Wealth shocks and portfolio choice

We use new euro area representative data from the Consumer Expectations Survey (CES) to elicit household-specific propensities to invest and consume out of positive wealth shocks. Using a randomized assignment of hypothetical lottery gains ranging from <euro>5,000 to <euro>50,000 and a realistic menu of consumption, saving and asset choices, we estimate the causal effect of wealth shocks on risky asset ownership and conditional asset shares. Wealth shocks have a positive effect on stockholding (between 8.4 and 12.8 percentage points increase in participation for the largest wealth shock). The majority of households do not participate in the stock market, even after a large increase in wealth. The conditional asset share invested in risky assets is constant for wealth shocks up to <euro>20,000, and edges up slightly (by at most 2 %) for larger prizes. Our evidence is consistent with constant relative risk aversion for the majority of risky asset investors, while we also find important heterogeneity in the level of risk aversion across individuals.