Homework 5
Homework 5
Eco 4306 Economic and Business Forecasting
Spring 2020
Due: Thursday, March 12, before the class
Problem 1
Obtain the annual data for percentage change in Per Capita Personal Income in Texas for the period 1947
to today from FRED, code TXPCPI. Note that this time series on FRED is in dollars, so you either need to
switch the units to percentage change directly on FRED or construct the percentage change yourself.
(a) Change your sample to XXXXXXXXXXUse this data to recreate the two panels in Figure 7.7, but for Texas
XXXXXXXXXXinstead of California XXXXXXXXXX.
(b) Compare your results in (a) with those in Figure 7.7. Based on the shape of AC and PAC, does the
growth in per capita personal income in Texas follow a similar process as the growth in per capita
personal income in California?
(c) Use AC and PAC functions from (a) to determine the order p of the suitable AR(p) model and estimate
this AR(p) model for the period XXXXXXXXXX.
(d) Use the model from (c) to create a forecast for period 2011 to today. Also generate the standard e
ors
for this forecast and use it to construct the lower and upper bounds of the 95% confidence interval fo
the forecast. Plot the actual data together with the forecast and its 95% confidence interval. Change
the options for the chart (line colors, text in legend entries) so that is is clear what the chart shows to
any person who would look at it.
(e) Comment on the accuracy of the forecast.
1
https:
alfred.stlouisfed.org/graph/?g=mZGh
Problem 1
FIGURE 7.7 Per Capita Income Growth (California, XXXXXXXXXX)
12 Sample: XXXXXXXXXX
Included observations: 33
10
Autoco
elation Partial Co
elation
8
6
2
0
XXXXXXXXXX XXXXXXXXXX
- Personal income growth in CA I
§=?
I
1
2
3
4
5
6
7
8
9
10
1 1
AC PAC
XXXXXXXXXX
XXXXXXXXXX
XXXXXXXXXX
XXXXXXXXXX
XXXXXXXXXX
XXXXXXXXXX
XXXXXXXXXX
XXXXXXXXXX
XXXXXXXXXX
XXXXXXXXXX
XXXXXXXXXX
autoco
elations (negative for odd lags and positive for even lags) are consequen~e-... of
the negative sign of the persistence parameter, which also causes the zigzaggy behavior
that we observe in the time series plot.
REAL DATA: Per Capita Personal Income Growth in California, XXXXXXXXXX