How do consumers evaluate risk in
financial products?
Received (in revised form): 20th January, 2005
Kathleen Byrne
has over 20 years experience in the insurance industry, spanning life insurance, investment and general insurance. Her
current focus is on lump sum investment products including guaranteed bonds and structured products. She joined Cardif
Pinnacle in 1994, with responsibility for Actuarial, and was Group Actuarial Director prior to her appointment as Managing
Director  Investments in October 2002. She is currently responsible for all aspects of Pinnacle s investment business from the
design of new products through to service delivery.
Abstract Decision-making processes consumers use in investing lump sums are
reviewed, focusing on how investment risk is perceived and assessed. Primary research
was undertaken with investment customers to explore the role played in evaluation of
investment risk by risk perceptions and risk propensity. Both the literature review and
the research findings indicate the central role risk perceptions play in financial
decisions. Sitkin and Weingart s risk model is used as a research framework. Risk
propensity and risk perception were found to be negatively correlated, however, deposit
accounts were selected for investment irrespective of how risky a respondent considered
them to be.
Risk perceptions and expected return were positively correlated for all asset types
apart from property. Further investigation revealed that experts exhibited positive
correlation in risk return judgments but novices showed no correlation. There was no
correlation between risk and return for either novices or experts for property.
Return expectations were positively correlated with investment allocation. Provision of
past performance information appears to create an expectation for future returns around
the same level as past returns. Research findings suggest that outcome history is a
predictor variable, with a Positive outcome history leading to higher risk Propensity. The
level of risk customers are assuming shows a significantly increasing trend.
Keywords Decision making, risk perceptions, financial products, consumer behaviour,
behaviour finance
need to investigate how they evaluate
INTRODUCTION
financial products, including how
Falling stock markets over recent years consumers perceive investment risk, to
have led to some financial products ensure that they understand the products
returning less than the initial investment to they buy. Decision making within
Kathleen Byrne
Cardif Pinnacle,
consumers when they have matured. financial products has not received much
Pinnacle House,
A1 Barnet Way, Despite product literature including research attention and consumer
Borehamwood,
information about risk to capital, understanding of risk within financial
Hertfordshire
WD6 2XX, UK.
consumers appear not to have understood products even less so. These topics are of
Tel: [phone];
Fax: [phone];
the risks they took on. practical importance for consumer
e-mail: kathy.byrne@
Providers of financial services products understanding of financial products.
cardifpinnacle.com
Henry Stewart Publications 1363 0539 (2005) Vol. 10, 1 21 36 Journal of Financial Services Marketing 21
Byrne
Literature on decision making and its to them. While different processes may be
relevance to financial decisions is reviewed. used to evaluate products and form
Risk is defined in relation to financial preferences, decisions are rational and
products and different facets of risk are cognitively based.
discussed. The research methodology Risk perceptions play an important role
includes a research model, constructs used in evaluating competing products and
and details of the research sample. Results behavioural finance authors5 10 have
are presented and practical implications for demonstrated that risky decision making
product providers are discussed. can be irrational rather than rational and
cognitively based. In complex decision
making consumers must make choices
LITERATURE REVIEW
between different alternatives and use
Consumer decision making literature in heuristics, or rules of thumb, to simplify
the context of financial products is the choice they make and in doing so
surprisingly scarce. This was noted by introduce bias into their decisions.
Harrison1 in this Journal in 2003 and Heuristics and biases include
Goodman2 explained the actuarial representativeness,5,11 16 availability,6,17,18
profession s recent research into consumer anchoring,6,16 overconfidence,19 loss
understanding of risk. Both economists aversion,11,17,20 status quo bias,21,22
and psychologists have studied risky hindsight bias,23 confirmation bias17 and
decision making. However, much research mental accounting.24
relates to simple choices, often using Representativeness leads to bias because
gambling devices. There is little coverage people ignore objective information that
of consumer understanding of financial does not fit with their stereotype and place
risk involved in the evaluation of financial more weight on information confirming
products. stereotypes. Jordan and Kass16 investigated
the role of judgmental heuristics in private
investors evaluation of risk and return.
Decision-making models
Anchoring, representativeness and the
Decision making for financial products is affect heuristic are used by investors and
complex because products are intangible, lead to biases in risk/return judgments.
outcomes are uncertain and financial risk They found that anchoring effects occur in
following a poor decision is significant. expected returns, whereas the affect and
Marketing textbooks3,4 present a five-stage representativeness heuristics affect
model of complex decision making. perceived investment risk. Both informed
Evaluation of competing products is the and uninformed groups show judgmental
stage at which investment risk is evaluated heuristics, with uninformed investors
and quantified and this is the focus of this showing larger biases.
paper. The way that information is presented
or  framed means that alternative
Evaluation of alternatives descriptions of a decision problem give rise
Marketing textbooks3,4 state that to different preferences when the same
consumers seek certain benefits from the problem is framed differently.25,26
alternatives they consider. Products are a
series of attributes with different abilities
What is risk?
to produce benefits to satisfy the need
identified. Consumers decide which The classical economist view of risk is a
attributes are relevant and most important situation where the future outcome is
22 Journal of Financial Services Marketing Vol. 10, 1 21 36 Henry Stewart Publications 1363 0539 (2005)
How do consumers evaluate risk in financial products?
unknown but a probability can be placed mediate the effects of problem framing
on each possible outcome. Dean and and outcome history on risky decision-
Thompson27 put forward two major making behaviour. They conclude that a
concepts of risk, the positivist and mediated model of risk behaviour is more
contextualist. A positivist concept of risk is powerful than one in which the direct
defined in terms of probabilities based on effects of a large number of antecedent
objective, verifiable experience and a variables are examined individually.
contextualist concept of risk is based on
the context in which it is used. Olsen28
Risk propensity
suggests that risk is a continuum taking
into account both the emergent and Several researchers30,32,33 have investigated
multidimensional nature of risk and the link between personality and risk
concludes that experts tend to focus on propensity to test whether risk propensity
probabilistic models whereas non-experts is a stable personality trait. Their findings
use contextual models. suggest that risk propensity or risk
Kahneman and Tversky6 were the first preference is not a personality trait and
to bring behavioural aspects into people s risk perceptions are important in
economically based risk models. They their propensity to take or avoid risks.
found that people appear irrational in Ensuring people have the right perception
decision making and utility theory does of the level of risk should mean that they
not fully explain how people make understand the risks they are taking on.
decisions. They offered an alternative
theory, prospect theory, which assigns
Investment risk
values to gains and losses rather than to
final assets and replaces probabilities with Olsen28 found that professional investment
decision weights. They observed two managers and experienced individual
effects, first the certainty effect, which investors share a common conception of
leads to risk aversion in choices for sure investment risk. Jordan and Kass16 found
gains and risk seeking in choices for sure that investors risk perceptions have four
losses. Secondly, the isolation effect where dimensions:
people generally discard components
shared by all prospects leading to  downside risk
inconsistent preferences when the same  upside risk
choice is presented in different formats.  volatility
Other models consistent with prospect  ambiguity
theory include Einhorn and Hogarth s29
ambiguity model and Harvey s17 heuristic While there is overlap between these and
judgment theory. Olsen s findings, there are also differences.
Sitkin and Pablo30 identify several Olsen did not include the ambiguity
studies that present contradictory results to dimension of Jordan and Kass, although he
those predicted by prospect theory. They considered economic uncertainty.
present a model, based on theoretical
analysis, reconciling these contradictions
Risk and return
by examining the usefulness of placing risk
propensity and risk perception in a more Ganzach34 found that for familiar assets,
central role than previously recognised. risk/return judgments tend to be derived
Sitkin and Weingart31 test this model in from past performance as proxies for
which risk propensity and risk perception current risk perceptions. The relationship
Henry Stewart Publications 1363 0539 (2005) Vol. 10, 1 21 36 Journal of Financial Services Marketing 23
Byrne
between risk and return judgments is propensity, past experience and framing
positive. For unfamiliar assets, risk and are also important. Experts appear better
return judgments are derived from global at evaluating investment risk than novices
preferences and the relationship between and the impact of expertise on risk
risk and return is negative. perceptions is likely to be important. (See
Muradoglu35 looked at experts and Table 1.)
novices predictions of stock prices.
Investors are positive feedback traders
METHODOLOGY
when presented with a time series
excluding contextual information,
Research context
supporting previous research. With
contextual and real-time information, The research sample was drawn from
however, optimism is the norm, bullish investment consumers so that the results
trends are extrapolated and mean reversion would be directly applicable to investment
is expected in bear markets only. products. The research instrument was
Differences are observed in return designed to simulate investment decisions
expectations and perceived risks due to that consumers might expect to make
contextual information, stock market when considering an investment rather
trends and expertise. than using simplified gambling examples
that have often been used to research risky
decision making.
Summary
People appear far from rational in making
Research model
decisions under risk because the complex
nature of financial decisions means that Role of risk perceptions
heuristics used to simplify complex Sitkin and Weingart s model31 was
decisions bring bias into the evaluation. developed for management decisions under
Risk perceptions appear to play a central risk but appears potentially applicable to
role in financial decisions. The role of risk financial decisions as risk perceptions are a
Table 1 Literature summary
Main research findings Author(s)
Financial decisions use complex decision making Assael, 19953; Kotler, 20034
Evaluation of alternatives
Decisions are rational and cognitively based Assael, 19953; Kotler, 20034
Risky decision making can be irrational Tversky & Kahneman, 197415; Kahneman &
Tversky, 197916; Thaler, 19807; Slovic, 19918;
Camerer, 19989; Rabin, 199810
Framing Khaneman, 200025; Johnson et al, 199326
Risk
Risk is multidimensional and emergent Dean & Thompson, 199527; Olsen, 199728;
Jordan & Kass, 200216
Risk propensity is not a personality trait Sitkin & Pablo, 199230; Sitgin & Weingart,
199531; Weber & Milliman, 199732; Weber, Blais
& Betz, 200233
Risk/Return relationship is positively correlated for Ganzach, 200034; Muradoglu, 200235
experts/familiar assets and negatively correlated for
novices/unfamiliar assets
Risk propensity and risk perception are key mediating Sitkin & Pablo, 199230; Sitkin & Weingart,
variables in risky decision making 199531
24 Journal of Financial Services Marketing Vol. 10, 1 21 36 Henry Stewart Publications 1363 0539 (2005)
How do consumers evaluate risk in financial products?
+
+ Risk
Outcome History
Propensity
H1
H1
How much to
H2
invest


Risk
+
+
Problem Framing
Perception
Novices

Experts
H4
+
+
H3
Return
Expectations
H5
Source: Adapted from Sitkin and Weingart (1995)31
Figure 1 Model of the determinants of investment decisions
key mediating variable in both types of The higher the level of perceived risk, the
decision. For financial decisions, the less likely someone is to take that risk and
expected return and investor expertise are vice versa. In the investment context, the
also expected to be important. This model higher the level of perceived risk the less
is used as a research framework, adapted likely someone is to invest and, if they do
to include return and expertise variables invest, the higher the perceived risk the
(Figure 1). A positive or negative sign smaller the investment. Conversely, the
indicates the direction of each relationship. lower the perceived risk, the more they
The purpose of the research is to test will invest.
whether the relationships indicated in the
model exist and their strength. Therefore H3: There is positive correlation between
the research looks at pairs of relationships risk perception and expected return for
rather than testing the whole model. The expert investors
hypotheses to be investigated are also
shown and are now discussed. H4: There is negative correlation between
risk perception and expected return for
H1: Outcome history is positively novice investors
correlated with risk propensity
The link between risk perception and a
Successful risk-taking experience is decision to invest is complicated by an
expected to lead to a higher risk propensity additional variable, which is the expected
and unsuccessful experience to risk aversion return from the investment. The literature
in future decisions. review shows the rational view is that the
higher the level of risk associated with an
H2: Risk propensity is negatively investment then the higher the return the
correlated with risk perception investor should expect. This is where
Henry Stewart Publications 1363 0539 (2005) Vol. 10, 1 21 36 Journal of Financial Services Marketing 25
Byrne
investment expertise may differentiate or considered a lump sum investment. By
behaviour, with experts following the using this customer base bias is introduced,
rational view of positive correlation as Pinnacle s customers may not be
between risk and return and novices representative of the whole population of
showing negative correlation. UK consumers with a lump sum to invest.
Research was conducted by
H5: Return expectations will be driven by
questionnaire mailed to selected customers.
past performance information provided
In order to be able to investigate decision
as an anchor
making across different customer groups,
the customer base was segmented by
Framing effects are difficult to measure in
customers past investment experience with
complex situations and so the effect of
Pinnacle. Two main product types had
anchoring on return expectations is the
been offered in the past, a guaranteed
only framing effect investigated.
insurance bond (GIB) and structured
capital at risk products (Scarps) where the
Constructs
return was linked to the performance of
specified stock market indices. Falling
Experience
stock markets have reduced returns under
Several authors looked at investors
Scarps and equities but have not affected
expertise in their research.16,28,35 Jordan
GIB returns as GIBs are fixed rate
and Kass s construct using knowledge ( I
products. GIB customers have had good
know a lot about investing money ) and
experience with their investment, while
experience ( I am very experienced at
Scarp customers may have been
investing ) was used as this was most
disappointed with their returns. A third
relevant to financial products.
customer group, direct customers, was
identified. Some direct customers have
Risk perception
policies with Pinnacle and some do not, so
This construct encapsulated components
this group has unknown past experience.
covering both probabilistic and contextual
Three customer groups were identified as
elements of risk perception:
follows:
 Upside risk (potential for good returns)
 Downside risk (potential for loss, not Premier customers
meeting investment objectives, strength These customers had structured capital
of regulation) at risk (Scarp) policies ( Premier
 Volatility (returns varying over time) Bonds ) that matured between
 Feelings (uncertainty, worry) November 2003 and March 2004.
GIB customers
Risk propensity
These customers had guaranteed
This was defined as the likelihood to
insurance bonds (GIBs) that matured
engage in a particular activity and in the
between June 2003 and June 2004 and
investment context was measured by the
were subject to similar economic
amount invested in a specified product.
conditions to Premier customers.
Direct customers
Research sample
These customers have registered to
Investment customers of Pinnacle receive investment offers and invest
Insurance plc (Pinnacle) were selected as a directly rather than using an
suitable group to study as they had made independent financial adviser (IFA).
26 Journal of Financial Services Marketing Vol. 10, 1 21 36 Henry Stewart Publications 1363 0539 (2005)
How do consumers evaluate risk in financial products?
Two versions of the questionnaire were paragraph preceding these. Questions were
used; version two included past structured with a four-point scale so that
performance information to test the effect respondents had to make a choice on
of anchoring on return expectations. Each whether or not they agreed or disagreed
of the sample groups needed to have with the statement. The choice was
approximately 50 respondents so that limited to four, as customers were not
results compared across groups would be expected to be able to make very fine
statistically significant. A 10 per cent distinctions. In some questions a  don t
response rate was expected, therefore know response was allowed so that
questionnaires were sent to roughly 1,000 respondents were not forced to make a
customers in each customer group so that choice when they genuinely did not know
around 50 responses for each version of how to answer. Appendix 1 shows the
the questionnaire would be obtained. questions relevant to this paper.
An extract from the customer database
for the three customer groups was
RESULTS
obtained at 30th June, 2004. A report was
run against this database to remove the There were 371 useable questionnaires, an
following customers: overall response rate of 11.24 per cent, with
53 per cent responding on version 1 and 47
 Customers who had opted out from per cent on version 2. Table 2 summarises
mailings the responses. Statistical tests were based on
 Private bank customers, since these 95 per cent confidence intervals.
IFAs might object to questionnaires
being sent to their clients
Demographics
 Customers who had invested over
Ł100,000 in a single investment Figure 2 shows the age profile of the
because the aim was to look at the respondents compared with the age profile
views of the mass affluent rather than of the database that was mailed with the
ultra high net worth customers. questionnaire. Age was not available in the
Direct database, so comparisons are shown
for the GIB and Premier database against
The research instrument
GIB and Premier respondents and against
Questions were based on previously tested all respondents. The old age profile of
questionnaires where possible because these respondents reflects the age profile of the
questions had already been tested for database.
validity. The questionnaire was divided Gender information was available for
into five sections with an introductory both respondents and the full mailing
Table 2 Questionnaire responses
Premier GIB customers Direct customers Customer
customers group
unknown
Number mailed 622 618 501 500 530 530
Questionnaire version 1 2 1 2 1 2 1 2
Responses 54 66 56 61 63 29 24 18
received
Responses by 120 117 92 42
group
Response rate 9.68% 11.69% 8.68%
Henry Stewart Publications 1363 0539 (2005) Vol. 10, 1 21 36 Journal of Financial Services Marketing 27
Byrne
40.00%
35.00%
30.00%
25.00%
20.00%
15.00%
10.00%
5.00%
0.00%
18 to 29 30 to 39 40 to 49 50 to 59 60 to 69 70 to 79 80 and over
Age Group
GIB & Premier Database GIB & Premier Respondents All Respondents
Figure 2 Age profile
70.00%
60.00%
50.00%
40.00%
30.00%
20.00%
10.00%
0.00%
Male Female Unknown
Gender
Database All Respondents
Figure 3 Gender profile
database. Figure 3 shows a comparison
Investment expertise
between the database and respondents,
which shows that males are over- Respondents ratings of their knowledge
represented in the sample and females are and experience with investments were
under-represented compared with the used to calculate an expertise score as the
database population. arithmetic mean of these scores.
28 Journal of Financial Services Marketing Vol. 10, 1 21 36 Henry Stewart Publications 1363 0539 (2005)
% of population
% of population
How do consumers evaluate risk in financial products?
The level of investment expertise was
Risk perception and propensity
compared between the three customer
groups. One-way ANOVA tests, followed Respondents completed risk assessments
by the Scheffe procedure, identified and risk likelihoods (likelihood of
significant differences. Direct customers engaging in each activity) for nine
have significantly higher expertise levels gambling and simple investment decisions.
than other customers but GIB and Premier For most cases there was negative
customers have similar expertise. correlation between risk likelihood and
A grouping variable was set according risk perception; only one gambling
to expertise score. Experts (49 per cent of activity did not show significant
sample) included respondents who agreed correlation between risk likelihood and
that they were both knowledgeable and risk perception. This means that the higher
experienced with investments, while the level of perceived risk the less likely a
novices (51 per cent) disagreed with these respondent is to engage in it.
statements.
Investment risk perception
Risk profile
Respondents provided their perception of
The financial products investors hold in risk inherent in four defined financial
their portfolios, both now and previously, products (deposits, property, equities and a
were investigated. This allowed the actual structured capital at risk product) by
level of risk taken on to be measured for answering a series of eight perception
comparison with risk perceptions and questions for each product. An average
investment choices. For each respondent risk perception score was calculated by
two variables were calculated by product, Cronbach s alpha ranged from
weighting their investment holdings 0.60 to 0.81. Product definitions are shown
according to the level of risk, as defined in Appendix 2.
by an expert, to calculate a current risk
score and a past risk score. These variables
Future returns
can be compared to see whether past risk
propensity is correlated with current risk Expected future returns under the four
propensity, enabling Hypothesis 1 to be financial products assessed for risk were
tested. There was strong correlation investigated. Respondents using version 2
between current and past risk scores (78 of the questionnaire were given past return
per cent), which confirms Hypothesis 1 information for each product.
and suggests that past risk behaviour is an Independent sample t tests were used to
indicator of future risk behaviour. test Hypothesis 5 to establish whether
A paired sample t test was used to test there was a difference in means between
the null hypothesis of no difference in the two questionnaire groups. There was a
current and past investment risk scores. significant difference for deposit accounts,
This showed that differences were property and Scarps but not for equities.
statistically significant. The mean past risk This may be because the past performance
score is lower than the mean current risk for equities has been poor and produced
score, which suggests that the level of risk negative returns and performance
in investment portfolios is increasing over information is discarded as it does not fit
time. This increase in risk propensity with respondents future return
supports the proposition that successful perceptions. In all cases, the group given
risk taking leads to higher risk propensity. past performance data has a higher mean
Henry Stewart Publications 1363 0539 (2005) Vol. 10, 1 21 36 Journal of Financial Services Marketing 29
Byrne
score than the group not provided with test was carried out to test significance.
this information. This suggests that for Differences were significant for deposit
some investments, providing past and Scarp allocation but were not
performance information creates significant for property or equity
expectations of higher returns than if no allocations. Using a Scheffe procedure, the
past performance information were differences in means for deposit allocation
provided. Therefore Hypothesis 5 is are significant between Premier and GIB
proved for some, but not all, products customers for deposit allocation, with
tested. Premier customers allocating the least to
Correlations between risk perception this product type.
and future return expectations were Interestingly, the difference in allocation
investigated. For all product types except for Scarps was between Premier customers
property, there was significant correlation. and Direct customers, with Premier
Experts had significant positive correlation customers allocating more than any other
between risk assessment and future return customer group. These customers could be
expectations for all product types except considered to have had bad previous
property, while novices showed no experience with this product and might
correlation. These results support therefore have been expected to be less
Hypothesis 3 for all product types except likely to invest. Conversely, Premier
for property, but do not support customers, having invested in Scarps
Hypothesis 4. previously, are more likely to be familiar
Individual investors who do not have with this investment.
expertise in property may explain the lack The relationship between risk
of correlation. Property is the third least perceptions and investment allocation (risk
likely investment to be held, only 26 per propensity) was used to test Hypothesis 2
cent of respondents currently hold for complex products. Negative correlation
property investments while a further 10 between risk perception and investment
per cent have held this type of investment allocation was significant for all investment
previously. A chi-squared test shows no types except for deposits. This means that
difference in the propensity to hold investors allocate least to the assets they
property between experts and novices. perceive as most risky. Hypothesis 2 is
proved for all products except deposits.
Further investigation into deposits
Investment decision
showed no significant differences among
Respondents were asked to allocate a experts and novices. Individuals using
Ł10,000 investment between the four deposit accounts for  rainy day money
financial products used in previous might explain this and therefore most
questions. Table 3 shows the mean amount investors allocate some money to a deposit
allocated between products. account irrespective of their risk
The amount allocated differed between assessment. Deposits feature in 89 per cent
customer groups, so a one-way ANOVA of respondents portfolios.
Correlation between future return
Table 3 Mean investment allocations expectations and investment allocation was
significant for all investment types except
Product Mean amount allocated
for deposits. The higher the expected
Deposits Ł4,723.30
return, the higher the investment allocation.
Property Ł2,264.56
Equities Ł2,259.55
No significant differences were found
Scarps Ł781.23
between experts and novices for deposits. A
30 Journal of Financial Services Marketing Vol. 10, 1 21 36 Henry Stewart Publications 1363 0539 (2005)
How do consumers evaluate risk in financial products?
decision to invest in deposits does not consumer knowledge, understanding and
appear to take account of future return expertise should be welcomed. This
expectations or level of perceived risk. research highlights the importance of the
FSA s consumer education remit.
DISCUSSION
Risk model
Investment expertise
The research results generally support the
Expertise is important for consumer relationships identified in the model of
understanding of risk and the link between determinants of investment decisions
risk and return. Direct customers had presented in Figure 1.
significantly higher expertise scores than The portfolio holdings of respondents
other customers. This may be expected as support the link between a positive
direct customers have an active outcome history leading to a higher risk
involvement with investments and buy propensity. Premier customers, however,
direct rather than using an IFA. In who could be considered as having had an
contrast, most other customers are unsuccessful risk-taking experience, have a
introduced by an IFA and therefore seek higher propensity to invest in a similarly
external advice on investment decisions. risky product than other customer groups.
Improving investment expertise, The returns under Premier policies were
measured by knowledge and experience, below less risky products such as deposits
should enable consumers to make better but were above higher risk products such
investment decisions. Expert consumers as equities. These customers could consider
more accurately assess risk and have a the experience to have been successful, but
better understanding of risk/return this is unlikely.
relationships. Any initiatives to improve Both literature review and research
+
+ Risk
Outcome History
Propensity
How much to
invest


Risk
+
+
Problem Framing
+
+
Perception
Experts
+
+
Return
Expectations
Past Performance
+
+
Information
Figure 4 Updated model of the determinants of investment decisions
Henry Stewart Publications 1363 0539 (2005) Vol. 10, 1 21 36 Journal of Financial Services Marketing 31
Byrne
findings indicate the central role that risk may be biased towards low risk investors.
perceptions play in financial decisions. This The customer age profile is skewed
is important because perceptions can be towards older ages, which is unlikely to be
influenced by how information is representative of the age profile of
presented. As expected, risk propensity investors in general.
and risk perception were negatively
correlated with each other for both simple
CONCLUSIONS
decisions and also more complex
investment allocation. However, deposit The research has shown that the more
accounts were selected for investment expertise a consumer has, the more
irrespective of how risky a respondent accurately they perceive risk and the better
considered them to be. This suggests that they understand the link between risk and
this asset class is seen as a suitable return. Expertise appears to be important
investment by all investors, at least for in consumers understanding of the
part of their portfolio. products they have purchased.
Return expectations were positively Trends in the UK, such as the move of
correlated with investment allocation and pension schemes from defined benefit to
this relationship could be added to the defined contribution, mean that
model. A revised model (Figure 4) is increasingly financial decisions will be
proposed that takes account of this result placed in the hands of consumers who are
and the lack of correlation between risk and likely to be inexperienced and novice
return expectations for novices. The investors. Consumers need financial
findings confirm outcome history as a knowledge and experience to be properly
predictor variable and expected return as an equipped to make these types of financial
additional mediating variable, which in turn decision. This means that financial
is mediated by risk perception for experts. education will become even more
important in the future. The FSA, the
media and product providers all
Limitations
potentially have roles to play. The
The research sample may not be possibilities of advisers mis-selling or
representative of investment customers in consumers mis-buying products should
general because Pinnacle offers a limited be reduced if consumers are better
investment product range. These are low informed and understand the risks they
to medium risk products, so the sample are taking on.
32 Journal of Financial Services Marketing Vol. 10, 1 21 36 Henry Stewart Publications 1363 0539 (2005)
How do consumers evaluate risk in financial products?
APPENDIX 1: SELECTED QUESTIONS
Q3 Please indicate the extent to which you agree with the following statements
about your knowledge and experience with investments:
Strongly Agree Disagree Strongly Don t
agree disagree know
I know a lot about investing money
I am very experienced in investing
Q5 Which of the following types of investment products do you currently hold
or have held in the past?
Deposit account National Savings
Government bonds (gilts) PEP
Guaranteed insurance bond GIB (Life
Insurance bond) Property (excluding your own house)
ISA  mini cash Stocks and shares
ISA  mini stocks and shares Structured product (no risk to capital) *
ISA  mini insurance Structured product (capital at risk) **
ISA  maxi stocks and shares TESSA
Investment trusts Unit trusts
With profits bond
Questions 6 and 7 used the same statements but required a different choice of responses.
The statements were:
Betting a day s income at the horse races
Investing 10 per cent of your annual income in a moderate growth unit trust
Lending a friend an amount equivalent to one month s income
Investing 5 per cent of your income in a very speculative stock
Betting a day s income on the outcome of a sporting event (eg football match)
Investing 5 per cent of your annual income in a building society deposit
Investing 10 per cent of your annual income in government bonds (gilts)
Gambling a week s income at a casino
Taking a job where you get paid exclusively on a commission basis
Q6 For each of the following statements, please indicate your likelihood of
engaging in each activity.
Responses: Very unlikely, unlikely, likely, very likely
Q7 For each of the following statements, please indicate your initial reaction of
how risky each situation is.
Responses: Not at all risky, a bit risky, moderately risky, extremely risky
Q8 For each type of investment indicated, please answer the following questions
assuming that you have invested in it.
Henry Stewart Publications 1363 0539 (2005) Vol. 10, 1 21 36 Journal of Financial Services Marketing 33
Byrne
(Question required responses for each of four products, which were Deposits, Property,
UK Stocks & shares, Structured capital at risk product)
a) This investment bears a high risk of losing money
b) This investment bears a high risk of missing personal investment objectives
c) I feel uncertain about investing in this investment as I feel uninformed about it
d) Investing in this investment also entails good chances to realise higher, above average
returns
e) I think there will be significant variations in performance over time
f) Investors who experience losses with this type of investment are very likely to lose
most or all of their investment
g) I would worry very much if I had invested in this type of investment
h) This type of investment is very well regulated
Responses: Strongly agree, Agree, Disagree, Strongly disagree, Don t know
Q9 For each of the investment types in Q8, please estimate the amount you
would expect back at the end of the next five years assuming you have
Ł10,000 to invest.
(Question required responses for each of four products, which were Deposits, Property,
UK Stocks & shares, Structured capital at risk product)
(Version 2 of the Questionnaire included the following note:
Note: Over the last five years, if you had invested Ł10,000 you would have got back on
average Ł11,716 from a deposit account, Ł19,918 from property, Ł8,528 from UK
stocks & shares and Ł10,420 from a structured capital at risk bond.)
Responses:
Less than Ł6,000 Ł6,000 to Ł7,749 Ł7,750 to Ł8,799 Ł8,800 to Ł9,999
Ł10,000 to Ł11,500 Ł11,501 to Ł12,750 Ł12,751 to Ł16,000 More than Ł16,000
Q10 Please allocate an amount of Ł10,000 between the following investment
types according to your investment preferences:
Investment type Investment amount
Deposit account
Property
UK Stocks & shares
Structured capital at risk bond
Total Ł10,000
Q12 Please indicate your age group
18 29, 30 39, 40 49, 50 59, 60 69, 70 79, 80 and over
Q13 Please indicate your gender: Male Female
34 Journal of Financial Services Marketing Vol. 10, 1 21 36 Henry Stewart Publications 1363 0539 (2005)
How do consumers evaluate risk in financial products?
APPENDIX 2: INVESTMENT DEFINITIONS
Deposit account: A bank or building society account that allows you to withdraw
your money without penalty at any time.
Guaranteed insurance bond: An insurance policy that pays a fixed income each year
and returns your investment at the end of a fixed term. Monthly income, annual income
or growth options are usually available.
Property: A direct investment in property, which excludes your own house but
includes second homes, buy to let property and commercial property.
Stocks and shares (Equities): Stocks and shares in UK companies such as BP, Marks
and Spencers, BT etc, also known as equities.
Structured product: A product where the return is linked to the performance of one
or more stock market indices.
Structured capital at risk product (Scarp): A product that offers fixed income but
the return of your capital at the end of the term will be reduced if the stock market is
below its starting level at the end of the investment term.
# Kathleen Byrne
12 Tversky, A. and Kahneman, D. (1982)  Judgments of
REFERENCES
and by representativeness , in Kahneman, D., Slovic, P.
1 Harrison, T. (2003)  Understanding the behaviour of and Tversky, A. (eds)  Judgment Under Uncertainty:
financial services consumers: A research agenda , Journal Heuristics and Biases , Cambridge University Press,
of Financial Services Marketing, Vol. 8, No. 1, p. 6. Cambridge, UK, pp. 84 100.
2 Goodman, A. (2004)  Consumer Understanding of 13 Kahneman, D. and Tversky, A. (1972)  Subjective
Risk , The Actuarial Profession, London, UK. probability: A judgment of representativeness ,
3 Assael, H. (1995)  Consumer Behavior and Marketing Cognitive Psychology, Vol. 3, pp. 430 454.
Action , 5th edn, South-Western College Publishing, 14 Kahneman, D. and Tversky, A. (1973)  On the
Cincinnati, USA. psychology of prediction , Psychological Review, Vol.
4 Kotler, P. (2003)  Marketing Management , 11th edn, 80, pp. 237 251.
Pearson Education Inc., USA. 15 Bar-Hillel, M. (1982)  Studies of representativeness , in
5 Tversky, A. and Kahneman, D. (1974)  Judgment Kahneman, D., Slovic, P. and Tversky, A. (eds)
under uncertainty: Heuristics and biases , Science, Vol.  Judgment Under Uncertainty: Heuristics and Biases ,
185, pp. 1124 1131. Cambridge University Press, Cambridge, UK, pp. 69
6 Kahneman, D. and Tversky, A. (1979)  Prospect 83.
theory: An analysis of decision under risk , 16 Jordan, J. and Kass, K. (2002)  Advertising in the
Econometrica, Vol. 47, No. 2, pp. 263 91. mutual fund business: The role of judgmental heuristics
7 Thaler, R. H. (1980)  Towards a positive theory of in private investors evaluation of risk and return ,
consumer choice , Journal of Economic Behavior and Journal of Financial Services Marketing, Vol. 7, No. 2,
Organization, Vol. 1, pp. 39 60. pp. 129 140.
8 Slovic, P. (1991)  The construction of preference , 17 Harvey, J. T. (1998)  Heuristic judgment theory ,
American Psychologist, Vol. 50, No. 5, pp. 364 71. Journal of Economic Issues, Vol. 32, No. 1, pp. 47 64.
9 Camerer, C. (1998)  Prospect Theory in the Wild: 18 Combs, B. and Slovic, P. (1979)  Causes of death:
Evidence from the Field , Social Science Working Paper Biased newspaper coverage and biased judgments ,
1037. Journalism Quarterly, Vol. 56, pp. 837 843.
10 Rabin, M. (1998)  Psychology and economics , Journal 19 Camerer, C. and Lovallo, D. (2000)  Overconfidence
of Economic Literature, Vol. XXXVI, pp. 11 46. and excess entry  An experimental approach , in
11 Tversky, A. and Kahneman, D. (1971)  Belief in the law Kahneman, D. and Tversky, A. (eds)  Choices Values
of small numbers , Psychological Bulletin, Vol. 2, pp. and Frames , Cambridge University Press, Russell
105 110. Sage Foundation, Cambridge, UK.
Henry Stewart Publications 1363 0539 (2005) Vol. 10, 1 21 36 Journal of Financial Services Marketing 35
Byrne
20 Baz, J., Briys, E., Bronnenberg, B. J., Cohen, M., Kast, 28 Olsen, R. (1997)  Investment risk: The expert s
R., Viala, P., Wathieu, L., Weber, M. and perspective , Financial Analysis Journal, Vol. 53, No. 2,
Wertenbroch, K. (1999)  Risk perception in the short pp. 62 66.
run and in the long run , Marketing Letters, Vol. 10, 29 Einhorn, H. and Hogarth, R. (1986)  Decision making
No. 3, pp. 267 283. under ambiguity , Journal of Business, Vol. 59, No. 4,
21 Kahneman, D., Knetsch, J. and Thaler, R. (1991)  The  Part 2: The behavioural foundations of economic
endowment effect, loss aversion and status quo bias , theory (October 1986), S225 S250.
Journal of Economic Perspectives, Vol. 5, No. 1, pp. 193 30 Sitkin, S.B. and Pablo, A.L. (1992)  Reconceptualizing
206. the determinants of risk behaviour , Academy of
22 Samuelson, W. and Zeckhauser, R. (1988)  Status quo Management Review, Vol. 17, pp. 9 39.
bias in decision making , Journal of Risk and Uncertainty, 31 Sitkin, S.B. and Weingart, L.R. (1995)  Determinants
Vol. 1, pp. 7 59. of risky decision-making behavior: A test of the
23 Fischhoff, B. (1982)  For those condemned to study the mediating role of risk perceptions and propensity ,
past: Heuristics and biases in hindsight , in Kahneman, Academy of Management Journal, Vol. 38, No. 6, pp.
D., Slovic, P. and Tversky, A. (ed)  Judgment Under 1573 1582.
Uncertainty: Heuristics and Biases , Cambridge 32 Weber, E. and Milliman, R. (1997)  Perceived risk
University Press, Cambridge, UK, pp. 335 351. attitudes: Relating risk perception to risky choice ,
24 Thaler, R. H. (1999)  Mental accounting matters , Management Science, Vol. 43, No. 2, pp. 123 144.
Journal of Behavioral Decision Making, Vol. 12, pp. 183 33 Weber, E., Blais, A.-R. and Betz, N. (2002)  A
206. domain-specific risk-attitude scale: Measuring risk
25 Kahneman, D. (2000)  Preface, in Kahneman, D. and perceptions and risk behaviors , Journal of Behavioral
Tversky, A. (eds)  Choices, Values and Frames , Decision Making, Vol. 15, No. 4, pp. 263 290.
Cambridge University Press, Russell Sage Foundation, 34 Ganzach, Y. (2000)  Judging risk and return of financial
Cambridge, UK, pp. ix xvii. assets , Organizational Behavior and Human Decision
26 Johnson, E., Hershey, J., Meszaros, J. and Kunreuther, Processes, Vol. 83, pp. 353 370.
H. (1993)  Framing, probability distortions and 35 Muradoglu, G. (2002)  Portfolio managers and
insurance decisions , Journal of Risk and Uncertainty, novices forecasts of risk and return: Are there
Vol. 7, pp. 35 51. predictable forecast errors? , Journal of Forecasting, Vol.
27 Dean, W. and Thompson, P. (1995)  The Varieties of 21, No. 6, pp. 395 416.
Risk , Working Paper, University of Alberta, Canada.
36 Journal of Financial Services Marketing Vol. 10, 1 21 36 Henry Stewart Publications 1363 0539 (2005)