QUESTION 1 1ST PDF Consider the simple AR(1) process: y_t=a+?y_(t-1)+e_t, e_t~i

QUESTION 1 1ST PDF
Consider the simple AR(1) process:
y_t=a+?y_(t-1)+e_t, e_t~i

QUESTION 1 1ST PDF
Consider the simple AR(1) process:
y_t=a+?y_(t-1)+e_t, e_t~iidN(0,1)
z_t=a+?z_(t-1)+e_t, e_t~iidN(0,1)
Generate a sample of 1000 obs (N=1000) using the AR(1) process above, assuming e_t~iidN(0,1). Choose the following values of theta: .9, 1, 1.002, 1.5, -.9. For which value of theta, assign a =0 and a =1. This gives a total of 10 specifications in total. Which of these are weakly stationary? How do you know?
For each value of theta, alpha and for both y and z, estimate the AR(1) regression conforming to the data generating process. Also, regress y on z.
Save the estimates from all three regressions (y_t on y_(t-1), z_t on z_(t-1) y_t on z_t) and the standard error.
Compute the theoretical variance of each observation and then weight each observation in order to estimate the optimal GLS estimator. Then run each of the three regression with optimal GLS weights.
Plot the distribution of the estimates of ?. What shape do the distributions have? What shape would you hope that they have for testing? What problems are introduced when the distributions of the ? do not have the required shape?
Compute the mean standard error of theta for each set of 1,000 regressions. Then compute the bootstrapped standard error ?_(i=1)^N¦v(1/(N-1) ? ^-? ¯ ). Compare the bootstrapped standard errors with the average standard errors from the regressions. What do you notice and thus conclude? Please include graphics and screenshots of the R-code in the answer file if it was used for solving the problem.
Hi. I will send you a file with the answers to similar questions. I hope it will be helpful for you https://files.transtutors.com/cdn/uploadquestions/… . QUESTION 2 SECOND PDF (1.)You have a two equation system:
y_t=a_10+a_11 y_(t-1)+?a_12 z?_(t-1)+e_yt
z_t=a_20+a_21 y_(t-1)+?a_22 z?_(t-1)+e_zt
a.Show that the solution for can be written as
y_t=[(1-a_22 L) e_yt+(1-a_22)a_10+a_12 Le_zt+a_12 a_20]/[(1-a_11 L)(1-a_22 L)-a_12 a_21 L^2]
a.Find a solution for z_t.
b.Suppose that y_t and z_t are CI(1,1). Write the error correction model and show that it contains an intercept term. QUESTION 3 THIRD PDF
Consider ARMA(1,1) process:
y_t=?y_(t-1)+e_t+??e?_(t-1), e_t~iidN(0,?s_^2?^ )
1.For what values of the vector(?,?) is the ARMA process stationary? 2.Show graphically in a two-dimensional graph of in the(?,?) plane, which values are stationary and which are not.