• No results found

The influence of margin trading and securities lending on the performance of stocks in China

N/A
N/A
Protected

Academic year: 2021

Share "The influence of margin trading and securities lending on the performance of stocks in China"

Copied!
35
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

MASTER THESIS

The influence of margin trading and securities

lending on the performance of stocks in China

Name: Junliang Shen

Student nr.: S2542943

Email: j.shen.4@student.rug.nl

Study Program: MSc Finance

Supervisor: Mr. Sibrand Drijver

Faculty of Economics and Business University of Groningen

(2)

Abstract

On 31 March 2010, the China Securities Regulatory Commission (CSRC) permitted margin trading and securities lending on the pilot stocks for the first time, and this project contained 900 stocks by the end of February 2015. This paper works on the influence of margin trading and securities lending on the returns and volatilities of eligible stocks. The study on the return performance is through the analysis of the Jensen's alpha and average abnormal daily return (AADR), and the beta of the capital asset pricing model (CAPM) and realized volatility (RV) are utilize to measure the performance of volatility. By means of comparing the calculated returns and volatilities between pre- and post-period, the results show that margin trading and securities lending overall reduces both the daily return and volatility for eligible stocks.

Key words: margin trading, securities lending (short selling), return, volatility,

(3)

1. Introduction

The Chinese capital market experienced rapid growth during last two decades, and became one of the most important financial markets in the world. As the largest emerging market, the lack of leverage instruments and short mechanism were its obvious shortcomings. However, on 31 March 2010, the China Securities Regulatory Commission (CSRC) permitted margin trading and securities lending on the pilot stocks for the first time. Under this project, the CSRC approved 90 blue-chip securities, which include 50 constituent stocks of the Shanghai Stock Exchange (SSE) 50 Index and 40 from the Shenzhen Stock Exchange (SZSE) Component Index, for margin trading and short selling. In particular, 4 stocks from the Small and Medium-size Enterprise Market (SMEM) were involved since they were the constituent stocks of the SZSE Component Index. According to the announcement of the CSCR, the purpose of implementing this project is to form more proper stock price by incorporating more information into it, because investors can conduct margin trading or short selling when they think certain stock price is too low or too high.

According to the regulations promulgated by the CSRC, only those investors with capital of no less than 500,000 Yuan (54,2001 euros or 73,2502 U.S. dollars) can participate in margin trading and securities lending, and this rule should be strictly implemented. Furthermore, short selling should follow the up-tick rule and naked short selling is strictly prohibited in the Chinese capital market (Chang, Luo, and Ren, 2014).

The list of eligible securities has been revised several times since the pilot scheme became a routine practice in 2011. This list has been expanded to include

1

The exchange rate is 0.1084, which was the close price on March 31st, 2010, for the Chinese Yuan to Euro and it is collected

from the software Dazhihui. Dazhihui software (http://www.gw.com.cn/) is a leading securities information platform in China, and it contains real-time quotes of stock markets, currencies, commodities and futures. It also supply market analysis and all open market information. All the other exchange rates are collected from the software Dazhihui as well.

2

(4)

900 eligible stocks and 14 exchange-traded funds (ETFs) by the end of 2014. Especially, in 2013, 6 stocks from the Chinese Growth Enterprise Market (ChiNext) became the eligible securities for the first time, which indicates that the mechanism of the ChiNext became more consummate. Table A shows the expanded amount of eligible securities, and Table B shows the distribution proportion for each time. Although the proportion of the stocks from the main board decreases as time goes on, they still dominate in all eligible securities.

Table A. Amount of eligible securities

SSE SZSE SMEM ChiNext ETFs Total

Mar-2010 50 36 4 / / 90 Dec-2011 180 80 18 / 7 287 Jan-2013 300 128 66 6 10 510 Sep-2013 400 143 123 34 14 714 Sep-2014 500 171 172 57 14 914 Mar-2015 498 170 171 57 16 912

Table B. Proportion of eligible securities

SSE SZSE SMEM ChiNext ETFs

Mar-2010 55.56% 40.00% 4.44% / / Dec-2011 62.72% 27.87% 6.27% / 2.44% Jan-2013 58.82% 25.10% 12.94% 1.18% 1.96% Sep-2013 56.02% 20.03% 17.23% 4.76% 1.96% Sep-2014 54.70% 18.71% 18.82% 6.24% 1.53% Mar-2015 54.61% 18.64% 18.75% 6.25% 1.75%

At the end of August 2012, the CSRC implemented the scheme of "refinancing", which allows banks, mutual funds, and insurance companies to lend out money or securities to investors. The brokers also loosened the margin requirements for investors. This led to rapid growth of margin trading but low growth rate of securities lending. The remaining balance of margin trading grew from 66.21 billion Yuan (8.293 billion euros or 10.434 billion U.S. dollars) to 1152.38 billion Yuan (163.985 billion euros or 183.756 billion U.S. dollars) whereas that of securities lending only increased

3

The exchange rate is 0.1252, which was the close price on August 31st, 2012, for the Chinese Yuan to Euro.

4

The exchange rate is 6.3486, which was the close price on August 31st, 2012, for the U.S. dollar to Chinese Yuan.

5

The exchange rate is 0.1423, which was the close price on February 27th, 2015, for the Chinese Yuan to Euro.

(5)

from 1.29 billion Yuan (0.16 billion euros or 0.20 billion U.S. dollars) to 5.55 billion Yuan (0.79 billion euros or 0.88 billion U.S. dollars) during September 2012 to February 2015. Chang, Luo, and Ren (2014) explains that this situation might be due to the short-selling mechanism is quite new to the Chinese investors and the strict up-tick rule also leads to the difficulties of short-selling. Moreover, the supply of securities lending and the high cost of short selling might be two other potential reasons. To be specific, the interest of securities lending is not lower than that of margin trading and brokers or institutions in China are used to seek for short-term profits so that they do not want to lend many stocks to investors.

Figure 1 illustrates the changes of margin trading and short selling by the end of every month. It is distinct that the remaining balance of margin trading is much larger than that of short selling and there are three different growth rates of the remaining balance of margin trading during three different period. The growth rate become larger and larger. Appendix A shows the list of monthly remaining balance of margin trading and securities lending.

Figure 1. Monthly remaining balance of margin trading and short selling with the

independent coordinate axis unit

(6)

markets, the SSE main board, the SZSE main board, the SMEM and the ChiNext. These four different boards stand for the firms with different sizes as well. The SSE and SZSE main board contain firms with large market capitalization, and the SMEM includes enterprises with medium market capitalization. Most listed firms in the ChiNext have small market capitalization.

I study the daily return through two methods, the Jensen's alpha and average abnormal daily return (AADR), and utilize both the beta of the capital asset pricing model (CAPM) and realized volatility (RV) to test the changes of volatility. Besides the work on overall analysis, I test and conclude the influences on four different boards separately. In addition, due to the period with different growth rate of remaining balance, I divide the post-period into three duration, and test the corresponding changes of daily return and volatility as well.

The aim of this paper is to indicate the overall conclusion of the influence of margin trading and securities lending on the eligible stocks, as well as the conclusions of different markets in order to indicate the different impacts of margin trading and securities lending on the stocks with different market capitalization. From the separated duration analysis, I intend to summarize the impact of rapid growth of the remaining balance of margin trading and short selling on the performances of eligible stocks. The main results of my study are that margin trading and securities lending overall reduces both the daily return and volatility for eligible stocks. In the return analysis, margin trading and short selling has more impacts on the stocks with medium and small market capitalization, and its influence becomes more obvious when the growth rate of remaining balance increases. In the volatility analysis, the volatilities of stocks with high market capitalization would reduce more under the project of margin trading and short selling.

(7)

2. Literature review

Margin trading means an investor builds up a leveraged long position on certain securities by borrowing capital from registered brokers, and it promotes the investors with a positive view on a stock but lack of funds to trade with a higher cost than ordinary trading as there is interest charged on the borrowed funds. Normally, margin trading implies the prices of eligible stocks should be higher than their peers with the change of the regulation. However, margin traders are generally blamed for producing excess volatility and destabilizing the market because they often receive positive inside information about a firm and purchase its stock on margin in order to enlarge their returns (Chang, Luo, and Ren, 2014; Sharif, Anderson, and Marshall, 2014; Zhao et al., 2013). Short selling is considered as a natural proxy for the level of information asymmetry, because the short sellers take advantage of private information which is not integrated into the stock price and trade before negative information reaches the public (Diamond and Verrecchia, 1987). Christophe, Ferri, and Angel (2004) also prove that short sellers are usually assumed to be more sensible and informed than other investors.

2.1. Impact on the returns of stocks

As what is mentioned above, margin trading usually increases the returns of stocks. Hirose, Kato, and Bremer (2009) find a significant cross-sectional relationship between margin trading and stock returns and margin trading leads to positive subsequent returns of stocks in Japan.

The theoretical studies of Miller (1977) and Chen, Hong, and Stein (2002) show that short-sale constraints result in overvaluation of stocks due to the pessimistic investors are not allowed to trade in the market, which hints that short selling decreases the price.

(8)

Table C lists five former studies which work on the changes in stock return after being permitted to short sell. Only Lamba and Ariff (2006) finds that short selling would increase the return of corresponding stock. They research on the Malaysian capital market with the trading data of 1996 and focus on the event study about the short-term return of approved stocks after being permitted to short sell. They find the mean and median cumulative abnormal returns (CAR) over 10 days are both 0.9% at 5% significance level.

Table C. Changes in the returns after short selling

Author(s) Year Country Period Preformance Jones and Lamont 2002 U.S 1926-1933 Decrease Ofek and Richardson 2003 France 1998-2000 Decrease Lamba and Ariff 2006 Malaysia 1996 Increase Chang, Cheng, and Yu 2007 Hong Kong 1994-2003 Decrease Sharif, Anderson, and Marshall 2014 P. R. China

&Hong Kong 2009-2010 Decrease Jones and Lamont (2002) find that the size-adjusted monthly returns are 1-2% lower for the new approved stocks in the U.S and Ofek and Richardson (2003) describe that the daily excess return of all observations is -0.11% in the full period from 1998 to 2000. The researches on Hong Kong and P. R. China are more close to our study. There are statistically significant post-period declines in cumulative abnormal returns for stocks without short-sales constraints in the Hong Kong capital market (Chang, Cheng, and Yu, 2007). Sharif, Anderson, and Marshall (2014) study on the firms with both stocks in the Chinese and oversea capital market, and they finds that the premium between these two stock price falls by 6.1% with the 1% significance level. This implies that the prices of Chinese stocks decrease in the post-period compared with the pre-period.

(9)

trading predominates in the market. Therefore, the hypothesis 1 is:

Margin trading and securities lending have significant positive influence on the returns of eligible stocks in China.

Figure 2. Monthly remaining balance of margin trading and short selling with the

unified coordinate axis unit

2.2. Impact on the volatilities of stocks

As what is mentioned in Table D and Table E, there is no unitive conclusion about the change of volatility between the pre- and post-period of the scheme of margin trading and securities lending.

Table D. Impact of margin trading on the volatilities of stocks

Author(s) Year Country Period Preformance Ferris and Chance 1988 U.S. 1945-1974 Null

Seguin 1990 U.S. 1976-1987 Increase

Hardouvelis and

Peristiani 1992 Japan 1951-1988 Decrease Hsieh and Miller 1990 U.S. 1936-1974 Null

Salinger 1990 U.S. 1934-1989 Null

Lee and Yoo 1993 Japan, Korea, and

Taiwan, 1975-1990 Null Sharif, Anderson, and

Marshall 2014

P. R. China

(10)

volatility, which means that margin trading would increase the volatility. The majority of evidence in Sharif, Anderson, and Marshall's (2014) research points to a similar conclusion that there is a general decline in relative volatility. On the contrary, Hardouvelis and Peristiani (1992) find a small decrease in the volatility of stocks when margin requirements are increased in Japan.

Most former studies on the post-period, including Ferris and Chance (1988), Hsieh and Miller (1990), Salinger (1990), find that there is no relation between the changes in margin requirements and subsequent volatilities of stocks in the United States. Lee and Yoo (1993) also conclude that the changes in margin requirements do not influence the volatility of market in Japan, Korea, and Taiwan.

Table E. Impact of short selling on the volatilities of stocks

Author(s) Year Country Period Preformance Scheinkman and Xiong 2003 U.S. 1996-2000 Decrease

Chang, Cheng, and Yu 2007 Hong Kong 1994-2003 Increase Boehmer, Jones, and

Zhang 2009 U.S. 2008 Decrease

Scheinkman and Xiong (2003), and Boehmer, Jones, and Zhang (2009) describe that short-sale constraints may result in the increase of volatility in the United States. This implies that short selling would reduce the volatility of eligible stocks. Nevertheless, Chang, Cheng, and Yu, (2007) find the volatilities of individual stocks become higher when short selling is practiced. Specifically, they find that the standard deviations of raw returns and abnormal returns are both significantly increased when stocks can be sold short.

As so many opposite results are presented and based on the status quo of the Chinese capital market that the margin requirements decrease and low trade value of short-sale, the hypothesis 2 is:

(11)

3.Research design

I intend to analyze the influence of margin trading and securities lending on the performance of stocks in China through two aspects, the changes in daily return and volatility. The Jensen's alpha and AADR are utilized to explain the changes in return, and the beta from CAPM and RV are applied to describe the changes in volatility.

For the sake of stringency and robustness, except the test for all observations, all observations are divided into four groups according to from which board they are. This classification shows the difference in market capitalization as well. The groups of the SSE and SZSE main board stand for the observations with large market capitalization, and the group of the SSE main board is a bit larger than the group of the SZSE main board. The group of the SMEM represents the observations with medium market capitalization, and the group of the ChiNext represents the observations with small market capitalization.

Moreover, due to the huge difference among the growth rates of remaining balance of margin trading and securities lending in different periods, I separate the time line into three periods, which are before the end of 2012, between the beginning of 2013 to the end of June, 2014, and after the beginning of July, 2014. Figure 3 expresses the different growth rates and identifies the three different

(12)

periods in three colours.

3. 1. Data collection

The samples analyzed in this study consist of the constituent stocks from the SSE main board, the SZSE main board, the SMEM and the ChiNext. The SSE A-stock Index, the SZSE A-stock Index, the SZSE SME Price Index and the ChiNext Price Index are the proxies of the market portfolio relatively. I use the data of daily returns from as early as January 1st, 2000 to the day before being approved stocks of margin trading and short selling to make the estimation of pre-period. Using such an early date as the start of pre-period is because I try to take more factors into account in the estimation of pre-event return and volatility, especially for the constituent stocks from the SSE and the SZSE main boards. Part of these samples were listed in 1990s and their share prices experienced not only the smooth and steady movement between 2000 and 2005 and after 2010, but the undue and drastic movement during 2006 to 2010. The estimation results of pre-period got from such a large amount of data might be more rational and comparable. However, due to the different date of being listed and being eligible stock, each sample has different starting date of pre-period and post-period estimation in this analysis. All data and information are collected from the Dazhihui7 software, the SSE and the SZSE8 and the website Eastmoney 9.

The return and volatility estimations are based on these two assumptions: 1. The securities price has no drift, which would lead to the problem of overestimation, because of the dividends payments and share split.

2. There is no price jump at opening which means the closing price is the same

7

Dazhihui software supplies the trading data of all observations and relative indices.

8

The SSE and SZSE supply the data of margin trading and securities lending, such as the date of being approved for all observations and the elimination information. Moreover, they provide the information of reverse takeover as well.

9 Website Eastmoney (http://www.eastmoney.com/) is a major financial information supplier in China, it supplies not only

(13)

as the next trading day’s opening price because the price jump would result in the problem of underestimation.

Therefore, the daily return is calculated as natural logarithm return between two trading days' closing price and all collected data of closing prices should be processed by the rehabilitation10 of historical share split and dividends payment.

The criteria of data collection and sample selection are as follow:

1. The natural logarithm daily return should be calculated as early as January 1st, 2000, and the trading data should be collected from one day before.

2. The trading data of samples from the SSE and SZSE main boards is collected from as early as December 31st, 1999.

3. The trading data of samples from the SMEM and the ChiNext is collected from as early as June 7th, 2005 and August 19th, 2010, relatively, which were one day before the establishment of the SZSE SME Price Index and the ChiNext Price Index.

4. The selected sample did not experience a reverse takeover between January 1st, 2000 and February 27th, 2015.

5. The selected sample was never removed from the list of eligible securities. 6. There is no negative close price (data error) after the rehabilitation of historical share split and dividends payment between January 1st, 2000 and February 27th, 2015.

Figure 4 shows that 82 observations from the SSE main board, 49 observations from the SZSE main board and 5 observations from the SMEM experienced reverse takeovers; 32 observations from the SSE main board and 4 observations from the SZSE main board were removed from the eligible list; negative close price11 (data

10

Rehabilitation means to eliminate the gap between the prices before and after dividends payments and share split. For instance, if one share is splitted into two shares, the price of this stock would become half; if there is a one Yuan dividend payment per share, the stock price would drop by one Yuan after it. Rehabilitation can help to eliminate such influences in order to keep all trading data continuous and avoid over- or underestimation.

11 The potential reason of the appearance of negative close price (data error) might be the database, which the SSE and SZSE

(14)

error) appears in 10 observations from the SSE main board and 1 observation from the SZSE main board.

Figure 4. The distribution of observations with different conditions

After filtration, there are 374 samples from the SSE main board, 116 samples from the SZSE main board, 166 samples from the SMEM and 57 samples from the ChiNext left in this study.

3.2. Methodology

3.2.1. Natural logarithmic daily return

Natural logarithmic daily return is utilized for all observations in this study, and the formula is as follow:

1 ln( t ) i t P R P  (1)

where Pt is the underlying reference price at time t; Pt−1 is the underlying reference

price to the time period preceding time t.

(15)

The RV provides a relatively accurate measure of volatility, and it is quite useful in the evaluation of volatility. Andersen et al. (2001) argue that the RV is an unbiased and very efficient estimator of return volatility. I calculate the annual RV according the following formulation which is mentioned in Pilbeam and Langeland's article (2015). 2 1 1 252*( ) 1 N it i R RV N    

(2)

where Rit is the natural logarithmic daily return of stock i on day t; N is the number of

trading days in the period; 252 is a constant representing the approximate number of trading days in a year.

3.2.3. Capital asset pricing model (CAPM)

The CAPM is used in the studies on stock performance, and it is based on single-index model.

( )

it f i i mt f it

RR   RR  (3)

where the intercept (i) is the Jensen's alpha (Jensen, 1968), which is typically

interpreted as a measure of out- or under-performance relative to a market proxy;

i

 is the sensitivity of the risk-adjusted stock return to the risk-adjusted benchmark index returns; Rit is the daily return of stock i on day t; Rf is the average daily

return of one-year Chinese government bond during last 13 years12; Rmt is the daily return of the relevant benchmark index on day t; and it is the error term.

12

(16)

3.2.4. Average abnormal daily return (AADR)

AADR is utilized to test whether margin trading and securities lending has impact on the return of eligible stocks, and it is also applied to check the robustness of return analysis by the test of Jensen's alpha from the CAPM. After the regression analysis of CAPM, I get the value of  and  for each observation in pre-period. I check the difference between real daily return in post-period and corresponding expected daily return with pre-period CAPM regression functions as the abnormal daily return (ADR), the formula is as follow:

( ) ( )*( )

it it f i pre i pre mt f

ADRRR   RR (4)

where Rit is the daily return of stock i on day t; Rf is the average daily return of one-year Chinese government bond during last 13 years; Rmt is the daily return of the relevant benchmark index on day t; i pre( ) is the alpha of pre-period CAPM on

stock i; i pre( ) is the beta of pre-period CAPM on stock i.

AADR is the average ADR for all observations in different periods, its formula is as follow: 1 D it t i ADR AADR D  

(5)

where ADRit is the abnormal daily return of stock i on day t; D is the duration of each

period.

3.2.5. Jarque-Bera test

(17)

2 1 2 ( 3) 6 4 n JB SK    (6)

where n is the number of observations; S is the sample skewness; K is the sample kurtosis.

3.2.6. One sample T-test

One sample T-test is adopted to test the significance of difference and proportions of Jensen's alpha, AADR, beta and RV between pre- and each post- period for all observations. T test statistic is as follow:

2 1 1 ( ) it it it DR T t DR DR T T  

(7)

where DRit is average difference between post-period return of stock i and

expected return with pre-period regression model on day t; T is the number of days.

4. Results

4.1. Return analysis

Table F indicates the statistical results of the changes of returns during the full period for all classifications. Both the Jensen's alpha and AADR show that the permission of margin trading and securities lending reduces the overall daily return for eligible stocks. The difference between the Jensen's alpha in pre- and post-period is 0.0003 (0.03%) and the AADR is 0.0006 (0.06%) with 1% significance level.

(18)

Table F. Statistical results of the changes of returns during the full period

Mean St. Dev. Skewness Kurtosis

Jensen's alpha All -0.0003* 0.00156 0.412 10.085 SSE -0.0001 0.00128 1.063 7.390 SZSE -0.0003* 0.00123 1.551 6.583 SMEM -0.0007* 0.00165 -0.910 10.911 ChiNext -0.0010* 0.00270 1.361 6.833 AADR All -0.0006* 0.00159 0.373 6.166 SSE -0.0005* 0.00142 0.124 5.197 SZSE -0.0004* 0.00136 1.784 7.544 SMEM -0.0006* 0.00164 -0.060 7.066 ChiNext -0.0009* 0.00258 0.817 3.737

* Significant at the 1% level. ** Significant at the 5% level. *** Significant at the 10% level. Not significant if there is no star.

margin trading and securities lending has more impacts on the stocks with medium and small market capitalization in the Chinese capital market since the volume of margin trading and short selling of the stocks with high market capitalization might occupy lower proportion to the total trading volume than the stocks with low market capitalization.

(19)

Table G. Statistical results of the changes of returns during the separate period

2010.03-2012.12 2013.01-2014.06 2014.07-2015.02

Mean St. Dev. Skewness Kurtosis Mean St. Dev. Skewness Kurtosis Mean St. Dev. Skewness Kurtosis

Jensen's alpha All -0.0003* 0.00112 -1.026 9.555 -0.0005* 0.00159 0.731 10.002 0.0000 0.00211 0.924 8.510 SSE -0.0003** 0.00115 -1.644 11.820 0.0000 0.00131 0.296 4.587 0.0004* 0.00186 1.115 5.194 SZSE -0.0004* 0.00087 -0.005 3.193 -0.0010* 0.00125 0.297 3.188 0.0002 0.00163 0.468 2.806 SMEM -0.0004 0.00157 0.540 3.427 -0.0008* 0.00190 2.564 18.596 -0.0006* 0.00209 -0.347 6.176 ChiNext / / / / -0.0019* 0.00217 0.024 3.250 -0.0005 0.00356 1.993 8.658 AADR All -0.0003* 0.00116 -0.824 9.119 -0.0005* 0.00160 0.762 10.236 -0.0006* 0.00225 1.241 7.965 SSE -0.0003* 0.00117 -1.645 11.524 -0.0001 0.00129 0.255 4.671 -0.0008* 0.00216 1.054 4.802 SZSE -0.0002 0.00102 0.695 5.065 -0.0011* 0.00130 0.366 3.413 -0.0001 0.00185 0.783 3.394 SMEM -0.0004 0.00161 0.632 3.537 -0.0008* 0.00190 2.663 19.402 -0.0006* 0.00206 0.214 4.943 ChiNext / / / / -0.0018* 0.00222 0.084 3.147 -0.0004 0.00359 2.010 9.162

(20)

securities lending becomes more when the growth rate of remaining balance increases.

Inconsistent results appear in the SSE classification. Although both the Jensen's alpha and the AADR show that the return declines 0.0003 (0.03%) during the first period and no significant impact during the second period, the Jensen's alpha expresses a 0.0004 (0.04%) increase in return whereas the AADR tells a 0.0008 (0.08%) drop during the third period.

The results of other three classifications, the SZSE, the SMEM, and the ChiNext, show consistency. There is a small drop during the first period, about 0.0010 (0.1%) drop during the second period, and no significant influence of margin trading and short selling during the third period for the SZSE. The SMEM experiences 0.0008 (0.08%) and 0.0006 (0.06%) decrease in daily return during the second and third period relatively but no significant change during the first period. The ChiNext gets a prominent drop, about 0.0018 (0.18%), in daily return during January 2013 to June 2014. This indicates that the margin trading and short selling have more impacts on the stocks with low market capitalization.

4.2. Volatility analysis

Two items, difference and proportion, are tested in both the beta of the CAPM and the RV sections. Difference presents the difference between the value of pre- and each post-period (ValuepostValuepre), and proportion means the ratio between

two period ( post

pre

Value

Value ). The results of difference are used in the explanations

and the results of proportion are only added as supplementary notes. All results of proportion are significant at 1% level.

(21)

Table H. Statistical results of the changes of volatilities during the full period

Difference Proportion

Mean St. Dev. Skewness Kurtosis Mean St. Dev. Skewness Kurtosis

Beta of the CAPM

All -0.1297* 0.30663 -0.311 4.198 0.8931* 0.27647 0.032 4.706 SSE -0.2402* 0.29956 -0.458 3.633 0.7957* 0.25908 -0.120 3.928 SZSE -0.1261* 0.21822 0.458 3.311 0.8921* 0.18761 0.665 3.053 SMEM 0.0482** 0.26932 0.367 4.778 1.0497* 0.26487 0.555 5.880 ChiNext 0.0694*** 0.26550 -2.033 12.299 1.0785* 0.27330 -1.796 12.191 RV All -0.0770* 0.10776 0.313 3.626 0.8517* 0.21927 0.966 4.283 SSE -0.0938* 0.10707 0.344 4.053 0.8197* 0.21964 1.224 5.307 SZSE -0.1020* 0.10104 0.598 4.068 0.8011* 0.19156 1.217 4.439 SMEM -0.0581* 0.09479 0.301 3.284 0.8843* 0.19338 0.609 3.431 ChiNext 0.0286** 0.09083 0.381 3.317 1.0698* 0.20191 0.896 3.786

(22)

overall volatility by about 10-15%.

Except the result of the SMEM, results of the beta of the CAPM and the RV are consistent in other three classifications. Margin trading and short selling reduce the volatilities of the SSE and the SZSE while it increase the volatility of the ChiNext during the full period. To be specific, the beta of the CAPM declines 0.2402 and the RV declines 0.0938 for the SSE group at 1% significant level, and both change about 20%. The beta of the CAPM declines 0.1261 and the RV declines 0.1020 for the SZSE group, which changes about 10% and 20% relatively, at 1% significant level. The ChiNext group gains a 0.0694 increase for the beta of the CAPM at 10% significant level and 0.0286 increase for the RV at 5% significant level, and both change about 7%. Putting the SMEM group aside, we conclude that margin trading and short selling would reduce the volatilities of stocks with high market capitalization and the more market capitalization is, the more decrease in the volatility occurs. Moreover, margin trading and short selling would increase the volatilities of stocks with low market capitalization.

Table I presents the statistical results of the changes of volatilities during the separate period. There is little inconformity in this separate-period analysis. During March 2010 to December 2012, the result of beta of the CAPM shows that margin trading and short selling leads to an overall 0.0898 increase whereas the result of the RV indicates a 0.1499 drop for the volatility with 1% significance level. The results of the SSE group disagree with each other as well. There is a 0.1088 increase in the beta analysis while there is a 0.1558 drop in the RV analysis, and both results are significant at 1%. Nevertheless, the results of the SZSE and the SMEM show the same conclusion that margin trading and short selling leads to a lower volatility for eligible stocks, but the results of RV drop much more than the results of beta.

(23)

Table I. Statistical results of the changes of volatilities during the separate period

Difference Proportion

Mean St. Dev. Skewness Kurtosis Mean St. Dev. Skewness Kurtosis

2010.03-2012.12

Beta of the CAPM

All 0.0898* 0.16259 0.837 4.612 1.0789* 0.15399 0.851 6.024 SSE 0.1088* 0.26777 0.412 3.192 1.0919* 0.26474 0.732 5.446 SZSE -0.1196* 0.18210 0.057 2.829 0.8948* 0.16218 0.412 2.794 SMEM -0.0285 0.18117 0.544 3.471 0.9744* 0.18855 0.525 3.042 ChiNext / / / / / / / / RV All -0.1499* 0.07442 0.414 3.151 0.6936* 0.14915 0.533 3.418 SSE -0.1558* 0.07873 0.575 3.277 0.6758* 0.16200 0.714 3.506 SZSE -0.1329* 0.06211 0.155 3.030 0.7338* 0.11501 0.449 2.928 SMEM -0.1620* 0.07386 0.430 2.723 0.6928* 0.12909 0.589 3.263 ChiNext / / / / / / / / 2013.01-2014.06

Beta of the CAPM

(24)

SSE -0.1244* 0.10617 -0.014 3.234 0.7519* 0.20486 0.468 3.294

SZSE -0.1364* 0.09576 0.375 4.318 0.7297* 0.17812 0.898 4.138

SMEM -0.0637* 0.10020 0.060 2.695 0.8765* 0.20044 0.327 2.723

ChiNext 0.0512* 0.08613 0.125 5.162 1.1191* 0.20454 1.391 7.770

2014.07-2015.02

Beta of the CAPM

All -0.2171* 0.37751 -0.130 3.365 0.8209* 0.33445 0.317 3.659 SSE -0.4012* 0.34702 -0.035 3.020 0.6609* 0.29358 0.539 3.508 SZSE -0.1265* 0.25492 -0.167 3.621 0.8979* 0.21427 0.264 3.140 SMEM 0.0346 0.30449 0.620 4.491 1.0381* 0.30190 0.834 5.080 ChiNext 0.0778** 0.28963 -1.247 9.667 1.0859* 0.29886 -1.057 9.617 RV All -0.0739* 0.12157 0.196 3.992 0.8601* 0.25202 1.056 4.718 SSE -0.0780* 0.12582 0.057 4.398 0.8599* 0.26501 1.259 5.429 SZSE -0.0933* 0.11792 0.324 3.370 0.8125* 0.23692 0.826 3.629 SMEM -0.0785* 0.10713 0.499 3.002 0.8395* 0.21861 0.742 3.158 ChiNext 0.0065 0.11187 0.561 3.460 1.0181* 0.22803 0.643 2.831

(25)

margin trading and short selling reduces the volatility for these two groups. Meanwhile, the results of the ChiNext group indicates an increase in the volatility.

During the last period, the results perform better. The differences of beta and RV between pre- and this post-period are -0.2171 and -0.0739 relatively, which imply that the volatility reduce by about 15% for all samples. Getting rid of the insignificant results, margin trading and short selling reduce the volatilities for all classifications besides the group of the ChiNext. The results of the ChiNext group shows an increase in the volatility, which is as the same consequence in the second period.

Because the beta of the CAPM represents the sensitivity of the stock return to the benchmark index returns, it might not tell the exact changes of volatility between pre- and post-period for all observations due to such relative values. Compared with the beta of the CAPM in pre-period, the beta in each post-period would be over- or underestimated due to the different changes of the RV between observations and benchmarks. However, the RV is the intrinsic value, and it only takes the daily return of samples into account. Therefore, the results of the RV analysis are more reliable. We conclude that margin trading and securities lending could reduce the volatility of the SSE and SZSE main boards, and the SMEM while it increase the volatility of the ChiNext.

Furthermore, based on the results of the RV analysis, it is obvious that the decreases of the volatilities for the SSE, the SZSE and the SMEM become smaller as time goes on. This implies that the rapid growth of the quantity and remaining balance of margin trading and securities lending weakens its impact on the volatility. In other words, the rapid growth of the quantity of margin trading and securities lending aggravates the volatility again compared with the former period.

(26)

proportion values are not close to each other for the same group during the same period.

From the Table J and Table K, the RVs of both stocks and indices decrease in the SSE, SZSE and SMEM group. In this situation, if the stock return volatility drop more than that of benchmark return volatility, the beta would be underestimated relative to the beta in pre-period, and vice versa. The ChiNext group shows different results from the other three groups. If the RV of observations increases and that of indices decrease, the beta would be overestimated relative to the beta in pre-period. If the RV of observations increases more than that of indices, the beta would be overestimated relative to the beta in pre-period as well.

Table J. Ratios of the RV of observations (RVpost / RVpre)

Table K. Ratios of the RV of indices (RVpost / RVpre)

Hereby, I introduce a discount rate for all betas in the post-period. All discount rates are calculated by the following formulas (8) and (9) and results are shown in Table J:

Table L. Discount rate for the beta of the CAPM (RVpost / RVpre)

SSE SZSE SMEM ChiNext

Full period 0.81974 0.80105 0.8843 1.06978 2010.03-2012.12 0.67579 0.73376 0.69275 / 2013.01-2014.07 0.75194 0.72966 0.87649 1.11913 2014.07-2015.02 0.85995 0.81251 0.83946 1.01813

SSE SZSE SMEM ChiNext

Full period 0.71353 0.79895 0.65909 0.98403 2010.03-2012.12 0.71627 0.85172 0.70041 / 2013.01-2014.07 0.62435 0.74561 0.62404 1.0456 2014.07-2015.02 0.87022 0.6763 0.54743 0.84758

SSE SZSE SMEM ChiNext

(27)

index stock Ratio d Ratio  (8) * Discounted d Original    (9)

Table M shows the statistical results of beta of the CAPM after discount. The results indicate that margin trading and securities lending reduces the volatility of all samples in the full post-period by 0.2629, as 23.05%, at 1% significance level. Moreover, the volatilities of the SSE, the SZSE, and the SMEM groups are also reduced by 0.3567, 0.1288, 0.2291 relatively at 1% significant level. This agrees with the results of the RV analysis in the former part.

The results during March 2010 to December 2012 are still against the results of the RV. It shows a 0.1250 increase of the volatility for all samples and 0.1821 increase of the volatility for the SSE group with 1% significance level. No significant results are presented for the group of the SZSE and the SMEM.

(28)

Table M. Statistical results of beta of the CAPM after discount

Difference Proportion

Mean St. Dev. Skewness Kurtosis Mean St. Dev. Skewness Kurtosis

(29)

SMEM -0.3396* 0.21053 0.489 4.745 0.6770* 0.19753 0.836 5.080

ChiNext -0.1076* 0.24669 -1.116 8.811 0.9040* 0.24880 -1.057 9.617

(30)

5. Conclusion

In general, margin trading and securities lending reduces both the daily return and volatility for eligible stocks. To be specific, the results of both full post-period and separate post-periods show a decrease in daily return for all groups, the SSE, the SZSE, the SMEM and the ChiNext, whereas the results of volatility analysis is unitive in the full post-period but a bit ambivalent results in some post-period for certain groups. Besides the results which shows an increase of the volatility for all samples and the SSE group during March 2010 to December 2012, all samples, the SSE, the SZSE, and the SMEM groups experience decrease in the volatility. The ChiNext group experiences a gentle increase in the volatility in the meantime.

Moreover, I find that margin trading and short selling reduces more on the return of stocks with low market capitalization, and increases the volatilities of these stocks. In the meantime, margin trading and short selling reduces the volatilities of stocks with high market capitalization and the more market capitalization is, the more decrease of the volatility occurs.

Although the results of volatility analysis are corresponding with my hypothesis, that margin trading and securities lending would reduce the volatilities of eligible stocks in China, the outcomes of return analysis do not match. The potential reason is that the project of margin trading and securities lending was launched in 2010, and I study its influence on the performance of stocks during March 2010 to February 2015. Most trading days during this period were in the "Bear" market, which means that most stocks experienced decline trend, so that they got less returns in my research. Even though the margin trading is practiced and its remaining balance grew quickly, it cannot reverse the negative abnormal return to a positive value. In addition, a huge amount of foul margin financing from the Over The Counter (OTC) market is reported by the CSRC, and it would influence the results as well.

(31)
(32)

References

Andersen, T. G., Bollerslev, T., Diebold, F. X., and Labys, P., 2001. The Distribution of Realized Exchange Rate Volatility. Journal of the American Statistical Association 96, 42-55.

Boehmer, E., Jones, C. M., and Zhang, X., 2009. Shackling short sellers: the 2008 shorting ban. Review of Financial Studies 26, 1363-1400.

Chang, E. C., Cheng, J. W., and Yu, Y., 2007. Short-sales constraints and price discovery: evidence from the Hong Kong market. Journal of Finance 62, 2097-2121. Chang, E. C., Luo, Y., and Ren, J., 2014. Short-selling, margin-trading, and price efficiency: Evidence from the Chinese market. Journal of Banking & Finance 48, 411-424.

Chen, J., Hong, H., and Stein, J. C., 2002. Breadth of ownership and stock returns. Journal of Financial Economics 66, 171-205.

Christophe, S., Ferri, M., and Angel, J., 2004. Short selling prior to earnings announcements. Journal of Finance 59, 1845-1875.

Diamond, D., and Verrecchia, R., 1987. Constraints on short selling and asset price adjustment to private information. Journal of Financial Economics 18, 277-311. Ferris, S., and Chance, D., 1988. Margin requirements and stock market volatility, Economics Letters 28, 251-254.

Hardouvelis, G. A., and Peristiani, S., 1992. Margin requirements, speculative trading, and stock price fluctuations: the case of Japan. The Quarterly Journal of Economics 107, 1333-1370.

Hirose, T., Kato, H. K., and Bremer, M., 2009. Can margin traders predict future stock returns in Japan? Pacific-Basin Finance Journal 17, 41-57.

(33)

Journal of Finance 45, 3-30.

Jensen, M., 1968. The performance of mutual funds in the period 1945-1964. Journal of Finance 23, 389-416.

Jones, C. M., and Lamont, O. A., 2002. Short-sale constraints and stock returns. Journal of Financial Economics 66, 207-239.

Lamba, A. S., and Ariff, M., 2006. Short selling restrictions and market completeness: the Malaysian experience. Applied Financial Economics 16, 385-393.

Lee, S. B., and Yoo, T. Y., 1993. Margin regulation and stock market volatility: further evidence from Japan, Korea and Taiwan. Pacific-Basin Finance Journal 1, 155-174. Miller, E. M., 1977. Risk, uncertainty, and divergence of opinion. Journal of Finance 32, 1151-1168.

Ofek, E., and Richardson, M., 2003. Dotcom mania: the rise and fall of internet stock prices. The Journal of Finance 58, 1113-1137.

Pilbeam, K., and Langeland, K. N., 2015. Forecasting exchange rate volatility: GARCH models versus implied volatility forecasts. International Economics & Economic Policy 12, 127-142.

Salinger, M. A., 1990. Stock market margin requirements and volatility: Implications for regulation of stock index futures. Journal of Financial Services Research 3, 121-138.

Scheinkman, J. A., and Xiong, W., 2003. Overconfidence and speculative bubbles. Journal of Political Economy 111, 1183-1219.

Seguin, P. J., 1990. Stock volatility and margin trading. Journal of Monetary Economics 26, 101-121.

(34)
(35)

Referenties

GERELATEERDE DOCUMENTEN

Front page image is an example of a Geodiveristy Index (with equation 2, cell size 750m) of Fiordland National Park The left logo is of the University of Amsterdam and the right

Dit liet volgens hem zien dat er door het Westen meer macht werd uitgeoefend door middel van bilaterale hulp en dat dit enkel zorgde voor economische groei in het westerse land

throline or an active derivative thereof is useful in the treatment of several conditions, such as ischemia, wound healing and tendon damage, the invention further provides a method

Thus, we expected the capacity of 3D stimulus class to be lower compared to high-contrast (HC) and isoluminant (ISO) stimuli, since afterimages could not benefit the capacity of IM

The developments of the determinants of the interest margin over the period 1995-2005 are presented in Table A.1 in Appendix C. Looking at its most important determinants, the

The red (95% criti- cal value sequence) and the blue line (BSADF sequence) depict the detection and subsequent dating of mildly explosive behavior under PSY method (Phillips et

This study will research the relations of firm performance and size with foreign ownership based on data of the Ho Chi Minh City Stock Exchange, thus in effect including the recent

H1: Low bonuses of executives have a higher effect on the brand loyalty of the customer than high bonuses H2: A positive financial period of three years has a better effect on