105 /304
105
臺大管理論叢
第
28
卷第
1
期
coverage, since the information content of news reports may affect the trading behavior of
investors through the media channel (Hypothesis 4); the alternative models are examined
below.
Model 1 (M1) examines whether CSR winners can promote stock market
performance, whilst Model 2 (M2) similarly examines whether media reputation around
CSR award announcement periods can promote stock market performance, with Model 3
(M3) subsequently further examining whether the media reputation of CSR winners can
promote stock market performance.
(14)
(15)
(16)
where
RET
i
are the stock returns and CARs of the
i
th
firm during the period from the
CSR announcement date (
T
0
) to
T
4
days afterwards;
CSR
i,T
0
is the CSR dummy for the
i
th
firm in the current year; the CSR dummy is equal to 1 if the firm is a CSR award
recipient; otherwise 0;
X
i,k
are control variables for the
i
th
firm, including
TURN, MV
and
P/
B
, during the period from
T
4
days prior to the CSR announcement date to the CSR
announcement date (
T
0
), where
K
is the total number of control variables.
MP
i,l
are the
l
th
media proxies for the
i
th
firm during the CSR award announcement period, which include
MEDIA, SRso
and
MR
during the period from
T
1
days prior to the CSR announcement date
to the CSR announcement date (
T
0
) and we also incorporate
MP
i,l,[T
0
,T
1
]
from
T
0
to
T
1
days
afterwards;
l
equals 1 to
L
,
L
is the total number of media proxies.
INT
i,l
is the interaction term which includes three variables, with
CSR
i,T
0
interacting
with
MP
i,l
during the period from
T
1
days prior to the CSR announcement date to the CSR
announcement date (
T
0
) and we also incorporate
INT
i,l,[T
0
,T
1
]
from
T
0
to
T
1
days afterwards.
Time
T
0
refers to CSR award announcements in the current year; time –
T
1
refers to the