Evidence on rationality and
behavioural biases in investmentMarketing Research and Data Analysis
decision making
Satish Kumar and Nisha Goyal
Department of Management Studies,
Malaviya National Institute of Technology, Jaipur, India
Abstract
Purpose – The purpose of this paper is to investigate the relationship between rational
decision-making and behavioural biases among individual investors in India, as well as to examine the
influence of demographic variables on rational decision-making process and how those differences
manifest themselves in the form of behavioural biases.
Design/methodology/approach – Using a structured questionnaire, a total of 386 valid responses
have been collected from May to October 2015. Statistical techniques like t-test, analysis of variance
(ANOVA) and Fisher’s least signifcant difference (LSD) test have been used in this study. Structural
equation modelling (SEM) has been used to analyse the relationship between rational decision-making
and behavioural biases.
Findings – The fndings show that the structural path model closely fts the sample data, indicating
investors follow a rational decision-making process while investing. However, behavioural biases also
arise in different stages of the decision-making process. It further explores that gender and income have
a signifcant difference with respect to rational decision-making process. Male investors are more prone
to overconfdence and herding bias in India.
Research limitations/implications – The fndings of the study have signifcant implication for the
individual investors. It is recommended that if individuals are aware about the biases, they may become
alert before taking irrational investment decisions.
Originality/value – To best of the authors’ knowledge, the present study is a frst of its kind to
investigate the relationship between rational decision-making and behavioural biases among
individual investors in India.
Keywords ANOVA, Overconfdence, Herding, Disposition effect, Rational decision-making
Paper type Research paper
1. Introduction
Rational decision-making is combined with a structured or logical thought process to
achieve an effcient and optimal result. A theory has been proposed to achieve the
desired outcome of a decision-making process known as theory of rational choice. The
theory of rational choice asserts that decision makers consider a set of alternatives from
different scenarios before selecting a choice. To attain full rationality, people necessitate
unlimited cognitive capabilities. Natures of human beings are quite different with
limited cognitive capabilities. Because of this reason, decision behaviour of people
cannot follow full rationality. Afterwards, Simon (1956) proposed a new concept of
bounded rationality. It suggests that due to lack of information and memory errors
people make irrational decisions. Bounded rationality is a more realistic theory of the
human decision-making.
The current issue and full text archive of this journal is available on Emerald Insight at:
www.emeraldinsight.com/1755-4179.htm
QRFM
8,4
270
Received 27 May 2016
Revised 27 May 2016
Accepted 20 June 2016
Qualitative Research in Financial
Markets
Vol. 8 No. 4, 2016
pp. 270-287
© Emerald Group Publishing Limited
1755-4179
DOI 10.1108/QRFM-05-2016-0016
The standard fnance theories are based on the assumptions that investors make
decisions in a rational manner. Effcient market hypothesis also believes that security
prices reflect all the available information in the effcient market condition (Fama, 1970).
Traditional theory assumes that investors’ decisions are based on the expected utility
theory. Where the expected utility theory believes in the concept of rationality and states
that investors make consistent and independent decisions among various available
alternatives.
However, various research studies have documented that investors do not behave
rationally while making decisions. With this view, in 1980s, a new concept i.e.
behavioural fnance has emerged in the area of fnance and economics. Behavioural
fnance is based on two building blocks i.e. cognitive psychology and limits to
arbitrage (Thaler and Barberis, 2002). Cognitive psychology refers to the how people
think, perceive and remember, while limits to arbitrage is the arbitrage opportunity
that appears in the market and, arbitrageurs may not be able to make proft from
market dislocations because of their irrational behaviour.
Kahneman and Tversky (1979) developed the prospect theory for
decision-making under the uncertainty that is also a critique to the expected utility
theory. Prospect theory believes that some psychological factors influence the
investors’ decision-making and deviate them from the rationality, which supported
Simon’s (1956) argument of bounded rationality. These psychological factors
are termed as behavioural biases and would lead to decline in the investment
returns.
Investors’ rational decision-making process includes the procedures of
identifcation of demand, search for the information and evaluation of the
alternatives and then such investment decision will be considered as a rational
investment decision. Practically, investors do not follow the logical procedure due to
the availability of limited information. Research on investment behaviour of
individual investors in various countries has shown that most of the investors still
actually display behavioural biases.
Existing studies have either identifed the behavioural biases or analysed the
impact of these behavioural biases on the individual investors. But no prior attempt
has been made to analyse the relationship between behavioural biases and rational
decision-making process, especially in the Indian context.
The main objectives of the study are as follows:
• To establish the relationship between rational decision-making process and
behavioural biases, i.e. overconfdence, disposition effect and herding.
• To test the effect of various demographic variables on rational
decision-making process and how those differences manifest themselves in
the form of behavioural biases.
The remainder of the paper is structured as follows. Section 2 explains the literature
review related to behavioural biases and rationality in the decision-making process.
In Section 3, researchers introduce the research methodology opted in the study. In
Section 4, researchers outline the analysis and the main fndings of the study.
Finally, Section 5 summarises and concludes the study.
271
Rationality
and
behavioural
biases
2. Literature review
The review of literature is focused on the theoretical and empirical research studies on
rational decision-making process and behavioural biases that are considered for the
present study.
2.1 Rationality in decision-making
Rational decision makers follow a logic and reason for making an optimised decision
(Nozick, 1993). In the literature, research studies have proposed two basic models of
decision-making: rational model and bounded rationality model. For the rationality
in decision-making, Mintzberg et al. (1976) provided a three-stage model of strategic
decision process that includes identifcation of problem, development of alternative
solutions to the problem and fnally, the selection of the alternative for an optimal solution.
Further, Schoenfeld (2011) proposed a six-step model for the rational decision-making
process. Practically, it is always diffcult to take a rational decision because of
availability of limited information, inadequate time and cognitive limitations.
Therefore, Simon (1956, 1982) replaced the term rationality with the concept of bounded
rationality. Researchers suggest that bounded rationality is not the irrationality and,
agents are also not irrational; they are bounded rational. Generally, because of the lack
of complete information and knowledge, people use shortcuts in the form of adopting the
path of simple models that results in a suboptimal decision. In other words, individual
investors involved in investment activities also analyse and evaluate investment
options that seem like a rational decision-making process.
2.2 Behavioural biases in investment decision-making
Behavioural fnance deals with the behavioural and psychological aspects of the
investment decision-making. Researchers found various anomalies in the investors’
behaviour that deviates them from the rational and logical decisions and violate the
standard fnancial theory. These anomalies are the cognitive errors or the biases that
influence the investment decision-making. Kahneman and Tversky (1979) developed
the prospect theory and explained the human judgement and decision-making under
risk and uncertainty. The prospect theory states that people are risk-averse in the gains
but become risk-seekers in the losses. In this study, we have explained the relationship
between three behavioural biases, namely, overconfdence, disposition effect, the
herding behaviour and the rational decision-making process.
2.2.1 Overconfdence. Overconfdence is a cognitive bias in which people have
unwarranted faith in their intuitive reasoning, judgments and cognitive abilities
(Pompian and Wood, 2006). Overconfdent people become too confdent about their
skills and knowledge while underestimating the various risk associated with the
investment. Generally, overconfdent investors overreact to the private information
signals while ignoring the publicly available information (Daniel et al., 1998). Odean
(1999) proposed that overconfdent investors may trade even when their expected gains
are not suffcient to offset the transaction costs. Barber and Odean (2000) collected data
of 78,000 households for a period from 1991 to 1996 from a large discount brokerage
house in the USA. They documented that investors are overconfdent and trade
excessively and because of the excessive trading, gross returns (before accounting the
transaction cost) earned by the households were normal, while net returns were very
poor. Moreover, these results are empirically consistent with Statman et al. (2006) and
QRFM
8,4
272
Grinblatt and Keloharju (2000)). Furthermore, in a study, Barber and Odean (2001)
analysed the common stock investment of men and women by using a data set of 35,000
households from a large discount brokerage house. They proposed that men are more
overconfdent and trade excessively than women.
2.2.2 Disposition effect. Disposition effect is a phenomenon in which investors’
exhibit a tendency to realise the gains, while reluctant to realise losses (Shefrin and
Statman, 1985). Firstly, Shefrin and Statman (1985) developed a framework based on
different elements (i.e. mental accounting, regret aversion, self-control and tax
consideration) and formally analysed the disposition effect. Most of the empirical
studies referred to the Kahneman and Tversky’s (1979) prospect theory to explain the
disposition effect. The prospect theory states that people become more risk-averse after
experiencing gains while risk seekers after suffering from the losses. Odean (1998)
analysed the 10,000 customers’ accounts from a nationwide discount brokerage house
and also empirically supported the implications of the prospect theory. Further, they
also reported that because of tax consideration, the investors are engaged in loss realised
selling at the end of the year, i.e. December. A number of research studies have
supported the existence of disposition effect (Barber et al., 2007; Shapira and Venezia,
2001; Weber and Camerer, 1998; Grinblatt and Keloharju, 2000; Jordan and Diltz, 2004).
Barber et al. (2003) analysed the mutual fund’s purchase and sale decision of investors.
They found the evidence that investors sell those funds which have realised positive
returns and are reluctant to sell the loss-making funds. Furthermore, Shapira and
Venezia (2001) documented that individual investors are more prone to disposition effect
than the professional investors. Dhar and Zhu (2006) analysed the difference in the
disposition effect among individuals and reported that nonprofessional and the low
income group investors exhibit more disposition effect than others. In an experiment,
Shafran et al. (2009) investigated that investors are prone to disposition effect but also
affected by the trading conditions. Goetzmann and Massa (2008) found a negative
correlation between the disposition effect and returns, volatility and trading volume.
2.2.3 Herding. Herding refers to the tendency of the individuals to imitating the
judgements (rational and irrational) of others. Thus, herding behaviour of investors is
the primary cause of bubbles in fnance. Lee et al. (2004) suggested that individual
investors are noise traders and trade for the liquidity than the institutional investors.
Fernandez et al. (2011) proposed an interdependent relation between the information
availability and the herd behaviour. They found that when the information is uncertain,
investors are more prone to imitate the decisions of others or group. However,
Lakonishok et al. (1992) documented that during the trade of large stocks, the US
pension fund managers are less influenced to herd behaviour and have no impact on the
stock prices. Further, Grinblatt et al. (1995) also found weak evidence of herd behaviour
for US mutual funds. Wermers (1999) found less evidence of herding for the average
stocks, while high level of herding for the small and growth-oriented stocks. Trueman
(1994) reported that analysts exhibit herding behaviour; thus, ignore the available
information and release forecasts based on the other analysts’ previous decisions.
Similarly, Sias (2004) proposed that trading behaviour of institutional investors is
related to the position of the previous trading. Further, they also found that in large
stocks, herding behaviour weakens by the time. Nofsinger and Sias (1999) documented
that institutional investors’ herding affects the stock prices more than individuals’
herding in the USA.
273
Rationality
and
behavioural
biases
2.3 Influence of demographic variables on rational investment and behavioural biases
In addition to the behavioural factors, investors’ demographic characteristics have
signifcant effect on the rational investment and behavioural biases of investors. For
instance, Lin (2011) analysed that individual investors follow the rational
decision-making process to select their investment products and also prone to various
behavioural biases. Mathuraswamy and Rajendran (2015) found that family
composition, biological make-up, psychological factors and lifestyle of individual
investors influence the investment rationality. Zaidi and Tauni (2012) indicated that
both age and education do not have any signifcant impact on overconfdence bias.
Moreover, there is a signifcant association between investment experience and
overconfdence bias. By using a survey of about 2,000 defned contribution pension plan
members, Bhandari and Deaves (2006) found that men are more confdent than the
women. With respect to the relationship between demographic characteristics and
disposition effect, Da Costa et al. (2008) identifed that males are more prone to
disposition effect than females. Further, Dhar and Zhu (2006) found that individuals
employed in professional occupations and high-income earners have lower disposition
effect. With respect to the relationship between demographic characteristics and
herding bias, Lin (2011) found that females are more involved in herding behaviour than
males. Moreover, he identifed that young investors are more prone to herd behaviour
than older ones.
2.4 Linkage between rational decision-making and behavioural biases
Rational decision theory states that decision makers follow a logical path or sequence.
But based on the concept of bounded rationality or limited rationality in decisions,
Simon (1956) suggested that individuals consider some threshold of satisfaction rather
than maximising a utility function. A possible reason behind it could be that they
usually lack information on the defnition of problem and so on. After that, Kahneman
and Tversky (1979) extended the theory of rational decision by introducing the prospect
theory. The prospect theory includes the reference position on evaluating the optimal
decision. Tversky and Kahneman (1974) also presented that people rely on a limited
number of heuristics principles which can be quite useful, but sometimes lead to severe
and systematic error.
Although, it is possible that investors have evaluated the information objectively, it
is also diffcult to ignore the emotional and cognitive biases involved in each stage of the
decision-making process. Therefore, despite investors following a logical process for the
investment decision, behavioural biases still exist in the mind of the investors.
3. Methodology
3.1 Research model and hypotheses development
In the present study, to understand the rational decision-making process, the three-stage
model of Mintzberg et al. (1976) has been used. Following are the different stages of
rational decision-making process: problem identifcation, seeking essential information
and evaluating alternatives solutions. The research model postulates six constructs
(demand identifcation, search information, evaluating alternatives, overconfdence
bias, disposition effect and herding bias). This study also examines four demographic
variables (gender, age, income, occupation) that have varying influence on the primary
constructs. The research model tested in this study is shown in Figure 1.
QRFM
8,4
274
To achieve the objectives of the present study, the following hypotheses are framed:
3.1.1 Hypotheses

H1a. Demand identifcation has a signifcant relation to the overconfdence bias.
H1b. Information search has a signifcant relation to the overconfdence bias.
H1c. Evaluating alternatives has a signifcant relation to the overconfdence bias.

H2a. Demand identifcation has a signifcant relation to the disposition effect.
H2b. Information search has a signifcant relation to the disposition effect.
H2c. Evaluating alternatives has a signifcant relation to the disposition effect.

H3a. Demand identifcation has a signifcant relation to the herding bias.
H3b. Information search has a signifcant relation to the herding bias.
H3c. Evaluating alternatives has a signifcant relation to the herding bias.
H4. Gender has an effect on rational decision-making process.
H4a. Gender has an effect on demand identifcation.
H4b. Gender has an effect on information search.
H4c. Gender has an effect on evaluating alternatives.
H5. Gender has an effect on the behavioural biases.
H5a. Gender has an effect on overconfdence bias.
Demand
Identification

Evaluating
Alternatives
Herding bias
Disposition effect
Overconfidence
bias
Gender Age Income Occupation
Figure 1.
Proposed research
model
275
Rationality
and
behavioural
biases
H5b. Gender has an effect on disposition effect.
H5c. Gender has an effect on herding bias.
H6. Occupation has an effect on rational decision-making process.
H6a. Occupation has an effect on demand identifcation.
H6b. Occupation has an effect on information search.
H6c. Occupation has an effect on evaluating alternatives.
H7. Occupation has an effect on the behavioural biases.
H7a. Occupation has an effect on overconfdence bias.
H7b. Occupation has an effect on disposition effect.
H7c. Occupation has an effect on herding bias.
H8. Income has an effect on rational decision-making process.
H8a. Income has an effect on demand identifcation.
H8b. Income on has an effect on information search.
H8c. Income on has an effect on evaluating alternatives.
H9. Income has an effect on the behavioural biases.
H9a. Income has an effect on overconfdence bias.
H9b. Income has an effect on disposition effect.
H9c. Income has an effect on herding bias.
H10. Age has an effect on rational decision-making process.
H10a. Age has an effect on demand identifcation.
H10b. Age on has an effect on information search.
H10c. Age on has an effect on evaluating alternatives.
H11. Age has an effect on the behavioural biases.
H11a. Age has an effect on overconfdence bias.
H11b. Age on has an effect on disposition effect.
H11c. Age on has an effect on herding bias.
3.2 Questionnaire design
A survey questionnaire is used to identify the relationship between rational
decision-making process and behavioural biases in India. The questionnaire is divided
in to two parts. The frst part included the questions for the three-stage rational
decision-making process, i.e. demand identifcation, information search, evaluating
alternatives and three proposed behavioural biases, that is overconfdence, disposition
effect and herding. The other part included the questions for the respondent’s profle. In
the present study, 18 measure items are developed by referring to past research tools
(Lin, 2011). A six-point Likert scale with 1 set as “strongly disagree” and 6 set as a
QRFM
8,4
276
“strongly agree” is used for all questions in the questionnaire to collect the strength of
relationship between rational decision-making process and the behavioural biases in
investment decision-making.
3.3 Data collection and methods used
The population of the present study was the individual investors investing in the Indian
stock market. A pilot test with 70 respondents was done to evaluate the reliability and
the consistency of the questionnaire. The questionnaires were distributed through
personal contacts, e-mails and also through the brokerage houses in some cases. The
survey was conducted from May 2015 to October 2015 by using judgment and snowball
samplings. A total of 386 valid respondents has been collected after eliminating the
incomplete questionnaires. After data collection, data have been compiled and analysed
by using Statistical Package for the Social Science (SPSS) 20 and Amos 20 software. In
this study, cross-section analysis has been performed by using the structural equation
modelling (SEM). SEM constructs a comprehensive path and explores how the rational
decision-making process of investment and behavioural biases are related. In SEM,
relationships between theoretical constructs are represented by regression or path
coeffcients between the factors (Hox and Bechger, 1998). Further, to achieve the
research objectives, t-test, analysis of variance (ANOVA) and Fisher’s least signifcant
difference (LSD) test have been used in this study.
4. Analysis and results
4.1 Demographic profle
Table I summarises the respondents’ demographic characteristics, which indicates that
sample is composed of 266 males (68.9 per cent) and 120 females (31.1 per cent)
respondents. Ages 25-35 (50.3 per cent) account for the biggest portion of the sample,
followed by ages less than 25 years (26.9 per cent), ages 36-45 years (12.4), ages 46-55
years (6.2) and ages over 55 years (4.1 per cent). A total of 37.6 per cent respondents have
fnance-related occupations. Respondents with lower annual income (less than Rs 2
lakh) accounts for 24.1 per cent, followed by Rs 2 lakh-5 lakh (54.4 per cent), Rs 6 lakh-0
lakh (13.7 per cent) and less than Rs 10 lakh (7.8 per cent).
Table I.
Summary of
respondents
characteristics
(n 386)
Variable Investor grouping Frequency (%)
Gender Male 266 68.9
Female 120 31.1
Age (years) 25 104 26.9
25-35 194 50.3
36-45 48 12.4
46-55 24 6.2
55 16 4.1
Occupation Finance-related 145 37.6
Others 241 62.4
Annual income (Rs) 2 lakh 93 24.1
2-5 lakh 210 54.4
6-10 lakh 53 13.7
10 lakh 30 7.8
277
Rationality
and
behavioural
biases
4.2 Reliability and validity of constructed model
Table II represents the means, standard deviation and reliability statistics for the
constructs. The Kaiser–Meyer–Olkin measure of sample adequacy (KMO) was 0.76 and
satisfying the assumptions of exploratory factor analysis (EFA). The Cronbach’s alpha
for the three stages of rational decision-making process is 0.72, 0.81 and 0.76,
respectively, and the Cronbach’s alpha for the overconfdence, disposition effect and
herding is 0.74, 0.61 and 0.81, respectively, that indicates that the reliability of the study
is accepted (Hair et al., 2010).
To assess the constructs’ validity, confrmatory factor analysis (CFA) is performed
by using AMOS 20. CFA is used to test the hypothesis that a relationship between
observed variables and the latent variables exists (Suhr, 2006). For construct validity,
Hair et al. (2010) suggested that standardised factor loadings for each item should be at
least 0.5. Table II presents the standardised factor loading of individual item. In this
study, standardised factor loadings for all the items are more than 0.6 and the minimum
factor loading is 0.64 for the item O4. Moreover, Table II also presents the composite
reliability of the latent variables in the model. Bagozzi and Yi (1988) suggested that
value of composite reliability for the individual latent variable should be above 0.6. In
this study, the composite reliability for each latent variables is higher than 0.6. These
results show that constructs validity for the model is accepted.
4.3 Model ft
In the SEM, the model ft indices indicate that whether the model is acceptable. Bentler and
Hu (1991) suggested that value of 2/df should be between 5 to 3 for the model to be ft and
accurate. The overall model Chi-square is 372.19 with 108 degrees of freedom. The 2/df is
3.44that is closeto 3forthe goodft ofmodel. The overall value of GFI (goodness offtIndex)
is 0.91 and CFI (comparative fx index) is 0.93. Bentler (1990) recommended that the value of
CFI should be greater than 0.9 for goof ft of model. The value of RMSEA (root mean square
Table II.
Internal quality of
latent variables
Latent variables Item Mean SD Factor loading
Cronbach’s
alpha CR
Demand identifcation DI1 4.54 1.43 0.83
(DI) DI2 4.59 1.20 0.79 0.72 0.77
DI3 4.19 1.26 0.66
(IS) IS1 4.46 1.21 0.77
IS2 4.34 1.17 0.80 0.81 0.76
IS3 4.60 1.18 0.81
Evaluating alternative EA1 4.61 1.26 0.80 0.76 0.69
(EA) EA2 4.43 1.12 0.85
Overconfdence O1 3.89 1.43 0.73
(O) O2 3.75 1.42 0.81 0.74 0.83
O3 3.57 1.47 0.67
O4 4.08 1.53 0.64
Disposition effect D1 3.88 1.37 0.79 0.61 0.73
(DE) D2 3.73 1.26 0.85
Herding H1 3.33 1.40 0.87
(H) H2 3.84 1.25 0.80 0.81 0.78
H3 3.32 1.45 0.86
QRFM
8,4
278
error of approximation) is 0.07. The suggested value of RMSEA is between 0.05 and 0.08 for
an adequate ft (Browne and Cudeck, 1993; Hu and Bentler, 1995). These results prove that
model has an adequate ft for further investigation.
4.4 Results of structural equation modelling
Figure 2 presents the results of SEM by the path between latent variables. H1b and H2c are
strongly supported. However, H1a, H1c, H2a, H2b, H3a, H3b and H3c are not supported.
The regression weights indicate that demand identifcation have signifcant relation with
the information search (0.50, p-value0.05),meansifthevalue ofdemandidentifcation
increases by 1 unit, the information search also increases with 0.50 unit. The positive sign
indicates that the increase in demand also increases the search for information. Similarly,
information search also have a signifcant relationship with the evaluation of alternatives
( 0.51, p-value  0.05), implies that if the value of information search increases by 1 unit,
the evaluating alternative increases by 0.51 units. It indicates that different stages of the
Figure 2.
The relationship of
rational
decision-making
process and
behavioural biases
279
Rationality
and
behavioural
biases
decision-making process are associated in a sequential form. Additionally, analytical results
indicate the relationship between decision-making process and behavioural biases.
Regression weight indicates that only the second stage of the decision-making process, i.e.
information search, has signifcant relationship with the overconfdence bias ( 0.36,
p-value  0.05). Evaluating alternatives has signifcant impact on the disposition effect (
0.31, p-value  0.05). However, the stages of the decision-making process do not have any
signifcant impact on the herding bias.
4.5 Results of t-statistics
t-Test is used to analyse whether there are differences between respondents’
demographic profle (i.e. gender and occupation) and rational decision-making process
and, how those differences manifest themselves in the form of behavioural biases. For
the mean difference, 5 per cent level of signifcance is used. Table AI shows the results
of t-statistics for gender difference in rational decision-making process. It shows that
gender has a statistically signifcant difference during the information search and
evaluation of alternatives stages of rational decision-making and supports H4b and
H4c. Results further indicate that male investors search more information and evaluate
more alternatives than female counter parts. Table AII presents the results of t-statistics
on gender difference and behavioural biases. The results indicate that male investors are
more prone to overconfdence and herding bias than females and support H5a and H5c.
However, there is no signifcant difference among rational decision-making and
behavioural biases in terms occupation of investors and the results do not support H6a,
H6b, H6b, H7a, H7b and H7c.
4.6 Results of ANOVA and Fisher’s least signifcant difference
ANOVA is used to test whether there are differences between respondents’
demographic profle (i.e. income group and age) and rational decision-making process.
Additionally, Fisher’s LSD test is used for determining the strength of difference
between respondents5002 demographic profle. Table AIII presents the results of
ANOVA for respondents’ income difference and rational decision-making process. The
results show that F-statistics is 3.433, which is statistically signifcant at a confdence
level of 95 per cent and support H8a. It indicates that a signifcant difference exists in the
demand identifcation stage of rational decision-making with respect to income of
individual investors. However, no signifcant difference exists in information search and
evaluation of alternatives stages of rational decision-making process and H8b and H8c
are not supported. In addition to that, results of Fisher’s LSD test presented in Table AIV
show that two out of six pair-wise comparisons are signifcantly different at the 95 per
cent confdence level. Thus, both ANOVA and Fisher’s LSD tests show a distinctive
difference between demand identifcation and investors of different income group.
Table AV shows the results of ANOVA for respondents’ income difference and
behavioural biases. The results show that F-statistics is 3.357, which is statistically
signifcant at a confdence level of 95 per cent and support H9a. It indicates that a
signifcant difference exists in overconfdence bias with respect to income of individual
investors. However, no signifcant difference exists in disposition effect and herding
bias in terms of income of investors. Therefore, results do not support H9b and H9c. In
addition to that, results of Fisher’s LSD test shown in Table AVI show that two out of six
pair-wise comparisons are signifcantly different at the 95 per cent confdence level.
QRFM
8,4
280
Thus, both ANOVA and Fisher’s LSD tests show a distinctive difference between
overconfdence bias and income group of individual investors.
Age difference in terms of rational decision-making process is tested by using ANOVA
and results are presented in Table AVII. F-statistics is found to be insignifcant, that shows
no signifcant difference exists in rational decision-making process with respect to the
income of investors and H10a, H10b and H10c are not supported. Table AVIII shows the
results of ANOVA for respondents’ age difference and behavioural biases. The results show
that F-statistics is 5.083, which is statistically signifcant at a confdence level of 95 per cent
and support H11b. It indicates that a signifcant difference exists in disposition effect with
respect to age of individual investors. However, no signifcant difference exists in
overconfdence and herding biases in terms of age of investors. It indicates that H11a and
H11c are not supported. In addition to that, results of Fisher’s LSD test shown in Table AIX
show that four out of ten pair-wise comparisons are signifcantly different at the 95 per cent
confdence level. Thus, both the ANOVA and Fisher’s LSD tests show a distinctive
difference between disposition effect and age of individual investors.
5. Discussions and conclusion
The purpose of this study was to analyse the relationship between rational
decision-making process and behavioural biases of individual investors in India. India is
an emerging economy and investors are more prone to behavioural biases because of
lack of fnancial awareness. The results of the study indicate that investors follow a
rational decision-making process while investing. Empirical evidence provides that
there is a signifcant relationship between each stage of rational decision-making
process. First, investors identify their investment demand and analyse whether to invest
in fnancial product will increase their wealth. After the frst stage of decision-making,
investors move towards information search from various sources like suggestions from
their friends and relatives, information from newspapers or magazines and from their
past experiences. After collecting the information, investors evaluate all the available
alternatives and choose an option. Results confrmed the study hypothesis that rational
decision-making process has a signifcant relationship with the behavioural biases.
Results confrmed that the second stage of rational decision-making, i.e. information
search, has a positive relation with overconfdence bias. In other words, in the context of
the present study, this positive relation shows that once the investors identify demand
for the investment, they continue to search for the information. But based on the
availability of limited information and their past experience, investors become
overconfdent and behave irrationally. Literature also states that investors overreact to
the private information than public information (Daniel et al., 1998).
Furthermore, only the last stage of decision-making, i.e. evaluating alternatives,
signifcantly contributes to the disposition effect. From the fndings, it can be argued
that during the search for information, investors become overconfdent and rely on their
past experiences and limited information. Literature proposed that because of
overconfdence, investors get engaged in excessive trading (Odean, 1999) and often
leads to the wrong decision. Moreover, Sun and Hsiao (2006) argued that overconfdence
is positively and signifcantly associated with the disposition effect.
Additionally, fndings imply that each stage of the decision-making process has no
impact on herding bias. It implies that herding behaviour is not directly related to the
281
Rationality
and
behavioural
biases
decision-making process, but it could be related to some other factors like market
conditions etc. Such a result is also consistent with the fndings of Lin (2011).
Further, this study also examined the effect of investors’ background on the rational
decision-making process and those differences appear themselves in the form of behavioural
biases. Our results confrmed that gender is signifcantly different in terms of information
search and alternatives evaluation stages of rational decision-making process. Moreover, it
indicates that after the demand identifcation, male investors search more information from
newspapers, magazines and exchange the information with relatives and friends. But based
on limited information and private information, they become more confdent than females
and these results are consistent with previous studies (Barber and Odean, 2001; Bhandari
and Deaves, 2006; Lin, 2011). In case of herding bias, males tend to follow other investors, e.g.
friends and relatives, during the information search; however, these results contradict the
fndings of Lin (2011). Moreover, results indicate that income of investors is signifcantly
different in terms of the demand identifcation stage. It shows that investors with higher
income have more demand to invest in fnancial products that can help them to increase their
wealth.Similarly,thereisasignifcantdifferencebetweenoverconfdencebiasandincomeof
individual investors. Investors in the higher-income group are less confdent than investors
belonging to the low-income group. The results of this study show that age is signifcantly
indifferent with respect to rational decision-making process. However, that young and
middle-age investors (25-45 years) are prone to disposition effect than aged investors. It
clears that young age investors who are less experienced are often reluctant to realise losses
from their portfolio. Such a result is consistent with the previous study (Talpsepp, 2013).
However, occupation has no signifcant effect on rational decision-making process and
behavioural biases.
In conclusion, our results show that investors follow a rational decision-making
process but some psychological factors are also involved in their investment behaviour.
The present study is based on the cross-sectional data, so it can also be concluded that
investors’ behaviour can be changed according to the market conditions. For future
research, more psychological biases can be considered in different market conditions
and interrelationship between different psychological variables can also be analysed.
References
Bagozzi, R.P. and Yi, Y. (1988), “On the evaluation of structural equation models”, Journal of the
Academy of Marketing Science, Vol. 16 No. 1, pp. 74-94.
Barber, B.M. and Odean, T. (2000), “Investors, trading is hazardous to your wealth: the common stock
investment performance of individual”, The Journal of Finance, Vol. 55 No. 2, pp. 773-806.
Barber, B.M. and Odean, T. (2001), “Boys will be boys: gender, overconfdence, and common stock
investment”, The Quarterly Journal of Economics, Vol. 116 No. 1, pp. 261-292.
Barber, B.M., Odean, T. and Zheng, L. (2003), “Out of sight, out of mind: the effects of expenses on
mutual fund flows”, available at: http://papers.ssrn.com/sol3/papers.cfm?abstract_id496315
Barber, B.M., Lee, Y.T., Liu, Y.J. and Odean, T. (2007), “Is the aggregate investor reluctant to realise
losses? Evidence from Taiwan”, European Financial Management, Vol. 13 No. 3, pp. 423-447.
Bentler, P.M. (1990), “Comparative ft indexes in structural models”, Psychological Bulletin,
Vol. 107 No. 2, pp. 238-246.
Bentler, P.M. and Hu, L. (1991), “Fit indices in covariance structure modelling: sensitivity to under
parameterized model misspecifcation”, Psychological Methods, Vol. 3 No. 4, pp. 424-453.
QRFM
8,4
282
Bhandari, G. and Deaves, R. (2006), “The demographics of overconfdence”, The Journal of
Behavioral Finance, Vol. 7 No. 1, pp. 5-11.
Browne, M.W. and Cudeck, R. (1993), “Alternative ways of assessing model ft”, Sage Focus
Editions, Vol. 154, pp. 136-136.
Da Costa, N. Jr., Mineto, C. and Da Silva, S. (2008), “Disposition effect and gender”, Applied
Economics Letters, Vol. 15 No. 6, pp. 411-416.
Daniel, K., Hirshleifer, D. and Subrahmanyam, A. (1998), “Investor psychology and security
market under-and overreactions”, The Journal of Finance, Vol. 53 No. 6, pp. 1839-1885.
Dhar, R. and Zhu, N. (2006), “Up close and personal: investor sophistication and the disposition
effect”, Management Science, Vol. 52 No. 5, pp. 726-740.
Fama, E.F. (1970), “Effcient capital markets: a review of theory and empirical work”, The Journal
of Finance, Vol. 25 No. 2, pp. 383-417.
Fernandez, B., Garcia-Merino, T., Mayoral, R., Santos, V. and Vallelado, E. (2011), “Herding,
information uncertainty and investors’ cognitive profle”, Qualitative Research in Financial
Markets, Vol. 3 No. 1, pp. 7-33.
Goetzmann, W.N. and Massa, M. (2008), “Disposition matters: volume, volatility, and price impact
of a behavioral bias”, The Journal of Trading, Vol. 3 No. 2, pp. 68-90.
Grinblatt, M. and Keloharju, M. (2000), “The investment behavior and performance of various investor
types: a study of fnland’s unique data set”, Journal of Financial Economics, Vol. 55, pp. 43-67.
Grinblatt, M., Titman, S. and Wermers, R. (1995), “Momentum investment strategies, portfolio
performance, and herding: a study of mutual fund behavior”, The American Economic
Review, Vol. 85 No. 5, pp. 1088-1105.
Hair, J.F.Jr., Black, W.C., Babin, B.J. and Anderson, R.E. (2010), Multivariate Data Analysis, 7th ed.,
Prentice-Hall, Englewood Cliffs, NJ.
Hox, J.J. and Bechger, T.M. (1998), “An introduction to structural equation modelling”, Family
Science Review, Vol. 11, pp. 354-373.
Hu, L.T. and Bentler, P. (1995), “Evaluating model ft”, In Hoyle, R.H. (Ed.), Structural Equation
Modeling, Concepts, Issues, and Applications, Sage, London, pp. 76-99.
Jordan, D. and Diltz, J.D. (2004), “Day traders and the disposition effect”, The Journal of Behavioral
Finance, Vol. 5 No. 4, pp. 192-200.
Kahneman, D. and Tversky, A. (1979), “Prospect theory: an analysis of decision under risk”,
Econometrica: Journal of the Econometric Society, Vol. 47 No. 2, pp. 263-291.
Lakonishok, J., Shleifer, A. and Vishny, R.W. (1992), “The impact of institutional trading on stock
prices”, Journal of Financial Economics, Vol. 32 No. 1, pp. 23-43.
Lee, Y-T., Liu, Y-J., Roll, R. and Subrahmanyam, A. (2004), “Order imbalances and market
effciency: evidence from the Taiwan stock exchange”, Journal of Financial and
Quantitative Analysis, Vol. 39 No. 2, pp. 327-341.
Lin, H.W. (2011), “Elucidating rational investment decisions and behavioral biases: evidence from the
Taiwanese stock market”, African Journal of Business Management, Vol. 5 No. 5, pp. 1630-1641.
Mathuraswamy, P. and Rajendran, G. (2015), “Essence of rational investment in equity market: an
empirical study”, International Journal of Economic Research, Vol. 12 No. 2, pp. 439-449.
Mintzberg, H., Raisinghani, O. and Theoret, A. (1976), “The structure of unstructured decision
processes”, Administrative Science Quarterly, Vol. 21, pp. 246-275.
Nofsinger, J.R. and Sias, R.W. (1999), “Herding and feedback trading by institutional and
individual investors”, The Journal of Finance, Vol. 54 No. 6, pp. 2263-2295.
Nozick, R. (1993), The Nature of Rationality, Princeton University Press, Princeton.
283
Rationality
and
behavioural
biases
Odean, T. (1998), “Are investors reluctant to realize their losses?”, The Journal of Finance, Vol. 53
No. 5, pp. 1775-1798.
Odean, T. (1999), “Do investors trade too much?”, The American Economic Review, Vol. 89 No. 5,
pp. 1279-1299.
Pompian, M.M. and Wood, A.S. (2006), Behavioral Finance and Wealth Management: How to
Build Optimal Portfolios for Private Clients, John Wiley & Sons, Hoboken, NJ.
Schoenfeld, A.H. (2011), How We Think: A Theory of Goal-Oriented Decision Making and its
Educational Applications, Routledge, New York, NY.
Shafran, S., Benzion, U. and Shavit, T. (2009), “Investors’ decision to trade stocks-an
experimental study”, The Journal of Behavioral Finance, Vol. 10 No. 2, pp. 81-88.
Shapira, Z. and Venezia, I. (2001), “Patterns of behavior of professionally managed and
independent investors”, Journal of Banking & Finance, Vol. 25, pp. 1573-1587.
Shefrin, H. and Statman, M. (1985), “The disposition to sell winners too early and ride losers
too long: theory and evidence”, The Journal of Finance, Vol. 40 No. 3, pp. 777-790.
Sias, R.W. (2004), “Institutional herding”, Review of Financial Studies, Vol. 17 No. 1, pp. 165-206.
Simon, H.A. (1956), “Rational choice and the structure of the environment”, Psychological Review,
Vol. 63 No. 2, pp. 129-138.
Simon, H.A. (1982), Models of Bounded Rationality: Empirically Grounded Economic Reason, MIT
Press, Vol. 3.
Statman, M., Thorley, S. and Vorkink, K. (2006), “Investor overconfdence and trading volume”,
Review of Financial Studies, Vol. 19 No. 4, pp. 1531-1565.
Suhr, D.D. (2006), “Exploratory or confrmatory factor analysis?”, Proceedings of SUGI 31,
available at: www2.sas.com/ proceedings/sugi31/200-31.pdf
Sun, P. and Hsiao, S. (2006), The Influence of Investor Psychology on Disposition Effect,
Department of Business Management, Tatung University, Taiwan.
Talpsepp, T. (2013), “Does gender and age affect investor performance and the disposition effect?”,
Research in Economics and Business: Central and Eastern Europe, Vol. 2 No. 1, pp. 76-93.
Thaler, R. and Barberis, N. (2002), “A survey of behavioral fnance, national bureau of economic
research”, Working Paper 9222, available at: www.nber.org/papers/w9222.pdf
Trueman, B. (1994), “Analyst forecasts and herding behavior”, Review of Financial Studies, Vol. 7
No. 1, pp. 97-124.
Tversky, A. and Kahneman, D. (1974), “Judgment under uncertainty: heuristics and biases”,
Science, Vol. 185 No. 4157, pp. 1124-1131.
Weber, M. and Camerer, C. (1998), “The disposition effect in securities trading: an experimental
analysis”, Journal of Economic Behaviour and Organisation, Vol. 33 No. 2, pp. 167-184.
Wermers, R. (1999), “Mutual fund herding and the impact on stock prices”, The Journal of Finance,
Vol. 54 No. 2, pp. 581-622.
Zaidi, F.B. and Tauni, M.Z. (2012), “Influence of investor’s personality traits and demographics on
overconfdence bias”, Institute of Interdisciplinary Business Research, Vol. 4 No. 6, pp. 730-746.
QRFM
8,4
284
Appendix
Table AI.
t-Test for gender
difference in rational
decision-making
process
Item t df
Signifcance
(two-tailed) Mean difference
Standard error
difference
95%
confdence
interval of the
difference
Lower Upper
Demand identifcation 1.411 384 0.159 0.162 0.114 0.063 0.388
Information search 4.111 384 0.000 0.447 0.108 0.233 0.662
Evaluating alternative 2.858 384 0.04 0.332 0.116 0.103 0.561
Table AII.
t-Test for the gender
difference among
behavioural biases
Item t df
Signifcance
(two-tailed) Mean difference
Standard error
difference
95%
confdence
interval of the
difference
Lower Upper
Overconfdence 3.149 384 0.002 0.376 0.119 0.141 0.611
Disposition effect –1.581 384 0.115 –0.191 0.121 –0.429 0.046
Herding 3.149 384 0.002 0.376 0.119 0.141 0.611
Table AIII.
ANOVA for income
difference in rational
decision-making
process
Sum of squares df Mean square F Sig.
Demand identifcation

Note: *Signifcant at 95% confdence level
285
Rationality
and
behavioural
biases
Table AIV.
Signifcance levels
for income mean
differences of
demand
identifcation
(Fisher’s LSD)
2 lakh 2-5 lakh 6-10 lakh 10 lakh

Note: *Signifcant at 95% confdence level
Table AV.
ANOVA for income
difference among
behavioural biases
Sum of squares df Mean square F Sig.
Overconfdence

Note: *Signifcant at 95% confdence level
Table AVI.
Signifcance levels
for income mean
differences of
overconfdence bias
(Fisher’s LSD)
2 lakh 2-5 lakh 6-10 lakh 10 lakh

Note: *Signifcant at 95% confdence level
QRFM
8,4
286
Corresponding author
Satish Kumar can be contacted at: [email protected]
For instructions on how to order reprints of this article, please visit our website:
www.emeraldgrouppublishing.com/licensing/reprints.htm
Or contact us for further details: [email protected]
Table AVII.
ANOVA for age
difference in rational
decision-making
process
Sum of squares df Mean square F Sig.
Demand identifcation
Between groups 5.467 4 1.367 1.250 0.289
Within groups 416.512 381 1.093
Total 421.978 385
Information search
Between groups 3.443 4 0.861 0.841 0.500
Within groups 390.037 381 1.024
Total 393.481 385
Evaluating alternative
Between groups 2.392 4 0.598 0.521 0.720
Within groups 437.249 381 1.148
Total 439.641 385
Table AVIII.
ANOVA for income
difference among
behavioural biases
Sum of squares df Mean square F Sig.
Overconfdence
Between groups 10.577 4 2.644 2.216 0.067
Within groups 454.740 381 1.194
Total 465.318 385
Disposition effect
Between groups 23.706 4 5.927 5.083 0.001
Within groups 444.221 381 1.166
Total 467.927 385
Herding
Between groups 4.794 4 1.198 875 0.479
Within groups 521.921 381 1.370
Total 526.715 385
Table AIX.
Signifcance levels
for age mean
differences of
disposition effect
(Fisher’s LSD)
25 years 25-35 years 36-45 years 46-55 years 55 years

Note: *Signifcant at 95% confdence level
287
Rationality
and
behavioural
biases