We compared the environmental performance and financial performance between green fund and non-green fund to explore the balance between environmental goals and financial objectives of green funds by in the context of China government’s strong advocacy of “developing green finance and setting up green development fund”. In this paper, the open-ended active fund with the shareholding ratio of over 75% is taken as the object of study. By the PSM method, 28 green funds are matched by 1:5 to obtain 140 non-green funds. Through the large sample data of 2010-2016, the results show that: 1) the green fund investment portfolio is cleaner than non-green fund. From the long-term environmental performance of the holding enterprises, the green and the non-green fund’s environmental performance and financial performance are tested empirically. Fund environmental performance is better than non-green fund. From the perspective of the short-term environmental performance of the holding enterprises, there is no difference between the environmental performance of the green fund and the non-green fund. 2) The excess return of the green fund is higher than that of the non-green fund. In the period of China’s economic transformation, green funds not only assume the responsibility of investing in environmental performance of enterprises, but also take into account its financial performance, to achieve a win-win result of low pollution and high yield.
China has developed rapidly since the reform and opening up. But due to deforestation and arbitrary emissions, environmental pollution is increasingly serious in China. Only 28% of the 500 monitored sites in China reported potable water quality, where one third of the water is contaminated and not suitable for drinking, or used as agricultural or other domestic water. In October 2011, the US embassy data showed that the PM2.5 index was more than 300 times, marked “toxic”, in Beijing. In order to solve the above problems, according to the Chinese Ministry of Environmental Protection estimates, “thirteen” green industry during the annual investment of at least 2 trillion Yuan, but in the past two years, China’s central and local finance can only come out more than 200 billion Yuan Support for environmental protection, energy saving, new energy and other green investment. It is urgent for China to use green financial tools leveraging social capital and accelerate green transformation of China’s economics.
It is undoubted that the emergence of green funds provides a viable solution to satisfy needs of funds for the green industry development. The People’s Bank of China, the Ministry of Finance and other seven ministries and commissions jointly issued the “Guidance on the construction of green financial system”, stressing that the main purpose of building a green financial system is to mobilize and stimulate more social capital into Green industry, and more effectively curb polluting investment on August 31, 2016. The State Council issued the “thirteen” eco-environmental protection plan, clearly putting forward the establishment of market-oriented operation of the various types of green development fund. It took “developing green finance and establishing green fund” as its theme on December 5, 2016.
The Green Fund does not currently have a unified definition. Fernando et al. [
In the background of vigorously promoting the development of green finance
Green Fund | Non-green Fund | |||
---|---|---|---|---|
Year | Size (billions) | quantity | Size (billions) | quantity |
2010 | 1.75 | 3.00 | 2480.00 | 776.00 |
2011 | 3.16 | 5.00 | 2140.00 | 1010.00 |
2012 | 2.79 | 6.00 | 2790.00 | 1345.00 |
2013 | 5.25 | 9.00 | 2990.00 | 1852.00 |
2014 | 13.40 | 15.00 | 4560.00 | 2328.00 |
2015 | 50.40 | 36.00 | 8380.00 | 3417.00 |
2016 | 38.90 | 46.00 | 9080.00 | 5120.00 |
Data source: http://www.resset.cn/.
and the establishment of green funds, the green fund on the one hand implement its protection of the environment. On the other hand, it is required to achieve the economic goals for investors to create revenue. So, is the green fund taking into account its two goals? This paper presents the following two questions: (1) Is green fund more green than non-green fund? 2) Is the green fund financial performance better than non-green fund?
In this paper, 28 gold funds are matched with 140 funds according to the ratio of 1:5 by the nearest neighbor orientation score matching method. First of all, this paper examines the differences between green fund and non-green fund environmental performance, to test whether the green fund to achieve its environmental goals. Secondly, this paper uses the Carhart four-factor model to measure the financial performance of the fund to test the difference between the financial performance of the green fund and the non-green fund.
The academic contribution of this paper is to enrich the literature of social responsibility institutional investors. Most of the research has focused on the financial performance of socially responsible investment funds, and there are few studies on green funds. However, the green funds are different from those of other institutional investors. For example, the industry the green fund investment in is highly correlated with environmental investment, such as low-carbon, new energy industry. In the context of China’s unique, green fund environmental performance is even more important. As the green fund in the country also belongs to the relatively new topic in China, the empirical research about Chinese green funds is almost blank. It is also of great theoretical and practical significance to focus on the investment strategy of green fund, guiding the investment decision of the green investment fund reasonably and promoting the development of green industry.
The remainder of the paper is organized as follows. In Section 2, we review the literature on the environmental performance and financial performance of green mutual funds. In Section 3, we provided 2 hypotheses according to the theory analysis. In Section 4, we describe the sample selection and the variables in the study. In Section 5, we compared the environmental performance and financial performance between green funds and conventional funds. In section6, we did a robustness test. We conclude in Section 7.
With the rapid increase in green industry, there triggered more and more academic discussion about green investment funds. Bollen [
The impact of environmental screening on fund performance is contradictory. JD Diltz [
The current opinions on whether social responsibility funds and green funds are able to achieve good financial performance are not consistent. On the one hand, companies that invest in environmental performance are likely to reduce their investment opportunities [
Some scholars have compared the Green Fund with the market returns of other funds. Climent and Soriano [
From the macroeconomic impact of the macroeconomic policy, in the context of China’s current system, the environmental performance of poor enterprises is facing greater legitimacy pressure [
H1: Environmental performance of green fund is better than non-green fund.
Is the green fund outperforming the regular fund? At present, there is no consensus on the academic community.
On the one hand, based on the theory of portfolio, environmental screening narrows the investment area, resulting in the green fund failing to take the optimal portfolio. For traditional funds, there is no industry limit, they will not miss the opportunities brought about by the cyclical nature of the financial market [
On the other hand, the financial performance of the Green Fund may be superior to other funds in the long term, as financial markets underestimate social responsibility in the short term [
H2: There is no significant difference in market excess returns between green funds and conventional funds.
The green fund, which is defined in this paper, is the open funds whose name or investment standards including key words of “environmental protection”, “green”, “low carbon”, “sustainable”, “new energy”, “low carbon”, “ecology” and “environment”. Because the first set of green funds-HSBC Jinxin low carbon pioneer stock securities investment fund was established in June 2010, so this sample period covers the period of 2010 to 2016. Besides, the fund holding ratio is greater than 75%. The funds included are active funds, that is, filter out the passive index funds. We choose all of China’s open-end fund from 2010 to 2016 screening out the proportion of less than 0.75 or index funds to get 1051 funds, including 1023 non-green funds, 28 green funds. This sample is the matching sample of this article.
The size of the fund, age, type of investment, monthly rate of return data from the Risi Financial Research Database (RESSET). Rui Si database founded by the Beijing JuyuanRui Si Data Technology Co., Ltd., covering stocks, foreign exchange, bonds, futures, funds, macro statistics, industry statistics and other data. Enterprise industry data from the CSMAR database, environmental performance data from the annual report of the enterprise manual collection.
The research variables include fund environmental performance, fund financial performance, fund size, fund age, fund investment type, enterprise industry attribute, enterprise sewage charges, enterprise environmental investment, risk- free interest rate, market factors, scale factors, value factors, and momentum factors (
This paper draws on the practice of Ghoul and Karoui [
E P 1 = ∑ i = 1 N t , i w i , j , t ∗ E P i , t (1)
Variable name | Variable symbol | Variable definitions |
---|---|---|
Fund grouping | Green | Fund classification, if the green fund is assigned to 1, otherwise assigned to 0 |
Fund environmental performance | EP1 | According to the enterprise industry attributes calculated fund environmental performance |
EP2 | According to the enterprise sewage charges calculated fund environmental performance | |
EP3 | According to the enterprise environmental investment calculation of the fund environmental performance | |
Fund financial performance | Yg | Green Fund Group Monthly average market excess returns |
Yq | Non-green fund group average monthly excess returns | |
Fund size | Formsize | Fund formation size |
Fund age | age | The age of fund at the end of 2016 |
Fund invest type | Investtype | Fund investment type classification, 0 for the value type, 1 for the balance, 2 for the growth |
Risk-free rate | Rf | “Three-month central bank bills” coupon rate |
Market factors | MKT | CSI 300 Index monthly rate of return and the average monthly risk-free rate of difference |
Scale factor | SMB | The difference between the monthly rate of return of small-scale companies and large-scale companies |
Value factor | HML | High book market value than the stock and low book value than the stock monthly rate of return difference |
Momentum factor | UMD | Monthly momentum effect |
Enterprise environmental performance | indu | Enterprise industry attributes, heavily polluting industries 1, non-polluting industries 0 |
sewage | Log (enterprise sewage +1) | |
epi | Log (enterprise environmental protection investment +1) |
where wi,j,t represents the share of stocki at the end of year t in fund j; Nj,t represents the number of shares held by fund j at the end of year t; EPi,t represents environmental performance of stock i at the end of the year.
As there is not unified environmental performance score in China, we try to evaluate the environmental performance of enterprises according to government regulations or previous literature, and then calculated according to formula (1) to obtain the fund environmental performance.
1) Enterprise industry-weighted fund environmental performance
In this paper, the company’s environmental performance measurement first reference to the 2008 published “listed companies environmental verification industry classification management list” and the China Securities Regulatory Commission 2012 edition of “listed companies industry classification guidelines”, which divided enterprises into heavily polluting industries and non-polluting industries. The “List of Management Regulations on Environmental Protection Industry of Listed Companies” states that the heavily polluting industries include thermal power, steel, cement, electrolytic aluminum, coal, metallurgy, building materials, mining, chemical, petrochemical, pharmaceutical, light industry (brewing, paper making, Textile and tanning. The industry attribute can be used as the initial screening of enterprise environmental performance. According to the Formula (1), fund environmental performance EP1:
E P 1 = ∑ i = 1 N t , i w i , j , t ∗ I n d u i , t (2)
where wi,j,t represents the share of stock i at the end of year t in fund j; Nj,t represents the number of shares held by fund j at the end of year t; Indui,t represents stock i at the end of the year t industry, 1 if it is in heavy polluting industry, else is 0.
2) Environmental performance of the enterprise’s environmental protection investment
Second, based on Li Wenjing and Lu Xiaoyan [
According to the Formula (1), fund environmental performance EP2:
E P 2 = ∑ i = 1 N t , i w i , j , t ∗ e p i i , t (3)
where wi,j,t represents the share of stock i at the end of year t in fund j; Nj,t represents the number of shares held by fund j at the end of year t; epii,t represents the expenditure of stock i at the end of the year t.
3) Corporate emissions charges weighted fund environmental performance
According to the Regulations of the Chinese State Council on the Administration of the Collection and Use of Sewage Charge (Chinese State Council Decree No. 369), it is necessary for the units and individual industrial and commercial households that pull directly pollutants into the environment to pay the sewage charges. Sewage charges include sewage charges, solid waste and hazardous waste discharge charges, excessive noise pollution charges. Sewage charges are measured according to the type of pollutants and the amount of pollution equivalent and each pollution equivalent is 0.7 Yuan. Enterprise sewage charges can be used as short-term environmental performance of enterprises, which is objective to reflect the level of pollution.
According to the Formula (1), fund environmental performance EP2
E P 3 = ∑ i = 1 N t , i w i , j , t ∗ s e w a g e i , t (4)
where wi,j,t represents the share of stock i at the end of year t in fund j; Nj,t represents the number of shares held by fund j at the end of year t; sewagei,t represents the sewage charges of stock i at the end of the year t.
Based on Carhart [
R t − R f , t = ∂ + β M K T ( r t m − r f , t ) + β S M B r t s m b + β H M L r t h m l + β U M D r t u m d + η t (5)
where Rt is the average yield of fund at month t, which means Rf,t average risk- free rate of return in month t, Rm represents the average yield of the market portfolio, ∂ representing the excess return on the fund’s investment, η t indicating heterogeneous gains; r t s m b , r t h m l , r t u m d representing market factors, scale factors and momentum factors,.
Based on Mallin et al. [
First, the size of the fund. Mallin et al. [
Second, the fund age. Fund age is the age of the fund as of December 31.
Third, the type of fund investment. Gregory et al. [
First, we enter the PSM first stage regression. We use the following log it regression model (model 1) to calculate the propensity score to pair.
Logit ( g r e e n i ) = ∂ 0 + ∂ 1 F o r m s z i + ∂ 2 A g e i + ∂ 3 I n v s e t t y p e i + ε i (6)
Secondly, we calculate the average treatment effect, referred to as “ATT”: The theoretical framework of the tendency analysis is the “counterfactual inference model”, which is mainly used to solve the causal inference problem of sample selectivity deviation. Suppose that each individual i receiving an intervention has two potential outcomes (Y0i, Y1i) that represent the potential results that are not intervened and intervened, respectively. Let Di = 0 indicate that the sample has not been intervened, Di = 1 indicates that the sample is intervened, The counterfactual framework model is:
Y i = D i Y 1 i + ( 1 − D i ) Y 0 i = Y 0 i + ( Y 1 i − Y 0 i ) (7)
where ( Y 1 i − Y 0 i ) is the processing effect of i. Because it is a random variable, so we are concerned about its expectations, that is “ATT”:
ATT = E ( Y 1 i − Y 0 i ) = E ( Y 1 i − Y 0 i | D i = 1 ) (8)
Thirdly, this paper takes k nearest neighbor matching to match. Since the propensity score calculated by model (5) is a continuous variable, we cannot directly match exactly the same tendency score to calculate ATT. In this paper, we take k nearest neighbor matching to match.
Let Pi and Pj be the inclination values of the intervention group and the members of the control group, respectively. I0 and I1 are the set of control group and intervention group members respectively. The nearest neighbor 1:n matches means that for each intervention member i, n members are found to fall d(Pi), d(Pi) is the distance between the intervention group and the control group:
d ( P i ) = min ‖ P i − P j ‖ , j ∈ I 0 (9)
When the distance between the intervention group and the control group is less than d(Pi), the control group is regarded as the successful matching object of the intervention group. At this point we can calculate the estimated value of ATT:
τ = 1 N T ∑ i ∈ T Y i T − 1 N C ∑ i ∈ C w j Y j C (10)
ATT variance is:
V a r ( τ ) = 1 N T V a r ( Y i T ) + 1 ( N C ) 2 ( w j ) 2 V a r ( Y j C ) (11)
The difference between ATT and the variance is obtained, and the difference between the matching group and the control group can be obtained.
First, we uses the logit model to estimate propensity score. In
see whether it is green fund is not significant correlated with the investment type, and positively correlated with the initial size of the fund at the 5% significance level, And negatively correlated with fund age the significance level of 5% we use 28 green fund to match 140 non-green fund with the ratio of 1: 5.
green | |
---|---|
cons | −9.167*** |
(−3.030) | |
lnformsize | 0.306** |
(2.080) | |
age | −0.039** |
(−2.020) | |
Investtype 1 | 0.345 |
(0.72) | |
Investtype 2 | 0.103 |
(0.16) |
Data source: Data for this article.
groups becomes very close, indicating that the characteristics of the green fund and the non-green fund become similar after matching.
The fund yield data for this paper are the monthly sequence values for a total of 78 months from July 2010 to December 2016. It can be seen that the average
Variable | Sample | Mean | Bias% | Decrease% | T test | |
---|---|---|---|---|---|---|
Treated | Controlled | |||||
lnformsize | Before matching | 20.74 | 20.25 | 37.2 | 98.3 | 1.72* |
After matching | 20.74 | 20.75 | −0.6 | −0.14 | ||
age | Before matching | 0.853 | 1.041 | −33.4 | 97.6 | −1.72* |
After matching | 0.853 | 0.858 | −0.8 | −0.09 | ||
Investtype | Before matching | 0.928 | 0.842 | 13.6 | 66.6 | 0.69 |
After matching | 0.928 | 0.957 | −4.6 | −0.17 |
Data source: Data for this article.
Variable | Sample | Treated | Controlled | difference | SD | T score |
---|---|---|---|---|---|---|
EP1 | Before matching | 0.241 | 0.268 | −0.0269 | 0.0207 | −1.3 |
ATT | 0.241 | 0.278 | −0.0373 | 0.0173 | −2.15 | |
EP3 | Before matching | 11.56 | 11.57 | −0.0182 | 0.575 | −0.03 |
ATT | 11.56 | 11.64 | −0.0783 | 0.59 | −0.13 | |
EP2 | Before matching | 41.73 | 40.52 | 1.207 | 0.839 | 1.44 |
ATT | 41.73 | 39.72 | 2.001 | 0.852 | 2.35 |
Data source: data for this article.
yield of green funds is higher than non-green fund in
Due to the time series data, we first do unit root test, see
Variable | Sample | Means | SD | Min | Max |
---|---|---|---|---|---|
yg | 78 | 0.0089 | 0.0676 | −0.2330 | 0.1910 |
yc | 78 | 0.0054 | 0.0699 | −0.2430 | 0.1800 |
smb | 78 | 0.0155 | 0.0460 | −0.1720 | 0.1960 |
hml | 78 | −0.0014 | 0.0392 | −0.1570 | 0.1510 |
umd | 78 | −0.0137 | 0.0668 | −0.2170 | 0.1480 |
rf | 78 | 0.0035 | 0.0008 | 0.0021 | 0.0051 |
Data source: data for this article.
ADF | 1% Critical value | PP | 1% Critical value | conclusion | |
---|---|---|---|---|---|
yg | −7.44 | −3.542 | −7.381 | −3.542 | sTable |
yc | −7.475 | −3.542 | −7.394 | −3.542 | sTable |
mkt | −7.353 | −3.542 | −7.282 | −3.542 | sTable |
smb | −8.633 | −3.542 | −8.652 | −3.542 | sTable |
hml | −9.934 | −3.542 | −10.067 | −3.542 | sTable |
umd | −8.771 | −3.542 | −8.779 | −3.542 | sTable |
Data source: data for this article.
¶ | mkt | smb | hml | umd | R2 | |
---|---|---|---|---|---|---|
yg | 0.0043 | 0.7383*** | 0.0939 | −0.4710*** | 0.1558*** | 0.896 |
yc | −0.0017 | 0.8339*** | 0.1468*** | −0.3362*** | 0.0603*** | 0.967 |
yq−yc | 0.0054*** | −0.0956 | −0.0529 | −0.1348 | 0.0955*** | −0.071 |
*p < 0.1, **p < 0.05, ***p < 0.01. Data source: data for this article.
Non-green intercept item is −0.0017, but it is not statistically significant, that is non-green fund is below the average market returns, but this difference is not significant. The non-green fund market risk factor (MKT) is 0.8339, which is positive at the significant level, indicating that the green fund has a significant positive correlation with the overall market trend, and its volatility is less than the market. The Green Fund size factor (SMB) is 0.1468, which is positive at a1% significance level, indicating that non-green funds are more susceptible to small-cap stocks. Non-green fund value factor (hml) is −0.3362, negative at 1% level of significance, indicating that non-green fund investment performance mostly by the impact of high-growth stocks. The non-green fund’s momentum factor (umd) was 0.0603, which was positive at the 1% significance level, indicating that the non-green fund as a whole was positively correlated with the trend of high-yield stocks.
Further using Suest test method, we compare the green fund and non-green fund four factors model coefficient difference. We can see that the difference between the green fund and the non-green intercept is 0.0054, which is positive at the 1% significance level, indicating that the green fund’s market returns is higher than the non-green fund, that is, the financial performance of the green fund better than non-green funds. In addition, there is no significant difference between the green fund and the non-green fund in the market factors, the scale factor and the value factor, which shows that there is no significant difference between the green fund and the non-green fund in the above factors. In the momentum factor, the green fund is higher than the non-green fund 0.0955, positive at the 1% significance level, indicating that the green fund is more likely to be affected by the high-yield stock trend than the non-green fund.
We use nuclear matching and radius matching method to do robust test. as shown in
We use the Jensen index and the three-factor model to test the financial performance of funds, as shown in
This paper compares the environmental performance and financial performance
variable | sample | Treated | Controlled | Difference | SD | T test |
---|---|---|---|---|---|---|
EP1 | before matching | 0.241 | 0.268 | −0.0269 | 0.0207 | −1.300 |
ATT | 0.241 | 0.268 | −0.0271 | 0.0152 | −1.790 | |
EP3 | before matching | 11.56 | 11.57 | −0.0182 | 0.575 | −0.0300 |
ATT | 11.56 | 11.58 | −0.0185 | 0.509 | −0.0400 | |
EP2 | before matching | 41.73 | 40.52 | 1.207 | 0.839 | 1.440 |
ATT | 41.73 | 40.53 | 1.199 | 0.701 | 1.710 |
Data source: data for this article.
variable | sample | Treated | Controlled | Difference | SD | T test |
---|---|---|---|---|---|---|
EP1 | before matching | 0.241 | 0.268 | −0.0269 | 0.0207 | −1.3 |
ATT | 0.241 | 0.273 | −0.0325 | 0.0157 | −2.07 | |
EP3 | before matching | 11.56 | 11.57 | −0.0182 | 0.575 | −0.03 |
ATT | 11.56 | 11.41 | 0.148 | 0.522 | 0.28 | |
EP2 | before matching | 41.73 | 40.52 | 1.207 | 0.839 | 1.44 |
ATT | 41.73 | 40.26 | 1.467 | 0.722 | 2.03 |
Data source: data for this article.
¶ | mkt | R2 | |
---|---|---|---|
yg | 0.0039 | 0.8113*** | 0.785 |
yc | 0.0004 | 0.8308*** | 0.844 |
yq-yc | 0.0035* | −0.0195 | −0.059 |
*p < 0.1, **p < 0.05, ***p < 0.01. Data source: Data for this article.
¶ | mkt | smb | hml | R2 | |
---|---|---|---|---|---|
yg | 0.0025 | 0.7565*** | 0.0706 | −0.4592*** | 0.873 |
yc | −0.0022 | 0.7649*** | 0.1580** | −0.4019*** | 0.946 |
yq-yc | 0.0047** | −0.0084 | −0.0874 | −0.0573 | −0.073 |
*p < 0.1, **p < 0.05, ***p < 0.01. Data source: Data for this article.
of 28 green funds and 140 matching non-green funds in 2010-2016, and finds that:
1) Green fund portfolio industry configuration is cleaner than non-green fund. From the long-term environmental performance of holding companies, green fund environmental performance is better than non-green fund. This shows that the Green Fund not only invested in the cleaner industry, and actively concerned about the enterprises’ long-term environmental performance.
2) The market excess return of green fund is higher than non-green fund, indicating that the green fund financial performance is superior to non-green fund. In addition, there is no significant difference between the green fund and the non-green fund in the market factors, the scale factor and the value factor, which shows that there is no significant difference between the green fund and the non-green fund in the above risks.
The findings of this paper have the following two aspects contributions: first, the findings of this article enriched the literatures of green funds from the Chinese market on the green fund research; second, this is the first empirical study of Chinese green funds, providing the evidence that Chinese green fund is much greener than non-green fund and achieves higher market returns.
Yuan, Y. (2017) Environmental Performance and Financial Performance of Green Mutual Fund― Evidence from China. Open Journal of Business and Management, 5, 680-698. https://doi.org/10.4236/ojbm.2017.54057