Sitonce
Country: US
Show exams for United States Hong Kong
Sign in

FRM Part I Sample Questions and Answers

Updated 7 min read
Key takeaway

These original FRM Part I questions test conditional probability, duration and convexity, and option profit.

  • Each explanation shows formulas, assumptions, units, and why a plausible alternative is wrong.
  • Review each solution, then try a new version with different values and assumptions.
On this page11 sections
  1. Question 1: conditional probability
  2. Question 2: duration and convexity approximation
  3. Question 3: long call profit at expiration
  4. How to use practice questions
  5. Question 4: normal one-day VaR
  6. Question 5: forward price with a cash benefit
  7. Question 6: expected loss under a two-state model
  8. Question 7: combine mutually exclusive outcomes
  9. Question 8: interpret an exceedance
  10. Turn a worked answer into transferable practice
  11. Use these items with official materials

Question 1: conditional probability

Example question

In a hypothetical group of 1,000 exposures, 100 are distressed and 900 are healthy. A warning signal appears for 80 distressed exposures and 90 healthy exposures. Given that the signal appears, what is the probability an exposure is distressed?

  1. 8%
  2. 47.06%
  3. 80%
  4. 90%
Answer: B. There are 170 signals in total: 80 distressed plus 90 healthy. Therefore P(distressed | signal) = 80/170 = 47.06%. The 80% figure is P(signal | distressed), which reverses the conditioning. The base rate matters.

Question 2: duration and convexity approximation

Example question

A bond has modified duration 4.2 and convexity 25. If its yield rises by 0.005, estimate the fractional price change using the duration-convexity approximation.

  1. +2.07%
  2. -2.07%
  3. -4.33%
  4. +4.33%
Answer: B. Approximate ΔP/P = -DmodΔy + 0.5C(Δy)^2. Substitution gives -4.2(0.005) + 0.5(25)(0.005)^2 = -0.0206875, or about -2.07%. This uses the stated modified-duration and convexity convention and a small yield change.

Question 3: long call profit at expiration

Example question

A long call has strike 60 and premium 4 per unit. Ignoring commissions and financing, what is the profit per unit at expiration if the underlying price is 70?

  1. -4
  2. 0
  3. 6
  4. 10
Answer: C. The call payoff is max(70 - 60, 0) = 10. Profit subtracts the premium: 10 - 4 = 6. Payoff and profit are not interchangeable. The example ignores contract multipliers, transaction costs, and financing.

How to use practice questions

Write the requested conditional probability before inserting counts. This prevents reversing P(A given B) and P(B given A). A frequency table is often the clearest way to make the denominator explicit.

For bond sensitivity, convert a basis-point move to a decimal yield change before multiplying. A 50 basis-point increase is 0.005, not 0.50. Keep the first-order duration term and second-order convexity term separate.

For options, calculate terminal payoff first, then add or subtract the premium to find profit. For a long call, breakeven at expiration is strike plus premium, ignoring costs. Being in the money does not guarantee a net profit.

These are original practice items, not official or recalled GARP questions. Use them to practice method selection and explain assumptions, then work the current GARP Learning materials.

Question 4: normal one-day VaR

Example question

A linear portfolio is worth $500,000. Assume a zero mean, normally distributed one-day return with standard deviation 1.2%. Use a one-sided 99% normal critical value of 2.326. What is the approximate one-day loss VaR?

  1. $6,000
  2. $13,956
  3. $27,912
  4. $116,300
Answer: B. Multiply portfolio value by volatility and the positive loss quantile: $500,000 × 0.012 × 2.326 = $13,956. The result is a positive dollar loss threshold. It does not cap losses, and it depends on the stated normal, zero-mean, one-day assumptions. $6,000 omits the quantile; $27,912 doubles it.

Question 5: forward price with a cash benefit

Example question

A non-income asset has spot price 100. With continuous compounding, the financing rate is 5% and a known continuous yield is 2%. Under the simple carry model, what is the one-year forward price?

  1. 97.04
  2. 100.00
  3. 103.05
  4. 107.25
Answer: C. Use F = S₀e^((r-q)T) = 100e^((0.05-0.02)×1), approximately 103.05. The income yield lowers the carry relative to financing alone. The calculation assumes the stated continuous rates and no other costs or frictions. Do not use a commodity storage formula unless the asset and question supply those terms.

Question 6: expected loss under a two-state model

Example question

A credit exposure has a 3% probability of default. If default occurs, loss given default is 40% of a $2 million exposure. Ignore discounting and other states. What is the expected loss?

  1. $2,400
  2. $24,000
  3. $60,000
  4. $800,000
Answer: B. Expected loss equals probability of default × loss given default × exposure: 0.03 × 0.40 × $2,000,000 = $24,000. The conditional loss in the default state is $800,000, but it is not the unconditional expected loss. Keep the probability as a decimal and distinguish exposure from loss given default.

Question 7: combine mutually exclusive outcomes

Example question

An event A has probability 0.30. Given A, event B has probability 0.20. Given not A, event B has probability 0.10. What is P(B)?

  1. 0.06
  2. 0.10
  3. 0.13
  4. 0.20
Answer: C. Use total probability: P(B) = P(B|A)P(A) + P(B|not A)P(not A) = 0.20×0.30 + 0.10×0.70 = 0.13. The 0.06 answer is only the joint probability of A and B. The second branch must be included because B can occur when A is false.

Question 8: interpret an exceedance

Example question

A model reports a one-day 95% VaR of $4 million. Which interpretation is most consistent with the definition, assuming the model is correctly calibrated?

  1. The portfolio cannot lose more than $4 million in one day
  2. The portfolio is expected to lose exactly $4 million every 20 days
  3. About 5% of modeled one-day outcomes exceed a $4 million loss threshold
  4. The average loss in the worst 5% of outcomes is $4 million
Answer: C. A 95% VaR is a loss quantile: about 5% of modeled outcomes lie beyond the threshold under the model assumptions. It is not a maximum, a schedule for exceedances, or the average tail loss. The final option describes an expected-shortfall idea, not VaR.

Turn a worked answer into transferable practice

After checking an answer, change one condition and solve again. For the VaR question, change the confidence quantile or volatility and predict the direction before recalculating. For the forward question, add an income yield and explain why it offsets part of financing. For expected loss, change exposure or loss given default. A variation tests whether you understand the relationship rather than the specific numbers.

Explain distractors in terms of a recognizable error. A VaR answer may omit the quantile, use a two-sided critical value, or mistake a percentage for a decimal. A forward calculation may add a yield instead of subtracting it. A conditional probability may use the wrong denominator. Naming the mistake helps you catch it in a new question under time pressure.

Keep the question conditions visible while you calculate: units, time horizon, sign convention, compounding rule, and whether the result is conditional or unconditional. Re-read the final sentence before selecting an option. Many multiple-choice errors are not failures of algebra; they are correct computations for a quantity the question did not ask for.

Check the direction before calculating: increasing volatility should raise VaR, a call below its strike at expiration has no payoff, and halving default probability halves expected loss when the other inputs stay fixed. You can also reverse a question, for example solving for implied carry from a forward price. Revisit altered versions after a delay to test the method rather than memory.

Use these items with official materials

These are original practice questions written for explanation, not official or recalled GARP items. They sample several concepts but do not reproduce the exam's full breadth, difficulty distribution, or scoring process. Use the current Learning Objectives and official GARP Learning materials to cover the syllabus, then use fresh mixed practice to test whether the methods transfer.

The questions do not establish a passing percentage. GARP does not publish a fixed raw cutoff in the public information reviewed here, and a short set cannot predict a result. Review your reasoning and use the patterns of error to decide what to revisit.

Question patternKey setupPlausible error
BayesTarget conditional and denominatorReverse the condition
DurationDecimal yield changeBasis-point scale error
Option profitPayoff then premiumReport payoff only
Expected lossPD times LGD times exposureOmit LGD

Common questions

Are these official GARP questions?

No. The items here are original practice questions written for explanation, not official or recalled exam questions.

Do these questions cover the whole syllabus?

No. They illustrate selected methods such as VaR, derivatives, expected loss, and probability. Use the current Learning Objectives and official materials for complete coverage.

Can the set estimate my passing score?

No. A short practice set is not equated to the live exam, and GARP does not publish a fixed raw cutoff in the cited exam information.

How should I review a wrong answer?

Identify the first failed step, explain why the correct method applies, then solve a new version with changed values or assumptions after a delay.