Umeå University's logo

This thesis consists of an introductory part and four self-contained papers related to individual investors’ subjective expectations and their financial behavior. 

Paper [I] analyzes multiple measures of individual investors’ expectations of risk and return using survey data on a random sample of individual investors in Sweden. The results indicate that, even though expectations from different measures are correlated, the magnitude of especially the risk expectations varies considerably between measures. The variations in the expectations mainly relate to differences in the responses to the questions underlying the different measures, rather than to the methods used to obtain them. Evaluation of the measures using three different comparisons indicates that the expectation measure proposed by Dominitz and Manski (2011) is the only measure for which it is possible to distinguish between individuals’ expectations using all of the considered comparisons.

Paper [II] addresses the relationship between sophistication and the expectations of individual investors with respect to risk and return. The findings show that sophisticated investors have lower (higher) risk (return) expectations that are closer to objective measures than less sophisticated investors. These results are important, since they enhance the understanding of the underlying mechanisms through which sophistication could influence individuals’ financial decisions.

Paper [III] provides new evidence for the sources of information individual investors’ use when making financial decisions and the relationship between how frequently investors use this information and their expectations of the risk and return in a stock market index, their confidence in these expectations, and their portfolio risk and return. The findings indicate that individual investors use different sources of filtered financial information (e.g., information packaged by a professional intermediary) more frequently than unfiltered financial information (e.g., information from annual reports and financial statements). However, an increase in the frequency with which investors use filtered financial information is positively related to their confidence in their stock-market expectations and to the risk in their stock portfolios. For investors who instead use unfiltered financial information more frequently than filtered financial information, the results indicate that they have more accurate stock-market expectations, lower portfolio risk, and higher portfolio return.

Paper [IV] links individual investors’ financial risk-taking to the behavior of peers within their community. By using detailed data at the individual level, it is found that the risk-taking among peers affects individuals’ choice of portfolio risk. The results hold for the full sample of individuals concerning their choice of the overall proportion of risky assets and stock market participants’ choice of total and systematic stock-portfolio risk. Overall, the results stress that interaction with peers is an important channel through which individual risk-taking is affected.