A student has submitted her final dissertation in which she has estimated the following model using Zambian data:
Dln Const= B0+ B1Dln GDPt+ B2Dln CPIt+ B3DRt+ et
WhereDis the first difference operator,
Consis private consumption
GDP is gross domestic product,
CPI is consumer price index,
R is the three-month Treasury bill rate.
The sample is: 1969 (2) to 1994 (1)
|
Variable
|
Coefficient
|
Std.Error
|
t-value
|
HCSE
|
|
Constant
|
0.01692
|
0.00338
|
5.001
|
0.004597
|
|
Dln GDPt
|
0.36086
|
0.09173
|
3.934
|
0.093796
|
|
Dln CPIt
|
0.25917
|
0.16066
|
1.613
|
0.241880
|
|
DRt
|
0.00131
|
0.00103
|
1.269
|
0.001131
|
R2= 0.174366
F(3,96) = 6.7581 [0.0003]
DW = 2.35
RSS = 0.02298197499 for 4 variables and 100 observations
AR 1- 2 F( 2, 94) = 1.8497 [0.1630]
ARCH 4 F( 4, 88) = 2.6102 [0.0408] *
Normality X2(2)= 4.4048 [0.1105]
RESET F( 1, 95) = 0.94744 [0.3328]
a) Identify the problems in the estimated model.
b) What can be done to improve the estimation?