CFA Level 2 Exam  >  CFA Level 2 Test  >  Quantitative Methods  >  Practice Test: Quantitative Methods - 2 - CFA Level 2 MCQ

Quantitative Methods - 2 - Free MCQ Practice Test with solutions, CFA Level


MCQ Practice Test & Solutions: Practice Test: Quantitative Methods - 2 (30 Questions)

You can prepare effectively for CFA Level 2 Quantitative Methods with this dedicated MCQ Practice Test (available with solutions) on the important topic of "Practice Test: Quantitative Methods - 2". These 30 questions have been designed by the experts with the latest curriculum of CFA Level 2 2026, to help you master the concept.

Test Highlights:

  • - Format: Multiple Choice Questions (MCQ)
  • - Duration: 80 minutes
  • - Number of Questions: 30

Sign up on EduRev for free to attempt this test and track your preparation progress.

Practice Test: Quantitative Methods - 2 - Question 1

An analyst runs a multiple regression with monthly equity excess returns as the dependent variable and three macroeconomic factors as independent variables. The regression output is presented below:


VariableCoefficientt-Statistic
Intercept0.851.45
GDP Growth2.304.12
Inflation-1.75-2.05
Interest Rate Change0.551.80

The critical t-value at the 5% significance level (two-tailed) is 1.96. Based solely on this output, which variables are statistically significant at the 5% level?

Detailed Solution: Question 1

GDP Growth (|4.12|>1.96) and Inflation (|-2.05|>1.96) exceed critical value; others do not.

Practice Test: Quantitative Methods - 2 - Question 2

An analyst reviewing a multiple regression model notes that the variance of the model's error terms appears to change systematically with the level of one of the independent variables. Which of the following statements regarding heteroskedasticity is most accurate?

Detailed Solution: Question 2

Conditional heteroskedasticity causes biased standard errors and unreliable inference; unconditional does not.

Practice Test: Quantitative Methods - 2 - Question 3

A fixed income analyst estimates a regression model to explain monthly bond returns. The Durbin-Watson (DW) statistic from the output equals 0.85. The lower critical value (dL) is 1.40 and the upper critical value (dU) is 1.70. Which conclusion regarding serial correlation is most appropriate?

Detailed Solution: Question 3

DW=0.85 < dl="1.40," reject="" null="" of="" no="" serial="" correlation;="" positive="" serial="" correlation="" present.="" dl="1.40," reject="" null="" of="" no="" serial="" correlation;="" positive="" serial="" correlation="">

Practice Test: Quantitative Methods - 2 - Question 4

A portfolio analyst estimates a regression model explaining fund returns using four independent variables. The regression output shows a high R-squared of 0.88, yet most individual slope coefficients are statistically insignificant with very wide confidence intervals. Pairwise correlations among the four independent variables range from 0.82 to 0.94. Which of the following best explains this pattern of results?

Detailed Solution: Question 4

High R² with low t-stats and high inter-predictor correlations are hallmarks of multicollinearity.

Practice Test: Quantitative Methods - 2 - Question 5

A quantitative analyst estimates a multiple regression model with three independent variables (k = 3) and 50 observations (n = 50). The R-squared of the model is 0.72. Which of the following is closest to the F-statistic for this regression?

Detailed Solution: Question 5

F = (0.72/3)/((1-0.72)/46) = 0.24/0.006087 ≈ 39.43.

Practice Test: Quantitative Methods - 2 - Question 6

An analyst compares two regression models for forecasting equity risk premia. Model A has three independent variables and an R-squared of 0.68. Model B adds two more variables and has an R-squared of 0.69. Which of the following statements regarding model selection using adjusted R-squared is most accurate?

Detailed Solution: Question 6

Adjusted R² penalizes added variables; if gain in R² is trivial, adjusted R² for Model B may be lower.

Practice Test: Quantitative Methods - 2 - Question 7

An equity researcher is building a regression model to explain quarterly stock returns. She wants to control for seasonal effects by including a categorical variable representing four calendar seasons: spring, summer, autumn, and winter. How many dummy variables should she include to properly capture all seasonal effects without inducing perfect multicollinearity?

Detailed Solution: Question 7

For 4 categories, use 4-1=3 dummy variables; one season serves as reference category.

Practice Test: Quantitative Methods - 2 - Question 8

A risk analyst detects conditional heteroskedasticity in her regression model using the Breusch-Pagan test. She decides to apply White-corrected standard errors. Which of the following best describes the effect of this correction on the regression output?

Detailed Solution: Question 8

White correction adjusts standard errors for heteroskedasticity; coefficient estimates remain unchanged.

Practice Test: Quantitative Methods - 2 - Question 9

An economist models monthly inflation using an AR(1) specification: xt = b0 + b1xt-1 + εt, where b0 = 0.40 and b1 = 0.75. Which of the following is closest to the mean-reverting level of this process?

Detailed Solution: Question 9

Mean-reverting level = b₀/(1-b₁) = 0.40/(1-0.75) = 0.40/0.25 = 1.60.

Practice Test: Quantitative Methods - 2 - Question 10

A time-series analyst is testing whether a macroeconomic indicator follows a unit root process before including it in a regression model. She applies the Dickey-Fuller test. Which of the following most accurately describes the Dickey-Fuller test and its implications?

Detailed Solution: Question 10

DF test: H₀ is unit root present; reject when test statistic is sufficiently negative; series then stationary.

Practice Test: Quantitative Methods - 2 - Question 11

A quantitative analyst is examining whether the residuals from a regression model exhibit time-varying volatility. She finds that large residuals tend to cluster together and small residuals tend to cluster together. Which of the following most accurately describes the model best suited to capture this phenomenon and its key characteristic?

Detailed Solution: Question 11

ARCH models capture time-varying volatility; conditional variance depends on past squared residuals.

Practice Test: Quantitative Methods - 2 - Question 12

A machine learning model designed to classify corporate bonds as either defaulting or non-defaulting produces the following confusion matrix:


Predicted DefaultPredicted No Default
Actual Default8040
Actual No Default2060

Which of the following correctly calculates the model's precision and F1 score?

Detailed Solution: Question 12

Precision=80/100=0.80; Recall=80/120=0.667; F1=2×(0.80×0.667)/(0.80+0.667)≈0.73.

Practice Test: Quantitative Methods - 2 - Question 13

A data scientist is building a predictive model for credit spreads and is deciding between LASSO and ridge regression as regularization approaches. Which of the following most accurately distinguishes these two methods?

Detailed Solution: Question 13

LASSO (L1) shrinks coefficients to exactly zero enabling selection; ridge (L2) shrinks toward but not to zero.

Practice Test: Quantitative Methods - 2 - Question 14

An analyst is evaluating a machine learning model's out-of-sample performance. She applies k-fold cross-validation with k = 5. Which of the following best describes the procedure used in k-fold cross-validation?

Detailed Solution: Question 14

Dataset split into k folds; model trained on k-1 folds, validated on remaining fold, repeated k times.

Practice Test: Quantitative Methods - 2 - Question 15

A portfolio manager constructs a machine learning model to predict next-month equity returns. The model achieves an in-sample R-squared of 0.92 but produces an out-of-sample R-squared of only 0.14 when applied to new data. Which of the following best characterizes this result and the appropriate corrective action?

Detailed Solution: Question 15

High training R² with poor out-of-sample performance signals overfitting; regularization or simplification needed.

Practice Test: Quantitative Methods - 2 - Question 16

A quantitative researcher is classifying two machine learning tasks: (1) estimating the probability of sovereign debt default using historical data where each observation is labeled either defaulted or not defaulted; and (2) grouping sovereign issuers into clusters based on macroeconomic similarity without predefined categories. Which classification of these tasks is most accurate?

Detailed Solution: Question 16

Task 1 uses labeled outcomes → supervised learning; Task 2 finds structure without labels → unsupervised learning.

Practice Test: Quantitative Methods - 2 - Question 17

A macroeconomic analyst models quarterly GDP growth using the following AR(2) specification: xt = 0.30 + 0.50xt-1 + 0.20xt-2 + εt. Given that xt-1 = 4.0 and xt-2 = 3.5, what is the forecast for xt?

Detailed Solution: Question 17

x̂ₜ = 0.30 + 0.50(4.0) + 0.20(3.5) = 0.30 + 2.00 + 0.70 = 3.00.

Practice Test: Quantitative Methods - 2 - Question 18

An econometrician is reviewing two regression models. In Model A, a variable known to influence the dependent variable is excluded from the specification. In Model B, an irrelevant variable with no true relationship to the dependent variable is included. Which of the following correctly describes the consequence of omitted variable bias in Model A?

Detailed Solution: Question 18

Omitting a relevant variable correlated with included regressors causes bias and inconsistency in coefficient estimates.

Practice Test: Quantitative Methods - 2 - Question 19

An analyst detects both serial correlation and heteroskedasticity in the residuals of her time-series regression model. A colleague recommends applying the Hansen method to correct for these issues. Which of the following best describes the Hansen method's effect on the regression output?

Detailed Solution: Question 19

Hansen method produces HAC (heteroskedasticity and autocorrelation consistent) standard errors; coefficients unchanged.

Practice Test: Quantitative Methods - 2 - Question 20

An analyst regresses a non-stationary equity price index on a non-stationary commodity price index without testing for cointegration or applying any data transformation. The regression output shows an R-squared of 0.91 and a highly significant t-statistic for the slope coefficient. Which of the following most accurately characterizes the likely problem with this regression and the appropriate corrective action?

Detailed Solution: Question 20

Two unit-root series regressed directly produce spurious results; first-difference or test for cointegration.

Practice Test: Quantitative Methods - 2 - Question 21

A credit analyst builds a binary classification model to identify distressed firms. The confusion matrix for the model is as follows:


Predicted DistressedPredicted Healthy
Actual Distressed12050
Actual Healthy30100

The analyst is specifically concerned about the model's ability to identify all truly distressed firms. Which metric and value best addresses this concern?

Detailed Solution: Question 21

Recall (sensitivity) = TP/(TP+FN) = 120/(120+50) = 120/170 ≈ 0.71.

Practice Test: Quantitative Methods - 2 - Question 22

A research analyst presents regression results for a five-factor model explaining hedge fund returns. The F-statistic is 28.4 with a corresponding p-value of 0.0001. A colleague interprets this as evidence that every individual factor is a significant predictor of returns. Which of the following best evaluates this interpretation?

Detailed Solution: Question 22

F-statistic tests joint significance of all slopes; individual significance requires separate t-tests.

Practice Test: Quantitative Methods - 2 - Question 23

A factor model analyst observes the following regression diagnostics: R-squared = 0.87, F-statistic p-value = 0.0002, yet t-statistics for all four individual slope coefficients are below 1.50 in absolute value. The pairwise correlation matrix shows correlations between 0.85 and 0.95 among all four independent variables. Which condition is most consistent with these observations?

Detailed Solution: Question 23

High R², significant F-test, low individual t-stats, and high inter-predictor correlations indicate multicollinearity.

Practice Test: Quantitative Methods - 2 - Question 24

An analyst suspects that her cross-sectional regression of stock returns on firm characteristics exhibits conditional heteroskedasticity. She applies a formal statistical test. Which of the following correctly describes the Breusch-Pagan test for conditional heteroskedasticity?

Detailed Solution: Question 24

Breusch-Pagan regresses squared residuals on independent variables; significant result indicates conditional heteroskedasticity.

Practice Test: Quantitative Methods - 2 - Question 25

A fixed income analyst estimates an AR(1) model for monthly changes in the 10-year Treasury yield. The estimated model is: xt = 0.12 + 0.95xt-1 + εt. A colleague claims that because the lag coefficient is close to 1, the series has a unit root and cannot be used for forecasting. Which of the following best evaluates this claim?

Detailed Solution: Question 25

AR(1) is covariance stationary if |b₁|<1; here=""><1, so="" it="" is="" stationary="" and="" mean-reverting,="" though="" slowly.="" so="" it="" is="" stationary="" and="" mean-reverting,="" though="">

Practice Test: Quantitative Methods - 2 - Question 26

A quantitative analyst estimates the following multiple regression model for annual fund returns:

Ŷ = 1.20 + 0.35X1 − 0.18X2 + 0.55X3

where X1 = market excess return = 5%, X2 = inflation surprise = 3%, and X3 = credit spread change = 2%. What is the predicted fund return?

Detailed Solution: Question 26

Ŷ = 1.20 + 0.35(5) - 0.18(3) + 0.55(2) = 1.20 + 1.75 - 0.54 + 1.10 = 3.51%.

Practice Test: Quantitative Methods - 2 - Question 27

A risk model developer is building a regression model to forecast credit default swap spreads using 50 candidate predictor variables. She believes only a small subset of these variables has true predictive power and wishes to produce a sparse, interpretable model. Which of the following best describes the most appropriate regularization approach and its rationale?

Detailed Solution: Question 27

LASSO (L1 penalty) can zero out irrelevant coefficients, producing sparse models ideal when few predictors are relevant.

Practice Test: Quantitative Methods - 2 - Question 28

An economist estimates the AR(1) model xt = 0.60 + 0.70xt-1 + εt for quarterly output growth. She wants to identify the long-run equilibrium level toward which the series reverts. Which of the following correctly states the mean-reverting level and the condition required for mean reversion to exist?

Detailed Solution: Question 28

Mean-reverting level = b₀/(1-b₁) = 0.60/0.30 = 2.00; requires |b₁|=0.70<1.>

Practice Test: Quantitative Methods - 2 - Question 29

A researcher regresses a non-stationary consumer price index (CPI) series on a non-stationary equity index series. Both series are confirmed to have unit roots via the Dickey-Fuller test, and the researcher does not test for cointegration prior to estimation. The regression produces an R-squared of 0.88 and a slope t-statistic of 9.4. Which of the following best characterizes this result?

Detailed Solution: Question 29

Two non-cointegrated unit-root series regressed directly produce spurious high R² and significant t-stats.

Practice Test: Quantitative Methods - 2 - Question 30

A quantitative analyst applies 5-fold cross-validation to evaluate a neural network model predicting equity factor returns. The training fold accuracy averages 94% across all folds, while the validation fold accuracy averages 51%. She considers applying regularization to address the observed performance gap. Which of the following most accurately characterizes the model's condition and the role of regularization?

Detailed Solution: Question 30

Large train-validation accuracy gap indicates overfitting; regularization penalizes complexity to reduce it.

10 docs|3 tests
Information about Practice Test: Quantitative Methods - 2 Page
In this test you can find the Exam questions for Practice Test: Quantitative Methods - 2 solved & explained in the simplest way possible. Besides giving Questions and answers for Practice Test: Quantitative Methods - 2, EduRev gives you an ample number of Online tests for practice
Download as PDF