Quantitative Methods
What's Inside
- Ch 1 — Rates & Returns (75 Questions)
- Ch 2 — Time Value of Money (75 Questions)
- Ch 3 — Statistics & Quantitative Methods
- Ch 4 — Probability Concepts
- Ch 5 — Portfolio Mathematics
- Ch 6 & 7 — Simulation & Estimation
- Ch 8 — Hypothesis Testing
1. Holding Period Return (HPR)
- Measures total return over any holding period (not annualized).
- For multi-period HPR: chain-multiply sub-period gross returns — HPR = (1+r₁)(1+r₂)···(1+rₙ) − 1.
2. Annualizing Returns
- Days: raise to (365/days). Months: raise to (12/months). Weeks: raise to (52/weeks).
3. Arithmetic vs. Geometric Mean
- Arithmetic mean — simple average; has known statistical properties; overstates compound growth.
- Geometric mean — GM = [(1+r₁)(1+r₂)···(1+rₙ)]^(1/n) − 1; best for multi-period compound growth; always ≤ AM.
- Harmonic mean — HM = n / Σ(1/rᵢ); used for dollar-cost averaging (average price paid); always ≤ GM.
- Key identity: AM × HM = GM²; and AM ≥ GM ≥ HM.
4. Time-Weighted vs. Money-Weighted Return
- Time-Weighted Return (TWR) — chains sub-period returns; eliminates effect of external cash flow timing; standard for evaluating manager skill.
- Money-Weighted Return (MWR) — equals the IRR of all portfolio cash flows; reflects impact of contribution/withdrawal timing; preferred when investor controls cash flows.
5. Continuously Compounded Returns
- For a price move from P₀ to P₁: r_cc = ln(P₁/P₀).
- CC EAR = e^(stated rate) − 1 (always > discrete compounding EAR).
- For positive returns: HPR > CC return.
6. Return Components
- Real return = (1 + nominal) / (1 + inflation) − 1 → measures purchasing power gain.
- T-bill yield = nominal risk-free rate (includes inflation premium; no default/liquidity risk).
- Real risk-free rate = pure time preference; no inflation, no risk premiums.
- Risk premium = equity return − risk-free rate.
- Gross return (before fees) is used for manager-to-manager comparison.
- Net return = gross return − management fees.
Click Show Answer to reveal the answer and explanation.
A. 3.7%
B. 5.0%
C. 15.0%
A. 22.0%
B. 24.0%
C. 27.5%
A. 3.0%
B. 3.7%
C. 9.3%
A. 0.06%
B. 6.35%
C. 11.24%
A. 0.06%
B. 5.29%
C. 6.35%
A. 9.4%
B. 10.8%
C. 15.0%
A. 9.4%
B. 10.8%
C. 7.7%
A. 5.5%
B. 7.0%
C. 10.4%
A. 23.2%
B. 51.4%
C. 51.7%
A. 16.1%
B. 48.9%
C. 51.4%
A. 6.00%
B. 6.08%
C. 26.27%
A. 4.01%
B. 4.50%
C. 17.03%
A. 0.23%
B. 1.16%
C. 3.08%
A. Arithmetic mean return
B. Geometric mean return
C. Holding period return
A. 4.01%
B. 4.50%
C. 17.03%
A. 3.73%
B. 3.80%
C. 3.76%
A. 1.0%
B. 2.0%
C. 2.5%
A. 8.33%
B. 13.33%
C. 18.33%
A. 1.5%
B. 9.0%
C. 18.5%
A. 6.0%
B. 12.1%
C. 18.7%
A. 18.1%
B. 20.4%
C. 25.5%
A. 31.1%
B. 40.8%
C. 49.4%
A. 3.6%
B. 15.6%
C. 16.8%
A. 9.2% higher than Jennet's
B. 11.7% lower than Jennet's
C. 22.7% higher than Jennet's
A. −0.1250
B. −0.1000
C. −0.1178
A. −5.17%
B. −5.04%
C. +5.17%
A. −14.86%
B. −13.84%
C. −16.09%
A. −5.7%
B. −6.3%
C. −6.7%
A. 9.42%
B. 9.20%
C. 9.67%
A. The same
B. Higher
C. Lower
A. e^R − 1
B. ln(1 + R)
C. ln(1 + R) − 1
A. The continuously compounded return
B. The holding period return
C. Neither; they are equal
A. Leveraged return
B. Real return
C. Nominal return
A. 12.5%
B. 15.0%
C. 6.25%
A. 17.19%
B. 23.44%
C. 32.42%
A. 24.00%
B. 28.50%
C. 31.25%
A. 1.20%
B. 0.94%
C. −0.50%
A. Arithmetic mean return
B. Geometric mean return
C. Money-weighted return
A. Net return
B. Gross return
C. Pre-tax return
A. Inflation
B. Fees
C. Taxes
A. Approximately the nominal risk-free rate plus expected inflation
B. Approximately the nominal risk-free rate reduced by expected inflation
C. Exactly the nominal risk-free rate reduced by expected inflation
A. Real risk-free interest rate
B. Nominal risk-free interest rate
C. Total rate of return
A. Nominal risk-free rates because they contain an inflation premium
B. Nominal risk-free rates because they do not contain an inflation premium
C. Real risk-free rates because they contain an inflation premium
A. Real risk-free rate + inflation + default risk premium + liquidity premium + maturity premium
B. Nominal risk-free rate + inflation + default risk premium + liquidity premium + maturity premium
C. Real risk-free rate + default risk premium + liquidity premium + maturity premium
A. 4.30%
B. 6.75%
C. 7.6%
A. 4.30%
B. 4.98%
C. 6.80%
A. 5.34%
B. 6.90%
C. 7.76%
A. 6.97%
B. 7.93%
C. 15.70%
A. After-tax return
B. Real return
C. Holding period return
A. 9.82%
B. 13.08%
C. 15.67%
A. 17.86%
B. 20.20%
C. 43.51%
A. Discount rate
B. Financing cost
C. Opportunity cost
A. 4.0% — required rate of return
B. 4.5% — discount rate
C. 4.5% — required rate of return
A. Discount rate
B. Opportunity cost
C. Required rate of return
A. 2011–2015
B. 2015–2020
C. Neither; both periods earned equivalent returns
A. Applying a discount rate approach
B. Using historical returns as proxies for expected returns
C. Applying a required return framework
A. −14.86%
B. −16.09%
C. −13.84%
A. 3.73%
B. 3.80%
C. 3.76%
A. Less than 12.75
B. Between 12.75 and 13.25
C. Greater than 13.25
A. AM > GM > HM
B. GM > AM > HM
C. HM > GM > AM
A. The square root of the geometric mean
B. The square of the geometric mean
C. The geometric mean itself
A. 8.12
B. 8.63
C. 9.30
A. $11.50
B. $11.75
C. $11.20
A. Arithmetic mean
B. Geometric mean
C. Harmonic mean
A. Take all cash inflows and outflows into account using IRR
B. Result in higher returns than money-weighted returns
C. Are not affected by the timing of cash flows
A. Money-weighted return will tend to be depressed
B. Money-weighted return will tend to be elevated
C. Time-weighted return will tend to be elevated
A. Money-weighted return
B. Net present value
C. Time-weighted return
A. The return earned on the money invested
B. A comparison of returns between similar portfolios
C. A return comparison between different investment opportunities
A. It accurately estimates compound growth over multiple periods
B. It has well-known statistical properties
C. It can test whether the mean return is statistically different from zero
A. Arithmetic mean
B. Geometric mean
C. Harmonic mean
A. Adamson's annualized return exceeds Jennet's by about 9.2%
B. Adamson's annualized return is about 11.7% lower than Jennet's
C. Their annualized returns are approximately equal
A. 0.06%
B. 5.29%
C. 6.35%
A. 0.06%
B. 6.35%
C. 11.24%
A. TWR is preferred when the manager controls cash flow timing
B. MWR is preferred when evaluating manager skill across different portfolios
C. TWR reflects the impact of investor contributions and withdrawals
A. 9.2%
B. 12.5%
C. 15.0%
1. Interest Rates — Interpretation
- Required rate of return — minimum return an investor must receive to accept an investment.
- Discount rate — rate used to find the present value of future cash flows.
- Opportunity cost — return forgone by choosing a particular investment.
2. Interest Rate Components
3. Effective Annual Rate (EAR)
4. Future Value & Present Value
5. Annuities & Perpetuities
- Ordinary annuity — cash flows at END of each period.
- Annuity due — cash flows at START of each period (BGN mode on BA II Plus).
- Perpetuity PV = PMT / r
A. Required rate of return.
B. Real risk-free interest rate.
C. Inflation rate.
A. The rate needed to calculate present value.
B. Opportunity cost.
C. The maximum rate of return an investor must receive to accept an investment.
A. 2.5%
B. 4.0%
C. 9.0%
A. Default risk premium.
B. Inflation premium.
C. Real risk-free interest rate.
A. Risk of loss relative to an investment's fair value if converted to cash quickly.
B. Increased sensitivity of debt market value to interest rate changes as maturity extends.
C. Possibility that the borrower will fail to make a promised payment.
A. Inability to sell a security at its fair market value.
B. Locking funds for longer durations.
C. A risk that an investment's value may change over time.
A. 20%
B. 15%
C. 11%
A. Inflation premium.
B. Maturity premium.
C. Nominal premium.
A. 23%
B. 24%
C. 25%
A. Semi-annual.
B. Quarterly.
C. Monthly.
A. CD1
B. CD2
C. CD3
A. 11.00%
B. 11.57%
C. 11.63%
A. 18.00%
B. 19.56%
C. 20.12%
A. 3.00%
B. 12.00%
C. 12.55%
A. 7.19%
B. 7.47%
C. 7.73%
A. 16.30%
B. 18.50%
C. 19.50%
A. −6.10%
B. −4.63%
C. 6.53%
A. $1,500
B. $1,523
C. $1,541
A. $6,387
B. $6,499
C. $6,897
A. £1,677
B. £1,712
C. £1,716
A. $2.150 million
B. $2.600 million
C. $2.671 million
A. $5.625 million
B. $5.649 million
C. $5.661 million
A. ¥22.21 million
B. ¥27.98 million
C. ¥35.25 million
A. $86,151
B. $86,628
C. $107,591
A. $3.003 million
B. $3.122 million
C. $3.562 million
A. $51,832
B. $55,702
C. $55,596
A. Interest rate.
B. Present value.
C. Future value.
A. 11.9 years
B. 18.9 years
C. 21.3 years
A. 95 months
B. 225 months
C. 250 months
A. $18,840
B. $26,438
C. $26,967
A. $216,116
B. $240,129
C. $264,706
A. $1,287
B. $1,305
C. $1,396
A. $13,333
B. $13,795
C. $13,887
A. $556
B. $1,000
C. $1,199
A. €776.72
B. €803.43
C. €923.13
A. €22,038.74
B. €22,365.36
C. €22,430.59
A. 29
B. 42
C. 50
A. $894,713
B. $1,094,713
C. $1,294,713
A. Annuity amount.
B. Future value annuity factor.
C. Interest rate.
A. $12,000
B. $12,148
C. $13,903
A. $406,104
B. $408,350
C. $886,473
A. Option 1 — greater PV.
B. Option 2 — greater PV.
C. Equal PV for both.
A. $316
B. $3,158
C. $3,185
A. $487
B. $550
C. $616
A. $3,105
B. $3,654
C. $3,921
A. $2,337
B. $2,467
C. $2,726
A. $1,137
B. $1,440
C. $1,623
A. $22,000
B. $18,500
C. $16,500
A. $989,300
B. $1,009,080
C. $1,220,390
A. 7.29%
B. 7.58%
C. 7.87%
A. $160,000
B. $170,000
C. $180,000
A. $13,365
B. $11,087
C. $22,857
A. $2,924,191
B. $3,158,126
C. $7,363,610
A. $683
B. $751
C. $1,000
A. $8,370
B. $8,539
C. $8,730
A. 9.6 years
B. 12.7 years
C. 16.6 years
A. $40.00
B. $62.50
C. $55.56
A. $6,802
B. $7,360
C. $8,111
A. $7,360
B. $7,802
C. $8,111
A. $28,755
B. $31,392
C. $33,540
A. $4,128
B. $4,564
C. $5,132
A. $1,169
B. $1,185
C. $1,192
A. PV increases as the discount rate increases.
B. PV decreases as the discount rate increases.
C. PV is unaffected by changes in the discount rate.
A. $58,468
B. $60,272
C. $61,039
A. $17,258
B. $17,934
C. $18,416
A. $1.160
B. $1.170
C. $1.172
A. END mode; N = number of payments.
B. BGN mode; N = number of payments.
C. BGN mode; N = number of payments + 1.
A. 5.5%
B. 6.1%
C. 6.5%
A. FV decreases as N increases for any positive interest rate.
B. FV increases as N increases for any positive interest rate.
C. FV is independent of N when the interest rate exceeds 10%.
A. $157,566
B. $168,221
C. $178,540
A. $959
B. $982
C. $1,012
A. $4,565
B. $5,009
C. $5,217
A. $1,250
B. $2,000
C. $3,333
A. Bank A — EAR = 8.24%.
B. Bank B — EAR = 8.19%.
C. Both have the same EAR.
A. $625,000
B. $747,258
C. $792,094
1. Measures of Central Tendency
- Arithmetic mean X̄ = ΣX / n. Most widely used; sensitive to outliers.
- Median — middle value when sorted; robust to outliers.
- Mode — most frequent value; can be unimodal, bimodal, etc.
- Geometric mean G = (X₁×X₂×…×Xₙ)^(1/n); used for compound growth. Always G ≤ X̄.
- Harmonic mean H = n / Σ(1/Xᵢ); best for averaging P/E ratios with equal dollar amounts.
- Trimmed mean — excludes a stated % from both tails.
- Winsorized mean — replaces extremes with percentile boundary values before averaging.
2. Measures of Dispersion
- Range = Max − Min.
- MAD = Σ|Xᵢ − X̄| / n.
- Sample Variance s² = Σ(Xᵢ − X̄)² / (n−1).
- CV = s / X̄. Risk per unit of return; lower CV = better.
3. Skewness & Kurtosis
Positive skew: Mean > Median > Mode. Negative skew: Mean < Median < Mode.
Kurtosis: Normal = 3 (excess = 0). Leptokurtic (excess > 0): fat tails. Platykurtic (excess < 0): thin tails.
4. Correlation & Covariance
A. 10.0%
B. 11.0%
C. 12.5%
A. 23%
B. 22.8%
C. 21%
A. 25%
B. 28%
C. 30%
A. 5.0%
B. 4.3%
C. 3.9%
A. Arithmetic mean = 15
B. Harmonic mean ≈ 13.6
C. Geometric mean ≈ 14.4
A. Top and bottom 10% each
B. Top and bottom 5% each
C. All values beyond 1 standard deviation
A. 4.8%
B. 6.0%
C. 5.5%
A. 44.5%²
B. 36.8%²
C. 40.0%²
A. 6.67%
B. 7.2%
C. 5.5%
A. T-bills (CV=1.44)
B. S&P 500 (CV=6.70)
C. They are equal
A. Uses the full sample in the numerator
B. Only counts deviations below a target return
C. Divides by n instead of n−1
A. 50th percentile
B. 75th percentile
C. 25th percentile
A. Q2 − Q1
B. Q3 − Q1
C. Q4 − Q2
A. Negatively skewed
B. Symmetric
C. Positively skewed
A. Positively skewed
B. Negatively skewed
C. Uniform
A. Platykurtic
B. Mesokurtic
C. Leptokurtic
A. 0
B. 3
C. 1
A. Lower probability of extreme losses than normal
B. Greater probability of extreme outcomes than normal
C. Perfect symmetry
A. Calculate variance directly
B. Display the relationship between two variables
C. Find the median of a dataset
A. Zero
B. Maximally positive
C. Negative
A. No linear relationship
B. Perfect positive linear relationship
C. Perfect negative linear relationship
A. 0.25
B. 0.40
C. 0.60
A. Very high positive correlation
B. Correlation caused by a third variable or chance, not a true relationship
C. Correlation exactly equal to +1
A. It is always positive
B. Its value depends on units, making interpretation difficult
C. It cannot be negative
A. 8.4%
B. 8.0%
C. 7.8%
A. 0.0061
B. 0.0025
C. 0.0041
A. σ²
B. σ²/2
C. 2σ²
A. The weighted average of individual standard deviations
B. Less than the weighted average
C. Zero
A. Maximum return
B. Zero variance
C. Infinite Sharpe ratio
A. 2.0
B. 1.8
C. 1.4
A. 1.96
B. 1.56
C. 2.00
A. 8.75%
B. 10.0%
C. 7.5%
A. 79.7%²
B. 50.0%²
C. 55.0%²
A. Comparing assets with very different mean returns
B. All assets have the same mean return
C. Returns are normally distributed
A. Median
B. Arithmetic mean
C. Mode
A. All observations are positive
B. All observations are equal
C. The sample is large
A. Simplify computation
B. Produce an unbiased estimator of population variance
C. Reduce the effect of outliers
A. 68% of returns
B. 95% of returns
C. 99% of returns
A. Display mean and variance
B. Visualise distribution using quartiles and range
C. Show the relationship between two variables
A. Return per unit of total risk
B. Risk per unit of return
C. Excess return per unit of total risk
A. 14%
B. 10%
C. 12%
A. 9.0%
B. 8.7%
C. 10.0%
A. Leptokurtic
B. Mesokurtic
C. Platykurtic
A. Mean < Median < Mode
B. Mode < Median < Mean
C. Median < Mode < Mean
A. Cov(X,Y) = ρ × σ_X × σ_Y
B. Cov(X,Y) = ρ / (σ_X × σ_Y)
C. Cov(X,Y) = σ_X² + σ_Y²
A. 8.8%
B. 9.0%
C. 10.0%
A. The mean of the dataset
B. Specified percentile boundary values
C. Zero
A. 1
B. 0
C. 3
A. 13
B. 40
C. 0
A. Standard deviation only
B. CV — normalises risk by return for fair comparison
C. Variance only
1. Fundamental Definitions
- Random variable — a quantity whose outcome is uncertain in advance.
- Mutually exclusive — events that cannot both occur: P(A∩B) = 0.
- Exhaustive — events that together cover ALL possible outcomes.
2. Probability Rules
3. Bayes' Formula
4. Counting Principles
- Combination nCr = n! / [(n−r)! × r!] — order does NOT matter.
- Permutation nPr = n! / (n−r)! — order DOES matter.
- Labelling: n! / (n₁! × n₂! × … × nₖ!)
A. 40 times out of 100
B. 4 times out of 10,000
C. 60 times out of 100
A. Empirical probability
B. A priori probability
C. Subjective probability
A. 3 to 7
B. 7 to 3
C. 3 to 10
A. 90%
B. 10%
C. 50%
A. 0.70
B. 0.11
C. 0.80
A. Yes, because P(AB) = P(A)×P(B)
B. No, because P(AB) ≠ P(A)×P(B)
C. Cannot determine from given information
A. 0.21
B. 0.30
C. 0.40
A. 0.20
B. 0.25
C. 0.30
A. 0.2875
B. 0.40
C. 0.175
A. P(A|B) = P(A)
B. P(AB) = P(A)×P(B)
C. P(A) = P(B)
A. 4.7
B. 5.0
C. 4.1
A. 3.81
B. 4.41
C. 2.95
A. 11.4%
B. 11.0%
C. 10.8%
A. 0.60
B. 0.75
C. 0.50
A. Joint probabilities
B. Prior probabilities using new information
C. Marginal probabilities only
A. 0.632
B. 0.500
C. 0.400
A. 8
B. 12
C. 16
A. 49
B. 5,040
C. 720
A. 27,720
B. 3,150
C. 479,001,600
A. 210
B. 5,040
C. 720
A. 120
B. 720
C. 210
A. Conditional probability
B. Marginal probability
C. Joint probability
A. P(B)
B. P(A)
C. P(AB)
A. 0.59
B. 0.50
C. 0.65
A. 1/6
B. 2/6
C. 3/6
A. P(A)+P(B)
B. P(A)×P(B)
C. P(A|B)
A. A priori
B. Empirical
C. Subjective
A. 2.9%
B. 3.5%
C. 4.0%
A. 62.1%
B. 45.0%
C. 50.0%
A. 0
B. c
C. c²
A. E(X)+E(Y)
B. E(X)×E(Y)
C. Cov(X,Y)
A. Portfolio weights must sum to more than 1
B. Assets may be correlated, affecting combined risk
C. Returns are always negative
A. Maximum portfolio variance
B. A zero-variance portfolio is achievable
C. No risk reduction
A. 0
B. 1
C. Number of outcomes
A. 36
B. 720
C. 120
A. 0.40
B. 2.50
C. 40%
A. A and B are independent
B. A cannot occur if B has occurred (mutually exclusive)
C. P(B) = 0
A. 0.32
B. 0.40
C. 0.26
A. Permutation
B. Labelling (multinomial)
C. Combination
A. E=3.6, σ=2.94
B. E=3.6, σ=8.64
C. E=2.0, σ=3.0
A. Computing variance directly
B. Visualising sequential conditional probabilities
C. Ranking investments by return
A. 0.24
B. 0.00
C. 1.00
A. A priori
B. Empirical
C. Subjective
A. 20%
B. 80%
C. 25%
A. Must sum to exactly 1
B. Can sum to more than 1
C. Must each equal 0.5
A. 126
B. 15,120
C. 3,024
A. w₁²σ₁² + w₂²σ₂²
B. (w₁σ₁+w₂σ₂)²
C. w₁σ₁²+w₂σ₂²
A. 20
B. 40
C. 25
A. P(X=x) must be positive and sum to more than 1
B. 0 ≤ P(X=x) ≤ 1 and ΣP(X=x) = 1
C. All outcomes must be equally likely
A. Independent
B. Mutually exclusive and exhaustive
C. Dependent and overlapping
A. Asset i's expected return
B. Total portfolio market value
C. Asset i's standard deviation
A. 10.4%
B. 11.6%
C. 12.0%
A. Each asset's deviation from its own mean only
B. Expected product of both assets' return deviations from their respective means
C. Sum of each asset's variance
A. Zero
B. Correlation of the asset with itself (= 1)
C. Variance of the asset σ_A²
A. When X is above its mean, Y also tends to be above its mean
B. When X is above its mean, Y tends to be below its mean
C. They always move in exactly opposite directions
A. 0.50
B. 0.75
C. 0.30
A. 0 and +1
B. −1 and +1
C. −∞ and +∞
A. w₁σ₁² + w₂σ₂²
B. w₁²σ₁² + w₂²σ₂² + 2w₁w₂Cov(R₁,R₂)
C. w₁²σ₁² + w₂²σ₂²
A. 15.0%
B. 11.18%
C. 12.5%
A. w₁σ₁ + w₂σ₂ (weighted average)
B. √(w₁²σ₁² + w₂²σ₂²)
C. Zero
A. Simple average of the two variances
B. Zero — a risk-free portfolio is achievable
C. Always positive
A. 6
B. 12
C. 16
A. n
B. n−1
C. n²
A. ρ(A,B) × σ_A × σ_B
B. ρ(A,B) / (σ_A × σ_B)
C. ρ(A,B) × (σ_A² + σ_B²)
A. Equal to zero only
B. Less than +1
C. Greater than the portfolio's return
A. 9.0%
B. 10.2%
C. 11.0%
A. Zero
B. The average pairwise covariance
C. The average individual asset variance
A. Perfectly positively correlated
B. Uncorrelated (ρ=0)
C. Perfectly negatively correlated
A. Exceeds the risk-free rate
B. Falls below a specified target level
C. Equals the expected return
A. [E(Rp) − R_L] / σp
B. [E(Rp) − Rf] / σp
C. σp / [E(Rp) − R_L]
A. Portfolio A — higher expected return
B. Portfolio B — higher SFRatio
C. Both equally preferred
A. −0.75
B. −0.60
C. −0.50
A. 25
B. 15
C. 10
A. Asset weights
B. Individual asset expected returns
C. Correlation between asset returns
A. 10.0%
B. 10.6%
C. 11.2%
A. Always positive
B. Dimensionless and bounded between −1 and +1
C. Only applicable to normal distributions
A. Cov(A,B) is always positive
B. Cov(A,B) = Cov(B,A) — it is symmetric
C. Cov(A,B) is bounded between −1 and +1
A. Maximises expected return for any given variance
B. Has the lowest possible standard deviation across all feasible weight combinations
C. Always assigns equal weights
A. The risky asset's variance
B. Zero
C. The risky asset's expected return
A. w_p²σ_p²
B. w_f²σ_f² + w_p²σ_p²
C. (w_f σ_f + w_p σ_p)²
A. Both systematic and unsystematic risk
B. Unsystematic (firm-specific) risk but not systematic risk
C. Only systematic risk
A. 0.0288
B. 0.0432
C. 0.0576
A. 100
B. 55
C. 45
A. Diagonal elements are variances
B. All off-diagonal elements must be negative to ensure diversification
C. Cov(A,B) = Cov(B,A)
A. 0.583
B. 0.833
C. 0.250
A. Have equal expected returns at all risk levels
B. Maximise expected return for a given level of risk, or minimise risk for a given expected return
C. Maximise risk for any given expected return
A. 3
B. 6
C. 9
A. 60
B. 59
C. 30
A. Increase both portfolio return and variance
B. Reduce portfolio expected return but likely also reduce portfolio variance
C. Increase portfolio variance because negative correlation amplifies risk
A. Portfolio expected return only
B. Covariance of returns from probability-weighted scenarios
C. The Sharpe ratio
A. (σ_A−σ_B)²
B. Zero
C. σ_A²+σ_B²
A. 15.0%
B. 12.5%
C. 13.4%
A. Zero
B. Σwᵢ²σᵢ² (not the weighted average of variances)
C. Σwᵢσᵢ²
A. 13.5%
B. 15.0%
C. 12.0%
A. 4.0%
B. 20.0%
C. 0.2%
A. Portfolio becomes riskier
B. Excellent diversification — when one falls, the other tends to rise
C. The two stocks are from the same industry
A. Stock with ρ=+0.9
B. Stock with ρ=+0.2
C. Stock with ρ=−0.5
A. All weights must be strictly positive
B. Weights must sum to 1.0
C. Each weight = 1/n
A. 12.0%
B. 13.0%
C. 11.0%
A. 0.0058
B. 0.0082
C. 0.0034
- Lognormal distribution — if X~N, then e^X is lognormal. Always positive; right-skewed.
- Monte Carlo simulation — repeated random draws from assumed distributions to model security values.
- Bootstrap resampling — draw repeated samples of size n WITH replacement from observed data.
- Jackknife — remove one observation at a time; compute statistic for each reduced sample.
- CLT — for n≥30, sample mean ~N(μ, σ²/n) regardless of population distribution.
- Standard error = σ/√n (σ known) or s/√n (σ unknown).
A. Uniformly distributed
B. Normally distributed
C. Exponentially distributed
A. It is symmetric and simpler mathematically
B. Asset prices cannot be negative, and a lognormal variable is always positive
C. It assumes returns follow a uniform distribution
A. Symmetric and bell-shaped
B. Right-skewed with a lower bound of zero
C. Left-skewed with an upper bound of one
A. Multiplicative and cannot be negative
B. Additive over time, making their sum approximately normal by the CLT
C. Always positive
A. Mean and variance do not change over time
B. Past returns are not useful for predicting future returns
C. Returns follow a normal distribution
A. An analytic method deriving exact closed-form solutions
B. A technique that repeatedly generates random values for risk factors to build a distribution of security values
C. A method limited only to historical data
A. It provides exact analytical solutions with no estimation error
B. Its inputs are not limited to the range of historical data — it can test unobserved scenarios
C. It requires no distributional assumptions
A. It cannot be used for complex derivative securities
B. It is a statistical method, and results are no better than the assumptions used
C. It requires all variables to follow a normal distribution
A. Calculating Value at Risk (VaR)
B. Valuing complex securities
C. Computing exact intrinsic values free from model error
A. Removing one observation at a time (jackknife)
B. Drawing repeated samples of size n WITH replacement from observed data
C. Selecting every nth observation
A. Provides exact error-free estimates
B. Can estimate distributions of complex statistics that have no closed-form solution
C. Requires no data assumptions
A. Repeatedly drawing samples with replacement
B. Removing one observation at a time and computing the statistic for each reduced sample of size n−1
C. Selecting random subsets of pre-defined clusters
A. The sample is divided into subgroups before selecting
B. Each member of the population has the same probability of being selected
C. The researcher uses judgment to select observations
A. Guarantees every bond in the index is selected
B. Ensures each maturity/coupon risk stratum is proportionally represented in the sample
C. Requires no knowledge of population characteristics
A. Cluster selects from all subgroups; stratified selects from only some
B. Cluster assumes subsets represent the whole population; stratified selects from every subgroup
C. Cluster is always more accurate
A. Selecting observations based on ease of access, using readily available data
B. Using researcher judgment to select informative observations
C. Selecting every nth member from a population list
A. Researcher experience to select specific observations from a larger dataset
B. Randomly generated subsets representing each stratum
C. Equal probability selection from the full population
A. Exactly the same as the population distribution
B. Approximately normal with mean μ and variance σ²/n
C. Always perfectly normal regardless of sample size
A. s/√n
B. σ/√n
C. σ/n
A. 3.65%
B. 0.67%
C. 6.67%
A. Increases
B. Decreases — estimates become more precise
C. Remains unchanged
A. Simple random sampling
B. Stratified random sampling
C. Convenience sampling
A. Each population member has a known probability of being selected
B. Selection is based on researcher judgment
C. The probability that the sample mean equals the population mean is calculated
A. Tells us the exact distribution of the population
B. Allows normal distribution-based inference about the population mean regardless of the population's shape
C. Proves all financial returns are normally distributed
A. One-stage
B. Two-stage
C. Both identical
A. It can only be applied to normal data
B. Its inputs are limited by the distribution of actual observed outcomes
C. It requires the population standard deviation to be known
A. 39
B. 40
C. 41
A. Past returns predict future returns with certainty
B. The mean and variance of the return distribution do not change over time
C. Returns are always positive
A. 1.41%
B. 0.14%
C. 3.65%
A. The sample standard deviation s
B. The population mean μ
C. One specific sample mean x̄
A. The dataset is too small
B. Post-2008 regulatory changes may have altered the distribution, violating stationarity
C. Pooling data always produces unbiased estimates
A. σ² (population variance)
B. σ²/n
C. s² (sample variance)
A. Selecting every nth member after a random start
B. Classifying the population into strata before sampling
C. Selecting entire clusters at random
A. Returns on assets are always positive
B. Asset prices modelled as lognormal can never fall to zero or below
C. The distribution is bounded above by 1
A. Jackknife
B. Bootstrap
C. Systematic resampling
A. Only the risk-free rate
B. The probability distributions and parameters of each risk factor
C. The exact future stock price
A. Number of observations
B. Mean and variance
C. Normal distribution
A. Uses only historical data
B. Can generate scenarios beyond the range of historical observations
C. Always produces more accurate estimates
A. The standard deviation of individual returns within the sample
B. How much the sample mean varies across repeated samples from the same population
C. The maximum possible estimation error
A. 8
B. 10
C. 16
A. Stratified random sampling
B. Simple random sampling
C. Judgmental sampling
A. The use of a pricing model to value securities
B. Data source — bootstrap uses actual historical data; Monte Carlo uses assumed parametric distributions
C. The number of simulations required
A. Computing the range of bootstrap sample means
B. Calculating the standard deviation of the many bootstrap sample means
C. Taking the average of all bootstrap sample means
A. 20%
B. 2%
C. 0.2%
A. σ/√n
B. s/√n
C. s/n
A. The guaranteed funding level at end of 30 years
B. The distribution of funding ratios and probability of insolvency across thousands of scenarios
C. Exact optimal asset allocation free from error
A. It is cheaper and faster
B. It guarantees proportional representation of each sub-group characteristic
C. It eliminates random selection
A. Uniformly distributed
B. Normally distributed
C. Lognormally distributed themselves
A. An exact mathematical requirement
B. A general guideline — highly non-normal distributions may need larger n; near-normal distributions may need smaller n
C. Only applicable to financial data
A. Bootstrap generates scenarios from assumed parametric distributions; Monte Carlo uses actual data
B. Bootstrap resamples from observed data; Monte Carlo generates from user-specified theoretical distributions
C. Both are identical — they differ only in number of simulations
- H₀ always contains "=" and is what we test. Hₐ is what we conclude if H₀ is rejected.
- Type I error (α) — reject true H₀. Type II error (β) — fail to reject false H₀.
- Power = 1 − β. p-value = smallest α at which H₀ is rejected.
- t-test — population mean (σ unknown). z-test — large sample or σ known.
- χ²-test — single population variance. F-test — equality of two variances.
- Paired comparisons — use when samples are dependent.
A. The hypothesis the researcher hopes to prove true
B. The hypothesis the researcher wants to reject; it is the one actually tested
C. Any statement about a sample statistic
A. H₀: μ > 0
B. H₀: μ = 0
C. H₀: μ ≠ 0
A. Failing to reject H₀ when it is actually false
B. Rejecting H₀ when it is actually true
C. Accepting H₀ when it is true
A. Rejecting H₀ when it is actually true
B. Failing to reject H₀ when it is actually false
C. Setting the significance level too low
A. P(Type II error)
B. P(Type I error) — probability of rejecting a true H₀
C. 1 − Power of the test
A. P(fail to reject H₀ | H₀ is false) = β
B. P(reject H₀ | H₀ is false) = 1 − β
C. P(reject H₀ | H₀ is true) = α
A. 0.30
B. 0.70
C. 0.05
A. The probability that H₀ is true
B. The smallest significance level at which H₀ can be rejected
C. The probability of a Type II error
A. ±1.645
B. ±1.96
C. ±2.576
A. (x̄ − μ₀) / (s/√n)
B. (x̄ − μ₀) × √n
C. x̄ / σ
A. 0.40
B. 6.33
C. 1.96
A. Decrease P(Type II error) and increase power
B. Increase P(Type II error) and decrease power
C. Have no effect on any errors
A. Use a one-tailed instead of two-tailed test
B. Increase the sample size
C. Decrease α significantly
A. Both can be true simultaneously
B. They cannot both be true, and together they cover all possible outcomes
C. Neither hypothesis can ever be proven
A. The parameter could differ in either direction from H₀
B. Prior theory specifies a directional hypothesis (e.g., μ > 0)
C. The sample size is small (n < 30)
A. Population σ is known
B. Population σ is unknown and estimated from the sample
C. Sample size exceeds 30
A. n
B. n−1
C. n−2
A. The difference between two population means
B. The value of a single population variance
C. The equality of two population means from independent samples
A. The value of a single population variance
B. The equality of two population variances from independent samples
C. The difference between two population means from dependent samples
A. 21.66
B. 20.76
C. 18.36
A. 11.689 and 38.076
B. ±1.96
C. 3.247 and 20.483
A. Symmetric and centred at zero
B. Asymmetric and bounded below by zero
C. Always bell-shaped
A. Paired comparisons t-test
B. Difference in means t-test with pooled variance
C. Chi-square test
A. Fail to reject H₀
B. Reject H₀ — returns differ significantly between merger types
C. Inconclusive — t is negative
A. Independent of each other
B. Dependent — both influenced by a common factor
C. Both larger than 30 observations
A. 39
B. 38
C. 76
A. Fail to reject H₀
B. Reject H₀ — statistically significant change in betas
C. Insufficient data
A. Accept or reject H₀
B. Reject or fail to reject H₀
C. Prove or disprove Hₐ
A. Fail to reject H₀
B. Reject H₀ — p-value < α
C. Reject H₀ — p-value > α
A. t-statistic
B. z-statistic
C. F-statistic
A. Reject H₀ — variance has changed significantly
B. Fail to reject H₀ — no significant change in variance
C. Cannot conclude
A. The two samples are independent
B. Both populations are normally distributed
C. Both samples have exactly the same size (n₁=n₂)
A. We fail to reject H₀
B. We reject H₀
C. The statistic equals the critical value
A. 1.645
B. 1.96
C. 2.576
A. Population variances are assumed unequal
B. Population variances are assumed equal
C. Both samples exceed 100 observations
A. Sum of the two sample standard deviations
B. Standard error of the difference in sample means
C. Pooled sample mean
A. Reject H₀ at 5% but not 10%
B. Fail to reject H₀ at 5%; would reject at α=0.10
C. Reject H₀ at all conventional levels
A. F-test for equality of variances
B. Paired comparisons t-test
C. Chi-square test for population variance
A. Normal (z) when σ is known; t-distribution when σ is unknown
B. Chi-square when σ is known; F when σ is unknown
C. t-distribution always, regardless of whether σ is known
A. Statistical significance always implies economic significance
B. The effect size is too small to be practically meaningful, even if p < α
C. Only relevant when the sample size is small
A. Test statistic < −critical value
B. Test statistic > +critical value
C. |Test statistic| > critical value
A. 3.247 and 20.483
B. 11.689 and 38.076
C. ±1.96
A. Reduces power by making rejection harder
B. Increases power by enlarging the rejection region
C. Has no effect on power
A. Reject H₀ at 5% — p-value is close to 0.05
B. Fail to reject H₀ at 5%; reject at α=0.10
C. Reject H₀ at all conventional levels
A. 68
B. 70
C. 69
A. It gives the exact probability H₀ is true
B. It shows the exact significance level at which H₀ is rejected, allowing assessment at any α
C. It does not require specification of a significance level
A. (n−1)s²/σ₀²
B. s²/σ₀²
C. n×s²/σ₀²
A. t and z (symmetric, centred at zero)
B. χ² (based on squared values)
C. F (ratio of variances)
A. Difference in means t-test (independent samples)
B. Paired comparisons t-test (dependent samples)
C. F-test for equality of variances
A. t for means (σ unknown), χ² for single variance, F for two variances, paired-t for dependent samples
B. z for all means, F for all variance tests
C. χ² for means, t for variances