The Relationship between Social Responsibility and Majority Shareholders’ Stock Selling: Considering the Mediation of Investors Tendency

Based on the research object of majority shareholders’ stock selling at the post period of equity division reform, this paper takes 8019 A-share listed companies in Shanghai and Shenzhen stock market from 2007-2013 as sample, empirically analysis the relationship between social responsibility and majority shareholders’ stock selling, and consider mediation effect of the investor tendency. The study found that from the point of internal governance effect, the better of social responsibility, the lower possibility of majority shareholders’ stock selling. At the same time, the lower of market reduction premium, the lower possibility of majority shareholders’ stock selling, after considering the influence of investors’ tendency, the inhibition effect of social responsibility will be magnified, which means securities market identity with the inhibition effect that comes from the social responsibility caused by listed companies’ internal. Further study found that majority shareholders will avoid stock selling during the period of the social responsibility reporting, and market investors will hold different attitudes toward majority shareholders for their first and continuous stock selling, and on their view, continuous stock selling is a “bad” event.


Social Responsibility and Continuous Reduction of Majority Shareholders
The behavior of fulfilling the social responsibility by the enterprises is a kind of self-conscious action which is their own request derived from the internal development, that is, the self-conscious act of the enterprise, which is the actual contribution to the promotion of social welfare through the sublimation of self-demand. Li and Xiao (2011) argued that the logical starting point of behavior of fulfilling the social responsibility by the enterprises is derived from voluntary charitable behavior, the social expectation of enterprises' behavior, the social impact of enterprises, the compliance with the spirit of the contract, the match of social rights and obligations, the enterprises' response to social pressure, the enterprise management of social risks and the balance of integrated business goals, then ultimately derived from the contribution to the maximization of social welfare from enterprises. In other words, from the outside, the implementation of corporate social responsibility is to meet the requirements of society, then achieve social expectations for the enterprise. But from the inside, it is derived from perfection and self-regulation of the internal corporate governance system. Enterprises integrate the concept of social responsibility into development and bear the responsibility of the enterprise that concerned about internal and external stakeholders. It will make a self-regulatory effect in the enterprises. Under the regulatory pressures, companies can naturally protect themselves from curbing the behavior that are detrimental to their own development, the interests of stakeholders, social development or the interests of the securities market and its investors. For example, those that perform better social responsibility will not easily hide the related information when they take the interests of stakeholders into account, which will increase the transparency of information (Chih et al., 2008). It can be seen that as an internal governance system, social responsibility will play an important role in natural regulatory capacity, which can control and standardize the relevant behavior of the enterprise management and shareholders.
First of all, from the moral point of view, the implementation of social responsibility will enable listed companies to protect the legitimate interests of the securities market and market investors under the guidance of moral conscience. As social responsibility is more manifested in the moral dimension, so companies will divide a reduction of the shares into several parts, which increases the volatility of the company's stock and the companies will acquire more additional benefits (Cai & Wei, 2011), it also reduces the degree of market regulation and investors' attention. Obviously, this is not a "responsible" behavior. It can be seen that under the moral constraints, the companies with implementation of good social responsibility will serve maintaining market stability and the interests of ordinary investors as their own obligations and they will not maliciously reducing shares continuously. But the companies with implementation of poor social responsibility will select a number of special ways to obtain more self-interest through earnings management or selection of holding time and other acts in the process of continuous reduction (Lin & Cao, 2012;Cao & Lin, 2014).
Second, from the perspective of shareholders, their equity itself is the interest that needs to be maintained in the process of social responsibility. On the one hand, listed companies need to safeguard the interests of shareholders in the process of fulfilling social responsibility, because the fundamental goal of the development of the company must be profitable, and direct beneficiaries is the majority shareholder, so the large beneficiaries of which to fulfill social responsibility is the majority shareholder, majority shareholders do not need to obtain short-term gains from the secondary market through continuous reductions once they can get a steady return. Kong et al. (2013) found that listed companies will fulfill the social responsibility better under the supervision of the media, which makes the earnings management of the majority shareholders, associated transactions which were hollowed out and irregularities were significantly reduced. It shows that social responsibility will bring more benefits for the majority shareholders, this kind of income may be actual money or invisible reputation, so that majority shareholders will balance between long and short term interests, even if they need a short term reduction. It will not easily cause unnecessary attention or destroy the established good image because of continuous reduction. On the other hand, the interests of shareholders not only include the interests of majority shareholders, but also the interests of ordinary market investors. In a company with good social responsibility, it is not easy to allow the occurrence of damage to the interests of small and medium investors. So that the behaviors of continuous reduction that will cause the stock price a greater volatility and damage the interests of ordinary investors will be reduced.
Especially in the period of post-share split reform, because the reduction of majority shareholders is the core factor of affecting the stable development of capital market, making the reduction of majority shareholder and stable development of capital market bounded together (Jia et al., 2009). Especially this process is accompanied by a strong behavior that majority shareholders will encroach on the interests of small and medium shareholders by taking advantage of the reduction (Liu & Li, 2010). So, that the implementation of social responsibility behavior may become the cost of the enterprise (Harrison & Freeman, 1999). But after the transition from "cost" to "opportunity", social responsibility will give enterprises more opportunities in the long run and bring better financial performance (Ruf et al., 2001). So in this process, the majority shareholders will take their long-term interests into account, and will not take the initiative to undermine the company's image or undermine the possibility of future greater gains, so there is no need and no reason for continuous reduction. At the same time, the continuous reduction of majority shareholders is contrary to the social responsibility thought advocated by the listed companies. Under the governance of the social responsibility supervision, the cost of the majority shareholders to reduce the holdings may be far greater than the reduction of earnings, and thus they will not easily reduce the share holdings continuously in the short term.
In addition, from a social point of view, the community will be more concerned about the company with the implementation of better social responsibility, which is one of the reasons for actively fulfills social responsibility. Xu et al. (2011) found that media attention and public opinion play an active role in the fulfillment of social responsibility of listed companies, and the media are more interested in the companies with better social responsibility, and this concern, in turn, enable the company to fulfill its social responsibilities better. At the same time, the securities market will also pay enough attention and support to companies that fulfill their social responsibilities (Godfrey, 2005). So in the case with many concerns, the invisible external regulatory power is formed. Although this regulation does not have a coercive effect, it can give a certain constraint on the listed company, because the listed companies have formed a good impression on the community and market investors, once the companies continuously make malicious reduction, they will cause even more impact on the heart of social and market investors, and the adverse effects will even be greater than the general behavior of listed companies which behave the same. In the poorly performing companies, because of the lack of social concern, the majority shareholders have more opportunities to achieve continuous reductions through various channels, and because of the lack of sufficient attention, the majority shareholders of these companies can try and make profit many times. Therefore, these arguments lead to our first hypothesis.

H1.
There is a significant negative correlation between social responsibility and continuous reduction of majority shareholders, that is, the lower the probability of continuous reduction of its majority shareholders in the listed companies with better social responsibility, and if the listed companies with worse behavior of fulfill the social responsibility, the probability of its majority shareholder holdings will be greater.

Social Responsibility, Tendency of Investors and Continuous Reduction of Majority Shareholders
The market value of listed companies will be affected by many factors, but the market investor's tendency is an important factor. The market is largely determined by the investors, when investors believe that a listed company has good development potential and future prospects, they will be sought after the company's stock, which prompted the stock price rise, and even more than its actual book value. But if the investors are not optimistic about a listed company and they think the company does not have a good future development potential, or investors find that their own interests are occupied by majority shareholders or management, they will lose confidence in the company, and then take the "foot vote" approach to punish it (Shleifer & Vishny, 1986). In other words, the securities market will give listed companies the opportunity to accompany the punishment, but whether the opportunity or punishment depends entirely on the cooperation between the listed company's own performance and the tendency of the market investors, or depending on the presented state of listed company in accordance with the attitude of the market investors. For example, Baker and Wurgler (2004)  that rational corporate management would choose cash dividends based on whether there was a dividend premium or a discount in the market. Therefore, for majority shareholders with reduced tendencies, in order to obtain more secondary market returns, they need to find ways to deliver good information to the market, such as the most direct earnings management behavior (Cai, N. & Wei, M., 2011), thus gaining short-term high market value. It can be seen that market behavior or market conditions will affect the choice of majority shareholders' holdings.
First, from the perspective of the market itself, Zhang Tao (2008) found that after the split share structure reform, China's stock market has changed from inefficient market to weak efficient market, indicating that the stock market has been able to contain stock prices and some transaction information.
In other words, rational market investors and rational majority shareholders or management will game through the stock between the buy and sell. When the market investors give the company stock trust, the majority shareholders will be able to find this trust, and will not easily make damage to the company's image; and when the listed companies issued a benign business signal, the market investors will give the company enough support. Of course, this market investors and listed companies between the "good" relationship once broken, the market investors and the interests of listed companies will be lost.
Second, from the perspective of market investors, because it is difficult to determine the true intention and reasons for the reduction of majority shareholders, so more market investors will see the reduction as a "bad" behavior, that is, when the listed company had a reduction, the market investors will think that there are some problems within the company, thus they will treat the company's stock carefully.
Empirical evidence also shows that the reduction has had an impact on the securities market, there will be a long time in the reduce effect, bringing the decline of market price , and the reduction will also bring investors the forecast that the company's profitability about pessimistic future, then accelerated the stock price fell (Fan, J., 2008). At the same time, if the majority shareholder reduce the share holdings after a period of time, and not reduced once again, the market will gradually ease the impact of the previous reduction. However, if the reduction event occurs again after the first reduction, especially in a relatively short period of time occurs the reduction event again by the same majority shareholder, it will accelerate the volatility of stock prices and result in a more significant reduction effect (Johnston et al., 2005). In addition, such as external audits and other regulatory actions will be through their own behavior (such as different audit opinion), give a corresponding signal to the company financial status represented by the majority shareholders' continuing reduction and so on (Chen et al., 2016). On the one hand, the market investors will be more cautious on the continuous reduction of the event, because the market investors think that the majority shareholders grasp more that is, the market investor does not have a special preference for the company that has a reduction, and the market value of the holding company is not significantly lower (or even higher, because there will be speculation in the market or in the short-term the stock price will be untrue because of the asymmetry of information) than the company that did not occur in the reduction, the majority shareholders will use this opportunity to continuously reduce the shares in order to maximize the benefits. However, if there exist opposite situation in the market, that is, when the stock market price in the company with significant reduction is much lower than the stock price of the company without reduction, the majority shareholders in order to avoid greater losses of their own interests, or to protect the company's image and long-development, they will not blindly reduce the share holdings continuously. It can be seen that the rational majority shareholders will give the market investors a tendency to respond, because rational majority shareholders need to consider the relationship between short-term reductions and long-term earnings even if they choose reduction. Especially because of the existence of information asymmetry, the majority shareholders can better choose the reduction time, then the market will not let investors abandon, but obtain the results of maximizing the proceeds.
Hence, we posit the following hypothesis.
H2. There is a significant negative correlation between the investor's tendency and the continuous reduction of the majority shareholders, that is, the lower the probability that the majority shareholders will continue to reduce the holdings in the listed companies with more inclined investors. While in the listed companies with less market tendency, the probability of continuous reduction by the majority shareholders will be higher.
The behavior of listed companies to fulfill their social responsibilities will be found and given the corresponding response by the securities market. Empirical evidence suggests that social responsibility is directly sought by investors in the securities market, regardless of whether the behavior of social responsibility is short-term or long-term (Gardberg & Fombrun, 2006;Ye & Wang, 2013), which means that market investors will value the behavior of listed company to fulfill their social responsibility. They assume that the conduct of social responsibility will protect their legitimate rights and interests, thus showing supportive attitude. Similarly, the majority shareholders of listed companies will also feel the pursuit of abandonment and abandonment from market investors, they will find that the behavior of social responsibility can be recognized by market investors. In particular, this behavior will also generate value in the stock price and enhance the company's share price (Kim et al., 2014). As a result, under the strength of securities investors, the inhibiting effect of social responsibility on the main reduction of majority shareholders will be amplified invisibly. On the one hand, when the majority shareholders find the market investors sought after the company's stock because the company fulfilled the social responsibility, the majority shareholders need to cooperate with the company to support the company's development, especially the majority shareholders who do not reduce all the shares, they will not destroy the good image of themselves and the company in the eyes of investors through continuous reduction. On the other hand, when the market investors recognize that listed company has fulfilled their social responsibility, they will show the pursuit of the company's share price. But when the majority shareholders continue to reduce their holdings, the market investors will reduce the trust of the company, which will offset the positive impact of social responsibility, and in this case, the rational management will find the company's market value difference, then they can only continue to strengthen the implementation of fulfilling social responsibility to reduce the adverse effects of continuous reduction of majority shareholders. And rational majority shareholders will also find changes in market investors, in the process of fulfilling social responsibility they will not deliberately reduce holdings continuously. It can be seen that in the suppression of social responsibility, once the majority shareholders find the behavior that market investors sought after company's stock, then they will not deliberately reduce share holdings continuously, especially when they need not cash the current majority shareholders immediately. Because in the case that the market investors recognize and sought after the company's stock, the company's future earnings will show a better development trend, and the company's stock value and price in the next period of time may show a better situation, so that the majority shareholders will get more revenue, thus there is no need to break this trend through the behavior of continuous reduction, and if they do that, even the basic gains will not be available. Hence, we put forward the following hypotheses.

H3.
After considering the influence of investors' tendencies, the impact of social responsibility on continuously reducing shareholdings by majority shareholders will be significantly enlarged.

Specification of Variables
Explained variable: Continuous Reduction of majority shareholders (CR). Lin and Cao (2014a) designed the index of majority shareholders' continuous reduction based on whether the company has two or more shareholders who reduce shareholdings or whether appear two or more reduction events in the company, but this indicator can't measure the "malicious" mentality of shareholders in the process of continuous reduction, because the reduction of a shareholder may not be affected by other shareholders. Therefore, the design of the index in this article is based on whether the same shareholder has a continuous reduction in the same fiscal year, that is, if it occurred two or more reduction events by the same shareholder in the listed company, then CR=1. Otherwise, if it didn't occur reduction events by majority shareholders that year, or did not occur a number of reduction events by the same shareholders (even if there are multiple majority shareholders reduce holdings once a time), then CR=0.
Explanatory variable: Social Responsibility (CSR). The index of early measurement of social responsibility are mostly based on whether the listed company disclose the information in the form of publish a report on social responsibility. However, this form can't distinguish the differences in social responsibility among different companies. And according to Stakeholder Theory, the company's daily behavior itself contains the implementation of social responsibility, or social responsibility should be a subtle behavior. Shen (2005) provides an analytical approach that takes the interests of the stakeholders into account, which by using the balance sheet and the income statement for analysis. The paper argues that the balance sheet liabilities and the owner's equity report the source of the company's funds, which include the funds provided by the suppliers, employees and government, as well as the resulting benefits. Owners' equity is considered to represent the shareholder's claim to the company in the narrow sense, but it does not fully reflect the funds from the shareholders. So the company has three types of funds, namely, liabilities, shareholders' equity provided by the shareholders and the company's own funds. The last one is the "main interests". And the main interests are actually the result of a joint effort from the shareholders, creditors, customers, suppliers, employees, government and other stakeholders. Therefore, with reference to Shen (2005)  In calculating the weights above, apart from the identified variables, EQ refers to the equity, that is the sum of the total share capital of the company and the capital reserve; DT refers to the debt, that is the sum of short-term borrowings, non-current liabilities due within one year, long-term borrowings and bonds payable. Size refers to the company's total assets; MR refers to the main interests, that is, the rest of the company's assets after deducting equity and debt from the total assets of the company.

Explanatory variables: the tendency of investor (Market): investors tend to show the market for
investors to pursue or not pursue a class of listed companies, so different companies will be different from the market value. The change of the market value of the company may also be due to the profitability of the company. Therefore, this paper uses the market book value ratio to measure the difference of the company value while considering the change of the book value of the company. At the same time, taking the differences in market value between the companies with reduction of holdings and the companies without reduction of holdings in into account, this paper adding an adjusted factor about value difference in the companies with and without reduction of holdings, namely:

 
The larger the market value is, the higher the market investor's tendency will be; while the smaller the market value is, the lower the market investor's preference will be.

Control variables:
Reference to Helwege et al. (2007) and other studies, control variables are listed in Table 1. Return on assets (ROA) (Net profit/Average total assets)×100% Ownership concentration (H10) The sum of square of the proportion of the top 10 shareholders of the company Market process (Index) Regional marketization index provided by Fan Gang et al (2011) Whether it is the largest shareholder reduce holdings (WFBS) In the company where the reduction occurred, if the reduction of majority shareholders as the largest shareholder, then WFBS=1, otherwise WFBS=0; In companies that do not have a reduction, then WFBS=0

Empirical Model
According to the previous research hypothesis and the design of the research variables, this paper constructs the empirical model: Since the explanatory variable is a dummy variable, the empirical model is a binary selection logit model. The estimate of the model is completed by the coefficient maximum likelihood function, where  is the parameter to be estimated, C is a constant term and  is a random interference term.   Table 3 lists the descriptive statistical status of the majority shareholders' continuous reductions.

Descriptive Statistics
Overall, about 7.43% of the sample companies occurred that a majority shareholder reduced holdings twice or for more than two times, accounting for 54.98% of the number of companies with the reduction of holdings, indicating that more than half of the companies with reduction have continuous reduction, that is, among the companies with reduction, there will appear more of the emergence of multiple reduction events. From the annual situation, the proportion of reducing holdings by majority shareholders in 2007 is highest, there are nearly 70% of the companies whose majority shareholders reduce shareholdings have continuous reduction in the situation, which is related to the split share reform in 2007; But the proportion of companies with continuous reduction in 2009 is highest. In the remainder of the year, the proportion of the company's total holdings was below 10%, and the proportion after 2011 was even below 5%, and after 2010, the proportion of companies holding reductions was under 50%. This situation also showing a gradual downward trend, indicating that over  Table 4 lists the descriptive statistical results of the main variables. In addition to the explanatory variables, the mean of explanatory variables CSR is 0.29, but there were significant differences in the social responsibility status of the listed companies, and some corporate social responsibility values were even negative; the mean of market is -0.75, but the same market for different listed companies investors tend to a greater degree of difference. In the control variable, the mean of variable Debt is 0.52, indicating that the total liabilities of the sample company account for more than half of the total assets; the mean of ROA is 0.03, indicating that the net profit of the sample company accounts for about 3% of the total assets; the mean of H10 is 0.18, indicating that the sample of company's equity concentration is not high, but different companies have a large difference in the concentration of equity.

Multivariate Regression Test
***Indicate significant difference at the 1% level (two-sided). Table 7 lists some of the results of the multivariate regression test. Firstly, the overall test value of the regression results can be checked by the significance level of the conventional confidence level, indicating that there is no significant difference between the actual probability distribution of the explanatory variables and the predicted distribution fitted by the explanatory variables and the control variables, the result is credible.
Specific to the explanatory variables of the test results, (1) shows that the coefficient between the explanatory variable CSR and CR is significantly negative (significant level is 1%), which indicates that the fulfillment of the social responsibility of the listed company is the same as that of the majority shareholder. There is a significant negative correlation between the holdings, that is, the better the social responsibility of the listed companies, the corresponding occurrence of the majority shareholder of the probability of reduction is even lower, which for the previous study hypothesis 1 provides empirical evidence support; regression results (2) shows that the explanatory variable Market is significantly negative as the coefficient value between the explanatory variables CR (the level of significance is 1%), indicating that the listed companies with more inclined investors, the probability of continuous reduction by majority shareholders is even lower. Regression result (3) takes the results of the investor's reciprocity into account, the coefficient between the variable CSR and the explanatory variable CR is significantly negative (the level of significant level is 1%), and the coefficient between the variable Market and the explanatory variable CR is also significantly negative (significant level of 1%), the interaction term can be tested by the significance of the conventional confidence level, indicating that after considering the investor's influence, the behavior of listed companies to fulfill their social responsibility will restrict the continuous reduction of majority shareholders, and compare the regression results (3) with the regression results (1) in the variable CSR coefficient values found that the absolute value of coefficient in regression result (3) is significantly larger than the regression result (1), indicating that after considering the influencing factors of investors' tendencies, the effect of social responsibility on the continuous reduction of majority shareholders has been obviously strengthened.
This conclusion provides empirical evidence for the previous hypothesis 3.

Robustness Test
In order to prove the reliability of the research conclusion, this paper carries out the robustness test.
First, the regression tests were performed in the regression model to control annual factors and industry factors. Second, considering the possibility of multicollinearity caused by the inter-variable correlation in the regression model, the ridge regression test will be performed. Third, taking the possible endogeneity between variables into account, that is, the endogenous problem of social responsibility and investors' tendencies will make the sample self-selection bias, such as investors deliberately choose high social responsibility or low social responsibility to seek special interests. Therefore, this paper uses Heckman two-stage regression method to solve the endogenous problem of social responsibility and investors' tendencies. In the two-stage regression of the social responsibility variables, the first stage is the social responsibility estimation model, join the exogenous variables such www.scholink.org/ojs/index.php/jbtp Journal of Business Theory and Practice Vol. 6, No. 1, 2018 as the previous period of social responsibility, accounting firm audit and so on, and the resulting IMR is used as a control variable for the second stage regression. In the two-stage regression of the investor's tendency variable, the first stage adds the exogenous instrument variable, the accounting firm's audit and other exogenous instrument variables, and the resulting IMR is also used as the control variable for the second stage regression. Fourth, considering a variety of measurement methods of social responsibility, different literatures differ in the way they measure social responsibility. In addition to the method of calculating the degree of social responsibility of the listed company, this paper further uses the dummy variable of the listed company "whether to disclose the social responsibility report" instead of the original social responsibility variable, and then carries on the regression test. The results of the robustness test are basically consistent with the previous empirical results.

Reduction?
Empirical evidence shows that the majority shareholders will choose the time to reduce holdings (Lin & Cao, 2012;Cai, 2012), which shows that the majority shareholders will choose some time without attention to reduce shareholdings, so that they can avoid some extra time with concern. But the previous empirical evidence shows that social responsibility will inhibit the continuous reduction. Does this mean that majority shareholders will avoid a continuous reduction in the disclosure of social responsibility report period? This article also found that the events of reduction in the second half of the year are far more than the first half of the year. There for, in the point of SCR, the majority shareholders will avoid reducing share-holders in the period of disclosing SCR, for the reduction is contrary to the fulfillment of social responsibility.

Research Conclusion
After the split share structure reform period, the reduction of holdings by majority shareholders had a greater impact on the stock market. On the one hand, a significant increase in the market supply after the non-tradable shares lifted, leading to oversupply, leading to oversupply and market equilibrium prices are naturally falling. On the other hand, market investors also follow the majority shareholders to sell shares, resulting in the second decline of market prices. In particular, it is the malicious holdings of majority shareholders, not only further influence the securities market, but also make the original normal fluctuation of shares becomes "abnormal". Therefore, this paper analyzes the effect of social responsibility on the continuous reduction of the majority shareholders by using the annual sample of 8019 listed companies in Shanghai and Shenzhen during the period of 2007 to 2013. The paper analyzes the effect of social responsibility on the continuous reduction of the majority shareholders with considering the mediating role of investor's tendency. The study finds that there is a significant negative correlation between social responsibility and continuous reduction of majority shareholders from the view of internal governance effect, that is, the better the social responsibility is performed, the lower probability of continuous reduction. There was also a significant negative correlation between the degree of the tendency of investors and the continuous reduction of the same majority shareholders, namely in the listed companies with the investors propensity, the lower probability of majority shareholders reduction. And after considering the impact of investor's tendency, the inhibitory effect of social responsibility was amplified, indicating that the market will agree the inhibitory effect arising from the social responsibility within the listed companies. Further study found that majority shareholders will avoid the period of disclosing social responsibility report to reduce shareholdings.
In addition, as the regulation of the reduction of majority shareholders, relying solely on the regulatory policy is not enough, because even the perfect policy may have loopholes, the source of the idea that majority shareholders reduce holdings can't be removed. So in addition to the proposed social responsibility and the attitude of the securities market investors, how to improve the management system through the regulation of majority shareholders reductions should be the follow-up study.