Турунцева М.Ю. Анализ временных рядов, 2003

-
-
( )
2003
I.
EViews
......................................................................................................... 3
1. EViews ............................................................ 3
1.1. EViews ........................................................ 4
......................................................... 9
1.2. EViews ........................................... 9
....................................................... 12
2. ..................................................... 12
2.1. ....................................................... 12
....................................................... 16
2.2 ................................................ 17
....................................................... 20
2.3. ...................................... 20
....................................................... 23
2.4. ...................... 23
....................................................... 31
2.5. (
)..................... 31
....................................................... 33
2.6. ARIMA( p, d , q ) .............................. 33
....................................................... 45
2.7. ( -) .......................... 45
....................................................... 47
2.8. ARMA(p, q)....................... 48
....................................................... 51
3. ................................................... 51
3.1. (
) ....................................................................................................................... 51
....................................................... 51
3.2. ...................... 52
....................................................... 54
3.3. ..................................................................... 54
3.4. ................................................... 57
3.5. ...................................................... 58
3.3-3.5 ........................ 62

EViews 4.1 .
,
:

.
,
EViews (1), (2)
(3).
, EViews
.
,
,
ARIMA(p, d, q),
.
, ,
,
, ,
.
1. EViews
EViews :
EViews
File Edit Object View Procs Quick Options Window Help
Welcome to EViews
Path=c:\eviews\
DB=macro
WF=none
. Eviews Eviews
, .
Eviews 9
(File, Edit, Object, View, Procs,
Quick, Options, Window, Help),

,
.
. EViews ,
, , ,
.
Enter .
() EViews,
. , ,
, , ,
..

Eviews.
1.1. EViews
1. , ,
,
File/New/Workfile, :
Workfile Range
Frequency
Annual
Weekly
Semi-annual
Daily [5 day weeks]
Quarterly
Daily [7 day weeks]
Monthly
Undated or irregular
Range
Start date
94:1
OK
Cancel
End date
2003M6
:
, (,
, , , ,
()) .
(Annual). (Start date) (End date)
,
19**. , 1994 2002 .,
1994 94.

, 2002.
(Quarterly):
19** .
, 3- 1998 1998:3 98:3
1998Q3. , (4
), Q
( 1 4).
(Monthly). :
M ( 1 (
01) 12). , 1998 1998:08,
1998:8, 98:08 1998M8.
(Semi-annual):

S, , 94:1 ( 1994 ) 2003S1
( 2003 ).
(Weekly Daily).
,
()
:: (mm:dd:yyyy).
07:10:1974,
, 10 1974
, , 10 1974.
(Undated or irregular).
( ),
.
,
94 2002 ,
, :
1994-2002
1994:1-2002:2
1994:1-2002:4
1994:01-2002:12
01/01/1994-12/28/2002
01/01/1994-12/31/2002
94-2002
,
, , EViews
( ) :
Wokrfile: UNTITLED
View Procs Object
Save Label+/Range: 1994:01 2003:06
Sample: 1994:01 2003:06
Show Fetch Store Delete Genr Sample

Filter: *
Default Eq: None
c
resid
, Range ,
, Sample
. 1.2
EViews.
2. EViews
. ,
, Microsoft Excel .
EViews File/Import/Read Text-LotusExcel
? X
Open
Look in:
C:\
File name
Files
type
Open
of
Excel(*.xls)
Cancel
Update default directory
Microsoft Excel, ,
. ,
Excel Spreadsheet Import
Data order
Upper-left data cell
B2
By Observation series in columns
E
E
m
Exxxccceeell5
l55+++ssshhheeeeeettt nnnaaam
meee
By Series series in rows

Names for series or Number if named in file
5
6
Import sample
1994:01 2003:06
Reset sample to:

Current sample
Workfile range
E
E
Exxxpppooorrrttt oooppptttiio
ioonnnsss
W
r
i
t
e
d
a
t
W
Wrriittee ddaatteee///ooobbbsss
E
V
w
m
E
EV
Viie
ieew
wsss dddaaattteeefffooorrrm
maaattt
FFFiirirrsssttt cccaaalle
leennndddaaarrr dddaaayyy
LLLaaasssttt cccaaalle
leennndddaaarrr dddaaayyy
W
W
m
Wrrriititteeessseeerrriie
ieesss nnnaaam
meee
A
S
C
I
I
T
e
x
t
d
e
l
i
m
i
t
A
AS
SC
CIIII--TTeexxtt ddeelliim
miitteeerrr
TTTaaabbb
S
S
Spppaaaccceee
C
m
m
C
Cooom
mm
maaa
To end pf range
OK
Cancel
,
Microsoft Excel,
EViews, , .
,
,
EViews. -,
Microsoft
Excel 4.0, , Microsoft Excel.
Excel5+sheet name.
Microsoft Excel, ,
, , .
, ,
.
-, ,
, Microsoft Excel (

), , ,
. ,
Microsoft Excel ,
EViews
, Microsoft Excel,
EViews.
-, EViews
Microsoft Excel. , , Microsoft
Excel, ,
.
(Data order). Microsoft Excel
, By Observation
series in columns ( ). ,
, By Series series in rows.
,
(Upper-left data cell).
, .
2, ,
, ,
.
Microsoft Excel (Excel5+sheet name).
Microsoft Excel,
,
Microsoft Excel. Microsoft Excel
4.0, .
()
, Microsoft Excel
(Names for series or Number if named in file).
Microsoft Excel,
()
, :
, .
Microsoft Excel,
,
.
(Import sample).
, ,
() , (
, ,
, EViews NA).
,
OK, EViews,
, , :
Wokrfile: UNTITLED
View
Procs
Object
Range: 1994:01 2003:06
Sample: 1994:01 2003:06
Save
Label+/Show
Filter: *
Fetch
Store
Delete
Genr
Default Eq: None
_
Sample
resid
y1
y2
y3
y1, y2, y3 .

1. EViews 3.1 EViews 4.1

2. EViews,

1000
1.2. EViews
.
EViews Windows,
, , Windows,
, Microsoft Word, Excel ..
EViews .
EViews.
, EViews,
File/Save File/Save as ,
Save (
).
, ,
File/Open EViews
.
.
EViews 1,
() :
@all , ;
@first ;
@last .
.
, .
,
smpl -1 -2
,
1993 1998 .

smpl @all
smpl @first 1996:12
smpl 1995:08 @last,
, ,
,

1996 ., , ,
1995 . .
-
,
, , :
smpl 1993:09 1997:1.

1993 1997 .
, EViews
.
, 1993 1997 .

smpl @first+9 @last-10.
.
Sample
,

, , EViews, @.
10
Sample
Sample range pairs (or sample object to copy)

2
1993:01 1998:07
OK
IF condition (optional)
Cancel
,
.
Quick EViews.
Quick
EViews Sample,
.
EViews
EViews
(),
:
(Coefficient Vector)

(Equation)
=
(Graph)
(Group)
G
(LogL)
#
() (Scalar)
() (Series)
(State space)
ss
S
(System)
(SYM Symmetric Matrix)
(Matrix)
[]
(Model)
(Pool)
P
(Sample)
Table
(Table)
TXT
(Text)
VAR
2
,
.
11
(VAR Vector Autoregression)

( ) (Vector/Row Vector)
,
Object/New Object EViews
New Object
Type of object
Name for object
Sample
Untitled
Equation
Graph
Group
LogL
Matrix-Vector-Coef
Model
OK
Pool
Sample
Series
Sspace
Cancel
System
Table
Text
VAR
,
OK, , ,
,
(, ),
, .

1.
2.
, ,
DEMO.wf.

y1, y2, y3, y4, y5.
2.
2.1.
EViews
, EViews.
, ,
.
12
3 :
Series: y1
Workfile: CP-2
View Procs Object
Print Name Freeze
Transform Edit+/- Smpl+/- Label+/- Wlde+- InsDel Title Sample Genr

Y1
Last updated: 03/20/03 - 12:31
Modified: 1 1000 // y1=-2

Modified: 2 1000 // y1=0.5*y1(-1)+nrnd
1
-2
-0.9566483
1.3842802
1.0306127
-0.3917438
-1.6905064
View :
SpreadSheeet ;
Graph (
. 2.2);
Descriptive Statistics
: , .. (
. 2.3);
Tests for Descriptive Statistics
( . 2.4);
Distribution ,
;
One-Way Tabulation
, :
,
,
, %, ,
( %);
Correlogram (
. 2.5);
Unit Root Test
( . 2.7);
BDS Independence Test
.
,
, -
( ),
.
13
. :
, ;
Conversion Options EViews

;
Label , :
, ,
, , , , .
Procs :
(Generate by Equation
Genr );
(Resample),
;
(Seasonal Adjustment )
: Sensus X12; X11
(Historical); Tramo/Seats; Moving Average Methods;

(Exponential Smoothing);
-
(Hodrick-Prescott Filter).
Object ,

:
Store to DB EViews;
Update from DB ,
EViews;
Copy Object ;
Name ( Name
);
Delete EViews (
Delete );
Freeze Output (
Freeze );
Print ( Print
);
View Options
, :
Sample +/-;
Edit +/-;
Label +/-;
Wide +/-;
InsDel;
Title
Print . Print
.
Name . Name
.
Freeze .
.
14
Transform4 .
(Level) :
1 Period Difference :
yt yt 1
1 Year Difference
:
yt yt 12
. ..
.
1 Period % Change :
yt yt 1
100%
yt

:
1 Period % Change (Annual Rate)
12
y y
t 1
100%
t
1
1
+
yt
1 Year % Change
:
yt yt 12
100% ,
yt
. ..
.
Edit+/- . /

. ,
Enter.
Smpl+/- .
,
(),
.
Label+/- . Label+/-
( )
, ..
Wlde+- . Wlde+-
( ).
InsDel .
/ .
, Insert
obs and move subsequent obs down one. ,
, ( ),
, .
EViews 3.1.
15
,
.
- , Delete obs
and move subsequent obs up one.
.
Title .
(
) :
.
Sample .
( . .1.2).
Genr .

,
EViews. ,

:
Generate Series by Equation
Enter equation
ind=
Sample
1994:01 2003:06
OK
Cancel
Sample ,
.

1. DEMO.wf :
wn=nrnd*
y1=a0+a1*y1(-1)+wn
y2=a0+a2*y2(-1)+wn,
a0, a1, a2 ,
,
.
* nrnd EViews ,
,
16
y1(-1) ,
.,
y1 y2, -,
( ,
, , Edit ).
-,
Generate Series by Equation
2 1000.
2. .
2.2
, ,
. View
Graph. :
5 (Line)
. ( ),
.
110
100
90
80
70
60
50
93
94
95
96
97
98
99
00
01
02
IN D
(Bar) .
,
.

(IND) 1993 . 2003 ( 1993 . =
100)
17
110
100
90
80
70
60
50
93
94
95
96
97
98
99
00
01
02
IN D
(Spike)
, (
).
110
100
90
80
70
60
50
93
94
95
96
97
98
99
00
01
02
IN D
(Seasonal Stacked Line)

. :
, , ..
. ,
.
18
IND by Season
110
100
90
80
70
60
50
Jan Feb Mar Apr May Jun

IND
Jul
Aug Sep Oct
Nov Dec
Means by Season
(Seasonal Split Line)

. .
.
IN D b y S eason
110
100
Jan
F eb
M ar
Apr
M ay
Ju n
Ju l
Aug
S ep
O ct
N ov
D ec
90
80
70
60
50
93
94
95
96
97
98
99
00
01
02
03
,
Freeze
. Name
, ( )
EViews.

Object/New Object/Graph EViews.
,
. ,
, () EViews.
19
graph,
.
()
graph < >.< ()> <
( )>
Enter.
EViews :
bar
errbar
,
.
,
. , ,
,
.
hilo
4- .
errbar.
.
line
( ,
)
pie
(
)
scat
xy, ( )
spike
xy
(x, y). (OX)

,
. ,
y(x)
xyline
xy
xypair
xy
1.
wn, y1,
y2, 2.2.

2.3.
Descriptive Statistics View
,
.
Histogram and Stats
Stats Table
Stats by Classification
20
(Histogram and Stats)

( ,
, )
24
Series: IND
Sample 1993:01 2002:02
Observations 110
20
16
12
8
4
0
50
60
70
80
90
Mean
Median
Maximum
Minimum
Std. Dev.
Skewness
Kurtosis
64.91535
62.88370
107.1960
50.82430
10.64339
1.744585
6.157356
Jarque-Bera
Probability
101.4897
0.000000
100
: (Mean),
(Median), (Maximum),
(Minimum), (Std. Dev.),
(Skewness) (Kurtosis), - (JarkeBera) - (Probability).

yt , .
( ) ,
,
/
, .
:
:
T
(y
t =1
T 1
;
:
3
1 y
S = t
,
T
T 1
T
,
. ,
, , .
21
,
, ;
:
4
1 T y
K = t
.
T t =1
3.
3,
3.
, , - ,
. - ,
,
:
T k 2 ( K 3)
S +
6
4
S , K ,
k ( ,
.. ). -
2 (2) .
2
JB =
(Stats Table).
( )
: (Mean),
(Median), (Maximum),
(Minimum), (Std. Dev.),
(Skewness) (Kurtosis), - (Jarke-Bera)
- (Probability), (Sum)
(Sum Sq. Dev.).

(Stats by Classification).
.
, ,
( , , ,
)
1995 .
1998 . 1998 . 2003 .
0, t 1998 : 08,
DTt =
1, .
View/Descriptive Statistics/Stats by Classification
:
Stats by Classification
Statistics
3
X
Series/Group for Classify
Mean
dt
Sum
NA handling
Median
Maximum
Output Layout
Table Display
Treat NA as category
3 Show Row Margins

3 Show Column Margins
3 Show Table Margins
22
Minimum
Group into bins if
Std. Dev.
# of values >
100
List Display
Avg. count <
S
S
w
S
M
Shhhooow
wS
Suuubbb---M
Maaarrrgggiin
innsss
S
p
a
r
s
e
L
a
b
e
l
s
S
Sppaarrssee LLaabbeellss
Quintile
000...555
Skewness
Kurtosis
Max # of bins
# of Nas
3
Observations
OK
Cancel
Options
,
:
Descriptive Statistics for IND
Categorized by values of DT
Date: 08/08/03 Time: 13:04
Sample: 1995:01 2003:05
Included observations: 101
DT
0
1
All
Mean
58.59149
65.57914
62.53502
Median
58.39150
66.40950
62.33490
Max
69.26350
75.67090
75.67090
Min.
51.70410
50.83030
50.83030
Std. Dev.
4.019442
5.516713
6.009558
Obs.
44
57
101
,
, Series/Group for Classify
, .

1. wn, y1, y2.

,
?
2.
.
?
2.4.
Eviews
. View/Tests for
Descriptive Stats/Simple Hypothesis Tests,
,
,
:
Series Distribution Tests
Test Value
X
Mean test assumption
Mean:
65
Median:
63
Mean test will use a known

standard deviation if supplied
Variance:
100
Enter s.d. if known
23
OK
Cancel
, ,
(
1993 .) 1993 . 2003 . 65, 63,
81. :
.
, m=65,
:
H0 : = m
HA : m
yt ,
t-:
m
,
t=
T
&& yt ,
. yt , t-
T-1 .
, EViews
z-:
m
z=
&& yt ,
, . yt
, z-
.

: ,
, 0,05, , 5%-
m ;
,
,
.
. ,
m=63, :
H 0 : med ( yt ) = m
H A : med ( yt ) m
EViews ,
.
24
Binomial sign test.
,
,
,
. EViews P-
, (
).
Wilcoxon signed ranks test. ,

,
( ). ,
, , , ,
. EViews -
t- (
ties) (correcting for both continuity and ties).
Van der Waerden (normal scores) test. ,
, . smoothed ranks.
, ,
.
. ,
=81,
H 0 : var( yt ) = 2
H A : var( yt ) 2
EViews :
2 =
(T 1) 2
2
, . yt
, 2 (T 1) .
,
,
.

:
Hypothesis Testing for IND
Date: 08/08/03 Time: 17:18
Sample: 1993:01 2003:05
Test of Hypothesis: Mean = 65.00000
Sample Mean = 65.66241
Sample Std. Dev. = 10.21339
Method
t-statistic
Value
0.725120
Probability
0.4697
Value
159.6894
Probability
0.0170
Test of Hypothesis: Variance = 81.000000

Sample Variance = 104.3133
Method
Variance Ratio
25
Test of Hypothesis: Median = 63.00000

Sample Median = 64.22410
Value
68
0.894427
1.749391
2.172257
Method
Sign (exact binomial)
Sign (normal approximation)
Wilcoxon signed rank
van der Waerden (normal scores)
Probability
0.3712
0.3711
0.0802
0.0298
Median Test Summary

Category
Count
Mean Rank
68
57
0
68.3529412
56.6140351
Obs > 63.00000

Obs < 63.00000
Obs = 63.00000
Total
125
, 5%-
,
1993 . 2003 . 65 ;
81 ;
.
, View/Tests for Descriptive Stats/Equality Tests by
Classification,
,
.
,
,
: 1995 . 1998
1998 . 2003 . ,
() () () (Series/Group for
classify),
: (Mean),
(Median) (Variance)
Tests By Classification
Series/Group for classify
X
NA handling
dt
Treat NAs as category
Test equality of
Group into bins of
Mean
# of values >
100
Median
Avg. count <
Max # of bins
Variance
OK
Cancel
. ,
(ANOVA)
: , ,

, ,
26

.
i- g as y g , i , i=1,, n g
g=1, 2,,G. (Between) (Within)
:
2
SS B = n g ( g )
G
g =1
ng
SSW = ( y g , i g ) ,
G
g =1 i =1
,
, g ,
g=1,, G. F-
:
SS B (G 1)
F=
,
SSW (T G )
. F- F-
(G-1, -G) , ,
o.
Test for Equality of Means of IND
Date: 08/11/03 Time: 13:33
Sample: 1995:01 2003:05
Method
df
Value
Probability
99
(1, 99)
7.073519
50.03467
0.0000
0.0000
Source of Variation
df
Sum of Sq.
Mean Sq.
Between
Within
1
99
1212.464
2399.015
1212.464
24.23248
100
3611.479
36.11479
Mean
58.59149
65.57914
62.53502
Std. Dev.
4.019442
5.516713
6.009558
Std. Err.
of Mean
0.605954
0.730707
0.597973
t-test
Anova F-statistic
Analysis of Variance
Total
Category Statistics
DT
0
1
All
Count
44
57
101
G=2, EViews t-,

F- T-G .
:

(Between) (Within)
27
Between =
Within =
SS B
df B
SSW
,
dfW
df B df W . F-

Between
F=
.
Within
. EViews ,
,
, ,
, .
,
.
, ,
, ,
.
Wilcoxon/Mann-Whitney signed ranks test. ,
,

- (. 2.3).
Chi-square test for the median. ANOVA
,
, .
(the median test). 2
2 (G 1) . EViews
(Yate), ,
.
Kruskal-Wallis one-way ANOVA by ranks.
- (Mann-Whitney test)
. ,
( 1,
). 1 2.
, .
Test for Equality of Medians of IND
Date: 08/12/03 Time: 13:02
Sample: 1995:01 2003:05
Method
Wilcoxon/Mann-Whitney
Wilcoxon/Mann-Whitney (tie-adj.)
Med. Chi-square
Adj. Med. Chi-square
Kruskal-Wallis
df
Value
Probability
1
1
1
5.948351
5.948351
30.60075
28.42072
35.42363
0.0000
0.0000
0.0000
0.0000
0.0000
28
Kruskal-Wallis (tie-adj.)
van der Waerden
1
1
35.42363
31.91995
0.0000
0.0000
Median
58.39150
66.40950
62.33490
> Overall
Median
8
42
50
Mean Rank
31.25000
66.24561
51.00000
Category Statistics
DT
0
1
All
Count
44
57
101
Mean Score
-0.617909
0.476983
-9.51E-17
EViews U-
( tie) -
. ,
. EViews 2 ( tie ).
2 (G 1) .
Van der Waerden (normal scores) test.
- ,
. EViews ,
2 (G 1) .
,
: ,
, (Count);
(Median);
, (> Overall Median);
6 (Mean Rank);

(Mean Score).
.

,
.
F-test (G=2).
s L2 s S2
. F-
:
F=
s L2
.
s S2
F- F- (TL 1, TS 1)
, .
6
:
(
1, ), , ,
, .
29
Siegel-Tukey test. ,
. ,
.
, (. ), .
,
1. 2, ,
3. , ,
4, 5, . EViews
-
.
Bartlett test. weighted average
variance . :
. ,
, 2 (G 1) .
, ,
. EViews
.
Levene test (ANOVA)
. F-statistic F (G 1, T G ) .
,

.
(robustness) .
Brown-Forsythe (modified Levene) test
Test for Equality of Variances of IND

Date: 08/12/03 Time: 18:07
Sample: 1995:01 2003:05
Method
F-test
Siegel-Tukey
Bartlett
Levene
Brown-Forsythe
df
Value
Probability
(43, 56)
1
(1, 99)
(1, 99)
1.883776
0.715720
4.623970
4.142851
3.613701
0.0263
0.4742
0.0315
0.0445
0.0602
Std. Dev.
4.019442
5.516713
6.009558
Mean Abs.
Mean Diff.
3.291818
4.448680
3.944700
Category Statistics
DT
0
1
All
Count
44
57
101
Mean Abs. Mean TukeyMedian Diff. Siegel Rank

3.291564
53.38636
4.411904
49.15789
3.923835
51.00000
Bartlett weighted standard deviation: 4.922649
30
7 (Category Statistics),
.

1. wn, y1, y2 ,

.
2.5. (
)

Correlogram View View/Correlogram:
X
Correlogram Specification
Correlogram of
Level
1st difference
OK
2d difference
Lags to include
Cancel
36
, (Correlogram
of) (Level), (1st difference)
(2d difference), 8,
9 (Lags to include).
, .
Series: Y1 Workfile: CP-2
View Procs Object Print Name Freeze
Date: 07/31/03 Time: 15:38

Sample: 1 1000
Autocorrelation
.|****
.|**
.|*
.|*
.|
.|
.|
.|
.|
|
|
|
|
|
|
|
|
|
Sample Genr Sheet Stats Ident LineBar

Correlogram of Y1
5
Partial Correlation
.|****
.|
|
.|
|
.|
|
.|
|
.|
|
.|
|
.|
|
.|
|
1
2
3
4
5
6
7
8
9
AC
PAC
Q-Stat
Prob
0.548
0.313
0.162
0.094
0.045
0.027
-0.002
0.012
-0.003
0.548
0.017
-0.023
0.013
-0.012
0.006
-0.025
0.031
-0.022
301.56
399.80
426.25
435.08
437.14
437.85
437.86
438.00
438.01
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
.
, /4 ,
( ).
9
, ,
36 .
8
31
.|
.|
.|
.|
|
|
|
|
.|
.|
.|
.|
|
|
|
|
10
11
12
13
-0.012
-0.018
-0.022
0.007
-0.009
-0.006
-0.011
0.038
438.15
438.47
438.98
439.02
0.000
0.000
0.000
0.000
,
(Y1),
(CP-2). :
View ProcsObject
Print Name Freeze
SampleGenr Sheet Stats Ident Line Bar
Correlogram of Y1 ,
Y1.
(Date:
Time: ), (Sample: )
(Included observations: ).
Autocorrelation Partial Correlation

( ).
, k-
yt
T
rk =
(y
t = k +1
)( yt k )
(y
t =1
yt , yt .
k-

r1 ,
k 1
rk k 1, j rk j
kk = k =
j =1
,
k 1
1 k 1, j rk j
j =1
k =1
k >1
rk k-
k , j = k 1, j k k 1, k j ,
.

2
. k-
T
( )
, ,
() 5%-
.
32
AC PAC

, .
, , Q-Stat Prob Q-
- (Ljung-Box) P- . k
r j2
QLB (k ) = T (T + 2 )
j =1
Tj

k:
k
H0 :
HA :
r
j =1
k
2
j
r
j =1
2
j
=0
>0
-
yt , QLB (k ) 2 (k ) .
ARIMA( p, d , q ) , QLB (k )
2 (k p q ) .

1.
2.
DEMO.wf
(y3),
(y4=a0+wn+b1*wn(-1))
(1, 1) (y5).
wn, y1-y5, .
?
, , , ,
?
2.6. ARIMA( p, d , q )
EViews
ARIMA( p, d , q ) ,
.
AR ( p ) .

EViews.

:
yt = 0 + 1 yt 1 + K + p yt p + t ,
(*)
33
t ~ WN (0, 2 ) . EViews
( ). , ,
y110,
.
EViews
ls y5 c y5(-1).
EViews,
Equation: UNTITLED Workfile: CP-2
View Procs Objects Print Name Freeze
Estimate Forecast Stats Resids
Dependent Variable: Y5
Method: Least Squares
Date: 08/13/03 Time: 17:18
Sample(adjusted): 2 1000
Included observations: 999 after adjusting endpoints
Variable
Coefficient
Std. Error
t-Statistic
Prob.
C
Y5(-1)
5.055556
-0.503753
0.096585
0.027201
52.34281
-18.51944
0.0000
0.0000
R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
Durbin-Watson stat
0.255953
0.255207
0.987497
972.2257
-1403.950
1.994097
Mean dependent var

S.D. dependent var
Akaike info criterion
Schwarz criterion
F-statistic
Prob(F-statistic)
3.363014
1.144243
2.814714
2.824538
342.9695
0.000000
.
,
:
Estimate / ;
Forecast
;
Stats ;
Resids
(Residual), (Actual) ,
(Fitted).
EViews ar(1), ar(2),, ar(p),

( ):
yt = 0 + ut ,
(**)
ut = 1ut 1 + K + put p + t .
10
y5 , 5,
, (-0,5).
34
, i i (*) (**)
. ,
, :
ls y5 c ar(1),
:
Dependent Variable: Y5
Date: 08/13/03 Time: 19:24
Convergence achieved after 3 iterations
Variable
Coefficient
Std. Error
t-Statistic
Prob.
C
AR(1)
3.361958
-0.503753
0.020777
0.027201
161.8136
-18.51944
0.0000
0.0000
R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
Durbin-Watson stat
Inverted AR Roots
0.255953
0.255207
0.987497
972.2257
-1403.950
1.994097
Mean dependent var

S.D. dependent var
Schwarz criterion
F-statistic
Prob(F-statistic)
3.363014
1.144243
2.814714
2.824538
342.9695
0.000000
-.50

: , ,
, ,
.
, (**) :
, Convergence achieved after 3
iterations - 3 . ,
,
(Inverted AR Roots).
( )
EViews. ,
Object/New Object/Equation,
Equation Specification
Equation Specification
Dependent variable following by list of regressors including ARMA
and PDL terms, OR an explicit equation like Y=c(1)+c(2)*X
Y5 c y5(-1)
5
5
5
6
6
6
Estimation Settings
Method:
LS Least Squares (NLS and ARMA)
OK
6
Cancel
Sample:
1 1000
5
5
5
6
6
6
Options
35
(*). , Method
:
LS ;
TSLS ;
ARCH ;
GMM ;
BINARY (logit, probit, extreme value);
ORDERED ;
CENSORED , ;
COUNT , ,
.

:
.
EViews , Quick/Estimate
Equation EViews. ,
, ,
.
,
ARMA( p, q ) , .
1.
, ,
, .
2.
, , ,
yt = c + a1 yt 1 + a2 yt 2 + t ,

, .. ,
,
ls y c y(-1) y(-2)
(
ls y c ar(1) ar(2)).
, ,
ls y c y(-2),

yt = c + a 2 y t 2 + t ,
,
.

.
36
3. ,
,
,
EViews.
, :
R 2 (R-squared).
R2 = 1
( y ) ( y )
yt , -
.

yt ,
.
,
0 1.
(, )
.
2
Radj
(Adjusted R-squared).
2
= 1 1 R2
Radj
)TT 1k ,
k .
R 2
,
.

,
.

,
,
.
s.e.regr. (S.E. of regression).
s.e.regr. =
T k
RSS (Sum squared resid).
,

.
2
RSS = SSR = = yt X t b ,
t =1
X t , b

37
l (Log likelihood).

T
l = 1 + log(2 ) + log
.
T
2
,
.
- DW (Durbin-Watson stat).
( )
.
T
DW =
(
t =2
t 1 )
t
t
2
T
T =1
(Mean dependent var S.D. dependent var).

:
T
=
t =1
yt
T
y =
t =1
( yt )2
T 1
AIC (Akaike info criterion).

l
k
AIC = 2 + 2 ,
T
T
l , k
. ,
,

, .
BIC SC (Schwarz criterion).
l k log T
BIC = SC = 2 +
.
T
T

, ,
. ,
,
.
F- (F-statistic). F- ,
,
, .. ,
, ,
, .
38
R2
F=
(k 1)
(1 R )
2
(T k )
k , ..
( ). EViews P F- (Prob(F-statistic)). P- ,
, ,
, ,
. ,
, .
4.
Residual Tests View .
:
Correlogram Q-statistics
( . 2.5);
Correlogram Squared Residuals
;
Histogram Normality Test
( . 2.3);
Serial Correlation LM Test
-. ,
p:
t = 1 t 1 + 2 t 2 + K + p t p + ut ,
u t ~ WN (0, u2 ). ,
:
H0 : 1 = K = p
:
H A : 12 + K + p2 > 0 .
TR 2 , T
, R 2

et = 0 + 1 x1 + K + k xk + 1et 1 + 2 et 2 + K + p et p + u t ,
et yt = 0 + 1 x1 + K + k xk + t . TR 2
2 ( p ) .
,
ARCH;
White Heteroskedastisity (no cross terms)
(
);
ARCH LM Test
39
White
Heteroskedastisity
(cross
terms)

).
q MA(q )

ma(1), ma(2), EViews. ,

yt = c + t + 1 t 1 + 2 t 2 + 3 t 3 ,

ls c ma(1) ma(2) ma(3).
,
ma(q),
.
ma3
Dependent Variable: MA3
Date: 09/09/03 Time: 17:24
Backcast: 1 3
Variable
Coefficient
Std. Error
t-Statistic
Prob.
C
MA(1)
MA(2)
MA(3)
1.028819
-1.383183
0.640902
-0.186233
0.017601
0.031186
0.049815
0.031133
58.45392
-44.35218
12.86569
-5.981893
0.0000
0.0000
0.0000
0.0000
R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
Durbin-Watson stat
Inverted MA Roots
0.669620
0.668620
7.697083
58711.88
-3440.477
1.915246
.90
Mean dependent var

S.D. dependent var
Schwarz criterion
F-statistic
Prob(F-statistic)
.24+.39i
1.022033
13.37097
6.923572
6.943281
669.5251
0.000000
.24 -.39i

. , Inverted
MA Roots ,

.

(p, q) ARMA( p, q )
EViews
, . ,

. ,
ARMA (2, 1), , ,
40
ls c ar(1) ar(2) ma(1)11.

,
, -:
Dependent Variable: ARMA21
Date: 09/10/03 Time: 17:11
Backcast: 2
Variable
Coefficient
Std. Error
t-Statistic
Prob.
C
AR(1)
AR(2)
MA(1)
3.746622
1.495641
-0.894885
-0.524807
0.184022
0.016269
0.014565
0.032300
20.35962
91.93400
-61.44169
-16.24770
0.0000
0.0000
0.0000
0.0000
R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
Durbin-Watson stat
0.832956
0.832452
4.878129
23653.37
-2995.689
1.951439
Inverted AR Roots
Inverted MA Roots
.75+.58i
.52
Mean dependent var

S.D. dependent var
Schwarz criterion
F-statistic
Prob(F-statistic)
3.728876
11.91745
6.011401
6.031063
1652.174
0.000000
.75 -.58i
, ,
, EViews
, ,

(
):
(1 L K
1
)(
Lp 1 L12 ut = t
)(
ut = 1 + 1 L + K + q Lq 1 + L12 t .

yt = c + u t ,
ut .
,

, :
ar(12)12 ma(12)
11
ar(1) ar(2) .
4: ar(4)
ma(4).
12
41
. ,
( s_ar),
EViews
ls s_ar c ar(1) ar(12).
,
:
Dependent Variable: S_AR
Date: 07/15/03 Time: 14:34
Variable
Coefficient
Std. Error
t-Statistic
Prob.
C
AR(1)
SAR(12)
2.006013
0.574613
-0.620051
0.245655
0.026095
0.025378
8.165992
22.02035
-24.43280
0.0000
0.0000
0.0000
R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
Durbin-Watson stat
Inverted AR Roots
0.728619
0.728068
5.318454
27833.38
-3048.447
1.963491
.93 -.25i
.57
-.25 -.93i
Mean dependent var

S.D. dependent var
Schwarz criterion
F-statistic
Prob(F-statistic)
.93+.25i
.25 -.93i
-.68 -.68i
.68 -.68i
.25+.93i
-.68 -.68i
2.017917
10.19894
6.183277
6.198155
1320.952
0.000000
.68+.68i
-.25+.93i
-.93 -.25i
-.93+.25i
ls s_ar ar(1) sma(12).

,
Date: 07/15/03 Time: 14:36
Backcast: -10 1
Variable
Coefficient
Std. Error
t-Statistic
Prob.
C
AR(1)
MA(12)
1.994558
0.698247
-0.485910
0.302074
0.022740
0.027931
6.602889
30.70528
-17.39699
0.0000
0.0000
0.0000
R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
Durbin-Watson stat
Inverted AR Roots
Inverted MA Roots
0.702925
0.702328
5.532175
30482.54
-3124.888
1.983751
.70
.94
.47 -.82i
-.47+.82i
Mean dependent var

S.D. dependent var
Schwarz criterion
F-statistic
Prob(F-statistic)
.82+.47i
.00 -.94i
-.82+.47i
.82 -.47i
-.00+.94i
-.82 -.47i
2.012653
10.13974
6.262037
6.276772
1178.342
0.000000
.47+.82i
-.47 -.82i
-.94
42

, ..
yt = c + ut ,
(1 a1 L )u t = (1 + L12 ) t .

. ,
, ..
yt = c + ut ,
(1 a1 L )(1 L12 )u t = t .
, ..
yt = c + ut ,
1 a1 L L12 ut = t ,
ls s_ar c ar(1) ar(12).
, ,
:
Date: 07/15/03 Time: 14:37
Variable
Coefficient
Std. Error
t-Statistic
Prob.
C
AR(1)
AR(12)
2.007593
0.475595
-0.513877
0.149227
0.018907
0.019092
13.45325
25.15399
-26.91633
0.0000
0.0000
0.0000
R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
Durbin-Watson stat
Inverted AR Roots
0.772290
0.771827
4.870138
23362.47
-2964.534
1.962955
Mean dependent var

S.D. dependent var
Schwarz criterion
F-statistic
Prob(F-statistic)
.97 -.24i
.28+.90i
-.64 -.66i
.97+.24i
.28 -.90i
-.64+.66i
2.023927
10.19552
6.007153
6.022019
1670.336
0.000000
.71 -.66i
-.21 -.91i
-.88 -.24i
.71+.66i
-.21+.91i
-.88+.24i

.
, ar(1) ar(12)
s_ar(-1) s_ar(-12)
, .
M ARIMA( p, d , q )
ARIMA( p, d , q )
L :
(1 a L K a L ) y
p
= c + 1 + 1 L + K + q Lq t ,
43
= 1 L . EViews ,
:
d(y)
, .. yt yt 1 = (1 L ) yt = yt ;
d(y, n, s)
,
..
(n, s)
n
s
n
s
(1 L ) 1 L yt = 1 L yt ;
dlog(y, n, s)
(n, s) , ..
(1 L )n 1 Ls log yt = n 1 Ls log yt .
ARIMA( p, d , q )
d, ..
d yt = c + a1 L + K + a p Lp d yt + 1 + 1 L + K + q Lq t .
, ARIMA(1, 1, 0)
(1 a1 L )yt = c + t ,
EViews :
ls d(y) c d(y(-1))
ls d(y) c ar(1).
:
Dependent Variable: D(Y)
Date: 07/15/03 Time: 18:53
Variable
Coefficient
Std. Error
t-Statistic
Prob.
C
D(Y(-1))
2.106930
0.473341
0.190197
0.027914
11.07765
16.95697
0.0000
0.0000
R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
Durbin-Watson stat
0.224020
0.223241
4.874090
23661.73
-2995.865
1.956180
Mean dependent var

S.D. dependent var
Schwarz criterion
F-statistic
Prob(F-statistic)
3.992921
5.530322
6.007746
6.017577
287.5388
0.000000
:
Dependent Variable: D(Y)
Date: 07/15/03 Time: 18:57
Variable
Coefficient
Std. Error
t-Statistic
Prob.
C
AR(1)
4.000559
0.473341
0.292955
0.027914
13.65589
16.95697
0.0000
0.0000
R-squared
0.224020
Mean dependent var
3.992921
44
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
Durbin-Watson stat
Inverted AR Roots
0.223241
4.874090
23661.73
-2995.865
1.956180
S.D. dependent var

Schwarz criterion
F-statistic
Prob(F-statistic)
5.530322
6.007746
6.017577
287.5388
0.000000
.47
ARIMA( p, d , q ) ,

: ma(1), ma(2) ..

1.
wn, y1-y5
ARMA( p, q ) p, q 2 .
2.

.
.

.
3.
2.7. ( -)
EViews 4.1 :
- (Augmented Dickey-Fuller Test),
-13 (Phillips-Perron Test), --
(Dickey-Fuller GLS Test & Elliot-Rothenberg-Stock Point-Optimal Test),
--- (Kwiatkowski-Phillips-Shmidt-Shin Test
KPSS Tests) - (Ng-Perron Test).
-.

-
View/Unit Root Test
Unit Root Test
Test type
Augmented Dickey-Fuller
Test for unit root in
Level
1st difference
2d difference
Include in test equation
13
Lag length
Automatic selection
Schwartz Info criterion
Maximum lags:
User specified 4
- - EViews 3.1.
45
Intercept
Trend and intercept
None
OK
Cancel
:

(Level), (1st difference),
(2d difference);

(Intercept), (Trend and intercept),
(None);
(Lag lendgh).
,
, (Automatic selection)
( 14 (Schwartz Info criterion))15,
User specified ( ,
, 4 ).
, OK.
-
( ):
Series: Y1 Workfile: CP-2
Augmented Dickey-Fuller Unit Root Test on Y1
Null Hypothesis: Y1 has a unit root
Exogenous: None
Lag Length: 0 (Automatic based on SIC, MAXLAG=21)
Augmented Dickey-Fuller test statistic
Test critical values:
1% level
5% level
10% level
t-Statistic
Prob.*
-17.08816
-2.567279
-1.941140
-1.616486
0.0000
*MacKinnon (1996) one-sided p-values.

Augmented Dickey-Fuller Test Equation
Dependent Variable: D(Y1)
Date: 07/25/03 Time: 18:40
Sample (adjusted): 2 1000
Included observations: 999 after adjusting
endpoints
Variable
Coefficient
Std.
Error
t-Statistic
Prob.
14

(Akaike) - (Hannan-Quinn),
, -.
15
EViews 3.1
46
Y1(-1)
R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
-0.451460 0.02641
9
0.226358
0.226358
1.054303
1109.331
-1469.846
-17.08816
Mean dependent var

S.D. dependent var
Schwarz criterion
Durbin-Watson stat
0.0000
0.002017
1.198658
2.944637
2.949549
2.018780 6
,
(Y1), (CP-2).
:
Augmented Dickey-Fuller Unit Root Test on Y1 ,

- Y1.
,
(Null Hypothesis: Y1 has a unit root) ,
, (Exogenous: None),

(Lag Length: 0 (Automatic based on SIC, MAXLAG=21)), ..
:
Y 1t = Y 1t 1 + et ,
:
H0 : = 0 .
- (Augmented
Dickey-Fuller test statistic, 0 ) P-,
1, 5 10%- :
Test critical values:
1% level
5% level
10% level
-2.567279
-1.941140
-1.616486

EViews,
.

1. DEMO.wf :
y6=0,99*y6(-1)+wn
y7=y7(-1)+wn
y8=0,7+y8(-1)+wn
2. .
.
3. -,
y2, y6-y8.
47
4. -:

? ,
?
2.8. ARMA(p, q)

EViews ARMA(p, q)
16
(IND1).
1998 . 2002 .
(, .)
Forecast
2002 . 2003 . ( ,
1998 . 2003
.):
Forecast
Forecast of IND1
Method
Series name:
Forecast name:
Dynamic
Static
Ind1f
Structural
S.E. (optional):
(ignore ARMA)
G
G
A
R
C
H
GA
AR
RC
CH
H (((oooppptttiin
innaaall)l)):::
Output
Forecast sample
2002:10 2003:04
Insert actual for out-of-sample

OK
Do graph
Forecast
evaluation
Cancel
:
(Forecast name), (S.E. (optional)),
, (Forecast sample),
(Method: Dynamic, Static Structural) ,
(Output: Do graph Forecast evaluation).
:
. ,

h . ,
,
.
2 yt 1
16
1993 .
48
,
, yt 2 .
3 ,

,
, , :
80
Forecast: IND1F
Actual: IND1
Forecast sample: 2002:10 2003:04
78
76
74
Root Mean Squared Error

Mean Absolute Error
Mean Abs. Percent Error
Theil Inequality Coefficient
Bias Proportion
Variance Proportion
Covariance Proportion
72
70
68
66
2.137304
1.886911
2.654802
0.014775
0.092529
0.492510
0.414961
64
02:11
02:12
03:01
03:02
03:03
03:04
IND1F
.
, ..
, 1 T,
ARMA(p, q) ~
yT +1 .
7 ,
:
80
Forecast: IND1F
Actual: IND1
Forecast sample: 2002:10 2003:04
76
Root Mean Squared Error

Mean Absolute Error
Bias Proportion
Variance Proportion
72
68
2.462476
1.944386
2.717071
0.017053
0.025005
0.175971
0.799024
64
02:11
02:12
03:01
03:02
03:03
03:04
IND1F
49
,
, ,
:
Root Mean Squared Error :
T +h
RMSE =
( y t yt )2
h +1
t =T +1
h , y t
, yt ;
Mean Absolute Error
:
T +h
MAE =
t =T +1
y t yt
;
h +1
( %):
MAPE =
T +h
t =T +1
y t yt
yt
;
h +1
:
T +h
TIC =
( y t yt )2
h +1
t =T +1
2
T +h
y t
+
t =T +1 h + 1
T +h
yt
t =T +1 h + 1

:
Bias Proportion
y t
y
h
,
2
( y y )
tht
y ;

:
Variance Proportion
(s
sy )
( y t yt )2
h
s y s y
;
:
50
2(1 r )s y s y
( y t yt )2
h
r y t yt .

1. wn, y2, y4-y8 1

100 ARIMA( p, d , q ) .
2. 101, , 110

3.
?

?
3.
3.1. (
)
:
Ctrl
. ,
,
Open/as Group.

. ,
, Label+/- Wide+-,
2.1.

Object/New object/Group EViews
group < > < >,
.

1. DEMO.wf
rw=rw(-1)+nrnd
rw1=rw1(-1)+nrnd
rw2=rw+nrnd
rw3=1+rw+nrnd
rw4=-1+0,5*@trend+rw+nrnd.
2. RW, .
51
3.2.
View
Graph Multiple Graph.
Graph,
, :
730
720
710
700
690
680
670
660
960
970
980
Y8
990
1000
Y9
, Multiple Graph,
:
52
730
725
720
715
710
705
700
695
960
970
980
990
1000
990
1000
Y8
710
700
690
680
670
660
960
970
980
Y9
2.3.
53
1. DEMO.wf
RW.
2.
,
RW.
3.3.
Procs
Make Vector Autoregression
VAR Specification
Basics
? X
Cointegration
VEC Restrictions
VAR Type
Endogenous Variables
ind1 oilex
Unrestricted VAR
Vector Error Correction
Lag Intervals for Endogenous

12
Estimation Sample
Exogenous Variables
1995:01 2003:04
OK
, .. ,
(Endogenous Variables),
(Exogenous Variables),
, .. (Lag
Intervals for Endogenous), ,
(Estimation Sample).

: IND1
17
, OILEX

( 1995 . 2003 .):
Vector Autoregression Estimates
Date: 07/22/03 Time: 19:57
Sample: 1995:01 2003:04
Standard errors in ( ) & t-statistics in [ ]
IND1(-1)
17
IND1
OILEX
0.835523
(0.10137)
[ 8.24255]
0.614035
(0.09675)
[ 6.34644]
1993 .
54
IND1(-2)
-0.098109
(0.10096)
[-0.97172]
-0.612509
(0.09637)
[-6.35589]
OILEX(-1)
-0.362132
(0.08288)
[-4.36909]
0.050454
(0.07911)
[ 0.63776]
OILEX(-2)
0.545722
(0.08279)
[ 6.59160]
0.981794
(0.07902)
[ 12.4244]
1.127190
(3.00032)
[ 0.37569]
-2.389568
(2.86374)
[-0.83442]
0.826194
0.818875
614.2458
2.542783
112.8962
-232.6551
4.753102
4.883360
62.45052
5.974760
0.917102
0.913611
559.5961
2.427032
262.7452
-227.9961
4.659922
4.790181
84.16891
8.257463
R-squared
Adj. R-squared
Sum sq. resids
S.E. equation
F-statistic
Log likelihood
Akaike AIC
Schwarz SC
Mean dependent
S.D. dependent
Determinant Residual Covariance

Log Likelihood (d.f. adjusted)
Akaike Information Criteria
Schwarz Criteria
33.14213
-458.8280
9.376560
9.637077

() t [], ,
. 4
,
:
Determinant Residual Covariance
:
= det 1 t t ,
T p t
T , p ;
Log Likelihood (d.f. adjusted)
,
:
T
,
l = k (1 + log 2 ) + log
2
(Akaike Information Criteria Schwarz
Criteria):
2l 2n
AIC = +
T T
2l n log T
,
BIC = +
T
T
n=k(d+pk) ,
, k , d
( ).
55

, . -,
Residual Tests View
, ,
.
, Lag Structure View
:

(AR Roots Table AR Roots Graph):
Roots of Characteristic Polynomial
Endogenous variables: IND1 OILEX
Exogenous variables: C
Lag specification: 1 2
Date: 07/24/03 Time: 19:07
Root
Modulus
1.020817
0.811875
-0.473357 - 0.251051i
-0.473357 + 0.251051i
1.020817
0.811875
0.535811
0.535811
Warning: At least one root outside the unit circle.

VAR does not satisfy the stability condition.
,
( ) ,
, ..
;
,

(Pairwise Granger Causality Tests).

(Lag Exclusion Tests). (. )
2

,
( Joint):
VAR Lag Exclusion Wald Tests
Date: 07/24/03 Time: 18:21
Sample: 1995:01 2003:04
Chi-squared test statistics for lag exclusion:
Numbers in [ ] are p-values
IND1
OILEX
Joint
Lag 1
68.47932
[ 1.33E-15]
71.34502
[ 3.33E-16]
113.9743
[ 0.000000]
Lag 2
54.47254
[ 1.48E-12]
154.9752
[ 0.000000]
170.7268
[ 0.000000]
df
56
,
;
,
(Lag
Length Criteria).
, (
13):
VAR Lag Order Selection Criteria
Endogenous variables: IND1 OILEX
Exogenous variables: C
Date: 07/24/03 Time: 18:53
Sample: 1995:01 2003:04
Lag
LogL
LR
FPE
AIC
SC
HQ
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
-617.0574
-508.0099
-453.6986
-441.4990
-431.2965
-410.5607
-397.0080
-386.4080
-382.2395
-371.9166
-338.1122
-322.3627
-312.2120
-284.6735
-280.7090
NA
211.5521
103.1914
22.69131
18.56853
36.90985
23.58158
18.02011
6.919574
16.72324
53.41095
24.25423
15.22602
40.20621*
5.629569
816.8005
99.93104
36.53920
31.02164
27.41672
19.63368
16.23819
14.25299
14.23500
12.57770
6.953008
5.519332
4.904419
3.080743*
3.103997
12.38115
10.28020
9.273973
9.109980
8.985931
8.651213
8.460160
8.328159
8.324791
8.198331
7.602243
7.367253
7.244240
6.773470*
6.774180
12.43325
10.43651
9.534490
9.474704
9.454861
9.224351
9.137505
9.109710
9.210549
9.188296
8.696415
8.565631
8.546825
8.180261*
8.285179
12.40223
10.34346
9.379409
9.257591
9.175715
8.883172
8.734294
8.644467
8.683273
8.598988
8.045074
7.852258
7.771419
7.342824*
7.385708
* indicates lag order selected by the criterion

LR: sequential modified LR test statistic (each test at 5% level)
FPE: Final prediction error
AIC: Akaike information criterion
SC: Schwarz information criterion
HQ: Hannan-Quinn information criterion
6 ,
,
*.
3.4.
, .. ,
xt yt ,
EViews : View
Granger Causality ,
: IND1
18
, OILEX

( 1995 . 2003 .). , , , (. . 3.1).
18
1993 .
57
View/Granger Causality .
,
X
Lag Specifaication
Lags to include
OK
13
Cancel
(
13). :
Pairwise Granger Causality Tests
Date: 07/18/03 Time: 18:13
Sample: 1995:01 2003:04
Lags: 13
Null Hypothesis:
Obs
F-Statistic
Probability
OILEX does not Granger Cause IND1

IND1 does not Granger Cause OILEX
100
7.53840
1.32699
3.6E-09
0.21759
: 5%-
, OILEX
IND1,
, IND1 OILEX.
, EViews
, ..
:
yt = 0 + 1 yt 1 + K + p yt p + 1 xt 1 + K + p xt p + t
xt = 0 + 1 xt 1 + K + p xt p + 1 yt 1 + K + p yt p + u t
,
1 = K = p = 0
.
, ,
, .
3.5.

,
,

. ,
, , ,

.
58
,
View/Cointegration Test... ,
. ,
,

:
Johansen Cointegration Test
Cointegration Test Specification:

Deterministic trend assumption of test
Exog variables
Assume no deterministic trend in data:

1) No intercept or trend in CE or test VAR
2) Intercept (no trend) in CE no intercept in VAR
Allow for linear deterministic trend in data:
3) Intercept (no trend) in CE and test VAR
Do not include C or Trend

Critical values may not be
valid with exogenous
variables
4) Intercept and trend in CE no trend in VAR

Allow for quadratic deterministic trend in data:
5) Intercept and trend in CE linear trend in VAR
Lag intervals
14
Summary:
6) Summarize all 5 sets of assumptions
Lag spec for differenced

endogenous
OK
LR

, EViews :
1) No intercept or trend in CE or test VAR
;
2) Intercept (no trend) in CE no intercept in VAR
, ,
;
3) Intercept (no trend) in CE and test VAR
;
4) Intercept and trend in CE no trend in VAR
,
;
5) Intercept and trend in CE linear trend in VAR
.
,
( ) ( )
;
( )
( ) ( ) (
); .
59
, ,
,
6) Summarize all 5 sets of
assumptions. ,
:
Date: 07/19/03 Time: 17:46
Sample: 1995:01 2003:05
Series: IND1 OILEX
Lags interval: 1 to 13
Data Trend:
None
None
Linear
Linear
Quadratic
Rank or
No. of CEs
No Intercept
No Trend
Intercept
No Trend
Intercept
No Trend
Intercept
Trend
Intercept
Trend
Selected (5% level) Number of Cointegrating Relations by Model (columns)

Trace
Max-Eig
0
0
0
0
0
0
0
0
0
0
-286.5035
-281.8382
-280.7090
-286.5035
-280.3765
-276.4083
-281.4094
-276.4220
-276.4083
6.810070
6.787529
6.808166
6.748187
6.728440*
6.808166
8.216862
8.324580
8.475475
8.207083
8.291543
8.475475
Log Likelihood by Rank (rows) and Model (columns)

0
1
2
-288.6771
-285.4032
-283.7022
-288.6771
-282.5282
-280.7090
Akaike Information Criteria by Rank (rows) and Model (columns)

0
1
2
6.813542
6.828063
6.874044
6.813542
6.790565
6.854180
6.810070
6.796763
6.854180
Schwarz Criteria by Rank (rows) and Model (columns)

0
1
2
8.168231*
8.286958
8.437146
8.168231*
8.275512
8.469385
8.216862
8.307762
8.469385
, , ,
:
, . ,
: ( trace
Trace max Max-Eig)
5%- .
, , , ,
:
Date: 07/22/03 Time: 19:24
Sample(adjusted): 1995:01 2003:04
Trend assumption: Linear deterministic trend
Series: IND1 OILEX
Lags interval (in first differences): 1 to 12
Unrestricted Cointegration Rank Test
Hypothesized
No. of CE(s)
Eigenvalue
Trace
Statistic
5 Percent
Critical Value
1 Percent
Critical Value
None
At most 1
0.087027
0.013317
10.44561
1.340697
15.41
3.76
20.04
6.65
*(**) denotes rejection of the hypothesis at the 5%(1%) level

Trace test indicates no cointegration at both 5% and 1% levels
Hypothesized
No. of CE(s)
Eigenvalue
Max-Eigen
Statistic
5 Percent
Critical Value
1 Percent
Critical Value
None
At most 1
0.087027
0.013317
9.104913
1.340697
14.07
3.76
18.63
6.65
60
*(**) denotes rejection of the hypothesis at the 5%(1%) level

Max-eigenvalue test indicates no cointegration at both 5% and 1% levels
Unrestricted Cointegrating Coefficients (normalized by b'*S11*b=I):
IND1
0.277611
-0.293166
OILEX
0.015040
0.451230
Unrestricted Adjustment Coefficients (alpha):

D(IND1)
D(OILEX)
-0.221737
0.168112
1 Cointegrating Equation(s):
0.088506
0.096102
Log likelihood
-285.3438
Normalized cointegrating coefficients (std.err. in parentheses)

IND1
OILEX
1.000000
0.054175
(0.43561)
Adjustment coefficients (std.err. in parentheses)
D(IND1)
-0.061557
(0.03391)
D(OILEX)
0.046670
(0.03211)

.
:
Unrestricted Cointegration Rank Test
trace = T log(1 i ) ,
i = r +1
H 0 : rank r , H A : r = n ,
n 19,
max = T log(1 r +1 )

H 0 : rank r ,
H A : rank r + 1 .
(Trace Statistic
Max-Eigen Statistic), 5%- 1%- . ,
: Trace test indicates no cointegration

at both 5% and 1% levels Max-eigenvalue test indicates no cointegration at both 5% and 1%
levels.

. , ,

.
19
EViews 10 ,
.. n 10 .
61
p
:
Yt = A1Yt 1 + A2Yt 1 K + ApYt p + U t .
,
(Error Correction Model):
Yt = B1Yt 1 + B2 Yt 1 K + B p Yt p +1 + U t ,
p
i =1
i= j
B1 = I + Ai , B j = Ai ,
j = 2, p . .. Yt , K , Yt p+1
, Yt 1 .
, B1
. r.
B1 = T ,
n r T r n .
T Yt 1 r , T
,
.
, Unrestricted
Cointegrating Coefficients (normalized by b'*S11*b=I) ,
T . , ,
, r
1 r n 1 (
1 Cointegrating Equation(s): Normalized cointegrating
coefficients (std.err. in parentheses)).
,
, ..
.
,
,
, , ..
, p,
1 p-1, ..

1 p1.
, ,
1,
0 0.
3.3-3.5
1. DEMO.wf RW1, RW2, RW3, RW4,

: rw rw1, rw
rw2, rw rw3, rw rw4.
62
2.
, -
,
.
3.
, ,
(. 3.1)?
?
63

Турунцева М.Ю. Анализ временных рядов, 2003

Загружено:

Сведения о документе

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

Турунцева М.Ю. Анализ временных рядов, 2003

Загружено:

Авторское право:

Доступные форматы

-

File Edit Object View Procs Quick Options Window Help

Quick, Options, Window, Help),

Daily [5 day weeks]

Daily [7 day weeks]

Show Fetch Store Delete Genr Sample

Update default directory

Upper-left data cell

By Observation series in columns

By Series series in rows

Reset sample to:

1. EViews 3.1 EViews 4.1

Sample range pairs (or sample object to copy)

(VAR Vector Autoregression)

Name for object

View Procs Object

Print Name Freeze

Transform Edit+/- Smpl+/- Label+/- Wlde+- InsDel Title Sample Genr

Modified: 1 1000 // y1=-2

(Seasonal Stacked Line)

Jan Feb Mar Apr May Jun

Aug Sep Oct

(Seasonal Split Line)

(x, y). (OX)

(Histogram and Stats)

3 Show Row Margins

Group into bins if

Avg. count <

1. wn, y1, y2.

Mean test will use a known

Enter s.d. if known

Binomial sign test.

Test of Hypothesis: Variance = 81.000000

Test of Hypothesis: Median = 63.00000

Median Test Summary

Obs > 63.00000

Treat NAs as category

Group into bins of

Avg. count <

G=2, EViews t-,

Test for Equality of Variances of IND

Mean Abs. Mean TukeyMedian Diff. Siegel Rank

Bartlett weighted standard deviation: 4.922649

Date: 07/31/03 Time: 15:38

Sample Genr Sheet Stats Ident LineBar

Print Name Freeze

SampleGenr Sheet Stats Ident Line Bar

Estimate Forecast Stats Resids

Mean dependent var

Mean dependent var

LS Least Squares (NLS and ARMA)

(Mean dependent var S.D. dependent var).

AIC (Akaike info criterion).

Mean dependent var

ls c ar(1) ar(2) ma(1)11.

Mean dependent var

Mean dependent var

ls s_ar ar(1) sma(12).

Mean dependent var

Mean dependent var

Mean dependent var

Mean dependent var

S.D. dependent var

*MacKinnon (1996) one-sided p-values.