Investors are not perfect.

Like everyone else, they can be too optimistic and too pessimistic. Yet, many theoretical models assume that investors are completely rational.

In a new paper in the August issue of the American Economic Review, authors Klaus Adam, Albert Marcet, and Johannes Beutel come up with a more realistic model of how investors think. In the model, investors use their feelings about an asset’s recent performance to predict its future price.

For example, if a stock price is rising, investors will expect it to keep rising. Even if the price is about to plummet, investors will remain optimistic. These beliefs then factor into stock prices. If investors believe that a stock will do well, they will buy more of that stock, making the price increase even more.

This self-fulfilling prophecy can only last for so long. When prices deviate too much from the stock’s actual value, they are bound to crash. The authors point out that, unlike other models, accounting for these subjective beliefs does a good job of explaining boom-bust cycles as well as differences between asset prices and their true values.  

Figure 6 from Adam et al. (2017)

Figure 6 illustrates how investor optimism evolves given two pieces of information: the investor’s current belief about an asset’s price and her belief before that. If you follow the arrows, you will eventually land on the black dot, which represents the true price.

Investors may ultimately get the price right, but only after making persistent mistakes. This doesn’t mean investors are bad at their jobs. It just means that they’re human ー and that maybe economists’ models should be more human, too.