Microsoft Word - Sensation seeking submitted to JF.doc

Sensation Seeking, Overconfidence,

and Trading Activity

Mark Grinblatt

UCLA Anderson School of Management and NBER

Matti Keloharju

Helsinki School of Economics and CEPR

September 14, 2006

We would like to thank the Finnish Vehicle Administration, the Finnish Armed

Forces, the Finnish Central Securities Depository, and the Finnish Tax

Administration for providing access to the data, as well as the Office of the Data

Protection Ombudsman for recognizing the value of this project to the research

community. Our appreciation also extends to Antti Lehtinen and Juan Prajogo,

who provided superb research assistance, and to Ivo Welch, Samuli Knüpfer,

Juhani Linnainmaa, Tyler Shumway, and seminar participants at the Hong Kong

University of Science and Technology, University of Illinois, London Business

School, London School of Economics, the University of Mannheim, University of

Michigan, Oxford University, and the University of Texas, who generated many

insights that benefited this paper. We also thank Seppo Ikäheimo for his help in

obtaining the data and Markku Kaustia, Samuli Knüpfer, Lauri Pietarinen, and

Elias Rantapuska for participating in the analysis of the Finnish Central Securities

Depository data. Financial support from the Academy of Finland, the Foundation

for Economic Education, and the Paulo Foundation is gratefully acknowledged.

Sensation Seeking, Overconfidence,

and Trading Activity

Abstract

This study analyzes the role that two psychological attributes—sensation seeking

and overconfidence—play in the tendency of investors to trade stocks. Equity

trading data are combined with data from an investor’s tax filings, driving record,

and psychological profile. We use the data to construct measures of

overconfidence and sensation seeking tendencies. Controlling for a host of

variables, including wealth, income, age, number of stocks owned, marital status,

and occupation, we find that overconfident investors and those investors most

prone to sensation seeking trade more frequently.

JEL classification: G10, G11

1. Introduction

The extraordinarily high volume of equity trading that takes place on stock exchanges

represents one of the greatest puzzles in finance.

The classic asset pricing models suggest

that investors should optimally buy a single risky portfolio and hold it in perpetuity.

Rebalancing motivations generate only negligible trading for most plausible return generating

processes, even when the motive is a proper asset allocation between cash and risky assets.

Researchers have offered a variety of explanations for why trading volume is so large,

but these are not fully satisfactory. The obvious candidate, heterogeneous information, does

not immediately generate rational trading. Milgrom and Stokey (1982) argued that in the

presence of common knowledge, there is no rational trade, even with heterogeneous

information signals. Models with disagreement require noise trading to generate trading

volume.

Without a satisfactory explanation of what drives the noise trading and a calibration

that accounts for observed trading, trading volume remains a puzzle.

As a consequence of the failure of traditional models to explain trading, empiricists

have begun to study and document how behavioral factors might explain trade. Odean

(1999), Grinblatt and Keloharju (2001), and Grinblatt and Han (2005) argue that trading can

arise as a consequence of a disposition effect. Odean (1998, 1999) suggests that

overconfidence may drive excessive trading. Barber and Odean (2001) find that gender is

related to trading—the portfolios of males exhibit greater turnover—and that this is due to

their greater overconfidence. Gervais and Odean (2001) develop a model in which successful

chance experiences can generate overconfidence and lead to excessive trading. Graham,

Harvey, and Huang (2005) contend that competence drives trading.

See, for example, Odean (1999) and Barber and Odean (2000, 2002).

See Weiss (1979), Grossman and Stiglitz (1980), and Pfleiderer (1984).

This paper employs data on the psychological traits of each investor to show that

trading may be driven by two emotional considerations: The first is sensation seeking, which

is the search for novel, intense, and varied experiences generally associated with real or

imagined physical, social, and financial risks; the second is overconfidence, which is the

tendency to place an irrationally excessive degree of confidence in one’s abilities and beliefs.

Using a comprehensive dataset from Finland, which offers a remarkable number of control

variables, we show that investors who are most prone to sensation seeking and those who are

most overconfident trade the most. To our best knowledge, this is the first study to

specifically focus on sensation seeking as a motivation for trade and the first that employs

comprehensive data to directly measure overconfidence and study its relationship to trading.

A potential link between sensation seeking and trading activity should be apparent to

anyone who has spent time in a casino. Gambling, which the psychology literature has tied to

sensation seeking,

is a worldwide industry, generating over $2 trillion in revenues each year.

There also is widespread unreported and often illegal gambling activity that most everyone is

acquainted with and which frequently is reported on in the popular press. Given the

consumptive utility enjoyed in all sorts of activities where an element of risk and thrill is

involved, it would truly be surprising if some consumers did not enjoy the thrill from trading

in the stock market. A recent paper by Kumar (2005) concludes that investor-types with

characteristics associated with an attraction to gambling prefer lottery-like stocks.

A link between overconfidence and trading activity is also plausible, as recent

literature has noted.

When one’s private valuation of a stock differs from that of the market,

For a review of the literature, see Raylu and Oei (2002).

Source: BBC News World Edition August 22, 2005 reporting on “Place Your Bets, Part 3.”

Barber and Odean (2001) test whether overconfidence drives trading using gender as a proxy for

overconfidence. Glaser and Weber (2004), using data on 215 online investors who responded to a survey, find

that the better than average effect is related to trading frequency. Using experimental data, Deaves et al. (2003)

the overconfident investor places more validity on his private valuation and less on the

market’s valuation. This can occur for two related reasons. The first is hubris or what is

sometimes referred to as the “better than average effect.” The other is “miscalibration.” This

arises when the confidence interval around the investor’s private signal is tighter than it is in

reality. Both forms of overconfidence lead the overconfident investor to form posteriors with

excessive weight on private signals. This generates larger trades than would be observed in a

less confident investor.

We measure sensation seeking as the number of automobile speeding convictions

earned by an investor over a multi-year period. Zuckerman (1994), one of the pioneers of the

concept, as well as Jonah (1997), suggest that driving behavior may be one of the best

observed behaviors for assessing sensation seeking. Data on speeding tickets from Finland is

particularly pertinent with respect to the financial risks associated with this trait. In Finland,

the fine for substantive automobile violations is a function of income. Thus, those who risk

breaking the law do so under severe financial penalty as well as enduring possible physical

risks.

We derive the overconfidence measure from a standard psychological test. This test is

given to all Finnish males at approximately the age of 19 or 20 (generally, many years prior to

observation of an investor’s trading activity). One of the scales from the test measures self-

confidence. As this confidence measure is a combination of talent and overconfidence, we

use regression analysis to control for talent and obtain overconfidence as the residual effect.

Because of the mandatory and comprehensive nature of the psychological examination, the

responses lack the bias typically associated with the decision of whether to answer a survey.

observe that overconfidence is positively related to trading activity, while Biais et al (2005) find that

overconfidence reduces trading performance.

The correlation between our sensation seeking and overconfidence measures is very

low, so both behavioral attributes have relatively independent influence on trading. Sensation

seeking is less related to the decision of whether to trade at all and more related to the

decision of how much to trade. Although the number of trades is influenced by

overconfidence, there does not appear to be a relationship between overconfidence and

turnover. In light of the Barber and Odean (2001) conclusion that turnover is related to

overconfidence, this finding points to the importance of direct measurement of behavioral

attributes.

The paper is organized as follows: Section 2 offers motivation for the paper and

describes the data. Section 3 presents the results on sensation seeking, overconfidence, and

trading activity. Section 4 concludes the paper.

2. Motivation and Data

The literature in finance is ripe with stylized facts about investor behavior. One of the

most prominent is that trading propensity is related to gender.

Figure 1 Panel A plots the

average number of trades per year as a function of age and gender. Consistent with earlier

findings, men trade more than women within all age groups. Panel B effectively offers the

same plot but takes out the effect of income, wealth, and the number of stocks in the portfolio.

It reports coefficients on dummies for birth year and the sum of those coefficients and those

on the product of a male gender dummy and the same birth year dummies, controlling for

income decile dummies and wealth decile dummies. The pattern is a bit different. Now,

except for those who are under age 23 at the start of our stock trading sample period, the gap

See, for example, Barber and Odean (2001) and Agnew et al (2003).

between the trading inclinations of men and women seems to diminish with age. Still, males

trade more than females, irrespective of age.

What lies behind the greater tendency of males to trade? One possibility is that males

enjoy the thrill of trading to a greater extent than females. Barber and Odean (2001) argued

that one factor that might explain gender-related differences in trading activity is that the

entertainment value of trading has a consumptive appeal that is similar to the thrill obtained

from gambling. In the psychology literature, the gambler’s thrill derives from an attribute

known as “sensation seeking.” This literature has shown that this attribute varies across

investors by age and gender. Ultimately, Barber and Odean (2001) dismissed entertainment

as an explanation for their results in favor of gender-based differences in overconfidence.

This is because trading volume, as measured by turnover, was largely invariant to the fraction

of an investor’s net worth invested in common stock.

We think this dismissal is premature.

Panel C plots the number of speeding tickets, a proxy for an investor’s degree of susceptibility

to sensation seeking, as a function of age and birth year. Except for those under 23 at the start

of the trading sample period, there is a marked similarity between the two graphs in Panels B

and C.

Irrespective of whether sensation seeking ultimately accounts for the trading pattern

in Panel B’s graph, we find it difficult to wholly dismiss it as a motivation for trading activity

when the evidence supporting this dismissal is so indirect. It would clearly be interesting to

run a horse race between sensation seeking and overconfidence if one had direct measures of

these attributes for each investor. We are fortunate to be able to analyze such data.

This argument hinges on the assumption that some fixed fraction of one’s wealth must be invested

prudently, even if entertainment from the act of trading drives much of observed trading activity.

One has to be cautious about drawing conclusions from this similarity. As Ameriks and Zeldes (2004)

and others point out, however, it is very difficult to disentangle cohort, age, and time effects from each other.

2.1 Sensation seeking

The classic characterization of sensation seeking is found in Zuckerman (1994, p. 27).

He labels sensation seeking as “… a trait defined by the seeking of varied, novel, complex,

and intense sensations and experiences, and the willingness to take physical, social, legal, and

financial risks for the sake of such experience.“

With respect to trading activity, sensation seeking is distinct from the magnitude or

sign of the risk aversion parameter. For example, the willingness to take on an undiversified

trading strategy may be encouraged by the consumptive value associated with sensation

seeking, yet deterred by a high degree of risk aversion. The mix of these two competing

forces may determine the degree of diversification. However, as Barber and Odean (2001)

observe, an investor’s risk aversion parameter has little bearing on desired trading frequency.

To the extent that markets are efficient, trading frequency has costly financial consequences,

but these are not tied to the risk aversion parameter. The mere act of trading and the

monitoring of a constant flow of “fresh stocks” in one’s portfolio may create a more varied

and novel experience than a buy and hold strategy, but it does not increase volatility per se.

There is reason to believe that males are more prone to sensation seeking behavior.

As Zuckerman (1994) points out, males are more prone to risky sporting activities. While

some of this may be explained by physical traits, there also is a greater tendency among males

towards violence, alcohol, drugs, gambling, and most forms of illicit activity that is not as

easily explained. Even relatively safe sensation seeking behaviors, like high speed

amusement park rides, are more popular among males.

A review article by Jonah (1997)

documents that sensation seeking is significantly related to risky driving.

See, for example, Zuckerman, Eysenck, and Eysenck (1978) and Ball, Farnhill, and Wangeman

(1984).

See Begg and Langley (2001).

Men also differ from women with respect to the type of gambling they do. Potenza et

al (2001) find that men prefer action-oriented forms of gambling, like blackjack, craps, or

sports betting, as opposed to passive escape-oriented gambling (e.g., slot machines, lotteries).

Biaszcynski et al (1997) as well as Vitaro et al (1997) suggest that action-oriented gambling

reflects a higher level of sensation seeking among males. Comings (1998) shows that

pathological gambling behavior may be transmitted genetically. Pavalko (2001, p. 34) likens

trading (as opposed to investing) to action-oriented gambling.

2.2 Overconfidence

The second explanation we investigate for the greater trading of males is

overconfidence. The literature offers differing views on whether males actually are more

overconfident than women. Lundeberg, Puncochar, and Fox (1994) argue that men are more

overconfident than women, particularly for tasks that are perceived to be in the masculine

domain. Pulford and Colman (1996) find that men are slightly more overconfident.

However, Deaux and Farris (1977), Beyer (1997), and Beyer and Bowden (1998) find that

men have higher self perceptions than women but also tend to be better—not less—calibrated.

Lichtenstein and Fishhoff (1981), Lundeberg et al (2000), Deaves et al (2003), and Biais et al

(2005) find no difference in overconfidence between men and women.

To assess whether overconfidence explains trading, it would be useful to directly

observe a measure of overconfidence, rather than a measure that is tied to a gender-based

instrument. We have overconfidence measures on a large sample of subjects from an

extensive psychological profile of those subjects. Our data also offer the possibility of a

much cleaner test of whether overconfidence causes excessive trading. Ideally, in a controlled

experiment of whether overconfidence affects trading activity, all other attributes of the

subjects would be identical and only overconfidence would vary. In a social science

experiment, this ideal is not attainable. However, in our study, all of the subjects for whom

we have a direct measure of overconfidence happen to be male. Moreover, the age at which

we measure overconfidence is approximately the same across subjects (about 20). To

demonstrate a link between such a measure of overconfidence and trading activity would

indeed be remarkable, as it may imply that overconfidence is a stable characteristic that

influences economic behavior throughout one’s lifetime. We also have data on a large

number of control variables that allow us to use traditional regression analysis to assess

overconfidence, with fewer concerns about omitted variables than one typically has in studies

of economic behavior.

2.3 Data Sources

Our paper’s analysis requires us to combine information from a number of datasets:

• FCSD data. This dataset records the portfolios and trading records from January 1,

1995 through November 29, 2002 of all household investors domiciled in Finland. The

daily electronic records we use are exact duplicates of the official certificates of

ownership and trades, and hence are very reliable. Details on this data set, which

includes trades, holdings, and execution prices, are reported in Grinblatt and Keloharju

(2000, 2001). We study trading data from July 1, 1997 on for those individuals who

held stocks at some point between January 1, 1995 and June 30, 1997. The latter

requirement allows us to focus on the determinants of trading activity rather than on

whether an investor participates in the stock market in the first place. (The results are

qualitatively similar if we use all individuals in lieu of individuals who have invested

in the market before.) In addition to trading data, we use this dataset to measure

financial wealth and number of stocks held.

• HEX stock data. Closing transaction prices are obtained from a dataset provided by

the Helsinki Exchanges (HEX). In combination with the FCSD data, this dataset is

used to measure financial wealth.

• FVA driver data. Data from the Finnish Vehicle Administration (FVA) were used to

obtain a set of subjects who have a normal vehicle driving license (as opposed to a

motorcycle or professional driving license) as of July 1, 1997. The FVA data contain

all driving-related final judgments on each motorist in the provinces of Uusimaa and

East Uusimaa between July 1, 1997 and December 31, 2001. (These provinces

contain Greater Helsinki and represent the most densely populated areas in Finland.)

The judgments contain paragraphs about the nature of the violation that we coded

either as “speeding related” or “not speeding related.” Thus, we have comprehensive

records of tickets for speeding that were finalized over a four and a half year period.

We use this data to measure differences in the sensation seeking attribute across

investors.

• FAF psychological profile. This dataset, from the Finnish Armed Forces, helps us to

measure cross-sectional variation in overconfidence among investors. Around the

time of induction into mandatory military duty in the Finnish armed forces, typically

at ages 19 or 20, males in Finland take a battery of psychological tests. It includes a

leadership inventory test for which we have comprehensive data beginning January 1,

1982 and ending December 31, 2001. The leadership inventory exam, which includes

Non speeding offenses are fewer in number, varied across many categories, and difficult to interpret.

For example, tickets do get issued for driving too slowly on a freeway. For these reasons, we focus only on

speeding offenses in the sample. When we pool speeding with all other driving offenses as our measure of

sensation seeking, we obtain highly similar results.

218 “agree” or “do not agree” questions, provides eight scales for leadership. One of

these scales is self-confidence, which is reported as a number from 1 to 9 (and is

designed to approximate a stanine in the overall sample of test takers). We convert

this measure to an overconfidence measure using regression techniques described later

in the paper for all shareholders who have driver’s licenses prior to the start of July 1,

1997. The psychological profile also contains an intellectual ability score. The test

measures intellectual ability in three areas: mathematical ability, verbal ability, and

logical reasoning. FAF forms a composite ability score from the results in these three

areas. We use the composite ability score in our analysis.

• FTA dataset. This dataset, from the Finnish Tax Administration, contains annual data

from the 1998 and 1999 tax returns of Finnish investors in the provinces of Uusimaa

and East Uusimaa, as well as data from a population registry. Variables constructed

from this source include income, age, gender, marital status, occupation, and

homeownership status. These variables are used as controls in regressions that explain

trading activity and regressions used to construct a measure of overconfidence for an

individual. We use 1998 data for all of the variables except for employment status,

which is first reported in 1999.

2.3 Variable Description and Summary Statistics

Our analysis largely consists of cross-sectional regressions, with some measure of

trading activity as a left hand side variable. The variables and the data sources for them are

described in Table 1 Panel A. The remainder of the table provides summary statistics on the

data. Panel B describes means, medians, standard deviations, and interquartile ranges for

most of the variables. Panel C provides detailed summary statistics on the self-confidence

measure. Panel D presents the correlation matrix for relevant variables.

As can be seen from Table 1, Panel B, stock trades and speeding tickets are rare.

Panel C’s distribution of the self-confidence measure indicates that the highest and lowest

measures of self-confidence (1 and 9) also are relatively rare. Our sample of male drivers is a

bit more self-confident than the universe of males taking the exam. Some of this may have to

do with the fact that we limit our sample to individuals who own stocks between January

1995 and June 1997. Thus, our sample is wealthier than the population at large. Panel D

indicates that the number of speeding convictions, self-confidence, and gender all have a

relatively large correlation with various measures of a subject’s trading activity, but self-

confidence, described later, is close to being uncorrelated with the number of tickets earned.

Consistent with Figure 1 Panel A, age (without controls for income) does not display an

obvious relationship with trading activity. Panel D also indicates that gender per se (with a

dummy value of one being male) is more correlated with all measures of trading activity than

are measures of sensation seeking and self-confidence. However, gender also is highly

correlated with the sensation seeking attribute, as we hypothesized earlier.

3. Results

Our analysis has two parts to it. The first part studies sensation seeking and the role it

plays in trading activity. This analysis makes use of both males and females. The second part

jointly focuses on sensation seeking and overconfidence as explanations for trading activity.

Because our overconfidence score can only be computed for young and middle-aged males, it

contains fewer observations.

The correlations of the variables in the table with overconfidence, which is derived from self-

confidence with a procedure described later, are similar to their correlations with self-confidence.

3.1 Sensation Seeking Results

Earlier, we mentioned that our proxy for sensation seeking is the number of final

convictions for speeding. Admittedly, speeding convictions are not a perfect instrument for

speeding because not all violators are caught. However, in Finland, where many fines are tied

to income, there is less reason to believe that the motivation for traffic violations is a rational

calculation based on the cost of one’s time. For example, Jussi Salonoja, a wealthy

businessman, received a 170,000 euro fine for driving 80km/hour in a 40km/hour zone, while

Anssi Vanjoki, a Nokia executive, received an 80,000 euro ticket for driving 75km/hr in a

50km/hr zone.

Moreover, because of the extreme cost of being caught, compliance with

traffic laws is likely to be greater in Finland than in the United States and most parts of

Europe. Speeding convictions are not a signal that one is simply the unlucky driver who is

almost randomly “fished out” from a sea of violators.

Table 2 reports regressions that explain three different measures of trading as a

function of this measure of sensation seeking and a host of control variables. The first

column, which uses probit estimation to study the decision of whether to trade or not, employs

all investors in the sample. The second column employs investors who trade at least once

and uses the natural logarithm of the number of trades over the sample period as the

dependent variable.

Because this sample is censored to exclude those who do not trade, we

use Heckman’s two stage procedure to estimate the coefficients. The first stage obtains a

Mill’s ratio from the probit regression in the first column. The second stage, estimated with

ordinary least squares, adds Mill’s ratio as an additional regressor to obtain consistent

estimates on the remaining variables. The third column uses the Barber and Odean (2000,

Source: Finn’s speed fine is a bit rich, BBC News, February 10 2004. Mr. Vanjoki’s fine was later

reduced by 95% due to a drop in his executive stock option income.

We also used Poisson estimation to obtain coefficients for a regression with the number of trades

(rather than the log of trades) as the dependent variable. The t-statistic on the speeding conviction coefficient

was 5.48.

2001) measure of turnover as the dependent variable.

The coefficients in this column are

estimated with ordinary least squares.

The rightmost three columns report the

corresponding t-statistics for the coefficients. All t-statistics and standard errors in the paper

are robust, in that they are computed using White’s heteroskedasticity-consistent standard

error estimation procedure.

The regressors for Table 2 include the number of ticket convictions as a predictor of

trading activity. As can be seen from the bottom row, this measure of sensation seeking has

coefficients that are highly significant for all of the measures of trading activity. The first

column indicates that the probability of trading increases by 4.7% for each additional

speeding ticket. The second column indicates that the number of trades increases by 9.8% for

each additional speeding ticket. Annualizing the coefficient in the third column (multiplying

by 12) implies that each additional speeding ticket tends to increase turnover by about 3.6%

per year. Starting from the annualized average turnover rate of about 40% per year among

those who traded stocks (Table 1’s monthly average of .019 times twelve, divided by the 56%

who traded stocks), each additional ticket increases turnover by a factor of 1.09. These effects

This is the average of buy turnover plus sell turnover. Buy turnover for a given month is the

investor’s portfolio weighted average of the ratio of shares bought of a stock to shares owned in the stock at the

end of the month (or one if the ratio exceeds one). Sell turnover is the investor’s portfolio weighted average of

the ratio of shares sold of a stock to shares owned in the stock at the beginning of the month (or one if the ratio

exceeds one). We average monthly buy turnover and sell turnover over all months to obtain an investor’s overall

buy turnover and sell turnover ratios. Months for which there is no end of month holding (for buy turnover) or

beginning of month holding (for sell turnover) are excluded from the average. The number of observations for

this measure of trading activity is slightly smaller than the sample for number of trades because of the absence of

computable portfolio holdings.

Although not reported formally, adjusting our turnover measure in each month by subtracting the

average turnover across all investors for that month, before averaging across months, yields approximately the

same results as we report here. This robustness applies, irrespective of whether the subtracted average for the

month equally weights all investors or weights them in proportion to their portfolio value.

As in the log number of trades specification, we first use the Heckman two-stage procedure to

account for self selection in the trading decision. The inverse Mill’s ratio does not significantly differ from zero,

so we only report the results from the more parsimonious OLS specification for observations with strictly

positive turnover. The reported results are very similar to the results from the Heckman estimation.

control for age dummies and dummies for the number of stocks held in addition to the

controls reported in Table 2.

The sensation seeking coefficients for the second and third columns in Table 2 are

highly similar when we run the regressions separately for males and for females. For males,

the coefficients on sensation seeking for the number of trades and turnover regressions are

.084, and .003, while for females, they are .092, and .002, respectively. The probit regression

in the first column has a coefficient on the tickets variable of .033 for males and .067 for

females. All of these coefficients are highly significant.

We obtain similar coefficients on the speeding tickets variable when we run the

regressions in the first two columns separately for buys and sells. For example, the probit

regression in the first column generates a coefficient of .045 (t = 5.75) when the buy dummy

is the dependent variable and .053 (t = 6.78) when the sell dummy is the dependent variable.

The fact that these are similar and that the regression with the buy dummy as the dependent

variable is highly significant dispels the notion that Table 2’s results are driven by asset sales

to finance high fines for speeding.

Another indication that cash needs do not drive the link between speeding tickets and

trading activity comes from regressions based on the type of speeding ticket. In Finland, there

are two types of speeding tickets. Mild violations receive a flat fine and more severe

violations receive a fine related to income. When the Table 2 regressions employ the number

of flat fine tickets as the proxy for sensation seeking, the coefficient is highly similar to the

corresponding regressions in which the number of income-related speeding violations is the

The coefficients for a turnover regression specification that employs dummies for one speeding

ticket, two speeding tickets, and three or more speeding tickets are .003, .008, and .011, respectively. Because so

few subjects have four or more tickets, these coefficients are consistent with the reported regression in Table 2,

which has a .003 coefficient on number of speeding tickets. For the other two regressions as well, we obtain

similar results to those reported in Table 2 when we employ dummies for tickets in lieu of number of speeding

tickets.

proxy for sensation seeking. For example, each additional income-related fine increases the

number of trades by 10.1% while each additional flat fine increases the number of trades by

9.7%.

Greater degrees of sensation seeking should not only be associated with greater

numbers of both mild and severe speeding violations, they should also be associated with

some of the more severe violations being even more severe. Fines for the more severe

violations in Finland, known as “day fines,” are assessed (approximately) as a number of half

days of foregone income. The number of half days assessed, referred to as “days fined,” is

based on the severity of the infraction. The mean days fined, averaged only across day fine

penalties earned by each driver who has earned at least one day fine, has a significantly

positive coefficient when it replaces number of speeding tickets in the Table 2 specifications.

Compared to speeding tickets in the same regression, mean days fined also is a more

significant predictor of the decision to trade and the log of the number of trades when added

as an additional regressor to Table 2’s specifications.

The reported coefficients on the control variables are interesting in their own right and

sensible. Financial wealth, income, and whether one is employed in a finance-related

profession are positively related to trading activity even after controlling for the number of

stocks in the investor’s portfolio. Also, being unemployed is positively related to trading

activity. This may be a retiree effect; it may be that independently wealthy individuals trade

for their own account rather than work; finally, it may simply reflect that those who lose their

jobs have to liquidate their financial wealth to consume.

The gender effect in Table 2’s regression—men trade more—is extremely strong, even

more so for single or widowed men. Moreover, as can be seen in Figure 2, the relationship

between age, gender, and trading remains about the same as in Figure 1, Panel B, even after

controlling for all of the remaining non-gender related regressors in Table 2, including our

sensation seeking proxy. This is to be expected even if the sensation seeking trait fully

explains the relationship between age, gender, and trading activity. Speeding tickets represent

an imperfect proxy for sensation seeking. As a consequence, other sensation seeking

correlates, such as age and gender, would have marginal explanatory power for trading

activity (or any other behavior that true sensation seeking generates).

3.2 Overconfidence, Sensation Seeking, and Trading Activity

Barber and Odean (2001) have argued that the relationship between gender and

trading activity is due to the greater overconfidence of men. We investigate this by

controlling for gender (focusing only on males) and looking at how variation in a direct

measure of overconfidence influences their trading activity. Our analysis also controls for

number of speeding tickets, a sensation seeking proxy, to assess whether overconfidence has

any marginal explanatory power for trading activity.

Overconfidence is derived from the FAF self-confidence scale, which is interpreted by

the FAF as follows:

“A person with a high score believes in himself. He views himself at least

as intelligent as others and believes he will manage in life, if necessary,

even without the help from others. He does not feel nervous or anxious in

social situations; he does not expect others’ approval and is not afraid of

others’ possible critique. A person with a low score is uncertain of

himself and he may hate himself and his outlook. He gives up easily when

facing difficulties and can even blame others for his failures: ‘he has been

given too difficult tasks.’ As a result of lack of self-confidence he feels

himself unsure and anxious in social situations, and can therefore avoid

particular individuals who are self-confident and view him critically.”

The self-confidence measure for an individual is transformed into an overconfidence measure,

which is a residual from a regression that uses controls for talent from the FAF, FTA, and

FCSD datasets. Table 3 Panel A reports the coefficients and test statistics for this regression.

The controls include the regressors from Table 2 (except for number of speeding tickets and

the finance professional dummy) as well as the composite intellectual ability score from the

FAF exam, which measures verbal, mathematical, and logical ability. (Our results are

virtually identical if we enter these dimensions of ability as separate scores in lieu of the

composite score.) Table 3 Panel A indicates that individuals’ self-confidence scores are

somewhat prescient about their future life success. Those who have greater FAF ability

scores, who later in life achieve greater income, marry, and hold down jobs, tend to be the

most self confident.

There also is an age pattern to the exam. As age dummies represent an age range of

the subject in 1997, higher age dummies generally correspond to those who took the exam in

the more distant past (and to a small extent, those who entered military service at a later

age).

Those who took the exam most recently exhibit the greatest self-confidence. One can

only speculate about the reasons for this. On the one hand, it may reflect generational

differences and economic changes in Finland. The successful economic growth of Finland

and the waning influence of the Soviet Union (and later Russia) may have produced ever

growing confidence among army recruits. On the other hand, our sample is filtered for those

who own at least one stock during the 2 ½ years that precede our sample period. This may

select more confident subjects among the very young.

The residual from Panel A’s regression is our measure of overconfidence. The idea

behind this is that self-confidence, as measured by a scale from the Finnish Armed Forces

Entrance to military service generally is from ages 18-20, but is never later than age 28.

leadership assessment, is a combination of talent and overconfidence. Panel A’s regression

controls for talent and the residual represents overconfidence.

The last two rows of Table 3 Panel B provide direct evidence on the joint impact of

sensation seeking and overconfidence on trading activity. Sensation seeking is highly

significant except when measuring whether someone trades or not. The number of trades and

turnover are significantly related to sensation seeking, even after controlling for

overconfidence and the other regressors listed in the table, as well as unreported dummies for

birth year and number of stocks in the portfolio.

Overconfidence also is significantly related to trading (at the 5% level), except when

turnover is the dependent variable. Barber and Odean (2001) found that the turnover of males

exceeded that of females and attributed that to the greater overconfidence of males. However,

within the male sample that took the FAF exam, the sensation seeking proxy appears to be

better at explaining turnover than overconfidence. Since we know that males are more prone

to sensation seeking than females, it is possible that the Barber and Odean (2001) gender

difference in turnover is driven more by the gender difference in the sensation seeking trait

than by the gender difference in the overconfidence trait. One might also argue, as with any

insignificant result, that the lack of a relationship between overconfidence and turnover is

driven by error in measuring true overconfidence. However, if that were the case, it would be

difficult to explain the other highly significant coefficients for overconfidence in the same

table.

Why is it that sensation seeking has little effect on the decision of whether to trade or

not, but overconfidence has such a large effect? One possibility is that sensation seekers

achieve stimulation with each trade; a single trade offers very little stimulus. However, this is

more likely a sample-specific finding: Restricting the sample so that it excludes older citizens

as well as women significantly weakens the predictive power of sensation seeking on the

decision to trade. This occurs even without the addition of the overconfidence variable. It

appears that women and older male investors who receive few speeding tickets do not trade.

The same cannot be said for the relatively younger and exclusively male group who are in the

FAF data sample. Even though their turnover ratios and number of trades are low, these

young males still trade on rare occasions.

4. Summary and Conclusion

This paper has shown that, for some investors, trading is driven by behavioral

attributes. Those who are sensation seekers (as measured by the number of speeding tickets

received) and those who exhibit more overconfidence (as measured by a psychological

assessment of each male entering the armed forces) trade more. The overconfidence proxy is

more related to whether an investor trades at all. The sensation seeking proxy is more related

to the turnover ratio. These findings are derived from a dataset that has several advantages

over those used to study related issues in behavioral finance in the past. It is fairly

comprehensive in the subjects it covers, lacks response bias, and allows us to control for a

number of other variables that might explain trading activity.

As a consequence of these controls, as well as additional tests, we do not believe that

our results on the relation between sensation seeking and trading activity are driven by

investor differences in risk aversion. First, our trading activity regressions control for the

degree of diversification in the investor’s portfolio by employing dummy variables for the

number of stocks held and additionally control for both income and wealth. Second, a proxy

for an investor’s risk aversion—the ratio of his equity wealth to his total wealth—appears to

be unrelated to that investor’s sensation seeking attribute. Therefore, it is not surprising that

employing this risk aversion proxy (in unreported regressions) as an additional control

variable in our trading activity regressions has little impact on the coefficients or test statistics

for the sensation seeking variable.

We have tried to be exhaustive in assessing whether other alternative rational

explanations, besides risk aversion, account for our findings. Examination of these

alternatives goes beyond what is reported in the paper. For example, our sensation seeking

findings might be due to an endogeneity bias if the probability of getting a speeding ticket

(but not the proclivity to speed) is tied to trading activity via a third factor. One such

possibility is that active traders locate in urban areas where speeding enforcement is high.

However, when we run Table 2’s regression separately for urban, suburban, and rural

locations of residence, we find that there is a significant relation between speeding tickets and

trading activity in all areas, including the rural areas. Another possibility is that income is not

properly controlled for with our decile proxies. However, we have run the same regressions

with finer categorizations of income at the extreme tails, as well as direct measures of income.

Alternative income and wealth controls do not seem to affect our results. A third possibility is

that inside information motivates trades and this information is somehow related to the

sensation seeking or overconfidence attribute. However, if this was the case, the trades of

sensation seekers and overconfident individuals would exhibit superior performance. There is

no evidence of this in the data. The weekly and monthly returns of portfolios consisting of

the past buys of investor-groups sorted by number of speeding tickets and overconfidence

quintiles do not significantly differ from the returns of their corresponding sell portfolios.

This absence of performance holds for multiple portfolio formation periods, ranging from the

past week to the past year.

Of course, it is theoretically possible that market making or other shorter-term motives

drive the volume difference between low and high sensation seekers and between the least

and most overconfident investors. We doubt this is true, in part because we obtain strong

results even when controlling for whether the investor works in a finance-related profession.

However, further study of this alternative hypothesis, while beyond of the scope of this paper,

might be an interesting topic for future research—particularly if it can be conducted with high

frequency data.

It also seems difficult to defend against the argument that our measures of sensation

seeking and overconfidence are better labeled by some other psychological attribute or some

mix of other psychological attributes than the ones we highlight here. However, we doubt the

science of labeling psychological concepts is so precise that sensation seeking and

overconfidence would be totally orthogonal to the newly-labeled concepts. We started with

our two concepts because they seemed plausible as explanations for trading and then tried to

find ways of measuring them. For example, to further assess the validity of the sensation

seeking explanation, we analyzed a second metric of sensation seeking that has been explored

in the psychology literature: sports car ownership. Regressions (containing the usual set of

controls) of trading activity on a dummy variable for whether one owns a sports car indicate

that sports car ownership is significantly related to trading activity, albeit to a lesser degree

than the number of speeding tickets. In the end, however, it is not terribly important to defend

the labeling of what we are measuring. If it turns out that speeding tickets and the FAF self-

confidence measure capture other non-neoclassical traits, the key point of this paper is still

valid: that psychological factors, outside the neoclassical realm, drive some component of

trading activity.

Despite the surprisingly strong results here, it is important to emphasize that the

degree of trading activity in financial markets remains an anomaly. We have not calibrated

our findings to suggest that sensation seeking and overconfidence explain a large proportion

of observed trading activity. Rather, what we have learned is that fairly stable behavioral

traits explain some cross-sectional differences in cross-sectional trading activity. This adds to

the evidence suggesting that rational motivations, like rebalancing, cannot explain the volume

of trade, because some of that volume appears to be clearly driven by behavioral motivations.

Whether better measurement of the behavioral motivations we have analyzed or whether

some other behavioral motivation can explain most of the observed trading activity is an open

question for future research.

REFERENCES

Agnew, Julie, Pierluigi Balduzzi, and Annika Sundén, 2003, Portfolio choice and trading in a

large 401(k) plan, American Economic Review 93, 193-215.

Ameriks, John, and Stephen Zeldes, 2004, How do household portfolios vary with age?

Columbia University working paper.

Barber, Brad and Terrance Odean, 2000, Trading is hazardous to your wealth: The common

stock investment performance of individual investors, Journal of Finance 55, 773-806.

Barber, Brad and Terrance Odean, 2001, Boys will be boys: Gender, overconfidence, and

common stock investment, Quarterly Journal of Economics 116, 261-292.

Barber, Brad and Terrance Odean, 2002, Online investors: Do the slow die first? Review of

Financial Studies 15, 455-487.

Ball, S., D. Farnhill, and J. Wangeman, 1984, Sex and age differences in sensation seeking:

some national comparisons, British Journal of Psychology 75, 257-265.

Begg, Dorothy and John Langley, 2001, Changes in risky driving behavior among young

adults, Journal of Safety Research 32, 491-499

Beyer, Sylvia, 1998, Gender differences in self-perception and negative recall biases, Sex

Roles 38, 103-132.

Beyer, Sylvia and Edward Bowden, 1998, Gender differences in self-perceptions: convergent

evidence from three measures of accuracy and bias, Personality and Social Psychology

Bulletin 23, 157-172.

Biaszcynski, A., Z. Steel, N. McCongaghy, 1997, Impulsivity in pathological gambling: the

antisocial impulsivist, Addiction 92, 75-87.

Biais, Bruno, Denis Hilton, Karine Mazurier, and Sébastian Pouget, 2005, Judgmental

overconfidence, self-monitoring and trading performance in an experimental financial market,

Review of Economic Studies 72, 287-312.

Comings, David, 1998, The molecular genetics of pathological gambling, CNS Spectrums 3,

20-37.

Deaves, Richard, Erik Lüders, and GuoYing Luo, 2003, An experimental test of

overconfidence and gender on trading activity, Working paper, McMaster University.

Deaux, Kay and Elizabeth Farris, 1977, Attributing causes for one’s own performance: The

effects of sex, norms, and outcome, Journal of Research in Personality 11, 59-72.

Gervais, Simon and Terrance Odean, 2001, Learning to be overconfident, Review of Financial

Studies 14, 1-27.

Glaser, Markus and Martin Weber, 2004, Overconfidence and trading volume, Working

paper, Mannheim University.

Graham, John, Harvey, Campbell, and Hai Huang, 2005, Investor competence, trading

frequency, and home bias, Working paper, Duke University.

Grossman, Sanford and Joseph Stiglitz, 1980, On the impossibility of informationally

efficient markets, American Economic Review 70, 393-408.

Grinblatt, Mark and Bing Han, 2005, Prospect theory, mental accounting, and momentum,

Journal of Financial Economics 78, 311-339.

Grinblatt, Mark and Matti Keloharju, 2000, The investment behavior and performance of

various investor-types: A study of Finland’s unique data set, 2000, Journal of Financial

Economics 55, 43-67.

Grinblatt, Mark and Matti Keloharju, 2001, What makes investors trade? Journal of Finance

56, 589-616.

Jonah, Brian, 1997, Sensation seeking and risky driving: A review and synthesis of the

literature, Accident Analysis and Prevention 29, 651-665.

Kumar, Alok, 2005, Who gambles in the stock market? Working paper, University of Notre

Dame.

Lichtenstein, Sarah and Baruch Fischhoff, 1981, The effects of gender and instructions on

calibration, Decision Research Report 81-5, Decision Research (Eugene, OR).

Lundeberg, Mary, Paul Fox, and Judith Punćochaŕ, 1994, Highly confident but wrong: gender

differences and similarities in confidence judgments, Journal of Educational Psychology 86,

114–121.

Lundeberg, Mary, Paul Fox, Amy Brown, and Salman Elbedour, 2000, Cultural influences on

confidence: Country and gender, Journal of Educational Psychology 92, 152-159.

Milgrom, Paul and Nancy Stokey, 1982, Information, trade and common knowledge, Journal

of Economic Theory 26, 17-26.

Odean, Terrance, 1998, Volume, volatility, price, and profit when all traders are above

average, Journal of Finance 53, 1887-1934.

Odean, Terrance, 1999, Do investors trade too much? American Economic Review 89, 1279-

1298.

Pavalko, Ronald, 2001, Problem Gambling and its Treatment, Springfield, IL, Charles, C.

Thomas Publisher, Ltd.

Pfleiderer, Paul, 1984, The volume of trade and the variability of prices: a framework for

analysis in noisy rational expectations equilibria, Working paper, Stanford University.

Potenza, Marc, M.A. Steinberg, S.D. McLaughlin, et al, 2001, Gender-related differences in

the characteristics of problem gamblers using a gambling helpline, American Journal of

Psychiatry 158, 1500-1505.

Pulford, Briony and Andrew Colman, 1997, Overconfidence: feedback and item difficulty

effects, Personality and Individual Differences 23, 125-133.

Raylu, Namrata and Tian Oei, 2002, Pathological gambling: a comprehensive review, Clinical

Psychology Review 22, 1009-1066.

Vitaro, F., I. Arnseneault, and R.E. Tremblay, 1997, Dispositional predictors of problem

gambling in male adolescents, American Journal of Psychiatry 154, 1769-1770.

Weiss, 1979, Information aggregation and policy, Cowles Foundation discussion paper #528,

Yale University.

Zuckerman, Marvin, 1994, Behavioral expression and biosocial bases of sensation seeking,

Cambridge University Press.

Zuckerman, M., Eysenck, S., and Eysenck, H., 1978, Sensation seeking in England and

America: Cross-cultural, age, and sex comparisons, Journal of Consulting and Clinical

Psychology 46, 139-149.

Table 1

Variable descriptions and descriptive statistics

Table 1 describes the variables used in this study and provides summary statistics on them. Panel A provides detailed descriptions of the variables used, date or

interval of measurement, and the source for the data used to construct the variable. Panel B reports means, medians, standard deviations, and interquartile ranges

for the variables used in the study. Panel C contains the histogram for the scores reported on the self-confidence measure. Panel D is the correlation matrix for

key variables used in the study. The sample is restricted to drivers in the province of Uusimaa or East Uusimaa who got their AB or B license before July 1,

1997, who owned stocks between January 1, 1995 and June 30, 1997, and for whom there is tax data from 1998.

Panel A. Variable description

Variable Data source Measuring time More details

Age FTA + FCSD Measured at 1997 Determined based on social security code

Male FTA + FCSD Does not change Determined based on social security code

Married FTA / Pop. Register End 1998

Cohabitor FTA / Pop. Register End 1998

Unemployed FTA Year 1999 Drew unemployment benefits for at least one day in 1999

Homeowner FTA End 1998 Declared either real estate or apartment wealth at end 1998

Finance professional FTA End 1998 Empoyment in finance-related profession in 1998

Total income FTA Year 1998 Declared total ordinary income + total capital income from 1998

Value of stock portfolio FCSD 6/30/1997 Market value of stock portfolio

# stocks in portfolio FCSD 6/30/1997 Number of different stock exchange listed stocks

# stock trades FCSD 7/1/1997-11/29/2002 Number of open market trades of stocks

Portfolio turnover FCSD 7/1/1997-11/29/2002 Computed as in Barber and Odean (2001) for stocks

# of speeding tickets FVA 7/1/1997–12/31/2001 Total number of speeding tickets

Self confidence FAF When test taken Psychological test self-confidence scores. The test scores

are (approximately) stanine scores and vary between

1 (lowest) and 9 (highest). 0 denotes an unreliable score.

Ability score FAF When test taken Psychological test ability scores. Each test score combines

results from three separate tests that measure mathematical

ability, verbal ability, and logical reasoning. The test scores are

(approximately) stanine scores.

Explanations of abbreviations:

FTA = Finnish Tax Administration

FCSD = Finnish Central Securities Depositary

FTA / Pop. Register = Tax authorities have obtained the information from the Finnish Population Register

FVA = Finnish Vehicle Administration

FAF = Finnish Armed Forces

Represents one of the following professions (# in the sample): Portfolio manager or professional investor (117), dealer (FX and money market, 47), bank

manager (mostly commercial banking, manager of branch, 297), stockbroker (61), stockbroker or portfolio manager assistant (29), investment advisor (generally

low level, in bank branches, 20), miscellaneous investment baking or other higher level finance professional (68), financial manager (corporation, 45), equity

analyst (33), miscellaneous low level investment banking related job (33), loan officer (commercial banking, 138), retired bank manager (23), CFO (227), analyst

(may be other than equity analyst, 104). The tax authorities do not update the profession information often, as there was very little change in the profession data

between 1998 and 2000.

Panel B. Means, medians, standard deviations, and interquartile ranges of variables

Mean Std. Deviation Percentiles N

25 50 75

Total income, EUR 35,559 73,337 14,888 25,474 40,425 95,804

Portfolio value, EUR 24,620 289,019 221 1,571 7,844 95,804

# stocks 2.0502.54611295,804

# stock trades 14.67 119.4001695,804

Monthly portfolio turnover 0.019 0.056 0 0.002 0.014 90,467

Age 44.40 14.39 32 46 54 95,804

# speeding tickets 0.2000.58600095,804

Ability score 6.425 1.74956812,466

Fraction of sample individuals

Traded stocks 0.5570.49701195,804

Male 0.4710.49900195,804

Married 0.5840.49301195,804

Cohabitor 0.0240.15200095,804

Unemployed 0.0600.23700095,515

Homeowner 0.7590.42711195,804

Finance professional 0.0140.11600091,129

Panel C. Distributions of the self-confidence measure (males only)

This sample Full sample

Stanine score # observations % of reliable scores % of reliable scores Stanine distribution

No reliable result 207

1 (low self confidence) 135 1% 3% 4%

2 203 2% 5% 7%

3 432 3% 9% 12%

4 844 6% 14% 17%

5 1,941 15% 20% 20%

6 1,903 14% 14% 17%

7 2,961 23% 16% 12%

8 3,620 28% 15% 7%

9 (high self confidence) 1,114 8% 4% 4%

Totals 13,360 100% 100% 100%

Average 6.54 5.48 5.00

Panel D. Correlations between key variables

Trade dummy ln(#trades) Turnover Male Age # tickets Self confidence

Trade dummy 1.000 N.A. 0.278 0.146 0.026 0.047 0.100

ln(#trades) 1.000 0.252 0.243 -0.007 0.095 0.058

Turnover 1.000 0.113 -0.076 0.068 0.040

Male 1.000 0.005 0.206 N.A.

Age 1.000 -0.108 -0.020

# tickets 1.000 0.012

Self confidence 1.000

Table 2

Regressions of trading activity on sensation seeking and control variables

Table 2 reports coefficients and robust test statistics for a probit regression (column 1), a

Heckman two-stage regression (column 2, which also reports the correlation coefficient between

the residuals in the two stages), and an OLS regression (column 3). These regressions explain

three measures of trading activity as a function of the number of speeding tickets and a host of

control variables. Income and other socioeconomic data are from 1998. Unreported are

coefficients on a set of dummies for the number of stocks in the investor's portfolio and birth

year dummies. The sample is restricted to drivers in the province of Uusimaa or East Uusimaa

who got their AB or B license before July 1, 1997, who owned stocks between January 1, 1995

and June 30, 1997, and for whom there is tax data from 1998.

Coefficient

-value

Dependent variable Dependent variable

Trade Trade

Independent variables dummy ln (#trades) Turnover dummy ln (#trades) Turnover

(Constant) -0.325 0.603 0.0847 -5.32 4.45 23.29

Total income dummies

Lowest -0.133 -0.105 -0.0037 -6.21 -3.30 -2.21

2 -0.039 -0.047 -0.0012 -1.90 -1.60 -0.79

3 -0.030 -0.077 -0.0004 -1.52 -2.72 -0.23

4 -0.031 -0.021 0.0000 -1.58 -0.75 -0.03

6 0.072 0.128 -0.0004 3.74 4.69 -0.29

7 0.093 0.201 -0.0013 4.74 7.34 -0.89

8 0.127 0.301 -0.0001 6.40 10.79 -0.04

9 0.200 0.454 -0.0016 9.75 15.41 -1.17

Highest 0.394 0.863 0.0041 17.60 25.28 2.94

Financial wealth dummies

Lowest -0.994 -0.740 -0.0189 -48.96 -8.67 -15.04

2 -0.788 -0.735 -0.0149 -42.38 -10.90 -11.85

3 -0.504 -0.645 -0.0084 -28.91 -14.37 -6.68

4 -0.335 -0.526 -0.0013 -19.27 -15.56 -1.07

6 -0.037 -0.168 -0.0012 -2.16 -7.91 -1.20

7 0.064 0.041 0.0015 3.67 2.01 1.54

8 0.193 0.240 -0.0022 4.84 7.70 -2.11

Highest 0.361 0.406 0.0021 12.85 13.05 2.09

Other dummies

Male 0.347 0.762 0.0147 23.08 25.45 14.53

Married 0.029 0.062 0.0038 2.19 3.41 4.56

Cohabitor -0.070 -0.034 0.0016 -1.79 -0.57 0.49

Male * married -0.107 -0.351 -0.0061 -5.55 -13.49 -4.67

Male * cohabitor 0.022 -0.082 -0.0054 0.37 -1.03 -1.21

Unemployed 0.083 0.166 0.0107 4.28 6.11 6.15

Homeowner 0.111 0.094 -0.0061 8.84 4.99 -5.83

Finance professional 0.539 0.426 0.0129 12.11 8.37 5.05

# speeding tickets 0.047 0.098 0.0032 5.75 9.68 5.39

Inverse Mill's ratio 0.476 3.88

ρ 0.358

Pseudo R

0.153

0.236 0.094

Number of observations 90,868 50,713 50,224

Table 3

Regressions of trading activity on both sensation seeking and overconfidence

Table 3 reports coefficients and robust test statistics for regressions. Panel A’s cross-sectional regression

uses ordered probit to estimate talent as the predicted value from a regression of self-confidence (from the

FAF leadership assessment) on control variables that measure success in later life. Overconfidence is the

residual from the regression. Panel B’s probit, 2-stage Heckman (which also reports the correlation

between the residuals of the two stages), and OLS regressions explain three measures of trading activity

as a function of overconfidence, the number of speeding convictions, and a host of control variables.

Income and other socioeconomic data are from 1998. Unreported in Panel B are coefficients on a set of

dummies for the number of stocks in the investor's portfolio and birth year dummies. The sample is

restricted to male drivers in the province of Uusimaa or East Uusimaa who got their AB or B license

before July 1, 1997, who owned stocks between January 1, 1995 and June 30, 1997, and for whom there

is tax data from 1998.

Panel A. Parsing out talent from self-confidence to derive overconfidence

Independent variables Coefficient t -value

Total income dummies

Lowest -0.031 -0.75

2 0.027 0.63

3 -0.061 -1.35

4 -0.128 -2.74

6 0.013 0.29

7 0.102 2.30

8 0.181 4.08

9 0.268 5.70

Highest 0.367 6.75

Portfolio value dummies

Lowest -0.032 -1.10

2 -0.080 -2.25

3 -0.020 -0.56

4 0.002 0.05

6 0.014 0.36

7 0.033 0.85

8 -0.004 -0.05

Highest -0.020 -0.32

Other dummies

Married 0.156 6.37

Cohabitor -0.117 -2.17

Unemployed -0.238 -4.85

Homeowner 0.009 0.43

Ability score 0.119 20.82

Age dummies

23-29 0.420 5.59

30-34 0.358 4.97

35-39 0.129 1.82

40-44 -0.012 -0.16

Pseudo R

0.021

Number of observations 12,379

Panel B. Sensation seeking, overconfidence, and trading activity

Coefficient t-value

Dependent variable Dependent variable

Trade Trade

Independent variables dummy ln (#trades) Turnover dummy ln (#trades) Turnover

(Constant) -0.639 1.502 0.1025 -1.67 2.17 5.21

Total income dummies

Lowest -0.023 -0.203 -0.0077 -0.40 -2.37 -1.48

2 0.075 0.051 -0.0061 1.30 0.59 -1.25

3 0.135 -0.015 -0.0046 2.25 -0.17 -0.90

4 0.029 -0.122 -0.0013 0.48 -1.31 -0.24

6 0.267 0.303 0.0043 4.47 3.13 0.81

7 0.307 0.386 0.0031 5.22 3.95 0.59

8 0.493 0.679 0.0057 7.96 5.97 1.09

9 0.564 0.777 0.0025 8.63 6.37 0.51

Highest 0.868 1.293 0.0178 10.36 8.65 2.95

Financial wealth dummies

Lowest -1.142 -1.010 -0.0039 -21.39 -4.37 -1.12

2 -0.892 -0.923 -0.0164 -16.91 -5.25 -5.02

3 -0.522 -0.812 -0.0126 -9.88 -7.39 -3.86

4 -0.321 -0.588 -0.0044 -5.87 -6.99 -1.29

6 0.023 -0.067 0.0030 0.39 -1.01 0.91

7 0.206 0.157 0.0042 3.20 2.20 1.23

8 0.275 0.317 -0.0189 1.40 2.32 -3.31

Highest 0.362 0.346 -0.0097 2.95 3.17 -1.93

Other dummies

Married -0.092 -0.324 0.0011 -2.74 -6.94 0.39

Cohabitor -0.134 -0.183 -0.0027 -1.80 -1.78 -0.48

Unemployed -0.072 0.050 0.0219 -1.12 0.50 2.71

Homeowner 0.144 0.134 -0.0029 4.77 2.86 -1.21

Finance professional 0.692 0.577 0.0191 6.47 4.60 3.17

# speeding tickets -0.001 0.070 0.0042 -0.05 3.48 3.24

Overconfidence 0.037 0.037 0.0004 4.81 2.93 0.61

Inverse Mill's ratio 0.816 2.45

ρ 0.532

Pseudo R

0.156

Adjusted R

0.154 0.112

Number of observations 11,521 7,359 7,271

Figure 1

The joint effect of age and gender on trading activity and sensation seeking

Figure 1 plots trades and speeding tickets as a function of age and gender. Panel A plots

number of trades from 7/1/1997-11/29/2002. Panel B effectively plots number of trades over

the same period, controlling for income, wealth, and number of stocks in the portfolio. It

reports coefficients from a regression of number of trades on birth year dummies (Females

line) as well as the sum of the former coefficients and the product of birth year dummies and a

male gender dummy (Males line). Regressors for income deciles, wealth deciles, and number

of stocks are also controlled for. Panel C plots the number of speeding tickets from 7/1/1997-

12/31/2001. The sample is restricted to drivers in the province of Uusimaa or East Uusimaa

who got their AB or B license before July 1, 1997, who owned stocks between January 1, 1995

and June 30, 1997, and for whom there is tax data from 1998.

Panel A. Average number of trades as a function of gender and birth year

100

1920 1930 1940 1950 1960 1970

Birth year

Average # of trade

Females Males

Panel B. Marginal effects of gender and birth year on average number of trades with effects of

control variables taken out

-0.5

0.5

1.5

1920 1930 1940 1950 1960 1970

Birth year

Coefficien

Females Males

Panel C. Speeding convictions as a function of gender and birth year

0.1

0.2

0.3

0.4

0.5

0.6

0.7

1920 1930 1940 1950 1960 1970

Birth year

Mean # speeding ticket

Females Males

Figure 2

The joint effect of age and gender on trading activity controlling for sensation seeking

Figure 2 effectively plots the number of trades from 1997-2002 for males and females,

averaged for each birth-year cohort controlling for the non-gender regressors in Table 2. It

reports the coefficients on birth year dummies (Females line) and on the sum of the former

coefficients and the product of age and a male gender dummy (Males line) are given in the

graph. The trades are from 7/1/1997-11/29/2002 on the vertical axis. The sample is restricted

to drivers in the province of Uusimaa or East Uusimaa who got their AB or B license before

July 1, 1997, who owned stocks between January 1, 1995 and June 30, 1997, and for whom

there is tax data from 1998.

-0.5

0.5

1.5

1920 1930 1940 1950 1960 1970

Birth year

Coefficien

Females Males