Trading Performance, Disposition Ef

Trading Performance, Disposition Effect,
Overconfidence, Representativeness Bias,
and Experience of Emerging Market Investors
GONGMENG CHEN1, KENNETH A. KIM2*, JOHN R. NOFSINGER3
and OLIVER M. RUI4
1School of Economics andManagement, Shanghai Jiaotong University, Hong Kong
2School ofManagement, State University of NewYork at Buffalo, and PACAP, USA
3College of Business,Washington State University, USA
4Faculty of Business Administration, Chinese University of Hong Kong, Hong Kong
ABSTRACT
Using brokerage account data from China, we study investment decision making in an
emerging market. We find that Chinese investors make poor trading decisions: the
stocks they purchase underperform those they sell. We also find that Chinese investors
suffer from three behavioral biases: (i) they tend to sell stocks that have appreciated in
price, but not those that have depreciated in price, consistent with a disposition effect,
acknowledging gains but not losses; (ii) they seem overconfident; and (iii) they appear
to believe that past returns are indicative of future returns (a representativeness bias). In
comparisons to prior findings, Chinese investors seem more overconfident than U.S.
investors (i.e., the Chinese hold fewer stocks, yet trade very often) and their disposition
effect appears stronger. Finally, we categorize Chinese investors based on proxy
measures of experience and find that ‘‘experienced’’ investors are not always less
prone to behavioral biases than are ‘‘inexperienced’’ ones. Copyright # 2007 John
Wiley & Sons, Ltd.
key words disposition effect; investor behavior; overconfidence; representativeness
bias
This study considers the extent to which Chinese investors make poor trading decisions and exhibit three
particular behavioral biases: (a) the disposition effect, (b) overconfidence, and (c) the representativeness bias.
These cognitive errors are forms of heuristic simplification, which stem from the brain’s tendency to make
mental shortcuts rather than engaging in longer analytical processing. The tendency for people to suffer from
these biases is well-established in the psychology and behavioral science literature (e.g., Grether, 1980;
Joseph, Larrick, Steele, & Nisbett, 1996; Kahneman & Tversky, 1973; Lichtenstein, Fischhoff, & Phillips,
1982; Ritov, 1996; Tversky & Kahneman, 1974; Yates, 1990). Recently, the economics and finance literature
Journal of Behavioral Decision Making
J. Behav. Dec. Making, 20: 425–451 (2007)
Published online 9 February 2007 in Wiley InterScience
(www.interscience.wiley.com) DOI: 10.1002/bdm.561
* Correspondence to: Kenneth A. Kim, School of Management, Jacobs Management Center, SUNY–Buffalo, Buffalo, NY, 14260, USA.
E-mail: [e-mail]
Copyright # 2007 John Wiley & Sons, Ltd.
also has found that U.S. investors suffer from these same behavioral biases (e.g., Barber & Odean, 2000,
2001; Odean, 1998a). In this study, we examine Chinese investors because Chinese culture is very different
from U.S. culture, and culture can breed overconfidence at varying levels (e.g., Yates, Zhu, Ronis, Wang,
Shinotsuka, & Toda, 1989). Some evidence exists that suggests that people raised in Asian cultures exhibit
more behavioral biases than people from the United States (e.g., Yates, Lee, & Bush, 1997).
We also consider the degree to which trading performance and behavioral biases are related to five specific
investor characteristics: investor’s investing experience, age, trading frequency, personal wealth, and the
location in which he or she resides. We specifically consider the possibility that experienced, middle-aged,
active, and wealthy investors living in cosmopolitan cities are less prone to suffering from behavioral biases.
To conduct our analyses, we analyze the trades of 46,969 individual investor brokerage accounts from a
brokerage firm in China. Before we present our results, we discuss the behavioral biases examined in the
paper, we highlight the cultural differences between China and Western countries, and then we describe the
investor characteristics that we consider.
INVESTOR BEHAVIOR
Investors may be inclined toward various types of behavioral biases, which lead them to make cognitive
errors. People may make predictable, nonoptimal choices when faced with difficult and uncertain decisions
because of heuristic simplification. We test for the existence of trading errors and consider three behavioral
biases, which are described as follows.
Overconfidence
One outcome of heuristic simplification (i.e., self-deception) occurs when people tend to think that they are
better than they really are (Trivers, 1991). The psychology and behavioral science literature characterize
people that behave as if they have more ability than they actually possess as being overconfident (e.g.,
Campbell, Goodie, & Foster, 2004; Lichtenstein et al., 1982; Yates, 1990). Investors who attribute past
success to their skill and past failure to bad luck are likely to be overconfident. An investor who is
overconfident will want to utilize his perceived superior ability to obtain large returns. Thus, overconfident
investors trade often and underestimate the associated risks of active stock investing (e.g., Kyle & Wang,
1997; Odean, 1998b). When analyzing investor behavior using their stock brokerage data, trading frequency
is commonly used as a proxy for overconfidence. Supporting these conjectures, Barber&Odean (2000, 2001)
and Odean (1999) find that U.S. individual investors trade excessively, expose themselves to a high level of
risk, and make poor ex post investing decisions (i.e., their investment portfolio returns are inferior to a
benchmark market-wide portfolio return). Of course, trading frequency is a noisy measure of overconfidence.
Investors with superior information and trading skills will utilize this ability by trading often to capture high
returns. So people with actual high ability and people who wrongly believe they have high ability will both
trade excessively. It is generally assumed that there are few truly high ability investors compared to the
number of overconfident ones. Therefore, the trading frequency proxy is often believed to represent the
behavior of overconfident investors on average.
Disposition effect
Consider the situation in which an investor wishes to sell a stock. The investor can sell a stock that has
increased in price or one that has decreased in price. People avoid actions that create regret and seek actions
Copyright # 2007 John Wiley & Sons, Ltd. Journal of Behavioral Decision Making, 20, 425–451 (2007)
DOI: 10.1002/bdm
426 Journal of Behavioral Decision Making
that cause pride. Nofsinger (2005) illustrates that selling the ‘‘winner’’ (the stock that has increased in price)
validates a good decision to purchase that stock in the first place and stimulates pride. Selling the ‘‘loser’’ (the
stock with a loss) causes the realization that the original decision to purchase it was poor, and thus stimulates
regret. Shefrin& Statman (1985) state that the propensity to avoid regret and seek pride causes investors to be
predisposed (thus the behavioral effect’s namesake) to selling winners too early and riding losers too long.
This disposition effect is one implication of extending Kahneman & Tversky’s (1979) prospect theory to
investment decision making. In prospect theory, the value function is concave in the area of gains and convex
in the area of losses, implying risk aversion in the area of gains and risk seeking in the area of losses. An
investment decision-making application of mental accounting is the process in which the mind keeps track of
gains and losses on each stock held rather than at the portfolio level (Thaler, 1980). The disposition effect
combines the prospect value function with the focus on individual stock gains and losses to predict that
investors are relatively more willing to sell the winners than the losers in their portfolios. There may be times
when selling the winner is the optimal wealth-maximizing decision. But this is generally not the case when
stock price changes move in trends (called return momentum) and when capital gains are taxed. Interestingly,
Shefrin and Statman also find that the disposition effect reverses in December when individual investors’
attention is more focused on minimizing capital gains taxes by engaging in tax-loss selling.
Consistent with the disposition effect, the economics and finance literature finds that investors often sell
stocks that have performed well so that they can feel good about themselves. At the same time, investors tend
not to sell their poorly performing stocks because they are not ready to acknowledge that they made a mistake
and because they are afraid that the stocks may recover (i.e., they wish to avoid regret; Shefrin & Statman,
1985). Odean (1998a) finds that U.S. individual investors are more willing to sell stocks that have done well
than those stocks that have done poorly. Grinblatt & Han (2005) derive an equilibrium model that explains
how prospect theory and mental accounting (i.e., the disposition effect) creates a momentum return pattern.
Frazzini (2006) empirically tests the model and concludes that when investors display the disposition effect,
it induces a stock price underreaction to news announcements and a post-announcement price drift.
Representativeness bias
One mental shortcut, the representativeness bias, involves overreliance on stereotypes (Shefrin, 2005).
Representativeness leads people to form probability judgments that systematically violate Bayes’s rule (see
Grether, 1980; Kahneman & Tversky, 1973; Tversky & Kahneman, 1974). The representativeness bias has
several implications to investment decision making. Investors may misattribute good characteristics of a
company (e.g., quality products, capable managers, high expected growth) as characteristics of a good
investment. This stereotype would induce a cognitive error as Lakonishok, Shleifer, & V

Hasil (Inggris) 3:

交易业绩，处置效应，
过度自信，代表性偏差，新兴市场投资者
公1
和经验，Kenneth A. kim2 *，约翰R. nofsinger3
和奥利弗M.锐4
经济管理学院，上海交大管理学院，香港
，布法罗，纽约州立大学和PACAP，美国
3业务，华盛顿州立大学，美国
4faculty工商管理香港大学，中国，香港
摘要
经纪帐户数据来自中国，我们研究了在新兴市场的投资决策
。我们发现，中国的投资者做出错误的买卖决定：他们购买的股票表现不佳的
他们卖。我们还发现，中国投资者有三个行为偏差：（我）他们倾向于卖出，在
价格升值的股票，而不是那些有折旧价，与处置效应相一致，
承认收益，而不是损失；（ii）他们似乎过于自信；和（iii）他们出现
相信过去的收益是指示未来的收益（一个代表性偏差）。在比较之前的调查结果，中国的投资者似乎比美国
投资者过度自信（即，中国持有较少的股票，但经常交易）及其配置
效应似乎更强。最后，我们把中国投资者基于代理
措施的经验和发现的'experienced”投资者并不总是那么
容易出现行为偏差比'inexperienced '的'。版权2007约翰威利#
&儿子，公司
关键词处置效应；投资者行为；过度自信；代表性
偏置
本研究考虑的程度，中国的投资者做出错误的买卖决定，表现出三
特定行为偏差：（一）处分效果，（b）和（c）的过度自信，代表性
偏见。这些认知错误是一种启发式的简化形式，它来源于大脑的倾向，而不是从事较长的分析处理，而不是从事心理的捷径。倾向的人们遭受
这些偏见是建立在心理学和行为科学文献（例如，格雷特，1980；
约瑟夫，拉里克，斯梯尔，&Nisbett，1996；Kahneman&Tversky，1973；李奇登斯坦，&Fischhoff，菲利普斯，
Ritov，1982；1996；1974；Tversky&Kahneman，Yates，1990）。近年来，经济学和金融学文献
杂志行为决策行为
J.。月，20：425–451（2007）
发表9二月2007威利跨学科
（www.interscience。威利。COM）DOI: 10.1002／BDM。561
*通讯作者：Kenneth A.基姆，管理学院，雅可布管理中心，苏宁–水牛，水牛，NY，14260，美国
电子邮件：kk52 @水牛。edu
版权2007约翰威利#&儿子，公司
也发现美国投资者遭受这些行为偏差（例如，理发&Odean，2000，2001
；Odean，1998a）。在这项研究中，我们研究中国投资者，因为中国的文化是非常不同的，从美国文化，