Sustainable stock, low return, or high responsibility? The case of Vietnam stock market

(1)

Sustainable stock, low return, or high responsibility? The case of Vietnam stock market By Lai Thi Thanh Phuong ( S1008302 )

Abstract: “Sustainable stock” has been emerging as a new trend of financial investment channel. Nonetheless, limited research papers are focusing on sustainable stock as it is still in its infancy. The main intention of my paper is to examine the performance of the sustainable stocks in the Vietnam stock market from 24th July 2017 to 28th December 2018. I make a performance comparison between sustainable stocks and the regular stocks using the ordinary least squared with dummy variables, and the Sharpe ratio. Then, I observe the impact of the establishment of the sustainable index – VNSI on the stock price of the included stocks in two different event windows. The number of the stock shows a significant result in the event study increases period by period, but still seems limited. I also find out that sustainable stocks outperform the regular stocks in one out of three periods. Implication and further research direction are also discussed.

Supervisor: Dr. Jianying Qiu

(2)

Table of Contents

Abbreviations: ... 3

1. Introduction: ... 4

2. Literature review: ... 7

3. Data and Methodology ... 9

3.1 Data ... 9

3.2 Methodology ... 10

3.2.1 CAPM ... 11

3.2.2 Fama and French ... 12

3.2.3 Time- Series Analysis ... 13

3.2.4 Event- Study Analysis ... 15

4. Result ... 17

4.1 Hypothesis 1 – VNSI and VN30 performances comparison ... 17

4.1.1 The mean-‐comparison test ... 17

4.1.2 Variance – comparison test ... 18

4.1.3 Sharpe ratio comparison ... 19

4.1.4 OLS regression result ... 19

4.2 Hypothesis 2 -‐ There is a positive effect on the stock prices of being included in the VNSI .. 22

4.2.1 Event Study 1 – 24/07/2017 ... 22

4.2.2 Event study 2 – 23/07/2018 ... 24

5. Discussion and Conclusion ... 26

5.1 Robustness test ... 26

5.2 Main findings and their implications ... 26

5.3 Limitation and Future Research ... 27

5.4 Conclusions ... 27

6. Bibliography ... 28

(3)

Abbreviations:

SRI Socially Responsible Investment

SR Socially Responsible

CAPM Capital Asset Pricing Model

VNSI Vietnam Sustainability Index

OLS Ordinary Least Squared

GIZ German Cooperation Agency

SSC State Securities Commission of Vietnam

GRI Global Reporting Initiative

CAR Cumulative Abnormal Return

AR Abnormal Return

FF Fama and French

SKI Sri Kehati Index

JCI Jakarta Composite Index

(4)

1. Introduction:

People are increasingly concerned about environmental problems such as climate change, greenhouses effect and so on. Similarly, investors also care more about global problems and want to contribute more to the future generation. In the financial market, we can see the contribution to sustainable development by the launch of various green financial products and green funds, which is grouped as a socially responsible investment (SRI). There are several definitions of SRI. Munoz et al. (2004) share the definition of SRI provided by The United Kingdom Social Investment Forum (UKSIF), which states that SRI is investment caring both financial and ethical objectives. Williams (2010) states that “SRI covers all types of investments, including direct share-ownership, hedge funds, corporate and country bonds, and various others.”. Apart from green bonds, the investors can also choose sustainable stocks as a sustainable investment channel, where the investors buy the stocks of the companies contributing their profits to the sustainable purpose. Sparkes (2001) reveals that SRI overgrows in the area of finance, and since 3rd July 2000, all the UK private pension funds have been considered as SRI. Similarly, Consolandi et al. (2009) find that there is a rapid growth of SRI in the USA, with its share in total mutual funds up to 11%. In the report of Williams (2010), OWW Consulting also indicates an increase of 23% of the SRI funds over two years in Asian. In the same paper, Williams also finds that the value of clean-technology launchings in Vietnam, Thailand, Indonesia, and the Philippines increases to US$ 13.9 billion. To show support for sustainable development in the commercial sector, governments develop a national wide green growth strategy. For example, the prime minister of Vietnam encourages all parties to make use of available instruments to establish new green products and green economy (Vietnam prime minister Nguyen Tan Dung, 2012).

As a further attempt in encouraging SRI activities, on July 2017, Vietnam stock market launched the Vietnam sustainability index (VNSI), which is believed to encourage the sustainable development of the listed firms (VOV, 2017). The representative of the HOSE (VOV, 2017) said that the VNSI is the collaboration of the HOSE, German cooperation agency (GIZ) and the state securities commission of Vietnam (SSC). Additionally, the main criteria for this index are built on the Global Reporting Initiative (GRI) standards and the organization for Economic Cooperation and Development (OECD) principle of corporate governance along with Vietnam current regulation on the corporate governance (VOV, 2017). The VNSI is the list of 20 trading stocks, selected from the top 100 largest companies, which has the highest sustainable ratio (Sustainable

(5)

stock exchanges initiative, 2019). This sustainable ratio is the combination of 60% financial criterion, 30% socially responsible criterion, and the last 10% for the environmental criterion (Hochiminh stock exchange, 2018).

Figure 1: The VNSI development

Note: The figure describes the change in the VNSI index value in percentage Source: Hochiminh Stock Exchange

With the increase in value after one and a half year of implementation, there are more and more concerns about the green financial product not only from the Vietnam government but also from the stock investors specifically. The mission of this study is to provide the very first quantitative study about sustainable stock in the Vietnam stock market by answering for the following research

(6)

“In the Vietnam stock market, does sustainable stocks have higher returns than the regular stocks?”

In this paper, I want to examine this question from 2 separate perspectives: from the investor’s perspective and the firm’s perspective. From the perspective of investors, maximizing profit is their primary objective, so do they earn more profit when investing in sustainable stocks comparing with investing the regular stocks? I build the first hypothesis to find the answer to the above question:

H1: Sustainable stocks perform similarly to the non-sustainable stocks.

Meanwhile, from the perspective of the firm, the second hypothesis is built to test whether does it help to increase the market capitalization of the firm after being included in the VNSI index?

H2: There is a positive effect on stock prices of being included in a VNSI index.

My research aims to contribute to the general literature on the Vietnam financial market. As there is still no official study exclusively focus on the sustainable stock in Vietnam, my analysis draws a general idea of the investor’s response toward the creation of sustainable index – VNSI. Secondly, by conducting the comparison between the sustainable and the regular stock’s performances, my analysis provides the investors another key information to consider before investing in a particular stock.

The remainder of this paper is put in the following order. The second section is the literature review, where I provide a general background of the available research on a sustainable stock. In the following section, I illustrate models also the statistical techniques applied in my research. The results will be presented in the fourth part, and in the final chapter, I will come up with the discussion also the conclusion of my research.

(7)

2. Literature review:

When testing the market reward on the firm when going green, Puopolo (2015) finds no evidence for the relationship between implementation of the sustainable standards on the operation or the increase in required earning by the investors. Meanwhile, Galema and Scholtens (2008) find the opposite result when examing SRI portfolio in the same country – the USA, in which, they conclude that SRI has a significant influence on the stock return by reducing the book-to-market value. Oppositely, Lockwood & Prombutr (2010) show that there is a negative relationship between the return of a firm and its past sustainable growth, in which the high sustainable growth firms end up with low book-to-market values and low risks. Along with those controversy conclusions about the impact of SRI on the change in stock return, the incentive of this paper is to find the answer to the question of whether sustainable stocks under or over-perform comparing with the regular stocks. In order to examine the performance of the sustainable stock, using the asset pricing model is necessary. Two well-known asset pricing models will be used in this paper due to their high validity, which is conclusively proved by the amount of research in the financial field. The first model will be used is the capital asset pricing model (CAPM), in which Cuthbertson and Nitzsche (2005) state that the CAPM model allows you to examine the expected excess return on asset. The second model is the Fama and French three-factor (FF) model, which takes the size factor into account. As Fama & French (1995) prove that the size factor helps to explain the return. In which, within the same book-to-market group, the bigger stock is more potentially profitable compared with the small one.

Jagannathan & Wang (1993) insist that although the CAPM model is built on the unrealistic assumption when the real world is quite complex, they still show their strong advocate for the validity of the CAPM model. As they explain, when there is an allowance of beta value variation over time, the CAPM can help explain 57% the variation in the average return of the portfolio. In like manner, Bello (2008) restates that in the CAPM model, they use only the proxy of the market portfolio to estimate the systemic risk of security.

Besides using classical CAPM model to examine the performance of sustainable stocks, I use FF models to take the firm’s size and book-to-market factors into account. Phong and Hoang (2012) make a study in the Vietnam stock market by applying FF three-factors models and CAPM

(8)

scenario of the Vietnam stock market but with the main observations are sustainable stock, I expect that the FF three-factor model will provide the explanation power is superior to the CAPM model. Despite a large number of papers focusing on the performance of the SRI mutual fund, there is still a limitation of the research on the performance of the sustainable index since almost all the sustainable indices have been established recently (Consolandi et al., 2009). In their paper, Consolandi et al. (2009) observe the difference of the performance between the Dow Jones sustainability stock index (DJSSI) and the benchmark Dow Jones STOXX 600 index. After conducting the time-series analysis and also the event study, they hardly find the difference in the performance between DJSSI and DJ Stoxx 600. Likely, Consolandi et al. (2009) also conclude that the sustainably responsible firms, in any case, perform indistinguishably from the other firms. Analogously, Cortex, Silva and Areal (2009) when examining the performance of the European socially responsible (SR) fund also find out the similarity in the performance between the SR funds and the conventional funds. On the contrary, Jones et al. (2008) claim that the SR funds under-perform comparing with the market in Australia.

In Southeast Asia, Zulkafli and Ahmad (2017) recently examine the performance of the Indonesia sustainable index – Sri Kehati index (SKI) by comparing it to the performance of the market index. They conclude by measuring the difference in effectiveness between the sustainable index and the market index. The Sharpe ratio is one of the effectiveness assessment they use to make the comparison. They find out that there is a slight underperformance of the SKI index against the Jakarta composite index (JCI), and the Jensen’s alpha is the only proof for the SKI’s out-performance in the same period.

Lastly, Revelli and Viviani (2015) conduct a meta-analysis of 85 studies and 190 experiments testing on the SRI and its relationship with the financial return. Heterogeneously, the result indicates that the sustainable portfolios might overperform or underperform comparing with the conventional portfolios. Also, I find hardly any research on the Asia SRI in their empirical corpus.

Inspired by the diverse literature, the main objective of my research is to fill in the literature gap of SRI in the Vietnam stock market by examining the reaction of the market toward the new sustainability index formation.

(9)

3. Data and Methodology 3.1 Data

I obtain all the data from the Thomson Reuters Eikon database in the form of daily data. My data cover from 1st June 2016 to 28th February 2019. Although the Ho Chi Minh Stock Exchange (HOSE) was officially launched in 2007 (Sustainable stock exchange initiative, 2019), I refrain from the data between 2007-2015 and use the data from 2016 to 2019 in order to match with the establishment of Vietnam sustainable index. For each estimation period, along with 20 sustainable stocks, in order to answer the research question from the perspective of the investor, I also use the data of enterprises listed in the index VN30 as a benchmark to compare with sustainable stock performance. Similar to VNSI, VN30 is the general index of the HOSE, which is the group of top 30 stocks in the market basing on the market capitalization and liquidity trading.

As both VNSI and VN30 is taken from the top 100 largest stocks in the HOSE, there might be some stocks are included in both VNSI and VN30 index. So, for the VN30, I only take the stock non-included in VNSI to see the difference between the performance of the sustainable stocks and the regular stocks.

In my research, the stocks included in VNSI stand for the sustainable stocks, whose list will be adjusted annually. The regular stocks are all the stocks composed of VN30 index only. Unlike the VNSI index, the HOSE will update the list of stock of the VN30 index twice a year. Finally, I let the return on VN-index represent for the return on the whole stock market in both CAPM and FF model. Table 1 summarizes all the details of variables as below.

Table 1: Data description

Note: The table summarizes all the data information used in this paper

Data name Quantity Data description Data type Data source

Sustainable Stocks 23 The stock is listed in

the VNSI index. Daily data

Thomson Reuters Eikon

(10)

Regular stocks 18 The stock is listed in

the VN30 index. Daily data

database

VN-index 1

The index is calculated as the value-weighted

return of all stocks in the HOSE. Daily data Thomson Reuters Eikon database Vietnam government 1-year bond yield

1 The proxy for the

risk-free rate. Daily data

database

3.2 Methodology

Inspired by the research of both Consolandi el al. (2009) about the US sustainable index and the paper of Zulkafli and Ahmad (2017) about the Indonesia sustainable index, I will apply the quantitative method in this study. I use two different models and two types of analyses to test the hypotheses. I will focus on two main dimensions to compare the performance of the VNSI and VN30 index, namely the risk and the return of the stocks. I first model the relationship between the stock returns and excess market return, then, use the intercept ( Jensen’s alpha) to generally compare the performance of VNSI and VN30 portfolio. I observe risk and return of the firms through the value of the standard deviation and the mean respectively. Since Cuthberson and Nitzsche (2005) contend that the Sharpe ratio is a better instrument to compare the performances of different asset classes then just solely average return when taking the variance in the risk into account. I also use Sharpe ratio to make the performance comparison.

After obtaining the standard deviation and the mean value of both sustainable and regular stocks, the statistical test ( t-test) will be conducted to test whether the difference in the standard deviation and the mean value between two groups of stocks is significant or not. Moreover, in order to have a more precise look on the performance of the sustainable index and the VN30 index,

(11)

I create the dummy variables with return on VN30 index as the reference to examine the significant difference between those two indices.

Along with the standard time-series analysis, I use the event study to examine the abnormal return. This is inspired by the paper of Galai & Masulis (1976), in which they reveal that ‘in an efficient capital market, any new information reaching the market concerning asset values is immediately impounded into security price’. This finding raises the question that if the information on the sustainable stock is released, does it affect the change in stock price? If yes, this effect will be positive or negative?. I will obtain the normal expected returns of both sustainable stocks and non-sustainable stocks by using CAPM model and Fama and French 3-factor model for the event estimation.

3.2.1 CAPM

Cortez, Silva, and Areal (2009) choose the well-known Jensen’s alpha as an unconditional benchmark of the performance when examining the performance of the European social responsible fund. A positive and statistically significant alpha value represents for the higher-skilled performance of the stocks or the fund manager (Cuthbertson and Nitzsche, 2005). In this analysis, I use the Jensen’s alpha as the criterion to compare the performance between the sustainable stock portfolio and the regular one. The symbol “α” stands for the Jensen’s alpha in the CAPM formula below:

(Rit –Rf) =αi +βi(Rmt –Rf)+εit

With R_it: Return of stock (portfolio)

R_f: Return on risk-free asset

Rmt: Return on market portfolio

The normal return of the stock is set as the lognormal of the change in price ( Cuthbertson and Nitzsche, 2005).

(12)

𝑅_"# = ln ( 𝑃"# 𝑃_"#*+) With Pit : the price of stock at time t

Pit -1 : the price of stock at time (t – 1)

3.2.2 Fama and French

The Fama and French three-factor model is the expansion of the CAPM, in which they add two more factors to help to increase the explanatory power of the model. Along with the excess market return factor, the size factor and the book-to-market are proved as capturing more 35% of the return variation over the CAPM in the case of the Vietnam market. The formula for the FF model is described below:

(Rit – rf) = αi + β1iRmt + β2iSMBt + β3iHMLt+ εit With

SMB: the size factor

HML: the book-to-market factor

I follow the construction of the paper of Fama and French (1992). I firstly use the market value of the stock in June of the current year to divide the group of sustainable stocks into two groups: stock with small market value denoted as “S” and the stock with big market value denoted as “B”. Second, I use the book-to-market ratio of the stock in December of the previous year to form three groups: stocks with low book-to-market ratio denoted as “L”, stocks with medium book-to-market ration denoted as “M”, and stocks with high book-to-market ratio denoted as “H”. I thirdly construct six portfolios namely, S/L, S/M, S/H, B/L, B/M/ B/H basing on those two intersections.

Fama and French (1992) describe the size factor – SMB as the gap of the return between small stock portfolio and big stock portfolio. Meanwhile, they characterize the HML factors as the difference between the high book-to-market stock portfolio and the low one. The formulas to

(13)

calculate SMB and HML are presented below:

SMB = [(S/L + S/M + S/H)/ 3] – [(B/L + B/M + B/H)/3] HML = [(S/H + B/H)/ 2] – [( S/L + B/L)/ 2]

3.2.3 Time- Series Analysis

Time-series analysis is a popular technique to examine the statistic aspect of the finance world. There are many researchers choose this type of analysis to test their hypothesis. For example, Le (2015) chooses the time-series analysis when testing the presence of the size and value risk factors in Vietnam stock market. Formerly, Fama and French (1993) choose the time-series regression to study the effect of the size and book to market ratio to the return of stocks and bonds. In the same paper, they highlight the convenience of using time-series analysis method in studying asset-pricing subjects (Fama and French, 1993). So, by running the time-series regression in this paper, I can make a comparison between the sustainable stocks and the regular stocks by looking at the difference in the intercept value of the regression result. This intercept value is also called Jensen’s alpha. Besides, Cuthbertson and Nitzsche (2005) prove that it is more effective to use the Sharpe ratio when comparing the performance of the assets as it provides the excess return for each unit of risk. So that I also so calculate the Sharpe ratio for all portfolios to verify the profitability of the sustainable portfolio.

Different from the VNSI, the index committee plans to review and make the adjustment for the VN30 index every six months ( Dao B, 2013). Within my sample size, I find out that the component of the VN30 index is different for every period. In total, I will run three time-series regression separately for each period. The table below provides the information of each period and the name of stock included in the VNSI and VN30 index of that period.

Table 2: Time-series data description

Note: The table provides the start date, end date of each period, and the abbreviation of stocks for each index in each period.

(14)

From 27/07/2017 to 19/01/2018 BID – BMP – BVH – CNG – CTD – DCM – DHG – DPM – FPT – HSG – MBB – NSC – NT2 – PNJ – PVD – REE – SBT – VCB – VIC – VNM CII – CTG – GAS – GMD – HPG – KBC – KDC – MSN – MWG – NVL – ROS – SAB – SSI – STB

From 20/01/2018 to 20/07/2018 BID – BMP – BVH – CNG – CTD – DCM – DHG – DPM – FPT – HSG – MBB – NSC – NT2 – PNJ – PVD – REE – SBT – VCB – VIC – VNM CII – CTG – GAS – GMD – HPG – KDC – MSN – MWG – NVL – PLX – ROS – SAB – SSI – STB

– VJC From 23/07/2018 to 28/12/2018 BMP – CTD – DCM – DHG – DPM – FPT – GAS – HCM – HSG – MBB – NSG – NT2 – NVL – PNJ – PVD – REE – SBT – SSI – VIC – VNM CII – CTG – GMD – HPD – KDC – MSN – MWG – PLX – ROS – SAB – STB – VCB – VJC – VPB – VRE

Wooldridge (2012) emphasizes how important to keep the time-series analysis stay stationary, which means the unchangeable of the joint probability distribution when we shift the sequence of variable ahead h period. Since the non-stationary variables can make the analysis output biased, I will conduct the specific test to identify the non-stationary problem of my dataset by using the Dickey-Fuller test. This test allows me to test the data set in three different types of time-series stationary: constant mean, constant variance, and covariance structure. I provide the result of these tests in the appendix.

(15)

3.2.4 Event- Study Analysis

As stated in the introduction part, I will approach the research question from 2 separate perspectives: the investor’s perspective and the firm’s perspective. For the firm’s point of view, increasing the value of the firm’s stock is considered as one of the most important goals. Mackinlay (1997) argues that we can see the influence of the event immediately through the change in stock price. Likewise, Cuthbertson and Nitzsche (2005) insist on the benefit of the event study in evaluating the quantitative impact of the events on the firm’s value. Through the change abnormally in return, I can demonstrate the effect of the event, in this case, is the establishment of the sustainable index – VNSI, on the stock value of the sustainable firm. Moreover, this analysis even reveals the propensity of this impact, whether it is negatively or positively influence. Within my research scope, I plan to provide the firms with a precise picture of how the Vietnam investor response to the establishment of the sustainable index.

In the event study, I will calculate the abnormal return (AR) and the cumulative abnormal return(CAR) to conclude whether there is the impact of the event or not. I compute the abnormal return by using the following formula:

ARi,t = (Ri,t –rf) - E(Ri,t - rf )

Abnormal return is exactly the difference between the actual return and the expected return of the stock over the event window (MacKinlay, 1997).

CARi( t1,t2) = #+#.𝐴𝑅"#

Cumulative abnormal return is the sum of all abnormal return included in the event window (Mackinlay, 1997). With the value of the CAR, I can test the null hypothesis that the CAR equals zero. With the 95% confidence, if the value of the t-testCAR equal to or higher than 1.96, I can conclude that there is an existence of the abnormal return. Vice versa, with the value of t-testCAR is smaller than 1.96, I will accept the null hypothesis that the abnormal return does not exist in that certain period.

As mentioned before, 24th July 2017 is the first date HOSE introduced the VNSI index to the stock market. So, I will choose this date as the event date for this analysis. The second event study is on 23rd July 2018, when the HOSE re-estimate the VNSI index after one year. Running two

(16)

In the research of Bartholdy, Olson, & Peare (2007), they state that the estimation period is around one year of trading before the event window, which is around 200 to 250 trading days. In my analysis, the number of trading in the estimation period is 272 trading days for the first event study and 269 trading days for the second event study.

Bartholdy, Olson, and Peare (2007) also mention in their paper that the most used even window is the event date along with one day before and one day after. Moreover, the HOSE deputy general director shares about the increase in concerns of the investors toward the VNSI establishment day (VOV, 2017). As I expect the effect might happen before the event date because of getting a lot of public attention. I set the event window of three days in total with one day before the event date, the event date, and one day after the event date. I present the summary of both event estimation and event window in the table below:

Table 3: Event-window description

Estimation window Event window 1st_{event date: 24/07/2017} _{From 01/06/2016}

to 30/06/2017 From 23/07 To 25/07/2017 2nd_{event date: 23/07/2018} ( VNSI re-estimation) From 01/06/2017 to 30/06/2018 From 22/07 To 24/07/2018

Finally, I will run one more event study with different event window to check whether if the presence of abnormal return change due to the change in the event window length or not. Because there is still no certainty about how many days should be set as the event window. For example, when conducting the event study to examine the abnormal returns on the Asia-Pacific markets, Thang Long Pham et al. (2007) use the event window of prior 20 days and post 20 days from the event date as the event window. Distinctly, the event window of ±10 days around the event day is the chosen window for the event study of Consolandi et al. (2009). Meanwhile, the length of the event window might affect the result of the event study. I will present the result of this additional study in the robustness check part.

(17)

4. Result

I present the result separately to answer for two hypotheses. First, I present the result of the time-series analysis to examine the performance of the sustainable portfolio versus the regular portfolio. The second part is the result of the event-study to verify the impact of the VNSI index establishment on the sustainable stock prices. As the Vietnam stock committee will re-calculate the list of VN30 twice a year, the result of my analysis will be presented in three different periods 4.1 Hypothesis 1 – VNSI and VN30 performances comparison

Before running time-series regression for three separate periods, I first run the Dickey-Fuller test for the stationarity problem of my dataset. The result indicates that the dataset of both VNSI and VN30 daily return are qualified to be run with all the Z-value of the test statistic smaller than 5% critical value. The result is attached in the appendix.

4.1.1 The mean-‐comparison test

I first conduct the significant test for the average return of both the sustainable stocks and the regular stocks. The mean of VNSI is composed of the average return of 20 stocks included. The mean of VN30 is the combination of 14 stocks included. The null hypothesis for testing the difference between the average return of the sustainable stocks and the regular stocks is zero.

Table 4: Mean-comparison t-test result

Note: The sample period is from 24.07.2017 to 28.12.2018 and divided into three phases. The table displays the mean values of the VNSI and VN30 index and the p-values of the test

Hypothesis: mean(VNSI_r) - mean(VN30_r) = 0

Significance levels: *** indicates 1%, **indicates 5%, *indicates 10%

Period Two-‐sample t-‐test

VNSI VN30 T-‐Test 24.07.17 -‐ 19.01.18 0.001317 0.0021046 0.2969 20.01.18 -‐ 20.07.18 -‐0.0018425 -‐0.0025166 0.4154 23.07.18 -‐ 28.12.18 -‐0.0004459 -‐0.0008149 0.5005

(18)

All the t-test values turn out to be insignificant in all three periods. So, I cannot reject the hypothesis that the difference between the mean of VNSI and VN30 equal to zero. This means that in term of the average return, the sustainable stocks perform similarly to the regular stocks.

4.1.2 Variance – comparison test

For the second test, I examine the difference in the risk between sustainable stocks and regular stocks. The t-test is computed to test for the null hypothesis, in which, the fraction of the standard deviation of the VNSI and the VN30 will equal to one.

Table 5: the two-sample variance-comparison test result

Note: The sample period is from 24.07.2017 to 28.12.2018 and divided into three phases. The table displays the standard deviation values of the VNSI and VN30 index and the p-values of

the test.

Hypothesis: ratio = sd(VNSI_r) / sd(VN30_r) = 1

Period Variance ratio test

VNSI VN30 P-‐value 24.07.17 -‐ 19.01.18 0.0022499 0.0020402 0.7311 20.01.18 -‐ 20.07.18 0.0021982 0.0025193 0.5705 23.07.18 -‐ 28.12.18 0.0017062 0.001488 0.6092

Similar to the test on the return, the results of the test on the standard deviation are also insignificant in all three levels 1%, 5%, and 10%. Since I cannot reject the null hypothesis, I conclude that on average, both the sustainable and the regular stocks have similar standard deviation values.

(19)

4.1.3 Sharpe ratio comparison

Besides the statistical test on the return and risk difference, I also compute the average value of the Sharpe ratio of companies in the VNSI and VN30 index. The Sharpe ratio provides me another instrument to compare the performance. I first calculate the Sharpe ratio for 20 stocks contained in the VNSI index and 15 stocks of the VN30 index. Next, I compute the average Sharpe value of each index. This process is replicated for three periods.

Table 6: Sharpe ratio

Note: The sample period is from 24.07.2017 to 28.12.2018 and divided into three phases. The table displays the average Sharpe ratio of two portfolios: VNSI and VN30

Period

Sharpe Ratio

VNSI VN30 24.07.17 -‐ 19.01.18 0.078 0.105

20.01.18 -‐ 20.07.18 -‐0.068 -‐0.083 23.07.18 -‐ 28.12.18 -‐0.022 -‐0.039

The result shows that even both type of stocks having bad Sharpe ratio, the average Sharpe ratio of the sustainable stocks is still better than the average Sharpe ratio of the regular stocks in the last two periods. Since I observe the difference in the Sharpe ratio quite negligible, my findings go against the paper of Zulkafli and Ahmad (2017), where they find out that there is a slight underperformance of the sustainable index against the benchmark index.

4.1.4 OLS regression result

Along with comparing the risk and return, I long for looking deeper into the performance of the sustainable and the regular portfolios. So, I use the value-weighted technique to obtain the returns of two portfolios, namely the VNSI portfolio and the VN30 portfolio. The VNSI portfolio is the combination of 20 sustainable stocks listed in the VNSI index. The VN30 portfolio is the

(20)

Table 7: OLS regression result

Note: The sample period is from 24.07.2017 to 28.12.2018 and divided into three phases. The table displays the constant values of the regression with the dependent variables are

VNSI and VN30 respectively.

Constant

Period VNSI VN30 24.07.17 -‐ 19.01.18 0.000255 0.00031 20.01.18 -‐ 20.07.18 0.00132* -‐0.000452 23.07.18 -‐ 28.12.18 -‐0.0000993 0.0000814

In the first period, the OLS regression output shows Jensen's alpha of the VNSI index is 0.000255. Meanwhile, this value is 0.0031 in the case of the VN30. However, they all turn out to be insignificant. The last period, I examine an opposite sign of VNSI and VN30 Jensen’s alpha. In which, The VNSI index has a negative Jensen's alpha of 0.0000993. The VN30 index has Jensen's alpha of 0.0000814. However, Jensen's alpha of both indices in these two periods all turn out to be insignificant.

For the second period after the re-estimation of the VN30 index, the result has been more noticeable. In spite of the insignificant constant value of the VN30, the VNSI shows its superior performance comparing with the market benchmark. The constant value of the VNSI of 0.00132 turns out to be significant at the 90% levels.

It is still not appropriate for me to reject my first hypothesis that sustainable stocks perform as equivalently as regular stocks. Since there is only one period, which the sustainable stock’s performance is significantly superior to the market benchmark. In order to consider this hypothesis more seriously, I run a regression with dummy variables. The VN30 portfolio is set as the reference to examine the difference in the performance of the VNSI. The result of the regression is present below:

(21)

Table 8: The OLS regression with dummy variables

Note: The sample period is from 24.07.2017 to 28.12.2018 and divided into three phases. The table displays the coefficient of the dummy variable with

binary coding VNSI = 1 and VN30 = 0

The P-value is tested in three significance levels: *** indicates 1%, **indicates 5%, *indicates 10% Period Dummy Coefficient P-‐value 24.07.17 -‐ 19.01.18 -‐0.0000227 0.972 20.01.18 -‐ 20.07.18 0.0019375 0.024** 23.07.18 -‐ 28.12.18 -‐0.0001824 0.757

Similar to the result of the primary OLS regression, the regression result with the presence of the dummy variable shows a significant output in the second period at the 95% level. This indicates that in the second period, the VNSI portfolio not only outperforms comparing with the market benchmark but also being superior to the VN30 portfolio.

To conclude, for the first and the third period, my analysis results show the support for the Consolandi et al. (2009), Cortez, Siva and Areal (2009) when finding the performances of the VNSI and VN30 portfolios are narrowly different from each other. Noticeably, the VNSI portfolio tends to perform better than both the benchmark market and the regular portfolio in the second period. The P-value of the coefficient is significant at the 90% level in the first regression and be significant at the 95% level in the second regression. Since the VNSI portfolio outperforms the market in one out of three periods, I only can partly accept my hypothesis that the performance of the sustainable stocks and the regular stocks is similar.

(22)

4.2 Hypothesis 2 -‐ There is a positive effect on the stock prices of being included in the VNSI

4.2.1 Event Study 1 – 24/07/2017

My second hypothesis is testing the market reaction of the establishment of the sustainable index – VNSI. In detail, I examine the change in the stock price toward the stocks named in the VNSI index. Under the setting at 95% percent significant level, the result shows that there is only one over 20 variables having significant t-value of -5.93. This means that the establishment of the sustainable index only has an impact on one stock over the entire sample. However, this impact is negative.

Table 9: Event study for an event date - 24/07/17 using CAPM Note: The sample period is from 01.06.2016 to 25.07.2017

The table describes the CAR values and the t-values of 20 stocks of VNSI index _1.96

STOCK_ID STOCK_NAME CAR TEST SIGNIFICANT_4TEST

1 BMP 70.0322658 70.836973 NO 2 DCM 0.0076812 0.4114106 NO 3 DHG 70.0341384 71.092327 NO 4 DPM 0.0056489 0.412932 NO 5 FPT 70.0019823 70.1447286 NO 6 HSG 70.0107706 71.353714 NO 7 MBB 0.0579891 0.8216155 NO 8 NT2 0.0130656 0.413556 NO 9 PNJ 70.0001769 70.0146929 NO 10 PVD 70.0164411 75.933086 YES 11 SBT 0.0452886 1.390386 NO 12 VIC 70.0063261 70.2848393 NO 13 VNM 70.0011487 70.0938699 NO 14 REE 0.0248927 0.6873363 NO 15 CTD 70.0322658 70.836973 NO 16 BVH 0.0058061 0.2458148 NO 17 CNG 70.0039859 70.1424242 NO 18 VCB 0.0010272 0.0421817 NO 19 NSC 0.0019014 0.1289112 NO 20 BID 0.0487454 1.124927 NO

(23)

The result remains the same when I use the Fama three-factors model instead. Nonetheless, instead of having a negative impact, the FF model shows that the sustainable index influences positively on the return of the PNJ stock.

Table 10: Event study for an event date - 24/07/2017 using Fama and French model Note: The sample period is from 01.06.2016 to 25.07.2017

The table describes the CAR values and the t-values of 20 stocks of VNSI index _Fama$

STOCK_ID STOCK_NAME CAR TEST SIGNIFICANT_$TEST

1 BMP 0.0227186 0.4621016 NO 2 DCM =0.0056775 =0.2546312 NO 3 DHG =0.0012532 =0.0310412 NO 4 DPM 0.0069439 0.355017 NO 5 FPT 0.0022717 0.1935193 NO 6 HSG 0.0032453 0.4431704 NO 7 MBB 0.0374259 0.650195 NO 8 NT2 0.0145391 0.3938991 NO 9 PNJ 0.0139941 3.471144 YES 10 PVD =0.0235118 =1.327996 NO 11 SBT 0.0429563 1.406967 NO 12 VIC 0.0010417 0.057969 NO 13 VNM 0.0043585 0.6362281 NO 14 REE 0.0194509 0.7250695 NO 15 CTD 0.0227186 0.4621016 NO 16 BVH 0.0042581 0.1776156 NO 17 CNG 0.0044624 0.1975329 NO 18 VCB =0.0106824 =0.3479824 NO 19 NSC 0.0104279 0.6453475 NO 20 BID 0.0183682 0.7937033 NO

(24)

4.2.2 Event study 2 – 23/07/2018

For the second event study, the event date is set at the date that the HOSE published a new list of stocks included in VNSI index – 23/07/2018.

Table 11: OLS regression result for an event date - 23/07/2018 using CAPM Note: The sample period is from 01.06.2017 to 24.07.2018

The table describes the CAR values and the t-values of 20 stocks of VNSI index

1 BMP 0.0284455 1.193381 NO 2 CTD <0.0312587 <1.107515 NO 3 DCM <0.006866 <0.4018768 NO 4 DHG <0.0248155 <2.70009 YES 5 DPM 0.0236047 1.080362 NO 6 FPT 0.0043858 0.2480824 NO 7 GAS 0.023176 1.542248 NO 8 HCM <0.0214809 <0.3326635 NO 9 HSG <0.009532 <0.1176726 NO 10 MBB 0.0148318 0.2315772 NO 11 NSC 0.0140381 1.370309 NO 12 NT2 <0.0097184 <0.1708945 NO 13 NVL 0.0265976 0.6453971 NO 14 PNJ 0.0001255 0.0028119 NO 15 PVD 0.0398449 1.697926 NO 16 REE 0.0170535 0.3688409 NO 17 SBT 0.062254 1.125634 NO 18 VCB <0.0345991 <0.5857759 NO 19 VIC <0.0059944 <0.6344519 NO 20 VNM 0.0035066 7.191736 YES

In the second event, one more variable shows the presence of the abnormal return when having significant t-value. In which, the event has a positive influence on the return of VNM and negative influence on the return of the DHG. Comparing with the first event, there is an increase in recognition of the investors toward the formation of the sustainable index. Since the first event, there is only one stock influenced significantly and negatively by the creation of the VNSI. After

(25)

one year of the implementation, the market experiences the abnormal return on two sustainable stocks.

Table 12: Event study for 23/07/2018 using Fama and French model Note: The sample period is from 01.06.2017 to 24.07.2018

The table describes the CAR values and the t-values of 20 stocks of VNSI index

1 BMP 0.0263086 1.057793 NO 2 CTD =0.0326241 =1.111295 NO 3 DCM =0.0194291 =0.913111 NO 4 DHG =0.0198989 =1.646454 NO 5 DPM 0.0083793 0.3913502 NO 6 FPT 0.0014901 0.1056449 NO 7 GAS 0.0305449 1.820349 NO 8 HCM =0.0265048 =0.3936876 NO 9 HSG =0.0248061 =0.2949166 NO 10 MBB 0.0067877 0.10005 NO 11 NSC 0.0114955 1.540165 NO 12 NT2 =0.0111494 =0.210494 NO 13 NVL 0.0287765 0.648772 NO 14 PNJ 0.011528 0.2956767 NO 15 PVD 0.0227188 1.36464 NO 16 REE 0.0133742 0.2764543 NO 17 SBT 0.0397618 0.9383677 NO 18 VCB =0.0419673 =0.6754497 NO 19 VIC =0.0037905 =0.2196645 NO 20 VNM 0.0083506 4.373059 YES

Showing the same result in the first event study, the result using the FF model also indicates the increase in the identification of the sustainable index toward the investors. Since the t-value of the VNM stock is positive and significant, being named in the VNSI sustainable index helps to increase the market capitalization of the VNM stock.

(26)

This means that there is a positive influence of the formation of the sustainable index on the stock market, especially the investors caring about sustainability.

The results obtained from two event studies are quite heterogeneous as the formation of the VNSI index has a positive impact on some stocks, but, negative impact on some others. In the case of DHG and PVD stocks, they experience a decline in stock prices after joining the list of sustainable stock.

It is, therefore, difficult for me to accept my second hypothesis, in which being listed in the VNSI index will affect positively on the stock price. My result is against the findings of the Consolandi et al. (2009) when they find out the positive CAR around the date of the announcement. Still, I examine the upward trending of the presence of abnormal return, which represents the growth in the responsiveness of the investors toward sustainable stocks.

5. Discussion and Conclusion 5.1 Robustness test

In order to test the consistency of the result, I run the event study one more time with a different time range. I set the event window of 15 days and examine the appearance of abnormal return. For the first event, there are two stocks – HSG, and BID having the significant CAR values. For the second event, the number of stocks experiencing abnormal return is two stocks in both CAPM and FF model. The result of this robustness test confirms that even the result varies when changing the event window, the trend is still steady. The formation of the VNSI index still has a negative effect on some stocks and a positive effect on some others. This result is entirely in line with my previous event study result with three-days observation. The result of the robustness will be presented in the appendix.

5.2 Main findings and their implications

Generally, based on the result, I explore an increase in the impact of the VNSI index establishment on the stock price of sustainable stocks, whether it is a positive or negative effect. From which, I conclude since the time that the VNSI index was introduced, there is an effect on

(27)

the stock price of some sustainable stocks. However, this number is still limited, which is one over 20 stock in the first event and two over 20 stock in the second event when using CAPM. It is, therefore, hard for me to accept my hypothesis, to wit: there is a positive effect on the stock price of being included in the VNSI index.

By showing the presence of abnormal return, the research gives the enterprises one potential direction for future development. In which they can improve the market capitalization of their firm by performing more sustainably and taking more social responsibility. My result also helps in providing a glimpse of the investor’s viewpoint. The partly acceptance of the first hypothesis shows that sustainable stocks perform comparably with regular stocks in some cases. Whereby, the investor, when deciding to invest in sustainable stocks, they are not only about the profit but also social responsibility. Meanwhile, one out of three cases, the sustainable stocks show its superior performance to the regular one. This result can encourage the investors to invest more in sustainable stocks to obtain both higher return and higher social responsibility.

5.3 Limitation and Future Research

As the sustainable index of Vietnam stock market is entirely new with the historical data of nearly two years. The findings of this study are not as significant as the study with more extended historical data. Collecting data to research this index, specifically in Vietnam stock market, is another difficulty of this research. Nonetheless, sustainability in the financial market or the stock market is explicitly still becoming important. The further direction for this paper is to keep going with a broader dataset of VNSI. Moreover, when examining the trade-off between risk and return in Vietnam stock market, Fang & Nguyen (2017) insist that the Carhart four-factor model help explaining this trade-off relationship better. And, Carhart (1997) proves the potential influence of the momentum factor in the portfolio returns. By using the Carhart model, in future, I can examine whether there is a momentum effect on the sustainable stocks in the Vietnam stock market. Lastly, Vietnam plans to introduce its first green bond soon. The second direction for this paper is to apply an asset pricing model to evaluate the performance of the green bond versus the government bond. 5.4 Conclusions

(28)

My paper concentrate on the sustainable index of Vietnam stock market and how the firms and investors perceive it. In my research, I use two different techniques to approach this concern. The time-series analysis helps to explain how the investors perceive sustainable stocks. Whereas, the event-study explains the attractiveness of the sustainable index through the presence of the abnormal return. The result proves that there is an effect of the VNSI index formation on the change in stock prices of sustainable stocks. My research contributes to the general literature of sustainability in the financial market, especially Vietnam. Moreover, my paper also shed the light on future research on how to estimate the influence of the Vietnam green bond establishment.

6. Bibliography

Bartholdy, J., Olson, D., & Peare, P. (2007). Conducting event studies on a small stock exchange. The European Journal of Finance, 13(3), 227-252.

Bello, Z. Y. (2008). A statistical comparison of the CAPM to the Fama-French Three Factor Model and the Carhart's Model. Global Journal of Finance and Banking Issues, 2(2).

Carhart, M. M. (1997). On persistence in mutual fund performance. The Journal of finance, 52(1), 57-82.

Consolandi, C., Jaiswal-Dale, A., Poggiani, E., & Vercelli, A. (2009). Global standards and ethical stock indexes: The case of the Dow Jones Sustainability Stoxx Index. Journal of Business Ethics, 87(1), 185-197.

Cortez, M. C., Silva, F., & Areal, N. (2009). The performance of European socially responsible funds. Journal of business ethics, 87(4), 573-588.

Cuthbertson, K., & Nitzsche, D. (2005). Quantitative financial economics: stocks, bonds and foreign exchange. John Wiley & Sons.

Dao, B. (2013). VN30 Index: An Overview and Default Probability Analysis. Available at

SSRN 2543084.

Fama, E. F., & French, K. R. (1993). Common risk factors in the returns on stocks and bonds. Journal of financial economics, 33(1), 3-56.

Fama, E. F., & French, K. R. (1995). Size and book-‐to-‐market factors in earnings and returns. The journal of finance, 50(1), 131-155.

(29)

Fang, K., Wu, J., & Nguyen, C. (2017). The risk-return trade-off in a liberalized emerging stock market: evidence from Vietnam. Emerging Markets Finance and Trade, 53(4), 746-763.

Jones, S., Van der Laan, S., Frost, G., & Loftus, J. (2008). The investment performance of socially responsible investment funds in Australia. Journal of Business Ethics, 80(2), 181-203.

Galema, R., Plantinga, A., & Scholtens, B. (2008). The stocks at stake: Return and risk in socially responsible investment. Journal of Banking & Finance, 32(12), 2646-2654.

Hochiminh stock exchange (2018). There will be the update of the sustainable index in 2018. Retrieved from https://tinnhanhchungkhoan.vn/chung-khoan/nam-2018-se-nang-cap-chi-so-phat-trien-ben-vung-218371.html

Le, C. (2015). The Fama-French Factors: Evidence from the Vietnamese Stock Market. Available at SSRN 2610588.

Lockwood, L., & Prombutr, W. (2010). Sustainable growth and stock returns. Journal of Financial Research, 33(4), 519-538.

MacKinlay, A. C. (1997). Event studies in economics and finance. Journal of economic literature, 35(1), 13-39.

Munoz-Torres, M., Fernandez-Izquierdo, M., & Balaguer-Franch, M. (2004). The social responsibility performance of ethical and solidarity funds: An approach to the case of spain. Business Ethics -Oxford-, 13(2-3), 200-218.

Phong, N. A., & Hoang, T. V. (2012). Applying Fama and French Three Factors Model and Capital Asset Pricing Model in the Stock Exchange of Vietnam. International Research Journal of Finance and Economics, 95.

Puopolo, G. W., Teti, E., & Milani, V. (2015). Does the market reward for going green?. Journal of Management Development, 34(6), 729-742.

Sustainable stock exchanges initiative (2019). Hochiminh stock exchange. Retrieved from http://www.sseinitiative.org/fact-sheet/hochiminh-stock-exchange/

Sparkes, R. (2001). Ethical investment: Whose ethics, which investment? Business Ethics: A European Review, 10(3), 194-205. doi:10.1111/1467-8608.00233

Thang Long Pham, V., Quoc Tho Nguyen, D., & Tô, T. D. (2007). Chapter 10 Abnormal returns after large stock price changes: Evidence from Asia-Pacific markets. In Asia-Pacific financial

(30)

markets: integration, innovation and challenges (pp. 205-227). Emerald Group Publishing

Limited.

Vietnam Prime Minister – Nguyen Tan Dung (2012). Approval of the national green growth strategy. Retrieved from https://www.giz.de/de/downloads/VietNam-GreenGrowth-Strategy.pdf

VOV. (2017). Vietnam sustainability index to be launched in July. Retrieved from https://m-english.vov.vn/economy/vietnam-sustainability-index-to-be-launched-in-july-346216.vov

Williams, G. (2010). Socially responsible investment in Asia. Social Space, 20-27.

Wooldridge, J. M. (2012). Introductory econometrics: A modern approach: Cengage Learning. A Figures, 18.

Zulkafli, A. H., & Ahmad, Z. (2017). The Performance of socially responsible investments in Indonesia: a study of the Sri Kehati Index (SKI). Gadjah Mada International Journal of

(31)

7. Appendix

• Test for stationarity

Dickey fuller test period 1

Z(t) -10.513 -2.596 -1.950 -1.612 Statistic Value Value Value Test 1% Critical 5% Critical 10% Critical Interpolated Dickey-Fuller Dickey-Fuller test for unit root Number of obs = 127

. dfuller VNSI_r, reg noconstant

MacKinnon approximate p-value for Z(t) = 0.0000

. dfuller VNSI_r, reg constant

. dfuller VNSI_r, reg constant trend

. dfuller VN30_r, reg noconstant

Statistic Value Value Value Test 1% Critical 5% Critical 10% Critical Interpolated Dickey-Fuller Dickey-Fuller test for unit root Number of obs = 127

(32)

Dickey fuller test period 2

. dfuller VN30_r, reg constant trend

. dfuller VNSI_r, reg noconstant

. dfuller VNSI_r, reg constant trend

(33)

Dickey fuller period 3

. dfuller VN30_r, reg constant

. dfuller VN30_r, reg constant trend

> dfuller VNSI_r, reg noconstant

(34)

. dfuller VN30_r, reg noconstant

. dfuller VN30_r, reg constant

(35)

• Robustness Test – Event study with 15 days window

Event study using CAPM – 1st period

1 BMP 40.0447272 40.7724447 NO 2 DCM 40.0254327 40.4347363 NO 3 DHG 40.0429645 40.608002 NO 4 DPM 40.0325938 41.082756 NO 5 FPT 0.0122057 0.2983407 NO 6 HSG 40.1958719 42.046266 YES 7 MBB 0.0748264 0.7952546 NO 8 NT2 40.0124108 40.2254752 NO 9 PNJ 0.032035 0.7640685 NO 10 PVD 0.0961775 1.260689 NO 11 SBT 0.1233253 1.647907 NO 12 VIC 0.0326699 0.7823561 NO 13 VNM 40.0376356 41.459763 NO 14 REE 40.0134789 40.179584 NO 15 CTD 40.0447272 40.7724447 NO 16 BVH 40.0204687 40.5253636 NO 17 CNG 40.0047294 40.0983052 NO 18 VCB 40.0080677 40.2555631 NO 19 NSC 0.1013286 1.298764 NO 20 BID 0.1410278 2.026836 YES

(36)

Event study using FF model – 1st_period

SIGNIFICANT_*TEST STOCK_ID STOCK_NAME CAR TEST SIGNIFICANT_*TEST

NO 1 BMP 0.0554652 0.9480364 NO NO 2 DCM =0.0350541 =0.7485369 NO NO 3 DHG 0.0069495 0.0950143 NO NO 4 DPM =0.0223854 =0.8358319 NO NO 5 FPT 0.0186012 0.4541335 NO YES 6 HSG =0.1622301 =1.723236 NO NO 7 MBB 0.0358056 0.4477137 NO NO 8 NT2 =0.0024563 =0.0440783 NO NO 9 PNJ 0.0513163 1.102808 NO NO 10 PVD 0.0968593 1.625272 NO NO 11 SBT 0.1251311 1.788441 NO NO 12 VIC 0.0440675 1.094009 NO NO 13 VNM =0.0302865 =1.186101 NO NO 14 REE =0.0154472 =0.2194276 NO NO 15 CTD 0.0554652 0.9480364 NO NO 16 BVH =0.0255609 =0.6723239 NO NO 17 CNG 0.0167543 0.3486893 NO NO 18 VCB =0.0332733 =0.8695102 NO NO 19 NSC 0.1238298 1.471999 NO YES 20 BID 0.0817329 1.797134 NO

(37)

Event study using CAPM – 2nd period

1 BMP 0.0734731 1.518122 NO 2 CTD 0.1012615 1.715831 NO 3 DCM 0.0023657 0.0816548 NO 4 DHG >0.0271966 >0.7155549 NO 5 DPM 0.0715542 1.668147 NO 6 FPT 0.016382 0.2897792 NO 7 GAS 0.0264235 0.4305064 NO 8 HCM 0.0725331 0.682113 NO 9 HSG 0.1349796 1.03223 NO 10 MBB 0.0829027 0.9844136 NO 11 NSC 0.0575739 1.017505 NO 12 NT2 >0.1368935 >0.9830565 NO 13 NVL 0.1838372 2.844796 YES 14 PNJ 0.063941 0.6468472 NO 15 PVD 0.0535023 0.4563394 NO 16 REE 0.0957258 1.404462 NO 17 SBT 0.0966031 1.342145 NO 18 VCB 0.0263066 0.3152836 NO 19 VIC >0.0899912 >2.658454 YES 20 VNM >0.0389134 >1.082159 NO

(38)

Event study using FF model – 2nd_period

1 BMP 0.1080007 2.140358 YES 2 CTD 0.0621086 1.043778 NO 3 DCM >0.0368465 >1.001059 NO 4 DHG >0.0161514 >0.4577638 NO 5 DPM 0.0042496 0.093968 NO 6 FPT >0.0155811 >0.3174978 NO 7 GAS 0.0416596 0.6182396 NO 8 HCM 0.0930452 0.8427307 NO 9 HSG 0.0633975 0.5264099 NO 10 MBB 0.0333922 0.3639278 NO 11 NSC 0.0703497 1.109304 NO 12 NT2 >0.0871636 >0.7450936 NO 13 NVL 0.1482337 2.06252 YES 14 PNJ 0.0971393 1.065444 NO 15 PVD >0.0211178 >0.2300839 NO 16 REE 0.071621 1.091058 NO 17 SBT 0.0028741 0.0396523 NO 18 VCB 0.0065105 0.0720997 NO 19 VIC >0.0402281 >0.950098 NO 20 VNM >0.0009774 >0.0288705 NO

(39)