Factors Influencing Individual Investor Behavior: The Case of the Karachi Stock Exchange
Chapter 1: Introduction
Substantial amount of attention has been given by researchers to the behavior and portfolio performance of institutional investors in the past whereas less attention has been given to the individual investor behavior (Venter, 2006; Prowse, 1990), (Nagy & Obenberger, 1994; Baker & Haslem, 1974). This study is an attempt to give insight into the behavior of individual investors i.e. which factors influence them to purchase stocks as did by (Nagy & Obenberger, 1994; Merikas, Andreas, George & Prasad, 2004; & Al-Tamimi, 2006).
Individual investors participate in the stock market by purchasing and selling different stocks and it is very important to identify various economic and behavioral motivations that affect their purchasing decisions. Thus it is important to identify the factors which have the greatest influence on the individual stock investor.
The individual investment decision in economic utility theory is viewed as a tradeoff between instant consumption and late consumption. The individual investor evaluates the benefits of consuming today against the benefits that would be gained by investing unconsumed funds in order to obtain greater consumption in the future. If the individual chooses to delay consumption he will select the portfolio that will maximize his enduring satisfaction. The essence of the utility theory axiomated by (Neumann & Morgenstern, 1947) state that investors are completely rational, deal with complex choices, are risk-averse and want to maximize their wealth. According to utility theory individual investors select the portfolio that increases their expected utility measured in expected return while decreases their risks or losses. There are other theories (Kahneman & Tversky, 1979; Tversky & Kahneman, 1986) which have made less strict suppositions about how investors make choices. These theories argued that investors look for investments that maximize geometric return, are focused on avoiding "terrible" outcomes and they make investment decisions free of assumptions about utility functions or probabilities e.g. stochastic dominance in which one outcome can be as superior to another (Nagy & Obenberger, 1994).
Get your grade
or your money back
using our Essay Writing Service!
According to (Nagy & Obenberger, 1994) the literature on economic utility theory does not cater to the individual investor's decisions. Instead it focuses on macroeconomic models that explain aggregate market behavior. Thus then in an economic utility theory the investment decisions are treated as a macroeconomic aggregate in which individuals capitalize on their utility by wealth maximizing criteria (Merikas, et al. 2004). But in a less than perfect world, investors are bounded in their rationality. They do not have all relevant information, unlimited cognitive and mathematical capacities, besides their knowledge and experience is also limited. (Hoffmann, Eije, & Jager, 2006)
Behavioral finance uses this body of knowledge, rather than ignoring these facts. Behavioral finance is the paradigm where financial markets are studied using models that are less narrow than those given by (Neumann & Morgenstern, 1947) expected utility theory and arbitrage assumptions. Behavioral finance is a response to the difficulties faced by the traditional models in financial markets which argues that some financial phenomena can be understood using models in which agents (individual investors) are not fully rational, either because of preferences or because of mistaken beliefs. Behavioral finance focuses on how investors translate and act on information to take investment decisions. It also examines the investor behavior which leads to various market abnormalities. It is a rapidly growing field which focuses on the effect of psychology on the behavior of financial practitioners. The sub discipline of behavioral finance included theories related with analyzing the attributes and attitudes of individual investors and exploring their choices under uncertain conditions. (Merikas, et al. 2004; Al-Tamimi, 2006)
This study aims at exploring Pakistani investor's behavior, representing the first attempt to be undertaken in Karachi, Pakistan. It will give an insight to individual local investors; investment professionals/planners and companies listed in Karachi stock exchange.
Understanding of behavioral processes of investors is essential for financial planners because an understanding of how investors generally respond to market movements will help investment advisors plan appropriate asset allocation strategies for their clients. Investment professionals who deal with retail clients may benefit by incorporating the factors which might turn out to be important when gauging and addressing individual investor concerns.
This study will also help companies in making their future policies and strategies to attract investors with focusing on those factors which influence them to invest. Besides, it would also give more support to market efficiency.
Always on Time
Marked to Standard
The above evidence presents the relative importance of various economic and behavioral motivations that affect investors purchasing decisions. Thus this research attempts to find out which factors influence their behavior while purchasing stocks using behavioral finance theory to decision making.
Chapter 2.1 discusses the background work of traditional finance, particularly, utility theory, its developments and its violations; Chapter 2.2 discusses the development of behavioral finance while Chapter 2.3 gives the literature review on the work done by previous researchers in the field of behavioral finance for individual investors. Research Methods is presented in Chapter 3, Results in Chapter 4 while Conclusions are given in Chapter 5.
Chapter 2.1: Traditional finance: Utility theory, Evolution & Criticism
The analysis of decision making under uncertainty has been dominated by expected utility theory. The expected utility principle was originated in the 18th century by (Bernoulli, 1738), it was first axiomated by (Neumann & Morgenstern, 1947) and it was further developed by (Savage, 1954) who included the notion of subjective probability into expected utility theory. The utility theory or expected utility theory has been used in economics as a descriptive theory to explain different phenomenon such as the purchase of insurance and the relationship between expenditure and saving. Besides, utility theory has also been used as a normative theory in decision analysis to determine best possible decisions and policies (Tversky, 1975).
(Tversky & Kahneman, 1986) There are four assumptions revealed by the axiomatic analysis of the expected theory, cancellation, transitivity, dominance, and invariance, besides more technical assumptions of comparability and continuity.
The basic assumption which is also the foundation block of expected utility theory is the cancellation or exclusion of any state of the world that gives the same result regardless of one's choice. This idea has been captured by different formal properties such as the substitution assumption of (Neumann & Morgenstern, 1947), the extended sure thing principle of (Savage, 1954) and the independence condition of (Luce & Krantz, 1971).
Transitivity of preference is another basic assumption in models of both risky and risky less choice. This condition is necessary and adequate for the illustration of preference by an ordinal utility scale U such that A is chosen over B whenever U (A) > U (B). Therefore transitivity is accepted only if it is possible to allocate a value to each option that does not depend on the other available options. Transitivity is true when the options are assessed independently but not when the consequences of an option depend on the other option to which it is compared for e.g. due to regret.
The third assumption called dominance states that if one option is better than another in one condition and at least as good in all other conditions, the dominant option should be selected. Stochastic dominance, a slightly stronger condition emphasizes that for one dimensional risky prospects, A is chosen over B if the cumulative distribution of A is to the right of the cumulative distribution of B. Dominance is both uncomplicated and more compelling than cancellation and transitivity, besides it also serves as the groundwork of the normative theory of choice.
The principle of invariance is an important condition for the normative theory of choice. Even though the representations of the same options would be different but it should yield the same preference or in other words, preference between options should be independent of their description. The decision maker on reflection of the different descriptions of alternatives should prefer the same option even without the benefit of such reflection. This principle of invariance or extensionality (Arrow, 1982) is so fundamental that it does not need to be stated as an axiom but is tacitly assumed during judgment. For example, decision models that describe the objects of choice as random variables all assume that alternative representations of the same random variables should be treated alike. In short, invariance states that variations of form that do not affect the actual outcomes should not affect the choice.
The four principles underlying expected utility theory can be arranged by their normative appeal starting from invariance and dominance which seem essential, transitivity which could be questioned and cancellation which has been rejected by previous researchers (Ellsberg, 1961). Most of the models use transitivity, dominance, and invariance e.g. (Hansson, 1975; Machina, 1982; Quiggin, 1982; Fishburn, 1983; Schmeidler, 1984; Luce & Narens, 1985; Segal, 1984). Other developments abandon transitivity but maintain invariance and dominance e.g. (Bell, 1982; Fishburn, 1982, Loomes & Sugden 1982). These theorists responded to observed violations of cancellation and transitivity by weakening the normative theory in order to keep its status as a descriptive model. However, this strategy cannot be extended to the failures of dominance and invariance because invariance and dominance are normatively essential and descriptively invalid; a theory of rational decision cannot provide an adequate description of choice behavior.
This Essay is
a Student's Work
This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.Examples of our work
The principle of invariance and dominance has been violated and the factors responsible for it are discussed by two descriptive examples. (McNeil, Pauker, Sox, & Tversky, 1982) First example is related to the study of choices among medical treatments in which information related to two treatments of lung cancer in terms of two frames, particularly, survival and mortality frame was given to the respondents. The respondents then indicated their preferred treatment.
In the a) survival frame in 1) Surgery treatment, out of hundred people who used surgery as a treatment, 90 people lived during the post operative phase, while 68 lived by the end of first year and 34 lived at the end of 5 years. While in 2) Radiation therapy, out of hundred people all lived through the treatment, 77 lived at the end of first year while 22 lived at the end of 5 years.
In the mortality frame in 1) Surgery, out of hundred people 10 died during surgery or through the post operative phase while 32 died at the end of the 1st year while 66 died at the end of 5 years. Under 2) Radiation therapy in mortality frame, out of 100 people, no one died during the treatment but 23 died at the end of first year and 78 died at the end of 5 years.
The unimportant difference in the example formulation showed a clear effect. The overall % of respondents who preferred radiation therapy grew from 18% in the survival frame, (N=247) to 44% in the mortality frame (N=336). The benefit of radiation therapy over surgery seems greater when it is given as a decrease in the risk of sudden death from 10% to 0% rather than as an increase from 90% to 100% in the survival rate.
In the next example decisions between combinations of risky prospects with monetary outcomes are evaluated. Each respondent made two choices, one between favorable prospects and one between unfavorable prospects (Tversky & Kahneman, 1981). It was assumed that the two selected prospects would be played separately.
Respondents examined two pairs of simultaneous decisions presented to them and then indicated the options preferred. Decision one included 1) a sure gain of $240, 2) 25% probability to gain $1000 and 75% probability to gain nothing, out of which respondents had to choose one. While decision two included 1) a sure loss of $750 and 2) 75% probability to lose $1000 and 25% probability to lose nothing, again out of which respondents had to pick one.
The percentage that chose each option in decision one are 1) 84%, 2) 16% while in decision two are 1) 13% and 2) 87%. The majority of choice in decision one is risk averse, while the majority of choice in decision two is risk seeking. This is a common pattern where choices involving gains are usually risk averse while choices concerning losses are usually risk seeking except when the probability of winning or losing is small (Fishburn & Kochenberger, 1979; Kahneman & Tversky, 1979; Hershey & Schoemaker, 1980). As the respondents considered both decisions simultaneously they preferred the portfolio A and D over B and C. Conversely, the preferred portfolio is in fact dominated by the rejected portfolio. The combined options for portfolio A & D were 25% chance to win $240 and 75% chance to lose $760 while B & C were 25% chance to win $250 and 75% chance to lose $750. This clearly explains that when the options are presented in the aggregate form, the dominant option is chosen.
In problem two, 73% respondents chose the combination A and D which was dominated, and option B and C was chosen by only 3%. The difference among the two formats demonstrates the violation of invariance. These findings also support that the failures of invariance also generate violations to stochastic dominance and vice versa. Clearly the respondents took the decisions separately in problem 2 where they displayed the typical pattern of risk seeking in losses and risk aversion in gains. The respondents were astonished to find out that the combination of two preferences that they thought were quite sensible led them to choose a dominated option.
Thus it can be said that variations in the framing of decision problems generate systematic violations of invariance and dominance which cannot be defended on the basis of normative model.
The principle of invariance would hold if all formulations of the same view were transformed to a standard canonical representation because then the different versions would all be assessed in the same way. For example in problem two invariance and dominance would both hold if the results of the two decisions were combined prior to evaluation. In the same way, the same choice would be made in both frames of the medical problem if the results were given in one dominant frame e.g. rate of survival. The above observed failures of invariance indicate that people do not aggregate simultaneous prospects or convert all results into a common frame. Thus normative models of choice which use invariance cannot provide a sufficient descriptive explanation of choice behavior. (Tversky & Kahneman, 1986)
(Tversky, 1975) To conclude, there is significantly less conformity concerning the descriptive validity of the above mentioned axioms. The experimental analysis of utility theory does not produce clear results. In quite a few studies most axioms of utility theory are violated for example, circumstances under which transitivity is violated are described in (Tversky, 1969; Raiffa, 1968). However, other experimental studies support utility theory, e.g. (Mosteller & Nogee, 1951; Tversky, 1967). Thus it remains unclear that utility theory provides a sound estimate to the behavior of individuals under uncertainty or not
(Tversky & Kahneman, 1986) Different models of risky choice used to explain the observed violations of utility theory have been developed in past ten years.
Table 1 below gives a summary of the assumptions of utility theory, its empirical violations and explanatory models. Column one represents the assumptions of utility theory, column two represents their empirical violations and column three presents the models which explains these violations.
All the models are consistent with the violations of cancellation created by the certainty effect. Besides, bivariate (non transitive) models are required to explain observed in transitivities and only prospect theory hold the observed violations of (stochastic) dominance and invariance.
Table 1: Summary of Empirical Violations and Explanatory Models
Summary of Empirical Violations and Explanatory Models
Certainty effect (Kahneman and All models Tversky 1979)
Lexicographic semi order (Tversky 1969)
Preference reversals (Slovic and Lichtenstein 1983)
Contrasting risk attitudes (Problem 2)
Framing effects (Problem 1)
Chapter 2.2: Behavioral finance
The essence of the utility theory axiomated by (Neumann & Morgenstern, 1947) state that investors are completely rational, deal with complex choices, are risk-averse and want to maximize their wealth. Utility theory also assumes that investors select the portfolio that maximizes expected utility measured in expected return while minimizing risks or losses (Nagy & Obenberger, 1994). Efficient market hypothesis, one of the holistic theories of traditional finance also states that humans are rational when making their decisions while behavioral finance attempts to find out and understand psychological decision processes for financial markets. The main approach to behavioral finance is that investors are not rational and are under influence. Therefore it focuses on the need for a new approach which does not ignore the decision processes of investors.
A new financial sub discipline called behavioral finance has ignited a wave in explaining the behavioral aspects of investment decisions. It examines choice under uncertainty. Behavioral finance takes into account the research that eliminates the traditional assumptions of expected utility maximization with rational investors in efficient markets. Rationality means that agents (individual investors) update their beliefs correctly when they receive new information and they take normative decisions consistent with the notion of subjective expected utility (SEU) but it is difficult to understand the facts about the aggregate stock market and individual trading behavior. Thus behavioral finance is a response to the difficulties faced by the traditional models in financial markets which argues that some financial phenomena can be understood using models in which agents (individual investors) are not fully rational, either because of preferences or because of mistaken beliefs.
It is assumed in behavioral finance that information structure and characteristics of market participants influence individual's investment decisions in addition to market outcomes. According to behavioral finance, investor market behavior is obtained from psychological principles of decision making which explains why do people buy or sell stocks. (Al-Tamimi, 2006)
(Al-Tamimi, 2006) Behavioral finance focuses on how investors translate and act on information to take investment decisions. It also examines the investor behavior which leads to various market abnormalities or irregularities (anomalies). It is a rapidly growing field which focuses on the effect of psychology on the behavior of financial practitioners. Behavioral finance gives light to issues related to investors who are able to beat the market even when evidence shows they cannot, why is the stock market volatile and volume of trading in financial markets excessive, why do investment analysts have difficulty in identifying under and overvalued stocks and why does bad news under react stock prices.
Behavioral finance is the paradigm where financial markets are studied using models that are less narrow than those given by Neumann & Morgenstern expected utility theory and arbitrage assumptions. The two building blocks on which behavioral finance resides are cognitive psychology and the limits to arbitrage. Cognitive means the way people think and that people make systematic errors in the manner in which they think for e.g. they are overconfident and put too much weight on recent experiences, etc. Their preferences also create misrepresentations. Behavioral finance uses this body of knowledge, rather than ignoring these facts. Limits to arbitrage refer to predicting in what circumstances arbitrage forces will be effective, and when they won't be.
The sub discipline of behavioral finance included theories related with analyzing the attributes and attitudes of individual investors and exploring their choices under uncertain conditions; namely Prospect Theory, Heuristics and Regret Aversion (Merikas et al. 2004).
The groundwork of prospect theory can be dated back to 1970's when psychologists Daniel Kahneman and Amos Tversky investigated how decision making heuristics (rules of thumb) bring biases in individuals behavior, therefore identifying whether the principle of idealized rationality is a useful approximation or not. The most significant part of their work, called prospect theory, summarized results in a way that can be seen as an alternative to utility theory (Kahneman & Tversky, 1979). Prospect theory explains how individuals make decisions under risky circumstances. It is an effort to express several of the principles of perception and judgment which restrict the rationality of choice. In recent times, studies related to individual choice behavior has given attention to not only choices concerning risky alternatives but also choices which are made over time, with studies focusing on strong irrational behaviors which lead to complexities of procrastination or self control, like, increase in savings and consumption behavior (Laibson, 1997; Loewenstein & Elster, 1992 and Thaler, 1987)
Prospect theory is important because it identifies the deficiencies and violations of traditional finance and the way it explains the behavior of individuals toward risk on the basis of “rational human being”. The rational investor refrains from risk but will take risks in exchange of more returns while prospect theory states that decision making process is not completely a rational one and individuals are more inclined to take risks to avoid losses rather than to get high returns. (Sevil, Sen, & Yalama, 2007) found that investors are far from being rational and support the risk seeking tendency of investors with prospect theory's frame work.
Individuals have the tendency of making quick judgments are called heuristics. These are simple strategies to deal with complex problems and limit the explanatory information. People are inclined to give higher probabilities to events which they are familiar with. (Sevil, et al, 2007) found that the greater companies investors know better are less risky than the small companies. Representative heuristics means that as the investors learn from past price movements their future expectations will be similar to their past experiences and they will not evaluate current information in its own situation.
Regret is an emotional condition which is related to information about the past pertaining to a decision in the past which lead to a worst outcome than an alternative decision or than a decision of someone else. (Statman, 1999) regret is also defined as the frustration which occurs due to the consequence of bad choice. The opposite of regret is gratification in positive; both regret and gratification are important. The desire for gratification and aversion of regret results in the realization of profits and retardation of losses. (Sevil, et al, 2007) found that the joy of gratification is not equal to the pain of regret. The pain of regret is greater than the joy of gratification thus investors avoid the pain of regret would decrease their personal responsibility in their investment decisions.
Chapter 2.3: Literature Review
Empirical studies of the behavior of individual investors first appeared in the 1970s, (Lease, Lewellen, & Schlarbaum, 1974) determined demographic characteristics, investment strategy patterns, information sources, asset holdings, market attitudes and perceptions, of the individual investor. They found that securities ownership was heavily concentrated within the upper age bracket and income segments of the community and those individual investors were male, married and had annual income of $25,000. They worked in professional and managerial occupations, and more than half achieved their bachelor's degree. Their investment strategy patterns asserted that they were fundamental analysts who perceived to hold a balanced and well diversified portfolio of income and capital appreciation securities. They claimed that they invested primarily for the long run and were prone to use market indexes as the benchmark by which they judged their personal investment performance results. Half the sample spent less than five hours a month and less than $15 a year on collecting information for and making the decisions about the securities in their portfolios, thus, conveyed an image of low involvement and low monitoring. They considered private messages they got from their account executives, and the public ones from their journals and newspapers. Their portfolio amounts suggested predominance of equity capital under personal administration. This group's passion for direct market participation was due to fun as well as profit. The respondents claimed to enjoy investing and felt that they would sacrifice their pleasure if they let institutions administer their funds, and they also planned on a larger relative commitment to equity securities under personal management in the future.
(Baker & Haslem, 1974) studied the factors that cause investors to differ in their opinions of the attractiveness of specific common stocks and whether these factors are related to their socioeconomic and behavioral characteristics. Their survey contained thirty-four decision variables measured on a five point importance scale and an investor profile containing variables of a socioeconomic and behavioral nature. Factor analysis and stepwise multiple regression was applied, factor analysis revealed three significant factors which were Dividends, Future Expectations and Financial Stability and they accounted for 23.7, 9.6, and 3.4% of the total sample variance. These 3 factors have the greatest differences in perceptions of importance among the investors. Under multiple regression the relationships between 21 socioeconomic and behavioral variables surveyed and the three scored factors were examined. These findings suggested that investors were of two distinct types, one who seek dividends and the others who seek capital appreciation. Investors concerned with income from dividends use decision variables represented by Factor 1 and Factor 3 i.e. Dividends and Financial Stability. Investors who gave importance to dividends were older, females, and risk averse and thus did not seek a large increase in the value of their stock. While the second type of investors who were concerned with capital appreciation used the decision variables represented by Factor 2, Future Expectations and thus were willing to sacrifice current dividends for future price appreciation.
(Cohn, Lewellen, Lease & Schlarbaum, 1975) suggests a strong pattern of decreasing relative risk aversion as investor wealth increases.
(Winsen, 1976) Efficient Market Hypothesis (EMH) asserts that stock market prices reflect all publicly available information so that it is impossible to consistently attain abnormal returns using such information. That is, prices adjust quickly and adequately to new information. According to the hypothesis, fluctuations in a certain stock price can be taken as indications of the flow of publicly available information into the stock market related to that specific stock. (Winsen, 1976) studied whether investor behavior is associated with such a flow of information or not. A linear model was formed to study the relationship between investor behavior and fluctuations in the stock price of a certain firm. The relationship was investigated between a) the error term representing behavior not adequately related to information flow and b) publicly available data about the firm which included four data items, particularly, earnings per share, cash flow per share, dividends per share, and available for common per share. An association was identified which varied across publicly available data and across firms. The findings supported the argument that investors in some firms misunderstand and/or misuse certain publicly available data items which results in their behavior not being an adequate function of the flow of information coming in the stock market.
(Falk & Matulich, 1976) They examined the relationship between some personal characteristics of a group of investors and a group of investment advisors, and the degree of risk attributed by them to various types of financial investments. Personal characteristics investigated included the investors' age, education, investment experience, degree of concern for others, and the extent to which they use financial journals. For both groups investigated, the degree of concern for others was significantly related to the degree of risk attributed to investments. Age and experience generally are not related to the degree of risk attributed to investments and thus no difference exists between young and old advisors and investors in the degree of risk attributed to various types of investments. The correlation analysis implies that risk attributed to investments does not increase with education. Investment advisors seem to place more reliance on financial journals than do investors, particularly when dealing with unsecured investments. In the group of investors, a high correlation exists between the respondents' degree of concern for others and the risk they attribute to an investment. The correlation is strongest for unquoted unsecured investments and weakens as one move to secured investments. In the advisors group, the greater the degree of concern the more highly is the correlation with negotiability of investments, with security a secondary consideration. The advisors' degree of concern for others is uncorrelated with the linkage of an investment and negotiability is more important to those who need investment advice than to those who can make their own investment decisions.
(Baker, Hargrove, & Haslem, 1977) found that investors behave rationally, taking into account the investment's risk/return tradeoff. There is a common dilemma faced by the investors and portfolio managers regarding the tradeoff preference between risk and return. There is an agreement that generally a positive relationship exists between risk and expected return. Previously studies have been conducted on ex post risk return relationships of portfolio managed by institutional or professional investors. His study analyzes the ex ante risk return preferences and expectations of individual common stock investors. The purpose of his study was to examine the relationship between acceptable risk levels and expected annual rates of return as well as examine the nature of this relationship between risk and the components of total return i.e. income from dividends and capital appreciation. Chi square and Somer's D test (+0.13) indicated a strong positive relationship between expected total return and acceptable level of risk. When assessing individually (Baker, et al, 1977) found that extremely strong relationship exists between acceptable levels of risk and dividend income indicated by chi square statistic while Somer's D statistic (-0.226) concludes that a strong but negative association exists between risk and dividends. It means that lower risk investors consider high dividend paying stocks which are less risky than those growth oriented stocks paying small dividends. Also, there is a strong relationship between acceptable levels of risk and capital appreciation based on large chi square test. The Somer's D test revealed that this strong relationship is positive (+0.255) means that lower risk investors have relative tendency to seek lower capital appreciation while there is stronger tendency for higher risk investors to seek higher capital appreciation.
Thus to sum it up the Somer's D statistic for the risk-total return is positive but smaller than the statistic for risk-capital appreciation relationship. It has been reduced by the negative Somer's D statistic for the risk-dividends relationship. As dividends and capital appreciation together sums to total return therefore the presence of a positive risk-total returns relationship even after negative risk-dividends means that the positive association between risk and expected return appears to be due to the impact of capital appreciation in investor expectations of total return. Nonetheless it gives a clientele effect which means that lower risk investors seek high dividends while higher risk investors seek higher capital appreciation in growth stocks.
(Lewellen, Lease, & Schlarbaum, 1977) gives importance to individual investor because of the individual investor's withdrawal from the direct participation from the American equity market even though which has been sufficiently documented but the causes of that observable fact remain almost entirely hypothetical. Almost all the individual investor's situations and decision processes has been inferred from broad trading statistics, general security price movements, or portfolio simulations. But it is important to find out the reasons to the withdrawal of individual investor's participation from the market so that remedial steps could be taken rapidly to counter act before his withdrawal from the market. Strong indications of systematic changes in investment objectives and risk preferences across age brackets and, to milder extent, income classes are apparent. These are mirrored in differences in investment tactics, portfolio composition, and environmental attitudes. Though analytical styles are diverse, especially between the genders, the ultimate goals and resulting decisions have an underlying harmony. Therefore, age, sex, income and education affect investor preferences for capital gains, dividend yield and overall return.
(Barnewell, 1987) found that individual investor behavior can be anticipated by lifestyle characteristics, occupation, risk aversion and control orientation. Barnwell characterized individual investors as belonging to either two extremes- active or passive in her lifestyle analysis.
(LeBaron, Farrelly, & Gula, 1992) argued that individuals' risk aversion is largely a function of intuition rather than rational considerations.
According to (Warren, Stevens, & McConkey, 1990) demographics are always used to segment the market for financial and economic services but lifestyle characteristics can further help in identifying individual investor's financial needs more precisely. Investors who are in the same income categories or age may have entirely different investment needs, which can be more fully analyzed with the help of lifestyle analysis. Lifestyle dimensions do not only help differentiate between investor behavior types (active/passive), they may also be useful in differentiating between light and heavy investors in particular investments (i.e., stocks and bonds). The analysis of the lifestyle dimensions revealed that two of the lifestyle characteristics were associated with the level of concentration in stocks and bonds. Respondents who had a light concentration of their investments in stocks and bonds could be described as volunteers and as dress- conscious. The tendency of heavy stock/bond investors not to get involved in community organizations and volunteer work may make them less accessible to the financial services marketer.
(Riley & Chow, 1992) found that as wealth, income, education increases risk-aversion decreases and it also decreases with age but only up to a certain point. After 65 of age i.e. retirement, risk aversion maximizes with age.
( Nagy & Obenberger, 1994) examined the factors that have the greatest influence on the individual stock investor. 34 variables were evaluated based on their importance which was identified by extensive testing as potentially influencing equity investment decisions. The variables included some from the traditional sphere (e.g. expected dividends, perceived risk, diversification needs) while others addressed to more contemporary concerns (firm's environmental record, perceived firm ethics, etc.). Their findings suggested that classical wealth maximization criteria are important to investors, even though investors employ diverse criteria when choosing stocks. Contemporary concerns such as local or international operations, environmental track record and the firm's ethical posture appear to be given only cursory consideration. The recommendations of brokerage houses, individual stock brokers, family members and coworkers go largely unheeded. Many individual investors discount the benefits of valuation models when evaluating stocks. Seven relatively homogenous groups of variables were formed that influence individual investor behavior which were neutral information, self image / firm image coincidence, classic, social relevance, accounting information, advocate recommendation, and personal financial needs. Thus one can say that investment decision process appears to incorporate a broader range of items than previously assumed.
The Wharton Survey (Bodnar, Hayt, & Marston, 1996) which is one of the most comprehensive studies of investor behavior examines how demographic variables influence the investment selection and portfolio composition process.
(Riley & Russon, 1995) Asset allocation is dependent upon expected capital market returns and the individual client's desire and ability to tolerate risk. (Riley & Russon, 1995) determined the risk tolerance of the client by making the client risk tolerance to be a function of time horizon, salary, expected salary growth, age, gender, marital status, and number of children. The relationship between perceived client risk tolerance and horizon, client salary, and projected salary growth were all positive. The longer the planning horizon, the greater the client's salary, the higher the projected salary growth, the more risk the client should be able to tolerate. They used Nonlinear Estimation Regression which showed that the explanatory power of the model is high with over 89 percent of the variation in the perceived risk being explained by the independent variables. The results clearly indicate that the assumed time horizon for a client appears to have the greatest explanatory power of risk tolerance and is an essential element for a proper asset allocation while variables such as age, marital status, gender, and number of children were observable and should be further investigated.
(Merikas, et al. 2004) studied the factors that appear to exercise the greatest influence on the individual stock investor in the Greek stock exchange. Participants were asked to evaluate the importance of 26 variables, identified from the literature and personal interviews as potentially influencing decision variables. The most important variables were related to classic wealth maximization criteria such as “expected corporate earnings”, “condition of financial statements”, or “firm status in the industry”. Environmental criteria like “coverage in the press”, “statements from politicians and government officials”, “ease of obtaining borrowed funds” and “political party affiliation” were totally unimportant to most stock investors and they are self-reliant ignoring inputs of family members, politicians, and coworkers when purchasing stocks. Factor analysis was used to categorize the similar variables based on correlations and group them into identifiable categories. Varimax algorithm of orthogonal rotation identified factor categories as Accounting Information, Personal Financial Needs Subjective/Personal, Advocate Recommendation, and Neutral Information. Accounting information includes variables from wealth maximizing criteria such as expected corporate earnings and condition of financial statements.
While Subjective/Personal factor includes feelings for a firm's products & services variable having the highest loading, thus being the most important variable and Neutral Information include current economic indicators and recent price movements in a firms stock as most important variable representing this factor. Advocate Recommendation having opinions of the firm's majority stockholders as most important variable and finally attractiveness of non-stock investments as the most important variable due to highest factor loading in Personal Financial Needs factor.
(Boye, 2005) studied the extent to which some selected dimensions of culture influenced the investment decisions of institutional and individual investors based on frequency distribution.
Cultural factors such as uncertainty avoidance, time conception, cultural logic, decision rules and oral tradition identifies the extent to which investors invest their funds. Besides choices of banks, stock investment e.g. decisions to invest in Treasury bill or fixed property is mostly influenced by friends.
(Hoffmann, et al, 2006) used theories of needs and conformity behavior on investors in their research in Netherlands. The results indicated that investors besides satisfying the financial needs also strive to satisfy other more socially oriented needs. Investors that report less investment related knowledge and experience display more informational and normative conformity behavior while investors rating social needs as more important display more informational and normative conformity behavior. Besides, the effect of socially oriented needs is larger than the effect of investment-related knowledge and experience and socially oriented needs are a stronger predictor of normative conformity behavior than investment-related knowledge and experience of the investor. They also found that even though individual investors give importance to the need for financial gain but they also give importance to social interaction with other investors, and therefore enjoy investing as a free-time activity. Thus, this study follows an “extended” utility approach which supports the claims from recent behavioral finance literature, which states that investing offers both utilitarian and expressive benefits and therefore investors do not only care about financial aspects of investing, but rather display a palette of different needs. (Fisher & Statman, 1997; Statman, 1999; Statman, 2002; Statman, 2004)
(Al-Tamimi, 2006) researched factors which influence the UAE investor behavior on the Dubai Financial Market and Abu Dhabi Securities Market. They first used regression and then factor analysis. Demand on common stocks was the dependent variable represented by past performance of the firm's stock and recent price movement in a firm's stock while independent variables were religious factor, market share and reputation of the firm, accounting information, publicly available information, advocate recommendation and personal financial needs. These independent variables were represented by the 25 variables asked in the form of questions in the questionnaire. Their regression model was insignificant because it only explained 20% of the variation in the dependent variable explained by the independent variable. By making the frequency table they identified the most influencing factors were past performance of the firm's stock, expected corporate earnings, government holdings, stock marketability, get rich quick, and the creation of financial markets (i.e. Dubai Financial Market and Abu Dhabi Securities Markets). The least influencing factor were expected losses in international financial markets, expected losses in other local investments, family member opinions, minimizing risk, and gut feeling on the economy. Factor analysis made 5 factors: neutral information, accounting information, advocate recommendation, self-image / firm-image coincidence, and personal financial needs.
(Sevil, et al, 2007) aimed at understanding the decision processes of small investors trading in Istanbul stock exchange and found out that investors are not completely rational as perceived by traditional finance theories.
Chapter 3: Research Methods
Specifically, two research questions have been addressed in this research.
* First, what relative importance do decision variables have for individual investors making stock purchase decisions?
* Second, are there homogeneous groups of variables that form identifiable constructs that investors rely upon when making equity investment decisions?
In order to answer the above research questions, 30 variables which has been previously used by (Nagy & Obenberger, 1994 and Al-Tamimi, 2006) given in appendix, table 2, were used according to their significance in the Pakistani market, particularly in Karachi. These variables included few from the traditional sphere i.e. from utility theory or wealth maximization criteria e.g., expected dividends, expected corporate earnings, perceived risk, diversification needs while others addressed more modern concerns such as firm's environmental record, perceived firm ethics, etc along with few variables concerned with financial information such as Condition of Financial Statements and Recent Price Movements of Firm's Stock. Besides these 30 items, respondents were also asked about their demographics which included age, income, marital status, gender, education, field and designation. Respondents were also asked about an additional variable which was to identify their common stock holdings in terms of their rupee value.
The study aimed at analyzing the behavior of individual investors/shareholders in Karachi Stock Exchange using these 30 variables. The information was gathered from individual investors who purchase and sell stocks in the Karachi Stock Exchange. The Karachi Stock Exchange also known as KSE is the oldest exchange in Pakistan which was established in 1947 and after few years it became a registered company limited. It has experienced an outstanding growth with only 5 companies listed and 90 members in 1950s while 663 listed companies and 200 members in 2006. KSE is the most liquid and biggest exchange in Pakistan with market capitalization of US $ 54.28 billion and an average daily turnover of 525.15 million shares. KSE was recognized as the top performing world stock market in 2002 by international magazine 'Business Week'.
There are around 200 brokerage houses in Karachi, out of which 142 are active while the rest are inactive. Each active brokerage house has at least 400 individual investors thus giving a total population of around 57000 (142*400). The variables were used to identify important variables which influence individual investors when making stock purchase decisions and whether these variables can be grouped in homogenous sets that form identifiable constructs on which they rely when making equity investment decisions.
Convenience based sampling was the technique used in this research in which respondents were selected based on convenience. It was a primary research thus data was collected through a questionnaire. In order to get responses on the research questions 153 questionnaires were distributed to individual investors who invested in Karachi stock exchange and the response rate was 100%.
Participants were asked to evaluate the importance of 30 variables which provided as potentially influencing on their equity investment decisions. Respondents noted whether each variable was (1) A important item used to make investment decisions ("Act On"), (2) A secondary item ("Consider") or (3) An item ignored in the investment decision process ("No Influence").
In order to test the reliability of the instrument used, Cronbach Alpha was applied. Cronbach alpha measures the reliability of the different categories and consists of estimates of how much variation in scores of different variables is due to chance or random errors (Al-Tamimi, 2006). A coefficient greater than or equal to 0.5 is acceptable and a good indication of construct reliability. Table 3.1 represents Cronbach alpha's result for 33 variables altogether which also included age, income and their common stock holdings, all measured on ordinal scale. Table 3.2 in appendix represents the Cronbach Alpha values for individual variables. The sample size chosen for the reliability test was 40. The overall significance level of Cronbach's alpha came out to be 0.761, thus reliability is 76% and the instrument was reliable to be used further in the study. It could have been further improved by removing some variables but these variables were important for research findings so all the variables were held at a reliability of 0.761.
N of Items
The variables were ranked according to how frequently they were found in each response category and used factor analysis to examine how they interacted. Factor analysis technique was used to determine whether there were underlying constructs that represented a combination of investor concerns. Factor analysis takes large numbers of variables (30 items in this study) and identifies similarities between them making explanation of the results easier. The sample size condition of factor analysis is that there should be at least 5 responses per variable for factor analysis to produce sufficient results, thus 30*5 gives the total of 150 responses which was an appropriate sample size. Besides, all variables should be metric, therefore all the variables included in the questionnaire was based on ordinal scale.
Chapter 4: Results
In response to the first research question of this study, significant variables based on their frequency distributions were identified which influence individual investor's behavior while making stock purchase decisions. Table 4 in appendix lists 30 variables by their frequencies with which respondents considered them to have significant influence on their stock purchase decisions. Some observations made were that most of the variables ranked significant were Classic wealth-maximization criteria such as Expected Dividends, Expected Corporate Earnings and Diversification Needs. Besides another criterion which was significant included the Performance of Stocks such as Expected Stock Market Performance, Recent Price Movements of Firm's Stock, Past Performance of Investor's Stock Portfolio, Current financial position, Condition of Financial Statements and Past Performance of Stock. Lastly, the sample respondents were more self reliant when considering which stocks to choose and ignore family members and friends/coworkers opinions but consider stock broker advices. This confirms that investors employ different decision criterion when selecting stocks. It is also evident that modern concerns i.e. Social Relevance & Image such as Perceived Ethics of Firm, Environmental track record, Local Operations and International operations, are given only minimum consideration by stock investors.
Table 5, in appendix, ranks the frequency distribution of variables that respondents ignore or in other words that least influence the investor's behavior. First, Social Relevance & image is apparently not important to stock investors which include Environmental Record, International Operations, Perceived Ethics of Firm and Local Operations. Second, they ignore inputs from family members and friends/coworkers when selecting stocks. While Data in Reports/Prospectuses and Exchange listings of companies were given only cursory considerations. The recommendations of stock brokers considered by respondents but lesser than their self reliance.
As it is difficult to identify which variables are related to wealth-maximization criteria (economic utility theory) and which are not related, but few observations can be made that it is evident that investors rely mostly on decision criteria predicted by classic economic utility theory. However, at the same time, it is also clear that investors use diverse criteria, rather than a single approach. Many respondents do not take investment decisions in a traditional manner.
The second focus of this research was to identify whether the variables most important to investors form homogenous groups or not. As there is no single set of variables investors used consistently to make stock purchase decisions, factor analysis was applied to determine whether there are underlying constructs that signify a combination of investor concerns. Varimax Algorithm of Orthogonal Rotation was used to analyze 34 variables. Evaluation of the resulting constructs is mainly subjective. The labeling of the variables and the empirical factor formation and identification are rarely perfect, thus endurance is encouraged. The underlying assumptions to run factor analysis is that KMO (Kaiser-Meyer-Olkin Measure of Sampling Adequacy) value should be greater than 0.5 and Bartlett test of sphericity should be rejected (i.e.) its significance value should be less than 0.05. Bartlett test of sphericity tests the null hypothesis that the correlation matrix (based on which factor analysis makes factors) is an identity matrix means there is no correlation or multicollinearity among variables.
Two variables were removed from factor analysis because their Anti Image values (Measure of Sampling Adequacy; an extension of KMO, which gives partial correlations of individual variables) were less than 0.5 (i.e. 0.410 and 0.430) which should have been equal to or more than 0.5. These variables were You/Yourself and Family Member Opinions. Removing these variables also improved KMO (Kaiser-Meyer-Olkin Measure of Sampling Adequacy) from 0.769 to 0.785 and Bartlett test of sphericity is rejected which means that the correlation matrix is not an identity matrix and thus there is an underlying structure among the variables. Table 6 shows the KMO and Bartlett values.
Table 6: Assumptions of factor Analysis: KMO & Bartlett tests.
KMO and Bartlett's Test
Kaiser-Meyer-Olkin Measure of Sampling Adequacy.
Bartlett's Test of Sphericity
Table no. 7, in appendix, represents Communalities which identify the most important variable among the set of 30 variables based on their extractions. The most important variables are Past Performance of Investors Stock Portfolio with 0.770, Past Performance of Stock with 0.702 and Stock Broker Recommendation with 0.702 extractions.
Seven factors/components were extracted based on “Principle Components” extraction method and threshold of Eigen value 1. Table 8 in appendix, represents the total variance explained by seven factors extracted. In the column of Extraction Sums of Squared Loadings, the first component being the most important component explains 22.29% of variation in the data with 6.242 Eigen value. The second component explains 10.185 % variation with Eigen value of 2.852, third component explains 7.664% variation with 2.146 Eigen value. While fourth component explaining 6.864% of variation with 1.922 Eigen value, fifth component explaining 4.911% variation in the data with Eigen value of 1.375, sixth component with 4.636% variation and 1.298 Eigen value and last component explaining 3.794% variation and 1.062 Eigen value.
Extraction Sums of Squared Loadings provides the initial loadings in each component with highest factor loading in first component along with highest Eigen value and variation explained by the component and then the rest of the Eigen value and variation in ascending order in other components. While Rotation Sums of Squared Loadings rotates all the loadings using Varimax Algorithm of Orthogonal Rotation and distributes all the loadings almost uniformly across all the components, thus variation explained and Eigen values are also bifurcated nearly even across the factors. Thus component one explaining 12.8% variation with 3.591 Eigen value, component two, 10.6%, 2.98, third component with 10.3% and 2.909, fourth component with 7.537% variation and 2.11 Eigen value, fifth component with 7.494% variance explained and Eigen value of 2.098, sixth component with 6.109% and 1.71 Eigen value and seventh component with 5.35% variation and 1.498 Eigen Value. Figure 1 below represents the Eigen values of the seven components based on Extraction Sums of Squared Loadings in the graphical form.
Table 9: Rotated Component Matrix identifying seven factors with the highest factor loadings of each variable in each factor.
Rotated Component Matrix
Perceived Ethics Of Firm
Competing Financial Needs
Current Financial Position
Condition of Financial Statements
Data In Reports & Prospectuses
Past Performance of Investors Stock Portfolio
Past Performance Of Stock
Recent Price Movements Of Firms Stock
Expected Stock Market Performance
Gut Feeling On Economy
Expected Corporate Earnings
Friend or Coworker Recommendation
Attractiveness of Non Stock Investments
Use of Valuation Equations
Current Economic Indicators
Time Before Funds are Needed
Affordable Share Price
Stock Broker Recommendation
Feelings For Firms Products And Services
Table 9 represents the Rotated Component Matrix which identifies seven factors with highest factor loadings of each variable in each factor. The first factor/component bearing 6 variables, tax consequences being omitted because it is cross loading in another component as well, thus 5 variables remaining which include Local Operations, International Operations, Institutional Holdings, Environmental Record, Perceived Ethics Of Firm with .726, 0.716, 0.669, 0.639, and 0.500 factor loadings, respectively. Factor/ Component two include variables such as Current Financial Position, Condition of Financial Statements, Exchange Listing, and Data in Reports & Prospectuses with factor loadings of 0.794, 0.715, 0.632, and 0.591. Factor 3 comprises of Past Performance of Investors Stock Portfolio, Recent Price Movements of Firms Stock, Expected Stock Market Performance, and Gut Feeling on Economy with 0.780, 0.694, 0.675, 0.635, and 0.533 loadings.
Factor 4 consists of Friend or Coworker Recommendation and Attractiveness of Non Stock Investments with factor loadings of 0.717 and 0.677. While Factor 5 has variables Use of Valuation Equations, Current Economic Indicators, and Time before Funds are needed containing 0.692, 0.690 and -0.601. This negative sign indicates importance of the variable but in the opposite direction of the status of the variable i.e. it indicates inverse relationship among these 3 variables. Competing Financial Needs and Diversification Needs have no factor loading because they have no important role to play in any factor. Affordable Share Price, Minimizing Risk and Expected Dividends are a part of sixth factor/ component with 0.718, 0.558 and 0.476 loadings. Finally, the last factor includes Stock Broker Recommendation and Feelings for Firms Products and Services variables with 0.810 and 0.451factor loadings in the rotated component matrix.
Table 10: Hypotheses Assessment Summary
Factor 1: Social relevance & Image alpha= (0.762)
Perceived Ethics Of Firm
Factor 2 : Accounting Information alpha= (0.752)
Current Financial Position
Condition of Financial Statements
Data In Reports & Prospectuses
Factor 3: Stock Performance alpha= (0.774)
Past Performance of Investors Stock Portfolio
Past Performance Of Stock
Recent Price Movements Of Firms Stock
Expected Stock Market Performance
Gut Feeling On Economy
Factor 4: Friend/Coworker Influence alpha= (0.679)
Friend or Coworker Recommendation
Attractiveness of Non Stock Investments
Factor 5: Evaluation alpha= (0.132)
Use of Valuation Equations
Current Economic Indicators
Time Before Funds are Needed
Factor 6: Classic alpha= (0.463)
Affordable Share Price
Factor 7: Stock Broker Influence alpha= (0.502)
Stock Broker Recommendation
Feelings For Firms Products And Services
Table 10 shows the hypotheses assessment summary to give clarity to the factors extracted. Each factor shows its own reliability score i.e. alpha along with the factor loadings of each variable in each of the seven factors.
Table 11: Seven Factors identified through Factor Analysis.
Factor 1: Social relevance & Image
Perceived Ethics Of Firm
Factor 2 : Accounting Information
Current Financial Position