use E-views

ECON 402

Problem Set 7

Gonzalo García – Spring 2019

Due on ELMS at 6 pm on April 18, 2019.

No late submissions are accepted.

For this problem set, I allow and in fact encourage working in groups of no larger

than 3. If you work in a group, everyone needs to submit the problem set

individually and at the beginning of the problem set list the group members.

You will write Eviews codes that perform certain analyses. You are allowed to use

parts of the codes we used in class. You are not allowed use any other code you

have not written yourself. Copy paste the content of the code in your submission

and also upload your prg file as a separate attachment. In your submission briefly

explain your results.

1. Fetch the monthly data for All Employees: Total Nonfarm Payrolls (NSA)

from Fred for the period 1995:m1 2019:m3 (Fred id: paynsa). Then, do the

following:

a. Using the sample 1995:m1 2019:m3, go through the motions of

figuring out the best model that captures the trend, seasonality and

the cycle as we did in class (we used the same variable but a

different subsample – so no reason to assume that the best model

will be identical)

b. Produce three forecasts for the period 2019:4 to 2021:12 using three

approaches:

i. Full model

ii. Only trend and seasonality (for this re-estimate the model you

identified in (a) leaving out the cycle terms and use it for the

forecast)

iii. Only trend and cycle (for this re-estimate the model you

identified in (a) leaving out the seasonal terms and use it for

the forecast)

This exercise will show us the contribution of seasonality and

cycle to the forecasting power. You should plot each of the forecasts

along with their uncertainty bands (three separate plots).

The Eviews code you write should produce all the output you use in one run. As

usual, you’ll need to make some decisions along the way. So do this

incrementally. Write the code to the point where you need to make a decision.

Look at what you need to look at and continue. You need to show me what

decisions you made and how you made them.

2. Using the full model above, produce an Excel table (just copy paste the

table to your write-up) that has the 1-,6- and 12-month ahead forecasts

using data up to 2018:3, 2018:9 and 2019:3. The table should look like this

Forecast Date Forecast_1M Forecast_6M Forecast_12M

2018M3

2018M9

2019M3

For this question, you do not need to write a code and can produce the

results interactively.