Scribd
Upload
63 views

Presentation of Econometrics: Topic Granger Causality Test Submitted by Qarsam Ilyas Roll No 7

The document discusses the concept of Granger causality and how it can be used to determine if there is a causal relationship between two time series variables. It provides an example analysis investigating the potential predictability relationship between stock market prices and GDP growth in the US. The results of the empirical analysis using Granger causality tests suggest that past stock prices contribute to predicting present GDP values but not vice versa.

Uploaded by

Qarsam Ilyas
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PDF, TXT or read online on Scribd
63 views

Presentation of Econometrics: Topic Granger Causality Test Submitted by Qarsam Ilyas Roll No 7

The document discusses the concept of Granger causality and how it can be used to determine if there is a causal relationship between two time series variables. It provides an example analysis investigating the potential predictability relationship between stock market prices and GDP growth in the US. The results of the empirical analysis using Granger causality tests suggest that past stock prices contribute to predicting present GDP values but not vice versa.

Uploaded by

Qarsam Ilyas
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 15

PRESENTATION OF ECONOMETRICS

TOPIC GRANGER CAUSALITY TEST SUBMITTED BY QARSAM ILYAS ROLL NO 7

What is causality in economics ?


Correlation: de enden!e o" one #aria$le on ot%er #aria$le& it doe' not ne!e''aril( i) l( !a*'ation+ Ca*'ation: !a*'e and e""e!t relation'%i
E''entiall(& t%e notion o" !a*'alit( in ti)e 'erie' e!ono)etri!' i' $a'ed on , redi!ta$ilit(-+ T%e 'tri!t r*le e) lo(ed i' t%at onl( t%e a't !an redi!t t%e "*t*re $*t not #i!e #er'a+ T%*'& i" A %a en' $e"ore B& it i' not ad)i''i$le t%at B !a*'e' A+ .o/e#er& A !an otentiall( !a*'e' B+

GRANGER CAUSALITY
Gran0er 123435 ro o'ed a ti)e6'erie' data $a'ed a roa!% in order to deter)ine !a*'alit(+ In t%e Gran0er6'en'e 7 i' a !a*'e o" ( i" it i' *'e"*l in "ore!a'tin0 ( + In t%i' "ra)e/or8 -*'e"*l - )ean' t%at 7 i' a$le to in!rea'e t%e a!!*ra!( o" t%e redi!tion o" ( /it% re' e!t to a "ore!a't& !on'iderin0 onl( a't #al*e' +

GRANGER CAUSALITY
Gran0er

!a*'alit(- i' a ter) "or a

' e!i"i! notion o" !a*'alit( in ti)e6 'erie' anal('i': Te) oralit( E7o0eneit( Inde enden!e

E9AMPLE+++++++++++
T%e relation'%i $et/een 'to!8 )ar8et ri!e' and 0ro/t% 1GDP5+ A Gran0er6!a*'alit( anal('i' %a' $een !arried o*t in order to a''e'' /%et%er t%ere i' an( otential redi!ta$ilit( o/er o" one indi!ator "or t%e ot%er+

EQUATIONS
Thus! the "ollo#in$ t#o e%uations can &e s'eci"ie( )

CONTINOUE*
A''*) tion: t%e error ter) i' a''*)ed to $e i+i+d+ /it% )ean :ero and !on'tant #arian!e+ In ti)e 'erie' !onte7t& t%e error ter) )*'t $e 'tationar( or inte0rated o" order ;+ T%i' a''*) tion /ill %old onl( i" $ot% 9 and Y are I1;5 or $ot% are I125 and t%eir linear !o)$ination i' I1;5+ It i' /ell noted t%at )a!roe!ono)i! #aria$le' are I125 ro!e''e' or are non6'tationar(+

Uni(i+ectional G+an$e+, causality "+om S- to G*.i(i+ectional causality

.YPOT.ESIS ABOUT T.E RELATION S.IP BET<EEN GDP AND SP

Uni(i+ectional G+an$e+, causality "+om G*- to SIn(e'en(ence &et#een G*- an( S-

*ATA ANALYSIS
T%e !o*ntr( t%at %a' $een !%o'en "or t%i' e) iri!al te't %a' $een t%e United State'& /it% no)inal #al*e' )ade real t%ro*0% t%e *'e o" t%e I) li!it GDP Pri!e De"lator+

GRA-/ICALLY RELATION .ETWEEN S- AN* G*-

E0-IRICAL ANALYSIS AN* RESULTS


C%e!8 t%e 'tationarit( o" GDP and SP 'erie' T%e A8ai8e 1AIC5 and S!%/ar: In"or)ation Criteria %a#e $een *'ed in order to "ind an a ro riate n*)$er o" la0'+ A"ter t%at t%e'e re=*ire)ent' %a#e $een 'ati'"ied& Gran0er6!a*'alit( te't' are !o) *ted+ Ta8in0 e=*ation 125 a' an e7a) le& t%e t/o 'te ' ro!ed*re in te'tin0 /%et%er SP !a*'e' GDP i' a' "ollo/'+

STEPS+++++++++++++
1 ) G*- is +e$+esse( on its 'ast 2alues e3clu(in$ S- in the +e$+esso+s 4) Thus! a secon( +e$+ession is com'ute( inclu(in$ the la$$e( SThe test statistics is)

RESULTS

RESULTS
The 2alues o" 5 statistic su$$est that S- G+an$e+, causes G*- !an( G*- (oes not cause S- Thus! it can &e a+$ue( that 'ast 2alues o" S- cont+i&ute to the '+e(iction o" the '+esent 2alue o" G*5o+ the e%uation 647 the associate( 5 tests $i2e the o''osite +esult! in "act the+e seems to &e no G+an$e+,Causality "+om 'ast 2alues o" G*- "o+ "utu+e 2alues o" S-

CONCLUSION
G+an$e+ causality is a statistical conce't &ase on '+e(iction I" la$$e( 2alues o" 8 hel' '+e(ict cu++ent 2alues o" Y in a "o+ecast "o+me( "+om la$$e( 2alues o" &oth 8 an( Y then 8 is sai( to G+an$e+ cause Y

An a''lication 9 Which came "i+st 9 the chic:en o+ the e$$?; .ecause sometimes G+an$e+ Causality Is mo+e than co++elation< It=s causation

You might also like