China Policy Journal Volume 1, Number 1, Fall 2018 | Page 50

Assessing the Implementation of Local Emission Trading Schemes in China
Note: The dependent variable is the price returns (lnReturn t
) of CEA in the ETS. *, ** and *** denote significance at 1%, 5% and 10% levels.
Standard errors are in parentheses.
ly granger cause the changes in
CEA price of Guangdong. In Hubei,
however, we found no significant
short-term effects or Granger
causality between energy prices
and CEA prices.
The results also show that
there was a slight decrease of
Guangdong CEA price after June
2014 and a slight decrease of Hubei
CEA price after June 2015.
Guangdong ETS started to operate
from the end of 2013, whereas
Hubei ETS was established in
April 2014. So, both ETSs had
higher CEA prices at the initial
operation stage, but the CEA prices
decreased after the compliance
period in the second calendar
year.
3.4 Multivariate Regressions
for City-Level ETSs
We ran multivariate regressions for
the three better operated city-level
ETSs, Beijing ETS, Shanghai ETS
and Shenzhen ETS, as shown in
Table 9. The dependent variables
are the CEA price returns (lnReturn
t
) of Beijing, Shanghai, and
Shenzhen respectively.
It was found that the one
week lagged Brent oil price return
(D.lnBrent t
) had a short-term positive
effect on Beijing CEA price at
5% significance level. It indicates
that the increase of oil price led
to the short-run increase of Beijing
CEA, as enterprises would
substitute coal for oil to some extent,
resulting in an increase in
47