Can Prospect Theory Explain the Disposition Effect? A New Perspective on Reference Points

Abstract

There has been recent debate about whether prospect theory can explain the disposition effect. Using both theory and simulation, this paper shows that prospect theory often predicts the disposition effect when lagged expected final wealth is the reference point under the principle of preferred personal equilibrium, regardless of whether the reference point is updated or not. When initial wealth is the reference point, however, there is often no disposition effect. Models that use a reference point with no lag under the principle of preferred personal equilibrium or that determine the reference point using the principle of disappointment aversion cannot explain why the investor bought a stock in the first place. Reference point adjustment weakens the disposition effect, leads to more aggressive initial stock purchase strategies, and predicts history dependence in stock holding. This paper was accepted by John List, behavioral economics.

Behavioral Household Finance

“…A common preference-based explanation for the disposition effect appeals to prospect theory (Kahneman and Tversky, 1979;Shefrin and Statman, 1985;Odean, 1998b;Meng and Weng, 2016). Because the prospect theory value function is convex in the loss domain and concave in the gains domain, investors will be risk-seeking in underwater positions and risk-averse in appreciated positions, causing them to be prone to hold onto losing stocks and sell winning stocks.…”

“…Since a stock must have had a positive expected return in order for the investor to have bought it, the distance from the kink will tend to be larger after a gain than a loss, causing a reverse disposition effect-the investor is more prone to sell after a loss than a gain. Meng and Weng (2016) are able to restore prospect theory's ability to generate a disposition effect in the Barberis and Xiong (2009) setting by assuming that the reference point equals the lagged expectation of end-of-year wealth rather than initial wealth, which moves the reference point closer to the stock's price after a gain. Barberis and Xiong (2009) show that a model where gain-loss utility is experienced only at the time a position is sold ("realization utility") more robustly produces a disposition effect.…”

“…The disposition effect is the tendency of investors to sell stocks that have appreciated since purchase and hold stocks that have declined since purchase (Shefrin and Statman, 1985;Odean, 1998b;Weber and Camerer, 1998;Grinblatt and Keloharju, 2001b;Shapira and Venezia, 2001;Feng and Seasholes, 2005;Brown et al, 2006;Barber et al, 2007;Chen et al, 2007;Calvet et al, 2009). Although it could be rational for an investor to sell some of an appreciated stock in order to rebalance her portfolio, Odean (1998b) and Brown et al (2006) show that investors are also more likely to completely liquidate a position when its price has fallen since purchase.A common preference-based explanation for the disposition effect appeals to prospect theory (Kahneman and Tversky, 1979;Shefrin and Statman, 1985;Odean, 1998b;Meng and Weng, 2016). Because the prospect theory value function is convex in the loss domain and concave in the gains domain, investors will be risk-seeking in underwater positions and risk-averse in appreciated positions, causing them to be prone to hold onto losing stocks and sell winning 23 stocks.…”

Losers and prospectors in the short‐term options market

“…Most laboratory studies assume a status quo reference point. However, empirical results such as those about the disposition effect are shown to be contingent on a reference point that is picked (see Barberis & Xiong, 2009;Meng & Weng, 2017). For example, Heath, Haddart, and Lang (1999) investigate employee stock option exercise decisions and find that the exercise decision depends on recent price movements.…”

A News-Utility Theory for Inattention and Delegation in Portfolio Choice

“…11 Heidhues andKoszegi (2008, 2014) and Herweg and Mierendorff (2013) explored the implications for consumer pricing, which were tested by Karle, Kirchsteiger, andPeitz (2011), Herweg, Muller, andWeinschenk (2010) did so for principal-agent contracts, and Eisenhuth (2012) did so for mechanism design. An insufficient list of papers providing direct evidence for Rabin (2006, 2007) preferences is Sprenger (2015) on the implications of stochastic reference points, Abeler, Falk, Goette, and Huffman (2011) on labor supply, Gill and Prowse (2012) on real-effort tournaments, Meng and Weng (2016) on the disposition effect, and Ericson and Fuster (2011) on the endowment effect (which, however, is not confirmed by Gneezy, Goette, Sprenger, and Zimmermann (2017) and Heffetz and List (2014)). Suggestive evidence was provided by Crawford and Meng (2011), Pope andSchweitzer (2011), andSydnor (2010) among others.…”