Tyler Cowen poses the following question about stocks, and what he says used to be the conventional behavioral economics answer.
Let’s say you bought two stocks last year. One has tanked and looks likely to fall further. One has gone up and you expect it to keep rising. (Hey, it’s not completely impossible.) Which are you more apt to sell?
Behavioral economists used to think they knew the answer: neither. Studies have shown that people tend to value things more – whether shirts, stereos or stocks – once they own them, no matter what has happened to their actual worth. This phenomenon is called the endowment effect. If it were the only psychological factor at work, you’d be reluctant to sell both losers and winners simply because they’re already tucked into your portfolio.
Cowen’s story is incomplete, and therefore unfair, even to old behavioral economists. In the scenario Cowen describes, two biases, each reinforcing the other, would be in effect: The endowment effect and loss aversion. The endowment effects for both stocks (assuming you bought them at the same price) would cancel each other out, but this would not necessarily mean investor paralysis. For more than twenty years, behavioral economists have been citing something called the disposition effect, which is an implication of prospect theory and the component of loss aversion). The status quo purchase price serves a reference point. Gains and losses are perceived relative to some other aspirational level different from the status quo – say, what you thought the stock would rise to. As the winner is closer to this aspiration, you, as the investor, become more risk-averse and therefore more likely to sell it, while holding on to the loser in the hopes of a roaring comeback, even one with a small probability.
But this isn’t the only explanation for identical behavior. An alternative is a commonly mistaken belief among average investors that stocks will revert to their mean. Stocks that have risen will fall; stocks that have fallen will rise. This story also predicts the selling of winners on the expectation that it will fall. Yes, Cowen’s scenarios says you, the ordinary investor, would expect the winning stock to keep rising. Old behavioral economics says you’d be quite extraordinary for believing this. Both of these potential explanations are laid out in Terrance Odean’s classic paper “Are Investors Reluctant to Realize Their Losses?” His data does allow him to distinguish which of the two stories makes more sense.
Addendum: Cowen’s column is actually an appreciation of a paper by Nicholas C. Barberis and Wei Xiong with yet another explanation for why investors sell winners and hold onto losers: That it’s the pleasure of actual (or what stock traders would called realized) gains – the good feeling you get from making a seemingly smart decision – and the pain of actual losses that leads to selling winners. Read the full paper.