Behavioural Biases in Investment Decision Making: A research study on snakebite and house money effects on Indian individuals
Udayan Das1*, Shakti Ranjan Mohapatra2
1Professor, Asian School of Business Management, Bhubaneswar and Doctoral Research Scholar, Utkal University
2Principal, College of IT and Management Education, Bhubaneswar and Dean, Faculty of Management Studies, Biju Patnaik University of Technology, Odisha
*Corresponding Author E-mail: prof.udayandas@gmail.com, shakti.r.mohapatra@gmail.com
ABSTRACT:
Investor’s previous experience appears to be one of the major determining factors in investment decision making. It is observed that investors avoid those stocks which resulted in losses earlier. The painful experience known as snakebite effect makes the investor more risk averse when the repeat behaviour comes into the picture. A completely opposite behaviour is observed when the repeat behaviour is related to reinvestment of their earlier earned profit. It is like a gambler who takes greater risk while playing with the house’s money, an investor also becomes willing to take greater risk if she/he plans to reinvest their earlier earned profit. Loss of the earlier earned profit is not considered as loss of their own money which insists them to undertake greater risks. A risk averse investor becomes increasingly risk seeker due to this house money effect. In this research study, it is strongly evidenced that both the snakebite and house money effects influence the investors in their decision making process which may lead to a serious error in judgement in many occasions.
KEY WORDS: Bias, Snakebite, House money, risk averse, risk seeker.
In investment, repeat behaviour is of great significance. Investor’s previous experience appears to be one of the major determining factors. It is observed that investors tend to repurchase those stocks which resulted in a gain for them but they avoid those which resulted in losses earlier. They do it because they do not want the repetition of an experience of pain but want to have the repetition of experiences of pleasure. These gain and pain experiences are evident in their subsequent behaviour. As a result, they generally avoid all the earlier losing stocks when they take a decision of repeat buying.
PREVIOUS STUDIES AND SCOPE OF RESEARCH:
Strahilevitz, Odean, and Barber (2005)8 find that the investors’ willingness to repurchase a stock is greatly dependant on the previous experience attached with it. Investors are ready to repurchase such stocks which were sold earlier resulting a gain. The same decision is not seen in case of those for which the investors suffered a loss and stocks those have risen in price subsequently after they sold those. The said behaviour is a reflection of investors’ emotional reactions. Their disappointment for a particular stock makes them distant from it. On the contrary, a gainful previous experience results reinforcement for purchasing the same stock again. Here, positive emotions are the responsible factor. Hence, positive emotions lead to repurchase and negative emotions lead to avoidance. The old proverb “once bitten, twice shy” is still relevant in decision-making of investors. It is “snakebite effect”, one of the psychological biases on decision-making of investors.
Chin (2012)3 finds significant relationship between regret and the investor’s decision making. Based on a primary data collected through a group of four questions and analysed at a 4 point likert scale, she finds the responses highly supports the relationship. She observes the likert scores as 3.18 (I don’t want to take the high risk although high risk brings high return), 3.12 (I try to avoid buying losing stock in which I had made losses), 3.07 (When the price drops temporarily, I will sell the stock to prevent from getting more losses) and 3.21 (I worry about the influence of financial crisis).
Kuo, Huang and Jane (2013)7 conduct a study on the Taiwanese stock investors with the backdrop of global financial tsunami 2008. The financial tsunami resulted serious looses on investors’ wealth and their study reveals that almost 54% lost significant money and out of those, approximately 57% could not recover even after 15 months. Authors find that after the tsunami the willingness of an average investor to take risk became significantly decreased. Moreover, their trading got decreased and their use of stop-loss mechanism got increased. The authors find that the individual investors who suffered due to this debacle became more risk-averse and their trading volume got decreased. Moreover, the authors identify another significant relationship that more the investors lost, more risk-averse they became. It establishes the snakebite effect proved to be in force.
Kartasova, Gaspareniene and Remeikiene (2014)6 examine the behaviour of Lithuanian investors based on their decisions on buying and selling stocks. They observe that when an investor faces big losses, it results in disappointment as well as loss of confidence on self. In other words, they become scared to consider the risky securities for investment which is none other than the feeling of “snakebite” effect. Their loss of confidence and fear to take risks causes influence on investors’ subsequent decisions in an adverse manner which results an impact on profit margin or return on investment.
Ghelichi, Nakhjavan and Gharehdaghi (2016)4 conduct a study on influence of psychological factors on investors’ financial decision making in Iran’s stock exchange in Tehran. They describe the snakebite effect as a danger that may threaten the investors. Snakebite serves a major role by providing with the forecasting feature while explaining the behaviour of decision-making under uncertainty. Their study reveals that beliefs and confidence impact the investors’ decision positively while regret and snake bite effect impact the investors’ decision negatively.
From the risk perspective, the effect of “house money” is considered to be opposite to the “snakebite”. Investors are found to take higher risks when they re-invest their earned income from previously held investments. This is because the investors show the tendency not to consider it as their own income rather a winning result of gambling.
Thaler and Johnson (1990)9 present their opinion based on their research on the effects of prior gain or loss on individual investor’s willingness for risk taking. They find that after a gain, if there are subsequent losses that are smaller than the original gain, are integrated with the prior gain, mitigating the influence of loss aversion and facilitating risk-seeking. The intuition behind this effect is captured by the expression in gambling parlance of "playing with the house money." It refers to increased risk seeking due to the presence of a prior gain. The authors describe also the contradictory picture when there is an increased risk seeking in presence of prior losses. It is known as “break even effects”.
Ackert, Charupat, Church and Deaves (2003)2 demonstrate that there is evidence that risk-taking behaviour is influenced by prior monetary gains and losses, People become more risk taking, when investors are endowed with house money, perhaps because the earlier gain cushions subsequent losses. The authors follow an experimental method and their results show traders’ bids, price predictions, and market prices are influenced by the amount of money that is provided prior to trading. They confirm that market prices are consistent with a house money effect in all treatments.
Weber and Zuchel (2005)10 address the issue of apparent contradiction in investors’ behaviour while making sequential decisions. It has been found that following a gain, a large percentage of subjects are increasingly risk seeking but there are examples of increasingly risk seeking following a loss. Their study reveals that a greater risk taking is exhibited following losses than following gains if the problem is presented as a portfolio decision making. This behaviour is consistent with escalation of commitment. On the contrary, if the problem is presented in the shape of two-stage betting, risk taking is greater following gains than following losses. This behaviour is consistent with house money effect.
Wen, Chao and Liu (2012)11 explain the house money effect as the psychological tendency of investors to become increasingly risk-seeking following prior gains. The authors track the behaviour of the whole stock market data of 14 countries or regions as one entity. Their empirical research shows a clear influence of prior losses or gains on current risk taking attitude. They find that prior gain makes the investors less risk averse while prior losses make them more risk averse. It confirms the influence of house money effect.
Hsu and Chow (2013)5 conduct a study of the house money effect on the risk taking behaviour of investors. They observe that when gains are more substantial, investors take greater risk. The effect gets diminished with time as the willingness of risk taking following a gain gets reduced. The find strong empirical evidences of house money effect in the real world financial market.
OBJECTIVES:
1) To observe the existence of snakebite and house money effects on individual in investment decision making process.
2) To analyse such existence based on multiple demographic parameters.
HYPOTHESIS:
H-1: “Snakebite and house money effects exist on Indian individuals in investment decision making process.”
RESEARCH METHODOLOGY:
Questionnaire:
This study is a primary research and exclusively based on data collected through structured questionnaire. It was framed by adopting the questionnaire used by Chin, L. L. A. (2012)2. Those were improvised and reshaped to match with Indian perspectives. To judge the effectiveness and validity, two professional brokers, two investors and three academicians were consulted before the final structure of the questionnaire was framed. Thus, face validity method has been used for validation purpose.
There are two distinctly segmented sections in the questionnaire:
A. Personal and Demographic information
B. Questions for Hypotheses testing
Question numbers 1-5 are for personal information and answers are open-ended.
Q-1. Name
Q-2. Address
Q-3. City
Q-4. Contact No.
Q-5. Email id
Question numbers 6-10 are specifically kept for five demographic variables which are gender, age, occupation, annual income and investment frequency. Responses have been collected through a closed options mode.
Q-6. Gender:
Two options: Male and Female.
Q-7. Age :
Four options: 26-35 Years, 36-45 Years, 46-55 Years and 56-65 Years.
Individuals upto the age of 25 and above the age of 65 are excluded because of their lack of exposure to invest and limited requirements for investment respectively.
Q-8. Occupation:
Three options: Salaried, Self- Employed and Professional.
Q-9. Annual Income:
Four options: INR 3,00,000 to INR 6,00,000, INR 6,00,001 to INR 9,00,000, INR 9,00,001 to INR 12,00,000 and INR 12,00,001 to INR 15,00,000.
The annual income group below INR 3,00,000 and the annual income group above INR 15,00,000 both have not been considered due to limited exposure towards investment and higher income status respectively.
Q-10. I invest in stock market (Investment frequency):
Four options: (I invest in stock market) Regularly, Very often, Sometimes and Never.
As an individual who is not having any exposure to the stock market cannot really contribute with a valid opinion relating to investment decision-making, all the responses with “Never” option have not been considered.
Question numbers 11-15 are related to Hypothesis-1
To test this hypothesis, respondents are asked five individual questions which are as under:
Q-11. I try to avoid buying such stocks in which I had made losses earlier.
Q-12. I search for opportunities of ‘repeat buying’ of such stocks in which I had made gains earlier.
Q-13. When the price drops temporarily, I sell the stock to prevent from getting further losses.
Q-14. I don’t want to invest in high risk stocks although high risk stocks bring high return.
Q-15. I invest in high risk stocks only when I reinvest my earlier made profits.
Data collection:
Responses have been collected in the form of primary data through online questionnaire (Surveymonkey.com) and physical collection through hard copy of the questionnaire. A total number of 360 responses have been collected through online and 140 responses have been collected through physical collection during the period 05.01.2015 to 02.01.2016. The total numbers of 500 responses have been received from 97 different places of 20 states of India namely Andhra Pradesh, Assam, Bihar, Chhattisgarh, Delhi, Gujarat, Haryana, Jharkhand, Karnataka, Kerala, Madhya Pradesh, Maharashtra, Odisha, Punjab, Rajasthan, Tamil Nadu, Telangana, Uttar Pradesh, Uttarakhand and West Bengal.
Methodology and Tools:
At the outset, reliability analysis of the data is performed to identify how well the questions grouped are positively correlated to one another. As stated by Nunnally and Bernstein (1994)1, Cronbach’s alpha value of 0.60 and above is considered to be reliable (To see that the items are homogenous and measuring the same construct). To establish the reliability, Cronbach’s alpha for the grouped questions Q-11-15 is calculated.
The hypothesis is tested through those five questions stated above which are basically five different statements. Likert scale has been used for the measurement of the degree of their agreements or disagreements on each of the five statements. A five point scale has been used but forced choice method is applied. The five points of measurements are “Strongly Disagree”, “Disagree”, “Neither Agree nor Disagree”, “Agree” and “Strongly Agree”. The assigned values are -2, -1, 0, +1, +2 respectively. The objective is to know the degree of agreement or disagreement and so, forced choice method has been used. The respondents are not given the opportunity to exercise the option “Neither Agree nor Disagree”. It does not allow the respondents to become indifferent and forces them either to give opinion for agreement or disagreement.
All these five statements are so prepared that each of those indicates the existence of the bias. A particular statement is considered as agreed if the value of the answer is greater than zero and the statement is considered as disagreed if the value is less than zero. As already stated earlier, the option for zero value is not allowed to the respondents.
The mean score of all responses for a particular statement also reflects the overall degree of agreement or disagreement. If the mean score is greater than 0 and upto +1, it confirms agreement while a mean score greater than +1 proves strong agreement. In the similar way, a mean score which is less than 0 and upto -1 proves disagreement and any mean score which is less than -1 confirms strong disagreement.
Acceptance or rejection of the hypothesis is made by cumulating all the values of all five questions grouped for the hypothesis. A positive mean score makes the hypothesis accepted and a greater than +1 mean score makes the hypothesis strongly accepted. Likewise, a negative mean score makes the hypothesis rejected and a less the -1 mean score makes the hypothesis strongly rejected.
As explained above, due to restriction over the middle option, respondents were provided with only four options “Strongly Disagree (-2)”, “Disagree (-1)”, “Agree (+1)” and “Strongly Agree (+2)”. Hence, for all purposes including data collection, data analysis and hypothesis testing, the scale appears to be a four point scale.
The demographic analysis is carried out based on five distinct parameters, gender, age, occupation, annual income and investment frequency. Two way ANOVA has been used as a tool to verify whether the results are equal across the different parameters.
Data Analysis – Descriptive Statistics:
A question wise analysis is made. The responses received are explained below:
Q-6: The gender wise distribution of the respondents
Male respondents (79.80%) are much higher than the female respondents (20.20%). It is an indication of higher involvement of male respondents into investments.
Q-7: The age wise distribution of the respondents
The largest respondent age group covering 72.60% of the respondents is 26-35. The age group 36-45 holds 15.20%, age group 46-55 is 8.80% and the age group 56-65 is 3.40%.
Q-8: The occupation wise distribution of the respondents
Salaried respondents are 77.80% of the total respondents while Self-Employed and Professionals are 12.40% and 9.80% respectively.
Q-9: The income wise distribution of the respondents is as follows
68.80% of the total respondents belong to the income group of INR 3,00,000 - 6,00,000, 17.80% belong to income group INR 6,00,001 – 9,00,000, 6.00% belong to income group INR 9,00,001 – 12,00,000 (6.00%) and 7.60% belong to income group INR 12,00,001 – 15,00,000.
Q-10: The investment frequency wise distribution of the respondents
Respondents who had opted for “Never” option are not considered for the study as they are non-investors (not included in 500). Out of the valid respondents, 71.80% invest “Sometimes”, 17.80% invest “Very often” and 10.40% invest “Regularly”.
A question wise analysis for Q-11 to Q-15 to test Hypothesis-1 is furnished below:
Strength of Agreement by respondents for Q-11 to Q-15 (H-1)
Q-11. I try to avoid buying such stocks in which I had made losses earlier.
Strongly disagree (-2): 1.60% Disagree (-1): 7.60%
Agree (+1): 64.60% Strongly agree (+2): 26.20%
Q-12. I search for opportunities of ‘repeat buying’ of such stocks in which I had made gains earlier.
Strongly disagree (-2): 1.20% Disagree (-1): 6.80%
Agree (+1): 69.60% Strongly agree (+2): 22.40%
Q-13. When the price drops temporarily, I sell the stock to prevent from getting further losses.
Strongly disagree (-2): 1.80% Disagree (-1): 8.80%
Agree (+1): 61.80% Strongly agree (+2): 27.60%
Q-14. I don’t want to invest in high risk stocks although high risk stocks bring high return.
Strongly disagree (-2): 1.60% Disagree (-1): 6.40%
Agree (+1): 71.60% Strongly agree (+2): 20.40%
Q-15. I invest in high risk stocks only when I reinvest my earlier made profits.
Strongly disagree (-2): 1.40% Disagree (-1): 8.20%
Agree (+1): 60.60% Strongly agree (+2): 29.80%
Data Analysis – Test of Hypothesis:
Before analysing the data, the reliability test has been performed for the grouped questions Q-11 to Q-15 in connection with the test of Hypthesis-1. The Cronbach’s alpha value is 0.739 for these grouped questions. As the Cronbach’s alpha value is more than 0.60, data is considered to be reliable. Hence, the reliability test so performed proves a positive correlation among the questions. The reliability is also hereby established.
The answers given by the respondents for Q-11 to Q-15 are analysed based on five demographic parameters. The findings are presented in Table1 to Table-5.
Table – 1. Strength of Agreement by respondents for Q-11
|
I try to avoid buying such stocks in which I had made losses earlier. |
|||||||||||
|
Answer Options |
Response Percent |
Response Count |
Score |
||||||||
|
Strongly disagree (-2) |
1.60% |
8 |
-16 |
||||||||
|
Disagree (-1) |
7.60% |
38 |
-38 |
||||||||
|
Agree (+1) |
64.60% |
323 |
323 |
||||||||
|
Strongly agree (+2) |
26.20% |
131 |
262 |
||||||||
|
Total Score |
531 |
||||||||||
|
Mean Score |
1.06 |
||||||||||
|
Standard Deviation |
0.840 |
||||||||||
|
Demographic analysis of the responses |
|||||||||||
|
Demographic Parameters |
No |
Strongly Disagree (-2) |
Disagree (-1) |
Agree (+1) |
Strongly Agree (+2) |
Total Respondents |
Mean Score |
||||
|
Gender |
Female |
101 |
2 |
8 |
67 |
24 |
101 |
1.02 |
|||
|
Male |
399 |
6 |
30 |
256 |
107 |
399 |
1.07 |
||||
|
Total |
500 |
8 |
38 |
323 |
131 |
500 |
1.06 |
||||
|
Age Group |
26-35 |
363 |
8 |
23 |
239 |
93 |
363 |
1.06 |
|||
|
36-45 |
76 |
0 |
9 |
46 |
21 |
76 |
1.04 |
||||
|
46-55 |
44 |
0 |
5 |
27 |
12 |
44 |
1.05 |
||||
|
56-65 |
17 |
0 |
1 |
11 |
5 |
17 |
1.18 |
||||
|
Total |
500 |
8 |
38 |
323 |
131 |
500 |
1.06 |
||||
|
Occupation |
Salaried |
389 |
8 |
30 |
252 |
99 |
389 |
1.04 |
|||
|
Self-Employed |
62 |
0 |
3 |
43 |
16 |
62 |
1.16 |
||||
|
Professional |
49 |
0 |
5 |
28 |
16 |
49 |
1.12 |
||||
|
Total |
500 |
8 |
38 |
323 |
131 |
500 |
1.06 |
||||
|
Income Range |
INR 3,00,000 - 6,00,000 |
343 |
8 |
25 |
222 |
88 |
343 |
1.04 |
|||
|
INR 6,00,001 - 9,00,000 |
89 |
0 |
7 |
58 |
24 |
89 |
1.11 |
||||
|
INR 9,00,001 - 12,00,000 |
30 |
0 |
1 |
22 |
7 |
30 |
1.17 |
||||
|
INR 12,00,001 - 15,00,000 |
38 |
0 |
5 |
21 |
12 |
38 |
1.05 |
||||
|
Total |
500 |
8 |
38 |
323 |
131 |
500 |
1.06 |
||||
|
Investment Frequency |
Regularly |
52 |
0 |
4 |
38 |
10 |
52 |
1.04 |
|||
|
Very often |
89 |
3 |
9 |
53 |
24 |
89 |
0.97 |
||||
|
Sometimes |
359 |
5 |
25 |
232 |
97 |
359 |
1.09 |
||||
|
Total |
500 |
8 |
38 |
323 |
131 |
500 |
1.06 |
||||
Male respondents have higher level of agreement. Age group 56-65, Self-Employed group, Income group INR 9,00,001-12,00,000 and Sometimes group have the highest level of agreements in four other parameters.
Table – 2. Strength of Agreement by respondents for Q-12
|
I search for opportunities of ‘repeat buying’ of such stocks in which I had made gains earlier. |
|||||||||||
|
Answer Options |
Response Percent |
Response Count |
Score |
||||||||
|
Strongly disagree (-2) |
1.20% |
6 |
-12 |
||||||||
|
Disagree (-1) |
6.80% |
34 |
-34 |
||||||||
|
Agree (+1) |
69.60% |
348 |
348 |
||||||||
|
Strongly agree (+2) |
22.40% |
112 |
224 |
||||||||
|
Total Score |
526 |
||||||||||
|
Mean Score |
1.05 |
||||||||||
|
Standard Deviation |
0.775 |
||||||||||
|
Demographic analysis of the responses |
|||||||||||
|
Demographic Parameters |
No |
Strongly Disagree (-2) |
Disagree (-1) |
Agree (+1) |
Strongly Agree (+2) |
Total Respondents |
Mean Score |
||||
|
Gender |
Female |
101 |
2 |
8 |
72 |
19 |
101 |
0.97 |
|||
|
Male |
399 |
4 |
26 |
276 |
93 |
399 |
1.07 |
||||
|
Total |
500 |
6 |
34 |
348 |
112 |
500 |
1.05 |
||||
|
Age Group |
26-35 |
363 |
6 |
20 |
256 |
81 |
363 |
1.06 |
|||
|
36-45 |
76 |
0 |
6 |
52 |
18 |
76 |
1.08 |
||||
|
46-55 |
44 |
0 |
7 |
29 |
8 |
44 |
0.86 |
||||
|
56-65 |
17 |
0 |
1 |
11 |
5 |
17 |
1.18 |
||||
|
Total |
500 |
6 |
34 |
348 |
112 |
500 |
1.05 |
||||
|
Occupation |
Salaried |
389 |
6 |
26 |
269 |
88 |
389 |
1.05 |
|||
|
Self-Employed |
62 |
0 |
6 |
40 |
16 |
62 |
1.06 |
||||
|
Professional |
49 |
0 |
2 |
39 |
8 |
49 |
1.08 |
||||
|
Total |
500 |
6 |
34 |
348 |
112 |
500 |
1.05 |
||||
|
Income Range |
INR 3,00,000 - 6,00,000 |
343 |
6 |
23 |
238 |
76 |
343 |
1.03 |
|||
|
INR 6,00,001 - 9,00,000 |
89 |
0 |
5 |
66 |
18 |
89 |
1.09 |
||||
|
INR 9,00,001 - 12,00,000 |
30 |
0 |
1 |
22 |
7 |
30 |
1.17 |
||||
|
INR 12,00,001 - 15,00,000 |
38 |
0 |
5 |
22 |
11 |
38 |
1.03 |
||||
|
Total |
500 |
6 |
34 |
348 |
112 |
500 |
1.05 |
||||
|
Investment Frequency |
Regularly |
52 |
0 |
2 |
41 |
9 |
52 |
1.10 |
|||
|
Very often |
89 |
2 |
4 |
58 |
25 |
89 |
1.12 |
||||
|
Sometimes |
359 |
4 |
28 |
249 |
78 |
359 |
1.03 |
||||
|
Total |
500 |
6 |
34 |
348 |
112 |
500 |
1.05 |
||||
Male respondents have higher level of agreement. Age group 56-65, Professional group, Income group INR 9,00,001-12,00,000 and Very often group have the highest level of agreements in four other parameters.
Table – 3. Strength of Agreement by respondents for Q-13
|
When the price drops temporarily, I sell the stock to prevent from getting further losses. |
|||||||||||
|
Answer Options |
Response Percent |
Response Count |
Score |
||||||||
|
Strongly disagree (-2) |
1.80% |
9 |
-18 |
||||||||
|
Disagree (-1) |
8.80% |
44 |
-44 |
||||||||
|
Agree (+1) |
61.80% |
309 |
309 |
||||||||
|
Strongly agree (+2) |
27.60% |
138 |
276 |
||||||||
|
Total Score |
523 |
||||||||||
|
Mean Score |
1.05 |
||||||||||
|
Standard Deviation |
0.888 |
||||||||||
|
Demographic analysis of the responses |
||||||||||||
|
Demographic Parameters |
No |
Strongly Disagree (-2) |
Disagree (-1) |
Agree (+1) |
Strongly Agree (+2) |
Total Respondents |
Mean Score |
|||||
|
Gender |
Female |
101 |
2 |
6 |
65 |
28 |
101 |
1.10 |
||||
|
Male |
399 |
7 |
38 |
244 |
110 |
399 |
1.03 |
|||||
|
Total |
500 |
9 |
44 |
309 |
138 |
500 |
1.05 |
|||||
|
Age Group |
26-35 |
363 |
9 |
30 |
214 |
110 |
363 |
1.06 |
||||
|
36-45 |
76 |
0 |
9 |
51 |
16 |
76 |
0.97 |
|||||
|
46-55 |
44 |
0 |
4 |
29 |
11 |
44 |
1.07 |
|||||
|
56-65 |
17 |
0 |
1 |
15 |
1 |
17 |
0.94 |
|||||
|
Total |
500 |
9 |
44 |
309 |
138 |
500 |
1.05 |
|||||
|
Occupation |
Salaried |
389 |
9 |
34 |
239 |
107 |
389 |
1.03 |
||||
|
Self-Employed |
62 |
0 |
7 |
37 |
18 |
62 |
1.06 |
|||||
|
Professional |
49 |
0 |
3 |
33 |
13 |
49 |
1.14 |
|||||
|
Total |
500 |
9 |
44 |
309 |
138 |
500 |
1.05 |
|||||
|
Income Range |
INR 3,00,000 - 6,00,000 |
343 |
8 |
25 |
208 |
102 |
343 |
1.08 |
||||
|
INR 6,00,001 - 9,00,000 |
89 |
1 |
10 |
58 |
20 |
89 |
0.97 |
|||||
|
INR 9,00,001 - 12,00,000 |
30 |
0 |
5 |
21 |
4 |
30 |
0.80 |
|||||
|
INR 12,00,001 - 15,00,000 |
38 |
0 |
4 |
22 |
12 |
38 |
1.11 |
|||||
|
Total |
500 |
9 |
44 |
309 |
138 |
500 |
1.05 |
|||||
|
Investment Frequency |
Regularly |
52 |
0 |
7 |
34 |
11 |
52 |
0.94 |
||||
|
Very often |
89 |
3 |
8 |
53 |
25 |
89 |
1.00 |
|||||
|
Sometimes |
359 |
6 |
29 |
222 |
102 |
359 |
1.07 |
|||||
|
Total |
500 |
9 |
44 |
309 |
138 |
500 |
1.05 |
|||||
Female respondents have higher level of agreement. Age group 46-55, Professional group, Income group INR 12,00,001-15,00,000 and Sometimes group have the highest level of agreements in four other parameters.
Table – 4. Strength of Agreement by respondents for Q-14
|
I don’t want to invest in high risk stocks although high risk stocks bring high return. |
||||||||||
|
Answer Options |
Response Percent |
Response Count |
Score |
|||||||
|
Strongly disagree (-2) |
1.60% |
8 |
-16 |
|||||||
|
Disagree (-1) |
6.40% |
32 |
-32 |
|||||||
|
Agree (+1) |
71.60% |
358 |
358 |
|||||||
|
Strongly agree (+2) |
20.40% |
102 |
204 |
|||||||
|
Total Score |
514 |
|||||||||
|
Mean Score |
1.03 |
|||||||||
|
Standard Deviation |
0.777 |
|||||||||
|
Demographic analysis of the responses |
||||||||||
|
Demographic Parameters |
No |
Strongly Disagree (-2) |
Disagree (-1) |
Agree (+1) |
Strongly Agree (+2) |
Total Respondents |
Mean Score |
|||
|
Gender |
Female |
101 |
2 |
8 |
70 |
21 |
101 |
0.99 |
||
|
Male |
399 |
6 |
24 |
288 |
81 |
399 |
1.04 |
|||
|
Total |
500 |
8 |
32 |
358 |
102 |
500 |
1.03 |
|||
|
Age Group |
26-35 |
363 |
8 |
23 |
264 |
68 |
363 |
0.99 |
||
|
36-45 |
76 |
0 |
4 |
58 |
14 |
76 |
1.08 |
|||
|
46-55 |
44 |
0 |
4 |
24 |
16 |
44 |
1.18 |
|||
|
56-65 |
17 |
0 |
1 |
12 |
4 |
17 |
1.12 |
|||
|
Total |
500 |
8 |
32 |
358 |
102 |
500 |
1.03 |
|||
|
Occupation |
Salaried |
389 |
8 |
24 |
286 |
71 |
389 |
1.00 |
||
|
Self-Employed |
62 |
0 |
5 |
42 |
15 |
62 |
1.08 |
|||
|
Professional |
49 |
0 |
3 |
30 |
16 |
49 |
1.20 |
|||
|
Total |
500 |
8 |
32 |
358 |
102 |
500 |
1.03 |
|||
|
Income Range |
INR 3,00,000 - 6,00,000 |
343 |
7 |
19 |
248 |
69 |
343 |
1.03 |
||
|
INR 6,00,001 - 9,00,000 |
89 |
1 |
6 |
63 |
19 |
89 |
1.04 |
|||
|
INR 9,00,001 - 12,00,000 |
30 |
0 |
3 |
21 |
6 |
30 |
1.00 |
|||
|
INR 12,00,001 - 15,00,000 |
38 |
0 |
4 |
26 |
8 |
38 |
1.00 |
|||
|
Total |
500 |
8 |
32 |
358 |
102 |
500 |
1.03 |
|||
|
Investment Frequency |
Regularly |
52 |
0 |
5 |
31 |
16 |
52 |
1.12 |
||
|
Very often |
89 |
3 |
10 |
61 |
15 |
89 |
0.84 |
|||
|
Sometimes |
359 |
5 |
17 |
266 |
71 |
359 |
1.06 |
|||
|
Total |
500 |
8 |
32 |
358 |
102 |
500 |
1.03 |
|||
Male respondents have higher level of agreement. Age group 46-55, Professional group, Income group INR 6,00,001-9,00,000 and Regular group have the highest level of agreements in four other parameters.
Table – 5. Strength of Agreement by respondents for Q-15
|
I invest in high risk stocks only when I reinvest my earlier made profits. |
||||||||||
|
Answer Options |
Response Percent |
Response Count |
Score |
|||||||
|
Strongly disagree (-2) |
1.40% |
7 |
-14 |
|||||||
|
Disagree (-1) |
8.20% |
41 |
-41 |
|||||||
|
Agree (+1) |
60.60% |
303 |
303 |
|||||||
|
Strongly agree (+2) |
29.80% |
149 |
298 |
|||||||
|
Total Score |
546 |
|||||||||
|
Mean Score |
1.09 |
|||||||||
|
Standard Deviation |
0.862 |
|||||||||
|
Demographic analysis of the responses |
||||||||||
|
Demographic Parameters |
No |
Strongly Disagree (-2) |
Disagree (-1) |
Agree (+1) |
Strongly Agree (+2) |
Total Respondents |
Mean Score |
|||
|
Gender |
Female |
101 |
2 |
7 |
70 |
22 |
101 |
1.02 |
||
|
Male |
399 |
5 |
34 |
233 |
127 |
399 |
1.11 |
|||
|
Total |
500 |
7 |
41 |
303 |
149 |
500 |
1.09 |
|||
|
Age Group |
26-35 |
363 |
7 |
21 |
220 |
115 |
363 |
1.14 |
||
|
36-45 |
76 |
0 |
13 |
41 |
22 |
76 |
0.95 |
|||
|
46-55 |
44 |
0 |
7 |
30 |
7 |
44 |
0.84 |
|||
|
56-65 |
17 |
0 |
0 |
12 |
5 |
17 |
1.29 |
|||
|
Total |
500 |
7 |
41 |
303 |
149 |
500 |
1.09 |
|||
|
Occupation |
Salaried |
389 |
7 |
30 |
240 |
112 |
389 |
1.08 |
||
|
Self-Employed |
62 |
0 |
7 |
36 |
19 |
62 |
1.08 |
|||
|
Professional |
49 |
0 |
4 |
27 |
18 |
49 |
1.20 |
|||
|
Total |
500 |
7 |
41 |
303 |
149 |
500 |
1.09 |
|||
|
Income Range |
INR 3,00,000 - 6,00,000 |
343 |
7 |
21 |
208 |
107 |
343 |
1.13 |
||
|
INR 6,00,001 - 9,00,000 |
89 |
0 |
8 |
57 |
24 |
89 |
1.09 |
|||
|
INR 9,00,001 - 12,00,000 |
30 |
0 |
4 |
15 |
11 |
30 |
1.10 |
|||
|
INR 12,00,001 - 15,00,000 |
38 |
0 |
8 |
23 |
7 |
38 |
0.76 |
|||
|
Total |
500 |
7 |
41 |
303 |
149 |
500 |
1.09 |
|||
|
Investment Frequency |
Regularly |
52 |
0 |
7 |
33 |
12 |
52 |
0.96 |
||
|
Very often |
89 |
2 |
10 |
42 |
35 |
89 |
1.10 |
|||
|
Sometimes |
359 |
5 |
24 |
228 |
102 |
359 |
1.11 |
|||
|
Total |
500 |
7 |
41 |
303 |
149 |
500 |
1.09 |
|||
Male respondents have higher level of agreement. Age group 56-65, Professional group, Income group INR 3,00,000-6,00,000 and Sometimes group have the highest level of agreements in four other parameters.
Demographic analysis:
Table – 6. Demographic analysis of respondents’ choices on Grouped Questions for H-1
|
Demographic Parameters |
Mean Score |
||||||
|
Q-11 |
Q-12 |
Q-13 |
Q-14 |
Q-15 |
H-1 |
||
|
Gender |
Female |
1.02 |
0.97 |
1.10 |
0.99 |
1.02 |
1.02 |
|
Male |
1.07 |
1.07 |
1.03 |
1.04 |
1.11 |
1.07 |
|
|
Total |
1.06 |
1.05 |
1.05 |
1.03 |
1.09 |
1.06 |
|
|
Age Group |
26-35 |
1.06 |
1.06 |
1.06 |
0.99 |
1.14 |
1.07 |
|
36-45 |
1.04 |
1.08 |
0.97 |
1.08 |
0.95 |
1.02 |
|
|
46-55 |
1.05 |
0.86 |
1.07 |
1.18 |
0.84 |
1.00 |
|
|
56-65 |
1.18 |
1.18 |
0.94 |
1.12 |
1.29 |
1.14 |
|
|
Total |
1.06 |
1.05 |
1.05 |
1.03 |
1.09 |
1.06 |
|
|
Occupation |
Salaried |
1.04 |
1.05 |
1.03 |
1.00 |
1.08 |
1.04 |
|
Self-Employed |
1.16 |
1.06 |
1.06 |
1.08 |
1.08 |
1.09 |
|
|
Professional |
1.12 |
1.08 |
1.14 |
1.20 |
1.20 |
1.15 |
|
|
Total |
1.06 |
1.05 |
1.05 |
1.03 |
1.09 |
1.06 |
|
|
Income Range |
INR 3,00,000 - 6,00,000 |
1.04 |
1.03 |
1.08 |
1.03 |
1.13 |
1.06 |
|
INR 6,00,001 - 9,00,000 |
1.11 |
1.09 |
0.97 |
1.04 |
1.09 |
1.06 |
|
|
INR 9,00,001 - 12,00,000 |
1.17 |
1.17 |
0.80 |
1.00 |
1.10 |
1.05 |
|
|
INR 12,00,001 - 15,00,000 |
1.05 |
1.03 |
1.11 |
1.00 |
0.76 |
0.99 |
|
|
Total |
1.06 |
1.05 |
1.05 |
1.03 |
1.09 |
1.06 |
|
|
Investment Frequency |
Regularly |
1.04 |
1.10 |
0.94 |
1.12 |
0.96 |
1.03 |
|
Very often |
0.97 |
1.12 |
1.00 |
0.84 |
1.10 |
1.01 |
|
|
Sometimes |
1.09 |
1.03 |
1.07 |
1.06 |
1.11 |
1.07 |
|
|
Total |
1.06 |
1.05 |
1.05 |
1.03 |
1.09 |
1.06 |
|
The opinions are almost similar in case of both Female and Male respondents in the context of grouped questions for Hypothesis-1. While the mean score is 1.06 for the entire respondents, mean score for Female is lower at 1.02 but for Male respondents, it is 1.07. There is a strong agreement for both the genders.
Age group 56-65 shows the strongest agreement for Hypothesis-1 with a mean score of 1.14. The other three age groups also exhibit strong agreement with respective mean scores of 1.07, 1.02 and 1.00. However, the overall mean score is at 1.06 showing strong agreement across the all age groups.
Professionals exhibit the strongest agreement with a much higher mean score of 1.15. Self-Employed and Salaried groups are having the respective mean scores of 1.09 and 1.04. The analysis reveals that strong agreement in favour of Hypothesis-1 exists irrespective of occupational variability and unequal level of agreements across the occupational groups.
The group having income range of INR 12,00,001-15,00,000 shows just a little bit less than the strong agreement with a mean score of 0.99. Other three groups exhibit almost identical strong agreement with respective mean scores 1.06, 1.06 and 1.05. In case of individual questions, INR 9,00,001-12,00,000 exhibits highest level of agreement in two cases and rest three groups have highest mean score one each. However, all the income groups show clear agreement in favour of Hypothesis-1.
Sometimes group with a mean score of 1.07 shows the strongest agreement. The Regular group exhibits a mean score of 1.03 and the Very often group has a mean score of 1.01. It means there is a strong agreement in favour of Hypothesis-1 irrespective of their degree of involvement in investment.
Let us now examine if the results are equal across the demographic parameters. We hypothesise that there is no significant difference across the different groups and statements. Following is the two way ANOVA table:
Table – 7. ANOVA table on demographic analysis for Hypothesis-1
|
Source of Variation |
SS |
df |
MS |
F Values |
P Value |
|
|
Observed |
Tabulated (.05) |
|||||
|
Between Genders |
0.0048 |
1 |
0.0048 |
2.11 |
7.71 |
0.22 |
|
Between Statements |
0.0046 |
4 |
0.0011 |
0.50 |
6.39 |
0.74 |
|
Error |
0.0092 |
4 |
0.0023 |
|
||
|
Total |
0.0186 |
9 |
|
|||
|
|
|
|
|
|||
|
Between Age groups |
0.0579 |
3 |
0.0193 |
1.45 |
3.49 |
0.28 |
|
Between Statements |
0.0171 |
4 |
0.0043 |
0.32 |
3.26 |
0.86 |
|
Error |
0.1602 |
12 |
0.0134 |
|
||
|
Total |
0.2352 |
19 |
|
|||
|
|
|
|
|
|||
|
Between Occupation groups |
0.0293 |
2 |
0.0146 |
7.81 |
4.46 |
0.01 |
|
Between Statements |
0.0062 |
4 |
0.0015 |
0.82 |
3.84 |
0.55 |
|
Error |
0.0150 |
8 |
0.0019 |
|
||
|
Total |
0.0504 |
14 |
|
|||
|
|
|
|
|
|||
|
Between Income groups |
0.0172 |
3 |
0.0057 |
0.44 |
3.49 |
0.73 |
|
Between Statements |
0.0311 |
4 |
0.0078 |
0.59 |
3.26 |
0.68 |
|
Error |
0.1581 |
12 |
0.0132 |
|
||
|
Total |
0.2064 |
19 |
|
|||
|
|
|
|
|
|||
|
Between Investor groups |
0.0111 |
2 |
0.0055 |
0.66 |
4.46 |
0.54 |
|
Between Statements |
0.0138 |
4 |
0.0035 |
0.41 |
3.84 |
0.79 |
|
Error |
0.0667 |
8 |
0.0083 |
|
||
|
Total |
0.0915 |
14 |
|
|||
Since the observed values of F for genders, age groups, income groups and investor groups (frequency of investment) are lower than the corresponding tabulated values [df (1,4) (3,12) (3,12) (2,8)], these are significant at 5% level. Similar observations are available in terms of P value where, all are greater than .05.Therefore, it may be concluded that results are equal across these four parameters. On the contrary, there is significant difference in results among the occupation group as the observed value is larger than the corresponding tabulated value [df (2,8)]. Similarly, P value is also less than .05. It leads to rejection of the hypothesis that there is no significant difference across the different groups. Hence, the results are not equal across the occupation groups. Here, statements are related to the repeat behavior and risk attitude. The attitude towards risk varies significantly among the occupational groups which might have caused the greater variability
The observed values of F for statements are lower than the corresponding tabulated values [df (4,4) (4,12) (4,8) (4,12) (4,8)], these are significant at 5% level. P values are also greater than .05 in all cases. Therefore, it may be concluded that results are equal across the statements.
To summarise, the mean scores of Q-11 to Q-15 are 1.06, 1.05, 1.05, 1.03 and 1.09 respectively. It establishes a strong agreement for all the statements as all the mean scores are greater than +1. The mean score is 1.06 for the grouped questions 11-15. It is again greater than +1 which signifies an overall agreement for the hypothesis.
Table – 8. Response Analysis for Hypothesis-1
|
Particulars |
Q-11 |
Q-12 |
Q-13 |
Q-14 |
Q-15 |
H-1 (Total Score) |
|
|
Strongly Disagree (-2) |
8 |
6 |
9 |
8 |
7 |
38 |
7 (<-5) |
|
Disagree (-1) |
38 |
34 |
44 |
32 |
41 |
189 |
9 (<0, >=-5) |
|
Neither Agree nor Disagree (0) |
NA |
NA |
NA |
NA |
NA |
NA |
NIL (=0) |
|
Agree (+1) |
323 |
348 |
309 |
358 |
303 |
1641 |
214 (<=5, >0) |
|
Strongly Agree (+2) |
131 |
112 |
138 |
102 |
149 |
632 |
270 (>5) |
|
Total Respondents |
500 |
500 |
500 |
500 |
500 |
500 (x 5) |
500 |
|
Total Score |
531 |
526 |
523 |
514 |
546 |
2640 |
--- |
|
Mean Score |
1.06 |
1.05 |
1.05 |
1.03 |
1.09 |
1.06 |
--- |
|
Disagreement Zone (%) |
9.20% |
8.00% |
10.60% |
8.00% |
9.60% |
9.08% |
3.20% |
|
Indifferent Zone (%) |
NA |
NA |
NA |
NA |
NA |
NA |
0.00% |
|
Agreement Zone (%) |
90.80% |
92.00% |
89.40% |
92.00% |
90.40% |
90.92% |
96.80% |
Mean scores have been calculated based on the assigned values of -2, -1, 1 and 2 for strong disagreement, disagreement, agreement and strong agreement. It means, the strong disagreement zone lies between -2 to <-1, disagreement zone is -1 to <0, agreement zone is >0 to 1 and strong agreement zone is >1.
The total score for Q-11 is thus (8 x -2) + (38 x -1) + (323 x 1) + (131 x 2) = 531.
The mean score is thus, 531 ÷ 500 = 1.06.
In the same way, the mean scores for Q-12 to Q-15 as 1.05, 1.05, 1.03 and 1.09 respectively. Hence, strong agreement is exhibited in case of all five questions.
A respondent has the option to chose any one out of four options where two belong to wider agreement zone (strong agreement zone + agreement zone, assigned value range >0) and other two belong to wider disagreement zone (strong disagreement zone + disagreement zone, assigned value range <0).
In case of Q-11, 8 respondents have strong disagreement and 38 are having disagreement. So, the wider disagreement zone is (8+38) ÷ 500 x 100% = 9.20%. Similarly, the wider agreement zone is (323+131) ÷ 500 x 100% = 90.80%.
In the same way, the disagreement-agreement combination for Q-12 is (8.00%, 92.00%), for Q-13 is (10.60%, 89.40%), for Q-14 is (8.00%, 92.00%) and for Q-15 is (9.60%, 90.40%) respectively.
Respondents are in the wider agreement zone with huge majority for all five questions.
The Hypothesis-1 is tested in the light of mean scores.
The total score of the grouped questions (Q-11 to Q-15) = 531 + 526 + 523 + 514 + 546 = 2640.
The mean score is then 2640 ÷ (500 x 5) = 1.06.
The total no of respondents in the wider disagreement zone has been calculated by adding total no of respondents with “Strongly Disagree” options and “Disagree” options. The total no of respondents in the wider agreement zone has been calculated by adding total no of respondents with “Strongly Agree” options and “Agree” options. The total numbers in wider disagreement zone and wider agreement zone are 227 (9.08%) and 2273 (90.92%) respectively.
It proves the Hypothisis-1 to be true.
Another analysis has been performed to prove the hypothesis. The assigned values aggregating all five questions for a respondent ranges between -10 to +10. This range may be classified in five zones, which are, strong disagreement zone (-10 to -6), disagreement zone (-5 to -1), indifferent zone (0), agreement zone (1 to 5) and strong agreement zone (5 to 10). It is found that, 7 respondents have the total score between -10 to -6, 9 respondents have the total score between -5 to -1, no respondent has the score of 0, 214 respondents have the total score between 1 to 5 and 270 respondents have the total score between 6 to 10. In a broader way, (7+9) = 16 (3.20%) respondents belong to wider disagreement zone and (214+270) = 484 (96.80%) respondents belong to wider agreement zone.
It re-affirms that Hypothesis-1 to be true.
Hence, it is proved that “Snakebite and house money effects exist on Indian individuals in investment decision making process.”
Hypothesis-1 (H-1) is accepted.
CONCLUSION:
In this research study, it is strongly evidenced that both snakebite and house money effects are existing on individuals in investment decision making. The risk attitude of an investor appears to be variable depending upon different sets of experiences. It is observed that the respondents go for repeat buying in case of winners but avoid the losers because the earlier shock makes them shaky. It clearly demonstrates the evidence of snakebite effect working in the mind of respondents. On the other hand, a risk averse respondent becomes a risk seeker when the subject matter is “earlier made profits”. This transition from risk aversiveness to risk seeking clearly demonstrates that house money effect is working in the minds of the investors. They do it because for them, loss of capital and loss of profit are two different entities. They consider one as loss but not the other one which is definitely an irrational behaviour leading to a behavioural bias. To conclude, it is evident that both the snakebite and house money effects are existing which may lead to a serious error in judgement in many occasions.
REFERENCES:
1. Nunnally, J. C. and Bernstein, I. H. (1994), Psychometric Theory, McGraw-Hill, New York.
2. Ackert, L.F., Charupat, N., Church, B.K. and Deaves, R. (2003), An Experimental Examination of the House Money Effect in a Multi-Period Setting, Federal Reserve Bank of Atlanta, Working Paper 2003-13.
3. Chin, L. L. A. (2012), Psychological Biases and Investor Behaviour: Survey Evidence from Malaysian Stock Market, International Journal on Social Science, Economics and Art, Vol. 2, No. 2, 74-80.
4. Ghelichi, M.A., Nakhjavan, B. Gharehdaghi, M. (2016), Impact of psychological factors on investment decision making in stock exchange market, Asian Journal of Management Sciences and Education, Vol. 5, No. 3, Page 36-44.
5. Hsu, Y.L. and Chow, E.H. (2013), The house money effect on investment risk taking: Evidence from Taiwan, Pacific-Basin Finance Journal, Volume 21, Issue 1, Pages 1102–1115.
6. Kartasova, J., Gaspareniene, L. and Remeikiene, R. (2014), Influence of ‘Snake-Bite’ Effect on Investment Return Rate: Lithuanian Example, Mediterranean Journal of Social Sciences, Vol 5, No 27, Page 1769-1773.
7. Kuo, M.H., Huang, P.K. and Jane, W.J. (2013), The impact of the financial tsunami on stock investment behaviour: the case of individual investors in Taiwan, Investment Management and Financial Innovations, issue 1, page 207-217.
8. Strahilevitz, M., Odean, T. and Barber, B. (2011), Once Burned, Twice Shy: How Naive Learning and Counterfactuals Affect the Repurchase of Stocks Previously Sold, Journal of Marketing Research, Vol. 48, Special Issue 2011, pp. S102-S120.
9. Thaler. R.H. and Johmson, E.J. (1990), Gambling with the House Money and trying to Break Even: The effects of prior outcomes on risky choice, Management Science, Vol. 6, No.6, Pages 643-660.
10. Weber, M. and Zuchel, H. (2005), How Do Prior Outcomes Affect Risk Attitude? Comparing Escalation of Commitment and the House-Money ffect, Decision Analysis, Vol. 2, No. 1, pp. 30–43.
11. Wen, F.H., Chao, Y.C. and Liu, Z.F. (2012), Research on House Money Effect at the Market Level, Journal of Management, Changsha University, Volume (4): 45-51.
Received on 19.05.2017 Modified on 28.06.2017
Accepted on 10.07.2017 © A&V Publications all right reserved
Asian J. Management; 2017; 8(3):460-470.
DOI: 10.5958/2321-5763.2017.00074.9