Professional Documents
Culture Documents
Vectorautoregression
yt c A1 yt 1 A2 yt 2 ... Ap yt p t
where c is a k x 1 vector of constants (intercept)
Ai is a k x k matrix (for every i = 1, ..., p) and
t is a k x 1 vector of error terms satisfying the conditions.
Properties:
Example: VAR(1)
Suppose {y1t}tT denote real GDP growth, {y2t} tT denote inflation
y1t c1 A11
y c A
2t 2 21
A12 y1,t 1 1t
A22 y2,t 1 2t
VAR(1) PROCESS
Example:
1.1 0.3
Yt
Yt 1 Z t
0.6 0.2
1.1 0.3
I
0
.
6
0
.
2
2 1.3 0.4 0
1 0.8, 2 0.5
B0 yt c0 B1 yt 1 B2 yt 2 ... B p yt p et
where c0 is a k x 1 vector of constants, Bi is a k x k matrix, i = 0, ..., p, and
et is a k x 1 vector of error terms.
The main diagonal terms of the B0 matrix (the coefficients on the ith
variable in the ith equation) are scaled to 1.
The error terms et (structural shocks) satisfy the conditions and
particularity that all the elements off the main diagonal of the covariance
matrix E(etet') = are zero. That is, the structural shocks are
uncorrelated.
yt c t 1 t 1 2 t 2 ... ( B) t
yt l
l
'
t
the row i , column j element of
GRANGER CAUSALITY
In time series analysis, sometimes, we would
like to know whether changes in a variable will
have an impact on changes other variables.
To find out this phenomena more accurately,
we need to learn more about Granger
Causality Test.
12
GRANGER CAUSALITY
In principle, the concept is as follows:
If X causes Y, then, changes of X happened
first then followed by changes of Y.
13
GRANGER CAUSALITY
If X causes Y, there are two conditions to be
satisfied:
1. X can help in predicting Y. Regression of X on Y
has a big R2
2. Y can not help in predicting X.
14
COINTEGRATION
Cointegration
In many time series, integrated processes are
considered together and they form equilibrium
relationships.
Short-term and long-term interest rates
Income and consumption
SPURIOUS REGRESSION
If we regress a y series with unit root on
regressors who also have unit roots the usual t
tests on regression coefficients show statistically
significant regressions, even if in reality it is not
so.
The Spurious Regression Problem can appear
with I(0) series
In a Spurious Regression the errors would be
correlated and the standard t-statistic will be
wrongly calculated because the variance of the
errors is not consistently estimated. In the I(0)
case the solution is:
t t - distributi on , where (long - run varian ce of )1/2
17
SPURIOUS REGRESSION
Typical symptom: High R2, t-values, F-value, but low DW
1. Egyptian infant mortality rate (Y), 1971-1990, annual data,
on Gross aggregate income of American farmers (I) and
Total Honduran money supply (M)
Y ^ = 179.9 - .2952 I - .0439 M, R2 = .918, DW = .4752, F = 95.17
(16.63) (-2.32) (-4.26) Corr = .8858, -.9113, -.9445
2. US Export Index (Y), 1960-1990, annual data, on Australian
males life expectancy (X)
Y ^ = -2943. + 45.7974 X, R2 = .916, DW = .3599, F = 315.2
(-16.70) (17.76)
Corr = .9570
18
Cointegration
If two or more series are themselves non-stationary, but a linear combination
of them is stationary, then the series are said to be cointegrated.
Example:
A stock market index and the price of its associated follow a random walk by time. Testing
the hypothesis that there is a statistically significant connection between the futures price
and the spot price could now be done by testing for a cointegrating vector.
The usual procedure for testing hypotheses concerning the relationship between nonstationary variables was to run Ordinary Least Squares (OLS) regressions on data
which had initially been differenced.
Although this method is correct in large samples, cointegration provides more
powerful tools when the data sets are of limited length, as most economic time-series
are.
The two main methods for testing for cointegration are:
The Engle-Granger three-step method.
The Johansen procedure.
Granger Causality
According to Granger, causality can be further subdivided into long-run and short-run causality.
This requires the use of error correction models or
VECMs, depending on the approach for determining
causality.
Long-run causality is determined by the error
correction term, whereby if it is significant, then it
indicates evidence of long run causality from the
explanatory variable to the dependent variable.
Short-run causality is determined as before, with a test
on the joint significance of the lagged explanatory
variables, using an F-test or Wald test.
20
Granger Causality
Before the ECM can be formed, there first has to
be evidence of cointegration, given that
cointegration implies a significant error
correction term, cointegration can be viewed as
an indirect test of long-run causality.
It is possible to have evidence of long-run
causality, but not short-run causality and vice
versa.
In multivariate causality tests, the testing of longrun causality between two variables is more
problematic, as it is impossible to tell which
explanatory variable is causing the causality
through the error correction term.
21
Engle-Granger Approach
Estimation of parameters can be done by OLS estimation of
linear regression equation:
Yt 0 1Y2t .. MYMt t
Dickey-Fuller t test is applied to the OLS residuals
Three-step approach
Determine the I(d) for every variable
Dickey Fuller, Perron tests H0: series is non-stationary
Estimate the cointegration relation by OLS regression
Test the residuals for stationarity
y1t 0 1 y2 t t t y1t 0 1 y2 t
y y
t
1t
2t
y1t
and
y 2t
p 1
I (0)
p 1
Johannsen Test
The approach of Johansen is based on the maximum likelihood
estimation of the matrix ( - I) under the assumption of normal
distributed error variables. Following the estimation the
hypotheses
H0: r = 0, H0: r = 1, , H0:r = M-1
are tested using likelihood ratio (LR) tests.
Example: Exchange rate, interest rates, S&P 500(GLOBAL) index, ISE index
Trace Test
Lags interval (in first differences): 1 to 4
Hypothesized
Trace
0.05
No. of CE(s)
Eigenvalue
Statistic
Critical Value
Prob.**
None *
0.065285
156.7717
47.85613
0.0000
At most 1 *
0.017048
38.96042
29.79707
0.0034
At most 2
0.005118
8.955273
15.49471
0.3695
At most 3
1.20E-06
0.002096
3.841466
0.9599
Max-Eigen
0.05
No. of CE(s)
Eigenvalue
Statistic
Critical Value
Prob.**
None *
0.065285
117.8113
27.58434
0.0000
At most 1 *
0.017048
30.00514
21.13162
0.0022
At most 2
0.005118
8.953177
14.26460
0.2901
At most 3
1.20E-06
0.002096
3.841466
0.9599
1 Cointegrating Equation(s):
Log likelihood
-46009.85
GLOBAL
EXCHANGE_R
ATE
INTEREST_RAT
E
ISE
1.000000
-0.000120
-16.55210
-0.026394
(0.00014)
(1.44457)
(0.00218)
Log likelihood
-45994.85
2 Cointegrating Equation(s):
GLOBAL
EXCHANGE_R
ATE
INTEREST_RAT
E
ISE
1.000000
0.000000
-15.66567
-0.024910
(1.30729)
(0.00194)
7382.376
12.36210
(1649.84)
(2.44689)
0.000000
1.000000
Lags: 5
Null Hypothesis:
Obs
F-Statistic
Probability
174
5
28.3482
1.1E-27
32.1459
2.0E-31
58.2545
3.6E-56
2.31559
0.04151
21.4690
6.7E-21
1.16105
0.32611
12.2991
9.7E-12
0.10286
0.99158
2.98831
0.01084
1.48645
0.19105
20.7727
3.3E-20
1.91422
0.08894
174
5
174
5
174
5
174
5
174
5
ARDL APPROACH TO
COINTEGRATION
F-statistics
F(CON|GDP,IND,LOS,PRICE,URB)
3.1012**
F(GDP|CON,IND,LOS,PRICE,URB)
6.3478*
F(IND|CON,GDP,LOS,PRC,URB)
7.2093*
F(LOS|CON,GDP,IND,PRICE,URB)
1.8595
F(PRC|CON,GDP,IND,LOS,URB)
5.5008*
F(URB|CON,GDP,IND,LOS,PRC)
0.88845
significance at 1%, ** at 2.5% levels with respect to Pesaran and Pesaran (1997)
critical values.
*
Bounds Test results indicates that there are four cointegated relations when
dependent variables are selected as annual electricity consumption, GDP,
industry value added and mean adjusted annual average electricity prices
REFERENCES
Pesaran, M.H., Shin, Y., Smith, R.J.,2001. Bounds testing approaches to the analysis of
level relationships. Journal of Applied Econometrics. 16, 289-326.
Pesaran M.H., and Pesaran B.,1997. Working with Microfit 4.0: interactive econometric
analysis. Oxford University Press.
Hamilton J.D.A., 1994. The time series analysis. New Jersey, Princeton University Press.
Engle, R.F., Granger, C.W.J, 1987. Co-integration and error correction: Representation,
estimation and testing. Econometrica. 55, 251-276.
World Bank Statistics Service (data resource). Available at: http://data.worldbank.org/
(Accessed: 4.29.2013).
TEIAS Electricity Statistics (data resource). Available at:
http://www.teias.gov.tr/istatistikler.aspx (Accessed: 4.29.2013).