Does circuit breaker in Chinese stock exchange encourage individual investors’ irrational herd behavior?

(1)

Does Circuit Breaker in Chinese Stock Exchange Encourage

Individual Investors’ Irrational Herd Behavior?

Hanlin Shen

10621776

Faculty of Economics and Business

Supervisor: Erasmo Giambona

(2)

Table of content

1. Introduction Section I

2. Circuit Breaker

2.1 History and Introductory Concept 2.2 Theoretical Review on Circuit Breaker

2.3 Actual results after introducing circuit breaker to Chinese stock market

3. Herd Behavior

3.1 History and Introductory Concept 3.2 Herd Behavior in Financial Market

3.3 Economic Psychological Approach to Herd Behavior

Section II

4. Survey and Result

4.1 Background, Assumptions and Hypothesis 4.2 Methodology

4.3 Sample Description 4.4 Data and Results 4.5 Discussion

Section III 5. Conclusion 6. Reference Appendix

I. Questionnaire: Survey Concerning Circuit Breaker and Herd Behavior in Chinese Stock

(3)

Abstract

This paper mainly studies the influence brought by circ uit breaker in Chinese stock exchange to Chinese individual investors. Circuit breaker and herd behavior literatures are reviewed in the first part. In the second part, a survey referring to the behavioral f inance literature is conducted, with a specific focus on whether circuit breaker in Chinese stock market encourages individual investors ’ more severe irrational herd behavior. A group of 60 randomly selected Chinese individual investors participate in the survey. The test result reverses the public opinion and indicates that more severe irrational herding is not triggered by circuit breaker. Being not able to obtain satisfying solution from the crowd is the main reason why individual investors choose not to follow the majority.

1. Introduction

Chinese stock market experienced an agitated year in 2015. The CSI 300 index reached the peak at the first half year, following with the popping of the stock market bubble in summer. The fluctuation of the stock exchange was so drastic that the Chinese Securities Regulatory Commission (CSRC) decided to implement certain mechanism to smooth out the extreme volatility. Thus, on January 1st 2016, circuit breaker was introduced to the market. This mechanism intended to help stabilize the market, provide investors time to calm down and reduce the abnormal intraday price volatility.

However, this mechanism turned out to be a failure. The worst-ever beginning to a year for Chinese stocks triggered a trading pause in equity shares, futures and options, placing the nation’s new market circuit breaker under the test on the first trading day of 2016. The drama has not answered the curtain call yet. A more dramatic story occurred on January 7th 2016. Chinese stock market was open for only 29 minutes before the market declines triggered automatic circuit breaker, and shut down the market for the day. Circuit breaker seemed to encourage the volatility, somehow, instead of pacifying the investors and decreasing the extreme price movement. According to the observation of the index movement, it only took 7 and 3 minutes for the second halt to be triggered as shares tumbled after first halting ended, respectively. The media and the public argue that individual investors, which are from the largest group of market participants in Chinese market, may have irrational and panic behavior under the pressure of circuit breaker.

A phenomenon, which can be labeled as “a race to sell”, might spread over the market participants. Investors, who could not calm down and obtain the complete information to re-judge the market rationally, might mimic others to execute the similar selling decision. The acceleration of the index drop may be the consequence of this contagion-like phenomenon. Theoretically, this contagion among groups can be identified as irrational herd behavior.

Social media and the public accuse the circuit breaker as the culprit to encourage more severe irrational herd behavior of Chinese individual investors. However, this argument is not valid unless it is examined. The intended purpose of this paper is to discover the influence, brought by the circuit breaker, to the behavior of Chinese individual investors. Since the circuit breaker is scarcely triggered in the history due to lack of extreme volatility, the researches refer to circuit breaker are still limited at a relatively theoretical level. Hence, the influence of circuit breaker is

(4)

still not clear to the policy makers and researchers when it is implemented in practice. Therefore, the research based on real events can undoubtedly bridge the gap between academic study and practical implement. Moreover, the evidence-based study can, to some extent, examine the hypothetical arguments, and help both professional researchers and investors to learn the mechanism more thoroughly. In the case of Chinese stock exchange, this is a growing opportunity to learn the actual effect brought by the circuit breaker to a certain group of investors and help people learn whether this mechanism will trigger more severe irrational herding among the individual investors.

To discover the influence of circuit breaker to Chinese individual investors, academic literature review and a survey will be the support. The literature review on herd behavior applies the underlying assumption to the following survey. An economic psychological approach to herd behavior, which is based on two assumptions, is introduced to elaborate the reason why people sometimes follow the majority. “Cognitive miser”, which is developed by Fiske and Taylor (1991), explains that people prefer depending on heuristics such as “the majority is correct” when facing difficult situations. Another assumption refers to “incomplete information” (Abrahamson and Rosenkorf, 1993), which demonstrates that people seek for consensus when they tackle with uncertainty caused by incomplete information. The two statements stated above are also the underlying assumptions of the research. The study in this paper will be executed by a survey to a group of 60 randomly selected Chinese individual investors. The survey concentrates on investors ’ perceptions and behaviors. Sample investors are motivated to have a comparison to their own perceptions to the market and behaviors between pre-and post-mechanism-implement period. A sign test with the signif icance level of 5% will be performed to examine whether their perceptions and behaviors change signif icantly, and greater irrational herding appears after circuit breaker is introduced. The sign test result of questions which directly access if Chinese individual investors have stronger incentive to follow other investors is the proof to demonstrate whether circuit breaker will cause more severe irrational herding in Chinese stock exchange. Other questions attempt to support and interpret the ultimate result.

This paper studies whether circuit breaker mechanism encourages more severe irrational herd behavior among the individual investors. The test would be executed by conducting an anonymous survey among a group of randomly selected Chinese individual investors. First section describes the general literature review of circuit breaker and herd behavior. Second section presents the analys is and result of the survey among individual investors in China. The last section summarizes the discoveries and concludes.

Section I 2. Circuit Breaker

2.1 History and Introductory Concepts

The year after the 1987 Crash, the New York Stock Exchange and several regional stock exchanges all adopted the trading halt rules (NYSE rule 80B, 1991). Harris (1998) introduces that the circuit breakers restrict trading activity in various ways, including trading halts, price limits , transaction taxes, etc. Proponents of circuit breakers commonly hope that these limitations will decrease the equity price volatility.

(5)

Chinese stock exchange (CSI 300) adopted the circuit breaker, which is in the form of trading halts, in 2016. Chinese stock market experienced a roller-coaster-like trading year in 2015. The stock market crashed from June and the stock market bubble popped. The index plunged from 5178 points to 2850 points in only two months. After this crash, the market lost almost 45 percent of its value. Under this situation, Chinese government determined to introduce certain mechanism to the market in order to reduce the unusual extreme volatility and systematic risks. On December 4th 2015, Chinese Securities Regulatory Commission (CSRC) published the rules and purpose of circuit breaker in its bulletin, and announced that the mechanism would be implemented on the first trading day of 2016. According to the bulletin (CSRC, 2015), first-level circuit breaker will be automatically triggered when CSI 300 index plunges with 5 percent. The trading will be paused for 15 minutes. After the 15-minute trading halt, the market will re-open. However, when the index plunges with 7 percent, the trading will be stopped and the markets shut down for the rest of day automatically.

2.2 Theoretical Review on Circuit Breakers

The effectiveness of circuit breaker is still debatable. Harris (1998) points out that because of lacking extreme volatility, no satisfactory results have emerged to answer the question whether circuit breakers can protect the market. There has not been enough extreme price movement and evidence to conclusively determine whether this mechanism is effective.

The intended purpose of implementing this mechanism is to mitigate the growing concern among individual investors and institutional market participants about the excessive price movement (Subrahmanyan, 1994). Some empirical researches and papers support this statement. Bertero and Mayer (1990) study the “1987 Crash” declines of 23 major stock exchanges in the world and find that those markets with circuit breaker in operation perform 7 percent to 9 percent better than those without this mechanism. When it comes to the trading halt, Greenwald and Stein (1988, 1991) suggest that the information and confidence should be restored by trading halt followed by an orderly re-open procedure. Lauterbach and Ben-Zion (1993) demonstrate the internal logic behind the circuit breaker. The great price swings happening within several minutes create severe asymmetric information between the floor investors and the public. Most market participants can no longer be certain about the order execution price. Thus, the investors may avoid trading altogether. As a result, the market would be more chaotic. In this case, a trading halt could offer investors an opportunity to re-evaluate the current situation, obtain more information, and reconsider about the trading plan. Kodres and O’ Brien (1994) also argue that such trading halts may provide traders more time to response so that transitory volatility might be dec reased and more liquidity might be offered.

However, there are arguments suggest that circuit breakers may encourage price volatility. Subrahmanyan (1994) builds a theoretical model that demonstrates the effect, so-called magnetic effect, which describes the situation that traders may submit their orders earlier to increase the possibility of execution since they fear that a pause may occur before they are able to submit the orders. Therefore, the triggering threshold attracts the order from fearful traders and this may cause greater volatility.

(6)

2.3 Actual results after introducing circuit breaker to Chinese stock market

Theoretically, empirical evidence on the effectiveness of circuit breaker is scarce and the conclusions appear mixed. However, this mechanism obviously is proved to be a failure in Chinese stock exchange. In the first trading week in 2016, this mechanism was triggered twice. On the first trading day, trading was paused at 1:34 PM local time after CSI 300 index plunged 7 percent. Graph 1 illustrates that an earlier 15-minute trading halt at the 5 percent level failed to prevent the index from dropping; the shares kept losing their values as soon as the market re-opened and the market headed for shutting down.

Graph 1: China halts stock trading after 7% plunge on January 4th 2016, Bloomberg* On January 7th 2016, the situation was even more chaotic. Graph 2 illustrates that CSI 300 index started a “free fall” right after the market opened and kept dropping toward to the threshold of triggering the circuit breaker. Finally, the circuit breaker was triggered and the market shut down after only 29-minute trading.

(7)

Both graphs above illustrate that the market volatility was not reduced by the implement of circuit breaker. More liquidity and stability, which is supposed to be provided by the circuit breaker, can hardly be observed after re-opening the market; the index kept plunging to reach the second threshold and the market is mandated to close for the rest of day. It is obvious that the result is opposite to the authority’s intended purpose. The earlier 15-minute suspension failed to provide time for the market participants to re-evaluate the market, to gather more information or to reconsider trading plans. The index curves also imply that market participants might be

performing the similar behaviors, which is liquidating the equities in this case. According to the graphs, the index only took several minutes to touch the button line and the circuit breaker was triggered quickly after 15-minute halt. The public and media claim that large group of individual investors might have panic and irrational behaviors; hence, they rushed to dump their shares, tried in vain to cash out and escaped from the market as soon as possible since the circuit breaker brought trading to an abrupt halt. In this case, it is reasonable that public opinion blame that circuit breaker is the culprit to trigger more severe irrational herding.

3. Herd Behavior

3.1 History and Introductory Concepts

Herd behavior is identified as the phenomenon of people following the major ity in a certain period of time, and sometimes people disregard their own individual information, which even suggest something else (Banerjee, 1992). Baner jee (1994) constructs a modeling approach to demonstrate the phenomenon that even if an individual’s private information and motivation suggest something else, the individual still is likely to follow other ’s behavior. Van Ginneken (1992) mentions that herd behavior were first studied in the field of social psychology. Early researches applied this behavior to rapid shifts in consumer actions like fads and fashions (Simmel, 1957).

Initially, herd behavior referred to an unconscious and irrational process. Le Bon (1895) suggests that imitation is considered as some type of “collective hypnosis” called “social somnambulism” in which people follows the majority’s action “in mental unity”. The idea behind this collective hypnosis, which is recognized as a sort of social contagion, provided a means to prove that herd behavior is irrational and unconscious. After decades, research of convergence in groups made the insight of herd behavior head to another direction. Sherif (1935) states that individual attempts rationally to make sense of social reality. In this case, the concept of reference group is vital. People set the reference group as standard to examine their perceptions of social reality when facing uncertainty (Festinger, 1950). Recent studies concentrate on the interaction between major ity influence and minority influence. Moscovici (1980) concludes that when the individual is motivated to follow heuristics, then majority inf luence is in the dominance; whereas when the individual is motivated to examine statements thoroughly, the minority influence is stronger.

There is a general category for the literatures about herd behaviors. Graham (1999) subdivides the herd behavior literatures into the following four categories: (1) informational cascade, (2)

reputational herding, (3) empirical herding and (4) investigat ive herding. When individuals choose to disregard or downplay their personal information and hop onto the “bandwagon” (see

Leibenstein, 1950) by imitating the behaviors of others who performed previously, the first two sorts of herd behavior happen. In this paper, these two types of herding will be discussed in detail.

(8)

Informational cascade happens when people face very strong majority inf luences. The aggregate information or the decision of the crowd is too overwhelming for the individuals to reverse by their own private information. Hence, individuals prefer following the decision of crowd, rather than performing the decision-making process by own personal information. Reputational herding is similar to the cascading; it occurs when an individual choose to ignore his or her personal information and mimic the action of others. The difference between reputational herding and informational cascade is that the reputational-herding individuals mimic the actions of positive reputational externalities, such as famous experts, reputational analyst, etc.

3.2 Herd Behavior in Financial Markets

In the realm of finance, herd behavior could potentially be universe (Devenow and Welc h, 1996). In the secondary market, herding can be generally defined as the positive correlation among the behavior patterns of the individual market participants. Investors, in stock markets, predict the future share price based on different sources of unc ertain information. The price prediction may be based on some market signals (e.g., asset price movement, index movement, etc.) or the

observation of other ’s prediction (e.g., peer investors’ prediction, analysts or experts’ speculation, etc.). From the perspective of finance, two polar views on herd behavior are widely admitted, which are rational herding and irrational herding. In other words, not all herd behaviors occur in the financial market are irrational. The former considers herding as a tendency of investors who irrationally disregard their own personal information, evaluation or analysis and bow to the crowd consensus, even if they hold different opinions (Christe and Huang, 1995). Imagine that many investors are trading a “hot” corporate stock actively, which is for instance significantly undervalued or the representing listed company wins the historical high profit; it could just because independently acting investors receive correlated information. Investors may perform the similar purchasing behavior after scrutinizing the information of the stock carefully. In this case, this phenomena is not irrational herding.

This paper mainly put the eyesight on the non-rational herd behavior. The irrational herding basically concentrates on investor psychology, including following others blindly and abandoning rational analysis. According to the research conducted by Hirschleifer (2003), there are

undoubtedly some phenomena which are evidently caused by irrational herd behavior. For instance, anecdotes of abnormal asset price movement without justifying news; overestimate the company and overpricing the stocks by mistake; the tendency that investors enamor the particular section at different periods. A great amount of theoretical and empirical researches, which irrational investors’ errors lead to market misevaluation of assets, have been conducted. The empir ical studies conclude that irrational herding is one of the reasons of mispricing.

3.3 Economic Psychological Approach to Herd Behavior

The economists and psychologists share the same view on what herding refers to, however, the development of the concept indicates not only the theoretical distinctions with regard to underlying assumptions, and shows the major differences among methodologies as well (Rook, 2006). An integration of economical and psychological approach to herd behavior will help to bridge the gap between these differences. The example of groupthink, which is a special case of herding in decision-making groups, brings the theoretical work into practice (Janis, 1972).

(9)

One underlying assumption stated by economists and psychologists is that institutional and human being limitations force people to make the decisions which are satisfying themselves instead of solving the problems optimally. Human beings are described as “Cognitive miser” by Fiske and Taylor’s seminal work (1991). They suggest that people have tendency to behave rationally and consistently, however, they also love less effort more than much effort. This phenomenon leads to the result that people often prefer simple solutions rather than complicated solutions. One of the simple solutions is to depend on heuristics such as “the majority is always correct”, “the crowd is right” or “consensus is great.” Groupthink is the result of such heuristics. Initially, Janis gives the definition to the groupthink as “a mode of thinking that people engage in when they are deeply involved in a cohesive in-group, when the members’ strivings for unanimity override their motivation to realistically appraise alternative courses of action” (1972, 9). After years of developing, the phenomenon is considered as “an excessive form of concurrence seeking within groups, eventually leading to poor decisions and /or fiascos” (Rook, 2006, 85). The inherence of groupthink demonstrates that disadvantageous decision making in groups sometimes is the result of approving solutions, as individuals pay less attention to scrutinize the arguments.

Another powerful underlying assumption related to herd behavior in economic perspective is the existence of complete information. Theoretically, individuals can always figure out rational and optimal solutions with complete information. Therefore, individuals are anticipated to make their decisions in independent and rational ways. Complete information, however, is rare in real economic world. Abrahamson and Rosenkorf (1993) demonstrate that herd behavior, is merely a result of incomplete and ambiguous information, which creates uncertain situations that can be mitigated by observing the behaviors of others for uniformity. Early study of Festinger (1950) also shows that in difficult situations, people tend to rely on the behaviors and decisions of reference groups since they wish to share and confirm their own actions and decisions w ith other peer individuals. However, the tendency for uniformity and consensus with peer individuals can lead to a quite detrimental or dangerous consequence without being based on actual information.

Summarizing, the integration of economic and psychological approach to herd behavior helps to elaborate the reasons why people sometimes follow the actions of crowd. The limitations of human being make less effort is more appealing to much effort for individuals, although people have tendency to act in rational and consistent ways. Thus, people seek for simple solutions more often than complex solutions. In this case, depending on heuristics and believing in the major ity is always a good choice for individuals. As a result, people behave similar ly or even the same with the actions of the crowd. However, studies blame that this phenomenon may produce detrimental outcomes or lead to disasters. Additionally, incomplete information also plays a role in causing herd behavior. The uncertainty, which is caused by incomplete and ambiguous information, could be reduced by looking at the actions of others for consensus. People are intended to seek for similar ity and consensus with others when they have to confront with difficult situations, thus they expect to share and confirm their ideas and beliefs with peer individuals. Hence, there is higher possibility that people follow each other in difficult situations because consensus makes them comfortable. In financial market, investors mimic each other for consensus because it decreases the uncertainty and meets their need to perceive confidence and comfortable ( Vaughan and Hogg, 2005).

(10)

Section II 4. Survey and Results

4.1 Background, Assumptions and Hypothesis

According to the observation of the index graph of those two trading days that circuit breaker was triggered, it is apparent that the index dropped significantly in astonishing speed. One of the potential reasons why the index plunged is that the majority of individual investors, who drive more than 80 percent of the trading, rushed to sell their holdings and cash out in the same period of time. The similar behavior patterns among investors dried up the liquidity of the market, and as a result, the circuit breaker kicked in when the market lost 7 percent of its value. It seems that the 15-minute trading halt does not help to restore investors’ calm and prevent the market from further dropping. Investors are likely to mimic each other’s behavior and follow each other ’s decision under the pressure of circuit breaker. The large group of individual investors follows the same action at the same period of time; hence, it is reasonable that this phenomenon can be identified as herd behavior.

Many columns, TV programs and even the public accuse the circuit breaker as the culprit of this stock market drama. They complain that the implement of circuit breaker stimulates the panic and irrational behavior of individual investors. However, this argument is not valid until it is tested. Thus, a survey will be conducted in this paper to figure out whether circuit breaker encourages the irrational herd behavior of the individual investors in Chinese stock exchange. Moreover, the survey also tries to explain the reason why circuit breaker may or may not encourage the investors’ irrational herd behavior through the view of economic psychology. This can help people get deeper understanding of circuit breaker.

The survey is based on two underlying assumptions, which have already been mentioned in the previous part. One is the terminology, which is called “cognitive miser”. This assumes that when investors are facing a more complicated situation and they feel it hard to deal with, they are likely to rely on heuristics such as “the majority is always right. Another assumption is that the

information investors obtain from the market or other channels is incomplete. The incomplete information may produce uncertainty in the market, and the uncertainty could be reduced by looking at or following the majority’s behavior.

For this test, the null hypothesis is that circuit breaker does not encourage the irrational herd behavior of the individual investors in Chinese stock exchange. In other word, under the null hypothesis, the investors’ behavior should be indifferent between pre-implement and

post-implement of the circuit breaker mechanism. Hence, this implies that the impact of circuit breaker which leads investors to act irrationally is expected.

4.2 Methodology

An anonymous survey is conducted among a group of 60 randomly selected Chinese individual investors. The sample investors, who have different educational background, cover from rooster investors to professional and experienced investors. The sample selection aims to overcome the

(11)

selection bias. The questions motivate the sample investors to compare their own perceptions and behaviors before and after the implement of circuit breaker in the Chinese stock market. The answers are assigned the score from 1 to 5, indicating from “firmly disagree” to “firmly agree”. After collecting the answers from sample investors, answer “agree” and “firmly agree” represent the positive difference, and a sign test with the significance level 5% will be performed to test whether there is a difference between pre-implement and post-implement of the circuit breaker.

4.3 Sample Description

A group of 60 randomly selected investors who have confronted with the circuit breaker

mechanism participate in this survey. Panel 1 and Panel 2 illustrate the distribution of educational background and investment experience of sample investors, respectively.

Panel 1 illustrates that the majority of sample individual investors are highly educated. This indicates that they are likely to have more working knowledge in investing in a secondary market instead of “gambling” in the market.

Panel 2 demonstrates that a large amount of selected sample investors have much experience in investing in Chinese stock market.

4.4 Data and Results

Respondents of this questionnaire are initially asked that via which channel they get information about the circuit breaker. After that, sample investors are requested to provide their perceptions after the new mechanism is introduced to the market by answering question 7 and question 13.

0 0 11 38 11 0 5 10 15 20 25 30 35 40 Primary School or Eelow Junior High School Senior High School

Bachelor Master or Above

Panel 1: Educational Background

12 6 4 3 35 0 5 10 15 20 25 30 35 40

1 Year or Below 1 to 3 Years 3 to 5 years 5 to 7 years 7 Years or Above

(12)

The aim of these two questions is to prove that after injecting the new mechanism, the stock market becomes more complicated to the investors; sample investors face more uncertainty.

Panel 3 illustrates from which channel the investors get the information about circuit breaker. Most investors obtain the information from the Internet (45) and TV program (30). Only several respondents check the concept of this mechanism in reference books or journals.

Panel 4 and Panel 5 illustrate that 63.37% and 68.33% of respondents feel that they face a more complex situation after the implement of circuit breaker, respectively. The results of sign test also prove that most investors agree that the market becomes more complicated to them.

Question 8, question 9, question 14 and question 15 make attempts to discover the interac tion between sample investors and their reference group. Question 8 and question 14 tests generally whether selected investors are more aware of other investors’ behavior or decision after

implementing the circuit breaker; whereas question 9 and question asks investors to provide their

45 30 23 17 17 6 1 0 10 20 30 40 50 Internet TV program Bullet from CSRC Newspaper From Other Investors Reference Others

Panel 3: Informational Channel

4 2 16 28 10 0 10 20 30 Firmly Disagree

Disagree Neutral Agree Firmly Aagree

Panel 4: Question 7

2 3 14 30 11 0 10 20 30 40 Firmly Disagree

Panel 5: Question 13

Sign test Observed

Positive 38 Negative 6 Neutral 16 Test statistics: z 4.824 z_α(α = 5%) 1.645 Table 1

Sign test Observed

(13)

experience under specific premise: during the 15-minute trading halt on both trading days.

With the result provided here, it indicates that respondents are more aware of peer investors’ decisions or behaviors than usual in the market with circuit breaker.

The results suggest that sample investors, during 2 times 15-minute trading halt, have more

1 8 18 26 7 0 10 20 30 Firmly Disagree

Panel 6: Question 8

1 10 18 25 6 0 10 20 30 Firmly Disagree

Panel 7: Question 14

1 11 16 25 7 0 10 20 30 Firmly Disagree

Panel 8: Question 9

2 11 19 24 4 0 10 20 30 Firmly Disagree

Panel 9: Question 15

Sign test Observed

Positive 33 Negative 9 Neutral 18 Test statistics: z 3.703 zα(α = 5%) 1.645 Table 3

Sign test Observed

(14)

incentive than usual to communicate with other investors. More intensive interactions occur between respondents and their reference group when the circuit breaker kicks in.

In question10 and question16, sample investors are examined whether they feel more comfortable after looking at or following the actions of their peer investors. This question is based on the underlying assumption developed by Festinger (1950), which argues that following the reference group for uniformity can mitigate the anxiety brought by the uncertainty and difficult situation.

The evidence shows that looking at or following other investors’ actions is not a useful tool to make sample investors feel more comfortable. Two-thirds of sample investors disagree that they could feel better when following the majority for consensus.

In order to assess directly whether Chinese individual investors have stronger incentive to follow fellow investors’ action when they are under the pressure of circuit breaker, question 11 and 17 are designed to examine the samples. The respondents are requested to compare their pre-and post mechanism-implement perceptions and behaviors.

8 11 21 16 4 0 5 10 15 20 25 Firmly Disagree

Panel 10: Question 10

4 12 23 18 3 0 5 10 15 20 25 Firmly Disagree

Panel 11: Question 16

Sign test Observed

(15)

Panel 12 and Panel 13 describe the results of question11 and question 17, respectively. Table 9 and Table 10 present the observations and sign test results for both questions. 35% of respondents agree that they are motivated to follow other investors’ behavior when the circuit breaker is triggered for the first time; whereas, 36.67% of sample investors vote to the choice that they rely on the decisions and actions of peer investors on the day that circuit breaker is triggered for the second time. Obviously, nearly the same amount of respondents disagree that they are willing to follow the crowd after implementing the circuit breaker. Almost one-third of the respondents for each question suggest that circuit breaker has no effect on their decisions and they act as usual. The results of sign test demonstrate that both null hypothesis cannot be rejected, which means that there is no signif icant differences between pre-and post mechanism-implement behaviors and decisions of sample individual investors. The results of this survey imply that the circuit breaker does not encourage more severe irrational herd behavior.

4.5 Discussion

The result of this study turns out surprisingly to reverse the newspaper headlines, the column opinions, the publics’ arguments and the prejudice to circuit breaker. There is no sufficient evidence to reject the null hypothesis; in other words, circuit breaker does not encourage individual investors’ irrational herd behavior according to this survey. Although circuit breaker fails to smooth out the extreme volatility and protect the Chinese stock market, it yet cannot be accused to be the culprit of this stock market drama. The evidence demonstrates that the implement of circuit breaker does bring uncertainty to the stock market which forces individual investors to deal with a more complex situation (See table 1, 2). Individual investors’ response is as the same as in the early study of Festinger (1950). They tend to have more interactions, and share and confirm their decisions and actions within the reference group when they find

5 16 18 17 4 0 5 10 15 20 Firmly Disagree

Panel 12: Question 11

6 17 15 18 4 0 5 10 15 20 Firmly Disagree

Disagree Neutral Agree Firmly Agree

Panel 13: Question 17

Sign test Observed

Positive 22 Negative 23 Neutral 15 Test statistics: z -0.149 z_α(α = 5%) 1.645 Table 10

(16)

themselves in a more difficult situation (See table 3, 4, 5 and 6). Having had more interactions, it does not finally lead to the consequence that individual investors just follow each other blindly. The result of the research disapproves that individual investors simply seek for simple solutions instead of complicated solutions and believe in the heuristics such as “the majority is always correct” when they face the difficult situation (See Table 9, 10).

Except for answering the survey questions, several respondents share their own experience and opinions in narrative at the end of survey. Some of the respondents point out that no matter what the majority’s opinion suggests, they will ins ist their own decision. They claim that they would rather believe in themselves instead of the majority, although sometimes they inevitably get involved into the self-overconfidence bias. The reason why they act independently for the most of time refers to a psychological factor called “regret aversion”. As long as they make their own decisions independently, they will not feel regret to the outcomes no matter it is good or bad. This type of investors, according to their stories, insists in acting independently on the trading days that circuit breaker is triggered. They also admit that they have more intensive interactions and

frequent communications with other investors; however they emphasize that regret aversion plays an important role and they still are independent in the process of decision making. Therefore, the implement of circuit breaker has no influence to this type of investors.

Another issue which is worth being noticed is that there is no sufficient evidence to prove that looking at or following other investors’ action would make sample investors feel more

comfortable. Several investors mention that they do follow other investors’ behavior and fire sell their shares at a lower price but it does not make them feel better. They claim that after following others’ action, they still feel confused about what happens at that moment and what would happen in the future. Some investors argue that their reference group even does not have deep

understanding to this mechanism and cannot provide a satisfying solution. This argument is proved by question 6 in the survey. As the matter of fact, most of people get rough idea of circuit breaker from the Internet and TV program. Only 6 investors (10% of the sample) read the

reference literature to get more complete knowledge about this totally new mechanism (See Panel 3). What the majority of investors obtain from the social media is s ometimes ambiguous and incomplete. When circuit breaker is really triggered, the reference group may also have no idea on how to confront with the situation. Individual investors cannot get satisfying solution from the reference group and feel comfortable. Therefore, following the crowd blindly is worse than useless to better-off individual investors’ utility. This is the main reason why irrational herd behavior does not appear among individual investors: they do have more communications and observations among each other when circuit breaker is triggered; however they find it useless to mitigate the uncertainty and tackle with the upcoming situation.

This paper is not a proof that no irrational herd behavior exists in Chinese stock market. Tan, Chiang, Mason and Nelling (2008) perform a test and conclude that there is herd behavior within Chinese stock market, which is dominated by domestic market participants. With the result of the survey among a group of 60 individual investors in China stated in this paper, there is sufficient evidence to prove that the implement of circuit breaker does not stimulate individual investors to act in a more “crazy” way. Still, approximately one-third of investors in the sample are more

(17)

motivated to follow the crowd after the introduction of circuit breaker (See Panel 12, 13).

Moreover, the sample selection also has some biases in this research. As illustrated in Panel 1 and Panel 2, the majority of the samples are highly-educated and experienced investors. It can be speculated that they have deeper understanding to the function of market and more working knowledge on how to tackle with the uncertainty than others. They might use their own experience to make the optimal decisions rather than simply to chase the majority to get satisfying solutions, which implies that they might act more rationally than under-educated investors or “rooster” investors. It is recognized that this paper fail to include low-educated or “gambler-like” investors, which is also a large group of investors in Chinese stock market, into the research due to the limited accessibility to these investors. Whether circuit breaker triggers their irrational herd behavior is not clear in this research. Therefore, this paper is not sufficient to draw the definitive conclusion whether a more severe irrational herd behavior is stimulated by circuit breaker among the whole population of Chinese market participants.

The result of the survey also provides some implications to policy makers. Policy makers should rethink the failure of circuit breaker. It turns out that the irrational herding among Chinese individual investors may not be the largest problem leading to the market dysfunction. Instead of simply accusing the irrationality of individual investors, policy makers should concentrate more on the institutional investors and technical issues. Unlike professional investors, Chinese individual investors can hardly hedge their position by futures or options. Moreover, individual investors in Chinese stock market are forbidden to take the short position or have OTC

(over-the-courter) trading because of the trading regulations. However, the institutional investors may see and seize the opportunity of arbitrage and take the short position in the immature market with extreme volatility. Thus, individual investors become the victims, and slaughtered by institutional investors when the index drops greatly. The hostile short-selling by some institutional investors may also be one of the potential reasons of this stock market drama. Since Chinese stock market has already introduced the price-limit mechanism for decades, the implement of circuit breaker may provide the “clear target” for hostile short position to attack. This may accelerate the plunge of index and lead to market dysfunction. Although this technical issue is not studied in this paper, it may warn the policy makers to pay more attention to the maturity and trading regulations of the market. Otherwise, the mechanism, which is intended to protect the market from volatility, may finally become the “weapon” of some hostile investors.

Section III 5. Conclusion

Circuit breaker is a debatable mechanism in the financial market. The empirical studies fail to provide a definitive answer whether circuit breaker can protect the market effectively. No sufficient evidence due to lacking extreme volatility can finally draw a convincing conclusion. Betero and Mayer (1990) test 23 major indexes after “1987 Crash” and find that the markets under the protection of circuit breaker perform better than those without such mechanism. The statement which trading halt can restore investors’ calm and provide investors more time to collect

information and response to the complex situation, is widely accepted (Greenwald and Stein, 1988; Kodres and O’Brien 1994). However, subrahmanyan (1994) constructs a theoretical model called “magnetic effect” and disapproves the previous arguments. He emphasizes that triggering

(18)

threshold of circuit breaker may attract fearful investors to submit their orders earlier in order to avoid being stuck in the sudden pause of trading.

The early introduction of circuit breaker to Chinese stock market is a failure. The social media and plenty of investors blame that the circuit breaker encourages investors ’ irrational behavior, produces more volatility and makes the situation worse. Truly, the graphs reveal that the index experiences sudden and sharp drop after the circuit breaker is triggered. The 15-minute trading pause fails to provide any liquidity to the market and finally the index reaches an abrupt stop. Irrational herd behavior, to be exactly “group fire sale”, can be considered as a potential reason to cause market dysfunction.

The history of herding in economics and psychology obviously demonstrates that various ar eas started to study the phenomenon in early years. Le Bon (1895) defines herding as “collective hypnosis”, which is a sort of social contagion. Sherif (1935) later proposes the concept of reference group, which is the standard that people are willing to meet when facing uncertainty. Moscovici (1980) explains the confrontation between majority and minority. After that, Graham (1999) categorizes the literature about herding into four categories. In the realm of finance, Devenow and Welch (1996) point out that herding is universal. Hirschleifer (2003) designs several experiments that find the evidence to prove the previous statement. Thanks to Janis’s work (1972), economical and psychological approaches to herd behavior are integrated. One of the underlying assumptions is “cognitive miser”, which is discussed in Fiske and Taylor ’s paper (1991). Another refers to incomplete information. Abrahamson and Rosenkorf (1993) and Festinger (1950) argue that the uncertainty produced by incomplete information would be mitigated by looking at or following others for consensus.

The survey aims to provide evidence to prove whether circuit breaker triggers more severe irrational herd behavior. A group of 60 randomly selected investors participate in this survey. It turns out that although the implement of circuit breaker brings great uncertainty and difficult situation to the stock market, there is no sufficient evidence to prove that this will directly lead to more severe irrational herd behavior. More intensive interactions and frequent communications rise among the investors. However, individual investors are not able to dispel concern by merely looking at or following other investors’ behavior. Incomplete information plays a vital role in this case. Many sample investors claim that the majority cannot provide satisfying solution due to superficial understanding of the new mechanism. Following crowd cannot provide much utility and make investors comfortable. Still, there is a type of investors who for the most of time insist their own beliefs. They make the decisions independently in order to averse regret afterwards.

However, this paper cannot draw a definitive conclusion if circuit breaker stimulates more severe irrational herd behavior among the whole population of individual investors in China because of the sample selection bias. The limited accessibility to low-educated, inexperienced and

“gamble-like” investors makes the analysis omit the influence from this group of investors. Therefore, the result of the survey cannot reflect the whole population of Chinese individual investors. In short, the panic and irrational behavior caused by the implement of circuit breaker may not play the most important role in this stock market dysfunctional drama. The hostile attack from short position or the cash-out selling from institutional investors such as hedge funds, which

(19)

are not discussed in this paper, may be the real culprit which make great “contribution” to the index sharp plunging. Policy makers, therefore, should put more effort to regulate the market and design a better mechanism to protect the market and minimize the frequency of market

dysfunction.

6. Reference

Abrahamson, E and Rosenkopf, L., “Institutional and Competitive Bandwagons: using mathematical modeling as a tool to explore innovation diffusion,” The Academy of

Management Review, 1993, v18, 487-517

Banerjee, A. V., “A Simple Model of Herd Behavior.” The Quarterly Journal of Economics*, 1992 v107, 797-817

Bertero, E., and C. Mayer., “Structure and performance: Global interdependence of stock markets around the crash of October 1987” European Economic Review, 1990, v44, 1155-1180

Bloomberg, 4.1.2016, “China halts stock trading after 7% rout triggers circuit breaker”, http://www.bloomberg.com/news/articles/2016-01-04/chinese-stocks-in-hong-kong-extend-annual-slump-as-yuan-declines

Bloomberg, 7.1.2016, “China’s 29 minutes of chaos: stunned brokers and a race to sell”, http://www.bloomberg.com/news/articles/2016-01-07/china-s-29-minutes-of-chaos-stunned-brokers-and-a-race-to-sell

China Securities Regulatory Commission. Bulletin, December 4, 2015

Christie, W. G. and Huang, R. D., “Following the pied paper: do individual returns herd from market?”, Financial Analysts Journal, 1995, v51, 31-37

Devenow, A and Welch, I., “Rational herding in financial economics,” European Economic

Review, 1996, v40, 603-615

Festinger, L., “Informal social communication,” Psychological Review, 1950, v57, 271-282

Fiske, S. T. and Taylor, S. E., “Social cognition,” 1991, New York, McGraw-Hill

Graham, J., “Herding among investment newsletters: theory and evidence,” Journal of Finance, 1999, v54, 237-268

Greenwald, B., and J. Stein., “The task force report: the reasoning behind the recommendations, ”

Journal of Economic Perspectives, 1988, v2, 3-23

Greenwald, B., and J. Stein., “Transactional risk, market crashes, and the role of circuit breaker,”

(20)

Harr is, L., “Circuit breakers and program trading limits: what have we learned?”

Brookings-Wharton Papers on Financial Service*, 1998

Hirschleifer, D. and Siew, H, T., “Herd behavior and cascading in capital markets: a review and synthesis,” European Financial Management, 2003, v9, 25-66

Janis, I. L., “Victims of Groupthink,” 1972, Boston, Houghton Mifflin.

Kodres, Laura E and Daniel P. O’Br ien., “The Existence of Pareto-Superior Price Limits,”

American Economic Review*, 1991, v11 (2), 213-238

Lauterbach, B. and Ben-Zion, U., “Stock market crashes and the performance of circuit breakers: Empir ical evidence,” Journal of Finance, 1993, v48 (5), 1909-1925

Le Bon, G., “Les lois psychologiques de l’evolution des peuples,” 1895, Translation: “The Crowd:

a study of the population mind,” 1947, London, Ernest Benn

Libenstein, H., “Bandwagon, snob and Veblen effects in the theory of consumers’ demand,”

Journal of Economics, 1950, v64, 183-207

New York Stock Exchange, Inc., office of Research and Planning, “The Rule 80A Index Arbitrage Tick Test,” Report to the U. S. Securities and Exchange Commission, May 31, 1991

Rook, L., “An economic psychological approach to herd behavior,” Journal of Economics Issues, 2006, v40, 75-95

Sherif, M., “A study of some social factors in perception,” Archives of Psychology, 1935, v27, 1-60

Simmel, H., “Fashion,” American Journal of Socialogy ,1957, v62, 541-558

Subrahmanyam, A., “Circuit breakers and market volatility: a theoretical perspective. ” Journal of

Finance, 1994, v49 (1), 237-254

Tan, L, Chiang, T, C, Mason, J, R and Nelling, E., “Herd behavior in Chinese stock market: an examination of A and B shares,” Pacific-Basin Finance Journal, 2008, v16, 61-77

Van Ginneken, J. “Crowds, psychology, and politics.” 1992, New York: Cambridge Press

(21)

Appendix

I. Questionnaire: Survey Concerning Circuit Breaker and Herd Behavior in Chinese Stock Exchange on January 4th and January 7th, 2016

THIS QUESTIONNAIRE SHOULD TAKE NO MORE THAN TEN MINUT ES OF YOU TIME

Instructions. On January 4, 2016 and January 7, 2016, Chinese stock exchange experienced two

dramatic days in the history. The circuit breaker in stock market was triggered twice and the market lost 7% and 7.21% of its value, respectively. On January 8, 2016, China Securities Regulatory Commission (CSRC) announced that this mechanism would be suspended. Could you please help me to try to figure out what occurred by introducing me about your own experience and behavior then? Please give your answers by circling numbers. You can also elaborate with remarks on any question. I want your stories since they will help me to analyze your answers. Please be candid; this survey is anonymous. Please finish this survey as much as you can. Please notice: this is a non-profit survey which is for the study on how circuit breaker influences the behavior of individual investors.

Part 1: General Information

1. Did you experience the first-time trigger of circuit breaker on January 4, 2016? (CIRCLE ONE NUMBER)

YES NO

1 2

If you circle 2 (you did not experience the first trigger of circuit breaker) you do not need to finish neither Part 2A.

2. Did you experience the second-time trigger of circuit breaker on January 7, 2016? (CIRCLE ONE NUMBER)

YES NO 1 2

If you circle 2 (you did not experience the second trigger of circuit breaker) you do not need to finish Part 2B.

Please notice: If you circle 2 for both questions above (you never experienced the trigger of circuit breaker), you have completed this questionnaire. Please return the questionnaire.

3. What is your age? ( )

4. How many years have you invested in Chinese stock market (till the trigger of circuit breaker)? 1 year or below 1 to 3 years 3 to 5 years 5 to 7 years 7 years or above

1 2 3 4 5

5. What is your educational background?

Primary school or below Junior high school Senior high school Bachelor Master or above 1 2 3 4 5

(22)

6. . Through what channel did you get information about the circuit breaker? (Multiple choices) Newspaper Internet TV program Bulletin from CSRC Reference Other investors 1 2 3 4 5 6

Others:

Part 2: Herd Behavior

Part 2A: First-time Trigger of Circuit Breaker (January 4, 2016)

7. You felt that you were facing a more uncertain and difficult situation than usual after the introduction of circuit breaker to the stock market.

(CIRCLE ONE NUMBER)

Firmly Disagree Disagree Neutral Agree Firmly Agree 1 2 3 4 5

8. You were, than usual, more aware of what your peer individual investors ’ behaviors or decisions after the introduction of circuit breaker to the stock market.

(CIRCLE ONE NUMBER)

9. During the 15-minute trading halt, you had more incentives than usual to communicate with your peer individual investors to discuss about what you felt or what your further decision was(for example: continued to hold the shares after re-opening the market or fire sell the shares)?

(CIRCLE ONE NUMBER)

10. You felt that you were facing less uncertainty and feeling more comfortable after looking at the actions of your peer individual investors for consensus or following their actions.

(CIRCLE ONE NUMBER)

11. Comparing the trading days without circuit breaker (stork index also plunges), you were more strongly or easily inf luenced by your peer individual investors; thought less about your own individual information and had more incentive to chase your peer group when the circuit breaker was triggered.

(CIRCLE ONE NUMBER)

12. Could you please leave some remarks that how your peer individual investors ’ actions influence your behavior when circuit breaker is triggered? (Optional)

(23)

Part 2B: Second-time Trigger of Circuit Breaker (January 7, 2016)

Please notice: if you have not experienced the first-time trigger of circuit breaker (January 4, 2016), then please skip the questions from 17 to 21:

13. You felt that you were facing a more uncertain and difficult situation than usual after the introduction of the circuit breaker to the stock market.

(CIRCLE ONE NUMBER)

14. You were, than usual, more aware of what your peer individual investors’ behaviors or decisions after the introduction of the circuit breaker to the stock market.

(CIRCLE ONE NUMBER)

15. During the 15-minute trading halt, you had more incentive than usual to communicate with your peer individual investors to discuss about what you felt or what your further decision was(for example: continued to hold the shares after re-opening the market or fire sell the shares)?

(CIRCLE ONE NUMBER)

16. You felt that you were facing less uncertainty and feeling more comfortable after looking at the actions of your peer individual investors for consensus or following their actions.

(CIRCLE ONE NUMBER)

17. Comparing the trading days without circuit breaker (stork index also plunges), you were more strongly or easily inf luenced by your peer individual investors; thought less about your own individual information and had more incentive to chase your peer group when the circuit breaker was triggered.

(CIRCLE ONE NUMBER)

18. Could you please leave some remarks that how your peer individual investors’ actions influence your behavior when circuit breaker is triggered? (Optional)

End of survey.