China Policy Journal Volume 1, Number 1, Fall 2018 | Page 42

Assessing the Implementation of Local Emission Trading Schemes in China
In the equation, lnReturn i,t
is
the price return of CEA at period t in
the ETS pilot i. Price return series are
used because they are all stationary. L
denotes the lag operator. D denotes the
first difference. µ i,t
is the error term.
Eq.3 is the main equation we want to
estimate, while Eq.1 and Eq.2 are the
restricted equations. The equations are
adjusted from ADL model. Coefficients
on lagged log differences of a variable
reflect the short-run effects, while the
joint F-test on all values of a variable
tells the Granger causality.
3. Empirical Analysis
3.1. Descriptive Analysis
Table 4 displays the descriptive
statistics of CEA price, price returns
and trading volumes. Beijing
and Shenzhen ETSs had the highest
CEA prices on average, which were
7.805 and 6.911 $/ton CO 2
e respectively.
Shenzhen CEA price is highly skewed,
with a large right-handed tail, meaning
that the majority of the price data
is lower than the average price. The distribution
of Beijing CEA price, on the
other hand, is both positively skewed
and leptokurtic. The largest CEA price
happened at Shenzhen which was
18.360 $/ton CO 2
e in week 42 of 2013.
Chongqing ETS had the lowest average
CEA price, 3.32 $/ton CO 2
e. Based
on a two-region dynamic CGE model,
Wang et al. (2015) suggested that the
carbon price should be about 38$ CO 2
e
for reaching the Copenhagen target
of 40%–50% reduction of CO 2
emission
intensity toward 2020 in relative
to 2005 level. However, even in Shenzhen,
its highest carbon price was less
than 20$/ton CO 2
e, far from the ideal
price suggested by Wang et al. (2015).
For Chongqing ETS, its price once decreased
to 0.163 $/ton CO 2
e, which provided
very poor incentives for emitters
to make environmental changes.
In most piloting ETSs, CEA prices
were at a high level in the initial operation
stage and exhibited a general
decrease trend over time, which may
explain the positive skewness in most
cases (see Table 4). Take Shenzhen ETS
as an example. Its carbon prices were
at a level of more than 10 $/ton CO 2
e
during June 2013 to June 2014. Then,
the CEA price gradually decreased
to about 5 $/ton CO 2
e by the end of
2014. From January 2015 to June 2017,
the CEA price reached a relatively stable
status, ranging between 3 and 8 $/
tonCO 2
e. Hubei CEA prices decreased
from about 4 $/ton CO 2
e in 2014 to
less than 3$/ton CO 2
e in 2017, but the
distribution showed a slight negative
skewness. Chongqing CEA price, however,
had a peak in the end of 2016 due
to stricter CEA allocation, but it decreased
to less than 1$/ton CO 2
e in the
second quarter of 2017.
We can see the descriptive statistics
of the spot trading volumes from
Table 4 and Figure 1. Volume t
denotes
the total trading volume of CEA in each
week. It was quite different across ETS
pilots, ranging from 16422 tons CO 2
e/
week in Chongqing ETS to 242541 tons
CO 2
e/week in Guangdong ETS. The
two provincial level ETSs, Guangdong,
and Hubei, obviously have the higher
amount of weekly trading volumes
than the five city-level ETSs. But, the
maximum trading volume happened
39