The underlying assumption of the work is that if an investor buys a stock that goes up in price and they sell it, he or she will get a burst of ‘positive utility’ right then at the moment of sale, depending on how big the realised gain is. Similarly if they buy a stock and it goes down, there will be a burst of ‘negative utility’ right then at the moment of sale, depending on big the realised loss is. “In our view this is largely a result of what people call ‘mental accounting’, Barberis explains. “When people think about their investing, they don’t think about it in terms of the overall return on their portfolio, they think about it as a series of investing ‘episodes’, where each episode is characterised by three things: the name of the investment; the purchase price; and the sale price.”

For example, he says, one episode might be: “I bought IBM at $80 and sold it at $100, or I bought my house for $280,000 and sold it for $320,000. People think about investing in a series of episodes; some of them are good, in the sense that you sold for more than the purchase price, some of them are bad, in the sense that you sold for less than purchase price. When you sell an asset at a gain you feel good because you are creating a positive new investing episode, one that you can look back upon and talk about with pleasure.

By contrast if you sell a stock at a loss you are creating a negative investing episode, one that will be painful to look back on and talk about.” Barberis admits this phenomenon probably applies more to individuals than institutions, who he thinks are trained to think about their investing history in terms of portfolio return, not a series of episodes. The theory probably applies most readily to the trading of individual stocks by an ordinary investor. Essentially what the model finds is that if an investor buys a stock he will sell it only if its price rises a significant amount above the purchase price, and he will never voluntarily sell at a loss.

Because the investor is desperate to avoid the negative feeling associated with crystallising a loss, he will postpone selling indefinitely until he is forced to do so by a liquidity shock. That means investors are willing to buy a stock with a negative expected return, so long as its liquidity is sufficiently high. Even though Barberis supposes realisation utility is linear – that is, the pleasurable feelings increase in line with size of the gain, he says investors are actually risk seeking.