Market Phenomena. Investors and the Disposition Effect

Seminar Paper, 2017

28 Pages, Grade: 1,3


Table of contents


1. Introduction

2. Basic discussion of the disposition effect
2.1. The main elements of the disposition effect
2.2. Empirical evidence of the disposition effect
2.3. Interaction of the effects

3. Experimental analysis of the the disposition effect for single investors
3.1. Theoretical background for the experiments
3.3. Results
3.4. Discussion of the results

4. Experimental analysis of the the disposition effect for team investors
4.1. Theoretical background for the experiments
4.2. Conceptual Background
4.3. Hypotheses and Experimental Procedures
4.4. Inferences
4.5. Conclusions

5. Discussion



This paper discusses aspects of disposition effects in several ways and perspectives. There is evidence, that investors sell winners earlier and hold losers longer. Theories from mental accounting, prospect theory, self-control, decision dependent emotions, internal locus of control, and many more relate to the disposition effect. After discussing them shortly, we investigate experiments in the laboratory and empirical evidence to come to the conclusion, that disposition effects exist for single investors and more pronounced for team investors. Tax considerations, automatic selling and decision dependent emotions change the proportion of how much investors are prone to disposition effects. The following mindmap shows an impression of the most important connections between the different effects.

Making it easier to read, enumerations are underlined and special effects are written italic. For this reason, also personifications are written in male standard, but of course women are equally meant not to discriminate them. You find more detailed statistical data in the Appendix, it is not always completely listed in the text. Unfortunately there is not enough space to discuss every aspect deeply, so I had to cut out the less relevant parts.

Figure 1 The disposition effect - referring theories and effects

Abbildung in dieser Leseprobe nicht enthalten

1. Introduction

To get an impression how much people know about the disposition effect and what they think of it, I asked[1] some former colleagues working as bank advisors for wealthy customers and business customers three questions. The first question is: “Do you think the disposition effect has an huge impact on the decisions of investors?” They think there is an impact especially for private investors. However, some of them made the experience private investors hold winner stocks too long.

The second question is: “Which methods do you use for yourself trying to resist the disposition effect?” Interestingly, they answered very different. One tries not to control his performance too often, while another checks it regularly. They have in common to use cost-average-effects by investing periodically, which is a typical argument in this sector. Furthermore, they try to investment long-term which is a typical strategy of value investors.

The third question is: “Do you thing the disposition effect is stronger or weaker for team decisions than for individual decisions?” All of them think team investing is more rational and harms the disposition effect. One of them is referring to the DISC-Model[2] trying to find an analytical partner (C-type).

It will be interesting to read more detailed about the disposition effect and its impact on investors. This can help us to understand how market phenomena works or rather how several aspects are related to each other. Of course it might also be useful to reflect disposition effects for yourself.

2. Basic discussion of the disposition effect

2.1. The main elements of the disposition effect

2.1.1. Prospect theory

This theory is based on the article from (Kahneman & Tversky, 1979) who did the first pioneering steps towards a descriptive theory of choice under uncertainty which investigates the fact, that investors sell winner too early and ride losers too long. This behavior is not rational according to maximize profit, because the investor behaves to increase his utility. In this case his mental utility function (happy by generating a gain or sad by generating a loss) weights the realization of a gain or loss not perfectly rational. (Shefrin & Statman, 1985) show, that prospect theory is a combination of several features. They say there is an “editing stage” where decision makers frame all choices to a fixed reference point like the purchase price. Then comes the “evaluating stage” where the investor decides based on his utility preferences. Prospect theory exhibits a S-shaped valuation function, whereas standard theory would suggest a linear function for an absolut rational investor. We can visualize this S-shaped valuation function easily with a trivial exponential function where alpha is bigger than zero and smaller than one.

Abbildung in dieser Leseprobe nicht enthalten

(Shefrin & Statman, 1985) conclude in case of the S-shaped valuation function risk aversion for the domain of gains and risk seeking for the domain of losses. The smaller alpha, the more we can observe this effect.

2.1.2. Mental Accounting

If we assume a case, where investors can generate profit from exploiting the difference between short- and longterm tax rates, they would “swap” the stock with an identical return distribution to neutralize the regulations for tax advantages, (Shefrin & Statman, 1985) exhibit. In this framework, investors would not behave like the normatively optimal strategy of (Constantinides, 1983) suggests, but select dominated lotteries where prospect theory applies. In conclusion (Thaler, 1984) named this constructed framework: mental accounting. The main idea is that decision makers tend to segregate different types of gambles into separate accounts. They just ignore possible interactions in the portfolio, which would lead to another strategy. For each purchased stocktype a new mental account is opened, whereas the natural reference point is the asset purchase price. In fact, the fundamental loss realization and implicitly the closure of the corresponding mental account are unpleasant. (Gross, 1982) argues by contrast, the real act of faith is an excuse for selling at loss, where the decision maker justifies his loss realization by investing immediately in another asset from which he expects more profit. “Transfer your assets” overcomes the need to close the mental account, according to (Gross, 1982).

2.1.3. Seeking pride and avoiding regret

The effects of mental account ing will be exacerbated by the fact, that the decision maker propably has to admit the mistake to others, as (Shefrin & Statman, 1985) point out. In this context they exhibit: “The quest for pride, and the avoidance of regret lead to a disposition to realize asymmetry between the strength of pride and regret (regret is stronger) leads to inaction to be favoured over action.” Furthermore, they claim continouing monitoring a stock after selling could lead to inaction is favoured over action for decision makers who are prone to this effect. In that case they are reluctant to realize gains as well as losses.

2.1.4. Self-control

(Glick, 1957) exhibits that reluctance of realizing losses is a self-control problem. Also (Thaler & Shefrin, 1981) talk about an intrapersonal agency conflict in their planner-doer-model. The “doer” takes the primitive, emotional part which is influenced by the effects of prospect theory, mental acconting and pride & regret, so he wants to have chosen correctly in the past. The “planner” takes the rational part which is not prone to this effects, but exhibits willpower instead. The fact that traders were aware of riding losers was not rational led them to set rules that mandate the realization, e.g. a predetermined percentage or a stop-loss order, (Glick, 1957) points out. At some point less willpower is needed to realize a loss because of important (real or contrived) personal reason, e.g. medical expenditure, (Shefrin & Statman, 1985) claims. (Dyl, 1977) exhibits a low trading volume in December for stocks, which were appreciated during the year and comes to the notion, that there might be a tax lock-in effect. The empirical evidence shows, that the volume of loss realization is higher in December, (Shefrin & Statman, 1985). One explanation of this fact might be, that they can influence their year-end-performance positively by tax-loss selling, so that they feel proud for their one-year-performance[3]. Also self-motivation is easier, if there is a deadline characteristic such as the end-of-the-year-month December, (Shefrin & Statman, 1985) state. (Constantinides, 1983) theory exhibits, that a rational investor would consider the fact of tax-loss selling during the year and realizes continuously.

2.2. Empirical evidence of the disposition effect

The first interesting question (Shefrin & Statman, 1985) investigates empirically is: Do investors time the realization of their losses differently from their realization of gains and why? We know, that is is efficient to realize the losses in short term and the gains in long term, because of tax considerations[4]. Nevertheless, the more the investor is prone to the disposition effect, the more he operates in the opposite direction.

Main part of the analysis are the realized returns on common stock investments which are divided in three classes of roundtrip duration[5]: I) 0 – 1 month II) 2 – 6 months III) 7 – 12 months. The expected ratio of realizations for a gain relative to all realizations should be I) low (bigger influence of transaction costs) II) very low III) high for trades, primarily motivated by tax option. This displayed pattern is mainly influenced by transaction costs and tax advantages. In contrast to that, the actual values of the evaluation are nearly the same, each slightly among 60%, (Shefrin & Statman, 1985) prove. They come to the conclusion, that tax-induced trades are the minor proportion of all trades, but there might be other interesting characteristics for traders like liquidity or inflation.

The second part of the empirical analysis from (Shefrin & Statman, 1985) is the realization of gains and losses in mutural funds where transaction costs are negligible. It is divided in I) Broker/Dealer funds II) Direct Seller funds III) No-load funds and surveys a 252-month period from 1961 by showing the monthly purchases and redemtions. (Shefrin & Statman, 1985) investigate the ratio of redemptions of a fund in month t related to purchases of the fund in the preceding month and assume, that tax motivated traders will redeem losses in t which leads to a high level ratio whereas gains will deter redemption in t which leads to a low level ratio. They classify the 30 highest gains and 30 highest losses in categories 1) gain/loss in t only 2) gain/loss in t–1 only 3) gain/loss in t–1 through t, because the exact time when a purchase or redemption was made is unknown. As a result (Shefrin & Statman, 1985) points out, that the mean ratio of redemtions to purchases associated with gains is greater than the ratio associated with losses in two of the three classes, but the differences are not statistically significant.

2.3. Interaction of the effects

One of the main investigations of the article from (Shefrin & Statman, 1985) is the aversion of loss realization. The disposition effect in general includes all of the effects we discussed at the beginning: mental accounting, regret aversion, self-control and tax considerations, also the concentration of loss realizations in December is consistent with our descriptive theory. The evidence for this effect in real world financial markets, shows tax considerations alone can not explain the observed patterns of loss and gain realizations, (Shefrin & Statman, 1985) infer.

3. Experimental analysis of the the disposition effect for single investors

3.1. Theoretical background for the experiments

3.1.1. Known investigations in this context

(Weber & Camerer, 1998) exhibit an example for reference point effects. If a decision maker has to choose beween a sure 5 DM[6] gain or a lottery of 0 or 10 DM gain with equal probability, a risk averse decision maker would choose the clear 5 DM gain. But if the decision maker has to choose between a clear 5 DM loss or a lottery of 0 or 10 DM loss with equal probability, he would choose the lottery in this case. As a result (Weber & Camerer, 1998) point out, that the risk attitudes depend on the reference point, because no capital loss seems to be an important concern for this kind of decision makers. (Weber & Camerer, 1998) state: “The disposition effect can be explained by investors judging gains and losses relative to their initial purchase price, and being risk-averse toward gains and risk-seeking toward losses.”

The main aspect from the study of (O' Dean, 1998) are different tests, whichs all verify the disposition effect is significant. He informs more detailed: “[…] investors realize about 24% of the gains they could realize by selling, but they realize only 15% of their losses”, while the exceptional month December shows boths values close to 20%, because of the nearby tax advantage to sell. Following the given argumentation, we can interprete the decrease of the gain realizations in December analogous, because a realization of a gain would increase tax disadvantages in December. Thus an interpretation for the shift in December might be the accumulated effects for loss realization, inhibited by the disposition effect.

The disposition effect is not only significant for the stock market, (Shiller & Case, 1988) infer, it is also substantial for the housowner market. (Kroll, Levy, & Rapoport, 1988a) (Kroll, Levy, & Rapoport, 1988b) come to the solution by reporting from their experiments, that subjects behavior is different then portfolio theory or capital asset pricing model would suggest, because they do not diversify properly.

3.1.2. Theory of reference points

In general the disposition effect includes referent point effects, (Weber & Camerer, 1998) report. The reference point effect, in the sense of prospect theory, shows that people values their gains and losses from a specific reference point. This reference point is most obviously the purchase price, the previous-period price or some weighted average price. Investors see their decision as a clear loss realization or a chance for accepting a gamble to get back to their reference point. In the domain of losses, the investor is risk seeking, because the pain of an additional loss is less than the joy to recover. Vice versa, an investor behaves risk avers in the domain of gains, because he wants to lock in the success. If the investor set a new reference point, he creates new disposition effect s which intensify the overall effect[7].

3.1.3. Hypothesis

In this setting, (Weber & Camerer, 1998) states the main hypothesis, that in perfect competition the number of sold shares will be smaller for losing stocks than for winning stocks as a consequence of the disposition effect. For further investigations, we divide this hypothesis and specify two possible reference points: the purchase price (=H1) and the previous period’s price (=H2) which sells more shares when the sale price is above the reference point, vice versa. As a third hypothesis (=H3) we set an automatic selling condition. As the fourth Hypothesis (=H4) we claim, that the trading volume is positively correlated with the size of price changes.

3.2. Experimental Design

The experiment, accomplished by (Weber & Camerer, 1998), has fourteen periods in which each subject can buy and sell six risky assets[8]. They set prices not determined by trading actions and exclude the possibility to borrow or sell short. To test H3, they divide the experiment in session one with deliberate selling and session two with automatic selling, where all shares are sold at the end of each period and subjects can buy their favoured ones back. Not invested money is held in cash and pays no interest. The subjects know about the process, that stocks have different chances of rising and falling in price, but were independent across the given assets and they even know the chances of a price increase as well. More precisely the stocks increase or decrease by 1, 3 or 5 DM equal likely and independent for all stocks. They created this setting such that the expected value of price change for a randomly choosen stock was zero. It still makes sense to play this setting, because finding out the right trend gives us a positive expectation of a profit. After some periods, the subjects were asked to classify the actual trends. In this context, the optimal bayesian method is to count the frequencies a share rises in price to become an idea of the trend which they had at the beginning. So (Weber & Camerer, 1998) can ensure the disposition effect is clearly a mistake in this setting.


[1] Of course this is not representative, but gives a good idea from this point of view.

[2] Based on the theory of psychologist William Moulton Marston

[3] It could be interesting to investigate the fact, that private investors have to pay taxes for their gains immediately in contrast to mutual funds. In general, is there an effect when realizing a smaller nominal gain after the tax allowance has been claimed?

[4] We talk about the taxation system of the USA in 1985.

[5] “Roundtrip duration” is the length of time an investor is holding the stock before selling it.

[6] Currency: “D eutsche M ark”

[7] More abstract, it is another S-shape in the S-shaped valuation function.

[8] Figure 3 Time series of stock prices used in the experiment from (Weber & Camerer, 1998)

Excerpt out of 28 pages


Market Phenomena. Investors and the Disposition Effect
University of Göttingen  (Behavioral Finance)
Bachelorseminar for Behavioral Economics
Catalog Number
ISBN (eBook)
ISBN (Book)
File size
1774 KB
Mental Accounting, Prospect Theory, Self Control, Deadline Characteristic, Seeking Pride, Avoid Regret, Risk, Reference Point Effect, Group Polarization, Emotions, Focal Point
Quote paper
Julian Fischer (Author), 2017, Market Phenomena. Investors and the Disposition Effect, Munich, GRIN Verlag,


  • No comments yet.
Read the ebook
Title: Market Phenomena. Investors and the Disposition Effect

Upload papers

Your term paper / thesis:

- Publication as eBook and book
- High royalties for the sales
- Completely free - with ISBN
- It only takes five minutes
- Every paper finds readers

Publish now - it's free