Week 34 Day 3: The Disposition Effect: Selling Winners, Holding Losers

You own Stock A (up 20%) and Stock B (down 20%). You need cash. Most investors sell Stock A ('lock in the win') and hold Stock B ('wait for it to come back'). But Stock A's fundamentals may be strong (hence the gain), and Stock B's fundamentals may be weak (hence the loss). You are selling your best performer and keeping your worst.

The disposition effect by the numbers (Odean, 1998): Individual investors are 1.5x more likely to sell a winner than a loser. The winning stocks they sold outperformed the losing stocks they held by approximately 3.4% over the next 12 months. In other words: the stocks they sold were the wrong ones to sell. The three costs of the disposition effect: (1) Return cost: selling winners and holding losers reverses the momentum premium. Winners tend to continue winning for 3-12 months (Jegadeesh and Titman, 1993). By selling them, you exit a profitable trend. (2) Tax cost: selling winners realizes capital gains (taxable). Holding losers defers tax losses (which are valuable). The optimal tax strategy is the exact opposite: sell losers (harvest the tax loss) and hold winners (defer the tax gain). (3) Opportunity cost: the money from selling a winner typically goes into a new, untested position rather than remaining in the proven winner. How to fight the disposition effect: (1) Never sell individual positions; hold a diversified index and avoid the single-stock dilemma entirely. (2) If you hold individual stocks, set time-based sell rules ('I will evaluate in 12 months') rather than price-based rules ('I will sell when it recovers'). (3) Tax-loss harvest annually: sell ALL positions with losses in December, buy similar replacements immediately.

The disposition effect was formally documented by Shefrin and Statman (1985) and empirically confirmed by Odean (1998) using trading records of 10,000 individual investors. The magnitude: the 'proportion of gains realized' (PGR) exceeds the 'proportion of losses realized' (PLR) by approximately 50% -- investors are approximately 1.5x more willing to sell a position at a gain than at a loss. Grinblatt and Keloharju (2001) confirmed the finding in Finnish stock trading data and showed that the effect is stronger for less sophisticated investors. The theoretical explanation combines prospect theory (loss aversion makes loss realization painful) with mental accounting (investors evaluate each position separately rather than as part of a portfolio). Barberis and Xiong (2009) provided a formal model: under narrow framing (evaluating each stock independently) and prospect theory preferences, the optimal strategy is indeed to sell winners and hold losers -- because realizing a gain provides utility (coding a mental account as 'successful') while realizing a loss provides disutility. This is individually rational under the wrong model (prospect theory utility) but collectively suboptimal under the correct model (total wealth maximization). The disposition effect is one of the strongest arguments for index investing: by holding VTI (the total market), the investor never faces the single-stock sell decision and therefore never triggers the disposition effect. The index investor's only decision is the contribution amount and the rebalancing schedule -- both of which can be automated.