Title: Why%20Do%20Individuals%20Exhibit%20Investment%20Biases?
1Why Do Individuals Exhibit Investment Biases?
Henrik Cronqvist Claremont McKenna
College Stephan Siegel University of Washington
National Taiwan University International
Conference on Finance December 6, 2012
2Genes and Household FinanceResearch Agenda So Far
More risk
Participate in stock market
Investment biases
Today
Save
Less risk
Dont participate in stock market
Investment styles value vs. growth
Consume
- We use twin study research design to examine, for
each of the above choices, the effects of i)
genetic and environmental influences, and ii) GxE
moderators. - Parallel to molecular genetics and
neuroscientific studies by Camerer, Bossaerts,
Kuhnen, Laibson, Zak, and others.
3Investment Biases
- Long list of investment behaviors that cannot be
explained by standard preferences and belief
formation - Underdiversify
- Prefer local securities home bias
- Avoid realizing losses
- Trade a lot
- Chase past performance
- Prefer lottery-type stocks
- Some of these behaviors have been shown to be
- Wide-spread present also among professional
investors - Related to fundamental psychological mechanisms
- Costly
- But, degrees of behavior vary across investors
4Biological Basis for Investment Behaviors
5Existing Evidence
- Experimental evidence (previous slide)
- Capuchin monkeys exhibit loss aversion
- Capuchin monkeys prefer gambles with good
outcomes framed as bonuses over identical pay-off
gambles with bad outcomes framed as losses - Loss aversion is part of decision-making process
that evolved before humans and capuchins
separated (Chen et al. (2006), Lakshminarayanan
et al. (2011)) - But No empirical evidence on the genetics of
investment biases based on real world financial
decisions and data
6Research Methodology Twin Researc
Identical Twins
Fraternal Twins
Josefin and Elin Nordegren
The Hodgson Twins
7Intuition of Methodology
- Use identical fraternal twins to decompose the
variation in investment behaviors - Identical (monozygotic, MZ) twins share 100 of
their DNA - Fraternal (dizygotic, DZ) twins share on average
50 of their DNA - Twins who grew up in same family share a common
environment - Each twin has his/her individual (non-shared)
environment
If genes matter, then identical twins should be
more similar than fraternal twins in terms of,
e.g., their investment behaviors.
7
8Methodology
- Random effect model with genetic effect a, common
effect c and individual-specific effect e - Covariance structure implied by genetic theory
9Methodology, contd
- Estimate parameters s2a, s2c, and s2e via maximum
likelihood estimation (MLE) with bootstrapped
standard errors - Derive the variance components
A-share genetic component C-share common
environment (parenting) E-share individual
environment measurement error
10Data
- Swedish Twin Registry
- Matched with annual financial data (including
holdings of assets and sales transactions) and
socioeconomic data from Statistics Sweden (1999
2007) - Filters
- At least 18 years old
- Both twins hold some equities (directly or
indirectly) in one year - Average all variables over the years that
individual is in sample
11Measuring Investment Biases
- Diversification
- Number of stocks in portfolio
- Home Bias
- Proportion of equity portfolio in local
(Swedish) equities - Turnover
- Annual sales volume scaled by value of portfolio
at beginning of year - Disposition Effect
- Proportion Gains Realized (PGR) Proportion
Losses Realized (PLR) (Odean (1998)) - Performance Chasing
- Proportion of equities acquired with raw returns
in top two deciles - Skewness Preference
- Proportion of lottery securities (Kumar (2009))
12Evidence from Correlations
13Variance Decomposition
14Variance Decomposition
15Robustness
- Opposite-sex twins
- Model misspecification
- Allowing for negative variance components
- Communication
- Identical twins communicate more with one another
- Financial decisions are influenced by
communication (e.g. Shiller and Pound (1998),
Hong, Kubik, and Stein (2004)) - Sort pairs into 10 communication intensity bins
and randomly drop identical/fraternal pairs until
both types are equally often present per bin.
Estimate model across all 10 bins. - A component is somewhat reduced, but overall,
results are robust. - Equal environments assumption
16Identical Twins
Unrelated Look-Alikes www.francoisbrunelle.com
17Two Additional Results
- Moderators of genetic investment biases
- Behavioral consistency Investment biases and
behaviors in other, non-investment, domains
18Moderating Genetic Effects
- Environment can enhance or reduce the effects of
genetic predisposition - Example Education seems to reduce genetic
variation in health outcomes (Johnson et al.
(2009)) - Examine how years of education interact with the
genetic effects - No significant evidence that years of education
reduces genetic predispositions to investment
biases
19Moderator Years of Education
20Financial Experience
- Does work experience in the financial industry
or, e.g., in a corporate treasury department
reduce genetic predispositions to investment
biases? - Repeat analysis for individuals working with
finance - A components reduced (generally by gt50).
- Experience in the finance industry seems to
reduce the genetic predisposition to investment
biases.
21Behavioral Consistency
- Behavior across different domains is often
consistent - If genetic factors matter, source of consistency
should be genetic - Correlate Home Bias with
- Distance to birthplace
- Indicator whether spouse is from same home region
22Behavioral Consistency
23Conclusions
- A long list of investment biases are human in
the sense that investors are born with
pre-dispositions - ?25-50 of variation explained by genetic
variation - Provide empirical support for evolutionary models
of investment biases (e.g., Brennan and Lo
(2009)) - Education does not reduce the genetic
predisposition to investment biases. But finance
industry experience reduces genetic effects. - Genetic factors influencing investment biases
affect behaviors in other, non-investment,
domains.
24Why Are Investment Behaviors Genetic?Why Are
Some More Genetic Than Others?
- Nature selects behaviors that are fitness
maximizing. - Several recent models show that non-standard
preferences are indeed fitness maximizing e.g.,
Rayo and Becker (2007), Brennan and Lo (2009).
25Public Policy Implications
- Do the results of significant genetic effects
make public policy initiatives irrelevant in the
domain of household finance? - No.
- But policy should be designed assuming that
individual investors exhibit investment biases,
rather than based on a notion that biases may be
easily educated away or otherwise eradicated in
markets (Cronqvist and Thaler (2004)).