Question 1

Which of the following best describes a dynamic causal effect in econometrics?

Accepted Answer

The effect of a change in X on Y over multiple time periods.

Question 2

In the context of the Orange Juice data, what does FDD represent?

Accepted Answer

The number of freezing degree-days during the month in Orlando, FL.

Question 3

Suppose you estimate the following distributed lag model: $$Y_t = \$$beta_0$$ + \$$beta_1$$ X_t + \$$beta_2$$ X_{t-1} + \$$beta_3$$ X_{t-2} + u_t. $$What is the 2-period cumulative dynamic multiplier?

Accepted Answer

$$\$$beta_1$$ + \$$beta_2$$ + \$$beta_3$$

Question 4

Which assumption is required for OLS to yield consistent estimators of dynamic multipliers in the distributed lag model?

Accepted Answer

Exogeneity of X: $$E(u_t | X_t, X_{t-1}, X_{t-2}, ...) = 0$$

Question 5

Why are conventional OLS standard errors incorrect in time series distributed lag models?

Accepted Answer

Because the errors and regressors can be serially correlated.

Question 6

What is the purpose of HAC (Heteroskedasticity and Autocorrelation Consistent) standard errors?

Accepted Answer

To provide standard errors robust to both heteroskedasticity and autocorrelation.

Question 7

In the Newey-West estimator for HAC standard errors, what does the truncation parameter $$m$$ represent?

Accepted Answer

The number of autocorrelation terms considered in estimating the variance.

Question 8

Given the distributed lag model $$Y_t = \$$beta_0$$ + \$$beta_1$$ X_t + \$$beta_2$$ X_{t-1} + u_t$$, how can you rewrite the model to directly estimate the cumulative multiplier and its standard error?

Accepted Answer

$$Y_t = \$$beta_0$$ + \$$beta_1$$ \Delta X_t + (\$$beta_1$$ + \$$beta_2$$) X_{t-1} + u_t$$

Question 9

Which of the following is NOT an assumption of the distributed lag model for estimating dynamic causal effects?

Accepted Answer

Y and X are strictly exogenous.

Question 10

If the QLR statistic for testing stability of regression coefficients is 21.19 and the critical value at the 1% significance level for 18 restrictions is 2.53, what can you conclude?

Accepted Answer

There is evidence of instability in the regression coefficients at the 1% significance level.

Question 11

In the Orange Juice example, what is the effect of a unit increase in FDD on OJ prices according to the initial regression?

Accepted Answer

A statistically significant increase in OJ prices.

Question 12

Which formula represents the variance of the OLS estimator $$\hat{\beta}_1 in $$time series data with autocorrelation?

Accepted Answer

$$\frac{1}{T} \frac{\sigma_v^2$}{(\sigma_X^2$)^2$$} 	imes f_T$

Question 13

What is the rule of thumb for choosing the Newey-West truncation parameter $$m$$?

Accepted Answer

$$m = 0.75 	imes T$$^{1/3}$$

Question 14

Why is strict exogeneity rarely plausible in time series applications such as the Orange Juice example?

Accepted Answer

Because future values of X can affect current errors.

Question 15

In the distributed lag model, what does $$\$$beta_1$$ represent?

Accepted Answer

The impact effect of a change in X on Y at time t, holding past X constant.

Question 16

If $$Y_t$$ and $$X_t$$ are stationary, what does this imply for the distributed lag model?

Accepted Answer

The coefficients do not change within the sample, ensuring internal validity.

Question 17

Which of the following best describes the cumulative dynamic multiplier in the context of the Orange Juice example?

Accepted Answer

The sum of distributed lag coefficients representing the effect through date t+h.

Question 18

What is the main reason for using the distributed lag estimator in time series applications?

Accepted Answer

Because we only have one subject (e.g., the US OJ market) observed at different times.

Question 19

Which of the following statements about HAC standard errors is correct?

Accepted Answer

They are robust to both heteroskedasticity and autocorrelation.

Dynamic Causal Effects and Distributed Lag Models in Time Series Econometrics

스터디 가이드 - 연습 문제

학습 가이드 - 플래시카드