Risk and Decision Making by Finance Executives
Author: Les Coleman
Mailing Address: Department of Finance
The University of Melbourne
Parkville. Victoria. 3010
phone: +61413 901085/+613 8344 3696
fax: +613 8344 6914
Electronic copy available at: http://ssrn.com/abstract=975133
Purpose – When finance managers face decisions, they do not always make clinical
evaluations using rational methodology, but systematically depart from utility
maximisation. This article addresses biases that are related to risk propensity, and
categorises them under five headings: decision makers’ characteristics and
perception; reference levels; mental accounting and the assumption of mean
reversion; the longshot bias or overconfidence; and the desire for immediate
gratification. The research reported in the paper seeks to understand the
mechanisms of these biases using a study of decision making by Australian finance
executives in a setting that is representative of a typical business decision.
Design/methodology/approach - This paper uses a case study that was designed
to identify why decision makers facing choices will prefer a risky alternative. Data
were collected using e-mail contact and an electronic survey. Respondents (n=67)
provided demographic data, and answered questions that probed their attitudes and
decision styles. Risk propensity was quantified by respondents’ attitude towards a
risky decision, and was explained using independent variables related to decision
Findings – Just over half the executives proved willing to take a risk, and almost half
the variance in their risk propensity was explained roughly equally by respondents’:
endowment, perception of risk’s role in decisions, assessment of alternative choices,
and expectation of the decision’s outcome. Manipulation of the cases along four
dimensions varied the decision’s facts, but they proved only marginally significant to
Electronic copy available at: http://ssrn.com/abstract=975133
Originality/value - The study provides a practical explanation of the risk taking
behaviour of finance executives; confirms that context is more important to decisions
than their content; and adds to the growing body of applied behavioural research in
Keywords: decision, risk, manager decision making, applied behavioural finance
Risk and Decision Making by Finance Executives♣
Qualitative aspects of decision making in finance – which form part of what is now
termed applied behavioural finance - have long been seen as important to the actions
of institutions and individual investors. For instance, Slovic (1972: 779) wrote: “Many
aspects of investment analysis are said to be psychological in nature”, and then
provided a catalogue of biases that remains a good description of the field. Evidence
has continued to grow that financial decision makers (both managers and investors)
do not make clinical calculations using rational methodology, but systematically
depart from utility maximisation.
Even though it seems intuitively important to understand the processes followed by
finance executives when they make decisions in the face of risk, the subject has
seen limited real-world examination. Financial economists have largely modelled
asset pricing without examining how investors make their choices (Barberis and
Thaler, 2003), and this is reflected in finance education. For example, Damodaran
(2001: vii) began his authoritative text on corporate finance by defining the subject
as: “encompass[ing] all of a firm’s decisions that have financial implications.” But, in
keeping with most other treatments, he only provided a normative depiction of how
financial decisions should be made.
There is, however, a growing body of descriptive material. Corporate behavioural
finance (Baker, Ruback and Wurgler, 2004) has developed studies covering topics
ranging from bank lending practices (McNamara and Bromiley, 1997) to decisions by
institutional investors (Shapira and Venezia, 2001) and equity and bond traders
(Willman, Fenton-O'Creevy, Nicholson and Soane, 2002).
Other researchers have sought more specific explanations for managerial decision
making. Statman and Sepe (1989) analysed the announcement effect of termination
of poorly performing investments and found (in conformance with the disposition
effect which is discussed below) that returns rose because shareholders rewarded
managers for decisive behaviour. Lintner (1956) showed that managers’ decisions to
change dividends are related to historical payments (typically targeting a steady
increase) and signal permanent changes in earnings; managers’ dividend strategy
has changed little over the years (Benartzi, Michaely and Thaler, 1997). Grable
(2000) evaluated risk taking in personal financial decisions using university faculty
Numerous authors have post-audited key strategic decisions – particularly mergers
and acquisitions – and generally concluded that only a small proportion are
financially successful (Jarrell, Brickley and Netter, 1988; and Rau and Vermaelen,
1998). Other studies have found that managers are prone to use rules-of-thumb or
heuristics such as targets for capital structure and payback period for investment
hurdles (Graham and Harvey, 2001). They also suffer from a herd mentality that
leads to waves of similar decisions such as IPOs, mergers and investment bubbles
(Loughran, Ritter and Rydqvist, 1994).
Despite this considerable body of work, few studies have used subjects who are
experienced managers; and most of the analyses have employed univariate test
statistics to examine the significance for risk propensity of individual aspects of
decision makers and decisions. Thus there remain three significant research gaps.
First relatively little is known about how finance managers actually reach decisions,
particularly what decision stimuli they place most weight upon. The second gap is a
limited understanding of how personal attributes and cognitive processes lead to the
heuristics and biases that managers seem to employ. As a resuly surprisingly little
attempt has been made to explain managerial risk taking. The third gap is that no
integrated model has been developed that shows the interaction and relative
importance of the many known influences on risk and decision making.
This paper makes two contributions to the literature. The first is to examine the
decision processes of practising finance managers using a representative decision.
The second is to incorporate a respondent questionnaire with strong grounding in the
behavioural literature and sufficient granularity to provide reliable measures of
respondents’ characteristics and decision modes. This specifically incorporates risk
propensity – not solely job-related competencies – as a key feature of decision
making; and enables development of an integrated model of decision making by
finance managers when facing risk. Results will add to the growing body of work in
behavioural finance that seeks to understand the sources and effects of managers’
The balance of the paper proceeds as follows: the first section provides a review of
literature on decision making by managers who are facing risk; the second section
describes the objectives and design of the empirical research; section three
discusses the results, whilst section four addresses their implications for researchers
I. Review of Financial Decision Making Under Risk
There is now an extensive literature on behavioural finance, with surveys provided by
Barberis and Thaler (2003), Camerer (1995), Rabin (1996), and Ricciardi (2004).
Shefrin (2001) has collated some of the most influential papers. Although much of
this field examines biases at the market level, there is a growing body of applied,
micro-level work that examines financial decisions by managers, including studies by
Baker, Ruback and Wurgler (2004) and de Bondt and Thaler (1994).
It has long been recognised that managers adversely impact firm performance
because they do not unconditionally seek to maximise shareholder value. Although
this has identified a large number of sub-optimising behaviours, most attention has
been paid to mathematically tractable factors, particularly managers’ conflicts in their
role as agents (Jensen and Meckling, 1976). Scholars have largely ignored the
salience of recognised, but more qualitative, biases in managers’ decision making
such as hubris (Roll, 1986), overconfidence (Kahneman and Lovallo, 1993) and firm
outlook (Singh, 1986). The contention of this paper is that the latter attributes are of
considerable importance to maximisation of value, and hence merit more detailed
understanding so that they can be actively managed.
The goal of this section is to summarise the literature in finance (and, where
appropriate, related disciplines) that relates to the real world behaviour of finance
managers who face risks.
Given the many behavioural anomalies that are known to affect managerial risk
propensity, it is desirable to provide some structure. In this section, the anomalies are
grouped under five headings: decision makers’ characteristics (especially
demography and personality) and perception; reference levels; mental accounting
and the assumption of mean reversion; the longshot bias or overconfidence; and the
desire for immediate gratification. This approach is consistent with recent studies
such as Kahneman (2003) that seek to understand the mechanisms of biases, rather
than just catalogue them.
1.1 Decision Maker Characteristics and Perception
A variety of studies of real-world decision making have shown that up to a quarter of
the variation in individuals’ risk taking is explained by personality factors. According
to Trimpop (1994), risk takers are psychologically flexible. They are usually better
educated, have a history of successful risk taking, tolerate ambiguity, seek novel
experiences, and rapidly respond to stimuli. They are described as adaptable,
adventurous, aggressive, informal, optimistic and sociable. Risk takers have a need
to be better and faster, and agree with statements such as `I set difficult goals for
myself which I attempt to reach’; they are typical Type-A personalities (see also:
Smith and Friedland, 1998 and Williams and Narendran, 1999).
Also important to risk propensity are a variety of relatively stable personal traits.
Byrnes, Miller and Schafer (1999) conducted a meta-analysis of 150 studies to
compare male and female risk taking and found that the proportion of women
accepting any risk is an average of six percent less than the proportion of men
offered the same risky choice. The general consensus from other studies is that
increasing risk propensity is linked to higher income and education (Hartog, Ferrer-i-
Carbonell and Jonker, 2000). Risk takers are likely to be younger, single, and in a
professional occupation (Grable, 2000); they also exhibit multiple risk behaviours and
so are more likely to pursue dangerous occupations and sports, and adopt risky
personal habits such as smoking, reckless driving, and sexual activity (Zuckerman
and Kuhlman, 2000).
An explanation for the mechanism whereby personality determines risk propensity is
the proposal by Lopes (1987: 275-6) that
“Risk-averse people appear to be motivated by a desire for security, whereas risk-
seeking people appear to be motivated by a desire for potential. The former motive
values safety and the latter, opportunity … Risk-averse people look more at the
downside and risk seekers more at the upside.”
Significantly, this mechanism is consistent with physiological evidence that is
obtained from magnetic resonance imaging (MRI) studies which use radiation to
examine humans at the molecular level. These found that different parts of the brain
– and hence different decision criteria (e.g. emotion or logic) – are used when
decision features vary (Greene, Sommerville, Nystrom, Darley and Cohen, 2001).
Other studies detected a “switch” in the brain that alternates decision maker
perspective between the optimistic, big picture and a cautious, detail focus (Miller,
Liu, Ngo, Hooper, Riek, Carson and Pettigrew, 2000).
The perception of decisions is important. Framing, for instance, involves presenting
identical data with a different emphasis, which shifts a decision maker’s expectation
of the outcome. Thus a positive frame (i.e. projecting gains) induces greater
weighting for a successful outcome (Kühberger, 1998). Decision makers also place
their own frame on a decision by editing the materials to be analysed through
instinctive perception of the costs and benefits of each outcome (Finucane,
Alhakami, Slovic and Johnson, 2000), and by self framing (Wang, 2004).
Another subjective influence is perceived control: people are more willing to take a
risk when they have a competence through relevant skill or knowledge (Heath and
Tversky, 1991), or when they are able to exercise some control over the outcome
(Slovic, 2000). Although most analyses of financial expertise conclude that its value
is marginal at best (De Bondt, 1991), support from experts adds to perceived control
and is especially valuable to risk-neutral decision makers (Eeckhoudt and Godfroid,
A well recognised influence on decision making is bounding (Simon, 1959), where
individuals are simply unable to access or process all available information.
Economically, this implies that the costs of gathering additional knowledge about a
decision are not expected to provide a reasonable return (Conlisk, 1996). Decision
makers become more risk averse when limits are placed on them (Mano, 1990);
similarly they do not like ambiguity (Ellsberg, 1961), and so will avoid outcomes with
1.2 Reference Levels
As discussed above, decision makers’ personality, training and experience incline
them to be instinctively risk averse or risk seeking. This section introduces more
transient influences which alter decision makers’ perspective in light of needs and
decision setting (e.g. recreation or retirement savings). This second group of decision
influences relate to changes in risk propensity around a reference level.
A useful model is risk-sensitive foraging that is used to explain risk taking by animals,
and assumes they choose foraging strategies in light of food needs (Bateson, 2002).
An excellent example was provided in a breakthrough experiment where researchers
established birds in a laboratory habitat, and taught them to feed by pecking on keys
that delivered either a fixed number of pellets, or a variable quantity (Caraco,
Blanckenhorn, Gregory, Newman, Recer and Zwicker, 1990). The experimenters
then threatened the birds’ survival by reducing the laboratory’s temperature. The
birds proved to be risk sensitive as their state changed: risk averse when food was
adequate and temperatures warm, and risk embracing as the temperature dropped
and food supplies became critical.
The essence of risk-sensitive foraging is that decision makers alter risk preference
around their satisficing level, or endowment which meets requirements at the time.
This means that - when endowment is inadequate to sustain the decision maker - risk
propensity is high; however, when endowment is adequate and survival is probable,
a lower risk strategy is preferred.
Despite its intuitive appeal, risk-sensitive foraging has only seen limited explicit
application to human behaviours; one of the few examples is the explanation of
political lobbyists’ strategies by Gray and Lowery (1998). However, there is
considerable evidence that the mechanism is relevant to humans and organisations
who – like animals - will take greater risks when their survival is threatened. For
instance, Prospect Theory proposes domain-sensitive risk propensity so that decision
makers are risk averse above an endowment reference point and risk embracing
below that point (Kahneman and Tversky, 1979). An example of this in practice was
provided by Singh (1986) who obtained responses to a questionnaire and publicly
available data for 64 medium to large North American firms; he concluded they have
a satisficing level of performance, and increase risk when results are below that level.
An important element of risk-sensitivity is loss aversion. Because losses can have
serious consequences, any loss is valued more than an equal gain, and the utility of
any loss relative to the same gain increases with the size of the loss. In other words,
the pain of loss relative to an equivalent gain is directly proportional to the amount,
and inversely proportional to risk propensity. This is most obvious in laboratory
studies where decision makers are offered a variety of choices (Schneider and
Lopes, 1986): as the size of the potential loss increases, people increasingly prefer a
low risk outcome.
Taken together, these behaviours evidencing risk sensitivity challenge the
assumption in finance theories of constant risk aversion (Parrino, Poteshman and
Reference levels can also serve as an anchor. For instance Lovallo and Kahneman
(2003) point out that executives typically begin a decision with a forecast outcome,
often prepared by a sponsor of the proposal. Whilst decision makers will adjust their
expectations, this is generally not enough. So the reference anchor provides an
overly optimistic assessment of the outcome. This is consistent with studies of
decision makers’ forecasts which show that their `error bars’ are far too narrow: thus
actual outcomes frequently fall outside the range of possibilities, even when experts
are involved (Camerer, 1995).
1.3 Mental Accounting and Mean Reversion
The third group of influences on decision makers comprise mental accounting and
the assumption of mean reversion.
Thaler (1985) developed the concept of mental accounting in which decision makers
apportion their wealth, knowledge and other resources into discrete and non-fungible
mental accounts. In economic terms this leads consumers to over-weight sunk costs
and current, cash outlays. But it also leads to a number of behavioural anomalies.
For instance, recent experience becomes important because it provides a
personalised sampling of outcomes and changes endowment relative to the
satisficing level: thus sequential outcomes have synergetic impacts on risk propensity
and hence on decisions.
One particularly relevant feature of mental accounting is what Tversky and
Kahneman (1971) called the law of small numbers. Because it is too demanding to
collect and process a statistically robust sample, decision makers overgeneralise
from small samples and tend to overweight personal experience and striking
observations (e.g. crises, highly publicised incidents, freakish calamities). Thus
recent experience carries more weight than population-based distributions, and –
because of self-framing - is especially likely to be over-weighted when it supports a
preferred outcome (Zackay, 1984).
The net result is that people place more emphasis on the consequences of decision
outcomes than on their probabilities. Thus decisions involving risk turn on
expectations of how alternative outcomes will impact endowment, rather than on
probabilities of the outcomes. This explains why a number of studies (Forlani, 2002
and Mullins et al, 1999) find that the facts of a decision are frequently ignored.
Another important decision influence is the assumption by decision makers that
mean reversion will apply unless there is reason to believe otherwise (Heath,
Huddart and Lang, 1999). An obvious example is the disposition effect, or tendency
for investors to sell assets that have risen in value in preference to those that have
made a loss (Odean, 1998); further support comes from evidence of the gamblers’
fallacy, which assumes that the recent occurrence of an outcome (e.g. win by red,
heads or favourite) lowers the probability of re-occurrence in an identical, statistically
independent event (Morrison and Ordeshook, 1975). This assumption of mean
reversion exerts Bayesian influences so that successful decision makers expect a
run of wins to be followed by losses, and – in the absence of overconfidence - will
tend to become less risk prone; whilst unsuccessful decision makers expect a turn for
the better and can become more risk prone.
1.4 Longshot Bias in Decision Makers
Overconfidence in managers is part of a pattern which psychologists term self-
enhancing biases. According to Rabin (1996: 50):
“We are over-optimistic regarding our health and other aspects of our life; we feel
we are less vulnerable to risk than others; and we are more responsible for our
successes than we are for our failures. We think that we are superior to others in
all sorts of ways: we are better at controlling risk, better drivers, and more ethical.”
There is much evidence of overconfident behaviour by finance managers who prefer
low probability outcomes that cannot be justified by their statistical record. For
instance, a variety of studies have shown that the average failure rate for common
business strategies lies in the range between 70 and 90 percent. Examples include
acquisitions (Rau and Vermaelen, 1998), research and development projects
(Palmer and Wiseman, 1999), company formation (Camerer and Lovallo, 1999),
mineral exploration (Mackenzie and Doggett, 1992) and new product launches
(Roskelly, 2002). According to one study of managers in US corporations, at least
half of all their decisions fail (Nutt, 1999).
One of the best-known depictions of the longshot bias relates to acquisitions where
Roll (1986) argued that managers of acquiring firms are over-optimistic in the
valuation of targets and over-confident in their ability to monetise potential merger
synergies. Selective analysis that induces a level of overconfidence bordering on
hubris explains why firms overpay for acquisitions and – through the winner’s curse –
suffer poor returns.
1.5 Desire for Immediate Gratification
The final striking feature of risk taking behaviour is the preference of managers for
immediate gratification (Angeletos, Laibson, Repetto, Tobacman and Weinberg,
2001). Although finance assumes that elapsed time is the only factor separating the
same decision now and in the future, actual behaviour reflects a bias towards
immediately achieving a desirable outcome. This effectively underweights risk
probabilities and accepts high opportunity costs. It is equivalent to use of a higher
discount rate for costs than benefits, which is consistent with behavioural evidence
(Sagristano, Trope and Liberman, 2002). This leads to a preference by managers for
investments with high early payouts, and a delay in cost-saving projects (present
value of costs saved will be reduced relative to present value of opportunity cost)
including risk management (which avoids costs rather than producing an incremental
II. The Model and Research Methodology
This section discusses the objectives and methodology used to research the how of
risk and decision making by finance executives, using a descriptive model built up
from the literature survey above.
2.1 Decision Model
The behavioural and managerial evidence in section I shows that individual risk
taking responds to a combination of personal, environmental, situational and
definitional aspects of the decision. These can be described by five groups of stimuli
that could be expected to drive the decision of a finance executive who faces a risky
Three of the stimuli are relatively obvious empirical parameters that can be reliably
measured: the facts or features of the decision; the population of decision events
which is made up of the previous outcomes of similar decisions; and experts’
opinions of the likely outcome of this event.
A fourth – and more qualitative – group of decision stimuli can be thought of as the
decision maker’s paradigm which comprises a pattern of personal features that are
relatively stable across different decision types. These include competencies,
personal attributes (especially demography and personality), endowment, experience
in previous decision making, and future aspirations.
The fifth group, which is more transient, comprise the decision maker’s perception
(how the decision is self-framed), its institutional setting, and the relevance of the
decision maker’s skill which indicates how much control they have over the outcome.
These various factors can be distilled into a decision model as shown in figure 1. The
principal objective of this study is to validate this depiction of managerial decision
[Insert Figure 1 Here]
2.2 Survey Materials
The research here is intensive in the form of a survey involving senior finance
executives as this gives sufficient granularity in responses to critically examine the
influences on decision making. The Survey is of the in-basket style Gill (1979), and
takes the form of a case study involving a decision with two alternative choices, one
of them risky and the other relatively safe. Participants are asked to indicate which
alternative they would recommend and provide personal details by answering a
variety of questions on their demography and personality.
The subject of the case study is a Grand Prix racing team, Carter Racing, and it is
loosely modelled on events leading up to the 1986 Challenger space shuttle disaster.
The case was designed to evaluate the reasons why experienced decision makers
select a risky alternative, and is commonly used to illustrate managerial decision
making (Sitkin and Weingart, 1995).
Original material was sourced from the copyright holders of the case study. This was
shortened to one page, and formatted as a memo seeking a typical business
decision. The memo described the experience of Carter Racing, which had
developed a unique engine design and was in its first year of Grand Prix racing. The
engine was prone to fail at high cost, but Carter’s initial success had attracted
attention, and the team needed to perform well in its next race. Respondents were
given details of Carter’s recent performance and some technical information, and
asked if they would recommend that the team should race or not race in the next
event. In keeping with the business style, no irrelevant information was included.
Using an approach that proved successful on several occasions (e.g. Forlani, 2002;
Mullins et al, 1999; and Sitkin and Weingart, 1995), the case was internally
manipulated to provide varying levels of risk by incorporating opposing values to four
facts that intuitively seem essential to the decision: finishing position in the last ten
races (either one or five top ten finishes); number of blown engines in the last ten
races (one or five); expert opinion on the cause of engine failure (ambient
temperature is, or is not, the cause of engine failure); and the anticipated
consequences of a wrong decision (almost certain bankruptcy, or some financial
pressures). Thus the case study came in 16 versions, and subjects were randomly
assigned to one.
The questionnaire had three parts. The first obtained the dependent variable through
the critical question: `If you were the owner of Carter Racing, what is the probability
that you would decide to race tomorrow?’. The second section contained 18
questions designed to elucidate reasons for the choice, which incorporated tests of
framing, endowment, expert opinion and outcome expectations.
The third section contained 51 questions. The first 15 measured personal
competencies including education, income and decision experience (years in
workforce, industry, type of job); and demographic features including age, gender,
marital status, and nationality. Other questions related to personality: locus of control
(by powerful others, internal control and chance) (Levenson, 1974); tolerance of
ambiguity (Budner, 1962); sensation seeking (seeking novel experiences and willing
to take risks to have them), impulsivity (rapidly responds to cues; not inhibited from
risk taking), aggression and sociability (Zuckerman and Kuhlman, 2000); egalitarian
preferences; extraversion, emotional stability, and conformity to social norms
(Robinson and Shaver, 1973); Type A personality (Williams and Narendran, 1999);
autonomy orientation, flexibility and competence; anxiety and susceptibility to
boredom; need for tension, risk and adventure (Keinan, 1984); lack of inhibition,
feelings of self-efficacy, and self-discipline; worldviews (hierarchic, egalitarian,
individualist, and fatalist) (Slovic, 2000); achievement motivation (McClelland, 1961);
and personal traits (Goldberg, 1990).
To ensure validity and reliability of the questions, the majority were drawn from
studies that had been previously published in peer-reviewed journals. Responses
were measured on a five-point Likert scale where 1=strongly agree and 5=strongly
disagree. Whilst altering the setting of questions (i.e. their frame) can elicit different
responses, the strategy of using a previously published case study and questions
was designed to facilitate validation of resultsi. The questionnaire was available on a
research website, and testing showed that it took an average of 20 minutes to
2.3 Survey Participants
Participants came from the Finance and Treasury Association, which is a
professional body whose members are active in corporate finance and typically
employed by banks and major companies; it publishes a directory of members’
names, positions and e-mail addresses. A total of 530 potential finance executives
were contacted to participate, with a follow up after two weeks.
Given that target respondents were practicing managers, e-mail was used to solicit
responses as it is a common medium of business-to-business contact, and a
common mode for decision making; it was not expected to introduce a bias in the
sample. The e-mail provided a hot link to the questionnaire website. The advantages
of this approach are that it is cheap, typically generates quick responses, and is
perceived as environmentally friendly. As responses come electronically, they can be
accurately compiled and are obtained anonymously which promotes completion of
sensitive questionsii. E-mail also avoids the shortcomings of personal interviews,
where bias can be induced by interviewer guidance (e.g. in cognitive processes) or
propensity to agree with questions (Dillman, Phelps, Tortora, Swift, Kohrell and
Berck, 2001). Conversely, e-mail surveys can produce response rates that are low, or
vary widely for no apparent reason (Smee and Brennan, 2000).
2.4 Limitations of Methodology
Critical assumptions underpinning this research are that: responses to a hypothetical
decision collected by an on-line survey can be generalised to actual finance decision
making; responses are representative of the decision making processes of Australian
finance executives in general; parameters can be quantified without error; and the
absence of real risks does not diminish the validity of the results. Other limitations are
the size of the sample and a practical limit to the number of questions, which
inevitably restrict the extent of research.
As questionnaires can do no more than recognise patterns in subjects’ reports, they
rely upon the goodwill and accuracy of participants. In a sensitive area such as risk
and decisions, there is no guarantee that responses will reflect true preferences.
Another deficiency of intensive techniques is possible contamination of the results
through what has been called the Hawthorne Effect since experiments by Mayo
(1933) suggested that simply observing behaviour can change itiii. Such concerns are
further complicated by the ethical research requirement for informed consent: even
describing the research proposal can frame responses.
Despite these limitations, the research strategy has important strengths, especially:
strong grounding in the literature, including empirical studies; linkage between
questions to ensure internal consistency of findings; emphasis on the real world so
that decision makers are operating in a familiar environment; and use of
heterogeneous samples of experienced decision makers. This should test the
research questions in a more realistic environment than using students in a
laboratory which is the approach of most decision studies.
III. Survey Results
This section reports the results from analysis of the survey responses.
3.1 Responses and Respondents
The summary of respondents’ demographic traits provided in table 1 shows they are
not a random sample of the population as they are: predominantly male (84 percent),
tertiary qualified (100 percent), in professional or executive roles (96 percent), with
considerable work experience (almost 70 percent have 16 or more years in
employment) and relatively high incomes (59 percent earn over $100,000 per year).
Conversely the sample group provides a good spread of ages and industry, and
appears a broad cross-section of decision making finance executives.
[Insert Table 1 Here]
There were a total of 67 useable responsesiv, which gives a response rate of 12.5
percent. Although this is expected for an unsolicited e-mail survey (Smee and
Brennan, 2000), it is modest and makes it desirable to validate the
representativeness of respondents. The approach suggested by West and Berthon
(1997) was employed and uses a two-tailed t-test to compare the first and last
quarters of responses. The methodological assumption is that late respondents are
closer in sentiment to the non-responding pool than are the early respondents. The
last quarter of respondents proved slightly less likely to race than the early
respondents (55 percent probability of racing versus 61 percent, respectively), but
the variation is not significant (p>0.6; see panel A of table 2). This gives confidence
that the survey data are not biased.
[Insert Table 2 Here]
The cases were designed to be `risk-neutral’ on average, and 56.7 percent of
respondents chose the risky alternative of racing. Assuming a binomial distribution of
race-don’t race responses, the overall result is not significantly different (p>0.27) to
that expected from a randomly chosen, risk-neutral samplev. Thus these executives
are risk neutral or slightly risk-prone on average.
A simple univariate analytical strategy was chosen to identify which effects on
decision making were most significant. Only respondents who gave a clear decision
(i.e. percentage probability of racing >60 or <40) were included, and this reduced the
sample size to 58. For each of the 69 questions, responses were divided into two
groups, one of which strongly agreed or agreed with the statement and the other
which disagreed or strongly disagreed. The sample means of each of the 69 paired
groups were compared, and significant differences reported in table 2.
Panel B examines risk propensity in light of the four manipulated facts of the case,
and shows that only two proved significant (p<0.10). The first was reliance on
experts: respondents who were told that experts knew the cause of engine failures
were considerably more likely to take the risk and race than those who were told
experts did not know the cause (61.4 and 48.6 percent, respectively). Thus support
from experts strengthens risk propensity.
The second significant fact related to the outcome from failure. Those who were told
that failure after taking the risk would result in bankruptcy were far less likely to race
than those who expected the worst case outcome was only some financial distress
(49.6 and 62.2 percent, respectively). Thus respondents proved loss averse.
The next step was to identify influences on risk propensity from respondents’
perception of the decision and their personal traits. Those that proved statistically
significant (p<0.10) are reported in panels C and D of table 2.
Results in panel C showed statistically significant support for a longshot bias as the
probability of racing was higher for respondents who foresaw large potential for gain
and for those who believed that racing was too good an opportunity to pass up. The
results also suggest that respondents self-framed the decision so that – even though
virtually all saw it as a risky activity – those who took the risk considered Carter’s risk
was low and was likely to prove successful.
Panel D shows that respondents who take a risk are younger, more controlled and
Type-A personalities, which confirms risk prone traits found in other studies (Grable,
2000; and Trimpop, 1994).
The table also reveals two negative findings that have significant implications for
managers’ decision making. The first is the absence of several intuitively important
factors. Gender, for instance, did not prove significant to risk propensity; nor did
perception of control, and respondents’ self-judgement of their risk propensity.
The second finding is that executives do not have a stable risk propensity that is
generalisable across different settings. Respondents were asked about their
tendency to take risks in their personal life, personal investment strategy and in
working for clients or employers. Only one of the three paired combinations was
statistically significant: the link between risk propensity in personal finances and
business. None of these self-reports of risk propensity was significantly correlated to
decisions in the case study.
As a key research objective was to identify which personality traits influence risk
taking in finance, the analysis above was extended by use of bivariate correlations
between probability of racing and self-report of 51 personal and psychological
measures. The statistically significant (p<0.05) relationships are listed in table 3, and
demonstrate strong agreement with previously published studies. Specifically, no
significant relationship found in this study contradicted the original relationship
between personal attribute and risk propensity.
[Insert table 3 here]
Results in table 3 characterise risk-prone finance managers as confident and
sociable; calm and relaxed, although hard driving. They believe that risk is required
to get ahead, and that risk taking contributes to success. An interesting finding is that
risk-prone executives believe that chance plays an important role in success, which
implies that they consider outcomes of risky actions are at least partly beyond their
The next step was to more comprehensively explain respondents’ risk propensity by
using multiple regression to derive an expression that is parsimonious and logical. It
was also desirable to show the extent to which subjects’ decision making matches, or
fails to match, the processes in the proposed decision making model.
The result is shown in table 4 using a layout proposed by Hair, Anderson, Tatham,
and Black (1998: 212). This shows that the probability of racing is increased by
agreement with the following statements:
• `If this opportunity is passed up, there will never be another as good’
• `Carter Racing is likely to succeed tomorrow’
And the probability of racing is decreased
• in proportion to the value of investments; and
• by agreement with the statement: ‘Risk is higher when facing situations we do
The solution identified variables which determine a person’s innate traits, propensity
to select a risky alternative when making a decision, and their decision making style.
[Insert Table 4 Here]
These results can then be used to identify which of the proposed relationships (set
out in figure 1) are important. As shown in figure 2, it is possible to describe
executives who take risks along four dimensionsvi.
The first dimension is personal attributes: risk propensity is higher for those with a
lower value of investments, which is correlated to lower income, fewer years of
employment and younger age. A second dimension is decision making style which
assumes that risk is not higher when facing decisions that we do not understand.
These risk takers describe themselves as `controlled’ and consider they have
received more breaks in life than most. The third dimension is an assessment of the
situation which concludes there is no alternative to taking the risk. The final decision
dimension is a judgement that risk will bring success. These people describe
themselves as more willing to take risks than their colleagues and as `confident’.
Table 4 and figure 2 indicate that almost half the variability in risk propensity by
finance managers can be explained by four measurable parameters. However, the
most interesting aspect of this finding is that none of the quantitative measures that
were proposed in figure 1 as influences on decision making – experts’ opinion,
decision facts and features and previous outcomes – had a significant, direct
influence on the decision to take a risk or not. By contrast, the explanatory variables
were all qualitative. This is particularly important given the relative neglect of these
parameters in most previous studies.
IV. Discussion and Implications
This study examined the reasons why finance executives select a risky alternative
when making decisions, and did not consider the effectiveness of these decisions.
Even so, it raises four particularly significant conclusions:
i. Just over half the executives surveyed proved willing to take a risk
ii. Executives’ risk propensity is strongly influenced by demographic and
personality characteristics which explain 24 percent of the variance in risk
iii. Factual elements in the case had limited impact on the decision. The most
important transient influences on decision making are judgements about the
outcome, which explain a quarter of the variation in risk propensity.
iv. Results are consistent with risk-sensitive foraging and Prospect Theory.
Consider each in turn.
The conclusion that finance executives are risk neutral or slightly risk prone on
average contradicts most finance assumptions (Bodie, Kane and Marcus, 2005: 144),
but is consistent with several previous studies. For instance, Levy and Levy (2002)
studied investment-type decisions by students, faculty and professionals and found
that between 47 and 67 percent of their subjects were risk prone. Similar results were
obtained for managers by MacCrimmon and Wehrung (1984) and Williams and
The second important conclusion is that relatively stable, personal characteristics
explain around a quarter of the variation between executives in their risk propensity.
This closely matches published data. In this study, risk prone executives proved to be
younger, and this attribute was associated with lower income and wealth. They are
Type-A personalities (‘I regularly set deadlines for myself’); sociable, confident and
calm; and believe in the need to take risks to be successful. The decision making
style of risk takers is driven by belief that they are luckier and more capable than their
The third important conclusion relates to the how of risk taking, which is driven by
contextual and process variables that can be thought of as decision making style,
alternatives to taking a risk, and expectation about the outcome. Risk-prone
respondents do not see a viable alternative to taking a risk, and are confident of a
successful outcome. Respondents conform to the conclusion of West and Berthon
(1997: 30) that “successful risk taking individuals are likely to believe that they can
beat the odds, that nature is good to them, and that they have special abilities.”
The executives’ risk propensity was increased by expert support, which is consistent
with studies that find experts play important facilitating roles in high risk decisions
such as mergers and acquisitions (Lewis and Zalan, 2004). Executives also proved
loss averse in their decision making and less likely to take a risk when it could result
Findings of this study contradict the normative assumption that decisions are
determined solely by their facts. Overall, executives tend to place much less weight
on the quantitative content of a risky decision than they do on its qualitative aspects.
Using their own unique perspective of the decision context, executives make risky
choices by looking to the future, virtually independently of apparent facts. This is
consistent with the finding by Forlani (2002) and Mullins, Forlani and Walker (1999)
that risk propensity is influenced by contextual factors such as a history of successful
Framing of the decision, especially its expected outcome, is also important for its
influence on affect. This is why managerial risk propensity can be encouraged by
rewarding individuals’ initiative and success, rather than forcing adherence to
process (Pablo, Sitkin and Jemison, 1996). Longer term, there is a suggestion that
successful experience in risky decision making and education about the process will
help modify behaviour. All this is consistent with the observation by Kahneman
(2003) that “the central characteristic of agents is … that they often act intuitively.
And the behavior of these agents is not guided by what they are able to compute, but
by what they happen to see at a given moment.”
The fourth conclusion of the study is to support risk-sensitive foraging and Prospect
Theory. Respondents with low incomes and less wealth have higher risk propensity,
and this is consistent with an inverse relationship between risk propensity and
These results have significant implications for finance management, particularly
investment decisions and agency theory.
The finding that at least half the sampled finance managers are prepared to make a
risky choice is clearly important based on the model developed by Parrino,
Poteshman and Weisbach (2005). They reached the intuitively obvious conclusion
that manager risk propensity is critical to selection of investments: managers who are
risk-averse will reject attractive, but risky, investments; whereas risk-neutral decision
makers find risky projects more attractive. Thus the assumption in many models of
myopic risk aversion will mischaracterise decisions by managers and investors.
The basis for the longshot bias in managerial decision making is apparent in the
result that variables with strongest links to risk propensity were belief that the
situation has large potential for gain and confidence in a successful outcome. This
points to one possible explanation for high risk strategies which is that managers
recognise the skewness in returns and prefer the small chance of a high payout
despite their low expected return (Golec and Tamarkin, 1998). The temptation to
gamble may be fostered by executives’ belief in their luck, and confidence that their
skill can increase the probability of success. It is encouraged, too, by experts’
The underweighting of risk probabilities is consistent with the desire for immediate
gratification. Moreover, there is strong evidence of self-framing in the high risk
propensity of respondents who foresee the potential for gain, low risk for Carter and
likelihood of success.
The study’s results also provide insights into contributors to agency problems. Much
of the variance in executives’ risk propensity arises from personal traits and
expectations about the outcome. Thus simply expanding on data and analysis will not
change decisions. Managers decide to take a risky alternative because of their innate
features, learned decision making style, and expectation of a successful outcome.
This means that a significant shift in firm risk propensity requires a change in its
managers, not just their compensation or the organisation’s structure. This explains
why so many organisations will make major changes to their senior staff, despite the
costs, disruption and loss of corporate memory.
The executives surveyed here are risk prone, although far from homogenous in their
risk propensity; and they do not rely solely on facts when taking risks. This, too, has
important implications for modern finance theory, which assumes financial decision
makers are rational, homogenous and risk averse, with diminishing marginal utility for
money. Finance executives and (if the agent-principal relationship holds true) their
firms do not follow the logical decision making process assumed in strategy
textbooks: collect the facts, weight them by probability, evaluate each outcome, and
choose the highest value adding alternative (Dearlove, 1998). Thus decision theory
needs to specifically recognise affect (with its psychological meaning as associated
feeling or emotion) and the concept of decision maker utility (Samson, 1987), even
though it has waned in popularity in the last decade.
This study’s results point to two gaps in the finance research agenda. The first is the
need to examine decisions (and, by implication, other finance practices) close to a
real-world context where subjects follow their natural decision styles, rather than
conforming to norms imposed by experimental settings. Given that risk propensity
can explain half the variance in executives’ decision making, a further area for
research is the impact of risk on firm performance. Although there has been
considerable effort using risk with its meaning of variance (Nickel and Rodriguez,
2002), less research has used risk in the way executives think of it. This is
particularly important as existing strategy paradigms have proven unable to explain
firms’ financial performance (Campbell-Hunt, 2000).
Abdellaoui, M. (2000). Parameter-Free Elicitation of Utility and Probability Weighting
Functions. Management Science 46 (11): 1497-1512.
Angeletos, G.-M., Laibson, D., Repetto, A., Tobacman, J. and Weinberg, S. (2001).
The Hyperbolic Consumption Model: Calibration, Simulation, and Empirical
Evaluation. Journal of Economic Perspectives 15 (3): 47-68.
Austin, E. J., Deary, I. J. and Willock, J. (2001). Personality and Intelligence as
Predictors of Economic Behavior in Scottish Farmers. European Journal of
Personality 15 (S1): S123-S137.
Baker, M., Ruback, R. S. and Wurgler, J. (2004). Behavioral Corporate Finance.
Working Paper. Tuck School of Business at Dartmouth, Hanover NH.
Barberis, N. and Thaler, R. (2003). A Survey of Behavioral Finance. Published in G.
M. Constantinides, M. Harris and R. M. Stulz (ed). Handbook of the
Economics of Finance. Elsevier/North-Holland, Boston.
Bateson, M. (2002). Recent advances in our understanding of risk-sensitive foraging
preferences. Proceedings of the Nutrition Society 61 (4): 1-8.
Benartzi, S., Michaely, R. and Thaler, R. (1997). Do Changes in Dividends Signal the
Future or the Past? The Journal of Finance 52 (3): 1007-1034.
Bleichrodt, H. (2001). Probability Weighting in Choice under Risk: An empirical test.
The Journal of Risk and Uncertainty 23 (2): 185-198.
Bodie, Z., Kane, A. and Marcus, A. J. (2005). Investments (6th Edition). McGraw Hill,
Budner, S. (1962). Intolerance of ambiguity as a personality variable. Journal of
Personality 30 (3): 29-50.
Camerer, C. (1995). Chapter 8 - Individual Decision Making. Published in J. H. Kagel
and A. E. Roth (ed). The Handbook of Experimental Economics. Princeton
University Press, Princeton.
Camerer, C. and Lovallo, D. (1999). Overconfidence and Excess Entry: An
Experimental Approach. American Economic Review 89 (1): 306-318.
Campbell-Hunt, C. (2000). What Have We Learned About Generic Competitive
Strategy? A meta-analysis. Strategic Management Journal 21 (2): 127-154.
Caraco, T., Blanckenhorn, W., Gregory, G., Newman, J., Recer, G. and Zwicker, S.
(1990). Risk-sensitivity: ambient temperature affects foraging choice. Animal
Behaviour 39: 338-345.
Conlisk, J. (1996). Why Bounded Rationality? Journal of Economic Literature 34:
Damodaran, A. (2001). Corporate Finance. John Wiley & Sons, Hoboken NJ.
De Bondt, W. F. M. (1991). What do economists know about the stock market? The
Journal of Portfolio Management 17: 84-91.
De Bondt, W. F. M. and Thaler, R. H. (1994). Financial Decision Making in Markets
and Firms: A behavioral perspective. Working Paper No 4777. National
Bureau of Economic Research, Cambridge MA.
Dearlove, D. (1998). Key Management Decisions. Financial Times, London.
Dillman, D. A., Phelps, G., Tortora, R., Swift, K., Kohrell, J. and Berck, J. (2001).
Response Rate and Measurement Differences in Mixed Mode Surveys Using
Mail, Telephone, Interactive Voice Response and the Internet. Accessed: 13
Eeckhoudt, L. and Godfroid, P. (2000). Risk Aversion and the Value of Information.
Journal of Economic Education 31 (4): 382-388.
Ellsberg, D. (1961). Risk, Ambiguity, and the Savage Axioms. The Quarterly Journal
of Economics 75 (4): 643-669.
Finucane, M. L., Alhakami, A., Slovic, P. and Johnson, S. M. (2000). The Affect
Heuristic in Judgements of Risks and Benefits. Published in P. Slovic (ed).
The Perception of Risk. Earthscan, London.
Forlani, D. (2002). Risk and Rationality: The influence of decision domain and
perceived outcome control on managers' high-risk decisions. Journal of
Behavioral Decision Making 15 (2): 125-140.
Fox, C. R. and Tversky, A. (1998). A Belief-based Account of Decision Under
Uncertainty. Management Science 44 (7): 879-895.
Gale, E. A. M. (2004). The Hawthorne studies-a fable for our times? QJM 97 (7): 439-
Gill, R. W. T. (1979). The in-tray exercise as a measure of management potential.
Journal of Occupational Psychology 52 (6): 185-197.
Goldberg, L. R. (1990). An Alternative `Description of Personality': The Big-Five
Structure. Journal of Personality and Social Psychology 59 (6): 1216-1229.
Golec, J. and Tamarkin, M. (1998). Bettors Love Skewness, Not Risk, at the Horse
Track. Journal of Political Economy 106 (1): 205-225.
Grable, J. E. (2000). Financial Risk Tolerance and Additional Factors that Affect Risk
Taking in Everyday Money Matters. Journal of Business and Psychology 14
Graham, J. R. and Harvey, C. R. (2001). The Theory and Practice of Corporate
Finance: Evidence from the field. Journal of Financial Economics 60: 187-243.
Gray, V. and Lowery, D. (1998). To lobby alone or in a flock: Foraging behaviour
among organized interests. American Politics Quarterly 26 (1): 5-34.
Greene, J. D., Sommerville, R. B., Nystrom, L. E., Darley, J. M. and Cohen, J. D.
(2001). An fMRI Investigation of Emotional Engagement in Moral Judgment.
Science 293 (5537): 2105-2108.
Hair, J. E. J., Anderson, R. E., Tatham, R. L. and Black, W. C. (1998). Multivariate
Data Analysis. Prentice Hall, Upper Saddle River NJ.
Hartog, J., Ferrer-i-Carbonell, A. and Jonker, N. (2000). On a Simple Measure of
Individual Risk Aversion. Rotterdam.
Heath, C., Huddart, S. and Lang, M. (1999). Psychological Factors and Stock Option
Exercise. The Quarterly Journal of Economics 114 (2): 601-627.
Heath, C. and Tversky, A. (1991). Preference and belief: Ambiguity and competence
in choice under uncertainty. Journal of Risk and Uncertainty 4 (1): 5-28.
Jarrell, G. A., Brickley, J. A. and Netter, J. M. (1988). The Market for Corporate
Control: The evidence since 1980. Journal of Economic Perspectives 2: 49-68.
Jensen, M. C. and Meckling, W. (1976). Theory of the firm: Managerial behavior,
agency costs, and ownership structure. Journal of Financial Economics 3 (4):
Kahneman, D. (2003). Maps of Bounded Rationality: Psychology for Behavioral
Economics. The American Economic Review 93 (5): 1449-1475.
Kahneman, D. and Lovallo, D. (1993). Timid Choices and Bold Forecasts: A cognitive
perspective on risk taking. Management Science 39 (1): 17-31.
Kahneman, D. and Tversky, A. (1979). Prospect Theory: An analysis of decision
under risk. Econometrica 47 (2): 263-291.
Keinan, G. (1984). Measurement of Risk Takers' Personality. Psychological Reports
Kilka, M. and Weber, M. (2001). What Determines the Shape of the Probability
Weighting Function Under Uncertainty? Management Science 47 (12): 12-26.
Kühberger, A. (1998). The Influence of Framing on Risky Decisions: A meta-analysis.
Organizational Behavior and Human Decision Processes 75 (1): 23-55.
Levenson, H. (1974). Activism and Powerful Others. Journal of Personality
Assessment 38 (4): 377-383.
Levy, H. and Levy, M. (2002). Arrow-Pratt Risk Aversion, Risk Premium and Decision
Weights. Journal of Risk and Uncertainty 25 (3): 265-290.
Lewis, G. and Zalan, T. (2004). The CEO-Advisors Nexus: Toward an explanation of
merger preference in mergers and acquisitions. MBS Working Papers.
University of Melbourne, Melbourne.
Lintner, J. (1956). Distribution of Incomes of Corporations among Dividends,
Retained Earnings and Taxes. American Economic Review 46 (2): 97-113.
Lopes, L. L. (1987). Between Hope and Fear: The psychology of risk. Published in L.
Berkowitz (ed). Experimental Social Psychology. Academic Press, San Diego.
Loughran, T., Ritter, J. and Rydqvist, K. (1994). Initial Public Offerings: International
insights. Pacific Basin Finance Journal 2: 165-169.
Lovallo, D. and Kahneman, D. (2003). Delusions of Success: How optimism
undermines executives' decisions. Harvard Business Review 81 (7): 56-63.
MacCrimmon, K. R. and Wehrung, D. A. (1984). The Risk In-Basket. Journal of
Business 57 (3): 367-387.
Mackenzie, B. W. and Doggett, M. D. (1992). Economics of Mineral Exploration in
Australia. Australian Mineral Foundation, Glenelg SA.
Mano, H. (1990). Anticipated Deadline Penalties: Effects on goal levels and task
performance. Published in R. M. Hogarth (ed). Insights in Decision Making.
The University of Chicago Press, Chicago.
Mayo, E. (1933). The human problems of an industrial civilization. Macmillan, New
McClelland, D. C. (1961). The Achieving Society. The Free Press, New York.
McNamara, G. and Bromiley, P. (1997). Decision Making in an Organizational
Setting: Cognitive and organizational influences on risk assessment in
commercial lending. Academy of Management Journal 40 (5): 1063-1088.
Miller, S. M., Liu, G. B., Ngo, T. T., Hooper, G., Riek, S., Carson, R. G. and
Pettigrew, J. D. (2000). Interhemispheric switching mediates perceptual rivalry.
Current Biology 10 (7): 383-392.
Morrison, R. S. and Ordeshook, P. (1975). Rational Choice, Light Guessing and the
Gambler's Fallacy. Public Choice 22: 79-89.
Mullins, J. W., Forlani, D. and Walker, O. C. (1999). Effects of organizational and
decision-maker factors on new product risk taking. Journal of Product
Innovation Management 16 (3): 282-294.
Nickel, M. N. and Rodriguez, M. C. (2002). A Review of Research on the Negative
Accounting Relationship Between Risk and Return: Bowman's Paradox.
Omega 30 (1): 1-18.
Nutt, P. C. (1999). Surprising but True: Half the decisions in organizations fail.
Academy of Management Executive 13 (4): 75-90.
Odean, T. (1998). Are Investors Reluctant to Realize Their Losses? The Journal of
Finance 53 (6): 1775-1798.
Pablo, A. L., Sitkin, S. B. and Jemison, D. B. (1996). Acquisition decision-making
processes: the central role of risk. Journal of Management 22 (5): 723-746.
Palmer, T. and Wiseman, R. M. (1999). Decoupling risk taking from income stream
uncertainty: A holistic approach. Strategic Management Journal 20 (11): 1037-
Parrino, R., Poteshman, A. M. and Weisbach, M. S. (2005). Measuring Investment
Distortions when Risk-Averse Managers Decide whether to Undertake Risky
Projects. Financial Management 34 (1): 21-60.
Rabin, M. (1996). Psychology and Economics. Accessed: 4 May 2003.
Rau, P. R. and Vermaelen, T. (1998). Glamour, Value and the Post-Acquisition
Performance of Acquiring Firms. Journal of Financial Economics 49: 223-253.
Ricciardi, V. (2004). A Risk Perception Primer: A Narrative Research Review of the
Risk Perception Literature in Behavioral Accounting and Behavioral Finance.
Accessed: 17 February 2005. http://ssrn.com/abstract=566802.
Robinson, J. P. and Shaver, P. R. (1973). Measures of Social Psychological
Attitudes. Institute for Social Research, Ann Arbor.
Robinson, J. P., Shaver, P. R. and Wrightsman, L. (1991). Measures of personality
and social psychological attitudes. Academic Press, San Diego.
Rohrmann, B. (1997). Risk Orientation Questionnaire: Attitudes towards risk
decisions. University of Melbourne, Melbourne.
Roll, R. (1986). The Hubris Hypothesis of Corporate Takeovers. Journal of Business
59 (2): 197-216.
Roskelly, N. (2002). If I Ran Your Company. New Products Magazine (November).
Sagristano, M. D., Trope, Y. and Liberman, N. (2002). Time-Dependent Gambling:
Odds now, money later. Journal of Experimental Psychology 131 (3): 364-371.
Samson, D. A. (1987). Corporate risk philosophy for improved risk management.
Journal of Business Research 15 (2): 107-122.
Schneider, S. L. and Lopes, L. L. (1986). Reflection in Preferences Under Risk: Who
and when may suggest why. Journal of Experimental Psychology: Human
Perception and Performance 12 (4): 535-548.
Shapira, Z. and Venezia, I. (2001). Patterns of behavior of professionally managed
and independent investors. Journal of Banking and Finance 25: 1573-1587.
Shefrin, H., Ed. (2001). Behavioral Finance. Edward Elgar Publishing, Cheltenham
Simon, H. A. (1959). Theories of decision-making in economics and behavioural
science. American Economic Review 49: 253-283.
Singh, J. V. (1986). Performance, Slack, and Risk taking in Organisational Decision
Making. Academy of Management Journal 29 (3): 526-585.
Sitkin, S. B. and Weingart, L. R. (1995). Determinants of risky decisionmaking
behavior. Academy of Management Journal 38 (6): 1573-1592.
Slovic, P. (1972). Psychological Study of Human Judgment: Implications for
investment decision making. The Journal of Finance 27 (4): 779-799.
Slovic, P. (2000). The Perception of Risk. Earthscan Publications, London.
Smee, A. and Brennan, M. (2000). Electronic Surveys: A Comparison of E-mail, Web
and Mail. ANZMAC 2000 Visionary Marketing for the 21 st Century: Facing the
Challenge, Gold Coast.
Smith, S. L. and Friedland, D. S. (1998). The Influence of Education and Personality
on Risk Propensity in Nurse Managers. Journal of Nursing Administration 28
Statman, M. and Sepe, J. F. (1989). Project Termination Announcements and the
Market Value of the Firm. Financial Management 18 (4): 74-81.
Thaler, R. (1985). Mental Accounting and Consumer Choice. Marketing Science 4
Trimpop, R. M. (1994). The Psychology of Risk Taking Behavior. Elsevier Science,
Tversky, A. and Kahneman, D. (1971). Belief in the law of small numbers.
Psychological Bulletin 76 (2): 105-110.
Wang, X. T. (2004). Self-framing of Risky Choice. Journal of Behavioral Decision
Making 17 (1): 1-16.
West, D. and Berthon, P. (1997). Antecedents of risk-taking behavior by advertisers:
Empirical evidence and management implications. Journal of Advertising
Research 37 (2): 27-40.
Williams, S. and Narendran, S. (1999). Determinants of Managerial Risk: Exploring
personality and cultural influences. The Journal of Social Psychology 139 (1):
Willman, P., Fenton-O'Creevy, M., Nicholson, N. and Soane, E. (2002). Traders,
managers and loss aversion in investment banking: a field study. Accounting
Organizations and Society 27: 85-98.
Zackay, D. (1984). The influence of perceived event's controllability on its subjective
occurrence probability. Psychological Record 34: 233-240.
Zaleskiewicz, T. (2001). Beyond Risk Seeking and Risk Aversion: personality and the
dual nature of economic risk taking. European Journal of Personality 46:
Zuckerman, M. and Kuhlman, D. M. (2000). Personality and Risk taking: Common
biosocial factors. Journal of Personality 68 (6): 999-1029.
Distribution of Respondents (Percent)
Gender Male Female
Age (years) ≤ 25 26-35 36-45 46-55 55-65
6 24 33 32 5
Marital Status Never married Married Separated Divorced
18 76 3 3
Education Level Diploma Bachelor Postgraduate
7 35 58
Work Experience <5 5-10 11-15 16-20 > 20
(years) 3 15 15 19 48
Industry Manufacturing Wholesale or Agriculture or Finance Services Government
Retail Trade Resources
8 7 3 48 26 8
Occupation Clerical Professional Executive Student
2 57 39 2
Income ($K PA) < 25 25-50 50-75 75-100 100-150 > 150
4 2 11 24 22 37
Investments ($K) < 25 25-100 100-250 250-500 500-999 > $1 mill
17 15 15 29 20 4
Table 2: Analysis of Decisions to Take a Risk
This table compares the characteristics of finance executives by univariate analyses that test for differences between the
mean risk propensity (measured here as the probability of taking the risk to race) of different samples of respondents.
In each panel, t-tests were used to compare the sample means. Let the respective means of two samples of size n1 and n2
be mean1 and mean2, with standard deviations stdev1 and stdev2. In comparing the means, t> 1.28 for p<0.10, t> 1.67 for
p<0.05, and t> 2.39 for p<0.01, where t = (mean1 - mean2)/√[(stdev12/n1)+(stdev22/n2)]. Significances at the ten, five and
one percent levels are denoted by *, ** and ***, respectively.
Panel A compares the first and last quarter of responses that were received. This assumes that late respondents are close
to non-respondents and so tests for biases from non responses. The t-statistic for the difference between the means of the
two samples is: (0.613-0.550)/√(0.3072/16+0.4162/16) = 0.487; which – using a two-tailed test- has a significance of 0.625.
Panel B reports results that partition respondents according to facts presented to them. Four facts were varied between
cases to give 16 versions, and differences in risk propensity for each are reported here.
Results in panel C classify respondents by their perceptions of the decision and its frame, whilst panel D classifies
respondents by their demographic or personality traits. Only statistically significant (p<0.10) relationships are reported in
panels C and D.
Probability of Racing (i.e. Risk Propensity) by Decision Maker Feature
INDEPENDENT VARIABLE Sample 1 Sample 2
Sample 1 and Sample 2 Mean Std dev n Mean Std dev n
Panel A: Test of Respondent Bias
First and last quarters of responses 61.3 30.7 16 55.0 41.6 16 0.5
Panel B: Internally manipulated facts of the case
Recent Results: zero or five top five finishes in last ten races 50.4 38.0 24 60.9 34.8 34 -1.1
Recent Vehicle Performance: one or five blown engines in last
58.6 35.4 28 54.7 37.5 30 0.4
Experts’ Opinion: knows cause and does not know 61.4 35.8 36 48.6 36.3 22 1.3
Outcome of Racing: almost certain bankruptcy or some
49.6 38.5 26 62.2 33.8 32 -1.3
Panel C: Respondent Perception of Decision (Strongly Agree&Agree versus Strongly Disagree&Disagree)
If this opportunity is passed up, there will never be another as
82.6 21.6 19 36.3 32.2 30 6.0***
The situation faced by Carter has large potential for gain 65.1 33.8 43 34.0 29.9 10 2.9***
Carter Racing can tolerate large risks 80.6 24.8 18 35.6 35.7 25 4.9***
Carter Racing is in a positive situation 68.8 33.6 16 47.6 37.2 29 1.9
Carter has a record of making the right decisions 65.0 35.0 14 33.8 37.8 13 2.2**
Although Grand Prix racing is risky, Carter Racing can expect
78.8 21.5 17 54.8 35.0 23 2.7***
The average person would make the same decision as me 42.5 38.1 20 82.5 21.4 12 -3.8***
Carter Racing is likely to succeed tomorrow 79.0 26.5 20 26.7 28.9 18 5.8***
Panel D: Respondent Trait or Response (Strongly Agree&Agree versus Strongly Disagree&Disagree)
I am calm and relaxed when participating in group discussions 64.5 34.2 33 33.6 38.3 11 2.4
I regularly set deadlines for myself 60.4 35.3 47 25.7 30.5 7 2.8**
Age (35 or less, >35 years) 67.8 28.8 18 51.5 38.3 40 1.8
Income (<$100K PA, >$100K PA) 73.3 27.7 12 52.2 37.1 46 2.2
Investments (<$100K, >$100K) 70.5 28.6 20 49.2 37.9 38 2.4
Table 3: Personality and Risk Propensity
This table compares the influences of personality on risk propensity as identified in this
study with the results from the earlier study from which the question was sourced.
Relationships are shown where significance in this study exceeds 20 percent;
relationships that are significant at the ten and five percent levels are denoted by * and
**, respectively. Positive correlations mean that risk prone decision makers agree that
the phrase describes their behaviour or beliefs.
Personality Links to Risk Propensity
Author Measure of Risk Propensity Sign of Relationship
Austin, Deary and o Successful people take risks + +
Levenson (1974) o Locus of control: When I get what I want, it’s + +
usually because I’m lucky
Robinson, Shaver, and o I am calm and relaxed when participating in + + (**)
Wrightsman (1991) group discussions
o In general I am very confident of my ability + + (**)
Rohrmann (1997) o I’ve not much sympathy for adventurous - -
Williams and o Type A: I regularly set deadlines for myself + + (**)
Zaleskiewicz (2001) o To achieve something in life, one has to + + (*)
Zuckerman and o My behaviour can be described as Sociable + +
Table 4: Mulitivariate Regression of Impacts on Risk Propensity
This table reports the results of multivariate regression of the dependent variable Probability of Racing (which is a proxy
for risk propensity) against 65 independent variables related to respondents’ perception of the decision and personal
characteristics. The dependent variables were measured using Likert scales and are reported so that a positive co-
efficient means agreement with the statement.
Statistics Associated with Probability of Racing
Summary of Model
R squared 0.494 Standard Error of Estimate 1.344
Adjusted R squared 0.462 Observations 67
Variables in the equation
Term Coefficient Standard Standardised Regression t value Significance
Error Coefficient (beta)
Intercept 8.125 0.766 10.61 0.00
Value of investments, excluding -0.240 0.103 -0.214 -2.338 0.023
Risk is higher when facing situations we -0.439 0.186 -0.219 -2.367 0.021
do not understand
If this opportunity [for Carter to race] is 0.492 0.149 0.328 3.309 0.002
passed up, there will never be another
Carter Racing is likely to succeed 0.687 0.175 0.387 3.926 0.000
Figure 1 : Proposed Links between Decision Components
Decision Maker’s Paradigm Experts’ Opinion
Historical decision outcomes
Aspirations Decision Facts or Features
Population of Outcomes
Decision Maker’s Risk Profile Results
Relevance of own Skill
Figure 2. Determinants of Risky Decision Making
[Correlation coefficients are shown beside arrows. Positive correlations with numeric
measures mean they increase with the other variable or risky decision; positive
correlations with statements mean agreement with risky decisions. Overall R2 = 0.60]
Years of 0.56
I have gotten more Decision Making - 0.31
breaks in life than 0.29 Style
most people I know Risk is higher when
facing decisions we - 0.24
“Controlled” 0.19 do not understand
The situation faced Racing (%)
by Carter has large 0.43
potential for gain
If this opportunity
is passed up, 0.53
describes my there will never be
behaviour 0.21 another as good
Carter has a record
of making the right
I am less willing to Carter Racing is
take risks than my likely to succeed
colleagues - 0.34
I am grateful for the advice and suggestions of numerous colleagues, and feedback from participants at
a workshop at University of Melbourne. The comments of three anonymous reviewers and the Journal’s
editor have significantly improved the paper. Remaining errors and omissions are my responsibility.
Copies of the case study and questionnaire are available on request from the author.
Although the Perseus software that was used to develop the questionnaires provides the ability to
surreptitiously record details of respondents’ servers, this was not done.
Note, though, that this study involved only five workers and the results were never published. Gale
(2004) derides it as an “anecdote” and “fable”.
Although a larger sample may be desirable, a wide variety of well-accepted studies has used much
smaller samples, and these have been of homogeneous students. Examples include: Abdellaoui (2000) –
64 economics students; Bleichrodt (2001) – 66 health economics students; Fox and Tversky (1998) – 50
students interested in basketball; Kilka and Weber (2001) – 55 graduate finance students; and Sitkin and
Weingart (1995) – 38 MBA students.
Assume for simplicity a binomial distribution of race-don’t race responses, with a mean of 0.5 and
standard deviation of [√(0.5*0.5)/67 =) 0.061. The observed pattern would occur by chance in 27.0 percent
of equivalent random samples.
Unfortunately the sample size is too small to permit structural equation modelling; however, the
dimensions above could be extended using basic path analysis to propose a more integrated model of risk