Depression Babies

Personal experience determines how much investment risk people are willing to take

It makes intuitive sense. If you've seen the stock market tank, you might be leery about putting your savings in company shares. But that proposition runs against conventional wisdom in economics. Standard economic models assume that experience doesn't change a person's appetite for risk. Those models also assume that people take into account all available information about the past when making investment decisions, not just what they've witnessed themselves.

Is this standard model wrong? Ulrike Malmendier, a finance professor at UC Berkeley's Haas School of Business, thinks so. In a 2009 paper published by the National Bureau of Economic Research, Malmendier and co-author Stefan Nagel tested the idea that personal experience affects how much investment risk people are willing to take on. Their analysis is a classic work in behavioral economics, the subdiscipline that draws on the field of psychology to understand how people make economic decisions.

Malmendier and Nagel designed an innovative way of examining the relationship between personal experience and willingness to take on risk. They used consumer survey data going back to the early 1960s to develop four measures of risk tolerance: consumer responses to questions about how much risk they were willing to take in their investments; whether households owned any stock; whether they owned bonds; and the proportion of household liquid investments held in stock. They then matched those risk tolerance measures with inflation-adjusted stock and bond returns over an investor's lifetime.

The results were striking. "We find that households' risk taking is strongly related to experienced returns. Households with higher experienced stock market returns express a higher willingness to take financial risk, participate more in the stock market, and … invest more of their liquid assets in stocks," Malmendier and Nagel wrote.

Specifically, the authors found that households that had seen higher stock market returns than 90 percent of other households, that is, those at the 90th percentile, were 10.2 percentage points more likely to own stock than households at the 10th percentile. And those 90th percentile households held 7.9 percentage points more of their liquid investments in the form of stock than 10th percentile households. The pattern was similar for bonds.

Malmendier and Nagel measure of investment returns was nuanced. For each household, they calculated "experienced returns," based on stock and bond market performance over the life of the household head. They used a scale in which recent results carried more weight than those from the distant past, rooted in the idea that new memories are vivid, while old ones fade. Thus, an event like the stock market crash during the Great Depression was contradictory. It was a searing experience that deeply traumatized those who lived through it. But it happened many decades ago, diminishing its impact.

What is the mechanism by which experience affects risk appetite? The authors looked at household survey data from 1998 to 2003 on expected returns from stocks and found evidence that experience shaped people's beliefs about how stocks would perform. They determined a 1 percentage point gain in lifetime experienced returns was linked with a 0.5 to 0.6 percentage point increase in stock market returns expected in the year ahead. Other factors might be at play, but the results suggest, when it comes to investments, people extrapolate from the past to predict the future. And if they are more optimistic about financial markets, they might be more willing to invest.

Malmendier and Nagel speculate that experienced returns may also have important overall effects on stock and bond market price levels. If past returns alter households' appetite for risk, then the aggregate demand for stock could shift over time. All else equal, higher demand would boost asset prices, while lower demand would depress them.