Вы находитесь на странице: 1из 672

. .

1
,



2011
330.43(075.8)
65673
84

:
.. , ,
, - , ,
. -

:
,
- ,
-
. .. , .
.
1994 .,
. .. ,

. .. .
60

..
84 . . 1 . 4 . 1, 2: / .. . .:
, 2011. 672 . (.
.)
ISBN 978-5-7749-0654-3

. ,
. .. , -
. ..
.
( ): . 1
; ,
; . 2
,
, ;
,
(, .).
.
, , ,
.
330.43(075.8)
65673
ISBN 978-5-7749-0654-3

, 2011

................................................................................................................. 6
.....................................................................................8

1
,

1. .
.................................................... 11
1.1. ;
, .............................. 11
1.2. .
............................. 26
1.3.
. ........................... 45
1.4. ....................51

2. .
.................................................................... 74
2.1.
.
....................................................................................... 74
2.2.
.
............................... 90
-2. ....................... 109
-26. ....................... 111

3. ,
........................... 113
3.1.
..............113
4

3.2. F -
.
....................................................................... 127
3.3. .
.
...........................................................................149

4.
................................ 170
4.1. ............................................................................170
4.2. ............................................ 184

5.
...............................................................................................203
5.1. ............................... 203
5.2. ................................................................ 215
5.3. .............................................. 224

6.

....................................................................................... 234
6.1.
............................................................234
6.2. ............................................ 243

,
................................................................ ............... 261
. .......................... 287
....................................................................................................................291
......................................................................................................................292

7. .
ARMA...................................................................................... 307
7.1. ARMA............................................................307
7.2. ARMA
........................................................................................... 340
- 7. ......................................... 369
5

8.
....................................................................................... 377
8.1. .......... 377
8.2. . .......................... 383

9. .
ARIMA.....................................................................................423
9.1. ARMA ........................................................423
9.2. TS- AS-.
.................................................................................448

10. TS- ZXS-............................ 454


10.1. .
............................................ 454
10.2. ................................................ 489

11.
. .
........................................................520
11.1. .
. .............................. 520
11.2.
................................................................................558
11.3.
.............................................................579

,
................................................................................605
. ............................ 637
................................................................................................................... 647
.......................................................................................................................651
............................................................................................. 665


,
. .. , -
. ..

.
, .
,
,
, ,
.

,
, ,
,
,
, , .
(
, .),
.

(2 2 ).
,
. ,
.

.

Econometric Views,
Stata.
.

,
7

. ,
, .


-,

.


( -
. .. ).
,
, , ,
.
, , ,

,

.
.

.
6

,
,
,

,

,
.

1.

.

:
.
,

,
.


.

, 2.

1 .. .
. .: , 2004.
2
: .., .., -
.. . . 7- ., . .: , 2005.
9


,

.


.

, -

.
: ,

(..

),
,
.

,
,
.
,

; ,
(-),
,
(DS-).
,
.
^-.
, -

,
. ,
,
,
.

6 12
3.

3 .. : . .:
, 2004.
10

,


.

-

(
). ,

.
1

_______ 1.1________
;
,

(Econometrics)
()

.

, ,
, ,
.

.

.
,
.
(
, , (Favero, 2001) (Cameron,
Trivedi, 2005)).

, .
,
.
,
, !

12 1. ,

,
. ,
(Samuelson, Koopmans, Stone, 1954): The application o f
mathematical statistics to economic data to lend empirical support to models con
structed by mathematical economics and to obtain numerical estimates1.

(data
mining),

, -
, .
,
,
. ,
,
,
.

, ,
. , ,
, ,

,
..

. ,

,
(, DGP
data generating process).
, ,
(disposable personal income) DPI
(personal consumption expenditures) CONS.
(Keynes, 1936)
( , )
,
, . ,

CONS = f (DPI),
, :

1
-
{.). . .
1. .. 13

f(D P l) ;
, ..
( ), 1.
,
,
.
DPI CONS
, (linear
relation) CONS DPI:
CONS = + ft -D P I,
, 0 < /3 < 1,
(propensity to
consume), ;
(autonomy consumption).

(, ,

0.9)
.

, DPIt CONS,
/-
. (,
,

).

(DPI,-, CONSt) DPIt CONSt (
scatter plot, scatter diagram, scatter
graph), , ,
CONS = + DPI, .
(scatter cloud),
CONS = + - DPI.
(. 1.1),

( 1999 ., ) 20
(. 1.1).

s i =CONSi - ( a + P D P I , )
CONSt
+ /? DPIf,
14 I. ,

/- , DPI
CONS\
,
, DPIt CONSt
(<observation model):
CONSi =(a + / 3 D P I i) + s i , = 1, ..., .
st - CONSt - ( + /? DPIt)
(observation error, disturbance), , j-m -


,

( ),
,
DPIt CONSt,
,
.

()
,
. 1.1 ,
:
CONSt = f(DPI,) + * i= ,
,
0:
E(ei\DPIl) = Q, / = 1, .
1.1
20

i DPI CONS i DPI CONS


1 2508 2406 11 2435 2311
2 2572 2464 12 2354 2278
3 2408 2336 13 2404 2240
4 2522 2281 14 2381 2183
5 2700 2641 15 2581 2408
6 2531 2385 16 2529 2379
7 2390 2297 17 2562 2378
8 2595 2416 18 2624 2554
9 2524 2460 19 2407 2232
10 2685 2549 20 2448 2356
1. . 15

:
E(CONSi\DPIl) = f ( D P I ,.), i = 1, ,
f ( D P f )
, DPI
:
CONSi = (a + j3-DPIi) + i9 i = 1, w,
DPIX, ..., DPIn () ;
cx, . . . , s n , E(sj\DPIi) = 0,
E(CONSi |DPI; ) = + f t - DPIt ;
/3 ,


.

f(DPI).
f( D P I ),
(statistical model)
CONSj = f (DPIj) + v,, / = 1, ..., w,
V], ..., v .
(econo
metric model), :
()

CONSi DPI,),
;
,
, .
(..
,
,
vl5 ..., vn)
(specification o f an econometric model).

= CONS, - f (D P Ii) = (a + p D P I,) + - f ( D P I t),

E(vi\DPIj) = (a + JB-DPI;)- f ( D P I i). Eiy^DPI;)


, E ^ O N S ^ D P I ^ ^ f ( D P I j ) , .. f(DPI)
16 1. ,


, DPIt.
,
(misspecifled model).
,
, , :
CONS, = {a + /3-DPI, )+,., / = 1, .

/? . ,
, ..
(DPIi9 CONSt), i = 1,..., n.
, .
b /?
- , (fitted model)
CONS = a + b DPI

, DPI. ,
:

\
, ( )
, ,
.

,
.
, ,
, ()
CONS =f(DPI) CONS = + SlnDPI, > , 8 > 0. ,
-
dCONS 8
, : ^ p j .= 5

. ( DPI > 8
,
.)


(, ).
,
, ,
( ).
1. .. 17

,
:


( );

( );
( ) ;

( ),
() ;
() ;

() (
);
;

(

); ;
( ..);
.

(, - parsimony princi
ple): , ,
.
(encompassing principle ):
,
(
,
).
(general-to-specific approach), ..
,
.

:
. 1.2 (BEL)
(.ZVET) 1968 . 1969 . ( ).

(. 1.2). :

,
.
18 1. ,

1.2
(BEL) (ZVET) , %

1 BEL ZVET i BEL ZVET


1968 . 1969 .
1 3.2 6.9 11 3.0 6.0
2 3.1 6.7 12 2.9 5.7
3 3.2 6.5 13 3.1 6.0
4 3.3 7.1 14 3.1 6.9
5 3.3 6.8 15 3.1 6.5
6 3.2 6.4 16 3.0 7.0
7 3.2 6.6 17 3.2 6.4
8 3.1 7.3
9 3.0 6.5
10 3.0 6.5

17
, 1
. ,
: -
, .
= 17 : (,).

00

00 00

00 ) ) )
s> <8
0 5
) 8
) -
)


> (
) )
0) >s X , .
2 2

.

.
.
.
Q
I-.

5
. S < X

20) X .
0Q 2
< X 0
0) :

. 1.2
1. . 19

,
,, \, (means)

* = ' - , = ' +" * - . = +


~ 17 /=1
(sample variances)

Var(x) = (,- - ) 2 , Far () = - (, ~ )2 ,


- 1 -1 ,=,
1 ? (1?
( ), (
variability) .
Var (variance). ,
,
. ,
1 (Std.Dev. standard deviation) ( ),

Std.Dev.(x) = yJVar(x) ( Std.Dev.(y) = ^V a r(y) ).
,

, - 1.
. ,
jcj, ...,
ju 2, , ,
1 1
= ] xi /, a Var(x) = ------V (, - )
n tt ~1
2.

Econometric Views ( Views),
.

: = 6.576, Std.Dev.(x) = 0.416, = 3.118, Std.Dev.(y) = 0.113.
,
2
.
4 : 1 , (
) .
: 7.3 - 5.7 = 1.6 3.3 - 2.9 = 0.4.

1 (The Oxford dictionary


of statistical terms, 2003).
20 1. ,


,
, ,

x i9y i9i = 1,..., ,

(

correlation diagram).

. 1.3.
. 1.3

,
1 9

= + /3, /3*0.
:
( + /
i = y i - ( a + p x i),

>>,. = ( + /? Xi) + et, i = l,...,n.
(linear
observation model),
= + fix
(linear relation)
( mathematical model, dependence
model), (dependent) ,
(independent) .
, ,

.
, -,
,
,
4 (
. 1.3).
Econometric Views.

1 ,
() ,
(. 1.3.4, 1.3).
1. . 21

-, - ,

,
.

4
(. 1.4).

()
(, ) (7, })
,
X/ - ,
.
, (,, j;,)
,
; - ,

.
++ = 4, +_ = 4, . 1.4
_+ = 3 (,
6 17, ), __ = 6
(, 9 10,
),
X/ - y t - +++ __ = 10, ,
, +_ + _+= 7.

( (,- - )(( - ) )
10/17 = 0.59, .. 59% ,
,
.

( (,- - x)(v, - ) ),

, .

.
(. 1.5)
7
15
(
. 1.4; ,
).
22 1. ,




,
xt 19 i = 1, ,

(sample correlation coefficient)
(,)
^ ^JVar(x) y jv a r(y )

. 1.5

(xt - x)(yi - ),
. Cov(x 9)

1
Cov(x,y) = ---- - ( * , - ~ )
-1
(sample covariance) .

Cov(x 9) = Var(x), Cov(y 9) = Var(y) .
, Cov(x, ) = Cov(y, ) = .

, 9 9z , /-
x i9y i9z i9 i = 1,..., n (n ).
, /-
{ = 9

Cov(x, ) = (*,. - )(,. - ) = - (, - )( - ) ,


-1 -1
Cov(x 9a) = 0. ,

Cov(x 9) = Cov(a 9) Cov(x 9) - Var(x).


,
1 1
Cov(ax, ) = ---- - ](, ~ <)(1 - ) = ---- - (,- - x)(yi - ) ,
- 1 ,=) - 1

Cov(ax,y) = aCov(x,y).
I. .. 23

Cov(x,y + z) = --2(;
1
~ 1 ,=1
+ z i ) ~( + ^))=
= - Z ( * < *) ((^ _ )+ (z' " * >)=
- 1 ,=1

=
-X(*,
- 1 ?
~)(, ~) + ] ( xi - x ) ( Z i - z ) ,
1 /=1

C ov(x,y + z) = C ov(x,y) + Cov(x,z).

, ,

Var(a) = 0, Var(ax) = a 2Var(x), Std .Dev.(ax) =\ a | Std.Dev.(x)

( > 0
),
Var(x + ) = Var(x)

( ).
,
Var( + ) = Cov(x + , + ) = Cov(x, ) + Cov(x, ) + C ov(y, ) + Cov(y, ) ,

..
Var(x + ) = Var(x) + (^) 4- 2C ov(x,j)
(
,
).
^,,
, , :,
,...,
x j = a + bxn / = 1,..., , (> 0)
= + ,

C o v(x,y) Cov(a + bx,y) _ bC ov(x,y) _


^JVar(x) yJVar(y) ]Var(a + bx) yjVar(y) yjb2Var(x) ^V a r(y) ^

,
.
1
,
:
24 1. ,

( * , _ >;/ - )
i=1 (=1 (=1


, ^ 1 (
+ 1, ,
, - 1,
, ).
+1 , (,, 1),..., (, )
, .
-1 , (,, _,), ..., (, )
, .
Var(x) = 0.1732, Var(y) = 0.0128, Cov(x, ) = 0.0217,
:

00217 =0.4608,
/. 1732 /0.0128
.. ,
0 1.

: ' = -0.717.
{positive corre
lation), (negative
correlation) .
,




.




UNJOB

.1.6

. , ,
UNJOB INF
1961 1969 . ( . 1.23
1.4).
-0.848.
(. 1.6)
1. .. 25

,
.

-
.
,
- 2, = -2 , -1 , 0, 1, 2.

.

J 1.1.1. Var Cov


- 1. , ,
(, 2004)
- 1, . , Var Cov
,
^ ,
Var Cov,
.

J 1.1.2. ,
,
.


1.
? ?

?
2. ?
() ?
?
3. ?
4. , ?
5. , ?
6. ,
?
7. ?
8.
() ?
9.
?
10. ()
?
11. Cov(x, )
?
26 1. ,

12. ^
!
13.
?

------------- 1 .2 _________
.

,
( )
, . ,
,
/?
,
, /?
(
).
,
()
= + /,
y t, i = 1, ..., ,

y i =(a + /3xi) + s i , / = 1, ...,.
,
s, = y i - ( a + j3Xi)
(error, disturbance) /- . ,
intercept,
slope.
,
, ,
, (, ) -
(, \ ..., (, )
() .

= + *
( + * = ) .
/-
1. .. 27

**= - ( * + /? * /)
\ , .
/ *, ..., *.
, * /3* , \ = ... = s*n - 0.
, ({, ),..., (, )
. * /,
,
.
, ,

= 4- /3 ,
/=1
1.
* /3* /3 .
- + (3* (, ),
- - /3* .
= - / 3* ,

= + /
, - - ^ . /

^ ( * ) 2 , , , /?
/=1
. , ,
, = ... = =,

=1 - ( * +*) = -

) 2 = 2 > , - * ) 2.
/=1 i=i
* *
^( *)2 - + /3 , -
=1
(} ).
= + f3
, = +
^ . -

1
. ,
, , .
28 1. ,

. 1.7

,- (fitted
value) i = a t + .
, .

ei = y t - y i = y i - ( a +
(residual) i- . ,
, ,
, .
,
, :

{ = ( + fi x t) + . ( );

y i - ( + xt) + et ( ).

/?
( ),
a + f i X i ^ a + ftXi. , et * ei9 .. /-
. 1.7
,
.
1? ..., ,
= + /3 ,
(least squares principle). -
1. .. 29

a*, J3*,
**, /*,

- * -** ,)2 = min Y j (yi - -*,)2.


/ =1 , =]

, **, /3**,
.
(-) ZS--
(least squares estimates). ,

'= , 0~=.
,
, (, );
.
(. -1.2 ),
, .
,
- xh y t,
i = 1,..., . ,

Q ( a , /?*) = (, - - p x i )2
/ = 1

z = Q(a , )
, , z ,
,
.
-1.2,
:

Y d(xi - x ) { y l - y )
=
2>,-)2
-----------------------------------,

/ = 1

-- .

,

30 I. ,

( * , . - X)2 * 0.
( =1

. ,
, ..., 1.
( 1,), i = 1,..., ,
- .

,

Var(x) = -----------
-1

Cov(x,y) = ------------------- .
-1

~ _ {,)
Var(x)

, , , /3
Cov(x, ), Var(x) > 0, /? , -
.
/3 ,
Var(x).


( ) ( )
1948 1961 . (. 1.3).
, /? /3 = -24.925.
, ,
/? = -0.24925.
, (
), /?
. /3

.

1 , .
1. .. 31

1.3
1948 1961 .

, , , ,
.
.

1948 67.8 0.5370 1955 66.6 0.4256


1949 67.7 0.4726 1956 67.4 0.4111
1950 69.2 0.4556 1957 61.5 0.4523
1951 71.9 0.4655 1958 60.2 0.4996
1952 72.4 0.4735 1959 67.6 0.4183
1953 63.5 0.5047 1960 65.2 0.4433
1954 60.0 0.5165 1961 62.2 0.4448

/3

Var ()
= * Var(x)

,
Cov(x,y) _ r ^ V a r j x ) yjVar(y)
=
Var(x) Var(x)
.

(standardized variables).
,

.
,
/ \ \
= 1, Var(yCT) = Var
Var(x) = Var = 1,
^JVar(x) JVar(y)

,1 = ( + ,:) + ,

:
Var(ye)
' -f V'LI I 7/ __ /\
Var(x )

/3 ,

.
32 1. ,


, -
.

,
y i =(a + j3xi) + ei , i = l,...,n.

/?

= 0.020415/0.162976 = 0.125,

a = y - j 3 x = 3.118-0.125-6.576 = 2.294.
,

= 2.294 + 0.125.

. ,
1 ( )
0.125 (^).
, : = 0.461, = 20.280.
,
0.461
.
- + = 0 (Cov(x,y) = 0)
0 (Cov(x 9y ) * 0 )

{ = y t - a +
. 1.3
= 2.294 + 0.125
= (7.1, 3.3),

1968 . (. 1.8).

.
= - (7.1, 3.118)
. 1.8 = + = (7.1, 3.183),

= , = 3.3 - 3.118 = 0.182,

= + .
1. .. 33

,
- + fi x
= - = 0.182 - (3.183 - 3.118) = 0.117.
, (,,^,) :
yl - y = (yi - y i) + ( y , - y )
y t = + /3xt , ,.

i :

Z (/ ~ )2= Z(A ~ + ZO7/ - ,-)2+2 - ,--Pi)-


/ =1 i=l / =1 /=1
(. -1.26),
,

( - ? = (-)2+ZCv<-Pi)2- (l.i)
/=1 i=i /=1
,
:
y j = (a + /3xi) + ei, i = l,

( proportional relation)
, = X, / = 1,
(1.1) .
1.3.
, (1.1),
(total sum o f squares) TSS.
,

/ =1
(1.1)
, {explained sum o f squares),
ESS,

ess = Z ch - y ) 2-
(=i
(1.1)

Z (>>,-)2 = 2 > 2
1=1 1=1
34 1. ,

(residual sum o f squares)


RSS'.
, (1.1)
, ,
:
TSS = ESS + RSS.
, /3 = 0 , = = . , /? =

Z 0 1 - 1)2 = ( , - ) \
; =1 1=1
.. RSS= TSS nESS = 0.
* 0 , = + ,

i ( y , - y ,)2 < Y d(yi - y ) \


i =1 /=1
.. RSS < TSS ESS * 0.
,
, RSS
TSS, , , ESS
TSS, ,
,
ESS/TSS.
(coefficient o f determination) /?2,

_ ESS _ / =1

2 > ,- 5 0
;=1
( 1.1)

iiy .-y if
R2 = l _RSS_ = l _^i_
= 1-
TSS
i =1

RSS TSS.
0

1 , , (, 2004).
: SSR, ESS (error sum o f squares),
(, , , 2005), SSE.
,
.
1. .. 35

RSS = TSS.
, /? = 0 ESS = 0 (
^).
1
RSS = 0.
: (^^), / = 1, , -
= + f i x . ESS = TSS ( -
).
,

0 < 2 <1.
,
. RSS
, (3 = 0 = ,
.

.

TSS = RSS ESS = 0.


TSS = ESS + R S S ESS 0,
:
TSS ESS RSS
Var(y)
n -1 -1 -1

ESS fT, / =i
= Var(y),
n- 1 n -1 /7 - 1

, i- (.

1 = 0> 4X0
1=1

ZCVi - = > Z ,= =
1=1 / =1 / =1

1 . -1.2 1.2.
36 1. ,

<->! 2 > / - >2


= ^ = = ------------------- = V a r(e),
- 1 VI 1 -1 -1
, i- et.

, = = 0.


Var(y) = Var(y) + Var(e),
, (
Var ())
( Var(y)).

( Var(e)).
Var(y)
(<analysis o f variance ANOVA).
,
.
,
R2,
R 2 = V ar{y)= l Varje)
Var(y) Var(y)

, , ,
,
.
(<explained variable),
(explanatory variable).
,
() ,
.


1.
.
,
y i =(a + f i x i) + ei, / = ,

1 . 1.3.4 ( 1.3).
1. .. 37


,
.

ESS = 0.043474
+ RSS = 0.161231
TSS = 0.204705,

Var(y) = 0.043474/16 = 0.002717
+ Varje) = 0.161231/16 = 0.010077
Variy) = 0.012794,

R 2 = 0.043474/0.204705 = 0.212374.
.
, R2,

.

^ ,
= + /3, ji /?
. ,
- + ( ei = . - . ), :

. Cov(y,y) _ Cov(y + e,y) Cov(y,y) + Cov(e,y)


^ ^() y/Variy) ^Var(y) ^Var(y) ^Var(y) <JVar(y)

TSS

( , - ) ( , - , ) = ,
1= 1


^ ( . - .) = 0
/=i

~ 1

, Cov(e,y). ,

() _ IVar(y)
39 <JVar(y) yjVar(y) \ Var(y)
38 1. ,


2 () >2
* Var(y)

,
^ .
,
R2, 1, 1
^ .
, 2.1,
(multiple-R, -/?).
, , ,

. , R 2
.
, ,

Cov(y, ) _ Cov(y, + /? )

ylVar(y) \ 2 Var(x) JVar(y) V Var(x)

sign(z) = -1 z < , sign(z) = 0 z - 0, sign(z) = 1 z > 0.



_ C ov(x,y)
Var(x)

, :
R 2
(
);
R 2
(
),

, ^
1. .. 39

(
)
.

1.2. < 0, sign(Cov(y, )) = -1 ^ > 0 ;


= 0, sign(Cov(y, )) = 0 ^ = 0; > 0, sign(Cov(y, )) = 1
> 0 , > 0 .

-1.2


Q(a,P) = Y J(yi - a - ,)2
/ =1
/3 () , ,
19 ..., .
z = Q(a, )
.. Q(a, )
J3:
dQ(a,p)

Q(a, /3) /3 \
(1 =0

(normal
equations)
8Q(a,P)
= 0

dQ(a,P) _ 0

, . ,

dQia.fi)= 2^ > , . - - / ? , ) ( - 1),


1 , -
.
40 1. ,

,)
= 2 Z ( .y ,- - / * ,)(-*/)>

Y i {yi - a - j 3 x l) = 0, Y j ( y , - a - p x i)xi = 0 ,
/=1 / =1

( , - , ) = 0 X , = 0.
i =1 / =1
, 0.

-
/ n \
( >
X * + X - 2 /=1
v -1 ) v 3=1 J

, , , , .
:

a = ~ Y j yi - - P Y Jxi = y - p x ,
nf?\ frt
(, ) = + .

X - X ; + X -2
v/sl A ,=1 wv =1 v =1 /=1

1 w II "
X w - r ^ X -
/? = '
2 2
X *,2
X -.2- - *1 /=1

( x, - x f = x ? - 2x x i + nx> = ^ ? - ,


X (, - )(, -)= ^-/->*,-**+5*=
i =1 / =1
-
1. .. 41


( ):

(, -)(, - )
= ----------------.
- _2
/ =1

- - fix
, /3
, 19..., .

-1.26
:

<>/ - -) = Y,(y,-, -. = ( -,)( +,)~^, =


/=1 / =1 / =1 / =1 1=1

= , + 0>/- > - 1
i=1 /=1 i=1

e i = ( y , - ( d + Px i)) = 0
1=1 1=1
(. ).

( , - ,), = ( , - ( + *,))*, =
1=1 /=1
(. ). ,

( , - ( 1 - ) = -


.
- :
42 1. ,


= + .
-1.2

(., - - / ? , ) = , ( . V i - a - p xi)xi =
1= 1



( ^ , 1) = 0, 1 = 0, = 0.

,
1 jc . ,
, , 1
L2( 1, ) w-
, 1.
1 ,
2(\,). 2( 1, ),
, 2( 1 , ), , jy.
, = + - ,
. 1.9 ABC .
L2( 1, ).
, ...,
. = 1,
1 . I 2( 1, ),
D. BCD.
DC 1,
L2( 1, ). L C D ( L2{ 1, )),
BCD . :
|BD \2 = \CD\2 + \\2, ..
\ - \2 = \ - \2 + \ - \ 2>


- ? = ( - ) 2 + ;<-- )2- (1.2)
/=1 i =i

1 2(1, ) ,
1 .
1 .
1. .. 43


TSS = ESS + RSS.
, :

\* -W 1
R 2 = ----- -------------------- ------ - = cos 2 Z B D C .
TSS . .
1= 1

, X CD , ( - ) _L(7 - ) ,

/ =1
, -1.26.
, = 0 (1.2) ,

Z->'i2 = Z A 2 + Z O ' / - A ) 2
1=1 1=1 1=1

ABC.
, , . 1.9
, 0 /3 0. , = 0, (3 0, -
44 1. ,

, , ;
. /? - 0, 0,
, 1. ,
, = y t = ,
= - 1. 0, D,
. (1.2) ,

- )1 = ~)2- , = 0, -
/=i 1=1
/0 = 0, D ,
2(1, ), = 0 . (1.2)

^ y f = ^ y f .
i =1 i=l

1.
?
2. ?
?
, ?
3. ,
?
4.
= + J3x
Cov(x, )?
5.
- + / !
6. ,
,
?
?
7. ? ?
?
?
8. , ?
9. ,
- = ( + /far,) + ,, i = 1, ..., ,

!
10.
?
11. ,
,
?
1. .. 45

12. ,
,
?

_______ 1 .3 --------------

.

1.3.1
. 1.4

15 ( , ).

1.4
15

I , , . / , , .

1 12 0.54 9 12 0.44
2 10 0.51 10 13 0.44
3 13 0.49 11 13.5 0.43
4 11.5 0.49 12 14 0.42
5 12 0.48 13 13.5 0.41
6 13 0.48 14 14.5 0.40
7 12 0.48 15 13 0.39
8 12 0.47


. 1.10.
,
y t = + ({ + et, i = 1, ..., ,
y t ---- /- , :,
i- ,


: = 21.100, = -18.559.
,
= 21.100 - . 1.10
- 18.559*.
TSS= 17.6, RSS = 8.562, ESS = 9.038,
46 1. ,

R 2 = 0.514, ..
51.4% .
,
.!

1.3.2
. 1.5
( ) ( )
1948 1961 .
1.5
1948 1961 .

, , , ,

. .

1948 67.8 0.5370 1955 66.6 0.4256


1949 67.7 0.4726 1956 67.4 0.4111
1950 69.2 0.4556 1957 61.5 0.4523
1951 71.9 0.4655 1958 60.2 0.4996
1952 72.4 0.4735 1959 67.6 0.4183
1953 63.5 0.5047 1960 65.2 0.4433
1954 60.0 0.5165 1961 62.2 0.4448


. 1.11.
v, = + /3+ ,, i = 1,..., ,
y t /-
, ,
i- ,


: = 11.552,
= -24.925. ,
. 1.11 =
= 77.552 - 24.925.
TSS = 208.194, RSS - 196.701, ESS = 11.493,
R 2 = 0.055.
5.5%
.
1. .. 47

1.3.3

1970 1979 .
(. 1.6). ( . ).

1.6

1970 1979 ., .

1970 695.2 621.7


1971 751.9 672.4
1972 810.3 737.1
1973 914.0 811.7
1974 998.1 887.9
1975 1096.2 976.6
1976 1194.3 1084.0
1977 1313.5 1204.0
1978 1474.3 1346.7
1979 1650.5 1506.4

, . 1.12.
,
, = + /?, + ,, i = 1, ..., ,
(
/-
, x t
,




: = -30.534, /3 = 0.932.
,
= -30.534 + 0.932.
TSS =791138.545, RSS = 740.320, ESS= 790398.225,
R 2 = 0.9995.
99.95%
.
48 1. ,

,
1 R 2 ,

.
.

1.3.4
1957 1966 .
:
( . /),
( )
( ).
. 1.7,
. 1.13.
1.7

, ,

, /

1957 34.9 716 478


1958 35.9 724 478
1959 37.9 797 478
1960 41.1 844 481
1961 43.5 881 483
1962 46.7 946 493
1963 48.9 1011 520
1964 52.0 1083 528
1965 56.1 1157 528
1966 62.6 1249 534

,
,
. , ,
= -2625.5 + 7.131, R 2 = 0.900;
= -129.30 + 0.350, R 2 - 0.871;
= 23.90+ 19.950, R 2 = 0.993;
= -0.860 + 0.0498 , R 2 = 0.993.
, ,
, ,
1. .. 49

ELECTRO USA CONS_ ILAND WORLD_RECORD

70 - - 540

1956 1967 1956 1967


. 1.13

,
: 19.950 0.0498 = 0.993 R 2
.

. ,

.
,

, (0.9)
.

, ,
(, spurious) ,
.
,
( ),
1.3.4.
, ,
R 2 =
2 . , 1
1
.

Covjy, )
^ JV ar(y)JV ar(x)
1 _ _
Cov{y,x) = ---- - (>>,.- y ) ( x t - ).
- -i
50 1. ,

Var(x) Var(y) 1,
Cov(y, ) > 0.
(, - ) (, - )
, ,

. (
, ,
. ,
, , ,
, , .)
^ -1,
Cov(y, ) < 0.
(, - ) (, - )
, ,
, . (
, ,
, , , .)
, 1

,
.

, (-
). .

1.3.5

1970 1983 . (. 1.8).
, 1972 .;

. ( 1972 . 100%.)
()
. 1.14. : = 1.743 + 0.0023.
R2 = 0.9398.
(. 1.15)
= 3.758 - ..
R2 = 0.0353.

1.3.3 ,


.
1. .. 51

1.8

1970 1983 ., .

1970 695.2 3.1 92.0 751.6 3.4


1971 751.9 3.3 96.5 779.2 3.4
1972 810.3 3.4 100.0 810.3 3.4
1973 914.0 3.6 105.7 864.7 3.4
1974 998.1 4.0 116.4 857.5 3.5
1975 1096.2 4.4 125.3 874.5 3.5
1976 1194.3 4.7 131.7 906.4 3.6
1977 1313.5 5.0 139.3 942.9 3.6
1978 1474.3 5.5 149.1 988.8 3.7
1979 1650.5 6.2 162.5 1015.7 3.8
1980 1828.7 6.3 179.0 1021.6 3.5
1981 2040.9 6.2 194.5 1049.3 3.2
1982 2180.1 6.6 206.0 1058.3 3.2
1983 2333.2 6.6 213.0 1095.4 3.1

. 1.14 . 1.15


(. 1.9).
. 1.16.
- -67.655 + 0.979*.

: 2 = 0.9918. , -
52 1. ,

1.9
1.6

1970 751.6 672.1 92.0


1971 779.2 696.8 96.5
1972 810.3 737.1 100.0
1973 864.7 767.9 105.7
1974 857.5 762.8 116.4
1975 874.5 779.4 125.3
1976 906.4 823.1 131.7
1977 942.9 864.3 139.3
1978 988.8 903.2 149.1
1979 1015.7 927.6 162.5

. 1.16

J 1.3.1.

1.3.1 1.3.2 (
) .
, ,
1 .
,
. , , ,


1. .. 53

(),
-
.
1.4.


, , ,

,
. 1.3.4.

()
().
: R 2 = 0.900.

,
.
:

= 613.333 + 59.539/, = 459.067 + 7.461/,


t t- .
0.9812,
0.8705. ,
/,
.
//,
,
.

eE(t) = Et - (613.333 + 59.5390,
,

eH(t) = H t - (459.067 + 7.461/),


. ,
eE(t) eH(t) , t = 1, ..., 10,

.
, , , ,
.
54 1. ,


=0.0000 + 1.420 .
R 2 = 0.2454 0.900
.

re =V 0.2454= 0.4954,

.. - /0.900 = 0.9487
.
^
(partial correlation coeffi
cient) t.
,
= 0.4954 = 10 ,
,
= + 5 0.



,
- .
, ,
,
.

(;, ,), / = 1,..., ,
.
y t = + fix t + st, i = 1,..., .
,
Xi = a + P y i + ei, / = 1, .. ., ?
, y t =
= + /?, + Sj, i = 1, ..., , ,/ /?
,, = + , + (, i - 1,..., .
( 2
0 _ Cov(x,y) C ov(y,x ) = ( , )
** ** Var(y) Var(x) {^() jVar(x) J
1. .. 55

( Cov(y, ) = Cov(x, )), .. * = , , R 2 = ^ ,

K K = Rl-

= +,

= +.

Cov(x, ) 0 (
,
), /3^ = 0 ^ = 0,
: = , =
.
Cov(x, ) 0, :

= ^ - + ^ X.


- a ^ - v p ^ x , ,
,
1 *
.. = 1,

, , R 2 = 1.
,

/ ^ *
= ~ -> = -

^ - - ^ ,

= ( - ) = ~ * -

R 2 = 1
= ~ .

- ^ - + ^ ,
56 1. ,

ZVET
7,5

6,5

5,5
1,8 3,0 3,2 3,4 BEL

. 1.17 . 1.18

R 2 = 1 . ,
, ,
( / = 1,..., .
, ,
, ,
(X/, ,), i = 1, ..., , ( = ... - = 0),
R 2 =1. R 2 1,
.
(, ), R 2 * 1 .
R 2 = 0 .
,
( , = + 1 +
/ = 1 , ..., ) , . 1.17.

x = 1.291 + 1.695j,

R 2 =0.212374.
0.125265 1.695402
0.212374
R 2.

, :

, , ,
.
. 1.18 ( ,
).
1. .. 57


, ..
=,
,
.
, ,
(capital asset pricing model).
, , :

~ = PjtTml ~ rfi) + j i i =

/- '- ;
rmi /- ;
rfi ;
Pj ( , ) (
) /- .
( ,), i = 1, ..., , ,

=
( ),

y^P xi+ Si, i .

/=1

( ( ,), i = 1,..., ),
.
Q(P) :

2 (., - Pxi)(-xi) = 0 ( ).
1= 1

=*1
i-1 / =1
58 1. ,

1>

}
/=1

, /? ^ (, )
Var(x)
, , = /.

^ e t = 0,
/=1
(.
1.3.4 .)
,

Z 0 - )2 * ZCPi - y ) 2 + 1t ( y i - y t)2,
(=1 1=1 1=1
, = /3xj , ..
TSS * RSS + ESS,
,
R 2 ,
.
ESS RSS
----- ----- 1.
TSS TSS
ESS
, R 2 = -----
TSS
1, -
RSS
R 2 = 1------------ .
TSS
1.3.6
,
. 1.10.
. 1.19.
,
^ = fix t + si9i= 1, 2, 3, 4.

/? = ----- = 0.7217.
2
/= 1
1. .. 59

1.10

1 1 2 3 4

xi 0 0.2 0.4 3

yt 0.5 0.8 1.2 2

RSS = 1.5377, TSS = 1.2675,


ESS = 4.0088. R 2
ESS
R 2= R 2 = 3.1627 > 1,
TSS
. 2 , RSS
R = 1 -------- -
TSS
R 2 = -0.213138.
, -

=1.9017. . 1.19
/ =1

,
-
(uncenteredR2):

2 > ,2
/=1

( ).
R 2

= Z ^ 2+ 1 > , -,)1. (1.3)


/ =1 / =1 i =l
,
,
.

^ ?9
/=1

Z .F ,2 Z U ' - )2
/ = 1 / =1
i=
X *2 L *2
/ =1 /=1
60 1. ,

Z ( 1 - 1)2 ?
Rt,= 1 - ------------ ,=
1
2
? Z*
/=1 /=1

0<
2 <1.
(1.3) . ,

! = ( - + ) 2 = Z ( 1 - Pi )2+Z P f +2Z (1 - Pi )Pt


;=1 i=1 ;=1 i=l /=1
Ho

Z - &)&=Z ( y t - P xi ) P xi =/?Z ( / =0
1=1 1=1 1=1
(. ), .
1.3.6
1 5377
RI = 1 1------ = 0.7571.
6.33

J 1.3.2.

Z ? = ' ? + 1(, 1- -. 1) 2
1=1 1=1 !=1
, -
0 1
. R2U
: R 2U
,
() .

/ 1.3.3. , 1.2,

= Pi + ( ~ Pi )> i = \ , .. .,n ,

= + .

2 2 > , - / - * , ) = 0,
1= 1
1. .. 61

, _L .
,
_L 1.
, pa
rt
2 > / = 0 , -
/=1
.

TSS = ESS + RSS ,
= J3x
(:), 9
2(\9),
1 9
. . 1.9 BCD
.

1. ?
?
2. ?
3.
, ?
4.
?
,
? ?

_______ 1.4--------------


. ,
DPI ,
(,
),

C = f ( D P I ),

, DPI.
. 1.20.
62 1. ,



.


(marginal propensity to consume MPC)
( ),

DPI0
. 1.20
:

lim / < .+ ) -
ADPI-+0 ADPI
,

MPC(DPI0) = dC =f\D P h ).
dDPI DPI=DPIq

f(D P I)
MPC{DPI) DPI. ,
DPI .
DPI
(power relationship):
= f (DPI) = a ( D P I /,
> 0, 0 < /?< 1.
MPC(DPI) = a p (D P I)p- \

DPI.
,
-
( !) (,
). , ,
:
log = log + \% DPI,
, logC = \ log = a , log DPI - DPI*,
* = + PD PI'.

:
* = * + PD PI*+ i = \ ,...,n ,
.
1. .. 63

: -
X Y
Y =f( X ),

MPY(X) = ^ = f \ X )

Y X.

(elasticity function) /{), X - 0
Y = f{X)
X , 0 :
f ( X 0 + A X ) - f ( X 0)
( _ JimQ
tj(X0) ,: * )__________
100

)(0) > 1 ](0) < -1 ( \rj(X^\ > 1), , Y


(elastic) X X = 0. | tj(Xq)\ < 1,
, Y (non-elastic) X
X = Xq. //("0) = 1 )(0) = -1
( (unit elasticity)).
, ,
:
Y dY
jj(X) = = MPY(X).
' Y dX Y
,
d\n f ( X )
d\nf{X) dX X dY
d\nX dInX Y dX
dX

/V d \n X Y Y X
MPY(X0)
Y =f( X ) X = 0, tj(X0)
In \ X = 0.
, MPY(X), .. MPY{X) =
:
Y =a +,
64 1. ,

ij(X) =

1 = + ft \ ,


Y = ( + In X ) = Const ,
( > 0 ) ( < 0 )
Y X.
: /() =
Y -
X 1%. , , Y - y f x , - 0.5,
X = 4 1% , .. X = 4.04. -
Y 2 Y =.04, .. 0.498%, 0.5%.
> 1 < -1 ( || > 1), Y
X. || < 1, Y
X. = 1 = -1
.
, Y = + fiX

> 0 0 1 X 0 .
= 0, rj(X) = 1. < 0 rj(X) + 1,
X +.

, , = f(D P I) - a (DPI)13
:
logC = logo; + log DP I.
dlogC
= -
d log DPI
(


( ) .
1. .. 65

1.4.1
(DPI)
() .
, 1972 . (. 1.9).
1970 1979 .

= -67.66 - 0.98DP/
( , , ).
0.98
, .

1970 1979 .

0,98' 751-6 .=1.10


-67.66 + 0.98-751.6

0.98-1015.7
= 1.07.
- 67.66 -h 0.98 1015.7

.
, ,
,
.


, .
, Y - , a Z -
, ,
:
Y =a + > 0, > 0,

, . 1 .2 1 .
X =

Y = - . Y Z 1
:

) .* * . '
dZY P +a Z
+
Z
( ).
66 1. ,

. 1.21 . 1.22

,
, ,
MPC(DPI) DPI, ,
.
, ,
( ) .
Y Z
Y = a + p \n Z , > 0, /?> 0
(. . 1.22, = 5, /?= 10).
,

* Z ) dZ
- ^ Y >0.
Z ) a + p \n Z
Y
Z.

(
In Y = a ---- , > 0 , .. Y = -
Z Z

( (
)). , Y
().

. 1.23 Y = -

= 3, = 10.
1. .. 67

() = 20.09,
Y = 20.09

= ( - | ) .





.
. 1.23

J 1.4.1. X Y
,
Y = + level-level .

Y X 1;
InY = + \ log-log .

Y X
Y
X 1 %;
1 = + log-level .
-
(semi-elasticity).
X 1 Y
1 0 0 /?%;
Y = + \ level-log .
X 1 %
.


\nYt = a + \ , + ^
, ,
Yi =ea X ? e Si, Yi = a X ? v i,
.. v,
).
, , ,
Yj = a X f + v;,
68 1. ,

.. . X - &
, /3 ,
. )!
, X f .
,
.
/?
,

Q(a,b) = f j (Yi - a X ? ) 2.
i= 1
,
.
(nonlinear least
squares NLLS).
,
,
b , Q(a, b).
,
Q(a, ). , ,
( ),
, ..
. :
)
;
) , ,
.

.

,



,
.
.
1 ,

1 1961 1963 .,
1963 1969 .
1. .. 69

. ,
(..
). ,
. , ,
1961 1969 . (. 1.11).
1.11
1961 1969 ., % *

1961 1962 1963 1964 1965 1966 1967 1968 1969

INF 1.0 1.1 1.2 1.3 1.7 2.9 2.9 4.2 5.4

UNJOB 6.5 5.4 5.5 5.0 4.4 3.7 3.7 3.5 3.4

* UNJOB ; INF .

. 1.24 UNJOB
INF , , ,

, ..
INFi = + /3 UNJOBs + s x?, i = 1,..., .

R 2 = 0.7184,
,


1.



1961 1969 .
( ).

. 1.24


,
.
,
.
. 1.24
,

1 ,
.
70 1. ,

I N F = a + f i ---- ----- + ,, / = 1,
UNJOBj '
INF UNJOBINV =
= ---- - .
UNJOB
1
INF = -3.90 + 27.47-
UNJOB
: R2 = 0.8307.
INF UNJOBINV (. 1.25)
.
INF
UNJOB 1961 1969 . (. 1.24), ,
INF UNJOB, -,
INF = 3. -
(Michaelis-Menton model):

- tK-UNJOB
2 + u n j o b

90 2
INF = <9, -
, + UNJOB
.

Y = -, X = ---- ----- . ,
INF UNJOB

2 0
1+
_ 2 + UNJOB _ 1^ u n j o b 1^
= = 2 ---------- = = + -1 = a + j3X,
INF -UNJOB 6 , UNJOB

2
= , =
. ,

Y = 5 , X ------^---- -
1NF UNJOB
. 1.26.

, :
Y = 1.947 - 5.952; 2 = 0.9914.
1. .. 71

. 1.25 . 1.26

= 1.947, /? = -5.952, = = 0.514, 2 = - -3.057,



:
0.514 UNJOB
INF -----------------------.
-3.057 + UNJOB
,
.

ft
INF = 9, +-
, + UNJOB
.
,

03 = ~\ ' @2
,
.
, 2 ,

/ - V
ft
e(6?02,03) = z INF: - ,
i=i 2 + UNJOBi j

(, ;).
, , 2,
(, = 0.5 2 = -3.0),
3 1 .
72 1. ,

INFmodel
INFtrue

UNJOB 7 12 UNJOB

. 1.28


:
=0.581, 62 =-3.117, =1.370;

1.370
INF = 0.581 + -
//05-3.117
. 1.27 INF
{INFtrue) {INFmodel), .
,

.

.

,
.

J 1.4.2.
,
.


.
, ,
,

.
, UNJOB
INF 1958 1984 .,
,
. 1.28.
I. .. 73

,
R 2 = 0.0864,
INF UNJOB.
-

.

J 1.4.3.

,
, .
, INF ,
, Y =
,
, ,
, ,
.

1. DPI
DPI ?
2. D PI
DPI ?
3. DPI
D PI ?
4.

? ,
?
5.
?
6. ?
?
7.
?
?
8. ,
,
? ?
2

_______ 2.1_______

.


,
()
= {+... +
..., ,
,
(,
). ,
, ..., , , ...,
.
(intercept),
, 1 ,
.
, ,
Q { K ,L ) = A K al f ,
Q ;
;
L ,
:
\ogQ = \ogA + a \o g K + /3 \o g L .
2. . 75

,
, /? .
:
= + 2 + ,

. = log Q, X, = 1, 2 = log/:, 3 = logL, = log^, = , = (3.



( ),
= + ... + 0,
{ = "I" i I,..., , >,

y t i- ;
6j '- ;
xtj j - - ;
() i- ;
.
,, , ..., xip, i = 1,..., ,
(statistical data), (observations).
, ..., .
: ,
,..., .
.
,
, ,

0. ,
, ..
(^ , . . . , x ip) = Q, = 1,..., .


y t = + fiXj + ,,

E (yt\xi) = a + fix i,

/? y i
jc, 1 .
, ,
(, , )
. Xj
1
76 1. ,

9j. , -
, ,

. 6j
Xj 1 ,
(
ceteris paribus).

:
log Qi, = log + a log K i, + ft log L,:+ s t , i = 1,..., n ,
, E(si\Ki,) = 0, i = 1,..., n . T
(log Qi\Ki,Li) = log A + a \o g K i + /?logL,,
:
log Qt
log^- 1 log/,,

;
logg,
logL, 1 logKt

.

{least squares)
0, ...,

Q(ex, . . . , e p) = Y j iyi - - . . . - e pxipf .


1=1

\ = ,..., = ,

min Q(e{, . . . , e p).

RSS,

R S S ^ ( y i - 9 xxn - . . . - e p xip)2,
1=1
.
2. . 77

R2

TSS

r a s ^ o ^ ) 2-
/=1

9 i ~ @\x i\ * X ip / = 1, , ,

( fitted ,
) (residual)
;-

= ( - ) 2=2>,2.
/ =1 /=1

/ = 1

, (explained sum o f squares),


= 2,
TSS = RSS + E SS ,

= ESS
TSS '

.
R 1 = *} ,

..
- ^ .
(multiple-R),
9
, ,
, .
, . . . 909

Q(.ex, . . . , e p) = f j (yi - e lxn - . . . - e pxip) \


1=1
78 1. ,

(
, ..., ) , ..., 9.
:

Z 2 O'/ - 0 - . . . ~ , )(-*,.,) = ,
1=1

Z 2 O'/ ~ ^ --,) (-,2) =


1= 1

Z 2,- - 4^,1 - - 1)(~*,) =


/=1

4 + Z X/1X<2
v/ =i
f \
2
/2 $2 ~
/ =1 . / =1 / =1

/ \ ( ^
I v n + Y . xipxn 4+.. Z 4
\| =1 ) V ~ l ) V'=' / 1= 1

..., 0^.
,
. -

= ,
X
* *12 \

21
&
>

2
=
-

2 ..
np J


*11 *21 .. 1'

*12 22 .. 2

II

X
2 . . np J
2. . 79

9 -
-

"V
2 2
=

II
<*; ,
(
), :
, .. 0:
&etXTX * 0,
:
X .
( )
( )~1.
9 = ( )~\
:
9 = ( 1 .
:

(*) "V
92 2 2
9= , = , =
II

, ; <,

y t 9]1 + ... + 9 +Sj,
| i 1, . . . , ,

-
=X 9 +S .

=9,

- - = - 9 .
80 1. ,

,

.
,

( 11 > ( X12 ( \

2\ 22

... $
II

II
V Xl ) 2 ) )

.. L(X) = L(xl9x2 9 )
Rn9 1? 2, ...,
.
- :

= \ + - - - + ,

.. 192, 9
, () .
= - ,
= + 9
, :
L (X);

,
L(X) /?".
,
= = { ' , - - = - = (/ - ( )~ ),

= {)~ , :
= , = (1 - ) ,

= ( )~ R"
L(X); (In - ) R" LL{X)
().
:

RSS = \\2 = = ( - ) ( - ) = ( - ) ( - ) =
= - - + .
2. . 81

, = ( ) = ,

RSS = - 6 + - = - + ( - ).

- ( )
,
RSS = y Ty - e TX TX 6 ,
:
RSS = \y \ 2 - \ \ 2.
| \2 = \ |2 + 1 \2,
R.
(
) 1923
1939 ., . 2.1.
; DPI
; .
.
2.1
1923 1939 ., % 1925 .

T DPI P T DPI P

1923 99.2 96.7 101.0 1932 153.6 105.3 65.4


1924 99.0 98.1 100.1 1933 158.5 101.7 61.3
1925 100.0 100.0 100.0 1934 140.6 95.4 62.5
1926 111.6 104.9 90.6 1935 136.2 96.4 63.6
1927 122.2 104.9 86.5 1936 168.0 97.6 52.6
1928 117.6 109.5 89.7 1937 154.3 102.4 59.7
1929 121.1 110.8 90.6 1938 149.0 101.6 59.5
1930 136.0 112.3 82.8 1939 165.5 103.8 61.3
1931 154.2 109.3 70.1



DPI, .

,
, .


lg = a ,+ j3{ lg DPI, lgT = a 2 +/32 lg P.
82 1. ,

,
(
)
( )
.
Tj, DPI,, ( , DPI, -
, / = 1 , ..., 17 , , i = 3
1925 .

:
\gTt = + p^XgDPIt + , lg7) = 2 + 2 lg7> +e2i, i = 1......17.

:
: , = 1.442, , = 0.348,
lgjT = 1.442 + 0.348 lg DPI,
ESS = 0.000959, RSS = 0.099185, TSS= 0.100144, R2 = 0.0096;

: 2 - 3.564, 2 = -0.770,
lg r = 3.564 - 0.7701g/\
55 = 0.087729, RSS = 0.012415, TSS = 0.100144, R2 - 0.8760.
. ,
, :
,
?
?

DPI , :
lgT = + p \g D P I + y\gP.


,
.

\%Tt = a + /3lgDPI, +y\%Pi + s i, i = 1,..., .



:
= 1.374, = 1.143, = -0.829,
2. . 83


lg = 1.374 +1.143 lg DPI - 0.829 lg ,

ESS = 0.097577, RSS = 0.02567, R2 = 0.9744.
: TSS = 0.100144
,
.
,
DPI

, 0.8760
0.97441. ,
0.9744
0.8760 0.0096
DPI .
1.143

, -0.829

.
,
. ,
, .


(Frisch-Waugh-Lovell theorem). FWL-
. (
(Davidson, MacKinnon, 1993).)
= +
,

= [ 2\ =

( - 1);
2 ( 1 )-;
fix ( - 1) 1 -;
/?2 ();

1 , -
.
, .
. 3.
84 1. ,

- \\ + 2(32 + 9 (2.1)
2 .
, , ,
.
R
L ( X x), - {9 a M l
R ( ).
, ..
,..., _ 19
= \ + ,

= -
,..., _ j.

9..., _
Xp=PiXp + M lxp ,

= -
,..., _ .
,
,
:
= (52 + . (2.2)

.

2 (2 . 1) (2 .2 ) .

/ 2.1.1.
, ,
v .

,
lg Tt - + /?lgDPIt + /lg /5, + i = 1,..., 17
= 1.374, = 1.143, = -0.829.
, FfTL-,
In.
\% lg7^ =
= , + /?[ \gDPIj + s u, i - 1,..., 17,
2. . 85

, = 1.442, = 0.348. ,
, lg : lg - lg Tt, - (1.442 + 0.348 lg DPIj).
lg lg/5, = 3 +
+ /?3 lg DPIt + % , /' = 1, ..., 17. 3 = -0.0586,
= 0.960, /, lgP
: lgP, = lgP, - (-0.0586 + 0.960lg DPIt).
, lg Tt = yM t lgf, + eAi, i = 1, ..., 17.
= -0.829.
, lgP

,
lg lgP, lg DPI.
>> z:
= + fiz + y z 1.
:
y i = a + fizi +yz? +;, i = l,..., n.
2 =z, jc3 = z2, :
y i = a + fixi2+ y x n +,, / = 1,..., .
,
{
1 x i3.
, /3 = ?2.
/-
y i = + P x i2 + )2
, = a + fi( xi2+A) + y ( x i2+A)2 = ( a + fix i2+yx?2) + /3A + 2 y x i2& + yA2,
y t ,2

+ 2 + ?.
( 0),
/ + 2 + A2 * { fi + 2y ,2) ,
x i2 + 2 yxi2, .
,
,

3 = z 2 2 = z.
86 1. ,

2 = z.
:
. = +xi3 + ,
- 2 + / + 2.
. - + x i3, xi2 = 2 + /32xi3
. = . - ., xi2 = xi2 - x i2, .. ,
. FWL-

,
y t = y i - y t xi2 = xi2 - x i2.

X
^ = + ... + 9pxip + ei9 i = 1 , .
,
,
,..., _.
,
, .
, , ..., X
, X j X k = 0 , k j = 1,...,/?.
:
-1
' \ .. '

2 2
2 (
2 2 '
2
6 = ( ' = =
( 1 )
;
\ K x Pj ,
, .
( X jTX j ) ~ x
^

$ = 1+ i = U ,n.

J 2.1.2.
, xiX = 1, - ,

- 2, ..., , , ^ x ik = 0 = 0 , ..
/=1
29 (centered variables).
2. . 87

X T
j X k = 0 * j , , j = 2, ,

Yux vxik ) {xik - ) = 0. ,
/=1 i=i
Cov(xj, ) = 0. = 1, = 1 Cov(x{9 ) = 0
= 2,..., .
,
X
19 29 ()
. R 2,
,

. ,
,

.

,
^.

, = 3, ..

: y i = + 02xi2 + + s i9 i = 1, . . ( ),
2: , = 1+ 2 + 1,
3: y i = 6 , + + i = \,...,n.
:
RSS = 2 = = ( ~ ) ( - ) = ( - ) ( - ) =
= - - + .

RSS = y ry - e rX TX e .
RSS,
, = ( ):
R S S = y Ty - - 7 + = - = - () =
rt

= - ^ - - ^ -03 =
/=1 1=1 1=1

= ( - 2) - 4 Z x/2y t - 4 Z Xn. = TSS - I-


/=1 1=1 /=1 1=1
88 I. ,

, R2 = R\ + R\, R\
*, ..
RSS RSS2 RSS3
1- 1- + 1-
TSS TSS TSS
RSSk ,
*.

(RSS2 + RSS3)~ RSS = TSS.
2 -L , 3 _L , 3 _L2.
:

RSS = TSS - 4 Z , 2, - 4 Z i
1= 1 / =1

RSS2 = T S S - e 2 x t3y RSS3 = T S S - e x ay,


i =1 i=l
( *
),
.


.

:
1)
y i =0lxti +... + 0p xip+i, i = \ , ..., , >,

y t i- ;
xtj j -
/- ;
6j j - ;
() i- .
- :
= + .

2) sl9 , ,
N ( 0, 2)
2. . 89

<2 > 0.
:
\ ?5 ~ i>i>d. 7V(0, (j ),
i.i.d. , (
independent, identically distributed).
, s - (19 ..., ) -
1 (
, , (0 ,..., )7)
Cov(s) = 2 1
\
3 ) ,
, 1 ,
( ) :
= i 1

4) :
d e t;rr X * 0 ,
: X .
(2),
.

1.
?
2. ,

?
3. ?
?
4. ,
?
?
5.
?
?
6.
?
7.
? ?

1
. -2 .
90 1. ,

_______ 2.2_______


.

, ...,

7j, ..., Y,
:
E(Yi) = exxn +... + epxip, (,) = <2,

Y j - N ^ +... + !,2), i = \,...,n.

Yl9 ..., Yn 19 ...,


. Yl9 Yn
Y = (Yl9 ..., Yn)T ,
w- 1.

Y = + 9 E(Y) = + () = 9 Cov(y) = 21.

,

{9...9 9...9
(
).
, = ( X)~lX TY
3 = CY9 = ( )~ 9

Y , ,
.
:

() = ( ( 1X TY ) = ( X TX)~l X TE(Y ) = ( )~ 1 ( + ) =
= ( )~ = 9

1 . -2 .
2. . 91

...,
:

0j ^ j -
.
,
,

:

Cov(3) = Cov(CY) = CCov(Y)CT = (( ' X T)cov(Y) ({)~ J =


= (( )1{( ] = 2( ( =
= 2( .


() = , ( )~ 1 ,
( ). , , 6 j :

D 0 j ) = ct2 ( ,

( X)j ' j - ( )~1.


(regression models), :
Yi = f ( X n, . . . , X ) + e i = \, .. .,n ,

Xix,..., Xip, i - 1,..., n, Yx, ..., Y, ,


Y)
= , ..., X ip = xip :

E(Yi\ x n =xi], . . . , X ip= xip) = f ( X n, . . . , X ip), i = l , ..., .

:
(, , ) = , ...,
Xip = xip, ) f ( X n ,..., X ip).
,
= , ..., Xjp - xip, ()
s = {ex,...,sn)T,
, f ( X n, ..., X ip).
92 1. ,

, , et =Yt - f ( X n , X ip), i = 1 ,..., n,


E(Yt \ x n = xn ,..., X ip =xip) = f ( X n , . . . , X ip), i = l,..., ,

E ( s \ x n = xa, . . . , X ip= x ip) = 0.

. 2 5 , , ..., X ip, i = 1, ..., ,


, . 6 ,
,
.
-
(regression analysis).
, = / ( , ..., X ip) + st E(Y^Xa = ;1, ..., X ip = xip) =
= f ( X il9 ), i = 1, n ,
Yt 1 9 .
, = f ( X i l 9 X ip) +
(regression equation), X l9
(iregressors).
/ (regression
function ). (
nonparametric regression)
,
( parametric regression ).
(linear) (nonlinear)
.

,
.
.
,
Qt = A K ? L f + e i, i = \,...,n ,
: f( K , L) - f(K , L; , , ) = AKaLp\
, , /?.
df(K, L; , , 0 ) _ 0 df(K,L; , , ) _ y a - i j P

/( , L; , , ) _ 1/-\


. ,
.
2. . 93


:
In Qj = \ + \ K t + /? In Lt + sr
: f(K , L) = In + \ + (3\nL.
,
:
df(K,L;A,aj3) ^

.
In,
In :
d f{ K ,L -,A ,a ,P ) _
din
.
, (1) (2) 2.1,

:,, \ . (mul
tiple)
, .
(normal)
. (..
),
(
multiple linear regression). (regression)
.
, - .

2.2.1
.

, ,
( 1 0 0 -
). , 38 ,


= 17.5 + 0.789*
R2 = 0.823.
= 17.5 + 0.789jc = = 82.94
, = . ,
82 ,
, ,
94 1. ,

.
, .
83 ,
, ,
.
, . ,
,
.!

(simple linear regression)

(two-variable
linear regression), (straight-line
regression).
( )~1 :

2 > ,2 -1 * 1
1=1 1=1
(. )~ 1

1= 1
,

I * ,2- = ^ ( . , - ) 2,
i=1 V =1 J =1
:
n

D(a) = [cr2( X T^ ) - 1] 11= _ - i 1^= 1 ------


Z ( * ,-* )2

D(p) = [a2( X TX y x\ 12 -----


Z ( x ,- x )2
/= 1
2. . 95

Cov(a,^) = [T2(X r ^ ) - 1]12 = -----^ ------,


( * ,.- * ) 2
/ =1

, Cov(a,/3) = 0, ^ / = 0 ,
/=i

* 0 J3 .
/=1

.

(Gauss-Markov theorem).
:
y i =6xxi1 + ... + 0 ^ + * , , / = 1,..., ,
;
9..., ;
9..., sn ,
, 2 ;
- : Y = X 0 + s, E(s) = 0, Cov(s) = <21.
, X ,
,
X 0.
= ( )~1

(BLUE best linear unbiased estimate) ,
9 = (
)9 Cov(6>) - Cov(0)
( ) .
, - Cov(6) - Cov(6)
D(9j) - D(0j)9 :

, 6j Oj

, .. ,
.
96 1. ,

. -
,\ .. ( ) = ,
= ( ) = () = () = .
6= , = 1.
= ( )~1 ( ), :
Cov(6)-Cov(6) = C o v ( C y - ( X TX y lX Ty ) = Cov(By) =
= BCov(y)BT =2 = a 2A,
= .
z 1 :
z TA z = z TBBTz = (BTz )TB Tz = wTw > 0,
w = B Tz p x 1 , Cov(6) - Cov(d)
,
.
,
, ( Gauss-Markov conditions).
,

.

, = ( )~
( ) ,
. (
(, 1968).)
,
- ^
}. , ,
, ,
, ..

.
()
, Y = {) + 6*, ()
.
. Y = {) +
. 1[{) {)^ 0 ( tr ),
[) .
2. . 97

. () (>
, .

jp- {)| .

, , .. tr(.Ar() ()) ,
0 , 0 >.
<9j") } :

D { e (p ) = <r2{ x (n)T () -> 0 ,


^ ' jj

j = 1, ..., .
E (e f{ )) = 6j, ,
, ( ( - O j ) >0, 6 {"} j = 1, ..., .
, {) ->0 , .. 0 {) .

, ()
, () 1>
(. (Amemiya, 1985)).

:


, = +... + epxip+i, i =\ , ,

j~ i.i.d. N(0, 2), , Qj


Gj j -
,
E 0 j ) = 6j , D(9j ) = 2 ( ) , = 1,.. .

j - 0 j t

<9,-
6j (..
6j).
98 1. ,


,
.
, :

( ~ ,
, , , .)
,
Qj , -
, ..
.

() <p(z):

q, 0 < q < 1, zq ,
() = q, Z
,
[ Z < z q\=q.
, q
.
. 2.1
zq q = 0.95;
z095 = 1.645. ,
z = 1.645, 0.95, : \ - q = \ - 0.95 =
= 0.05. , Z,
, ,
1.645.
2. . 99

. 2.1 . 2.2

- 0.5 1 (0.5 < < 1) -


/
, , . 2 .2 .
zi- ,
V 2 2 j

, . 2.2.
1 - I=
1 1 ---- , ,
. 2 2
. , , ,
, Z ,

- z ,
1-
2

Z Z
~2 2

,
/ \
, 1
2 2J
(), .. :
N
= 1- .

, Z,
,
1:

1 , <
< , > > ( ),
.
100 1. ,

p y z l- - Z - V " j 1--

^ ~)

Oj- ,
-Z < J 3 <z = 1- .
1-
v! "
*
2

, 1 - ,
6 ,-6 .
-^ ^ - -^ *
-2 p (6 j) 1~1

..
6j - Z [ < 6j < 6j + z J d (3~) .
2 2

, , 1 - ,

6 j-z &j + z )
2 2

-.
^ (
) 1 - , (1 - )- ,
100(1 - )% - ,
2 s l9..., .
= 0.05.
(X
() = 0.025,

0.05,
/
1 - 0.05 = 0.95.

, z0975 = 1.960,

\ . 9 6 ^ \ ^ \ . 9 /
2. . 101

95%-
2-1.96 J E 0 j )
( ) 6j.
, ,
-
. ,

D(dj ) = < j \ X TX)]]}, 7= 1,..., ,
2.

:

,
6j
= 6\ +. -+ Xip + Sj, / - 1,..., ,
j~ i.i.d. N(, 2)

D 0 j ) = (j2{ X TX y ^ j = \,...,n,

2.

1 - ,
.
(statistics).

2

-

. RSS = 2 > , - )2 -
1= 1

..., ,
, ,
S 2.
,
,
(jn - p ) S 2 _ RSS
102 1. ,

( ) ,
( - ) ,
( ,
, (,
, , 2005)).


--
( ) .
p(z)
- = 15
, . 2.3.

-
z 2(K)-
. 2.3
, -
(n - p ) S 2 RSS
----- 1 --

( n - ) , ..
~ P ) S 2 2,
------ 5------- X ( - ) .

( - ),
E ( S 2) = cr2,
.. S 2 2.

> / 2.2.1. = 1
:
1= -1, 1 = 1,..., ,
_Vj, ..., (,<2).
2
5 2 _ RSS
-1
9 = ,

RSS = - )2 - TSS ,
; =i

Z ( ^ - y )2
5 2 = ^ ----- ------= Var(y).
- 1
2. . 103

, Var(y) ,
TSS ( - 1) ( ),
2
, 2.

, 2
},
D(ej) = (j2{ X TX )-'

i = S \ X TX ) j l .

J d 0 j)

%

,
, .

, /-
( - ) ( , ,
(, , , 2005)).

t(K). q tq(K).

0 .
, = 10 , . 2.4.
. 2.5
. ,
.
:
^.95 = 1*645, ^0.95 ( 1 )= 1*812;
^0.975 = 1.960, ^0.975( 1 )= 2.228;
^0.99 = 2.326, /099( 10) = 2.764;
^0.995 =2.576, ^0.995( 1 )= 3.169;
104 1. ,

. 2.4 . 2.5

.
, , ,
,
1.645 0.05. 0.05
, 10
, , 1.812.
, >0
t(K )
7V(0, 1).
,
0 ,- 0 ,

6
-t ( - ) < -L---- L z t ( - ) = 1 - ,
1- - 1- -
2 2

, 1 - ,

0
( ~ )* ---- - (~),
'~1 S0, 1~1
..

~ ( ~ ) % ~ 6j ~ ^ +\ ! (~ ) %
2 2

, , 1 - ,
2. . 105

9}, .. 100(1 - )%-


9j , <2
s t, ..., s .
,
, 2.

2.2.2.

.
1 - ,
, t ( - ) -
2

(X
1---- , .. t ( - ) . -
2 1_
f ( - ). -
2

,
( ).
( - ) > 0 ,
t> A n - p ) = z ,
1 2 2

zq q .
, 1 - 0.9, 0.95
0.99, t ( - ) , -
1 2

: z095 = 1.64, z0975 = 1.96, z0995 = 2.58. ,


0.9 0.95
1.2 ,
0.99
1.3 .


-
.

2.2.2

.

106 1. ,

=6+62ZVETi +,, / = 1,
2 RSS 0.161231 1
: S = ------= ------------= 0.010749. ^
-2 1 7 -2
/ /\

0 2 = 0.125265, D ( 0 2)
s j =(0.062286)2. 95%- 62

1--^^- = 0.975

15 ( - = 1 7 - 2 ) : ^0975(15) = 2.131.
95%- 62\

@2 ~ ^0.9750 5) Se2 ^ 2 <2 + tQ915(15) ,


..
-0.0075 < 2 <0.2580.

0 = 2.293843, =0.410396. 95%-


6:
^1 ^0.975 0 S^ ^1 - ^1 ^0.975 (1 ^ >

..
1.4193 <3.1684.

.
1. 62
,
.
2. 0.95,
, 0.95
2.
, ,
0.90.
^0.97(15) = 2.131 f095(15) = 1.753,
2
0.0164. ,
02
90, 95 100.
,

, 02 ,
2. . 107

{ 02, } =
= 1 { 6? =
= 1 - [ { } +
+ { 2 } -
- { ^ }] =
= 1 - [ + - { }} \ > \ - - -
- 1 - 2.

, 2

1 - * , * = - ,

1 - 2 = 1 - .
, ,
, 62
0.95, * = 0.025.

(15) = / 0975(15) = 2*131


2

* ^ ^ = ^ 0025 ^ ^ = ^-9875^ ^ = 2*49,


!- ^
2.49
= 1.17 . ,
,
69 2.

*/ 2.2.3. ( . 3)

.

.


. ,
,
,
.
108 1. ,

2.2.3
( 1972 .)
(DPI)
(Q 1970 1979 . (. 1.3).
: = -67.555 + 0.979DP/.

^ =0.031454, t0975 (10-2) = t0975 (8) = 2.306,
95%- DPI
0.907 < < 1.052
/ = 1.052 - 0.907 = 0.145.
Ct = + PDPIt + ei9 , = PDPIt + st
(
),
/? = 0.903, =0.033633, t0975 (-1) = t0975 (9) = 2.262,

95%- DPI
:
0.895 < <0.911,
: / = 0.911 - 0.895 = 0.016.

9 . ()
0, , ,
,
. ( ,
,
0, 0.05 .
3.1.5 3.1.)

1. ?
?
2. ?
3. ?
, , ?
4. () ?
?
5. ?
?
6. ?
7.
?
2. . 109

8.
?
?
9.
,

, ?

-2

(, 2,
Xn) 1919 ..., 9
,
F(vl, v 2,...,v) = P{Xl < 2 <y2, . . . , X n<v\ - co < v l, v 2, . .. ,v ll<co.
,
1929 ...,
(, )9
*>12
F(vx, v 2,...,v)= JJ... jp(xl, x 2,..., xn)dxl dx2 ---dx.
-00-00
oo

F( vx, v2,..., vn)=P{Xx< v/,}-P{X2 < v2) P{Xn < ^} = ),


1
n
(,...,) = \ \ { }),

(XX9 Xl9 ..., Xn) F,


().
F(vl9 v2, v ) (). :
j 9j = 1, ..., 9
Fj(Vj ) = P{Xj <Vj } ; , = F(vl9 o o , o o ) .
p x(xx) ;
j 9k9 1 < j < k < n 9
Fj k(vj9 vk) = < vj9Xk < v*}; , FX2(yX9 v2) =
F(vx, v2, o o , ..., oo). p X2(xx, x2)
2.

-

= = ( , , 2, ...,).

Jnj
110 1. ,

X = (19 2, ..., )
, ,
(, ):
()

()=


,
:
,
(), ;


( ). ,
,
X
;
Y
X , Y = , ,
( Y) = () ( ).
X D{Xx\ ...,
D(Xn)
Cov(Xj, ) = E{XJ - E ( X j ) ){ - ( )),

Cov(Xj, )

Cov(Xj9 \ 1 <j\ < 9

(Cov{Xx, X x) - ( , )
Cov(X) =
[Cov(Xn, ) - Cov(X,X)J
X ,
X -
9 , Cov(Xj9 Xj) = D(Xj)9
:
'
D(Xx) - Cov(Xx, Xnf
cov(x)= : :
yCov(X,Xx) - D(Xn) ,
2. . 111

, Cov(Xj9 ) = Cov(Xk9Xf),
. , X
(
0),
. , -
- Z, ,
Z TCov(X)Z> 0.
, -
:
, ,
;
,
,
( ).
, :
Y
^, Y = , ,
Cov(Y)= Cov(CX) = Cov(X)CT.
, X -
, Cov(X) .
D(Xj) = a 2,j = 1 , 9
Cov(X) = a 2l n,
1 .

-26

..

/
/i-

p(xt, x 2,..., xn) = k -12 exp ( - -1( - ) ?:~1( - ) ^ , - 0 0 <,,2,..., < 0 0 , (2.3)
V 2
= (,,2,...,);
/ = (jix, /j^,..., /) ;
.
112 1. ,

= ( ]92, ) (2.3),
= (2 ) d e tZ ;

E(X) = ju, Cov(X) = Z.


, X - Nn(ju, I )
ju S.
:
S ).

, X - Nn(n, X),
() :

. . j - I.

(

):
X ~ Nn(ju, I), Y - X - / ~ Nn(0 , 1) ( X );
X , -
Nn(ju, I), ,
I ,
D(Xj) = 2- = 2, = 1,..., ,
X ~ N n(M, ).
, X
,
X ~ N n{0,(J2In).
X ~ Nn(/j, ), ,
Y = ~ Np(Cp, ).
( ).
3

_______ 3.1_______


2.2.2 95%-
2 :
^0.975 @2 @2 ^ ^0.975 0 ^) S ^ ,
..
-0.0075 < 2 < 0.2580.
:
2
0.95.
2 = 1, .
2 1,
: 1 - 0.95 = 0.05. , 2 - 1
( , 2 = 1)
, 0.05.
, 92 = 1.
\ ,
95%- :
, 2 - \ , .

(statistical hypothesis).
114 1. ,

(null),
0.
0 :2 =2.

(), \ 95%-
2, .. [-0.0075, 0.2580].
, , , \
,

4 -^ 2
> *0.975 0 5 ),
2
0,-0,
..

.
2 \ 2
Sg2 -Jd (32) .
, 0 : 2 - 2,

2 - \
> *0.975 0 5 ). (3.1)

, 0
, .. 02 - 2. : 1 - 0.95 = 0.05.
, 2 = \ , (3.1)
, 0
1- (error o f the first kind).
1- 0.05, ..
5 100.
1 - ,
0 : \ = \

02 - 0 2
>*, *(15),
2

1- 100# 100. ,
1- :
{#0 | 0 } = .
3. , .. 115


0
(statistical test o f hypothesis) 0,
(significance level).

= 0.05,
(, = 0.01 = 0.10).

0. , = 0.05 = 0.01 = 0.01 ,
0,
, .
- 0.05 ,
0
.
:
1- ,
0 : 02 = \ ,
2 = \ \ . (power)
2 = 9 \.

,
, .
0
, ,
.

( , test sta
tistics), .. ,
( , )
( ).
0 : 2 = \
/- (/ statistic)

2 - 1

\
, \
, 2 Sg2 . ,
/-, /- (/ tests).

(critical region) R ,
116 1. ,

.
^-,
t (15).
1
, :
) 0;
) ;
) ;
) R.
, ) ) ,
R
.
,
R
0 ( ).


: 9 = 0

\ = \ + + + * / / = 1, . . . . ,

~ i.i.d. N( 0, 2).
, j -

, .
:
) # : 9 = 0;
) 0.05;
) :

A ifl.fL .
%
0 : 6>j = 0 , /-
( - ) :

0/ t ( n - p ) \
------
%
/- (/ statistic) /-
(t ratio);
3. , .. 117

) :

,
>t. - ) -
2

(..
) :
0. ^ (Coefficients);
s * /-
(Std. Error);
0,-
/- (/ statistic).
S'

- ( P-value), .. ,
, ( - )
, , ,

, P- (P-value Probability).

P - .

1. P- , -
Q
, /-

0 : 6j = 0, .. > * ( - )
1--2

0 .

2. P - , -
Q
, /-

,
HQ : 0j = , .. < / ( - ) . -
1_
0 .
118 1. ,

3. ( ) P-
, # 0 : 6} = 0
.
, 0 : 9 j - ( 1), ,
^
{statistically significant estimate).
, j -

.
, 0 : 9j = 0 ( 2), ,
9j 9j
(statistically non-significant estimate).

, 0f = 0.
, j -

, ,
.
0 : 9j = 0
(hypothesis o f the statistical signifi
cance o f the coefficient ; , hypothesis that the coefficient 9j is equal to zero).
, ( )

. ,
,

. :
,
, .

3.1.1

R2 = 0.212375 ,
Econometric Views (. 3.1).
3.1
BEL

/- -

1 2.294 0.410 5.589 0.0001
ZVET 0.125 0.062 2.011 0.0626
3. , .. 119

Econometric Views
(. 3.2).
3.2
Dependent Variable: BEL

Variable Coefficient Std. Error t-Statistic Prob.

2.293843 0.410396 5.589344 0.0001


ZVET 0.125265 0.062286 2.011120 0.0626

= 0.05
ZVET (--
). = 0.10, -
,
ZVET .

3.1.2
(
. 1.7) R 2 = 0.513548 . 3.3.
3.3
SPROS


/- P-

1 21.100 2.304 9.158 0.0000

CENA -18.559 5.010 -3.705 0.0026

CENA
= 0.01,
.

3.1.3

( . 1.8)
R2 = 0.054483 . 3.4.
3.4
CONS


/- / -

1 77.484 13.921 5.566 0.0001

CENA -24.775 29.794 -0.832 0.4219


120 1. ,

CENA

( = 0.01, = 0.05, =0.10).

J 3.1.1.


.

.

3.1.4



(, ) (, -).

: R2 = 0.900.
(. 3.5).
3.5


/- -

1 -2625.497 420.840 -6 .2 3 4 0.0000
7.131 0.841 8.483 0.0000


, ()
7/,
() .

3.1.5
2.2.3, ( 1972 .)
(DPI)
(CONS) 1970 1979 .
. 3.6.
3.6
CONS


/- -

1 -67.65523 27.77342 -2.435970 0.0408
DPI 0.979441 0.031454 31.13836 0.0000
3. , .. 121

2.2.3, ,
,
5%- .



F-
. 3.7
1949 1960 . ( , 1959 .):
Y ;
2 ;
.

3.7
1949 1960 .,



Y *2 *3 Y *2 *3

1949 15.9 149.3 108.1 1955 22.7 202.1 146.0

1950 16.4 161.2 114.8 1956 26.5 212.4 154.1

1951 19.0 171.5 123.2 1957 28.1 226.1 162.3

1952 19.1 175.5 126.9 1958 27.6 231.9 164.3

1953 18.8 180.8 132.1 1959 26.3 239.0 167.6

1954 20.4 190.7 137.7 1960 31.1 258.0 176.8

:
. "I"02Xj2 1^5

} i- (/-
(1948 + /)- , = 1 ( ,
1).
, ..., % ~ i-i.d. N ( 0, 2) 2 .
R2 = 0.9560 ,
. 3.8.
f -.
0 : 2 = 0, 0 : = 0
( 0.10) -
.
122 1. ,

3.8
Y


/- -

-8.5 7 0 2.869 -2.988 0.0153

*2 0.029 0.110 0.267 0.7953


*3 0.177 0.166 1.067 0.3136

( 2 )
.
.


0 : 2 3 ,
, .

(linear hypotheses)
M : y i =6xxn +... + ep xip+ei, / = 1,...,, >,

, ..., ~ i.i.d. N( , 2),


. :
\ + - + =
<
aqX6x+... + aqpep =cq,

akJ (= 1 , q, j = 1, ...,) .
,
, , (, ..., ), ..., {aqX, ..., aqp)T
. , q (linear
restrictions) , ..., ( 6= {, ..., )).
:
0 : A Q - ,
(q ), , = q,
\ / \
" >
= ' 5 =


3. , .. 123

9 6, ,
, 0 : 9= ,
9 = . , 9 -
,
9 -
0 : 9= . :
6 /5- Np{9, 2( '),
9 - ,
, , rank = q,
^- .
9 - :
( 9 - ) = (9 )- = 9 - ,

Cov(A9 - ) = Cov(A9) = ACov(9)Ar = 2 ( {.

0 :9= , (9 - ) = 0,

9 - ~ Nq(, <2 ( 1).

V = ()~1 . <2
^- , V
. V
, V~\
( )- , V~' = .
, 9 -
( - )* - ( 9 - ) .
^ *
( - ) - 0 ( - )

Cov((A 9 -)* ) = Cov(P(A 9 - ) ) = PCov(A 9 - ) = a 2PVPT.

V= = ( )~*,

Cov((A 9 - )") = 2( 1= 2 ( = <j2l q,

( 9 -)* ~ N q(0, cr2I q).

, ( - )* ,
,
2.
, , q
124 1. ,

, -
. -
q . ,
(9 - )*

. <'2|( - )*\2 = ~2\{ - ) |2 = -"2 \{ - ^ [ ( 0 - )]=

= ' 2 ( - ? ( -)-< 7~ 2 ( - c f V~x( - ) =


= ' 2 ( - ) ( ( 1 ( - ) .
,

( - ) (( 1 ( - ) 2
2
( - )
() .
, 0 : = ,
\-() 2()>
.
, , , 2
. ,
/'-: 2
S 2.
(/^- F ratio, F-
F statistics):

_ ( - ) (( ' ( - ) /
F
S2
:

~2( - )( ( )~1 \ \ - )/
~2 RSS/(n )
0 :6= ,
q , 2( \
( - ) , 2( ~ )-
, (
. (, , , 2005)).
/'- q ( - )
:
F ~ F (q ,n -p ).
3. , .. 125

F - (F test)
0 : =
:
0 : = ,
F
, FKp= F l _ a(q, - \ .. F > F X_ a(q, - ).
F - ( Wald statistic)

w ^( - )({ ' ( - ) '


S
( Wald test)
0 : = . , ->
S 2 2.
, W
,
( 6 - )( ( 1 )*( - )
2
, , 2()-
0 : = ,
W > \- (?)
\-() -> ()
W9
.
(asymptotic test). F-,
, , (non-asymptotic
test) , F-
> F-pac-
F(q, - )9 .
, # 0 : = ( q
),
( - ) (( ' ( - ) !
S2
F(q, - ) , , 1 -

F < F x_a(q, - ) ,
..
( 9 - )( ( 1 ) ( 9 - ) < q S 2Fx_a(q, - ) .
126 1. ,

,
' 9 = 6 , .. 0 : 9 - ,
q = = 1 ( ), 1 -

( 9 - 9 0) ( 9 - 9 0) < p S 2Fl_a(p, - ) ,
100(1 - )%- 0 (
= 1 - ).
= 2 .

1. ? .
2.
? ?
3. ( )
?
4. ( )
?

?
5. = 0.10 = 0.01
? = 0.10 - 0.01
?
6. /-
.
7.
?
8. P-?

?
9.

?
10.
? ?
11. F- (F-)
?
,
?
12.
?
F-?
?
13. ( )
?
3. , .. 127

_________ 3.2 _________


F-
.

^-,
q ,
:
r (RSSHo - RSS)/q
RSS/ ( - )
RSS

: ,- = 0ixn +... + 6pxip + si9 i =

( unrestricted model), a RSSH


, 0, ..
= .
F-


# 0, 6Hq
0 ,
0 (restricted least squared estimate).
,

+---+e Pxip) f
i=1

0*=(*9...9*) * = ,
. ,
:

= - ( ) - (( )~1 ( - ) =
= ( 1 - (. ' [{ (( 1 - )=
= ^(^)1 - ( 1(( 1 (1 +

+ ( 1[ ( ' .
128 1. ,

, -, ,
, ..
, 0 := ,

(0) = ( ) - ( ' (( ) - 1 ( (0)-)=

= 9 - ( )~[{ { 9 - ) = ,

.
-, :

= By + d,
;
d .

9 ,
, 9 - 9,
9 = By + d, , Cov(9) - Cov(9H(j),
.

9 = ( ' - ,
9 = By + d, 9
. , Cov(9) - Cov(9Ho)
, j-x 9 9H<j
:
D(9j)>D{9Hoj), j = l,...,p .
:

RSSHo = ~ 9 = ( - 9 )( - 9 ),

RSS = - = (- 9 )(- 9).

RSSHo - RSS = -
- 9 = ( 9 - 9 )

= 9 = , :
3. , .. 129

\ - 0\ - \ + |>)- ^ |
, - - 0 .

> / 3.2.1. ,
|>>|2 =|.|2 + |._ .|2 -
2 1 I2 I I2
| | = \0 + \~ 0 0, .

\|
, 1.
2
( ),
0 : 9+ 2 = 1
- + = 2, - 3 ,
"0 2"
f 2l
= 2 , = 1 3
,3 4,
,
RSSHo -RSS = \ x ( e - 0 Ho)\2 = ( - ) 0 - ) =

= [ ( jr r -)-1 (( )"' ( - )] X

X ^( ( ( ( - ) =

= ( - ) ( ( )-' ( 1 ( 1X
{( 1 ( - ) = ( - ) ( ( 1)~'(-),

(RSSHo-RSS)/q
RSS/(n-p)


y i 0 + ... + 0pxip + , i = 1,..., ,
= 1 F-,

0 : 2 =3 =... = =
130 I. ,

( hypothesis o f overall significance


o f a regression, hypothesis that all o f the slope coefficients, excluding the constant,
or intercept in a regression are zero).
,
F -
(RSSHo- R S S ) / ( p - 1)
R S S /(n -p )
RSS ,
( ,
);
RSSH ,
0 .
()
y i =9l + i = l,...,n ,

6\ = ,

XSSh0 =1 (y,-y,Y = t ( y , ~ y i f = TSS-


/= 1 i=l
,
_ (TSS - RSS)/(P -1) _ ESS - 1)
R S S /(n -p ) RSS/(n - p ) '
(, Excel)

( mean squares).
sx, ..., ~ i.i.d. N(0, 2), F-,
, 0 (..
2 = ... = 9 = 0) F(p - 1, - ), .. F-
( - 1 ) ( - ) .
ESS/RSS,
, 2, ...,
Y.

0 : 92 93 ... = 9 =0
F-, ,
.
3. , .. 131

( 1 - ) F(p - 1, - ),
F x_a(p - \ , - ) .
, 0 ,

e ssKp - 1) ,

# 0,
( 1- ), .
, ,
F F -
- (P-value, observed level o f significance), ..
- = P{F(p -1 , n - p ) > F ] .

H0 -
F x_a(p - 1, - )

- < .
,
() - :
F= 97.75, () 095(2, 9) = 4.26.
P -
0.000000. ,
0 : 2 = = 0,

#02 *@2 ~ 0 03 * =
, , .

.
3.3.
--
.

3.2.1

:
R2 = 0.212, = 4.0446, - = 0.0626,
= 0.05 0 , = 0.10
.
132 1. ,

3.2.2

:
R2 = 0.513, = 13.7241, - =0.0026,
0 ,
.

3.2.3

:
R2 = 0.900, F = 71.96, - =0.0000,
.

3.2.4

:
R2 = 0.054, F - 0.6915, - =0.4219,
0 - 0.10.

, , . 4
()
( = 2) - F- -
^-, 2 = 0.
.

3.2.5
. 3.9
:
DPI ;
;
.
:
I : Ct = + 02D P It + O^Af + 64D P It_x + , / = 1,... ,119
t (1965 + /)- . 4
:
3. , .. 133

3.9
, ., 1982 .



82 DPIS2 2 82 DPIS2 AS2

1966 1300.5 1433.0 1641.6 1971 1540.3 1730.1 1902.8

1967 1339.4 1494.9 1675.2 1972 1622.3 1797.9 2011.4

1968 1405.9 1551.1 1772.6 1973 1687.9 1914.9 2190.6

1969 1458.3 1601.7 1854.7 1974 1672.4 1894.9 2301.8

1970 1491.8 1668.1 1862.2 1975 1710.8 1930.4 2279.6

1976 1804.0 2001.0 2308.4

=1, X 2 =DPI, 3 =, X 4 = D P I(-1),


D PI(-1) ,
DPI, DPI0 = 1367.4.
:
2 = 0.904, - = 0.0028;

= -0.029, - = 0.8387;
#4 = -0.024, - = 0.9337;
RSS
RSS = 2095.3, TSS = 268 835, R 2 =1------- = 0.9922.
TSS
F- :
F = 297.04, - = 0.0000.
.

, , ,
.
/-,
, ,
3 .

.
:
2 : , = + 02DPIt + 03At i = 1 , . . . , 11,
134 I. ,

D PI(-1). :

2 = 0.893, - = 0.0001;
9= -0.039, - = 0.6486;
RSS = 2098.31, R2 = 0.9922.
F-
F = 508.47, - = 0.0000.

,

3 : , = 9+ 92DPIt + Sj, i = 1,..., 11.

9 2 = 0.843, - = 0.0000;
RSS = 2143.57, R2 = 0.9920.
F-
F = 1119.7, - = 0.0000,
.
, , (
/-)
,

, F-.
,

^3 = ^4 = 0 >
Mj 3.
0

(RSSHo-R S S )/q
RSS/(n - )
RSS ,;
RSSHo 3;
<7 = 2 ;
- = 11 - 4 = 7.
3. , .. 135


/7 _ (2143.57-2095.3)/2_ Q0g
2095.3/7
F095(2, 7) = 4.74.
F < F095(2, 7), 0 : &$ = 4 = 0
, ,1

3.2.2.
:
F- ( /-),
0 : 4 = 0 ,,
0 : 9- 0 2;
F-,
0 : j = = 0 ,.

. 3
, , ,
F-,
1. ,
0 : 3= 4 = 0

.
, ,
//, : #4 = 0 ,, ,
, 2 : = 0 2.

. :
{ 0} =
= { , 2} ^
( ),
{ 0} =
= { ,, 2} <
< { } + { 2} =
= + - 2.

1
3.3.
136 1. ,

,
*,

, = . (
, -
*

* , - .) ,

* = 0.05,
= 0.025.
0 0.025 0.05.
0
,
: 0 = 0.025
, 0.05 ( ).

, q = 1,
, ,
0 : 2 + 9= 1.

(RSSHo - RSS)/q ( - ) [ ( X)"1] ( - )
R S S /(n -p ) qS2
# 0 : - , = ( , ..., 1), ..
<7=1. ( - ) [()~1 ] ,
F- :
_ ( -)2
S 2[A (. 1] '
, F -
:
( -)
S[A (.X TX y xATf 2
0 : - , # = 1 (
), /-
( - ) .
, q - 1
F- /-, /-
. ,
. 3.2.7.
3. , .. 137

P-
.
<7=1, /* /-,
/? = 0, --
*, ( - 2)
, F = /*2 F-,
,
(1, - 2) - \

.
( )
IgT = 1.3739 - 0.8289 lg +1.1432 lg DPI,
;
;
DPI .
, ,
0 : 2 = -1 0 : - 1
.

,
# 0: 6j = 0. 0: 2 = -1
/-
02 - ( - ! ) _ 4 +1^
se2 se2
0: = 1 /-

4 -i

'

(, ,
) t ( n - p ) = /(14).
0.05 ,
/-

1 4 , 3 ( ).
138 1. ,

__(14) = t0975 (14) = 2.145 . , , -


2

, 0.05,
1- .
:
2 +1 _ -0.8289 + 1
= 4.740 >2.145,
sA ~ 0.0361

4 z i_ 0918 <2145.
Si 0.1560

, 2 2 = -1
0 : 2 = -1 .
-. = 1
, 0 : = 1 .

/ 3.2.3. ,
6j 6?,
6j, D{6j)
s . ,
:
|-0.8289 + 1| = 0.1711 11.1432- 1| = 0.1432
, .. . s$2
4.3 , s ,
s2 s3 , ,
0 : 2 = -1 0 : = 1.

, 0 : 2 = - \
. ?
0 : 2 = -1
0 : 2 -1,
, (
, 2 + 1
) ^- . -
se2
$2+1
= 4.740, 2 - (-1) = 0.1711, ,
se2
, 2 > -1.
3. , .. 139


, 0 : 62 = - \
: 2 > -1 (
0 : 2 -\).
0 : 2 = -1
: 92 > -1
2 - (-1), .. /-.
, tx_a(\A) = ^0.95(1 = 1-761,
, 1- (
) 0.05. -
0 +1
>1.761; 1.761, 2.145,

Q _|_1
. = 4.740, -
se2
# 0 : 2 = -1 : 2> - \.
2.
0.05 0 : 9= 1
0 : > 1.
^-, ?0.95(14) = 1.761.
_1
= 0.918 < 1.761 ,

0 : j = 1 0 : > 1.
,

0 : 3 = 1, : 3> 1

# 0 = 1, ,
, , ,
# 0 : 0< #3 < 1, : 9 > \.
: 0
(composite hypothesis), .. ,
,

0 {simple hypothesis).

?
140 1. ,

0 : - 0.5.
: > 1

0 : = 1
: #3 > 1

, ,
0 : 0 < < 1,
, ( 1- )
0.05.
= #3,
0 : 0 < < 1
.
.
0 < < 1 ,
- 0.05
, 0 < < 1
,
= 0.05.
,
, = 1.
1
, > 1.761, -

= 1,

- 0.5,
3. , .. 141

, 3 = #3 (0 < 03 < 1) ,
0.05.
= 0.05 -
, 1-
. 1- ,
,
.
:

# 0 : ] < -1 ( ^ < 0),
0 : -1 < 6j < 0 ( 0j < 0),
# 0 : 0 < 6j < 1 ( ^ > 0),
0 : 0j > 1 ( 6? > 0)

, ,
6? = -1, 6? = -1, ^ = 1, 0j= 1 .

J 3.2.4.
, 1 1
.

,
0 : 2 < - \

: - \ < 2< 0.
0.05,
0 : 2 = - \ : 02 > -1.
0

3 l!>1.761,
se2
.
.

: 0 < 03 < 1
142 1. ,


: 3> 1.
0.05,
0 : 3 = 1 : 3> 1.
0

U l.7 6 1 ,

.
.
, , , ,
.

0 < 2< 0, : 2< - 1.


0 : 2 = - 1, : 2< - 1,
0

4 1 < ^ ( 1 4 ) = /0.05(14) = -1.761



( /(14)).

41>.

0 : = -1, , 0 : -1 < 2 < 0,
: 2 < -1.
,
,
.
, ,
0 : 3> 1, : 0 < 03 < 1.
,
0 : 3= 1, \ 3< 1,
0 : 3> 1

^ < ^ ( 1 4 ) = /0.05(14) = -1.761.


3. , .. 143

= 0.918 >-1.761,
sey
0 : > 1 .
,
. ,
. ,
(<conflict among testing procedures).
.


,
().



: ,
.


, .
,

, ,
, ,

.
.
0 : Qj < 0, : > 60.
/- :

, *

0 ,
:~ 0
------>tl_a( n - p ) .
%
tl_a(n - ) > 0 < 0.5, , } < 90,
# 0 .
144 1. ,

,
6j
0 > ,
/- # 0 .
,
0 : Oj > 0, : < 0.
#0 < 6?, /-,
, 0 .
/-

0 : 0j - 0 0 = 6j
6?, ,
(/- ).
,
, 0
.
,
Qj
0.

3.2.6
, Y
(1970 1979 ., . 1972 .).

= -67.655 + 0.9797.
, 2 - 0.979, ( )
0 : 2 < 1 : 2 > 1,
0
. /--
4 -1
, -
S02
. ,
. ,
, 2 1, 2 < 1.
3. , .. 145

0 : 2 = 0.9
0 : 2 > 0.9 , 1970
1981 ., = 12 .
2 - 0.952, s& = 0.0261, /-

0 /-
= 0.05, 0,

| * | > tCrit - *.975() - 2.228.


/-
= 0.05, # 0,
* > tcrit ~ *0.95 (1 0 ) 1.812.
0 0.05.
, 2 - 0.95.
/(10)
2 -0.95

, / / 0 ?
:

{ 111> 2.228 \2 = 0.95 }= < 0 9 > 2.228 \2 = 0.95


=
[ Sb
= / >{ |4 -0 .9 > 2.2285^ | 2 = 0.95} = { 2 -0 .9 <-2.228

2 - 0.9 > 2.228s-^ | 2 = 0.95} = { 2 - 0.95 + 0.05 < -2.228

0.05

0.05
2 =0.95

= />{/(10) <-4.14
146 1. ,

7(10) > 0.312} = { (10) < -4.14}+ { /(10) > 0.312} =

= 0.001006 + (1 - 0.619276) = 0.3 817.

P {t > 1.81212 =0.95}= />< 01 0 9 >1.812|02 =0.95


I %

J 02-0-95
1

= 1- { f(10) < -0.104} = 1- 0.4596 = 0.5404.

, 0 : 2 = 0.9
, 2 = 0.95, :

0.3817 ,
0.5404 .

2 = 0.95.
(0.5404
0.3817 )
, 0.05. ,
2 = 9 \ \
2, : 2 > 0.9 (.. \ > 0.9).


: 2 > 0.9.


.
,
.

Y = AK al P ,
;
.
3. , .. 147

+ > 1, ,
.

# 0 : + < 1, : + > 1,


# 0 : + > 1, : + < \.
:
y i =0lxn +... + 0pxip+i, / = 1 , Si-i.i.d. N (0 ,a 2),

{q = 1 ),
, :
0 : 92 + $3 = 1

0 : 6 = , = ( , 1).
,

_ ( - )

s [ A ( X TX ) - 1AT]1
0 : - , q = 1 (
), /- ( - )
.

0 \<, : 0 > ,

0 : > , : < ,

3.2.7

1960 1972 . (. 3.10). Y
( .); ; L
- ().

,
:
In Yi =\nA + a \n K i + p \n L t + /9 / = 1,..., 13.
(. 3.11).
148 1. ,

3.10
1960 1972 .

Y L

1960 20171.2 269.7 18271.8


1961 20932.9 267.0 19167.3
1962 20406.0 267.8 19647.6
1963 20831.6 275.0 20803.5
1964 24806.3 283.0 22076.6
1965 26465.8 300.7 23445.2
1966 27403.0 307.5 24939.0
1967 28628.7 303.7 26713.7
1968 29904.5 304.7 29957.8
1969 27508.2 298.6 31585.9
1970 29035.5 295.5 33474.5
1971 29281.5 299.0 34821.8
1972 31535.8 288.1 41794.3

3.11
Ln Y


/- -

-2.026556 1.195492 -1.695165 0.1209

Ln 0.372618 0.052261 7.129920 0.0000

Ln L 1.481073 0.259434 5.708868 0.0002

R-squared 0.959293

Econometric Views F-
0 : + /3 = 1, F = 13.90. ,
7-
/13.90 =3.728. + /3 - 1 > 0 , /'-
t = 3.728. 0.95 /-
(13 - 3) = 10 ?095() = 1-812
0.99 /95(10) = 2.764,
0 : + < 1 0 : + /3> 1
1%- .
( ,
).
3. , .. 149

1.

? ?
2. F-

, ,
?
3. ?
?
4.
?
5. F-
?
6. F- /-

?
7.

?
8.
?
9.
?
10.
?
11. ,
?

------------- 3.3--------------
.
.

:
R2 0 ,
, 1?

,
R2, R2 ,
.
.
F -

150 I. ,

_ ESS/(p -1) _ ESS/TSS ( - ) R2 ( - )


~ RSS/(n - ) ~ RSS/TSS ( p - \ ) ~ l - R 2 ( - 1) '
:
( - 1)F(1 - R 2) = ( n - p ) R 2, ( - 1)F = (( -1 )F + { - p ))R 2,
r2_ (p -\)F
1 _
( p - \ ) F +( n - p ) n
n -p
{p - \ ) F
F
R2, 0 : 2 = = ... = = 0,
F > Fcrit - F\-a(P w
2 > 2.

1+ *
( -^ ,
0 -
.
R2crit = 0.05
.
( = 2),

Rcrit - Fcrit ~ ^0.95 0 ~ 2).


1+
F crit


R2crit (. 3.12).
3.12

3 4 10 20 30 40 60 120 500

Rlu 0.910 0.720 0.383 0.200 0.130 0.097 0.065 0.032 0.008

,
R2 0
, , ..
.
R2 = 2
()
,
(. 3.13).
3. , .. 151

3.13

3 4 10 20 30 40 60 120 500

D 0.360 0.311 0.254 0.179 0.089


xy\cfit 0.953 0.848 0.618 0.447

. 2
.

.
, , ,
Y ,
,
< .
..., , , , { = I.
:
DGP\ = + ,

X , ~N(0, <21).
, () (
) ,
..., Z,, ..., Zq,
q =K - p :
SM : = Wp + v = + Zy + v,
W , Z q,
(
W = [ XZ] , =
JJ

, W .
= 0,
v = , , v ~ N(0, 21)-
, ft :

=(wTw y lWT =(wTw y xWT(W p +).


:

() = , ..

{ ) = , .
S 2K = RSSK/(n - ) (
152 1. ,

) -
2.
D { 9 j ) > C T \ X TX r x, j = \,...,.

(, ,
, 2005).
:
DGP :y = W/3 + e = X e + Zy + e, *0, ~ N (0 ,a 2I n),
SM : = + rj ( ).
, \
= (. 1 = ( (X0 + TJ) = 0 + ( 1
].
/ = Zy+ , Cov(t]) = <21, E(rj) = Zy,
E{G) = 9 + { X TX y lX TZy.
, ,
, , ,
:
X TZ - 0 ( X Z).
(.
(, , , 2005)): ,
= + Zy + ,
.
S1
, (. (, , , 2005)):
19 ..., 9
= + 77,
, ^ - + Zy + \
= + /7
S 2p = RSSp/(n - ) , ,
2 : E(S2
p) > 2.
,
,

. ,
,
,
.
,
3. , .. 153

, ,
.
/?2,
(, )
R2 ,
, ,
, .
,

19 ..., 6.
, , - ,

1=1

9 ..., _9
9 ..., _
= 0.
, ,
9 ..., 9 .
, RSS ,
. ,
, R2.
,
R2
R 2 {adjusted R-squared, adjusted R2):
2 R S S /jn -p )
adl TSSjin - 1)
, , {penalty)
.

TSS TSS \ n - p J J
2 RSSf n - 1
TSS n - p J n - p TSS
n > p p > 1
154 1. ,

, R2
. ,
RSS - 1
> 1.
TSS -
, ,
_1 2 +1
R < > ----- .
2 2
R2adj
,
.

/ 3.3.1.

, ,
, R2adj,
R2adl S 2 RSS/(n - )
S = y]RSS/(n- ).
S 2 ( S).

3.3.1
3.2.5, R2adj
[, 2, 3:
, =0.9889,
2 <=> Radj = 0.9902,
3 ^ =0.9911.
R2adj 3,
, t- F-.




(information criteria),
,
.
(Akaikes information criterion AIC).

, ,
3. , .. 155

'R S S .) 2 , ,
/ = In + + 1+ 12,

RSSp ,
.

, .
( )
AIC,

.
{Schwarzs information criterion SC, SIC).

, ,
( RSSp) plnn , .
SC = In ------- --------- 1 + 12.
J

, .
( )
SC.

3.3.2
, 2
3 AIC SC (. 3.14).
3.14

AIC SC

, 8.8147 8.9594
2 8.6343 8.7428

3 8.4738 8.5462

3.

3.3.2.
R ^ , AIC SC .
.
:
,
,

156 1. ,

.

. (),
.
,
, ,
,
.

J 3.3.3. ,
.
,
, ,

. ,
, , , ,
, .

.
(
, , (, ,
, 2005), . 14, . 351 382).


,
, , , ,
,
.

. :
4) 0:
detX r X * 0 ,
: X .
,
, .. X ,
:
xiP =\ + --- + -,-\ i = h- . . , n,
(-
X).
3. , .. 157

()
.
,

,. = a + /?xi +i , / = 1 ,..., ,

jtf = *. = ,
( )
, (*9).
, (,multicollinearity),
,
4 ,
0,
.
:

;

;

, ;
,
.
, p -
,
(variance inflation factor)


. R2 p

= \ + + Vp-\ x i,p- 1 + 1 = 1,..., .

R2p = 0, (VIF)p = 1, -
. R2p * 0, ( VIF)p > 1,
- ,
-
.
158 1. ,


(VIF)j -
j = 1, - , -

(W f)' = 7 ^ r
R2
j -
.

,
:

, , , ,

/-. (
. 3.3.5.)

3.3.3
, :

(VIF)2 =(VIF)3 =----- ------ = 109.89.


1-0.9909
( 2 3

2 3 3
1 2.)

2 3 (
(2, 3) = 0.995).


,
.

.
,

, ,
. ,
.
, ,

3. , .. 159

,

.

3.3.4
,
,
2 3.
:
Y= -6.507 + 0.146,
R2 = 0.9504 - = 0.0000 2\
Y= -9.030 + 0.222
R2 = 0.9556 - = 0.0000 3.
2
.
2 95.04% Y,
3 95.56% Y.
2 3 ,

, Y
( 95.60% ),
2 3.
2
2 , 3,
3
3 Y, 2.
2 3 , ,
2 3
Y.
. 3.15 R2adj, S, AIC SC .
3.15

S 1 SC
| 0.9462 1.1788 3.379 3.500
5 3 0.9454 1.1870 3.332 3.413
X, 0.9512 1.1228 3.220 3.301


2.
160 1. ,

. 3.1

,

, ( ,

,
,
).
. ,
, ,


.

3.3.5
. 3.1 500 ,
, 2, 3, 4. :
{ = 1+ 5 xi2 + X;3 + 0.5 xi4 + et ,
x i2 = 1 + 0.1/ +0.l-/2, x i3 = 2 + 0.5/ + 0 .l /3, x i4 =3 + t + 0 .\si4,
st, si2, sa, si4, ~ i.i.d. N(0, 1), i = 1 , 5 0 0 .
:
( VIF)2 = 21 747.73, (VIF)3 = 441 819.28, (VIF)4 = 444 750.88.
,
(. 3.16).
3. , .. 161

3.16
Y

-
/-

0.888304 0.373803 2.376393 0.0179

5.355627 0.464932 11.51916 0.0000


*2
0.905120 0.419226 2.159026 0.0313

0.511572 0.210310 2.432463 0.0153


X*


, .

j- ,
(. (Gujarati, 2003)):

(VIFh

/=1

( , , ( Wooldridge, 2000)).
, ( VIF)j
:
2;

(*,. - )2.
i=i
.
, : , /-
2 5 /-
3, ,
2 - 53.

F - ,
0
.
0 0 (nested
within model ) ,
(nested models).
0 (encompasses model 0).
162 I. ,

Mj 2
,
,
.
:
Mj : y t = a x+ p xxi +si , / = 1,...,,
2 : y i = a 2 +P2 zi +vi , / = 1,...,,
( zf .
2 (nonnested models):
,
.
.

.
\
,
. Mj 2
,
*: . = +2 f + 03zt +ei9 i = 1,..., .
* Mj 2:
.
, =0,
2, 62 = 0.
Mj 2:
2 ,
/-. Mj 2 ,
,
F- (
F - (nonnested F-tests)).
.
,
,
(. 3.17).
3.17

: 0 : 2 = 0


# 0 : = 0
D
3. , .. 163

,
, .
, z, .

,

.
,,
2.
: ,
. , D
, 2 .

,
, ,

, ,
, , :
, , ,
(log-log model).
:
, : y t = ,+ /?, , + ,, i = l,...,n , ( )
2 : In i = 2 + /?2 lnx, + v,, i = 1,..., , ( ).
,
, , {MacKinnon,
White, Davidson, 1983) .
, 2,
^, 1, .
, , a log-log
, =
:

*: = , + + (\ , ~ 1>;,) + *= 1......-
log-log ,
yLOG= 0 :

2* : *: In ,. = 2+/32 In , + yLOG - ,-)j + v,., i = \,...,n.

, /-,
,
164 1. ,

.
4 ,
F- (. 3.18).
.
3.18

0 : yLIN = 0



:y io<?- 0
D

,
, , ,
(, 2001), (, 2004).

= +
, () = 0, C o v ( e ) = <2/,

, ( + 1)-
, (-)
+ . ,
( + 1)- + 1
,

y i = Y j e j xij + i * = ! > , w + l , l, . . . , n+l~ i . i . d N ( 0 >o-2).
3=i
P
: g(i90) = ^ jej xij,
7 = 1

+1= + +1 = ( + 1, ) + +1,
3=1

g(n + \,e ) = f dejxn+l'j =E{yn+,).


7 = 1

,
9 ...,
g(n + 1, ):
3. , .. 165

g(n +1, ) = g(n + 1, ) = 0J +uj = +, .


j =1

E(g(n +1,9)) = +, = g(n +1, ) ,
.
g(n +1, ) = 2^@jxn+\, j (+,),
7=1

>(( +1,6)) = Cov(g(w +1,6)) = Cov(xn+l ) = +| Cov(<9) xj+1 =


= <r2xll+1(X r Jr)-Ix ^ I =<T2v,
:
v = jc+1(X r X )-1xr+1 (v ).
p
, j>+l.
7=1
+ - g(n + \ , 0 ) . ,
,
(forecast error) . :

(+, - g(n +1 0)) = (+,) - E(g(n +1, )) = g(n + 1 ,0 )- g(n + 1,60 = 0,

D(y n + 1 - g(n +1,0)) = D((g(n + l,0) + e+i) - g ( n + \, 0)) =


= D(e+l - g ( n + \, 0)) = D(e+,) + D(g(n +1,0)) = a 2 + a 2v = a 2v ,

:
v * = i + v = i + * +1( r t r 4 r+1.

+1- g(n + l,0 )~ N(0,(T2v*),

y n+1 - g ( n + \ , 0) ^ ^
<TVV*
2 S 2 = RSS/(n - p )

Sy/v*
_+|
. = 1- , 100^%-
+1 ( forecast interval) :
166 1. ,

g(n + \ , e ) - t ( n - p ) s 4 7 <yn+l < g(n + \,) + 1 ( n- p) S J7.


2 1 2

g(n + 1, ) = +1
(forecast) +1.

+
+1, (+1\+1).
(+1 |+1)

(+1| *+| ) = ( + !>) = +10,


a 2v , a a 2v.
(+1 |+|) (1 - ) :

g(n + \ , e ) - t { - p )S 4 v < (+1|:+1)< g (n + \, ) +1 ( - p )SVv.


2 2

( = 2):
yt = a + Pxt +,., 1 = 1,.. .,,
+ jc+i

+1= + /+1,

,
.
(+\\+1).
^+1 (+1\+1) ,

. v :

- 1 ! +1-*)2
2> ,- )2
1=1

, , = .

, (yi9 xf), i = 1, 9 (
) +1 ,

v*. i +i +1 =

3. , .. 167

(+] - )2. ,
:+| = *, ,
jc, ,..., .
, +1 = \ ,
,
.

3.3.6

1970 1979 . ( .,
1972 .) :
= -67.555 + 0.979DP/.
, 1979 . 1980 .
( ) DPI* = 1030
.
1980 . :
*98 = -67.555 + 0.979 1030 = 940.815.

S'2 =66.46, Id PI = 879.16, ^ ( , - ) 2 = 68015.18,
1=1

S V = 95.340, S y [ 7 = 9.7642.
0.95,
0.05( 2) = ^0.975(8) = 2.306,
2

DPI* = 1030 Ci980


:
940.815 - 2.306 9.7642 < *1980 < 940.815 + 2.306 9.7642,
..
940.815-22.516 < \ <940.815 + 22.516,

918.299 < \ <963.331.


,

.
, 1980 .
1021 ., 931.8 .

168 1. ,

1940.815 931.8 J
100 = 0.97%.
931.8
DPImo = 1021,
DPI* = 1030, 1980 932.0

1932.0 931.81
100 = 0 .021 %.
931.8
,

95%-


DPI. . 3.2


.
, S4/v =5.374,
900 1000 DPI , 95%-
(*980|/}/* = 1030)
:

940.815 -2.306* 5.374 <( \DPI : 1030) < 940.815 + 2.306 5.374,
..
940.815-12.392<(* \DPI* =1030)<940.815+ 12.392,

928.423 <*980 \DPI* =1030) <953.207.

, ,
*1980.

1.
?
2.
:
, ;
, ?
3.
?
3. , .. 169

4.
?
5. ?
?
6. ?
7. ?
8.
?
9. ?
? ?
10. ?
11.
?
12. ?
?
13.
? ?
14. ? ?
4

_______ 4.1________



( ,

).
R2 (
1)
,
,

(<adequate model)
.
4
(. 4.1),


y i = a + p x i +ei9 / = 1,...,,
( )
, R2 /-.
4 :
>> = 6.00 + 0.50;
() s& = 1.12;
() Sp = 0.12;
4. _____171

/- 0 : = 0 2.67,
- 0.026;
/- 0 : = 0 4.24,
- 0.002;
R2 = 0.67.
. 4.1
(4 )

1 2 3 4
1
X X X X
1 20 16.06 20 18.28 20 14.92 16 13.16
2 16 13.90 16 16.28 16 13.54 16 11.52

3 26 15.16 26 17.48 26 25.48 16 15.42


4 18 17.62 18 17.54 18 14.22 16 17.68
5 22 16.66 22 18.52 22 15.62 16 17.94

6 28 19.92 28 16.20 28 17.68 16 14.08


7 12 14.48 12 12.26 12 12.16 16 10.50
8 8 8.52 8 6.20 8 10.78 38 25.00
9 24 21.68 24 18.26 24 16.30 16 11.12

10 14 9.64 14 14.52 14 12.84 16 15.82


11 10 11.36 10 9.48 10 11.46 16 17.98

(. 4.14.4).
4 ,
(. 4.1)

y i = +P x t + s i , i = l,..., .
(. 4.2)
;= + /3 Xj + y x f + st , i = l,...,n.
(. 4.3) (3-
),
.
(. 4.4)
,
.

,

172 1. ,

. 4.1 . 4.2

. 4.3 . 4.4

.
.
:
..., ,
, ;
. ..., (
), ,
;

..., ;
t- F -
, t- /^
. /- F -
/- F-

;
, ,
.
4. 173

, :
, ,
, (diag
nostic for model misspecification);
,

.
,
,
.
, ,

.
, ,
:
;
;
,
;
;
.

, , ,
(
, /-
, ).
,
(LSE approach; LSE London School o f Economics),
.
,

, , ,
.
,
,
,
.

: ,
(test, test and test) (. (Hendry, 2003), . 27 28).
,
, :
174 1. ,

;
;
.


, = 1 +... + 6pxip + ,, 1= 1,.. ,
..
= 6 + ,
(residual analysis), ..
e t= y t - y t , / = 1,..., .
y t - y t ,

Yj - Yj, et.

= - = - { { = (/ - ( )~1 ) = ( 1 - )
,
.

,

() = ( - ) = ( + - 6 ) = 6 - =

Cov(e) = Cov((I - )) = (/ - H)Cov(y)(I - )=


= ( \ 1 - ) ( 1 - ) = ( 2( 1 - ) .

, , ..
ei9 :
(,) = (), / = 1 , ,

D(ej) = cr2( l - h ii), i I,..., ,


hu /- ( )-

= ( ) - 1 .

, <?,
( 2), ,
, .
4. 175

.. , .

<2 ,
(studentized
residuals)

-
hu
dt ,
(standardized residuals)

S
( , , Excel Econometric Views).
, d{ ,
( !) .
(graphical
diagnostic analysis) ci9 i = 1, ..., .
= ( )~1 ,
.
( ),
,

^ = /> .
1= 1

( \ hH
.
()

. , ,
176 1. ,

,
ei9 i = 1, ..., ,
si9 i = 1, ..., . , ,
si9 i = 1,..., , ,
N(, 2), ,
ei9 i = 1, ..., 9
,
N(09 2).
,
,
N(, 1).
, .
, (
) ,
. , (
)
,
.

-

,
N(09 1).
( )

( ) :
y t = + ... + 0pxip9
;
,
.

,- = + ... + 0px ip
3
.

(outliers) ,
E(et)
0, D(&)) <2
.

. , ,
( 4 ) (. 4.5).
4. 177

Standard Residuals & Y3F Standard Residuals & YF


Resid Stand Resid Stand

. 4.5 . 4.6

, (heteroskedasticity),
, ,
D(Si) 1 + ... + &pxip. ,
. y t = xxix + ... + , . 4.6, , ,
6xxix + ... + 6pxip.

(regression error specification)
E(st) = 0, E(Yt) * + ... + dpxip. ,
, 4
(. 4.7).
Standard Residuals & Y2F Standard Residuals & X2
Resid Stand Resid Stand

10 20 30 X2

. 4.7 . 4.8

ci9 -
'-
. , 4
. 4.8.
178 1. ,

. 4.9 . 4.10


, (
). :
(. 4.9);
() ,
(. 4.10);

ei9 i = 1, ..., , .

, ,
(. 4.11)
(. 4.12).

. 4.11 . 4.12


( , ,
, ).

4. 179

( , ,
, ).

et,
/ 1,..., .
- (Q-Q plot).
,, i = 1, ..., ,
.

(1) < (2) < ' < ()

k = 1, ...,
()

-1 ___2
=

-
2
Qk - ---- ;

(*)- ,
Qk), - 1, ..., ,
,
, 1. ,
4
(. 4.13).

4.4.1.
, eh
i = 1, ..., , (e(jt), Qk), = 1, ..., ,
( )
, ,
1.
-
,

--
L(*) -1 ___ 2

J
180 1. ,

Q-Q Plot
Normal Quantile
24

2 Resid_Stand

. 4.13 . 4.14

,
Excel.
1
--

(DP-plot, DPP) -
- , Qk

() , Qk, .. <p(Qk)-

,

. ,
,
(CONS)
(DPI) ( .
1982 ., 1959 1985 .) (. 4.14).
,

.
. ,
.


.
(kernel density estimates)
,
. ,
, , Views
4. 181

8 , Kernel Density
(Epanechnikov, h = 0.5025)
. ,
,


(. 4.15).
,
,


, 0 1 2
RESID_STAND

( = 27). . 4.15

(instability) , ,

(
recursive coefficients).
DPI , . 4.16,
DPI
.

. 4.16

Views
(recursive residuals),


.
,
182 1. ,

, .
( + 1)-
wp+!. ( + 1) ,

( + 2)- . wp +2.
,
.
, (t + 1)- ,
, t , :
CT2(l + xt+l(X ^ X ty \ x t+{)r ),
X, t
;
1+, - (t + 1)-
.
(t + 1)-
(1 + xt+x(XjXt)~' (+1)).
, ,
wp+ 1, ..., w ,
,
2.
,
.
, , ,
2 .
, ,
, .

.

. 4.17.
EViews
{CUSUM cumulative sums),

Wk = V , S = I, a RSS
f +1 s \-
.
(
) Wk
, , (, 0.948 Jri-p)
(, 3*0.948yjn-p), 0.95.
4. 183

. 4.17

(. 4.18)
.

. 4.18

1. ,
, ?
2. ?
3. ?
4.
?
5. :
;
();

;
184 1. ,

;
;
;
?

------------- 4.2--------------

,

, .
.

H 0 :eu ...,s n ~ i.i.d N ( 0, 2).

,
,
.
(Goldfeld-Quandt test).

D(st)9.. , :
, ,
;
(

), ( - ) ;
( - )/2
( - )/2 ;
F = RSS2/RSS{ ,
( - )/2 (
RSS2) ( - )/2 (
RSSX) .
, D(si) = 2,
/ = 1, ..., , ( )
,
, F = RSS2/RSS{ F-pacnpe-

;
4. 185

# 0 : >($) = 2, i = 1, ..., , (,
homoscedasticity) ,
F - , ..
/ -r -
Fx_a - , ----- , -
V 2
.
(Durbin-Watson test) ,
,
, -
( )
. ,

Si =pSi_l + 6i9 / = 1,.
\\ < 1, a Si9 i = 1, ..., ,
, N(0 , | ) , St
6t_s s > 0 .
(Durbin-Watson statistic)

DW =
.2
i =l
el9..., en ,
.

# 0 := 0 ,
(
) s l9 ..., sn.


0 : > 0 ,

s l9..., (..
/- (/ + 1)- ).
DW 0 4.
, 0 : = 0 (..
186 1. ,

) (),
- 2 . - * > 0,
DW .


da (0 < da < 2 )
# 0 : = 0 0 : > 0 D W < da.

9 xij9j = 1,...,/?, i = 1, 9
,
.
. ,

, ( )
dLa dUa ,
da ,
, xij9j = 1, ..., , i = 1, ..., .
,
0 < dLa <da < dUa < 2 ,

dLa dUa xij9j = 1, ..., , i = 1, ..., 9


,
.
0 : = 0 :
: > 0, D W < dLa\
, DW> dUa.

dLa <DW< dUa9


0 : - 0 .

0 : = 0 .
(Jarque-Bera test)
(, Views)
0 , ,

# 0: ~ i.i.d. N( 0, 2)
( 2 ). ,

{sample skewness) (sample kurtosis - 3)


JB = n
24
4. 187

, - 2
\ 2 ), :
1 --
() = 2 , > 0 .
:
sample skewness

sample skewness = - - - - ;
(2) 2
sample kurtosis

sample kurtosis =j ;
m2
If k
mk = - 2 s ei>
,=i
e x, ..., en ,
.
,
0 ,
3.
0 ( 0 )
( ).

JB .
, ,

J B > x l a(2),
\- (2 ) j 2(2 ), 1 - .

J 4.2.1.
( -
asymptotic tests) ,
. ,
(asymptotic test):
JB j 2(2)
.

. ,

188 1. ,

(
, ,
).

(Breusch-Godfrey test)
(, Views)

1 = \ + + e p x ip + i> *= 1,.... -
(
)
si9 i = 1,..., . ,
,
et
e ^ p e ^ + S f, i = l,...,n ,
\\ < 1, a St, / = 1, ,
, N(, 2), 8t
et_s s > 0 .
st,

Sj, .

j,
Sj.
:
= , +... + Si_K +Sf, i = 1,. . . , ,
Sj, i = 1,..., , ,
N(0, cr2s ), St
s t_s . > 0 , \\ < 1
:
1 - atz - ... - aKz K= 0 ,
, 1 .
nR2, R2
,
, = ++ xi + +... + aKei_K+vi, i = \,...,n ,
e l9..., ,

4. 189

y t = + ... + epXjp + et,


0, _ .

//0 : | = ... = - 0 .
,
, -
. 0
, nR2 ,
(1 - ) , ..
nR 2 > (nR2)crit= z L ( K ) .
,
, ,
. , ,
,

, .. = 1;
,
.
.
( White test)
(, Views)

,- =\ +... + 1+1, i = 1,..., .
.
I.

$ = 1+ H aj xv + Y*Pjxi}+ yi> *= u . ,
7=2 j=2

,
,

: a j = J3j = 0, j = 2,..., p .

nR2, R2 , -
. ,

, - (2 - 2 ) .
0 ,
190 1. ,

nR2 ,
(1- ) , ..
n R 2 > (nR2)crit=X l a (2 p -2 ).
.

= 1 + Z aj + X Z Pi* i= 1>>">
j-2 j=2 k=j

e l9..., en ,
,

fa, = 0 , j = 2 ,...,p ,
: [ = 0 , 2< j < < .
nR2, R2 , -
. ,
-
_ 2+ - 2
, - - ------
. 0 ,
nR2 ,
( 1 - ) , ..
(
+, ~ 2_

n R 2 > (nR 2)crit = \-

,
,
.

/ 4.2.2.
, = 1.
, ,
,
j = 1 .

RESET (Ramsey's Regression Specification Error Test)



= { + ... + + i = 1,..., .

E(s{) = 0, / = 1,...,,
- .
4. 191

:
(..
);
(
) (,
);

( , ,
.
. 6 ).
. ,
Views.
,
y i= d xxa +... + 9pxip, i = l,...,n .

! = \ + + 0pxip + + r r-i7 i, I = 1, ,,

ffo'- = = -1 =-
,
, -
( - 1) . 0
, nR2
, (1 - ) , ..
n R 2 > (nR2)crit= x l a{ r - 1).
.

, {.
^.

(Chow tests).
. Chow
breakpoint test


. (Chow forecast test)
,

192 1. ,

, ,
. ,
:
/ = 1,..., 0 / = 0 + 1,..., .

, = &\ + + &pxip +, i = 1 , . . ,
,
tin), i = n0 + \,..., .
0
.

. i = n0 + l,...,n .
, (0)
^), i = n0 + 1,..., .

(R S S - R S S J /( n - n 0)
RSSnJ ( n 0 - p )
RSSn
;
RSSo
0 .
( ),
F -, F(n - 0, 0 - ).
,
F x_a(n - 09 0 - ).

4.2.1
1960 1985 .
:
DPI ;
CONS ;
ASSETS .
( ., 1982 .). . 4.2.

. 4.19.
4. 193

4.2
1960 1985 . ( 1982 ., .)

DPI CONS ASSETS DPI CONS ASSETS

1960 1090.9 1005.2 1203.1 1973 1914.9 1687.9 2190.6


1961 1122.5 1024.3 1226.7 1974 1894.9 1672.4 2301.8
1962 1168.7 1067.6 1293.7 1975 1930.4 1710.8 2279.6
1963 1208.7 1109.6 1374.9 1976 2 0 0 1 .0 1804.0 2308.4
1964 1289.7 1169.4 1464.5 1977 2067.9 1884.9 2421.6
1965 1367.4 1237.9 1544.0 1978 2166.5 1960.2 2554.9
1966 1433.0 1300.5 1641.6 1979 2211.4 2003.6 2666.2
1967 1494.9 1339.4 1675.2 1980 2214.8 2000.7 2704.3
1968 1551.1 1405.9 1772.6 1981 2249.0 2024.4 2682.8
1969 1601.7 1458.3 1854.7 1982 2261.4 2050.7 2741.8
1970 1668.1 1491.8 1862.2 1983 2332.5 2146.5 2850.1
1971 1730.1 1540.3 1902.8 1984 2470.5 2246.3 3038.1
1972 1797.9 1622.3 2011.4 1985 2527.3 2323.9 3267.7

. 4.19


CONSt ={+ e2DPIt + ASSETS, + t = 1 ,..2 6 ,
t (1959 + t)- . 3
:
=1, 2 = DPI, 3 = ASSETS.
: R2 = 0.9981,
194 1. ,

2 = 0.672, - = 0.0000;
# 3 = 0.174, - = 0.0069;
2 = DPI,
= ASSETS .
(CONSF)
(CONS) CONS (. 4.20), -

ct = (RESJSTAND) -
S
(CONSF) CONS (. 4.21).

CONS RES STAND

CONSF CONSF

. 4.20 . 4.21

. 4.20
. . 4.21
CONSF> 1600.

,

(
).
26 , 10 10
(, , ), 1960 1969 .
1976 1985 . ( = 6 ).

RSSX = 208.68 RSS2 = 1299.66
, F-
:
4. 195

RSS2 1299.66 22g


RSS, 208.68


,
F -
26_6 26_6 ^
- 3 , --------- 3 \ = F (1 ,1). , , -
. 2 2
= 0.05,
F- F 095(7, 7) = 3.79.
6.228 ,

D(et)
+ 92DPI + ASSETS. , ,
F{7, 7) 6.228 (-)
0.0138.

, .

nR2 = 8.884. = ,
- 2 - 2 = 4. ,
, , 8.884,
0.0641, nR2 = 8.884 , ,
.

nR 2 = 9.699. -

- -2+
-----=
~ 2 5.
,
, , 9.699, 0.0842,
nR2 = 9.699 , ,
.
,

:
,
. ?
,

.
,
-
196 1. ,

,
,

. :
, ,

,
(.. , ,
).
et
ct =
S
(. 4.22)
(. 4.23).
RESID STAND RESID STAND

1960 1965 1970 1975 1980 1985

. 4.22 . 4.23


, , ,
, . 26
= 3
- 0.05 ( ) :
^ l, .5 = 1*22, dv .5 = 1-55.

DW - 1.01,
dL .5 = 1-22. ,

.

.

(= 1),
4. ______ 197

Norm_Quantile DP

. 4.24 . 4.25

, -
nR2 = 6.068, -
, 0.014. ,
,
.
= 5, nR2 - 10.331, --
, 0.066.
= 0.05,
,
. = 6 (nR2 = 0.095), - 1
(nR2 = 0.127) .. : = 5, - 6, =7

Kernel Density
= 1,


, .
, ,

.
-
(Q-Q plot) (. 4.24),
(DP-plot)
(. 4.25)
RESID STAND
(. 4.26).
. 4.26

,
, .
-1.285,
198 1. ,

5.321.
. 12.997,
- 0.0015. ,
.
RESET . 4.3.

. 4.3
RESET

/ -

2 0.0015
2, 3 0.0046

2, 3, 4 0.0032


E(st) = 0, t=
, (Chow
Forecast Test), = 13, -
F = 11.037, - 0.0003
.
, ,
CONSt = +62DPI, +ASSETS,+ t = 1,..., 26,
,
,
, E(st) = 0,
t = 1,..., , .
ct ~~^


( 1972 .) (1973 1985 .).
, 1973 .
, --
,
. , ,
2,
.

4. 199

CONS, = yi(D\)t +y2(D2)t +y3(DPI\)t +yt(DPI2)t +


+5(ASSETSI), + 6 (ASSETS2), + 7= 1,..26,
(.D\), , 1 7 - 1, ..., 13 (
1960 1972 .) 0 / = 14, ..., 26 (
1973 1985 .);
(D2), = 1 - (D1), , 0 7= 1,..., 13
1 7 = 14,..., 26;
(.DPIl), = DPI, (D\), , (DPI), 7=1, ..., 13
0 7 = 14,..., 26;
(DPI2), = DPI, (D2), , 0 7 = 1, ..., 13
(DPI), t = 1 4 , 2 6 ;
(ASSETSI), = ASSETS, (Dl), , ASSETS, t = \ , 13
0 t = 14,..., 26;
(ASSETS2), = ASSETS, (D2), , 0 7 = 1 , 13
ASSETS, 7=14,..., 26.
,
(DPIl), + (DPI2), = DPI,, 7 = 1,..., 26,
(ASSETSI\ +(ASSETS!), = ASSETS,, t=1 ,...,26 .

0 : = 2, =, =(,-
q = 3 ,

'-\-=, 3~4 =0, 5~6 =0.
, F -.
F- (Chow break
point test). F - 10.490 - 0.0002,
0 ,
0, 2> .
5 (
- 0.1157 0.5599),
(q = 2), F-.
- 0.2412 ,
, ,
,
ASSETS
CONSt =yx(D\)t +y2{D2)t +y3(DPI\)t ^ y A(DPI2)t +n t = 1,..., 26.
200 1. ,

: R2 = 0.9992,
, - 58.786, - = 0.0119;
2 = -234.836, - = 0.0000;
3 = 0.864, - = 0.0000;
= 1.012, - = 0.0000.
# 0 : 3 = 4 (q = 1) (- = 0.0000),
0 : , = 2, ,
DPI.
DW = 2.06
.
= \, =2, = 3.
- = 0.508, ,
- = 0.469,
.
, , .
1960 1969 . 1976 1985 ., -
3.354, - = 0.0832,
.
RESET . 4.4.

. 4.4
RESET

i -

2 0.3080
2, 3 0.4132
2, 3, 4 0.3659


E(st) - 0, t = l,...,n .

,
1960 1985 .,
CONS, = 58.786(>1), - 234.836(^2), +
+ 0.864(Z)P/1), +1.012(DP/2), +et , t = 1,..., 26.

[58.786 + 0.864(DP/), +., t = 1,..., 13
CONS,=\ '
' [-234.836 + 1.012(D PI),+et , t = 14,...,26.
4. 201


{two-phase linear regression model)
{switching regression model). , ,
, DPI
,
, .. , D{st) = <j\ t - 1, ..., 13
D(et) = < j\ t = 14, ..., 26. < <2
8.887 14.886.

/ 4.2.3. , , ,
, ,


. ,

.
, (

) ( ),
19 2, ...,
Rl9..., RK9
, { RK} = a, k = 1, ..., .

, RK , = 1, ..., .
1- \ ..
, ,

. * , , ,

.
*.
:
*_ p{TK RK k9k= 1,..., } <

< f,P { T k e R K} = K a.
k =1

,
,
/ .
1- .
, {9 2, ...,
202 1. ,

1-
.
(Godfrey, 2005).

J 4.2.4. ,
, , ..
, ,

(time series analysis).
.

1.
? ?
2. ?
?
3.
?
4.
( , )?
5

_______ 5.1________

. 4
1960 1985 . .
, - 1973 .

CONS, = + e2DPI, + ASSETS, +,, t = I,..., 26,
(
). ,
1960
1972 . 1973 1985 .
:
CONS, = yx(DX),+y2{D2),+y,{DPI\), +y4(DPI2), +
+ y5(ASSETSl)t +y6(ASSETS2), +s t = 1,..., 26,
(D1), (D2)
(>2), = 1 - (D 1) (D1), = 1 t= 1,..., 13 ( 1960
1972 .) (>1), = 0 7=14,..., 26 ( 1973 1985 .).

{dummy variables -, dummies ),
(
) , -
, ,
,
204 1. ,

..
,
, , , ..
3 ,
6
.

,
4 , , ,
.
--
,
( ). .

5.1.1
. 5.1
DPI SALES
( 1964 1972 .; D PI .,
SALES ., 1972 ., . . 5.1).

SALESt = + PDPIt + st9 t = 1,_, 36,
, . 5.2.

. 5.1

DPI .
(
. 5.2 . 5.3)
.
,
: 1.966 2.
5. 205

5.1
DPI SALES
( 1972 .)

, DPIY912, SALES 1972, , DPI 1912 SALES 1972,


. . . .
1964:1 157.4034 53.43051 1968:3 184.8915 47.97185
1964:2 165.5068 48.21284 1968:4 185.3775 55.73602
1964:3 161.5369 44.02953 1969:1 185.6923 54.49402
1964:4 165.2709 53.99799 1969:2 186.6742 49.77978
1965:1 166.4503 52.75629 1969:3 188.8089 51.92244
1965:2 168.7129 44.42772 1969:4 189.7213 55.98525
1965:3 170.4000 47.48852 1970:1 190.0968 56.91452
1965:4 172.0651 52.86840 1970:2 192.7143 49.47460
1966:1 172.5916 54.79100 1970:3 195.0632 46.63579
1966:2 173.6561 47.48408 1970:4 193.6364 56.43117
1966:3 175.2532 47.04873 1971:1 196.1125 56.65473
1966:4 175.7250 53.51625 1971:2 196.9447 49.45025
1967:1 175.9565 55.43292 1971:3 197.6725 49.78809
1967:2 178.6025 45.90745 1971:4 197.3382 55.02500
1967:3 181.4444 51.01481 1972:1 197.0121 53.84310
1967:4 181.8293 56.75671 1972:2 197.6394 48.12981
1968:1 182.3892 56.37485 1972:3 196.3040 48.36770
1968:2 185.2722 50.91834 1972:4 196.0000 52.80000

5.2
SALES
(Method: Least Squares; Sample: 1964Q1 1972Q4; Included observations: 36)


| /- -

f
1 27.99904 6.954436 4.026069 0.0003
DPI 0.119943 0.040061 2.994036 0.0051

R-squared 0.208644 Mean dependent var 48.74395


Adjusted R-squared 0.185369 S.D. dependent var 3.970245
S.E. o f regression 3.583419 Akaike info criterion 5.444465
Sum squared resid 436.5904 Schwarz criterion 5.532438
Log likelihood -96.00037 Hannan-Quinn criter 5.475170
F-statistic 8.964254 Durbin-Watson stat 1.965655
Prob. (F-statistic) 0.005103
206 1. ,

IhJilkrilb.
"|| 1 I

- 2.0 - ) | 1 I 1 I 1 11 U l N U l t i
I4 - > eg
) G

)

)
)
) )
f4- h-
05


. 5.2 . 5.3

:
, ,

(. 5.4).
. 5.4 RES01 ,
RESO'l
8 , PisSO ^-l)
6"
,

4 --
RES01.
2 --
0
-2 ,
-4 (. 5.5),
-6 ,
(. 5.6).
-8
-8 -4 4 8

RESO'li-'l) ,
. 5.4 :

,
, - = 0.000009,
.
.
, I IV , , ,
II III , . , ,
,
- - ,
.. .
5. 207

RESO'l

4 8
R E S 0 1 (-2 ) RESO'li-A)
. 5.5 . 5.6


. ,
,
.
DUMMY, 1
I IV 0 II III .
.

SALES, = + DPI, + DUMMY, + , , t = 1,..., 36,
, . 5.3.
5.3
SALES 1972
(Method: Least Squares; Sample: 1964Q1 1972Q4; Included observations: 36)

-
/-

23.93338 2.781524 8.604413 0 .0 0 0 0

DPI 1972 0.124919 0.015933 7.840376 0 .0 0 0 0

DUMMY 6.410030 0.475062 13.49303 0 .0 0 0 0

6.410 DUMMY
, I IV
II III
6.41 . ( 1972 .). . 5.7
.
- (
1 2 ) 0.4389
208 1. ,

Residual
Actual
Fitted

. 5.7

0.000009,
, ..., .
,
DPI SALES:
SALES = 23.93338 + 0.124919DP7
- ;
SALES = (23.93338 + 6.410030) + 02499DPI
- .
( )
0.124919.
, ,
,

1960 1985 .
.
:
SALES, = + pDPI, + DUMMY, + (DUMMY, DPIl) + ,, t = 1,..., 36.
:
d SALES,
L = fi + SDUMMY, ,
d DPI,
.. DUMMY,'.
d SALES, \P, DUMMY, = 0
dDPI, "1 p + S, DUMMY, =1.
5. 209

, DPI DUMMY
( ),
DUMMY, DPI,
(interaction) DPI DUMMY.

.

/ 5.1.1. SALES, = + PDPI, + DUMMY, +


t = 1, ..., 36, -
DUMMY,
1 I IV 0 II III ,
DUMMY* = 1 - DUMMY, 0 I
IV 1 II III :
SALES, = DPI, + DUMMY, - 8 DUMMY", + t = 1 ,..3 6 .

8
- - ,

(differential effect)
- .

5.1.2. -
-
(dummy trap). , 5.1.1,
,
.
- ,
..
SALES, = + DPI, + DUMMY, + 8 DUMMY*, + s , , 7= 1, ...,36,
:
(\ DPI\ 1 "
1 DPI2 0 1
1 DPI3 0 1
1 DPI4 1
=
1 DPI 1 0
1 d p i 34 0 1
1 DPI35 0 1

,1 DPI36 1 ,
210 1. ,

1- 3- 4- ,
.. X ,
4, (det = 0)
, {)~ .
Econometric Views


{Error Message): Near singular matrix (
).

5.1.3. -
-
.

, , DUMMYX, DUMMY2, DUMMY4,
1 I, II IV
0 .

DUMMY!,
II III . ,
DUMMY DUMMYX.
,

DUMMY, .


() ,
, ( ),
, , .
, ,
, , ,
.


, ,

,
. -
(, 2004).

J 5.1.4. -

,
-, 5.1.2,
5. 211

( - 1) -
. , --
, 1 .

/ 5.1.5.
-
, ,
- 1,

.

-
{panel data), ..
(, ) (
, ). (
, ) (, , )
(cross-section data),
(, )
(, ) (time-series data).

5.1.2
. 5.4
3 (N = 3) (7 = 1 0 )
(. (Greene, 1993), . 481). Yif Xt /-
, i = 1, 2 , 3.
5.4
3 10

1 2 3
t
Y1 XI Y2 2 F3

, 1 13.32 12.85 20.30 22.93 8.85 8.65


2 26.30 25.69 17.47 17.96 19.60 16.55
3 2.62 5.48 9.31 9.160 3.87 1.47
4 14.94 13.79 18.01 18.73 24.19 24.91
5 15.80 15.41 7.63 11.31 3.99 5.01
6 1 2 .2 0 12.59 19.84 21.15 5.73 8.34
7 14.93 16.64 13.76 16.13 26.68 22.70
8 29.82 26.45 1 0 .0 0 11.61 11.49 8.36
9 20.32 19.64 19.51 19.55 18.49 15.44
10 4.77 5.43 18.32 17.06 20.84 17.87
212 1. ,

( Econometric Views)

= 1 + . , 10 ,

II
+
+ PiX2t + S2(? t 1, .. . , 10 ,

II
ft

2,
, = 3 + 33, + 3,, t - 1, .., 10 ,
:
1- (. 5.5):
5.5


/- -

-2.468913 0.980426 -2.518205 0.0359
\ 1.167170 0.058250 20.03737 0 .0 0 0 0

R-squared 0.9805

2- (. 5.6):
5.6
2


/- -

-1.384797 1.972680 -0.701988 0.5026
XI 1.014542 0.115314 8.798102 0 .0 0 0 0

R-squared 0.9063

3- (. 5.7):
5.7


/- -

0.455479 1.491604 0.305362 0.7679
1.076374 0.100360 10.72516 0 .0 0 0 0

R-squared 0.9350

2, 3
,
:
0 = /?2 = '
,
30 -
5. 213

(
). - D l, D2, D3
:
1, i = 1, 1, / = 2 , 1, i - 3,
D2it=\ D3=\
[, / * 1; [0, i ^ 2; [0, *3.

:
0 : = D \it + 2 D2it + <3D3it + (D 1 xit)+ 2( xit) + } (D3jt xit) + it,
i= 1,..., 3, 7=1,..., 10,
- :
/
' ' 1 0 0 0 0


1,\ 1 |10 0 0 0 0 ,
0 0 1 jc21 0 0 '21
2
+

2,10 0 0 2 0
1 0 ' 2,10

\ 0 0 0 0 1 31 '31

^ , V0 0 0 0 1 -^3,10j \ ;

- + ,

^ ( 1 *11 0 0 0 0 '

0

\, 1 * 0 0 0 ' 1,10

0

0 0 1 *21 0 '2 1

= , X - , = , =
3^2,10 0 0 1 * 2,10 0 0 ' 2,10

\ 0 0 0 0 1 *31 '31

1 -) 0 0 0 1 *3,10 ^.'

,
: = ( )~1 .
214 1. ,

,
,
0 : = = F-.
(. 5.8).
5.8


/- -

D\ -2.468913 1.388033 -1.778714 0.0880
D2 -1.384797 2.233064 -0.620133 0.5410
D3 0.455479 1.137109 0.400559 0.6923
D\ X 1.167170 0.082467 14.15324 0 .0 0 0 0

D 2X 1.014542 0.130535 7.772208 0 .0 0 0 0

D3 X 1.076374 0.076508 14.06874 0 .0 0 0 0

R-squared 0.950532 F-statistic 0.592788


Probability 0.560676

, 0 : = - ,
:
Mj : y it= a xD \it+ a2D2it+ a3D3it+ P xit + sin i = 1,...,3, t = 1,...,10.
M l9
,
-. , '0 : { = 2 = 3,
= Pi = 0.
.
.

1. -
?
2. -? -
, ?
3. - ,
?
4. - ,
?
5. - ,
?
6. , ?
- ?
5. 215

_______ 5.2________

,
(, heteroscedasticity),
, ,
, , (
).
(,
). 4.2.1
:
.
(-
):
19 ..., (
),
, ;

9 ...,
;
/- F -
, t- F-pacnpe-
, .
t- F-
t- F-

.

5.2.1

27
(. 5.9). :
xt /- ,
y t z- .

y t = a + p x t + si9 / = 1,..., 27.
R2 = 0.776 ,
. 5.10.
216 1. ,

5.9
X
Y 27

1 X Y 1 X Y

1 294 30 15 615 100

2 247 32 16 999 109


3 267 37 17 1022 114
4 358 44 18 1015 117
5 423 47 19 700 106
6 311 49 20 850 128
7 450 56 21 980 130
8 534 62 22 1025 160
9 438 68 23 1 021 97
10 697 78 24 1200 180
688 80 25 1250 1 12

12 630 84 26 1500 210

13 709 88 27 1650 135


14 627 97

5.10


/- -

1 14.448 9.562 1.511 0.1433

X 0.105 0 .0 1 1 9.303 0 .0 0 0 0

. 5.8
= 14.448 + 0.105, . 5.9
-
ct = j), = 14.448 + 0.105*;.
S
,
j),
D{t) = a f = cr2x f .
,
y t = + /? + s t :
5. 217

. 5.8 . 5.9

..
] = (J + * +*,

. yt 1 . t
,=>
xi
*, =
xt
> */=
xt
D{st) = o f = <2?,

( ) = 0, D(s;) = - ^ D ( ,) = a 2,
(.
.. .
. 5.11.

5.11
/


/- -

1 0 .1 2 1 0.009 13.445 0 .0 0 0 0

3.803 4.570 0.832 0.4131


= 3.803 + 0.121.

/?.
.
- 3.803 /? = 0.121. . 5.10
218 1. ,

RES STAND


*.
(
, ) .



. ,
. 5.11,

* = /5 + * + %
. 5.10 ..

/=1
(, )
:
' -'^ - - -
1 V = ^^ 1 ( , - - 1)2.
1= 1 xf =1 ,
1
wt, = - , , -


(-
) . wt ,
/- .
, ?,
D{t) = ? = 2 2 i-
. , ei9
, i-
.

( , . . ) 2,
/=1

,
. (, ,
, 2005) ,
5. 219

: WLS weighted least squares


OLS ordinary least squares.

/ 5.2.1.
(, Econo
metric Views)

:

/= 1

,
(, - - (5 xt). ,
: wi = .
Xi
,
.

,

Econometric Views. 5.2.1
.
,

w = .

. 5.12.
:
(weighted statistics) ,
, , ..
] =* - - * ;
(unweighted statistics) ,
ut = yt - a WLS - PWLSxi9 ..
,
,
, a WLS, PWLS, .

(0.02696)
.
,
F -, 180.7789,
- 0.0000 (. Weighted statistics). -
220 1. ,

5.12
Y
(Method: Least Squares; Sample: 1 27; Included observations: 27; Weighting series: 1/X)


/- -

3.803296 4.569745 0.832277 0.4131


0.120990 0.008999 13.44540 0 .0 0 0 0

( Weighted statistics)
R-squared 0.026960 Mean dependent var 74.04946
Adjusted R-squared -0.011961 S.D. dependent var 13.08103
S.E. o f regression 13.15902 Akaike info criterion 8.063280
Sum squared resid 4328.998 Schwarz criterion 8.159268
Log likelihood -106.8543 F-statistic 180.7789
Durbin-Watson stat 2.272111 Prob (F-statistic) 0 .0 0 0 0 0 0

(Unweightedstatistics)
R-squared 0.758034 Mean dependent var 9 4 .4 4 4 4 4

Adjusted R-squared 0.748355 S. D. dependent var 45.00712


S.E. o f regression 22.57746 Sum squared resid 12743.54
Durbin-Watson stat 2.444541

R2, ,
, (0.758).
, . 5.12
, 7-, -
, .
, , R2 = 0.758,
, R2 = 0.776,
() . ,
, , R2 = 0.776
ei = y ,- a - P X i ,
, /3 ,
.


. 5.2.1 , :
D(i) = a f - 1x f,
5. 221

ct = j), .

(Glejser test). ,

, .

,

. X ,
:

= +y2X i +Vi

| = +2/^7 + v

= 1 + 2 -^- + >'/>

0
2. ,
.

.
,

^ 1?
9 ..., . (
, . . 6 .)


(OLS-, ordinary least squares estimates) l9 ..., p
9 ..., -
s.
s&
,
Views.
( White estimator)
.
222 1. ,



, ,
/
,
Cov(e) = V,
V , , V * 21.
:
Cov(9)= Cov({XTX y lX Ty) = ( X TX y ' X TC o v(y)X (X TX ) - { =
= (. 1( X TV X \X T .
,

of

Cov(6) ,
.
C ov(6\ ,
s#.,
V 2, ..., 2
,
, .. ..., 2.
.

5.2.2
,
, EViews.

, = 14.448
= 0.105. sd Sp sd = 9.562
Sp = 0 .0 1 1 ,
, sa = 10.633 Sp = 0.018.

(3.803 14.448)
, ,
, (s = 4.570 s^ =
10.633 ).
5. 223

5.2.3
, 5.2.1 5.2.2,
Iny j= a +fiXj+Sj, / = 1, , 27.
, -

, ln^. . 5.11.

RESID_STAND RESID_STAND

. 5.11 . 5.12

,
xf.
,
2, ,
(. 5.12).
. 5.13.
5.13
Ln Y


- -

1 2.851 0.157 18.205 0 .0 0 0 0

X 0.003 0.000399 7.803 0 .0 0 0 0

X2 -1 .1 0 -0 6 2.24 - 07 -4.925 0 .0 0 0 1

, ,
:
= 3.803 + 0.121* In>>= 2.851+ 0.003.-1.~ V .

.
224 1. ,

1. (-
) ?
2.
?
3. ?
?
?
4.
, ?

------------- 5.3 --------------


, -
( ) (autocorrelation, serial correlation),
, (
).
, ,
.
:
RSS
S 2 =------
-

;
..., (
)

;
...,
:
,

;
t- F-
, t- F--
, .
t- F-
/- F-

5. 225

. t- F -

.



,
(, , ,
), -
ei /
=
S
.
...,
yt = +... + + (, / 1,..., ,
.
,
(first order autoregressive process):
= p s i_l +Si , / = 2 ,..., n,
0 < p < 1, a Si9 i = 2 ,..., n,
, N(0, 2),
t_s s > 0.

.
(Cochrane-Orcutt).
(/ - 1)- ,
- 1 = 6 \ , - \ , + *1-\, + P i- \ .


z- :
1 - i-\ = \ ( - P x i- 1 , 1 ) + ... + (* * , - P x i-hp ) + (S, - p j_ i )
( autoregressive transformation).

j = 0lxjl + ... + @pxip +
> i = 2 ,..., n,

\=1-1-\>

Xn = Xi\ ~ P Xi- 1,1 T - - X',p=Xi p - P Xi-l,p


e\ = e i - p s t_
226 1. ,


ei - p e i_x =5i , i = 2 ,...,n ,
, s fl9 ..., 9
,
N(, <2).
,
. ,

. ,

.
.
-
. ,
, s'29 ...,
,
, -
.

5 > |*

I * .
1= 2

,..., ,
.

* = y t - r y , - j,

* = ~ r x i- i , i . > x iP = x i p - r x i-\,p >

= , - S t_i ,

* = eixn + + \ + *, i 2 , .

,
0 j , ..., 19..., .
-
5. 227

,
..,

.
, , s 1.

* = -,_1=1 - ,

* = +62*2 +... + ' +*, i = 2,..., ,
* = 0,(1 - ). * *,
*

, = ----- .
1-

5.3.1

(CONS) (MONEY) 1952 1956 . (. 5.14).

5.14
(CONS)
(MONEY) 1952 1956 ., .

, MONEY CONS , MONEY CONS

1952:1 159.3 214.6 1954:3 173.9 238.7


1952:2 161.2 217.7 1954:4 176.1 243.2
1952:3 162.8 219.6 1955:1 178.0 249.4
| 1952:4 164.6 227.2 1955:2 179.1 254.3
1953:1 165.9 230.9 1955:3 180.2 260.9
! 1953:2 167.9 233.3 1955:4 181.2 263.3
1953:3 168.3 234.1 1956:1 181.6 265.6
1953:4 169.7 232.3 1956:2 182.5 268.2
1954:1 170.5 233.7 1956:3 183.3 270.4
1954:2 171.6 236.5 1956:4 184.3 275.6


, = + /3,, + S f , i = 1, ..., 20,
y t CONS;
X/ MONEY, . 5.15.
228 1. ,

5.15
Y


/- -

1 -154.719 19.850 -7.794 0 .0 0 0 0

X 2.300 0.114 20.080 0 .0 0 0 0

R-squared 0.957 Durbin-Watson stat 0.328

1,
,

. . 5.13
, . 5.14 .
, ,
, -
.

CONS RESIDO'i

ll llll I 'm IIH

n il

- 1 1 1 1 1 t 1 1 1 1 1 1 I l I l M 1 J -i
180 190 52:1 52:3 52:3 53:1 53:3 54:3 55:1 55:3 56:1 56:3
MONEY ,

. 5.13 . 5.14


. . .5,
(, 2004, . 403),
d005 = 20: dL 005 = 1.20.
DW = 0.328 ,
0 : - 0 : > 0.

.
5. 229

, \

i =2

, ,
. 5.16.

5.16
*


/- -

1 -30.777 14.043 -2.192 0.0426

X* 2.795 0.609 4.593 0.0003

R-squared 0.554 Durbin-Watson stat 1.667


, ,
dv 005 = 1.40
d005, - 19. (
, ,
.)

( ).
. 5.15
, . 5.16 .

.

, ( 5 !)
Sp9

.
Sp
5 .
95%-
2.058 < < 2.542,
1.516 < < 4.074.
230 1. ,

CONS TRANSFORMED RESID02

MONEY_ TRANSFORMED ,

. 5.15 . 5.16





,
.
, ,

.

y t = a + j3xt / = 1,...,.
, et - {_ - Si9 i = 2,..., ,
,

* =* + /?* +*, / = 2 ,..., ,

* = { - - 1> X* = (,- - ), i = 2,..., , = (\ - ).


* J3 *

:
* =* +fix*9 / = 2 ,..., .
:
& - ry t_x = (\ - )+ (/ - ), i = 2 ,..., ,
*

= - :
1-
5. 231

y t = a + /3xi +r(yi_l - d - P x t_x), i = 2,


+1,
+1 ,
+1

+ = + f a n+x + ( - - ) .
+
:
+ j3x,
- + , v+,
+ /+{,
> 0. , + /, +
, , + +.


y t = a + Pxl:+if i = \,
=St , / = 2,..., n.

y \= y i - p y t-i> A = * i - P xi-1

y't = a ' + P x ' i + d i , i-2 ,...,n ,


y i = a { \- p ) + p y iA + p (x i - p x i_x) + Si, / = 2 ,..., .
, 0 < < 1,
:
, = { \ - ) +y t_x - ( 1 - ),_ + (*,. - ,_ + (1 - ) ) + S,=
= y,.i + (1 - )( + ,_, - ,_) + (, - ) + 8,,,

A y i = P A xi + ( p - - - ) + S,,

; =yt - ,_ , A x i-x i - - 1 < ( - 1) < 0 .


, ,
()
= + .
232 1. ,

(7 - 1) y t_, + (3xi_, (,_, > +


+ [),
A, = ( - ,_ . ,_ + p x t_,
(,_, < + fixt_,),
; - y t - .

, /3, . ,
,
.
, (3, .

:
.1=2.915, -0.126(j,_! -244.262-2.795 ).
,

, ..., ,
, ...,
.

, ..., ,...,

Sg .

Sg. (Newey, West)
, Views.
(Newey-
West estimate) .
(
, , (, , , 2005),
Views
.

(. 5.17).
5.17
Y


/- -

1 -154.719 19.850 -7.794 0 .0 0 0 0

2.300 0.114 20.080 0 .0 0 0 0


5. 233

. 5.18.

5.18


/- -

1 -154.719 32.729 -6.520 0 .0 0 0 0

2.300 0.139 16.449 0 .0 0 0 0

,
.

1.
?
2.
?
3. ,
?
4.
?
5.
,
?
6

_______ 6.1________




( )

! = \ + 2 + + epxip +i> i =
, , , ...,
xip, / = 1,..., , ,, 2, ..., ()
,

(
).


.
-
:
= +8,
^ = (, 2, ,) -
;
X ( /?)-
, >;
6. ... 235

= (, 2,..., ) - ;
= ( ,, & !,..., ) - ()
.
, -

() = ((), (2),..., {)) = (0,0,...,0)
( : () = 0)

Cov{) = (Cov(si, j )) = <J2I ,
/ ( ).

Cov(t, j) = E(i - E(si))(j - E( j ))
j.


:
, ,
. , , :

, ^- ;
,
^- ;

, F -
;

,
.


,
, , ...,x ip,
i = 1, ..., . ( )
s l9 82, ,
, y l 9y 2, .


236 1. ,

,
.

. ,
( )
.
,
.
, -

= 6 +
.
X ,
, ( )~,
:
= ()- .
:

{) = {{ )' {+)) =
= {{1 )+{{)~ ) = +((' ).
X ,

{{ ) = (. {) = ,

{) = \ .. \
, , (stochastic)
, .. X
, X
. xij9 i = 1, ..., , j = 1,
,
. (()~1 )) * 0,
{6) ^ 9 . ,
,
.
,

,
, , (
).
.
6. ... 237

()
sk () , , ..., xip
/
s,, j, ,

2 > 0. ,
st ~ i.i.d. N(0, 2).
:
(( ' ) = (( 1 )() = 0,
.
( )
.
X;
(
).
,
X X.

, :

\ ~ , 2( ).
,
N(0, 2( )~1) X.
.
/- :
ej \X~N(ej , * 2(XTXyJ),

(XTX)J '- () ',

X ~ N ( 0,1).

RSS
------ ----- , S 2 = , RSS -
< -
, - { - )
:
238 1. ,

, ^- 0 : 6*

s J ( X TX)~J
, 0 ,
/- /- ( - )

11 ~ t( n - ).
X.
, X,
/- 0 : - *
t(n - ) .

F -
.

.
'
8 1X ~ N(0, 2/), / ( ).
:
X,. = (, , ...,1) -
/'- ;

.

() , , ..., 1
i ;
st ,
~ i.i.d., (;) = , D (,) = 2 > 0 E(ef) = //4 < ;
E(xjxf) = Q, , (1 /n)(Q1 + ... +
+ Q) -> Q , Q ;
( xik xis) < i,j, k , l , s ;
(1 /n)(XlXi + ... + xnx l ) = (1 In) XrX -> Q ft .

, .
- ()
6. ... 239

S2, t- F -
, .
,
,
, , {Hamilton, 1994).

{) ;
S 2, tn,F n S2, /, F, .
,
, :

4 S 2 - 2) - * N ( 0 , / 4 - -4);

qFn > 2()>


.
, ..
, ,
, .
:
{) N{0, <j2Q~x!n), {) * N{6, 2{^)~1) (
);
S 2~ N (a 2,( { iA- (4)/);
/ , 1);
qFn~x2iq)-

, t{n - )
^- ( N ( , 1)) F(q, - )
/- ( %2{q) qF),
( ,
).
, ,

t Fn.

, ,
, ,\ ~ i.i.d.
, :
240 1. ,

X
- N{, o 2V);
V
( ).
V ,
V~l.
( )- 9 V~l = . 9
s
* = Ps.
(*) = 0 ( X)

Cov{s \) = E(s*s*T|) = (()\) = ( \) = P a 2VPT.
V = (-'1= ( \
Cov(e* |X) = P a 2VPT = 2{)~= 21.

-+,
:
= +, =*+,
=, * = , = .

\~0,21),
,
. , , ,

* = *0+ .
,
* = (**'* *= ( X TPTP X y lX TPTPY = (X TV-lX y 1X TV~'y
, .. (*) - , (
X)
Cov{0* \) = 2('* 1= cr2(X TV-lX y l.

Z Z w ik ( i - \ x i i - - d P x iP )( - & ix ki - - ).
/ = 1 = 1

wik =v\kl) ~.
6. ... 241


(GLS generalized least squares). *
( GLS estimator),
GLS:
=&GLS = (X TV~lX)~l X TV~ly.
, * = X* + '
, t- f -.

/ 6.1.1. V , V = diag(/2, ..., 2),


ht, ..., h > 0 , hk ( ...,
),
V 1 = diagO//?,2, ..., 1/2),
- diag(1 /1? ..., 1//?).

*=1 l hi 4 =Xij/hi > j = 1= 1,..., ,

1-\*------
I ^ = ?
- - 0 '/ - * ! *<)>
/=1
.. ,
.

.
5.2.1.
= 2 , 9 / = 1, ,
h \ =/2-

J 6.1.2. 5.3

1- :
p&i\ ^ 2 ,..., ,
\ \ < 1 9 St91 = 2 , ..., ,
,
N(, 2), St st_s9 s > 0. ,
, > 5 (\) = ,

D(ei) = - 2
1- 1
242 1. ,

' 1 2 ... - 1
1 .. - 2
v = 1 1 .. - 3

, - - 2 1 >

1 - 0 0 0 "
- (1 + 2) - 0 0
0 - (IV) 0 0
- . . 1
1- 2
0 0 0 . (IV) -
0 J 0 - 1
( ),

1- 2 0 0 . . . 0 0

- 1 0 .. . 0 0
0 - 1 .. . 0 0
=

0 0 0 .. . 1 0
0 0 0 ..
- 1

\ =
= 1~
^ ' \ |, - -/-) xlj
, = ,- ,- 1 X,j = x tj -px,-hJ, j = !.

(Prais-Winsten transformation). :
,
( ),
, 5.3
.

V , ,
, V = V(P), (3 ,
.

, GLS- * = (X TV~lX)~lX TV~ly -
6. ... 243

V= V(J30) (<2( )
) V(/3n), , ,.

(feasible GLS). ,
,
. 4.5 :

/=2
..., ,
.
, ,
.
, ,
,
.

1. ?
2.
?
3.
,
/- F-?
4. ,
X -
N(, 2 V) F, ?
5. ?
?
6.
,
, ?

_________ 6.2 --------------


, , ' , 6.1,

E(ei \ X) = 0, / = 1,..., ,

E(si\xkj) = 0 j = 1, ...,/? in .
244 1. ,


(,) =

Cov(e,, ]) = ((*, - (, ))(xkj - (xkj))) = E{st(xkj - E(xkj )) = E(s, xk]) = 0


(, , E(xkj)
).
, z-
xkj9 , \
. , , i- -
.

,
. , ,

.

6.2.1

:
DGP\ yi = a + /3xi + si9 S i-id d . N{0,1), / = 1,...,100,
a - 10 , /3=2,
X,. =,.-0 .9 sM, i = 2, ..., 100 ,
Corr(xi9 = 0.743.
, y i9 xi9 i = 2, ..., 100,
.
. = + /?, + si
. , . 6.1
6.1
Y_FIXED
(Method: Least Squares; Sample (adjusted): 2 100


- -

1 10.13984 0.069148 146.6398 0.0000

X 2.553515 0.054971 46.45184 0.0000

/3 /? = 2.553,
.
6. ... 245

2, ..., 10> 499


. . . , ^}? = 2, ..., 500,
,
, -
[ {),..., } :

\) = a + p Xi+ s\k), i = 2 ,..., 100.


= 2,..., 500, y f \ xi9 i - 2,..., 100,
^ = + (3xt + ^\ /?(*}.
(2\ ..., (500) /?(2), ..., /?(500).
. 6.1
/?(2), ..., jff(500\

Series: SLOPE
Sample 2 500
Observations 499

Mean 1.999663
Median 1.994058
Maximum 2.238951
Minimum 1.740510
Std. Dev. 0.083992
Skewness 0.161028
Kurtosis 2.895625
Jarque-Bera 2.383013
Probability 0.303763

^ i i i l i i i I i i i l i i i l t i i I i i

1.8 1.9 2.0 2.1 2.2 P


. 6.1


(5 .
.
>&ioo} = 2, ..., 500,
..., X|qq|, {^,---, )\
:
\) = + /3 ?> + \), / = 2 ,..., 100.


\ s (2k), zloJ}.
246 1. ,

{4*\ > *100} ^ 100}


= 2, 500
y f ] = + /{) + s(k) *(), {).
*(2\ ..., *(> /3*{2\ ...,
. 6.2
...,/?*(500).

OU
Series: SLOPE RANDOM
Sample 2500
Observations 499
60
Mean 2.552114
Median 2.551333
Maximum 2.588107
Minimum 2.530200
40 -
Std. Dev. 0.007346
Skewness 0.754039
Kurtosis 4.608878
20 -
Jarque-Bera 101.1054

v I

2.53
i i
1 Hi
i

2.54
I i i i

2.55
l i i i

2.56
l i i
1 lL i
i

2.57
i i

2.58
l i i i l

2.59 P*
Probability 0.000000

. 6.2

j3*(k\ 2.552114,
/? = 2,
,
/? = 2
.
: xt ,
xi9 ,
,
. 0
, , ,
.
^, xi9 i = 2,..., 100, ,
. 6.3. Linear () ,
, .. = 10.13984 + 2.553515,
a Y THEOR >>= 10 +2. . 6.3,

, xt {
= 10 +2
6. ... 247


20
18 Y_THEOR
16 Linear ()

14

12

10
8

6
4

2
0
-4 -2 0 2
%
. 6.3

, ,

(missing variables). . 3


.
D GP:^, = + p z t + ,- + i = \,...,n ,
z , ) ~ i.i.d.,
(') = 0, D(Si) = 2, Cov(jc = Cov(Zj, Sj) = 0. ,
z ,

, /-
,
.
. 3,

.
,

S M : ,,= + P z t,+ ,,

, = yxi + st.
248 1. ,

Cov(z,, ,) = Cov(z,. ,yxi + i ) = r v(z,, xi) * 0,


z.
(errors-in-
variables models).
D G P: y i, = + /?z,- + ,, / = 1,__ 100,
z,
:
(,.) = 0, D(ut) = 2, E(ui |z ,) = 0,

E {y \z t) = a + p z t.
, z, ,
z,
* ,= z ,+ v
v, . :
. = 0,(,.) = 2;
ut v, :
vt zt.

zt { xt - vt zf .
:
y i =a + p x i + si,
, = , - /?v,
( ,j) = Cov(zj + v,, (. - /? vi) = -/? 2.
> 0, xt st ; < 0, xt
.
,
, , ..

. /?:

- T f a i - )

/ =1
^. :
6. ... 249

xi - + i - e){xt - )
1= 1 / =1
=
=+

( ,- ) 2

, /?
, 2 = 0, .. zt . -
] ,
.
.
{simultaneous equations).

, = + Yt + Gt ,
, ;
Yt ;
(0 < < 1), -
( ).

,
.

t + Yt + 81
Y' =Ct +I,

(structural form o f simultaneous equations),
,
, ,
( Yt , ,).
,
,
.
250 1. ,

, , /
{reducedform o f simultaneous equations) :
a ft 1
, --------1-------- 1, -------- ,,
' \- \- ' \-
v 1 1
Y. --------1--------1, -------- ..
' 1- 1 - ' 1-
, Y,
,
.
, ,
I, 1. Yt 1.
,
. , ,
Yt , 2, ..
Yt 1, , , ...
, 2, , ...
, Y,

1+ + 2+ + ... = 1
1-
,
, / +, ... _= -
+ ?2 +*
1-
/,
.
, et ~ i.i.d., ({) - 0, D(Sj) = 2 > 0 t
/, , .
:
1 2
Cov(Yt ,,) = Cov(s, ,,) = ,

, Yt -
.
( )
, :
Cov(Y )
\\ = + >*
-00 D(Yt )

6. ... 251

- ?
2
= + ( \ - ) - -
0 (1 ,)+ * :
<2 > 0 0 < < 1,
.

^ 6.2.1. \\ /} ,


D(It) t .
.

J3 ,
.

.
:
, = + /311+ 61,

a = al(\-p), p = p/(l- p) , t = s t/(l-P), (,) = ,


D{,) = al =<t2J( \ - P ) 2.
,
/3 ~.
:
= / ( \ + ), = /(1 + ), ] = |/(1 + )2.
,
.
. , ,
.
, :
Y, = +8 1 , +
= = /(1 ), = 1 /(1 -/? ).
,
8 ~ .
:
=(8-1)/8, = / 8 , <2 - a j / S 2.
252 1. ,

:
,
?
, , 2
, ,
r / S = 5 / ( l + fi), ( S - \ ) / S = f i / ( l + fi), crl / S2 = a l / ( \ + p ) 2.
,
,
.
,
.
,
,
. ,

.


,
,

y t = + fiXj + si9 t ~ , E(st) = 0, D(st) = 2, / = 1,..., .

Y j {yi - a - p x i) = 0
/ =1
5
ZCv/ - - / ? * , ) * , = 0
/= 1

= (19 ..., )9 { = y t -
- - , i- , 1 = (1,..., 1) = ( },..., ).
,
1 1

- 1 > = 0, - = 0
(=1 /-1

Cov(si, 1) = 0, Cov(s[, ,) = 0.
,
E(st) = 0 E(si ,) = 0.
6. ... 253

1 J
Cov(Sj, Xj) 0, lim Y e,:t, * 0 =0
, =1 ,=1
(; *,) = 0.
- zf,
Cov(j, z,) = E(jZt) = 0,

, ..

' E( y t - a - f i x , ) z l =0.
i=1

,
,
* /?*.

X & - " - ' >) = Z O '/ - ' ~ =


i=i /=i
/3*:

( * , -x X z, - z )
/ =1
:

(/? , - p X + t - f X z , , - 2 ) - f ) ( Z (. - Z )

z0* = ----- ;----------------------------= P + - J 1T--------------------


Z *)(Zi ~z ) Z (*l ~ * XZ- f )
/=1 W1=1

p lim - ](,- - ?)(z,. - z) = C ov(^,z;) = 0,


- 00 / = 1

1 _ _
p lim - V (xt - x)(z,- - z) = Cov(x,.,z,.),
-> / =1

, p lim /?*=/?, :
>00
Cov(xh Zj) 0.
z, :

Cov(eib Zj) = 0, Cov(Xj, z,) * ,


254 I. ,


(instrumental variable, instrument).
/3 , ,
, ,.
(exogenous variable) ,
,- = + /?, + (. ,
(endogenous
variable) () ,
, ()
, ,
, ( )
IV: a ,v, f3!V ( 1, /31), IV Instrumental
Variables ( ).
(IV method instru
mental variables method).
, :
|C f = a + /3Yt +,,
\y,=c,+it.
:

Cov(Y,,st) = 1s 0, Yt .

Cov(It, s,) = 0 (
), /, .
, :
1
Cov( Y.,I.) = Covl hi I + -^ t, I ,
[ l - p 1-/3 ' 1- p ' ' \~
-Mi,)* o,
I,
/3.

( C ,- C ) ( /,- /)
t =\
P lV -~ n
Y^(Yt - Y ) ( I t - I )
r=l

/F - f3
.
, = + f3Yt + , :

Cov(C, ,1,) = Cov(a,, I ,) + Cov(Jt ,1,) + Cov(s, ,1,).


6. ... 255


Cov(Ct,It) = /3Cov(Yt,I t),
:
Cov(C,,/,)
Cov(Yt,It)
/3 ,
:

1 ( ,- ) (/ ,-7) 2(,-)(/,-7)
- 11=1________
P lV ~ _ _
=!=!_________
_
(,-/,-7)
/-1 /=1

6.2.2
. 6.2 (Economic Report of the President, 2000,
Appendix ) (. . 1996 .,
)
:
CONS {personal consumption expendi
tures);
Y {disposable personal income),

1= Y - CONS.
6.2

, . . 1996 .

Y CONS / Y CONS /

1959 9.068 8.213 0.855 1979 15.942 14.073 1.869


1960 9.111 8.267 0.844 1980 15.944 13.918 2.026
1961 9.260 8.298 0.962 1981 16.154 13.973 2.181
1962 9.561 8.574 0.987 1982 16.250 14.038 2 .2 1 2

1963 9.779 8.799 0.980 1983 16.564 14.644 1.920


1964 10.342 9.197 1.145 1984 17.687 15.303 2.384
1965 10.842 9.655 1.187 1985 18.120 15.924 2.196
1966 11.288 10.088 1 .2 0 0 1986 18.536 16.448 2.088
1967 11.641 10.278 1.363 1987 18.790 16.867 1.923
256 1. ,

. 6.2

Y CONS I CONS /

1968 12.055 10.755 1.300 1988 19.448 17.397 2.051


1969 12.322 11.055 1.267 1989 19.746 17.682 2.064
1970 1 2 .6 8 8 11.180 1.508 1990 19.967 17.818 2.149
1971 13.044 11.429 1.615 1991 19.892 17.653 2.239
1972 13.492 11.972 1.520 1992 20.359 18.025 2.334
1973 14.269 12.428 1.841 1993 20.354 18.372 1.982
1974 14.099 12.259 1.840 1994 20.675 18.878 1.797
1975 14.236 12.414 1.822 1995 21.032 19.272 1.760
1976 14.653 12.960 1.693 1996 21.385 19.727 1.658
1977 15.010 13.364 1.646 1997 21.954 20.272 1.682
1978 15.627 13.842 1.785 1998 22.636 21.060 1.576


CONSt =a + /3Yt +st
, . 6.4.
. 1986 .
,
10 . .
,
1959 1985 .
, . 6.5.
. . 6.3.

. 6.4 . 6.5
6. ... 257

6.3
CONS
(Method: Least Squares; Sample: 1959 1985; Included observations: 27)


/- -

546.7669 105.9135 5.162389 0 .0 0 0 0

Y 0.841496 0.007803 107.8374 0 .0 0 0 0

, :
CONS, = 4012.195 + 4.949 1, +
Yt =4012.195 + 5.949/, +,,
:

= = 4012.195, = 4.949, = 5.949.



/3:

= - ^ = 0.832, = ^ = 674.5;
1+ 1+ 0

/9 = 0.831891 = 674.4849 .
,
:

= ^ - = 0.831891, = 4 = 674.4849.
8 8
, ,
, ,
.
(ILS indirect least
squares).
:

, ,
.

I, : = 1737.04, / = 1561.04, Y = 13298.07,
258 1. ,

(,-/,-/)
, = -------------------- = 0.831891.

t =1

/? ,
,

Yt .

Yt
It (
). S,

Yt =y + S I t ,
Yt
st.
Yt :

Yt =Yt +( Y' -Yt),


.

CONS t = + fiYt + Gt ,
Yt
.
CONSt =
= + j3Yt + et
(2SLS two-stage least squares). a 2SLs 02sls>
,

2 ( C O N S t oc2 s l s ~ P i s L s ^ t ) ~ 0 ^ ( C O N S t & 2sls P i s L s ^ t ) =


t =1 t =i
.. /F -.
2SLS
6.2.2 , . 6.4.
Views ,
.
6.2.2 . 6.5.
6. ... 259

6.4
CONS
(Method: Least Squares; Sample: 1959 1985; Included observations: 27)


/- -

674.4849 667.4401 1.010555 0.3219


YF 0.831891 0.049231 16.89762 0 .0 0 0 0

6.5

(Estimation Method: Two-Stage Least Squares; Sample: 1959 1985; Included observations: 27;
Total system (balanced) observations 27; Instruments: I C)


/- -

(1) 674.4849 112.2025 6.011319 0 .0 0 0 0

0 2 ) 0.831891 0.008276 100.5160 0 .0 0 0 0

J 6.2.2.
CONSt = + j3Yt + st
:
= 546.7669 / = 0.841496,
. ,
, ,
(
-
),
= 674.4849 /? = 0.831891,
.

J 6.2.3.

:


,
;


,
;
.
260 1. ,

, ,
,
,

. , /F -
, /^-.

. - ,
, ,
.
, ,

(weak instruments), ..
, /F -
, OLS-.


,
, .

1.

?
2.
?
3.
, ,
, ?
4.
?
5. , ?
?
6.
?
7. ?
,

1
.

1.

Y J(xi - x ) ( y i - y )
, /3 =
----------------.
* -* )2
/=1

2.
:
a) (0,1), (1,3), (3,3), (4,1);
b) (-2, 4), (-1,1), (0,0), (1,1), (2, 4);
c) (0,1), (1,0), (3,4), (4,3);
d) (0,3), (1,4), (3,0), (4,1);
e) (1,1), (2,3), (3,2), (4, 4);
()

.
.

. ?

3. ,
-
, TSS, ESS, RSS R2.
2 2 2
. , , - .
262 1. ,

4.
s
, /3 =

5.
,
, .

6.
= { + 2 :
= a i+ fiixi + > = 2 + Pixi + = a + fiXi + Sj, = 1 , .
, = + 2, - \ + -

7.
2
: r-y = R , = , 2
2
2 = * = ,2, ,
2 2

R 2, .

8.
, ,
.

9.
yt = + {+ &i9 i = 1,..., 4,
(1, 1), (2, 3), (3, 2), (4, 4). (
.)
,
xt = + Syt + wz.
,
xi9 ^, { * 9
,- .

10. Eviews
2
EViews.
.
:
;
, ;
;
;
... 263

. = + /?, + t
( Forecast);
, .
( ). ,
scatter with regression ,
;

*, ,-
;
, , *;
, * .
.
:
.

11. EViews:

ta b l_ l
1968 . 1969 .,
. 1.1 (. 1).
,
.
, 10.
,
BEL,
. ? (
BELt = + SBELt + .)
, RSS. TSS, ESS, R 2.
, R 2 = ~
2
= 2^ .
:
.

12.
, 11,
. :

;
,
/3 ,
- + /?0. RSS0
()
;
264 1. ,

/?0, /,
.
/?0, /?0,
RSSX. RSSX RSS0;
RSSX < RSS0, /?2,
( ),
, /?2. /?2,
/?0;

.
,
,
.

13.

, 1.3.1 1.3.5.
: tabl_2, tabl_7, tabl_8, tabl_9,
tabl_10.

.

14.
EViews ,
1.3.4 (tabl_9)
:
( );
( );
,

;
;
,

;
, .

15.
, ,
1.3, ,
capm.wfl : http://www.econ.kuleuven.be/gme/.
camp . TSS, ESS, RSS,
... 265

, . ,
TSS = ESS + RSS. -
.
Equation
. ,
.
, .

16.

EViews, 1.3.1 ( tabl_7),
,
(log-log ).

( TSS ).
R2, ,
.

17.
, 1.4
( . 1.5).

-1.

( )
4 X Y.
X 1 2 3 4
1 3 2 4

():
;

XnY;
X ,
;
R2, 3
; , 3
.
. ,
J3 . ,
0. ,
. ,
266 1. ,

. , .
:
) X Y;
) .

-2.

,
, = +, / =1,
- .
.
.

-3.
. -1
( 18 ,
3.7854118 ) (
, , 1000 )
1935 1945 .
?
. ,

.

-4.
INF,
Michaelis-Menton
0.514 UNJOB
INF = ----------------------,
- 3.057 + UNJOB
R2.
,
Y= + ,\ Y= 1/INF, = \ / UNJOB.

2
.

18.

. 2.1 (. 2)
( )
1923 1939 ., Views tab2_l
, 2.1.
... 267


, lg , a lg DPI.
, , 2.1,
(
0.434294).
, .

19.
EViews, ,
,
nz.

= + jSz + yz2;

, = + /3zj + yz2t + eit i=l,...,n.
.
,
.

20.

. -2
5 EViews
tabc_4 .
,
cosat, sin at. ,

X.
,
.
. ,
X cos at, sin at
,
0.

21.

EViews, ,
/3 yi= + + st, (%) =0,
>(;)= 2, /= 1 , \
1) ;

2) , >0.
/=1
268 1. ,

:
) /?;
) {, ..., ;
) z\l\ i = 1, ,
, ,
^1), i = 1, ..., , : = + fixt + {)
/? ), / = 1, ..., ,
(1) /^);
) ) \2\ / = 1, ..., ,
^2), / = 1,..., , : \2) = + J3x{+
(2) /^2);
)
{\ ..., ..., /7*);
) ..., ^
f i l\ ..., /?*).
.

22.

EViews, 90%-, 95%- 99%-
,
, tabl_7 (
), tabl_8 ( ) tabl_10 ( ).
. Equation,

: .
,
EViews (. Function Reference Help).

23.

EViews,
0.90, 0.95 0.99 ,
, .
. , ,
.

24.

1. EViews Excel, 95%-

,
tab2_l.
... 269

2.
,
0.95.
.
.

-5.

,

.

-6.

, , R\9R\ R2X2
,
- = + pxt + sXi9 i= 1,..., ,
yt = + yzt + i= 1,..., ,
yt = a + pxt + yzt + n, i> *= U > ^
, :
z
, R]2 = Rf + R\\
R\2 , Rf + R2.
. EViews
:
) 3 ,
(0, 1): xl92, z x\
) 4 ,
(0, 1): el9 el9 , s4.
) z2 = zx + 0.1 s x;
)
: Vj v2;
) ^ = + 2 + 2, w = zx + z2 + , v = vx + v2 + 4;
)
2


2 ( ,

);
270 1. ,

)
w zx z2
w
z x w
z2 ( ,

);
)
v v{ v2
v
VjHb v
v2 ( ,

).
,
.

-7.
.
1. ,


, .
2.
.
.
,
,

,

. ,
.

-8.
.
, ( , , , 2005, . 63 64).

-9.

,
,
,
... 271

(,n - p ) S 2 _ RSS

( ) ,
( - ) ,
.
. . ( , , , 2005, . 66).

-10.



,
,
,

/- ( - ) .
. . (, , , 2005, . 7071).

25.
t-
,
, tab l_ l,
tabl_2, tabl_7, tabl_8, tabl_10, EViews Excel.
:
EViews Equation;
Excel : > >
.
.

26.
t-
,
log-log
( ) , tab2_l,
EViews Excel.
272 1. ,

:
EViews Equations
Excel : -
.
.

27. t-
1959 1980 .

( 1982 .), . 4.4 . 4.
tab4_4.
:
,
0.9, ;
,
0.9, ,
0.9.
. ,
@qtdist(p, v). 5%-
-2.074, 5%- -1.717.

28. , t-,

/-, /

, tab2_l.
.
,
.

29.
. -7
1949 1966 . ( , 1959 .):
Y ;
2 ;
3 ;
4 .
1949 1959 .

,- =&1 +02 + + 4;4 + i = 1 , 1 1 , = 1.



2, 3 JC4.
... 273

/-.
.
. tab3_7.
.
?
F - ,

. ,
.
F - ,
EViews,
.

30. .
F-
. -3 24
(Erie, , ),
, (PRICE)
:
TAXES ( + + );
BATH ;
LOTSIZE ;
LIVSP ;
GAR ;
ROOMS ;
BED ;
AGE ;
FIRE .

.
F-,
.
F -,
.
. tab _5.
Equation,
: View - Coefficient Tests Wald Coefficient Restrictions.
,
, .
:
, ,
;
, .
F -.
274 1. ,

31. .
F-
30
(. . -4) ,
. :
;
Z1 ;
Z2 ;
Z3 ;
Z4 ;
Z5 ;
Z6 .

. 1
5 , 1 , 5 .
1 2 ,
3, 4, 5 .
7 .
. , .

32.
1.
(. -2),
( 12) ,
95%-
. ,
,
0.95.
2.
( 12)
, ( 12) .
3.

, , tab2_l.

33. F- t-



1960 1972 ., . 3.10 (. 3).
:
Y ( );
;
L - ().
... 275


,
.
.
,
:
In 7= In + \ + /31L.
0 : + (i > 1.
. Equation
: View > Coefficient Tests Wald Coefficient Restrictions.

. F -
/-,
.
, () -
-, F -. -
, @ctdist(/, - ).

34.

F -
, 5%-

:
= 3, 4, 10, 20, 30, 40, 60, 120, 500.
\
. F -
:
;
@qfdist(x, vl, v2) EViews.

35.


31 ( ):
,
;
, , ,
Z1
Z3. ,
;
, :
)
(, , )?
276 1. ,

)
Z2, Z4, Z5 Z6
(, , )?

. ) ).

36.
29 ( ) ,
1949 1959 .
?
?
.
2.

.
?


2, 3, 4.

37.
EViews, , 3.3.5.

.

.
.
yt = 1 + 5 + + 0.5 + xt2=\ + 0.1/ + 0.1%, = 2 4-0.5/+ 0.1%,
,4 = 3 + / + 0.1^4, , , t4 ~ i.i.d N(0, 1), /= 1,..., 500.

: 500, 100, 50, 25.

,
, , 4.

38.
,

29, 30 31, :
;
;
.
... 277

.
Equation.
, ,
, .

39. , , ,


:
,
1.3.1;
,
2.1.

40.

, 3.3.6,
, ,
EViews. , ,
, 3.3.6.

-11.
= 6. ,
, ,
, .
1) 2 = 0; 2) 2 = 0, #3 = 0; 3) 2 = 1, 3 = 0; 4) 2 = 1, 5 = -1; 5)4 = 6 = 0;
6)4 = 6; 7) 04 + 06 = 1; )2 = 4 = = 0; 9) 2 = 4 = 6 = 1;
1)02 = 04 = 06; II) 2 4 = 6; 12) 62 - 64 = 4.
.
.

-12.



=+ s = 3 :
'3 ' "3 0 1"
<9= 0 , S 2( X TX) - x = 0 1 0
0 1,
278 1. ,


# 0 : - 2 + 03?
.
,40= .

-13.

, yt = + si9
i - 1,..., , =.
.
.

-14.

4
. 24, 44
. 0.25,
0.20.
?

. F = ^ ^ ^

,
0 ,
, F F - ( - 1) ( - )
.

-15. ,

,
.
. . ( .., 2001, . 2, . 2.4.)

-16. ,
,
{ = + ^ + + dpxip + ei9 i = 1, ..., , xiX= 1 ?~ i.i.d. N(0, 2),
:

-2 1 ,...,6 > <2) RSS


------------------ -------= In + \ + \2,

.
... 279

. (19 ..., 9 2)
, 19 ..., 9 29
.

4
.

41.
,

EViews, 4
, . 4.3 . 4.
. ,
Equation. .
.

42.
,
.
24 (. -3),
, , .
.
,
. Series, -
.

43.
,

(. -4),
, ,
.
.
Z1 ,
. -
.

44.
,


:
(tabl_l);
( tabl__7);
280 1. ,

( tabl_8);
( );
( tab2_l);
( tab _4);
1949 1966 . (. -7);

1959 1985 . ( tab4_4);
27 , . 5.17
. 5 ( tab5_17);
, -
. 5.18 . 5 ( tab5__18).
. :
, .

45.
,

42 ,
, ,
.
. ,
.
?

46.
,

43 ,
, ,
.
. . 45.

47.
,

44 ,
.
. , 45,
.

48. :
,
EViews,
1949 1966 . (. -7)
.
... 281

. Y
2, , 4 .
. ,
CUSUM. ,
,
(1960 .).
- D 1, 1 1949 1959 .,
- D2, 1 1960 1966 .
-
X I, , . .
, D2, 2 , , , D2 2, D2 ,
D2 . ?

. .

49. :
,
EViews,
1959
1985 . ( tab4_4)
. , .
. .

. .

-17.

-3 (
. -1)
,
, .
.
.

5
.

50.
EViews,
(. 5.16), .
.
.
. ,
.
282 1. ,

DPI, 4 .
, .
DPI.
( DPI).
.
IIIII
I IV . .

51. -
,
15 , , Z
X , . -5.
Z , X .
.

.
15
.
-,

.
- .

52. -, -
51 , 5.1.2.
. Equation EViews.

53. -

( )
5.1.2 ( tab5_3) , Equation
EViews:
0 19 ,
,
F - 0 : = 2 = .
F -,
0.05;
'0 : = 2 = \
0 H q : = 2 = 3, = 2 ~ -
.
. ,
30 , ,
, 5.1.2, 6 ,
X , .
... 283

Equation
, 0,
, 6 (
), .
, 15
,
() 3 - ,
, 3 X: 2-, 4- 6-.
'0 Equation
Views, Coefficient Tests.
54. :
,

EViews, ,
27
( tab5_17).
. Y
X , .
X ,
, X.
,
. ,
.
. .
Y X,
EViews ( WLS).
( ).
,
.
WLS (
). R 2
. ?
log-level .
, .
. .

55. :

EViews, ,

1952 1956 . (. 5.18).
.
.
. ^4(1) ,
284 1. ,

, .
,
. , .

, !(1).
, .
, .
AR(1) 1952 1955 .

1956 . ( 1956 . ).
1956 .

56.
-

. -6
( ) 1951 . 1953 .
,
1951 . 1953 . .
- , ,
1953 .
. -.
.
1951 . 1953 .
, ,
; --
.
, 1953 .
.

57.

54 , .

58.
.
55 ,
.

-18. -

- (
) , 4.2.1 (
1960 1985 .).
... 285

. 3
, 6
4 .

-19.
- (
) , 5.1.1 ( ).
.
3 .

-20.
, 5.1.5. ,
-.
.
. ,
.
, , .
^ = + 16
17 ^ = + J3xt + yD \l t + ei9
D 17 -, 1 17- 0
16 .

-21.


, ?
. ,
.

-22.


(. -1) , , , .
.
.

6
.

59.
55 .
. .

.
286 1. ,

60.
1.
, .
DGP: ^ = 10+ 2Z; + 2eyi9
z= 10 + 10*@md,
= z,. + 2sxi,
sxi, syi ~ i.i.d. N(0, 1), i = 1 , 1 0 0 .
,
. (xi9yt) (zi9).
2. ,
, RESET ?

-23.

1. ,
DGP: ^ = 10+ 2, + 2ei9 = 10*@md,

2. ,
OLS { = + /+ si9
.
3. ,
.

OLS , , ,
EViews AR( 1) ,
.
4. ,
? ( , ?)
. . 4
.
:

1* = 1 * ~~ 1-1 X* = , - , _ t = 2,...,n.
y i, = + ftx{ +
ei9 : et = pet_x + v,.

.
... 287

-24.

, 6.2.1.
DGP: yf = + (5{+ si9 ~ lid. N( 0, 1), i = 1,..., 100,
- 10, = 2,
xt = i - 0.9_1# i = 2,..., 100.

-25. 2SLS
, 6.2.2.

-1

PIVO DET PIVO DET


1935 10.3 55.7 1940 12.4 47.0
1936 1 1 .8 57.1 1941 12.3 45.3
1937 13.3 54.4 1942 14.1 40.4
1938 12.9 51.0 1943 15.8 40.4
1939 12.3 48.0 1944 18.0 39.8
1945 18.6 38.3

-2
,

, , ,

1935:01 48.9 1936:01 48.3 1937:01 42.4


1935:02 43.4 1936:02 47.1 1937:02 41.4
1935:03 43.8 1936:03 52.4 1937:03 49.0
1935:04 50.8 1936:04 55.3 1937:04 50.8
1935:05 67.6 1936:05 64.7 1937:05 65.8
1935:06 83.7 1936:06 79.5 1937:06 85.9
1935:07 82.7 1936:07 62.6 1937:07 70.6
1935:08 60.8 1936:08 51.3 1937:08 55.8
1935:09 55.4 1936:09 51.0 1937:09 49.1
1935:10 48.4 1936:10 54.0 1937:10 45.7
1935:11 37.4 1936:11 45.2 1937:11 43.8
1935:12 41.0 1936:12 44.9 1937:12 46.7
288 1. ,

-3

AGE BATH BED FIRE GAR LIVSP LOTSIZE ROOMS PRICE TAXES

42 1 4 0 1 0.998 3.472 1 25.9 4.918


62 1 4 0 2 1.500 3.531 1 29.5 5.021
40 1 3 0 1 1.175 2.275 6 27.9 4.543
54 1 3 0 1 1.232 4.050 6 25.90 4.557
42 1 3 0 1 1 .1 2 1 4.455 6 29.9 5.060
56 1 3 0 1 0.988 4.455 6 29.9 3.891
51 1 3 1 1 1.240 5.850 1 30.9 5.898
32 1 3 0 0 1.501 9.520 6 28.9 5.604
32 1 3 0 2 1.225 6.435 6 35.9 5.828
30 1 3 0 1 1.552 4.988 6 31.5 5.300
30 1 2 0 1 0.975 5.520 5 31.0 6.271
32 1 3 0 2 1 .1 2 1 6 .6 6 6 6 30.9 5.959
46 1 2 1 0 1 .0 2 0 5.000 5 30.0 5.050
50 1.5 4 0 2 1.664 5.150 8 36.9 8.246
22 1.5 3 1 1.5 1.488 6.902 7 41.9 6.697
17 1.5 3 0 1 1.376 7.102 6 40.5 7.784
23 1 3 0 1.5 1.500 7.800 7 43.9 9.038
40 1 3 1 2 1.256 5.520 6 37.9 5.989
22 1 3 0 1 1.690 5.000 6 37.9 7.542
50 1 4 1 2 1.820 9.890 8 44.5 8.795
44 1.5 3 0 1 1.652 6.727 6 37.9 6.083
48 1.5 4 1 2 1.777 9.150 8 38.9 8.361
3 1 3 0 2 1.504 8 .0 0 0 7 36.9 8.140
31 1 4 0 1.5 1.831 7.326 8 45.8 9.142

-4

Y Z1 Z2 Z3 Z4 Z5 Z6
43 51 30 39 61 92 45
63 64 51 54 63 13 47
71 70 68 69 76 86 48
61 63 45 47 54 84 35
81 78 56 66 71 83 47
43 55 49 44 54 49 34
58 67 42 56 66 68 35
... 289

. -4

Z1 Z2 Z3 Z4 Z5 Z6

71 75 50 55 70 66 41
72 82 72 67 71 83 31
67 61 45 47 62 80 41
64 53 53 58 58 67 34
67 60 47 39 59 74 41
69 62 57 42 55 63 25
68 83 83 45 59 77 35
77 77 54 72 79 77 46
81 90 50 72 60 54 36
74 85 64 69 79 79 63
65 60 65 75 55 80 60
65 70 46 57 75 85 46
50 58 68 54 64 78 52
50 40 33 34 43 64 33
64 61 52 62 66 80 41
53 66 52 50 63 80 37
40 37 42 58 50 57 49
63 54 42 48 66 75 33
66 77 66 63 88 76 72
78 75 58 74 80 78 49
48 57 44 45 51 83 38
85 85 71 71 77 74 55
82 82 39 59 64 78 39

-5
15

1 X Z / X Z

1 1 .0 0 0 0 0 0 1.257072 9 9.000000 4.316424


2 2 .0 0 0 0 0 0 1.811984 10 1 0 .0 0 0 0 0 5.322512
3 3.000000 3.641244 11 1 1 .0 0 0 0 0 1.804023
4 4.000000 4.401497 12 1 2 .0 0 0 0 0 1.956286
5 5.000000 5.561015 13 13.00000 3.134308
6 6 .0 0 0 0 0 0 0.865034 14 14.00000 4.648561
7 7.000000 1.929552 15 15.00000 4.559242
8 8 .0 0 0 0 0 0 2.944438
290 1. ,

-6
,

, ,

1951:03 0.386 1952:05 0.381


1951:04 0.374 1952:06 0.381
1951:05 0.393 1952:07 0.470
1951:06 0.425 1952:08 0.443
1951:07 0.406 1952:09 0.386
1951:08 0.344 1952:10 0.342
1951:09 0.327 1952:11 0.319
1951:10 0.288 1952:12 0.307
1951:11 0.269 1953:01 0.284
1951:12 0.256 1953:02 0.326
1952:01 0.286 1953:03 0.309
1952:02 0.298 1953:04 0.359
1952:03 0.329 1953:05 0.376
1952:04 0.318 1953:06 0.416

-7

Y XI 4 Y 2 4

1949 15.9 149.3 4.2 108.1 1958 27.6 231.9 5.1 164.3
1950 16.4 161.2 4.1 114.8 1959 26.3 239.0 0.7 167.6
1951 19.0 171.5 3.1 123.2 1960 31.1 258.0 5.6 176.8
1952 19.1 175.5 3.1 126.9 1961 33.3 269.8 3.9 186.6
1953 18.8 180.8 1.1 132.1 1962 37.0 288.4 3.1 199.7
1954 20.4 190.7 2.2 137.7 1963 43.3 304.5 4.6 213.9
1955 22.7 202.1 2.1 146.0 1964 49.0 323.4 7.0 223.8
1956 26.5 212.4 5.6 154.1 1965 50.3 336.8 1.2 232.0
1957 28.1 226.1 5.0 162.3 1966 56.6 353.9 4.5 242.9

1. .. (2001). . . . 2. .:
.
2. . (2008). . .: .
3. . (2004). . 2- . .: -.
4. ., ., .. (2005). .
. 7- ., . .: .
5. .. (2004). .
. .: .
6. . (2004). :
. .: .
7. .. (1968). . .: .
8. Amemiya . (1985). Advanced Econometrics. Oxford: Basil Blackwell.
9. Cameron A.C., Trivedi P.K. (2005). Microeconometrics: Methods and Applications.
New York: Cambridge University Press.
10. Davidson R, MacKinnon J.G. (1993). Estimation and inference in econometrics. New
York: Oxford University Press.
11. Favero C.A. (2001). Applied Macroeconometrics. Oxford University Press.
12. Godfrey L.G. (2005). Controlling the overall significance level of a battery of least squares
diagnostic tests // Oxford Bulletin of Economics and Statistics. Vol. 67. P. 263 279.
13. Greene W.H. (2003). Econometric analysis. 5th ed. Upper Saddle River, Prentice Hall.
14. Gujarati D.N. (2003). Basic econometrics. 4th ed. Boston: McGraw Hill.
15. Hamilton J.D. (1994). Time series analysis. Princeton: Princeton University Press.
16. Hendry D.F. (2003). Econometrics: Alchemy or Science? Essays in Econometric Meth
odology. Oxford: Blackwell Publishers.
17. Keynes J.M. (1936). The General Theory of Employment, Interest and Money. London:
Macmillan.
18. MacKinnon J.G., White H., Davidson R. (1983). Tests for model specification in the
presence of alternative hypotheses: some further results // J. of Econometrics. Vol. 21.
P. 53 70.
19. Samuelson P.A., Koopmans T.C., Stone J.R. (1954). Report of the Evaluation Commit
tee for Econometrics, Econometrica. Vol. 22. P. 141 146.
20. The Oxford dictionary of statistical terms. (2003). New York: Oxford University Press.
21. Wooldridge J.M. (2000). Introductory econometrics: a modem approach. Cincinnati, OH:
South-Western College.

1
(<additive errors) ,
.
(variability)
.
(sample variance) ( - 1)
1? ...,
. 1?
, ()
.
(sample covariance) ( - 1)
...,
y l9 ...,
.
(sample correlation coefficient)

,
7 yi9 i - 1,..., .
(scatter plot, scatter diagram, scatter graph)
.
,
(, yt)9 i = 1, ..., 9
.
(analysis of variance)
.
(dependent variable)
,
.
(correlation diagram) . .
293

(<coefficient of determination)
, , .
(linear observation model) ,

.
(linear relation)
.
(, spurious)
,
.
(mathematical model, dependence model)
,

( )
( ).
(general-to-specific approach)
, ,
, .
(multiple-R)

.
- (Michaelis-Menton model)
,
.
(observation model) ,
,

, .
(proportional relation) ,

.
(multiplicative errors) ,
.
(independent variable)
, .
(nonlinear least squares NILS)

.

.
(uncentered R2)
,
.
294 1. ,

(scatter cloud) .
(inverse)
,
() .
(explained variable) ,
.
(explanatory variable) ,
.
i- (residual)
i- ,
i-
, .
(residual sum of squares RSS)

, .
(negative)
,
, ,
.
(least squares estimates), LS-
,

,

.
i- (error, disturbance)
i- ,
i-
, .
(fitted model) ,
.
(total sum of squares TSS)

.
(positive)
,
,
, .
(least squares principle) ,
.
(encompassing principle)
, ,
.
295

, , {parsimony principle)
, ,
.
/- {fitted value)
,
i- ,
.
(, DGP data generating process)
.
(normal equations) ,
.
(specification of an econometric model)


,
.
(mean, mean value)
..., .
(standardized variable) ,
, ..
.
(standard deviation)
. ,
.
(statistic model) . .
, (explained sum o f squares ESS)

.
(partial correlation coefficient)
,
.
(econometrics)
()

.
(econometric model) ,
,
, ,

.
,
,
( ).
296 1. ,

2
(covariance matrix)
,
, .
(expectation) ,
, .
(least squares)
,

,
.
(multivariate normal distribution)
,
.
(multiple linear regression)
,
.
(multiple regression) ,
,
.
(BLUE best linear unbiased
estimate)
.
(nonparametric regression) ,
().
(normal linear regression)
, ()
.

,
,
, .
(parametric regression) ,

, .
(ceteris paribus) ,

.
(simple) ( , (two-variable)
, , (straight-line) )
,
, .
297

, (regression model, regression equa


tion) ,
i-
.
(regression sum of squares) .
, .
(regression analysis) --
.
(regressors) .
(n-dimensional random vector)
,
.
5 2 (S2 statistic) ,
.
(statistical data), (observations)
, .
(Gauss-Markov theorem)

.
(Frisch-Waugh-Lovell theorem) ,

.
(Gauss-Markov conditions) ,
.
(regression function) ,

i- (
z- ) z- .
(centering)
.
, .
(centering)
.
.

3
(asymptotic test) ,
(
).
(hypothesis of the statisti
cal significance o f a coefficient, hypothesis that a coefficient is equal to zero)
,
.
298 1. ,

(hypothesis o f overall significance o f


a regression; hypothesis that all o f the slope coefficients, excluding the constant,
or intercept in a regression are zero)
,
.
(nested models)
,
, .
(forecast interval)
(
) ,
.
(information criteria)

;

, .
(conflict among testing procedures)
,
,
.
(variance inflation factor) ,


,
.
(Akaikes information criterion AIC) -
, ^ , -

, .
(Wald statistic)

.
(Schwarzs information criterion SC, SIC)
,
plnn
-------, ,

.
(linear hypothesis) ,

.
(unrestricted model) ,
.
299

(restricted model) ,
.
(multicollinearity)
.
F- (nonnested F-tests) F -,
.
(restricted least
squared estimate)
, # 0.
(forecast error)
,
,
.
(exact collinearity) ,
.
(reduced model) ,


.
, R 2
(adjusted R-squared, adjusted R2) ,

.
,
.
(Wald statistic)

,
- .
(test statistics)
,
( , )
( ).
(statistical hypothesis) ,
.
(statistically significant estimate)
,

.
(statistically non-significant estimate)
,

.
300 1. ,

, (forecast)
,
.
, (non-asymptotic test) ,

, .
F- (F-test)
F -
.
/^- (F-statistic) , F-pac-

- (P-value), (observed level of sig
nificance) /- ,
, /-
, , ,
/- .
- (P-value), (observed level of sig
nificance) F- ,
, F-
, , ,
F - .
/- (t-tests) ,
/-.
/- (t-ratio) /-,

.
/- (t-statistic) , /-
.

4
(autocorrelated errors)

, .
(adequate model) ,
,
.
: ,
.
(residual analysis)
,
.
(large sample tests)
, (
301

) ()
,
.
(outliers) ,
,

.
(heteroscedasticity, heteroskedasticity)
()
.
, (homoscedasticity)
()
.
(graphical diagnostic analysis)

.
(two-phase linear regression model)
,
.
(diagnostic for model misspecification)

.
- (Q-Q plot)

.
(DP-plot, DPP)

.
(goodness-of-fit tests)
, -
.
(Chow tests) ,
.
(Chow breakpoint test)

,
(Chow forecast test) -
,

, ,
.
(Breusch-Godfrey test)
,
.
302 1. ,

(Goldfeld-Quandt test)
,
.
(Durbin-Watson test)
,
.
(regression specification error test RESET)
,

.
,
,
.
(White test) ,

.
(Jarque-Bera test) ,

.
(CUSUM -
cumulative sums) ,
.
(switching regression model) .
.
(LSE approach)
,

,
,
, ,
.
(instability)
.
(forecast error)
,
,
.
(recursive coefficients)

.
(recursive residuals)


.
303

(standardized residuals)
,
.
(Durbin-Watson statistic) ,
.
(studentized residuals)
, .
, (non-asymptotic test) ,
.
(kernel density estimates)

, .

, (autocorrelation,
serial correlation)
,
, ( ).
(autoregressive transformation)

,
.
(interaction) ,
-.
(weighted statistics)
: , ,
.
(WLS weighted least squares)
,

, .
- (dummy trap) ,
--
.
- (dummy variables, dummies) ,
,
,
,
..
,
, , ,
..
304 1. ,

(Cochrane-Orcutt)

.
( Glejser test)

.
(unweighted statistics)
: ,
, , .
(<ordinary least squares estimates, OLS
estimates) ,
.
(OLS ordinary least squares)
,
;

.
(Newey-West estimator) -

.
( White estimator)

.
{panel data)
(, ) (, ).
, (<cross-section data)
(, )
(, , ).
{first order autoregressive process)
,
.

6
(two-stage least squares TSLS,
2SLS) ,



,
.
(FGLS feasible GLS)

305

,
,
,
.
, (instrumental variable, instrument)
,
, ,
.
(IV method instrumental variables
method)
,
.
(errors-in-
variables models) ,

( )
.
, GLS (GLS estimator)
,
.
(GLS generalized least squares)

,
, ,

.
(Prais-Winsten transformation)
, GLS ,

.
(reduced form of si
multaneous equations)
, ,
,
.
, (simultaneous
equations) ,


.
(weak instruments) ,
.
(stochastic)
, ,
306 1. ,

,
.
(structural form of
simultaneous equations)
,
, ,
.
, ,
.
(exogenous variable) ,
.
: ,
.
(endogenous variable) ,
.
: ,
.
2

.
ARMA

________ 7.1________
ARMA

{time series)
, .

(, , , ).
(, , , ),
, (, ...,
t = 1, ..., . ,

. ,
, .
1 ...,
.

(random sample).
(, ) X ,
jc, , 2, ..., (
) ,
(general population), ,, 2, ...,
( ) X t, 2, ..., ,
( ),
F(x) = ( < ), - < < .
F (,
, F).
308 2.

F ,

(), F(x) = Jp(x)dx. ,



00
- <9 < , ( < X < b) = F(b) - F(a),
, .

, 1 9
, ,
X l9 9

F(vx,v2,...,vn) = P{xx< vx,X 2 <v2,...,X n <vn}.
,
X l9 ...,
(,..., )9
v2 vn
F(vx,v2,...,v n) = J J... Jp(xx, x2,...,x )d x xdx2---dx.
-00 -oo -00
F(v l9 v2, ..., v) ().
t,t= 1,..., ,

Ft(vt) = P{Xt <vt},
, F^Vj) = F(v1? , ) ;
tl9 tl9 1 ^ tx < t2 < 9

Fh. <2 V = P {Xh < V'l Xh <Vh


, F 12(v,, v2) = F(vx, v2, o o ,o o );
tx, t2, ..., tm, m < n, 1 < tx < t2 < ... <tm <n,
w-
F,b (v, vf2, ..., v,m) = P{Xt] < vtl,X h < v,2, ...,Xtm <vtJ ,
, F 1j2j m(vx, v2, ..., v j = P {X x < vx,X 2 < v2, ...,Xm < vm, o o ,o o } .
, ...,
(, ..., ),
tx, t2, ..., tm, < , 1 < tx < t2 < ... < tm < n,
ph t , (x([, x,2, ..., xtJ
X, ,X t , ,X tm.
7. . ARMA 309

X l9 ...,
X t (<discrete-time stochastic process, discrete
time random process)
.
,
X l9 ..., .

F{vx,v 2,...,v n) = P { x i <vl} p { X 2 <v2}--P {X < v} = f \ F ( v t),


t=1

p{xx,...,x n) = Y \p ( x t),
t=1

{9 ..., (random sample)


, F(x)
().
(
) (randomness test):
0 : 19 2, ...,
.
,
- ,

.
-7.

,
,
Xt9
, .

.
Xt9
t~ 1, ..., 9 (strictly stationary)
(strict-sense stationary), ( < )
Xt 9 ..., Xt
, Xt +9 ..., Xt +9 tx <t2 < ... <tm n r 9, 1 < tl9
tl9 ..., tm<n 1 </j + ,..., tm + r < n ;
,
. , - 1
Xt ,
310 2.

Xt
t, , t
(, , ),
E{Xt) = D{Xt) = 2.
/ ,

.
,
, ( ),
, X,
t= 1, 2 ,..., 1 < < t2 < ... < tm<
Xt ,..., ,,
Xt
, 2, ...
, , Xt
, t = , 1, 2 , ...,
X, ,..., X,
- < tx <t2 < ... <tm<<x>.
,
, , ,
, ,
.
X, 1+

C ov(X ',X t+I)

Cov(Xt,X t+I) = E [(X '- ( , 1+ - ( ,+))\


, Cov(Xt, 1+) X, Xt+T
()
. X,
/ .
^ , Cov{X 1+)
t .
{)\
y ( T ) = Cov(Xn Xt+T).
,
D(Xt) = Cov(Xt,X t) ^ y { 0).
7. . ARMA 311


Corr(Xn Xt+T) .
():
() = (!+) = ^
( 0)
, /?(0) = 1.
Xt
..., .


E (X ,)Sju ;
D (Xt) = cT2;
Cov(Xt, X t+T) = () / .
, ,
(wide-sense stationary),
(weak-sense stationary, weakly stationary),
(second-order stationary) (co
variance-stationary) .
,
. ,
, , , (?) /.

,
/ .
,
,
. :

9 ,
.



. ,

312 2.

xt (
Xt Gaussian process),
tl9 t2, ..., tm
Xt 9 ..., Xtm m- .

.
,
, , (,
Xt) .
, xt
E(Xt) ^ //, D(Xt) = * 2() = Corr(Xt9Xt+T).
()
( ),
(
autocorrelation). () = Cov(Xt, Xt+T)
(autocovariance).
()
() (<autocorrelation function ACF).
, ..
. -1 +1,
/?(0) = 1. , xt ,
() = (-).

.
() = 1,2, ...
(<correlogram).
, .
: xt ,
xt (xt + ) .
,
,
X j9 ..., ( + 1) : ju9 (0)9
/(1),..., ( - 1) ( ju9 (0)9 ( 1),..., ( - 1)). ,
, , .

,
,
. ,

.
7. . ARMA 313

(white noise process)


(white noise) t = 0 , 1, 2 ,

(,) = 0, D(Xt) = 2 >

( ) = 0 .

, t s Xt Xs,
t s,
. , ,
X, Xs t s, m
tx, t2, ..., tm Xt ,..., Xtm
N(0, 2),
.
(Gaussian white noise process).
m /,, t2, ..., tm
,, ..., ,
, , Xt ,
X,
/ (
).
, ,
- t s
Xt Xs.
(NOISE) D(Xt) = 0.04 (. 7.1)1.

,
. , ,
,
.

, ,.
,
, ,
() 1984 . (
). . 7.2.

1
EViews (Econometric Views).
314 2.

NOISE

0 . 6 I I | i I I I | I i I I | I i | I I I I | i i I I | I i i I | I I I i | I I I I | I I I I | ^

50 100 150 200 250 300 350 400 450 t

. 7.1

DOW JONES TEMP

I II Ij ni'rp'TTT'f 1 I'J I'I 'r | I I I I | I I I I I I I I I | I I I I | I I I I I


50 100 150 200 250 t

. 7.2

, ,
xt (
),
.



(autoregressive process).
(
first-order autoregressive process, AR(1)):
X t = a X t_x +et9 / = 1,
ex ,
2;
0 ;
0 .
7. . ARMA 315


( , ) = ( , _ ),


(,) = 0 t = , 1 , .
,
X\ = a X t_x +,= (,_2 + ,_) + ,= a2X t_2 + at_x + ,= ...=
= ' 0+ '~{ + '~22 + ... + ,,

X t_x = t_2 + ,_ = a1 ^Xq + * 2 + 1 ^2 + ... + ,~\


,_2 = a X t_j + st_2 = ~2 0 + '~ + '~2 +... + ,_2,

ciX q + \.
0
\, ^5
Cov(X0,s {) = , Cov(X1,,2) = 0, Cov(Jr,_2,*M) = 0, Cov(XM,^ ) = 0,

D (X ,) = >( +,) = a2D(X,_x) + D(t), t=


, ,
D (Xt) = <
j 2x t = , 1,..., n,
:
2 =(
2 2 2
+CJZ.
2 < 1, .. \\ < 1.
:

=
~ \- 2 '
,
Cov(Xt, 1+) =
= Cov(a'X0 + ~ + '~2 2 + .. . + , ,+ 0 + ,+~ + +~2 2 + .. . + , ) =

( a2ta \ . (1 - a 2t)a \ 2
= a2l+TD (X0)+ (1 + 2+ ...+2(,~1)) <2=
\- 2

( 1+) = ,
316 2.

..
,
.
, ,

X t =aXt_x +st , t =
,
\\ < 1;
0
**
> ) = 0;
2
D (X(j)
Lo^ = <T
1- 1
Corr(Xn Xt+T) = () = .
( )
, .
y t
E(Yt) = //, ,
Xt = Yt - j u :
Yt -ju = a(Yt_{ -ju) + sn t=

Yt = a Yt_x + S + s t9 t = 1,..., n ,
8 - ju( 1 - a).

,
.
Xt (
), ,
( 1) = {,,_,) = [(,_ +,),_] = {0),

(0)

> 0 , 1,
,
.
< 0 , , ,

7. . ARMA 317

= 0.8 X
0.8
1.5 - 1.5

50 100 150 200 250 250 350 400 450 50 100 150 200 250 250 350 400 450

. 7.3 . 7.4

.
,
^ = aXt_x + st a 2 = 0.2 = 0.8 (. 7.3) = -0.8 (. 7.4).

. 0 = 0,
Xt = aXt_x + st
,

, , Xt = aXt_j + st
2. ,
t = 0 , 0
.
Xt \\ < 1.
.
,
t = 1, 0
. :
X t *0 + * ^+ * 22 +... + t >

{,) = '0 + at~lE (sI) + ~2(2) + ... + (,) = 0,

Cov(Xt,X l+T) = Cov(Xt - 0, ,+- ,+0) =


318 2.

,
Cov(Xn Xt+T) t.
, Xt
t= 1, 2 ,...
\\ < 1, t ->

{ ,) ^ 0, D( Xt) - > C o v ( X t, X t+T) = aT- ^ j ,


I- \ a
.. / -> oo
Cov(Xn Xt +T) ,
.
\\ < 1
, Xt = aXt_{ + st
0 =0.
t- 1
X, = *0 + ' ,_, \\ < 1,
=0

* , = 2 > -* . / = 1, 2 ,...
=0

k=t

t ->

k=t

, Xt Xt;
Xt Xt t > . X,
, 0 0.

Xt = aXt_ + et = 0.2
0 (. 7.5 7.8).
Xt
\
( ,) = X ,. = ^ (,_) = ,
= =0
7. . ARMA 319

0 = 0; = 0.5

I I I I | I III | I I 11| 11 I I | I I I 11I I | III I | I


10 20 30 40 50 60 70 80 90 100 t 10 20 30 40 50 60 70 80 90 100 t

. 7.5 . 7.6

0 = 0; = 0.9

10 20 30 40 50 60 70 80 90 100 t 10 20 30 40 50 60 70 80 90 100 t

. 7.7 . 7.8

_2

D{X,) = D X ,. = ^ a 2kD(l_k) = a 2 22* = ^
1,^ =
= 00 ) = =

/ 00
Cov(Xn X t+T) = E 2 > t+ J ^ a 2kE (s2_k) \ =
_V*= = \ = 1 -
Xt ( ).
,

= 1 - *

=
.. X,
X t a X t_j + t .
320 2.

st Xt_X9 Xt_2, ... ,


st Xt_x,X t_2, ... ,
Xt ,
,
.
, , Xt
, , Xt
, Xt = aXt_x + st ( \\ < 1)
t = 1, 2,..., t = 0, -1, - 2 ,. ..

X t X t_x + t (aXt_2 + t_x) + t ...
= a*Xq + a* * + * ^2 +... + t t + Qt-\ ^ ^2 ^^ ~
= t 4- af_| 4-... + * ^2 4- ^ + * ( X _ 4- q) =
00
= t 4- +... + * 22 4- * ^ + a*q 4- at+^X_x = ... = ^t_jc5
=

.
Xt - aXt_ 4- st
. (pth-order autoregressive
process AR(p))
X t = axX t_x + a2X t_2 +... 4- apX t_p +t9 0,
t D(t) = j .

, Cov(Xt_s, t) = 0 s > 0.
, et (
) (innovation sequence), et
(innovation) t.
, t


(,
) t, ^.

L (lag operator),

LXt =X,_x.

(backshift operator).
7. . ARMA 321

,
, .
L
t = 0, 1, 2, L
, Lk.
LkX t = X t_k.
9
LXt =l- X t = X t9
L L~x:
ITlX t =Xt+1.
, ,
(, 2002 ).
9
a\X t-\ +a2^t-2 + + apX t-p
:
axLXt + a2L2X t +... + apLpX t =(axL + a2L2 +... + apLp ) X t,
, p - , :
a(L)Xt = Gt,

a(L) = \ - ( a xL +a2l} +... +ap Lp).

- a(L)Xt = st
, 1 , ( )
a(z) = 0 ( )
\ z\<*\ .

, AR(1) a(z) = 1 - az9 a(z) = 0


z = 1/9 |z| > 1
| | < 1.

1 s t ,
, () Xt = et (. 9 . 7).
322 2.

a(L)Xt = ,
(., , (Hamilton, 1994, . 58 59)) :
1 00
.== ------S =
.

a(L) fa 1 ' 1

b0 = 1, Y ) bj l<Q0-
=
, ,

{,) = [ X = -
V-/=0 7=0
AR(p) /

a(L)(Xt -//)= ,, a(L)Xt = S + st ,

S = a(L){i = M(l - i 1
- - ) = / / ( ).

a(L)(Xt - /) = et :
1 00
* ,= //+ ,= //+
\> =0
, AR(p) a(L)Xt = S+e
, S,
<5
---------------------------------------- '
(1 ! 2 - - )
, 5 = 0, = 0. ,
1, .
AR(1) a(L) = 1 - aL, \ \ < 1,

X, ------- st (1 + aL + a2I? +...) st = st + as,_{ + a2st_2 + ...


1 -a L
,
() = Corr(X,, X t+k) = , = 0,1,2,...
0 < < 1 ( () - 1,2, ...)

, -1 < < 0
.
AR(1) = 0.8 = -0.8
. 7.9 7.10.
7. . ARMA 323

. 7.9 . 7.10

AR(/>) > 1 ,
( )
a(z) = 0. ()
( = zmin -
~ min
a(z) = 0 ),
, \ \
. > 0 ,
, 2, ..., .
, Xt
, Xt (X, + ) -
.
, AR(p).
Xt
.
Xt_k ( > 0)
X, =axX t_x +2,_ 2 +... + ,_ +,,
AR(p) ,

,,_ =,_,_ + a2X,_2X t_k +... + apX t_pX t_k + stX t_k.
:
{ , ,_) - (,_ ,_) + 2(,_2 ,_) + ... + apE(X,_pX t_k) + E{s,Xt_k).
,
E(e,Xt_k) = 0 ( , ),
r(s) = Cov(XtX,_s) = E (X ,X ,_S) ( E(Xt) = 0),
324 2.


() = , ( - 1) + 2 { - 2 ) + ... + { - ), >0.
;/(0 ),
(Yule-Walker equations):
() = ( - 1) + 2( - 2 ) + ... + ( - ), > 0 .

, ,
Oj ,

( . ).

7.1.1
AR(2):
X, =4.375 + 0.25 -0.125 X,_2 +et.
a(z) = 0 :
l-0.25z + 0.125z2 = 0, z 2 -2 z + 8 = 0,
z 12 = l iy fl. 1,
.
:
6 4.375 _5
~1 - - ~ 1-0.25 + 0.125 "
5.
.
=2,
() = 0 .2 5 ( -\)-0 2 5 ( -2 ), > 0.
{0) = 1. ( 1)
(1) = 0.25/0(0) - 0.125/?(-1) = 0.25 - 0.125(\),

^ ( 1) = ~25 = - = 0 .222 .
1+ 0.125 9

{2) = 0.25/7(1) -0.125 (0) = 0.25-0.222- 0.125 = -0.069,
/?(3) = -0.045, () = -0.003, / 7(5 ) = -0.005 ..
7. . ARMA 325

*
0.24--
0.20 -
0.16-
0.12-
0.08-
0.04-

-0 .0 4 - | ;
0.0 8------1-----1-----1-----1-----1-----1-----1-----1-----1-----1
1 2 3 4 5 6 7 8 9 10

. 7.11 . 7.12

,
,
.

. 7.11 7.12 .


q (qth-order moving average process MA(#)).

X t = st + bxst_x + b2et_2 + ... + bqst_q, bq * 0,

st D(t) = 2.
.
, //,

X t -/u = t + bxt_x + b2t_2 + ... + bqt_q,


..
X t = H + t +bxt_x +b2s t_2 + ... + bqt_q.

q = 0 ju = 0 . q = 0,
X t = ju + t + bt_x ((1)).


D{X, ) = (l + b2) a 2
e, [(X, - //)(*,_, - )] = 2
,
326 2.

Xt

( ,) = /, D(X,) = (1 + Z>2 )<t2,

(1 + &2)2, = ,
() = <I, = \,
, >
: > 1.
1.

'l, it = 0 ,

0 , > 1,
.. .
.
, b > 0, < 0.
(1) b > 0 ,
(1) < 0
. , (1)

.. ,
AR(1)
( \\9 1).
() :
X, - f i = b(L)et ,
b(L) = 1 + + ... + bqLq.

0, k> q .
, ()
//,
<4 = ( 1 +>2 + ...+ 2 2

7. . ARMA 327

q-k

I
-------, * = 0 ,1 ......q,
() =
=
, := # + 1, q + 2,..

q (.. q).

7.1.2
:
) (1) = 0.8 E(Xt) = 6 , .. X, = 6 +st + ;
0.8
{ 1) = - = 0.488;
1+ 0.8
) (1) b = -0.8 (,) = 6
- 0.8
= -0.488.
/41) =
1+ 0.8
. 7.13 7.14.
2 = 1
. 7.15 7.16;

. 7.13 . 7.14

) (2)Xt = 5 + , - 0.75,_, + 0.125s,_2

_ b ^ + b f o _ -0.75-0.75-0.125 _
Pi _ ___ ~ __
? +,2 + 6 2 1+ 0.75 +0.125
328 2.

= 0.8 = - 0.8

I I I | I I [ I | 11 I I | I I I I | 11I I | I I I I | I I 1I | I I I I '| I I I I | I I I I | I I I I | I I I I | I I | I I I I | 11 I 11 111 11 1*

50 100 150 200 250 250 350 400 450 t 50 100 150 200 250 250 350 400 450 t

. 7.15

| I |-| 11 j I I 11 | I I I I | I I | I I | I I | 11 | I I '
50 100150 200 250 250 350400 450 t

. 7.17 . 7.18

{ 2) = ^ ^ = 0 .0 7 9 .
1.578

. 7.17 7.18.


, bj = aJ9 0 < < 1,
, et ,
():

X ,= b{L )et = f dbJet_J,


j =0

b{L) = \ + Y b j L J 00.
;=i =
7. . ARMA 329

,
AR(1):
X, = aXt^ +,, < 1,
.. () AR(1).
, AR(p),

a(L)(Xt -ju) = s a{L) = 1- (axL + a2L2+... + apLp),


s, , ():
X, = + (),,

c(L)= 'L c J LJ = - j - | ,|< > .


j = a\L) =0
(., , (Hamilton, 1994, section 3.4)).
, ,
.
a(L)(Xt - ju) - t
,
- Xt - ju = b(L)st

?
Xt ()-

X t - p = b(L)el = Y dbj l-j, b0 = 1,
j =0

,
(invertibility condition):
b(z) = 0
\z\ < 1 .
Xt

d(L){Xt - / i ) = 6t ,

d(L) = \ - f ^ d j L j d0 = 1, .|< ,
= 1 \L ) 7=1


\
1 -^ d jU { ,- ) =
7= 1
330 2.

..
(X, ~M) = dx(X t_x - f i ) + d2(,_2 -//)+ + ,.
( dj ., ,
(Pollock, 1999).)
X, = / + , + ,_
: | 6 | < 1,
) ) 7.1.2 (1)- .
, )
, b(z) = 0 :

l-0.75z + 0.125z2 =0.


(2 4) ,
(2)- .
, , ( ^-
X, = 6 + , + 0 .8 ,_]
1
b(L) = 1+ 0.81, X , - 6 = (1 + 0.8L)s ( X ,- 6 ) = s
1+ 0.8 L

(1-0.8L + 0.82Z,2 ---- )(X t -6 )= t,

(X, - 6 ) = 0.8(X,_, - 6 ) - 0.82(X t_2 -6 ) ++

X, = 6 (l-0 .8 + 0.82 ---- ) + 0.8Xf_, -0 .8 2X ,_2 h + s, =

6 + 0 .8X ,_ ,- 0 .82X ,_2 + + ,.


1+ 0.8
, ( 1^ X, = ju + , + b s,_, | \ > 1?

1
t + T ;>+1 e,+ \-F s, 4 = ^ fl + - F

F = L 1 , FXt =Xt+l,
7. . ARMA 331

1 , 1 1 1 ) U 1 v 1 v ,
X l+l =\ !- - +-^2^ + - - \ +1 <+2 +- ^ <+4 -+ b,

X t+l = - ^ - + - X t+2- \ , + + \ x t+4- + ber


f+l 1 + b b ,+2 b2 b3
, Xt
, Xt^k, = 1, 2 ,...


(
)

Xt ,
,
q AR
ARMA(/>, q) (autoregressive moving average, mixed autoregressive moving aver
age). , X,
ARMA(p, q),

*, = Z ^ -y + 2 > , * 0 , \ ,
j=1 =

, , D(st) = <2, b0 = 1.
:
a(L)Xt = b(L)s

a(L) b(L) , AR()


MA(<jr). ,
a(L) b(L) (. 7.1.2 7.1.4).
/j ,
ARMA(p, q) :

'~ / = (x '-j ~ ) + & t-J-


7=1 7=0

ARMA(p, q) Xt E(Xt) = (.,


, (Hamilton, 1994, . 59 61)):
a(z) = 0
\z\ < 1 , Xt
()
332 2.

xt Y jcj st-j' =1> X l 1<00 X t =ii + c(L)en


7=0 7=0


c {(z)=
\ ^ c lz jJ = -----
2^1 h i2 )
- ' a (z )

b(z) = 0 < 1
( ),
X,
AR(oo)
00
X t - n = Y ,d j(X t-j-M ) + t d(L)(X t -ju) = sn
j =i

* ) -
7= 1
b (z)'

, ARMAQo, q)
,

.
ARMA(p, q)
, AR(/>) (). ,
> a(L)Xt = b(L)et ,
a(L)X, = sr ,
ARMA(1, 1)
( ) = ( - 1) = 2 , 3 , ...,

^, = ,_] + , , ( 1) .
ARMA
. ARMA(/?,, qx) X, ARMA(/?2, q2) Y,
Z, = Xt +
, Z, ARMA(p, q) ,

=\ +2,
q =\ + <72 , pi + q2>p2 + q i ,
q= P i + q \, + q i> p \+ q 2 .
, q
. ( ,
ax(z) aY(z), X, Y
.)
7. . ARMA 333

, AR(1),
, ARMA(2, 1).
.
,
AR, ARMA
. ,
AR, .
,
.
, , AR()
, ,
, .. (0).
ARMA(p, ).

J 7.1.1. , ARMA(p, q)
Xt
,
,
.
AR(p)
, ()
.
,
ARMA,
AR, . , ,
, ,
( )
. ARMA
(more parsimonious models).

7.1.2.
(common factor
problem). , ,
. ARMA(2, 2):
X t = 13 X t_x- 0AXt_2 + st - 03t_x- 0Ast_2.

:
a(L)Xt =b(L)et,

a(L) = 1-1.3 L + 0.4L2 = (1 - 0.81) (1 - 0.5L),
b(L) = 1- 0.3L - 0.4Z, = (1 - 0.8L)(1 + 0.51).
334 2.

, a(L) b(L)
(1 -0 .8 1 )
(1 - 0.8Z)(1 - 0.5 = (1 - 0.8L)(1 + 0.5L)ev

,
(1 - 0 .5 = (1 +0.5L)*,,
..
Xt 0.5Xt_ + 6t + 0.5^_ j .
X, ARMA(1, 1),
ARMA(2, 2).

a(L) i(L)
a(L)Xt = b(L)st ARMA
. , (. 7.1)

X t = axX t_x+ a2X t_2 + s t + bxt_x+ b2st_2
( 100 )
X t =l.3Xt_x - 0 A X t_2 +t -0.3et_x - 0Ast_2.

7.1
X
Method: Least Squares; Sample (adjusted): 3 100; Included observations: 98 after adjusting endpoints;
Convergence achieved after 16 iterations; Backcast: 1 2


/- -

AR(1) 0.188566 0.732636 0.257381 0.7974


AR(2) 0.088017 0.364477 0.241489 0.8097
MA(1) 0.809562 0.726516 1.114307 0.2680
MA(2) 0.183854 0.392497 0.468421 0.6406

R-squared 0.544602 Mean dependent var -0.142850


Adjusted R-squared 0.530068 S.D. dependent var 1.301768
S.E. o f regression 0.892383 Akaike info criterion 2.650117
Sum squared resid 74.85665 Schwarz criterion 2.755626
Log likelihood -125.8557 Durbin-Watson stat 1.989028

Inverted AR Roots 0.41 - 0.22


4^

Inverted MA Roots
1

-0 .4 0 + 0.14/
i


0
7. . ARMA 335

4
, .
. ARMA(2, 2),
ARMA(1, 1), (. 7.2).

7.2
X
Method: Least Squares; Sample (adjusted): 2 100; Included observations: 99 after adjusting endpoints;
Convergence achieved after 6 iterations; Backcast: 1


/- -

AR(1) 0.458982 0.115096 3.987819 0.0001

(1) 0.529106 0.108628 4.870793 0.0000

R-squared 0.542302 Mean dependent var -0.141407

Adjusted R-squared 0.537583 S.D. dependent var 1.295189

S.E. o f regression 0.880745 Akaike info criterion 2.603897

Sum squared resid 75.24399 Schwarz criterion 2.656324

Log likelihood -126.8929 Durbin-Watson stat 1.969688

Inverted AR Roots 0.46


Inverted MA Roots -0 .5 3


ARMA(1, ^-
, .

J 7.1.3. Xt ARMA(p, q \ a(L)Xt =


= b(L)sr ,
a(z) = 0 |z| < 1
, Xt .
, a(z) = 0
z c \z \ = \. .
Xt - sr . :
X t= X t-v
Xt - X t_x = et - e t_x.
:
( 1 - = ( 1 - 1 )*,,
.. a(L)Xt = b(L)sn a(L) = 1 - L b(L) = 1 - L. ,
^, ARMA(1, 1)
336 2.

Xt X t_\ + 8, St-\i
a(z) = 0 z - 1.

7.1.4. , ARMA(p, q) Xt,


a(L)Xt = b{L)s a(z) b{z) ,
a(z) =
\z\ < 1
Xt.

7.1.5.
X,
ARMA(j, q), ,
.
( Wolds decomposition). {Wold, 1938)
, X ]
:
00
X t ~ X cj s t - j 9
=
00
0 = 1 2 <,
j =
st ;
Zt
, Cov(Zn st ) = 0 j\ Z,

, : X t_l9X t_2,...
Xt
: (linearly indeter-
00
ministic) ^ Jcj t-j (line-
7=0

arly deterministic) Zr
, .. Z, = 0, Xt
(purely linearly indeterministic).

,
,

00
X, = X ; et-j = 1 X ; <00
=0 J=
7. . ARMA 337



:
Xt X q, t 1,2,...,
0 ,
.

.

ARMA,

,
ARMA, , ,

.

(SAR(l) first order seasonal
autoregression)
X t =aAX t_4 +t, \a4 |< 1,
(SMA(l) first order
seasonal moving average)
X t =et + b4t_4.

p(k) = akJ A k = 4m, m = 0 ,1 ,2 ,...,
p(k) = 0 > 0 .

*0) = 1, (4) = 4,
() = 0 > 0 .
SAR(l) = 0.8 SMA(l)
4 - 0.8 . 7.19 7.20.
, ,
ARMA((1, 4), 1):
X t = a \ X t-\ + # 4 ^ 0 -4 + t + t- 1

ARMA(1, (1,4)):
X t = axX t_j +st + bxt_x+ b4t_4.
338 2.

' 11111 I I I I |
1]II111 1III|III|III1|IIII[I
I I[ 11 I I |

10 20 30 40 50 60 70 80 90 100 t
. 7.19 . 7.20

-
2 1 1
= , 4 = ----- , 4 = (. 7.21)
3 48 5
= 0.4, = 0.3, 4 = 0.8 (. 7.22).

ARMA(1, (1,4))

-2 -

10 20 30 40 50 60 70 80 90 100 t
. 7.21 . 7.22

, a(z) = 0 :

1 - z + z 4 =0, .. z 4-32 z + 48 = 0;
3 48
z x = 2 , z2 = 2 , z3 = - 2 +i /8 , z 4 = - 2 - 1
/8
, .
a(z) = 0 :
1 -0 .4 z = 0;
z = 2 . 5 > 1, .
7. . ARMA 339

,
, :
(1 - a, L)X, = (1 + L) (1 + bAL*) ,
(1 - ,1X1 - 4L4)X , = (1 + bxL )s,.

X t = axX t_x +et + bxst_x + b4st_4 + bxb4st_5,

X t = axX t_x + a4X t_4 axa4X t_5 + st + bxst_x.

1 4 (.. st_x st_4),
1 4 (.. Xt_ Xt_4).
,
X t ~ a \X t-\ + t + b \ t-\ + b 4S t_4 +bst_s ,
X t = axX t_x+ a4X t_4 + a5X t_5 +st + bxst_x
b5 = bxb4, a5 = - a xa4.
( ,
),
,
(parsimony model). , -
(
) .
ARMA ,
, (Ghysels, Osborn, 1991).

1. ?
?
2. ( )?
3. ( )?
4.
?
5. ?
?
6 .
? ?
7.
?
8 . ?
?
9. ?
?
340 2.

10.
?
?
11.
?
12.
, , ?
13. ?
, , ?
14. !
15. !
?
16. !
17. ?
?
18. q?
, , ?
19. q ?
20. q?
?
21. ?
?
22. ARMA(p, q ) l
?
?
?
23. ARMA ?
24. ( ARM) ?
25. ARMA?
26. ARM?

_________ 7.2 _________


ARMA

, ...,
ARMA,
.
:
1) (identification stage);
2 ) (estimation stage);
3) (diagnostic checking stage).

ARMA, .. q.
,
7. . ARMA 341


.
19 2, 9 9 2, bq
.

.
,

19 2, ...

.

.
, ,
, , ..
, .
,
,
.
ARMA(, q)9
X t ~ ARMA(p, q).
AR(p), X t ~ AR(),
[q)9 X t ~ (^).

ARMA

ARMA
(ACF autocorrelation function,
PACF partial autocorrelation function) ,
ARMA.

ARMA .
(
) .
(PACF)
Ppart(k)
(partial autocorrelation)
Xt Xt+k9
Xt+,..., Xt+k_.
, {) Xt_k
Xt_l9 ..., Xt_k9
,
342 2.

(., , (Hamilton, 1994, . 111)), ppart(k)



p(s) = axp(s - 1) + a2p(s - 2 ) +... + akp(s - ), s = \,2 ,...,k ,
:
p(s - 1)0 , + p(s - 2)2+... + p(s - ) = /?(s), s = \,2 ,...,k ,
, , 2, ..., , ( 1 - ), ...,
( - 1) .
, ,
PACF :

/V () = 1

1 ( 1)
( 1) ( 2) ( 2 ) - 2(\)
Ppart ( 2 )
1 ( 1) 1 -V 0 )
( 1) 1

1
( 1)' Pi 1)
( 1) 1 ( 2)
( 2) Pi 1) ( 3)
Ppar,0) = 1 9
Pi 1) ( 2)
( 1) 1 ( 1)
( 2) ( 1) 1

1 ( ( 2) .. ( 1)
( 1) 1 ( 1) .. /7(2)
(2) ( 1) 1 .. /7(3)

( - 1) ( ~ 2 ) ( - ) .. />(*)
1 ( 1) ( 2 ) .. /7(^-1)
( 1) 1 /7(1) .. (~2)
(2) ( 1) 1 ( ~ 3)

( - 1) ( ^ - 2 ) /?(*-3) ... 1
7. . ARMA 343

,()

, { 1), ( 2 ),..., ().
, AR(p),

)*>
,() = >.
PACF

ARMA(p, q )c q > 0 .
, ACF q ().
, PACF
AR(p). ACF PACF
ARMA(p, q) * 0, q 0.

() ,()
ACF {sample ACF SACF),
{sample autocorrelations)

t =1
1
{) = ~ ^ { , - ){,+ - ) {) ,
I ~ t =1
PACF {sample PACF SPACF),
^ { ) {sample partial autocorrelations).
, 1{)
p{s) r(s). .
, ppart{k) (Xt_k - //)
Xt_, - /, ..., Xt_k - ju,
X, - ,

344 2.

( ut
ut = X t { + axX t_x + a2X t_2 +... + akX t_k),
).
rpart(k).
ARMA(p, q) a(L)Xt = b(L)et
a(L) = 0 b(L) = 0
, 19 ..., ~ i.i.d. {?) < , /2,
/(), () rpart(k) //, (), ()
pPart(k) (. (Hamilton, 1994, . 199)). ()
rpart(k) () ppart(k \
() ppart(k). ,
= = 2 , , () = 0 ppart(k2) = ,
, , ( ^ 0 2) * 0 ,
. ,


.
ACF PACF
- .
ACF PACF

ARMA. , . 7.3
ACF PACF
.
AR(p) (^)
r(k) rpart(k)
() ppart(k)
() ^ ^ ). ,
,
(.. st ).
1. Xt (),
lim E(r(k)) - p(k),
>00

1( q ^
1+ 2 p \ j ) k > q,
J=i
,
(), k> q.
7. . ARMA 345

7.3
ACF PACF

ACF PACF

, p(k ) = 0 k * 0 ) = 0 * 0
(0)

AR(1), PpartO) = \
> 0 () = a* ) = 0, > 2

AR(1), ,() = \
< 0 () = 1 ) = 0, > 2

AR () >

(1), = 1; ;
^ > 0 > 1 Ppar'i 1 ) > 0

(1), = 1;
^ < 0 > 1 ; (() < 0 > 1

(q) > q

ARMA(1, 1), 1; ( > 0)


> 0 1) ( + ) ( < 0 )
1

ARMA(1, 1), 1; ( < 0)


< 0 { 1) ( + ) ( > 0)
1

ARMA {, q)
, q ,

SA R (l) , ,
; ;

SM A (l) , ,
; ;

, > q
() . ,
0 : Xt
() , ,
346 2.

0.05.
, q = 0, Xt ~ (0) ,
0: Xt

2
> =, > 0 .
/77
2. Xt AR(p), >
,()

, () 0, ~1), D{rpart ()) -1.

, HQ:X ,~ AR(p)

,
0.05.



2
ACF PACF, =.
/
, 0.95, > 0 (),
X, , > 1(), X, ~ AR(p).
:
, 0.05, , 0
.
, , ?
, .
= 499 2, ..., 499. . 7.4

() = 1, 2 ,..., 36.
2
ACF , = = 0.0895
4
( 13) = 0.102. ,
H0:X t ? PACF ,
, .
( = 499),
rpart(k \
= 1, 2 ,...
7. . ARMA 347

7.4

(fc= 1..... 36)

ACF PACF ACF PACF ACF PACF

1 -0.0 19 -0.019 13 0.102 0.126 25 -0.053 -0.031

2 -0.013 -0.014 14 -0.071 -0.051 26 -0.015 -0.0 18

3 -0.083 -0.083 15 -0.044 -0.0 36 27 -0.064 -0.035

4 0.038 0.035 16 0.017 0.034 28 0.032 0.042

5 -0.0 47 -0.049 17 -0.083 -0.115 29 -0 .0 5 7 -0.075

6 0.017 0.009 18 0.035 0.028 30 -0.053 -0 .0 4 4

7 -0.024 -0.019 19 -0 .0 4 9 -0.085 31 0.011 -0 .0 0 6

8 0.062 0.053 20 0.069 0.032 32 0.034 0.021

9 0.061 0.069 21 0.041 0.022 33 0.029 0.034

10 0.074 0.073 22 -0 .0 1 4 -0 .0 5 7 34 -0.042 -0 .0 5 7

0.079 0.099 23 -0.035 -0 .0 1 8 35 0.013 0.064

12 0.021 0.034 24 0.034 0.012 36 0.046 0.055

, , rpart{k) -
2
=. 0.05. -

( 36) rpart(k), = 1,2, ..., 36,
:

2 = 260.052(1-0.05)36"2 0.052 -0.9534,

lg(^2) = lg(630) + 21g(0.05) + 34 lg(0.95) = - 0.560.


: 2 = 0.275,
PACF 36 .

ACF,
()9 - 1,2, ..., , Xt (
(#) q > 1 ).
ACF :
2 =\6 0.05(1 -0.05)35,
lg(Pj) = -0.780, : 1 = 0.166.
348 2.

,
ACF PACF 1.
, ,
() EViews, :
() ( - ), .
, ()
.

ACF PACF 0- (Q-statistics),
,
.
g -.
(Box-Pierce Q-statistic)
(Box, Pierce, 1970) :

0 = | > 2(:).
=\


r( 1), (2 ) , () , X, , ,
4 r(k) N (0,1), Tr2(k) [jV(0, I )]2 = %2( 1)
(, Xt . (,
1974)). ,

0 , , .
0.05,

Q > .95( ^ 0
Jq 95(M) 0.95 -
.
P - (
) Q - 1,2,...
0 ,
- 0.05.
, ,
%2() .

1
. 2 .


7. . ARMA 349


(Ljung-Box Q-statistic\ (Ljung, Box, 1979):
r2(k\
e=r<r+2) ( F ^ )
( ) %2(),
,
. EViews
P-,
%2{).
g -
. P- 0-

(. 7.5).
7.5
Q-

- - - -

1 0.670 10 0.243 19 0.044 28 0.072


2 0.873 11 0.146 20 0.033 29 0.065
3 0.292 12 0.187 21 0.037 30 0.061
4 0.348 13 0.064 22 0.049 31 0.076
5 0.349 14 0.045 23 0.056 32 0.084
6 0.455 15 0.049 24 0.064 33 0.096
7 0.539 16 0.066 25 0.061 34 0.099
8 0.438 17 0.037 26 0.077 35 0.119
9 0.360 18 0.044 27 0.063 36 0.119

-, = 14, 15, 1722, 0.05,



0 : Xt .
^- , 0.05,
0 .
- , ,
, ,
.
(.Kwan, 1996) (Kwan, Sim, 1996).
,
ARMA
.

350 2.

, .
,
.
.
AR , .. , X,
AR(&)

, ~ = Y*a(x t-J ~V) + ,


j =i

,
(Akaike
information criterion AIC) (. Akaike, 1973).
,
2
1 = 1<*2 + ,

;
< , AR - 1.

Ps ~ ^ ]P s- j
7=1

a ^ , j = 1..... , akJ
, , ,

, = (,- * ) - ](:t_j - ) ,
7 = 1


1
^ - 1V
~ t *
1 t=1

,
() 0
. ,

1
Views. , ,
.
, .
7. . ARMA 351

1<I + ,
= 0(7~1) (.. -> <> ,
).

(Schwarz information criterion
SIC) (. {Schwarz, 1978)):
__ , 2 , In
SIC = In ak + k^~ .

{-
nan-Quinn information criterion HQC) (. {Hannan, Quinn, 1979)),
cT = 27^111, > 1,
,^ 1 2 , 2 1 1
HQ = 10* + ---- ----- ,

0 >.

.

7.2.1
AR(2)
X t = \2 X t_x -0.36,_2 +r
a(z) = 0
l-1.2z + 0.36z2 =0

zx_2 = ^ > 1 , ,
, .
t = 1, 2, ..., 500
. 7.23.
352 2.


:
ACF PACF g - -
1******* I******* 0.000
1 0.899 0.899 406.25
1****** 0.000
2 0.732 -0 .3 9 6 675.97
1**** 3 0.561 -0.005 834.98 0.000
1*** 4 0.409 -0 .0 2 7 919.59 0.000
1
1** 5 0.277 -0.048 958.40 0.000
1
1* 1 6 0.167 -0.015 972.59 0.000
1* 1* 7 0.095 0.071 977.15 0.000

1 8 0.045 -0.045 978.19 0.000

1 9 0.014 0.011 978.29 0.000

1 10 - 0.001 0.020 978.29 0.000

1 11 0.003 0.055 978.30 0.000

1 12 0.019 0.001 978.49 0.000

j= = 0.089 PACF,
JT
- 1, 2. ,
, 0 : Xt ~ AR(2).

, AR 4-,
3-, 2 - 1- ,
AR (. 7.6).
AR(2).B
7.6

II

p=1 /> = 2
II

AIC 3.083264 2.91800 2.919244 2.924441


SIC 3.100148 2.94336 2.953116 2.966846

AR , ,
, ARMA(/;0, q0) ( p 0,q 0)
a(L)X, =b{L)st,
(0, q0),
(Kavalieris, 1991).

AR(Ar):

Xf ~ ^ akjXt_j + t,
7=1
7. . ARMA 353

^9 j - 1, ..., ,


4 (0 = * ,-Z < v ^ -r
=1


.
. (
, ,
.)
X, ,_ j - 1,..., , X,
(t - 7 ), '= 1 ,..., q.
7, .. 7 = 1, ...,, bj
bj, 7 = 1,..., q. a(z), b(z)

) = 1 - X " /zj ^ ~ Z h zj
j =1 j =0


i,= b - [(L)a(L)xt,

1 t=1

, ,
ARMA , st = b~l(L)a(L)xt,
.
0, q0 (, q \

1
S IC { 2 q ) = \ 2 q + ( + q ) ^ j - .

, SlC(apq) q,
p > p 0,q > q 0, (, q).

() ARMA,
..
, q ARMA(p, q), ,
.
354 2.

,
.
(, AR(1)),
, (
) , .

, . ,
()
()
() .
, AR(p), 9 2,


P(k) = Y , ajP (k ~ j ) = 1,..., ,
7=1

( 1) , {)
( ) r( 1) , ()
.
(^) (q > 0)
, .
(1)
X , - = , + bxs,_x, t = \,...,T .
19 ..., 9 9...,
() 0:
= X x- ju - b e 0,
2 = 2 - - bsx = 2 - - b(Xx - p .-b s 0) = ( 2 - ) - ( - ) + 20,

= - - _ = ( - ) - (_- ) + ( _2 - ) - . . . +
+ ( - l) T-'bT-'(X x- p ) + ( - l f b Te0.
( ) ,
9 ..., 09

Q(b) = s x + s2 + ... + %9
b.

, , , (
) . ,
.
7. . ARMA 355


0, .
, | 6 | < 1,
|6 | 1.
,
0 = 0. 0
, , 0 = 0 ,
,
0, .
() q > 1:
0 = _ = ... = _q+l = 0.
( EViews)
backcasting,
0 = _ - ... = _q+x
.
,
,
.
ARMA
, ,
(Hamilton, 1994, . 117 151). ,

,
.
.
AR()
a(L)Xt = S + t.
, /
Xt 8

= ------------------------->
1 1 2 - ...~
,, 2, S,

X t = + 1 /_1 +a2X t_2 + ... + apX t_p +

, , ..., 8, /
:
356 2.

, , ,
EViews. :
X, = //(1 - , - 2 - . . . - ) + ,_ +a2X t_2 +... + apX t_p +s,
19 2, , /.
.
, ,
,
(NLLS nonlinear least squares)
.

7.2.2

( , ). . 7.7,
. 7.24.

7.7
,

, , , ,

1946 10.8 1951 11.2 1956 10.4 1961 10.7

1947 10.3 1952 11.2 1957 10.2 1962 10.6

1948 11.1 1953 11.4 1958 10.6 1963 10.7

1949 10.9 1954 11.2 1959 10.9 1964 10.5

1950 11.8 1955 10.5 1960 10.3 1965 10.9

X
12.0 -

11.6 -

11.2 -

. 7.24
7. . ARMA 357

, , :
Autocorrelation Partial Correlation g - -
1*** *** 1 0.429 0.429 4.2694 0.039
1*** 2 0.366 0.222 7.5350 0.023
**| 3 0.059 -0 .2 0 4 7.6255 0.054
1
1 1 4 0.016 -0 .0 3 4 7.6324 0.106
*1 *1 5 -0.156 -0 .1 2 9 8.3498 0.138
**| ** 6 -0.255 -0.195 10.3930 0.109
***] 7 -0.321 -0.123 13.8790 0.053
*1
*1 1* 8 -0.133 0.175 14.5230 0.069

,
2 2
= / = = 0.447. 0.447
/T /20
.
,
(0)
0 = +

,
ACF 1 . ,

AR(1) (1). ,
: ,
, .
-
, ,
,
.

,
,
, -
,
.
,
1940 1945 . (
). 1940
1965 . . 7.25.
1942 1944 .
1940 1965 .
1946 1965 .
358 2.

X ^
12 - ........................................................................ ;

11 - i

- |

9 - i


id
7
o c \ I 4 ^
+>

+- -4-
T T T T T T T T T T T T
o I rso cn1 I ' s
05
3' cCD
i,n i0
r\
in
O)
oO1 O o CO

Gi
eo

o>
r \ c CD
O)
en
CNJ j '^'
r\ 3er%
)
*

^

. 7.25

AR(1), , , ( 1) = .
, ( 1) r( 1) = 0.429,
. ,
AR(1)
,
(. 7.8).
7.8
X
Method: Least Squares; Sample (adjusted): 1947 1965;
Included observations: 19 after adjusting endpoints; Convergence achieved after 3 iterations


/- -

10.81451 0.159261 67.904450 0.0000

AR(1) 0.430515 0.219257 1.963522 0.0662

R-squared 0.184864 Mean dependent var 10.81053


Adjusted R-squared 0.136915 S.D. dependent var 0.425434
S.E. o f regression 0.395238 Akaike info criterion 1.080645
Sum squared resid 2.655627 Schwarz criterion 1.180060


.
(1), ( 1) = -
1 + [
( 1) (1) = 0.429
1 = 0.429. 0.567 1.704.
1 +
7. . ARMA 359

(1) ,
(1) . (1)
(backcasting) (. 7.9).

7.9
X
Backcast: 1945


/- -

10.81379 0.113355 95.38726 0.0000

MA(1) 0.280610 0.228102 1.230195 0.2345

R-squared 0.117961 Mean dependent var 10.81000

Adjusted R-squared 0.068959 S.D. dependent var 0.414094

S.E. o f regression 0.399561 Akaike info criterion 1.097739

Sum squared resid 2.873684 Schwarz criterion 1.197313


.
P- t- F-
. ,
,
20 .

1945 ., (. 7.10).

7.10
X
Method: Least Squares; Sample (adjusted): 1946 1965; Included observations:
20 after adjusting endpoints; Convergence achieved after 24 iterations; Backcast: OFF


/- -

10.81515 0.112582 96.06431 0.0000

(1) 0.274024 0.229231 1.195405 0.2474

R-squared 0.116800 Mean dependent var 10.81000

Adjusted R-squared 0.067734 S.D. dependent var 0.414094

S.E. o f regression 0.399824 Akaike info criterion 1.099054

Sum squared resid 2.877464 Schwarz criterion 1.198628

Log likelihood -8.990543 F-statistic 2.380443

Durbin-Watson stat 1.778286 Prob(F-statistic) 0.140261


360 2.



, .. ,
( ).
.

, 0
, , , s,
.

ARMA(p, q)
a(L)X, -b ( L )s t,
..
X, = o , I M + a2X,_2 +... + apX t_p +et +bxs,_x + ... + bqsl_q,

a{L)X, = b(L)et,

a(L) = l - a xL - . . . - a pLp,

b(L) = \ + bxL + ... + bqLq.


ARMA(/;, q) ,


, a(L) b(L) a(L)
b(L) :

b(L)
, , ,
-
.
,
st. , ,
, .
, (Bartlett, 1946)
(. {Box, Pierce, 1970))
st:
7. . ARMA 361

rs{k) = ^ j -------
2
t=i


Q bp = ^ X () (- ).
=\

, QBP ,
/ 2( - - q) , ,
/ . ,
Qbp> 1.9$( - ~)-
,
QBP / 2( - - q).
, ,
( + 5)
E(QBP) * ( M - p - q ) -
2 + 2
_ ( + 5)
, - ,

2{ - p - q ) .
.

Q bp X (E (Q bp ))
E{QBP) (
).
D(rs(k)) -
1 -
------- . -
2 +27
2
Ql b =T(T + 2 ) 1
= -
%2( - - q),
Qbp> QBP
/ 2(/ - - q), Qbp-
,
; .
:
, st
362 2.

, ,
st . ,
D(et).
,
( , , ).
st (heavy-tailed
distribution), ..

.
EViews ARM
P-
0- .
AR(1) (1)
.
AR(1)
ACF PACF 2 - -

*l *1 1 -0 .0 9 6 -0.096 0.2033
i** 2 0.271 0.265 1.9334 0.164
*l *1 3 -0 .1 1 6 -0.078 2.2687 0.322
I* I 4 0.076 -0.008 2.4232 0.489
*1 I 5 -0 .0 9 9 -0.049 2.7024 0.609
*1 *1 6 -0.125 -0.175 3.1852 0.671
**! **| 7 -0.257 -0 .2 6 0 5.3801 0.496
I I 8 0.019 0.051 5.3928 0.612
I I* 9 0.047 0.189 5.4826 0.705
*1 **| 10 -0.178 -0.259 6.8945 0.648
I I 11 0.040 -0.050 6.9748 0.728
*1 *1 12 -0.187 -0.132 8.9579 0.626

(1)
ACF PACF 2 - -

I* \* 1 0.100 0.100 0.2306


|*** 2 0.365 0.358 3.4838 0.062
I *| 3 -0 .0 5 0 -0.127 3.5491 0.170
i *| 4 0.058 -0.068 3.6417 0.303
*1 I 5 -0 .1 3 9 -0.088 4.2051 0.379
*1 6 -0 .1 6 9 -0.182 5.1071 0.403
*1 **| 7 -0.277 -0 .1 9 9 7.6991 0.261
*1 1* 8 -0 .0 6 6 0.094 7.8603 0.345
I 1* 9 - 0.021 0.159 7.8772 0.446
*1 **| 10 -0 .1 8 7 -0.302 9.4191 0.400
i 1 11 0.002 -0.042 9.4192 0.493
*1 *1 12 -0 .1 8 7 -0.104 11.338 0.415
7. . ARMA 363

P - Qbp 0.05,
, s,
, . ,
,: -
0.480 0.608.

.
.
, ,
() .
, /,, ..., t , ...,
.
- ,
.
()
.
(Lomnicki, 1961).

, ~ ),
1 t =1

, Xt
, Gx G2


D(Gl) = ^ f dp \ k ) , Z>(G2) = ^ | > 4(*).
* =
=

()
(). EXG\), DiG2),

* _ ^1 ' - ^2
^1 I > 2 ;------
/ ) y]D(G2)
,
.
, >
(G;) 2 +(G*)2 * j 2(2).
364 2.

, ,
G2*
. , (
). ,
(G* )2 + (G*2)2,
AR , ..
.
,


ARMA ,
,
, N(, 1).

D(G1) = | , D(G2) ~ ,

( () = * 0 ),
Gj G2

= ^ <-)
1 t \


1
mb
Tf z


,2 , 2 J G,2 G 2 ^
(G,)z +(G 2 ) = 7 1 + -^ -
24

,
(Jarque, , 1980).
,
(
),
, , , ,
.
7. . ARMA 365

- EViews
( - ),
, .
, .
,
.
?

(, q) ARMA, ,
,
.

, ,
. ,

,

.

J 7.2.1.
ARMA
,
, .. et et
.

0 ,
.
ARMA
,
, .

j 2(2) (. (Yu, 2007)).

7.2.3
(1) AR(1)
, ,
0 : , = 0 AR(1) 0 : = 0 (1)
. , (0)
X t =/J + t.
, . 7.11.

( 7 . 1 .2 )
366 2.

7.11
(0)


/- -

10.81000 0.092594 116.7460 0.0000

R-squared 0.000000 Mean dependent var 10.81000

Adjusted R-squared 0.000000 S.D. dependent var 0.414094

S.E. o f regression 0.414094 Akaike info criterion 1.123258


Sum squared resid 3.258000 Schwarz criterion 1.173045
Log likelihood -10.23258 Durbin-Watson stat 1.138735

AR(1) (1). ,
( 1) ( \ \
,
,
.
, ,
.
, ,
, ,
.
(Sargan, Bhargava, 1983). ,
0.05 = 21 1.069.
, ,
(0).
(0), (1) AR(1)
(. 7.12).
7.12



MA(0) MA(1) AR(1)
AIC 1.123 1.098 1.081
SIC 1.173 1.197 1.180

AR(1),
(0).

: ,
.
7. . ARMA 367

7.2.4
7.2.1
, = 1.2,_, - 0.36,_2 + -
, , ()
. AR 4, 3, 2 1
AIC SIC .

(. 7.13).
7.13
X
Method: Least Squares; Sample (adjusted): 3 500;
Included observations: 498 after adjusting endpoints; Convergence achieved after 3 iterations

/- -


0.001015 0.330958 0.003067 0.9976
AR(1) 1.256580 0.041257 30.45723 0 .0 0 0 0

AR(2) -0.397095 0.041290 -9.617188 0 .0 0 0 0

:
ACF PACF g - -

1 1 1 -0.003 -0.003 0.0042


1 1 2 -0.005 -0.005 0.0165
1 1 3 0 .0 0 0 0 .0 0 0 0.0165 0.898
1 1 4 0.036 0.036 0 .6 8 6 6 0.709
1 1 5 0.037 0.037 1.3675 0.713
*1 *1 6 -0.086 -0.085 5.0736 0.280
1 1 7 0.005 0.005 5.0882 0.405
1 1 8 -0.004 -0.006 5.0977 0.531
1 1 9 - 0 .0 0 2 -0.004 5.0993 0.648
1 1 10 -0.054 -0.050 6.5887 0.582
1 1 11 -0.014 -0.008 6.6897 0.669
1 1 12 0.019 0.011 6.8676 0.738

P- QBP 0.05,
, st
, . t
(- 0.616).
Xt ,
AR(2)
. ,
, . 7.14.
368 2.

7.14
X
Method: Least Squares; Sample (adjusted): 3 500;
Included observations: 498 after adjusting endpoints; Convergence achieved after 2 iterations


/- -

AR(1) 1.256581 0.041215 30.48807 0 .0 0 0 0

AR(2) -0.397096 0.041248 -9.627056 0 .0 0 0 0

S.E. o f regression 1.036707 Akaike info criterion 2.91398


Sum squared resid 533.0816 Schwarz criterion 2.93089

Inverted AR Roots 0.63 - 0.05/ 0.63 + 0.05/

.
AR(2), AR(3) AR(4)
P - ,
AR(2)
(. 7.15).
7.15
AR



* ,-2 * /-3 * ,-4

AR(2) 1.26 -0.40


AR(3) 1.25 -0.39 P - 0.87
AR(4) 1.25 -0.40 P = 0.72 P = 0.56

,
,
, ,
.
AR(2) .
, , .
, AR(2),
, a(z) - 0, .. 1 - 1.2z + 0.36z2 = 0,
z = ^-1.67. 1,
, .
, ,
7. . ARMA 369

, ,
.
0.63 0.05/, z = 1.58 0.125/.
, , () a(z) = 0
, 1 ,
, AR(2) .

1. ARMA
? ?
2. ARMA?
3. ?
4. ?
5.
?
6. ,
AR(p)?
7. ,
(#)?
8.
?

-7

,
, ,
.
,
.
(randomness test),
.
() , , # 0 (
) (null hypothesis).
,
.
, ,
.

(runs test) .
F , F(M) = { < } = { > ) = 1 - F(M) = .
,,...,
.
370 2.

F, ,
.
, ..., , ..
( ): *(1), ..., {).
:
,

med = med(xx,...,) = < \


: xt +, xt > med, -, xt < med.
(run)
.
2, + 2= ,
zx z2 z = zx + z2 .
zx z2
Zj Z2. 0 , Z=ZX+ Z2


= 21{2,1 -)
2( - 1)
, 2 ,

y[D(Z)
N(0, 1), 0
.

(turning points test)


, S
xt
.
, ,
, .
(turning point).
,
, .
t= 1,2,..., - 2

{1 xt < xt+x (+ > xt+2 xt > xt+l xt+l<xt+2,


.
7. . ARMA 371

Z, = 1 , +1 ,
-2
S = ^ Z,.
/ =1
S :
-2 -2

0 , F ,


E(S) = (n - 2 ) |.

S 0

jD (s j
0 , k N(09 1).
0 ,
.

/ -7-1.
, , ,
, .
.
(
) .

().
,
.
, ,
.

(Kendall test)
. (/,/), 1 <i<j <,
0 xi < Xj,
1 xf > Xj,
372 2.

.. Htj = 1 , xi9 ,
, .. .

= I ,
\<i<j<n

,..., . 0 = 0
^ -C^2n -
< ... < , Q (-\)
- > ... > .

Q = Cl =
,
40 = 1.
( -1)

4Q
= 1-
( -1)
. 0

Z>(r ) = ^ ) ,
9(-\)

=

, > 11
. 0 ,
.

J -7-2. ,
,
.
:
,
. ,
0 ,
, ;

Htj = 0,
Q, () .

7. . ARMA 373

=1, Q.

*,
0.


.
, ,
, , 0
. ,
,
.
,
,
,
.
,

.
(goodness offit tests).
,
, ,
F ,
,
.
EViews ,
,

.
, ...,

1. < , 1, Xt < , 0
.

( ) (Kolmogorov
test, Kolmogorov-Smirnov test). Fn(x)

G(x). F(x) .
0 : G(x) = F(x).

Dn =sup|F(x)-F(x)|.

374 2.

0 Dn ,
. 0 Dn
1. 4 Dn,
. 0 -

, . G(x) = F(jc)
, >
D -> sup | G(x) - F(x)| > 0,
X

-> oo. : 0 ,
Dn .
, 9 ...,
F,
,
yf Dn > Kl-a >
_ ( 1 - ) .
Dn

rk - 1
D = max F(x,k)),
( ) /9 F(x.k))
1<
\

(1), ..., () {, ..., 9


19..., .

(Kuiper test). ,

-\
F =
1

(Lilliefors test)
, 19 ...,
, (
) .

,
,

.
, , ,
,
( ).
7. . ARMA 375

( -) (Cramer-von Mises,
W2-test).
Fn F

W2 = \\F(x)-F{xfdF(x).

,
2 -\
W2 = + Y
2 -*(*(*))

(Anderson-Darling test)
,
, < 25.
Fn F

2= f------- -------- \Fn(x)-F(x)\2dF{x).


JlF(x)(l-F(x))
F ( vYI F ( " 1

Xj, xn .
:
V -x
S

Ik 1 1
2 = - - - [in () + In (1- (+1_*))],
=\
(:) .
, , :


0.05
*2 >0.752.

() (Watson test).
Fn F

= ][f (x )-F (x )- ][f(x)-F (x)]^U )1 dF(x).


i J

F () ,
376 2.


.
:
*2 *2 N
JB = n iL + iL
6 24

,..., ;
( = - 3),



=----, * =
(*,)*
1
. = , - ) .
8

_______ 8.1________

, ,


.

. 6 . 1 , ,
,

( , ', , ),
.
,
.
AR(p)
,= + ,_ + 2,_2 +... + ,_ + t = 1,2,..., ,
, , D(et) = <rs2.

y , = x * 9 + et,

, =&,-\,,->--- ,-)> = ( , ,2, . . . , ).


378 2.

,
. ss x t s > t,
s = t - 1, st_x
y t_x, ,
.
, ' .
, s = (sl9 s2, sn)T
X
N (, cr2V) ()
V ( 9 ).
x t+1 = >y t-p+i)T et
.
AR(p) ,
. .
D
y t
y t = a + axy t_x + a2y t_2 + ... + apy t_p + et (st );
1 - axz - a2z 2 - ... - apz p = 0
;
et ~ i.i.d., E(st) = 0, D(st) = 2 > 0, E(ef) = ju4 < .

6= (, 9 2,..., ),
,
nl/\e - e ) ^ N ( 0 ,< T 2Q-1),
Q ,

t;
cr2Q~x , -
( 1
S 2 -- ,
I )
,
, 9 S 2{X^Xn)~x)
(
).
,
,
.
8. 379

,
, ,

.
.
,
, = a + fit +
, ~ i.i.d., (,) = 0, D(s,) = 2 > 0, (*) = < .
:
, = x j + s, xt = (\,t)T, = (, /3),
,
= (, /?),
; ( - ) 1/2, (/? - ,) 1.
,
/?. , /-
7V(0, 1). ,
,
.

,
.

,
, = + /3t + axy t_x + a2y t_2 + ... + ,_ + st (s, );

1 - axz - a2z 2 - ... - apz p = 0


;
,~ lid .,(,) = 0, D(et) = <
2 > 0, (,) = ju4 < .
/-
qF ( q , a F
F- )
N(0, 1) %2(q).
,
.
,
,
,
.
380 2.


= + , = ,
:
,= ] + t - \ , 2,...,,
, = (, , ..., ,)
/- .

- 2, ..., ), .
- :
\
1) \\ =0

( : ) = 0);
/, \
2) lim =Q
-1
( : plim(n lX^X) = Q), Q
;
' 1
3) \-*N {Q ,a2Q)
yjnt =1 J
( : (~^2*) -> N (0, <20 ) .
lim ;
.

, , D,

(Mann, Wald, 1943) :


/in
lim 1 X-1
=
V /' = 1

( : p\im(n X n) = Q), Q
;
s,~i.i.d.,E{st) = 0, D{et) = 2 > 0, E(\et\m) < m = 1, 2, ...;
(*, () = 0 ,/= 1,2, ...,,
8. 381


,
-> 00.
, (, ,) = 0, t = 1, 2, ..., , (,) =
,
Cov(xtk,s ,) = 0, = \,2 ,..., ,
.. ,
. E(\st\m) < 00 - 1,2, ...
, ,
.

F

y t = x je + st, t = \,2 ,...,n ,
*, = (x,i, , ..., ,)
t- ;
9={, , ..., 0),
.
:
1 "
lim =Q
( : \\( ' ^ ) = Q), Q
;
,~ i.i.d., (,) = 0, D(s,) = (J2> 0, {\,\) < = 1, 2 ,...;
Cov(xtk, t) = 0, k = 1, 2, ..., .

->
^ i 0 n - e ) ^ N ( O ,c r 2Q -').
, xt (Jf-)
(-dimensional stationary time series),
E(xt) = / - const, Cov(xt) = Q9 Cov(xtk,xt+sJ) = ykl (s)

/, s , I = 1, 2, ..., . ykl(s) -
(cross-covariance) - /-
s .
382 2.

s , ykI{s) s , ykl{s)
- (cross-covariance function) - /-
,. F

Cov(xt) = Q ,]
/= 1 ,2 ,..., .
F
ARX :
, = + 2,-2 + + ,- + zJp +

Z = ( Z, \ , Z, 2, . . . , Z, M ) T, = { , 2 , . . . , ) .

F,

, = (yt-\>yt-2> >yt-p>zi>z2>>zm)T>
6 = {0\,a2, ...,a p, P\,P2, iPm)
:
z, (-) ;
^
^lim ~1 VL 1 ztzt =&>
\ t=i
Q2 ;
,~ i.i.d., E(s,) = 0, D(st) = 2 > 0, E(\st\m) < 00 = 1 ,2 ,...;
Cov(zlm, ,) = 0 = 1, 2 ,..., ;
Cov(yt_j, st) = 0 = 1,2,..., ;
1 - axz - a2z 2 - ... - apz p = 0
.
F, >
4 { 9 - < ? ) - > N ( Q , (t 2Q ~
2 x).

,
a(z) = 0, ARX. ,
(.. t)
(long-run) ^,, zfl, z<2, ..., xtW
.


1.

?
8. 383

2.
, ,
?
3. ARX?

_________ 8.2 --------------


.


( )
ARX
(
, autoregressive distributed lag models ADL).
A D L (, ; s):
, = 0 + 1,_1+ 2,_2 +... + ,_ +
(\0\, t , f-1 + + \\, 1- ) ^ ^
"*"((,( PslXs,t-\ + + PsrXs,t-r)^ t
p ,;
x l t , 2< ..., xs t,
,',
s .
, ,
,
xL,.
ADL(, ; s) :
-bj {L)x1 , +... + bs(L)x$' t + st,

a(L) = 1- a xL - a 2L2 apLp,

bj(L) = 0 + Pt\L + + PirLr> i =


, , :
384 2.

, + + ... + cs{L)xSyt
a(L) a(L)

a(L)
, L =1
st = 0

, = ^ 7 ^ + q ( lK ,+ - - - + cs(l)x
(1)
, t
:
= -^ / + \{\) +... + ,(1),.
( 1)
^(1), ..., 1)
(long-run multipliers, equilibrum multi
pliers). ADL(1, 1; 1),
:
( \ - a {L)y, = /+ 0, +/?,*,_, +,.
\{\< \

( = ,----- ^ +
77
(1 - a xL )
-----77 ( + P \ x t - 1+ t \
(1 - a xL )
..
y t = (1 + + ...)// + (1 + + c%\I? +...) (J3qX( + /3\Xt_\ + t)?
:

oxt

_ , _
dxt ,_
dy,+2 _. tyt _ \\ + ct\ 0,
, ,_2
,+ _ tyt _

^ Xf_2
@ 1 1

.. (impact
multiplier, short-run multiplier),
8. 385

() ,
^,. ,

, , for , , , Qy, , _
dxt dxt_x dxt_2 dxt_3
= P q(1 + (Xx + (Xx 4- ) + /?! (1 + (Xx + CCX + ) =

= ( l - a 1)"1(A>+A)-
,
ADL(1, 1; 1).
y t
1 ,.
ADL ,
.
,
(
. 2.1, F):
/- N(0, 1) ;
F F -
q , qF
^- q ;

t qF
() (
/-, F-).

;


et , .
,
d
d - 2,

( ,).

8.2.1
ADL(3, 2; 1):
(1 - 0.51 - 0.1Z2 - 0.05L3) y t = 0.7 + (0.2 + 0.1L + 0.05L2)xt + ev

L = 1 st = 0:
386 2.

(1 - 0.5 - 0.1 - 0.05) = 0.7 + (0.2 + 0.1 + 0.05),


.. 0.35j = 0.7 + 0.35, = 2+.
. 8.1 , = 0.7:, +
sxt ~ i.i.d., ?(0, 1), .,,
ADL(3, 2; 1), , ~ i.i.d., N(0, 1), ,
ext.
: , = 0, y t =2 =3 = 0.
, ,
. , = + /, + ,
,
, = 1.789 + 0.577, + , .
. 8.2.

. 8.1 . 8.2

PACF g - -

1 0.696 0.696 49.981 0.000


**** I* 2 0.536 0.099 79.868 0.000
*** *1 3 0.364 -0.081 93.801 0.000
** 4 0.227 -0.056 99.279 0.000
I
I 5 0.130 -0.015 101.100 0.000
I I 6 0.057 -0 .0 20 101.460 0.000

,
- 0.0000. , ,
.
.
8. 387

xt :
ACF PACF g - -
**** ***** 1 0 .6 8 6 0 .6 8 6 48.468 0 .0 0 0

*1 2 0.429 -0.079 67.594 0 .0 0 0

1* *1 0.193 -0.132 71.527 0 .0 0 0

1 *1 4 0.024 -0.066 71.591 0 .0 0 0

1 1 5 -0.058 0.003 71.958 0 .0 0 0

*1 *1 6 -0.140 -0.107 74.090 0 .0 0 0

, AR(1).
, :
ACF PACF g - -
1****** \ 0.767 0.767 60.58 0 .0 0 0
|***** |* 2 0.629 0 .1 0 0 101.75 0 .0 0 0
!**** 0.494 -0.042 127.37 0 .0 0 0
1
1*** 0.399 0.019 144.32
1 4 0 .0 0 0
1** 0.318 -0.003 155.21 0 .0 0 0
1 5
* 1 6 0.257 0.004 162.38 0 .0 0 0

AR(1).
ADL
, 1:
, = + ,-\ + + /? -1 + , .
ADL(1, 1; 1) . 8.1.
(-
AR(1) 0.164),
, (-
0.267),
(- 0.159), ,
,
,
t- F-.
8.1
Y


/- -

0.558588 0.157276 3.55163 0.0006
- 1) 0.695204 0.066095 10.51828 0 .0 0 0 0

X 0.208971 0.126135 1.65673 0.1009


( - 1) 0.161690 0.132352 1.22166 0.2249
388 2.

0 : /?0 = = 0 --
- = 0.0032,
R%2(2) - = 0.0022.
.

-, , . 8.2.

8.2
Y

-
/-

0.517868 0.154098 3.360648 0.0011

Y (- 1) 0.719738 0.063131 11.40064 0 .0 0 0 0

X 0.310343 0.095241 3.258511 0.0015


Schwarz criterion 3.024602

,
.
x t_l9 a x t9
(. 8.3).
8.3
Y

-
/-

0.567821 0.158600 3.580215 0.0005
- 1) 0.692523 0.066673 10.38690 0 .0 0 0 0

( - 1) 0.305939 0.100582 3.258511 0.0030


Schwarz criterion 3.037497


x t_l9 . ,
.
,
(1 - 0.720L)yt = 0.518 + ., +
L {= 0, :
0.28, = 0.518 + 0.310
= 1.839 + 1.107*.
, , .
, , ADL(3, 2; 1)
(. 8.4).
8. 389

8.4


/- -

0.539617 0.174057 3.100239 0.0026
Y(r 1) 0.590293 0.105583 5.590795 0 .0 0 0 0

Y(r 2) 0.153936 0.120517 1.277303 0.2048


-3 ) -0.031297 0.099814 -0.313555 0.7546
X 0.205570 0.129483 1.587626 0.1159
(-1) 0.191959 0.153608 1.249666 0.2147
{-2) -0.024779 0.138389 -0.179053 0.8583

R-squared 0.643818 Mean dependent var 1.882787


Adjusted R-squared 0.620072 S.D. dependent var 1.706160
S.E. o f regression 1.051648 Akaike info criterion 3.008022
Sum squared resid 99.536670 Schwarz criterion 3.193826
Log likelihood -138.8891 F-statistic 27.11327
Durbin-Watson stat 1.999213 Prob. (F-statistic) 0 .0 0 0 0 0 0


, ,
= 1.882 + 1.300,
, ,
. ,
.


(VAR vector autoregression).
y 2t
:
= 1 + 7.\\, -1 + 7\2.\2^-\ + \

~ /^2 + ^21.1 y \ , t - 1 + ^22.1 2, t- 1 + 2


..
^ - \, 2 ,_i
2 s2t :
Cov(sjn els) = 0 t s , /= 1 ,2 ;
Cov(sjn t_r) = 0 > 1 , / = 1,2.
390 2.


s2t .

n y 2t
. , ,
. ,
VAR(p) (pth-order VAR).
1 2 ...,
,
:
= \ + - 1 + .2 -2 + +

+ \2.\ 2* - 1 + ^ 1 2 .2 ^ 2 , t- 2 + + \ 2 . 24 - +
+ ...+

+ 7\.\,1-\ + 7 \ .2 ; -2 + * * + \ . ^ - +

2( ~ t*2 + 2\.\\,-\ + 2\.2\,t-2 + * * * + 2\.~ +

+ ^*22.1^2,/-1 + 22.22, t - 2 + + 2 2 . 2 ^ - +

+ +
+ 72.\,1-\ + 2.2^-2 + 2 . ^ - + S 2 t

= V k + * \ - \ + -2 + + \ . - +

+ 2 .\ 2 * - \ + 2.2 2,t - 2 + + 7 2 . 2 ; - +

+ +
+ * -1 + 7 .2,1-2 + ' - + 7.,1- + k t

nij r ^ ,_
1 2 ..., kt ,
Cov(jn ls) = 0 t * s , / = 1,..., ;
Cov(jn y lt t_r) = 0 > 1 , / = 1 , ;
Cov(Jn lt) .
eXt9 2 ..., kt t =
= (1 2 ..., kt)T9
lt9 2 ...,
()

y t - 1 = ( y \ , t - \ 9 - 2 >> - > ^ - \ 9 ^ - 2 9 > ^ - ) '
8. 391

8.2.2
VAR(l) ( = 2, = 1):
= 0.6 + 0.771(_, + 0 -2^ 2, /-1 + i/>
21 0.4 + 0.2_1(_! + 0.7^2, -1+ 2
, 2
t = 2,3, ..., 100, . 8.3.
= 21 = 0; s2t
, N(0, 0.12).
. 8.4 ( - y 2t).

. 8.3

. 8.4

, :
. . 8.4
,
2, 0.4 (
( - y 2t) 0.403). y lt, y 2t
VAR.B
392 2.

,
VAR , , .
VAR(p) :
=\ + ^ ,-! + 2,_2 + ... + Tlpy t_p + t.

= 0 W 2 r *J '*,)r , = { \ , 2 , . . ; t * k ) T, t = ( l t , 2l, . . . , kt) T ,

= () ( ) y h t_r, y 2f ,_r, >


, t - r .

yt - ,7,_1 - 2y t_2 - ... - ,_ = +
( h ~ ,L - U 2L2 - ... - ), = + t,

AL)y, = jU+ t,

A(L) =I k - n xL - 2L2 - ... - pLp.

VAR :

det(/* - - z 2 2 - ... - z T lp ) = 0 (.. det A(z) = 0)
(..
1).
,
,
, 2, , .
L = 1 , = 0.
),=>

, =A~\l)ju.

8.2.3
VAR(l) :
= 0.6 + 0 . 7 ^ , + 0 .2^2 ! + ,
2, = 0.4 + 0.2 , + 0.72 , + 2
:
Jr = >W+ n iJ f_1 +
8. 393

' ^0.6^ "0.7 0.2^ '


, !* = 5 Hi -
,=
,; ,0-4, 0.2 0.7, . >=

(), = V + v

"1 0" '0.7 L 0.21? 1-0.7Z, - 0 .2


A{L) = I 2 - Y lxL =
,0 K0.2L 0.1 Lj V
-0 .2 L \-0 .1 L

0.3 -0.2
A( 1) =
,-0-2 0.3, ,
det A(z) = 0 :
( l- 0 .7 z -0.2z^l
det A(z) = =0,
^ -0 .2 z l-0 .7 z J
.. (1 - 0.7z)2 - (0.2z)2 = 0, (1 - 0.9z)(l - 0.5z) = 0. z = 1/0.9
z = 1/0.5 1, .. .
()
'6 4 0."| (5.2
,= ~ \ 1)// =
4 6 0.4 4.8
,
= 5.2, 2, = 4.8,
(yXt - y 2t)
\ 2 ~ 0.4.

5.2, y2t
4.8; ( - y 2t) 0.4.
VAR(l)
.

, , ,
VAR (closed VAR),
VAR (open VAR),
,
( endogenous variables),
( exogenous
variables).
394 2.

,
, ,
,
.
. ,
VAR .
VAR :
A(L)y, =/j + B(L)xt +e
A(L) B(L) ,
A(L) = I - ,! - 21'} - . . . - 1.

detA(z) =
,
,

y t = A -\L )n + C(L)xt + A -\L )e


C(L) = A~l(L)B(L) (transfer function);
C(L) ,
.

(, long-run)
,
L = 1 et = 0.

y t = ^ _1(1)// + C(l)xr
(1) .
(/, /)- ctj( 1)
xjt .

8.2.4
VAR(l)
VAR:
y lt = 0.6 + 0.7yl t_x + 0.2y2t_l + 0.1*! , _ j +0.22 + elt9
,

y 2t = 0A + 0.2yUt_l + 0 J y 2' t_l + 0.2xl't + 0A x2tt-i + b9


A(L) , ,

0.2 "


(N

^0.1 0' "0.1 L


+ L=
<N
(N

0-4 0.4 Z,
8. 395


0.1 0.2
(1) = 0+ 1=\
0.2 0.4
:
'6 4 0.1 0.2 '1.4 2.8^
(1) = -1(1)(1) = |
4 6 0.2 0.4 1.6 3.2

"5.2" "1.4 2.8"
= +
,1.6 3.2, <2/

,
..
= 5.2 + 1.4^ + 2.8:2,
2 = 4.8 + 1.6*! + 3.2*2-

, x2t AR(1) :
= 0.7, + vlf, 2, = 0.52 + v2l; v v2, ~ i.i.d. N(0, 1).

1: jcn =21 = 0, =2, =0 (. 8.5),
2: xu = 21 = 0, y lt = 5.2, 21 = 4.8 (. 8.6).

1 2

. 8.5 . 8.6

-
,
.
.
396 2.

VAR(1) :
= ^ ,- \+ 0 .2 2<,_1+ ,
y 2t = 0.2yl,t_l + 0 S y 2',_! + % .


"0.81 0.2L"
,0.2L 0.8L,
+
(


^1-0.8 L -0 .2
A{L) =
-0 .2 L 1-0.8 L

0.2 - 0.2
( 1) =
,-0 .2 0.2
, ^ (1) .
det A(z) = 0 (1 - 0.8z)2 - (0.2z)2 = 0, ..
(1 -z )(l -0.6z) = 0.

1. , 1,
0.6
.
? . 8.7.

. 8.7

. ,
- _ = 21 = 0.
,
: _ =21 - 5. -
8. 397

. 8.8

. 8.8. - ,

VAR .

.
,

, ,
ADL(1,1; 1)

, = + \^\ + + \,-\ + ,
, ,
.

= (19 /?0, )-
,
. ADL(1, 1; 1)
.
.

1. ( = = 0):

yt=M + Poxt + er
y t xt ,
t_ x t_x >>
398 2.

,
, , ,
, .
2. (0 = = 0):

,= + <*\,- 1+ ,-
, ,_, ,
t (t - 1) ,.

- ,
,

,.
3. (, = = 0):

y t = + \,_\ + ,.

,

.
.
,
, .
, -
.
4. ( = 1, = - ):
, = ^1 + 0, +
= 1 - L, , =y t - y t_u Axt =xt - x t_x.
, ,
( ).
,
.
, ,

.
5. ( - 0):

' = + + - 1 +
.
, ,
xt xt_x.
8. 399

6. ( = 0):

= ^ + - 1 + *, + ,.
.
, , .
* = + fix, ,
, , -y,_i
y t ~,- 1 = ) (t ~>-1) + 0 < < 1,
..
, = { \ - ) ; + ,_1 +

, ,
* . (
, , , , .)
yt =yt-i + (1 ~ )( + p x , - y t_x) + , = { \ - ) + ,_1+(1 - )/, +

, = + \!-\+ , +
= { \ - ), , = , , = (1 - )
-
, ,_,
{, 1 .
7. , (0 = 0):

y t = M + ^ iy ,- \+ Pix,-i + r
, , , = ,_, + ,.
, ADL(1, 1; 1)
, = + \{- 1 + < + fiixt-i + t = M + a i,- 1 + (A*- + ) */-1 + (t + >ut \

, = + 1,-1 + AV1 + *
(
) , .
, ,
( ADL(1, 1; 1)).
8. ( = - , 0):

, = + axy t. 1 + < - 1 <- 1 +


:
, - 1- 1 = (1 - \) + (* - 1) +
400 2.


,
:
,= + J30x, + u , = ,_ + | ,|< 1 .
9. (l^il < + \ {\ ]), b ^ 0):
&yt = M + fio&xt - ( l - a l)(yl_]- b x t_i) + et,

, = , , - (1 - )(,_ - - ,_ ) +

- , L=
+ Al .
1-, 1-

.

= +
, , ,_ - - ,_
.

> / 8.2.1. () ADL(1, 1; 1)


< + *,- 1 +
, = + , , _ +
:
yt- y t- 1=M-(l-a\)y,-i+0o(x,-xt- l) +(0o+ - 1 + ,
|i| < 1 ( ),

+
, =M+0o*xt - (1 - 1) ,-1 " -----L^-1
\- ,
0 + 0 .
, \\ < 1 0 + 0
.

.
,
(data generating
process DGP). , -

, (statistical
model SM), , ,
. ,
8. 401


,

.

.
SM
DGP, SM .
,
SM,
.
, ..
, ,
,
.

8.2.5
( = 100) ADL(1, 1; 1)
, = 0.5>,_, + 0.2, + 0.3,_, + s, ~ i.i.d. N(0, 0.12),
, = 0.5,_, + v v, ~ i.i.d. N(0, 0.52),
, v, .
. 8.9.

. 8.9

ADL(1, 1; 1)
(. 8.5).
,
, . 8.6.
402 2.

8.5
Y
Sample (adjusted): 2 100


1- -

0.014122 0.009556 1.477773 0.1428


(- 1) 0.555208 0.034143 16.26107 0 .0 0 0 0

X 0.188567 0.018421 10.23666 0 .0 0 0 0

( - \) 0.258377 0.020673 12.49808 0 .0 0 0 0

R-squared 0.913395 Mean dependent var 0.062869


Adjusted R-squared 0.910660 S.D. dependent var 0.310554
S.E. o f regression 0.092824 Akaike info criterion -1.876660
Sum squared resid 0.818547 Schwarz criterion -1.771806

8.6

Sample (adjusted): 2 100


/- -

- 1) 0.565569 0.033621 16.82186 0 .0 0 0 0

X 0.190325 0.018495 10.29043 0 .0 0 0 0

{ - \) 0.256578 0.020764 12.35668 0 .0 0 0 0

R-squared 0.911404 Mean dependent var 0.062869


Adjusted R-squared 0.909558 S.D. dependent var 0.310554
S.E. o f regression 0.093394 Akaike info criterion -1.874134
Sum squared resid 0.837363 Schwarz criterion -1.795494
Log likelihood 95.76965 Durbin-Watson stat 2.218619

.
:
;

, (- = 0.375 AR(1) 0.165
AR(2) );

(- = 0.689);
(- =
= 0.285).
8. 403

,
.
,
SM 8 .

SMj : y t = / + >, +
. 8.7.
8.7


/- -

0.271208 0.053356 5.082965 0 .0 0 0 0

R-squared 0.174472 Mean dependent var 0.062241

Adjusted R-squared 0.174472 S.D. dependent var 0.309046

S.E. o f regression 0.280794 Akaike info criterion 0.307561

Sum squared resid 7.805700 Schwarz criterion 0.333613

Log likelihood 14.37805 Durbin-Watson stat 0.839862


.
.
..
.

SM2 : y t = / + axy t_x +


. 8.8.
8.8
Y


/- -

0.013941 0.020679 0.674149 0.5018

(- 1) 0.764874 0.065621 11.65594 0 .0 0 0 0

R-squared 0.536633 Mean dependent var 0.062869

Adjusted R-squared 0.531856 S.D. dependent var 0.310554

S.E. o f regression 0.213344 Akaike info criterion -0.228643

Sum squared resid 4.415033 Schwarz criterion -0.176216

Log likelihood 13.98284 Durbin-Watson stat 1.349170


404 2.


,
.
st , . AR(1)
- 0.00003,
st . ,
.

SM3 : y t = ju + /3xxt_x + r
. 8.9.
8.9


/- -

0.049508 0.019722 2.510238 0.0137


(-1) 0.455497 0.037291 12.21457 0 .0 0 0 0

R-squared 0.652012 Mean dependent var 0.062869


Adjusted R-squared 0.648424 S.D. dependent var 0.310554
S.E. o f regression 0.183939 Akaike info criterion -0.525251
Sum squared resid 3.281850 Schwarz criterion -0.472824
Log likelihood 28.15718 Durbin-Watson stat 1.420075

AR( 1) - 0.0002,
st .
.

SM4 : Ayt = / + fi0Axt + st.


. 8.10.
8.10
D(Y)
Sample (adjusted): 2 100


/- -

-0.001126 0.021384 -0.052674 0.9581


D(X) 0.040538 0.033362 1.215078 0.2273

Log likelihood 13.74152 F-statistic 1.476415


Durbin-Watson stat 1.574116 Prob(F-statistic) 0.227286
8. 405

AR(1) - 0.029,
, .
.
SM5 : , = / + /?0, + ,_, +
. 8.11.
8.11
Y
Sample (adjusted): 2 100


/- -

0.046032 0.018096 2.543741 0.0126


0.156214 0.035435 4.408526 0 .0 0 0 0

{ -\) 0.414363 0.035435 11.69370 0 .0 0 0 0

AR(1) - 0.0000,
st .
.
SM6 : y t = + axy t_x + J30xt + er
. 8.12.
8.12
Y
Sample (adjusted): 2 100


/- -

0.007088 0.015431 0.459354 0.6470

- 1) 0.753212 0.048925 15.39514 0 .0 0 0 0

X 0.253493 0.028588 8.867013 0 .0 0 0 0

AR(1) - 0.012,
st .
.
SM7 : y t = / + axy t_x + J3xxt_x + sr
. 8.13.
- AR(1)
0.499, AR(2) 0.538.
st ,
.
(- = 0.937).
(- = 0.348).
406 2.

8.13

Sample (adjusted): 2 100


/- -

0.020438 0.013757 1.485648 0.1406


- 1) 0.517457 0.048968 10.56734 0 .0 0 0 0

( - \) 0.318058 0.028613 11.11579 0 .0 0 0 0

S.E. o f regression 0.133909 Akaike info criterion -1.153476


Sum squared resid 1.721440 Schwarz criterion -1.074836

,
.
,
. (. 8.14).

8.14


/- -

(- 1) 0.532005 0.048276 11.02001 0 .0 0 0 0

(-1) 0.316252 0.028765 10.99445 0 .0 0 0 0

S.E. o f regression 0.134740 Akaike info criterion -1.150947


Sum squared resid 1.761018 Schwarz criterion -1.098520

.

.

y t - axy t_x + J30xt + Pxxt_x + sn ,
xt (-1.874)
(-1.795) .
SM8 : y t = + axy t_x + J30xt - axfi0xt_x + sr
()
,
. 8.15.
AR( 1) - 0.0002,
st .
.
8. 407

8.15

= (1) + (2)*(-1) + ( 3 ) * - ((2)*(3))*(-1)


/- -

(1) 0.014489 0.020617 0.702743 0.4839


( 2 ) 0.749747 0.070184 10.68267 0 .0 0 0 0

( 3) 0.052577 0.036535 1.439066 0.1534

Schwarz criterion -0.269859

SM (
).

SM9 : , = ju + , - (1 - ^){,_ - ,_ + ,.
( )
. 8.16.
8.16

D () = ( 1) + C(2)*D(X) + ((3) - 1)*((-1) - (4)*(-1))


/- -

(1) 0.014122 0.009556 1.477773 0.1428


(2) 0.188567 0.018421 10.23666 0 .0 0 0 0

( 3) 0.555208 0.034143 16.26107 0 .0 0 0 0

( 4) 1.004839 0.078119 12.86299 0 .0 0 0 0

Schwarz criterion -1.771806

- 1 AR( 1) 0.130,
AR(2) 0.318,
st .
(- = 0.711).
(- = 0.380).
,
. , ,
(. 8.17).
:
Ayt = 0.190,- 0.434(yt_x - \.029xt_x) + et.
.
408 2.

8.17
()
Convergence achieved after 3 iterations; D(Y) = C(2)*D(X) + (C(3) - 1)*(Y(-1) - C(4)*X(-1))


/- -

(2 ) 0.190325 0.018495 10.29043 0 .0 0 0 0

( 3) 0.565569 0.033621 16.82186 0 .0 0 0 0

(4) 1.028710 0.080225 12.82279 0 .0 0 0 0

R-squared 0.812174 Mean dependent var - 0 .0 0 1 1 0 0


Adjusted R-squared 0.808261 S.D. dependent var 0.213288
S.E. o f regression 0.093394 Akaike info criterion -1.874134
Sum squared resid 0.837363 Schwarz criterion -1.795494
Log likelihood 95.76965 Durbin-Watson stat 2.218619

^ ,
= 0.566^.! + 0.190, + 0.253,_1+
:
DGP: y t = 0.5^_! + 0.2xt + 0.3jc,_j + et
:
y t = 0.565yt_l +0.190, + 0.257/_1+
y t
:
y t = 0.5jy, + 0.2x, + 0.3 { ->y = x.
,
SM, DGP:
y t = 0.565yt + 0.190jc, +0.257*, -* = 1.002.
, , SM9 (
):
y t = 0.566^ + 0.190, + 0.253, = 1.021.
, SM7 (
) ,
y t = 0.532yt +0.316, -> = 0.675.
,
SM,
DGP. , SM
8. 409


.
SMj SM6 SM8, SM7.B

, , ,
SM () DGP,
, DGP,
, .
DGP (1) (8),
SM ADL(1, 1; 1) :
, =H + a\yt-\ + Poxt + P\xt- 1 + f
(1) (8)
, ADL(1, 1; 1)
, - 0.5y,_i + 0.2, + 0.3x,_l + ,.
DGP , ~ i.i.d. N(0, 0.12).

DGP,:
, = 0.2; + ,.
. 8.18.

8.18
Y


/- -

-0.004647 0.010300 -0.451175 0.6529


Y(r 1) 0.102848 0.101833 1.009966 0.3151
X 0.186813 0 .0 2 0 2 2 2 9.238033 0 .0 0 0 0

( - 1) 0 .0 0 0 2 0 1 0.028272 0.007101 0.9943

S.E. o f regression 0.102190 Akaike info criterion -1.684398


Sum squared resid 0.992068 Schwarz criterion -1.579545

- AR(1) 0.760,
AR(2) 0.951,
st .
(- = 0.733).
(- = 0.770).
,
. ,
.
410 2.

yt_1 xt_{
F-, F = 0.738, qF = 2.214. --
F, - = 0.532.
^ 2(3) qF - = 0.529.

.
^, = , + st (. 8.19).
8.19

Sample: 1 100


/- -

X 0.190067 0.019291 9.852604 0 .0 0 0 0

S.E. o f regression 0.101512 Akaike info criterion -1.727139


Sum squared resid 1.020359 Schwarz criterion -1.701087

,
.
, .

DGP2:
= 0.5,_, + ,.
. 8.20.
8.20


/- -

-0.004519 0.010315 -0.438075 0.6623
Y(r 1) 0.576756 0.084134 6.855173 0.0000
X -0.013220 0.020253 -0.652774 0.5155
( - \) 0.021476 0.020228 1.061719 0.2911

R-squared 0.338422 Mean dependent var -0.007891


Adjusted R-squared 0.317530 S.D. dependent var 0.123820
S.E. o f regression 0.102290 Akaike info criterion -1.682441
Sum squared resid 0.994011 Schwarz criterion -1.577588

- AR(1) 0.600,
AR(2) 0.773,
st .
8. 411

(- = 0.654).
(- = 0.956).
xt
xt_x -, = 0.641, qF - 1.283.
P- , - = 0.529.
20 )
- = 0.527.
.
y t = axy t_ + st (. 8.21).
8.21


/- -

Y (-1) 0.575922 0.082705 6.963585 0 .0 0 0 0

S.E. o f regression 0.101482 Akaike info criterion -1.727825


Sum squared resid 1.009258 Schwarz criterion -1.701612

,
.

DGP3:
y t - .^! + st.
. 8.22.
8.22


/- -

-0.005202 0.010305 -0.504767 0.6149


(- 1) 0.052373 0.054196 0.966363 0.3363
X -0.012475 0.020310 -0.614213 0.5405
( - 1) 0.315962 0.020645 15.304330 0 .0 0 0 0

R-squared 0.736662 Mean dependent var 0.004035


Adjusted R-squared 0.728346 S.D. dependent var 0.196154
S.E. o f regression 0.102236 Akaike info criterion -1.683501
Sum squared resid 0.992959 Schwarz criterion -1.578647
412 2.

- AR(1) 0.614,
AR(2) 0.868,
, .
(- = 0.740).
(- = 0.804).
,
,_ F-, F - 0.577, nqF = 1.730.
F- F, - = 0.632.
/ 2(3) qF
- = 0.630.
.
, = j3xx,_x + , (. 8.23).

8.23
Y


/- -

X ( - \) 0.315777 0.019302 16.35987 0 .0 0 0 0

S.E. o f regression 0.101572 Akaike info criterion -1.726058


Sum squared resid 1.011044 Schwarz criterion -1.699844

,
.
sr

DGP4:

Ayt = M+ A,Ax, + r
. 8.24.
8.24
Y
Included observations: 99 after adjusting endpoints


/- -

-0.029026 0.017741 -1.636049 0.1051


Y (- 1) 0.959750 0.024228 39.613030 0 .0 0 0 0

X 0.184064 0.019993 9.206520 0 .0 0 0 0

Xi-X) -0.173461 0.020340 -8.528162 0 .0 0 0 0


8. 413

. 8.24

R-squared 0.944972 Mean dependent var -0.599911

Adjusted R-squared 0.943234 S.D. dependent var 0.425076

S.E. o f regression 0.101277 Akaike info criterion -1.702356

Sum squared resid 0.974412 Schwarz criterion -1.597502

Log likelihood 88.26661 F-statistic 543.7988

Durbin-Watson stat 1.745026 Prob(F-statistic) 0 .0 0 0 0 0 0

1 ^,_1?
, DGP : = 1.
. ,
^ , (. 8.10), ,
.

. 8.10

0: = 1
, . ,
^ 1,

,= + * + \ >-1+
()
, , (. 8.11),
, ,
(. 8.12), ,.
(. 8.25).
414 2.

. 8.11 . 8.12

8.25
D(Y)


/- -

-0.004915 0.010297 -0.477336 0.6342
X 0.185224 0.020163 9.186333 0 .0 0 0 0

(-1) -0.179782 0.020163 -8.916410 0 .0 0 0 0

, ,
, , xt xt_x
.

, = + , + \(_j + *,
0: = 0, /?0 = - . F -
F - - 2(2) qF = 2F,
- = 0.876. 0 ,
, ..
Ay, = j30Axt + . 8.26.

8.26
D(Y)


/- -

D(X) 0.182495 0.015882 11.49045 0 .0 0 0 0

- AR(1) 0.328,
AR(2) 0.605,
8. 415

, .
(- = 0.673).
(- = 0.988). ,
,
, DGP.
DGP5:
, = ju + 0.2xt + 0.3jc,_ , + ,.
. 8.27.
8.27
Y

-
/-

-0.003282 0.010206 -0.321584 0.7485


Y (-1) 0.000523 0.058294 0.008975 0.9929
X 0.181735 0.019977 9.097166 0 .0 0 0 0

( - \) 0.289502 0.024922 11.61638 0 .0 0 0 0

R-squared 0.804020 Mean dependent var 0.010665


Adjusted R-squared 0.797831 S.D. dependent var 0.225113
S.E. o f regression 0.101218 Akaike info criterion -1.703515
Sum squared resid 0.973283 Schwarz criterion -1.598661

- AR(1) 0.972,
AR(2) 0.826,
, .
(- = 0.689).
(- = 0.433).
0: / = 0, = 0 . P-
- - j 2(2) qF = 2F,
- = 0.950. 0 ,
ju = 0, , =0, .. , = , + , + ,.
. 8.28.
8.28


/- -

X 0.181461 0.019754 9.185972 0 .0 0 0 0

X { -\) 0.289367 0.019756 14.64736 0 .0 0 0 0


416 2.

DGP6:
;y, = // + 0.5yf_1+ 0.2x, + f.
. 8.29.
8.29
Y


/- -

-0.005099 0.010263 -0.496869 0.6204


Y (-1) 0.592041 0.071296 8.304016 0 .0 0 0 0

X 0.188766 0.020280 9.308077 0 .0 0 0 0

{ - \) 0.000973 0.025019 0.038898 0.9691

- AR(1) 0.904,
AR(2) 0.723,
st .
(- = 0.691).
(- = 0.533).
0: ju = 0, = 0. P-
- - \ 2) qF - 2,
- = 0.884. 0 ,
// = 0, =0, .. ^, = axy t_x + 0, + sv
. 8.30.

8.30
Y


/- -

7 ( - 1) 0.592666 0.056702 10.45233 0 .0 0 0 0

X 0.188468 0.019202 9.814814 0 .0 0 0 0

R-squared 0.690477 Mean dependent var 0.004719


Adjusted R-squared 0.687286 S.D. dependent var 0.180476
S.E. o f regression 0.100924 Akaike info criterion -1.728911
Sum squared resid 0.988000 Schwarz criterion -1.676484

DGP7:
, = 0.5,_ + 0.3,_1 + ,.
. 8.31.
8. 417

8.31
Y


/- -

-0.005886 0.010271 -0.573100 0.5679
(- 1) 0.559497 0.042762 13.08387 0 .0 0 0 0

X -0.010318 0.020320 -0.507807 0.6128


( - \) 0.316645 0.020229 15.65291 0 .0 0 0 0

.- AR(1) 0.701,
AR(2) 0.827,
, .
(- = 0.740).
(- = 0.586).
0: = 0, 0 = 0. -
F- - j 2(2) qF = 2F ,
- = 0.734. 0 ,
/ - 0, 0 = 0, ..
y t = axy t_x + ,_, + ,.
. 8.32.
8.32
Y


/- -

(- 1) 0.561389 0.041849 13.41452 0 .0 0 0 0

( - \) 0.313207 0.019269 16.25422 0 .0 0 0 0

DGP8:
y t = 0.5^_! + 0 . 2 ;, - 0 .1 jt,_ ! + sr
. 8.33.
8.33
Y


/- -

-0.004519 0.010315 -0.438075 0.6623

( - 1) 0.576756 0.084134 6.855173 0 .0 0 0 0

X 0.186780 0.020253 9.222532 0 .0 0 0 0

( - \) -0.093875 0.025414 -3.693770 0.0004


418 2.

- AR(1) 0.600,
AR(2) 0.773,
s, .
(- = 0.654).
(- = 0.682).

(. 8.34).

8.34
Y


/- -

- 1) 0.578309 0.083705 6.908866 0 .0 0 0 0

X 0.186426 0.020151 9.251413 0 .0 0 0 0

X { - \) -0.094531 0.025263 -3.741827 0.0003

, y t_x ,
0.108, ..
,_[. 0: = - , .
,
.
:

0: = - , 0; : 0 = - ^ ; : { = -& -.
1
P - ;^2(1)- 0.515,
0.514 0.506, 0 :
. ,
= -, 0, ..
, = {,_ | + f]()xt - ]{),_, + :
(. 8.35).

8.35
Y
Convergence achieved after 3 iterations; Y = (1)*(-1) + C(2)*X- (C(1)*C(2))*X(-1)


/- -

(1) 0.575812 0.083369 6.906747 0 .0 0 0 0

( 2) 0.182370 0.019110 9.543254 0 .0 0 0 0


8. 419


, = 0.576>,_1+ 0.182, - 0.105:,_, + ,.
, , DGP
,

( SM )
,
DGP. , ,
, ,
DGP,
.

, ..
,


.

,
.


. ,
, ( 8)
ADL(1,1; 1).

, = + 1,-1 + + \,-1 + ,
, ,
fix - - fi0.
fi0 0 ADL(1, 1; 1)
:

(1 - a lL)yt =fi0 ' , + / , + ,, a(L)y, = b(L) xt +et,


.

/
a(L) = \ - a xL, b(L) = , 1+ & -L
. .
( p. = 0.)
, = , :
420 2.

(1 - axL)y, = (1 - ,) ,+
a(L) b(L) (1 - axL).
,

, = +

(1 - axL) ut = st ut = ut_x + st.



, COMFAC (common factors).
-
= - /30. COMFAC
.

y t = , + ut

, , .
:
a(L)yt = b(L)xt + st
; , st
,
;
, a(L) b(L)
;
,
# 0: = 0 ( ).
,

.
.
0: et , ,
: st ~ AR(k) <, ..

t = p xt_x+ ...+ Ppt_p +V


v, ~ i.i.d. Pj 0.
, , :

, = 1,- 1 + + \,-i + P \ t-\ + - + Pps,-p + v


8. 421

0: = ... = 2 = 0
: \ + ... + 0. ,
.
,
( 1.2)
(., , (Kwan, Sim, 1996)).

: = - /30 \ - 0.
, ,
,
: = - a xfi|), = 0 = - , -
\
.
,
.
, ,
, , ,
2: = 0 .
,
, = 0.05.
? :
{ , 2} <
< { } + { 2} =
= + = 2.
, = 0.025,

0.05. , ,
.

1.
?
?
2.
?
3.
?
4. (VAR)?
?
5. ?
() ?
422 2.

6. VAR?
7. VAR?
?
?
8. ,
?
9. (
,
)?
10. ?

?
9

.
ARIMA

________ 9.1________
ARMA


.
.

(GNP gross national product)
I 1947 . IV 1961 . (
60 , .,
). (. 9.1) .

. 9.1

(NONDURABLE)

I 1974 . IV 1985 .
424 2.

NONDURABLE

I I | I | I I I | I I I | I I I | I I I | I I I | I I I | I I I | I I I | I I I | I 1 I

CD Is-
Is- - Is- Is-
) > >

. 9.2

( 48 ,
, ).
(. 9.2).

, -
, .
GNP :

A utocorrelation Partial Correlation g - -

******* 1* * * * * * * | 0.946 0.946 56.419 0 .0 0 0


******* 0.893 - 0.021 107.52 0 .0 0 0
1 2
****** 0.840 -0.024 153.55 0 .0 0 0
1
I****** 0.791 0.013 195.14 0 .0 0 0
1 4
****** 0.743 - 0.021 232.52 0 .0 0 0
1 5
1 6 0.696 - 0 .0 2 2 265.90 0 .0 0 0
***** 0.648 -0.030 295.41 0 .0 0 0
1 7
***** 0.599 -0.044 321.09 0 .0 0 0
1 $
|** ** 0.550 -0.033 343.13 0 .0 0 0
1 9
**** 0.498 -0.052 361.57 0 .0 0 0
1
|* * * *| 11 0.442 -0.073 376.44 0 .0 0 0
j*** 1 12 0.388 -0.034 388.08 0 .0 0 0
|*** 1 13 0.337 - 0 .0 0 2 397.06 0 .0 0 0

1 14 0.291 0.007 403.91 0 .0 0 0

1 15 0.253 0.041 409.21 0 .0 0 0


;** 0.218 - 0 .0 0 2 413.22 0 .0 0 0
1 16
9. . ARIMA 425

NONDURABLE:

ACF PACF A AC 2 - -
j******* 1 0.917 0.917 42.921 0 .0 0 0
j******* 2 0.843 0.014 79.976 0 .0 0 0
1
j****** 3 0.774 -0.004 111.92 0 .0 0 0
1
j***** 4 0.704 -0.040 138.99 0 .0 0 0
1
j***** 5 0.643 0.013 162.09 0 .0 0 0
1
|**** 6 0.581 -0.041 181.36 0 .0 0 0
1
j**** *I 7 0.510 -0.090 196.57 0 .0 0 0
*** 8 0.439 -0.050 208.14 0 .0 0 0
1
J*** *1 9 0.366 -0.066 216.39 0 .0 0 0
j** 10 0.300 -0.012 222.07 0 .0 0 0
1
j** 11 0.246 0.026 225.99 0 .0 0 0
1
1* 1 12 0.196 -0.006 228.56 0 .0 0 0

I* 1 13 0.150 -0.013 230.11 0 .0 0 0

1* 1 14 0.106 -0.025 230.91 0 .0 0 0

1* 1 15 0.072 0.036 231.28 0 .0 0 0

1 1 16 0.041 -0.016 231.41 0 .0 0 0

,
ACF PACF
AR(1). ,

X t = + fit + axX t_x +ur
( ut
.)
1.
GNP . 9.1.
: -
AR(1) 0.0000.
,
. 9.2.
, ,
0.7926,

1 (Q t ()
/
- - ) = 1 (X, - 1 - - r(t ~ 1)) + 2(,_ 2 - - r(t - 2)) + ,
(2 = 0 . 9.1).
. AR(1) , a AR(2) 2.
426 2.

9.1
G N P
Method: Least Squares; Sample (adjusted): 1947:2 1961:4;
Included observations: 59 after adjusting endpoints; Convergence achieved after 3 iterations


/- -

216.0630 11.30237 19.11661 0 .0 0 0 0

5.269279 0.281754 18.70170 0 .0 0 0 0

AR(1) 0.846976 0.072723 11.64665 0 .0 0 0 0

Inverted AR Roots 0.85

9.2
G N P
Method: Least Squares; Sample (adjusted): 1947:3 1961:4;
Included observations: 58 after adjusting endpoints; Convergence achieved after 3 iterations

-
/-

217.7399 5.054473 43.07865 0 .0 0 0 0

5.221538 0.140436 37.18089 0 .0 0 0 0

AR(1) 1.380274 0.109452 12.61078 0 .0 0 0 0

AR(2) -0.630066 0.109453 -5.756490 0 .0 0 0 0

Inverted AR Roots 0.69 - 39z 0.69 + 39/

GNP .
,
X f JLI + y t 4- U t .

. 9.3.
9.3
X


/- -

218.4825 2.640153 82.75373 0 .0 0 0 0

5.181995 0.075274 68.84144 0 .0 0 0 0

Durbin-Watson stat 0.316211 Prob(F-statistic) 0 .0 0 0 0 0 0


9. . AR1MA 427

, ,
:
Autocorrelation Partial Correlation g - -
****** 1****** 1 0.836 0.836 44.028 0 .0 0 0
2 0.531 -0.554 62.115 0 .0 0 0
**| 3 0.183 - 0 .2 1 0 64.294 0 .0 0 0
1*
1 1 4 - 0 .1 0 0 0.044 64.960 0 .0 0 0
** 5 -0.272 -0.004 69.949 0 .0 0 0
1
*** *1 6 -0.339 -0.082 77.846 0 .0 0 0
*** *1 7 -0.350 -0.169 86.446 0 .0 0 0
*** *1 8 -0.332 -0.072 94.332 0 .0 0 0
**| 9 -0.281 0.058 100.070 0 .0 0 0
1
**1 *1 10 -0.234 -0.177 104.160 0 .0 0 0
** ***| 11 -0.234 -0.321 108.320 0 .0 0 0
**| 1* 12 -0.226 0.103 112.260 0 .0 0 0

AR(2).
AR(2) (. 9.4) ()
X, , = X ,-218.4825-5.181995/.

9.4
X detrended


/- -

AR(1) 1.379966 0.107605 12.82435 0 .0 0 0 0

AR(2) -0.630426 0.107605 5.858722 0 .0 0 0 0

X, -218.4825-5.181995/ =
= 1.379966(*,_, -218.4825-5.181995(^-1))-
- 0.630426 (,_2 - 218.4825-5.181995(^-2)),

X, = [(1 -1.379966+0.630426) 218.4825 +


+1.379966 5.181995 - 0.630426 5.181995 2] +
+ (1 -1.379966 + 0.630426) -5.181995/ +13199,_ - 0.630426,_2 +, =
= 55.338375 +1.297882*+ 1 .3 7 9 9 6 6 * ,- 0.630426,_2 +,
428 2.


X t = + J3t + Q \X t_x + 12 ^ t-2

X t -217.7399-5.221538/ =
= 1.380274(JTM -217.7399-5.221538(/ -1 ))-
- 0.630066 (X t_2 - 217.7399 - 5.221538(/ - 2)),

X t = [(1 -1.380274 + 0.630066) 217.7399 +


+1.380274 5.221538 - 0.630066 5.221538 2] +
+ (1 -1.380274 + 0.630066) 5.221538/ +1.380274X,_1-0.630066X,_2 +et =
= 55.17011 +1.304298/ +1.380274X,_1-0.630066X,_2 +etm

, ,
, .

,
, Xt = +
+ fit + ClxXt_x + Cl2Xt_2-\- Wp
,
.
NONDURABLE.
( Xt /)
:
Autocorrelation Partial Correlation g - -

|* * * * * * |* * * * ** -1 0.793 0.793 32.083 0.000


j***** 2 0.632 0.011 52.942 0.000
1
*** **| 3 0.432 -0.195 62.887 0.000
|** **| 4 0.219 -0.193 65.515 0.000

1 5 0.090 0.062 65.965 0.000

*i *1 6 -0.067 -0.152 66.218 0.000


** ** 7 -0.242 -0.277 69.647 0.000
***| 8 -0.362 -0.084 77.505 0.000
*1
**| 9 -0 .5 1 0 - 0.211 93.500 0.000

1,
Xt = + p t + axXt_x +
(. 9.5).
9. . ARIMA 429

9.5
X


/- -

47962.75 2862.678 16.75451 0.0000

315.1909 76.44770 4.122961 0.0002

AR(1) 0.884803 0.080824 10.94727 0.0000

P -
0.05 1 20.
-,
0.05, AR(1) , AR(2),
AR(3) .. , -
0.648 ,

X, - 47962.75 -315.1909/ = 0.884803 (X t_x - 47962.75 -315.1909(/ -1)) + et.
, , .
0.884803 Xt_x 1,
0.080824
,

0.884803 2*0.080824 , 1.
.
,
0.884803 Xt_x
(.. -
AR(1) ),
,
.
,
,
.
1980- ., .

,
,
.
,
( )
ARMA ,
.
430 2.

AR(1)
Xt = axXt_x + et.
,
-1 < , < 1. Xt
?
= 0.5 (. 9.3), = 0.7 (. 9.4), = 0.9 (. 9.5), = 1 (. 9.6), = 1.05
(. 9.7), = 1.1 (. 9.8).

. 9.3 . 9.4

. 9.5 . 9.6
9. . ARIMA 431

. 9.7 . 9.8

= 0
s l9 ..., ,
, 1 (
. 9.9).


.
(. 9.6),

= 0.5
= 1.1. ,

Xt 0.
= 0
( ) = 1

. 9.9 ,
,
.
= 1, 500
(. 9.10),
Xt Xt_ + s n
, , t - 1
Xt = ( = 0),
( ) ,
( ),
.
432 2.

9.6

Noise ( ) 25 -0 .0 4 6
AR(1) ax = 0.5 14 -0.097
AR(1) a x = 0.7 8 -0.191
AR(1) ax = 0.9 8 -0.649
AR(1) a x = 1.0 1 -3.582
AR(1) a x = 1.05 1 -13.511
AR(1) aj = 1.1 1 -59.621

. 9.10

,

,
,

(. 9.11).
, AR(1)
= 1:
X t = X t_t + s t
(
random walk).
,


9. . ARIMA 433

X, \[1 + Et
= 1.05 (. 9.7) , = 1.1 (. 9.8).
(explosive) AR(1) , > 1:

.
-
.
AR(1)
:
X/ ,_ = QxX,_x ,_ + , = (Ct\ l)* f_i + ,,

AX,= (pX,_x +
AXt =Xt - Xt_j, (p = ax - 1.
, = 1 cp = a, - 1 = 0, AX, Xt
, ,
() ,_, = ,_ 0.
() ,_ - ,_
X, = , + ,_
,_!. ,
( ,
), X, - ,
, ,_ .
.
> 1 ( = - 1 > 0, ,
() ,_ =,_, {,\,_ =
= ,_ ) = (,_ , , ,_,. ,
,_ > 0, X, = ,
,_, ,_, < 0,
X, = , ,_ .

, AR(1) - 1.05
= 1.1.
, 0 < < 1 ( - - 1 < 0,
", () ,_, = ,_,,
(,\,_ = ,_,) = (,_, ,
, jc,_ > 0,
X, = , ,_ , ,_ j < 0, -
434 2.

. 9.12 . 9.13

Xt = , xt_,.

, (,) =
( = 0),
.
> 0,
.
Xt = ,_ + , , = - 1 (. 9.12) , = - 1.1
(. 9.13).


Xt =Xt_x + e t = l ,...,T ,
0 =0. X, :
X, = ,_ + S, = ( '_2 + ) + ,= X t_2 +st_1+s, = (,_3 + s,_2) + ,_,+ , =
= *,- +t-2 +et-i+ t =--- = X 0+(el +... + t),

.= 0 +
j =1


( t \ 0 = 0) = 0,

D{X, | 0 = :0) = D(et +... + ,) = D{s,) + ... + D(st) = tD (sx) = tc].


9. . ARIMA 435

,
Cov(Xt,X t_x | = 0) = [(, - 0)(,_{ - 0) | 0 = 0] =
= [{ +... + , )(, + )\ = (t -)<72
0,

,_,) = .
jD ( X t) j D ( X t_x) V /

t Corr(Xn X t_ x) t Corr(Xn X t_x)

1 0 6 0.913
2 0.707 7 0.925
3 0.806 8 0.935
4 0.866 9 0.943
5 0.894 10 0.949

.. X, ,_, ,
, t.
, .
,
0.

0 (
), .
,
,
0.
^ -

= 1

(
),
E(Xt) = 0, D(X,) = t a 2,
, .
(stochastic trend),


,
.
436 2.

,
,
, (, )

,
, , (,
) .
.

Xt = 4- (it + t, t = 1 , ..., ,
..
.

(random walk with drift), ..
( + Xf_ j 4- t, t 1, . . . , , X0 = Xq,

E(kXti) = oc*Q.

^ :

X t -b X 4- St = + ( 4- X f_2 + t_x) + t 2 4- X f_2 + + t

= 3a 4- X t_ 2 4- t_ 2 +f-\ +t = ... = Xq + at 4- (x 4-... 4- t),


t
X t =x0 +at + ^ j ,
j =1

X, , .

X ? = X t - ( a + fit) = et,
.

, = ,-( 0 + = 2 > ,
=1

.
.
Xt

AXt = X t - X t_v
9. . ARIMA 437


(differencing), (difference
operator).
:
, = , -,_= ( + pt + , )- (a + Pit - 1 ) + ,_) - + ,
:
, = , - ,_ = +

, .
(1)
. (0)
.
, ,
.
,
.


(., , (Hamilton, 1994, . 72 151)).
,
X, 0Q+ 0xt + + ,

Y, = + p t + y t1 +Z,,
Z, ,
Z, =, +2 ,_+ ,_2 +... + tx, t = l,...,T .

= ,- (0 + 6xt + e2t2) = ,.

Y ,= Y ,-(a + p t + y t2) = Z

D(Z,) = D{e, +2t_x +3f,_2 + ... + te,) = cr2(l + 2 + ... + /) = cr2 ^ ,

, .

, X,
A X t = 62 + w 2t + st t_i,
438 2.

:
AXt - ( - 0 2+ 2d2t) = t - t-\
, (1) .
Yn

AYt = ( f i - y + 2yt) + Zt - Z t_x =
- 2yt + (st + 2 st_x + _2 +... 4- t s ^)
- (,_+ 2et_2 + 3s,_3 +... + (t - l)f,) =
= P + 2yt + (fj + S2 + ... + t_\ + 8t)
:
AYt - ( - + 2yt) = ( 1 +2 + + ,)
.
,
.. A2Y 2 = (1 - L)2 = 1 - 2L + L2,
(0)
2, = 2 + tfZ , = 2+ Z, - 2ZM + Z,_2 =
= 2 + ( , + 2 s,_i + ,_2+ . . . + ts j)
2 (ff,_ j + 2 ,_2 + + . . . + ( / 1 )^ ]) +

+ ({_2 + 2ff,_3 + t_4 + . . . + (/ 2)ffj) = 2^ + , .


X,

2 ,= ( ') = ( ,- 1_1) = (, - , . ) - { _ - , _ 2) =
2,_ + X t_2 = 2 ^ + Et 2,_, + _2,

(2) 2,
. , b{z) = 0 :
1- 2z + z2 =0
z = 1.
,
X,
.
.
Xt
/(/), X, - f(t) . X,
, ,
9. . ARIMA 439

,
(trend-stationary time series),
5- (TS trend-stationary). -
, .
Xt (integrated o f
order ) 9 = 1,2, ..., :
Xt
, .. TS-;
^-
;
Ak~lXt, ( - 1)-
, TS-.

AXt = = 1 .
1().
Xt ,
X t ~ 1(). Xt ~ 7(0)
,
'-.
TS- Xt = + (3t + Yn Yt ,
. Xt
:
00
X ^ a + p t + Y j p f r . j , /0 =1, 2 > } < ,
= 0 =

s t .
(

.) ,

, = A f ( t ) + f jV,j(el_j - t_i_j ) = P + 'Zbit_j = + b{L)st,


7=0 j =

00
b(L) = bjLj , b0 = 1, bj= y/j - / H , j = 1,2,...
j =
,

6(1)= ][> ,.= ,


=

.. b(z) = 0 .
440 2.

, Z,
Z, =ju + b(L)et,
00
b(L) = b .U , b0 = 1, 6(1) = 0,
j=o
(overdifferencing) Zt \
- 5-.
= 1 ,2 ,...
, /^- (DS difference
stationary time series) . X, ,
DS-.

ARIMA.

X, . ^-
.
ARMA(p, q), , X,
(, , q), ARMA(p, q) (ARIMA
autoregressive integrated moving average). p - 0 q = 0,
:
ARIMA(p, , 0) = ARI(, ) , ARIMA(0, , q) = IMA(, q),
ARIMA(0, , 0) = ARI(0, ) = 1(, 0).
,
1) Xt - + fit + st -;
2) Xt - + X,_j + t ~ 1(1), Xt ARIMA(0, 1,0);
3) Xt = + 0xt + 02t2 + t TiS-;
4) Xt = a + fit + yt2 + t + 2.t_! + 3,_2 + ... + t{ ~ 1(2), Xt
ARIMA(0, 2, 0).
r.S'-,
DS-.
, ,
:
TRENDJ, = 1 + 0.5/ + , ~ N(0, 1) (. 9.14),
WALK, = 0.5 + WALK,_, + , ~ N(0, 0.52), WALK0 = 0 (. 9.15).
9. . ARIMA 441

. 9.14 . 9.15


(. 9.7).
,
. 9.16 9.17.
9.7


/- -

TREND1
0.796390 0.208085 3.827226 0.0002

0.501522 0.003577 140.1946 0.0000

WALK
-0.930832 0.249804 -3.726248 0.0003
0.437818 0.004295 101.9477 0.0000

. 9.16 . 9.17
442 2.

. 9.16 ,
. 9.17 .
. 9.17.
{Chan, Hayya, Ord, 1977):

, ( )

.
,
. 9.18 9.19.

X TREND DIF WALK DIF

11111 11111II11111 11111n |n 111111r| 1111111n 111n | 11 | 111111 111111111 11 111111111111111111


10 20 30 40 50 60 70 80 90 100 t 10 20 30 40 50 60 70 80 90 100 t

. 9.18 . 9.19

.
(.X_TREND_DIF):
ACF PACF k AC g - -

***! *** | 1 -0 .4 4 9 -0 .4 4 9 20.527 0.000

I **l 2 -0.045 -0.308 20.736 0.000

I "I 3 -0 .0 0 6 -0.2 36 20.740 0.000

I **l 4 -0.052 -0.266 21.021 0.000


I* 5 0.078 -0.157 21.676 0.001
*I
*l ** | 6 -0.074 - 0.221 22.258 0.001
I* 7 0.194 0.073 26.358 0.000
**| 8 -0 .2 1 6 - 0.120 31.473 0.000
*l
I* I 9 0.079 -0 .0 47 32.159 0.000

I *l 10 -0.0 56 -0.143 32.513 0.000


9. . ARIMA 443

( WALK DIF) :
ACF PACF g - -

1 1 1 0.035 0.035 0.1271 0.721


1 1 2 -0.044 -0.045 0.3271 0.849
1 1 3 -0.042 -0 .0 3 9 0.5099 0.917
1 1 4 -0.025 -0.024 0.5766 0.966
1 1 5 0.065 0.063 1.0215 0.961
1 1 6 -0.004 - 0.012 1.0231 0.985
1* 1* 0.107 2.1405 0.952
7 0.101
*1 * I -0.173 -0.181 5.4243 0.711
8

1 1 9 -0.041 -0.013 5.6087 0.778


1 * 1 10 -0.057 -0.073 5.9707 0.818

.
,

(1) . (1) = -0.449
( 1) ( 1) = ^ , - 0.449 = - - .
1+ 6, 1+ 6,
: -1.6036 -0.6236. ,
(1) . ,
6, -0.6236.
(1)
: . 9.8
(hackcasting . . 7, 7.2) . 9.9
.
9.8
X_TREND__D!F
Sample (adjusted): 2 100; Included observations: 99 after adjusting endpoints;
Convergence achieved after 10 iterations; Backcast. 1


/- -

0.501429 0.004483 111.8447 0.0000

(1) -0.977743 0.015377 -63.58459 0.0000


, 1, (1)
, . ,
,

444 2.

9.9
X_TREND_DIF
Method: Least Squares; Sample (adjusted): 2 100; Included observations: 99 after adjusting
endpoints; Convergence achieved after 18 iterations; Backcast: OFF


/- -

0.518807 0.007616 68.12224 0.0000
(1) -1.062852 0.042816 -24.82378 0.0000

. (Slutsky
effect . (Slutzki, 1937)).
:
X, = 0.01/2 + et9 ~ N(0, 52) (. 9.20),
Yt = 0.04/2 + t + 2s t_l + 3st_2 + ... + ts x, s t ~N(0, 0.12) (. 9.21).

. 9.20 . 9.21


( X
) (. 9.10).
9.10


/- -


2 0.009926 0.000114 86.93333 0.0000

Y
-2.273387 0.621988 -3.655036 0.0004
-0.119781 0.028427 -4.213709 0.0001

2 0.013087 0.000273 47.99344 0.0000


9. . ARIMA 445

. 9.22 . 9.23

. 9.22 9.23.
X_DETRENDED :
ACF PACF g - -

1 1 1 0.030 0.030 0.0936 0.760


1 1 2 -0 .0 4 9 -0 .0 5 0 0.3429 0.842
1 1 3 -0.042 -0.039 0.5278 0.913
1 1 4 -0.024 -0.024 0.5867 0.965
1* 1* 5 0.071 0.068 1.1208 0.952
1 1 6 0.000 -0.008 1.1208 0.981
1* 1* 7 0.101 0.107 2.2429 0.945
*1 *1 8 -0.171 -0.177 5.4698 0.706
1 1 9 -0 .0 3 9 - 0.011 5.6366 0.776
1 *1 10 -0.050 -0.068 5.9165 0.822

Y DETRENDED
AR(2) :
ACF PACF g - -
******** 1******** | 0.985 0.985 99.989 0.000
******* **** I 2 0.956 -0 .5 0 7 195.02 0.000
******* ** I 3 0.912 -0 .3 1 4 282.52 0.000
******* ** I 4 0.857 -0 .1 9 6 360.52 0.000
****** 0.000
*1 5 0.791 -0 .1 3 6 427.67
|****** *I 6 0.716 -0.073 483.36 0.000
|***** 1 7 0.635 -0.044 527.61 0.000
j**** *1 8 0.549 -0 .0 6 7 560.97 0.000
|*** 1 9 0.458 -0.043 584.49 0.000
|*** 1 10 0.365 -0.051 599.58 0.000
446 2.

. 9.24 . 9.25

Xt Yt . 9.24 9.25.
X_DIF2
:
ACF PACF g - -

*****| ***** 1 -0.635 40.777


1 -0.635 0.000
**** I
2 0.128 -0.463 42.440 0.000

* * I
1 3 0.038 -0.294 42.586 0.000
** I
4 -0.064 -0.273 43.014 0.000
*1
*I 5 0.081 -0.141 43.710 0.000

** I 6 -0 .1 3 7 -0.292 45.710 0.000
*1
** 7 0.243 0.037 52.065 0.000
1
**| 8 -0.254 -0.037 59.064 0.000
1
1 9 0.160 0.057 61.891 0.000

*1 1 10 -0 .0 9 9 -0 .0 5 6 62.976 0.000

(1) = -0.635
( 1) (1), ..
-0.5 < ( \ ) < 0.5.
-0.635 ( 1)


0.63562 + + 0.635 = 0,
.
9. . ARIMA 447

Y DIF2 :
ACF PACF g - / -

1 1 1 0.031 0.031 0.0942 0.759


1 1 2 -0.041 -0.042 0.2636 0.877
1 1 3 -0.038 -0.036 0.4127 0.938
1 1 4 -0.013 -0.013 0.4313 0.980
I* 1* 5 0.073 0.071 0.9925 0.963
1 1 6 -0.006 -0.012 0.9958 0.986
1* 1* 7 0.104 0.111 2.1713 0.950
*1 *1 8 -0.172 -0.178 5.3769 0.717
1 1 9 -0.031 -0.006 5.4804 0.791
*1 *1 10 -0.066 -0.084 5.9673 0.818

:
TS-
;
DS-
DS-;
75- ^-;
-
ARMA, 5-
, ;
^- X, ~ 1()
; 1()
ARIMA, ^-
ARMA.
, 5-
, /35-

( ) .
AR(1) *, = aXt_x + s t= 1, .
(. . 7, 7.1):
X t = a 'X q+ 1 + 1 22 + ... + ,,

V +h v I *+ -1_ I . , _ , , _
X t+^ ^ a X q + o , 2 + . . . + + &t+h'

,
() ()
st (shock innovation) :
448 2.

, h :
dXt.u ,
\\ < 1,
dst
dXt .h .
^ =1 = 1.
8st
, > 1 ( ),

0S,

.

.

1.
?
?
?
2. ?
?
? ?
3. ,
, , (,
) ?
4. -?
5. -?
6. ARIMA?
7.
?

------------- 9.2________
TS- DS-.



()
. ,
,
9. . ARIMA 449

( ), ^-,
, (,
)
^-1 , DS-.

, 75-
, DS-

- .
,
,

. TS- DS-
,
TS DS
,
( ,
, ).
,
>-, -
(cointegration) (. , . 11, 11.1),
, :
;
;

.
, /^-


(error-correction model),
,
.
.
(Maddala, Kim, 1998), (Enders, 1995), (Hamil
ton, 1994).
TiS- ,
, , ,
-. DS-,
( ),
, - DS-
(

1 .
450 2.

), , ,
,
, .
-
, DS- .


,


, -
.


,
(, ARMA) , , ..
.
ARIMA
()
.
, ,
,
.
, >-
() ,
( ),

. ^- -
, ,
,
,
(
). ,

(.,
, (Hamilton, 1994, . 4, 5)).
, ,

,

, , ..
(TS DS).
,
, , (Maddala, Kim, 1998),
9. . ARIMA 451


.
,
,
( ,
),
,
.

, , (Maddala, Kim, 1998), (Enders, 1995), (Hamil
ton, 1994).
,

rS- /35-.
, 14
(Nelson, Plosser, 1982)
(Perron, 1989). 14
TS, , ,
11 . ,
75-.
5- ,
, .
, ,
, (. (Zivot, Andrews,
1992)).
(. (Bierens, 1997)). , (Nunes,
Newbold, Kuan, 1997) :
(Zivot, Andrews, 1992)
14 , (Nelson,
Plosser, 1982).

,
, (TS
DS) .

TS- DS- ARIMA.



,
Xt TS ( )
DS ( )
.
. , ,
, ,
452 2.

() ,
. ,
,
,
.
,
,
.
TS DS ,
,
( ),
.
TS- AS-
.
, DS- 5-,
ARMA ( ).
Xt ARIMA(/?, k, q),
AkXt ARMA(/?, q)
,
a \L )A kXt = b(L)60
a*(L) b(L) Z,
q .
, AXt = (1 - L)Xn
AkXt = ( \ - L ) kXn

a \ L ) ( \ - L ) kXt = b(L)6n

a(L)Xt = b(L)sn
a(L) = a*(L)(\ - L)k ( + ).
AkXt , a*(z)
, a(z)
, ,
z = 1 .
, Xt ARMA(p+ , q \
() , 7,
1.
0 , ARMA Xt ^S'- (
), ,
a(L) , 1.
, , a(z)
9. . ARIMA 453

, .. .
0 (UR
unit root hypothesis),
.
- , ARMA .
, () TS, ,
, AS-.
a(z)
z = 1 b*(z) = 0, b*(L)
L
AXt = b*(z)st AXt - X t - Xt_x
Xv

1.
TS- AS-?
2. ARM?
10


TS- DS-

_______ 10.1________
-


.
,
,
.
TS-
xt = + J3t + t_x + t, | qx| < 1.

? + St , a y t ,
>>, = xt - - St, y t

,=*1-1+>
., y t_x ,
xt - - St = axxt_x S(t 1)) + sn

xt = ( y - + 8 ) + S(l - )t + axxt_x + et,

= - - 8 = S(l - ),
_ - f l i ( + )
1 - , (1-,)2
, ,

- ( + ) | f
(1-,)2 1-, '
10. TS- DS- 455

, (3 = 0 0


=
\-

= 1, ,
,.
, = + [3t + + st =
= ( + pt + et) + (a + P( t - \ ) + e,_x) + ... + (a + f3 + sx) + x0 =
\ 2 ,

= + + t + t +(sx+e2+... + s,).
2
= = 0
X, = ,_ + X, =0 +(, + s2 +... + S,).
0, /?=
xt = a + xt_x +st, xt =x0 + a t + (x +e2
.. 0 + at.

, 0, /

xt = + fit + jc(_| +t, , - 0 + +

, -

0 + ( ~ \ t .+r .
2

,
,= + p t + axxt_x + ,,
0, = 0, :
\\ < 1 , ;
t
= 1 , ^ f
/=1
0 + at;

0, :
, , , - - - + ) .
I . I < 1 . ------------ + ------ 1;
(1 - ,) 1-
456 2.

= 1 ,
j =1

0 + \\t,
. t
2J 2
,
, = + fit + axxt_x + st
, :

ST_1: , = 0.8, = 0, /3=0 (. 10.1).


ST 2: [ = 0.8, = 0.2, /3=0 (. 10.2).
ST_3: , = 0.8, = 0.16, /?=0.04 (. 10.4).
W A LK 1: , = 1, = 0, /3=0 (. 10.3).
WALK 2: = 1, = 0.2, /3=0 (. 10.5).
WALK 3: = 1, = 0.2, /?=0.1 (. 10.7).
, ,
, = + fit + y t2 + axxt_x + st :
ST_4: a, = 0.8, a = 0.12, fl=0.13, y= 0.01 (. 10.6).

. 10.1

_1, STJ2, WALK X


ST_3 WALK_2, ST_4 WALKJ3 ,

,
.
(
) .
10. TS- DS- 457

. 10.2 . 10.3

. 10.4 . 10.5

. 10.6 . 10.7
458 2.


, , (
, DGP data generating process), :
DGP: xt = axxt_x +et,
st ;
(1 < < 1),

SM: xt = axxt_x +et9 t - 1,..., .
xt , ,
. 8 ( 8.1),
,
,
. , ,
.
1958 . (White, 1958),

, .
, , , :
DGP :, = + ,, * = 1,2,..., , 0 =0,
( , ), .. = 1,
( )

. ,
.
, ,

0: = 1,
, , - /-,
( !).
. = 1,

I { [ ^ ( l) ] 2 -l}
(- 1 ) - > ^ ---------------,
j'[W(r)fdr

W(r) (standard Brownian motion),
-
10. TS- DS- 459

, ,
, .
W(r)

X, = ,_, +,.
, :
(0) = 0;
(2) - )), ..., (W(rk) - W{rk_,))
, 0 < < 2 < ... < ', W(s) - W(r) ~ N(0, s ) s > ;
) 1.
, , , 1) = 1) - ) ~ N(0, 1),
[W(l)]2 ~ !) ,
{ < 1} = {, -1 < 0} * { [W(I)]2 -1 < 0}= />{ j 2(1) < 1}= 0.68.
, DGP ( ), DGP:
, =(_, + SM: , = <,_, + , , < 1
2/ .
({ - 1)
= 1 (-
), (Fuller, 1976).
, , ~ N(0, 2), , ...,
0 = 0. :
0 ( - 1),
( - 1) .
, ( - 1), 0: = 1
/. < 1 5%-
( - 1), (] - 1), ,
5 . 10.1.
10.1

&1 1) ^1

25 -7 .3 0.708
50 -7 .7 0.846
100 -7 .9 0.921
250 -8 .0 0.968
500 -8 .0 0.998
00 -8.1
460 2.

. 10.1, 0
, 1.
.
, 0: , = 1
: < 1 , ,
1.
, ,
0: = 1 : < \
1 . ,
4
/ = (/-, /-ratio, /-),
5(,)
s(ax) .
, = 1
, /- .
/- >
(Fuller, 1976).
0: = 1 : < 1
. 5%-
, , 5%-
/-,
t( T - 1) ( - 1) (. 10.2).

10.2
f-

' () * ()

25 -1 .9 5 -1.71
50 -1 .9 5 -1 .6 8
100 -1 .9 5 -1 .6 6
250 -1 .9 5 -1.65
500 -1 .9 5 -1.65
00 -1 .9 5

7 1
10.2 ----
s(tfi)
- 1. , ,
,
,
.
10. TS- DS- 461


, = ,_, + s, (WALK I).
, = ,_ + ,,
50 (. 10.3).
10.3
WALK_ 1
Sample (adjusted): 2 50; Included observations: 49 after adjusting endpoints


/- -

W A L K J(-l) 0.970831 0.035729 27.17224 0.0000

/- :------------ = -0.816 -
J 0.035729
1.95.
5%- .
0.970831 ,
0.846.
ST_ 1
, = 0.8,_, + , ( )
, = ,_ + ,,
50 , , . 10.4.

10.4
ST_1
Method: Least Squares: Sample (adjusted): 2 50;
Included observations: 49 after adjusting endpoints', Convergence achieved after 2 iterations


/- -

AR(1) 0.790557 0.090501 8.735349 0.0003

/- -2.314 < -1.95,


. ,
0.791 1.
S T 2 AR(1)
, = 0.2 + 0.8(_, + , ( 1).
50 0.596.
, , ,
, , ,
. 0: , = ,_ + , (
) : , = + ,_ + < 1, 0.
462 2.


SM: xt = a + axxt_x + st.
{ - 1)
,
SM = 0, = 25 { < 1} = 0.95. . 10.5
.
10.5

1-1) ^1 ' ()

25 -12.5 0.500 -3 .0 0
50 -13.3 0.734 -2 .9 2
100 -1 3 .7 0.863 -2 .8 9
250 -1 4 .0 0.944 - 2.88
500 -1 4 .0 0.972 -2 .8 7
00 -14.1 - 2.86

SM: xt = + axxt_x + et
WALK_ 1 ( 50 ) :
/ = -2.143 > / =-2.92,
,
, =,_1 + .
S T 2 ( - 50)
/ = -2.245, ,
.
, , , , = 0.794
0.734.

/ 10.1.1. SM:
, = + Oj ,_ j + , ST I ( = 0),
, = 0.785 > 1 = 0.734, t = -2.298 > / = -2.92,
ST_\ .
, , ST_ 1
SM: , = , + st,
.
,

( ) ,
10. TS- DS- 463

,
.
, ,
.

ST 3
, = 0.16 + 0.04/ + 0.8,_, + ),
0.2/ (. 10.8).
WALK 2 (. 10.9)
0.2
xt = 0.2 + xt_x + st.

. 10.8 . 10.9

,

# 0: , = + ,_! + * 0, ( )

SM:x, = + fit + axx,_x + ,.
/-
= 1 .
DGP: , = + ,_ + , * 0, 5%-
(- 1) /- . 10.6.
SM: , = + fit + ,_ + ,
( = 50).
WALK. 7. = 0.858 > 1 = 0.604, / = -2.027 > / = -3.50,
.
464 2.

10.6

1 - 1) \ ' ()

25 -1 7 .9 0.284 -3 .6 0
50 -1 9 .8 0.604 -3 .5 0
100 -2 0 .7 0.793 -3.45
250 -21.3 0.914 -3.43
500 -21.5 0.957 -3.4 2
00 -2 1 .8 -3.41

ST_3 = 0.733 > , ^ = 0.604, t - -2.687 > / = -3.50.


t , WALKJ1,
,
.
, , ST 3,
, DGP .
t
SM: , = + j3t + ,_ + ,
,
DGP: , =*,_! + ,.
,
/,
DGP: , = + ,_, + 0,


SM: , - + jit + axxt_ + ,
t DGP: , - ,_ +
DGP: , = + ,_ , + 0 .
SM: , = + fit + ,_, + ,
DGP: , = ,_[ + s WALK_ 1.
, ( DGP: , = 0.2 + ,_, + ,)
= 0.858 > 1 = 0.604, t = -2.027 > tKpm = -3.50,
.
, , (MacKinnon,
1991),
/- . ,
10. TS- DS- 465

?( ) /- ,
,
^ ( ^ + ' + 2,
, , 2 , ,
.
= 0.01, 0.05, 0.10.

10.1.1
. 9 ( 9.1)

I 1974 . IV 1985 .

:
Xt - 47962.75 - 315.1909* = 0.884803(^_! - 47962.75 - 315.1909(*- 1)) +

Xt = 5804.037 + 36.30898 *+ 0.884803^_! +


, ,
0.884803 wp\iXt_x

(.. AR(1) ),
,
.
.
,

SM: xt - + fit + axxt_x + sr
0: xt = + xt_x +
EViews,
( ,
):

Test Statistic - 1.425277 1% Critical Value -4.1630


5% Critical Value -3.5066
10% Critical Value -3.1828

0 .
, ,
xt = + fit + axxt_x + xt = + xt_x + st.
Axt = + et (. 10.7).
466 2.

10.7


/- -

236.8958 80.80998 2.931517 0.0052


jc, = 236.8958+ jc,_! +,.


:
SM: , = ,_ +
SM: , = + ,_ +
SM: , = + fit + ,_ + s,

/- = 1
:
DGP: , = ,_! + , ( ),
DGP: , = + ,_ + , ( ).
, , SM: , = + ,_ + ,, , = *,_,+ st

0: = 0, = 1.

, F-
RSS0 - RSS

(T -l)-2
0 ^-
F(2, - 3) ( - 3) . , ,
# 0 , ,
, , ,
0 ( ) F(2, - 3).
(Dickey, Fuller, 1981).
, 0: = 0, = 1.
. 10.8 5%- ,, -
10. TS- DS- 467

10.8
(5%-)

4>i1 F

25 5.18 3.44
50 4.86 3.20
100 4.71 3.10
250 4.63 3.00
500 4.61 3.00
00 4.59 3.00

, ( ) 5%-
FKpiiT, F(2, - 3) (., (.Hamilton,
1994, . .7 Case 2) (Enders, 1995, . )).

10.1.2
WALK_ 1 ( ),
ST_2 ( AR(1)
), ST_ 1 ( AR(1)
) 50 .
SM: , = + ,_ + , WALK_l, = -0.579,
= 0.850, RSS = 48.0335. = 0, , = 1
, = ,_,

RSS0 = ( x f - ,_,)2 = 52.7939,
1=2
52.7939-48.0335
, = ------ 48 0335------ = ^ < .
5 0 -1 -2
ST 2 : = 0.181, = 0.777, RSS = 52.6618.
= 0, = 1 RSS0 = 59.0547,
59.0547-52.6618
, = ------ 3 ------ = 2.853 < 4.86 -> 0 .
52.6618
5 0 -1 -2
ST_ 1 : = -0.042, = 0.785, RSS = 52.7007.
= 0, = 1 RSS0 = 58.0671,
, = 2.662 < 4.86 0 .
468 2.

, ,
/-
, F -.


SM: xt = + fit + alxt_l + et
DGP: xt = xt_x + st
H0: a = P = 0, ax = 1,
DGP: xt = a + xt_, + s,
H0:/3 = 0, ax = 1.
F- 2, 5%-
. 10.9 (. (Enders, 1995, . )).
10.9
(5%-) F-

25 5.68
50 5.13
100 4.88
250 4.75
500 4.71
00 4.68

F- 3, 5%-
. 10.10 (. (Hamilton, 1994, . .7 Case 4)
(Enders, 1995, . )).
10.10
(5%-) F-

*
^3

25 7.24
50 6.73
100 6.49
250 6.34
500 6.30
00 6.25
10. TS- DS- 469

10.1.3
WALK_ 1 ( ), WALK_2 (
), S T 3 ( AR(1),
).
SM: , = + fit + ,_, + ,

0: = fi = 0, , = 1, 2;
0: = 0, , = 1, 3.
0: = = 0, , = 1
WALK X: = -0.854, = 0.012, , = 0.858, RSS = 46.7158.
RSS0 = 52.7939,
2 = 1.995 < 5.13 > 0 .
WALK_2: = -0.711, = 0.040, = 0.858, RSS = 46.7158.
RSS0 = 52.7939,
2 1.995 <5.13 > 0 .

S T J : = -0.345, 0 = 0.070, , = 0.733, RSS - 50.3928,


2 = 1.207 <5.13 > 0 .
0: = 0, = 1
W ALK J: RSS = 46.7158. RSS0 = 52.7282,
3 = 1.973 < 6.73 0 .
WALKJ2: 3 1.973 0 .
ST_3: RSS = 50.3928, 3 0.711 -> # 0 .

,


.

. , ,
, < 0.8, 1 > 1. 2
3 . , /-
( - 1).
,
.
470 2.


(
) .
, ,
.
,
.
, ()
.


50
:
ST X S T 2 :
DGP: , =,_, +
SM: xt = + ,_ + st\
ST 3 :
DGP: xt = + ,_ + , ( DGP: , = jcf_, + ,),
SM: , = a + /3t + axx,_x + ,.
,
, .
= 100. (
. 10.11 10%- ).
,
.

/1 = 50 /1 = 100

STJ / = -2.298 > / = -2 .9 2 , / = -3.238 < t = -2 .8 9 ,



ST2 / = -2.245 > / = -2 .9 2 , = -3 .2 1 7 < ^ = -2 .8 9 ,



STJ / = -2.68 7 > / = - 3 .1 8 , / = -3 .2 0 7 < / = -3 .1 5 ,



10%- 10%-
10. TS- DS- 471


(GNP) I 1947 . IV
1961 . . 9
:
X, - 217.740 -5.222/ = 1.380(,_, - 217.740 - 5.222(/ -1)) -
- 0.630(Xt_2 - 217.740 - 5.222(/ - 2)) + ,,

X, = 55.017 + 1.304/ + 1.380,_, - 0.630,_2 + ,.



, ?

. .

SM: , = + jit + ,_, + 2,_2 + ... + ,_ + ,.

:
, = a + /3t + pxt_l+ (<91,_1+... + _1'_+1)+ (10.1)

= +2 +... + , = -(a j+x + ... + ).

( GNP
X, = 55.017 + 1.304/ + 1.380,_,-0.630,_ + 0.630,_1-0.630X,_2 +st =
= 55.017 + 1.304/ + 0.750,., + ^^ + )
, a(z) = 0
z - 1, ( - 1)
, , , + 2 + ... + = 1,
.. - 1 (., , (Hamilton, 1994, . 517)). ,
AR(p)
0: = 1 (10.1).
,
( - 1) /-
- 1, (augmented)
(10.1) ( /? , ).
/- ADF (augmented Dickey-Fuller)
DF, AR(1).
472 2.

/-
/7=1, (10.1)
, = +fit + (,_ + (,_ + ... + _,_+) + (10.2)

= - 1, 0: - 1 (10.1)
0: - 0 (10.2).
0: - 1 (10.1)
: < 1. (10.1) (10.2)
: ( < 0.
/- # 0: = 1
(10.1) /-
0: = 0 (10.2).

10.1.4
GNP
Axt = + p t + (pxt_x + e{Axt_{ + et
( ) ,
. 10.12 (
).

10.12

ADF Test Statistic -4.117782 1% Critical Value 4.1219


5% Critical Value -3.4875
10% Critical Value -3.11718

( D(X))


/- -

- 1) -0.249792 0.060662 -4.117782 0.0001

D (X {- 1)) 0.630066 0.109453 5.756490 0.0000

56.32136 13.18303 4.272264 0.0001

@TREND(\941:l) 1.304300 0.315357 4.135949 0.0001

: /-
0: = 0 5%- ,
, 1%-
.
10. TS- DS- 473

,
,
, , ,
, .
,
-
.

- * , ,
/-
< = 0 ,
, ..., 0_, .
, 9*_ = 0,
/- /-
. ,
0*_j = *_2 - 0, F- F-
, ..
.

10.1.5
. * = 5,
(. 10.13) (
).

10.13
* = 5

ADF Test Statistic -2.873575 1% Critical Value -4.1314


5% Critical Value -3.4919
10% Critical Value -3.1744

( D(X))


^- -

- 1) -0.266169 0.092626 -2.873575 0.0060


D (X (- 1)) 0.546230 0.133521 4.090958 0.0002

D (X(- 2)) 0.183918 0.149711 1.228486 0.2253


D (X (- 3)) -0.020254 0.152201 -0.133077 0.8947
D (X(- 4)) -0.058683 0.148061 -0.396345 0.6936
59.45556 19.32396 3.076779 0.0035
@TREND( 1947:1) 1.397409 0.482120 2.898469 0.0056
474 2 .

t = -2.873575 > -3.1744,


10%- .

. - F -
0.44.
,
, .

,
(,
). , xt
ARMA(p, q )c q > 0 1
xt ~ ARMA(p, q) , a(L)xt = b(L)st, a(L), b(L)
p q, b(L) ,

c(L)xt = st,

c(L)X: = = 1+ + c2l} +...


' b(L) 1 2
(10.2)

Axt = + /3t + cpxt_x + (0Axt_x + 02Axt_2 +...) + st. (10.3)

. ?
(Said, Dickey, 1984, p. 599 607) ,
ARIMA(/7, 1, q) q
ARI(p*, 1) /7* < / .
(10.3)
.


, ,
(Dickey, 1976), (Fuller, 1976), (Dickey, Fuller, 1979), (Dickey, Fuller, 1981).
() ,
xt DS (/^-);
TS (-).
,
(,
). , -
10. TS- DS- 475


.
.
1. xt (
),

SM: Axt = + flt + (pxt_\+ et9 t = 2,..., ,
DGP: Axt = a + s t, *= 2,..., , * 0.
st ,

.

SM /- t9
0: = 0.
t , ,
DGP (
). DS- , ^
, ,
, (Fuller, 1976), (Fuller, 1996),
= 25, 50, 100, 250, 500.
,
, , (Mac
Kinnon, 1991).
2. xt (
) ,

SM: Axt = + <pxt_x + st, t - 2 , ..., ,
DGP: , = st, t = 2,..., .

SM /- t9
0: ( = 0.
^? ,
DGP (
). DS- , t9 < tx.
, ,
, (Fuller, 1976), (Fuller, 1996),
= 25, 50, 100, 250, 500.
,
, , (MacKinnon, 1991).
476 2.

3. , , (
)
,
SM: , = (pxt_ + sn t = 2 ,
j ..., ,
DGP: , = t - 2,..., .
SM
/- / 0: - 0.
/,
,
DGP ( ).
^- , f < f . ,
, ,
(Fuller, 1976), (.Fuller, 1996),
= 25, 50, 100, 250, 500.
, ,
, (MacKinnon, 1991).


. ,
,


, ,
(. (Perron, 1988)).
,
.


(Dolado, Jenkinson, Sosvilla-Rivero, 1990), . (Enders,
1995). ( (Enders, 1995, . 251 260))
.
a(L)xt = st ( ) , a(z) = 0

, (augmented)
.

Axt_j9j = 1, ..., - 1,
:
- 1
1) SM: , = + J3t + (pxt_x + ^ 6 j lsxt_j +*,, t = /? + 1,...,;
/=i
10. TS- DS- 477

- i
2) SM: , = a + q>xt_x + ^6^,_/ + ,, / = + 1,...,;
j =1
- i
3) SM: , =(px,_x + ^ j ej Ax,_j + ,, t - + \,..., .
7= 1


/- / 0: ( = 0
/, . DS--
, / < /.
,
,
. (Said, Dickey, 1984), ,, ...,
ARIMA(p, 1, q) q > 0,
ARI(p\ 1) = ARIMA(p*, 1, 0) * < l/
.
, ...,
AR(p) ,
,
,
.
- p mm ,
0 , , *
,
, .
, ,

( 10%- ) (GS-
) ()
< p mm
(SIC). (Hall, 1994) (Ng, Perron, 1995) , p mm > 0,
( ) SIC ,
GS > 0,

.
, , (Cheung,
Lay, 1995).
(GS SIC),
,

,
478 2.

. ,
(GDP)
1870 1994 . (. (Murray, Nelson, 2000)), = 8,
GS- = 6, SIC
= 1.

LM- (. (Holden, , 1994)).
, , (Taylor, 2000) ,
Ng Perron:

, . , ,

.

. (Dolado, Jenkinson, Sosvilla-Rivero, 1990)



,
(SM)
(DGP). (
(Enders, 1995)), ,
AR(1), , .
1
, ,
:
SM: Axt = + flt + p x ^ + st9 / = 2, ...,,
,

DGP: Axt = a + st, / = 2, ...,7\
: ().
DS
(UR Unit Root) .
0: = 0,
: << 0.
/-
, ,
.
, = 0 * 0.
10. TS- DS- 479

0: <= ,
. ,
0: <=0
DGP: , = + , ( = 0 0),

DGP: , = + fit + 0 0,
(
).
2. 1 0: <= 0 ,
:
, <=0;
0, 0: <= 0 - ,
DGP = 0,
DGP: , = + fit + 0.
2
0: = 0
SM: , = + p t + (,_ j + ,,

DGP: Ax, = a + p t + s .
/- (
) (Dickey, Fuller, 1981). . 10.14
5%- |/|
t .
10.14
(5%-) f-


( > 0 )

25 3.25 2.85
50 3.18 2.81
100 3.14 2.79
250 3.12 2.79
500 3.11 2.78
00 3.11 2.78

0: = 0 , , 1-
( = 0 -
, DGP = 0.
480 2.

0: = 0 ,
= 0
SM: Axt = + fit + cpxt_ +

DGP: Axt = + fit + sn 0.
/- ( * 0)
N(0, 1) .
(Kwiatkowski, Schmidt, 1990). , <= 0
, . ,

Axt = + p t + st9 0.
3. 3,

SM: Axt = + fit + <pxt_x + sr
, -
t.
3
SM: Axt = + (pxt_x + st
q> = 0 ( ( < 0)
SM. /-
( 2). ,
DGP: Axt = sr
0: - 0 ,
( , 1).
4. 3 = 0 ,
SM .
= 0
SM: , = + (pxt. x +

DGP: Axt = st.
/- (
) (Dickey, Fuller, 1981). . 10.15
5%- | / 1
/ .
= 0 , ,
(- 0 DGP = 0.
10. TS- DS- 481

10.15
f-



( > 0 )

25 2.97 2.61
50 2.89 2.56
100 2.86 2.54
250 2.84 2.53
500 2.83 2.52
00 2.83 2.52

= 0 ,
0: (- 0
SM: Axt = + (pxt_x +

DGP: Axt = a + s t * 0.
/- - 0 ,
, , 0: = 0 ,

Axt = + st 0.
, ,
, ,
- 0, .
5. , 4 0: - 0 ,

SM: , = (pxt_x + sr
/- 0: ( - 0
( 1).
:
0: (= 0 - ;
0: (- 0 >, = st
( Axt = Oxbxt_x + ... + _Axt_p+X + t).

J 10.1.2.
.
,
,
.
482 2.


, . 10.10.

I
# 0: = 0 ------------------------------------ A xt = a + fit + <(_
-1
+X 61/ Ax-i + ,
. j=l >
I

\

I
0: - 0 ( = 0 -------
!
0: #>= 0 / 3 * 0

-1
2 Axt = a + f i t + ,
=1

?=0- - ------- , = + + 6!'<- +


V j =l

1
0: - 0 <= 0 ------- ------- 0: (= 0 * 0

3 Axt- a +,
j=i
i
0: (= 0 ----------------------- ------- Axt = (pxt_x
=|

i
(~ 1
'*'-/ +i
0=> J

. 10.10
10. TS- DS- 483

:
- 1
1: Axt = + /3t + (pxt_x +yL ei Ax,-j + et\
v J=l
\
2: Axt = a + (pxt_x + t,O jA xt_j + tl
v l=x
' -i ^
3: Axt = cpxt_ + 'L eJAxM +
v J=l
.

10.1.6

1959 1985 . ( ., 1982 .).
. 10.11.

. 10.11


,
. .
1.
SM: Axt = + J3t+ (pxt_x + 6xbxt_x + ... + 0p-\Axt-P+\ + r

,
SM: xt = + J3t+ (pxt_x + st.
( = 26):
Axt = 461.338 + 25.857/- 0 . 4 4 8 * , + et\
484 2.

/- # 0: <= = -2.640.
(5%-)
DGP: , = + 0,
f = -3.59 ( = 26). t9 /
.
2. , DGP
DGP: , = + fit + sr

#:/? = 0.
= 2.680. (5%-)
(: /? > 0) 2.85 0: 0 = 0
.
3.
SM: , = + (,_ + ,;
DGP
DGP: , =
# 0: < - 0. : , = 47.069 +
+ 0.00522,_! + ,; /- t9 = 0.335. (5%-)
/ = -2.98 > .
4.
SM: , = + (,_ +
DGP
DGP: , = 0: = 0.
: , = 47.069 + 0.00522,_,+ ,; /- ta = 1.682 <
/ =2.61 > 0: - 0 .
5.
SM: , = <,_ +
DGP
DGP: , = 0: <= 0.
: , = 0.03070,_! + ,; /- t(p = 7.987 > / =
= -1.95 > 0: <=0 > :
X, =, _, + ,.
10. TS- DS- 485

.
(
, ),
:
.
0: =
2 (,
/- : Tpz = 2.680, / = 2.85.)
2
0: = 0, 0, 0: <=0

SM: , - + p t + <,_ + s, ,
DGP: Axt = + fit + st, 0.
/^ '(0, 1)
, 5%-
: / = -1.645. tv = -2.640
(. 1), ,
, :
, = 461.338 + 25.857/- 0.448jc,_, + ,.

DGP SM

: ,
SM: , = + ,_ +

DGP: , = + ,_ + 0 ( )?
/ DGP
t
-
/=1
xt_x SM
a(t - 1). ,
. , ,

, .. /-,
. * 0 ,
/- ,
= 0, .
486 2.

10.1.7
WALKJ2
0.2 SM: xt = + axxt_x + st
, . 10.16.
10.16
WALK_2


/- -

0.173019 0.160495 1.078031 0.2865


WALK_2 (-1 ) 0.991851 0.045395 21.84958 0.0000

, /-
0.991851-1
t =---------------= -0.180.5%- /- -
0.045395
( - ) = (49 - 2) = 47 -1.68,
5%- , DGP:
, =,_ + , ( = 0), -2.92,
. ,

-1.68 -2.92.
ST_2 , = 0.2 + 0.8,_, + ,
( , 1) = 50
, . 10.17.
10.17
ST_2


/- -

0.166899 0.159693 1.045128 0.3013


ST_2 (-1 ) 0.793680 0.091904 8.635959 0.0000

, /-
0.793680-1
/ = ---------------= -2.245. -1.68 -
0.091904
,
-2.92 .

.
ST_ 1 , = 0.8,_) + ,.
, , , ,
10. TS- DS- 487

SM: , = ,_ + (, DGP
), SM: , - + a]xt_l + , (, DGP
).
5%- /- ( = 50)
: -1.95.
,
/ = -1.68 ( , ),
/ = -2.92.
SM / = -2.314.
.
SM / = -2.298.
, DGP , / < / = -1.68,
. , DGP
, / > / = -2.92, .
,
(SM)
, , , .
SM,
. , DGP
,
, .
5% ,
DGP .

,

.
1
DGP:
) xt =xt_x +
) , - + ,_, + st, 0,
) , = + /?/ + ,_, + ,, /3 0.

,

.

1 .
, 10.2.
488 2.



SM: xt = + /?/ + axxt_ + * ,
/- 0: - 1 ,
, DGP.
, /-
) )9.. , - 0 0.
DGP ) /5 0, /-
, /- (,
N(0, 1)). (Kwiatkowski,
Schmidt, 1990).

10.1.8
WALK 3
:
DGP: xt = 0.2 + 0.\t + xt_x + st.
SM: xt = + /?/ + a xxt_x +
: = 0.989 / = -0.775. /-
/ = -1.68,
. ,
DGP = 0 , / = -3.50,
.
ST 4
DGP: xt = 0.12 + 0.13 / + 0.01 /^ + 0.8,_ +
, = 0.990
/ = -0.577 .
, .

1. TS- xt = + fit +
+ axxt_x+ st, 1^1 <1?
2.
/-?
3. /- ,

xt = axxt_x + et91= 1, ..., , , ,
, : xt =xt_x + et ?
10. TS- DS- 489

4.
?
5.
?
6. F-
?
7.

?
8. ?
9. ?
10.
?
11. ?
?
?
12. , xt
ARMA(/?, q) q > 0?
13.
?
14.
?
15.
? ?
16.
() ,
?

_________ 10.2 --------------


, (Phillips, Perron, 1988),


xt DS 0: = 0

SM: Axt = + p t + (pxt_x + t = 2, ..., ,

, ,
. , ,
ut -
(
), ()
( , E\ut\s < <
490 2.

8 > 2). , ,
.
/-
0: q> = 0 ,
Z,, -

Z, (long-run)

2 = \im T -V ar(Ul + = lim -\ 'ui +- + ul


...............
> [
; ( )
, = + fit + <px,_l + t = 2, ..., ,
(2)* 2 (Newey, West, 1994):

(2)*= 0* + 2 1- J /
=1
1+ 1

*= 1 **- '-
t=j+1
/
/, , , -1/4
-> 0, (2)

2 (. {Phillips, 1987))
Z,
tv .
,
/ ( /
window size). ,

( ).

. , -

.
(
, , (Phillips, Perron, 1988), (Schwert, 1989)), -
I .
(Schwert,
f rjl 1/4"
1989), I = , [] -


10. TS- DS- 491

, 4 12
. /, ,
f 2/9
, EViews, / = 4 {Newey,
lio o j

West, 1994).
, I
.
Z, (Fuller,
1976) (M a c K in n o n , 1991).
: , IMA(1, q),
q / .
q = 1, , = , + ble,_l, > 0
,
, .
< 0
,
< 0 (
DS).

10.2.1
GNP
Axt = + fit + <,_ + ,_ + ,
:
ADF Test Statistic -4.117782 1% Critical Value -4.1219
5% Critical Value -3.4875
10% Critical Value -3.1718

: /-
0: < = 0 5%- ,
, 1%-
.
AR = 5,
(. 10.18).
t = -2.873575 > -3.1744,
10%- .

. .- F-
0.44.
,
, .
492 2.

10.18
AR = 5

ADF Test Statistic -2.873575 1% Critical Value -4.1314


5% Critical Value -3.4919
10% Critical Value -3.1744

( D(X))

-
/-

(-\) -0.266169 0.092626 -2.873575 0.0060
D (X (-1)) 0.546230 0.133521 4.090958 0 .0 0 0 2

D (X (-2)) 0.183918 0.149711 1.228486 0.2253


D (X (-3)) -0.020254 0.152201 -0.133077 0.8947
D (X(-4)) -0.058683 0.148061 -0.396345 0.6936
59.45556 19.32396 3.076779 0.0035
@TREND{ 1947:1) 1.397409 0.482120 2.898469 0.0056

, .
-------------- -------------------(j\[eweyf West, 1994)
^ 2/9
/ = 4 = 3 , . 10.19.

10.19

Test Statistic -2.871178 1% Critical Value -4.1190


5% Critical Value -3.4862
10% Critical Value -3.1711

Lag truncation for Bartlett kernel: 3 (Newey-West suggests: 3)


Residual variance with no correction 29.28903
Residual variance with correction 54.87482

( D(GNP))
Method: Least Squares; Sample(adjusted): 1947:2 1961:4;
Included observations: 59 after adjusting endpoints

-
/-

G N P(-\) -0.153024 0.072723 -2.104212 0.0399
38.33211 15.97824 2.399020 0.0198
@TREND( 1947:1) 0.806326 0.378145 2.132322 0.0374
10. TS- DS- 493

,
,
(Newey, West, 1994), ,

.

10.2.2

ST_ 1, ST_2, ST 3 (. 10.20).
5_1 ST 2 :
DGP: , =,_ + st,
SM: , - + ,_ + ,;
ST_3 :
DGP: , - + xt_x + s, ( DGP: xt = xt_, + ,),
SM: x, - a + fit + alx,_l + r
DF ()
, / ,

(Newey, West, 1994).
10.20

/1= 50 = 100

STJ D F = -2.298 > % = -2.60, D F = - 3.238 < *5/0 = -2.89,


(3) = -2.394 > ^ 10/ = -2.60, (4) = -3.399 < ^5/ 0 = -2.89,

10 %- 5%-

STJ2 D F - - 2.387 > ^ % = -2.60, D F = -3.217 < / 5% 2.89,


(3) = -2 3 2 2 > tKpml0% =-2.60, (4) = -3.364 < / 5/ = -2 .8 9 ,

10 %- 5%-

STJ DF = -2.6S1 > ^ 1% = 3.18, DF = -3.201 < ^ 10% = _ 3.15,


(3) = -2.755 > ^ = -3.18, {4) = -3.368 < ^1% = -3.15,

10 %- 10 %-

, DF 9

.
494 2.

(Leyboume, 1995)
(DF) , ,
(..
), (DFmax).
>Fmax
- 25, 50, 100, 200, 400
() .

(). ,
.

.

10.2.3
ST_3 100
DF = -3.207.
-3.352.
, -3.207, 5%- -3.45,
. 5%-
( ) -3.15,
ST 3 5%-
.!

(Schmidt, Phillips, 1992)


DS ( )
x, = y/+%t + wt,
w, = fiw,_! + t = 2,..., .
, (/? = 1 /? = 1) /
, .
(DS), (TS)
/, % .
, ,
/ 1/2.
.
(Maddala, Kim, 1998, p. 85).
.
10. TS- DS- 495

10.2.4
ST 3 100 ,
, -3.12.
5%- -3.06.
5%- .

DF-GLS

, , ,
(.Elliott, Rothenberg, Stock, 1996). DF-GLS
(. (Maddala, Kim, 1998)) 0 - 0
? = 0? + <(_, + . . . + apAyi_p + error,
e y f ( .
).

10.2.5
, DF-GLS.
-3.246, 5%- -2.89.
5%- , ,
.


(KPSS)
, (Kwiatkowski, Phillips, Schmidt, Shin,
1992), TS.
:
_
~ +
,
, ,
, .
, ,
-.
LM-
.
,
, .
I ,

496 2.

/.
, (Schwert, 1989).
KPSS
(Maddala, Kim, 1998, p. 120 122).

10.2.6
STJ3 100
KPSS 1= 3 0.157. ,
7-, ,
. ,
, : 5%- 0.146,
- ^- .
,
, DF-GLS,
TS- /-, .

( )
, (Cochrane, 1998),
(VR variance ratio):

Vk = jD ( x ,- x ,_ k).

xt , VRk = 1, xt ,
( ), VRk 0
- > .
-

, ---------- -
- +1
VRk.
VRk = 1, ..., ,
TS DS,
.

VRk:

j
Axt - xt ~ x t_}.
10. TS- DS- 497

10.2.7
ST 3 ,
,
TS DS ,
.
. . 10.12
-.

. 10.12 . 10.13


WALK_2 (. 10.13).
, WALK_2
DS .


,
, . , ,
,
. ,
( )
. ,

,
DS (., , (Ghysels, Perron, 1992)),
. (Davidson, MacKinnon,
1993) -
.

498 2.

(dummy) DI, D 12 (
) D 1, D4 ( ).
,
.
(Dickey, Bell, Miller, 1986),
, /
.

,
,
, . ,
,
, , .
, , (Shiller, Perron,
1985) (Perron, 19896).



TS- DS -
TS DS
,
TS, DS, 4 ,
(. 10.21).

10.21

# 0: TS 0: TS

# 0: DS 1 2

0: DS 3 4

:
2 DS ;
3 TS ;
1 -
;
4 (DGP)
TS- ^-.
10. TS- DS- 499

{Fuller, 1976) {Dickey, Fuller, 1981)




, .


. ,

,
.
,
, 1.

.

2.
,
,
.
:
, ,
. ,
,
.
,
,/? > 1,
{Dickey, Pantula, 1987)

, .
,
, ( - 1)
..
AR(2)
a{L)x, = st,
..
(1 - axL - a2L2)x, = s (1 - aL){ 1 - bL)xt = s

a - , b = ,& zl,z 2 a{z) = 0.


z, z2
, ,
IzJ |z2| > 1, , \\ |Z>| < 1.
500 2.

,
, :
xt = ( + b)xt_x - abxt_2 + sr
xt_:
Axt = ( + b - l)xt_x - abxt_2 +
Axt_:
A2xt = Axt - Axt_x = -Axt_x +(a + b - l)xt_x - abxt_2 + st =
= (a + b - 2)xt_x + (1 - ab)xt_2 + r
:
A2xt = {a + b - 2)xt_x + [-(ab - l)xt_2 + (ab - l)xt_x] - (ab - l)xt_{ + s t =
= ( + b - ab - 1),_! + (ab - l)Axt_{ +

A2xt = ( - 1)(1 - b)xt_x + (iab - l)Axt_x + et.
,
2 , 1
:
- b = 1 ( ), A2xt = st;
= 1, \b \ < 1 ( ),
A2xt = (b - l)Axr_! + sn A2xt = q>Axt_x + st ( < 0;
\\ < 1 \b \ < 1 ( ),
A2xt = y/xt_j + (pAxt_I + et q>< 0 / < 0.

, , .
,

A2xt = + (pAxt_x + ut
/-
( 1 2
, = 0 0). ut
,
A2xt_x,..., A2xt_p+l.
(< = 0) ,

A2xt = y/xt_x + <pAxt_x + ut
10. TS- DS- 501

{// = 0 / < 0.
, xt ,
, xt ~ /(1).

10.2.8
AR(2)
(. 10.14 10.16). ,
.

. 10.14 . 10.15


SM: A2xt = + (pkxt_x + st
< = 0
< 0. (





.) ROOT2
,
2 ( 0),


. = 100
. 10.16
5%-
-2.89.
/- -1.64,
. ROOTO ROOT1
502 2.

, 1 { = 0),
.
= 100 5%- -1.95.
/- -7.83 ROOTO -5.50
ROOT1,
.

ROOTO ROOT1.
SM: 2 , = /,_ + (,_ + ,
/ = 0 / < 0.
/- -3.89 ROOTO -1.63 ROOT1,
/ = 0 ROOTO
ROOT1.
, DGP ROOT2
, DGP ROOT1
, DGP ROOTO :
DGP ROOTO: , = 1. 1jc/_ - 0.3,_2 +
(1 - 0.6Z,)(1 - 0.5L)xt = et\
DGP ROOT1: xt = 1.5jc,_! - 0.5x,_2 +
(1 - L)( \ - 0.5L)xt - s,\
DGP ROOT2: xt = 2xt_x - x,_2 + s
(1 - L ) 2 x t = s ,M

, ,
, ,
, {Patterson, 2000).


{Perron, 1989)
DS
,
( ),
( ),
.
: ^-
, ,
, ..
/)5- (., , {Engle, Granger, 1991)).
10. TS- DS- 503


(Nelson, Plosser, 1982),
13
( 62 111 )
GNP, (1948
1987 .). , ,
.
:
SM: xt = + fit + <,_ +
, AR(),
DGP: , = + , ( = 0 * 0),
.
, 13 14
.
.

, , ,
, /'-. ,
,
,
14 .
, (Perron,
1989).
GNP I 1958 .
IV 1979 . (. 10.17).
(
) ,
.
,
1974 . 1974 .

1958:1 1979:4, 19.086.
1958:1 1973:4
19.852.
1975:1 1979:4 31.995.

,
. ,
, 1975 .
, ,

.
504 2.

. 10.17

,
,
(,Perron, 1989),
,
.
z,
:
zt = Cl\Zt_x + t9
y t
, =f(t) + z
f(t) = 0 t< T B f(t) = // 0 t> T B.
E(zt) = 0, (,) = 0 t < (,) = t > .
, ,
(
).
,
AR(1) v,.
y t =At) + aly,_x + { < 1,
f{t) = 0 t< T B f(t) = // (1 - {) t> , ju^O.
, .
, ?
:

, = \- 1 + ( + t) = \1- 1 + vr
10. TS- DS- 505

t= T B + h :

+ = ,

TB +h - 1 h-1
yT
oB+h + I a \et. k + ^ [( \ - ).

( )
E(yt) = 0. h >

lim y V / / ( l -<*!) = //.


h-* k=0
t = y t
//,
.
f(t)
et ( ),
(innovation outlier).
(additive outlier).
(
),
, .
, :
. 10.18;
. 10.19;
. 10.20;
(
) . 10.21.

Y
16 -------- 1 ADD

Y'IJNNOV
12

-4 - iiii|ii ii|iii|iiii|iii if r rii|iiii|iiii |iiii|iiii|


10 20 30 40 50 60 70 80 90 100 f

. 10.18
506 2.

. 10.19

. 10.20

. 10.21
10. TS- DS- 507

(.Perron,
1989) :
A) :
0 : ,= + dDTB, + ,_, + s, (AS-),
: , = + 6DU, + fit + s, (rS-);
B) :
0 : xt =c + 0DU,+ ,_, + , (DS-),
: ,= +fit + yDTS, + s, (TS-,);
C) , :
0 : , = + 6DU, + dDTB, + xt_x + , (DS-),
: ,= + 6DU, + fit + SDT, + s, (TS-).

;
fl t = TB +l,
DTBt =\
[ ;
fl t> T B,
DUt =\
[ t< TB;
[t-Tn t> T B,
DTS, = \
[ t< TB;

DT ={* (>
[ t< TB;
s, ~ ARMA(/?, q) ( q ).


, . , ,
:

,= + 9DU, + fit+ st ;

, = + fit + yDTS, +, ;

,= + 0DU, + fit + SDTt + , .
508 2.



et = aet_x + vt
( vt )


et = ,_1+ /,_,.+1'.
7=1

, ,
, v, -
. ta /- = 1 -

; xt\ ^ ~ ~^ ^ ()
.
( ) ta
, , (
) / ta,
.
= , .. ,
.
,
,
,
,
.
,
ta , .. ta < /.
, , ,
. , , ,
, xt
:

f( t ) = c + 6DUt + (3t ,

f ( t ) = c + p t + yDTSt ,

f ( t ) = + 0DUt + fit + SDTt .


10. TS- DS- 509

, st .
et , et ~ ARMA(p, q),
2 + q2 > 0, ,
/(/) , -
^ xt ( ^-).
, 2 + q2 > 0,
, . ,
/- ta9 /-
- 1

, = + 0DUt + fit + t_x+ et ,

xt = +fit + yDTSt + a x t_x +t ,

xt =c + 0DUt + p t + 8DTt + axt_x+t .


(Perron, 1989), ,
ta9
. ,
ta9 ,
.
(Perron, Vogelsang, 1993). ,
.
ta
= 1

< = X 0}jDTB,-j +,_, + X ; , +v,
=0 7=0

t ^ m,
(Perron, 1989).
( )
ta = 1

, = aet_l + ^ cj Aet_J +vl,
j =i

et ,

xt =c + fit + yDTSt + t.
510 2.

(Perron, 1989).
14
,
.
11 14 , ..

. ,

.

10.2.9


, xt = M l, Ml
,
, .
1995:06 2000:07
. Xt = Ml . 10.22.

. 10.22


(. (
.., 2001))
,
. ,
1998 . 1999 .,
- 1998 .
10. TS- DS- 511

,
1998 .,
,
()
( ).
,

, - ju + p t + yDTS, +
DTS, t - t > 0
t. ,.
, ,_,
,_!, ...,,_:

et =,_ + ' c j Aet_J +et;
j =i

/- 0: = 1
,
(Perron, Vogelsang, 1993, p. 249).

,
.
= 42, 1998:08.
12
, ( 11 ) -
( AR(1) ), 0.0002
. ,
GS ( ) SIC,
,
( 10%-
) .
. 10.22.
. 10.22 ,

. ,
10%-
.

(SIC), .
- (P-values) LM-
. , P-
, ( )
.
512 2.

10.22


/-
- - -
SIC
/- White J-B

- ( 22.236 1 0.983 0.701 0.281 -1 .9 2


12 2 0.967
)

8 22.157 -2 .2 7
11 22.089 -2 .6 0
10 22.018 -2 .9 0
9* 21.986 1 0.590 0.372 0.223 -3 .2 7
2 0.844
3 0.954
4 21.974 0.040 -2 .7 8
5 21.935 0.035 -2 .5 9
3 21.898 0.016 -2 .2 2
* ( GS) 21.837 0.006 0.518 -2 .0 4

1 21.834 0.002 0.184 -1 .3 7


6 21.793 0.008 -1.31
2 ( SIC) 21.782 0.006 -0 .9 2

P- (White)
.
P-
.
/- (
) ,
( ) .
( 10%-
) 12
, 8, 11, 10, 9
(). ,
, 1 7 12 ;
, .
, ,
10%- ,
, ,
. ,
.
10. TS- DS- 513

, -

, ,
, .


, ;
.
= 42/62 = 0.667. 5%-
(
) -3.94 ( = 0.6) -3.89 ( = 0.7).

.
, 1998:08
1998 .
Ml.
/-
.

14
,
.
(Zivot, Andrews, 1992), . 9 ( 9.2),
, GNP
II 1973 . (
- ).
,
().
, (1964 .),
(1965 1974 .), 1980- .
1973 .,
GNP. , ,
( )
,
. , ,
(Perron, 1989), , ,
.
514 2.

,
,
( ),
,

/- ta 0: = 1;
, ta (/min).
?
, (1890 1970 .).
(Zivot, Andrews, 1992) (.
, , ), DTBt.
, GNP,
1929 . ( ). ta
ta = -4.95, = 40, = 40/81 = 0.49.
(5%-) ta -3.76,
. , ,
(1929 .), fmin = -4.95.
/- . tm[n
ta : 5%-
/min -5.26. tmin = -4.95 > -5.26,
(0: = 1) .

.
11 14 (
GNP ( ) (1860 1970)).

( ):
, .
, , ,
GNP,

(kurtosis): 5.68, 4.658, 4.324,
,
/min. (

. 1.)

1
( = - 3), 0
. - , ,
Econometric Views,
() .
10. TS- DS- 515

(
)
5%-
: -5.86, -5.81 -5.86 ( -5.38, -5.33 -5.63 ).
?min : -5.82, -5.30 -5.61, , ,
.
, ,
, D(e,) = ,
/min ,
5%-
.

{Perron, 1997). ,
/ mjn
, .
,
( ) DTBt. ,
t = TB +\,
+1.
, (.Perron, 1997),
PERRON97 RATS.
:
101 , ;
102 , ,
;
, .
:
UR ^-
= 1;
STUDABS ^-

, (
101) ( 102);
STUD ^-
,
( 101)
( 102).

,
.
516 2.

10.2.10 ( 10.2.9 Ml)


, ,
PERRON97 RATS, ,
(Perron, 1997). ,
,
, .
,
().
PERRON97 . 10.23.

10.23
PERRON97
break date = 1999:07; statistic ta (alpha = 1) = -3.34124

Critical values at 1% 5% 10 %

For 70 observations -6.32 -5.59 -5.29

Number o f lag retained: 12


Explained variable: Ml

Coefficient Student

CONSTANT 124786.79561 3.33345


DU -2506239.31872 -3.77751
D(Tb) 40455.79442 2.72347
TIME 9769.03708 3.44839
DT 23866.02686 3.78217
Ml {1} -0.91050 -1.59235


DUt = 1 t > DUt = 0 t\
D(Tb)t = 1 t = + 1 D(Tb)t = 0 /;
DT = t t > DTt = 0 t\

:


6 : CONST, DU, D(Tb\ TIME, DT
Ml {1}.
10. TS- DS- 517

PERRON97 1999:07,
/-
/=1, .
ta=l = -3.341, 5%- -5.59,
. ,
, 12
GS 10%- .

/- d DTn
, 1998:04. ta=l = -0.547,
5%- -5.33,
. (
11.) ,
/-
DT, ,
1998:04
(-).
,
( ) ().
PERRON97
. 10.24.

10.24
PERRON97
break date = 1999:02; statistic ta (alpha = 1) = -3.59417

Critical values at 1% 5% 10%

For 100 observations -5.45 -4.83 -4.48

Number o f lag retained: 12


Explained variable: Ml

Coefficient Student

CONSTANT 104939.65455 20.48279


TIME 4832.56930 26.73200
DT 14335.07564 21.11189
M1{1} -0.75752 -1.54915

:
.

CONST, TIME, DT,
518 2.


et et_x Aet_x, ..., Aet_p.
ta=x
1 et_
. 1999:02, ta=x = -3.594 (
12 ), 5%- -4.83,
//?- .
,
:
1.626
, 3. (Zivot, Andrews, 1992) (
),
,
, t/-
.
Ml 1995:06
2000:07, (
.., 2001).
. 10.25.

10.25
1 1995:06 2000:07

()

()
DS TS

()

DF-GLS

KPSS

DS


( )


( )

,
, : DS-
, TS- ;
^-.
10. TS- DS- 519

1. , ,
TS- DS-?
2.
?
3. ?
4.
TS /)-?
5. ,
?
6. (
)?
7.
TS- >-?
8. ?
9. ?
11


.
.

_______ 11.1________
.
.

(, spurious)
. (. 1.3.4
. 1) , 1

,
.

11.1.1

2 ,
, '(0, 1).
. 11.1 11.2.
DGP
DGP: , = 1 + 0.21+ ,
, = 2 + 0.4/ + 2

.
11. .. 521

. 11.1 . 11.2

, ,
SM: y t = + /?, + et
. , . 11.3.
.

. 11.3

. 11.1.
,
,
.
.
(. 11.2).
,
. -
522 2.

11.1


^- -

1.553866 0.685771 2.265868 0.0280


X 1.800255 0.102997 17.478780 0 .0 0 0 0

R-squared 0.864218 Mean dependent var 12.22809


Adjusted R-squared 0.861389 S.D. dependent var 5.953260
S.E. o f regression 2.206028 Akaike info criterion 4.459442
Sum squared resid 233.5948 Schwarz criterion 4.535923
Log likelihood -109.4860 F-statistic 305.507600
Durbin-Watson stat 2.150060 Prob. (F-statistic) 0 .0 0 0 0 0 0

11.2
,

-
^-

2.037450 0.294861 6.909879 0 .0 0 0 0

0.412232 0.028055 14.69394 0 .0 0 0 0

X -0.054186 0.133658 0.405410 0.6870

R-squared 0.975727 Mean dependent var 12.228090


Adjusted R-squared 0.974694 S.D. dependent var 5.925326
S.E. o f regression 0.942598 Akaike info criterion 2.777771
Sum squared resid 41.75908 Schwarz criterion 2.892492
Log likelihood -66.44428 F-statistic 944.6386
Durbin-Watson stat 2.249075 Prob. (F-statistic) 0 .0 0 0 0 0 0

xt
, , ,
^,.
xt
(. 11.3),
, .
,
^, = + (3xt + sr ,
SM 233.59,
11. .. 523

11.3
Y


^- -

1.990020 0.268291 7.417403 0 .0 0 0 0

0.401493 0.009157 43.84727 0 .0 0 0 0

R-squared 0.975642 Mean dependent var 12.228090


Adjusted R-squared 0.975134 S.D. dependent var 5.925326
S.E. o f regression 0.934357 Akaike info criterion 2.741262
Sum squared resid 41.90511 Schwarz criterion 2.817743
Log likelihood -66.53155 F-statistic 1922.583
Durbin-Watson stat 2.249658 Prob. (F-statistic) 0 .0 0 0 0 0 0

41.91.
, y t


, DGP
.
()

. ,
, .
.

11.1.2
:
DGP: xt = xt_x + s lt9
y t = y t- 1 + * 2P

s Xt elt , 11.1.1.
xt y t . 11.4.
, ,
50 ( 51- 100-).

SM: y t = + J3xt + st
(. 11.4).
524 2.

. 11.4

11.4

Sample: 51 100; Included observations: 50

-
/-

8.616496 0.748277 11.515120 0 .0 0 0 0

X 0.597513 0.077520 7.707873 0 .0 0 0 0

R-squared 0.553120 Mean dependent var 3.404232


Adjusted R-squared 0.543810 S.D. dependent var 3.354003
S.E. o f regression 2.265356 Akaike info criterion 4.512519
Sum squared resid 246.3283 Schwarz criterion 4.589000
Log likelihood -110.8130 F-statistic 59.41131
Durbin-Watson stat 0.213611 Prob. (F-statistic) 0 .0 0 0 0 0 0

DGP
y t xt

,


: 0.553. ,
,

(. 11.5).
, ,
100 ,

(. 11.5).
11. .. 525

11.5
Y
Sample: 1 100; Included observations: 100


/- -

1.490206 0.664538 2.242470 0.0272
X 0.055097 0.083978 0.656086 0.5133

R-squared 0.004373 Mean dependent var 1.120548


Adjusted R-squared -0.005786 S.D. dependent var 3.513463
S.E. o f regression 3.523613 Akaike info criterion 5.376648
Sum squared resid 1216.753 Schwarz criterion 5.428752
Log likelihood -266.8324 F-statistic 0.430449
Durbin-Watson stat 0.061638 Prob. (F-statistic) 0.513306

,
xt 0.0551 0.5975,
51- 100-.
DGP

.
(0.214 0.062
).

,
,
( !)
.
.
DGP: , = ,_ + , , = ,_ + 2 0 = 0, 0 = 0,

, ~ N(0, <,2), e2t ~ N(0, 22),
Cov{x ,) = 0. , ( ,), t - 1,2, ..., ,

SM: , = fix, + , ~ i.i.d. N(0, 2), Cov(x ,) = 0.
/3
:
526 2.

DGP
->
, .
SM ( ),
( DGP!)
Cov(xn yt) = Cov(xn J3xt + f) = fiCov(xt9 xt) = f3D(xt),
..
Cov(xn y t)
D(xt)

( DGP) Cov(xn y t) = 0, /3
, ( SM)
, /? >0 .
, -> /- tp 0: /3=
,
/-
, .. , xt y t
. -
1
/^, j= tp,
/71
.
(DW),
DW > 0 >,

.
,
.

11.1.3

DGP: , =xt_j + elt9
y t = y t-i + s 2t,
s lt e2t ,
,
N(0, 1).

SM: y t = + fixt + st
11. .. 527

51- 100-
0 = 0.598, tp = 7.708, DW= 0.214.
(. 11.6)
.




,

,


(
, .).
, , ,

. 11.6 , ,

( /? ).

,
.
,
,
11.2. ,
.

11.1.4
SM: y t = + J3xt + st
51- 100-
y t = 8.616 + 0.598, +
,
, ..
Aet = (pet_x + vt
/- t 0: - 0,
.
0 : < 0 (
), / < /.
/ ( ) :
528 2.

+ 2~2,
, , 2 . -9
... (MacKinnon, 1991).
5%-
/ -3.3377 - 5.967-1 - 8.982,
= 50 / = -3.46.
5%- (-2.92),
.

19 = -2.01. 5%-
, .

, () (
) xt
( yt ~ /(1)),
:
j , - , ?
, , x t y t
( x t y t) (cointegrated time series).
, x t y t ( x t y t)
.
y t = + fixt + ut
, /?, ,
- xt
y t (., , 11.1.6).
, xt y t , (3
^ - J3xt .
,
DGP: xt xt_! + s ln
y t =yt- 1
s u s2t
, .. Cov(su, s2t) 0, s2t
, xt y t . ,
yt = , + ut, ut .
y t_x = ,_! + _ Ayt = j3Axt + e2t - P sXt + Aut.
ut = t_x + 77,,
11. .. 529

I t = , ~ P s \, ~ i-i-d. N (, </).
, ut .
0 =0 = 0,

Cov{xyt) = Cov(su + ... + , s2l + ... + e2t) = tCov(su, e2l),


xt y t ,
.
,

X = (Cov(skl, sl)), k ,s= 1, 2.

, ( CRDW cointegrating regres
sion DW):
e , = y t - a T- fax,.
T = 50 5%- 0.78.
,
.
11.1.3 0.214,
.

1 xt .
.
1.
, ,
SM: y t = + yyt_x + J3xt + Sxt_x +
ut xt
.
:
) , = + , - 1 +/, + (j3+ S)x,_x +
) , = + yyt-\ +( + S)x, - SAx, +
~ /(1).
) /3

; y t_l9 xt_x ~ /(1), ut .
(Sims, Stock, Watson, 1990),
SM , -
530 2.

(3 . /-
0: /?=
N(09 1), ut .
) - 8
Axt9
; y t_l9 xt ~ /(1), ut .
8 SM ,
/- 0\ 8 = 0
N(, 1), ut .
8
, ut , .
N(, 1)
/-,
ut
, 10.2.
qF = 2F 0: /3= 8 =
%2(2),
SM ,
/? 8
,
(
).
2. xt
..
SM: Ayt = + /?, + ut9
ut .
, /?
. /-
N{0, 1), ut . ut
, ,
/-, .
3. -
:
SM: y t = + f3xt + ut - pxt_x + et9 ut - i.i.d. N(0, <r2),
9
,
, - pyt~\ = (1 - ) + *, - ,-) +
11. .. 531

> 1 ( ),

.

11.1.5
,

DGP: xt =xt_x + s ln
y t - y t-\ + 2n
s u s2t ,
,
7V(0, 1).
50 .
, ,
. 11 .6 .
11.6
Y

-
-

0.447271 0.550358 0.812691 0.4206

X -0.014458 0.123861 -0.116725 0.9076

(- 1) 0.970105 0.055989 17.32664 0 .0 0 0 0

-1 ) 0.033532 0.120061 0.279290 0.7813

P- xt xt_x
, xt
^,.
, ,
. 11.7.
,
6*2,.
11.7
D(Y)

-
/-

0.184523 0.117614 1.568884 0.1232

D{X) -0.033386 0.116361 -0.286915 0.7754


532 2.

,
, - , - 1 (1 ) + , - ,_ J + s,,
..
= * + -1 + * / - , - \ ) + ,-

(. 11.8).

11.8
Y
Method: Least Squares; Convergence achieved after 7 iterations; =(1)+(2)*(-1)+(3)*(-(2)*(-1))


/- -

( 1) 0.217398 0.159423 1.363650 0.1792


(2) 0.988791 0.035801 27.61920 0 .0 0 0 0

(3) -0.027306 0.119276 -0.228934 0.8199

R-squared 0.940380 Mean dependent var 3.404232

, upuyt_x 1,
.
- 0.367.

11.1.6
, xt
..
%t ~ **7- 1 ^*1
y , = y t-i + .
:
s \n sit ,
, N(0, 1.25);
Cov(su, s2s) = 0 t * s, Cov(ln 2t) = 1.
, , , Corr(sln s2t) = 0.8.
s2t . 11.7.
s2t
,
0.789. xt yt
, . 11.8. . 11.9
xt yt
2
11. .. 533

. 11.7

. 11.8 . 11.9


;
s2t 0.792.
y t - + J3xt 4-
, -2.112,
5%- -3.46.
.
, ,
. 11.9.
, s2t
,
, n x t_x.
, . 11.10.
534 2.

11.9
Y


/- -

0.548392 0.377080 1.454312 0.1526


X 0.718479 0.079943 8.987361 0 .0 0 0 0

(- 1) 0.913556 0.067811 13.47210 0 .0 0 0 0

-1 ) -0.641522 0.088805 -7.223976 0 .0 0 0 0

R-squared 0.987574 Mean dependent var -0.957402

11.10
Y


/- -

X 0.695862 0.079341 8.770553 0 .0 0 0 0

(- 1) 1.005257 0.025245 39.82053 0 .0 0 0 0

- 1) -0.707002 0.077447 -9.128837 0 .0 0 0 0

R-squared 0.987003 Mean dependent var -0.957402

ecibjy, = 1.005^.! + 0.695;, - 0.707xt_x + ^,_!


1, xt xt_x
,
DGP.
, ,
. 11.11.
11.11
D(Y)


/- -

0.100915 0.083851 1.203508 0.2347
D(X) 0.694000 0.077274 8.981039 0 .0 0 0 0

R-squared 0.626921 Mean dependent var 0.226857

DGP

s Xt s2t, .. ,
, ,
. 11.12.
11. .. 535

11.12

()

-
/-

D (X ) 0.709553 0.076533 9.271155 0 .0 0 0 0

, = 0.710Ajc, + , = ,-+ 0.710, - 0.710,_, + ,.


, ,
, = + ,_, + j3(x, - ,_x) + st.
. 11.13.
11.13

Convergence achieved after 8 iterations; Y=C( 1)+(2)*(-1 )+C(3)*(X-C(2)*X(-1))

-
/-

( 1) 0.329205 0.250398 1.314726 0.1950

0 (2 ) 0.941984 0.056946 16.54170 0 .0 0 0 0

( 3) 0.723593 0.079341 9.119982 0 .0 0 0 0

R-squared 0.987410 Mean dependent var -0.957402


. 11.14.

11.14

Convergence achieved after iterations; =(2)*(-1)+(3)*(-(2)*(-1))

-
/-

(2) 1.014411 0.020750 48.88608 0 .0 0 0 0

(3) 0.702102 0.078268 8.970448 0 .0 0 0 0


y t = 1.014,.! +0.702(x, - \.0\4xt_x) +

y t = 1.014yf_j +0.702,-0.712:,_1 +er


536 2.

, :
y t - 1.005^_! + 0.695, - 0.101xt_x + et ( 1);
yt = yt~\ + 0.7 10, - 0.7 lOx^j + et ( 2);
yt = + 0.702x, - 0.712x,_! + et ( 3).

:
Ayt = 0.7t + et.

( ):
1.985, -
0.344.

/ 11.1.1. ,
, ,
.
, ,
,
( )?

:

,
,
-;
1
,

.
,
,
.

y t ~ /(1), , ~ 7(0). yt xt
,
y t bxt ~7(1).
, , yt ~ 7(0), xt ~ 7(1).
y t - a - b x t -7(1)
11. .. 537

b =
y t - a - b x t ~ /( 0),

.
jy, ~ /(1), xt ~ /(1) .
b
,- , ~ /( 1),
, , , ,
.
, 0
, - , ~ 1(0) .
, , , ,
(1, -) (cointegrating vector).
, y t ~ /(1), , ~ /(1) (
, deterministic cointegration),
() /?= ( , ) * 0,

, + V, ~ 7(0) .
, = ( , )
, ,,
= ( , 2), 0 . -
, (normaliza
tion) , (1, -) (
(-, 1)).
, , ~ /(1),
, . ,
( ,)
( )
( ,) = + B(L)s
// = (//,, /2), = EiAx,), /2 = (,),
s, = (, 2,) ,
.. 2, ... ,
,
E(st) = (0, 0), D(eu) = ,2, D(s2t) - 2, Cov(su, s2t) = <
;
(*) .(*)
1 0"
B(L) = \ 12 L k.
v0 1 h(k) ()
V 21 22
538 2.

, ,
( ,)- / .
(. (Granger, 1983), {Engle,
Granger, 1987)), , 7(1)
, , ( ):
(I) ( ,) = / + B(L)s, 7?(1) 1;
(II) , , ARMA
A(L)(xt,y ,)T = c + d(L)e
, , (I);
= (,, 2), , 2 ;
A(L) ;
d(L) ,
(0) = 12 ( (2 2)),
rank (\) = 1 ( (2 2)- (1) 1),
d( 1) .
(2 2)- ( 1)
,
{reduced rank VAR).
(II) :
4
'= + +bXjy,_j)+
7 =1 k=


y t =^2 + ( 2jXt.j +b2jy t_j)+ 2 ^ 2 , , - *
7=1 =0

q
.
AR , d(L) = \ 9 < .
(III) xt y t
(error correction model )
00

, = + a xz,_x + ^ ( r XjAx,_j + SXJAyt_j )+ '^ j 0kl t _k,


7=1 *=0
00 00

Ay, = 2 + 2^-1 + Y M v bxt-j + S2j Ayt- j ) +


j =1 k =0

z, =y, - fix, - E(y, - fix,)


, z, ~ 7(0);
+ 2 > 0.
11. .. 539

(II) AR(p) ( < ),


:
- 1
, = + a1zt_l + ( , ,_ +SXJAyt_j ) + el t ,
=i
- 1

ky,=M 2 + a 2z,-\ + Z (j *, _j + S2JAy, _ j ) + e2 r


;=i
.
t ~ /(1) ,
.
( , , yt)T ~ / ( 1 ) ( ( A yt)T
) -, xt y t
. (, ,
zt_ , zt_{ .)
j ,

xt, y t ~ I( 1) , VAR
( , xt y t
) .

zt =yt - -/? = E(yt - j3xt),


, t (equi
librium), y t - a -J3xt = 0.
xt y t
zt_x. zt9 ,
,

(mean-reversion).

> / 11.1.2. xt

, y t
. y t

y t
,
1. (
.)

1 ,
.
540 2.

y t ~ 7(1) ,
( Granger causality ), ,
.
, + 22 > 0. x t_x
z,_ y t (..
xt ^,),
2 0. y t_ z,_
xt (.. y t
xt), 0.

J 11.1.3. y t -7 (1 ) wt ~ 7(0).
xt yyt_k + wt, 0.
xt -7(1), xt (,
xt - xt_k = , + Axt_x + ... + Axt_k 7(0)-,
7(0)-.)

, x n y t ~ 7(1) :
() t = + J3xt ;
,
.
, , ,
(
). ,
. ,
.
(Engle, Granger, 1987)
.
(3
y t ,
y t = + f3xt +

(3
(-),

^ = , - - fix,,
.. .
11. .. 541

(
!) :
-i
+ ( X \ Z, - \ + X f a + s \jb yl-J ) + v
7=1
P- 1
&y,=M2 + 2*- l + +S2jAyi _j ) + Wn
1 j =

.. VAR(p) xn y r
, (
xt y t)
( ) /?
(1, ) ( , (3
). , ,
, ,
zt_l9 ,
, zt_{
( ). -

.
, ,
z t ,

,
(, 1), xt
, ,
.

^ 11.1.4. ,
( ) ,

. ,
, ,
. , /?
.

,

. ,
(,
).
542 2.

J 11.1.5.

z t - y t - - fixt
,
xt
. ,
f ,
,
,
. xt y t
,
/,
, .. t9 < /. , /
f ,
,
. (.Mac
Kinnon, 1991); ,
, (Patterson, 2000) (Hamilton, 1994).
11.2.

11.1.7

DGP: xt =,_1 +
' = 2, + v
[ = 0;
, v,
, ,
'(0, 1).
, , . 11.10.
( ,)
X , = X t_ l + t ,

y t = 2x,_j + 77,,
Tjt =v, + 2s, ~ i.i.d. N{0, 5).
:
, -
, = -( ,- 1 - 2xf_i) + 7], = z,_ j + 7
z , y, 2
11. .. 543

. 11.10

Axt = (XxZ t _ x + s t9

Ay, = a 2z,_x + T]
a x - 0, a 2 = -1, a x + a 2 > 0.
, ,
, VAR DGP. ,
:
, = a xz,_x + ,-1 + ,-1 + v
Ay, = a 2z,_, + 2,_ j + S2lA y,.x + w
,
( = 2). 100 .
I. , = + fix, + ,.
. 11.15.
11.15
Y

-
/-

-0.006764 0.165007 -0.040992 0.9674

X 1.983373 0.020852 95.11654 0 .0 0 0 0

R-squared 0.989284 Durbin-Watson stat 2.217786

, = -0.006764 + 1.983373*, +
z, = , =, + 0.006764 - 1.983373*,.
544 2.

, VAR 2,
y t xt
:
Az,= (pzt_x +<91Ai,_1 + .
, . 11.16.

11.16
D(Z)


/- -

Z (-1) -1.153515 0.151497 -7.614088 0 .0 0 0 0

D (Z (-1)) 0.038156 0.100190 0.380837 0.7042

f = -7.614 5%-
-3.396 (. (Patterson, 2000, . 8.7)).
. (

,
. t = -11.423,
.)
, y t xt
.

II. Axt (. 11.17).

11.17
D{X)


/- -

-0.028016 0.100847 -0.277810 0.7818


Z ( - 1) 0.250942 0.176613 1.420858 0.1587
(- 1)) 0.639967 0.257823 2.482201 0.0148
D (Y (-\)) -0.258740 0.116654 -2.218019 0.0290


P -
(. 11.18) , ,
, - v
,.
11. .. 545

11.18
D(X)


/- -

D (X (-1)) 0.115141 0.100249 1.148554 0.2536

. 11.19.
11.19
D(Y)


/- -

-0.060101 0.211899 -0.283630 0.7773


Z ( - 1) -0.641060 0.371097 -1.727472 0.0874

- 1)) 1.313872 0.541733 2.425311 0.0172


((- 1)) -0.482981 0.245111 -1.970459 0.0517

,
. 11.20.
11.20
D(Y)


/- -

Z ( - 1) -0.638888 0.369218 -1.730381 0.0868


D (X (-\)) 1.317763 0.538932 2.445138 0.0163
D {Y {-1)) -0.483722 0.243908 -1.983217 0.0502


zt_X9
y t
, + 2 > 0.
zt_x ,,

,_15 (.
11.21) ,_1?
yt = a2zt_ + wr ,
. 11.22.
0: 2 = - 1 . 11.23.
546 2.

11.21
D(Y)


/- -

Z (-1) -1.186411 0.248876 -4.767072 0 .0 0 0 0

( - ) 0.331411 0.210732 1.572671 0.1191

11.22
D(Y)


/- -

Z ( - 1) -1.273584 0.247887 -5.137760 0 .0 0 0 0

11.23
( 1) = -1

F-statistic 1.218077 Probability 0.272441


Chi-square 1.218077 Probability 0.269738

, = Ay, =-,_j +wt,


=,_\ +0.006764- 1.983373,_,.
,_, ,

, = -0.0068 + 1.983x(_! + w

, = 2,_! + TJ
DGP.
, , wt = Ayt + zt_

4.62 ( DGP 5.00),
et - Axt
1.04 ( DGP 1.00).

AXt = n Ayt = - z t_l +wt,
,
zt_x, ..
11. .. 547

yt_i - (-0.0068 + 1.983,_,) > 0,


,
^,.
, ,_,
,.
,
.. ,
,. ,
,
,.

, Cov(v wt) 0,
,
.

DGP: , = ,_, + ,, , = 2, + v,
,
(Phillipss triangular system). ( )
:
, = fix, + vr>
,= ,_ +
(-,, vt)T ~ i.i.d. N2(0, I ) ,
,

.
(two-dimentional
Gaussian white noise).

, Cov(sn vt) = 0, xt
,
/3 .
Cov(st, v,) 0, xt
, Cov(xn vt) = Cov(xt_x + et9 vt) 0.
/?
.

,
.
548 2.

N y lt, ..., yNn


1. /? = (J3X,..., PN)T9

\ + + ~ ) ,
, ( ),
/3 .
= E{J3xy lt + ... + PNyNt)<>

(long-run equilibrium relation) :
\\ + + = '
t
(deviation) ,

= \ + +--
zt ,
,
, ..
.

.
1. .


\ + + *
,
.
N.
.
.
2. ..., yNt (,
E(Aykt) = 0 = 1 ,2 ,..., N).
2. (SM)
.
11. .. 549

SM: = 22, + + +

, = ~ (22, + + \

ut = (put_x + xtSiit_x + ... + - + 6t

0: = 0: << 0.
/- t9
N.
,
(Hamilton, 1994, . .9, 1).
.
2. (SM) .

SM: = + 22, + + , +

/ = ~ ( + 2, + + ),

, = <,_1 + ^ ,_ , + ... + ,_ + ,


# 0: <=0 0: << 0.
2.

, (Hamilton, 1994, . .9, 2).
,
(MacKinnon, 1991, . 1 ( trend)) (.
(Patterson, 2000)).
3. 2 y Nt ,
E(Aykt) 0 .
. .

SM: = + 2 2, + ... + yNyN, + -
2,
.
, (Hamilton, 1994, . .9,
550 2.

3).
, ( MacKinnon, 1991, . 1 (
with trend)) (Patterson, 2000).
. .

SM: = a + S t + y 2y 2t + ... + yNy Nl + ,.


, , ,
, N, (N + 1) .

-
8.
, ,
, 2

11.1.8
4 , 2 3 4

DGP: = y 2>t +.3>, + 4, + ;

= - 1 + 2

/= ,/-1 + % ;

At ~ A,t- 1
, e2t9 s 3t9 s At
, 1 s 2n 3 s 4t 2 .
= 200 . 11.11.
,
. 4
.
/-, -2 .1 8 , -1 .7 8 , -0 .5 7 , -1 .7 0 . 4
.
, y ln 2 3 y 4t
(1)- , ..
1 .
4 .
1.

=.2,+,

~ ~ ^ - 4 ,
11. .. 551

. 11.11

(. 11.12)
.
( /- -1 5 .0 7 ).
.

. 11.12

, -
.
2.
, . 11.24.

. 11.25.
/- -1 5 .1 7 ,
5%- - 3 .7 4 ( H am ilton, 1994,
. .9, 1). .
552 2.

11.24
Y1


/- -

72 0.996084 0.009973 99.88161 0 .0 0 0 0

73 0.992550 0.009578 103.6296 0 .0 0 0 0

74 1.002305 0.012393 80.87922 0 .0 0 0 0

11.25

Dependent Variable: D(RESID_2A)


/- -

RESID_2A(-\) -1.075552 0.070892 -15.17178 0 .0 0 0 0

2.
, . 11.26.

11.26
Y^


/- -

0.332183 0.373542 0.889279 0.3749


72 1.002583 0.012369 81.05843 0 .0 0 0 0

73 0.987369 0.011215 88.04048 0 .0 0 0 0

74 0.999022 0.012937 77.22129 0 .0 0 0 0


. 11.27.

11.27

Dependent Variable: D(RESID_2B)


/- -

RESID_2B(-\) -1.079049 0.070861 -15.22764 0 .0 0 0 0

/- -1 5 .2 3 5%-
-4 .1 1 (. (Hamilton, 1994, . .9, 2)).
.
11. .. 553

3. 1 ^*, = + 0.75/,
^ , . 11.13.
4 . 11.14.

200- 1
1 STAR
150 - . . ,,
.* '.'*'
100 -

50- .wV . - ** if.-.*#


.*-"--if.'f

I I I | I I I I | I I I I | I I I I I I I I I I I I I I I I I I I I I I I I-------

20 40 60 80 100 120 140 160 180 200 t

. 11.13

. 11.14

.
(. 11.28).
(. 11.15).
11.28
Y1_STAR


/- -

11.49053 2.704802 4.248195 0 .0 0 0 0

Y2 -1.333762 0.089561 -14.89224 0 .0 0 0 0

2.856952 0.081207 35.18115 0 .0 0 0 0

Y4 0.072630 0.093677 0.775323 0.4391


554 2.

. 11.15


, . 11.29.

11.29


/- -

RESlD_3A(-\) -0.119805 0.033630 -3.562431 0.0005

f- -3 .5 6 5%-
, -4 .1 6 (. (Hamilton, 1994,
. .9, 3)).
.

. (. 11.30).
(. 11.16) .
(. 11.31).

11.30
Y1_STAR


/- -

0.304068 0.390739 0.778187 0.4374


@ TREND 0.751890 0.007507 100.1621 0 .0 0 0 0

Y2 1.008470 0.026468 38.10166 0 .0 0 0 0

Y3 0.982658 0.021830 45.01453 0 .0 0 0 0

Y4 1.001356 0.015942 62.81247 0 .0 0 0 0


11. .. 555

RESID-

-8 1'1 I I | I I I I | I I I I | I I I I | I I I I | I I I I | I I I I | I I I I |

20 40 60 80 100 120 140 160 180 200 t

. 11.16

11.31

Dependent Variable: D(RESID_3B)


/- -

RESIDJB ( - \ ) -1.079492 0.070859 -15.23448 0 .0 0 0 0

/- -1 5 .2 3 4 5%-
, -4 .4 9 (. (Hamilton, 1994, . .9,
3)). .

,
.
. , y 2t, y 3t, y 4t
, (
).
. \ y 2t, y 3t9 y 4t
.
.
,

,
.
.
-
(stochastic cointegration). -
556 2.

,
.
( ).
,
.
, ,
, ,
.
, -

, , ..
, .

(<deterministic cointegration ).

*/ 11.1.6. ,


y t = + J3xt + ut

, (3
-
,
, ,
, / .
, (Entorf, 1992),
.

DGP: xt = / j x + x ,_ x + ,

,= + ,-1 +
e lt 2,
, * 0 , / 0.

SM: y t = + p x t + ,

, , >
, > <>

.
*
11. .. 557


SM: y t = + Pxt + y t +

( ^ ) juy, a f3T
,
.

1. ()
?
2.
?
3. ?

,
?
4. ,
TS- AS-?
5.
1 *, (?
6. ?
7. ,
yt ~ 7(1) ?
?
8. VAR , yt ~ /(1)
( )?
9. yt ~ 7(1) wt ~ 1(0).
( yyt_k + wt9 ?
10. xt ~ 7(1). xt xt_kl
11. t ~ 7(1) .
?
12.
?
13.
?
?
t ~ 7(1)
,
?
14.
?
15. -
?
,
?
558 2.

_______ 11.2________

N , ..., y Nn
1. /? = (J3l9 ..., PN)T9
,

\ + + ~ ) ,
( ); /? -
.

= ( + . . . + PNy Nt),

\ \ + + = .

t
,

^ = + ---+ - -
z t ,
,
, ..
.
, , ,
7(1)
.
y lt9 ..., y Nt ,
(cointegrating rank).
, N ,
= 1 ,..., N - 1. ( , = 0.
= N 9
N .)
7(1) -
,
(<cointegrating space).
,
,
, .
11. ..._______________________ 559

7(1) y ln ...,y Nt -
VAR()
\
A L ) y t =ju + s n
, = (, ...,y Nt)T\

A (L )= A 0 - A lL - . . . - A pLp;
0, , ..., ( N N);
0 = IN ( ),
..
y t = j u + A ly t_ l + ... + A py t_p + t.

( 1) rank (1) = ( N = 2)
V A R (
)
- 1
, , = / / , + a, , zu + . . . + a lrz r + (, Xj + / m j , , - j ) + u ,
j =i

p -1
Ay N t = MN a N\Z\,t-\ a NrZr,t -1 Y S Y N\ , j + + / NN,j S Nf>
j =1
z u, ..., z rJ 1(0) ,
/?(|), ..., );
( , ..., ), ..., ( |, ..., a Nr)T
.

Ayt =ju + ,-\ + CAyt-i + + - ,-+1+


,..., N N;
/? (N )- .

..., /?(>)
, a tj

z \,t-\ ~ 1)>'<-!> r r . / - i = !\ ) < \

( (t - 1)
, ...,yNt) , ..., AyNt.
560 2.

VAR
, /?(1), ..., /?()
.
.

7(1) .
, 0 < < N,
7(1) 1 ..., y Nt
( ):
\ f \ / \
+ Vl,
+c +
J rty \r J , y.t ,
/ \
Vr + U
_ +

V V n J

= () (N - );
v t = ( v i n > v N t ) T ( )
E (vt) = 0;
y r+ h ..., y Nt t .


~ \ \ + 1 , - 1, N. r y N, , = + V i

~ \+\, t ~ " ~ Cr, N- N, t ~ f i r + V r t >



{\) ~~ 0 , 0 , . . . , 0 , Cj j , . . . , _ ) ,

(2) ~~ (0? lj 0 , . . . , 0 , 2\ , . . . , 2^/_ ) ,

() ~ (0? 0 , 0 , . . . , 1, , . . . , ^N_r )

.
y lt9 ...,y Ntt

Z\ ,t - (1)( = ~ 1\+1, t ~ * * 1, N- N, t >

Z r, t - P( r ) y t - + 1 , t Cr, N- N, t *
11. .. 561

v lt9..., vrt vr+1? vNn


y r+ 1 ..., y Nj ,
. ctj
, (^ - ctj)
7 .
= 1

= \ + + + c lt N_xy Ni, + vu ,


( y 2ft, y N>t)
,
, t- F - (,
), ,
v u . , . 10,
v u
.
ctj
(N 1)- /?(1), ...,

= ( 0 ?0 , . . . , 0 , j, . . . , Cj N_r) ,

) - (, , ,
. . 1, rl,
- . . - _ ).

/?(1), ..., /?(),


:

Z \,t-1=Ai)JVi> zr t-\ = /3()-\-

, = ju + a z t_{ + A y t-i + + Cp-vtyt-p+i + t >



/ \


,
.
: ( )
, z l t , ..., zr t
562 2.

..., y Nt
,

(N + 1) ..., y Nt /,
.
^4(1) ,
, ,

\ \ + + P \ N y \ N + P\, N+\*9

\ + + + P r , N+\ t

((N + 1) 1)-

{\) = ( A i > > P i n 9 P \ , n + i ) 9

P( r) ~ ( \ 9 ***9 P r N 9 P r , N + 1)


-
.
-

(), , ,
.
7(1)

,
.
,

= + 2 + - + +
( = + 2 2, + ... + y Ny Nl + y N+1t + U').

, > 1, (1, - 2, - y N)T ( (1, - 2, ...,


- y N, - y N+iY )
,
.
- , > 1
.
:
11. .. 563

{\) ~ , ,..., , , . . . , clN _r ) ,

{2) ~ 1? , . . . , , 21, . . . , 2 N_r ) ,

P{r)= ( 0 , 0 , 0 , . . . , l , - c rl, . . . , - c rN _r )r .

(
) ,

.
( )
1 ..., y Nn /?(1), ..., /?()
.
( )
(identification o f the cointegrating vectors),
,
.

, (Johansen, 1988).

.
11.4, ,
= 1, .. ( ) -
.

, N = 2.
y 2t ,
1. ,
,
, (
). ,

.


, = f a , + v
Xf xt_ i + s n
t n v t .
564 2.

,
(West, 1988):
, - + /, +

, ~ 7(1), (,) = ( , ,
);
, ~ 1(0) ,
.



( , ) . , ,
/-,
,
.
/- /?
/? , :

s J ( X TX );l d T, s J ( X TX)- .,

X ( 2)- (1 xt)
;
S 2 ut , ut ~ i.i.d.,
1
s 2 = --- 2 , - y t - - f i x , .
- 2 /1

, ut ~ i.i.d., /-
(, N (, 1)) /- (
7), S 2 .
, ut
. yh = Cov(ut, ut+h). ,
S 2

2 = ,
h = -
00
<0-
h =-
11. .. 565

, , , 2 ,
. , ,
yh, h = 0 , 1 , 2 , ...,
, , . - , ,


2 yh = Cov(ut, ut+h).
, yh
h ,
() yh 2
h.
, ut
() q , yh = 0 | h \ > q,
yh h .


yh = yh.
t=h+1
(., , (Hamilton, 1994, . 513),
q = 0 ( 115) 2 (
).
EViews .
:
N ewey West (, ,
,
).
, ut
AR(p) ,
ut = a xut_x + a2ut_2 + ... + aput_p + s t ,

s t D ( s t) = 2,

(1 - - 2 - . . . - )2

566 2.

, 2, , 2, \
1
-1 _ 1 V .
* rp jLd

, ,.

S 2 2 /-

, , ^ .
X

11.2.1

DGP: , = 2, +

X, = 1 + ,_, + v

, = 0 . 4 , + 0.2,_2 + , AR(2) ;
t, vt ,
: C ov(s v,) = 0.8.

. 11.17.

. 11.17

SM: , = + fix, + ,
, . 11.32.
,
, DGP
, . (. 11.18)
A R (2). A R (2)
. 11.33.
11. .. 567

11.32
Y


/- -

-0.398071 0.172093 2.313111 0.0228

X 2.031938 0.007241 280.6336 0 .0 0 0 0

R-squared 0.998757 Mean dependent var 37.39809

Adjusted R-squared 0.998745 S.D. dependent var 30.23477

S.E. o f regression 1.071308 Akaike info criterion 2.995436

Sum squared resid 112.4748 Schwarz criterion 3.047539

Log likelihood -147.7718 F-statistic 78755.23

Durbin-Watson stat 1.080957 Prob. (F-statistic) 0 .0 0 0 0 0 0

. 11.18

11.33
RESIDS


/- -

RESIDS{-\) 0.363522 0.100494 3.617344 0.0005

RESIDS(-2) 0.205074 0.100427 2.042024 0.0439

R-squared 0.240364 Mean dependent var -0.008547

Adjusted R-squared 0.232451 S.D. dependent var 1.075097

S.E. o f regression 0.941891 Akaike info criterion 2.738343

Sum squared resid 85.16727 Schwarz criterion 2.791098


568 2.

0.941891
= ---------- ---------- = 2.183,
1 -0 .3 6 3 5 2 2 -0 .2 0 5 0 7 4

S 1.071
= 0.491.
2.183
/-
-:

ta : -2.313111 (- = 0.0228) -1.135738 (- = 0.2588);

tp\ 280.6336 (- = 0.0000) -> 137.791098 (- = 0.0000).

= 2
( ) N
7(1) y lt9 ...9y Nt.

= + 2 * + - ' + +

,
,
. , ,


.

.
N = 2 ,

= + + + + s \t >

2, t - \ ^ S 2t'>

N t ,1-\ + S Nt'>

s t = ( e U9 e2t, Nt) T TV- , ..


s l9 *2, ... ,
, -
11. .. 569

-
= ( ;,).
2 ..., s Nt ,
( GXj = 0 j = 2 ,..., N).
y 2t, ...,y Nt , /?= (1, - 2, ..., - y N)T
.

( - , 2 - 2, . . y N - y N) T\{y2 - - ; > t = l,...,T }

jV- , , F -
, 2, ..., yN
F -, ^- /'-.

= + 2 21 + - + 1 +

2, ~ 2, /-1 + 2>

= N,t- 1 + U N l >

, - (, 2 ..., uNt)T ^-
( /- ).
2 ..., uNt ,
, C ov(uu, uks) = 0 1
t, s.
. ( , ,

- .)

F- ^-
F- ^-,
S 2 X2
2 . F-
S 2 $
/- .

, , ,
>>2 ...9y Nt
.
(Stock, Watson, 1993) (Saikonnen, 1991)
,

(lags) (leads) per-
570 2.

. ( leads and la g s1.)




= + 22, + --- + , + 4 2, ,- ,+ + e Nj > 4 ,-j ) + ut-
j= ~ P

( ),
^2, ..., yN
(,
),
t- F -, ut .

, 2 ..., y Nt
.
, ,

.
, :

) ~ Yiyit ~ YNyNt ;
) - 22, -- ! .

, ) ),
, .

11.2.2
DGP: , = 5 + z t + ut,

+ vo
Zj = 0 ;
ut - v, + 0.25vf_j + 0 .2 5 v ,+ l + 0.1v ,_2 + 0.1v f+2 + O .l^;
s v ,
.
, v v(_ ,, v(+1, v,_2, vr+2,

.
DGP (. 11.19 100
). y t z, 100
.
.
, - + fizt + /, , . 11.34.

1 DOLS ( OLS).
11. .. 571

. 11.19

11.34
Y


/- -

4.851262 0.133152 36.43410 0 .0 0 0 0

Z 1.088870 0.047570 22.88977 0 .0 0 0 0

0: /3= 1, /-, ,
() AZj) * 0 t, s.
-
, Azn , ,
= + f3zt + ]

Included observations: 96
Correlations are asymptotically consistent approximations
e, AZ(-i) e, AZ(+i) i lag lead
1********* I********* 0.9017 0.9017
0

i 1* 1 -0.0217 0.0830
i *1 2 -0.0956 -0.0413
i 1 3 0.0064 0.0341
*1 1 4 -0.0510 0.0118
*1 1 5 -0.0824 -0.0228
i 1 6 -0.0171 0.0150
**i ** I 7 -0.1858 -0 .1579
i 1 8 -0.0292 -0.0272
1* 1* 9 0.0833 0.0701
i 1 10 0.0125 0.0216
572 2.

- ( ,_;)
/ = 0, 1 ,2 , - lag.
- C orr(et, A zt+i) / = 0, 1 ,2 ,
- lead.
-
- DGP 7- .

z, 7- (. 11.35).

11.35
Y


/- -

4.987362 0.020874 238.9236 0 .0 0 0 0

Z 1.000689 0.007818 127.9955 0 .0 0 0 0

D(Z) 1.006216 0.013125 76.66298 0 .0 0 0 0

D (Z (- 1)) 0.237875 0.012764 18.63643 0 .0 0 0 0

D (Z (- 2)) 0.089302 0.012810 6.971105 0 .0 0 0 0

Z)(Z(-3)) -0.008934 0.012368 -0.722323 0.4726


D (Z (- 4 -0.002997 0.012391 -0.241901 0.8096
D (Z (- 5)) -0.011646 0.012179 -0.956245 0.3423
D (Z(-6)) - 0 .0 1 0 0 1 2 0.011925 -0.839615 0.4041
D (Z(-1)) -0.003586 0.011634 -0.308269 0.7588
D{Z{ 1)) 0.262537 0.013373 19.63226 0 .0 0 0 0

D{Z{ 2)) 0.116863 0.013365 8.744236 0 .0 0 0 0

D(Z( 3)) -0.010921 0.013219 -0.826184 0.4116


Z)(Z(4)) 0.003903 0.013276 0.294017 0.7696
Z>(Z( 5 0.021536 0.013232 1.627644 0.1082
D(Z(6)) -0.008452 0.012699 -0.665583 0.5079
D(Z(7)) 0.002945 0.012199 0.241376 0.8100

: P -
0.252 ( = 1) 0.427
( = 2). 0: /3= 1
/- ; :
1 .0 0 0 6 8 9 -1
/=- = 0.0081,
0.007818
0: /?= 1 .
11. .. 573

11.2.2 ()
DGP ,
7(1) .

DGP: , - 5 + , +

, = 1 + x,_i + vn

, = 0 , v, , .

DGP . 11.20.

. 11.20

, = + fix, + ij,
. 11.36.

11.36
Y


/- -

5.172117 0.201812 25.62837 0 .0 0 0 0

0.997147 0.003444 289.5306 0 .0 0 0 0

-
, , .
,
7 7 .
. 11.37.
574 2.

11.37


/- -

3.224675 0.098061 32.88423 0 .0 0 0 0

X 1.000621 0.000520 1923.269 0 .0 0 0 0

D(X) 1.009585 0.012922 78.12803 0 .0 0 0 0

D (X (-\)) 0.241440 0.012760 18.92225 0 .0 0 0 0

D (X {- 2)) 0.093472 0.012938 7.224562 0 .0 0 0 0

D (X(-3)) -0.004479 0.012656 -0.353864 0.7245


D (X (- 4)) 0.001328 0.012648 0.104986 0.9167
D (X (- 5)) -0.007856 0.012272 -0.640121 0.5242

- 6 -0.006808 0.011934 -0.570501 0.5702


D (X(-1)) -0.001175 0.011513 -0.102067 0.9190
D (X (\)) 0.266271 0.013016 20.45699 0 .0 0 0 0

D(X( 2)) 0.120232 0.012948 9.286042 0 .0 0 0 0

D(X( 3)) -0.007953 0.012873 -0.617748 0.5388


D(X(4)) 0.006576 0.012866 0.511164 0.6109
D(X(5)) 0.023956 0.012814 1.869580 0.0658
D(X( 6 )) -0.007153 0.012294 -0.581813 0.5626
D(X(1)) 0.003367 0.011821 0.284813 0.7767

,
0: /3= 1 ^-
. :

,.1 .0 0 0 6 2 1 - 1 ,,
0.00520
0: /?= 1 .
(. 11.38).

11.38


/- -

3.223725 0.101508 31.758480 0 .0 0 0 0

@TREND -0.000317 0.007802 -0.040686 0.9677

X 1.000938 0.007798 128.3544 0 .0 0 0 0


11. .. 575

. 11.38


/- -

D(X) 1.009461 0.013372 75.49241 0 .0 0 0 0

D (X (-\)) 0.241345 0.013061 18.47841 0 .0 0 0 0

D (X (- 2)) 0.093380 0.013231 7.057897 0 .0 0 0 0

0{(-)) -0.004552 0.012877 -0.353495 0.7248

D {X {- 4)) 0.001255 0.012867 0.097531 0.9226

D (X (-5)) -0.007941 0.012541 -0.633231 0.5287

D (X (-6)) -0.006888 0.012181 -0.565473 0.5736

D (X (- 7)) -0.001256 0.011767 -0.106726 0.9153

D (X (\)) 0.266438 0.013736 19.397180 0 .0 0 0 0

D(X{ 2)) 0.120397 0.013657 8.815884 0 .0 0 0 0

D(X( 3)) -0.007809 0.013441 -0.58098 0.5632

* )) 0.006723 0.013451 0.499805 0.6189

D(X( 5)) 0.024098 0.013370 1.802394 0.0760

D(X(6)) -0.007035 0.012719 -0.553156 0.5820

D(X{7)) 0.003469 0.012172 0.285016 0.7765

0: = 1 ,
. .
,
.

11.2.3
DGP:
W, = 5 + 1 + rw
V, = 1 + t + 0.5rw , + 0.12

rw t = rw ,_ x + 0.53, ;
2 ,
.
. 11.21.
Vt = + p W t + 7 , (. 11.39).
(. 11.22) (
-2 .2 2 5%- -3 .4 6 ),
V, Wt .
1 -
576 2.

. 11.21

11.39
V


/- -

-2.657128 0.325116 -8.172855 0 .0 0 0 0

W 1.021898 0.011165 91.52908 0 .0 0 0 0

Durbin-Watson stat 0.290480 Prob(F-statistic) 0 .0 0 0 0 0 0

. 11.22

, Vt Wt.
(Vt - Wt) ,
, . 11 .2 2 ,
.

(. 11.40).
11. .. 577

11.40
V


/- -

-1.493506 0.035792 -41.72787 0 .0 0 0 0

@TREND 0.503277 0.007131 70.57274 0 .0 0 0 0

W 0.496897 0.007519 66.08810 0 .0 0 0 0

Durbin-Watson stat 2.275079 Prob(F-statistic) 0 .0 0 0 0 0 0

. 11.23

(. 11.23)
( -7 .0 9 ). -
Wt :

, W_DIF{-i) , W_DIF(+i) / lag lead

1 1 0 0.0353 0.0353
1 1* 1 -0.0237 0.1217
*1 1 2 -0.0846 0.0115
1 *1 3 0.0052 -0.1083
*1 1* 4 -0.0776 0.1174
1* I* 5 0.1352 0.1018
i** 6 0.1986 0.0347
1
1* 1** 7 0.1093 0.1669
1* 8 -0.1751 0.0614
**1 **** I 9 -0.2456 -0.3565
1* *1 10 0.1177 -0.0421

,
9
Wt.
578 2.

,
W, (. 11.41).

. , .
0.5
Wt.

11.41
V


/- -

@TREND 0.497313 0.023984 20.73533 0 .0 0 0 0

W 0.505495 0.025512 19.81437 0 .0 0 0 0

V, - 0.5/ - 0.5 W, . 11.24,


.
. V, - Wt
, . Vt - 0.5 W, (
. 11.25) ,
.

. 11.24 . 11.25

, Vt - 0.5/ - 0.5 fV, ,


-1 .4 9 3 0.104.
,
. , DGP

Vt - 0 .5 t- 0 .5 W , = (1 + / + 0.5 rw, + 0.1 n2t ) - 0.5 / - 0.5 (5 + / + rw,) =


= -1 .5 + 0.12,.
11. .. 579

1. ?
?
2. 7(1) , ..., yNt
VAR(p).
?
?
3.
7(1) ?
?
4.

= + 2 + - - - + +

N 7(1)
y lt, ...,yNt ( , leads and lags)?

------------- 11.3 --------------



7(1) , ..., y Nt

y t = (y u, . . . , y Nt)T,

VAR(p)

A ( L ) y t = j u + e

A(L) = A 0 - A lL - ... - ApLp;


A 0, A u ...,A p ( N x N);
A 0 = IN ( ),

..
, = / + { , _ i + ... + A py t_p + s t.

* y t =M + ,-\ + ,-1 + + - ,-+1 + ,

Q = { + ... + - 1 ,
& = - ( A + i + + 4 ) , k = 1 , 2 , . . . , - 1.
580 2.

,
C o - A t + . . . + A p - I N = A (l),
rank <^0 = ra n k ^ (l).
, ..., y Nt ,
( 1) (rank ,4(1) < TV).
,
- rank <^0 = 0 , 1,..., N :
- 1, ..., N - 1 VAR (
...,yNt ~ 1) );
= 0 , ..., y Nt ;
= N, TV- ,
, , ( 1, 0 , ..., 0 ),
(, 1, ..., 0 ), ..., (, 0, ..., 1). , ..., y Nt
.
= rank ^0,
y lt, ...9y Nt ,
.


.
(Johansen, 1988; 1991).
. ,
,
.
, /(1 ) ...,y Nt
VAR rank ( \ ) - ,
VAR .
1

/ = + y t~\ + A y t- \ +.. + - \ AVf-p+i + t



) = \
(5 (N )- .
/?(1), ..., /?() /3
,

ZU- \ - {\)1-\9 > Zr,t-\ = ()(-\

1 7(1) .
, 7(2).
11. .. 581

(
, ...9y Nt , ..., AyNt).
, s t TV-
. s t = ( s Xt9 ..., s Nt)T TV-

C o v(st) = Q, C o v(skt9 sjs ) = 0 t * s k9 j - 1 ,..., TV.
,
, .
t- F - ( TV(0, 1)
t- 1
qF ) , VAR , ,
VAR.
,
VAR, .
,
y lt9 ...9y Nt91= 1 , 9
L (Q, ju9 > -\) Q, //,
<^0, , - 1 .
, 9 :

19..., X N ,

, 1 > > ... > X N > 0 .

Z,max(r), 9
.


.
() 0: = 9
: = + 1. 0
:
582 2.

(maximum eigenvalue
test)

( / ) = 2 ( I max(r* +1) - Lmax(r*)) = - f i n d - . +1).

( )
0 " N .
(., , (P atterson, 2000, . 14.3 14.7), (Enders, 1995, . )
(Hamilton, 1994, . .1 1)).
]|
# 0: - , Ar*+19 ..., XN .
, *+1 ,
Amax(r*)
. 0: =
: = * + 1 Amax(r*),
.
()
0: = / , : > *.
0 :
*

z /=1

^ ( ^ ) = - 7 1(1 - ).
Z 1=1

(trace test)

,race ( * ) - 2 (Z,max(N ) ~ L max (* )) = - f ln d - i,).


^ i = r *+1

( )
0 N.
(., , (Patterson, 2000, . 14.3 14.7), ( Enders, 1995, . )
(Hamilton, 1994, . . 10)).
0: = * , *+1, ..., .
, ,
Atrace( r )
. 0: -
\ > &( \
.
11. .. 583

, , , ,
.
,
.
,
.
, 0.05,
0: = 0 : > 0.
, = 0.
# 0: = 1 : > 1. 0: = 1
, = 1, 0: = 2
: > 2, ..
.
= 0,

{ = } >0, = 0 ,1 ,..., r0 - 1 ,
{ = 0} -> 1 - .

, , ,
.
,
. ,
TV,
, ,
/ ( ,
cointegrating equation).
5 (
, , EViews):
2()\ ,
, ;
*(): ;
, .
{(): ;
, .
*(): ;
.
()\ ;
.

5
:
#2 () { () , () *() ().
584 2.

,
, , ,
, .
-
( 2).
,
2() *() 1() *(),
( N - ) *() () *() ().
, 5 , ,
:
( 0) . . . () . . . (0;
\ 0) ... \ ) ... H*(N);
1( 0 ) . . . # 1( ) = . .. = 1(;
* ( 0 ) ... : *() ... H*(N);
2(0) = ... 2() ... H2(N).
^,
, 5 .
,
, .

(AIC) 5(/ 1)
. ,
. , ,
.

, .
5 , ,
/(1 ) .
DGP,: , = , +

, = , - \ + vr
:
y , - y , - i = - y , - i + P ( x , - i + v , ) = - ( y ,_ i - p x , _ ]) + ut ,
, = fiv, + ,.


A y , = - z t_\ +
, = v
z , = y , ~ , ( ).
11. .. 585

, , (
), 2().

DGP2: , = 0 + , +
X ,=X ,_i + V,.
:

4 v ,= - z ,_ i +
, = v
z, - , - 0 - , ( ).

, , (
), *().

DGP3: , = 0 + fix, + s
, = + ,-1 + v r

, ~-1 - ~>-1 + + ( +JC(-1 + v t) + t =


= ~( , - 1 - - P xt- 1 ) + 0 +

= ~ z, - i + +

, = 0 + v
z t = y t - 0 - J3xt ( ).

xt ( ).
{ ().

DGP4: , = 0 + P \t + , +
, = + ,-1 + vf.

,- ,-1 = - y ,- \ + P o + P \ t + P ( 7 o + x t- { + v t ) + e t =
= ~( , - 1 - ~ /V ~ ,- 1) + +

= ~ z ,-\ + +
, = + V,,

z, = , - p Q- p xt - , ( ).
586 2.

xt ( ).
*().

DGPS: , = 0 + P it + fix, +

, = Yo + Y\t + ,_ i + v ,.

=- y, -\ +P o+ Pi t + fi(ro+rit + x,_l +vt) + el =
= ~(y,-i - 0- p xt - fix,_) + fir + pT\t + U,,

Av = - v i + firo + f i U + u

Ax, = r0 + Y\t + v
z, = , - p 0 - p xt - fix, ( ).

, ( ).
().

11.3.1
,
5 DGP.
:

fi = 2, 0 = 5, Pi = 0.2, = 0.2, / ,= 0 .0 1 .

, v, = 400
, , 4 1
.
DGP,: , = 2, +

,= ,_ +V,.
. 11.26.

, EViews ( , VAR
2), . 11.42.
. 11.42 5 3
( = 0 , 1, 2 ) :
Lmm(r) (Log
Likelihood), ;
( 1C Akaike Informa
tion Criteria),
;
(Schwarz Criteria),
.
11. .. 587

. 11.26

11.42

Sample: 1 400; Included observations: 398; Series: Y1 X1; Lags interval: 1 to 1

Data Trend: None None Linear Linear Quadratic

Rank or No Intercept Intercept Intercept Intercept Intercept


No. of CEs No Trend No Trend No Trend Trend Trend

Log Likelihood
0 -1526.582 -1526.582 -1526.015 -1526.015 -1525.928
1 -1434.108 -1433.925 -1433.357 -1433.324 -1433.242
2 -1434.100 -1432.926 -1432.926 -1430.264 -1430.264

AIC
0 7.691369 7.691369 7.698565 7.698565 7.708180
1 7.246775 7.250880 7.253051 7.257911 7.262521
2 7.266836 7.270985 7.270985 7.267658 7.267658

Schwarz Criteria
0 7.731434 7.731434 7.758663 7.758663 7.788309
1 7326905 7.341026 7.353213 7.368089 7.382716
2 7.387030 7.411212 7.411212 7.427918 7.427918

L.R.Test: Rank = 1 Rank = 1 Rank = 1 Rank = 1 Rank = 2

(Rank or No. o f CEs)


. 5 5 ,
( 2(), *(), (), *(), ()).

588 2.

, . ,
4- - 1, 5-
() - 2.
,
2( 1) ( , ,
; 1)
( 7.246775). (
7.326905).
2( \
. 11.42, . 11.43.

11.43
2()
Sample : 1 400; Included observations : 398;
Test assumption: No deterministic trend in the data\ Series : X1; Lags interval: 1 to 1

Likelihood 5% 1% Hypothesized
Eigenvalue
Ratio Critical Value Critical Value No. o f CE(s)

0.371673 184.9642 12.53 16.31 None**


4.00E-05 0.015911 3.84 6.51 At most 1

* 5%- .
** 1%- .
5%-
.

(Eigenvalue) . 11.43 19 2,
. 184.9642
^( 0 ),
0: = 0 : > 0. 5%-
1%- ,(0)
.
, 0: = 0 \ > 0.
0.015911 ^(1)9
0: = 1 HA: r > 1.
5%- 1%-
^trace(l)- ( 1)
, 0: = 1 .
1,
.
DGP2: y t = 5 + 2xt + s n
xt = x t_x +v,.
11. .. 589

Y2

. 11.27

. 11.27.

. 11.44.

11.44

Sample: 1 400; Included observations: 398; Series: Y2 X2; Lags interval: 1 to 1

Data Trend: None None Linear Linear Quadratic

Rank or No Intercept Intercept Intercept Intercept Intercept


No. o f CEs No Trend No Trend No Trend Trend Trend

Log Likelihood
0 -1526.582 -1526.582 -1526.015 -1526.015 -1525.928

1 -1513.785 -1433.925 -1433.357 -1433.324 -1433.242

2 -1513.777 -1432.926 -1432.926 -1430.264 -1430.264

AIC
0 7.691369 7.691369 7.698565 7.698565 7.708180

1 7.247159 7,250880 7.253051 7.257911 7.262521

2 7.667223 7.270985 7.270985 7.267658 7.267658

Schwarz Criteria
0 7.731434 7.731434 7.758663 7.758663 7.788309

1 7.727289 7*341626 7.353213 7.368089 7.382716

2 7.787417 7.411212 7.411212 7.427918 7.427918

L.R.Test: Rank = 1 Rank = 1 Rank - 1 Rank = 1 Rank = 2


590 2.

,
# * (1 ) ( , ;
1) ( 7.250880).
(
7.341026).
*(),
. 11.42, . 11.45.

11.45

^()
Test assumption: No deterministic trend in the data] Series: Y2 X2; Lags interval: 1 to 1

Likelihood 5% 1% Hypothesized
Eigenvalue
Ratio Critical Value Critical Value No. ofCE(s)

0.372251 187.3130 19.96 24.60 None**

4.005008 1.998159 9.24 12.97 At most 1

* 5%- .
** 1%- .
5%-
.

187.310 ^(0),
0: = 0 : > 0,
. 0: = 0
: > 0. 1.998159 ^( 1 \
0: = 1 : > 1,
. 0: = 1 .
1,
.

DGP3: y t = 5 + 2xt + et,

xt = 0.2 + xt_ x + vt.

. 11.28.

. 11.46.
,
( 1) ( , ;
1) ( 7.252889).
(
7.353051).
11. .. 591

. 11.28

11.46

Sample: 1 400; Included observations: 398; Series: Y3 X3; Lags interval: 1 to 1

Data Trend: None None Linear Linear Quadratic

Rank or No Intercept Intercept Intercept Intercept Intercept


No. o f CEs No Trend No Trend No Trend Trend Trend

Log Likelihood
0 -1531.268 -1531.268 -1526.015 -1526.015 -1525.928
1 -1506.877 -1438.576 -1433.325 -1433.324 -1433.242
2 -1504.370 -1432.880 -1432.880 -1430.264 -1430.264

AIC
0 7.714915 7.714915 7.698565 7.698565 7.708180
1 7.612447 7.274253 7.252889 7.257911 7.262521
2 7.619950 7.270753 7.270753 7.267658 7.267658

Schwarz Criteria
0 7.754980 7.754980 7.758663 7.758663 7.788309
1 7.692577 7.364399 7*35305! 7.368089 7.382716
2 7.740144 7.410980 7.410980 7.427918 7.427918

L.R.Test: Rank = 2 Rank = 2 Rank = 1 Rank = 1 Rank = 2

(),
. 11.42, . 11.47.
0: = 0 : > 0.
H0: r = 1 : > 1.
592 2.

J 1.47
,,()
Sample: 1 400; Included observations: 398;
Test assumption: Linear deterministic trend in the data\ Series: Y3 X3; Lags interval: 1 to 1

Likelihood 5% 1% Hypothesized
Eigenvalue
Ratio Critical Value Critical Value No. ofCE(s)

0.372353 186.2692 15.41 20.04 None**


0.002234 0.890114 3.76 6.65 At most 1

* 5%- .
** 1%- .
5%-
.

1,
.
DGP4: y t = 5 + 0.21 + 2xt + s t9
xt = 0.2 + x t_x + vt.
. 11.29.

. 11.48.

. 11.29

,
# * (1 ) ( ,
; 1)
( 7.257911). (
7.368089).
*()9
. 11.42, . 11.49.
11. .. 593

11.48

Series: 4 4; Lags interval: 1 to 1

Data Trend: None None Linear Linear Quadratic

Rank or No Intercept Intercept Intercept Intercept Intercept


No, o f CEs No Trend No Trend No Trend Trend Trend

Log Likelihood
0 -1533.049 -1533.049 -1526.015 -1526.015 -1525.928
1 -1525.311 -1525.279 -1518.361 -1433.324 -1433.242
2 -1521.201 -1518.005 -1518.005 -1430.264 -1430.264

AIC
0 7.723863 7.723863 7.698565 7.698565 7.708180
1 7.705079 7.709944 7.680208 125 7.262521
2 7.704525 7.698520 7.698520 7.267658 7.267658

Schwarz Criteria
0 7.763928 7.763928 7.758663 7.758663 7.788309
1 7.785208 7.800090 7.780370 i3 6 m m 7.382716
2 7.824720 7.838747 7.838747 7.427918 7.427918

L.R. Test: Rank = 2 Rank = 2 Rank = 1 Rank = 1 Rank = 2

11.49
*()
Test assumption: Linear deterministic trend in the data] Series: 4 X4; Lags interval: 1 to 1

Likelihood 5% 1% Hypothesized
Eigenvalue No. ofCE(s)
Ratio Critical Value Critical Value

0.372355 191.5010 25.32 30.45 None**

0.015260 6.120353 12.25 16.26 At most 1

* 5%- .
** 1%- .
5%-
.

H 0: r = : > 0.
0: = 1 : > 1.
1 ,
.
594 2.

DGP5: , = 5 + 0 .2 1 + 2xt +

, 0.2 + 0.01/ + ,_ + v r
. 11.30.
'
. 11.50.

. 11.30

11.50

Series: Y5 5; Lags interval: 1 to 1

Data Trend: None None Linear Linear Quadratic

Rank or No Intercept Intercept Intercept Intercept Intercept


No. o f CEs No Trend No Trend No Trend Trend Trend

Log Likelihood
0 -1672.222 -1672.222 -1630.634 -1630.634 -1525.928
1 -1527.340 -1527.331 -1525.738 -1525.724 -1432.667
2 -1527.280 -1520.795 -1520.795 -1432.659 -1432.659

AIC
0 8.423224 8.423224 8.224289 8.224289 7.708180
1 7.715279 7.720258 7.717274 7.722231 7,259633
2 7.735077 7.712538 7.712538 7.279694 7.279694

Schwarz Criteria
0 8.463289 8.463289 8.284386 8.284386 7.788309
1 7.795408 7.810404 7.817436 7.832409 7379 7
2 7.855272 7.852765 7.852765 7.439953 7.439953
L.R. Test: Rank = 1 Rank = 2 Rank = 2 Rank = 2 Rank = 1
11. .. 595

,
( 1) ( ,
; 1)
( 7.259633).
( 7.379827).
(),
. 11.42, . 11.51.

11.51

()
Test assumption: Quadratic deterministic trend in the data: Series: Y5 X5; Lags interval: 1 to 1

Likelihood 5% 1% Hypothesized
Eigenvalue Critical Value No. ofCE(s)
Ratio Critical Value
0.374152 186.5374 18.17 23.46 None**
3.97-05 0.015819 3.74 6.40 At most 1

* 5%- .
** 1%- .
5%-
.

H 0: = 0 : > 0.
0: = 1 : > 1.
1 ,
.

,
,
.

11.3.2
,
.
DGP: L234, = 0.5 W2t + W3, + 2 W4t + s u,
L23t = W2t + 0.5 W3l + s 2
w 2t = W 2, t - i +
+
II

+
II

e u, e2t, s 3t, s 4l, e 5t


,
, 0.04.
596 2.

W2n W3t9 WAt


(common trends), ,
()
. L23t W2t W3t,
234, W2t9 W3t W4r
500 5
DGP Z,234 L23n W2t9 W3t W4t. . 11.31 11.32
.

. 11.31

. 11.32

5 V A R (l)
( )
,
11. .. 597

5 . , Z234,
:
Autocorrelation Partial Correlation g - -

I* *1* 1 -0 .0 4 4 -0.044 0.9530 0.329


I- -I- 2 -0 .0 3 7 -0 .0 3 9 1.6395 0.441
I- -1- 3 0.031 0.028 2.1276 0.546
I* -1- 4 - 0 .0 2 0 -0 .0 1 9 2.3355 0.674
.|. .|. 5 0.031 0.031 2.8073 0.730
.|. .|. 6 0.036 0.037 3.4726 0.748
1* -I- 7 0.008 0.015 3.5055 0.835
.|. .|. 8 0.039 0.041 4.2674 0.832
1* -|. 9 0.003 0.007 4.2734 0.893
.|. .|. 10 0.001 0.005 4.2743 0.934

VAR(1)
.
,
.

. 11.52.
11.52

Series: /.234 L23 WALK2 WALK3 WALK4 ; Lags interval: No lags

Data Trend: None None Linear Linear Quadratic

Rank or No Intercept Intercept Intercept Intercept Intercept


No. o f CEs No Trend No Trend No Trend Trend Trend

Log Likelihood
0 99.73833 99.73833 103.2311 103.2311 104.7734
1 279.21780 279.4377 282.9301 283.2575 284.7997
2 457.04190 457.6821 461.1675 462.6611 464.2006
3 461.17540 463.9622 466.3818 469.2587 470.4661
4 464.54380 468.0578 469.2034 472.1756 473.0952
5 464.77990 470.6760 470.6760 473.8191 473.8191

Akaike
0 -0.400556 -0.400556 -0.394502 -0.394502 -0.380616
1 -1.081960 -1.078063 -1.076025 -1.073323 -1.063453
2 -1.755188 -1.749727 -1.751677 -1.749643 -1.743777
3 -1.731628 -1.730772 -1.732457 -1.731963 -1.728780
4 -1.704995 -1.703044 -1.703628 -1.699501 -1.699178
5 -1.665783 -1.669382 -1.669382 -1.661924 -1.661924
598 2.

. 11.52

Data Trend: None None Linear Linear Quadratic

Rank or No Intercept Intercept Intercept Intercept Intercept


No. o f CEs No Trend No Trend No Trend Trend Trend

Schwarz
0 -0.400556 -0.400556 -0.352227 -0.352227 -0.296066
1 -0.996646 -0.985058 -0.949199 -0.938043 -0.894352
2 -1.563717 -1.54301 -1.521357 -1.490127
3 -1.477977 -1.451756 -1.436531 -1.410672 -1.390579
4 -1.366794 -1.331023 -1.323152 -1.285205 -1.276427
5 -1.243032 -1.204355 -1.204355 -1.154623 -1.154623

L.R.Test: Rank = 2 Rank = 2 Rank = 2 Rank = 2 Rank = 2

2.
( )
,
; 2 .

W2 W3/ W4t (. 11.53).

11.53

Series: WALK2 WALK3 WALK4 ; Lags interval: No lags

Data Trend: None None Linear Linear Quadratic

Rank or No Intercept Intercept Intercept Intercept Intercept


No. o f CEs No Trend No Trend No Trend Trend Trend

Akaike
0 -1.209358 -1.209358 < 4 .21131* -L 2 I1 -1.205445
1 -1.201837 -1.206417 -1.208105 -1.209616 -1.206429
2 -1.191311 -1.194747 -1.195348 -1.193234 -1.192934
3 -1.168162 -1.177195 -1.177195 -1.171753 -1.171753

Schwarz
0 >130938 ^ ?358 -1.185951 -1.185951 -1.154715
1 -1.151107 -1.147232 -1.132009 -1.125065 -1.104968
2 -1.089850 -1.076376 -1.068523 -1.049499 -1.040743
3 -1.015971 -0.999639 -0.999639 -0.968832 -0.968832

L.R.Test: Rank = 0 Rank = 0 Rank = 0 Rank = 0 Rank = 0


11. .. 599

0,
DGP.
,
, DG P.B


( )
.
EViews 5 ,
( = 2), (. 11.54).

. * ,
5 , 5
. 5
,
. ,
, ..

11.54


Test assumption : No deterministic trend in the data',
Series : L234 L23 W ALK! WALK3 WALKA] Lags interval: No lags

Unnormalized Cointegrating Coefficients:

1234 L23 WALK2 WALKS WALK4

-0.079261 -0.198108 0.236127 0.178603 0.159704

-0.202709 0.079211 0.022787 0.161363 0.406370

0.001194 0.000453 -0.014625 -0.034465 0.037834

-0.002101 0.001543 0.019423 -0.024621 0.007077

-0.000206 0.000771 0.011197 -0.009764 -0.012244

- 1, -
,
, ..

(-0 .0 7 9 2 6 1, -0 .198108, 0.236127, 0.178603, 0.159704),

.
, , .. ,
600 2.

,

(1, 2.499451, -2.979119, -2.25363, -2 .0 1 4 9 1 6 ).
= 2,

, , .. :

[ ) = (-0 .079261, -0.198108, 0.236127, 0.178603, 0.159704);

/?*2) = (-0 .202709, 0.079211, 0.022787, 0.161363, 0.406370).

, , ,

. ,

( )
(, ,
..). , ,
,
,
.
> 1,
q = - 1
(
, ).

,
, -
() .
- 2,
(,
).
.

.
, ,

L234 L23 W2t, W3t W4t
, L234n W2n W3n W4t
L23t, 23,, W2n W3n W4t
L234r
,
L234t, L23n ,
11. . 601

,
, -
:
{\) - 09 0, , 4, P\s)T\
{2) = (0> >* PisY-
,
:

(-0 .0 7 9 2 6 1, -0 .198108, 0.236127, 0.178603, 0.159704)7-;


(-0 .2 0 2 7 0 9, 0.079211, 0.022787, 0.161363, 0.406370)
,
:
( 1 , 0 , / 7 , 3 , /? 1 4 , \ ( 0 , 1 , / ? 2 , / ? 2 4 , 5 ) -

EViews
(. 11.55).

11.55

Normalized Cointegrating Coefficients: 2; Cointegrating Equation(s)

L234 WALK2 WALKS WALK4

1.000000 0.000000 -0.499995 -0.993065 -2.006077


(0.005410) (0.008680) (0.008520)

0.000000 1.000000 -0.991867 -0.504230 -0.003537


(0.005490) (0.008810) (0.008650)


= ( 1 ,0 , -0.499995, -0.993065, -2 .0 0 6 0 7 7 ),

(2) = (0, 1, -0.991867, -0.504230, -0 .0 0 3 5 3 7 ).

Z-234, = 0.499995 W2t + 0.993065 W3t + 2.006077 W4n

L23, = 0.991867 W2t + 0.504230ff3, + 0 .0 0 3 5 3 7 ^ ,,

,
DGP:
L234t = 0.5W 2, + W3t + 2W4[,

L23, = W2t + 0.5 W3r


602 2.

, /?*,> *2)
/?(1) /3{2):

\) = -0.079261/?(1) - 0.198108 2),

'{2) = -0 .2 0 2 7 0 9 # ,, + 0.079211 2).


. 11.55
( )
,
. ,

( ).

.
,
, ..
.
, . 11.56.

11.56

Z)(I234) D(L23) D(WALK2) D(WALK3) D(WALK4)

ecml -0.971220 -0.051197 -0.040351 0.005400 0.031996


(0.11006) (0.06585) (0.04248) (0.04468) (0.04231)
(-8.82480) (-0.77748) (-0.94983) (0.12086) (0.75621)
ecml -0.033474 -1.015363 -0.001894 0.010505 -0.024292
(0.10788) (0.06455) (0.04164) (0.04380) (0.04148)
(-0.31028) (-15.72980) (-0.04549) (0.23985) (-0.58571)

A(L234), = -0.971220(ecm l),_, - 0.033474(ecm 2),_, + e ]t,

A(L23), = -0.051 \9 7 ( \\_ - 1.015363(ecm2),_, +<?2

A(W2), = -0.040351 ( 1),_, - 0.001894(ecw 2)f_ 1 + 3

A(W3), - 0.005400(ec/w l),_,+ 0.010505(ecw 2)f_ ! + 4

A(W4), = 0.031996(ecm l),_, - 0.024292(ecw 2)f_ , + e 5t,

(ecm l), = L 2 3 4 , - 0.499995 W2t - 0 .9 9 3 0 6 5 ^ , - 2.006077 W4t,


(e c m 2 \ = 1 2 3 , - 0 .9 91867ff2f - 0.504230fF3, - 0 .0 0 3 5 3 7 0 V
11. .. 603


( )
, ,

)' Pis ~

0: 12 = 2 = 31 = 2 = 41 = 42 = 51 = 52 = ,

(e c m \)t_ x ,
, (e c m l)t_x ,
(
).

,
,
EViews,
. (,

.)
CATS (Cointegration
Analysis o f Time Series), RATS
(Regression Analysis o f Time Series).
,
, (Patterson, 2 0 0 0 ).
,
:

VAR ;
VAR;

,
.


-
T ~- NP
^ /
(

).
,
( )
.
604 2.

1. , ?
,
?
2. ?
?
3.
?
?
,

1. :
1897
1913 . ,
.
,
.

1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913

dj 45.5 52.8 71.6 61.4 69.9 65.4 55.5 55.1 80.3 93.9 74.9 75.6 92.8 84.3 82.4 88.7 79.2

.
-1 . 7. ,
.

2.
EViews dj_1984_daily.wfl.
http://www.measuringworth.org
( ) 28 31 1984 .
(Series) dj ( Index)
.
1. 28 7 (
1 71).
, .
2.
( Return),
, ,
. ,

606 2.

,
, EViews 4.0 (Kolmogorov
Smirnov, Lilliefors, Cramer von Mises, Anderson Darling, Watson),
.
3.
28 31 ( 1 214).

.
,
EViews.
: File/New/Program.
,
, Run.
, .
(,
, ). (
, Views,
; EViews6

. ,

.)

,
load dj_1984_daily.wfl ' ,

.
, , ,
load C:\program files\eviews6\econometrica\dj_1984_daily.wfl.
smpl 1 71 71 .
!=71 ' ,
series x=dj ' dj .
,
: series = d(log(dj)).
series m=@median(x) ' ,
.
series sign ' sign,
1 -1 ,

.
' sign.
'
,
scalar 1=0
scalar 2=0
for !i=l to ! ' !i = 1 Index !i = 2 Return.
... 607

if x(!i)>m(!i) then
n l= n l+ l
sign(!i)=l
endif
next
for !i=l to !n
if x(!i)<=m(!i) then
n2 =n2 +l
sign(!i)=-l
endif
next
, , ,
,
scalar z l =0
scalar z2=0
!i=l
if sign(!i)>0 then ' ,
.
z l= l
for !i=2 to !n
if sign(!i)<sign(!i-l) then
z2 =z2 +l
endif
if sign(!i)>sign(!i-l) then
z l= z l+ l
endif
next
else ,
.
z2 =l
for !i=2 to !n
if sign(!i)<sign(!i-l) then
z 2 =z2 +l
endif
if sign(!i)>sign(!i-l) then
z l= z l+ l
endif
next
endif
scalar z=zl+z2 ' z.
'!n=!n-l ' Return, ! 1.
scalar exp_z=(2*nl *n2/!n)+l ' Z
scalar var_z=2*nl*n2*(2*nl*n2-!n)/(!nA2*(!n-l)) ' Z

scalar z_star=(z-exp_z)/@sqrt(var_z) ' z*.


608 2 .

load dj_1984_daily.wfl '. .


!start= 1 ' .
!end=71 ' .
!n=!end-!start+1 ' .

series x=d(log(dj)) ' , series = dj,


,
series = d(\og(dj)).
series z=na ' z.
scalar 1=0
scalar 2=0
for !i=2 to !n-2 '!i = 2 Return, !i = 1 Index.
if x(!i+l)>x(!i) then c l= l
endif
if x(!i+l)>x(!i+ 2 ) then c 2 =l
endif
if x(!i+ l)<x(!i) then 1=0
endif
if x(!i+l)<x(!i+ 2 ) then c 2=0
endif
if c l-c 2=0 then z(!i)=l
else z(!i )=0
endif
next
scalar s=@sum(z)
' -, ! 1.
!n=!n-l
scalar exp_s=2*(!n-2)/3
scalar var_s=( 16* !n-29)/90
scalar s_star=(s-exp_s)/@sqrt(var_s) S*.

,

.
load dj_1984_daily.wfl
smpl 1 71
!n=71
series x=na .
series x=d(log(dj)) ' Return, dj Index.
scalar Q
Q=0
... 609

for !i=l to (!n-l)


for !j=!i+l to !n
if x(!i)>x(!j) then Q=Q+1 '
0 .
endif
next
next
scalar Q_low=Q
Q=o
for !i=l to (!n-l)
for !j=!i+l to !n
if (x(!i)>x(!j) or x(!i)=x(!j)) then Q=Q+1
1.
endif
next
next
scalar Q_up=Q
diog(dj), n :
!n=!n-l
scalar tau_up= 1-((4*Q_low)/( !n*( !n-1)))
scalar tau_low= 1-((4*Q_up)/( !n*( !n-1)))
scalar var_tau=(2 *(2 *!n+5 ))/(9*(! n* (! n - 1)))
scalar taustar_low=tau_low/@sqrt(var_tau) ' ,
scalar taustar_up=tau_up/@sqrt(var_tau) .

3.
http://www.rts.ru -1 (
) 9 1998 . 10 2000 .

, ,
, .

4.
4
.
4 .
.
. white_noise.wfl,
100 , Genr.
eps 1:
epsl=@nmd.
Genr,
eps2:
eps2 =@nmd.
610 2.

eps3 eps4.
Group - GroupOl ,
. View -> Multiple Graph -> Line
. GroupOl
View Graph > Line
.

5.

, 4
, ,
EViews 4.0 (Kolmogorov Smirnov, Lilliefors,
Cramer von Mises, Anderson Darling, Watson).

, 4?

6 .
( 100)
Xt = a0 + a xX t_x + s t
,
1) 0 = 0, ?! = 0 .1 , 0 = 0;
2) 0 = 0, = 0.5, 0= 0
3) 0 = 0, = -0 .5 , 0 = 0
4) 0 = 5, <^ =0.5, 0 = ;
5) ?q = 5, 0.5, X q 5;
6 ) 0 = 5, - 0.5, 0 = 0;
7) 0 = 0 , ^ = 0 .8 , = 10 ;
8 ) 0 = 0, = 0.8, 0 = 5;
9) 0 = 5, = 0.8, 0 = 0.

.
, . 1 3.
?
?
, . 4, , . 5 6 .
, . 4 7, 5 8 , 6 9.
.
eps.
1. Series x l, Genr
: Enter equation > xl=0. Model M l
: xl=0.1*xl(-l)+eps.
M l Solve. Stop
... 611

on missing data Iteration control .


\ 1. :
View: Descriptive Statistics Histogram and Stats.
Mean x l, Std. Dev.
.
2. Series 2, Genr
: Enter equation -> 2=0. Model 2
: x2=0.5*x2(-l)+eps.
M l Solve. Stop
on missing data Iteration control .
2 2. :
View: Descriptive Statistics > Histogram and Stats.
Mean 2, Std. Dev.
.
3. Series 39 Enter equation 3=0.
Model 3 :
x3=-0.5*x3(-l)+eps. 3.
4. Series 4 , Enter equation -> 4=10.
Model 4 :
4=5+0.5*4(-1)+eps. 4 4.
5. Series 5, Enter equation > 5=5.
Model 5 :
x5=5+0.5*x5(-l)+eps. 5 5.
6 . Series , Enter equation - 6=0.
Model :
x6=5+0.5*x6(-l)+eps. 6 .

7.
, xt
, xt (xt + ) .

8 .
( ) Xt
: LXt = X t_x. ,
Lk;
L kXt = Xt_kLk.
Z,0,
L% = ixt=xt,
L L~l ( lead operator ),
L - % = Xt+
612 2.

:
1) L kC s ;
2) L k(CXt) = CXt_k\
3) (Lk + L m)Xt = L kXt + LmX, = X ,.k + X t_m;
4) LkLXt = Xt_m_k = LmXt_k = L m+k X, = LmL % ;
5) A:> 0 L~kX t = Xt+k;
6 ) \a \< 1, (1 - aL)( 1 + aL + a2L2 + ...) X t = X t9
(1 + aL + a2L2 4-...) (1 - aL). -
X
: ---- = (\ + aL + a 2L2 + .. .)X t ;
1- a L
1) | | > 1, (1 - aL)( 1 + (aL)~l + (aL)~2 + ...) X t = - aLXt.

: .. . = -(<aL)~l (1 + (aL ) -1 + (aL)~2 + .. .)X t .


1- a L
. 1 5 .
6
X t - ( l - a L ) ( \ + aL + a 2L2 + ... + a nLn)X t ,
7
(~aL)X t - (1 - aL)( 1 + (aL)~l + (aL)~2 + ... + (aL)~n)X t
, .

9. AR(1)
AR(1) X t = aXt_x + s t ,
s t , s t
D (st) = 2 Cov(X t_s, s t) - 0 s > 0.
, \ \ < 1, Xt = aXt_x + s t
00
X , =
k=

, Cov(Xt_s , st) = 0 s > 0,


X t = aXt_ j + et | \ > 1.
.
. | | < 1
:
Xt = aXt_x + s t - a(aXt_2 + t_x) = ...
| | > 1

X - -1 t+l - I b, t + 1-

... 613

:
\

t+k ~ t 2^ *+ *
k=o\a J
/ 1 \ *-1+1
00 1
X t_x = &t+m 9 :
k=
= \\ \\ a )

Xt aXt_ j + t.

, E(Xt) = 0, , D(Xt) = const Cov(Xt, X t+k)


.

10.

1. ---- - .
1 -0 .5 1
2. xt = O.Ijc,.! + 0.2,_2 + ^
.


. s t ~ WN(0, ^2)
, t
. ,
AR s t
.

11. .

t ~ WN(0, 1),
) xt = 1 - 0.5-1 + s t ?
, .
5 .
,
;
) xt = 1 - \ .2xt_x - xt_2 + s t ?
) xt - 1 + O.lx^.j + 0 .2 ,_2 + t
? ,
. ,
. 5
.
. ) ,
,
614 2.

/ ( 0 ). ()- :
1 1
X, --------+ -------..
a(L ) a(L)

12.
et ~ WN(0, 1),
) , = -5 + et - 0.8^_j ?
, .
5 .
,
;
) xt - et + et_x ?
, .
5 .
,
;
) xt = - 1 + et - 0A t_x - 0.2st_2
? , .
5
. ,
.

13.
, X t AR (), p part(k) = 0 > .

14.
1. , 11 12,
p part(k)
k = 1, 2 3.
2. EViews,
, 11 12 ,
.

15. ARMA
s t ~ WN(0, 1), , = 1 4- 0.4,_! + s t - 0.8,_!.
1. ?
2. , ,
3- ;
3. ,
;
4. ( 100), -
.
... 615

16.

p part( 1) = 0.4
Pparti2 ) = 0.6 ( 1) ( 2 )
.

17. ARMA(p, q)

, q ARMA(p, q)
9 ..., , ..., bq , p part( 1) = 0.5, p part(2) = 0.5,
P Pa rt(k ) = : >3.

18. ARMA(p, q)

9 q ARMA(p, q)9 a l9 ...,
bq