One will think that when a stock is considered a winner the investor will be willing to hold this stock and try to get rid of the loser stocks. In contrast, studies have indicated that there is an anomaly affecting the investor's behavior on stock trading. In Hersh Shefrin and Mein Statman's report (1985) this anomaly is referred as a disposition effect and it is related to investors' tendency on selling assets whose price has substantially increased too quickly but also holding losers too long. In other words, investors have a propensity to recognize gains while in the same time are less willing to recognize losses. Due to this fact, several papers have been published to illustrate the outcome the disposition effect has on several aspects of stock trading. Taking this in consideration, this paper focuses on the features of this effect in the extend of the angles these economists have presented their empirical evidence.
Following the Prospect Theory raised by Amos Tversky and Daniel Kahneman (1986), Shefrin and Statman (1985) take a further step and try to examine the decisions that might have a crucial impact on the decision making on realizing gains and losses in a market setting. Through their empirical evidence they include elements such as mental accounting, self-control, regret aversion and tax consideration in order to perceive these results and widen the prospect theory. In addition, Shefrin and Statman (1985) taking into consideration the work of Constantinides(1985), who studied a strategic optimization of short term- gains and losses through US tax code, came up the December effect. Constantinides(1985) argued that the returns on an investment can be either taxed in the short run with an ordinary income or in the long run with a lower rate. Whenever transaction cost appear, the normative theory of Constantinides(1985) in controversy to the disposition effect, suggests to realize immediately any losses; thus, loss realization should increase steadily making December the peaking month. Extending these findings, Shefrin and Statman (1985) use the disposition effect to test the normative theory and introduce a fifth aspect, named the potential gain, to the prospect theory giving the interaction results with the other elements. In addition, the evidence of the phenomenon of the disposition effect can be observed in real-world financial markets and according to Statman's and Shefrin's (1985) report tax considerations cannot on its own explain the pattern but only when combined with the disposition effect. In a more recent paper conducted by Terrance Odean (1998), the empirical evidence has shown that investors are willing to postpone taxable gains and continue holding profitable assets in their portfolio and should capture taxable losses by clearing all securities in a non-constant rate.
Empirical evidence has mainly supported the disposition effect. The prospect theory, by Tversky and Kahneman, describes decisions in situations where there are several alternatives with uncertain results. This theory is descriptive and tries to explain investor behavior in real circumstances rather than optimizing decisions. Usually a reference point is set in order to compare the results, such as the value of a share when purchased. As mentioned above the disposition effect can be clarified by investors judging their gains and losses according to the initial price bought and the fact of being a rational investor with no abnormal behavior. In the prospect theory the value function is clarified over single outcomes. In order to value multiple outcomes Thaler (1985) introduced the mental accounting system, that explains how traders appraise the results of their investments and the way their financial decisions are framed. Thaler (1985) exploits the principles of mental accounting that can establish whether segregation or integration is favored. Thaler's empirical evidence illustrate that traders should segregate gains and integrate losses because of the fact that the value function reveals diminishing sensitivity as the scale of the gains or loss becomes larger. In other words, a loss can be compensated by a greater gain.
The disposition effect has also consequences in other trading parameters. Lakonishok and Smidt (1986) by observing the New York Stock Exchange found out that some stocks had irregular volume accordingly to the volatility of the prices in the previous months. He concluded that winners had higher trading volume than losers. Trying to test this theory Ferris (1988) employed US stock prices and compared the historical with the current trading volume. His results, which were in accordance to the disposition effect, indicated the negative correlation between the current volume and the volume of the trading days in which the price of the stock was higher substantially to the spot price. In Stephen Ferris, Robert Haugen and Anvil Makhija (1987) can empirical evidence illustrated strong support of the disposition effect and the trading volume but inconsistency with the tax- loss- selling. It is indicated that this effect is not only determinant of "year-end" volume, but can also determine trading volume levels during the year. In Dyl's report, in accordance to the December effect, it is argued that losing stocks, specifically those that had a drop in price about twenty percent, experienced very high trading volume in the end of the year, while the stocks that had over performed disclose low trading volume. These findings were supportive to the tax- loss- selling hypothesis. In contrast to Dyl's evidence, Statman and Shefrin interpreted differently this abnormal trading volume during December. They invoked self-control as an issue to support the disposition affect that forecasts such behavior, being against the tax- loss - selling hypothesis.
Obtaining 10.000 accounts as a database Odean concluded that the rationality of investors resulted in keeping underperforming assets and selling winners because investors had the belief that the losing stocks will gradually rebound while the winners will drop in price. Checking the results, these actions proved to be wrong because the losers that were held in the portfolio performed much worse than the winners sold. A possible explanation to this radical behavior could flow from the belief of investors that their underperforming assets will in future outperform the current winner stocks. The investors may sell the winners in order to rebalance their portfolio or due to high transaction costs of low price trading they may cease on selling losers. In cases where the data are controlled for rebalancing purposes and for stock price the disposition effect is still detected and the wining investors continue to perform better than the losers they hold. Another explanation given by Harris (1988) is that high transaction costs in the lower asset prices have an impact on investors. The idea on the analysis of Harris has to do with the scale of similar stocks, investors are eager to sell the winner stocks even when the trading costs are similar. Overall, in contrast to the prospect theory and to the mean reversion belief, it is proved that investors are more likely to sell either the larger gains or losses in a slower manner than the small gains and losses.
In terms of portfolio decisions, the disposition effect can also have a crucial impact. Through Kroll's (1988) empirical evidence this kind of situations were tested. According to portfolio theory the subjects of the tests could invest in the risk-free securities, in risky assets and leverage was allowed. The results of the empirical test were in controversy to portfolio theory and the CAPM. The subjects of the experiment tended to behave quite differently, as they did not diversify as ruled by the optimization theory neither hold the right weights in risky stock. Adding to this report, Kroll and Levy (1992) made further changes concerning the separation of a portfolio. Martin Weber and Colin F. Camerer (1998) tried to test this behavioral effect in stock picking by matching individuals' decisions in transactions of stocks with the prices of these assets and thus giving results in the selling behavior. The experiment results strengthened Kroll's (1988) findings. Specifically, there is a tendency on selling small amount of assets when there is depreciation in prices rather than when there is a rise. It has also been noticed that investors have a propensity to trade away shares whose price is lower than the purchased price. All in all, the empirical evidence verifies Shefrin and Statman's (1985) work but on the other hand is inconsistent with the portfolio theory. As in Kroll (1988) the subjects of the tests, were keen on knowing past performance of stocks, which were random via a probability distribution, that might explain this abnormal behavior.
Another aspect that the disposition effect induces is the under-reaction to news that leads to the prediction of future returns. In Andrea Frazzini's (2006) analysis, he suggests that the disposition effect can stimulate this reaction to news and thus a change in price. This depends on the type of information and the reference price in relation to the current price. Specifically, the empirical evidence indicates that bad news about trade trading at large losses can lead to negative change in price. Similar are the results of good news. The asymmetric pattern found is in consistency with the disposition effect because of the difference in the referenced and spot price. As a consequence, an investor that faces capital losses is reluctant to realize this loss, thus creating an under-reaction to negative news. In the same outline when there is positive news announcements a prevention of an asset pricing rise is occurred by active selling.
So if the disposition effect occurs on every stock exchange can it vanish when professional traders make all the trading decisions? This question is the main subject for a lot of economists. As the fundamental of this phenomenon is the irrational behavior of investors one could question if "discipline" strategies are introduced will this anomaly extinguish? In Steve Mann and Peter Locke's (2000) paper this question is investigated. They examined the behavior of professional investors and concluded that the most successful financier is the one that minimizes the time they hold the losers in their portfolio while the least successful trader is the one that holds the losers for the longest period. Extending this theory, Coval and Shumway (2000) account support of loss aversion between professional traders, concentrating in a more specific category that of "morning loser traders". In addition, Wermers (2003) indicates that even in funds many of the managers of underperforming funds tend to hold the
As there is a difficulty in measuring variables, such as the professionalism of a trader etc, it is almost impossible to conclude on one certain result. Also, there are no evidence of success being simultaneous related to an inclination to keep losses for a greater period than gains. Therefore, it is obvious that there is no evidence of costly disposition effect among professional investors. On the other hand, a lack in discipline in the realization of profit and losses is hazardous to the probability of a successful trade. It is understood by the evidence indicated that even professional traders cannot escape this financial behavior but surely can reduce it.
The disposition effect clearly has a huge impact in stock trading. Even though behavioral finance has recently become a popular topic in the finance field there is not a lot of the empirical evidence to support it. This is due to the fact that it is relatively complicated to test the behavioral models for different reasons. It is difficult to gather all the parameters simply because the variables are not always quantitative. The aim of this paper is to provide a detailed look in to the aspects and observations of the disposition effect. Other aspects analyzed in this report are trading volume, reaction to news, professionalism of the investors and portfolio theory. There are also extensive arguments presented concerning the tax implications and the result that the disposition effect has in the month of December. These are the major aspects and are significant in identifying whether behavioral biases are important in influencing the stock trading.
