Journal of Personal Finance
Volume 10, Issue 2
2011
The Official Journal of the International Association of
Registered Financial Consultants
Volume 10, Issue 2
3
CONTENTS
EDITOR’S NOTES ..................................................................................................8
RESEARCH & THEORY
Applying neuroscience to financial planning practic: A framework and
review .................................................................................................................10
Russell N. James III, Associate Professor, Texas Tech University
This paper presents findings from neuroscience, neuro-finance, neuroeconomics, behavioral finance, and behavioral economics in the context of a
two-system model of human decision-making, labeled as the “rider” and the
“elephant”. The rational “rider” system is characterized by overconfidence
and deficiencies in speed and endurance. The emotional “elephant” system
is characterized by time preference myopia, emotional marker processing,
and loss aversion. Application of this neural model of financial decisionmaking results in a variety of effective and practical suggestions for
financial planners.
Mutual Fund Tax Efficiency and Investment Selection ................................66
D.K. Malhotra, Professor of Finance, Philadelphia University
Rand Martin, Associate Professor of Finance, Bloomsburg University of
Pennsylvania
C. Andrew Lafond, Assistant Professor of Accounting, The College of New
Jersey
We examine six factors that may be important when looking for a more tax
efficient mutual fund. We consider the pre- and post-liquidation bases for
over 4,000 mutual funds in logical groupings. Our results for turnover show
that its effect on tax efficiency depends upon conditions in the securities
markets. A falling market leads to greater tax efficiency due to security
sales at depressed prices. We have a similar finding for expense categories
probably because increased sales lead to higher expenses. We find greater
tax efficiency if mutual funds have institutional status, no-loads, and no
12b-1plan.
4
Journal of Personal Finance
Student financial counseling: An analysis of a clinical and non-clinical
sample ................................................................................................................95
Sonya L. Britt, Kansas State University
John E. Grable, Kansas State University
Julie Cumbie, Kansas State University
Sam Cupples, Kansas State University
Justin Henegar, Kansas State University
Kurt Schindler, Kansas State University
Kristy Archuleta, Kansas State University
The purpose of this study was to determine what factors predict whether
students will seek on-campus peer-based financial counseling. An attempt
was made to determine if students who seek help differ significantly from
students who do not seek help. Findings provide a profile of college student
financial counseling help-seekers. College-age financial counseling help
seekers tend to be older, less satisfied with their income, less
knowledgeable, less wealthy, and more stressed. The results from this study
suggest that college financial counseling centers appear to be on target in
connecting with some of the students they were designed to reach.
Continued efforts to assist students with high financial stress may be a way
to increase financial well-being among college students.
Gender Differences in Risk Aversion: A developing nation’s case .............122
Binay K Adhikari, Graduate Student, University of Alabama
Virginia O'Leary, Professor emerita, Auburn University
This study used Hanna and Lindamood (2004)’s graphic-based survey
instrument to examine whether women who are employed in the Nepalese
banking sector show more risk aversion than men. Women indeed reported
their intention to take less risk and invested less of their wealth in risky
assets than men. However, the difference disappeared after controlling for
other relevant variables, notably their perceived knowledge of financial
markets. Our analyses suggested that women demonstrated more risk
aversion than men because they considered themselves to be less
knowledgeable about financial markets. Our findings support the need to
educate female investors to increase their confidence in their abilities to
succeed in the world of finance.
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
5
CALL FOR PAPERS
JOURNAL OF PERSONAL FINANCE
(www.JournalofPersonalFinance.com)
OVERVIEW
The new Journal of Personal Finance is seeking high quality
manuscripts in topics related to household financial decision
making. The journal is committed to providing high quality article
reviews in a single-reviewer format within 45 days of submission.
JFP encourages submission of manuscripts that advance the
emerging literature in personal finance on topics that include:
-
Household portfolio choice
Retirement planning and income distribution
Individual financial decision making
Household risk management
Life cycle consumption and asset allocation
Investment research relevant to individual portfolios
Household credit use
Professional financial advice and its regulation
Behavioral factors related to financial decisions
Financial education and literacy
EDITORIAL BOARD
The journal is also seeking editorial board members. Please
send a current CV and sample review to the editor. JPF is
committed to providing timely, high quality reviews in a single
reviewer format.
CONTACT
Michael Finke, Editor
Email: jpfeditor@gmail.com
www.JournalofPersonalFinance.com
6
Journal of Personal Finance
JOURNAL OF PERSONAL FINANCE
VOLUME 10, ISSUE 2
2011
EDITOR
Michael S. Finke, Texas Tech University
ASSOCIATE EDITOR
Wade Pfau, National Graduate Institute for Policy Studies (GRIPS)
EDITORIAL ASSISTANT
Britini Peoples, Texas Tech University
EDITORIAL BOARD
Steve Bailey, HB Financial Resources
Joyce Cantrell, Kansas State University
Dale Domian, York University
Monroe Friedman, Eastern Michigan University
Joseph Goetz, University of Georgia
Clinton Gudmunson, Iowa State University
Sherman Hanna, The Ohio State University
George Haynes, Montana State University
Karen Eilers Lahey, University of Akron
Doug Lambin, University of Maryland, Baltimore County
Jean Lown, Utah State University
Angela Lyons, University of Illinois
Ruth Lytton, Virginia Tech University
Lewis Mandell, University of Washington and Aspen Institute
Yoko Mimura, University of Georgia
Robert Moreschi, Virginia Military Institute
Edwin P. Morrow, Financial Planning Consultants
David Nanigian, The American College
Barbara O’Neill, Rutgers Cooperative Extension
Jing Xiao, University of Rhode Island
Rui Yao, University of Missouri
Tansel Yilmazer, University of Missouri
Yoonkyung Yuh, Ewha Womans University
Mailing Address:
IARFC
Journal of Personal Finance
The Financial Planning Building
2507 North Verity Parkway
Middletown, OH 45042-0506
© Copyright 2011. International Association of Registered Financial
Consultants. (ISSN 1540-6717)
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
7
Postmaster: Send address changes to IARFC, Journal of Personal Finance, The
Financial Planning Building, 2507 North Verity Parkway, Middletown, OH
45042-0506
Permissions: Requests for permission to make copies or to obtain copyright
permissions should be directed to the Editor.
Certification Inquiries: Inquiries about or requests for information pertaining
to the Registered Financial Consultant or Registered Financial Associate
certifications should be made to IARFC, The Financial Planning Building, 2507
North Verity Parkway, Middletown, OH 45042-0506.
Disclaimer: The Journal of Personal Finance is intended to present timely,
accurate, and authoritative information. The editorial staff of the Journal is not
engaged in providing investment, legal, accounting, financial, retirement, or
other financial planning advice or service. Before implementing any
recommendation presented in this Journal readers are encouraged to consult with
a competent professional. While the information, data analysis methodology,
and author recommendations have been reviewed through a peer evaluation
process, some material presented in the Journal may be affected by changes in
tax laws, court findings, or future interpretations of rules and regulations. As
such, the accuracy and completeness of information, data, and opinions provided
in the Journal are in no way guaranteed. The Editor, Editorial Advisory Board,
the Institute of Personal Financial Planning, and the Board of the International
Association of Registered Financial Consultants specifically disclaim any
personal, joint, or corporate (profit or nonprofit) liability for loss or risk incurred
as a consequence of the content of the Journal.
General Editorial Policy: It is the editorial policy of this Journal to only
publish content that is original, exclusive, and not previously copyrighted.
Subscription Rates:
Individual:
Institution:
$55 U.S.
$98 U.S.
$68 Non-U.S.
$115 Non-U.S.
Single Issue:
$19 U.S.
$25 Non-U.S.
Send subscription requests with complete mailing address and payment to:
IARFC
Journal of Personal Finance
The Financial Planning Building
2507 North Verity Parkway
Middletown, OH 45042
8
Journal of Personal Finance
EDITOR’S NOTES
Individual investors exhibit a number of behaviors that
confound scientists who expect people to make choices that are in
their long-run best interest. The worst behaviors often involve
risk. We seem to have an instinctive negative response to a loss
and avoid many activities that promise to make us better off (from
equity investing to flying) because we follow our emotions more
than our reason. We also tend to focus too much on investment
returns in the recent past, and are often motivated to take risks only
after the value of assets have risen to the point where they are
overvalued (for example housing in the 2000s and internet stocks
in the late 1990s). It seems that our emotions get in the way of a
reasonable approach to personal finance decisions.
This isn't news to anyone who has studied personal finance or
who has counseled families about making choices that give them a
better chance of achieving financial goals. This issue of the
Journal includes the most exhaustive review of an emerging
literature that provides insight into how the brain works when
processing complex financial decisions. After reading the article I
now have a completely different perspective on why we observe so
many puzzling financial decisions and why it can be so difficult to
change our behavior.
In "Applying neuroscience to financial planning practice: A
framework and review," author Russell James explains that we are
all involved in a constant conflict between parts of the brain that
operate more quickly using instinct and emotion, and the slower,
deliberate part of our brains that attempts to impose reason on the
emotional (gut) cognitive responses. This simple but significant
insight of reason attempting to control the much more powerful
emotion and instinct is characterized by the author as an elephant
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
9
being controlled by the rider. The rider may have very clear ideas
about what is rational, but he is no match for the elephant when
there is a difference of opinion.
Recognizing how and why the mind responds to financerelated stimuli can help explain any number of deviations from
what economists term normatively rational behavior. I often
encounter a dismissive attitude toward rigorous economic
theoretical treatment of personal finance questions when normative
theory does a poor job of predicting actual behavior. I don't find
this perspective appealing. If our financial decisions really do
involve a conflict between the rational and the emotional, then it is
important for our rational side to know where to steer the elephant,
and how to keep the elephant from getting distracted.
Understanding how and when our mind is liable to pull us in the
wrong direction gives us a better opportunity to anticipate our
weaknesses and avoid succumbing to temptations that compromise
our long-term goals.
~Michael Finke
10
Journal of Personal Finance
APPLYING NEUROSCIENCE TO FINANCIAL PLANNING
PRACTIC: A FRAMEWORK AND REVIEW
Russell N. James III *
Texas Tech University
This paper presents findings from neuroscience, neuro-finance, neuroeconomics, behavioral finance, and behavioral economics in the
context of a two-system model of human decision-making, labeled as
the “rider” and the “elephant”. The rational “rider” system is
characterized by overconfidence and deficiencies in speed and
endurance. The emotional “elephant” system is characterized by time
preference myopia, emotional marker processing, and loss aversion.
Application of this neural model of financial decision-making results in
a variety of effective and practical suggestions for financial planners.
Introduction
Technological advances in brain imaging and related sciences
have resulted in a variety of new findings uncovering information
about how humans make financial decisions. The task of
translating these neuroscience findings into financial planning
practice holds great promise, but is also highly problematic. The
numerous related studies are scattered throughout academic
journals from disparate fields. Incorporating these findings into
actual financial planning practice requires not only accurate
translation of individual findings, but also the construction of a
practically applicable framework for use in practice. Neither of
these steps is easily accomplished. Popular reports of individual
studies can over-simplify the reality of sometimes complex and
fragile results. The accurate communication of results can still
leave readers without clear guidance regarding practical
*
Russell N. James III, Associate Professor, Department of Personal Financial
Planning, Texas Tech University, Box 41210, Lubbock, TX 79409-1210;
(806)742-5050 x273; russell.james@ttu.edu
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
11
implications. Even when results and implications are both
appropriately described, practitioners may be left with only a
disjointed laundry list of scattered concepts. This paper attempts
to bridge this gap by presenting recent findings within a specific
framework that can be applied in financial planning practice. The
understanding and application of this general decision-making
framework can illuminate human behavior and provide practical
insight for the financial planner. The paper is organized as
follows:
1. Review of neuroscience methodologies
2. Introduction to the dual-self (“elephant and rider”) decisionmaking framework
3. Review of findings in the context of the framework
4. Application of the framework to financial planning practice
Review of neuroscience methodologies
The recent increasing interest in neural components of human
decision-making, including financial decision-making, may be
driven in large part by technological advances in our ability to
measure brain activity. For many generations, the “black box” of
decision-making in the brain was largely impenetrable to scientists.
Decision-making models could be deduced only by observing
actual behavior. But, the mechanics of such decisions were always
hidden. Rare exceptions occurred in patients with damage to
specific brain areas. By studying humans with naturally occurring
brain damage, and creating corresponding damage in animals,
neuroscientists gradually mapped the functions of several brain
regions.
Since then, a variety of technological advances have
dramatically improved the ability to non-invasively observe brain
activation during decision-making processes. One of the earliest
12
Journal of Personal Finance
was the Electroencephalograph (EEG) (Huettel, Song, &
McCarthy, 2008). The EEG records electrical signals generated on
the surface of the brain through electrodes attached to the subject’s
scalp. The EEG is still widely used today. It is particularly
important for its ability to track the timing of brain activation in
micro-seconds. But, because readings are taken only on the
surface of the brain, there is no way to determine exactly which
internal brain structures are being activated. This lack of “spatial
resolution” can be addressed with Positron Emission Tomography
(PET) scans or functional magnetic resonance imaging (fMRI)
(Huettel, Song, & McCarthy, 2008). In a PET scan, subjects are
injected with a small amount of radioactive glucose. As different
regions of the brain are activated, they use this glucose as fuel.
The relative difference in radioactivity of different regions can be
observed by the PET scan. In this way, specific activation of brain
structures can be observed in three-dimensions. Although still
used for some purposes today, PET scanning is now relatively rare
compared with the widespread application of fMRI. Unlike PET
scanning, fMRI requires no radioactive injections.
The basic concept of fMRI is that it attempts to measure
changes in blood oxygenation. It measures this change based upon
the magnetic properties of tissue. When blood without oxygen is
exposed to a magnet it becomes magnetic. When blood with
oxygen is exposed to a magnet it does not become magnetic. By
measuring these differences in magnetism, the fMRI machine can
estimate the relative oxygenation of different parts of the brain.
Because oxygen use is part of neural activity, the fMRI can reveal
areas of the brain that are more or less active during various tasks.
One would naturally think that oxygenation levels would be lower
in active brain areas. Neuronal firing does, in fact, burn oxygen.
But, in response to this small use of oxygen, the body sends a
relatively large amount of oxygenated blood to the area. It is this
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
13
rush of oxygenated blood to the active brain regions that can be
easily seen by the fMRI scanner (Huettel, Song, & McCarthy,
2008).
Like PET scans, fMRI has excellent spatial resolution. Blood
flow changes are measured in small cubed areas called voxels.
These voxels are usually about 3 cubed millimeters (about the size
of a peppercorn) but can be 1 millimeter or smaller. The blood
response, however, is relatively slow, taking about 4 to 6 seconds
for peak differences. So, unlike EEG recordings, fMRI scans
precision is measured in seconds, not milliseconds. (However,
new technology allows EEG recording simultaneously with fMRI
scanning to get the best of both worlds.)
Typical fMRI experiments allow subjects to see a computer
screen while lying in the fMRI machine. Subjects make choices
using buttons or a joystick. During these choice tasks, the fMRI is
constantly collecting images of the brain. Later data analysis can
reveal what regions were particularly active during different tasks
or prior to particular kinds of choices. However, locating relevant
active regions is always a matter of probability, not certainty. A
variety of factors can cause difficulties in the process. The brain is
always active in thousands of different areas. This makes brains
“noisy.” The activation that occurs during a particular task may
relate to that task or it may not. To overcome this, researchers
typically use many subjects and have them repeat the decision task
dozens of times. Ideally, as the number of task-related images
increase the random noise elements wash out, leaving only
consistent, relevant activations. An additional problem is that the
same brain areas can perform many different, seemingly unrelated,
tasks. Thus, it may not always be possible to clearly identify
which function or emotion is involved.
14
Journal of Personal Finance
Although much financial decision-making neuroscience
research involves fMRI analysis, there are a variety of other
methods available. Lesion studies involve experiments with
individuals who have damage to specific brain areas. The function
of these areas can be identified by observing differences between
normal subjects and patients with lesions in certain areas. Animal
studies involving decision making in primates have made use of
single-cell recording where electrodes can record the activation of
a single neuron (Glimcher, 2004).
Other techniques measure brain-relevant physiological
changes outside the brain itself. For example, hormone studies
have used blood draws before, during and after a task to measure
the levels of hormones related to various human emotions such as
stress or attachment. Other studies have introduced hormones to
measure effects on subsequent tasks (Barrazza & Zak, 2009). A
variety of studies have employed measurements of skin
conductance response. Skin conductance response measures
autonomic arousal, reflecting underlying emotional reactions
(Buchel, Morris, Dolan, & Friston, 1998).
As always, new methodologies are continually being
developed in neuroscience. However, several of these have yet to
be used with a specific focus on financial decision making. Thus,
this review contains no studies using magnetoencephalography,
single photon emission computed tomography, or transcranial
magnetic stimulation.
Introduction to the dual-self (“elephant and rider”)
decision-making framework
Brain activation is inordinately complex. The human brain
contains perhaps 100 billion neurons and 100 trillion synapses
(Williams & Herrup, 1988). Thus, any useful model of brain
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
15
activation must necessarily be a simplistic abstraction from reality.
Fortunately, there are some core realities that allow categorization
of brain activity. Camerer, Loewenstein, and Prelec (2004)
explained, “A crucial fact is that the human brain is basically a
mammalian brain with a larger cortex. This means human
behavior will generally be a compromise between… animal
emotions and instincts, and… human deliberation and foresight.”
This suggests the possibility of effectively modeling human
decision-making as the outcome of an interaction between two
different actors.
Such a “dual-self” model differs from the
traditional economic model of a single, rational, utility-maximizing
actor. This simple distinction may be the most critical difference
between standard economic approaches to financial decision
making and neuroscience approaches. Sanfey, Loewenstein,
McClure and Cohen (2006) concluded, “perhaps the single most
important perspective that neuroscience brings is to challenge the
core assumption in economics that behavior can be understood in
terms of unitary evaluative and decision-making systems.” While
this represents a challenge to traditional economic assumptions, a
variety of economists have found such “dual-self” models
particularly useful in explaining human decision-making.
Economic Models
Adam Smith is known as the author of The Wealth of Nations
(published in 1776) and the founding father of economics.
However, Adam Smith’s first book, The Theory of Moral
Sentiments (published in 1758) provided the underpinnings for his
later work. In it, Smith argued that the behavior was the outcome
of a struggle between two internal actors. Smith labeled these
actors the “passions” and the “impartial spectator.” Ashraf,
Camerer, and Lowenstein (2005, p. 131) summarized Smith’s
approach describing, “The passions included drives such as hunger
and sex, emotions such as fear and anger, and motivational feeling
16
Journal of Personal Finance
states such as pain .… The spectator, in contrast, ‘does not feel the
solicitations of our present appetites. To him the pleasure which
we are to enjoy a week hence, or a year hence, is just as interesting
as that which we are to enjoy this moment’.”
In 1981, Thaler and Shefrin introduced the “first systematic,
formal treatment of a two-self economic man” employing the titles
of “planner” and “doer” to represent the two sides. Thaler and
Shefrin (1981, p. 394) explained, “The planner is concerned with
lifetime utility, while the doer exists only for one period and is
completely selfish or myopic.” Similarly, Fudenberg and Levine
(2006) proposed a dual-self model of impulse control. They
described their model, “Our theory proposes that many sorts of
decision problems should be viewed as a game between a sequence
of short-run impulsive selves and a long-run patient self”
(Fudenburg & Levine, 2006, p. 1449). Benhabib and Bisin (2005)
developed a model of consumption-savings decisions in which the
division is between automatic processes and control processes.
They concluded, “Agents have the ability to invoke automatic
processes that are susceptible to impulses or temptations, or
alternative control processes which are immune to such
temptations” (Benhabib & Bisin, 2005, p. 464). In a review of
neuroeconomic theories, Sanfey, et al. (2006, p.111) described a
similar dual system model, “System 1 is automatic and heuristicbased; quickly proposing intuitive answers to problems as they
arise. System 2, which corresponds closely with controlled
processes, monitors the quality of the answer provided by System
1 and sometimes corrects or overrides these judgments.”
Like Adam Smith, other economists have colored their
particular dual self-models with a strong emphasis on emotional
components.
Loewenstein and O’Donoghue (2004, p. 1)
developed “a two-system model in which a person’s behavior is
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
17
the outcome of an interaction between… [an] affective system that
encompasses emotions such as anger and fear and motivational
drives such as those involving hunger and sex… [and a]
deliberative system that assesses options with a broad, goal-based
perspective (roughly along the lines of the standard economic
conception).”
The addition of neuroscience into the economic dual-self
model holds the potential of showing that the models are not
merely theoretically useful, but actually represent distinct
biological systems. It is probably conceptually sufficient to think
of the two systems as consisting of the more central mammalian
brain and the higher cognitive functions of the prefrontal cortex.
The exact location of associated functions generally follows this
model, but is naturally more detailed and specific. A summary of
these locational correlates by Sanfey, et al. (2006, p.112) described
the current state of knowledge,
“There is a general consensus that high-level,
deliberative processes, such as problem-solving and
planning, consistently engage anterior and dorsolateral
regions of prefrontal cortex as well as areas of posterior
parietal cortex. By contrast, automatic processes appear
to rely heavily on more posterior cortical structures, as
well as subcortical systems. Emotional processes, in
particular, seem reliably to engage a set of structures
classically referred to as the limbic system … and several
other areas such as the amygdala and insular cortex.”
Elephant and Rider
Each of the previously discussed economic labels for the two
actors in the dual-self model (e.g., planner-doer, spectatorpassions, affective-deliberative) emphasized characteristics
particularly important to the topic at hand. Rather than employing
18
Journal of Personal Finance
one of these more formal academic labels, Haidt (2006) adopted an
analogy that may be more useful for contemplating practical
applications with clients. Haidt explained, “The image that I came
up with… was that I was a rider on the back of an elephant. I’m
holding the reins in my hands, and by pulling one way or the other
I can tell the elephant to turn, to stop, or to go. I can direct things,
but only when the elephant doesn’t have desires of his own. When
the elephant really wants to do something, I’m no match for him”
(Haidt, 2006, p. 4).
A variety of neuroscience and behavioral results suggest that
this elephant and rider analogy has much to recommend it. It
visually communicates the concept of the dual-self model, and
even some sense of the relative advantages and disadvantages of
each player. At its core, the elephant and rider model is a neural
analogy, corresponding with the description of the human brain as
a mammalian brain with a larger cortex. The rider is rational,
representing the larger human cortex. The elephant, representing
the more central parts of the brain that are shared with other
mammals, is focused only on the immediate emotional impact of
any action. The following section reviews specific neurological
and behavioral findings corresponding with the framework of a
rational rider and emotional elephant.
Review of findings in the context of the framework
The rider
Because the rider represents traditional rational cognition,
there is no need to spend time defining the rider’s core decisionmaking processes. (For those desiring a full elaboration, any
standard microeconomics textbook can provide a range of
mathematical models for rational, utility-maximizing choice.) The
elephant and rider analogy, however, provides additional
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
19
information. As in real life, elephants are much faster and stronger
than their riders. Elephants can lift enormous weight with
relatively little effort. In contrast, work done by the rider is
relatively slow and effortful. Likewise, when compared with the
amazing endurance of elephants, riders are easily exhausted. The
following section reviews three relative imperfections of the ridersystem.
1. Speed deficiency (rider processes are slower than elephant
processes).
2. Endurance deficiency (rider processes are easily exhaustible).
3. Overconfidence (the rider misprojects control).
Speed deficiency (rider processes are slower than elephant
processes)
This deficiency can be most powerfully demonstrated in
individuals with a specific type of brain lesion that essentially
removes the elephant’s emotional input from decision-making
processes.
Antonia Damasio (1996) explained that the
ventromedial prefrontal cortex (VMPC) records linkages between
facts of a particular situation and emotions previously paired with
the situation in memory. For example, when shown emotionally
charged pictures (such as a social catastrophe or traumatic injury),
normal subjects generated strong responses in skin conductance
responses, indicating an emotional reaction. However subjects
with damage to the VMPC region failed to generate any such
changes in skin conductance response (Damasio, 1996). The
emotional input was simply lost.
Essentially, the VMPC serves as the communication bridge
that allows the elephant’s emotions (generated in the more central
parts of the brain) to influence the rider’s rational cognition
(operating in the cortex). Working with patients who have damage
in this VMPC region, Damasio (1996) found that this elimination
20
Journal of Personal Finance
of emotional input impaired only specific types of decision
making. Such patients’ intellect remained normal as measured by
conventional IQ tests. Learning and retention of facts and skills
were also unimpaired, as was the ability to logically solve
problems. However, patients had great difficulty in decision tasks
involving a large number of potentially important considerations
such as future planning or social interactions.
Damasio (1994) provided an example of such behavior in
attempting to schedule an appointment with such a patient. “I
suggested two alternative dates, both in the coming month and just
a few days apart from each other. The patient pulled out his
appointment book and began consulting the calendar… For the
better part of a half-hour, the patient enumerated reasons for and
against each of the two dates: previous engagements, proximity to
other engagements, possible meteorological conditions, virtually
anything that one could reasonably think about concerning a
simple date” (Damasio, 1994, p. 193). Without the emotional
markers that allow people to quickly choose one option over
another, these patients appeared to rely exclusively on the slow,
grinding machinery of raw cognitive processing. In other words,
decisions emanated purely from the rational rider without input
from the emotional elephant. Maruamtsu and Hanoch (2005)
explained that emotional decisions “are fast because they rely on
few cues, thus dispensing with much computational effort.”
According to the “somatic marker hypothesis”, these cues are
stored as emotional markers (Damasio, 1996). When a similar
scenario or scene emerges, the similarity triggers the marker,
which causes the body to physically respond. These physical
changes then influence the resulting decision making process.
One result of the relative speed of these emotional markers
(and the related physical changes) is that emotion-related physical
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
21
changes occur well before any conscious cognitive awareness. In
other words, when developing strategies for complex choice tasks,
there is evidence that the emotionally charged areas of the brain
identify and engage in appropriate strategies even before the
rational side is aware that a strategy is being employed. Bechara,
Damasio, Tranel, and Damasio (1997) reported results from a card
selection game. Subjects selected cards with financial rewards and
penalties from four different decks. Unknown to them, two of the
decks had larger rewards but also much larger penalties and were
thus less desirable. After about the 10th card selection, normal
participants began to generate high skin conductance responses
prior to selecting from the disadvantageous stack, suggesting a
high emotional response to the losing deck. Even after selecting
20 cards, however, none of the participants had any opinion about
the nature of the differences in the decks. It was not until the 50th
card that subjects, on average, began to express a “hunch” as to
which decks were riskier. By the 80th card, many were able to
clearly state why certain decks were undesirable. Apparently, the
riskiness of the decks was being signaled very early by the
emotions, but was not recognized by higher cognitive functions
until much later. Accordingly, patients with VMPC damage never
generated any increased skin conductance responses prior to
selecting from the disadvantageous decks. Neither did such
patients ever express a “hunch” about deck characteristics.
The ultimate concept developed through a variety of such
studies is that quickly and efficiently making potentially complex
decisions requires more than the rider’s pure rational cognition.
Instead, complex decision-making processes must rely on
emotional markers developed by the more central parts of the
brain. As Muramatsu and Hanoch (2005) argued, it is the “fast and
frugal” nature of emotional decision mechanisms that make them
so important and useful. Understanding the relatively slow and
22
Journal of Personal Finance
effortful nature of the rational rider’s processing helps to remove
the notion that the behavioral goal is somehow to eliminate or
quash the emotional elephant’s participation.
Endurance deficiency (rider processes are easily exhaustible)
Not only is the rational rider side relatively slow, but it is also
easily exhaustible. The cognitive resources of the rider side can be
quickly depleted with extended effort.
During the rider’s
interaction with the elephant, the rider can be thought of as a
representation of self-regulation. Self-regulation is cognitively
complex, but largely located in one brain region. As explained by
Banfield, Wyland, Macrae, Munte, and Heatherton (2004, p. 62),
“self-regulation refers not only to executive processes such as
working memory, attention, memory, and choice and decision
making, but also to the control of emotion (covering issues of
affect, drive, and motivation). The primary brain region
responsible for these control functions is the prefrontal cortex .…”
A series of experiments suggests that the capacity of the prefrontal
cortex to control basic emotional affective drives can be
substantially diminished by prior or simultaneous effortful tasks
involving the same brain region. In other words, the ability of the
rider to regulate elephant-side diminishes quickly with repeated or
simultaneous effort.
Roberts, Hager, and Heron (1994) conducted an early
experiment testing the ability of subjects to inhibit a natural
reaction through conscious control.
In this case, subjects
attempted to avoid looking at the side of a computer screen on
which a flashing box appeared.
When subjects were
simultaneously read single digit numbers and required to add them
together, the ability to inhibit the natural eye movement reaction
following the flashing box decreased substantially.
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
23
Shiv and Fedorikhin (1999) conducted a related experiment
with more relevance to real world choice. Subjects were asked to
memorize either a 2-digit or 7-digit number. While maintaining
that number in memory, subjects selected which desert they would
receive for lunch – either a healthy fruit salad or chocolate cake.
The group memorizing the longer number chose the chocolate cake
much more frequently (41% v. 63%). The results again suggest
that the memory tasks partially exhausted the control of the
affective drive for immediate gratification. Ward and Mann (2000)
conducted a similar experiment with undergraduate females
pursuing a calorie restricted diet. During a 10 minute session
participants were asked to sample from a variety of snack bowls
while completing either a simple reaction time task or a
challenging memorization task. The group of dieters engaged in
the memorization task consumed substantially more grams of
snacks than did the other dieters. The authors explain, “the high
cognitive load succeeded in narrowing restrained eaters' attention
away from normally present inhibitory pressures” (Ward & Mann,
2000, p. 758).
This potential for environmental pressures to periodically
exhaust rider control also explains a core phenomenon in negative
addictions and substance abuse: that constant availability is a key
predictor of consumption. Mann (2005, p.924) explained, “alcohol
problems vary with alcohol availability; this body of evidence is
among the strongest bodies of evidence in existence linking health
problems to determinants.” Room, Babor, and Rehm (2005, p. 526)
found that “Drinking and alcohol related problems can be affected
by restriction of the hours and days of alcohol purchasing and of
the numbers and types of alcohol outlets.” Similarly, Ahmed
(2005, p. 11) found “drug availability represents a major risk
factor. Increased drug availability can precipitate the transition to
addiction.” A rational decision not to consume may be fulfilled for
24
Journal of Personal Finance
an extended period of time. But, various factors leading to
periodic exhaustion of rider control, combined with persistent
availability of a temptation, inevitably increase the likelihood of
consumption.
Cognitive exhaustion can also result from the effort involved
in making multiple sequential decisions. This concept of cognitive
exhaustion can be seen in a variety of sales techniques involving
multiple sequential requests. Once earlier or preliminary requests
have exhausted cognitive resources, people tend to fall back on
more emotional or social oriented responses, often increasing the
likelihood of compliance. Examples of this include a sequence of
rejection-then-moderation requests, where a large initial request is
gradually reduced in a series of subsequent requests (O’Keefe &
Hale, 2001). In a similar approach, the “That’s-Not-All” technique
begins with an initial request followed by a series of subsequent
requests that are progressively more desirable (Pollock, Smith, &
Knowles, 1998). Alternatively, the “Foot-In-The-Door” technique
starts with a very small request, followed by more substantial ones
(Burger, 1999). In a review of all of these techniques, Fennis and
Janssen (2010) concluded, “We propose that all sequential request
techniques essentially trigger one underlying psychological
mechanism that accounts for their impact: that of self-regulatory
resource depletion.” Consistent with this notion, a review of 30
years of research on the “Foot-In-The-Door” sequential request
technique found that the approach was more effective whenever
the initial request was more cognitively demanding (Burger, 1999).
Such cognitive exhaustion techniques are effective in a
context where the salesperson’s social pressure creates a default
behavior of complying with a specific request. In the absence of a
simple, socially-preferable default compliance path, however, the
typical result of cognitive exhaustion is inaction. For example,
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
25
Iyengar and Lepper (2000) found that consumers sampling
premium jams purchased 30% of the time when there were 6 jams
to choose from, but only 3% of the time when there were 24 jams.
More relevant for financial planning, Botti and Iyengar (2006)
studied 401(k) participation among 800,000 eligible employees at
657 companies. They found the highest rate of participation at
companies with only two fund choices. As fund offerings
increased, participation rates generally fell. For example, in
companies offering 60 fund options, participation was nearly 15
percentage points lower than in companies with only 2 options.
Overconfidence (the rider misprojects control)
One of the greatest challenges in managing the relative
limitations of the rider side in terms of speed, strength, and
endurance, is that these limitations are persistently misperceived.
Retrospectively, the rider may have a tendency to construct stories
explaining actions in cognitively rational terms, regardless of the
original motivations. This behavior was seen in early studies of
patients following brain bisection where split visual screens could
give instructions exclusively to one side of the brain (Gazzaniga,
1983). For example, the non-language right hemisphere would be
given an instruction (such as “laugh”), followed by a request to
verbally explain why the action was taken (a left hemisphere
function). The language-producing left hemisphere would
construct a rational, but inaccurate, story to explain the behavior
(such as “because you said something funny”). Gazzaniga (1983,
p.536) explained this as “the left hemisphere’s unrelenting need to
explain actions taken from any one of a multitude of mental
systems that dwell within us.”
Although this retrospective evaluation of behavior is of
interest, a more problematic issue is that when unemotionally
predicting the future, people persistently over-project rational
26
Journal of Personal Finance
control relative to elephant-side desires and emotions. People
project that they will prefer things that are good for them in the
future, but just not right now. So, for example, when asked to
choose between a healthy snack and chocolate for delivery in one
week, 74% of subjects in Reed and van Leeuwen’s 1998 study
chose the healthy snack. But, when given the chance to change
their selection at the point of delivery, only 30% stayed with the
healthy choice. Similarly, only 15% of light smokers (less than
one cigarette per day) in high school predicted that they might be
smoking in 5 years, but 5 years later 43% were still smoking
(Johnston, O’Malley, & Bachman, 1993). Haidt (2007, p. 990)
summarizes a wide variety of studies by explaining “people are
generally more accurate in their predictions of what others will do
than in their (morally rosier) predictions about what they
themselves will do.” In a similar vein, Epley and Dunning (2000)
report, “Researchers have repeatedly demonstrated that people on
average tend to think they are more charitable, cooperative,
considerate, fair, kind, loyal, and sincere than the typical person
but less belligerent, deceitful, gullible, lazy, impolite, mean, and
unethical---just to name a few.”
The greatest mechanism by which the rider can influence
future choices is to change the future decision environment,
usually by a form of pre-commitment. However, to the extent that
people mis-project their own future rational control, such devices
will seem unnecessary and undesirable. So, it is not only the
weaknesses of the rider side that create undesirable choices, but it
is also the tendency to ignore these weaknesses when planning for
the future.
The elephant
In contrast to the rider’s cognitive deliberation, the elephant,
representing the more central parts of the brain shared with other
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
27
mammals, weighs only the immediate emotional impact of any
action. There is, of course, an extensive range of factors associated
with these elephant-side processes. However, for purposes of
application to decision making, there are three crucial distinctions
that characterize elephant-system thinking.
1) Myopia (the elephant focuses on now)
2) Somatic marker processing (the elephant activates by
tangible emotional/social imagery)
3) Loss aversion (the elephant fears relative loss)
The following sections review some evidence
implications for each of these three characteristics.
and
Myopia (the elephant focuses on now)
A key characteristic of the elephant-side system is its focus
on immediate, rather than future, rewards. The issue of trading
current for future rewards, or time discounting, has long been of
interest to economists (Samuelson, 1937). In the traditional model,
consumers have a consistent discount rate, modeled as an
exponential decay function. This consistent discount rate means
that all future delays of the same length are treated the same.
Behavioral results, however, do not support this consistency when
immediate rewards are available (Laibson, 1997; Prelec &
Loewenstein, 1991; Thaler, 1981). Typically people will choose
$10 now rather than $11 in a month. But, those same individuals
will prefer $11 in 13 months rather than $10 in 12 months. Both
cases involve a one-month delay, but the willingness to wait differs
significantly depending upon whether or not an immediate reward
is available. One popular way to represent this reality is to employ
a beta-delta (a.k.a., quasi-hyperbolic) model of time discounting
(Laibson, 1997). This beta-delta model represents the interaction
of two time discounting functions, one sharply focused on
immediate rewards, the other representing the time consistent
28
Journal of Personal Finance
rational discounting of earlier economic models. Such a two-part
model corresponds with a dual-self model of decision-making.
McClure, Laibson, Loewenstein, and Cohen (2004)
conducted the first fMRI study attempting to identify whether or
not separate brain regions actually represented the two parts of the
proposed dual-self interactive system. In their study, subjects
repeatedly chose between two different payments. All payoff sets
involved a choice between an earlier (smaller) and later (larger)
monetary payment. In some of the payoff sets the earlier payoff
was immediate. In others, the earlier payoff was in two weeks or a
month. For those payoff sets including an immediate potential
payoff, the fMRI showed much greater activation in the classic
limbic structures generally associated with emotional reward.
(These areas have heavy neuronal connections with the brain’s
dopamine system.) In contrast, the areas involved in higher-level
cognitive functioning were equally engaged in all decision sets.
This is consistent with a dual-self model where one side cares only
about immediate rewards.
This immediacy focused side
essentially doesn’t participate in decisions comparing two future
rewards. Conversely, the higher-level rational side is equally
concerned about all tradeoffs, regardless of when the earliest
payoff occurs. As further evidence of the interacting dual-self
model, McClure, et al. (2004) found that the higher cognitive areas
were relatively more active than the immediate reward areas when
the delayed choice was selected, and relatively less active than the
immediate reward areas when participants chose the earlier payoff.
McClure, et al. (2004, p. 503) concluded that “separate neural
systems value immediate and delayed monetary rewards.”
A later study found the same phenomenon to be true when
the rewards presented were not monetary but “primary”, i.e. fruit
juice or water given to thirsty subjects (McClure, Ericson, Laibson,
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
29
Loewenstein, & Cohen, 2007). Again, the limbic reward area
activation was greater for choices between an immediate reward
and delayed reward than for choices between two delayed rewards.
Higher cognitive areas responded similarly to either type of choice
sets. As before, relative activation of the limbic system as
compared with the higher cognitive system predicted whether the
earlier or later reward would be selected. McClure, et al., (2007, p.
5796), pointed out “Humans share this limbic reward system with
many other mammalian species, none of which respond to costs
and benefits delayed more than a few minutes, except in a rigid,
preprogrammed manner.” This coincides with a dual-self analogy
based upon the shared mammalian brain (“elephant”) and the
uniquely human higher-cognitive faculties (“rider”).
The beta-delta model is not universally accepted among
fMRI studies of time discounting. Although employing a betadelta model, Witmann, Leland and Paulus (2006) found no
significantly greater brain activation when subjects chose
immediate rewards over delayed. However, in that study all
rewards were purely hypothetical, and subjects knew the choices
were not real. Others have argued for a unitary representation of
reward valuation in the brain (Kable & Glimcher, 2007). Carter,
Meyer, and Huettel (2010), in a review of findings from all such
fMRI studies, argue that the evidence supports an intermediate
perspective, validating both views. In essence, this is possible in a
system where a beta-delta type model is engaged in the valuation
process, but the process later ends with a single value signal
representation.
Taking a different approach, Bechara and Damasio (2005)
argue for greater emotional involvement of immediate outcomes
based upon the structure of the previously discussed VMPC. They
explain, “The organization of the VM[PC] region in relation to
30
Journal of Personal Finance
time (i.e., “near future” is processed more posteriorly, whereas
“distant future” is more anterior) explains why information
conveying immediacy trigger stronger somatic [emotional]
responses, and therefore exert a stronger bias on decisions, than
information conveying delayed outcomes.”
In summary, there is neurological support for the use of a
dual-self model with a shared mammalian side responding strongly
to immediate visceral or emotional experiences competing with a
higher cognitive rational side coolly weighing all outcomes.
Somatic marker processing (the elephant activates by
tangible emotional/social imagery)
As discussed previously, elephant-side responses are
dramatically faster than the relatively slow processes of higher
level cognition. This speed comes from the use of emotional
markers. According to the “somatic marker hypothesis,” emotions
attach to particular scenarios or images (Damasio, 1996). With
relatively minimal cues linked to the marked scenario, a complex
assortment of emotions is instantly retrieved and physiologically
experienced (Muramatsu & Hanoch, 2005; Damasio, 1996).
Naturally, these emotions easily attach to scenarios based upon
actual experience. Bechara and Damasion (2005) labeled such
experiences as “primary inducers.” It comes as no surprise that
one’s personal life experiences can dramatically influence
subsequent decision making on an emotional level.
But, of critical importance is the existence of “secondary
inducers” (Damasio, 1995; Bechara & Damasio, 2005). These
secondary inducers include imagination of hypothetical emotional
events. For example, “The imagination of being attacked by a
bear, winning an award, or losing a large sum of money, are also
examples of secondary inducers” (Bechara & Damasio, 2005, p.
340). Emotional imagery, and corresponding emotional markers,
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
31
can thus be generated through the conscious effort of imagination.
Consequently a goal, such as a financial goal, can be consciously
built into an emotional goal through the generation of emotional
imagery.
Such images, however, must be of concrete or tangible
scenarios rather than of abstract concepts. It is instructive here to
conceptualize the “elephant” side as being shared mammalian
components. Other mammals have little capacity for language,
mathematics, or text. Instead, animal cognition likely relates to
images and circumstances, rather than abstract concepts.
The significance of tangibility may also relate to neural
structures. Again, returning to the importance of the VMPC, a
bridging structure incorporating emotional markers into decision
making, Bechara and Damasio (2005, p. 358) explained,
The organization of the VM cortex along the axis of
“concrete/tangible” to “abstract” may also explain why, for
instance, people have an easier time spending money on credit
cards as opposed to spending real money. Similarly, spending
money becomes no object when a disease threatens the life of a
loved one, and so on. This is because credit is more abstract
than money, and money is more abstract than losing a “bond”
from a loved one.
One might even think of tangibility in terms of translating it
to the experience of other mammals. In a laboratory setting
monkeys have been trained to use simple currency markers that
can be exchanged for food (Chen, Lakshminarayanan, & Santos,
2006), suggesting that tangible money is within the realm of
mammalian (at least primate) understanding.
No such
accomplishment has been achieved with analogies to credit cards,
retirement plans, insurance, or other sophisticated financial
32
Journal of Personal Finance
devices. Thus, such abstract concepts are unlikely to automatically
generate emotional meaning unless tied to more tangible
(“mammal-friendly”) imagery.
Perhaps the most plentiful source of tangible emotional
scenarios across mammalian experience is social relationships. In
the natural world, this is no small issue. For most social mammals,
access to food and reproductive opportunities depends largely upon
the individual’s relative social status within the group, and the
group’s relative social standing among other groups. In humans,
social relationships are the most frequent instigators of emotion.
Kemper (1991) explained, “when respondents are asked for
instances in which they experienced certain emotions, invariably
and with very high frequency they report contexts involving social
relations.” Consequently, the most powerful imagined scenarios
influencing elephant-side behavior are likely to be those involving
significant others.
Such social influences can have powerful effects on financial
outcomes. Social pressure has been used with dramatic effect in
microfinance programs in poorer regions of the world.
Microfinance deals mainly with poor borrowers who could not
qualify for conventional loans. Borrowers function in groups, each
being responsible both for their own and for other group members’
payments. Individually, these borrowers may have irregular
sources of income and may lack any meaningful collateral to
protect against default. Yet, the social nature of these obligations
often results in unexpectedly high repayment rates. Smets and
Bähre (2004, p. 216) explained, “Such conventional collateral has
to a large extent, if not completely, been replaced by nonconventional or social collateral. The strength lies in the social
constraint mechanism among participants; they exert pressure on
each other or on specific participants to encourage payments to be
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
33
made on time.” Similarly, Woolcock (1999) provides evidence
that microfinance repayment success is not simply a matter of
lending policies, but is instead dependent on the nature and extent
of social relations among and between group members and staff
members.
Less formal peer relationships can still exert significant
influence, even on more sophisticated financial choices. Duflo and
Saez (2002) studied tax deferred accounts of faculty members at a
large university and found substantial peer effects in investment
decisions. Controlling for other factors, when participation in a
retirement savings program by other members of a department
increased by 1.0 percent, this peer influence, on average, resulted
in a 0.2 percent increase in one’s own retirement savings.
Similarly, when the average share of the contribution invested in
one particular vendor increased by 1.0 percent among other
members of the department, the peer influence alone resulted in a
0.5 percent increase in one’s own usage of that vendor.
In sum, the elephant side appears to process information
through emotional markers attached to images or scenarios. Most
emotional content comes from social connections and
relationships. Scenarios or images that are more tangible (such as
social relationships and physical objects) will have greater impact
than abstract concepts that are not common to general mammalian
experience
Loss aversion (the elephant fears relative loss)
Having an aversion to losses can be perfectly rational.
However, an excessive aversion to losses occurs when the risk of
loss is heavily overvalued compared to an equivalent potential for
gain. (For the remaining paper, “loss aversion” will refer to this
myopic overvaluation of the pain from loss as compared with the
benefit from an equivalent gain.) For example, Benartzi and
34
Journal of Personal Finance
Thaler (1995) showed that when gambles involving possible loss
are presented one at a time, most people exhibit myopic loss
aversion. Thus, most people will not accept a 50-50 chance to gain
$200 or lose $150. This core asymmetry in valuation of equivalent
gains and losses results in a variety of persistent, non-maximizing
choice behaviors. Loewenstein, Weber, Hsee, and Welch (2001)
suggest that this behavior is rooted in emotional, rather than
cognitive, processing. If so, then excessive loss aversion may have
origins in related dual-system neural circuitry.
As mentioned previously, the elephant system represents
shared mammalian features. As such, it is not surprising that fear
of loss appears as the single most powerful emotion for the
elephant. An animal regularly giving precedence to other emotions
over well-founded fear will very quickly become someone else’s
lunch. (This fear may center on the potential for loss of life, loss
of bodily integrity, loss of social standing, loss of resources, or
other losses, but, in most circumstances fear is centered on the
potential for some loss relative to the status quo.) In support of the
shared mammalian nature of this elephant-system characteristic,
Chen, et al. (2006) reported loss aversion bias in Capuchin
monkeys that had been taught to trade small coin-like disks in
exchange for food rewards.
Fear of loss is urgent, powerful, and tends to dominate other
considerations. As Muramatsu and Hanoch (2005, p. 204)
explained, “elicitation of fear prompts agents to focus their
attention on the importance of the incoming stimulus (while
ignoring all other pieces of information).” Fear reactions are
automatic, coming from neural circuitry centered on the amygdala
(Öhman & Mineka, 2001). As discussed above in the somatic
marker hypothesis, this neural circuitry communicates to decision
making processes through the VMPC.
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
35
Shiv, Loewenstein, Bechara, Damasio, and Damasio (2005)
tested the origins of loss aversion with a group of individuals
having damage to the VMPC, amygdala, or other components of
the neural circuitry necessary to engage emotion in decision tasks.
In this experiment, subjects were give $20. In each of 20 rounds
they could decide to play a dollar or keep a dollar. If a subject
played a dollar, the experimenter flipped a coin in front of the
subject. If the coin landed on heads, the subject lost the dollar,
otherwise the subject received $2.50.
The compensation
maximizing behavior was to play every round.
Normal
participants played in 57.6% of the rounds, exhibiting strong loss
averse behavior. In contrast, subjects with damage to emotional
neural circuitry played in 83.7% of rounds, showing less loss
aversion. Although there was no rational reason to do so, normal
subjects were much more likely to play immediately after a win
(61.7%) than immediately after a loss (40.5%). Patients with
damage to emotional circuitry exhibited much less of this nonrational behavior, investing almost identically after a win (85.2%)
as after a loss (84.0%). For normal subjects the likelihood of
playing each round decreased consistently as the game progressed.
Presumably, as normal subjects accumulated more loss
experiences, they became less likely to play. Thus, it appears that
without the emotional elements of experiencing loss, the resulting
choices by patients with neural damage were more rational.
Results from Damasio (1994) support this difference in felt
emotional experiences resulting from a loss. Subjects selected
cards from different decks with various loss and gain outcomes.
To measure emotional, autonomic system activity, researchers
recorded subjects’ skin conductance responses. Normal subjects
responded with high amplitude skin conductance responses
whenever selecting from a deck with greater previous loss
experiences.
VMPC-damaged participants showed no such
36
Journal of Personal Finance
changes in skin conductance responses. Later, Bechara and
Damasio (2005) reported that amygdala damaged patients
performed similarly to VMPC-damaged patients on the card
selection task. These amygdala-damaged patients also failed to
generate high amplitude skin conductance responses when
selecting from a deck with greater previous loss experiences.
De Martino, Camerer, and Adolphs (2010) also tested
patients with damage to the amygdala. Although the patients
retained a normal ability to respond to changes in risk and
expected value, they displayed dramatically lower loss aversion
than normal comparison players. In contrast, in a subsequent test
involving a series of risky choices with no potential for loss, there
were no significant differences between the amygdala-damaged
and control groups.
One fMRI study did not find a detectable BOLD signal in the
amygdala with loss-averse behavior (Tom, Fox, Trepel, &
Poldrack, 2007). However, this may have been caused by the
much smaller and less frequent losses in this study as compared
with the previous examples. Another suggestion is that quick
spiking in the amygdala is less likely to be observed in the fMRI,
which more easily detects the longer processing in structures
receiving input from the amygdala (De Martino, Camerer, and
Adolphs, 2010).
Physiological differences in emotional reactions to equivalent
gains and losses were demonstrated in a study by Sokol-Hessner,
et al. (2009). They found, using normal subjects, that the average
skin conductance response per dollar lost was significantly greater
than the average skin conductance response per dollar gained.
Similar results were obtained by Hochman and Yechiam (2011)
who found that losses led to greater autonomic responses as
measured by pupil dilation and heart rate than did equivalent gains.
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
37
This core issue of loss aversion can be used to explain a wide
range of inconsistent behaviors demonstrated in behavioral studies.
For example, loss aversion can help to explain behaviors such as
the endowment effect (Thaler, 1980). The endowment effect is the
tendency for people to value items more highly once they have
taken ownership of them. Once a person has ownership of an item,
giving that item up involves a feeling of relative loss. In a trade,
the gain may ultimately outweigh the loss, but if equivalent gains
(e.g., the new item) are less important than losses (e.g., the owned
item), the tendency will be to keep the owned item. If instead, a
person is simply choosing between two items not yet owned, the
selection is between two potential gains and does not involve loss
aversion. Kahneman, Knetsch, and Thaler (1990) reported results
where students in every other seat were given university labeled
mugs to keep. All students were later asked to value the mugs.
Those who had been given the mugs to keep valued them, on
average, at twice the price of those who had not received the mugs.
Similarly, Knetsch (1989) reported that when students were given
a coffee mug at the beginning of class and had the opportunity to
trade for a chocolate bar at the end of class, 89% kept the coffee
mug. But, when students were given a chocolate bar at the
beginning of class and allowed to trade for a coffee mug at the end
of class only 10% traded for the coffee mug. Corresponding with
the concept that this phenomenon is driven by shared mammalian
characteristics, a similar endowment effect has also been
demonstrated in chimpanzees, in that case trading peanut butter
and popsicles (Brosnan, et al., 2007).
Loss aversion also relates to the sunk-cost effect. In this
phenomenon, people allow previous unrecoverable expenditures
(i.e., sunk costs) to influence subsequent decisions. If a previous
investment is abandoned, then it must be framed as a loss.
However, if additional resources are put into the same investment,
38
Journal of Personal Finance
then it is still possible that eventually the combined net result will
no longer be a loss. So, for example, Thaler and Johnson (1990)
found, “in the presence of prior losses, outcomes which offer a
chance to break even are especially attractive.” This desire to
avoid accepting a loss can influence an investor’s decision to sell
or hold stocks. Odean (1998), tracking 10,000 brokerage accounts
and 162,948 trades, found that in any one year 14.8% of all gaining
stocks were sold, but only 9.8% of all losing stocks. This occurred
even though such action was detrimental to investor gains. (The
subsequent performance of losing stocks held was 1.06% worse
than the market, while the subsequent performance of winning
stocks sold was 2.35% better than the market.)
In a similar fashion, loss aversion can be used to explain
status quo bias. Status quo bias is the tendency for people to
choose to maintain the current state of affairs, even if it wasn’t
chosen by them. A deviation from the status quo typically brings
with it the risk of loss or gains (compared with the current path).
But, to the extent that losses are felt more strongly than equivalent
gains, the potential gains must be much more substantial than
potential losses in order to motivate a deviation from the status
quo. Thus, individuals may be willing to maintain the status quo
even if the expected outcome is somewhat less advantageous than
another alternative.
Status quo bias is not necessarily an issue of risk aversion –
the status quo may contain a good deal of risk. Instead, the key
issue is using the present path as the comparison point for defining
gains and losses. If the alternative path is selected and it ends up
being worse that the status quo, then the decision results in a
relative loss. But, if the status quo is selected and the alternative
path ends up with a better outcome, then the decision results only
in a missed relative gain. Because an experienced loss is much
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
39
more distressing than a missed opportunity for gain, the status quo
is more emotionally safe.
Kahneman and Tversky (1982) reported that people
subjectively expected this emotional difference. Where stock A
subsequently outperformed stock B, most believed that an investor
who traded stock A for stock B would feel worse than an investor
who decided not to trade stock B for stock A. The missed
opportunity was the same in both cases. However, trading out of
stock A can be perceived as a loss from the prior status quo, where
holding stock B was only a missed gain opportunity.
Nicolle, Fleming, Bach, Driver and Dolan (2011) examined
this link between status quo bias and feelings of regret in an fMRI
study. Using a visual detection task with a default response, they
found subjects reported greater regret when making an error by
rejecting the status quo (default) than when making an error by
staying with the status quo (default). Analysis of fMRI data
indicated that the anterior insula and medial prefrontal cortex
showed increased activation after status quo rejection errors than
after status quo acceptance errors. Both areas are typically
involved in error processing, but the anterior insula is particularly
important for subjective feeling states.
Additionally, other
research has suggested that the insula plays a role in both fear and
anxiety (Shin and Liberzon, 2010). As such, the fMRI results are
consistent with a greater emotional or fear response to status quo
deviation errors than to status quo maintenance errors.
As further examples of status quo bias, Samuelson and
Zeckhauser (1988) had subjects choose between a moderate risk
investment and a high risk investment for hypothetical inherited
money. If the inherited money was already in the high-risk
company, most (56%) opted to leave it in the high risk company.
If the inherited money was already in a moderate-risk company,
40
Journal of Personal Finance
most (63%) opted to leave it in the moderate risk company. This
difference occurred even though the tax and brokerage charges for
changes were inconsequential and the precise nature of the risk
(percentage likelihood of various outcomes) was known.
In the context of insurance selection, Harvard employees who
started when there were few health insurance options tended to
stay with initial choices, even when new options became available.
Employees who began work later, and were thus not subject to this
same status quo bias, were much less likely to choose the older
plans (Samuelson & Zeckhauser, 1988). Similarly, employees
often accept a company’s default retirement plan even when it
leads to poorer investments (Thaler & Sunstein, 2008).
Perhaps more startling, Ritov and Baron (1990) presented
subjects with a vaccination scenario in which their child had a 10
in 10,000 chance of dying from the flu without vaccination. If the
proposed vaccine eliminated this risk, but had a 6 in 10,000 chance
of causing the child’s death, most subjects would choose not to
vaccinate. One possible explanation is that the vaccination had too
great of a chance of killing a child who would otherwise have been
healthy – a clear loss compared to the status quo. The potential for
saving a child with the vaccine might be viewed as a useful
opportunity for a gain as compared to the status quo, but not as
emotionally impactful as the opportunity for a loss. (Additionally,
this scenario could also relate to the time preference of the
elephant side, as death from vaccination would presumably be
more immediate.)
The general bias towards inaction also makes sense in the
context of the somatic marker hypothesis. If a proposed action is
connected with a loss image, then contemplating taking the action
will immediately trigger the loss image and related bodily
responses. But attaching emotion to inaction may be more
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
41
difficult. Inaction is a general default state, not specific to any one
scenario. Perhaps doing nothing cannot be meaningfully attached
to a specific emotion. Thus, loss aversion might be thought of
more specifically as an aversion to identifiable actions that
generate loss.
In an fMRI study, Fleming, Thomas, and Dolan (2010) tested
whether or not status quo preference increased with cognitive
complexity. (Such a result would be expected where a loss-averse
elephant interacts with an easily exhaustible rider.) Using a visual
detection task, they found that subjects tended to favor the default
only when making difficult decisions, but not when making easy
ones. Their neuroimaging analysis suggested that overcoming this
status quo bias in a high complexity scenario required increased
drive from the inferior frontal cortex to the subthalamic nucleus.
In a sense, overcoming the elephant’s status quo bias in a complex
task required special effort on the part of the rider (here
representing activity in the inferior frontal cortex).
In reviewing the various permutations of loss aversion on
human choice, it is clear that the conceptual framing of a scenario
is important. Behavior may change dramatically depending upon
whether or not a choice is framed in terms of potential gain or
potential loss. A classic example of this framing effect was
provided by Tversky and Kahneman (1981). Subjects were asked
to imagine that the U.S. was preparing for an outbreak of a disease
expected to kill 600 people. When the two options were
characterized as either (A) 200 people will be saved or (B) 1/3
chance that 600 people will be saved with 2/3 chance that no
people will be saved, 72% of subjects chose option A. When the
two options were framed as either (A) 400 people will die or (B)
1/3 chance that nobody will die with 2/3 chance that 600 people
will die, only 22% of subjects chose option A. In both cases,
42
Journal of Personal Finance
options A and B were identical (e.g, if 600 were expected to die,
saving 200 is the same as 400 dying). Only the framing of the
options changed. But that framing change led to a strong
behavioral shift. Loss aversion led to a preference for even a small
chance that no loss would be encountered, but only when the
options were described in terms of losses.
Reframing identical gambles as losses or gains was shown to
impact Capuchin monkey trading behavior as well. In Chen, et al.
(2006), monkeys could trade tokens with one of two
experimenters. In both cases monkeys had an equal probability of
getting either one or two apple pieces. The “gain frame”
experimenter always showed monkeys one apple piece. After
receiving the coin, the experimenter would give the monkey one
apple piece and, with 50% probability, also add a second apple
piece. The “loss frame” experimenter always showed two apple
pieces. With 50% probability this experimenter would only give
one of the displayed pieces to the monkey, dropping the other into
an opaque box. All monkeys strongly preferred trading with the
“gain frame” experimenter. (In the immediately prior experiment,
monkeys had strongly preferred the other experimenter, suggesting
the effect was not due to a general preference for one
experimenter).
Application of the framework to financial planning
practice
A persistent and recurring issue in financial planning practice
is the need to successfully motivate clients to take actions in their
own long term interests and to avoid actions that sabotage their
long-term goals and plans. The rider-elephant framework can be
used to create and analyze specific strategies to accomplish this
sometimes daunting task. Implementing this framework means
dropping the notion that clients are always perfectly rational actors
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
43
who will persistently do what is in their long-term interests.
Instead, client actions and decisions are understood as the outcome
of an interaction between a rational rider side and an emotional
elephant side. The rational rider side characteristics are speed
deficiency, endurance deficiency, and overconfidence.
The
emotional elephant side characteristics are myopia, somatic marker
processing, and loss aversion. Discussed below are a variety of
potential strategies resulting from the application of this
framework.
Use the outsider’s advantage
Financial planners begin with an advantage that clients don’t
have. This advantage is separate from any advanced technical
knowledge the planner has acquired. Planners have a major
advantage in identifying appropriate courses of action based
simply on the fact that they are not the clients. Each of us will
have elephant-system involvement in our own losses. Each of us
will have elephant-system temptations for our own immediate
consumption. Each of us will be emotionally influenced in our
decision for ourselves. However, these elephant-system influences
have little effect on our decisions on behalf of someone else. It is
easy for us to remain emotionally detached regarding someone
else’s loss. We can be completely objective and rational about
another person’s temptations. This is not to say that planners
should be unempathetic. The point is that planners are in the
rationally advantageous position of being removed from the first
person experiences. In other words, you are a better planner for
me than I am simply by virtue of the fact that you are not me.
Taking advantage of this position, planners not only can reduce the
negative influences of elephant excesses, but can also recognize
the potential fragility of a client’s stated rider-system intentions.
Fortunately, the rider system’s persistent overconfidence does not
apply to our evaluations of others. Planners can design for
44
Journal of Personal Finance
scenarios that will help clients even in moments of cognitive
exhaustion and temptation.
Seek illiquidity in investments
Seeking illiquidity, of course, contradicts all traditional
assumptions about investment choice. In traditional finance,
illiquidity is undesirable and requires offsetting compensation.
However, in the context of the elephant-rider interaction, illiquidity
can be extraordinarily valuable. Accomplishing long-term goals
requires that invested funds not be consumed prematurely. But,
highly accessible funds are a constant temptation for the elephantsystem, which is always focused on immediate, emotionallyinteresting experiences. Given that the rider system is easily
exhaustible, constant availability of the elephant’s temptation
increases the likelihood of large and small unplanned expenditures.
Two sources of illiquidity will help reduce elephant temptation.
First, delayed access makes funds less tempting. The elephantsystem is intently focused on immediate experience. Funds that
require administrative delays of several days to access, perhaps
with a series of forms to be completed, are of little interest to the
elephant. Second, attaching losses to withdrawals can also protect
against elephant intrusion.
One of the elephant’s core
characteristics is a fear of losses. Attaching an immediate loss to
the act of withdrawing funds prevents the elephant from being
tempted. The immediate emotional experience of taking an
unnecessary loss is not attractive to the elephant.
Consider how common it is for people to build wealth in
illiquid assets. The personal residence (30% of total wealth),
closely-held businesses (18%), other real estate (11%), and
retirement plans (9%) are dominant sources (68%) of household
wealth (Wolff, 1998). These assets make up an even larger portion
(83%) of all household wealth for those outside the top 20% of all
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
45
wealth holders (Wolff, 1998). Yet, none of these assets can be
quickly liquidated without taking major penalties (at least until
certain ages with retirement plans). None of these assets generate
elephant-system temptation to consume because of their illiquidity.
Laibson (1997) goes further to suggest that financial market
innovations that increased liquidity are partially to blame for the
decline in the U.S. savings rates.
Seek illiquidity in personal debt
When financial products allow immediate access to
borrowing against otherwise illiquid assets, such as the personal
residence, the advantage of illiquidity is damaged and temptation
for immediate consumption grows. As Laibson (1997) explains,
By enabling the consumer to instantaneously borrow
against illiquid assets, financial innovation eliminates the
possibility for partial commitment. This has two effects
on the welfare of the current self. First, the current self no
longer faces a self-imposed liquidity constraint and can
therefore consume more in its period of control. Second,
future selves are also no longer liquidity constrained and
may also consume at a higher rate out of the wealth stock
that they inherit. (p. 645)
When personal credit is instantly available for use in
consumption, it can become a persistent attractive temptation for
the elephant system. For some clients, reducing the total number
of credit cards and lines of credit can be a positive step in this
direction. However, so many sources of instant credit exist that
this may be of only limited help.
A more effective approach can be the use of a security freeze.
A security freeze (a.k.a. credit freeze) prevents access to the
consumer’s credit file. No new credit lines of any source will be
46
Journal of Personal Finance
available while the freeze is in place (although current credit lines
can still be used). The access can be lifted by using a PIN or, if the
PIN is lost, by mailing in proof of identity. For clients needing
intervention with ongoing consumer credit use problems, consider
keeping half of the PIN in your office and half with the client. Or
in extreme cases consider intentionally destroying the PIN, so that
any new access to credit requires the arduous process of gathering
and mailing in a variety of personal identification documents. The
delay created by a security freeze is not problematic for planned
credit transactions such as the purchase of a house or car. But,
especially where access to the PIN requires additional steps, it
prevents immediate elephant-side temptation for new, easy-credit
expenditures and additional credit cards. Further, such a technique
may be especially helpful in the case of married couples where one
partner has substantially greater problems with overusing
consumer credit.
In some cases, the rider-side will tend to over-project future
control and may not find these barriers necessary. A more
palatable approach can be to argue for protection against identity
theft. Security freezes prevent anyone, regardless of the amount of
detailed information they have about a person, from applying for
credit. Indeed, a security freeze where the client does not have
access to the full PIN would prevent even a well-informed burglar
from being able to engage in identity theft. The use of an identity
theft argument to justify the security freeze has the added
advantage of protecting the client from an elephant-motivating loss
scenario.
Seek imagined illiquidity
Real illiquidity can be imposed by contractual agreements
such as those discussed above. However, imagined illiquidity can
be created by the use of mentally separate accounts for saving and
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
47
spending. The classic economic assumption is that wealth is
fungible, i.e. all money is treated the same.
Bertrand,
Mullainathan, and Shafir (2004) found that this did not fit with
common behavior. They explained, “Also, contrary to standard
fungibility assumptions, people compartmentalize wealth and
spending into distinct budget categories, such as savings, rent, and
entertainment, and into separate mental accounts, such as current
income, assets, and future income” (p. 420). Obviously, setting up
three identical bank accounts for current spending, medium term
savings goals, and retirement savings actually adds no real
liquidity constraints.
It does, however, substantially affect
spending behavior. The tendency to spend money out of these
different accounts will differ widely (Betrand, et al., 2004).
One advantage of separate accounts is that they can alert the
rider side to stop spending. If all money was held in a single
account, there is no natural stopping point to consider the
cumulative effect of immediate spending. Mental accounts,
however, create stop signs that trigger rider attention. Also,
separate accounts allow for an attachment of different emotional
markers to different pools of funds. If the elephant has attached a
somatic (emotional) marker to the tangible savings goal associated
with the account (such as a car or vacation) and if breaking into the
account is seen as losing that emotionally-salient scenario, the lossaverse elephant will be less tempted. Even a piggy bank can serve
as an excellent barrier to liquidity if it is mentally categorized for a
particular expenditure. (Consider the visceral impact on the lossaverse, emotional elephant of having to actually break the piggy
bank!)
In a similar way, expenditure budgets create stop signs that
alert the rider to intervene in spending behavior within a particular
category. An additional $20 expense in an exhausted budget
48
Journal of Personal Finance
category can be reframed as threatening the stability of the budget,
the plan, and associated long-term goals.
Use “opt out” defaults
It is possible for financial planners to pre-select a wide
range of appropriate choices for a client’s comprehensive financial
plan and deliver it as one completed package. However, without
client participation in building each component, there is a danger
that clients will not see the plan as theirs. If the plan is not
mentally “owned” by the client, future commitment to plan
guidelines may be at risk. Conversely, clients who are heavily
involved in making detailed choices for every point in the planning
process will be more likely to take ownership of the plan as
something they have built. The advantage of this approach,
however, is offset by two problems. First, the process of
considering a long range of cognitively challenging questions is
exhausting for clients. Second, financial planners are experts who
should provide guidance to clients, not simply list all available
options.
Providing wisely-selected default options allows the planner
to influence outcomes while still maintaining client participation.
The power of default options comes from the elephant-side status
quo bias. As discussed above, the elephant prefers the status quo
because deviations are seen as having potential for loss. To the
extent that the default options are viewed as the standard, normal
choices, i.e., the status quo, then deviations will require special
rider-side cognitive effort to overcome the elephant’s loss
aversion. As discussed previously, because the rider is easily
exhausted, default options will become progressively more likely
in a long series of questions.
Perhaps the most powerful version of a default is the “opt
out” provision. In this scenario a single, implicitly-recommended
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
49
path is provided, and clients are given the choice to opt out of that
path. This sends the strongest signal that the provided path is the
status quo. The requirement to opt out builds the perception that
such actions are a bold violation of the accepted status quo. This
perception can be further heightened by the use of a detailed opt
out process highlighting the potential losses resulting from
rejecting the standard approach. Requiring physical effort such as
filling out a form may also increase the tendency to accept the
default (Samuelson & Zeckhauser, 1988).
Johnson, Hershey, Meszaros, and Kunreuther (1993) studied
this phenomenon by providing an auto insurance feature (regarding
the right to sue for pain and suffering) as either an “opt out” from
the full policy, or an “opt in” from the limited policy. Over half
(53%) choose to keep the right by not opting out of the full policy,
but less than a quarter (23%) chose to add the right by opting in
from the limited policy. (A common implementation of this can be
seen when renting a car as contracts typically require that
insurance be refused, and that the refusal be separately signed.) In
another example, organ donation rates across countries increase by
almost 60 percentage points when comparing an “opt in” nation
with an “opt out” nation (Gimbel, Storsberg, Lehrmen, Gefenas &
Taft, 2003; Johnson & Goldstein, 2003).
A more subtle form of creating defaults is by adding an
option that is similar to, but clearly inferior to, the desired choice.
Consumers are attracted to the value represented by easy relative
comparison through a process known as anchoring (Simonson &
Tversky, 1992; Simonson, 1999).
Consider the following three approaches to encouraging a
client to include disability insurance as part of a financial plan.
Using knowledge of the elephant and rider allows an evaluation of
the relative likelihood of success.
50
Journal of Personal Finance
Approach 1: Client is given full information including
proposals from 10 companies. Each includes pricing for
every available short-term and long-term alternative
including income replacement percentages from 40% to
70%, varying delays until payments start from two weeks
to 24 months, varying lengths of total coverage, several
alternative cost of living adjustment riders, varying
definitions of disabilities and when each will result in
payments, and the relative tax advantages and
disadvantages for company paid or employee voluntary
approaches.
Comment: Because there are so many options, the risk of
choosing the “wrong” one is high. The elephant would
prefer to avoid the embarrassment and perceived loss of
making the wrong choice by simply postponing the process
(indefinitely) and sticking with the status quo (doing
nothing). This choice scenario requires enormous rider-side
effort. Even with the best of intentions, given the volume
and complexity of required choices the rider’s endurance
will be quickly exhausted allowing the elephant’s status
quo bias (doing nothing) to prevail.
Approach 2: The client is provided with five options
representing different levels of protection with associated
prices, one of which is recommended by the planner.
Comment: The cognitive complexity of the decision is
substantially reduced making it manageable for the rider.
The elephant side is less concerned about making the
wrong choice as there are fewer options and the
recommendation provides a comfortable default.
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
51
Approach 3: The client is provided with a recommended
plan, with the knowledge that features could be adjusted if
any seem inappropriate. If the client prefers not to have
disability insurance, the planner requires a signed “release
of liability” form. The planner explains that when families
and children are left with no income, lose their house to
foreclosure, and have nowhere else to turn, they may turn
to a plaintiff’s lawyer to sue the financial planner for
malpractice. The planner explains that the form indicates
that she has tried to the best of her abilities to add disability
insurance to the client’s plan and that the client is
knowingly refusing that protection.
Comment: The cognitive complexity of the decision is
reduced to a simple yes or no. The client knows that
features could be adjusted, but has no specific prompts to
do so on any particular issue. Because of the “opt out”
structure, doing nothing requires positive action. Opting
out requires reviewing a separate document that is
intentionally designed to alert the elephant to emotionally
relevant loss. The action (signing the “release of liability”)
is associated with concrete loss images (children losing
their house and income) making it odious to the elephant.
Build imagination capital
As discussed previously, the elephant system processes
through somatic markers, i.e. attaching emotions to images. When
working with a client’s long-term goals it is important to make
sure that these goals are translated into elephant-friendly images.
Understanding the elephant system as shared mammalian
characteristics may help to clarify what concepts are or are not
relevant. Abstract financial and mathematical concepts are not
relevant here. For example, instead of just setting an abstract goal
52
Journal of Personal Finance
- like 80% of current income available at age 65 from retirement
assets - have clients identify potential emotional experiences
associated with the goal. Ask clients a question about what they
envision doing after their career. Where do they want to go? With
whom do they want to spend time? What activities would they like
to do together? Through this process, the planner can identify
emotionally-relevant goals. More importantly, the planner actually
helps clients build those goals, not by interjecting ideas but by
asking questions that cause clients to construct their own imagery.
Nobel prize winning economist Gary Becker used the term
“imagination capital” to explain this process (Becker, 1996, p. 11).
Becker and Mulligan (1997) presented an economic model based
on the standard idea that consumers maximize their lifetime
consumption experience. So, a simplified model of consuming
now (period 0) and later (period 1) is that people maximize the
function
V = f0(c0) + β(S)·f 1(c1)
In words, the formula indicates that consumers value both
current consumption and future consumption. However, future
consumption is less valuable than current consumption by the
factor of β. This β factor is the traditional time discounting of
standard economics.
Contrary to the traditional economic
approach, Becker and Mulligan (1997) suggested the time discount
rate could be altered through spending effort (S) on becoming more
future-focused. In other words, a consumer could “make future
pleasures less remote by spending resources (S) on imagining
them” (p. 734). Becker and Mulligan (1997) expanded on this
process by explaining,
“How can a person improve his capacity to appreciate the
future? What exactly is S? First, S is partially determined
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
53
by time and effort spent appreciating future pleasures.
While forming a mental picture of one’s future pleasures
may not be incredibly difficult, the process of anticipation
is not merely one of image formation but also one of
scenario simulation. Even image formation may not be
cheap because images of future pleasures have to be
repeatedly refreshed in one’s mind in order to compete with
current pleasures” (p. 734).
Far from being a fanciful notion irrelevant to hard financial
choices, Becker and Mulligan (1997) suggested that the most
central concept in all financial decision making, the rate of future
discounting, resulted from imagination. Further, the effort spent
imagining these future experiences is cumulatively valuable,
building up in the form of imagination capital.
From the perspective of the elephant-rider model, the ideal
way to develop elephant support of a long-term financial plan is to
(1) associate relevant emotional imagery with the goals, and (2)
conceptualize deviations from the plan in terms of risk of loss of
the emotional goals. The elephant side is focused on immediate
emotional experience, but the immediate emotional experience
from (potentially) losing a closely held future dream is quite
unpleasant. Conversely, if the goal is simply to hit some abstract
numerical target for lifetime income smoothing the elephant will
be unmoved.
Build social networks
As discussed previously, social connections and peer
influences are strongly felt by the elephant side. Strong social
relationships between clients and planners, and between clients and
office staff, can help to emphasize the elephant’s perceived social
importance of staying on track with planning goals. Where
possible, bringing together groups of clients with similar goals in a
54
Journal of Personal Finance
social setting may help to provide mutual encouragement. Similar
techniques are commonly used in nonprofit fundraising where
social events and board meetings are valued for building a
community of like-minded supporters. In the fundraising context,
the natural reinforcement from peer group influence encourages
donations more than any appeal from a development officer ever
could. Likewise, much of the success of mutual support groups
like Weight Watchers or Alcoholics Anonymous come from
counteracting temptation with the elephant-side sensitivity to
social approval. Although this sometimes puts the planner in loco
parentis (in the place of a parent), it can make a difference to the
emotional elephant that someone else is monitoring whether or not
the client is successfully pursuing long term goals.
Remove felt losses from good choices through precommitment
A dominant characteristic of the elephant system is fear of
loss. To encourage positive choices it is important that these
choices, as much as possible, not be associated with the felt
experience of loss. Otherwise, the elephant will strongly resist
taking these loss-associated actions. Removing felt losses can be
accomplished through pre-commitment that removes felt losses
associated with future choices or through reframing scenarios from
a loss perspective to a non-loss perspective.
One example of removing felt losses through precommitment is the Save More Tomorrow program (Thaler &
Benartzi, 2004). In this plan, people are allowed to commit in
advance to allocating a portion of their future salary increases
toward retirement savings. The timing and framing of this
decision avoids any elephant-system interference. First, the
decision is about two future options (saving more in the future
versus not saving more in the future). This avoids the hyperbolic
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
55
discounting problem arising when one of the options is immediate
consumption. Additionally, the decision poses no risk of loss from
the status quo. Only gains, i.e. future salary increases, are being
allocated here. The results are dramatic. Thaler and Benartzi
(2004) reported that program participants increased their savings
rate from an initial 3.5% all the way to 13.6% over the course of 40
months. A similar device was used by Chiou, Roe, & Wozniak
(2005) for low income families by using pre-commitments to put a
portion of the tax refund check into a savings accounts. In this
pilot program, participants saved 236% more than they projected
they would prior to hearing about the program.
The pain of savings (lost consumption) was never felt by the
elephant side in these programs because income never dropped
from the status quo. Also, in the Save More Tomorrow program,
because the savings deductions were made by the employer,
transfers occurred automatically. This concept of automatic
savings was made popular by David Bach’s (2005) Automatic
Millionaire series. The basic idea of making savings occur
automatically so that it does not require recurring conscious
decisions to postpone consumption fits well within the elephantrider framework. In essence, the elephant does not feel the lost
consumption due to the automatic nature of the savings. The initial
pre-commitment to begin automatic savings in the future is of less
interest to the elephant because it is a choice between two future
outcomes.
Remove felt losses from good choices through reframing
There are, however, circumstances where avoiding felt losses
is more difficult and pre-commitment may not provide a solution.
A particular challenge for financial planners is helping clients deal
with the experience of market volatility. Over the long-term,
investment in more volatile assets tends to produce greater returns.
56
Journal of Personal Finance
However, volatility may result in some interim loss experiences.
The elephant-system reaction to these losses may be extreme, often
leading to market withdrawal at precisely the wrong times.
A series of results suggests that these loss experiences are
subject to reframing based upon broad or narrow bracketing (Read,
Lowenstein & Rabin, 1999). Bracketing is the idea of thinking
about experiences separately (narrow bracketing) or in large sets
(broad bracketing). When investors focus on returns in narrow
brackets, they will inevitably experience many felt losses. In
particular for investments in publicly traded equity, many
individual days will produce negative results. Even on overall
positive days, many individual stocks or funds within a portfolio
will have losses. The probability of felt losses diminishes to the
extent that investors think in broader brackets. Today’s loss in one
stock is not felt as a loss if the investor focuses on today’s positive
results for the entire portfolio. Today’s loss in the entire portfolio
is not felt as a loss if the investor is focused on the overall positive
returns for the last 30 days. This month’s loss on the entire
portfolio is not felt as a loss if the investor is focused on the overall
returns for the last 12 months. As investors focus on larger and
larger brackets, the likelihood for an emotionally felt loss typically
becomes less. Conceptually, the only bracket that should matter to
an investor is the one starting at the introduction of new savings
and ending at the planned time of expenditure. However, in the
midst of a down market, it is difficult to maintain this sanguine
view. Indeed, investors would typically do better by simply not
following their returns during the interim period. As Read, et al.
(1999) explained,
Over brief periods, stock prices are almost as likely to fall
as to rise. For loss averse investors, the falls will be
extremely painful and the rises only mildly enjoyable, so
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
57
the overall experience might not be worth undertaking. By
this logic, if people could resist looking at their portfolios
for longer periods—i.e., bracket their investment choices
more broadly—the likelihood that they would see such
losses would diminish, and the clear benefits of stocks
would emerge. (p. 180)
Thaler, Tversky, Kahneman and Schwartz (1997) had
subjects make investment decisions between stocks and bonds
simulating returns that would be experienced based upon making
those allocation decisions twice per quarter, once per year, or once
every five years. Those making reallocation decisions twice per
quarter inevitably invested primarily in bonds, whereas those with
a longer time frame invested mostly in stocks. Reallocation
decisions made by the planner based upon client goals can be an
effective strategy. But reallocation decisions made by the investor
based on recent experiences of loss will rarely be strategically
sound.
Sokol-Hessner, et al. (2009), tested a form of reframing in a
study measuring physiological changes and loss aversion. Subjects
made a series of choices between certain gains and uncertain
outcomes with a potential for loss. In one choice series, subjects
were told to focus on each choice “as if it was the only one.” In
the other, subjects were told to think of all choices together, “as if
creating a portfolio.” The use of the larger “portfolio” bracket
resulted in a dramatic reduction in loss aversion in the subject’s
choices. Further, the experience of losses generated substantially
more skin conductance response arousal than the experience of
gains in the narrow-focus choice series, but not in the “portfolio”
scenario. Even though the choices remained the same in both
scenarios, a simple reframing led to both behavioral and
physiological changes reflecting reduced loss aversion.
58
Journal of Personal Finance
Although planners may not be able to convince clients not to
look at their returns for long periods of time, other forms of
framing using broad bracketing may be effective. During down
periods, planners may point to either the past (returns since
original investment) or the future (time remaining until the money
will be spent) to attempt a broadening of the client’s return
bracketing. Focusing on the future, a planner responding to
distress about a recent market downturn might ask a series of
reframing questions such as: “Were you planning on spending the
money today? Then why do we care what it did this month? We
only care where it will be at the end of the next X years, right?”
Planners may also consider reporting returns on a longer term basis
(e.g., cumulative returns since initial investment) to reduce the
likelihood of showing overall negative returns.
Add felt losses to undesirable choices
Because the elephant side experiences substantial fear of loss,
this reality can be used to discourage negative behaviors that
undermine long-term success. A prime example of adding felt
losses to undesirable choices is intentionally seeking illiquidity as
described above. Characterizing divergence from a plan as the loss
of a cherished and emotionally-salient future dream is another
powerful example. However, the same concept can be used in a
variety of other contexts. Suppose a client has a long investment
time horizon but has no interest in deviating from short-term
secure investments such as bank certificates of deposit. It is likely
that this mismatch of volatility and time horizon is being driven by
underlying fear of loss. Simply pointing to the long-term
difference in overall returns is unlikely to be convincing to the
immediacy-focused, loss-averse elephant side. Instead the planner
could present two plans, both with the same long-term financial
goal. The difference in the two plans is presented as the monthly
amount of savings that must be set aside to reach the goal based
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
59
upon projected returns. The elephant-system clearly experiences
the reality of less money now as compared with more money now.
This reframes the choice in terms relevant to the elephant-system
and encourages the optimal long-term behavior.
For clients with over-spending problems, especially related to
the use of credit cards, shifting to an all cash system will
dramatically influence spending behavior. The “mammalian
brain” elephant-side clearly feels the visceral loss associated with
handing over valuable cash and leaving the store without the
money. Swiping a credit card provides no such sense of loss as
perceived by the mammalian brain. This simple difference has
been shown to dramatically affect spending behavior. Prelec and
Simester (2001) reported, “In studies involving genuine
transactions of potentially high value we show that willingness-topay can be increased when customers are instructed to use a credit
card rather than cash. The effect may be large (up to 100%).”
Similarly, Soman (2003) analyzed receipts from a grocery store
and found the percentage of the total expenditure spent on nonessential treats and luxuries was 27% for those paying with cash
and 43% for those paying with a credit card.
There are, of course, a variety of other examples and
scenarios that one might construct to add felt losses to bad choices
(e.g., the making of a public pledge and the desire to avoid loss of
social standing by breaking the pledge). The goal of adopting the
elephant and rider model, however, is not to construct the perfect
list of techniques. Instead, it is to provide a neurologicallyinformed perspective to aid in the ongoing creation and evaluation
of strategies that will ultimately prove effective for financial
planning practice.
These strategies are made effective by
incorporating the known characteristics of both elephant and rider.
Thus, effective strategies will not seek to consistently overpower
60
Journal of Personal Finance
the elephant. (Such attempts are difficult, exhausting and, over
extended periods of constant temptation, ultimately doomed.)
Instead, effective strategies will seek to harness the elephant. By
creating plans that incorporate known rider weaknesses with
elephant-sensitive incentives, the financial planner can help clients
to accomplish their goals, providing long-term satisfaction to both
elephant and rider.
References
Ahmed, S. (2005) Imbalance between drug and non-drug reward availability: A
major risk factor for addiction. European Journal of Pharmacology, 526, 920.
Ashraf, N., Camerer, C., & Loewenstein, G. (2005). Adam Smith, behavioral
economist. The Journal of Economic Perspectives, 19(3), 131-145.
Bach, D. (2005). The automatic millionaire: A powerful one-step plan to live
and finish rich. New York: Crown Business.
Banfield, J., Wyland, C. L., Macrae, C. N., Munte, T. F., & Heatherton, T. F.
(2004). The cognitive neuroscience of self-regulation. In R. F. Baumeister
& K. D. Vohs (Eds.), Handbook of self-regulation: research, theory, and
applications (pp. 62-83). New York: Guilford Publications, Inc.
Barrazza, J. A., & Zak, P. J. (2009). Empathy toward strangers triggers oxytocin
release and subsequent generosity. Annals of the New York Academy of
Science, 1167(1), 182-189.
Bechara, A., & Damasio, A. R., (2005). The somatic marker hypothesis: A
neural theory of economic decision. Games and Economic Behavior, 52,
336-372.
Bechara, A., Damasio, H., Tranel, D., & Damasio, A. R. (1997). Deciding
advantageously before knowing the advantageous strategy. Science, 274,
1293-1295.
Becker, G. (1996). Accounting for tastes. Cambridge, Massachusetts: Harvard
University Press.
Becker, G., & Mulligan, C. (1997). The endogenous determination of time
preference. Quarterly Journal of Economics, 112(3), 729-758.
Benartzi, S. & Thaler, R. H. (1995). Myopic loss aversion and the equity
premium puzzle. Quarterly Journal of Economics, 110(1), 73-92.
Benhabib, J., & Bisin, A. (2005). Modeling internal commitment mechanisms
and self-control: A neuroeconomics approach to consumption-savings
decisions. Games and Economic Behavior, 52(2), 460-492.
Bernheim, B. D. & Rangel, A. (2004). Addiction and cue-triggered decision
processes. American Economic Review, 94(5), 1558-1590.
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
61
Bertrand, M., Mullainathan, S., & Shafir, E. (2004). A behavioral-economics
view of poverty. American Economic Review, 94, 419-423.
Botti, S., & Iyengar, S. S. (2006). The dark side of choice: When choice impairs
social welfare. Journal of Public Policy & Marketing, 25(1), 24-38.
Brosnan, S. F., Jones, O. D., Lambeth, S. P., Mareno, M. C., Richardson, A. S.,
Schapiro, S. J. (2007). Endowment effects in chimpanzees. Current
Biology, 17, 1704-1707.
Buchel C., Morris, J., Dolan, R., Friston, K. (1998). Brain systems mediating
aversive conditioning: an event-related fMRI study. Neuron, 20, 947–957.
Burger, J. M. (1999). The foot-in-the-door compliance procedure: a multiple
process analysis and review. Personality and Social Psychology Review,
3(4), 303–325.
Camerer, C., Loewenstein, G., & Prelic D. (2004). Neuroeconomics: Why
economics needs brains. Scandinavian Journal of Economics, 106(3), 555579.
Carter, R. M., Meyer, J. R., & Huettel, S. A. (2010) Functional neuroimaging of
intertemporal choice models: A review. Journal of Neuroscience,
Psychology, and Economics, 3(1), 27-45.
Chen, M. K., Lakshminarayanan, V., & Santos, L. R. (2006). How basic are
behavioral biases? Evidence from Capuchin monkey trading behavior.
Journal of Political Economy, 114(3), 517-537.
Chiou, A., Roe, S., & Wozniak, E. (2005) An evaluation of tax-refund splitting
as an asset-building tool for low-to-middle income individuals. Retrieved
November,
21,
2011,
from
http://www.d2dfund.org/system/files/publications/PAE+R2A2+FINAL.pdf
Damasio, A. R. (1994). Descartes’ error: Emotion, reason, and the human
brain. New York: HarperCollins.
Damasio, A. R. (1995). Toward a neurobiology of emotion and feeling:
operational concepts and hypotheses. Neuroscientist, 1, 19–25.
Damasio, A. R. (1996). The somatic marker hypothesis and the possible
functions of the prefrontal cortex. Philosophical Transactions of the Royal
Society of London B: Biological Sciences, 351, 1413-1420.
De Martino, B., Camerer, C. F., & Adolphs, R. (2010). Amygdala damage
eliminates monetary loss aversion. Proceedings of the National Academy of
Science, 107(8), 3788-3792.
Duflo, E. & Saez, E. (2002). Participation and investment decisions in a
retirement plan: The influence of colleagues’ choices. Journal of Public
Economics, 85(1), 121-148.
Epley, N., & Dunning, D. (2000). Feeling ‘holier than thou’: Are self-serving
assessments produced by errors in self- or social prediction? Journal of
Personality and Social Psychology, 79(6), 861-875.
Fennis, B. M. & Janssen, L. (2010). Mindlessness revisited: Sequential request
techniques foster compliance by draining self-control resources. Current
Psychology, 29, 235-246.
62
Journal of Personal Finance
Fleming, S. M., Thomas, C. L, & Dolan, R. J. (2010). Overcoming status quo
bias in the human brain. Proceedings of the National Academy of Sciences,
107(13), 6005-6009.
Fudenburg, D., & Levine, D. K. (2006). A dual-self model of impulse control.
American Economic Review, 96(5), 1449-1476.
Gazzaniga, M. S. (1983). Right hemisphere language following brain bisection.
American Psychologist, 38(5), 525-537.
Gimbel, R. W., Strosberg, M. A., Lehrman, S. E., Gefenas, E., & Taft, F. (2003).
Presumed consent and other predictors of cadaveric organ donation in
Europe. Progress in Transplantation, 13(1), 17-23.
Glimcher, P. W. (2004). Decisions, uncertainty, and the brain: The science of
neuroeconomics. Cambridge, MA: The MIT Press.
Haidt, J. (2006). The happiness hypothesis, New York: Basic Books.
Haidt, J. (2007). The new synthesis in moral psychology. Science, 316, 9981002.
Hochman, G., & Yechiam, E. (2011). Loss aversion in the eye and in the heart:
The autonomic nervous system’s response to losses. Journal of behavioral
decision making, 24, 140-156.
Huettel, S., Song, A. W., & McCarthy, G. (2008). Functional magnetic
resonance imaging (2nd ed.). Sunderland, MA: Sinanauer Associates, Inc.
Iyendgar, S. S., & Lepper, M. R. (2000). When choice is demotivating: Can one
desire too much of a good thing? Journal of Personality and Social
Psychology, 79(6), 995-1006.
Johnson, E. J., & Goldstein, D. (2003). Do defaults save lives? Science, 302,
1338-1339.
Johnson, E. J., Hershey, J, Meszaros, J, & Kunreuther, H. (1993). Framing,
probability distortions, and insurance decisions. Journal of Risk and
Uncertainty, 7(1), 35-51.
Johnston, L., O’Malley, P., & Bachman J. (1993). Illicit drug use, smoking, and
drinking by America’s high school students, college students, and young
adults, 1975-1987. Rockville, Maryland: National Institute on Drug Abuse.
Kable, J. W., & Glimcher, P. W. (2007). The neural correlates of subjective
value during intertemporal choice. Nature Neuroscience, 10, 1625–1633.
Kahneman, D., Knetsch, J., & Thaler, R. (1990). Experimental tests of the
endowment effect and the Coase theorem. Journal of Political Economy,
98(6), 1325-1348.
Kahneman, D., & Tversky, A. (1982). The psychology of preferences. Scientific
American, 246, 160-173.
Kemper, T. (1991). Predicting emotions from social relations. Social Psychology
Quarterly, 54, 330-342
Knetsch, J. (1989). The endowment effect and evidence of nonreversible
indifference curves. American Economic Review, 79, 1277-1284.
Laibson, D. (1997). Golden eggs and hyperbolic discounting. Quarterly Journal
of Economics, 112, 443–477.
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
63
Loewenstein, G. (2000). Emotions in economic theory and economic behavior.
American Economic Review, 90(2), 426-432.
Loewenstein, G., & O’Donoghue T. (2004). Animal spirits: Affective and
deliberative processes in economic behavior. Retrieved November 21, 2011
from http://ssrn.com/abstract=539843
Loewenstein, G. F., Weber, E. U., Hsee, C. K., & Welch, N. (2001). Risk as
feelings. Psychological Bulletin, 127(2), 267-286.
Mann, R. E. (2005). Availability as a law of addiction. Addiction, 100(7), 924925.
McClure, S. M., Ericson, K. M., Laibson, D. I., Loewenstein, G., & Cohen, J. D.
(2007). Time discounting for primary rewards. Journal of Neuroscience, 27,
5796–5804.
McClure, S. M., Laibson, D. I., Loewenstein, G., & Cohen, J. D. (2004).
Separate neural systems value immediate and delayed monetary rewards.
Science, 306, 503–507.
Muramatsu, R. & Hanoch, Y. (2005). Emotions as a mechanism for boundedly
rational agents: The fast and frugal way. Journal of Economic Psychology,
26(2), 201-221.
Nicolle, A., Fleming, S. M., Bach, D. R., Driver, J., & Dolan, R. J. (2011). A
regret-induced status quo bias. The Journal of Neuroscience, 31(9), 33203327.
Odean, T. (UC-Davis). (1998). Are investors reluctant to realize their losses?
Journal of Finance, 53, 1775-1798.
Öhman, A., Mineka, S. (2001). Fears, phobias, and preparedness: Toward an
evolved module of fear and fear learning. Psychological Review, 108, 483–
522.
O’Keefe, D. J., & Hale, S. L. (2001). An odds-ratio-based meta-analysis of
research on the door-in-the-face influence strategy. Communication
Reports, 14(1), 31–38.
Pollock, C. L., Smith, S. D., & Knowles, E. S. (1998). Mindfulness limits
compliance with the that’s-not-all technique. Personality and Social
Psychology Bulletin, 24(11), 1153–1157.
Prelec, D., & Loewenstein, G. (1991). Decisionmaking over time and under
uncertainty: A common approach. Management Science, 37, 770–786.
Prelec, D. & Simester, D. (2001). Always leave home without it: A further
investigation of the credit-card effect on willingness to pay. Marketing
Letters, 12, 5-12.
Read, D., Loewenstein, G., & Rabin, M. (1999). Choice bracketing. Journal of
Risk and Uncertainty, 19, 171-197.
Read, D., & van Leeuwen, B. (1998). Predicting hunger: The effects of appetite
and delay on choice. Organizational behavior and human decision
processes, 76, 189-205.
Ritov, I., & Baron, J. (1990). Reluctance to vaccinate: Omission bias and
ambiguity. Journal of Behavioral Decision Making, 3, 263-277.
64
Journal of Personal Finance
Roberts, R. J., Hager, L. D., & Heron, C. (1994). Prefrontal cognitive processes:
Working memory and inhibition in the antisaccade task. Journal of
Experimental Psychology: General, 123, 374-393.
Room, R. P., Babor, T. & Rehm, J. (2005). Alcohol and public health. The
Lancet, 365, 519-530.
Samuelson, P. A. (1937). A note on measurement of utility. The Review of
Economic Studies, 4, 155-161.
Samuelson, W. & Zeckhauser, R. (1988). Status quo bias in decision making.
Journal of Risk and Uncertainty, 1, 7-59.
Sanfey, A. G., Loewenstein, G., McClure, S. M, & Cohen, J. D. (2006).
Neuroeconomics: Cross-currents in research on decision-making. Trends in
Cognitive Sciences, 10(3), 108-116.
Shin, L. M., & Liberzon, I. (2010). The neurocircuitry of fear, stress, and
anxiety disorders. Neuropsychopharmacology, 35, 169-191.
Shiv, B. & Fedorikhin, A. (1999). Heart and mind in conflict: The interplay of
affect and cognition in consumer decision making. Journal of Consumer
Research, 26(2), 278-292.
Shiv, B., Loewenstein, G., Bechara, A., Damasio, H., & Damasio, A. R. (2005).
Investment behavior and the negative side of emotion. Psychological
Science, 16(6), 435-439.
Simonson, I. (1999). The effect of product assortment on buyer preferences.
Journal of Retailing, 75(3), 347-370.
Simonson, I., & Tversky, A. (1992) Choice in context: Tradeoff contrast and
extreme aversion. Journal of Marketing Research, 29, 281-295.
Smets, P., & Bähre, E. (2004). When coercion takes over: The limits of social
capital in microfinance schemes. In H. B. Lont & O. Hospes (Eds.)
Livelihood and microfinance: Anthropological and sociological
perspectives on savings and debt (215-236). Delft, The Netherlands: Eburon
Publishers.
Sokol-Hessner, P., Hsu, M., Curley, N. G., Delgado, M. R., Camerer, C. F.,&
Phelps, E. A. (2009). Thinking like a trader selectively reduces individuals’
loss aversion. Proceedings of the National Academy of Sciences of the
United States of America, 106(13), 5035-5040.
Soman, D. (2003). The effect of payment transparency on consumption: Quasiexperiments from the field. Marketing Letters, 14, 173-183.
Thaler, R. (1980). Toward a positive theory of consumer choice. Journal of
Economic Behavior and Organization, 1, 39-60.
Thaler, R. (1981). Some empirical evidence on dynamic inconsistency.
Economics Letters, 8, 201-207.
Thaler, R. H., & Benartzi, S. (2004). Save More Tomorrow: Using behavioral
economics to increase employee saving. Journal of Political Economy,
112(1), S164-S187.
Thaler, R. H., & Johnson, E.J. (1990). Gambling with the house money and
trying to break even: The effects of prior outcomes on risky choice.
Management Science, 36, 643-660.
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
65
Thaler, R. H., & Shefrin, H. M. (1981). An economic theory of self-control,
Journal of Political Economy, 89(2), 392-406.
Thaler, R. H., Sunstein, C. R. (2008). Nudge: Improving decisions about health,
wealth and happiness. New Haven, CT: Yale University Press.
Thaler, R., Tversky, A., Kahneman, D., & Schwartz, A. (1997). The effect of
myopia and loss aversion on risk taking: An experimental test. Quarterly
Journal of Economics, 112, 647-661.
Tom, S. M, Fox, C. R., Trepel, C., & Poldrack, R. A. (2007). The neural basis of
loss aversion in decision-making under risk. Science, 315, 515-518.
Tversky, A. & Kahneman, D., (1981). The framing of decisions and the
psychology of choice. Science, 211, 453-458.
Ward, A., & Mann, T. (2000). Don’t mind if I do: Disinhibited eating under
cognitive load. Journal of Personality and Social Psychology, 78(4), 753763.
Williams, R. W. & Herrup, K. (1988). The control of neuron number. Annual
Review of Neuroscience, 11, 423-453.
Wittmann, M., Leland, D. S., & Paulus, M. P. (2007). Time and decision
making: Differential contribution of the posterior insular cortex and the
striatum during a delay discounting task. Experimental Brain Research,
179(4), 643-653.
Wolff, E. N. (1998). Recent trends in the size distribution of household wealth.
Journal of Economic Perspectives, 12(3), 131-150.
Woolcock, M. J. (1999). Learning from failures in microfinance: What
unsuccessful cases tell us about how group based programs work. American
Journal of Economics and Sociology, 58(1), 17-42.
66
Journal of Personal Finance
MUTUAL FUND TAX EFFICIENCY AND INVESTMENT
SELECTION
D.K. Malhotra*
Philadelphia University
Rand Martin
Bloomsburg University of Pennsylvania
C. Andrew Lafond
The College of New Jersey
We examine six factors that may be important when looking for a more
tax efficient mutual fund. We consider the pre- and post-liquidation
bases for over 4,000 mutual funds in logical groupings. Our results for
turnover show that its effect on tax efficiency depends upon conditions
in the securities markets. A falling market leads to greater tax
efficiency due to security sales at depressed prices. We have a similar
finding for expense categories probably because increased sales lead to
higher expenses. We find greater tax efficiency if mutual funds have
institutional status, no-loads, and no 12b-1 plan.
Introduction
Mutual funds continue to be popular as investment vehicles.
Evidence of this is the fact that, according to the Investment
Company Institute, the total dollars invested through mutual funds
increased from $3.028 trillion in 1995 to $10.701 trillion as of May
2009. Since 1981 when the Internal Revenue Service (IRS)
proposed regulations that opened the doors for companies to use
401(k) plans, many companies have moved away from pension
plans towards self-directed 401(k) type retirement plans fueling the
*
D.K. Malhotra, School of Business Administration, Philadelphia University,
School House Lane and Henry Avenue, Philadelphia, PA 19144; (215)-9512813; MalhotraD@philau.edu
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
67
tremendous
growth
in
the
mutual
fund
industry
(Gilliam, Dass, Durband, and Hampton 2010). This popularity
results largely from the low cost diversification afforded by mutual
funds (Weiss, 2005). But, with so many mutual funds available,
investors need the best criteria for making selections that fit their
needs.
Investors usually select mutual funds based on performance,
risk, and investment objective. Two other selection criteria have
recently been touted as important. One is the expense ratio which
has been the subject of extensive academic research (McLeod and
Malhotra, 1994, Malhotra and McLeod, 1997, Bogle, 2006,
Barber, Odean, and Zheng, 2005). In recent years tax efficiency
has also emerged as a selection criterion of great interest. Tax
efficiency is the subject of this study.
Is the attention paid to tax efficiency justified? It may be for
investors who are not invested in a tax-exempt fund or in a taxdeferred account such as an IRA. To answer the question for those
investors, consider some facts about the taxation of mutual fund
returns. Bruce (2003) reports that “the Securities and Exchange
Commission says that more than 2.5 percent of the average stock
fund's total return is lost each year to taxes, significantly more than
the amount lost to fees. The tax bite varies from zero percent for
the most tax-efficient funds to 5.6 percent for the least efficient.”
Bernard (2006) says that mutual fund investors paid an estimated
$15.2 billion in taxes according to a study conducted by data
tracker Lipper, Inc. According to Roger Ibbotson, “roughly 2
percent of pre-tax returns of mutual funds are lost to taxes for those
taxpayers in the highest tax bracket” (Tuve Investments, 2007).
Discussion of tax efficiency in the popular press has become so
common that in Kiplinger’s Personal Finance it was said that “taxefficient investing is the most over-covered topic in the financial
68
Journal of Personal Finance
press” (Gregory and Savage, 2003). Even with that coverage in
the popular press, this topic has received very little attention in
academic research. We think tax efficiency is an important topic
and that investors should have a better idea of how to look for tax
efficient mutual funds (Bergstresser & Poterba, 2002, Ivkovic´ &
Weisbenner, 2009, Sialm & Starks, 2011). Our purpose in this
study is to investigate the likely characteristics of tax-efficient
mutual funds that an investor can easily use to select funds.
Taxation of Returns and Avoidance Of Taxes
Taxes can be incurred on shares of mutual funds in three
ways. The first is when a shareholder sells shares for more than
the purchase price and thus realizes a capital gain. Incurrence of
this tax is under the control of shareholders since they decide when
to sell shares. The second way is taxation of interest and dividends
received when shares are not held in a tax-deferred retirement
savings account. Mutual funds could be selected that do not pay
interest or dividends due to their investment objective. But, if an
investor selects funds that generate interest and dividends, tax
liability will be incurred. The third type of taxation is caused by
mutual fund managers as they turn over the assets of the fund and
create capital gains income which may be distributed to
shareholders. This is out of the control of the shareholder and is
often overlooked by investors as a cause of taxation. Qualified
dividends and long-term capital gains get preferential tax treatment
at the maximum 15% tax rate; however, interest, nonqualified
dividends, and short-term capital gains can be taxed as ordinary
income.
Since mutual fund rules require that all income earned by a
fund be distributed each year to the shareholders, those funds that
earn more income through interest, dividends, or capital gains are
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
69
less tax efficient and create a higher tax liability for the investor.
But, there are three things that investors can do to avoid taxes
besides holding shares in tax-deferred retirement savings accounts.
One is using index funds. Index funds are more tax efficient than
actively managed mutual funds because their objective to invest
according to the representation of securities in indexes requires
less frequent trading of securities. Other attributes of index funds
is that they tend to have lower fees but lower earnings perspectives
(Loibl, Lee, Fox and Mentel-Gaeta 2007). Over a long period of
time, this buy and hold approach will enhance the overall return on
an index fund (Ellentuck, 2004). The second means for avoiding
tax is to use mutual funds that have tax efficiency as an objective.
These funds reduce a shareholder’s taxes by having a lower
turnover rate, by investing in non-dividend paying stocks, and by
using a strategy of matching losses with gains during a year to
minimize the amount of capital gains. These funds also attempt to
reduce the incurrence of taxes by discouraging share sales through
charging a redemption fee if shares are sold before a specified time
period elapses after they are purchased (Ellentuck, 2004). The
third means is by using exchange-traded funds. Gardner and
Welch (2005) describe these funds by saying that “exchangetraded funds are a basket of securities owned by a mutual fund
company, similar to a mutual fund, but differ in how their shares
are issued, traded, and redeemed.” These funds use a strategy of
creating and redeeming shares through in-kind transactions so that
sales aren’t made and capital gains and other taxable transactions
are not created.
In the current study, we do not limit ourselves to an
examination of mutual funds that are designed for tax efficiency.
Instead, we examine all mutual funds for which we have adequate
data in an effort to determine whether certain characteristics can be
used to select tax efficient funds.
70
Journal of Personal Finance
Literature Review
We could find only one academic study that examines aftertax mutual fund returns. Peterson, et al (2002) study the pre-tax
returns, after-tax returns, and tax efficiency of 1,170 mutual funds
that had diversified U.S. equity investment objectives over the
years 1981 to 1998. The authors hypothesize that the important
determinants of pre-tax returns would also be the important
determinants of after-tax returns. They find that historically tax
efficient funds outperform historically tax inefficient funds on an
after-tax basis. Funds that experience net redemptions, especially
in the case of large cap value funds, subsequently perform worse
after-tax than comparable funds that do not have net redemptions.
They find that the important determinants of after-tax and pre-tax
returns are the level of risk taken, investment style, past pre-tax
performance, and expenses. They find that turnover is not related
to after-tax returns. Finally, they find that funds with large cash
inflows, high expense ratios, or an emphasis on small cap stocks
tend to be more tax efficient and to have lower pre-tax returns.
The authors take their findings to mean that investors should focus
on after-tax returns rather than tax efficiency.
Our study differs from that of Peterson, et al (2002) in three
ways. First, we include funds with debt as well as equity
investment objectives. Second, we study tax-efficiency on the
basis of earnings distributed and on the basis of earnings after
distribution and sale of fund shares. And, third, we use a different
sample period.
Method of Investigation
Tax efficiency is typically measured in one of two ways.
One is by dividing the after-tax return by the pre-tax return to
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
71
obtain a tax-efficiency ratio. The other is computation of the
return lost to taxes by subtracting the fund’s after-tax return from
its pre-tax return. This second method is more relevant since the
tax-efficiency ratio is not meaningful when the numerator or
denominator is negative (Riepe, 2000). As a result we use the
second method. When using this method, a positive difference
(pre-tax return minus after-tax return) will mean less tax efficiency
since more return is lost to taxes. A negative result can be had and
will be discussed in our empirical results section below. The most
desirable result would be a difference of zero, meaning that no
return is lost to taxes.
We use a three-pronged approach to analyze the tax
efficiency of mutual funds. Firstly, we calculate tax efficiency on
a pre-liquidation status and on a post-liquidation status for six fund
characteristics. They are (1) turnover, (2) investment style, (3)
expense category, (4) institutional or retail status, (5) load status,
and (6) 12b-1 plan status. Many institutional investors such as
endowment funds and pension funds are tax exempt and would not
care about tax efficiency. But, we investigate the effect of the
institutional/retail variable because purchase of institutional fund
shares is not limited to tax exempt institutions. The only
requirement for purchase of those shares is meeting the minimum
investment amount.
Pre-liquidation tax efficiency is the total return for 12 months
minus the return after tax on distribution. Post-liquidation tax
efficiency is the total return for 12 months minus the return after
tax on distribution and sale.
To calculate return after tax on distribution and sale,
Morningstar says that it “applies the appropriate historical tax rate
72
Journal of Personal Finance
based on the date of distribution.” The tax rates that Morningstar
uses are as follows.
• 35 percent for interest income and dividends that do not qualify
for taxation at the lower long-term capital gains tax rate
• 15 percent for dividends that do qualify for the lower long-term
capital gains tax rate
• 35 percent for short-term capital gains
• 15 percent for long-term capital gains
We calculate tax efficiency for what we consider to be
logical groupings of mutual funds under each of the six data items
mentioned above. Listed below is an explanation of how we
placed mutual funds into the groupings.
1. Turnover Level (Table 2): According to Morningstar, the
turnover ratio “loosely represents the percentage of the
portfolio’s holdings that have changed over the past year.” We
use five turnover percentage ranges to categorize mutual funds:
Very Low ≤ 20%, Low > 20% and ≤ 40%, Average > 40%
and ≤ 80%, High > 80% and ≤ 150%, Very High > 150%.
2. Investment Style (Table 3): Morningstar says that mutual
funds with an equity style invest in stocks and mutual funds
with a fixed-income style invest in bonds. Those with a style
of “equity and fixed income” invest in both stocks and bonds
and are sometimes called balanced funds.
3. Expense Category (Table 4): Morningstar defines an expense
ratio as “the percentage of fund assets paid for operating
expenses and management fees, including 12b-1 fees,
administrative fees, and all other asset-based costs incurred by
the fund, except brokerage costs.” They use expense ratios to
put mutual funds into four expense categories as follows.
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
•
•
•
•
73
Category A: Mutual funds whose expense ratios fall within
the cheapest quintile of its type of share class within its
comparison group
Category B: Mutual funds in the second-cheapest category
Category C: Mutual funds with expense ratios falling
between the 40th and 60th percentiles of the peer group
Category F: Mutual funds with higher expense ratios than
the funds in category C
We combine the Morningstar expense category
classifications A and B to form our Low expense ratio category
and we combine the C and F classifications to form our High
expense ratio category.
4. Institutional / Retail Status (Table 5): Morningstar defines
an institutional fund as one that meets one of three criteria: (1)
has the word "institutional" in its name, (2) has a minimum
initial purchase of $100,000 or more, or (3) states in its
prospectus that it is designed for institutional investors or those
purchasing on a fiduciary basis. We consider all other funds to
be retail funds.
5. Load and No-Load Groupings (Table 6): Morningstar
describes sales fees in this way: “Also known as loads, sales
fees represent the maximum level of initial (front-end) and
deferred (back-end) sales charges imposed by a fund.” To
form our Load status categories, we combine the front-end and
back-end sales charges reported by Morningstar.
6. 12b-1 Plan Status (Table 7): Morningstar defines a 12b-1 fee
as “a fee used to pay for a mutual fund’s distribution costs. It is
often used as a commission to brokers for selling the fund.”
74
Journal of Personal Finance
For each group of funds, we calculate the mean and variance
of the tax efficiency results. Then, we compute t-statistics for twotailed tests of the difference in sample means. We compare the
means of tax efficiency measures for each group of funds to the
mean tax efficiency of the group that is likely to have the highest
tax efficiency. In effect we are testing the following hypotheses
about the mean tax efficiency of groups of funds.
Secondly, in order to evaluate the impact of fund
characteristics on tax efficiency of a fund, we model tax efficiency
(pre-liquidation as well as post liquidation) as a function of
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
75
investment style, load, 12b-1 plan, fees, share class, and turnover
ratio. Equation 1 describes our regression model:
Tax Efficiency = ß0 + ß1 Equity Style + ß2 Fixed Style + ß3
Fees + ß4 Load + ß5 12b-1 Plan + ß6 Share Class + ß7 Turnover
Ratio + e
(1)
Where
•
•
•
•
•
•
•
Tax Efficiency refers to pre-liquidation and post-liquidation
tax efficiency of a mutual fund.
Equity Style is a dummy variable that equals 1 if the
mutual fund follows equity style of investing and 0
otherwise.
Fixed Style is a dummy variable that equals 1 if the mutual
fund follows fixed style of investing and 0 otherwise.
Load is the actual percentage load (front-end or deferred)
of a fund.
12b-1 Plan is a dummy variable that equals 1 if the mutual
fund has a 12b-1 plan charge and zero otherwise.
Fees is a dummy variable that equals 1 if the mutual fund
got a rating of A or B on expense ratio and zero otherwise.
Turnover ratio is the actual percentage turnover ratio of the
mutual fund.
We estimate the model specified in equation 1 using linear
regression for each of the four years from 2006 to 2009. According
to Peterson et al. (2002) mutual funds typically do not distribute
capital gains more often than once a year and this can create a
problem in identifying tax inefficient mutual funds. For instance, a
mutual fund may decide not to distribute capital gains in a
particular year due to poor performance. Such a fund’s after-tax
return will be roughly the same as its pretax return, which makes it
tax efficient for that particular year. Peterson et al. also argue that a
76
Journal of Personal Finance
fund can manage the realization of capital gains and, therefore,
taxable distributions and after-tax returns in the short run.
Therefore, in order to analyze the true impact of fund
characteristics on tax efficiency of a fund, we need to lengthen the
measurement period for after-tax returns beyond one year.
Therefore, the third part is an estimation of coefficients for
the above tax efficiency models using the panel data approach,
which allows for pooling of observations on a cross-section of
open-end funds over four years. The most obvious advantage of
panel data is that the number of observations is typically much
larger in panel data, which will produce more reliable parameter
estimates and thus enable us to test the robustness of our linear
regression results. Panel data also alleviate the problem of
multicollinearity, because when the explanatory variables vary in
two dimensions (cross-section and time series), they are less likely
to be highly correlated.
The most intuitive way to account for individual and/or time
differences in the context of panel data regression is to use the
fixed effects model. The fixed effect model assumes that
difference across mutual funds can be captured in differences in
the constant term. The regression coefficients (the slope
parameters) across groups in this model are unknown but fixed
parameters. It is also known as least square dummy variable
(LSDV) model, and we use LSDV fixed-effect model to estimate
cost efficiencies in the mutual fund industry.
Furthermore, we include a dummy variable to account for
time trend in the tax efficiency of mutual funds since 2006.
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
77
Data
Our data is from Morningstar Principia and it covers the
years 2006 through 2009. We use nine Morningstar data items to
investigate tax-efficiency: turnover, investment style, expense
ratio, status as an institutional or retail fund, front-end load,
deferred load, 12b-1 plan status, after-tax return on distribution for
one year, and after-tax return on distribution and after sale of
shares on a one-year basis. We combine front-end load and
deferred load in order to put funds in one of two categories: load
and no-load. Our data is summarized in Table 1. *
*
We are using live funds in this paper and, as a result, our study does suffer
from survivorship bias. However, when we use panel data approach by
combining the data for four years, we address the survivorship bias issue to
some extent.
78
Journal of Personal Finance
Table 1: Summary Statistics for the Data of this Study
2006
2007
2008
2009
12Month
Total
Return
Retur
n after
Tax
on
Distri
bution
1Year
Return
after
Tax on
Distrib
ution
and
Sale
1Year
2.63
7.35
3.99
4.72
2.52
2.22
5.88
2.93
Std.Dev.
2.61
3.62
8.87
Mean
1.73
-10.51
Median
1.09
Std.Dev.
PreLiqui
dation
Tax
Effici
ency
PostLiquid
ation
Tax
Efficie
ncy
Mean
3.36
Median
FrontEnd
Load %
Plus
12b-1
Plan
%
Turnov
er
Ratio
1.64
0.42
79.02
3.66
0.00
0.25
53.00
9.04
6.02
2.19
0.38
116.98
-30.94
-32.67
-20.43
1.56
0.40
102.31
-12.07
-35.74
-37.31
-23.51
0.00
0.25
57.00
1.67
5.75
16.03
15.64
10.46
2.16
0.38
292.18
Mean
2.55
11.59
30.27
27.72
18.68
1.50
0.39
105.53
Median
1.17
11.18
29.43
26.65
17.87
0.00
0.25
59.00
Std.Dev.
2.84
6.08
16.00
16.04
10.34
2.13
0.38
62.00
Mean
2.69
16.73
44.71
42.03
27.99
1.49
0.39
105.65
Median
1.18
17.71
47.93
45.27
29.81
0.00
0.25
60.00
Std.Dev.
3.21
8.51
22.79
22.68
14.62
2.13
0.38
160.43
Deferred
Load %
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
79
Empirical Results
Our findings are as follows for the six variables that we
investigate with regard to tax efficiency.
Turnover Ratio
Table 2 gives our tax efficiency results by turnover level.
80
Journal of Personal Finance
Inspection of Table 2 reveals that mutual funds with lower
turnover rates generally lose smaller percentages of return to taxes.
For 2006, 2008, and 2009, almost every turnover category has a
smaller return loss due to taxes than the next higher turnover
category on both the pre-liquidation and post-liquidation bases.
We use t-statistics to compare the mean tax efficiency measures for
each of the low, average, high, and very high turnover groups to
the tax efficiency measures for the very low turnover group for the
same year. For these three years, this requires calculating 24 tstatistics. Twenty of those t-statistics show statistically significant
differences in means. Each of the higher turnover groups is less
tax efficient than the very low turnover group. This result is
consistent with our expectation that the fewer the transactions a
mutual fund does, the lower the total capital gains tax.
For 2007, we have negative tax efficiency measures on the
post-liquidation basis for all levels of turnover. Negative tax
efficiency measures mean that there is a gain when shares are sold.
This was triggered because the stock market fell in 2007. These
capital losses reduced taxes. Therefore, in years where the stock
market goes down, funds with higher turnover ratios can be more
tax efficient than funds with lower turnover ratios. We find that
funds with low, average, and high turnover levels in 2007 have tax
benefits that are great enough to produce a statistically significant
difference compared to the mean tax inefficiency measure of the
very low turnover group. The very high turnover group goes in the
opposite direction with a loss compared to the very low group.
This
difference
is
also
statistically
significant.
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
81
Investment Style
Table 3 gives our tax efficiency results by investment style.
82
Journal of Personal Finance
The results in Table 3 show that for all years on the pre-liquidation
basis, funds with a fixed income style are more tax efficient than
funds with the equity and fixed income style and two of those
differences are statistically significant. But, funds with a fixed
income style are less tax efficient than funds with the equity style
for 2007, 2008, and 2009. All three of those differences are
statistically significant.
On the post-liquidation basis, funds with a fixed income style
are more tax efficient than funds with the equity and fixed income
style for 2006, 2008, and 2009 and those differences are
statistically significant. Compared to funds with the equity style,
those with the fixed income style are even more efficient for those
same three years. For 2007, the results are reversed with both
differences being statistically significant.
To explain these results, consider that equity style funds can
generate both dividend and capital gain income that must be
distributed to investors. The dividends, if qualified and the capital
gains distributions are taxed at a maximum preferential tax rate of
15 percent, result in less tax for the shareholder. Fixed income
funds typically hold bonds that generate interest income. Interest
income is taxed at the ordinary income tax rate which typically
results in a higher tax liability and therefore lower tax efficiency.
This last fact would explain the greater tax efficiency of equity
funds on a pre-liquidation basis. We expect that the lower tax
efficiency of equity funds on the post-liquidation basis results from
having to pay tax on both dividend income and capital gains
income. The reversal of tax efficiency results for 2007 compared to
the other years resulted from the decline in the stock market that
year. Capital losses on shares that were sold reduced taxes due.
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
83
Expense Category
Table 4 gives tax efficiency results for our low and high
expense ratio categories.
Table 4: Tax Efficiency of Mutual Funds by Expense Category
Groupings
Morningstar defines an expense ratio as “the percentage of fund assets paid for
operating expenses and management fees, including 12b-1 fees, administrative
fees, and all other asset-based costs incurred by the fund, except brokerage
costs.” We combine the Morningstar expense category classifications A and B
to form our Low expense category and we combine the C and F classifications
to form our High expense category.
Mean values are percentages of return lost due to taxes. They are computed by
subtracting after-tax return from total return. T-statistics are for two-tailed tests
for the difference in sample means assuming that the sample variances are
unequal.
*Statistically significant at the 5 percent level
84
Journal of Personal Finance
Tax efficiency results for the expense ratio categories are
mixed. One would expect that funds with low expense ratios would
have lower turnover in their portfolios and therefore would be
more tax efficient. That seems to be true for 2006 on the pre- and
post-liquidation bases, for 2008 on the pre-liquidation basis, and
for 2009 on the post-liquidation basis. T-statistics show that all of
these differences in efficiency are statistically significant.
For 2007 on the pre- and post-liquidation bases and for 2008
on the post-liquidation basis, high expense ratio funds are more
efficient than low expense ratio funds and the t-statistics show that
all of those differences are statistically significant. We believe this
reversal compared to 2006 is also the result of the fall in the stock
market in 2007.
Institutional or Retail Status
Table 5 gives our tax efficiency results by the institutional
and retail status of groups of mutual funds.
Table 5: Tax Efficiency of Mutual Funds by Institutional /
Retail Status
Morningstar defines an institutional fund as one that meets one of three criteria:
(1) has the word "institutional" in its name, (2) has a minimum initial purchase
of $100,000 or more, or (3) states in its prospectus that it is designed for
institutional investors or those purchasing on a fiduciary basis. We consider all
other funds to be retail funds.
Mean values are percentages of return lost due to taxes. They are computed by
subtracting after-tax return from total return. T-statistics are for two-tailed tests
for the difference in sample means assuming that the sample variances are
unequal.
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
85
*Statistically significant at the 5 percent level
Our results show that retail funds are less tax efficient than
institutional funds on both the pre- and post-liquidation bases for
all years of the sample period. Those differences in efficiency are
statistically significant for all but the 2007 post-liquidation basis.
A likely explanation for our results is found in the differences
in retail and institutional funds especially as they affect turnover of
securities in their portfolios. These funds usually have a minimum
investment requirement and their investment strategy typically
involves buying and holding securities. This works to the
advantage of their customers which are pension funds,
endowments, and high net worth individuals. Those investors will
86
Journal of Personal Finance
redeem shares less frequently than the public in general. Thus the
buy-and-hold approach leads to a lower turnover rate, lower
operating costs, and thus greater tax efficiency. Retail funds cater
to individuals and financial advisors. They are characterized by a
more active management style, higher risk, a higher turnover rate,
and thus lower tax efficiency.
Load Status
Table 6 gives our tax efficiency results according to the load
status of groups of mutual funds. We separate mutual funds into
two broad groups here: load and no-load. A load fund is one for
which the buying and/or selling of its shares causes the investor to
incur a sales fee which is called a load. For a no-load fund there is
no sales fee involved in transacting its shares.
Table 6: Tax Efficiency of Mutual Funds for Load and NoLoad Groupings
Morningstar describes sales fees: “Also known as loads, sales fees represent the
maximum level of initial (front-end) and deferred (back-end) sales charges
imposed by a fund.” To form our Load status categories, we combine the frontend and back-end sales charges reported by Morningstar.
Mean values are percentages of return lost due to taxes. They are computed by
subtracting after-tax return from total return. T-statistics are for two-tailed tests
for the difference in sample means assuming that the sample variances are
unequal.
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
87
*Statistically significant at the 5 percent level
Inspection of Table 6 shows that mutual funds with loads are
on average less tax efficient than no-load funds for every year
during the sample period on both the pre- and post-liquidation
bases. The t-statistics for differences in sample means are all
statistically significant. Apparently, mutual funds with sales
charges are on average more willing to engage in turnover of their
portfolios which in turn reduces tax efficiency.
12b-1 Plan Status
Table 7 gives our tax efficiency results for funds grouped
according to whether they have a 12b-1 plan. If a mutual fund has
a 12b-1 plan, it charges customers for marketing expenses which
are often commissions to brokers who sell shares of the fund.
88
Journal of Personal Finance
Table 7: Tax Efficiency of Mutual Funds by 12b-1 Plan Status
Morningstar defines a 12b-1 fee as “a fee used to pay for a mutual fund’s
distribution costs. It is often used as a commission to brokers for selling the
fund.”
Mean values are percentages of return lost due to taxes. They are computed by
subtracting after-tax return from total return. T-statistics are for two-tailed tests
for the difference in sample means assuming that the sample variances are
unequal.
*Statistically significant at the 5 percent level
Our results show that mutual funds that do not have a 12b-1
plan are more tax efficient than funds with a 12b-1 plan on both
the pre- and post-liquidation bases for all years of the sample
period. The t-statistics for differences in sample means are all
statistically significant. Mutual funds that are willing to pay this
additional marketing expense apparently have a higher turnover of
securities and thus lower tax efficiency.
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
89
Table 8 summarizes regression results for the model
specified in equation 1.
Table 8: Summary of regression results for tax efficiency of
mutual funds.
Dependent variable is pre-liquidation tax efficiency and post-liquidation tax
efficiency. Independent variables are equity-style, fixed-income style, fees, load
charges, 12b-1 plan charges, turnover ratio, and share class (institutional or
retail).
90
Journal of Personal Finance
On a pre-liquidation basis, mutual funds with equity style of
investing are more tax efficient in three out of four years. Panel
data also shows that there is lower return lost due to taxes if the
fund is an equity style. Mutual funds with a fixed income style of
investing are tax efficient in all the five regressions. On a postliquidation basis, mutual funds with an equity style of investing
lose more return due to taxes and, therefore, are less tax efficient.
Mutual funds with a fixed style of investing continue to be more
tax efficient even on a post-liquidation basis and panel data
analysis confirms this result.
According to panel data analysis, mutual funds with lower
fees (rated A or B on fees) are more tax efficient on a postliquidation basis and less tax efficient on a pre-liquidation basis.
On a pre-liquidation basis, the relationship between fees and tax
efficiency is statistically weak.
Load mutual funds report significant loss return due to taxes
and therefore are less tax efficient on a pre- and post-liquidation
basis for each of the four years. Panel data analysis also confirms
that load mutual funds are less tax efficient.
Funds with a 12b-1 plan lose less return due to taxes. Panel
data shows that on a pre-liquidation basis, 12b-1 plans are more tax
efficient. On a year to year basis, the relationship between 12b-1
plan and tax efficiency is negative, but either it is weakly
statistically significant [2007 (pre- and post-liquidation), 2008
(pre-liquidation), and 2009 (pre-liquidation)] or not significant
(2006).
On a pre- and post-liquidation basis, higher turnover ratio
results in higher loss of return due to taxes for each of the four
years except for the year 2007. Mutual funds with a higher
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
91
turnover ratio report higher loss of return due to taxes in 2006,
2008, and 2009 on a pre- as well as post-liquidation basis. For the
year 2007, higher turnover ratio results in more tax efficiency on a
pre-liquidation basis and the results are weakly statistically
significant. On a post-liquidation basis also, higher turnover results
in less loss of return due to taxes, but the results are not statistically
significant. Panel data analysis shows that higher turnover ratio
funds are less tax efficient on a pre-liquidation basis and the results
are highly statistically significant. On a post-liquidation basis,
higher turnover funds are more tax efficient and the results are
weakly statistically significant.
Institutional share class funds have a lower loss of return due
to taxes, but this relationship is either not statistically significant or
weakly statistically significant.
The last two columns in Table 8 capture the time trend in
mutual fund tax efficiency. On a pre-liquidation basis, tax
efficiency shows improvement in 2007, 2008, and 2009 with
reference to 2006. On a post-liquidation basis, tax efficiency shows
improvement for the year 2007 in comparison to the year 2006.
However, in 2008 and 2009 tax efficiency has deteriorated in
comparison to 2006.
Implications for Financial Planners
The findings of this study provide advice for financial
planners in selecting tax efficient mutual funds for their clients.
Financial planners need to consider tax efficiency in building their
client’s mutual fund portfolio if they are going to have a successful
financial plan. In building a portfolio of mutual funds, financial
planners need to manage the portfolios of their clients by taking a
comprehensive approach and examining all retirement accounts,
nonretirement accounts, and nontaxable accounts at the same time.
92
Journal of Personal Finance
By taking a more comprehensive view, financial planners should
employ a strategy of putting more tax-efficient mutual funds in
nonretirement accounts and less tax efficient mutual funds in
retirement accounts. Financial planners must understand that asset
allocation is very important in employing a tax efficient strategy
and mutual funds should be selected based upon after-tax return
not pre-tax return.
The study finds that more tax efficient funds include those
funds that have a lower turnover rate, fixed income investment
style, institutional status, and no-load status. Financial planners
desiring tax efficiency should seek mutual funds with these
attributes. In evaluating mutual funds that have low expense ratios,
financial planners need to keep in mind that these funds will be
more tax efficient when the securities markets are rising.
However, if markets fall, funds that are more active and, therefore,
have a higher expense ratio will be more tax efficient as they will
create more capital losses selling shares at depressed prices, thus
creating a tax advantage. Other tax-efficient strategies that can be
taken include investing in index funds which employ a buy and
hold strategy, receive qualified dividends taxed at 15%, and
generate little if any capital gains. On the contrary, investing in
high yield bond funds which generate interest income which is
taxed at ordinary income tax rates is not a tax-efficient strategy.
The study also finds that the turnover rate of a fund’s impact
on tax efficiency depends upon conditions in the securities
markets. In a rising market the opportunity is obvious - realize
profits/capital gains and the investor accumulates wealth.
Although there is less tax efficiency in a rising market, as taxable
income is generated there is potentially a return for the investor,
which is the primary objective of investing in the market. On the
other hand, no financial planner or investor desires a falling
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
93
market. However, in a falling market there is opportunity for a
financial planner or investor to create capital losses and offset
capital gains, if any, and reduce the tax burden or increase tax
efficiency. At the same time the investor benefits by reinvesting
back in the market at lower prices.
All of these implications need to be prefaced by stating that
no investment decision should be solely based upon the tax impact
of a decision or tax efficiency. The tax impact of an investment
decision is just one criterion that needs examination in the
selection of a particular mutual fund. At the end of the day, the
only return that matters to an investor is the return that he or she
has left over after taxes.
Summary and Conclusions
We investigate six factors which we have reason to believe
could characterize or affect mutual funds as to their tax efficiency.
Our measure of tax efficiency here is the percentage of return lost
to taxes. One of the factors we investigate, turnover, is actually a
determinant of tax efficiency since a higher turnover of securities
causes the incurrence of more taxes which in turn reduces return.
We evaluate the impact of the other five factors on tax efficiency.
We find tax efficiency to be greater on average for mutual
funds that have a lower turnover rate, a fixed income investment
style, institutional status, and no-load status. Mutual funds with
low expense ratios will be more tax efficient when the securities
markets are rising. This will reverse if markets fall because a tax
advantage will be created by capital losses incurred when investors
decide to sell shares at depressed prices. Mutual funds whose
investment style is a combination of equity and fixed income will
94
Journal of Personal Finance
have tax efficiency between those of the fixed income and equity
styles.
References
Barber, B.M., Odean, T. and Zheng, L. (2005). Out of Sight, Out of Mind: The
Effects of Expenses on Mutual Fund Flows, Journal of Business, 78 (6),
2095-2119.
Bergstresser, D. and Poterba, J. (2002). Do after-tax returns affect mutual fund
inflows? Journal of Financial Economics 63, 381–414.
Bernard, T.S. (2006). Fund Investors can expect a tax increase. Lipper Inc. study
Bogle, J. (2010). On mutual funds, cheaper is better, The Wall Street Journal.
August 27
Bogle, J. (2006). Whose capital is it? Putting owners back in control. Business
Economics, 41(2), 47-52.
Bruce, L. (2003). Mutual fund turnover and taxes. Retrieved on January 4, 2010
from http://www.Bankrate.com.
Ellentuck, A.B. (2004). Investing in tax efficient funds. The Tax Adviser, 35(11),
716-717.
Gardner, R. and J. Welch, J. (2005). Increasing after-tax return with exchangetraded funds. Journal of Financial Planning, 18(6), 30-35.
Gilliam, J., Dass, M., Durband, D., & Hampton, V. (2010).
The role of
assertiveness in portfolio risk and financial risk tolerance among married
couples. Financial Counseling and Planning, 21(1), 53-78.
Gregory, K. & Savage, S. (2003). Invest with an edge. Kiplinger’s Personal
Finance, 57(2), 60.
Investment Company Institute, Statistics, Mutual Fund Trends, Retrieved on
December 31, 2009 from http://www.ici.org.
Ivkovic´, Z. and Weisbenner, S. (2009). Individual investor mutual fund flows,
Journal of Financial Economics 92, 223–237.
Loibl, C., Lee, J., Fox, J., & Mentel-Gaeta, E. (2007).
Women's
HighConsequence Decision Making: A Nonstatic and Complex Choice
Process. Financial Counseling and Planning, 18(2), 35-47, 98-100.
Malhotra, D. K. and McLeod, R. (1997). An empirical analysis of mutual funds
expenses, Journal of Financial Research, 20(2), 175-190.
McLeod, R. and Malhotra, D.K. (1994). A re-examination of the effect of 12b1plans on mutual fund expense ratios, Journal of Financial Research, 17(2),
237-244.
Peterson, J.D., Pietranico, P.A., Riepe, M.W., Xu, F. (2002). Explaining aftertax mutual fund performance, Financial Analysts Journal, 58(1), 75-86
Tuve Investments Inc., 2007, What makes for a tax efficient mutual fund?
http://www.tuveinvestments.com .
Sialm, C. and Starks, L. (2011). Mutual Fund Tax Clienteles, Journal of
Finance, forthcoming.
Weiss, Ira S. (2002). An empirical examination of tax factors and mutual funds’
stock sales decisions, Review of Accounting Studies, Boston, 7(2), 343-347.
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
95
STUDENT FINANCIAL COUNSELING: AN ANALYSIS OF A
CLINICAL AND NON-CLINICAL SAMPLE
Sonya L. Britt*
John E. Grable **
Julie Cumbie*
Sam Cupples*
Justin Henegar*
Kurt Schindler*
Kristy Archuleta*
Kansas State University*
The purpose of this study was to determine what factors predict
whether students will seek on-campus peer-based financial counseling.
An attempt was made to determine if students who seek help differ
significantly from students who do not seek help. Findings provide a
profile of college student financial counseling help-seekers. Collegeage financial counseling help seekers tend to be older, less satisfied
with their income, less knowledgeable, less wealthy, and more stressed.
The results from this study suggest that college financial counseling
centers appear to be on target in connecting with some of the students
they were designed to reach. Continued efforts to assist students with
high financial stress may be a way to increase financial well-being
among college students.
*
Contact Author: John Grable, Institute of Personal Financial Planning, School
of Family Studies and Human Services, 318 Justin Hall, Kansas State
University, Manhattan, Kansas; Phone: (785)532-1486; Fax: (785)532-5505;
Email: jgrable@k-state.edu
96
Journal of Personal Finance
Introduction
Financial literacy is an issue of national importance. The
relevance of the issue was highlighted in 2009 when the U.S.
Department of Treasury convened a meeting of researchers and
policy makers to help establish a national agenda for financial
literacy and education. The general consensus concerning
outcomes from the meeting, and resulting research priorities, was
that financial well-being appears to be strongly associated with
educational outcomes. It was also concluded that helping
consumers become more financially literate should result in a
decline in financial stress at the aggregate and household level.
Federal and state policy makers and educators have known this for
years, although formal preparation on personal financial matters
has only recently been integrated into the public education system.
Today, 44 states offer some form of mandated or recommended
personal financial education in elementary and high schools,
although only 13 states require a course for graduation (Council for
Economic Education, 2009).
Albeit proportionally small in number, some American
universities have begun to address the association between
financial literacy and overall well-being by establishing campusbased financial education/counseling centers (Chamberlin, 2011).
Students have increasingly voiced their opinion that they want and
would utilize financial education services if offered at their
university (Cude, Lawrence, Lyons, Metzger, LeJeune, Marks, &
Machtmes, 2006; Goetz, Cude, Nielsen, Chatterjee, & Mimura,
2011). Over the past two decades, student financial counseling
centers have developed as higher education administrators and
educators have come to realize that college students may not be as
financially astute as necessary in the modern financial
marketplace. According to Chamberlin’s research, 14% of large
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
97
four-year public American colleges and universities have websites
devoted entirely to financial literacy. Many of these websites also
allow students to schedule an appointment to meet with a financial
counselor. One-third of large public universities and schools (33%)
offer a link on their financial aid website directing students to
external sources of financial education, such as CashCourse®.
Another 21% of schools provide financial education web links on a
different website besides the financial aid office site. Although
these statistics are impressive, it is important to note that slightly
over one-third (38%) of large public universities and colleges
provide no financial education to students.
Cunningham (2000) made the case for preventive financial
education based on her findings that nearly 70% of college
students carry at least one credit card in their own name with the
majority of those students being solely responsible for making
payments on the credit card. Students with more debt spend
additional time at work and consequently less time in school
related activities. Had students received preventive financial
education, they might have learned how to live within their means
without going into credit card debt or other forms of predatory
lending debt. More recent research adds support to this assertion in
that student dropout rates may be related to financial problems they
are experiencing (Joo, Durband, & Grable, 2008-2009).
Perhaps to make matters worse, universities and colleges
have jumped on the credit and debt bandwagon by issuing branded
credit cards. Often, these credit offers have been extended to
current students. Some have even argued that colleges or
universities have become accomplices to the growing student debt
problem by focusing on the portion of transaction fees transferred
to the university, rather than assessing the true cost to students
(Johnson, 2005). Easy credit is often used by students to fulfill
98
Journal of Personal Finance
immediate consumer needs and desires, thus creating financial
stress for many students. In extreme cases, this debt can result in
students taking their own lives (Johnson). New debt is in addition
to student loans taken to complete studies. The Credit CARD Act
of 2009 may have addressed some of these issues for college
students by establishing stricter rules under which credit may be
issued to college students (The White House, 2009), but for many
students the Credit CARD Act of 2009 came too late, and there
really is little evidence to suggest that the Act will stem the
growing tide of student consumer debt in the future.
Credit card debt, as suggested above, is a major factor of
stress in the lives of some students (Grable & Joo, 2006). Although
certain types of stress can be beneficial (i.e., eustress), few students
are prepared to effectively manage their debt situation (Palmer,
Bliss, Goetz, & Moorman, 2010). This often leads to distress or
negative stress (Borden, Lee, Serido, & Collins, 2008; Goetz,
Durband, Halley, & Davis, 2011). Students who report high levels
of financial stress are known to be more likely to report other
negative outcomes such as poor academic performance (Ross,
Cleland, & MacLeod, 2006) and physical and mental health
problems (Northern, O’Brien, & Goetz, 2010).
There are other issues that affect college student stress and
financial capacities beyond credit and debt issues. For example,
college students tend to be caught in an inflationary environment
(particularly with the rising costs of tuition) while concurrently
facing weaker employment outlooks. Low incomes and the ever
changing job market are factors college students face on a daily
basis. The result is potential stress among college students. This is
a worry for college and university administrators and those who
oversee public and private education systems. Specifically, there is
evidence to indicate that high levels of stress, specifically related
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
99
to financial difficulties, can drive students either to drop out of
school or to reduce their yearly credit load (Grable & Joo, 2006).
Both are negatives when viewed from an institutional perspective.
Not only do colleges and universities lose revenue but society loses
the human capital gain derived from a higher education degree.
Given the high level of financial sophistication required by
college students to effectively manage daily financial matters
(Lusardi, Mitchell, & Curto, 2010), in conjunction with
maintaining healthy relationships and academic achievement, it is
reasonable to conclude that some degree of financial knowledge is
needed among college students so that they can meet the demands
of the marketplace. This, in addition to the growing availability of
on-campus financial counseling across the U.S., provides the
impetus for this research. The purpose of this study was to
determine what factors predict whether students will seek peerprovided assistance as a way to reduce financial stress. In other
words, an attempt was made to determine if students who seek help
differ significantly from students who do not seek help. The results
from this study can be used to develop a profile of college students
who are likely to seek help for financial questions and problems.
The results also provide an insight into the types of financial
stressors college students face today. As a preview, findings lend
support to the hypothesis that there are groups of students who
desperately need consumer-focused financial interventions as a
way to reduce stress and maximize gains of a college education.
Literature Review
Through the identification of financially at-risk students,
college financial counseling centers can design marketing
strategies to meet the needs of these students. Further, both campus
and legislative policies can be designed to ensure that the most at-
100
Journal of Personal Finance
risk students are provided assistance. The following review of
literature addresses some of the most important consumer financial
issues facing college students today and how these issues influence
a college student’s likelihood of seeking help.
Financial Resources
There is evidence to suggest that college students worry
about their financial resource situation. Joo and her associates
(2008-2009) used a convenience sample of college students at a
large southwestern university to estimate that (a) 38% of college
students worry about their debt load, (b) greater than 50% of the
students are somewhat (34%) to extremely (17%) stressed about
their financial situation, and (c) 42% of students feel that financial
issues interfere with their academic performance. Certain
characteristics can be used to identify college students who have
financial problems. According to Lyons (2004), financially at-risk
college students hold $1,000 or more of debt other than credit
cards, are of a minority status (i.e., female, black, and/or Hispanic),
and receive need-based financial assistance. These characteristics
can also be used to predict who may be more likely to seek
financial help. For instance, Worthy, Jonkman, and Blinn-Pike
(2010) identified sensation-seeking, risky behavior, and use of
credit as characteristics predicting which college students were
likely to exhibit financial problems. Adams and Moore (2007)
found that driving after consuming alcoholic beverages, taking
amphetamines, experiencing depression in the past 30 days, having
a high body mass index, and a low grade point average were all
associated with high-risk credit behavior as measured by the
amount of unpaid credit card debt. Given these findings, it seems
apparent that about half of college students may actually be in a
position of needing financial counseling while in college to help
lessen their financial stress. Whether or not a student will ever seek
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
101
help or how consumer financial educators can predict who might
seek help are questions as yet unanswered in the literature.
Financial Attitudes
Financial attitudes, including related factors such as financial
knowledge and stress, are known to influence college students’
ability to function in the financial marketplace (Lusardi et al.,
2010). Recent studies reporting on the financial knowledge of
college students indicate that students, in general, lack
comprehension of basic financial concepts, although there is some
evidence to suggest that current college students outperform high
school students on tests of financial literacy (Mandell, 2008).
According to results from the 2008 Jump$tart Coalition survey,
high school students’ financial literacy remains low—in fact, it is
the lowest level ever. Financial literacy does appear to increase for
college students and continues to increase as students progress
through college. This may not be attributable solely to increased
classroom instruction, since the Jump$tart Coalition found that
high school students who took a personal finance course did no
better on financial literacy assessments than students who did not
take a course. Rather, results appear to indicate that experiential
learning is at least as important in increasing financial literacy as
classroom instruction, although it is important to note that students
have an interest in both learning approaches. According to Masud,
Husniyah, Laily, and Britt (2004), using a sample of college
students enrolled internationally, students stated that they wanted
to learn more about personal finance topics. In their study, 90% of
college students desired more information about saving, investing,
insurance products, budgeting, and general financial management.
As this brief review of literature suggests, there are large
groups of college students who face financial resource constraints.
Few of these students have the necessary skills to effectively
102
Journal of Personal Finance
manage both long- and short-term financial stressors. As Mandell
(2008) pointed out, financial literacy on America’s college
campuses is low. Fortunately, many leading colleges and
universities have stepped into the fray to help students obtain
information, knowledge, and resources in a way that both reduces
stress and increases financial management skills (Chamberlin,
2011). The development and growth of on-campus financial
counseling centers is, on the one hand, an improvement in helping
students; on the other hand, little is known about the types of
students who use financial counseling services. The remainder of
this paper describes the conceptual framework used to address this
important question and subsequent findings and implications.
Conceptual Framework
This study’s methodology was based on a modified
conceptual framework conceptualized by Suchman (1966), which
was developed for assessing health care help-seeking behavior and
decision-making processes. Grable and Joo (1999) later modified
and refined the model to fit into a consumer finance framework as
a way to incorporate help-seeking behavior within consumer
behavior studies. According to Grable and Joo, individuals go
through five decision-making stages to determine whether they
will seek financial help (see Figure 1).
During Stage 1 of the help-seeking process, a person may
exhibit various behaviors—both positive and negative—related to
their financial situation. Possible positive behaviors for college
students include paying bills on time and balancing checking
accounts. Examples of negative behaviors might be over drafting
checking accounts or using payday loans. Joo (1998) found that
these types of financial behaviors may be influenced by
demographic and socioeconomic characteristics, such as gender,
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
103
age, education, and income. Financial stressors and financial
knowledge may also influence a person’s financial behavior
(Grable & Joo, 1999). Stage 2 of the help-seeking process consists
of a self-evaluation of financial behavior to determine what actions
may result in positive and/or negative outcomes. In general,
women are within the model predicted to evaluate their behavior
more harshly than men (Lytton & Grable, 1997).
Next, in Stage 3 of the process, a person identifies the
cause(s) of certain financial behaviors. At Stage 4, a person must
make the decision to seek help or not to seek help related to their
financial behavior. The choice to seek help ought to be associated
with a positive outcome. Elliehausen, Lundquist, and Staten (2007)
found that participating in counseling was associated with a
positive change in borrowers’ credit profiles. This decision to seek
help may be influenced by the same factors associated with Stage 1
of the process (i.e., demographic and socioeconomic
characteristics, financial stressors, and financial knowledge). If a
person decides to seek help the process moves to Stage 5 where
different options for help are explored (e.g., financial counselor,
financial planner, retirement specialist, or friend). Again, the
expected outcome associated with seeking help is an increase in
economic, social, and emotional well-being, which has been
defined as contentment with one’s material and non-material
financial situation (Joo, 1998; Williams, 1983).
The purpose of this study was to examine Stage 4 of the
framework. Specifically, the choice to seek help via an on-campus
financial counseling center was tested. The choice dilemma at
Stage 5 was, by default, defined as seeking on-campus financial
counseling or not seeking on-campus financial counseling.
104
Journal of Personal Finance
Figure 1
Framework of college student help-seeking behavior
Method
Data
Data used for this study were obtained from two sources.
Each dataset was comprised of college students from one large
midwestern U.S. university. The first dataset included students
Feedback
NO
Feedback
who sought free financial counseling from an on-campus financial
counseling center (i.e., the clinical group). The counseling services
were offered to all students at the university and marketed through
the campus school newspaper, posters around campus, and word of
mouth. Some of the students in this sample may have made a
follow-up appointment after listening to a group presentation. The
second dataset was obtained from a research study designed to
evaluate financial planning behavior and attitudes of college
students (i.e., the non-clinical group). This random group of
students learned about the research study through posters around
campus and word of mouth. Students in the non-clinical sample
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
105
volunteered their time to complete a survey in return for a cash
incentive.
Outcome Variable
Seeking financial counseling at the on-campus financial
counseling center was the outcome variable of interest in this
study. The variable was measured dichotomously where a student
was classified as seeking financial counseling help (coded 1) or
participating in a university research study but not seeking
financial counseling (coded 0). There were no overlaps between
the datasets; that is, each data point was an independent
observation.
Independent Variables
Financial Resources. Several financial variables were
available in the clinical and non-clinical datasets, including selfperceived net worth, amount of credit card debt, amount of noncredit card installment debt, and net income. The net worth
variable was measured by asking respondents to answer the
following question: “Suppose you were to sell off your major
possessions (including your home), turn all of your investments
and other assets into cash, and pay all your debts. Would you be in
debt, break even, or have something left over?” A 10-point stairstep scale was used, with a response category of 1 indicating the
student would be in serious debt, a response of 5 or 6 indicating the
student would about break even, and a response of 10 indicating
the student would have money left over. It was hypothesized that
students with a lower level of net worth would be more likely to
seek financial counseling.
The amounts of credit card and installment debt were
measured continuously for the clinical sample, but categorically
for the non-clinical sample. The clinical sample was recoded to
match the non-clinical dataset with the following categories for
106
Journal of Personal Finance
credit card debt: 1 = $0; 2 = less than $500; 3 = $500 – $999; 4 =
$1,000 – $1,499; 5 = $1,500 – $1,999; 6 = $2,000 – $2,499; 7 =
$2,500 – $2,999; 8 = $3,000 – $3,499; 9 = $3,500 – $3,999; and 10
= $4,000 or higher. Installment debt was coded in the following
categories of 1 = $0; 2 = less than $1,000; 3 = $1,000 – $4,999;
and 4 = $5,000 or higher. It was hypothesized that students with
higher levels of credit card and installment debt and a lower level
of income would be more likely to seek financial counseling given
their reduced discretionary cash flow. Net monthly income data
were coded continuously for both data sets. Missing income scores
were replaced with the sample median income level.
Financial Attitudes. The following attitudinal variables were
used to assess common characteristics thought to influence a
student’s decision to seek financial counseling: (a) confidence in
one’s ability to meet a financial emergency, (b) satisfaction with
one’s income, (c) level of financial stress, and (d) self-perceived
financial knowledge. All financial attitude items were measured on
a 10-item Likert-type scale with 10 being the most confident,
satisfied, stressed, and knowledgeable, respectively. Students’
confidence to meet a financial emergency was measured with the
following question: “How confident are you that you could find the
money to pay for a financial emergency that costs about $1,000?”
The satisfaction question asked respondents to rate their
satisfaction with their “present financial situation.” The financial
stress question asked, “How stressed do you feel about your
personal finances, in general?” The knowledge question asked,
“How knowledgeable do you think you are about personal finances
compared to others?” It was hypothesized that students with less
ability to meet a financial emergency and those who were less
satisfied with their income would be more likely to seek financial
counseling. Students with lower levels of financial stress and
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
107
higher levels of perceived financial knowledge were hypothesized
to be less likely to seek financial counseling.
Mental Health Status. Questions were asked in the clinical
dataset as to students’ level of anxiety, sleeping difficulty, ability
to concentrate on school and/or work, level of irritability, difficulty
in controlling worries, level of muscle tension, and level of fatigue
experienced as a result of their financial situation. Questions were
asked using a scale of 1 to 7, where 1 meant they never
experienced the above symptoms and 7 meant they always
experienced the above symptoms. The non-clinical sample was
asked similar questions about their level of distress; however, the
questions were asked in a general sense, not merely related to
students’ financial situation. Moreover, the response categories
were coded 1 = not at all, 2 = several days, 3 = more than half the
days, and 4 = nearly every day. To address the problems associated
with different methods of measurement, a principal component
analysis was separately conducted for the clinical and non-clinical
samples to obtain a standardized mental health score for each
respondent. A principal component analysis score always has a
mean of 0 with a standard deviation of 1, eliminating the problem
of different measurements between the two data sets. The unique
score for each participant was then determined to be representative
of their level of distress, with a higher scores representing greater
distress. For the purposes of this study, the variable was coded
“mental health.”
Demographic Characteristics. Two demographic variables
were measured in the clinical and non-clinical datasets and used
for this study: age and sex of the student. These control factors
were chosen based on their use in previous studies (e.g., Borden et
al., 2008; Goetz et al., 2011; Palmer et al., 2010). Age was
108
Journal of Personal Finance
measured as a continuous variable. Sex was coded 1 = male and 0
= female.
Data Analysis Method
Based on the exploratory nature of this study, a non-parametric
classification and regression tree (CR&T), using AnswerTree in
SPSS, was used to determine the key characteristics separating the
clinical and non-clinical group. CR&T trees are a useful
exploratory tool because the method can accommodate nominal,
ordinal, and continuous data in the same analysis to return the
statistically significant predictors of the dependent variable. CR&T
works on a binary tree algorithm system. The algorithm partitions
data and produces accurate homogeneous subsets. According to
SPSS (1998), “C&RT partitions data into two subsets so that the
cases within each subset are more homogeneous than in the
previous subset” (p. 184). The algorithm works on a recursive
process by repeating until a “homogeneity criterion is reached or
until some other stopping criterion is satisfied” (p. 184). It is
possible for one variable to be used multiple times in a model. In
this study, C&RT was chosen as a mechanism for splitting each
node “such that each child node is more pure than its parent node.
Here purity refers to the values of the target variable” (SPSS, p.
187). In a completely pure node, all of the cases have the same
value for the target variable. C&RT measures the impurity of a
split at a node by defining an impurity measure. Statistically
significant predictors from the CR&T tree model were then used in
a post-hoc confirmatory logistic regression model to determine
each item’s impact in predicting whether a student sought financial
help.
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
109
Results
The full sample characteristics are shown in Table 1. The
sample was fairly evenly split between the clinical and non-clinical
groups (48% and 52%, respectively). Forty-seven percent were
female and most students did not hold any credit card or
installment loans. For the most part, respondents said they would
nearly break even if they were to sell all of their major possessions
and convert all of their assets into cash to pay off debt. On average,
students were moderately satisfied with their current financial
situation and net income.
Table 1
Sample Characteristics (n=140)
Sample Characteristic and Code
Clinical Sample
Non-Clinical Sample
Age
Gender
Female =0
Male =1
Monthly Income
Credit Card Debt
Installment Debt Loans
Net Worth
Emergency Preparedness
Income Satisfaction
Financial Situation Satisfaction
Financial Stress
Perceived Financial Knowledge
Mental Health Status
%
47.86
52.14
Mean
Range
23.95
18-60
$752
<$500
$0
4.80
5.78
4.19
4.86
5.42
5.28
0.00
$0-$7,440
$0-$4,000
$0-$15,000
1-10
1-10
1-10
1-10
1-10
1-10
-1.24-3.10
52.86
47.14
When the clinical and non-clinical samples were compared
(Table 2), the clinical sample was older and carried higher credit
card debt but similar amounts of installment debt. The clinical
110
Journal of Personal Finance
sample had lower net worth, was less prepared for a $1,000
financial emergency, had lower financial and income satisfaction,
and reported lower levels of perceived financial knowledge. In
regards to financial stress, the clinical sample’s (M = 5.66) stress
level was only slightly higher than that of the non-clinical sample
(M = 5.21).
Table 2
Clinical vs. non-clinical sample characteristics
Sample Characteristic and
Code
Age
Female
Male
Monthly Income
Credit Card Debt
Installment Debt Loans
Net Worth
Emergency Preparedness
Income Satisfaction
Financial Situation Satisfaction
Financial Stress
Perceived Financial Knowledge
Mental Health Status
Clinical
Mean
25.12
53.73%
46.27%
$754.66
$500-$999
$0
3.15
4.78
3.84
4.49
5.66
4.54
0.00
Non-Clinical
Mean
22.84
52.05%
47.95%
$749.89
<$500
$0
6.29
6.70
4.51
5.19
5.21
5.96
0.00
Statistical Differences Between Samples
Independent t tests were performed to determine if there was
a statistically significant difference between groups, based on each
of the variables included in the study. Four variables were found to
be significantly different between the two groups. First, age of the
clinical group (M = 25.12, SD = 6.67) was significantly higher than
the non-clinical group (M = 22.84, SD = 6.09), t140 = -2.07, p < .05.
Second, self-reported net worth of the clinical group (M = 3.15, SD
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
111
= 1.58) was significantly lower than the non-clinical group (M =
6.23, SD = 2.94), t140 = 7.68, p < .001. Third, the ability of the
clinical group (M = 4.78, SD = 2.43) to be able to pay for an
emergency of $1,000 was significantly less than the non-clinical
group (M = 6.70, SD = 3.00), t140 = 7.68, p < .001. Finally, the
perceived financial knowledge of the clinical group (M= 4.54, SD
= 1.76) was found to be significantly lower than the nonclinical
group (M = 5.96, SD = 1.92), t140 = 4.61, p < .001.
Classification and Regression Tree Results
A classification and regression tree (C&RT) and logistic
regression analysis were used to address the research question
stated at the outset of this paper (i.e., to determine what factors
predict whether students will seek financial counseling). C&RT
was used to predict the most prominent characteristics associated
with seeking services in a clinical setting. Demographic variables,
such as gender and age, as well as all other variables assessed,
including financial resources, financial attitudes, and mental health
status, were entered into the model. The CR&T, illustrated in
Figure 2, shows that net worth was the most important variable
influencing those who sought financial help. Net worth produced a
level of prediction improvement of .02. If a respondent scored 5.5
or lower (Node 1) on net worth (maximum of 10), then they were
predicted to be more likely to seek services from the counseling
center than those who scored 5.5 or higher. For those who scored
5.5 or lower on the net worth measure, mental health status became
the next most important factor, producing a prediction
improvement of .04. Those who scored more than 1.23 on mental
health status (Node 3), indicating some type of mental health
distress such as a depressive symptom, were more likely than those
who scored less than 1.23 (Node 4) to seek help from a counseling
center. For those who were more distressed, age became the next
important predictor of seeking help from a counseling center. Age
112
Journal of Personal Finance
resulted in a prediction improvement of .03. Those who were over
the age of 22.5 years (Node 6) were more likely to seek services
from the counseling center than those who were under the age of
22.5 (Node 5). For those who were under the age of 22.5 years,
perceived financial knowledge became an important predicting
factor with a prediction improvement of .03. Those who scored
less than 6.5 (maximum score of 10) were more likely than those
who scored above 6.5 (Node 8) to seek help. Income satisfaction
was also an important predicting factor of those who were likely to
seek financial counseling. Income satisfaction had a prediction
improvement of .02. Those who scored lower than 6.5 (maximum
score of 10) on income satisfaction were more likely to seek
financial help than those who scored above 6.5 (Node 10).
Post-Hoc Confirmation. A logistic regression was used to
confirm results from the C&RT model by utilizing the significant
predictors found in the C&RT output. Net worth, perceived
financial knowledge, and age were found to be significant
variables associated with whether a person would seek financial
help. In confirmation with the C&RT, the logistic regression
results showed that net worth was the most important determinant
explaining who sought financial counseling, with students
reporting a higher level of net worth being 39% less likely to seek
financial counseling (OR = .61, SE = -.77, p < .001). The next two
important determinants associated with financial help seeking
included reporting a lower level of perceived financial knowledge
and being older, respectively. Students who reported higher levels
of perceived financial knowledge were 28% less likely to seek
financial counseling services (OR = .72, SE = -.35, p < .01). Older
students were 9% more likely to seek financial counseling services
(OR = 1.09, SE = .31, p < .05). Table 3 summarizes the results
from the logistic regression model.
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
113
Figure 2
Decision tree predicting who sought financial counseling (1 =
sought help)
Table 3
Logistic regression results predicting who will seek financial
counseling
CLINICAL
Node 0
Category
%
n
1
51.27
81
0
48.73
77
Total
(100.00) 158
NW
Improvement=0.2093
<=5.5
>5.5
Node 1
Category
%
n
1
71.68
81
0
28.32
32
Total
(71.52) 113
Node 2
Category
%
1
0.00
0
100.00
Total
(28.48)
n
0
45
45
MH
Improvement=0.0351
<=-1.2329241500000001
Node 3
Category
%
1
0.00
0
100.00
Total
(3.16)
>-1.2329241500000001
Node 4
Category
%
n
1
75.00
81
0
25.00
27
Total
(68.35) 108
n
0
5
5
AGE
Improvement=0.0300
<=22.5
>22.5
Node 5
Category
%
1
59.62
0
40.38
Total
(32.91)
Node 6
Category
%
1
89.29
0
10.71
Total
(35.44)
n
31
21
52
FINKNOW
Improvement=0.0311
<=6.5
>6.5
Node 7
Category
%
1
67.39
0
32.61
Total
(29.11)
Node 8
Category
%
1
0.00
0
100.00
Total
(3.80)
n
31
15
46
INCOME
Improvement=0.0223
<=6.5
Node 9
Category
%
1
76.32
0
23.68
Total
(24.05)
>6.5
n
29
9
38
Node
Category
1
0
Total
10
%
25.00
75.00
(5.06)
n
2
6
8
n
0
6
6
n
50
6
56
114
Journal of Personal Finance
Standardized
Beta
Estimate (SE)
-.77
-.11
Odds Ratio
(OR)
Variable
Beta
Perceived Net Worth
Mental Health Status
-.49***
-.20
Age
.09*
.31
1.09
Perceived Financial Knowledge
Income Satisfaction
-.32**
-.08
-.35
-.10
.72
.93
.61
.82
*p < .05, **p < .01, ***p < .001
Pseudo R2 = .48
Discussion
This study was conducted to evaluate the characteristics of
students who sought financial counseling in comparison to the
general student population (in this study, those who were likely to
volunteer for a university sponsored research project). Once the
comparison was complete, five significant factors were identified
for use in anticipating who was likely to seek on-campus financial
counseling: (a) perceived net worth, (b) mental health, (c) age, (d)
perceived financial knowledge, and (e) income. In general, help
seekers were older, had a lower asset base, less satisfaction with
income, less knowledge, and elevated levels of mental health
distress. These findings add to the literature that describes at-risk
students (e.g., Bliss, 2010; Palmer et al., 2010). Results from the
logistic regression analysis confirmed these general themes;
however, mental health and income were not found to be
statistically significant in the post-hoc test.
Overall, the clinical sample demonstrated a much lower
perceived net worth than the non-clinical sample. Perception of net
worth could be influenced by the overall amount of debt a student
holds, including credit cards, installment loans, and student loans.
Naturally, students accrue more debt as they progress through
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
115
school with the accumulation of additional student loans.
Therefore, targeting senior seminar classes to deliver financial
education may prove beneficial in reaching those most in need, as
well as prompting some students to seek additional financial
counseling (Goetz et al., 2011). Identifying students who receive
student loans is another way to target the potentially low net worth
students.
In this study, students who sought out financial counseling
felt as though they had a lesser understanding of financial concepts
than their peers. The level of objective knowledge of participants
or their previous course experience was not known, therefore it
was not possible to test whether increased objective knowledge
(possibly through a personal finance course) results in an increased
likelihood of seeking financial counseling, as suggested by Goetz
and his associates (2011). Rather, the perceived lack of financial
knowledge among help seekers does suggest that financial
counseling centers may be attracting some students who feel they
need more information regarding personal finances.
This study also found that those seeking financial counseling
services were older than the non-clinical sample by an average of
two years. This seems counterintuitive, as older students would
seem to be more mature and responsible with their spending and
would need less financial counseling. Several suggestions come to
the forefront to explain the age difference. First, as students get
closer to graduation, they may begin to grasp their personal
financial realities. In other words, these students might be
reevaluating their personal financial position, including debt,
savings, and income, as they look to enter the “real world.” It may
be this reevaluation that drives them to seek help. Other external
factors might also be motivating help-seeking behavior, such as
self-reflection. As they begin to interview for jobs, separate from
116
Journal of Personal Finance
parental support, or perhaps begin to get married and take on the
financial dependency of a family, the realization that there is more
to be learned may come to mind. Additionally, age may simply be
associated with experience. It is likely that some older students
have amassed more debt in the latter part of their post-secondary
education than the former and thus have more debt to pay off,
creating a sense of urgency to seek financial guidance. That is,
they have gained experience through behavior that prompts a helpseeking response. In either case, targeting marketing efforts toward
senior seminar classes may aid in reaching students who have
much to benefit from financial education.
Although not statistically significant predictors in the logistic
regression, the C&RT results showed that students seeking
financial counseling were less satisfied with their income situation
and more likely to exhibit mental health distress. When a financial
emergency arises the tendency to use credit increases in response
to the liquidity crisis, especially if sufficient cash is not available to
cover the necessary expenses. These factors may work in tandem
since a person who is not able to meet a financial emergency must
find alternative cash sources (which are likely to charge high rates
of interest and fees), leading to a dissatisfaction with their financial
situation. This, in turn, may cause mental health distress in the
form of anxiety, fear, and depression. Financial and emotional
issues tend to be highly related, as confirmed by Aniol and Snyder
(1997) who found that individuals struggling with a similar issue
sometimes sought help from a financial counselor and sometimes
from a mental health counselor with no apparent reason for
choosing one over the other. As such, it is important for those
working in on-campus centers to engage in joint marketing efforts
with university mental health/counseling centers, since contact
with these outreach offices may prove the catalyst that prompts
students decide to seek help.
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
117
Conclusion
The results from this study are important for three reasons.
First, the results provide a profile of college students who are
likely to seek help for financial questions and problems via oncampus financial counseling. Student help seekers tend to be older,
less satisfied with their income, less knowledgeable, less wealthy,
and more stressed. Second, findings provide insights into the types
of financial stressors college students face today. Specifically,
issues related to debt, new job benefit information, sources of
increased income, and other factors that directly impact young
adults seem to be variables underlying search behavior. This is
somewhat of a conjecture however. Additional research is needed
to test Stage 5 of the theoretical help-seeking framework to
determine what alternatives individuals evaluate in their decision
to seek financial help. Finally, results lend support to the
hypothesis that there are groups of students who need consumerfocused financial counseling as a way to reduce stress and
maximize gains associated with a college education.
Future research efforts are needed to expand upon the results
from this study. It would be useful to document who, among those
seeking help, have completed a personal finance course while in
college. It is possible that financial awareness is a factor that
influences a student’s choice to seek help. Other areas to be
explored include conducting a longitudinal study over several
years to include years of school enrollment and post-graduation
years to assess whether or not help-seeking characteristics change
over time and to determine if the outcomes associated with
counseling are effective. Such a study would be beneficial in
testing the effectiveness of free on-campus financial counseling
centers versus financial counseling assistance that must be paid for
outside of the college environment. Additionally, replications of
118
Journal of Personal Finance
this study at other universities that provide on-campus financial
counseling are needed to ensure the validity of findings.
The limitations of the current study should also be
considered. First, it would be ideal to have an identical measure of
mental health distress available for clinical and non-clinical
samples. This study used a principal component analysis to obtain
a usable score for each respondent, although the clinical sample
was asked specifically about mental health distress caused by
financial issues whereas the non-clinical sample was asked about
general mental health distress. The sequence of the mental health
distress questions followed several financial questions, so it is
possible that non-clinical sample respondents were thinking about
finances when answering the mental health distress questions.
Secondly, all students were sampled from one midwestern
university. The students self-selected into either the help seeking
group by receiving free financial counseling, or into the nonclinical group by participating in the research study in exchange
for a small cash incentive. Students from other more liberal or
conservative schools may report different responses. Also, as an
exploratory study, the sample size was relatively small, although
representative of the university in which the sample was taken and
sufficiently large enough for the statistical analyses used in this
study. For example, Field (2009) noted that a sample size of 85 is
needed to detect medium effect sizes, whereas 10-15 participants
are needed for coefficient stability in a regression. Under both
rules, this study’s sample size was suitably large.
In summary, the results from this study suggest that college
financial counseling centers are reaching some of the students they
were designed to reach—i.e., those with fewer financial resources
and lower levels of perceived financial knowledge. An interesting
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
119
finding is that as students get older (thus closer to graduation) they
become more likely to seek financial counseling. Further research
should be conducted to determine what motivates this behavior.
These behavioral concepts could help colleges and universities
understand the emotional and financial stress students face and
lead to constructive solutions.
References
Adams, T., & Moore, M. (2007). High risk health and credit behavior among 18
to 25 year old college students. Journal of American College Health, 56(2),
101-108.
Aniol, J. C., & Snyder, D. K. (1997). Differential assessment of financial and
relationship distress: Implications for couples therapy. Journal of Marital
and Family Therapy, 23, 347-352.
Borden, L. M., Lee, S., Serido, J., & Collins, D. (2008). Does participation in a
financial workshop change financial knowledge, attitudes, and behavior of
college students? Journal of Family and Economic Issues, 29, 23-40.
Chamberlain, L. (2011). Dollars and sense: How colleges and universities
promote financial literacy. NASFAA Student Aid Transcript, 22(1).
Council for Economic Education (2009). Survey of the states, economic,
personal finance and entrepreneurship education in our nation’s schools in
2009:
A
report
card.
Retrieved
from
http://www.councilforeconed.org/about/survey2009/CEE_2009_Survey.pdf
Lawrence, F., Cude, B., Lyons, A., Marks, L., & Machtmes, K. (2006). College
students’ financial practices: A mixed methods analysis. Journal of
Consumer Education, 23, 13-26.
Cunningham, J. (2000). College student credit card usage and the need for oncampus financial counseling and planning services. Undergraduate
Research Journal for the Human Sciences. Retrieved from
http://www.kon.org/urc/cunningham.html.
Elliehausen, G. E., Lundquist, E. C., & Staten, M. E. (2007). The impact of
credit counseling on subsequent borrower behavior. Journal of Consumer
Affairs, 41(1), 1-28.
Field, A. (2009). Discovering statistics using SPSS (3rd ed.). Los Angeles: Sage.
Goetz, J., Cude, B. J., Nielsen, R. B., Chatterjee, S., & Mimura, Y. (2011).
College-based personal finance education: Student interest in three delivery
methods. Journal of Financial Counseling and Planning, 22(1), 27-42.
Goetz, J., Durband, D., Halley, R., & Davis, K. (2011). A peer-based financial
planning & education service program: An innovative pedagogic approach.
Journal of College Teaching & Learning, 8(4), 7-14.
120
Journal of Personal Finance
Grable, J. E., & Joo, S. (1999). Financial help-seeking behavior: Theory and
implications. Financial Counseling and Planning, 10(1), 14-25.
Grable, J. E., & Joo, S. (2006). Student racial differences in credit card debt and
financial behaviors and stress. College Student Journal, 40, 400-408.
Gross, K. (2009, September 20). New credit card rules may hurt financially
insecure students. The Chronicle of Higher Education. Retrieved from
http://chronicle.com/article/New-Credit-Card-Rules-May-Hurt/47039/
Johnson, C. (2005). Maxed out college students: A call to limit credit card
solicitations on college campuses. New York University Journal of
Legislation
and
Public
Policy,
8.
Retrieved
from
http://ssrn.com/abstract=925234.
Joo, S. (1998). Personal financial wellness and worker job productivity.
Unpublished doctoral dissertation. Virginia Polytechnic Institute and State
University, Blacksburg.
Joo, S., Durband, D., & Grable, J. (2008-2009). The academic impact of
financial stress on college students. Journal of College Student Retention,
10(3), 287-305.
Lusardi, A., Mitchell, O. S., & Curto, V. (2010). Financial literacy among the
young. Journal of Consumer Affairs, 44(2), 358–380.
Lyons, A. C. (2004). A profile of financially at risk college students. The
Journal of Consumer Affairs, 38(1), 56-80.
Lytton, R., & Grable, J. E. (1997). A gender comparison of financial attitudes.
Proceedings from the annual meeting of the Eastern Family Economics and
Resource Management Association, Athens, GA: 1-14.
Mandell, L. (2008). The financial literacy of young American adults: Results of
the 2008 National Jump$tart Coalition Survey of High School Seniors and
Students.
Retrieved
from
College
http://www.jumpstart.org/assets/files/2008SurveyBook.pdf
Markovich, C. A., & DeVaney, S. A. (1997). College seniors' personal finance
knowledge and practices. Journal of Family and Consumer Sciences, 21-28.
Masud, J., Husniyah, A., Laily, P., & Britt, S. (2004). Financial behavior and
problems among university students: Need for financial education. Journal
of Personal Finance, 3(1), 82-96.
National Institute of Health (2009, July 23). Clinical research & clinical trials.
Retrieved from http://www.nichd.nih.gov/health/clinicalresearch
Palmer, L., Bliss, D. L., Goetz, J. W., & Moorman, D. (2010). Helping
undergraduates discover the value of a dollar through self-monitoring.
American Journal of Business Education, 3, 103-108.
Suchman, E. A. (1966). Health orientation and medical care. American Journal
of Public Health, 56(1), 97-105.
SPSS (1998). AnswerTree 2.0 User’s Guide. Retrieved from
http://www.uic.edu/classes/idsc/ids422/trees.pdf
The White House (2009). Fact sheet: Reforms to protect American credit card
holders. Retrieved from http://www.whitehouse.gov/the_press_office/FactSheet-Reforms-to-Protect-American-Credit-Card-Holders/
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
121
Williams, F. L. (1983). Money income, nonmoney income, and satisfaction as
determinants of perceived adequacy of income. In M. Dunsing (Ed.),
Proceedings of the Symposium of Perceived Economic Well-Being, Urbana,
IL: University of Illinois at Urbana: 106-125.
Worthy, S., Jonkman, J., & Blinn-Pike, L. (2010). Sensation-seeking, risktaking, and problematic financial behaviors of college students. Journal of
Family Economic Issues, 31(2), 161-170.
122
Journal of Personal Finance
GENDER DIFFERENCES IN RISK AVERSION: A
DEVELOPING NATION’S CASE
Binay K Adhikari *
University of Alabama
Virginia O’Leary, Ph.D.
Auburn University
This study used Hanna and Lindamood (2004)’s graphic-based survey
instrument to examine whether women who are employed in the
Nepalese banking sector show more risk aversion than men. Women
indeed reported their intention to take less risk and invested less of their
wealth in risky assets than men. However, the difference disappeared
after controlling for other relevant variables, notably their perceived
knowledge of financial markets. Our analyses suggested that women
demonstrated more risk aversion than men because they considered
themselves to be less knowledgeable about financial markets. Our
findings support the need to educate female investors to increase their
confidence in their abilities to succeed in the world of finance.
Introduction
Risk aversion is a concept in economics, finance, and
psychology related to the behavior of consumers and investors
faced with uncertainty. In the context of personal financial
decisions, Dalton and Dalton (2008, p. 898, as cited in Hanna,
Waller & Finke, 2008) define risk tolerance as “the level of risk
exposure with which an individual is comfortable; an estimate of
the level of risk an investor is willing to accept in his or her
*
Binay K Adhikari, University of Alabama; bkadhikari@ua.edu
The authors want to thank Kathmandu University School of Management for
partially funding this study and James Ligon, an anonymous referee and Michael
Finke, the editor for useful comments. Remaining errors are our own.
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
123
investment portfolio.” Risk tolerance is the inverse of risk aversion
(i.e. a higher risk aversion implies lower risk tolerance).
Extensive empirical evidence suggests that women are more
risk averse than men. One study conducted by Weber, Blais and
Betz (2002) involved a survey to quantify five distinct risk
domains: financial risks, health and safety risks, recreational,
ethical and social risks. Their results indicated that women were
more risk-averse in all domains except social risk. Hinz, McCarthy
and Turner (1997), Sunden and Surette (1998), and Olsen and Cox
(2001) found that increased risk aversion affects the investment
choices of women.
In this paper we examine gender differences in financial risk
aversion among employees of the Nepalese banking sector using
Hanna and Lindamood (2004)’s graphic-based survey instrument.
To our knowledge, ours is the first study of this kind in the context
of a developing country. We find that knowledge of financial
markets and products explains the difference in risk aversion more
than the participants’ gender. Evidence of seemingly higher risk
aversion exhibited by women than men is muted once the
participants’ perceived level of knowledge of financial markets and
products is introduced as an explanatory variable.
Literature Review
Several studies have sought to determine whether the
difference in risk preferences between men and women found in
the psychological literature translates into a difference in
investment choices. Two of the early empirical studies to identify
women as more risk averse than men when making investment
choices were conducted by Cohn, Lewellen, Lease and Schlarbaum
(1975) who obtained data from a mail questionnaire survey among
124
Journal of Personal Finance
customers of a large retail brokerage firm, and Riley and Chow
(1992) who used data on actual asset allocation of a random
sample of the US population. Later studies by Hinz et al. (1997)
and Bajtelsmit and Vanderhei (1997) specifically examined the
pension choices of U.S. investors and concluded that, even after
controlling for income and age, women generally chose less risky
pension fund options. Similarly, Sunden and Surette (1998) used
household data from Survey of Consumer Finances (SCF) of 1992
and 1995 and reported that even after marital status was added to
their list of control variables women still chose less risky pension
funds.
Using the 1983 SCF data, Hawley and Fujii (1993) found that
education, income and debt were positively related to risk
tolerance. Married couples and households headed by a single male
were more risk tolerant than otherwise similar households headed
by a single female. Age was not statistically significant in the
analysis. Warner and Cramer (1995) explored the saving behaviors
of baby boomers and fond similar results.
Using the 1983 SCF risk tolerance data, Sung and Hanna
(1996) found that income and education were positively related to
risk. Generally, risk tolerance decreased with age after 45. Selfemployed persons were significantly more willing to take financial
risks than those employed by others.
In a theoretical paper Jaggia and Thosar (2000) noted an
inverse relationship between age and risk taking. However, Sunden
and Surette (1998) did not find any impact of age on risk
preferences in their empirical research.
A number of studies have investigated the role of
knowledge/education on risk taking. In a study using data from a
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
125
national survey of nearly 2000 mutual fund investors’ investing
behavior, Dwyer, Gilkeson and List (2002) found the impact of
gender on risk taking is significantly weakened when investor
knowledge of financial markets and investments is statistically
controlled. Other studies that find significant role of
knowledge/education include those by Hawley and Fujii (1993),
Sung and Hanna (1996) and Haliassos and Bertaut (1995). For
instance, in a theory cum empirical paper, Haliassos and Bertaut
(1995) showed that those who have not attended college are
significantly less likely to hold stock than those with at least a
college degree. However, using US sample data of household
holding of risky assets, Jianakoplos and Bernasek (1998) found no
relationship between knowledge and risk taking.
Daly and Wilson (2001) suggested that the increased
responsibilities accompanying marriage and children will make a
man less tolerant of risk. Supportive of this theory is the finding by
Sunden and Surette (1998) that marriage makes both men and
women more risk averse in their choice of pension plans.
Similarly, Yao and Hanna (2005) found that risk tolerance was
highest for single males, followed by married males, then
unmarried females, then married females. Säve-Söderbergh (2003)
argued, however, that marriage might encourage a couple to invest
in riskier assets because each person now has a second income
stream insuring against the loss of his or her own income.
Furthermore, Säve-Söderbergh suggested that marriage has the
potential to affect the risk preferences of men and women
differently. Jianakoplos and Bernasek (1998) also found that the
presence of children significantly increases the risk tolerance of
married couples in their investments but significantly decreases the
risk tolerance of single women. Therefore, the literature indicates
that marriage may significantly affect the risk preferences of
126
Journal of Personal Finance
individuals, even though there are potentially offsetting theories
explaining the direction of such effects.
In summary, age, income, knowledge, and marital status are
the four variables most commonly controlled for in the literature
concerned with the relative risk preferences of men and women.
Other variables, of course, may also be influencing the results. For
example, race has been shown to influence risk perceptions (Flynn
et al. 1994). Finucane et al. (2000) found white men to be the most
risk tolerant. Moreover, Riley and Chow (1992) noted that the
geographical location of an investor might influence risk
preferences, although they suggest that income is probably driving
this result.
Rationale for Studying Risk Aversion in a Developing
Country like Nepal
Studies related to gender and risk in the Western world are
abundant. It is of interest to explore whether the gender differences
observed in the developed world relate to the developing world.
This becomes especially interesting since studies including Asian
population suggest that there exists an evident cultural difference
in overconfidence. For example, Hsee and Weber (1999) found
that Chinese were significantly more risk-seeking than Americans
(in investment domain). They explained this phenomenon in terms
of “cushion hypothesis” which suggests people in a collectivist
society such as China are more likely to receive financial help if
they are in need, and they are consequently less risk averse than
those in an individualistic society such as the USA. A few other
studies led by Yates and his associates (e.g. Yates et al., 1996)
indicated that respondents in Asian cultures (e.g. in China and
Taiwan) exhibited markedly higher degrees of overconfidence than
did respondents in Western cultures (e.g. in the United States).
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
127
Consequently, the question that whether there are cultural
differences in the structure of risk aversion (such as gender
difference) is pertinent. Despite Asia’s increasing influence in the
world’s trade and financial markets, very few studies on risk
aversion have been conducted in Asian countries and especially in
south Asia. In fact, Yao (2008) noted that risk aversion studies that
include Asian population are very limited. This paper attempts to
fill this gap in the literature.
Nepal is a developing country situated between the emerging
economic powers, India and China. All of these countries have an
increasing presence of woman in the workforce. Women, because
they usually have lower working-life incomes than men (Bajtelsmit
& Bernasek, 1996), are likely to have less wealth than men when
they retire. The potential effect of this difference on the retirement
benefits of men and women is compounded by the fact that women
typically have longer life spans over which their retirement
benefits must be allocated and they also tend to retire earlier than
men (Blondal & Scarpetta 1998 as cited in Watson & McNaughton
2007). If women are indeed more risk averse in their investment
choices, this characteristic will magnify the problems associated
with their lower work-life incomes, lower retirement ages, and
longer life expectancies.
The importance of women as financial decision makers is
highlighted by a World Bank (2001)’s report that focused on
gender issues and their broad economic and social implications.
This study concluded, among other things, that it is critically
important to take gender into account in the field of social
protection and the design of public programs. The positive effect
of increased household income on the children’s welfare - their
education, health, and nutrition - is stronger if that increase is
controlled by, or channeled through, the mother. There may be a
128
Journal of Personal Finance
case, from a development effectiveness perspective, for targeting
larger funds to women or designing investment products and
schemes targeted to the attitudes and values of women, who are
likely to be more productive mobilizers of their households’
resources than men.
The worldwide literature has repeatedly found evidence for
the greater financial conservatism of women compared to men. A
conservative investment strategy results in less income on average
than a more aggressive strategy since higher returns generally
come at the cost of higher risk. This relationship between risks and
returns implies that women, who choose to bear lower risk, will
generally earn lower returns - in the long run. Life expectancy at
birth has been increasing for both males and females in Nepal. It
has increased from 42 years for males and 40 years for females in
1971i to 62 years for both males and females in 2005. Women’s
life expectancy is projected to exceed that of the men in the near
future (World Health Report, 2005 and population projection for
Nepal 2001-2021). As put by Bajtelsmit and Bernasek (1996, p. 1)
“Women’s greater longevity implies that, even with the same
investment strategy and pension accumulation, retirement wealth
must support a longer period of retirement. Women have lower
lifetime earnings, less growth in earnings, less wealth, less pension
coverage and lower pension participation rates.” To date, no
empirical study of gender differences in financial risk preference in
Nepal has been conducted. This research was undertaken to
provide baseline information of value to those working to establish
forward looking policies and procedures to accommodate the
financial interests and needs of working women and men.
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
129
Hypothesis
We hypothesize that after other factors (age, marital status,
income, investment knowledge) known to influence an individual's
risk preferences are controlled, Nepalese women will demonstrate,
on average, more risk aversion than Nepalese men.
Method
Hanna, Gutter and Fan (2001) observed that there are at least
four methods of measuring risk tolerance: asking about investment
choices, asking a combination of investment and subjective
questions, assessing actual behavior, and asking questions based on
hypothetical scenarios. They argue that inferring risk aversion
based on observing actual portfolio allocations has many
limitations, including the fact that many households have no
portfolio to allocate so that nothing can be inferred about their risk
aversion from their allocation.
We employed hypothetical questions because it has been
shown to be the firmest link to the theoretical concept of risk
aversion (see e.g. Hanna et al., 2001). Barsky, Juster and Shapiro
(1997) showed that if the respondent chooses to take an uncertain
risk that could result in a decrease in income (or a significant gain)
instead of one that is certain although less advantageous, the
expected utility of the income resulting from the riskier choice
exceeds the expected utility of having the current income stream
with certainty.ii
Instrument
We used a modified version of the Hanna et al. (2001)
pension choice measure of risk aversion which is an improvement
over Barsky et al. (1997) that follows the expected utility model.
The modified pension choice questions (Hanna & Lindamood,
130
Journal of Personal Finance
2004) include graphical illustrations to represent the quantity of the
change in the pension. These illustrations increase the respondents’
understanding of the consequences of the hypothetical alternative
outcomes and thus more accurately estimate their true risk level.iii
In addition to the series of pension choice questions, the survey
also included the SCF Investment Risk question for comparison
purposes.iv
Participants
Two hundred and six employees of 12 banks and financial
institutions in Nepal participated in this study. The sample size
reported varies in different tables since participants were free to
skip any questions they chose not to answer. Employees of the
banking sector were chosen because the questions demanded a
certain level of financial knowledge and a sense of probability and
risk. Participants could choose either Nepali or English version of
the questionnaire. MBA students of Kathmandu University
pretested the instrument.
Model and Variable Definition
Respondents’ subjective risk tolerance ranks were derived
from the risk tolerance questions and assigned a numeric rank in
ascending order coded 1 (extremely low), 2 (very low), 3
(moderately low), 4 (moderate), 5 (high), 6 (very high) and 7
(extremely high) risk tolerance. Similarly, Investment Risk
Tolerance (IRT) responses were also obtained using Survey of
Consumer Finances questions. Since both dependent variables of
interest, SRT and IRT, are ordered responses, an ordered probit
regression model was applied for the analysis. The central idea of
the model is that, underlying the ordered response is a latent,
continuously distributed random variable representing relative risk
aversion. Additionally, we also employed OLS regression as
follows since OLS estimates have a more straightforward
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
131
interpretation and well-known model diagnosis tools are available
for it.
where SRT is the respondent’s subjective risk tolerance rank
derived from Hanna and Lindamood’s risk tolerance questions;
coded from 1 for extremely low to 7 for extremely high risk
tolerance; IRT is investment risk tolerance as it appeared in the
SCF ranging from 1 to 4 (reverse coded to so that higher number
represents higher risk tolerance); FEMALE is a dummy variable
coded 1 for women and 0 for men; AGE is the respondent’s age
range coded 1 (24 and below), 2 (25-29), 3 (30-34), 4 (35-39), 5
(40-44), 6 (45-49), 7 (50 and above); MARRIED is a dummy
variable coded 1 for married and 0 for not married; INCOME is
respondent’s monthly income range coded 1 (Up to Rs.19,999), 2
(Rs. 20,000-39,999) , 3 (Rs. 40,000-59,999), 4 (Rs. 60,00079,999), 5 (Rs. 80,000-99,999), 6 (Rs. 100,000-119,999) and 7
(Rs. 120,000 and above); KNOWLEDGE is the respondent’s selfreported rating of knowledge of investment market and products
ranging from 1 to 7. 1= very little knowledge, 7 = very good
knowledge; AGE*FEMALE is the interaction of age and sex;
FEMALE*MARRIED is the interaction of sex and marital status; ε
and v are independent random disturbance terms. In addition, the
respondents also indicated an approximate percentage of wealth
they had invested in stock and business (STOCK) which we used
to tentatively assess their actual investment risk taking behavior.
132
Journal of Personal Finance
Results and Discussion
Table 1 shows the frequency distribution and means of all
variables of interest. About 37 per cent of our respondents were
females. Seventy per cent were married. The average age was
about 30 years and most made between Rs. 20,000 to 40,000
(about US $370 to $740, not adjusted for purchasing power parity)
per month. The average IRT and SRT scores were 2.64 and 4.71
respectively both of which are on the upper half of the respective
rating scales. Most of the respondents indicated that they had
invested some share of their wealth in risky assets (stock and
business). However, there were many (30) who had not invested
anything on stock or businesses. This is not uncommon in Nepal
where the stock market is small and stocks are not popular vehicles
of investment. Table 5 shows correlations among these variables.
Table 2 presents a quick comparison of the findings of Hanna
et al. (2001) and the present study. Two obvious differences
emerge. First the average relative risk aversion is lower for both
males and females in this study. This is primarily driven by more
men and women reporting extreme risk tolerance.v Second, unlike
the results of Hanna et al. the difference in risk taking propensity
of male and female is statistically significant (p-value of chisquared <.01) in this study.
Table 3 reports the results of univariate OLS regressions for
both SRT (Panel A) and IRT (Panel B). The results suggested that
sex and knowledge were significantly associated with the level of
risk tolerance (for both SRT and IRT). In particular, women
demonstrated the intention to take less risk than men and
participants with more perceived knowledge of investment markets
tend to take more risk than those with less knowledge. However,
the multivariate analysis in table 4 shows that the gender effect
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
133
disappeared after controlling for other variables. The effect of
knowledge remained significant for both SRT and IRT (from both
OLS and ordered probit models). The dy/dx columns indicate that
the probability of being in the highest risk tolerant group (SRT=7,
IRT=4) increased as knowledge increased. Age was not significant
for SRT but was weakly significant for IRT and had the expected
sign for IRT suggesting older employees intended to take less risk
than younger ones. We found no significant problem of
heteroscedasticity (chi-squared statistic is not significantly large)
and multicollinearity (tolerance for none of the variables is less
than .10) with OLS.
Although our findings mainly suggest that only knowledge
has a significant impact on risk tolerance, the univariate analysis
indicates that sex also might also have impact on risk aversion.
Thus, gender affects risk taking propensity although not in the
magnitude hypothesized based on the review of the literature.
Directionality of the coefficient is consistent with the hypothesis:
men tend to be more risk tolerant than women. Perceived
knowledge of the investment market and financial products has
been shown to have a significant impact on risk aversion. The
positive sign of the regression coefficient suggests that risk
tolerance increases with the increase in perceived knowledge of
investment related information.
The cross correlations presented in table 5 explain some of
the findings further. The significant positive relation between age
and knowledge (r = .28, p<.01) indicates that older bank
employees are more knowledgeable about financial matters than
their younger colleagues. Likewise, there is a positive relationship
between knowledge and risk tolerance (SRT: r = .28, p < .01, IRT:
r = .284, p<.01), which suggests that people with more investment
knowledge endorse more risky options. However, the direct
134
Journal of Personal Finance
relationship between age and risk tolerance is at most weakly
significant both in univariate and multivariate settings.
Furthermore, both reported risk tolerance measures, the IRT
(r = .18, p <.05) and the SRT (r = .13, p <.10), for the employees
are positively related to the proportion of wealth kept in risky
assets (STOCK) i.e. stocks and direct investment in business.
Those who rate themselves as high risk takers invest a larger
portion of their wealth in risky assets (stocks and business) than
those who rate themselves as more risk averse. Table 6 shows that
there is a significant difference in the proportion of risky
investment held by men and women.
Women, who generally consider themselves less
knowledgeable about finance than men exhibit greater risk
aversion. However, the relationship is stronger between knowledge
and risk tolerance than between gender and risk tolerance.vi It is
therefore not surprising that in a review of a wide range of
literature (see e.g. Lyons et al., 2008) conclude that the research on
risk tolerance indicating women are more risk averse than men is
far from clear.
All in all, our findings are consistent with the stream of
literature which concludes that knowledge, not gender, might be
the key determinant of risk-taking propensity. Consistent with this,
Elder and Rudolph (2003) employed the “final say” question in the
Health and Retirement Study (HRS) to identify the sources of
decision-making power within households. They found that
decisions were more likely to be made by the spouse with greater
financial knowledge, more education, and a higher wage,
irrespective of gender.
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
135
We did not obtain significant interaction effect of either sex
and marital status or sex and age on risk tolerance. This is
plausible since traditionally, in Nepalese households like in most
South Asian households, husbands and/or fathers make most of the
financial decisions so wives and daughters remain less
knowledgeable about finances and less active in financial decisionmaking before and after marriage.
Conclusion
The major finding of the current study, that women exhibit
less financial risk tolerance than men is apparently occasioned by a
disparity in perceived knowledge about investments. Our results
are consistent with the stream of existing literature that argue that
the difference in risk aversion is attributable to task familiarly and
knowledge rather than gender. In fact, Deaux and Emswiller
(1994) and Beyer and Bowden (1997) have noted that women are
less confident and more risk averse in domains considered
masculine, regardless of their (equal) ability to perform. Men are
represented in greater numbers in financial markets than women
(Merrill Lynch, 1996). In addition, women are likely to be
perceived as more conservative investors and therefore offered less
risky investments by brokers (Wang, 1994).
Psychological research has shown that men tend to be more
overconfident than women in areas like finance. Theory predicts
that overconfident people trade excessively. Barber and Odean
(2001) tested this prediction using account data from a discount
brokerage firm and found that men traded 45% more than women.
Trading reduced men's (risky) net returns by 2.65 percentage
points a year compared to 1.72 percentage points for women.
136
Journal of Personal Finance
Our results from a developing nation reinforce the need of
specific and targeted financial education to help men and women
with their investment decisions. Men may need to be cautioned
about the pitfalls of overconfidence, while women may need
guidance on how to make investment choices that carry a
calculated risk to obtain adequate growth.
Endnotes:
_____________
i
Source:
Country
Health
System
Profile,
http://www.searo.who.int/EN/Section313/Section1523_6868.htm
Nepal
ii
Let U be the utility function and C be permanent consumption. An expected
utility maximizer will choose the 50-50 gamble of doubling lifetime income as
opposed to having it fall by factor 1 - λ if: .5 U (2C) + .5 U (λ C) ≥ U(C)
Assuming the utility function is constant relative risk aversion (CRRA), the
following shows the relation between the Arrow-Pratt measure of relative risk
aversion A and λ: λ = (2 – 2(1-A))[1/(1-A)]. This equation holds if A≠ 1, and λ = .5
when A = 1. Therefore, by asking questions with different levels of λ, the
Arrow-Pratt coefficient of relative risk aversion can be directly estimated.
iii
Illustrative of the questions contained in the questionnaire is the following:
Suppose that you are about to retire, and have two choices for a pension.
Pension A gives you an income equal to your pre-retirement income. Pension B
has a 50% chance your income will be double your pre-retirement income, and
a 50% chance that your income will be 20% less than your pre-retirement
income. You will have no other source of income during retirement, no chance
of employment, and no other family income ever in the future. All incomes are
after-tax. Which pension would you choose?
Subsequent questions pose different percentage reductions in income. There
were six questions out of which the respondent is required to answer a
maximum of four questions. The respondent who accepts the possibility of
largest cut in income for a possibility of doubling the income gets the higher
possible point (on a scale of 1 to 7) indicating an extremely high subjective risk
tolerance (SRT).
iv
The Survey of Consumer Finances (SCF) is used to gather data on assets,
liabilities, financial attitudes, and financial behaviors of individuals and families.
The SCF risk assessment item has been widely used as a proxy for financial risk
tolerance. The question appears as follows:
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
137
Which of the following statements on this page comes closest to the amount of
financial risk that you are willing to take when you save or make investments?
1. Take substantial financial risk expecting to earn substantial returns
2. Take above average financial risks expecting to earn above average returns
3. Take average financial risks expecting to earn average returns
4. Not willing to take any financial risks
(Note: In this paper we reverse coded the above responses (4 being substantial
financial risk to make it comparable to SRT measure)
v
It seems unlikely that such a large portion of respondents (24.7%) had a true
relative risk aversion (RRA) of less than 1. RRA of 1 or less suggests that the
household should invest more than their wealth in risky assets and buy on
margin (see e.g. Hanna and Chen, 1997) – an implication contrary to most
empirical evidence. Precise calculation of RRA is not critical for our purpose.
However, for robustness, we repeated the analyses that follow by discarding the
observations with implied RRA of less than 1. Our key findings were not
different from the ones with full data. We, therefore, decided to report the results
that come from all available data. We thank a reviewer for pointing out this
important issue.
vi
It is credible that knowledge is endogenous since a person’s attitude towards
risk may influence the willingness to acquire knowledge about financial assets.
We realized that it is a very important issue and decided to do a second round of
analysis to see if there exits an endogenity problem. For the regression with
SRT, we used age as the instrumental variable for knowledge and for IRT, used
marital status as the instrument. These variables were carefully chosen so that
we have an instrument which is correlated with knowledge but is not likely to be
correlated with error terms and does not have any independent explanatory
power on the dependent variables (SRT and IRT). Marital status is likely to be
endogenous in western cultural context but unlikely to be so in Nepalese context
where most marriages are arranged and happen within a certain age range. Data
limitations prevented us from doing more rigorous analysis. The Hausman test
obtained no evidence of endogenity for our model.
138
Journal of Personal Finance
References
Bajtelsmit, V. L. & Bernasek, A. (1996). Why do women invest differently than
men? Financial Counseling and Planning, 7, 1-10.
Bajtelsmit, V.L., & Vanderhei, J.L. (1997). Risk aversion and pension
investment choices. In M.S. Gordon, O.S. Mitchell, & M. M. Twinney
(Eds.), Positioning pensions for the twenty-first century (pp. 55-66).
Philadelphia: University of Pennsylvania Press.
Barber, B. & Odean, T. (1999). Boys will be boys: Gender overconfidence and
common stock investment. Quarterly Journal of Economics, 116, 261-292
Barsky, R.B., Juster, F.T., Kimball, M.S. & Shapiro, M.D. (1997). Preference
parameters and behavioral heterogeneity: An experimental approach in the
health and retirement study. Quarterly Journal of Economics, 112 (2), 537579.
Bernasek, A., & Shwiff S. (2001). Gender, risk, and retirement. Journal of
Economic Issues, 35(2), 345-356.
Beyer, S & Bowden, E. (1997). Gender differences in self-perceptions:
Convergent evidence from three measures of accuracy and bias. Personality
and Social Psychology Bulletin, 23, 157-172.
Bruce, A. & Johnson, J. (1994). Male and female betting behavior: new
perspectives. Journal of Gambling Studies, 10, 183-98.
Byrnes, J.P., Miller, D.C. & Schäfer, W.D. (1999). Gender differences in risk
taking: A meta- analysis. Psychological Bulletin, 125(3), 367-383.
Cohn, R. A., Lewellen, W.G., Lease, R.C., & Schlarbaum, G.G. (1975).
Individual investor risk aversion and investment portfolio composition.
Journal of Finance, 30(2), 605-620.
Daly, M., & Wilson, M. (2001). Risk-taking, intra-sexual competition, and
homicide. Nebraska Symposium on Motivation, 47, 1-36.
De Bondt, W.F.M. (1998). A portrait of the individual investor. European
Economic Review, 42(3-5), 831-844.
Deaux, K., & Emswiller, T. (1974). Explanations of successful performance on
sex-linked tasks: What is skill for the male is luck for the female. Journal of
Personality and Social Psychology, XXIX, 80–85
Dwyer, P.D., Gilkeson, J.H., & List, J.A. (2002). Gender differences in revealed
risk taking: Evidence from mutual fund investors. Economics Letters, 76(
2), 151-158.
Elder, H. W., & Rudolph, P. M. (2003). Who makes the financial decisions in
the households of older Americans? Financial Services Review, 12(4), 293308.
Finucane, M.L., Flynn, J., Mertz, C.K., Satterfield, T.A. & Slovic, P. (2000).
Gender, race, and perceived risk: The 'White Male' effect. Health Risk and
Society, 2(2), 159-1 72.
Flynn, J., Slovic, P. & Mertz, C.K. (1994). Gender, race and perception of
environmental health risks. Risk Analysis, 14(6 ), 1101-1108.
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
139
Goldsmith, E. B., & Goldsmith, R. E. (1997). Gender differences in perceived
and real knowledge of financial investments. Psychological Reports,
80(11), 236-238.
Goldsmith, E. B., Goldsmith, R. E., & Heaney, J.-G. (1997). Sex differences in
financial knowledge: A replication and extension. Psychological Reports,
81, 1169-1170.
Grable, J. E., & Lytton, R. H. (1999). Financial risk tolerance revisited: The
development of a risk assessment instrument. Financial Services Review, 8,
163-181
Hanna, S. & Chen, P. (1997). Subjective and objective risk tolerance:
Implications for optimal portfolios. Financial Counseling and Planning,
8(2), 17-25.
Hanna, S. D. & Lindamood, S. (2004). An improved measure of risk aversion.
Financial Counseling and Planning, 15 (2), 27-38.
Hanna, S.D., Gutter, M.S., & Fan, J. X. (2001). A measure of risk tolerance
based on economic theory. Financial Counseling and Planning, 12 (2), 5360.
Hanna, S. D., Waller, W., & Finke, M. (2008). The concept of risk tolerance in
personal financial planning. Journal of Personal Finance, 7 (1), 96-108
Haliassos, M. & Bertaut, Carol C, 1995. Why Do So Few Hold Stocks?,
Economic Journal, Royal Economic Society, 105(432). 1110-29
Hawley, C.B. & Fujii, E.T. (1993). An empirical analysis of preferences for
financial risk: Further evidence on the Friedman-Savage model. Journal of
Post Keynesian Economics, 16(2). 197-204.
Hersch, J. (1996). Smoking, seat belts, and other risky consumer decisions:
Differences by gender and race. Managerial and Decision Economics,
17(5), 471-481.
Hinz, R. P., McCarthy, D. D., & Turner, J. A. (1997). Are women conservative
investors? Gender differences in participant-directed pension investments.
In Mitchell, O. (Ed.) Positioning Pensions for the Year 2000 (pp. 91–103).
Philadelphia: University of Pennsylvania Press.
Hsee, C. K. & Weber, E. U. (1999). Cross-national differences in risk
preferences and lay predictions for the differences. Journal of Behavioral
Decision Making, 12, 165-179
Jaggia, S., & Thosar, S. (2000). Risk aversion and the investment horizon: A
new perspective on the time diversification debate. Journal of Psychology
and Financial Markets, 1(3-4), 211-215.
Jianakoplos, N.A., & Bernasek, A. (1998). Are women more risk averse?
Economic Inquiry, 36(4), 620-630.
Johnson, J. E.V. & Powell, P. L., (1994). Decision making, risk and gender: Are
managers different? British Journal of Management 5 (2), 123-38.
Lewellen, W., Lease, R., Schlarbaum, G. (1977). Patterns of investment strategy
and behavior among individual investors, Journal of Business, 50(3), 296333.
140
Journal of Personal Finance
Lyons, A. C., Neelakantan, U. & Scherpf, E. (2008). Gender and marital
differences in wealth and investment decisions: Implications for
researchers, financial professionals, and educators. Networks Financial
Institute Working Paper No. 2008-WP-02.
Merril Lynch (1996). You and your money. A financial handbook for women
investors, New York.
Olsen, R.A., & Cox, C.M. (2001). The influence of gender on the perception and
response to investment risk: The case of professional investors. Journal of
Psychology and Financial Markets, 2(1), 29-36.
Powell, M. & Ansic, D. (1997). Gender differences in risk behaviour in financial
decision-making: An experimental analysis. Journal of Economic
Psychology. 18, 605 – 628.
Riley, W., & Chow, K.V. (1992). Asset allocation and individual risk aversion.
Financial Analysts Journal, 48(6), 32-37.
Säve-Söderbergh, J. (2003). Essays on gender differences in economic decisionmaking. PhD thesis, Stockholm University.
Sunden, A.E., & Surette, B.J. (1998). Gender differences in the allocation of
assets in retirement savings plans. American Economic Review, 88(2), 207211.
Sung, J. & Hanna, S. (1996). Factors related to household risk tolerance: An
ordered probit analysis. Proceedings of the 42nd Annual Conference of the
American Council on Consumer Interests, 221-228.
Wang, P. (1994). Brokers still treat men better than women. Money, 23 (6), 10810.
Warner, N. & Cramer, S. (1995). Saving behaviours. Journal of Consumer
Studies and Home Economics, 19, 57-67.
Watson, J., & McNaughton, M. (2007). Gender differences in risk aversion and
expected retirement benefits. Financial Analysts Journal. 63(4)
Watson, J., & Robinson, S. (2003). Adjusting for risk in comparing the
performances of male- and female-controlled SMEs. Journal of Business
Venturing, 18(6), 773-788.
Weber, E.U., Blais, A.R. & Betz, N.E. (2002). A domain-specific risk-attitude
scale: Measuring risk perceptions and risk behaviors. Journal of Behavioral
Decision Making, 15, 263–290
World Bank (2001a). Engendering Development: Through Gender Equality in
Rights, Resources, and Voice. Policy Research Report: Oxford University
Press.
Yao, R. (2008). Financial issues of Asian Americans. In J. J. Xiao (Ed.)
Handbook of consumer finance research (pp. 303-316). NY, NY: Springer.
Yao, R. & Hanna, S.D. (2005). The effect of gender and marital status on
financial risk tolerance. Journal of Personal Finance, 4 (1), 66-85.
Yates, J. F., Lee J. W., & Shinotsuka, H. (1996). Beliefs about overconfidence,
including its cross-national variation. Organizational Behavior and Human
Decision Processes, 65,138-147.
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
141
Table 1: Frequency Distribution of Key Variables
1
7
1
31
Male (0)
126
Age
(years)
24 and
below (1)
2
81
2
5
Female (1)
75
25-29 (2)
76
3
65
3
8
30-34 (3)
46
4
29
4
31
35-39 (4)
18
5
38
40-44 (5)
24
6
33
45-49 (6)
50 and
Above (7)
11
IRT
Freq.
SRT
Freq.
Sex
Freq.
Freq.
20
7
48
Total
182
Total
194
Total
201
Total
201
Mean
2.64
Mean
4.71
Mean
0.37
Mean
3.03
Freq.
Knowledge
Rating
Marital
Status
Not
Married
(0)
Married
(1)
71
129
Freq.
Monthly
Income
(in Rs.)
6
Freq.
%Invest
Stock &
Business
None (0)
30
1
31
Up to
Rs.19,999 (1)
54
Freq.
2
5
20,000-39,999 (2)
70
1 - 20% (1)
58
3
4
8
31
40,000-59,999 (3)
60,000-79,999 (4)
28
11
21 - 40% (2)
41 - 60% (3)
48
31
5
38
8
33
61 - 80% (4)
80 -100%
(5)
13
6
7
48
80,000-99,999 (5)
100,000-119,999
(6)
120,000 and
above (7)
13
6
4
Total
190
Total
194
Total
190
Total
184
Mean
0.70
Mean
4.35
Mean
2.57
Mean
2.73
Numbers in parenthesis represent coded values for analysis. Means should be interpreted
based on the coded values
See the text (model and variable description section) for more details about the key
variables.
Totals do not match because respondents were free to skip any question they did not want
to answer.
142
Journal of Personal Finance
Table 2: Comparison of Preliminary Findings with Hanna et
al. (2001)
Our sample*
Male
n
Extremely High,
accept 50% cut
(A< 1.0)
Very High, reject
50% cut (1.0≤A <
2.0)
Moderately High,
reject 33% cut
(2.0≤A< 3.8)
Moderate, accept
10% cut (3.8 ≤A<
7.5)
Low, accept 8%
cut (7.5 ≤A< 9.3)
Very Low, accept
5% cut (9.3 ≤A
14.5)
Extremely Low,
reject 5% cut
(A>14.5)
Entire sample
Mean Relative
Risk Aversion
Hanna et al.(2001)**
Female
%
Male
Female
n
%
n
%
n
%
41
33.3%
7
9.9%
0
0.0%
2
1.4%
22
17.9%
11
15.5%
12
5.8%
6
4.1%
17
13.8%
21
29.6%
48
23.2%
30
20.5%
16
13.0%
15
21.1%
94
45.4%
60
41.1%
4
3.3%
4
5.6%
14
6.8%
21
14.4%
4
3.3%
1
1.4%
18
8.7%
10
6.8%
19
15.4%
12
16.9%
21
10.1%
17
11.6%
123
100.0%
71
100.0%
207
100.0%
146
100.0%
4.84a
5.72a
6.55b
6.88b
*This study: n = 194. ** Web Survey: n = 353.
a. Weighted average based on midpoints of each range, except value of 0.9 used for
“Extremely High” category, and 16.0 used for “Extremely Low” category. Chi-square for
difference in male and female significant at .01.
b. Weighted average based on midpoints of each range, except value of 0.9 used for
“Extremely High” category, and 16.0 used for “Extremely Low category”. Notice, Hanna
et al. say they used 16.0 used for “Very Low category” which is a typo (We confirmed
with one of the coauthors). Chi square statistic not significant.
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
143
Table 3: Univariate Analysis
Unstandardized Coefficients
Variables (Univariate)
B
Std. Error
T
Sig.
SRT and FEMALE
-.625
.303
-2.062
.041
SRT and AGE
.110
.097
1.139
.256
SRT and MARRIED
-.147
.235
-.625
.533
SRT and KNOWLEDGE
.538
.125
4.322
.000
SRT and INCOME
-.024
.087
-.281
.779
IRT and FEMALE
-.243
.121
-2.013
.046
IRT and AGE
-.022
.039
-.561
.575
IRT and MARRIED
.038
.093
.407
.685
IRT and KNOWLEDGE
.191
.048
3.979
.000
IRT and INCOME
.011
.034
.316
.752
Panel A: SRT
Panel B: IRT
144
Journal of Personal Finance
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
145
146
Journal of Personal Finance
©2011, IARFC. All rights of reproduction in any form reserved.
Volume 10, Issue 2
147
Table 6: Independent Samples Test for Differences of Men and
Women
Levene's Test for
Equality of Variances
AGE
STOCK
Equal
variances
assumed
Not assumed
Equal
variances
assumed
Not assumed
t-test for Equality
of Means
Sig.
(2-tailed)
F
Sig.
t
df
27.64
0.000
5.533
6.133
204
203
0.000
0.000
2.27
0.000
2.127
2.315
182
182
0.0348
0.0217