Why do households not hold stocks? A growing literature advocates that factors such as stock market return experience, lack of trust, educational attainment, financial sophistication and cognitive ability, explain the stock market non-participation puzzle.
This article sets up a theoretical framework and empirically assesses the distinct channels of stock market literacy and trust that simultaneously explain households’ stock ownership decisions. Additionally, we investigate whether the previously documented evidence for sociability is in fact capturing the role of stock market literacy and hence whether it is literacy, rather than sociability, that matters for understanding stock market participation.
For the empirical analysis, we use data from the American Life Panel (ALP) survey, consisting of over 340 diverse surveys and 6,000 representative samples of US consumers. ALP surveys capture a rich set of information of scientific and policy interest, such as expectations, opinions, financial participation and circumstances,cognition, and demographics. Hence, we are able to measure stock market literacy,
sociability, and trust in the stock market, and also take into account for a wide range of personal characteristics that can explain the heterogeneity observed in stock market participation, including past economic shocks, future expectations, optimism, risk aversion, self-confidence, sense of commitment, and time preference.
The results indicate that stock-market-literate and trusting households are more likely to participate in stocks and invest a higher proportion of their wealth in the stock market. These two independent household characteristics concurrently remain significant, even after accounting for several other behavioral variables. Changing stock market literacy by one standard deviation varies the probability of participation by 11%, while the equivalent change for trust in the stock market is around 17%. We find that past economic shocks and future expectations explain households’ probability of participation and, conditional on participation, several other behavioral characteristics such as self-confidence and time preference, along with past economic shocks and future expectations, explain households’ portfolio choice decision of how much to invest in stocks.
Next, we find no association between sociability and participation, once we account for stock market literacy. The additional analysis finds that sociability is insignificant even among highly sociable households if they have low stock market literacy; conversely, a significant relation holds between stock market literacy and participation even among low sociable households. These results indicate that households with low sociability invest in stocks if they are stock-market-literate; hence participation is explained by households’ level of stock market literacy rather than their level of sociability.
Our findings will be of interest to policy makers in their strategic endeavors to promote stock market participation. For instance, since stock market literacy and trust concurrently explain participation, this relation should be taken into account while designing financial literacy programs for encouraging stock market participation. Also, our results show that social interaction and peer group effects cannot explain stock ownership decisions per se; what matters is stock market literacy.