Abstract

We analyze stock market participation in 19 European countries, providing a composite view of the interplay and relative importance of established participation drivers. We jointly control for nearly all relevant drivers found in prior studies, which tend to introduce one novel variable at a time and often omit risk-aversion. Excellent full model predictive power decomposes into institutional (country) fixed effects (about 30% of total), traditional individual-level variables (50%), and more recently identified behavioral variables (20%). We sketch a hierarchical framework where factors’ effects vary by agents’ proneness to participate. We also challenge and complement existing interpretations given to sociability, IQ, trust, and life experiences.

Introduction

Limited stock market participation has been the quintessential topic in the emerging field of household finance.1 Differences in stockholding propensity across countries, and between people of comparable wealth, show that country factors related to economic or cultural institutions are important. On the individual level, wealth, income, and education are highly influential. Besides these traditional factors, more recent studies have uncovered an interesting set of behavioral factors that explain stock market participation. What is missing, however, is a composite view of the interplay and relative importance of the various drivers of participation.

In this paper, we take a novel, big-picture approach. Whereas the literature has tended to focus on testing the effect of a single new factor after controlling for traditional ones, we combine an extensive set of these variables in one model and explore the relative contribution of three types of factors – institutional, traditional, and new behavioral ones. We jointly estimate the effects of all variables while including a directly queried financial risk aversion measure known to predict actual financial risk-taking well (Dohmen et al., 2011; Halko et al., 2012; Guiso et al., 2018; Laudenbach et al., 2020). Then, based on the contribution of individual explanatory variables in various regulatory environments and subsamples of varying sophistication, we sketch a hierarchical model that describes the likely importance of each factor in different contexts.

Conclusion

The stock market participation literature has progressed steadily over the past few decades, but in step-by-step fashion, usually by adding one novel explanatory variable at a time. We take a consolidating approach by examining the roles of institutional, traditional, and behavioral factors, independently and jointly. We use data from SHARE, which comprehensively cover traditional drivers of participation, including risk aversion, and nearly all of the behavioral drivers from the recent stock market participation literature. We take full advantage of the data, exploiting cross-country variation in regulatory quality and varying levels of wealth and education across individuals.

Institutions clearly matter. Of the explanatory power, institutional factors captured by country fixed effects account for about a third. The traditional individual-level variables capture about 50%, and less than a fifth of the variation is explained by the new behavioral variables. We show that the effects of traditional variables and new behavioral variables change across levels of regulatory quality and across levels of wealth and education. In environments where participation rates are higher, such as high regulatory quality countries and among the wealthy and highly educated, variables reflecting values and attitudes have larger effects. In environments with lower levels of participation these variables have smaller or even non-significant effects. To complete our comprehensive approach, we also take a look at individual behavioral drivers and document complementing and contrasting results with previous literature.

Based partly on our results, and partly on prior research, we suggest a hierarchical model of participation drivers. In this model, similar in spirit to Maslow’s (1943) hierarchy of needs, the low-level factors, such as wealth, have to be on a sufficient level before high-level, e.g., psychological, factors can come into play. Further progress in explaining non-participation, particularly among the well-off, would likely benefit from a focus in the high end of the model.