Assignment -

Answer all 4 questions. For the STATA questions, I need to see the do-files and a log file demonstrating that your do-file/program actually works.

1. For the Nerlove paper, in a lab session we conducted a test of structural change on all the parameters for the five regimes. Carry out the corresponding test if all we are concerned about is the constancy of the returns to scale parameter and have nothing to say about the other parameters in the regression equation.

2. Suppose we postulate that the variable y is a continuous and (possibly piecewise) linear function of the variable x (apart from an additive iid disturbance term). We admit the possibility that the slope of y with respect to x may take one value if x < x₀ and may take another value if x > x0. How will you test for this against the null hypothesis that the slope is constant throughout? For the dataset Takehome18.xlsx posted on the course site conduct the test if x₀ = 10.

3. Every week a manger receives a transaction file. For simplicity, suppose the file has observations on only two variables. Each observation is a record on a purchase by a customer. The first variable (called code) is a customer code and the second variable (called amount) is the dollar amount of that transaction. Note that the same customer may be involved in several transactions. Write a do-file that will compile a list (to be displayed as a matrix) where for each unique customer code it will show the total amount of dollar transactions by that customer during the week. Create a mock transaction file and show that acting on it, your do file does the job it is supposed to do. Explain in detail the logic behind your do file.

4. Consider the model: y_t = α + βt + ε_t; t = 1, . . . , T. Let ε_t be iid with mean 0 and variance σ². Let b_T be the least squares estimator of β when we have T observations. Is b_T consistent for β? Provide arguments.

Attachment:- Assignment Files.rar