th 2009 International Conference on Management Science & Engineering (16 ) September 14-16, 2009 Moscow, Russia
An Experimental Study on Investors’ Status Quo Bias
LI Jian-biao，REN Guang-qian，LIU Lv-ke Research Center for Corporate Governance / Business School, Nankai University, P.R.China, 300071
Abstract: Status quo bias is an important factor management area. Samuelson and Zeckhauser (1988) which affects investors’ decision-making. The existing defined status quo bias as the perpetrator tending to research on this issue only considers Prospects Theory. maintain the existing choices in the face of In fact, framing effect, investor sentiment and decision-making or choices [2]. Status quo bias is an information structure are also important variables which important factor which affects the decision-making of affect status quo bias. But the existing papers are less investors. The researches on status quo bias of investors involved to study status quo bias from the three aspects are in favor of grasping the investment rule of investors above. This paper researches the impact on investors and in favor of improving the effectiveness of status quo bias from the three aspects of framing effect、 decision-making of the relevant departments. Status quo investor sentiment and information structure in the bias of investors mainly comes from the obtaining phase, experimental method. The results show that: The investor the editing phase and the evaluation phase of information status quo bias in the margined frame of the investment in the investment decision-making process. The current portfolio is higher than that in the ratio frame of the studies focus on the evaluation phase which is observed investable portfolio. There is status quo bias of investors easily. And the studies research status quo bias from the in the three emotions conditions. In the positive emotions perspectives of prospect theory, regret evading and other investors status quo bias is lower and in the negative perspectives. emotions investors status quo bias is higher. The amount At present, the prospect theory is the most accepted of choose the determined probability is five times that theory in explaining status quo bias（Kahneman and choose uncertainty probability .The level of status quo Tversky, 1979）[3]. Prospect theory figures out that status bias in self-status assignment group is higher than that in quo bias is due to loss aversion of investors or their value external status assignment group. function is S type.1 In the decision making process, Keywords: framing effect, information structure, people endow more weight with the loss than the gain investor sentiment, status quo bias significantly. And the loss has greater impact on people's feeling than the contour gain. In their 1991 paper, 1 Introduction Tversky and Kahneman clearly states that the value [4] appraisement on the loss is twice of the gain . People It is urgent to improve the diathesis of investors in detest any form of loss and make the losses not recur as the mature stock market. Reducing status quo bias level far as possible. Prospect theory is an important theory in of investors is the important aspect to improve the behavioral finance research, but it also has limitations, diathesis of investors. It should be an important problem for example, it assumes that the effectiveness of people for how and from which perspectives to reduce status relies mainly on the given baseline based on the past data quo bias level of investors. Bias is a prior or and considers less uncertainty about the future. And pre-judgment. Bourne (1993) considered that bias is the people only care about gain and loss, neglecting the portfolio of feelings, attitudes and behavioral tendencies impact on the decision-making of perpetrators from and is a preconceived view [1]. The studies focused on the emotions, information etc. The researches from the essence of bias, the structure of bias and the elimination of bias in the areas of psychology and sociology. And in 1 The loss and the gain have different value functions. The the areas of economics and management, the studies are value function of the loss is convex and relatively steep; and mainly concentrated in the areas of the impact on the value function of the gain is concave and relatively flat. decision-making from bias, the source of bias and the This shows that individual feeling on the loss is strong than the exhibition of bias. Status quo bias of investors is a contour gain. Kahneman found that the ratio of effectiveness typical bias problem in the economics area and the brought to the conductors is about 2:1 which is from the loss and the contour gain. Therefore, in the decision making process, people have no-balanced judgments between the loss and the Supported by the National Natural Science Foundation of gain. The consideration of avoiding the loss is much more than China (70672029, 70532001), Project of key base for Studies the consideration of preference on the gain. Because of the fear of Social Science of Ministry of Education of China of the loss, people always ask too high prices when they sell (05JJD630023), Social Science 985 Project-“Management and out their goods. As the great fear of loss, the owner of the stock institution innovation of Corporate in China” item. tends to become risk appetite when the stock price fell.
- 1582 - 978-1-4244-3971-3/09/$25.00 ©2009 IEEE perspective of regret evading indicated that the change in they can increase the sensitivity of status quo bias from the status quo may lead to greater sense of regret than perpetrators. maintaining the status quo（Kahneman,2003[5]; Gilovich The current studies found the prevalence of status and Medvec,1994[6]; Bomnger et al, 1994[7] ） . The quo bias. But the studies on investors’ status quo bias in perpetrators are preferred to those decision-making the evaluation phase could only partly explain the source which can minimize their expected sense of regret and the degree of status quo bias. But the deep-seated （Kahneman and Miller, 1986[8]; Loomes and Sugden, reasons are existence in information obtained phase and information edited phase. In these two phases, framing 1982[9]; Loomes,1988[10]; Roese and Olson, 1995[11]）. effect, investor sentiment and information structure are Gleicher et al (1990) divided regret into two categories in three typical sources of bias. There are some studies on accordance with whether or not to take action, and one is framing effect at present. Moxey et al (2002) researched the regret which comes from some kind of negative the impact degree of framing effect on decision-making results with some actions; the second is the regret which of doctors [23]. Chinese scholars WANG Chong-ming and comes from some kind of negative results with no LIANG Li (1995) got the conclusion that the dynamic actions [12]. Bomnger et al (1994) considered that for the characteristics of framing effect were impacted by the same degree of negative results, the regret sense coming nature, the content and the scene of the mission [24]. from the former is more strongly and more durable than Papers on investor sentiment were mainly researched the latter [7]. In studies, Inman and Zeelenberg (2002)[13], combining with prospect theory at present. Schwarz and Ritov and Baron (1995)[14], as well as Johnson et al Clore (2003) found good sentiment had greater (1993)[15]got the results: perpetrators always pay likelihood of success in decision-making [25]. In the attention to those regret senses which come from the research of Fishbach and Trope (2005), they found failure of their missions, but don’t pay too much positive emotions could increase the wishes of the attention to the regret senses which come from the errors transaction and reduce the price margin between buyers as a result of negligence. In the study, Gilovich and and sellers, but the negative emotions had the opposite Medvec (1994) showed that the regret sense which came effect [26]. Studies on information structure: Ryan and from the results of changing the status quo was stronger Bate (2001) found perpetrators preferred the status quo than not changing the status quo; and perpetrators were mainly due to the relatively little information they had more compunctious on the negative outcomes which for changing their options [27]. DuPont and Lee (2002) came from their action in the short time, but in long found asymmetric information was the reason of period, they were more compunctious for the negative inconsistent of the WTA and WTP prices by empirical outcomes that they did not make the conduct2 [6]. Thaler study [28]. These studies mainly researched the existence and Shefrin (1981) believed that regret is a feeling that of these three sources and their impact factors from the perpetrators must be responsible for the loss besides the view of Psychology. But papers were also rare which loss, so, regret is more painful than loss [16]. In China, researched the deep-seated reasons of investors’ status LAI Zhi-gang and SHI Kan (2006) studied the quo bias in the phases of information obtaining and relationship between regret responses on behavior of editing in the decision-making of the investors subjects and risk preference [17]. SHI Jun-qi, WANG Lei combining the three aspects. and PENG Kai-ping (2004) researched the behavior of Existing researches had little regard to the effect of subjects in symbolic situation and benefit situation and framing effect, investor sentiment and information whether the regret degree was different in the two structure in investors’ status quo bias, but the three situations, and then they found subjects in symbolic factors played an important role in theoretical analysis situation preferred as a test and in benefit situation and practical application of status quo bias. Framing subjects tended to choose not to as [18]. There are some effect is one of the important factors which impact the other perspectives of researches on status quo bias in decision-making of investors. The current studies have evaluation phase, such as fuzzy offensive perspective confirmed the existence of framing effect and impact on (Tversky and Kahneman, 1974[19]; Gilboa and the decision-making of perpetrators. Investor sentiment Schmeidler, 1995[20]). Studies from this perspective has been one of the research emphases in behavioral indicated investors preferred certainty gain in relation to economics. Different sentiment can bring different uncertainty gain, and they preferred uncertainty loss to decision-making results and different level of status quo certainty loss; Researches on the perspective of impact bias for the investors. As one of necessary conditions for from experience （ Burmeister and Schade, 2007[21]; [22] decision-making of perpetrators, information has very Kahneman et al, 1991 ) thought perpetrators preferred important influence on the decision-making of the status quo because of the impact of past experiences, perpetrators, and under different information structure, and experiences had unilateral impact on status quo bias, the level of status quo bias is different. Based on these and with the purpose of finding the factors of reducing 2 Gilovich and Medvec collected 77 long-term regrets of the investors status quo bias, this paper researches investors subjects. In the 213 regrets(an average of 2.77 regret for status quo bias in experimental methods from the three everyone), the 63% is the regret for not actions and only 37% is perspectives of framing effect, investor sentiment and regret for the actions(10 regrets are excluded for they can not information structure and researches the level of these be classification as done or not done).
- 1583 - three factors impacting on investors status quo bias. The a higher level of status quo bias. In the experiment, the results provides decision-making basis for investors and questions are divided into two groups and each is policy makers. component with one decision-making question and each subject only completes one question. In order to ensure 2 Experimental design and experimental the experiment quality, subjects are informed there are process non-standard answers to all the questions and they can answer the question by individuals in accordance with 2.1 Experimental design their own tendency. In order to not cause any This paper designs framing effect and investors psychological pressure to the subjects, all the status quo bias experiment using different expressing experiments are anonymous. In order to be able to fashions of stock returns. Stock returns can be expressed express their true preferences completely, this by the margin of bid-ask price and can also be expressed experimental design makes subjects rank the four options by the return rate. But the same perpetrator may have and the option with strongest will is ranked first, latter is different feelings on them. So, this paper assumes that: if the second and the third, the fourth is weakest. Specific using margin to express the stock returns, perpetrator has experimental frameworks are showed in Tab.1.
Tab.1 Experimental category of investors’ status quo bias
General situation of subjects Conditions Environment Time Payoff Questions Number Source Experience Grouping In the Laboratory experiment, Experiment: Senior level The group we give 9：00, 13:30 Laboratory undergraduate expressed cash to and 15:30 in Experiment: and graduate by the subjects as Laboratory April 18, 60 subjects; students of difference reward. In Framing Experiment 2008; Field Field Nankai No of price laboratory See Fig.1 Effect and Field Experiment: Experiment: University and the experiment and Fig.2 Experiment April 15, 250 who enroll group every 2008—April subjects. voluntarily in expressed subject can 20, 2008. the by return get ￥10 experiment. rate. and in field experiment, subjects are no income. Positive Every Undergraduate sentiment subject can and graduate group, get ￥10 as Investor Laboratory 9:30, 14:00 60 subjects. students of No Negative participate See Fig.3 Sentiment Experiment and 16:00 in Nankai sentiment fee and has and Fig.4 April 18, University group and the chance 2008. who enroll Neutral to get voluntarily. sentiment Olympic group souvenirs. Subjects Freedom touch a Undergraduate choice pingpong and graduate group, and guess Information Laboratory 9:00, 13:30 students of External its color. If Structure Experiment and 15:30 in 60 subjects. Nankai No status the guess is See April 18, University assignment right, he Fig.5, 2008. who enroll group and will get Fig.6 and voluntarily. Self-status ￥3 . And Fig.7 assignment every group subject can get ￥10 as participation fee. Source: Finished by the author.
- 1584 -
Sentiment has very important influence on the Selten laboratory of Nankai University (the other two decision-making of investors. Positive sentiment can experiments are as the same). increase the wishes of transaction, reduce the price In the experimental process, first of all, the margin between buyers and sellers and reduce the level organizer reads and explains the experimental of status quo bias of perpetrators. But negative sentiment introduction. Secondly, give the subjects three minutes to has the opposite effect. In the experiment of investor understand the experimental introduction and answer the sentiment and status quo bias, we divide 60 subjects into questions that the subjects ask separately. And the three groups, positive sentiment group, negative principle to answer the questions is clarifying the sentiment group and neutral sentiment group. And in experimental process and not providing any new each group, there are 20 subjects and the 20 subjects are information, especially not discussing the experimental divided into two types of A and B. Specific experimental aims or the expected results so as to ensure the subjects frameworks are showed in Tab.1. only know the public information. Then, subjects fill out Status quo bias can reduce the ambiguity elusion the experimental questions as showed in Fig.1 and Fig.2. degree of perpetrators. When the motivation of Finally, at the end of the experiment, experiment ambiguity elusion is stronger, the level of status quo bias organizers take the record cards back and pay cash to the is higher. When the subjects have no initial option, they subjects as reward privately. Then, interview some prefer the game of identifying probability. The degree of subjects based on the experimental results and record the ambiguity elusion is different between subjects with interview results. Field experiments are carried in initial ambiguous option and subjects with no option. students’ dorms which are selected randomly and its Subjects with initial ambiguous option prefer to choose process is much the same as in the laboratory. this ambiguous option. The status quo bias level of (2) The experimental process of sentiment and subjects in self-status assignment group is higher than investors’ status quo bias experiment. This experiment is the level of subjects in External status assignment group. carried after the framing effect experiment. First of all, In this paper, we studies the status quo bias level of experimental presider announces that the subjects should perpetrators under different information structure watch a short film as an necessary rest in the condition referring to the experimental design of Ellsberg experimental process. After the end of the short film, experiments in which the authors measure the status quo experimental presider immediately announce that the bias level with the rate between the number of changing laboratory prepares 10 presents for the subjects as options and the number of maintaining the options. souvenirs for they taking part in the experiment. Then Specific experimental frameworks are showed in Tab.1. subjects play the finger-guessing game between every two. The winner subjects of A type can get a 2.2 Experimental process souvenir—Crystal Olympic Fuwa. And then let subjects (1) The experimental process of framing effect and pass to see the souvenir one to one in very short time. investors’ status quo bias experiment. This is manual Then experimental organizers distribute type A and type experiment and it includes laboratory experiment and B record cards to the subjects (the whole process isn’t field experiment. The laboratory experiment is carried in more than 120 seconds).
Decision 1: Assuming that you are an investor in the stock market, you use all your assets to buy stocks: When the price of stock A is￥18 per share, you buy 100 shares and now the price is￥21, your profit is￥300; when the price of stock B is￥17 per share, you buy 200 shares and now the price is￥15.5, your loss is￥300. How will you do the decision now? A. Sell stock A and hold stock B B. Hold all the stocks C. Sell stock B and hold stock A D. Sell all the stocks Please rank above decision-makingSupported optionsby the National in accordance Natural with Science the strength Foundation of your of wishesChina (from strong to weak).
Fig.1 Record card of the price differential expressing group
Decision 2: Assuming that you are an investor in the stock market, you use all your assets to buy stocks: When the price of stock A is￥18 per share, you buy 100 shares and now your profit rate is 16%; when the price of stock B is￥17 per share, you buy 200 shares and now your loss rate is 9%. How will you do the decision now? A. Sell stock A and hold stock B B. Hold all the stocks C. Sell stock B and hold stock A D. Sell all the stocks Please sort above decision-making options in accordance with the strength of your wishes (from strong to weak).
Fig.2 Record card of the return rate expressing group
- 1585 -
Serial number of subject: ______Welcome to take part in the experiment and congratulations on your souvenir! Now you have the opportunity to sell the souvenir to the experimental presider in a certain price, please fill out the price you are willing to accept on the following line. ______(￥).
Fig.3 Record card of type A subjects
Serial number of subject: ______Welcome to take part in the experiments! Now you have the opportunity to purchase the souvenir from the experimental presider which is displayed just now in a certain price, please fill in the price you are willing to pay on the following line.______(￥)
Fig.4 Record card of type B subjects
After subjects finish the record cards, experimental reads the experimental instruction after all subjects have organizers take them back, and then experimental arrived. After subjects understand the experimental presider announced the experiment is end. Finally process fully, the organizers distribute experimental experimental organizers extend the participation fee and record cards. The record card is showed in Fig.5. Then interview some subjects in accordance with the content let subjects choose the game based on bag A or bag B on the experimental record cards. Specific organizational and complete the record cards. And then subjects walk processes of the three experimental groups are as on the stage and take a ball from the box which is held follows: by the experimental assistants. At the same time, the ① Positive sentiment group. Play the short film organizers collect the experimental record cards back and
- 1586 -
Serial number of subject: ______Please make a tag in the parentheses in front of the options that you choose: Bags you choose: （ ）Bag A （ ）Bag B （ ） （ ） Colors you guess: Yellow White Fig.5 Record card of subjects from freedom choice group
Serial number of subject: ______Results of drawing lots: （ ）Bag A （ ）Bag B Colors you guess: （ ）Yellow （ ）White
Fig.6 Record card of subjects from external status assignment group
Serial number of subject: ______Please make a tag in the parentheses in front of the options that you choose: Bags you choose: （ ）Bag A （ ）Bag B Colors you guess: （ ）Yellow （ ）White
Fig.7 Record card of subjects from self-status assignment group opportunity to change the results of their initial choice. one is higher than the proportion of decision-making two Subjects can take balls from the boxes that they re-select significantly. Therefore, different ways to describe in accordance to their serial number after they mark the impact status quo bias significantly. The status quo bias changed options on the record cards (Subjects who degree of subjects on the profit and loss expressed by the change the options should mark on the blank space at the margin is higher than expressed by the rate. This shows top of the record cards). The organizers collect the framing effect has a significant impact on investors’ experimental record cards back and record the income of status quo bias. In this experiment, the expression of subjects. After all subjects have taken the balls, profit and loss by the margin in decision-making one has experimental presider announced this experiment is end. higher accessibility than the expression by the ratio in decision-making two. Therefore, subjects in 3 Analysis of experimental results decision-making one can know their profit and loss very directly. Without taking into account the time value of 3.1 Results analysis of framing effect and investors’ money, subjects are aware that they have neither loss nor status quo bias experiment profit, so they tend to maintain the status quo that is to (1) Descriptive statistical analysis. There are 107 continue to hold all stocks. Because the profit and loss of copies of decision-making questionnaire 1 which are subjects is expressed by rate in decision-making two, its collected back and there are 32 copies which choose accessibility is weak. So, subjects have difficult to option A, 51 copies which choose option B, 15 copies understand their situation of profit and loss intuitively. which choose option C and 9 copies which choose option And under the action of disposition effect, subjects tend D. There are 112 copies of decision-making to sell profit stocks to lock in profitable revenue and questionnaire 2 which are collected back and there are 44 continue to hold the loss stock which can be observed copies which choose option A, 35 copies which choose from the decision-making two (There are 44 copies option B, 22 copies which choose option C and 11 copies accounted for about 40% which rank option B —“Sell which choose option D. So subjects rank the option of stock A and hold stock B”—at the first place). hold all the stocks at the first place in the first experimental frame and subjects rank the option of sell 3.2 Results analysis of sentiment and investors’ status stock A and hold stock B at the first place in the second quo bias experiment experimental frame. In this part, we define option B for 1 (1) Descriptive statistical analysis. Descriptive and non-option B for 0, define decision-making group statistical results of WTA values and WTP values under one for 1 and decision-making group two for 2, then the three states of sentiment are showed in Tab.4 and analyze the proportion of option B in the total number Tab.5. The WTA / WTP value is about 1.64 in positive through crosstabs. Statistical results are showed in Tab.2. sentiment group, in negative group the value is about (2) Significant test analysis. We can see the pearson 2.35 and neutral group is about 2.15. So we can see that value is 6.182, accompanied probability is 0.013 and it is the level of investors’ status quo bias under positive less than significant level of 0.05 from Tab.3. So we sentiment is lowest and the level under negative group is think the proportion of option B in the decision-making highest.
- 1587 -
Tab.2 Descriptive analysis results of crosstabs of framing effect and investors’ status quo bias experiment Option B Decision-making types Items Total 0 1 Amount 56 51 107 Expectation Amount 65.0 42.0 107.0 1 % within decision-making types 52.3% 47.7% 100.0% % within option B 42.1% 59.3% 48.9% % of Total 25.6% 23.3% 48.9% Residual -9.0 9.0 Std. Residual -1.1 1.4 Adjusted Residual -2.5 2.5 Amount 77 35 112 Expectation Amount 68.0 44.0 112.0 2 % within decision-making types 68.8% 31.3% 100.0% % of Total 35.2% 16.0% 51.1% Residual 9.0 -9.0 Std. Residual 1.1 -1.4 Adjusted Residual 2.5 -2.5 Amount 133 86 219 Total Expectation Amount 133.0 86.0 219.0 % within decision-making types 60.7% 39.3% 100.0% % of Total 60.7% 39.3% 100.0% Source: The data collected by authors based on the statistical output of SPSS 16.0.
Tab.3 Chi-square statistics results of framing effect and investors’ status quo bias experiment
Value df Asymp. Sig. Exact Sig. Exact Sig. (2-sided) (2-sided) (1-sided) Pearson Chi-Square 6.182a 1 0.013 Continuity Correctionb 5.513 1 0.019 Likelihood Ratio 6.210 1 0.013 Fisher's Exact Test 0.018 0.009 Linear-by-Linear Association 6.153 1 0.013 N of Valid Cases 219 Source: The data collected by authors based on the statistical output of SPSS 16.0. a. 0 cells (.0%) have expected count less than 5. The minimum expected count is 42.02. b. Computed only for a 2x2 table.
(2) Significant test analysis. Tab.6 gives the variable. From Fig.8 we can see there are some points variance homogeneity test results of WTA prices and which deviate from the normal distributing slash WTP prices which are separated in the three experiments seriously in the normal probability graph of WTA price of sentiments. We can see the significant level of four variable and the prices which subjects are willing to indicators are all less than 0.05. So we think there is accept are scattered. From Fig.9 we can see most points significant difference between the two prices. The WTA are close to normal distributing slash in the normal prices are higher than WTP prices significantly and WTA probability graph of WTP price variable and the prices / WTP value is about 2.05. which subjects are willing to pay are concentrated. So we In order to study the fluctuation range of WTA can get the conclusion that WTA prices are higher than prices and WTP prices, Fig.8 and Fig.9 give normal WTP prices significantly and the fluctuation range of probability graphs of WTA price variable and WTP price WTA prices is larger than WTP prices.
- 1588 -
Tab.4 Descriptive statistics results of WTA price under the three states of sentiment 95% Confidence Interval of the Mean Types Amount Mean (￥) Min Max Lower boundary Upper boundary Positive 10 8.85 5.23 12.47 1 15 Negative 10 8.00 2.11 13.89 3 30 Neutral 10 8.50 4.15 12.85 2 30 Total 30 8.45 6.36 11.87 1 30 Source: Authors finish based on experimental results. Tab.5 Descriptive statistics results of WTP price under the three states of sentiment 95% Confidence Interval of the Mean Types Amount Mean (￥) Min Max Lower boundary Upper boundary Positive 10 5.40 3.29 7.51 1 10 Negative 10 3.40 1.16 5.64 0 10 Neutral 10 3.95 2.20 5.00 0 6 Total 30 4.25 3.09 5.17 0 10 Source: Authors finish based on experimental results. Tab.6 Variance homogeneity test results of WTA and WTP Levene Statistic df1 df2 Sig. Based on Mean 9.120 1 58 0.004 Based on Median 7.797 1 58 0.007 Based on Median and with adjusted df 7.797 1 37.532 0.008 Based on trimmed mean 8.163 1 58 0.006 Source: Authors finish based on experimental results.
Normal Q-Q Plot of 价格 Expected Normal for ?型= WTA 2
1
0
-1
-2 -10 0 10 20 30 Observed Value Fig.8 Normal probability graph of WTA price variable Source: Authors finish based on the statistical output of SPSS 16.0. Normal Q-Q Plot of 价格 Expected Normal for ?型= WTP 2
1
0
-1
-2 -2 0 2 4 6 8 10 Observed Value Fig.9 Normal probability graph of WTP price variable Source: Authors finish based on the statistical output of SPSS 16.0.
- 1589 -
3.3 Results analysis of information structure and self-status assignment group and this proves the above investors’ status quo bias experiment assumption again. All of the 4 subjects who choose (1) Descriptive statistical analysis. Freedom choice vague option initially retain the vague option of bag B, group, External status assignment group and Self-status but there are 2 subjects who change their options from assignment group are on behalf of three groups of certain A to vague B of 16 subjects who choose certain information structure. Experimental results of the three options of bag A initially. groups are showed in Tab.7. From Tab.7 we can see there Comparing the experimental results we can find the are only 5 subjects who choose vague option of bag B in rate of maintaining the status quo is 75% in external freedom choice group which has 20 subjects. This status assignment group and the rate is 90% in. So the indicates that majority of subjects have the tendency to status quo bias level of subjects from external status elude vague options when they face the vague options. assignment group is lower than self-status assignment And the rate to choose the options of which the group. The experimental results are consistent to probability is certain is five times as the options of which previous assumptions. That let subjects of investors the probability is vague. In the experiment of external choose options freely can strengthen their sense of status assignment group, the ratio of reserving options is ownership and then improve the status quo bias of three times as changing options and there are 9 subjects subjects. This paper analyzes whether the rate of retain the vague bag in ten subjects who have been choosing vague options is different significantly between endowed with vague option. These prove the previous subjects of freedom choice group and subjects of assumptions. external status assignment group using crosstabs analysis. The place which is worth to explore in this We define option A as 0 and B as 1. The results are experiment is that there are 4 subjects who change their showed in Tab.8. From Tab.8 we can see the chi-square options to the vague option B of the 10 subjects who are value is 6.465 and accompanied a probability is 0.011 endowed the option A. We think there are two main which is less than significant level of 0.05. So we can get reasons: one is that subjects don’t trust the experimental the conclusion that the degree of eluding vague options is organizers; the other is the study between subjects. In the different significantly between freedom choice group and experiment, the subject of No.7 earns ￥3 after he external status assignment group. And we think the status change his option from A to B and this may lead other quo bias can reduce the degree of eluding vague. These subjects to follow. The interviews after the experiment confirmed the existence of investors’ status quo bias and confirm this conjecture. There are only 4 subjects who its impact on the decision-making of perpetrators from choose the vague option of bag B of 20 subjects in some profiles.
Tab.7 Statistical results of investors’ status quo bias experiments under different information structure External status assignment group Self-status assignment group Serial number Freedom choice of subject group Endowed option Final option Given option Final option 1 A A A A A 2 A A B B A 3 A A A B A 4 A A B B B 5 B B A A A 6 A A B B A 7 A A A B A 8 A A B B A 9 A A A A B 10 B B B B B 11 B B A A A 12 A A B B A 13 A A A B A 14 A A B A A 15 B B A A A 16 B B B B A 17 A A A B A 18 A A B B B 19 A A A A B 20 A A B B B Source: Authors finish based on experimental results.
- 1590 -
Tab.8 Chi-square statistical results of freedom choice group and external status assignment group experiments Value df Asymp. Sig. Exact Sig. Exact Sig. (2-sided) (2-sided) (1-sided) Pearson Chi-Square 6.465 1 0.011 Continuity Correction 4.949 1 0.026 Likelihood Ratio 6.660 1 0.010 Fisher's Exact Test 0.025 0.012 N of Valid Cases 40
Source: Authors finish based on the statistical output of SPSS 16.0.
Tab.9 The fitting effect of iterative results for neutral sentiment-WTA-external status assignment group Predicted Observed y Percentage Correct 0 1 0 6 1 85.7 y 1 2 1 33.3 Overall Percentage 70.0 Source: Authors finish based on the statistical output of SPSS 16.0.
Tab.10 Regression results of neutral sentiment group-WTA-external status assignment group
95.0% C.I.for EXP(B) Coefficients S.E. Wald df Sig. Lower Upper
X1 (Gender) -1.097 2.140 0.263 1 0.608 0.005 22.126
X2 (Education) 0.945 2.089 0.205 1 0.651 0.043 154.319
X3 (Asking Price) -0.173 0.317 1.738 1 0.390 0.409 1.418 Constants 0.978 2.081 0.221 1 0.638
Source: Authors finish based on the statistical output of SPSS 16.0.
(2) Regression analysis of experimental variables bias degree of female is higher than male. The coefficient and the status quo bias. Based on the above two of education is positive and this shows that the higher of experiments, this part analyze the relationship between education, the lower of the status quo bias degree. gender, education, WTA, WTP and the status quo bias Tab.11 and Tab.12 show the results of neutral using logistic regression comprehensively. Because the sentiment-WTP- external status assignment group. information structure experiment and the investor Tab.11 shows the fitting effect of iterative results. The sentiment are completed by the same group of subjects, accuracy is 87.5% for y=0, 50% for y=1 and 80% for all so this part will cut the neutral sentiment-external status cases. Tab.12 shows the coefficients of all variables and assignment group to study. The definition of gender the corresponding Wald statistical values and their variables: 1 for male and 0 for female; The definition of accompanied probability values. The coefficient of WTP educational variables: 1 for undergraduate and 0 for is negative and this shows that the higher of the bid, the master; The definition of dependent variables (the status easier to change the status quo. The coefficient of gender quo bias variables): 0 for maintaining the option and 1 is also negative and this shows that the status quo bias for changing the option. Statistical results are showed in degree of female is higher than male. The coefficient of Tab.9. Tab.9 shows the accuracy is 85.7% for y=0, education is positive and this shows that the higher of 33.3% for y=1 and 70% for all cases. education, the lower of the status quo bias degree. Tab.10 shows the coefficients of all variables and From the regression results of neutral the corresponding Wald statistical values and their sentiment-WTA-external status assignment group and accompanied probability values. The table shows the neutral sentiment-WTP-external status assignment group, Wald statistical value of X3 is largest and its we can draw the following conclusions: the higher of the accompanied probability is smallest, so the variable is bid, the easier to change the status quo; the lower of very important in the model. The coefficient of WTA is asking price, the easier to change the status quo; the negative and this shows that the lower of asking price, degree of women's status quo bias is higher than men and the easier to change the status quo. The coefficient of the higher of education, the lower of the status quo bias WTA is also negative and this shows that the status quo level.
- 1591 -
Tab.11 The fitting effect of iterative results for neutral sentiment-WTP-external status assignment group Predicted y Observed Percentage Correct 0 1 0 7 1 87.5 y 1 1 1 50.0 Overall Percentage 80.0 Source: Authors finish based on the statistical output of SPSS 16.0.
Tab.12 Regression results of neutral sentiment-WTP-external status assignment group Coefficients S.E. Wald df Sig.
X1 (Gender) -1.112 1.103 0.275 1 0.612
X2 (Education) -0.748 2.014 1.305 1 0.271
X3 (Bid) 0.105 1.007 0.836 1 0.394 Constants 2.018 2.341 0.337 1 0.326 Source: Authors finish based on the statistical output of SPSS 16.0.
4 Conclusion subjects under negative sentiment have lower transaction willingness and higher status quo bias level. In the This paper reviews the research literatures on experiment of information structure and investors’ status investors’ status quo bias in decision-making information quo bias, we measure investors’ status quo bias level evaluation phase from the perspective of prospect theory, respectively using 2×3 experimental setting. And the regret evading and other perspectives. Then this paper results show that: in the basic experiment, subjects have points the status quo biases in decision-making the tendency to elude vague options and the probability information obtaining phase and editing phase are more of certain options they choose is five times as the deep-seated problems. Framing effect, investor sentiment uncertain options. The status quo bias level of subjects in and information structure are three typical problems in external status assignment group is lower than in decision-making information obtaining phase and editing self-status assignment group. There are 75% subjects of phase. And they have important impact on investors’ investors who choose to maintain the status quo in status quo bias. This paper designs three experiments external status assignment group and 90% in in which include framing effect, investor sentiment and self-status assignment group. information structure impacting on investors’ status quo bias using the principles and methods of behavioral References economics and experimental economics and analyzes the impact that the three factors cause on investors’ status [1]G.R.Bourne. Proximate costs and benefits of mate quo bias in different experimental settings. acquisition at leks of the frog ololygon rubra[J]. Animal In the experiment of framing effect and investors’ Behaviour, 1993, 45(6):1051-1059. status quo bias, we focus on the investors’ status quo bias [2]W. Samuelson, R. Zeckhauser. Status quo bias in level under the frames of profit and loss expressed by decision making[J]. Journal of Risk and Uncertainty, price gaps and return rates. And the results show that: the 1988, 1(1): 7-59. investors’ status quo bias level in the frame of profit and [3]D. Kahneaman, A. Tversky. Prospect theory: An loss expressed by price gaps is significantly higher than analysis of decision making under risk[J]. in the return rate frame. In the experiment of investor Econometrica, 1979, 47(3): 63-91. sentiment and status quo bias, we induce positive [4]A. Tversky, D. Kahneaman. Loss aversion in riskless sentiment and negative sentiment using the images of choice: A reference-dependent model[J]. The Quarterly stories and we investigate their status quo bias level. At Journal of Economics, 1991, 106(4): 1039-1061. the same time the experimental group with no sentiment [5]D. Kahneaman. Maps of bounded rationality: induced is made as the base for compare. And the results psychology for behavioral economics[J]. American show that: the value of WTA/WTP under positive Economic Review, 2003, 93(5): 1449-1475. sentiment is the smallest as 1.64 and the value under [6]T. Gilovich, V. H. Medvec. The temporal pattern to negative sentiment is 2.35. These show that there is the experience of regret[J]. Journal of Personality and difference for WTA and WTP under positive sentiment Social Psychology, 1994, 67(33): 357-365. and negative sentiment for the subjects of investors. The [7]D. S. Bomnger, F. Gleicher, A. Strathman. subjects under positive sentiment have higher transaction Counterfactual thinking: From what might have been to willingness and lower status quo bias level and the what may be[J]. Journal of Personality and Social
- 1592 -
Psychology, 1994, 67(22): 297-307. Psychological Science, 2004, 27(4): 1016-1018. (in [8]D. Kahneman, D. T. Miller. Norm theory: Comparing Chinese) reality to its alternatives[J]. Psychology Review, 1986, n [19]A. Tversky, D. Kahneman. Judgment under 93(2):136-153. uncertainty: Heuristics and biases[J]. Science, 1974, [9]G. Loomes, R. Sugden. Regret theory: An alternative 185(9): 1124-1131. theory of rational choice under uncertainty[J]. The [20]I. Gilboa, D. Schmeidler. Case-based decision Economic Journal, 1982, 92(368): 805-824. theory[J]. The Quarterly Journal of Economics, 1995, [10]G. Loomes. Further evidence of the impact of regret 110(3):5-39. and disappointment in choice under uncertainty[J]. [21]K. Burmeister, C. Schade. Are entrepreneurs’ Economica, 1988, 55(217):47-62. decisions more biased? An experimental investigation of [11]N. J. Roese, J. M. Olson. Outcome controllability the susceptibility to status quo bias[J]. Journal of and counterfactual thinking[J]. Personality and Social Business Venturing, 2007, 22 (3): 340-362. Psychology Bulletin, 1995, 21(6): 620-628. [22]D. Kahneman, J. L. Knetsch, R. Thaler. The [12]F. Gleicher, K. A. Kost, S. M. Baker, A. J. Strathman, endowment effect, loss aversion and status quo bias[J]. S. A. Richman, S. J. Sherman. The role of counterfactual Journal of Economic Perspectives, 1991, 5(1):193-206. thinking in judgments of affect[J]. Personality and [23]A. Moxey, D. O. Connell, P. McGettigan, D. Henry. Social Psychology Bulletin, 1990, 16(2):284-295. Describing treatment effects to patients--How they are [13]J. J. Inman, M. Zeelenberg. Regret in repeat expressed makes a difference?[J]. Journal of General purchase versus switching decisions: The attenuating role Internal Medicine, 2003, 18(11): 948-959. of decision justifiability[J]. Journal of Consumer [24]WANF Chong-ming, LIANG Li. A study on Research, 2002, 29(6): 116-128. dynamic framing effects in risky decision[J]. Acta [14]I. Ritov, J. Baron. Outcome knowledge, regret, and Psychologica Science, 1998, 30(4): 394-400. (in omission bias[J]. Organizational Behavior and Human Chinese) Decision Processes, 1995, 64(2):119-127. [25]N. Schwarz, G. L. Clore. Mood as information: 20 [15]E. J. Johnson, J. Hershey, J. Meszaros, H. years later[J]. Psychological Inquiry, 2003, 14(3): Kunreuther. Framing, probability distortions, and 296-303. insurance decisions[J]. Journal of Risk and Uncertainty, [26]A. Fishbach, Y. Trope. The substitutability of 1993, 7(1): 35-51. external control and self-control[J]. Journal of [16]R. H. Thaler, H. M. Shefrin. An economic theory of Experimental Social Psychology, 2005, 41(3): 256-270. self-control[J]. The Journal of Political Economy, 1981, [27]M. Ryan, A. Bate. Testing the assumptions of 89(2): 392-406. rationality, continuity and symmetry when applying [17]LAI Zhi-gang, SHI Kan. A study on the relationships discrete choice experiments in health care[J]. Applied among regret orientation, regret response and risk Economics Letters, 2001, 8(1): 59-63. preference[J]. Management Review, 2007, 19(3): 3-7. (in [28]D. Y. DuPont, G. S. Lee. The endowment effect, Chinese) status quo bias and loss aversion: Rational alternative [18]SHI Jun-qi, WANG Lei, PENG Kai-ping. Symbolic explanation[J]. The Journal of Risk and Uncertainty, and economic factors in action effect: An 2002, 25(1): 87-101. economic-psychological analysis of regret theory[J].
- 1593 -