Are individual stock investors overconfident? Evidence from an emerging market

Are individual stock investors overconfident? Evidence from an emerging market

Journal of Behavioral and Experimental Finance 5 (2015) 35–45 Contents lists available at ScienceDirect Journal of Behavioral and Experimental Finan...

464KB Sizes 0 Downloads 52 Views

Journal of Behavioral and Experimental Finance 5 (2015) 35–45

Contents lists available at ScienceDirect

Journal of Behavioral and Experimental Finance journal homepage: www.elsevier.com/locate/jbef

Full length article

Are individual stock investors overconfident? Evidence from an emerging market Bülent Tekçe ∗ , Neslihan Yılmaz 1 Bogazici University, Department of Management, 34342, Bebek, Istanbul, Turkey

article

info

Article history: Received 8 December 2014 Received in revised form 4 February 2015 Accepted 7 February 2015 Available online 18 February 2015 Keywords: Behavioral biases Overconfidence Stock market Emerging markets Turkey

abstract This paper investigates overconfidence among individual stock investors. We focus on Turkey in order to use a unique nationwide dataset and study how common overconfidence is, what factors affect overconfidence and how overconfidence relates to investor return performance. Our findings show that overconfident behavior is common among individual stock investors. Male, younger investors, investors with a lower portfolio value, and investors in low income and low education regions exhibit more overconfident behavior. Moreover, we find that overconfidence has a negative effect on portfolio wealth. To the best of our knowledge, our study is one of the few studies in the literature and the first in Turkey focusing on nationwide data to analyze overconfidence. We extend the findings of the behavioral finance literature to the Turkish market based on this dataset. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Empirical evidence in the behavioral finance literature shows that individuals do not behave rationally. Barberis and Thaler (2003) provide a summary of models that try to explain the equity premium puzzle, excess volatility, excessive trading, and stock return predictability using both the Prospect Theory of Kahneman and Tversky (1979) and beliefs. Daniel et al. (2002) support the view that markets are not efficient and investor biases affect security prices substantially. There are numerous studies2 showing that investors are not rational or markets may not be efficient and hence prices may significantly deviate from fundamental values due to the existence of irrational investors. One of the most important deviations from rationality is overconfidence which has a significant impact on markets.



Corresponding author. Tel.: +90 5303990198. E-mail addresses: [email protected] (B. Tekçe), [email protected] (N. Yılmaz). 1 Tel.: +90 5335661488. 2 Black (1986), De Long et al. (1990), Shleifer and Vishny (1997), Barberis et al. (2001), Hirshleifer (2001), Daniel et al. (2002), and Subrahmanyam (2007) to name a few. http://dx.doi.org/10.1016/j.jbef.2015.02.003 2214-6350/© 2015 Elsevier B.V. All rights reserved.

It affects level of trading volume as well as price formation in the stock markets. Overconfidence results in aggressive trading behavior which may lead investors to pay a significant amount of commissions. In addition, overconfident investors may hold riskier portfolios than they should tolerate due to their underestimation of risk. Overconfidence not only affects financial markets and prices, but also individuals as they may make suboptimal investments resulting in deterioration of their wealth. Hence, it is important to examine the prevalence of overconfidence in addition to its determinants and consequences. Unfortunately, due to data availability, most of the research in the behavioral finance literature depends on the data that is restricted to the subsamples of overall investor groups in the countries of focus which limits us from generalizing the findings of these studies. To the best of our knowledge this is one of the few studies to focus on nationwide data to analyze overconfidence. Studies specific to Turkey are also limited, Korkmaz and Çelik (2007) to name one. Overconfidence may depend on many factors including individual characteristics. For example, Vissing-Jorgensen (2004) uses the investor optimism survey data conducted by UBS and Gallup from 1998 to 2002 and finds that irrational behavior is weaker for more sophisticated investors.

36

B. Tekçe, N. Yılmaz / Journal of Behavioral and Experimental Finance 5 (2015) 35–45

The author uses wealth and investor experience as proxies for investor sophistication. Moreover, overconfidence may depend on cultural differences. Fan and Xiao (2005) and Statman (2010) show that individuals in different societies and cultures may have different behavioral biases which may affect their financial decisions. For example, individualism is a more evident trait in western countries, such as the USA and the UK, as collectivism is in eastern countries like China and India. Hofstede (2001) finds that Turkish people are more collectivist compared to the USA, the UK and Western Europe. Fan and Xiao (2005) and Statman (2010) argue that individuals in collectivist societies tend to be more risk tolerant, therefore, we may expect collectivism to have an effect on overconfident behavior, as well. In this study, our dataset allows us to look at the effect of different demographic factors, namely age, gender, portfolio wealth, experience and region of residence on overconfidence. Besides, the collectivistic nature of the Turkish culture will help us examine how culture affects overconfidence. The majority of the behavioral finance literature analyzes individual investors in developed markets such as the USA, the UK and Western Europe. However, Turkey as an emerging market, and the Istanbul Stock Exchange (ISE) in Turkey, have important characteristics that would appear to be worth analyzing.3 ISE is a member of the World Federation of Exchanges (WFE) and the Federation of European Securities Exchanges (FESE). As a leading/advanced emerging market stock exchange, ISE is recognized as an investable market according to the US Securities and the Exchange Commission (SEC) and the Japan Financial Services Agency. Moreover, ISE has one of the highest turnover ratios among world stock markets, which may be related to the overconfidence among Turkish stock investors. According to the World Bank, in 2011, ISE ranked fifth in terms of turnover ratio after Italy, the Republic of Korea, China and the USA. Trading volume in ISE is relatively high and provides a liquid market for investors. Although foreign investors hold around 65% of free float in ISE, they constitute only around 15% of the trading volume.4 Local individual investors trade more aggressively and hence, trading volume and liquidity is mostly provided by these investors. Our database helps us focus on Turkey and the Turkish stock market in order to have a better understanding of overconfidence in an emerging market. We use transaction data for the year 2011 in order to analyze how prevalent overconfident is among investors, what factors affect overconfidence, and how it relates investor return performance. First, we find that overconfidence is common among Turkish individual investors. Overconfidence is higher among males, young investors, investors with low portfolio values and investors in less developed regions. Next, we analyze the trade performance of

3 Istanbul Stock Exchange became Borsa Istanbul in April 2012. 4 Foreign investors have a different investment style. They mostly prefer ISE30 and ISE100 (a major benchmark) stocks, which have high market capitalization, high liquidity and are representative of the sectors they operate in. Moreover, Ülkü and İkizlerli (2012) argue that foreign investors may be more sophisticated and maybe rationally adjusting their investment styles.

investors in order to understand whether overconfidence impacts return performance. Return calculations are based on matched sell transactions with one or more buy transactions within 2011. This methodology allows us to calculate realized returns taking into account the stocks chosen for sale. Our findings show that overconfidence is detrimental to portfolio wealth since mean return decreases with increasing turnover. Investment decisions of overconfident investors do not seem to be justified. The findings of this paper contribute to the behavioral finance literature in a number of ways. First, we use a unique nationwide dataset unlike most of the studies in the literature which are limited to a subset of investors. We also focus on an emerging market with high collectivist attitudes as opposed to the widely studied developed countries which are markets with individualistic attitudes in order to have a better understanding of overconfidence of investors in such countries. 2. Literature review and hypothesis development Overconfidence is the unmerited confidence in one’s self judgments and abilities. Odean (1998) describes overconfidence as the belief that a trader’s information is more precise than it actually is. This is equivalent to narrow confidence intervals in predictions. Daniel et al. (1998) define an overconfident investor as one who overestimates the precision of his private information signal, but not of information signals publicly received by all. Overconfidence may stem from different reasons. Miller and Ross (1975) and Kunda (1987) argue that self-attribution bias means attributing successful outcomes to one’s own skill but blaming unsuccessful outcomes on bad luck. Langer (1975) states that the illusion of control is the tendency for people to overestimate their ability to control events that they have no influence over. Unrealistic optimism is the confidence about the future or successful outcome of an event. It is the tendency to take a favorable or hopeful view as discussed by Weinstein (1980) and Kunda (1987). Russo and Shoemaker (1992) define confirmation bias as the tendency of people to favor information that confirms their arguments, expectations or beliefs. Svenson (1981) discusses that better than average effect implies that people think that they have abilities superior than an average person. Hence, individuals tend to believe that they are in the best class among peers. Calibration refers to how individuals assess the correctness of their estimates. Deaves et al. (2010) argue that a miscalibrated agent assumes that s/he has made a lower level of mistakes than is true. Different forms of overconfidence reveal that overconfident investors believe that their decisions will prove to be correct and they will expect higher than average returns. However, this is not necessarily the case. There are numerous examples in stock markets which prove that investors often make wrong decisions and deteriorate wealth. Barber and Odean (2000, 2001, 2002) argue that overconfidence can explain poor trading performance related to high trading levels and conclude that trading is hazardous to wealth. Hence, we hypothesize that overconfidence diminishes trading performance of Turkish individual stock investors. ISE has one of the highest turnover ratios among world stock markets. Since trading volume in ISE is mostly provided by local individual investors, we may argue that

B. Tekçe, N. Yılmaz / Journal of Behavioral and Experimental Finance 5 (2015) 35–45

overconfidence that is common among individual investors lead to increased trading activities in stock market. Findings in the psychology and finance literature also show that overconfidence is common among individuals; Fischhoff et al. (1977), Russo and Shoemaker (1992), Griffin and Tversky (1992), Kahneman and Riepe (1998) to name a few. Odean (1998) presents a good summary of overconfidence in different professional fields such as investment bankers and managers. The author also finds that overconfidence increases expected trading volume. Thus, we hypothesize that overconfidence is common among Turkish individual equity investors. Women tend to be more risk averse compared to men. This stylized fact is valid in different domains including financial investment decisions. Hence, it can be argued that men may exhibit higher degree of overconfidence compared to women. To test this hypothesis, Barber and Odean (2001) use data from a nationwide brokerage house between 1991 and 1996 by focusing on common stock investments of households. They find that men trade 45% more than women. Findings of Barber and Odean (1999), Chen et al. (2007), Acker and Duck (2008), Graham et al. (2009), Grinblatt and Keloharju (2009), and Hoffmann et al. (2010) also support the view that men are more overconfident than women, In line with the findings, we hypothesize that Turkish male investors are more overconfident than female investors. Fan and Xiao (2005) and Statman (2010) show that individuals in different societies and cultures may have different behavioral biases which may affect their financial decisions. Hofstede (2001) shows that Turkish people are more collectivist, which is also a more evident characteristic in eastern societies such as China, compared to the USA, the UK and Western Europe.5 Individuals in collectivist societies emphasize the group more than the individual. Hence people who belong to a group feel more comfortable and become less risk averse in their behaviors and decisions. Knowing that the group that the individual belongs to will support and back the individual, he or she may take riskier decisions and exhibit overconfidence more than those who are not a part of a group. Although the study of Hofstede dates back to decades, since culture only changes very slowly, the results can be considered up to date. There is no consensus in the literature whether individualism or collectivism leads to overconfidence. Chui et al. (2010) link individualism with overconfidence and conclude that individuals in less individualistic cultures act less like overconfident by looking at trading volume. On the other hand, there are studies which show that individuals in collectivist cultures exhibit higher degree of overconfidence, are less risk averse and/or trade more. Chen et al. (2007) use transaction data of a large Chinese brokerage house to analyze overconfidence among Chinese investors. The authors find that individual investors in China trade more frequently than those in the USA. Acker and Duck (2008) use a stock market game and predictions of examination marks to measure overconfidence among Asian and British students. They find that Asian students

5 Based on data available on http://geert-hofstede.com/.

37

are more overconfident than British students. Weber and Hsee (1998) find that Chinese individuals are less risk averse compared to American individuals. Li et al. (2009) also summarizes a list of studies which find that Chinese are more risk seeking and/or more overconfident than Americans. Since individuals in collectivist societies feel more secure within a group, we think that they will be more inclined to take riskier decisions and exhibit higher degree of overconfidence. Hence, we hypothesize that Turkish individual stock investors are more overconfident than individual investors in the US. There is no consensus in the literature about the relationship between overconfidence and investor sophistication, in the form of investor wealth, experience or financial literacy. For example, Graham et al. (2009) find that wealthy and highly educated investors are more likely to perceive themselves as competent, implying overconfidence. However, sophisticated investors have more access to information, are more competent with objectively assessing the information, and usually have more experience in the stock market. Hence, they may exhibit less irrational behavior in their investment decisions. Ekholm and Pasternack (2007) find that investors with smaller portfolios are more overconfident compared to investors with larger portfolios as investors with larger portfolios are more experienced and wealthy. We think that sophistication may help individuals make rational investment decisions. Hence, we hypothesize that sophisticated investors are less prone to overconfidence. 3. Data and methodology 3.1. Data We use nationwide data to look at the investment activities of Turkish individual stock investors during 2011. The data is from the Central Registry Agency (CRA) which consists of all buy and sell transactions as well as monthly stock only and total portfolio positions (stocks, funds, private sector bonds and warrants). We also have investor information on the demographics, region of residence and account opening date. According to the CRA monthly statistics as of January 2011, the total number of Turkish individual stock investors was approximately 1 million. However, a significant portion of these investors are either dormant or have a very low stock portfolio value. When the dataset is limited to individual stock investors whose total stock portfolio in any month in 2011 is above 5000 TL (approximately USD 3000), the number of investors reduces to 432,085. Of these, 74,051 investors do not have any buy or sell transactions, in other words they are dormant, during 2011.6 In addition to the dormant investors, a portion of the remaining investors do not have either any buy or any sell transactions. These investors are also excluded from the dataset for further analysis. Hence, the dataset is limited to those investors who have at least 1 buy and 1 sell transaction, resulting in 305,546 investors (labeled as the expanded investor set). Table 1 shows that the total trading volume of the expanded investor set is 517.9 billion TL (buy and sell 6 75% of the dormant investors are 50 years old or older and have been in the stock market for a relatively longer period of time. 66% of the dormant investors opened their accounts before 2002. Female investors constitute 41% of the dormant investors and 18% of the active investors.

38

B. Tekçe, N. Yılmaz / Journal of Behavioral and Experimental Finance 5 (2015) 35–45 Table 1 ‘‘Transaction data for the expanded investor set and the analysis investor set’’. This table displays both expanded and analysis investor sets. The expanded investor set consists of buy and sell data of investors whose total stock portfolio in any month in 2011 is above 5000 TL with at least 1 buy and sell transaction. The analysis investor set consists of buy and sell data of investors whose total stock portfolio in any month in 2011 is above 5000 TL with at least 1 buy and sell transaction excluding abnormally high turnover investors (investors who shift their portfolios more than 100 times annually).

Number of investors Number of buys Total value of buys (m TL) Mean buy value Median buy value Standard deviation buy Number of sells Total value of sells (TL) Mean sell value Median sell value Standard deviation sell

Expanded investor set

Analysis investor set

305,546 92,560,232 519,154 5,609 790 21,587 92,238,569 516,590 5,601 770 20,679

244,146 31,563,471 148,199 4,695 990 19,516 31,433,535 145,822 4,639 865 16,891

amounts divided by two), constituting 76% of the total trading volume in ISE in 2011. It indicates that the sample has a significant influence on price formation in the stock market. Of the remaining trading volume, 15% is attributable to foreign investors and the rest is attributable to the low portfolio value individual Turkish stock investors (investors with a monthly average stock portfolio value lower than 5000 TL in 2011) and to local institutional investors. However, a portion of these investors have very high annual turnover values at levels of millions TL, some have as high as 10 billion.7 One possible explanation is that these high turnover investors have their portfolios managed by professional money managers or they act like day traders and scalpers: therefore we exclude these investors. In our analyses, we cap the turnover value at 100 in comparison with international benchmarks. Besides this, according to ISE and CRA, the average holding period of individual investors is around 30 days, implying an annual turnover of 12. The mean turnover is 11.2 when turnover is capped at 100. If we increase the cut off (up to 1000), the results do not change. Moreover, the main results do not change when we include high turnover investors in the analysis using natural logarithm of turnover. Hence, the turnover is capped at 100, decreasing the sample size to 244,146 investors (labeled as analysis investor set). The demographics of high turnover investors (investors with turnover higher than 100) are slightly different. Compared to the analysis investor set, the high turnover investors are more male than female (88% versus 83%), younger than older (investors up to 35 years old are 21% versus 27%), not as wealthy (investors with portfolio wealth up to 10,000 TL – approximately USD 6000 – are 41% versus 34%) and more experienced (account open date 2002 or earlier are 49% versus 36%).8 There is no difference in terms of the region of residence. However, as expected, investors with high turnover have significantly more buy and sell transactions than those in analysis dataset. 12%

7 Median turnover is 7 where as the mean turnover is 1.15 m and standard deviation is 278 m, implying high dispersion. 8 GDP per capita in Turkey was USD 10,469 in 2011.

of high turnover investors have 1000 or more buys and sells (versus 2% in the analysis investor set). Investors with 1.5 m TL (approximately USD 800 thsd) or higher total value of buys and sells constitute 20% of the high turnover investors (versus 6% in the analysis investor set). Table 1 shows that the total trading volume of the analysis investor set is 147.0 billion TL, which is 22% of the total trading volume in ISE in 2011. The investors made 31.6 million buy transactions that is worth 148.2 billion TL and 31.4 million sell transactions that is worth 145.8 billion TL. The average buy value is 4695 TL (median 990 TL), slightly higher than the average sell value of 4639 TL (median 865 TL). Table 2 presents the demographic breakdown of the 244,146 investors. Due to confidentiality reasons, CRA provided categorized data for age, experience and portfolio wealth. Age is the age of investors as of 2011. Portfolio wealth is the average of 12 end-of-month investment portfolios consisting of equity, funds, warrants and corporate bonds. Experience is the opening date of account of the investor (if more than one account is available, we consider the opening date of the oldest one). Region is the geographical region of residence of the investor registered in the CRA database. Male investors are 83% of the total investor base whereas the 30–55 age group accounts for 76% of all investors. Of the investors, 76% have a portfolio wealth of 50,000 TL or less and 90% of the investors opened their accounts before 2008 (three or more years of investment experience in stock exchange). Almost half of the investors (45%) reside in the Marmara region, mostly in Istanbul, which is the largest city in Turkey. Next is the Central Anatolia with 17%, mainly in Ankara, which is the 2nd largest city followed by the Aegean (15%) mainly in İzmir, the 3rd largest city. The Marmara region is the most developed and the Southeast Anatolia region is the least developed among the regions in terms of welfare, income and education. ISE100 index, which consists of the largest 100 companies, decreased from 67,608 at the beginning of the year to 51,267 at the year end. From this standpoint, the stock market performance in 2011 can be said to bearish. However, out of 253 trading days, ISE100 index had positive returns on 129 days and negative returns on 124 days. Hence,

B. Tekçe, N. Yılmaz / Journal of Behavioral and Experimental Finance 5 (2015) 35–45

39

Table 2 ‘‘Investor demographic data’’. This table displays demographics of 244,146 investors in the analysis investor set. Age is the age of investor as of 2011. Portfolio wealth is the average of 12 month end portfolios consisting of equity, funds, warrants and corporate bonds. Experience is the account open date of investor (if more than one account is available, the opening date of oldest account is taken into consideration). Region is the geographical region of residence of investor registered in CRA database. N/A indicates data not available. N

%

Cum %

Gender

1 2

Male Female

203,051 41,095

83% 17%

83% 100%

Age

1 2 3 4 5 6 7 8 9

25 < 25–29 30–34 35–39 40–44 45–49 50–54 55–59 60 ≥

3,412 19,600 42,013 46,719 40,633 31,278 23,759 16,273 20,459

1% 8% 17% 19% 17% 13% 10% 7% 8%

1% 9% 27% 46% 62% 75% 85% 92% 100%

100–1,000 1,000–5,000 5,000–10,000 10,000–20,000 20,000–50,000 50,000–100,000 100,000–250,000 250,000–500,000 500,000–1,000,000 1,000,000–5,000,000 5,000,000+

2,704 38,176 43,212 47,745 54,004 27,117 18,935 6,607 3,085 2,170 391

1% 16% 18% 20% 22% 11% 8% 3% 1% 1% 0%

1% 17% 34% 54% 76% 87% 95% 98% 99% 100% 100%

After 2008 2006–2008 2003–2005 2000–2002 Before 2000

24,877 53,744 76,671 43,434 45,420

10% 22% 31% 18% 19%

10% 32% 64% 81% 100%

816 36,321 17,749 40,657 6,919 109,948 24,095 7,641

0% 15% 7% 17% 3% 45% 10% 3%

0% 15% 22% 39% 42% 87% 97% 100%

Portfolio wealth

Experience

Region

1 2 3 4 5 6 7 8 9 10 11 1 2 3 4 5

N/A Aegean Black Sea Central Anatolia East Anatolia Marmara Mediterranean Southeast Anatolia

it can be argued that the stock market was slightly bearish throughout the year. 3.2. Methodology Using a theoretical model, Harris and Raviv (1993) show that differences in opinions lead to trading among investors. Hence, trading volume is related to different expectations among investors. Differences in opinions are a result of different interpretations of the same signal by investors. As they rely more on their beliefs and decisions, overconfident investors’ interpretations of the same signal will significantly differ from those of rational investors. This difference may cause the increased trading volume for overconfident investors. De Bondt and Thaler (1995) state that the key behavioral factor in understanding the trading puzzle is overconfidence. Kyle and Wang (1997) and Benos (1998) argue that overconfident investors trade more aggressively because they believe that they have better information. Kahneman and Riepe (1998) propose that overconfidence should be expected from those who

do not face similar problems every day, or who make explicitly probabilistic estimates and obtain feedback on outcomes of their decisions, implying that individual stock investors are likely to be overconfident. Odean (1998) develops a theoretical model in which overconfident investors overestimate the precision of their knowledge about the value of an asset. These investors overestimate the probability that their personal assessment of an asset’s value is more accurate than that of others. Thus, overconfident investors believe their valuations to be correct which increases the differences in opinions among individual investors. The author proposes that trading volume increases when investors are overconfident. Odean (1999) tests this hypothesis using data provided by a nationwide discount brokerage house in the USA. He argues that if traders are overconfident in the precision of information, then the average return of the securities they sell must outperform the average return of the securities they buy. He finds that the average return of the securities sold outperform the average return of the securities purchased over the horizons of four months,

40

B. Tekçe, N. Yılmaz / Journal of Behavioral and Experimental Finance 5 (2015) 35–45

one year and two years. The author looks for possible explanations to excessive trading resulting in losses and eliminates meeting liquidity needs, realizing tax losses and rebalancing the portfolio or moving to a less risky portfolio. He concludes that excessive trading resulting in losses may be due to overconfidence. A number of studies also confirm that overconfident investors trade more.9 Similar to Barber and Odean (2001), we use annual turnover to measure overconfidence. Annual turnover is twelve times the average monthly turnover. For each month, monthly turnover is calculated based on the following formula. n 

Table 3 ‘‘Correlation of proxies’’. This table shows the correlation of overconfidence proxies (turnover, ISE30 ratio, diversification and Small Mcap ratio). Turnover is twelve times monthly turnover which is one half of the total buy and sell amounts in any month based on beginning of month prices divided by beginning of month portfolio value. ISE30 is the twelve month average of percentage of ISE30 stocks in the month end portfolios in 2011. Small Mcap is the twelve month average of percentage of small market capitalization stocks (less than USD100 m) in the month end portfolios in 2011. Diversification is the twelve month average of naive diversification of month end portfolios in 2011. Turnover ISE30 Diversification Small Mcap ***

Xit ∗Pit

ISE30

Diversification

***

−0.128 −0.052*** 0.114***

0.002

−0.477***

0.007***

Indicates correlation is significant at 1% level.

i=1

Monthly Turnover t =

Wt

2

where Xit is the amount of stock purchased or sold in month t, Pit is the beginning of the month price of the stock purchased or sold and Wt is the total stock portfolio value of the investor at the beginning of the month. Higher turnover implies higher overconfidence. Since both theoretical and empirical findings for turnover are robust, it is used as the main proxy to measure overconfidence. Josephs et al. (1992) argue that low self-esteem individuals take less risk than high self-esteem individuals. As Campbell (1990) shows, high self-esteem people have higher confidence. Hence, it can be inferred that overconfident investors tend to take more risk. Chuang and Lee (2006) find that overconfident investors trade more in riskier securities. They measure riskiness of a security as return volatility and firm specific risk (return volatility minus market component). Glaser and Weber (2009) also argue that after high portfolio values investors become more overconfident and buy high risk stocks. These findings imply that overconfidence can also be measured using portfolio riskiness. Consistently, we use two different measures as proxies of portfolio riskiness. The first measure is the percentage of stocks from ISE30 since these stocks have high market capitalization and high liquidity; thus we assume them to be less risky. The second measure is the percentage of small market capitalization stocks in the portfolio as we assume that smaller firms are riskier. Moreover, the average return volatility of ISE30 stocks was smaller and small market capitalization stocks were larger than rest of the stocks during 2009–2011. Hence, taking also return volatility into account, ISE30 stocks turned out to be less risky and small company stocks turn out to be more risky. ISE30 ratio is the twelve-month average of the monthly ISE30 ratio. Monthly ISE30 ratio is calculated based on the month-end portfolios. Monthly ISE30 Ratiot =

ISE30 stocks in the portfolio Total stocks in the portfolio

.

9 An incomplete list of papers include Barber and Odean (1999, 2000, 2001, 2002), Hirshleifer and Luo (2001), Gervais and Odean (2001), Chuang and Lee (2006), Statman et al. (2006), Korkmaz and Çelik (2007), Glaser and Weber (2007), Graham et al. (2009), Glaser and Weber (2009), Grinblatt and Keloharju (2009), and Hoffmann et al. (2010).

A lower ISE30 ratio implies riskier portfolios. For example, if stocks A and B are in ISE30, and a portfolio consists of stocks A, B and C then the ISE 30 ratio is 67%. Likewise, small Mcap ratio is twelve-month average of monthly small Mcap ratio. Monthly small Mcap ratio is calculated based on month-end portfolios. Monthly Small Mcap Ratiot

=

Small Mcap stocks in the portfolio Total stocks in the portfolio

.

Firms with market capitalization lower than USD 100 m are labeled as small. As of the end of 2011, almost 50% of the stocks have market capitalizations lower than USD 100 m. The maximum market capitalization value is USD 13,119 m. A higher small Mcap ratio implies riskier portfolios. Graham et al. (2009) find that competent investors trade more. Since higher level of trading is associated with overconfidence, we can argue that competence may lead to overconfidence. Heath and Tversky (1991) use the competence hypothesis and argue that overconfident investors may forego the advantage of diversification and concentrate on a small number of companies with which they are more familiar. Odean (1998) finds that overconfident traders hold under-diversified portfolios. Similarly, Goetzmann and Kumar (2008) find that high portfolio turnover, which is a sign of overconfidence, is related to underdiversification. The authors argue that more overconfident investors hold under-diversified portfolios similar to the investors with a tendency towards local stocks, in other words the investors with familiarity bias. Glaser and Weber (2009) argue that with increased portfolio turnover individuals reduce the number of stocks in their portfolio. These findings imply that overconfidence can be measured using diversification. In line with the literature, we use the average number of stocks in the portfolio as a naïve way of measuring the diversification level. Table 3 presents the correlations among proxies for overconfidence. Turnover is negatively correlated to the ISE30 ratio and diversification and is positively correlated to the small Mcap ratio. The small Mcap ratio is by definition negatively correlated to the ISE30 ratio and not to diversification. Correlation of diversification with other overconfidence proxies is either low or insignificant.

B. Tekçe, N. Yılmaz / Journal of Behavioral and Experimental Finance 5 (2015) 35–45

We also run a regression analysis to determine how each demographic factor affects overconfidence. Overconfidence = α Age + β Male + γ Experience

+ δ1 Wealth_Low + δ2 Wealth_High + θ1 Marmara + θ2 Southeast . In this regression model, age is the age of the investor and is a continuous variable. Male is a dummy variable for gender, which equals one for male investors. Experience is the year the account has been opened and is a continuous variable. Wealth_Low is a dummy variable, which equals one for portfolio wealth levels up to 10,000 TL and Wealth_High is a dummy variable and is equal to one for portfolio wealth levels higher than 100,000 TL. Marmara is a dummy variable, which equals one for the Marmara (most developed) region and Southeast is a dummy variable, which equals one for the Southeast Anatolia (least developed) region. Turnover is the main measures of overconfidence. Other proxies such as the ISE30 ratio and the small Mcap ratio are also used for robustness checks. Since explanatory variables are categorical, three additional regression models have been utilized for robustness checks. In these models, wealth is a dummy which equals one either for each portfolio wealth level presented in Table 2 or for low and high portfolio wealth levels presented above. In these models, experience is either continuous or is a dummy variable which equals one for each experience level presented in Table 2. However, for brevity, the results of these regressions are not presented. Return calculations are based on matched sell transactions. This approach allows us to calculate realized returns of each investor taking only into account stocks chosen for sale. By this way, we can better judge whether the investor is overconfident (i.e. the decision turns out to be wrong as return attached to this transaction is low or negative) or rational (i.e. the decision is justified by return level). That is why we believe, realized return better reflects the overall trading performance of investors as it takes both buy–sell as well as the timing of the sell decision. With a backward looking approach (for each investor starting from last trading day of 2011 and going back to beginning of 2011), we match each sell transaction with one or more buy transactions (depending on the amount of stocks sold). We then determine if the sell is a gain or loss by comparing the sell price with the purchase price. If there are more than one buy transactions for matching, we use the average purchase price as the reference price. For each investor, return is the weighted average of return of each matched transaction. For 7% of the investors, no return is calculated since no buy–sell matching is possible. Table 5 shows that the mean return for the remaining 93% of investors is 1.37% (median 0.56%) whereas the average daily return of ISE100 in 2011 is −0.10%. Although the average daily performance of ISE100 is lower, 43% of the investors have negative returns. Besides, comparing investor performance using realized return with ISE100 may lead to wrong conclusions. As the approach we employ allows us to calculate realized return of each investor based on each transaction, the realized return is in part driven by disposition effect (phenomenon of selling

41

Table 4 ‘‘Descriptive statistics for overconfidence’’. This table displays descriptive statistics for overconfidence proxies (turnover, ISE30 ratio, diversification and Small Mcap ratio). Turnover is twelve times monthly turnover which is one half of the total buy and sell amounts in any month based on beginning of month prices divided by beginning of month portfolio value. ISE30 is the twelve month average of percentage of ISE30 stocks in the month end portfolios. Small Mcap is the twelve month average of percentage of small market capitalization stocks (less than USD100 m) in the month end portfolios. Diversification is the twelve month average of naive diversification of month end portfolios in 2011. N

Min

Max

Mean

Median

Std. Dev.

Turnover 244,146

0.00

100.00

11.23

4.29

17.16

ISE30 ratio 244,146

0.00

1.00

0.30

0.16

0.34

Small Mcap ratio 244,064 0.00

1.00

0.28

0.22

0.27

Diversification 244,146 0.00

347.30

3.43

2.00

5.93

winners and holding on to losers). Of course, if stocks not sold were also taking into account, the average return of the investors would be lower and more comparable with ISE100 daily return. 4. Results Table 4 shows that Turkish individual stock investors have high turnover values which imply that the investors exhibit overconfident behavior. The mean annual turnover equals 11.2, more than 10 times computed by Barber and Odean (2001). The median equals 4.3. Both the mean and the median turnover levels are high compared to the literature.10 Individual investors in Turkey, being an emerging market, trade much more compared to developed markets such as US. The mean annual turnover level including the investors with turnover levels higher than 100 increases to 1.15 million, which is extremely high for a typical individual investor.11 Both the standard deviation in Table 4 and the histogram in Fig. 1 confirm that turnover is highly dispersed. Table 5 shows the mean turnover for each gender, age, portfolio wealth, experience and region. Table 6 reports the regression results. The coefficient on age is negative and statistically significant, suggesting that overconfident behavior decreases with age. However, as Table 5 shows, the effect of age is

10 Barber and Odean (2001) find that for a sub sample of US investors, the mean turnover ratio is 0.77 for men and 0.53 for women, implying that Turkish individual stock investors have a higher turnover than US investors. Chen et al. (2007) find that for Chinese investors, the mean turnover is 3.27, significantly higher than US investors, yet still lower than Turkish investors. 11 One possible explanation for these investors with high turnover values is that they have their wealth managed by professional money managers or they act like day traders and scalpers: therefore we exclude these investors. In our analyses we cap the turnover value at 100 in comparison with international benchmarks. Besides this, according to ISE and CRA, the average holding period of individual investors is around 30 days, implying an annual turnover of 12. The mean turnover is 11.2 when turnover is capped at 100. We have similar results with or without the high turnover investors.

42

B. Tekçe, N. Yılmaz / Journal of Behavioral and Experimental Finance 5 (2015) 35–45 Table 5 ‘‘Mean values’’. This table displays mean values for return and overconfidence at each gender, age, portfolio wealth, experience and region category based on the analysis investor set (244,146 investors). Return is significantly different at 1% level between male and female, lowest and highest age, portfolio wealth and experience categories and between the Marmara and the Southeast Anatolia regions. Turnover is twelve times monthly turnover which is one half of the total buy and sell amounts in any month based on beginning of month prices divided by beginning of month portfolio value. Total

Return 1.37%

Overonfidence (Turnover) 11.23

Gender

Male Female

1.18% 2.36%

11.86 8.10

Age

25 < 25–29 30–34 35–39 40–44 45–49 50–54 55–59 60 ≥

1.02% 0.48% 0.81% 1.13% 1.33% 1.76% 1.90% 2.17% 2.47%

10.72 11.84 12.30 12.09 11.80 11.06 10.34 9.43 8.10

−0.52% −0.24%

Portfolio wealth

100–1000 1000–5000 5000–10,000 10,000–20,000 20,000–50,000 50,000–100,000 100,000–250,000 250,000–500,000 500,000–1,000,000 1,000,000–5,000,000 5,000,000+

0.59% 1.28% 1.94% 2.45% 2.86% 3.04% 2.89% 3.37% 3.57%

7.47 12.75 12.17 11.76 10.80 10.05 9.50 9.33 8.96 8.61 6.55

Experience

After 2008 2006–2008 2003–2005 2000–2002 Before 2000

0.57% 0.94% 1.45% 1.63% 2.00%

11.85 11.35 10.98 11.36 11.02

Region

N/A Aegean Black Sea Central Anatolia East Anatolia Marmara Mediterranean Southeast Anatolia

1.49% 1.38% 1.36% 1.44% 0.54% 1.51% 1.11% 0.54%

9.97 11.06 11.38 11.42 13.00 10.77 11.96 13.40

Table 6 ‘‘Regression results’’. This table displays regression results for overconfidence. Dependent variable is annual turnover, ISE30 ratio and small Mcap ratio. Independent variables are age, gender (in the form of male dummy), experience, portfolio wealth (in the form of dummy variables for low and high portfolio wealth categories) and geographical region of residence (in the form of dummy variables for Marmara and Southeast Anatolia regions). Wealth_Low is dummy for portfolio value less than 10,000 TL and Wealth_High is dummy for portfolio value above 100,000 TL. Turnover Constant Age Male Experience Wealth_Low Wealth_High Marmara Southeast Obs R2 Adjusted R2 ***

ISE30

Small Mcap

−0.058*** (−25.1)

0.106*** (45.8)

−0.085*** (−36.5)

0.075*** (37.1) 0.027*** (11.5) 0.027*** (12.4) −0.025*** (−11.9) −0.016*** (−7.6) 0.013*** (6.3) 244,146 0.014 0.014

−0.046*** (−22.7) −0.003 (−1.3) −0.007*** (−3.4)

0.047*** (23.3) 0.011*** (4.6) 0.017*** (8.0) −0.061*** (−28.8) −0.037*** (−18.2) 0.009*** (4.5) 244,064 0.019 0.021

0.056*** (26.4) 0.021*** (10.5) −0.006*** (−3.0) 244,146 0.021 0.021

Indicates coefficient is significant at 1%. t-values in parenthesis.

B. Tekçe, N. Yılmaz / Journal of Behavioral and Experimental Finance 5 (2015) 35–45

43

Fig. 1. ‘‘Turnover histogram’’. This figure displays the frequency of annual turnover which is used to measure overconfidence. A significant portion of investors have high turnover levels, implying that these investors exhibit high degree of overconfidence.

nonmonotonic; age has a positive effect on overconfidence up to the 30–34 age group, but has a negative effect for the higher age groups. The nonlinear relationship is also confirmed by the quadratic age term in the regression, which we omit for brevity. Male investors exhibit more overconfidence than female investors which confirms the findings in the literature. On the other hand, experience increases overconfident behavior contrary to the expectations. However, this is valid only for investors with low portfolio value. Experience decreases overconfidence for investors with high portfolio value. One possible explanation is the definition of experience. The account opening date does not necessarily measure experience. An investor may gain experience in a shorter period of time with frequent trading. However, when an alternative definition of experience is used (the number of years after the first time the investor’s stock portfolio reaches 5,000 TL), the results do not change. The coefficient on wealth is negative and statistically significant suggesting that portfolio wealth decreases overconfidence. However, as Table 5 shows, the only exception is the lowest portfolio wealth group which has the second lowest turnover among all wealth categories. This is probably mainly due to low amount of available funds to trade. Portfolio wealth may be related to financial sophistication as wealthy investors have better access to information and can leverage on professional portfolio management. Investors in the Marmara region have lower and investors in the Southeast Anatolia region have higher overconfidence. The turnover difference between the two regions is not related to gender, age, experience or portfolio wealth. The Marmara region is economically more developed than the Southeast Anatolia region. Besides this, the percentage of university graduates is higher in the Marmara region.12 Therefore, financial sophistication in the Marmara region is more likely to be higher than that in the Southeast Anatolia region. Overall, portfolio wealth and region imply that sophisticated investors are less prone to overconfidence. Table 7 shows the negative relationship between return and turnover. As presented in Panel A, the mean return in

12 13% in the Marmara region versus 6% in the Southeast Anatolia region. Data based on Turkish Statistical Institute.

the lowest quartile of turnover is 2.79% and is significantly higher compared to that of the highest quartile which is 0.12%. The mean return for analysis investor set is 1.37% with a median of 0.56%. Although not presented here, the mean return of investors with turnover higher than 100 is 0.35% with a median of −0.07%. Regression result in Panel B also confirms the negative relation between return and overconfidence. Moreover, Table 5 shows that men have a lower return than women, 1.18% versus 2.36%. Hence, in line with the literature, we find that trading is hazardous to wealth and men’s trading performance is worse than women’s due to men’s higher turnover levels. In order to check the robustness of our findings, we experiment with some alternative variable definitions and subsamples. Our findings of the analyses are virtually unchanged when we use these alternative measures. First, we use the ISE30 ratio, the small Mcap ratio and diversification as alternative proxies for overconfidence. Table 4 shows the descriptive statistics for these proxies. Table 3 shows that turnover is negatively correlated to the ISE30 ratio and diversification and positively correlated to the small Mcap ratio. Small Mcap ratio is by definition negatively correlated to the ISE30 ratio and not to diversification. Correlation of diversification with other proxies is either low or insignificant, implying that diversification is not as good as other proxies in measuring overconfidence, or does not necessarily measure overconfidence. Hence, further regression analysis for overconfidence robustness checks is based on portfolio riskiness (measured by ISE30 ratio and small Mcap ratio). Table 6 shows regression results for the ISE30 and the small Mcap. Results are in line with those of turnover. Although for brevity they are not presented here, regression results are further confirmed using different robustness checks. Results do not change when we include high turnover investors or look at different sub samples including male only, female only, low/high age, low/high experience, and low/high portfolio wealth regressions. Our findings are also robust for different regression models such as using categorical versus continuous explanatory variables. 5. Conclusion Empirical studies in the behavioral finance literature show that individuals may not behave rationally. The behavioral biases may influence investor decisions and af-

44

B. Tekçe, N. Yılmaz / Journal of Behavioral and Experimental Finance 5 (2015) 35–45

Table 7 ‘‘Return performance and overconfidence’’. Panel A displays mean returns at quartile level based on analysis investor set (244,146 investors). Panel B displays regression results for return and overconfidence. Overconfidence is measured by turnover. Turnover is twelve times monthly turnover which is one half of the total buy and sell amounts in any month based on beginning of month prices divided by beginning of month portfolio value. Panel A

Overconfidence* (Turnover)

Return

Total

1st quartile lowest 25%

2nd quartile

3rd quartile

4th quartile highest 25%

Mean Median Std. Dev.

1.37% 0.56% 10.87%

2.79% 2.29% 17.81%

2.03% 1.87% 11.94%

0.92% 0.62% 6.52%

−0.14%

0.12% 4.21%

Panel B

Constant Overconfidence (Turnover)

Return

t values

−0.066**

−31.4

*

In Panel A indicates 1% significance between first and fourth quartile mean returns. ** In Panel B indicates coefficient is significant at 1%.

fect financial markets. However, these studies are limited to the subsamples of the overall investor groups studied and mainly focus on the USA and Europe. In this paper, we focus on Turkey, an emerging market with a collectivist culture, in order to use a unique nationwide dataset to analyze overconfidence. We examine how prevalent overconfidence is, what factors affect overconfidence, and how overconfidence relates to investor return performance using transaction and demographic data. Using nationwide data for buy and sell transactions of individual investors, we report several interesting findings. First, we find that overconfidence is highly common among investors. Turkish individual investors exhibit more overconfident behavior compared to that observed in the behavioral finance literature. Male investors exhibit more overconfidence than female investors, age and portfolio wealth decreases overconfidence, and investors in more developed regions are less overconfident than those in the less developed regions. Portfolio wealth and region results imply that sophisticated investors are less prone to overconfidence. Moreover, overconfidence is hazardous to wealth as the mean return decreases with increasing turnover. Our findings are line with the findings for China and Taiwan showing that individual investors in an emerging economy with collectivist characteristics are prone to behavioral biases as well, generally to a higher extent, compared to their counterparts in the developed and mostly individualistic societies. Consequently, we may expect similar results in the other emerging markets, most of which are observed to be collectivist countries, too, such as Brazil and Russia. References Acker, D., Duck, N.G., 2008. Cross-cultural overconfidence and biased self attribution. J. Socio Econ. 37, 1815–1824. Barber, B.M., Odean, T., 1999. The courage of misguided convictions. Financ. Anal. J. 55 (6), 41–55. Barber, B.M., Odean, T., 2000. Trading is hazardous to your wealth: the common stock investment performance of individual investors. J. Finance 55 (2), 773–806. Barber, B.M., Odean, T., 2001. Boys will be boys: gender, overconfidence, and common stock investment. Quart. J. Econ. 116 (1), 261–292. Barber, B.M., Odean, T., 2002. Online investors: do the slow die first? Rev. Financ. Stud. 15 (2), 455–487. Barberis, N., Huang, M., Santos, T., 2001. Prospect theory and asset prices. Quart. J. Econ. 116 (1), 1–53.

Barberis, N., Thaler, R., 2003. A survey of behavioral finance. In: Constantinides, G.M., Harris, M., Stulz, R. (Eds.), Handbook of the Economics of Finance. Benos, A.V., 1998. Aggressiveness and survival of overconfident traders. J. Financ. Mark. 1, 353–383. Black, F., 1986. Noise. J. Finance 41 (3), 529–543. Campbell, J.D., 1990. Self-esteem and clarity of the self-concept. J. Pers. Soc. Psychol. 59 (3), 538–549. Chen, G., Kim, K.A., Nofsinger, J.R., Rui, O.M., 2007. Trading performance, disposition effect, overconfidence, representativeness bias, and experience of emerging market investors. J. Behav. Decis. Mak. 20, 425–451. Chuang, W., Lee, B.S., 2006. An empirical evaluation of the overconfidence hypothesis. J. Bank. Finance 30, 2489–2515. Chui, A.C.W., Titman, S., Wei, K.C.J., 2010. Individualism and momentum around the world. J. Finance 65 (1), 361–392. Daniel, K., Hirshleifer, D., Subrahmanyam, A., 1998. Investor psychology and security market under and over reaction. J. Finance 53 (6), 1839–1885. Daniel, K., Hirshleifer, D., Teoh, S.H., 2002. Investor psychology in capital markets: evidence and policy implications. J. Monetary Econ. 49, 139–209. Deaves, R., Lüders, E., Schröder, M., 2010. The dynamics of overconfidence: evidence from stock market forecasters. J. Econ. Behav. Organ. 75, 402–412. De Bondt, J.B., Thaler, R., 1995. Financial decision making in markets and firms: a behavior perspective. In: Jarrow, R.A., Maksimovic, V., Ziemba, W.T. (Eds.), Handbooks of Operations Research and Management Science. De Long, J.B., Shleifer, A., Summers, L.H., Waldmann, R.J., 1990. Noise trader risk in financial markets. J. Polit. Econ. 98 (4), 703–738. Ekholm, A., Pasternack, D., 2007. Overconfidence and investor size. Eur. Financ. Manag. 14 (1), 82–98. Fan, J.X., Xiao, J.J., 2005. A cross-cultural study in risk tolerance: comparing Chinese and Americans. SSRN Paper No.: 939438. Fischhoff, B., Slovic, P., Lichtenstein, S., 1977. Knowing with certainty: the appropriateness of extreme confidence. J. Exp. Psychol. 3 (4), 552–564. Gervais, S., Odean, T., 2001. Learning to be overconfident. Rev. Financ. Stud. 14 (1), 1–27. Glaser, M., Weber, M., 2007. Overconfidence and trading volume. Geneva Risk Insur. Rev. 32, 1–36. Glaser, M., Weber, M., 2009. Which past returns affect trading volume? J. Financ. Mark. 12, 1–31. Goetzmann, W.N., Kumar, A., 2008. Equity portfolio diversification. Rev. Financ. 12, 433–463. Graham, J.R., Harvey, C.R., Huang, H., 2009. Investor competence, trading frequency and home bias. Manage. Sci. 55 (7), 1094–1106. Griffin, D., Tversky, A., 1992. The weighing of evidence and determinants of confidence. Cogn. Psychol. 24, 411–435. Grinblatt, M., Keloharju, M., 2009. Sensation seeking, overconfidence and trading activity. J. Finance 64 (2), 549–578. Harris, M., Raviv, A., 1993. Differences of opinion make a horse race. Rev. Financ. Stud. 6 (3), 473–506. Heath, C., Tversky, A., 1991. Preference and belief: ambiguity and competence in choice under uncertainty. J. Risk Uncertain. 4, 5–28. Hirshleifer, D., 2001. Investor psychology and asset pricing. J. Finance 56 (4), 1533–1597.

B. Tekçe, N. Yılmaz / Journal of Behavioral and Experimental Finance 5 (2015) 35–45 Hirshleifer, D., Luo, G.Y., 2001. On the survival of overconfident traders in a competitive securities market. J. Financ. Mark. 4, 73–84. Hoffmann, A.O.I., Shefrin, H., Pennings, J.M.E., 2010. Behavioral portfolio analysis of individual investors. SSRN Paper No.: 1629786. Hofstede, G., 2001. Culture’s Consequences: Comparing Values, Behaviors, Institutions and Organizations Across Nations, second ed. Sage Publications. Josephs, R.A., Larrick, R.P., Steele, C.M., Nisbett, R.E., 1992. Protecting the self from the negative consequences of risky decisions. J. Pers. Soc. Psychol. 62 (1), 26–37. Kahneman, D., Riepe, M.W., 1998. Aspects of investor psychology. J. Portf. Manag. 24 (4), 52–65. Kahneman, D., Tversky, A., 1979. Prospect theory: an analysis of decision under risk. Econometrica 47 (2), 263–291. Korkmaz, T., Çelik, E.İ., 2007. The validity of overconfidence hypothesis in behavioral finance models: an application on ISE. İkt. İşlet. Finans. 22 (261), 137–154. Kunda, Z., 1987. Motivated inference: self-serving generation and evaluation of causal theories. J. Pers. Soc. Psychol. 53 (4), 636–647. Kyle, A.S., Wang, F.A., 1997. Speculation duopoly with agreement to disagree: can overconfidence survive the market test? J. Finance 52 (5), 2073–2090. Langer, E.J., 1975. The illusion of control. J. Pers. Soc. Psychol. 32 (2), 311–328. Li, S., Bi, Y.L., Zhang, Y., 2009. Asian risk seeking and overconfidence. J. Appl. Soc. Psychol. 39 (9), 2706–2736. Miller, D.T., Ross, M., 1975. Self-serving biases in the attribution of causality: fact or fiction? Psychol. Bull. 82 (2), 213–225.

45

Odean, T., 1998. Volume, volatility, price and profit when all traders are above average. J. Finance 53 (6), 1887–1934. Odean, T., 1999. Do investors trade too much? Amer. Econ. Rev. 89 (5), 1279–1298. Russo, J.E., Shoemaker, P.J.H., 1992. Managing overconfidence. Sloan Manage. Rev. 33 (2), 7–17. Shleifer, A., Vishny, R.W., 1997. The limits of arbitrage. J. Finance 52 (1), 35–55. Statman, M., 2010. The cultures of risk tolerance. SSRN Paper No.: 1647086. Statman, M., Thorley, S., Vorkink, K., 2006. Investor overconfidence and trading volume. Rev. Financ. Stud. 19 (4), 1531–1565. Subrahmanyam, A., 2007. Behavioral finance: a review and synthesis. Eur. Financ. Manag. 14 (1), 12–29. Svenson, O., 1981. Are we all less risky and more skillful than our fellow drivers? Acta Psychol. 47, 143–148. Ülkü, N., İkizlerli, D., 2012. The interaction between foreigners’ trading and emerging stock returns: evidence from Turkey, 13, 381–409. Vissing-Jorgensen, A., 2004. Perspectives on behavioral finance: does irrationality disappear with wealth? Evidence from expectations and actions. NBER Macroeconomics Annual 2003, Vol. 18. Weber, E.U., Hsee, C.K., 1998. Cross-cultural differences in risk perception, but cross-cultural similarities in attitudes towards perceived risk. Manage. Sci. 44 (9), 1205–1217. Weinstein, N.D., 1980. Unrealistic optimism about future life events. J. Pers. Soc. Psychol. 39 (5), 806–820.