Environmental Engineering Reference
In-Depth Information
Table 4 TVAR model for EUAs and EU industrial production
Diagnostic tests
Statistics
p
-value
LR Test of linearity (1 versus 2)
39.2378
0.0000
LR Test of linearity (1 versus 3)
77.3049
0.0000
LR Test of TVAR(
1
) against TVAR(
2
) (2 versus 3)
38.0671
0.0000
Best unique threshold (Delay = 1)
0.0072
Second step threshold (Delay = 2)
0.0302
Lower Regime
D(EUAFUT)
D(PRODIND)
Constant
0.0346***
0.0555***
(0.0007)
(0.0006)
D(EUAFUT)-1
7.4624e-06***
5.4364e-05
−
(4.1612e-07)
(1.1715e-01)
D(PRODIND)-1
−
2.5196e-02***
−
0.9221***
(7.3649e-06)
(
0.0004)
−
D(EUAFUT)-2
1.6839e-02***
5.5582e-05
(2.4155e-05)
(1.9330e-01)
D(PRODIND)-2
5.8173e-02***
0.9687***
−
−
(
6.9728e-06)
(
0.0008)
−
−
Higher Regime
D(EUAFUT)
D(PRODIND)
Constant
0.09878***
0.09020***
(0.0003)
(0.0002)
D(EUAFUT)-1
5.0073e-03***
−
0.0938
(2.7073e-06)
(
−
0.1770)
D(PRODIND)-1
5.8413e-02***
0. 0270***
−
(3.225e-06)
(
0.0001)
−
−
0.0824***
−
0.4474
(0.0009) (0.7693)
D(PRODIND)-2 0.0187*** 0.3778***
(0.0007) (0.0003)
Note
EUAFUT stands for the ECX EUA Futures price, and PRODIND for the EU Industrial
Production Index. The operator D(.) refers to the log-return transformation of the time series. ***
indicates statistical signi
cance at the 1 % level, ** at the 5 % level, and * at the 10 % level. The
values between parentheses denote the standard errors of the estimated coef
cients
D(EUAFUT)-2
5 Results
Estimation results are presented for a TVAR model containing the CO
2
futures
price and the EU industrial production index. The goal of this procedure is to assess
the sensitivity of carbon prices relative to macroeconomic activity in the EU, as
proxied by the in
uence of industrial production.
The results from the LR linearity tests are shown in the top panel of Table
4
.
During the
rst step, the null hypothesis of linearity is clearly rejected in favor of
Search WWH ::
Custom Search