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



Table IV: Ex-post model diagnostics and ex-ante autocorrelation

Heteroskedasticity

LIN

THM

SPA

ZERO

A1

H2

H3

H4

ST

AR1

AR12

AR1

AR12

AR1

AR12

AR1

AR12

AR1

AR12

Exports

AT

1

2

0

1

3

0

0

1

1

1

2

0

1

1

2

BE

2

2

1

3

3

1

1

1

5

2

5

0

5

1

5

CA

2

5

0

6

1

2

5

2

9

2

10

1

8

2

8

FI

2

5

0

3

2

1

0

2

4

2

4

2

4

2

4

FR

2

2

0

4

2

0

2

1

5

0

4

1

3

1

6

GE

4

3

0

3

0

0

2

3

0

3

0

3

1

3

0

GC

7

5

1

3

0

1

1

2

3

1

3

2

3

1

3

IR

1

3

1

4

2

1

0

3

1

3

1

2

2

3

2

IT

3

10

2

10

6

1

2

2

6

1

6

0

6

2

5

JP

1

2

0

2

7

1

4

2

8

2

7

2

5

2

8

NL

2

3

1

2

1

0

4

1

5

1

4

1

2

1

4

NO

3

2

0

2

2

0

0

0

2

0

1

0

2

1

2

SW

1

2

3

3

4

2

3

3

5

2

4

2

6

2

6

UK

2

4

4

3

1

0

3

2

4

2

5

3

5

2

4

US

2

7

1

7

1

0

3

0

3

0

3

0

2

0

3

agg

35

57

14

56

35

10

30

25

61

22

59

__19

55

24

62

Imports

AT-

1

0

1

1

1

0

3

2

6

2

6

1

6

1

6

CA

1

2

1

2

1

1

5

0

6

0

6

1

5

1

6

FI

1

3

0

1

0

1

0

0

3

0

3

0

2

0

3

FR

0

4

1

3

2

1

0

2

1

2

1

1

2

2

2

GE

1

2

0

0

2

2

5

3

5

3

5

2

3

3

6

GC

2

7

0

7

2

1

2

0

3

1

4

1

4

0

3

IR

1

3

1

5

2

3

3

3

5

3

5

3

5

3

5

IT

4

6

2

8

5

1

3

2

6

2

6

2

5

2

6

JP

2

3

0

3

1

0

0

3

4

2

4

0

1

3

3

NL

2

1

1

1

0

1

4

1

5

1

5

1

2

1

5

NO

0

1

0

1

1

1

0

1

2

0

1

1

1

1

2

SW

4

2

2

3

2

1

4

1

4

2

4

0

1

1

3

UK

0

2

0

3

1

0

3

1

5

1

6

1

5

1

5

US

3

8

2

8

3

1

2

4

8

4

9

5

9

5

8

agg

22

44

11

46

23

14

34

23

63

23

65

19^

51

24

63

Notes: Absolute frequencies of rejecting particular null hypotheses at the 5% significance
level for various in-sample diagnostics (left hand side panels) and serial correlation of one
step ahead forecast errors (right hand side). The homoskedastic model is tested against
ARCH(1), shift in variance, variance governed by volatility, and trend in variance (H1 to
H4). ST is short for an F-test on structural stability. LM1 and LM12 are LM-tests against
first order autocorrelation and autocorrelation up to order 12. LIN, THM, SPA and ZERO
are short for four competing forecasting schemes, the linear regression (11), a threshold
model (16), the semiparametric model (12) and unconditional ”zero” forecasts. Ten sectors
are modelled for each country such that 10 is the maximum entry per cell.



More intriguing information

1. National curriculum assessment: how to make it better
2. The name is absent
3. Macroeconomic Interdependence in a Two-Country DSGE Model under Diverging Interest-Rate Rules
4. Human Rights Violations by the Executive: Complicity of the Judiciary in Cameroon?
5. The name is absent
6. The name is absent
7. The name is absent
8. The name is absent
9. PEER-REVIEWED FINAL EDITED VERSION OF ARTICLE PRIOR TO PUBLICATION
10. AN ANALYTICAL METHOD TO CALCULATE THE ERGODIC AND DIFFERENCE MATRICES OF THE DISCOUNTED MARKOV DECISION PROCESSES