Psychology and Investing: New Research Findings

As Alan Greenspan has rather memorably put it, "there is one important caveat to the notion that we live in a new economy, and that is human psychology."

We have noted before that psychological researchers have found that human beings tend to behave in ways that are predictably irrational. In other words, most people’s thinking is reliably biased in a consistent manner.

Why is this important to an index investor? Because, despite their understanding of the twin dangers of high expenses and high turnover (both of which generate costs, and make it impossible to ‘beat the market’), we too are human and therefore occasionally tempted to trade more actively than is good for our financial health. As is the case with our physical health, we can benefit from the occasional "booster shot" that strengthens our resistance to these dangers.

With that in mind, we’d like to briefly review the results of two important academic research studies.

In one that has now become something of a classic, Terrance Odean (then at Berkeley, now at the University of California at Davis) analyzed a data set consisting of the trading records of 60,000 households that maintained an account at a large discount brokerage firm between February, 1991 and December, 1996. In his article titled "Are Investors Reluctant to Realize Their Losses?" (Journal of Finance, 1998), Odean found that over this period, the average investor in his study earned a compound annual return on his or her account of 15.3%. However, the 20% of households that traded most frequently realized a compound annual return of only 10.0%. Most importantly, Odean found that it was the cost of frequent trading (commissions, bid/ask spreads, and taxes), and not differences in investment selection that accounted for the difference in return between the high traders and the rest of the group.

More recently, Odean and a colleague, Brad Barber, have authored a new working paper, entitled "The Courage of Misguided Convictions". In this case, they analyzed the performance of 10,000 investor accounts between January, 1987 and December, 1993. Their study focused on the question of why investors tend to sell their winners too early, and hold their losers too long. There are two logical explanations for why people might do this.

On the one hand, they might believe that over time, the performance of most shares "reverts to the mean." This means that a company with poor recent performance is more likely to improve than to get worse, while a company with above average recent performance is more likely to decline than to get better. Arguably, "mean revision" is the single most important belief held by true value investors (and its opposite, "mean aversion" is the single most important belief held by momentum investors).

On the other hand, investors might sell their gains faster than they realize their losses because the latter causes them twice as much pain as the former causes them pleasure. Or, as the country song says, "losing hurts twice as much as winning feels good." More formally, this is known as either "prospect" or "loss aversion" theory.

In the study they conducted, Odean and Barber started out with the reasonable hypothesis that if investors were motivated by mean revision, then the losing stocks they held on to would tend to outperform the winning stocks they sold in the months following their sale. Unfortunately, the data they studied did not support this. Over the year following their sale, investors’ winning stocks outperformed their losers by 3.4% per year (relative to a market index). In other words, "investors who sell winners and hold onto losers because they think the latter will outperform the former are, on average, mistaken."

Moreover, if investors’ decisions were actually motivated by a belief in mean revision, when adding to their portfolios they would tend to buy stocks that had previously performed poorly. Unfortunately, Odean and Barber found that this wasn’t true for the investors they studied, who bought stocks that, on average, had outperformed the market index by 25 percent over the previous two years.

In summary, these studies provide us with three important insights:

• We seem to have a natural tendency to overtrade.
• When we buy, we often do so on the basis of momentum rather than value.
• We tend to sell our winners too soon and hold our losers too long.

Next month, we will further explore these findings.

| Psychology and Investing: New Research Findings | Choosing an Index for an Asset Class | Recommended Portfolio Performance |