Другие предметы \ Проектирование

Model-free evaluation of directional predictability in foreign exchange markets, страница 19

Table III. (Continued)

Currency

c D 0

Positive direction

c D 0.5

c D 1.0

Negative direction

c D 0

c D 0.5

c D 1.0

MZID1,4

AD

0.26

0.68

0.17

0.39

0.38

CD

1.13

1.52

0.94

1.37

0.48

0.66

BP

4.10

13.72

5.95

3.84

0.52

8.37

JY

0.19

0.37

1.69

0.11

0.46

SF

1.87

0.07

2.82

2.51

9.11

9.00

DM

0.16

0.27

6.38

0.23

1.52

5.85

FAD

0.64

1.41

0.36

0.43

0.41

FCD

2.83

0.68

2.12

5.74

1.22

FBP

1.07

11.26

9.60

4.82

0.04

4.64

FJY

0.71

0.92

1.16

FSF

0.58

9.04

11.24

0.97

0.09

2.30

FDM

1.26

4.52

3.66

0.89

0.85

1.49

MZZID1,0

AD

4.53

0.94

6.63

4.59

4.80

1.01

CD

8.96

3.12

0.43

9.13

0.38

4.81

BP

0.33

15.52

1.03

0.05

5.56

18.09

JY

0.57

9.77

SF

4.70

0.67

4.31

5.27

11.57

DM

0.07

2.51

15.42

FAD

1.80

0.68

1.89

5.78

3.22

1.06

FCD

13.85

2.77

0.48

20.11

10.58

8.13

FBP

0.42

16.38

1.99

0.96

11.50

16.06

FJY

2.53

0.67

2.21

2.40

6.36

4.27

FSF

2.89

13.62

1.88

5.76

1.60

15.25

FDM

4.61

20.00

0.97

3.47

2.86

4.23

Notes:

1.  GCS tests are asymptotically one-sided N(0,1) tests and thus upper-tailed asymptotic critical values may also be used, which are 1.65 and 2.33 at the 5% and 1% levels, respectively. M1,l represents test statistics on the martingale test, serial correlation test, ARCH-in-mean test, skewness-in-mean test and kurtosis-in-mean test for l D 0,...4, respectively.

2.  Interest rate differentials are defined by rt  rtŁ where rt is the domestic (US) risk-free interest rate and rtŁ is the foreign risk-free interest rate.

3.  A preliminary bandwidth p is crucial to run GCS tests. We have computed GCS test statistics for p D 11,...,50, but reported only for the value of p D 21 to save space.

4.  A threshold value c is introduced to forecast bigger changes. A higher threshold value of c implies a bigger change in rate or price.

the directions of past interest rate differentials are more useful to predict the directions of the returns with zero threshold. Such differences are, however, attenuated with large thresholds

Скачать файл