction moves.

The parameter scores describe a separate feature: the proposed regime. A volatility score of +2 means volatility is two historical standard deviations above its historical mean. A correlation score of -4 means correlation is four historical standard deviations below its mean, not that correlation is negative. Its proposed value remains +0.20.

The scenario score is a nonnegative multivariate distance. Under the stated bivariate normal assumption, \(D^2\) follows a chi-squared distribution with two degrees of freedom. This supports conditional radial-tail calculations, not a univariate normal-tail interpretation or an unconditional probability for the return path and regime.

• A. The correlation z-score measures deviation from its historical mean, not the sign of dependence; the proposed positive correlation favors same-direction moves.
• B. Both absolute standardized-return sums equal 4.0, but that calculation ignores dependence; positive correlation produces different joint densities for co-movement and divergence.
• C. Divergence has the larger Mahalanobis distance and therefore lower conditional normal density; correlation remains positive at 0.20 despite its historical score of -4.0.
• D. Mahalanobis distance is not a univariate standard normal score, and a calculation conditional on the proposed regime does not establish an unconditional tail probability.

---

Question 27

Topic: Liquidity and Treasury Risk Measurement and Management

A bank’s treasurer evaluates a one-day liquidity stress at 09:00. Unavoidable payments must settle by 16:00, and the bank must retain its operating reserve throughout the day.

Resources:

• Cash: $35 million, including a mandatory $10 million operating reserve.
• Government bonds: $80 million in market value, including $30 million already pledged and unavailable for reuse today. The remaining bonds can support an overnight repo with a 10% haircut; proceeds settle at 14:00.
• Committed credit facility: $20 million undrawn. The facility remains accessible under the stress assumptions, all covenants are satisfied, and a draw settles at 14:00.
• Customer escrow: $15 million legally segregated and unavailable for the bank’s own obligations.
• Pending sale of other assets: $20 million of proceeds settle at 10:00 on the next business day.

Payments:

• Unavoidable withdrawals and margin payments total $95 million.
• A separate $12 million strategic investment payment can be deferred for one week without penalty.

The resource balances are separate except for the inclusions explicitly stated. Assume repo and facility proceeds settle as specified, with no other cash flows before 16:00.

After deferring the strategic payment, which assessment of usable liquidity and remaining funding needs by 16:00 is correct?

• A. Usable liquidity is $90 million, leaving a $5 million shortfall against required payments.
• B. Usable liquidity is $70 million, leaving a $25 million shortfall against required payments.
• C. Usable liquidity is $105 million, leaving a $10 million surplus against required payments.
• D. Usable liquidity is $110 million, leaving a $15 million surplus against required payments.

Best answer: A

Explanation: Usable liquidity depends on access and timing, not merely the balance-sheet value of resources. Operational reserves must be retained, pledged assets cannot be reused, and repo capacity reflects collateral haircuts. Contingent funding can contribute when the stress assumptions establish that it remains accessible before the payment deadline.

In millions of dollars, usable capacity is:

\[ (35 - 10) + (80 - 30) \times (1 - 0.10) + 20 = 90 \]

The escrow balance is restricted, and the asset-sale proceeds arrive too late. Deferring the strategic investment removes an optional payment but does not create additional cash. The bank therefore faces a $5 million same-day funding-liquidity shortfall against $95 million of unavoidable payments. This shortfall does not, by itself, establish insolvency.

• A. Retaining the operating reserve, applying the repo haircut, and including the accessible facility yields $90 million against $95 million of unavoidable payments.
• B. Excluding the committed facility understates capacity by $20 million because it remains accessible under stress and delivers cash before the deadline.
• C. Customer escrow cannot fund the bank’s obligations, so counting its $15 million overstates usable liquidity by $15 million.
• D. Asset-sale proceeds arrive on the next business day, so including them overstates the liquidity available for payments due by 16:00 today.

---

Question 28

Topic: Current Issues in Financial Markets

A bank applies the mechanisms discussed in the IMF’s 2025 Tokenization and Financial Market Inefficiencies when reviewing a proposal to reduce a bond portfolio’s stressed liquidation haircut following a settlement pilot.

Settlement arrangements:

• The bank purchases bonds for $12 million at 09:00 and sells other, already-owned bonds for $10 million at 09:05 on the same day.
• The conventional arrangement legally nets these payments and requires $2 million of opening cash, with no interim gross prefunding.
• The tokenized arrangement settles each transaction gross through atomic delivery-versus-payment at the stated times. Sale proceeds become available only at 09:05.
• No intraday borrowing or other cash inflows are available.

Pilot observations:

• Reconciliation time fell from 40 minutes to 4 minutes.
• Normal-market bid-ask spreads and order-book depth were unchanged. The pilot included no stressed-market observations.

Which assessment of the minimum opening cash under tokenized settlement and the proposed haircut reduction is best supported?

• A. Minimum opening cash is $2 million; the evidence supports reducing the stressed liquidation haircut.
• B. Minimum opening cash is $12 million; the evidence does not support reducing the stressed liquidation haircut.
• C. Minimum opening cash is $2 million; the evidence does not support reducing the stressed liquidation haircut.
• D. Minimum opening cash is $12 million; the evidence supports reducing the stressed liquidation haircut.

Best answer: B

Explanation: Tokenization can reduce asset-lifecycle frictions without necessarily reducing funding or market liquidity risk. Atomic delivery-versus-payment coordinates the securities and cash legs of each trade; it does not automatically preserve netting across separate trades.

The bank must pay $12 million at 09:00, before receiving $10 million at 09:05. Minimum opening cash is therefore $12 million, compared with $2 million under the conventional netting arrangement. This increase follows from the specified gross-settlement sequence, not from tokenization universally.

Faster reconciliation demonstrates an operational efficiency gain. A stressed liquidation haircut instead concerns potential losses when selling assets under adverse market conditions. Unchanged normal-market spreads and depth, together with no stressed observations, provide no basis for reducing that haircut.

• A. This carries conventional netting into the gross workflow and treats faster reconciliation as evidence of lower stressed liquidation losses.
• B. The purchase must be funded before sale proceeds arrive, while faster reconciliation provides no evidence of reduced stressed liquidation losses.
• C. The haircut conclusion is appropriately limited, but the tokenized arrangement requires the full purchase payment before the offsetting sale receipt becomes available.
• D. The cash requirement reflects gross settlement, but unchanged normal-market trading conditions and absent stress observations do not justify a lower stressed liquidation haircut.

---

Question 29

Topic: Risk Management and Investment Management

An investment committee will approve an active equity mandate only if the fundamental law of active management implies a maximum annual information ratio of at least 0.75.

Research and trading evidence:

• The estimated information coefficient is 0.05, assumed stable going forward.
• The manager issues 12 one-month forecasts each month. Bets based on different forecasts are independent within and across months.
• Each forecast is implemented through four perfectly correlated active bets, resulting in 576 implemented bets per year.
• The manager’s three-year record has a realized annualized information ratio of 0.80.

Assume optimal risk-weighted portfolio construction, no binding investment constraints, and no transaction costs.

Which committee conclusion follows from the fundamental law?

• A. Reject the mandate; the model-implied maximum annual information ratio is approximately 0.17.
• B. Approve the mandate; the model-implied maximum annual information ratio is 1.20.
• C. Approve the mandate; the model-implied maximum annual information ratio is 0.80.
• D. Reject the mandate; the model-implied maximum annual information ratio is 0.60.

Best answer: D

Explanation: The fundamental law gives \(\mathrm{IR}_{\max}=\mathrm{IC}\sqrt{\mathrm{BR}}\), where IC measures forecast skill and BR is the number of independent active bets over the information ratio’s measurement horizon. Four perfectly correlated implementations of one forecast contribute one independent bet, not four.

Annual breadth is \(12\times12=144\), so the maximum annual information ratio is \(0.05\sqrt{144}=0.60\). The mandate therefore fails the 0.75 approval hurdle.

This theoretical ceiling on expected risk-adjusted performance depends on stable forecasting skill, independent bets, optimal portfolio construction, and the stated absence of implementation frictions. It does not guarantee a realized annual result. The historical information ratio of 0.80 can reflect sampling variation and does not replace the forecast based on skill and breadth. Greater dependence, weaker skill, binding constraints, or transaction costs can reduce attainable performance.

• A. Using \(0.05\sqrt{12}\approx0.17\) includes only one month’s breadth, omitting the additional independent forecasts during the year.
• B. Using \(0.05\sqrt{576}=1.20\) counts correlated implementations as independent bets, overstating annual breadth.
• C. The historical 0.80 is a realized sample statistic, not the maximum expected information ratio implied by forecast skill and independent breadth.
• D. Independent annual breadth is \(12\times12=144\), yielding \(0.05\sqrt{144}=0.60\), below the approval requirement.

---

Question 30

Topic: Risk Management and Investment Management

A foundation’s portfolio manager proposes moving 12% of total assets from daily redeemable bond funds into a private-credit fund. Both investments fall within the mandate’s fixed-income category, and the proposal remains within approved asset-weight and reported volatility limits.

Mandate excerpt:

Managers may select investments within the approved strategy. Material changes to that strategy require investment committee approval.

Review findings:

• The committee’s approved portfolio design uses the bond allocation as a source of cash during market stress.
• The proposed fund permits annual redemptions subject to gates. Its low reported volatility is calculated from quarterly appraisal-based valuations.
• Independent risk analysis identifies a substantial reduction in liquidity available under stress.
• Operational due diligence finds satisfactory custody and reconciliation controls.

The manager considers the proposal a routine fund substitution because the asset-class designation is unchanged and operational due diligence is satisfactory. The foundation’s board receives regular portfolio risk reports.

Which governance response is most appropriate before capital is committed?

• A. Classify it as operational onboarding: the manager approves after due-diligence clearance, risk monitors compliance, the committee reviews it routinely, and the board receives updated risk reporting.
• B. Classify it as a tactical substitution: the manager approves and implements the allocation, risk monitors compliance, the committee reviews it routinely, and the board receives updated risk reporting.
• C. Classify it as a strategy change: the manager retains allocation responsibility, risk independently challenges the proposal, the committee decides before commitment, and the board receives updated risk reporting.
• D. Classify it as a risk exception: the manager retains allocation responsibility, risk authorizes the proposal, the committee reviews it routinely, and the board receives updated risk reporting.

Best answer: C

Explanation: Investment discretion operates within an approved strategy, not merely within asset-class weights or reported volatility limits. Replacing daily redeemable investments used for stressed liquidity with annual, gated redemptions materially changes the allocation’s function. Appraisal-based valuations can also smooth reported returns, so low measured volatility does not resolve the liquidity concern.

The portfolio manager remains responsible for the investment proposal, allocation decisions within authorized strategy, and subsequent implementation. Independent risk challenges assumptions and evaluates exposures; it does not take over business responsibility. The investment committee must decide whether to authorize this material strategic change before commitment.

Operational due diligence provides assurance about matters such as custody and reconciliation, but does not establish investment suitability or authorize a strategy change. Board reporting should communicate the changed liquidity and valuation profile. Reporting informs oversight; it does not replace the required approval.

• A. Satisfactory custody and reconciliation controls address operational integrity, not whether changing the portfolio’s strategic liquidity function can proceed under delegated investment authority.
• B. Remaining within asset-weight and reported volatility limits does not make the substitution tactical when gated redemptions materially change the allocation’s intended liquidity function.
• C. The loss of stressed liquidity changes the approved strategy, requiring committee approval while preserving the manager’s business responsibility and risk’s independent oversight role.
• D. Independent risk challenge cannot substitute for the committee’s required authorization of a material strategy change, even when the manager remains accountable for the allocation.

---

Question 31

Topic: Market Risk Measurement and Management

A credit trading desk’s USD-denominated corporate-bond portfolio has an approved one-day 99% VaR limit of $5 million but currently reports VaR of $5.825 million. Market Risk has been notified, and risk-increasing trades are frozen.

Approved limit policy:

A breach closes only when approved-model VaR is within the limit. Changes to the estimation window require independent validation and approval.

Approved risk model: The delta-normal model assumes zero mean and uncorrelated factor P&L components. The 99% normal quantile is 2.33.

An index CDS hedge can eliminate only the sector-credit component. A Treasury futures hedge can eliminate only the interest-rate component. Either hedge leaves the other components unchanged, and all proposed exposure adjustments are executable today.

The desk also proposes shortening the covariance estimation window from 250 days to 20 days, producing VaR of $4.8 million with unchanged positions. This change has not completed independent validation or approval.

Which remediation plan should Market Risk authorize to close the breach today?

• A. Retain existing positions, adopt the proposed 20-day calibration, and close the breach using the recalibrated model.
• B. Reduce every existing position, including hedges, by 20% and close the breach using the approved model.
• C. Neutralize the sector-credit factor with index CDS and close the breach using the approved model.
• D. Neutralize the interest-rate factor with Treasury futures and close the breach using the approved model.

Best answer: B

Explanation: Limit compliance must be assessed using the approved model. For zero-mean delta-normal VaR with uncorrelated factor P&Ls, variances add rather than standard deviations. With amounts expressed in millions of USD:

\[ \mathrm{VaR}_{0.99} = 2.33\sqrt{0.90^2 + 1.20^2 + 2.00^2} = 5.825. \]

Reducing every position, including existing hedges, by 20% multiplies each factor exposure by 0.8. Portfolio standard deviation and VaR therefore also fall by 20%, giving approved-model VaR of $4.66 million and restoring compliance.

Hedging a systematic factor leaves the substantial issuer-specific exposure unchanged. Neither individual hedge reduces total VaR sufficiently. Shortening the estimation window may change reported VaR, but it does not change positions or remove their risk. The proposed calibration also cannot replace the approved limit measure before independent validation and approval.

• A. The unapproved calibration cannot determine limit compliance, and its lower VaR estimate does not represent a reduction in the portfolio’s underlying exposure.
• B. A 20% proportional reduction scales every factor exposure and VaR by 0.8, reducing approved-model VaR to $4.66 million.
• C. Removing the sector-credit component leaves interest-rate and issuer-specific risks, producing VaR of approximately $5.11 million, still above the $5 million limit.
• D. Removing the interest-rate component leaves sector-credit and issuer-specific risks, producing VaR of approximately $5.43 million, still above the $5 million limit.

---

Question 32

Topic: Credit Risk Measurement and Management

A bank’s credit risk committee is revising the one-year loss model for a portfolio with $60 million of secured exposure and $40 million of unsecured exposure. The existing model retains state-dependent default frequencies but fixes each segment’s recovery at its probability-weighted average across economic states.

One-year economic scenarios:

• Expansion (90% probability): default frequency 1%; secured recovery 70%; unsecured recovery 40%.
• Slowdown (8% probability): default frequency 5%; secured recovery 50%; unsecured recovery 25%.
• Severe recession (2% probability): default frequency 15%; secured recovery 30%; unsecured recovery 10%.

Default frequencies apply equally to both segments. Treat the stated default frequencies and recoveries as exact outcomes within each state, with no additional loss uncertainty. Exposure at default remains constant, and discounting is ignored.

The committee will replace fixed recoveries with state-dependent recoveries. It defines economic capital as the 99th percentile of one-year portfolio loss minus expected loss. Which revised expected loss and economic capital should it approve, rounded to the nearest $1,000?

• A. Expected loss: $1,248,000; economic capital: $10,452,000.
• B. Expected loss: $852,000; economic capital: $7,136,000.
• C. Expected loss: $852,000; economic capital: $10,848,000.
• D. Expected loss: $707,000; economic capital: $5,917,000.

Best answer: C

Explanation: PD and LGD are positively dependent here: as default frequency rises, recovery falls. Losses must be computed jointly within each state before taking a probability-weighted average.

Using amounts in millions:

• Expansion: \(0.01[60(0.30)+40(0.60)]=0.420\).
• Slowdown: \(0.05[60(0.50)+40(0.75)]=3.000\).
• Severe recession: \(0.15[60(0.70)+40(0.90)]=11.700\).

Expected loss is \(0.90(0.420)+0.08(3.000)+0.02(11.700)=0.852\) million. The first two states account for 98% of outcomes, so the 99th-percentile loss is $11,700,000. Economic capital is therefore $11,700,000 minus $852,000, or $10,848,000.

Separately averaging default frequency and LGD understates expected loss when their covariance is positive. A default-weighted average recovery can preserve expected loss but still understate capital if it replaces the recession-specific recovery in the tail.

• A. Applying the severe-recession LGD of 78% in every state overstates expected loss and consequently understates capital measured as VaR minus expected loss.
• B. Expected loss is preserved, but this capital applies a default-weighted portfolio LGD of 53.25% to the tail instead of the severe-recession LGD of 78%.
• C. The probability-weighted loss is $852,000, and the 99th-percentile loss is $11,700,000, giving economic capital of $10,848,000.
• D. Fixing portfolio LGD at its state-weighted average of 44.16% produces these estimates but ignores higher loss severity when defaults cluster.

---

Question 33

Topic: Current Issues in Financial Markets

A bank is considering automatic suppression of credit-review alerts. An internal logistic model produces one-year PDs independently of a vendor-hosted GenAI service, which receives the PDs and interprets lending policy.

Numerical validation: Approval relied on pooled calibration. The same completed validation sample contains 10,000 borrowers in each segment. Testing with GenAI disconnected produces these results:

Workflow trace: During a test involving a borrower with PD of 4.5%, the GenAI input contains the following current approved rule as its only policy source:

Require credit review when the one-year PD is at least 4%.

The raw GenAI response states:

No credit review is needed because the policy exempts borrowers with PDs below 6%.

The dashboard displays this response unchanged. The vendor supplies a general-purpose performance certificate, while the bank’s validation has covered only the numerical model.

Which assessment best characterizes the risks the bank must address before enabling automatic alert suppression?

• A. A GenAI-induced PD-calibration failure and a GenAI implementation failure in policy interpretation, both subject to the bank’s model-risk governance.
• B. An ordinary PD-model validation weakness and a retrieval implementation failure involving stale policy, both subject to the bank’s model-risk governance.
• C. An ordinary PD-model validation weakness and a GenAI implementation failure in policy interpretation, both subject to the bank’s model-risk governance.
• D. An ordinary PD-model validation weakness and a dashboard implementation failure in output mapping, both subject to the bank’s model-risk governance.

Best answer: C

Explanation: Calibration must be assessed at a granularity relevant to the model’s use. Equal segment sizes make pooled predicted and observed default rates both 3%, concealing underprediction for corporate borrowers and overprediction for retail borrowers. This weakness exists with GenAI disconnected and is therefore an ordinary numerical-model validation issue.

A separate GenAI implementation risk arises when generated instructions depart from authoritative inputs. The service receives the 4% review threshold but invents a 6% exemption in its raw output. Consequently, the borrower with PD of 4.5% receives a policy-inconsistent recommendation. Correctly passing numerical PDs into a language service does not ensure faithful policy interpretation.

The bank’s obligations extend to end-to-end validation, source grounding and controls proportionate to automatic alert suppression. Numerical-model approval and a generic provider certificate do not establish the reliability of the combined workflow. More sophisticated or externally supplied technology does not transfer the bank’s governance responsibilities to the vendor.

• A. The segment-calibration errors persist with GenAI disconnected, identifying a weakness in conventional numerical-model validation rather than a GenAI-induced PD failure.
• B. The current approved rule was the only retrieved policy supplied to GenAI, so stale policy retrieval does not explain the invented exemption.
• C. Offsetting segment errors undermine pooled calibration, while the unsupported threshold in the raw response identifies a separate GenAI policy-interpretation failure.
• D. The invented exemption already appears in the raw GenAI response, which the dashboard displays unchanged; output mapping is not the demonstrated failure.

---

Question 34

Topic: Current Issues in Financial Markets

A financial supervisor reviews a securities settlement system and its three largest settlement banks. The immediate priority is to preserve critical settlement processing within an agreed 60-minute disruption tolerance during a common third-party outage.

Review findings:

• Each institution uses a different primary cloud provider and a second provider for recovery. Individual cloud-failover tests restored processing within the disruption tolerance.
• One external identity service authenticates settlement users and recovery administrators across all primary and backup environments. It remained available during every failover test. Existing sessions expire within 30 minutes; an identity-service outage would then block new settlement instructions and recovery activation.
• Institutions report direct cloud suppliers but do not exchange information about shared upstream dependencies.
• Delayed settlements deprive other participants of incoming cash, potentially spreading intraday liquidity shortfalls.

Which supervisory response best addresses the systemic weakness revealed by these findings?

• A. Require real-time threat-intelligence sharing and joint penetration tests that assess privileged-access security across each institution’s primary and recovery cloud environments.
• B. Require tighter cloud-provider concentration limits and joint failover exercises that transfer settlement workloads between each institution’s primary and recovery cloud providers.
• C. Require shared-dependency reporting and joint outage exercises that demonstrate independently operable authentication for critical settlement processing and recovery activation.
• D. Require higher prefunded settlement balances and joint liquidity stress tests that assess cash shortfalls caused by delayed receipts during a provider outage.

Best answer: C

Explanation: System-wide digital resilience requires identifying common upstream dependencies and testing correlated disruption. Cloud diversification does not provide independent recovery when every environment relies on one identity service. Here, sessions expire within 30 minutes, so a prolonged outage could disable both settlement processing and recovery, breaching the 60-minute disruption tolerance across multiple institutions.

Shared-dependency reporting gives the supervisor a sector-wide view of concentration. Joint outage exercises should demonstrate that independently operable authentication can sustain critical processing and activate recovery when the common service is unavailable. This combines institutional recovery arrangements with macroprudential oversight of simultaneous failures.

Payment interdependence amplifies the concern: one institution’s missed settlement becomes another’s missing cash receipt. Resilience assessment must therefore examine common failures and their propagation, rather than infer system-wide continuity from successful isolated cloud-failover tests.

• A. Threat intelligence and penetration testing address compromise and access weaknesses, not continuity when a shared authentication service becomes unavailable.
• B. Changing cloud providers leaves the common identity dependency intact, so diversified cloud recovery can still fail simultaneously when authentication is unavailable.
• C. Independent authentication addresses the shared operational bottleneck, while dependency reporting and joint outage exercises assess whether recovery remains effective across institutions.
• D. Prefunding can absorb liquidity consequences of missed receipts, but cash balances do not restore settlement processing or recovery access during the identity-service outage.

---

Question 35

Topic: Credit Risk Measurement and Management

A bank is considering an uncollateralized derivative portfolio with one year remaining. The quoted upfront price, payable by the bank, is $3,640,000. Its pricing policy permits entry only if this price does not exceed the bank-side bilateral value. Ignore funding, capital, and transaction-cost adjustments.

Valuation inputs:

• Default-free portfolio value to the bank: $3,800,000.
• Expected positive clean close-out exposure: $12,000,000.
• Expected negative clean close-out exposure magnitude: $8,000,000.
• These conditional exposure expectations apply in every first-default scenario below.
• Recovery rates: counterparty 40%; bank 25%.
• Discount factor for every close-out gain or loss: 0.95.

Close-out conventions: The portfolio terminates at the first default using default-free replacement value. The surviving party pays its liabilities in full and receives the defaulter’s recovery rate on its claims. Close-out amounts are settled at year-end. No simultaneous defaults occur.

The model provides these mutually exclusive risk-neutral probabilities over the remaining year:

What bilateral value and trade decision follow from these inputs?

• A. The bilateral value is $3,982,400; accept the quoted price.
• B. The bilateral value is $3,663,200; accept the quoted price.
• C. The bilateral value is $3,617,600; decline the quoted price.
• D. The bilateral value is $3,389,600; decline the quoted price.

Best answer: C

Explanation: Bilateral valuation stops at the first default. When both parties default within the year, only the earlier default affects close-out. The relevant probabilities are therefore 6% for counterparty-first default and 4% for bank-first default.

Counterparty default reduces the bank’s positive claims, while the bank’s own default reduces its payments on negative exposures. Each adjustment uses the relevant party’s loss-given-default. In millions of dollars:

• Counterparty loss adjustment: \(12 \times 0.60 \times 0.06 \times 0.95 = 0.4104\).
• Own-default benefit: \(8 \times 0.75 \times 0.04 \times 0.95 = 0.2280\).

From the bank’s perspective, value equals default-free value minus CVA plus DVA:

\[ V_{\text{bank}} = 3.8 - 0.4104 + 0.2280 = 3.6176 \]

The quoted price exceeds this value by $22,400, so the bank’s policy requires declining the trade.

• A. Adding counterparty losses and subtracting own-default benefits reverses both valuation-adjustment signs from the bank’s perspective.
• B. Using marginal default probabilities of 7% and 6% includes second defaults that cannot change an already closed-out portfolio.
• C. First-default counterparty losses of $410,400 and own-default benefits of $228,000 produce a value below the $3,640,000 price.
• D. Subtracting only the $410,400 first-default counterparty loss omits the $228,000 benefit from the bank’s own default.

---

Question 36

Topic: Operational Risk and Resilience

A bank rates an established domestic charity as low risk. Its registration, controllers, donors, and legitimate funding sources have been verified. The charity previously made occasional large transfers directly to established relief partners.

Recent activity:

• Following an expansion into disaster relief, the charity begins making frequent online cross-border payments of $900-$1,400 to two new intermediaries.
• The intermediaries operate in a jurisdiction the bank assesses as having elevated terrorist-financing risk.
• Sanctions screening produces no matches, but the charity has not provided documentation of the ultimate recipients or the intermediaries’ onward use of funds.
• Automated monitoring flags individual transfers above $10,000. None of the new payments generates an alert.

Under the Basel Committee’s July 2020 AML/CFT guidance, which recommendation best addresses the risks revealed by this change in activity?

• A. Raise the customer risk rating and add linked-payment analysis and checks on onward beneficiaries and fund use to the existing amount-based monitoring.
• B. Retain the low-risk rating and add checks on the intermediaries’ corporate registration and payment-provider authorization to the existing amount-based monitoring.
• C. Retain the low-risk rating and add more frequent sanctions screening of the new intermediaries to the existing amount-based monitoring.
• D. Raise the customer risk rating and add donor source-of-funds refreshes and incoming-donation reconciliation to the existing amount-based monitoring.

Best answer: A

Explanation: Risk-based AML/CFT assessment combines customer characteristics with product, geography, and transaction risks, and must respond to changes in activity. Verified charitable status and legitimate incoming funds do not neutralize the risks introduced by new intermediaries, an elevated-risk destination, and unexplained onward disbursements.

Terrorist financing can involve legitimately sourced funds and relatively small payments. Monitoring based solely on individual transfer amounts can therefore miss relevant patterns. Negative sanctions screening also does not establish that funds reach their stated beneficiaries or serve their stated purpose.

The bank should increase the customer’s risk classification, connect related payments in its monitoring, and obtain credible information about onward beneficiaries and fund use. These indicators support proportionate enhanced due diligence and investigation, not an automatic conclusion that terrorist financing has occurred.

• A. The changed payment pattern, destination risk, and undocumented onward disbursements require enhanced customer assessment and monitoring that connects smaller transfers.
• B. Legal registration or authorization cannot resolve the changed transaction pattern and unknown onward beneficiaries, so those checks alone do not support continued low-risk treatment.
• C. Sanctions screening does not establish the purpose or destination of onward payments, and negative matches do not justify retaining the existing customer risk assessment.
• D. The funding sources are already verified; repeating origin checks does not address the unexplained onward use of funds, which remains relevant to terrorist-financing risk.

---

Question 37

Topic: Risk Management and Investment Management

A fund allocates capital to a rates manager and an equity manager. Its investment committee requires the fund to use a 12.0% annual portfolio volatility budget in full, with half of that budget assigned to each manager through Euler component-volatility allocation.

Annual risk estimates:

The correlation between the managers’ returns is 0.125.

Implementation assumptions:

• Allocations are percentages of fund NAV, and portfolio returns are approximated by the linear weighted sum of manager returns.
• The volatility and correlation estimates remain unchanged when allocations are scaled.
• Allocations above 100% of NAV are financed by borrowing. Borrowing is limited to 30% of NAV, and its deterministic cost adds no volatility.

Which allocation meets both the total volatility budget and the manager risk split, allowing for rounding to one decimal place?

• A. Allocate 84.9% of NAV to the rates manager and 42.4% to the equity manager, borrowing 27.3% of NAV.
• B. Allocate 80.0% of NAV to the rates manager and 40.0% to the equity manager, borrowing 20.0% of NAV.
• C. Allocate 66.7% of NAV to the rates manager and 33.3% to the equity manager, with no borrowing.
• D. Allocate 51.2% of NAV to the rates manager and 51.2% to the equity manager, borrowing 2.4% of NAV.

Best answer: B

Explanation: Risk budgeting allocates portfolio risk rather than invested capital. For annual return covariance matrix \( \Sigma \), manager \(i\)’s Euler component-volatility contribution is \( \mathrm{CR}_i = w_i(\Sigma w)_i/\sigma_p \), where \(w_i\) is the allocation relative to NAV and \(\sigma_p\) is portfolio volatility. These contributions sum to total volatility.

With two managers, equal contributions require \(w_R^2\sigma_R^2 = w_E^2\sigma_E^2\), because the covariance terms in their contributions are identical. Thus, the rates allocation must be twice the equity allocation.

At allocations of 80% and 40%, each manager’s weighted standalone volatility is 8%. Including correlation gives:

\[ \sigma_p = \sqrt{0.08^2 + 0.08^2 + 2(0.125)(0.08)(0.08)} = 0.12. \]

Each manager contributes \( (0.08^2 + 0.125 \times 0.08^2)/0.12 = 0.06 \), or 6.0 volatility percentage points. Combined allocations of 120% require borrowing 20% of NAV. Under unchanged covariance assumptions, leverage scales total volatility and component contributions proportionally.

• A. This allocation approximately equalizes component risk but effectively ignores correlation when scaling exposure; the positive correlation raises annual portfolio volatility to about 12.7%.
• B. These allocations produce 12.0% annual volatility and a 6.0-percentage-point contribution from each manager, while borrowing remains within the 30% limit.
• C. The approximately 2:1 allocation produces equal component risk contributions, but annual portfolio volatility is only about 10.0%, leaving the total budget underused.
• D. Equal capital allocations produce approximately 12.0% total volatility, but component contributions are about 2.7 percentage points for rates and 9.3 for equity.

---

Question 38

Topic: Operational Risk and Resilience

A bank’s payment service permits a maximum interruption of 120 minutes, measured from loss of processing until processing resumes with mandatory integrity checks complete. Its maximum acceptable data loss is 10 minutes of instructions accepted before the disruption.

A controlled loss-of-region exercise isolates the primary region at minute 0, stopping payment processing.

Recovery observations:

The two component restorations start at minute 0 and run in parallel at the alternate site.

Payment processing can resume only after both components are ready and an additional 25-minute journal replay and mandatory reconciliation have finished.

Backup evidence:

• The independently stored database snapshot represents the state 20 minutes before disruption.
• An independently stored transaction journal contains all accepted instructions through 5 minutes before disruption. No newer recoverable records exist.
• The exercise confirms that both records remain accessible after regional isolation and that all journal entries can be successfully replayed against the snapshot.

Which service-level conclusion is supported by the exercise?

• A. The demonstrated interruption is 100 minutes and potential data loss is 20 minutes; only the data-loss limit is exceeded.
• B. The demonstrated interruption is 125 minutes and potential data loss is 20 minutes; both limits are exceeded.
• C. The demonstrated interruption is 100 minutes and potential data loss is 5 minutes; both limits are met.
• D. The demonstrated interruption is 125 minutes and potential data loss is 5 minutes; only the interruption limit is exceeded.

Best answer: D

Explanation: Recovery objectives must be assessed against the complete service outcome, including dependencies and mandatory recovery work. Meeting individual component targets does not establish that the service meets its interruption tolerance.

Because the component restorations run in parallel, their combined readiness time is the slower restoration, not their sum. End-to-end recovery therefore takes \(\max(80,100)+25=125\) minutes, exceeding the 120-minute limit by 5 minutes.

The recovery point depends on the latest consistently recoverable data, not merely the age of the base snapshot. Successful replay of the independently available journal advances the restored instruction history to 5 minutes before disruption. Potential data loss therefore meets the 10-minute allowance. The evidence supports the backup arrangement for this event, while showing that the combined recovery sequence is too slow.

• A. This omits the mandatory final recovery work and ignores the usable journal, understating the interruption while overstating potential data loss.
• B. The snapshot’s age does not determine data loss when the independently usable journal extends recoverable instructions to 5 minutes before disruption.
• C. The 25-minute replay and reconciliation must finish before processing resumes, so component readiness at minute 100 does not establish service recovery.
• D. Parallel restorations require 100 minutes, followed by 25 minutes of mandatory recovery work; journal replay limits potential data loss to 5 minutes.

---

Question 39

Topic: Risk Management and Investment Management

An open-end fund must fund redemptions payable at the end of business day 2. Its policy requires coverage of the modeled 99% redemption-at-risk amount while retaining a $2 million operational cash buffer. Current cash is $8 million.

Redemption model: Two investor groups’ two-day net redemptions are jointly normal. Their means are $4 million and $5 million, their standard deviations are $3 million and $5 million, and their correlation is 0.50. The estimates use 250 calm-market observations. Use a one-sided 99% standard normal quantile of 2.326 and round redemption at risk upward to the next $0.1 million.

Liquidation waterfall: Use cash above the buffer first, then securities whose proceeds settle by day 2, in increasing order of liquidation discount. For sales initiated today:

Discounts reduce cash proceeds. Partial sales are permitted in market-value increments of $0.01 million. Assume the stated discounts and settlement times remain fixed.

Intervention condition: Under the applicable fund rules, extraordinary postponement requires a formal systemic-liquidity emergency declaration and regulatory approval. Neither is in place.

Which plan meets the modeled coverage requirement at the lowest liquidation cost?

• A. Use $6 million of cash, sell $10 million of government bonds, and sell $9.55 million of listed equities.
• B. Use $6 million of cash, sell $10 million of government bonds, and sell $6.79 million of listed equities.
• C. Use $6 million of cash, sell $10 million of government bonds, and sell $12 million of listed equities.
• D. Use $6 million of cash, sell $10 million of government bonds, and sell $9.45 million of corporate bonds.

Best answer: A

Explanation: Redemption at risk must reflect the covariance between investor groups. In millions of dollars, the standard deviation of aggregate net redemptions is \( \sqrt{3^2+5^2+2(0.50)(3)(5)}=7 \). The 99% amount is therefore \( 4+5+2.326(7)=25.282 \), rounded upward to $25.3 million.

The cash buffer leaves $6 million available. Selling all $10 million of government bonds produces $9.95 million, leaving a $9.35 million funding gap. Corporate bonds cannot fill that gap because their proceeds settle after the deadline. Required equity sales are \( 9.35/0.98=9.5408 \) million, rounded upward to $9.55 million under the trading increment. Total available proceeds are $25.309 million, and liquidation costs are $0.241 million.

Coverage remains conditional on the normal model, estimated dependence, and fixed sale discounts. Calm-market data may understate stressed withdrawals and liquidation costs. Meeting the modeled quantile is not a guarantee of liquidity, and postponement cannot be assumed without the required declaration and approval.

• A. This plan provides $25.309 million by day 2, covering the $25.3 million requirement through the least-cost eligible sales.
• B. This plan provides only $22.6042 million by day 2; ignoring the positive correlation understates the required redemption coverage.
• C. This plan covers redemptions but incurs $0.29 million of liquidation costs; adding the groups’ individual tail amounts unnecessarily increases asset sales.
• D. Corporate-bond proceeds arrive on day 3, leaving only $15.95 million available by the redemption deadline; extraordinary postponement is unavailable.

---

Question 40

Topic: Risk Management and Investment Management

A pension plan must select one of four fully invested allocations for its $100 million portfolio. A $20 million benefit settlement is due in 30 days and must be funded by asset sales, with no external cash or new borrowing.

The mandate is to maximize forecast year-end assets after the payment, subject to an inception limit of $7 million for one-month 99% VaR. The allocations must be compared as presented. All VaR estimates use a common valuation basis and exclude liquidation costs.

Under the committee’s common liquidity-stress scenario, executable bids allow sufficient holdings in every allocation to be sold before settlement at the discounts shown.

For this comparison, assume no market-value changes before settlement. Discounts apply only to holdings sold; remaining assets earn the stated forecast returns over the following 11 months. Ignore other costs and cash flows.

Which allocation should the committee select under its stated criterion?

• A. Select the short-duration bond allocation.
• B. Select the liquid balanced allocation.
• C. Select the private credit allocation.
• D. Select the leveraged global allocation.

Best answer: B

Explanation: Investment risk management must connect the return objective to both the risk budget and the cost of funding liabilities. A discounted sale requires disposing of more book value than the cash payment itself.

With amounts in millions of dollars, forecast year-end assets are \( V = \left(100 - \frac{20}{1-h}\right)(1+r) \), where \(h\) is the sale discount and \(r\) is the post-payment return forecast.

For the liquid balanced allocation, the sale removes approximately $20.20 million of holdings. The remaining $79.80 million grows to approximately $86.18 million, the greatest forecast value among allocations satisfying the inception VaR limit.

VaR is a constraint here, not the objective to minimize. It also excludes the specified execution costs. The conclusion is conditional on the supplied discounts and return forecasts; neither a low measured VaR nor a higher expected return guarantees superior realized performance.

• A. With no sale discount, $80 million remains invested and grows to a forecast $83.20 million; minimizing VaR does not maximize assets under this mandate.
• B. Selling $20.20 million of holdings funds the payment and leaves forecast year-end assets of $86.18 million, the highest among allocations within the VaR budget.
• C. The 8% sale discount requires selling $21.74 million of holdings, leaving forecast year-end assets of $85.30 million, below the liquid balanced allocation’s $86.18 million.
• D. Its $8.5 million inception VaR exceeds the $7 million mandate limit, despite its higher forecast return and resulting year-end assets.

---

Question 41

Topic: Operational Risk and Resilience

A bank has recorded eight operational losses from corporate payments sent to superseded beneficiary accounts. Most principal was recovered, but investigation costs, customer compensation and client departures remained material.

Process and event review:

• At 08:00, the payment workstation receives an extract of the approved beneficiary master.
• Authenticated intraday amendments update the approved beneficiary master and take effect immediately.
• Batches are released at 15:00 after a preparer and an approver check payments against the 08:00 extract.
• In all eight incidents, amendments were approved between preparation and release. The released payments matched the extract, and both employees followed the documented checks.

The bank prioritizes reducing recurring payment errors and client disruption while continuing to accept intraday amendments. All proposals have comparable annual costs. Which control response best fits the evidence?

• A. Insure erroneous-payment losses and remediation expenses, retaining existing release checks and assigning incidents to a dedicated recovery team.
• B. Add an independent reviewer to compare batches with the preparation extract, holding reviewer disagreements for correction and renewed approval.
• C. Automate post-release reconciliation with the current approved beneficiary master, assigning mismatches for immediate recovery and customer remediation.
• D. Automate a pre-release comparison with the current approved beneficiary master, holding mismatches for correction and renewed approval.

Best answer: D

Explanation: Risk reduction should target the mechanism causing repeated losses. These payments were not mistyped: they matched a data snapshot that became outdated when later amendments took effect. Multiple reviewers using that snapshot share the same informational weakness, even when they perform their documented checks correctly.

An automated comparison against the current approved beneficiary master immediately before release is a preventive control. Holding mismatches for correction and renewed approval addresses the timing gap without abandoning intraday amendments. Automation still depends on reliable master data and sound authorization controls; it does not eliminate operational risk. Insurance can transfer covered monetary losses, but it does not prevent payment errors or remove reputational consequences.

• A. Insurance transfers covered financial consequences, but the retained checks still permit payments to superseded accounts and resulting damage to client relationships.
• B. The additional reviewer still uses the same outdated extract, so agreement among reviewers does not detect beneficiary amendments made after preparation.
• C. Post-release reconciliation detects mismatches after erroneous payments have occurred, when recovery and compensation cannot prevent the initial client disruption.
• D. Comparison with the current approved master before release detects the stale-data mismatch that caused the losses and prevents affected payments from proceeding.

---

Question 42

Topic: Market Risk Measurement and Management

A dealer holds a long corporate bond position and proposes shorting 500 Treasury futures contracts to match its total DV01. The risk manager must choose the number of short contracts that minimizes estimated daily net P&L variance.

Estimates from the same daily sample:

Corporate bond yield changes include both Treasury-rate and credit-spread movements. Futures yield changes refer to the reference Treasury yield. The correlation between these two yield changes is 0.80.

Assume constant DV01s, stable covariance estimates, and linear price changes. Ignore carry, convexity, trading costs, and futures basis changes. There is no post-hedge DV01 limit.

Which short futures position should the risk manager approve?

• A. Short 400 Treasury futures contracts.
• B. Short 600 Treasury futures contracts.
• C. Short 750 Treasury futures contracts.
• D. Short 500 Treasury futures contracts.

Best answer: B

Explanation: The minimum-variance cross-hedge scales the DV01 ratio by the regression slope of corporate yield changes on Treasury yield changes. With \(\rho\) denoting correlation and \(\sigma_B\) and \(\sigma_T\) denoting the respective yield-change standard deviations, the slope is \(\beta = \rho\sigma_B/\sigma_T = 0.80 \times 9/6 = 1.20\).

The required number of short contracts is therefore:

\[ N^* = \frac{45,000}{90} \times 1.20 = 600. \]

The larger short position reflects the corporate yield’s estimated sensitivity to Treasury movements. It is not parallel-DV01-neutral: the futures hedge has $54,000 of DV01 against the bond’s $45,000. That mismatch is compatible with the stated variance-minimization objective. Correlation below one means that corporate yield movements not explained by Treasury yield changes remain unhedged.

• A. Multiplying the DV01-matched 500 contracts by correlation alone omits the 9-to-6 yield-volatility ratio required for minimum-variance hedging.
• B. The regression sensitivity is \(0.80 \times 9/6 = 1.20\), giving a minimum-variance hedge of \(45,000 \times 1.20/90 = 600\) contracts.
• C. Scaling 500 contracts by the 9-to-6 yield-volatility ratio assumes perfect correlation rather than the observed correlation of 0.80.
• D. Matching DV01s targets equal-sized yield changes, whereas the estimated regression sensitivity of corporate yield changes to Treasury yield changes is 1.20.

---

Question 43

Topic: Current Issues in Financial Markets

A supervisory risk committee is reviewing AI-assisted payment screening at eight banks that process 70% of domestic retail payment value. It must prioritize the next review based on the potential for several banks to breach their two-hour screening-disruption tolerance simultaneously.

Review findings:

• Four banks use AI provider Alpha as primary and Beta as backup; the other four reverse these arrangements. Alpha and Beta independently developed their screening models.
• Both providers, including their backup endpoints, depend on CloudCo’s identity service. Freshly renewed authorizations remain valid for 60 minutes. When they expire, screening stops until the identity service returns; no alternative screening process is available.
• Each bank passed a 30-minute test that blocked the identity service immediately after authorizations were renewed. No test continued beyond authorization expiry.
• CloudCo’s severe but plausible scenario assumes a four-hour identity-service outage. No actual sector-wide screening outage has been recorded.

Which report summary provides the strongest basis for prioritizing the committee’s next supervisory response?

• A. Report that balanced primary-provider shares reduce the sector’s common-infrastructure outage exposure. Recommend four-hour provider-switching tests at each bank and continued monitoring of direct AI-provider concentration.
• B. Report that shared identity dependence leaves prolonged continuity untested and could breach disruption tolerances. Recommend quantifying exposed payment flows and conducting a coordinated four-hour identity-service outage exercise.
• C. Report that shared identity dependence creates correlated screening errors despite independent model development. Recommend comparing cross-bank screening decisions and coordinating tests of simultaneous model misclassification.
• D. Report that successful identity-service tests support continuity throughout the severe outage scenario. Recommend coordinating repeats of the 30-minute tests and adding real-time monitoring of identity-service availability.

Best answer: B

Explanation: AI-provider diversification does not eliminate concentration in shared infrastructure. The FSB’s 2024 AI report highlights third-party concentration, while Banco de Espana’s 2023 digital-resilience analysis addresses systemic transmission through common ICT dependencies.

Here, all primary and backup arrangements require the same identity service. Successful 30-minute tests establish continuity only while cached authorizations remain valid. Under the four-hour scenario, screening would stop after the first hour and remain unavailable for three hours, exceeding each bank’s two-hour tolerance.

The senior report should therefore connect the dependency to affected payment flows, identify the limitation of existing tests, and propose a coordinated exercise extending beyond authorization expiry. The absence of a recorded sector-wide outage is a caveat about observed experience, not evidence that the common-mode exposure is controlled.

• A. Balanced AI-provider shares diversify direct vendor exposure, but switching providers does not remove the shared identity-service dependency that threatens simultaneous screening continuity.
• B. A four-hour outage outlasts cached authorization by three hours, exceeding the two-hour tolerance across banks whose primary and backup arrangements share the dependency.
• C. Shared authentication infrastructure supports concern about correlated service interruption, not correlated misclassification; the findings provide no evidence of common screening errors.
• D. The tests ended before authorization expiry, so they demonstrate short-term continuity rather than the ability to withstand the four-hour outage.

---

Question 44

Topic: Current Issues in Financial Markets

A pension investment committee is reviewing a private-credit allocation using the BIS’s 2025 The Global Drivers of Private Credit. An investment memo attributes lower borrower-default risk to the fund’s rapid growth and low reported volatility. The figures below describe this allocation, not the global market.

Case evidence:

• The closed-end fund has seven-year institutional commitments and no investor redemption rights. Its growth followed new pension and insurance commitments while local banks tightened lending.
• Its floating-rate loans finance smaller, more leveraged private firms. The floating-rate bank-loan benchmark primarily contains larger public firms.
• Fund returns use quarterly manager-estimated valuations, and borrower information consists mainly of annual unaudited accounts. The benchmark uses dealer quotes and borrowers’ quarterly public reports.

Internal comparison:

Stress projections hold EBITDA unchanged and raise the common base rate by 200 basis points. Interest coverage is EBITDA divided by cash interest. Volatility uses quarterly returns over the same period.

Which conclusion is most defensible when comparing underlying credit risk?

• A. The lock-up reduces redemption pressure, while lower reported volatility supports reduced borrower-default exposure; that assessment remains provisional pending updated accounts for the private borrowers.
• B. The lock-up reduces redemption pressure, while weaker stressed coverage signals borrower vulnerability; valuation smoothing and differing borrower quality undermine the claimed credit-risk advantage.
• C. The lock-up reduces redemption pressure, while the spread premium supports lower borrower-default risk; that assessment remains provisional pending borrower-matched pricing comparisons.
• D. The lock-up reduces redemption pressure, while growth following bank retrenchment supports stronger borrower selection; that assessment remains provisional pending independently verified quarterly valuations.

Best answer: B

Explanation: Private-credit assessment must distinguish stable fund financing from borrower repayment capacity and reported valuation stability. Seven-year commitments limit investor-driven redemption pressure, but do not reduce borrowers’ floating-rate interest obligations.

Under the specified stress, median private-credit interest coverage is 0.9, meaning EBITDA covers only 90% of cash interest, versus 1.6 for the bank benchmark. This signals vulnerability, not certain default. Current, verified borrower information is necessary to assess available liquidity and repayment capacity.

Quarterly manager-estimated valuations can smooth reported returns, making low measured volatility an unreliable indicator of low economic credit risk. Comparisons should also control for borrower size, leverage, financing costs, and other credit characteristics. Growth following institutional inflows and bank tightening is consistent with investor demand and credit-supply substitution, not proof of safer lending. The BIS paper’s source-period findings do not establish a universal current-market credit-risk ranking.

• A. Manager-estimated marks can suppress measured volatility without improving repayment capacity, so updated accounts alone cannot validate the volatility-based default-risk comparison.
• B. Stressed coverage below 1 signals repayment vulnerability, while infrequent estimated valuations and different borrower populations weaken the apparent evidence of lower credit risk.
• C. Higher spreads are costs to borrowers and may compensate lenders for risk; they do not demonstrate lower borrower-default probabilities.
• D. Institutional allocations and tighter bank supply can drive growth without improving borrower quality; verifying valuations does not establish stronger borrower selection.

---

Question 45

Topic: Current Issues in Financial Markets

A fund buys a tokenized bond on a permissioned platform operated by a market-infrastructure provider.

Platform rules:

• The operator controls validator admission and contract upgrades. Participants share a synchronized transaction record.
• Validators check permissions, signatures, balances, and contract execution. They do not independently verify external custody or encumbrance records.
• A smart contract atomically exchanges bond tokens for cash tokens when a registrar-signed attestation reports that the bond is unencumbered.
• Cash-token payment becomes legally final upon ledger transfer. Bond title passes only when the external registrar updates the legally authoritative ownership register.

Completed trade:

The attestation was authentic but stale: the bond had already been pledged. The contract completed the token exchange, and all ledger copies agreed. The registrar subsequently rejected the ownership update.

A board member proposes replacing permissioned validation with public, permissionless consensus while retaining the same contract, attestation feed, custody arrangements, and legal rules.

Which assessment of the completed trade and proposed redesign is most accurate?

• A. The payment is final but bond title has not passed; public validation would remove registrar dependence by independently establishing the bond’s encumbrance status.
• B. Both payment and bond title are final; public validation would prevent the registrar from rejecting an ownership change already approved through ledger consensus.
• C. The payment is final but bond title has not passed; public validation would leave dependence on the registrar’s records and external attestations unchanged.
• D. Neither payment nor bond title is final; public validation would be unnecessary because legal settlement requires both tokens to be redeemed through off-ledger institutions.

Best answer: C

Explanation: Ledger sharedness, programmability, and trust address different problems. A synchronized record reduces interparty reconciliation needs. Programmability executes specified conditions automatically, but its results depend on the accuracy of inputs and the legal meaning of the resulting transfers.

Here, signature verification establishes who supplied the attestation, not whether the bond remains unencumbered. The contract correctly executed against inaccurate external information. Its atomic token exchange therefore did not produce legally effective delivery-versus-payment: cash payment became final, but bond title did not pass. The buyer consequently retains principal exposure.

Permissioning and governance determine who can validate transactions or modify contracts. Public consensus changes that validation model, but cannot establish off-ledger asset facts or override the authoritative ownership register. Technical agreement about a digital record is distinct from enforceable rights to the underlying asset.

• A. Public consensus can verify ledger transactions, but it cannot independently establish off-ledger encumbrances from the same registrar-supplied data.
• B. Consensus approval did not transfer bond title because the legally authoritative register was not updated; changing validators does not alter that legal condition.
• C. Cash payment is legally final, while rejected registration leaves the buyer without bond title; the redesign preserves the external verification dependency.
• D. The stated rules make cash-token payment legally final on the ledger, so cash redemption is not required even though bond delivery failed.

---

Question 46

Topic: Current Issues in Financial Markets

A bank’s sovereign-risk committee evaluates monetary-fiscal policy packages using the BIS concept of a joint region of stability. A country has persistent inflation of 6.0%, an inflation target of 2.0%, and opening government debt equal to 120.0% of GDP.

Scenario assumptions:

• Debt is entirely denominated in domestic currency, and a substantial share reprices during the year.
• Primary surpluses of 2.0% or 4.0% of year-end GDP are feasible. The committee applies the conditional projections below to either fiscal target.
• Interest accrues on opening debt. There are no valuation effects, asset sales, or other debt adjustments.

The committee’s stabilizing benchmark must return inflation to target by year-end without increasing the debt-to-GDP ratio. Which package satisfies this criterion?

• A. Maintain restrictive monetary policy and target a primary surplus of 2.0% of year-end GDP.
• B. Cut interest rates immediately and target a primary surplus of 4.0% of year-end GDP.
• C. Maintain restrictive monetary policy and target a primary surplus of 4.0% of year-end GDP.
• D. Cut interest rates immediately and target a primary surplus of 2.0% of year-end GDP.

Best answer: C

Explanation: Joint monetary-fiscal stability cannot be inferred from lower government financing costs alone. Inflation outcomes and debt dynamics must be assessed together.

Let \(b_0\) and \(b_1\) denote opening and closing debt-to-GDP ratios, \(r\) the effective nominal interest rate, \(g\) nominal GDP growth, and \(s\) the primary surplus divided by year-end GDP. Expressing all inputs as decimals:

\[ b_1 = b_0\frac{1+r}{1+g}-s \]

Under restrictive monetary policy and a 4.0% surplus:

\[ b_1 = 1.20\frac{1.07}{1.04}-0.04 \approx 1.1946 \]

Debt therefore falls to 119.46% of GDP while inflation reaches its target. A positive primary surplus alone is insufficient when interest costs exceed nominal growth.

Immediate easing improves near-term debt arithmetic but leaves inflation above target. The selected benchmark is conditional on the supplied projections, not a universally optimal policy recommendation. Fiscal consolidation can weaken growth, while monetary easing can reduce debt service at the cost of inflation persistence.

• A. The 2.0% surplus leaves year-end debt at 121.46% of GDP, meeting the inflation objective but failing to stabilize the debt ratio.
• B. The larger surplus lowers the debt ratio to 114.87%, but immediate easing leaves projected inflation at 5.0%, above the target.
• C. Restrictive policy returns inflation to 2.0%, while the 4.0% primary surplus reduces the debt ratio to 119.46%.
• D. Lower financing costs reduce the debt ratio to 116.87%, but projected inflation of 5.0% fails the committee’s price-stability criterion.

---

Question 47

Topic: Liquidity and Treasury Risk Measurement and Management

A risk committee is reviewing Bear Stearns’ March 2008 failure. Dependence on short-term secured funding, deteriorating counterparty confidence, and customer withdrawals contributed to its rapid liquidity collapse.

To assess the same rollover mechanism, the committee uses this illustrative one-day dealer stress, not Bear Stearns’ reported balances:

• All existing repos have a 20% haircut, and collateral market prices remain unchanged.
• Collateral excluded from replacement repos cannot be sold or refinanced before day-end.
• There are no other cash flows or financing sources. Ignore interest, fees, and intraday settlement timing.

Which assessment of the net end-of-day funding gap is correct?

• A. A $2 billion shortfall, reflecting released collateral as cash available against repo maturities and the other cash outflows.
• B. A $7 billion shortfall, reflecting the total reduction in repo proceeds and the other cash outflows.
• C. A $3 billion shortfall, reflecting the haircut increase on continuing repos and the other cash outflows.
• D. A $3.25 billion shortfall, reflecting refinancing of the entire collateral pool at the new haircut and the other cash outflows.

Best answer: B

Explanation: Repo collateral limits lenders’ credit exposure; it does not commit them to renew short-term financing. Bear Stearns’ experience illustrates how deteriorating confidence and withdrawals can interact with dependence on short-term funding.

Replacement lenders advance 75% of the $20 billion collateral they accept, providing $15 billion. The dealer must therefore supply $5 billion to repay maturing repos, alongside $3 billion of other outflows. After using its $1 billion cash reserve, the funding gap, in billions of dollars, is:

\[ 20 - (1 - 0.25) \times 20 + 3 - 1 = 7 \]

The remaining collateral cannot generate cash by the deadline. Thus, unchanged collateral prices and collateral value exceeding repo principal do not prevent a funding-liquidity shortfall. This cash gap alone does not establish insolvency.

• A. Released collateral is not same-day cash; counting its $5 billion market value improperly reduces the funding gap.
• B. Replacement repos raise $15 billion; repaying $20 billion plus $3 billion in outflows, less $1 billion cash, creates a $7 billion gap.
• C. The haircut increase requires $1 billion on continuing repos, but another $4 billion of existing borrowing loses refinancing altogether.
• D. This calculation refinances the entire $25 billion collateral pool, although counterparties will accept only $20 billion of collateral.

---

Question 48

Topic: Market Risk Measurement and Management

A bank backtests a 99% one-day VaR model over 1,000 daily observations containing 12 exceptions. Its statistical approval rule requires non-rejection of the joint conditional-coverage null.

The joint null states that exceptions occur independently over time, with an exception probability of 1% each day. Validation reports:

The conditional-coverage statistic is evaluated against a chi-square distribution with two degrees of freedom. Its critical values are 5.99 at 5% significance and 9.21 at 1% significance. Reject a null when the test statistic exceeds its critical value.

The committee proposes lowering the significance level from 5% to 1%, keeping the test design and sample size unchanged. Which conclusion about conditional coverage and the error trade-off is accurate?

• A. Fail to reject conditional coverage at both levels; the change reduces rejection of accurate models and increases failure to reject inaccurate models.
• B. Reject conditional coverage at both levels; the change reduces rejection of accurate models and increases failure to reject inaccurate models.
• C. Reject conditional coverage at 5% and fail to reject at 1%; the change reduces failure to reject inaccurate models and increases rejection of accurate models.
• D. Reject conditional coverage at 5% and fail to reject at 1%; the change reduces rejection of accurate models and increases failure to reject inaccurate models.

Best answer: D

Explanation: Conditional coverage requires both the correct exception probability and independent exception indicators. Here, the unconditional-coverage statistic does not reject the 1% exception probability, but the independence statistic rejects independence at 5%, indicating clustering.

The joint likelihood-ratio statistic adds the two components: \( LR_{CC} = 0.38 + 7.43 = 7.81 \). Because 7.81 exceeds 5.99, conditional coverage is rejected at 5%. Because it is below 9.21, conditional coverage is not rejected at 1%.

Lowering significance makes rejection harder. It reduces Type I error, rejecting a true null, but increases Type II error, failing to reject a false null, for a fixed inaccurate model and sample size. Non-rejection at 1% is therefore not proof of model adequacy; it reflects a less powerful test, and the clustering evidence remains relevant.

• A. Non-rejection of exception frequency does not establish conditional coverage: the combined statistic is 7.81, exceeding the 5% critical value of 5.99.
• B. The combined statistic of 7.81 is below the 1% conditional-coverage critical value of 9.21, so the joint null is not rejected at that level.
• C. Lowering significance reduces Type I errors, rejecting accurate models, while increasing Type II errors, failing to reject inaccurate models, rather than the reverse.
• D. The combined statistic of 7.81 exceeds 5.99 but not 9.21; the higher rejection threshold reduces false rejections while increasing missed model deficiencies.

---

Question 49

Topic: Credit Risk Measurement and Management

A bank evaluates a bond-CDS hedge using the combined fair-value change, including counterparty credit valuation adjustment (CVA).

Position and contractual terms:

• The bank holds $20 million of Alder Manufacturing’s senior unsecured bonds and has purchased CDS protection for the full notional.
• Both instruments have five years remaining. The CDS references Alder Manufacturing, covers the relevant credit events, and accepts the bond as a deliverable obligation.
• Enforceability has been confirmed. The CDS is uncollateralized.

One-day valuation evidence:

The bond and purchased CDS each have an absolute spread sensitivity of $8,000 per bp. Use first-order spread revaluation and apply CVA separately. Ignore interest-rate changes, carry, convexity, and other valuation adjustments.

Which assessment correctly attributes the portfolio’s one-day fair-value change?

• A. An $80,000 loss: $0 from cash-CDS basis risk and $80,000 from counterparty risk.
• B. A $560,000 loss: $480,000 from documentation risk and $80,000 from counterparty risk.
• C. A $560,000 loss: $480,000 from cash-CDS basis risk and $80,000 from counterparty risk.
• D. A $400,000 loss: $480,000 from cash-CDS basis risk, offset by an $80,000 counterparty-risk gain.

Best answer: C

Explanation: Matching reference entities, maturities, notionals, and credit-event terms can align default protection without eliminating mark-to-market basis risk. Equal spread sensitivities offset equal spread changes, not different changes in cash and CDS markets.

The bond spread widens by 80 bp, producing a $640,000 loss. The purchased CDS spread widens by 20 bp, producing a $160,000 gain before counterparty adjustment. The resulting clean hedge loss is $480,000. Equivalently, the CDS-bond basis, defined as CDS spread minus bond asset-swap spread, moves from -30 bp to -90 bp.

CVA is deducted from the counterparty-risk-free value of the CDS. Its $80,000 increase therefore creates an additional loss, bringing the combined loss to $560,000. Contractual matching does not force cash and CDS spreads to move together, and purchasing protection introduces exposure to the protection seller.

• A. Equal spread sensitivities do not eliminate basis risk when cash and CDS spreads move differently; the unoffset clean valuation loss is $480,000.
• B. Credit-event provisions govern settlement, not daily spread repricing; the matched and enforceable contractual terms do not explain the $480,000 clean hedge loss.
• C. The bond loses $640,000 and the CDS gains $160,000 before CVA; the $80,000 increase in CVA brings the total loss to $560,000.
• D. A larger CVA charge reduces the value of purchased protection; the $80,000 increase adds to the $480,000 basis loss rather than offsetting it.

---

Question 50

Topic: Liquidity and Treasury Risk Measurement and Management

A bank is sizing a new committed credit line to meet its projected 30-day cash requirement. It starts with $45 million of usable cash and must retain at least $10 million at day 30. Ignore interest, transaction costs, and within-period timing.

30-day sources-and-uses forecast: These are all projected cash flows, excluding draws on the new line. Positive amounts are inflows; negative amounts are outflows.

Separate 90-day structure-of-funds benchmark: Treasury applies the following internal liquidity-reserve factors to current funding balances and adds $50 million of projected 90-day net loan growth.

Which recommendation correctly sizes the line for the stated 30-day cash requirement and interprets the structural estimate?

• A. Set the line at $45 million using the 30-day sources-and-uses requirement; retain the structural estimate as a separate 90-day planning benchmark.
• B. Set the line at $35 million using the 30-day sources-and-uses requirement; retain the structural estimate as a separate 90-day planning benchmark.
• C. Set the line at $150 million using the combined sources-and-uses and net structural requirements; treat the estimates as separate, additional funding demands.
• D. Set the line at $105 million using the structural requirement net of opening cash; treat that balance as the 30-day funding requirement.

Best answer: A

Explanation: The sources-and-uses method estimates funding needs over a specified cash-flow horizon. Total inflows are $105 million and outflows are $185 million. Opening cash of $45 million therefore produces a projected closing balance of negative $35 million before financing. A $45 million draw restores the required $10 million cash floor.

The structure-of-funds benchmark is \(50 \times 0.80 + 100 \times 0.30 + 200 \times 0.15 + 50 = 150\) million dollars. It depends on funding-stability classifications, reserve factors, and projected loan growth, rather than matched 30-day cash flows. Subtracting opening cash gives $105 million, but does not convert that 90-day benchmark into a 30-day facility requirement. The estimates are not additive because they address overlapping funding needs. Relating them requires consistent horizons and assumptions about funding behavior and cash-flow timing.

• A. Projected cash falls to negative $35 million before a draw, so $45 million is needed to finish with the required $10 million.
• B. A $35 million draw eliminates the negative projected cash balance but leaves zero cash, rather than the required $10 million at day 30.
• C. Adding the $45 million cash requirement to the $105 million net structural requirement combines different horizons and double-counts overlapping funding needs.
• D. Subtracting opening cash from the $150 million structural requirement gives $105 million, but this 90-day estimate does not establish the 30-day facility requirement.

Questions 51-75

Question 51

Topic: Liquidity and Treasury Risk Measurement and Management

A bank treasurer must decide whether to increase an intraday credit limit before settlement begins.

Funding arrangements:

• Opening settlement cash is $35 million.
• The undrawn intraday credit line has a $65 million limit and is immediately available. Repaid amounts can be redrawn.
• Receipts repay outstanding credit before replenishing cash.
• Payments must settle at their listed times. Receipts cannot be used before their listed times, and no other liquidity sources are available.

The schedule includes all cash flows, in millions of dollars.

Which assessment correctly identifies the peak required credit draw and the minimum credit-limit change needed to settle every payment on time?

• A. Peak required credit draw of $90 million; increase the limit by $25 million.
• B. Peak required credit draw of $0 million; keep the limit at $65 million.
• C. Peak required credit draw of $80 million; increase the limit by $15 million.
• D. Peak required credit draw of $125 million; increase the limit by $60 million.

Best answer: A

Explanation: Peak intraday liquidity usage depends on the largest cumulative net outflow, not the end-of-day balance. In millions of dollars, cumulative payments less cumulative receipts are 55, 125, 75, 115, 25, 70, and 10 at the successive settlement times.

The maximum is $125 million at 09:15. After using opening cash, the required peak credit draw is \(125 - 35 = 90\) million dollars. The existing limit is therefore insufficient by \(90 - 65 = 25\) million dollars. Increasing it to $90 million covers every scheduled payment.

The final net outflow is only $10 million, leaving $25 million of cash after credit repayment. Monitoring only that closing balance would miss the earlier funding shortfall.

• A. At 09:15, cumulative payments total $125 million before any receipts; opening cash covers $35 million, leaving credit demand $25 million above the existing limit.
• B. Opening cash covers the day’s final net outflow, but payments due before receipts still create a substantial intraday borrowing requirement.
• C. The required draw is $80 million at 11:00, but the earlier funding requirement at 09:15 is higher and determines the necessary limit.
• D. The $125 million peak cumulative net outflow includes the amount funded by opening cash; deducting that cash reduces the credit requirement to $90 million.

---

Question 52

Topic: Credit Risk Measurement and Management

A bank is assessing a currently performing loan with a fixed one-year exposure at default of $10 million. Only default losses are modeled, with zero loss if no default occurs and no discounting.

Calibration evidence: A representative cohort of 200 comparable borrowers was observed for one year:

• Eight borrowers missed scheduled debt payments beyond their contractual grace periods.
• Four of those eight borrowers also entered legal bankruptcy.
• Six additional borrowers had liabilities exceeding assets but made every scheduled payment. All remaining borrowers also paid on time.

The model defines default as a payment failure beyond the contractual grace period and uses the observed default frequency as the loan’s one-year probability of default. Conditional on default, loss given default is either 40% or 80%, each with probability 50%.

Reporting conventions:

• Expected loss is the mean loss.
• Unexpected loss is the standard deviation of loss.
• Stand-alone economic capital equals the 99th-percentile loss minus expected loss. The percentile is the smallest loss amount whose cumulative probability is at least 99%.

Which combination should the risk manager report, rounded to the nearest $1,000?

• A. Expected loss: $240,000; unexpected loss: $1,242,000; economic capital: $7,760,000.
• B. Expected loss: $240,000; unexpected loss: $1,242,000; economic capital: $8,000,000.
• C. Expected loss: $240,000; unexpected loss: $1,176,000; economic capital: $7,760,000.
• D. Expected loss: $120,000; unexpected loss: $886,000; economic capital: $3,880,000.

Best answer: A

Explanation: Expected loss measures average credit loss, while unexpected loss measures loss dispersion under the stated standard-deviation convention. Both default occurrence and default severity contribute to dispersion.

The payment-default frequency is 8/200, or 4%. Legal bankruptcy is only a subset of the observed defaults. Balance-sheet insolvency alone does not meet this model’s payment-default definition.

The loan loses zero with probability 96%, $4 million with probability 2%, and $8 million with probability 2%. Measuring loss \(L\) in millions of dollars:

\[ E[L]=0.02(4)+0.02(8)=0.24 \]\[ \mathrm{Var}(L)=0.02(4^2)+0.02(8^2)-0.24^2=1.5424 \]

Unexpected loss is therefore \(\sqrt{1.5424}=1.241934\) million dollars, rounded to $1,242,000.

Cumulative probability reaches only 98% at a $4 million loss, so the 99th-percentile loss is $8 million. Economic capital is $8,000,000 minus $240,000, or $7,760,000, rather than either the standard deviation or the full adverse loss.

• A. The 4% default probability and both severity outcomes produce these mean, standard-deviation, and net-of-expected-loss capital measures.
• B. The $8,000,000 figure is the 99th-percentile loss; the specified capital convention requires deducting the $240,000 expected loss.
• C. The $1,176,000 estimate treats loss given default as fixed at its 60% mean, omitting the additional variance from uncertain severity.
• D. Counting only the four bankruptcy filings understates default probability; all eight payment failures qualify, giving a 4% probability rather than 2%.

---

Question 53

Topic: Market Risk Measurement and Management

An asset manager holds equity and corporate credit sleeves. Its risk engine combines separately estimated marginal return distributions with a Gaussian copula fitted to the full historical sample.

Independent validation supports the marginal models in both calm and stress regimes, including individual loss-tail severity. Within each regime, each sleeve falls below its forecast 5th percentile on approximately 5% of days and exceeds its forecast 95th percentile on approximately 5% of days.

However, the model understates portfolio expected shortfall during stress windows, which include the 2007–2009 global financial crisis. The manager observes these joint frequencies using each sleeve’s regime-specific return percentiles:

The committee will retain the marginal models and use stress-regime 97.5% expected shortfall to set portfolio limits. Which dependence-model revision best matches the observed patterns?

• A. Fit a Gumbel copula to return percentiles with separate parameters for calm and stress regimes.
• B. Fit a Gaussian copula to return percentiles with separate parameters for calm and stress regimes.
• C. Fit a Clayton copula to return percentiles with separate parameters for calm and stress regimes.
• D. Fit a Student t copula to return percentiles with separate parameters for calm and stress regimes.

Best answer: C

Explanation: Copulas separate marginal return behavior from cross-asset dependence. With the marginal models supported by validation, the important deficiency is the joint-loss structure. Joint downside frequency rises from 0.3% to 1.6% during stress, while joint upside frequency remains much smaller at 0.4%. This indicates regime-dependent, asymmetric downside dependence.

A Clayton copula applied to return percentiles has lower-tail dependence and zero asymptotic upper-tail dependence. Separate calm- and stress-regime calibration can represent the stronger concentration of joint losses. Finite-threshold joint upside events can still occur; zero asymptotic upper-tail dependence does not imply their absence.

The manager should recompute portfolio expected shortfall using the stress-calibrated dependence model and retained marginal distributions. The resulting estimate remains conditional on the model and calibration, rather than guaranteeing future tail-loss coverage.

• A. A Gumbel copula applied to returns emphasizes upper-tail dependence, placing stronger clustering in large gains rather than the observed large losses.
• B. Changing Gaussian correlation across regimes cannot reproduce unequal joint lower- and upper-tail frequencies at matching marginal percentile thresholds.
• C. A Clayton return copula captures asymmetric lower-tail dependence, while separate regime parameters accommodate the increased clustering of losses during stress.
• D. A Student t copula imposes symmetric lower- and upper-tail dependence, inconsistent with the substantially greater joint downside frequency in stress periods.

---

Question 54

Topic: Risk Management and Investment Management

A hedge fund holds crowded credit positions financed by dealers that also lend to peer funds. Its risk policy requires reliable liquidity to cover an independently modeled 30-day stress without forced asset sales.

Starting balances:

Stress and contractual terms:

• Repo financing remains available, but dealers increase haircuts from 10% to 25% and require immediate repayment to restore the permitted advance rate. Collateral prices remain unchanged. Both the existing 30-day repo and an offered 90-day extension permit these haircut resets.
• Collateral sold before the stress repays repo borrowing equal to 90% of the sale proceeds; the remaining 10% becomes unencumbered cash.
• Investors request $20 million of redemptions payable at day 30. A contractual fund-level gate can limit payments to 10% of starting NAV, deferring the remainder.
• The same repo dealer offers a $10 million credit line that it may cancel at its discretion.

Ignore transaction costs and other cash flows. Which package meets the stress-liquidity requirement?

• A. Sell $80 million of repo collateral and pay redemption requests in full, while retaining the existing repo maturity.
• B. Sell $50 million of repo collateral and enforce the 10% gate, while extending repo maturity to 90 days.
• C. Sell $75 million of repo collateral and enforce the 10% gate, while retaining the existing repo maturity.
• D. Sell $25 million of repo collateral and enforce the 10% gate, while relying on the $10 million dealer credit line.

Best answer: C

Explanation: Higher repo haircuts create funding liquidity demands even when collateral prices do not fall. Selling financed assets releases only their equity portion as cash because most sale proceeds repay borrowing.

Selling $75 million of collateral repays $67.5 million of repo borrowing and releases $7.5 million, increasing cash to $32.5 million. Remaining collateral is $125 million and remaining borrowing is $112.5 million. At a 25% haircut, permitted borrowing falls to $93.75 million, requiring repayment of $18.75 million. The gate limits redemption payments to $10 million, so total cash demands are $28.75 million and the surplus is $3.75 million.

Common dealer funding and crowded positions can transmit simultaneous haircut increases into forced selling, price declines, and further margin demands. Longer maturity does not protect against the specified haircut resets, and discretionary dealer credit is not dependable stress funding. Redemption gates defer investor payments; they neither eliminate financing demands nor promise unconditional liquid redemption.

• A. The sale leaves $33 million of cash, while $18 million of repo repayments and $20 million of redemptions create a $5 million shortfall.
• B. The sale leaves $30 million of cash against $32.5 million of outflows; extending maturity does not prevent the contractual haircut reset.
• C. The sale provides $32.5 million of cash against $18.75 million of repo repayments and $10 million of gated redemptions, leaving a $3.75 million surplus.
• D. The sale leaves $27.5 million of cash against $36.25 million of demands; the cancellable dealer line cannot reliably cover the resulting funding gap.

---

Question 55

Topic: Liquidity and Treasury Risk Measurement and Management

A risk manager applies the BIS’s 2009 analysis of the dollar shortage to a stylized euro-area banking group in October 2008.

Consolidated position:

• Multiyear dollar loans and securities are held across the euro-area parent and its U.S. branch.
• Funding combines short-term dollar borrowing from U.S. money-market funds with five-year euro debt converted into dollars through one-week FX swaps.
• Intragroup claims are eliminated on consolidation. Including derivatives, the group’s net dollar exchange-rate exposure is close to zero.

Market disruption:

• Money-market funds withdraw lending, and FX-swap dealers are unwilling to renew most maturing dollar funding.
• Next week’s dollar repayments exceed available dollar cash. Prompt sales of the long-dated assets would require substantial discounts.

Which assessment most accurately identifies the group’s principal funding vulnerability and the central-bank response that targets it?

• A. A consolidated dollar maturity mismatch is exposed; the group can obtain dollar funding by drawing directly on a Federal Reserve central-bank swap line.
• B. A consolidated dollar maturity mismatch is exposed; the Eurosystem can supply dollar loans using dollars obtained by the ECB through a Federal Reserve swap line.
• C. A consolidated euro maturity mismatch is exposed; the Eurosystem can supply euro loans to replace the group’s lost dollar borrowing and maturing FX-swap funding.
• D. An unhedged dollar currency position is exposed; the Eurosystem can supply dollar loans using dollars obtained by the ECB through a Federal Reserve swap line.

Best answer: B

Explanation: The dollar shortage is a consolidated funding-liquidity problem, not simply an open foreign-exchange position. Long-dated dollar assets require continuing dollar funding even when their exchange-rate exposure is hedged. Short-term dollar borrowing and FX swaps both require renewal: five-year euro debt does not make a one-week FX swap a five-year dollar funding source.

When money-market lenders retreat and FX-swap renewal is disrupted, the group must obtain replacement dollars or sell assets at distressed prices. Eliminating intragroup claims prevents transfers between offices from being mistaken for an independent source of external dollar funding.

Federal Reserve swap lines provide dollars to foreign central banks. In the euro area, the ECB obtains dollars and the Eurosystem lends them to eligible banks. This response relieves immediate dollar rollover pressure, but does not itself eliminate longer-term funding or credit risk.

• A. Federal Reserve central-bank swap lines have foreign central banks as counterparties; the banking group receives dollars through central-bank lending rather than drawing directly on those lines.
• B. Dollar assets mature long after dollar borrowings and FX swaps; ECB-obtained dollars can be lent by the Eurosystem to replace disrupted short-term funding.
• C. The euro debt is long-term, while imminent repayments are in dollars; euro liquidity does not itself replace dollar funding when FX-swap renewal is disrupted.
• D. The near-zero net dollar exchange-rate exposure reflects the swaps’ hedge; the vulnerability arises from their short maturity rather than a missing currency hedge.

---

Question 56

Topic: Credit Risk Measurement and Management

A bank’s credit committee must select a workout strategy for a manufacturer that has missed two scheduled payments and breached a cash-flow covenant. A partial accounting write-off reduced the loan’s recorded principal from $12 million to $8 million, but did not extinguish collection rights over the $12 million claim.

Governance policy:

• Distressed-exposure reporting measures contractual claims under active workout, including written-off amounts still being pursued.
• The risk-appetite limit is $25 million. Other distressed claims total $16 million.
• The committee may authorize a temporary appetite breach during a workout. It must select the feasible strategy with the highest expected present value of recovery, net of any new funding.

Validated recovery estimates:

• Immediate collateral enforcement: collect $6.7 million today, with no additional funding.
• Restructuring: advance $0.5 million today; collect $8.5 million with 80% probability or $4.5 million with 20% probability in one year. These collections include repayment of the new advance.
• Retaining original terms: collect $8 million with 60% probability or $6 million with 40% probability in one year, with no additional funding.

All three strategies are feasible. Collections are net of collection costs, and no additional recoveries are expected. Use a 10% annual discount rate. No strategy has been implemented and no repayment has been received.

Which recommendation and current appetite classification should the committee adopt?

• A. Retain the original terms and report a current risk-appetite breach.
• B. Approve collateral enforcement and report current compliance with risk appetite.
• C. Approve collateral enforcement and report a current risk-appetite breach.
• D. Approve restructuring and report a current risk-appetite breach.

Best answer: C

Explanation: Credit governance connects early-warning indicators to committee oversight, accurate appetite reporting, and evidence-based remediation. An accounting write-off reduces the recorded balance but does not itself extinguish collection rights or satisfy a limit based on contractual claims.

Current distressed claims equal $16 million plus $12 million, or $28 million, exceeding the $25 million limit.

The recovery comparison, with amounts in millions of dollars, is:

• Immediate enforcement: 6.7.
• Restructuring: \( (0.80 \times 8.5 + 0.20 \times 4.5)/1.10 - 0.5 = 6.50 \).
• Original terms: \( (0.60 \times 8 + 0.40 \times 6)/1.10 \approx 6.55 \).

Enforcement therefore maximizes expected discounted net recovery. Restructuring’s larger expected collections do not compensate for its funding requirement and delayed receipts. The committee should report the current breach and oversee the recovery process, reassessing exposure when collections or legally effective changes actually alter the claims.

• A. Discounting the probability-weighted collections under existing terms produces approximately $6.55 million, below the immediate enforcement recovery.
• B. The reporting measure includes the full $12 million contractual claim, so the write-off does not reduce current distressed claims from $28 million to $24 million.
• C. Enforcement provides the highest net recovery value, $6.7 million, while current distressed claims of $28 million exceed the $25 million limit.
• D. After discounting collections and deducting the new advance, restructuring provides $6.5 million, below the $6.7 million available through enforcement.

---

Question 57

Topic: Credit Risk Measurement and Management

A risk manager proposes replacing a mortgage-backed security model’s constant 12% conditional prepayment rate (CPR) with a rate calibrated to the most recent month. Assume the observed single-month mortality rate (SMM) persists. Annualize CPR using 12 identical monthly survival factors.

Most recent month:

• Opening pool principal: $100.00 million.
• Scheduled principal payments: $2.00 million.
• Unscheduled prepaid principal: $1.96 million.

Next-month forecast assumptions:

• Opening pool principal: $96.04 million.
• Scheduled principal payments: $1.04 million under either prepayment assumption.
• Opening senior tranche principal: $2.50 million.
• Opening junior tranche principal: $93.54 million.
• All scheduled and prepaid principal pays the senior tranche until it is retired; remaining principal pays the junior tranche.

In each month, prepayments occur after scheduled principal payments. There are no defaults, losses, fees, or other principal adjustments.

Which revised CPR input and next-month junior principal payment correctly implement the proposed revision? Round CPR to two decimal places and principal payments to the nearest $0.01 million.

• A. CPR of 21.14%; junior principal payment of $0.40 million.
• B. CPR of 24.00%; junior principal payment of $0.69 million.
• C. CPR of 38.42%; junior principal payment of $2.30 million.
• D. CPR of 21.53%; junior principal payment of $0.44 million.

Best answer: D

Explanation: SMM measures unscheduled principal repayments relative to the balance remaining after scheduled principal payments. Thus, \( \mathrm{SMM}=1.96/(100-2)=0.02 \). Annual CPR compounds monthly survival: \( \mathrm{CPR}=1-(1-0.02)^{12}=21.53\% \).

Next month’s prepayment base is $96.04 million minus $1.04 million, or $95.00 million. At 2% SMM, prepaid principal is $1.90 million. Adding scheduled principal gives $2.94 million available for distribution. The senior tranche receives $2.50 million and is retired; the junior tranche receives the remaining $0.44 million.

The original 12% CPR implies approximately 1.06% SMM and $2.05 million of total principal next month, insufficient to retire the senior tranche or pay junior principal. The revised assumption therefore accelerates senior retirement and brings junior principal forward. This is prepayment contraction and cash-flow timing risk, not a default loss. The forecast remains conditional on the observed SMM persisting.

• A. This rate divides prepaid principal by the $100 million opening balance rather than the $98 million balance remaining after scheduled principal payments.
• B. Multiplying the 2% monthly mortality rate by 12 ignores survival compounding; entering 24% CPR overstates next month’s prepayments.
• C. This rate treats the $3.96 million of scheduled and prepaid principal as prepayments, overstating mortality and the junior tranche’s principal receipt.
• D. Prepaid principal divided by post-scheduled principal gives 2% SMM; the resulting forecast distributes $2.94 million, leaving $0.44 million after senior repayment.

---

Question 58

Topic: Liquidity and Treasury Risk Measurement and Management

A U.S. bank must invest a $100 million liquidity tranche in one security. This tranche is separate from its regulatory high-quality liquid asset buffer, and all four securities meet the bank’s credit limits.

Investment objective: Maximize annualized after-tax income, subject to obtaining at least $99 million from an immediate sale on day 30 under the bank’s specified stress scenario.

Tax and valuation assumptions: Apply a 25% effective tax rate to Treasury, agency, and commercial-paper income. Municipal interest is fully tax-exempt for this bank; ignore other tax effects. The quoted yields are annualized income yields on the initial investment. Stressed sale cash includes accrued income and is net of transaction costs and applicable taxes.

Available securities:

• Treasury bill: 90 days to maturity; 4.4% income yield; U.S. government obligation. Executable stressed sale cash: $99.7 million.
• Agency note: Two years to maturity; 5.2% income yield; AA+ credit rating, with no explicit government guarantee. Executable stressed sale cash: $98.8 million.
• Municipal bond: Nine months to maturity; 3.6% income yield; AA credit rating. Executable stressed sale cash: $99.2 million.
• Commercial paper: 180 days to maturity; 5.6% income yield; A-1+ short-term credit rating. Executable stressed sale cash: $98.4 million.

Which security should the treasury manager purchase for this tranche?

• A. Invest the tranche in the commercial paper.
• B. Invest the tranche in the agency note.
• C. Invest the tranche in the municipal bond.
• D. Invest the tranche in the Treasury bill.

Best answer: C

Explanation: A liquidity portfolio must be assessed using realizable cash under stress, not just credit ratings or headline yields. All four securities mature after day 30, so contractual repayment cannot meet the potential cash need; executable sale proceeds determine eligibility.

The Treasury bill and municipal bond satisfy the $99 million cash floor. For taxable income, the after-tax yield is \(y_{\text{after-tax}} = y(1-t)\), where \(y\) is the income yield and \(t\) is the effective tax rate. The Treasury bill therefore earns \(4.4\% \times 0.75 = 3.3\%\), while the municipal bond retains its 3.6% yield.

The municipal bond maximizes the specified after-tax income objective while preserving sufficient stressed liquidity. Credit approval alone does not establish suitability for a liquidity tranche.

• A. The commercial paper offers a 4.2% after-tax yield, but its stressed sale cash falls $0.6 million short of the liquidity requirement.
• B. The agency note offers a 3.9% after-tax yield, but its stressed sale cash falls $0.2 million short of the liquidity requirement.
• C. The municipal bond provides $99.2 million under stress, and its tax-exempt 3.6% yield exceeds the after-tax yield of the other qualifying investment.
• D. The Treasury bill meets the cash requirement, but its after-tax yield is only 3.3%, below the qualifying municipal bond’s 3.6%.

---

Question 59

Topic: Liquidity and Treasury Risk Measurement and Management

A bank is selecting a five-business-day liquidity stress scenario. All amounts below are in USD millions.

Funding exposures:

• Uninsured, nonoperational deposits of 100 are withdrawable within the horizon.
• Unsecured wholesale notes of 300 mature within the horizon; the portion not renewed must be repaid.
• Repo borrowing of 900 matures after the horizon and is secured by bonds currently worth 1,000 at a 10% haircut. Daily margining requires immediate cash repayment sufficient to satisfy the stressed haircut; additional collateral is not accepted.

Liquidity resources:

• Unrestricted cash of 80 remains after all essential operating payments.
• Unencumbered bonds currently worth 120 can be sold within the horizon. They experience the same percentage price decline as the pledged bonds.
• Ring-fenced cash of 100 is legally unavailable to the bank.
• A strategic equity holding worth 200 takes at least 20 business days to sell.
• A standby credit line of 150 cannot be drawn following the downgrade assumed in every scenario.

All specified withdrawals, nonrenewals, price declines, and haircut changes occur within the five days. Scenario haircuts are final levels, not increases. Bond sales incur no cost beyond the stated price decline. No funding is available beyond the specified note renewals.

Which scenario produces the largest cash shortfall by the end of the five-day horizon?

• A. Deposit withdrawals of 50%, unsecured-note nonrenewal of 100%, a repo haircut of 12%, and a bond-price decline of 5%.
• B. Deposit withdrawals of 70%, unsecured-note nonrenewal of 70%, a repo haircut of 15%, and a bond-price decline of 10%.
• C. Deposit withdrawals of 20%, unsecured-note nonrenewal of 40%, a repo haircut of 25%, and a bond-price decline of 20%.
• D. Deposit withdrawals of 90%, unsecured-note nonrenewal of 60%, a repo haircut of 12%, and a bond-price decline of 10%.

Best answer: C

Explanation: A funding stress scenario must be evaluated against exposure sizes and liquidity available within the specified horizon. Higher withdrawal percentages alone do not establish the most challenging scenario.

Repo borrowing capacity depends on both stressed collateral value and the final haircut. With a 20% bond-price decline and a 25% haircut, required cash repayment is, in USD millions:

\[ 900 - 1000 \times 0.80 \times (1 - 0.25) = 300. \]

Deposit withdrawals require \(100 \times 0.20 = 20\), and note nonrenewal requires \(300 \times 0.40 = 120\). Total cash outflows are therefore 440. Available liquidity is \(80 + 120 \times 0.80 = 176\), leaving a shortfall of 264.

Operating payments have already been deducted from cash. Restricted cash, the unavailable contingent credit line, and the slow strategic asset sale cannot support this horizon. The dominant vulnerability is the large repo exposure: falling bond values and rising haircuts simultaneously require repayment and reduce proceeds from selling unencumbered bonds.

• A. Full note nonrenewal requires $300 million, while repo repayment is $64 million; total outflows of $414 million exceed usable liquidity by $220 million.
• B. Repo repayment is $135 million, bringing total outflows to $415 million; usable liquidity of $188 million leaves a $227 million shortfall.
• C. Repo repayment requires $300 million, bringing total outflows to $440 million against $176 million of usable liquidity and producing the largest shortfall, $264 million.
• D. Outflows total $378 million against $188 million of usable liquidity, leaving a $190 million shortfall; the severe deposit shock affects a relatively small funding base.

---

Question 60

Topic: Market Risk Measurement and Management

A market-risk validator is assessing 99% one-day historical VaR estimates for two trading portfolios. For limit reporting, the risk committee requires an approximate two-sided 95% confidence interval with a half-width no greater than $2 million. The assessment concerns sampling precision, not overall model validity.

Dependence evidence:

• Portfolio A exhibits serial clustering of tail losses. A dependence-preserving block bootstrap estimates its VaR standard error at $1.30 million.
• Portfolio B’s observations may be treated as independent and identically distributed for this assessment.

Use normal-approximation confidence intervals with a multiplier of 1.96, treating the estimated quantile densities as fixed.

Which conclusion should the validator report?

• A. Both portfolios meet the precision requirement; the confidence-interval half-widths are approximately $0.59 million for A and $1.12 million for B.
• B. Both portfolios meet the precision requirement; the confidence-interval half-widths are approximately $1.95 million for A and $1.54 million for B.
• C. Only Portfolio B meets the precision requirement; the confidence-interval half-widths are approximately $2.55 million for A and $1.54 million for B.
• D. Neither portfolio meets the precision requirement; the confidence-interval half-widths are approximately $5.10 million for A and $3.08 million for B.

Best answer: C

Explanation: VaR estimation precision concerns uncertainty in a loss quantile, not the volatility of daily portfolio losses. For independent observations, the asymptotic quantile standard error is

\[ \mathrm{SE}(\widehat{q}_p) \approx \frac{\sqrt{p(1-p)/n}}{f(q_p)} \]

Here, \(n\) is sample size, \(p\) is the quantile probability, and \(f(q_p)\) is the density at that quantile. Measuring losses in millions of dollars, Portfolio B’s standard error is \(\sqrt{0.99\times0.01/1{,}000}/0.004 \approx 0.787\). Its 95% confidence-interval half-width is therefore approximately $1.54 million.

Portfolio A’s serial dependence makes the independent-sample calculation inappropriate. Its block-bootstrap standard error gives a half-width of \(1.96\times1.30=2.548\), or approximately $2.55 million. Only Portfolio B meets the $2 million requirement.

The 99% VaR level and 95% estimation-confidence level describe different probabilities. Moreover, Portfolio B has only about 10 observations beyond its estimated 99% quantile, so the normal approximation remains potentially fragile despite satisfying the numerical precision threshold.

• A. These values use daily-loss standard deviation divided by the square root of sample size, which measures uncertainty in a mean rather than a quantile.
• B. The $1.95 million half-width applies an independence-based quantile formula to A, ignoring the serial clustering reflected in its bootstrap standard error.
• C. A’s dependence-aware bootstrap error gives a $2.55 million half-width, while B’s independent-sample quantile error gives $1.54 million, so only B meets the limit.
• D. These amounts are full confidence-interval widths, whereas the committee’s $2 million threshold applies to half-widths.

---

Question 61

Topic: Liquidity and Treasury Risk Measurement and Management

A bank enters a repo against securities initially worth $100.00 million, with a 5% haircut. The agreement defines the haircut as \(h=(V-L)/V\), where \(V\) is collateral market value and \(L\) is the cash loan principal. The principal has not changed; ignore interest and fees.

Margin call facts:

• The original securities now have a market value of $92.00 million.
• The lender increases the required haircut to 8%.
• The bank can post additional eligible securities worth $10.00 million, subject to the same haircut.

Treasury will post all additional securities and repay only the minimum cash needed to restore the required haircut.

Liquidation stress: Immediately after the margin requirement is restored, the bank defaults. Net liquidation proceeds equal 90% of the aggregate collateral market value used for the margin call. No other recoveries are available.

Which assessment correctly gives the required cash repayment and the lender’s resulting liquidation shortfall, rounded to the nearest $0.01 million?

• A. Repay $0.36 million; the lender’s liquidation shortfall is $2.84 million.
• B. Repay $1.16 million; the lender’s liquidation shortfall is $2.04 million.
• C. Repay $0.56 million; the lender’s liquidation shortfall is $2.64 million.
• D. Repay $1.16 million; the lender’s liquidation shortfall is $0.00 million.

Best answer: B

Explanation: With collateral value as the denominator, borrowing capacity is \(L=(1-h)V\). The original loan is \(100.00\times0.95=95.00\) million dollars.

After the additional securities are posted, aggregate collateral value is $102.00 million. The revised 8% haircut permits borrowing of \(102.00\times0.92=93.84\) million dollars. Treasury must therefore repay $1.16 million to restore the required ratio.

Stress liquidation produces \(102.00\times0.90=91.80\) million dollars. Against the remaining $93.84 million loan, this leaves a $2.04 million shortfall.

A haircut provides a buffer against adverse collateral price movements and liquidation costs. It does not guarantee a liquidation price or eliminate secured-credit exposure. Here, the 10% liquidation discount exceeds the 8% haircut, so the lender remains exposed despite full compliance with the margin requirement.

• A. Crediting the additional collateral dollar-for-dollar ignores its haircut; it contributes only $9.20 million toward the $10.36 million margin deficit.
• B. The haircut permits a $93.84 million remaining loan, requiring a $1.16 million repayment; net proceeds of $91.80 million leave $2.04 million uncovered.
• C. These amounts use borrowing capacity of \(102/1.08\), incorrectly treating the haircut as a percentage of the loan rather than collateral value.
• D. The repayment restores the contractual haircut, but the 10% liquidation discount exceeds the 8% buffer, leaving part of the loan uncovered.

---

Question 62

Topic: Market Risk Measurement and Management

A bank reviews the $12.0 million market-risk economic-capital allocation for a corporate-bond trading desk hedged with a liquid credit index.

Validation findings:

• The hedge leaves little common credit-spread sensitivity, but material bond-index basis sensitivity remains.
• A one-day 99% VaR backtest passes. Independent repricing confirms that basis movements, rather than pricing errors, drive the largest losses.
• The current ES calculation assigns every factor a 10-business-day liquidity horizon. Stressed execution analysis supports 10 days for curve and index factors, but 40 days for the basis factor.

Allocation policy: The minimum allocation equals liquidity-horizon-adjusted 97.5% ES, using Basel’s horizon-increment aggregation with a 10-business-day base horizon. No multiplier or separate capital add-on enters this internal calculation.

Validation produces the following correctly calibrated 10-business-day ES estimates. Other factors are held fixed in the basis-only rerun.

Which minimum allocation and capital action follow from the policy and validation evidence?

• A. The minimum allocation is approximately $16.0 million; increase the existing allocation to that amount.
• B. The minimum allocation is approximately $8.0 million; retain the existing allocation of $12.0 million.
• C. The minimum allocation is approximately $14.4 million; increase the existing allocation to that amount.
• D. The minimum allocation is approximately $13.1 million; increase the existing allocation to that amount.

Best answer: D

Explanation: A successful short-horizon VaR backtest does not establish that capital captures losses over a stressed liquidation horizon. The index hedge reduces common credit-spread exposure, but the material basis exposure remains for 40 business days.

Liquidity-horizon aggregation retains the all-factor ten-day ES and adds contributions for longer-lived risk factors over their additional horizon. With ES inputs expressed in millions of dollars:

\[ ES_{LH} = \sqrt{8^2 + 6^2 \times \frac{40-10}{10}} = \sqrt{172} \approx 13.1. \]

The all-factor estimate already includes basis risk during the first ten days. Consequently, the additional basis term covers only the subsequent 30 days. Scaling every factor to 40 days would also overstate the validated horizons of curve and index risk.

The existing $12.0 million allocation falls approximately $1.1 million below the policy minimum. Increasing it addresses the liquidity-horizon gap identified by validation.

• A. Multiplying the all-factor ES by \(\sqrt{40/10}\) incorrectly extends curve and index risk to 40 days, although their validated horizons remain 10 days.
• B. The ten-day ES omits the additional basis exposure during stressed liquidation; passing a one-day VaR backtest does not validate that liquidity-horizon assumption.
• C. This adds a full 40-day basis contribution to the all-factor ES, double-counting the first 10 days already included in that estimate.
• D. Horizon-increment aggregation gives \(\sqrt{8^2 + 6^2 \times 3}\), or approximately $13.1 million, exceeding the existing $12.0 million allocation.

---

Question 63

Topic: Risk Management and Investment Management

A risk manager reviews a multi-asset fund whose investor mandate allows at most 20% of NAV in private credit, measured on a look-through basis. Investments in pooled vehicles are permitted.

Current holdings:

• Direct private-credit investments: 16% of fund NAV.
• An unleveraged feeder fund: 10% of fund NAV. The feeder invests 40% of its NAV in private credit and 60% in traded government bonds, and permits quarterly redemptions.
• Remaining holdings: listed securities and cash, with no private-credit exposure.

Committee pack:

After replacing some listed bonds with additional direct private-credit holdings, the pack reports private-credit exposure of 16% and a decline in one-month 99% historical VaR. Private-credit NAVs are independently reviewed quarterly and carried forward between valuation dates. The feeder’s monthly NAV incorporates these values. VaR uses the reported monthly NAV returns without an adjustment for appraisal smoothing.

The committee accepts the results without challenging these assumptions and approves this investor statement:

“The portfolio remains within its mandate, and the new allocation has reduced investment risk.”

Which assessment of mandate compliance and the investor risk statement is most defensible?

• A. The mandate is met at 20% private-credit exposure, and the lower VaR does not substantiate the investor statement’s claim of lower investment risk.
• B. The mandate is breached at 26% private-credit exposure, and the lower VaR does not substantiate the investor statement’s claim of lower investment risk.
• C. The mandate is met at 20% private-credit exposure, and the lower VaR substantiates the investor statement’s claim of lower investment risk.
• D. The mandate is met at 16% private-credit exposure, and the lower VaR does not substantiate the investor statement’s claim of lower investment risk.

Best answer: A

Explanation: Look-through limits apply to underlying exposures rather than only direct holdings or entire legal vehicles. The feeder contributes \(10\% \times 40\% = 4\%\) of fund NAV in private credit. Total exposure is therefore \(16\% + 4\% = 20\%\), meeting the mandate exactly. Nevertheless, reporting only 16% gives the committee and investors an inaccurate picture of remaining headroom.

Quarterly appraisals carried forward between valuation dates can smooth monthly returns and suppress measured volatility and tail losses. Independent review of valuation levels does not eliminate this effect. A lower historical VaR therefore does not establish that allocating more to illiquid credit reduced overall investment risk. The consequential governance weaknesses are incomplete exposure reporting and inadequate challenge of the measurement assumptions, resulting in an unsupported investor risk statement.

• A. Look-through exposure reaches the permitted maximum, while unchallenged appraisal smoothing means the VaR decline may reflect valuation practices rather than reduced economic risk.
• B. Counting the entire feeder as private credit incorrectly includes its government-bond holdings; its private-credit contribution is only 4% of fund NAV.
• C. Independent valuation review does not remove appraisal smoothing, so lower VaR based on these returns cannot establish that overall investment risk has declined.
• D. The 16% figure omits private credit held through the feeder, understating exposure by 4% of fund NAV and overstating mandate headroom.

---

Question 64

Topic: Operational Risk and Resilience

A bank incurs compensation charges when manually repaired payment instructions are released after their currency-specific settlement cutoffs. At 13:00, the operations manager needs a normalized early-warning KRI to decide whether to redeploy repair staff before today’s cutoffs.

Operating facts:

• Payment submissions have doubled to 100,000 compared with yesterday’s 13:00 count, entirely because of straight-through payment growth. Outstanding manual-repair cases remain unchanged at 100.
• The general queue-aging alert threshold is 60 minutes.
• Estimated remaining clearance times include waiting, repair and independent release checks under current staffing.
• All manually repaired payments met their cutoffs yesterday.

13:00 repair-queue snapshot:

Which KRI best supports the manager’s decision?

• A. The percentage of today’s submitted payment instructions that have been routed to manual-repair queues.
• B. The percentage of yesterday’s completed manual-repair cases that were released after their respective settlement cutoffs.
• C. The percentage of outstanding repair cases whose elapsed queue time exceeds the 60-minute aging-alert threshold.
• D. The percentage of outstanding repair cases whose estimated clearance time exceeds the time remaining to their settlement cutoff.

Best answer: D

Explanation: An effective early-warning KRI links the monitored exposure to the specific loss-generating pathway. Here, the relevant threshold is whether remaining repair and verification can finish within the available settlement window, rather than whether queue age exceeds a general limit.

Group X requires an estimated 90 minutes but has only 45 minutes before cutoff. Group Y requires 10 minutes and has 180 minutes available. The current at-risk proportion is therefore \(40/100 = 40\%\).

Using outstanding manual-repair cases as the denominator keeps the measure focused on the exposed workflow rather than unrelated growth in straight-through payments. A calibrated escalation limit can translate this proportion into a staffing decision. The indicator signals projected cutoff risk under current staffing; it does not establish that the flagged payments will inevitably miss settlement.

• A. Straight-through growth dilutes this ratio, while routing frequency does not measure whether pending repairs can finish before their cutoffs.
• B. Yesterday’s zero breach rate describes realized performance and cannot identify today’s cases projected to overrun their cutoff windows.
• C. This flags the 60 older cases, although they need only 10 minutes to clear and have 180 minutes remaining before cutoff.
• D. This flags the 40 cases needing 90 minutes with only 45 minutes remaining, giving a 40% at-risk share under current staffing.

---

Question 65

Topic: Market Risk Measurement and Management

A market-risk team is valuing a default-free, three-year callable zero-coupon note.

Contract terms:

• The holder receives $100 at year 3 if the note remains outstanding.
• The issuer may repay $94 immediately at year 1 or $97 immediately at year 2, terminating the note.
• The issuer exercises optimally, and there are no other payments or path-dependent terms.

Annual pricing tree: Each branch has risk-neutral probability 0.5. Each displayed annual effective short rate discounts the following one-year interval. The tree is calibrated to the current discount curve.

Refinement review: The team plans to test a half-year grid for lower discretization error. The refined model will retain the call dates, use transition rules determined only by date and current short rate, and be recalibrated to the current discount curve. The committee requires the lowest computational cost without discarding information relevant to valuation.

Which annual-tree value and implementation of the refined grid should the committee approve?

• A. Value the note at $88.32; use a half-year recombining lattice that merges identical time-rate states.
• B. Value the note at $88.00; use a half-year nonrecombining lattice that retains every rate history.
• C. Value the note at $88.32; use a half-year nonrecombining lattice that retains every rate history.
• D. Value the note at $88.00; use a half-year recombining lattice that merges identical time-rate states.

Best answer: D

Explanation: At each call date, an issuer-callable note is worth the smaller of immediate repayment and its risk-neutral discounted continuation value.

Working backward, year-2 values at rates of 8%, 4%, and 0% are $92.59, $96.15, and $97.00, respectively. The final value is capped by the $97 call payment. Year-1 continuation values are $89.03 and $94.68; the $94 call payment caps the latter. The current value is therefore:

\[ V_0 = \frac{0.5(89.031339 + 94)}{1.04} \approx 88.00 \]

Recombination preserves valuation because surviving paths reaching the same date and short rate have identical future transition rules, cash flows, and exercise opportunities. A standard recombining binomial tree has total node growth of order \(n^2\), versus order \(2^n\) when every history is retained, where \(n\) is the number of intervals.

Smaller time steps can reduce discretization error but increase computational work and require renewed calibration and validation. Additional nodes provide numerical resolution, not new market information, and do not automatically improve economic accuracy.

• A. The $88.32 value applies the year-2 call limit but omits the year-1 call, where repayment of $94 is below continuation value of $94.68.
• B. The valuation reflects both call dates, but retaining every history adds unnecessary computation because cash flows and call decisions depend only on date and current rate.
• C. Omitting the binding year-1 call overstates value, while separate copies of identical time-rate states provide no additional valuation information for this contract.
• D. Backward induction gives $88.00, and merging identical time-rate states preserves continuation values and optimal call decisions with lower computational cost.

---

Question 66

Topic: Market Risk Measurement and Management

A trading desk proposes replacing its investment-grade sovereign credit-spread exposure with investment-grade corporate credit-spread exposure, while leaving its USD interest-rate exposure unchanged. All risk factors are modellable, and there are no other exposures.

Assigned FRTB liquidity horizons:

• USD interest rates: 10 days.
• Investment-grade sovereign credit spreads: 20 days.
• Investment-grade corporate credit spreads: 40 days.

Risk estimates: For both the existing and proposed portfolios, 10-day, 97.5% expected shortfall is $10.00 million with all factors varying and $6.00 million with only credit-spread factors varying.

Aggregation convention: For each threshold \(L_j\), let \(E_j\) be the 10-day expected shortfall with only factors whose assigned liquidity horizon is at least \(L_j\) varying. All remaining factors are held constant. Use \((L_1,L_2,L_3,L_4,L_5)=(10,20,40,60,120)\) days and \(L_0=0\):

\[ \mathrm{ES}_{LH}=\sqrt{\sum_{j=1}^{5}E_j^2\frac{L_j-L_{j-1}}{10}} \]

Approval rule: The capital-allocation committee approves the replacement if liquidity-adjusted expected shortfall does not exceed $14.00 million and rejects it otherwise. Stress calibration, capital multipliers, and other capital components are outside this review.

Which assessment of the proposed replacement is correct?

• A. Reject the replacement: liquidity-adjusted expected shortfall is $20.00 million, above the $14.00 million limit.
• B. Approve the replacement: liquidity-adjusted expected shortfall is $11.66 million, below the $14.00 million limit.
• C. Approve the replacement: liquidity-adjusted expected shortfall is $13.11 million, below the $14.00 million limit.
• D. Reject the replacement: liquidity-adjusted expected shortfall is $14.42 million, above the $14.00 million limit.

Best answer: D

Explanation: FRTB liquidity-horizon adjustment uses nested risk-factor subsets rather than scaling the entire portfolio to its longest horizon. Longer horizons extend the contributions of less liquid factors while preserving the shorter horizons assigned to other factors.

For the proposed portfolio, the 10-day threshold includes all factors, so its expected shortfall is $10.00 million. The 20-day and 40-day thresholds include only corporate spread factors, each with expected shortfall of $6.00 million. No factors remain at the 60-day or 120-day thresholds.

With amounts measured in millions of dollars:

\[ \mathrm{ES}_{LH}=\sqrt{10^2+6^2\frac{20-10}{10}+6^2\frac{40-20}{10}}=\sqrt{208}\approx14.42 \]

The existing portfolio’s adjusted expected shortfall is \(\sqrt{10^2+6^2}\approx11.66\) million. Thus, unchanged base expected shortfall does not imply unchanged liquidity-adjusted risk. Moving the spread exposure to a longer horizon raises the screening measure above $14.00 million, requiring rejection. This measure is not the bank’s total regulatory capital charge.

• A. Scaling the entire $10.00 million expected shortfall by \(\sqrt{40/10}\) incorrectly assigns a 40-day horizon to the USD interest-rate factors.
• B. This retains the sovereign exposure’s 20-day horizon and omits the additional 20-to-40-day contribution required for the corporate spread factors.
• C. This includes the 20-to-40-day contribution but omits the 10-to-20-day contribution, even though corporate spread factors belong to both nested subsets.
• D. The corporate spread factors contribute at both the 20-day and 40-day thresholds, producing \(\sqrt{100+36+72}=14.42\) million.

---

Question 67

Topic: Operational Risk and Resilience

A bank is considering a cash dividend while keeping its risk exposures unchanged. The board requires available risk-bearing resources to cover both internal economic capital and the unchanged board-approved risk-capital allocation. Eligible regulatory capital must also remain at or above the prudential minimum.

The bank uses a one-year, 99.9% loss-quantile approach to economic capital. Expected losses are fully provisioned, and available capital figures are net of those provisions.

Capital report: All amounts are in millions of dollars.

The board’s allocation reflects business units’ contributions to diversified economic capital plus management buffers. A dividend reduces available risk-bearing resources and eligible regulatory capital dollar for dollar. Ignore other payout restrictions.

What is the maximum dividend permitted by these capital tests?

• A. The maximum dividend is $15 million, with the board-approved risk-capital allocation as the binding requirement.
• B. The maximum dividend is $30 million, with the internal economic capital estimate as the binding requirement.
• C. The maximum dividend is $10 million, with the modeled total loss quantile as the binding requirement.
• D. The maximum dividend is $20 million, with the stated prudential capital minimum as the binding requirement.

Best answer: A

Explanation: Economic capital is an internal model estimate of capital needed for unexpected losses. Here it equals the $220 million loss quantile minus $20 million of provisioned expected losses, or $200 million.

Risk-capital allocation assigns risk-bearing capacity to business units. The board uses diversified risk contributions plus management buffers, producing a $215 million allocation that must remain funded. This allocation does not create additional capital. Regulatory capital instead depends on prudential eligibility rules and the applicable minimum requirement.

The dividend limits are $30 million against economic capital, $15 million against the board allocation, and $20 million against the regulatory minimum. The smallest limit governs. A $15 million dividend leaves $215 million of risk-bearing resources and $195 million of eligible regulatory capital, satisfying all three tests.

• A. The board allocation permits a $15 million payout, compared with $30 million under the economic-capital test and $20 million under the regulatory test.
• B. A $30 million payout would leave resources below the board allocation and eligible regulatory capital below the prudential minimum.
• C. Using the total loss quantile includes expected losses already covered by provisions; economic capital is $200 million rather than $220 million.
• D. Regulatory headroom is $20 million, but this payout would leave $210 million of risk-bearing resources against the fixed $215 million board allocation.

---

Question 68

Topic: Risk Management and Investment Management

An investment committee is reviewing an equally weighted allocation to two equity funds. Both funds have a market beta of 1.0. The committee’s skill-based sleeve requires strictly positive expected alpha after fees relative to its market-plus-value benchmark.

Annual inputs:

• Risk-free rate: 3.0%.
• Expected market risk premium: 5.0%.
• Expected premium on a zero-investment value factor: 4.0%.
• Fee drag on the combined allocation: 0.5 percentage points, with factor exposures unchanged.

Fund-specific shocks have positive variances, are mutually uncorrelated, and are uncorrelated with both factors. For this review, use the stated forecasts, premiums, and exposures as point estimates, not guaranteed outcomes.

Which assessment of the proposed allocation is correct under these inputs?

• A. The allocation does not qualify for the skill sleeve: net factor-adjusted alpha is 0.0%, and combining the funds eliminates value-factor exposure.
• B. The allocation does not qualify for the skill sleeve: net factor-adjusted alpha is 0.0%, and combining the funds leaves positive value-factor exposure.
• C. The allocation qualifies for the skill sleeve: net factor-adjusted alpha is 0.8%, and combining the funds leaves positive value-factor exposure.
• D. The allocation qualifies for the skill sleeve: net factor-adjusted alpha is 0.5%, and combining the funds leaves positive value-factor exposure.

Best answer: B

Explanation: CAPM’s expected-return restriction assumes a mean-variance-efficient market portfolio. A return above a market-beta-only benchmark can instead reflect omitted systematic risk.

The equally weighted allocation has a gross expected return of 9.3% and a net expected return of 8.8%. Its market beta remains 1.0, and its value loading is \(0.5(0.8)+0.5(-0.4)=0.2\). The market-plus-value benchmark therefore requires:

\[ 3.0\% + 1.0(5.0\%) + 0.2(4.0\%) = 8.8\% \]

Net factor-adjusted alpha is zero, so the allocation fails the sleeve’s strictly positive alpha requirement. Its 0.8% net CAPM alpha represents compensation for value-factor exposure rather than unexplained performance.

Uncorrelated fund-specific shocks provide idiosyncratic diversification. Opposite value loadings also partially offset, but their unequal magnitudes leave systematic value risk. These conclusions depend on the benchmark and point estimates: positive CAPM alpha alone does not establish skill or exploitable mispricing, and zero multifactor alpha does not prove market efficiency.

• A. Equal weights on loadings of 0.8 and -0.4 produce a loading of 0.2, so the opposite exposures only partially offset.
• B. The 8.8% net forecast matches the 8.8% factor benchmark, while equal weighting leaves a positive value loading of 0.2.
• C. The 0.8% is net CAPM alpha, fully explained by the allocation’s 0.2 value loading multiplied by the 4.0% factor premium.
• D. The 0.5% is gross factor-adjusted alpha; subtracting the annual fee drag of 0.5 percentage points reduces net alpha to zero.

---

Question 69

Topic: Credit Risk Measurement and Management

A bank is buying protection on a new one-year CDS. Its desk proposes paying $40,110 upfront. A settlement analyst may approve only a payment that makes the bank’s net value zero at inception.

Contract terms:

• Notional: $10,000,000; recovery rate: 40%.
• Fixed annual coupon: 600 basis points, paid semiannually using a 0.5-year accrual fraction.
• Scheduled premiums are payable only if the reference entity survives to the payment date. Accrued premium is payable on default.
• There is no accrued premium at inception. Ignore counterparty risk and transaction costs.

Risk-neutral valuation inputs:

Survival probability at inception is 1.

For valuation, assume defaults within each six-month interval occur at its midpoint. The discount factors at 0.25 and 0.75 years are 0.9950 and 0.9850, respectively. Both the protection payment and accrued premium are settled at the assumed default time.

Which upfront payment should the analyst approve?

• A. The buyer pays the seller $22,365 at inception.
• B. The seller pays the buyer $25,290 at inception.
• C. The buyer pays the seller $40,110 at inception.
• D. The buyer pays the seller $25,290 at inception.

Best answer: D

Explanation: From the protection buyer’s perspective, fair upfront payment equals the present value of protection minus the present value of all premiums. The interval default probabilities are \(1-0.97=0.03\) and \(0.97-0.90=0.07\).

The discounted default-probability sum is \(0.03\times0.995+0.07\times0.985=0.0988\). Therefore:

• Protection value: \(10{,}000{,}000\times0.60\times0.0988\) = $592,800.
• Scheduled premium value: \(10{,}000{,}000\times0.06\times0.5\times(0.97\times0.99+0.90\times0.98)\) = $552,690.
• Accrued premium value: \(10{,}000{,}000\times0.06\times0.25\times0.0988\) = $14,820.

Midpoint defaults imply one-quarter year of accrued premium in either interval. Total premium value is $567,510, leaving a positive protection-buyer value of $25,290 before upfront settlement. The buyer pays that amount to make the initial net value zero.

• A. This amount discounts protection and accrued-premium settlements at the semiannual payment dates rather than the stipulated default midpoints.
• B. The magnitude includes accrued premiums, but the direction is reversed: protection exceeds premium value, so the buyer must pay upfront.
• C. This amount subtracts only scheduled premiums from protection value, omitting the $14,820 present value of accrued premiums payable on default.
• D. Protection has a present value of $592,800 and total premiums of $567,510, requiring the buyer to pay the $25,290 difference.

---

Question 70

Topic: Credit Risk Measurement and Management

A bank is reassessing the next-12-month default risk of two corporate borrowers during a recession.

Credit-risk evidence:

• Both external issuer ratings remain BBB. The agency evaluates creditworthiness under normalized economic conditions. The bank maps BBB to a long-run average annual issuer default rate of 1.2%.
• A validated internal model maps granular scores to physical 12-month issuer default probabilities using current borrower financials and macroeconomic conditions. Higher scores indicate lower risk. Both borrowers scored 720 before the recession.

Bond update: Alpha’s bond issue rating was upgraded after additional collateral increased expected recovery in default. Alpha’s debt-service capacity and issuer rating were unchanged.

Which interpretation is most consistent with these systems for the bank’s next-12-month issuer default-risk assessment?

• A. Alpha’s higher next-year PD is compatible with the common BBB issuer grade; internal scores capture current conditions, while Alpha’s bond upgrade indicates lower loss given default.
• B. Alpha’s lower next-year PD is implied by its improved bond grade; internal scores capture cyclical stress, while the common BBB issuer grade reflects normalized conditions.
• C. Different next-year PDs indicate inconsistent calibration with the common BBB issuer grade; both systems reference annual defaults, while Alpha’s bond upgrade indicates lower loss given default.
• D. Equal next-year PDs are implied by the common BBB issuer grade; internal scores provide finer rank ordering, while Alpha’s bond upgrade indicates lower loss given default.

Best answer: A

Explanation: Point-in-time systems measure credit risk conditional on current borrower and macroeconomic conditions. Through-the-cycle systems emphasize normalized creditworthiness and generally show less cyclical rating migration. The distinction concerns economic conditioning, not simply prediction horizon: a long-run average annual default rate and a current 12-month PD both concern one-year defaults.

Here, the internal scores imply PDs of 3.6% for Alpha and 1.8% for Beta, even though both retain a BBB issuer rating. Granular scores can distinguish borrowers within a broad rating grade and, when calibrated as specified, produce default probabilities rather than only ordinal rankings.

Issuer ratings assess issuer-level creditworthiness. Issue ratings can additionally reflect collateral, seniority, and expected recovery. Alpha’s collateral-driven bond upgrade therefore indicates lower loss given default, not a reduction in its issuer default probability.

• A. The score model conditions on current information, whereas the issuer rating uses normalized conditions; the collateral-driven bond upgrade improves recovery rather than issuer default propensity.
• B. An issue-rating upgrade driven by collateral reflects loss severity after default, so it does not reverse the issuer PD ranking indicated by the internal model.
• C. A shared one-year measurement horizon does not equate a long-run average default rate with a default forecast conditioned on the current recession.
• D. A cycle-normalized BBB grade and its long-run annual default rate do not require equal conditional 12-month PDs for borrowers with different current risk profiles.

---

Question 71

Topic: Market Risk Measurement and Management

A market risk manager reviews a delta-hedged option book after a change in the implied-volatility curve. The desk’s limit is $350,000 for 99% one-day vega VaR.

Position and volatility data:

One volatility point is one percentage point of implied volatility. Use fixed first-order vegas for revaluation and VaR, and ignore all other P&L sources.

Next-day risk estimates:

• Changes in the one-month and six-month implied volatilities are jointly normal with zero means and daily standard deviations of 0.60 and 0.40 volatility points, respectively.
• Their correlation has declined from 0.90 to 0.25, while the daily standard deviations remain unchanged. Use 0.25 for the updated VaR.
• Use a 99% standard normal quantile of 2.33.

Which assessment correctly reports the approximate revaluation P&L and updated VaR limit status, with amounts rounded to the nearest $1,000?

• A. A $500,000 revaluation loss and a $316,000 VaR, within the limit.
• B. A $250,000 revaluation gain and a $316,000 VaR, within the limit.
• C. A $500,000 revaluation loss and a $406,000 VaR, above the limit.
• D. A $250,000 revaluation gain and a $406,000 VaR, above the limit.

Best answer: A

Explanation: Tenor-specific vega exposures determine revaluation; net vega captures only a parallel volatility shift. The one-month and six-month implied volatilities rise by five and two points, so P&L is \( -200{,}000(5) + 250{,}000(2) = -500{,}000 \) dollars.

For next-day risk, use the covariance of changes in implied volatility, not the implied-volatility levels. The vega-weighted daily shock magnitudes are $120,000 and $100,000. With opposite-signed vegas, the covariance term is negative. Daily P&L standard deviation, \( \sigma_P \), is:

\[ \sigma_P = \sqrt{120{,}000^2 + 100{,}000^2 - 2(0.25)(120{,}000)(100{,}000)} \approx 135{,}647. \]

Thus, \( \mathrm{VaR}_{0.99} = 2.33\sigma_P \) is approximately $316,000, using 90.3% of the $350,000 limit. The correlation decline weakens the hedge between the short and long volatility exposures, raising VaR despite unchanged daily volatility-change standard deviations.

• A. The five-point and two-point volatility increases cause a $500,000 loss; the negative covariance term gives approximately $316,000 VaR, below $350,000.
• B. The VaR incorporates the signed exposures correctly, but the $250,000 gain assumes a five-point parallel shift rather than the displayed term-structure change.
• C. The revaluation loss reflects the tenor-specific moves; $406,000 VaR incorrectly makes the covariance term positive despite the opposite-signed vegas.
• D. The gain treats the volatility shift as parallel, while the VaR calculation incorrectly uses a positive covariance term for opposite-signed exposures.

---

Question 72

Topic: Operational Risk and Resilience

A bank’s capital committee is evaluating a loan-portfolio transaction intended to achieve a 12% CET1-to-RWA target without raising capital.

Capital assumptions:

• CET1 capital is $15.60 billion and remains unchanged by the transaction.
• The bank’s applicable 2026 implementation specifies a transitional output-floor percentage of 65%.
• The bank uses approved internal models for credit and market risk and the standardized approach for operational risk.

Current RWA: All components are shown below, in USD billions.

The transaction reduces credit-risk RWA by $10 billion under the approved internal models and by $20 billion under the standardized approach. All other RWA remain unchanged.

Which assessment correctly reports the post-transaction binding RWA and CET1 surplus or shortfall against the committee’s target?

• A. Binding RWA is $137.75 billion, with a CET1 shortfall of $0.93 billion against the target.
• B. Binding RWA is $134.50 billion, with a CET1 shortfall of $0.54 billion against the target.
• C. Binding RWA is $110.00 billion, with a CET1 surplus of $2.40 billion against the target.
• D. Binding RWA is $123.50 billion, with a CET1 surplus of $0.78 billion against the target.

Best answer: D

Explanation: The Basel output floor applies to aggregate RWA. Binding RWA equals the greater of total RWA under the bank’s approved approaches and the applicable floor percentage of total standardized RWA.

After the transaction, approved-method RWA is \(60 + 30 + 20 = 110\) billion, while standardized RWA is \(130 + 40 + 20 = 190\) billion. Therefore:

\[ \mathrm{RWA}_{\mathrm{binding}} = \max(110, 0.65 \times 190) = 123.50\text{ billion} \]

The committee’s target requires CET1 capital of \(0.12 \times 123.50 = 14.82\) billion. Available CET1 of $15.60 billion exceeds this amount by $0.78 billion, so the transaction achieves the target without additional capital.

The floor is not a separate minimum for every risk category or individual exposure. The applicable transitional percentage, rather than the fully phased-in percentage, governs this calculation.

• A. This applies the fully phased-in 72.5% floor to $190 billion, rather than the applicable 65% transitional percentage.
• B. This separately floors each risk category, producing $84.50 billion plus $30 billion plus $20 billion, rather than applying the floor to aggregate RWA.
• C. This uses only approved-method RWA of $110 billion and omits the binding aggregate output floor of $123.50 billion.
• D. The 65% floor on $190 billion of standardized RWA binds, requiring $14.82 billion of CET1 and leaving a $0.78 billion surplus.

---

Question 73

Topic: Liquidity and Treasury Risk Measurement and Management

A bank is revising the liquidity transfer price for two undrawn credit-line portfolios. Each portfolio has $400 million in undrawn commitments. Over the next 30 days, portfolio A draws either $120 million or nothing, while portfolio B draws either $180 million or nothing.

Joint drawdown forecasts:

Treasury pricing policy:

• Continuously maintain the smallest liquidity buffer that covers aggregate drawdowns in at least 95% of scenarios. No other funding sources are credited toward this requirement.
• Fund the buffer at an annual cost of 3.0% and invest it at an annual yield of 1.8%.
• Recover the buffer’s full annual net carry through a uniform fee on total undrawn commitments. Funding charges after facilities are drawn are assessed separately.

Which revised buffer and annual contingent-liquidity fee should treasury adopt?

• A. Use a $30 million buffer and charge 4.5 bp annually on total undrawn commitments.
• B. Use a $180 million buffer and charge 27.0 bp annually on total undrawn commitments.
• C. Use a $270 million buffer and charge 40.5 bp annually on total undrawn commitments.
• D. Use a $300 million buffer and charge 45.0 bp annually on total undrawn commitments.

Best answer: D

Explanation: Contingent-liquidity pricing must reflect simultaneous drawdowns and the cost of maintaining resources against them. Under the revised forecast, a $180 million buffer covers only 92% of scenarios. The smallest buffer meeting the 95% target is therefore $300 million.

Annual net carry is 3.0% minus 1.8%, or 1.2%. Holding $300 million consequently costs $3.6 million annually. With $800 million of total undrawn commitments, the annual fee is:

\[ \frac{3.6}{800} \times 10{,}000 = 45.0\text{ bp} \]

Each portfolio’s individual draw probability remains 10%, and aggregate expected drawdowns remain $30 million. Nevertheless, the probability of simultaneous draws rises from 1% to 8%. Greater drawdown dependence increases the required buffer from $180 million to $300 million and the fee from 27.0 bp to 45.0 bp, despite unchanged expected utilization.

• A. A $30 million buffer prices expected drawdowns rather than the liquidity resources required to satisfy the 95% coverage policy.
• B. The revised probability of drawdowns no greater than $180 million is only 92%, so this buffer fails the 95% coverage requirement.
• C. Subtracting expected drawdowns of $30 million from the $300 million quantile leaves a buffer that still covers only 92% of scenarios.
• D. A $300 million buffer meets the revised coverage target, and its $3.6 million annual net carry requires 45.0 bp on $800 million.

---

Question 74

Topic: Operational Risk and Resilience

A bank’s model-risk committee is conducting a retrospective review under the FDIC’s June 2017 model-risk guidance. The bank proposes using a logistic model to estimate one-year default probabilities at the start of each year and set credit limits for performing commercial borrowers.

Review findings:

• The economic rationale, variable definitions, estimation method, and intended use are documented. Predictors include the internal borrower rating.
• Historical borrower-year records retain their original year labels, but the warehouse replaces historical ratings whenever a later credit review changes a rating, including after default.
• Estimation and a random holdout test use the same warehouse extract. The holdout shows strong discrimination and close agreement between average predicted and realized default rates.
• Independent implementation tests reproduce the documented equations. Production forecasts will use ratings available at the start of the forecast year.

Which remediation most directly addresses the principal weakness before approval for the proposed use?

• A. Rebuild the scoring code independently, reconcile predictions with the approved equations, and test the existing coefficients across the full rating range.
• B. Adjust the model intercept to match realized portfolio default rates, retain its ranking coefficients, and test calibration on a later historical period.
• C. Reconstruct forecast-date ratings, re-estimate using those records, and test performance on a later period assembled with the same information cutoff.
• D. Replace the random split with chronological partitions, re-estimate using the existing extract, and test performance on the latest historical year.

Best answer: C

Explanation: Under the historical June 2017 FDIC framework, model development must connect sound theory, suitable data, estimation, implementation, and testing to the intended use. Documentation and accurate coding do not establish fitness for purpose when the development data contain unavailable information.

The overwritten ratings introduce look-ahead bias: historical predictors incorporate credit reviews occurring after the forecast date, potentially including the default being predicted. A holdout drawn from the same extract preserves this defect, making strong discrimination and calibration unreliable evidence of prospective performance. Merely separating records by calendar year does not repair their information content.

The bank should reconstruct ratings as they were known at each forecast date, re-estimate the model, and evaluate it on a later period using the same information-availability rules. This directly aligns development and testing with production use.

• A. Matching documented equations establishes coding fidelity, not whether the estimation data represent information available for the intended prospective forecasts.
• B. Intercept recalibration can adjust average default frequency, but it cannot correct risk rankings estimated from ratings updated after the forecast date.
• C. Forecast-date ratings remove information unavailable when predictions would have been made, allowing re-estimation and later-period testing to assess genuine prospective performance.
• D. A chronological split does not remove post-forecast information from predictor values within each record, so later-year results remain contaminated.

---

Question 75

Topic: Operational Risk and Resilience

A banking group headquartered in Country H operates corporate-banking subsidiaries in Countries A and B. Its group-wide AML/CFT policy uses Country H’s stricter customer due diligence (CDD) standard.

Review findings:

• Country A permits group compliance to access customer, beneficial-owner and transaction records.
• Country B has a less demanding CDD minimum but permits subsidiaries to apply stronger CDD procedures.
• Country B prohibits identifiable records from being shared with other group entities or accessed by headquarters staff. Local compliance and local auditors have access; headquarters receives only anonymized summaries.
• Transfers between customers of the two subsidiaries are increasing. The current summaries do not allow group compliance to identify common beneficial owners or assess linked activity.
• An application for an exemption from Country B’s information-sharing restriction is pending.

Which response is most consistent with Basel’s group-wide AML/CFT guidance?

• A. Retain the group CDD standard, pursue the pending legal exemption, and postpone home-supervisor notification until Country B reaches a decision.
• B. Retain the group CDD standard, notify the home supervisor promptly, and accept local audit access as sufficient mitigation of the cross-border monitoring gap.
• C. Apply Country B’s CDD minimum, notify the home supervisor promptly, and assess lawful additional controls to mitigate the cross-border monitoring gap.
• D. Retain the group CDD standard, notify the home supervisor promptly, and assess lawful additional controls to mitigate the cross-border monitoring gap.

Best answer: D

Explanation: Basel’s group-wide approach requires consistent AML/CFT risk management, not identical local laws or unrestricted information transfers. Where home and host standards differ, the higher standard should be applied when host law permits it. Country B therefore provides no legal reason to lower the group’s CDD standard.

The information-sharing restriction must nevertheless be respected. Because it prevents effective consolidated monitoring under the current arrangements, headquarters should inform the home supervisor and assess lawful mitigating controls. Local access and audits can support those controls, but their existence alone does not resolve the inability to connect beneficial owners and activity across subsidiaries.

A pending exemption does not remove the present weakness. If lawful alternatives cannot provide adequate risk management, further measures, potentially including restrictions on or discontinuation of affected business, may be necessary.

• A. The restriction already impairs group-wide risk management, so the pending exemption does not justify postponing notification to the home supervisor.
• B. Local audit access is not equivalent to adequate consolidated monitoring when the group cannot identify common owners or assess linked activity.
• C. Country B’s lower statutory minimum does not justify weakening CDD when applying the higher home-country standard is legally permitted.
• D. Country B permits stricter CDD, while the unresolved information restriction requires home-supervisor notification and assessment of lawful risk-mitigation arrangements.

Questions 76-80

Question 76

Topic: Current Issues in Financial Markets

A geopolitical shock reprices a bank’s sovereign bond portfolio and triggers deposit withdrawals. Its initial balance sheet is shown below.

Shock and funding conditions:

• Sovereign bonds fall 15% in value. Loan values initially remain unchanged.
• Depositors require $15 million of withdrawals today.
• Of the revalued sovereign portfolio, $10 million is available for new repo borrowing at a 40% haircut. The remaining bonds are already encumbered.
• The bank proposes selling a loan block currently valued at $6 million for $4 million in cash.
• No other funding or equity is available today. All valuation and sale losses reduce equity immediately, and no other cash flows or valuation changes occur.

Assess balance-sheet solvency using the stated asset values, rather than regulatory capital thresholds.

After the shock and the maximum repo draw, which assessment correctly describes the bank before and after the proposed loan sale?

• A. Before the sale, the bank is solvent with no cash shortfall; the sale is unnecessary and would leave negative equity.
• B. Before the sale, the bank is solvent with a cash shortfall; the sale meets the withdrawals and leaves positive equity.
• C. Before the sale, the bank is insolvent with a cash shortfall; the sale meets the withdrawals and leaves negative equity.
• D. Before the sale, the bank is solvent with a cash shortfall; the sale meets the withdrawals and leaves negative equity.

Best answer: D

Explanation: Balance-sheet solvency concerns asset values relative to liabilities; liquidity concerns cash available when payments fall due. Sovereign repricing produces a $6 million loss, reducing equity from $7 million to $1 million. The bank therefore remains balance-sheet solvent before the loan sale.

The repo advances 60% of the $10 million collateral value, providing $6 million. Together with $5 million of cash, this covers only $11 million of the $15 million withdrawals, leaving a $4 million shortfall. The collateral haircut limits funding but does not itself reduce equity.

Selling the loan block supplies the missing $4 million but realizes a $2 million loss, turning the $1 million equity buffer into a $1 million deficit. Repo borrowing increases cash and liabilities equally, while deposit repayment reduces both equally. Neither changes equity. Thus, the sale resolves the immediate liquidity shortfall while creating balance-sheet insolvency, illustrating how funding stress and forced sales can transmit into solvency problems.

• A. The haircut limits borrowing against $10 million of collateral to $6 million, leaving a $4 million shortfall after using existing cash.
• B. The sale realizes a $2 million loss against only $1 million of post-repricing equity, leaving a $1 million equity deficit.
• C. Sovereign repricing leaves $1 million of positive equity; the repo haircut restricts borrowing capacity rather than creating an additional asset valuation loss.
• D. The $4 million sale proceeds fill the cash shortfall, but the $2 million sale loss exceeds the remaining $1 million equity buffer.

---

Question 77

Topic: Risk Management and Investment Management

A systematic equity fund has regressed its annual return forecasts on an intercept and market beta using equal observation weights. The table shows the resulting residual forecasts and inputs for four stocks. Specific variances are annualized using decimal returns.

Investment mandate: Final holdings must be net-dollar neutral and market-beta neutral. A positive value premium is part of the investment thesis, so value exposure should not be separately neutralized.

Sizing rule: Signed portfolio weights are \(w_i=k r_i/\sigma_i^2\), where \(r_i\) is the final forecast in decimal form, \(\sigma_i^2\) is the displayed specific variance, and \(k>0\) is a common scaling factor. No additional constraints or hedges are applied.

The team will use residuals from a further cross-sectional regression of the displayed forecasts as \(r_i\). Which regression specification meets the mandate?

• A. Regress on an intercept, market beta, and value loading using observation weights proportional to inverse specific variance.
• B. Regress on an intercept and market beta using observation weights proportional to inverse specific variance.
• C. Regress on an intercept and market beta using equal observation weights.
• D. Regress on an intercept and market beta using observation weights proportional to specific variance.

Best answer: B

Explanation: Neutrality must apply to implemented holdings, not merely to forecasts. The displayed forecasts sum to zero and have a zero beta-weighted sum. Inverse-variance sizing nevertheless produces holdings proportional to \((1,-0.5,-0.25,1)\), whose net weight and beta exposure are proportional to 1.25 and 1.30, respectively.

Let \(X\) contain the intercept and market beta, and let \(W\) be diagonal with entries \(1/\sigma_i^2\). Weighted least squares using these observation weights produces residual forecasts satisfying \(X^{\mathsf T}Wr=0\). Since holdings are \(w=kWr\), they satisfy \(X^{\mathsf T}w=0\), delivering both net-dollar and market-beta neutrality. These exposure conditions could also be imposed directly as portfolio constraints.

Adding value loading to the weighted regression would impose zero value exposure as well. That removes the intended value premium rather than only unwanted market exposure. Required neutralization may remove some forecast information overlapping the intercept or market beta, but it need not eliminate the entire value tilt.

• A. Including value loadings also forces the implemented portfolio’s value exposure to zero, eliminating an intended source of expected return.
• B. Inverse-variance regression enforces zero net and beta exposures after sizing because its orthogonality conditions use the portfolio’s implementation weights.
• C. Equal-weight residuals are neutral before sizing, but unequal inverse-variance position weights reintroduce both net-dollar and market-beta exposure.
• D. Variance-weighted regression neutralizes sums weighted by variance, whereas the implemented portfolio exposures depend on sums weighted by inverse variance.

---

Question 78

Topic: Credit Risk Measurement and Management

A bank currently holds industrial loans (segment A) and proposes adding commercial-property loans (segment B) in the same region. Approval requires combined portfolio economic capital to be no more than $40 million under a specified regional-downturn sensitivity.

The bank measures unexpected loss (UL) as the standard deviation of one-year credit losses. It defines economic capital as three times portfolio UL and allocates capital using Euler contributions to this risk measure.

One-year inputs (millions of dollars):

Dependence and recovery assumptions:

• The estimated base-case correlation between segment losses is 0.00. Segment B’s recovery rate is 60%.
• In the downturn sensitivity, loss correlation rises to 0.50 and B’s recovery rate falls to 50%. A’s UL remains unchanged.
• B’s exposure and marginal distribution of defaulted exposure are held fixed. Recovery is deterministic within each case.

Which assessment correctly compares base-case diversification and applies the downturn capital limit and allocation convention?

• A. Base-case capital per unit of exposure falls; downturn capital is $42.0 million, above the limit, with $27.9 million allocated to segment B.
• B. Base-case capital per unit of exposure falls; downturn capital is $35.0 million, within the limit, with $25.7 million allocated to segment B.
• C. Base-case capital per unit of exposure falls; downturn capital is $36.5 million, within the limit, with $21.7 million allocated to segment B.
• D. Base-case capital per unit of exposure rises; downturn capital is $48.0 million, above the limit, with $30.0 million allocated to segment B.

Best answer: A

Explanation: Covariance aggregation can increase absolute risk while reducing risk per unit of exposure. The existing portfolio’s capital is \(3 \times 6 = 18\) million dollars, or 18% of exposure. After expansion, base-case UL is \(\sqrt{6^2+8^2}=10\) million dollars, giving $30 million of capital and a 15% capital-to-exposure ratio. Base-case Euler allocations are $10.8 million for A and $19.2 million for B.

The downturn raises B’s LGD from 40% to 50%, so its UL becomes \(8 \times (50/40)=10\) million dollars. Combined UL is \(\sqrt{6^2+10^2+2(0.50)(6)(10)}=14\) million dollars. Capital is therefore $42 million, exceeding the $40 million limit.

Euler allocation assigns three times a segment’s covariance with total portfolio loss, divided by portfolio UL. B receives \(3(10^2+0.50 \times 6 \times 10)/14=27.86\) million dollars, and A receives $14.14 million. Shared regional drivers and lower recoveries weaken diversification in this sensitivity; low average correlation does not establish tail independence.

• A. The recovery reduction raises B’s UL to $10 million; correlation of 0.50 produces $42 million of aggregate capital and a $27.9 million Euler allocation.
• B. These values treat downturn losses as uncorrelated; the specified positive covariance increases both aggregate capital and B’s Euler allocation.
• C. These values apply the higher correlation but retain B’s $8 million base-case UL, overlooking the increase in LGD from 40% to 50%.
• D. Adding standalone capital ignores diversification; the specified correlations require covariance aggregation and Euler allocation rather than assigning B its standalone capital.

---

Question 79

Topic: Credit Risk Measurement and Management

A risk committee is reviewing a mezzanine tranche’s expected principal-loss estimate before approving a securitization’s pricing. Consider the following simplified two-year deal. Ignore discounting.

Capital structure and waterfall:

• Initial collateral principal is $100 million, financed by $70 million of senior notes, $20 million of mezzanine notes, and $10 million of equity.
• At the end of year 1, collateral prepays $20 million in every scenario. Prepayments redeem senior principal first.
• After prepayments, an overcollateralization test divides remaining collateral principal by outstanding senior plus mezzanine principal. If the ratio is below 1.05, all year-1 excess interest redeems senior principal; otherwise, it is distributed to equity.
• Year-1 excess interest is $4 million in every scenario, after all expenses and contractual note interest.
• At year-2 maturity, remaining collateral principal pays senior, then mezzanine, then equity. There are no other cash flows or credit support.

Model assumptions:

• Baseline: A diversified-loss approximation assigns a certain $15 million collateral principal loss immediately before year-2 maturity.
• Sensitivity: A clustered-default representation assigns a 50% probability to no collateral loss and a 50% probability to a $30 million loss immediately before the year-1 prepayment.
• Losses immediately reduce collateral principal, with no subsequent recoveries. Both models therefore have expected collateral principal loss of $15 million.

Which conclusion about expected mezzanine principal loss under the sensitivity model is justified?

• A. Expected mezzanine principal loss rises to $8 million; interest diversion partly offsets the larger shortfall in the clustered-default state.
• B. Expected mezzanine principal loss falls to $1 million; triggered interest diversion reduces the baseline shortfall above equity protection.
• C. Expected mezzanine principal loss rises to $10 million; clustered collateral losses exhaust the equity tranche’s first-loss protection.
• D. Expected mezzanine principal loss remains $5 million; unchanged mean collateral loss preserves the baseline tranche-loss estimate.

Best answer: A

Explanation: Tranche losses are nonlinear in collateral losses, and cash-flow waterfalls depend on event timing. Under the baseline, the year-1 coverage test passes, releasing excess interest to equity. At maturity, $65 million of collateral repays $50 million of senior principal and $15 million of mezzanine principal, producing a $5 million mezzanine loss.

In the sensitivity’s loss state, the $30 million loss and $20 million prepayment leave $50 million of collateral. Senior and mezzanine balances are $50 million and $20 million. The coverage ratio is \( 50/70 = 0.7143 < 1.05 \), so $4 million of excess interest reduces senior principal to $46 million. At maturity, mezzanine receives the remaining $4 million and loses $16 million. It incurs no loss in the no-default state, giving expected loss of \( 0.5 \times 16 = 8 \) million dollars.

Default clustering increases mezzanine exposure to severe outcomes, while earlier defaults activate protective interest diversion. Matching mean collateral loss alone does not validate tranche estimates; the model must represent the loss distribution and waterfall timing.

• A. Interest diversion reduces senior principal to $46 million, leaving $4 million for mezzanine and a $16 million shortfall; its 50% probability gives $8 million.
• B. Subtracting $4 million from the baseline $5 million treats state-dependent diversion as a certain offset and ignores the nonlinear allocation of clustered losses.
• C. The $10 million estimate averages a $20 million loss over the 50% default state, omitting the $4 million interest diversion that reduces its shortfall.
• D. The unchanged collateral-loss mean does not preserve tranche losses because the revised distribution concentrates losses and activates interest diversion before maturity.

---

Question 80

Topic: Operational Risk and Resilience

A bank’s capital committee is comparing the Basic Indicator Approach (BIA), the original business-line standardized approach (SA), and the Advanced Measurement Approach (AMA) under the historical Basel II framework. The comparison uses the same business portfolio and excludes transitional floors and national add-ons.

Gross income: All amounts are in millions of dollars. All other business lines have zero gross income.

The SA beta factors are 18% for corporate finance and 12% for retail banking. A supervisor-approved AMA produces a capital estimate of $18.0 million.

The CFO comments:

The lower AMA charge proves that our controls are more effective than the income-based approaches suggest.

Which assessment should the capital committee accept?

• A. BIA capital is $30.0 million and SA capital is $20.4 million; the lower AMA charge does not establish superior control effectiveness.
• B. BIA capital is $20.0 million and SA capital is $20.4 million; the lower AMA charge does not establish superior control effectiveness.
• C. BIA capital is $30.0 million and SA capital is $30.6 million; the lower AMA charge does not establish superior control effectiveness.
• D. BIA capital is $30.0 million and SA capital is $20.4 million; the lower AMA charge establishes superior control effectiveness.

Best answer: A

Explanation: Basel II’s BIA and original business-line SA use gross-income proxies, but treat negative years differently.

Let \(G_t\) be total gross income in year \(t\), and \(n_+\) the number of positive-income years. BIA capital is:

\[ K_{\mathrm{BIA}} = 0.15\frac{\sum_{t:G_t>0}G_t}{n_+} \]

Annual totals are 200, -20, and 200 million. Only positive years enter the numerator and denominator, giving \(0.15(200+200)/2=30\) million.

For SA, let \(G_{j,t}\) be business-line income and \(\beta_j\) its beta factor:

\[ K_{\mathrm{SA}} = \frac{1}{3}\sum_{t=1}^{3}\max\left(0,\sum_j\beta_jG_{j,t}\right) \]

The annual weighted amounts are 30, -4.8, and 31.2 million. Negative business-line income can offset positive income within a year, but each annual total is floored at zero. Capital is therefore \((30+0+31.2)/3=20.4\) million.

AMA permits bank-specific risk modeling. Its lower estimate does not establish stronger controls because model assumptions and calibration also affect capital. The later revised standardized approach instead uses a business indicator component and, where applicable, an internal loss multiplier.

• A. BIA excludes the negative-income year, SA retains three-year averaging after annual flooring, and the AMA capital difference cannot isolate control effectiveness.
• B. The BIA divides positive-year gross income by the two positive-income years, not all three years, producing capital of $30.0 million.
• C. The SA retains all three years in its denominator, including the year with a zero floored charge, producing $20.4 million rather than $30.6 million.
• D. AMA capital reflects bank-specific loss data, calibration, modeling assumptions, and control-environment inputs; a lower charge cannot isolate the contribution of controls.

After your attempt

Classify mistakes by fact, assumption, probability measure, method, units, arithmetic or interpretation. Rework the case under a changed input before moving on. Use the Cheat Sheet , Study Plan and official readings for focused follow-up.

Report an unclear item or see FRM Part II practice updates .