Exchange Rate Uncertainty and Trade Growth - A Comparison of Linear and Nonlinear (Forecasting) Models



Table III: (G)ARCH estimation and diagnostic results

Exports

Imports

e

kt

GARCH(1,1)

ξkt

e

kt

GARCH(1,1)

ξkt

A1

A5

δki

βk1

A1

A1

A5

δki

-------TT------------

βk1

A1

AT-

0.28

0.41

0.13*
(1.95)

0.83*

(8.11)

0.99

0.41

0.52

0.12*
(1.93)

0.84*
(8.74)

0.95

BE

0.28

0.27

0.16*

(1.87)

0.47

(1.52)

0.47

CA

0.50

0.76

0.05

(0.74)

0.53

(1.06)

0.85

0.59

0.89

0.05

(0.66)

-

0.86

FI

0.00*

0.01*

0.12*

(1.71)

0.79*
(6.04)

0.63

0.00*

0.02*

0.13*

(1.80)

0.78*
(6.06)

0.64

FR

0.09

0.31

0.11

(1.16)

0.86

0.20

0.56

0.10

(1.08)

0.95

GE

0.37

0.69

0.11

(1.38)

0.71

0.36

0.47

0.10

(1.20)

0.76

GC

0.00*

0.00*

0.18*

(2.66)

0.82*

(11.56)

0.09

0.00*

0.00*

0.19*

(2.76)

0.81*

(11.26)

0.12

IT

0.00*

0.00*

0.15

(1.61)

0.79*

(5.33)

0.85

0.00*

0.00*

0.15*

(1.67)

0.79*

(5.43)

0.93

IR

0.15

0.03*

0.04

(0.86)

0.70*

(2.21)

0.65

0.01*

0.11

0.11*
(1.68)

0.58*
(2.76)

0.86

JP

0.45

0.44

0.05

(0.77)

0.71*

(1.77)

0.94

0.58

0.46

0.05

(0.71)

0.71*

(1.68)

0.87

NL

0.01*

0.22

0.13

(1.31)

0.63

0.02*

0.15

0.10

(1.09)

0.28

(0.47)

0.71

NO

0.33

0.07

0.15*

(2.28)

0.74*

(6.17)

0.50

0.23

0.07

0.19*

(2.44)

0.72*

(6.10)

0.52

SW

0.01*

0.00*

0.06

(1.61)

0.91*

(18.34)

0.65

0.02*

0.02

0.05

(1.17)

0.92*

(14.29)

0.64

UK

0.01*

0.16

0.11*
(1.66)

0.71*

(4.40)

0.64

0.01*

0.13

0.10

(1.57)

0.70*

(4.09)

0.56

US

0.68

0.65

0.043
(0.85)

0.76

(1.64)

0.60

0.69

0.56

0.05

(1.12)

0.82*

(3.10)

0.51

Notes: p-values obtained from ARCH LM-tests of orders 1 and 5 (A1, A5) for ∆etk and LM-
tests on remaining ARCH effects in
ξkt (A1). Parameter estimates (t-ratios in parentheses)
for the GARCH(1,1) or, if indicated, for ARCH(1) processes. A star indicates significance
at the 5% level. Owing to parameter restrictions we use one sided tests for the (G)ARCH-
parameters. All
t-ratios are obtained from Quasi Maximum Likelihood estimation.39



More intriguing information

1. Volunteering and the Strategic Value of Ignorance
2. Can genetic algorithms explain experimental anomalies? An application to common property resources
3. Modelling the health related benefits of environmental policies - a CGE analysis for the eu countries with gem-e3
4. Improving Business Cycle Forecasts’ Accuracy - What Can We Learn from Past Errors?
5. Problems of operationalizing the concept of a cost-of-living index
6. The name is absent
7. An institutional analysis of sasi laut in Maluku, Indonesia
8. The name is absent
9. Language discrimination by human newborns and by cotton-top tamarin monkeys
10. Categorial Grammar and Discourse