On the Origins of Risk‐Taking in Financial Markets

• Date: 01 October 2017
• Published date: 01 October 2017
• Author: SANDRA E. BLACK,KAVEH MAJLESI,PETTER LUNDBORG,PAUL J. DEVEREUX
• DOI: http://doi.org/10.1111/jofi.12521

THE JOURNAL OF FINANCE •VOL. LXXII, NO. 5 •OCTOBER 2017

On the Origins of Risk-Taking

in Financial Markets

SANDRA E. BLACK, PAUL J. DEVEREUX, PETTER LUNDBORG,

and KAVEH MAJLESI∗

ABSTRACT

Financial investment behavior is highly correlated between parents and their chil-

dren. Using Swedish data, we find that the decision of adoptees to hold equities is

associated with the behavior of both biological and adoptive parents, implying a role

for both genetic and environmental influences. However, we find that nurture has a

stronger influence on the share of financial assets invested in equities and on portfolio

volatility, suggesting that financial risk-taking is substantially environmentally de-

termined. The parental investment variables substantially increase the explanatory

power of cross-sectional regressions and so may play an important role in understand-

ing cross-sectional heterogeneity in investment behavior.

INVESTMENT BEHAVIOR VARIES SIGNIFICANTLY across individuals. At the same

time, investment decisions, such as the decision to invest in the equity mar-

ket and the riskiness of portfolios, are similar across generations—parents

who hold riskier financial portfolios tend to have children who themselves

hold riskier financial portfolios.1But why are participation in risky markets

and financial risk-taking correlated across generations? Does this have to do

with prebirth characteristics that are correlated across generations (such as

a genetic predisposition toward risk-taking), or is it that children learn from

∗Sandra E. Black is with the University of Texas at Austin, NHH, IZA, and NBER. Paul J.

Devereux is with University College Dublin, CEPR, and IZA. Petter Lundborg is with Lund Uni-

versity, IZA, and Knut WicksellCentre for Financial Studies. Kaveh Majlesi is with Lund Univer-

sity, IZA, and Knut WicksellCentre for Financial Studies. The data used in this paper come from

the Swedish Interdisciplinary Panel (SIP) administered at the Centre for Economic Demography,

Lund University,Sweden. This paper has benefited from the valuable comments of conference par-

ticipants at the 2015 NBER Summer Institute, the 2015 NBER Behavioral Finance Fall meeting,

the 2016 Western Finance Association annual meetings, the 2016 European Finance Association,

Austin-Bergen-London Workshop, and SoLE/EALE, as well as from participants at seminars at

the Copenhagen Business School, Lund University,Southern Denmark University, The University

of Texas at Austin, and UC Davis. Wethank Nigel Jones Barradale, Henrik Cronqvist, Francisco

Gomes, Camelia Kuhnen, and Stephan Siegel for comments and discussion. We also thank the

Editor, Kenneth Singleton, and two anonymous referees for insightful comments that helped us

improve our paper. Erik Grenestam provided excellent research assistance. All errors are our own.

The authors have read the Journal of Finance’s disclosure policy and have no conflicts of interest

to disclose.

1For example, Charles and Hurst (2003) document similarities in asset holdings between

children and their parents.

DOI: 10.1111/jofi.12521

2229

2230 The Journal of Finance R

parents, and risk-taking behavior is acquired over one’s lifetime?2Is it nature

or is it nurture?

Understanding the intergenerational persistence of participation in risky

financial markets and the riskiness of financial portfolios is important, first

and foremost, because it sheds light on the origins of differences in investment

behavior across individuals. It is also important because of the potential welfare

implications. Historically, risky market investment has had a higher return

compared to safer financial assets such as bonds and money market funds.

Thus, to the extent that participation and risk-taking behavior are correlated

with wealth, these behaviors can exacerbate or mitigate wealth inequality over

time.

To evaluate the role of nature versus nurture, we take advantage of a unique

feature of the Swedish adoption system whereby we observe both the biological

and the adoptive parents of adopted children. We use administrative data on

the portfolio choices of a large sample of adopted children born between 1950

and 1980 merged with similar information for their biological and adoptive

parents, as well as corresponding data on own-birth children. We disentangle

the role of nature versus nurture by looking at how the behavior of adopted

children relates to that of both their biological and adoptive parents. Adoption

allows us to examine the effects of environmental factors in a situation where

children have no genetic relationship with their (adoptive) parents.

We find evidence for important effects of both biological and adoptive parents

on stock market participation, implying a role for both genetic as well as en-

vironmental influences.3However, while we find evidence that both genes and

environment affect portfolio volatility, there is little evidence for any genetic

effect on the share of financial wealth invested in equities. Overall, we gener-

ally find that nurture has a stronger influence than nature on risk-taking by

children irrespective of the precise outcome variable we study. Our conclusions

are robust to a variety of robustness and specification checks.

We also examine the mechanisms underlying the environmental effects that

we find. Importantly, the relationships between the risky share and portfolio

volatility of adoptive parents and their children remains largely intact even

when we control for parental financial and real estate wealth, earnings, edu-

cation, and entrepreneurship status. This suggests that parental investment

behavior is not simply a proxy for other adoptive parental characteristics and

plausibly has a direct effect on child behavior. We also find that the relation-

ships between the risky share and portfolio volatility of adoptive parents and

their children remains strong after controlling for child education and wealth,

indicating that an important mechanism may be parental influence on child

attitudes toward risk. Consistent with this and theories of role modeling, we

2Recent research in gene-mapping suggests that specific genes are associated with risk attitudes

and risk-taking behavior. See, for example, Kuhnen and Chiao (2009).

3We can only distinguish between prebirth and postbirth effects using adoptees, so what we

refer to as genetic or biological effects will also include the effects of the in utero environment.

On the Origins of Risk-Taking 2231

find suggestive evidence that men are more influenced by their adoptive father

and women by their adoptive mother.

While many covariates are strong predictors of stock market participation,

regressions for the share of financial wealth allocated to stocks and for portfo-

lio volatility can have relatively low explanatory power (e.g., Calvet and Sodini

(2014) report an adjusted R2for the risky share of about 12%).4Having estab-

lished that parental investment variables are important predictors of children’s

investment behavior, we also show that the parental investment variables sub-

stantially increase the explanatory power of cross-sectional regressions and so

may play an important role in understanding cross-sectional heterogeneity in

investment behavior.

Our paper relates to an active literature that documents intergenerational

correlations in both risk preferences and asset allocations. Charles and Hurst

(2003), Hryshko, Luengo-Prado, and Sørensen (2011), Dohmen et al. (2012),

and Kimball, Sahm, and Shapiro (2009) document similarities in self-reported

attitudes toward risk across generations. The same tendency in the choice of

assets across generations is documented by Chiteji and Stafford (1999)and

Charles and Hurst (2003), among others.

To distinguish between genetic and environmental determinants of risky

financial behavior, the literature has primarily focused on twin studies.5

Cesarini et al. (2010) and Barnea, Cronqvist, and Siegel (2010) both use differ-

ences between identical and fraternal twins to decompose the cross-sectional

variation in investor behavior. Interestingly, they find little role for shared

environmental factors, suggesting that parental influences on children oper-

ate predominantly through genetic channels. However, the twin approach is

fundamentally different from what we do with the adoption data in that it de-

composes the total variation in asset allocation into genetic and environmental

factors; it does so by making relatively strong assumptions about similarities

in environment across fraternal and identical twins. Our approach studies the

intergenerational association and relies on an entirely different set of assump-

tions. Because of the differences in the methodologies and goals, the approaches

can be seen as complements rather than substitutes.6

Adoption studies have been widely used by economists in other contexts to

determine the relative roles of genes and environment in influencing economic

behavior. Much of this research incorporates information on outcomes of adop-

tive parents but not on the biological parents of adopted children. Thus, it

does not compare the relative effects on adopted children of the behaviors or

outcomes of biological and adoptive parents, but instead compares the corre-

lation of adopted children with adopted parents relative to the correlation of

biological children with their biological parents and infers the role of nature

4Calvet and Sodini (2014) use a sample of twins and find greater explanatory power when they

include twin fixed effects.

5Also, Dohmen et al. (2012) use data from the German Socio-Economic Panel Study (SOEP)

and find evidence of environmental influences in the determination of child risk attitudes net of

parental risk attitudes, suggesting that both nature and nurture play a role.

6We discuss how our estimates compare to those from twin studies in more detail below.