(外文) ESG评级分歧、外部关注与股票回报
(外文) ESG评级分歧、外部关注与股票回报
Economics Letters
journal homepage: www.elsevier.com/locate/ecolet
article info a b s t r a c t
Article history: Using a comprehensive dataset comprising various ratings on ESG performance of A-listed companies
Received 6 June 2023 in China, we find that ESG rating disagreement exerts a significantly negative influence on both
Received in revised form 18 July 2023 stock returns and volatility. External attention serves as a crucial moderating factor influencing the
Accepted 23 July 2023
relationship.
Available online 7 August 2023
© 2023 Elsevier B.V. All rights reserved.
JEL classification:
G24
G38
Keywords:
ESG rating disagreement
Stock returns
Volatility
External attention
https://doi.org/10.1016/j.econlet.2023.111268
0165-1765/© 2023 Elsevier B.V. All rights reserved.
R. Tan and L. Pan Economics Letters 231 (2023) 111268
volatility after 2020 may be affected by the outbreak of Covid-19. Management −0.070 −0.890**
(−0.46) (−2.19)
Then we remove firm-year data with missing values, and obtain
20,927 firm-year observations. Roe 0.010** 0.003
(2.39) (0.98)
The regression equation we use in this paper is as follows: Board 0.007 −0.048
(0.13) (−0.28)
yi,t +1 = β0 + β1 ESGdisi,t + γ W it + δi + ρt + εit (1) Indep 0.167 0.276
(1.57) (0.81)
where the dependent variable yi,t +1 is the annual average level
Femboard −0.020 0.013
and volatility of weekly stock returns. We follow the measure-
(−0.32) (0.07)
ment of Avramov et al. (2022) to construct ESGdisi,t . We first
Growth 0.000 0.000
divide the ESG scores year by year into quartiles on the overall
(0.98) (1.33)
ranking, then calculate the standard deviation of pairwise per-
Risk 0.000*** 0.000***
centile ranking across all rater pairs. To denote the degree of ESG
(12.47) (21.73)
rating disagreement, we take the average standard deviation of
Tobinq −0.039*** −0.003
pairwise percentile ranking for each rater pair as the key inde- (−6.08) (−1.16)
pendent variable. We control for year fixed effects (ρt ) and firm
Constant 0.263*** 5.473*** 6.007*** 11.573***
fixed effects (δi ) to exclude confounding factors that are constant (37.54) (8.58) (236.50) (14.74)
across firms and across years. Standard errors are clustered at the
Firm FE Yes Yes Yes Yes
industry level. The definitions of variables and summary statistics Year FE Yes Yes Yes Yes
are presented Table A.1. Observations 14,534 14,534 14,534 14,534
R-squared 0.558 0.570 0.660 0.662
3. Results and discussions
Notes: *** p < 0.01; ** p < 0.05; *p < 0.1, t-values are in parentheses.
We present the estimates of the impacts of ESG rating dis- 3.3. The moderating effect of external attention
agreement on the level and volatility of stock returns in Table 1.
After controlling for the firm and year fixed effects, ESG rating dis-
agreement reduces both the stock returns as well as its volatility. ESG rating disagreement would erode investor confidence,
When ESG rating disagreement increases by one unit, the stock leading to a negative perception of the company’s ESG perfor-
returns and volatility decrease by 0.001 units and 0.004 units, mance (Hu et al., 2023; Darnall et al., 2022). The growing external
respectively. Gibson Brandon et al. (2021) find that stock returns attention makes it easier for information about ESG rating dis-
of firms in S&P 500 index between 2010 and 2017 are positively agreement to spread and become public knowledge, potentially
related to ESG rating disagreement, but they use the standard leading to a drop in the company’s stock price due to reduced de-
deviation of the available ESG rating from the seven different data
mand. Stock returns reflect the market sentiment and investors’
providers for a given firm at a given point of time. Anselmi and
confidence. Therefore, those firms with more external attention
Petrella (2022) find that ESG rating has no impact on stock returns
for firms listed in Europe and the United States in the period on its ESG rating disagreement may suffer more decline in stock
2010–2020. Compared to them, we get a different finding in the returns and volatility.
context of China. We use the number of analysts tracking the firm annually, the
total number of news that in the titles with the name of the firms
3.2. Robustness checks and Baidu Index to proxy for external attention from analysts,
media and public, respectively. Dum in Table 2 is an indicator
Given that different industries tend to hold distinct prefer- of whether the firm receive more attentions from analysts, me-
ences for ESG ratings and thus the variations in assessment of dia and the public respectively in columns (1)–(2), (3)–(4) and
ESG rating within the industry might matter more than the whole
(5)–(6). The coefficient of the interaction term (ESGdis*Dum) is
ranking. We conduct two kinds of robustness checks. First, we
measure it of each firm within different industries to account for statistically significant at least at the 5% level except the column
the industry-specific factors instead of measuring the ESG rating (2). The results that show all the variables are in Table A3. In
disagreement in the whole sample. Second, we control the fixed general, the results indicate that ESG rating disagreement reduces
effects of industry by year to eliminate the specific impacts of the level and volatility of stock returns more in firms that receive
different industries. The results remain robust in Table A.2. more media and public attention.
2
R. Tan and L. Pan Economics Letters 231 (2023) 111268
Table 2
The moderating effect of external attention.
Dependent (1) (2) (3) (4) (5) (6)
variable FRet FSigma FRet FSigma FRet FSigma
ESGdis 0.000 −0.003* 0.001* −0.002 0.002** 0.002
(0.08) (−1.94) (1.96) (−1.38) (2.38) (0.68)
ESGdis*Dum −0.002** −0.002 −0.004*** −0.003*** −0.005*** −0.011**
(−2.54) (−0.73) (−4.75) (−2.98) (−3.71) (−2.87)
Controls Yes Yes Yes Yes Yes Yes
Firm FE Yes Yes Yes Yes Yes Yes
Year FE Yes Yes Yes Yes Yes Yes
Observations 14,055 14,055 13,384 13,384 13,535 13,535
R-squared 0.573 0.663 0.580 0.674 0.555 0.673
Notes: *** p < 0.01; ** p < 0.05; *p < 0.1, t-values are in parentheses.
Our findings suggest that ESG rating disagreement has ad- Supplementary material related to this article can be found
verse effects on the level and volatility of stock returns. The online at https://doi.org/10.1016/j.econlet.2023.111268.
heterogeneous analysis shows that the adverse impact of ESG
rating disagreement on stock returns and volatility is larger in References
firms receiving more attentions from the analysts, media and the
public. Anselmi, G., Petrella, G., 2022. ESG ratings: Disagreement across providers and
effects on stock returns. Available at SSRN 4328468.
Overall, the findings suggest that higher levels of rating dis- Avramov, D., Cheng, S., Lioui, A., et al., 2022. Sustainable investing with ESG
agreement indicate increasing uncertainty and potential risks, rating uncertainty. J. Financ. Econ. 145 (2), 642–664.
and thus promoting standardized reporting frameworks can con- Berg, F., Koelbel, J.F., Rigobon, R., 2022. Aggregate confusion: The divergence of
tribute to more accurate and consistent ESG ratings, providing a ESG ratings. Rev. Finance 26 (6), 1315–1344.
Brandon, R.Gibson., Krueger, P., Schmidt, P.S., 2021. ESG rating disagreement and
more reliable foundation for investors and market participants. stock returns. Financ. Anal. J. 77 (4), 104–127.
Furthermore, the cross-national research into the disagree- Chatterji, A.K., Levine, D.I., Toffel, M.W., 2009. How well do social ratings actually
ment in ESG and their long-term impacts on corporate opera- measure corporate social responsibility? J. Econ. Manag. Strategy 18 (1),
tional efficiency remains a fertile ground for future studies. 125–169.
Christensen, D.M., Serafeim, G., Sikochi, A., 2022. Why is corporate virtue in the
eye of the beholder? The case of ESG ratings. Account. Rev. 97 (1), 147–175.
Data availability Darnall, N., Ji, H., Iwata, K., et al., 2022. Do ESG reporting guidelines and
verifications enhance firms’ information disclosure? Corp. Soc. Responsibil.
Environ. Manag. 29 (5), 1214–1230.
Data will be made available on request. Escrig-Olmedo, E., Muñoz-Torres, M.J., Fernández-Izquierdo, M.Á., et al., 2014.
Lights and shadows on sustainability rating scoring. Rev. Manag. Sci. 8,
559–574.
Acknowledgments
Hu, X., Hua, R., Liu, Q., et al., 2023. The green fog: Environmental rat-
ing disagreement and corporate greenwashing. Pacific-Basin Finance J. 78,
This paper is funded by MOE (Ministry of Education in China) 101952.
Project of Humanities and Social Sciences (No. 22YJC790112) and Widyawati, L., 2021. Measurement concerns and agreement of environmental
social governance ratings. Account. Finance 61, 1589–1623.
Fundamental Research Funds for the Central Universities, China Zhang, X., Zhao, X., Qu, L., 2021. Do green policies catalyze green investment?
(No. JZ2022HGTA0352). Evidence from ESG investing developments in China. Econ. Lett. 207, 110028.